0198778228
0198778228
0198778228
OXF O RD ST UD I E S
I N ANCIE NT
PHILOSO PH Y
EDITOR: VICTOR CA S TON
VO L U M E L
3
3
Great Clarendon Street, Oxford, ,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Except where otherwise stated, Oxford University Press,
The moral rights of the authors have been asserted
First Edition published in
Impression:
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence, or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this book in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
Madison Avenue, New York, NY , United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Oxford studies in ancient philosophy.—
Vol. l ().—Oxford: Clarendon Press;
New York: Oxford University Press, –
v.; cm. Annual.
. Philosophy, Ancient—Periodicals.
B.O .′—dc. –
AACR MARC-S
ISBN –––– (hbk.)
ISBN –––– (pbk.)
Printed in Great Britain by Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work
A D V I S O R Y B O ARD
. The programme
K. Reinhardt, Parmenides und die Geschichte der griechischen Philosophie [Par-
menides] (Bonn, ), , takes ἐχρῆν as a past historic. Mourelatos, Route, –,
argues for the counterfactual construction.
See most recently M. R. Cosgrove, ‘What Are “True” doxai Worth to Parme-
nides? Essaying a Fresh Look at his Cosmology’ [‘Fresh Look’], Oxford Studies in
Ancient Philosophy, (), –, for a survey of scholarship.
See Mourelatos, Route, .
Cf. J. H. Lesher, ‘Parmenides’ Critique of Thinking: The poludēris elenchos of
Fragment ’, Oxford Studies in Ancient Philosophy, (), – at , who argues
Parmenides’ Likely Story
as the one that being passes and so makes it an object of alēthēs pistis.
This creates a certain tension between the two terms dokounta and
dokimōs, which Parmenides exploits: what is believed falls short
of alēthēs pistis, yet it will be held to standards of credibility that
can properly be met only by genuine proof. How beliefs can do so is
at this point left intriguingly open, but the answer, I shall suggest,
will lie in the extent to which ta dokounta match the signs (sēmata)
which, as at B . –, ultimately single out being as the only genuine
object of enquiry.
Owen overstates his case when he argues that dokimōs serves to
exclude the possibility that any dokounta could be in any way reli-
able:
There can be no degrees of reality: what exists must πάμπαν πελέναι, on
pain of being nothing at all. But all such general issues apart, Wilamowitz’s
sense cannot be got from the Greek. Where δοκίμως is attested elsewhere
(Aeschylus, Persae , Xenophon, Cyr. i. . ) the lexica and the editors
rightly translate it ‘really, genuinely’, and the earlier editors of Parmenides
had no doubt that this was its sense in the present context. The δόκιμος is
the reliable man, not one who measures up to some standards but fails the
main test. So δοκίμως εἶναι is assuredly to exist; and this is what the pheno-
menal world can never do for Parmenides’ goddess. The same fact defeats
any attempt to read δοκίμως as ‘in a manner appropriate to δοκοῦντα’.
‘is simultaneously existential and predicative’. See also L. Brown, ‘The Verb “To
Be” in Greek Philosophy’, in S. Everson (ed.), Language, Companions to Ancient
Thought, (Cambridge, ), –.
For the need to observe the connection between the δοκ- words see Mourelatos,
Route, –.
Owen’s strategy of seeing the δοκίμως from the mortals’ perspective rather than
the goddess’s does not work in any case, since the mortals do not see their δόξα as
mere δοκοῦντα but as ὄντα, and surely δοκοῦντα has to be viewed from the same per-
spective as δοκίμως.
As noted by A. H.Coxon, The Fragments of Parmenides [Fragments], rev. edn.
(Las Vegas, ), .
Parmenides’ Likely Story
standable why no human would ever (ou mē pote) outstrip the young
man possessed of this account. On Owen’s reading, why should we
trust the way of doxa always to defeat competing appearances?
() The effect of combining the counterfactual chrēn with dokimōs
can be gauged only if we add in the clause dia pantos panta perōnta
(‘all pervading through everything’). Rather than simply under-
standing dokimōs as a complement to the subject, in which case
we might expect an adjective, we should see it as qualifying the
full complement: ‘should have been all (sc. ta dokounta) pervading
through everything’. The clause specifies the condition for the
beliefs’ being acceptable: only if we premiss ta dokounta on all
things pervading throughout will the beliefs be acceptable. The
counterfactual is used also because people have not based their
beliefs on this condition. This, then, is the bit that mortals do not
generally, but should, concern themselves with: how they ought to
have thought of things, considering them as pervading throughout
everything.
() The condition dia pantos panta perōnta anticipates Parmeni-
des’ own description of being later in the poem. By satisfying this
condition the dokounta will emerge as similar to being. So as we
shall see, the way of doxa will be premissed on plurality and change,
but will try to make its subject as similar as possible to the unitary
and changeless being described in the Way of Truth. Plurality is
anticipated in B by the plurals dokounta and panta, while perōnta
suggests the motion of everything passing through everything. The
passing through everything (dia pantos) will be seen to mirror the
saturation, homogeneity, and completeness of what is.
At the beginning of B the goddess says that there are very many
‘signs’ (sēmata) along the route of what is. ‘Signs’ is clearly a word
that fits well with the imagery of a route or path (hodos), and we
Mourelatos, Route, –, argues for the reading περ᾿ ὄντα, which he under-
stands as equivalent in force to ‘qua being’. The reading is less likely on linguistic
grounds, as one might expect the form ἐόντα. However, adopting it would rather re-
inforce the point that the δοκοῦντα are to be described in a way that strengthens their
similarity to being.
B . : πᾶν δ᾿ ἔμπλεόν ἐστιν ἐόντος; B . –: τετελεσμένον ἐστί | πάντοθεν; B . :
πάντοθεν ἶσον.
Thomas Kjeller Johansen
may understand the term in this context as a signpost or marker in-
dicating whether we are travelling the right way. This suggestion
does not, of course, exclude the possibility that the markers are also
properties of being. However, it seems possible in principle to dis-
tinguish between what it is to be a sign on the route and what it is
to be a property of being. For as markers the signs serve a further
epistemic function in relation to the enquirer, namely that of indi-
cating that he is travelling in a certain direction. As signposts they
tell us something about what we are looking for, and they offer us
information by which we can identify the object of our enquiry. It
is clear that the markers taken together identify nothing other than
what is. However, the importance of distinguishing the role of the
properties as markers is that one may allow, at least in principle,
that the markers are not just properties of being even if, by follow-
ing them all, one inevitably ends up at what is. Given that there is
a plurality of signs, ‘a great many’ of them as emphasized by the
goddess, it might be possible for another object to realize some of
the properties, even if not all. We need not exclude, then, the possi-
bility that we might find some of these properties also recurring as
signs along the route of doxa, even if not all the signs could do so.
It is after all not uncommon that at the beginning of a journey one
travels along a route that can lead to several destinations.
Here are the signs, as listed in B , in the order mentioned:
What is
(a) is ungenerated, imperishable (ll. , –);
(b) is whole, unperturbed, complete (ll. , );
(c) never was, will not be, is now (l. );
(d) is indivisible, all alike, and continuous (ll. –, );
(e) is changeless, motionless (ll. , );
(f) is steadfast, limited by Necessity (l. );
(g) has nothing else next [parex] to it (l. );
Cf. Mourelatos, Route, n. ; P. Curd, The Legacy of Parmenides: Eleatic
Monism and Later Presocratic Thought [Legacy] (; repr. Las Vegas, ),
n. , goes one step further and compares the σήματα to turnstiles.
Cf. Curd, Legacy, n. : ‘the signs in B function both as characteristics
of the natures that are reached by proper inquiry and regulating principles for that
inquiry’.
There are plenty of instances of this use of σῆμα in the recognition scenes in
Odyssey (e.g. the scar at line and the bed at line ).
Reading ἠδὲ τελεστόν in . . See Palmer, Parmenides, .
Parmenides’ Likely Story
(h) is complete and equal from every direction, ‘like a well-
rounded ball’ (ll. –, ).
The third path is one that combines what is and what is not. While
this is not a path that leads to true conviction, it is nonetheless
referred to as a ‘path’ (hodos/keleuthos) which is partly character-
ized by what is. It is reasonable, therefore, to expect that what is
For an analysis of the relationship between the σήματα see R. D. McKirahan,
‘Signs and Arguments in Parmenides B’ [‘Signs’], in P. Curd and D. W. Graham
(eds.), The Oxford Handbook of Presocratic Philosophy (Oxford, ), –.
I side with the majority of scholars in distinguishing three roads for the sorts
of reasons given by Palmer, Parmenides, –, and McKirahan, ‘Signs’. For an ar-
gument that there are only two roads see Curd, Legacy, –.
I retain Diels’s reading, against the suggestions of ἄρξω (A. Nehamas) or ἄρξει
(N.-L. Cordero). For helpful discussion of the options see Palmer, Parmenides, –.
Thomas Kjeller Johansen
described on the path to some extent will have the properties of the
signs.
The mortals who travel this path are described as utterly con-
fused. It is not clear whether this degree of confusion is entirely the
result of the road or whether it is also caused by the ignorance of the
people, that is all people so far, who travel it. There are indications
that part of the problem also lies with the mortals’ disposition. So,
we are told first, they wander the road ‘knowing nothing’. Simi-
larly, it is the helplessness in their breast which directs them, rather
than simply the contours of the route itself, and this helplessness
is explained by their reliance on perception. One might well won-
der, therefore, if knowing something, specifically knowing what is,
would not make a difference to how they would approach a path
characterized by what both is and is not. And correspondingly, one
might surmise that an approach not simply reliant on perception,
but based on logos, would allow them to make more headway. The
content of their thinking is next described as ‘that what is and what
is not are both the same and not the same’. This is clearly a way
of specifying the route, a route which is neither just of what is, nor
just of what is not, but combines in some manner the two. However,
the specification is perhaps unnecessarily befuddled, reflecting the
mortals’ general confusion when faced with a road that combines
two opposites. There may well be a way of understanding a route
that combines what is and what is not in a way that does not involve
one in a double self-contradiction, and it may well be an under-
standing available only to one such as the goddess, or now also the
young man, who has properly understood the nature of what is.
When we come to the goddess’s own cosmology, we shall see that
there is a way of talking about the world as both being and not be-
ing which is far less confusing. Finally, we should note that it is
not the path without qualification that is referred to as ‘backward-
turning’, but the path of all, again perhaps an indication that it is
the path as it has been travelled by mortals, which ultimately leads
nowhere.
One is reminded of the enigma in Republic , –, about the eunuch who
threw something at a bat, which is used by Socrates to illustrate how the object
of belief is what both is and is not. Behind the riddle there is a less confusing an-
swer.
κέλευθος, like ὅδος, can mean either the path or the journey; see LSJ s.vv.
Parmenides’ Likely Story
In Graham’s translation:
For they made up their minds to name two forms,
of which it is not right to name one—this is where they
have gone astray—
and they distinguished contraries in body and set signs
apart from each other: to this form the ethereal fire of flame,
being gentle, very light, everywhere the same as itself,
not the same as the other; but also that one by itself
contrarily unintelligent night, a dense body and heavy.
The mortals have posted signs of their own, characteristics that are
opposed to each other. However, there is no implication yet that
the mortals have posited these two forms as the fundamental prin-
ciples of everything. In B the goddess says that ‘humans laid
down a name (again katethent’), a mark for each’. Here the mistake
On the reading see Palmer, Parmenides, .
Following Diels’s bracketing: for metrical reasons one of the adjectives in this
line requires excision; cf. A. P. D. Mourelatos, ‘The Conception of eoikōs/eikōs as
Epistemic Standard in Xenophanes, Parmenides, and in Plato’s Timaeus’ [‘Concep-
tion’], Ancient Philosophy, (), – at .
Coxon, Fragments, , commenting on B . –, emphasizes the conventional
flavour of ἔθεντο: ‘the phrase sēmat’ easi contrasts with sēmat’ ethento [“assigned . . .
marks”] (l. ); the contrast indicates that, while the characteristics of the two Forms
into which P. analyses the physical world are, like the Forms themselves, empirical
and conventional in status, those of Being are objectively real.’
Thomas Kjeller Johansen
is similar but there is no indication that laying down a mark for
something makes it a basic principle. Clearly people, before and
after Hesiod’s Theogony, talked of night and day as different and
as opposites without necessarily constructing a cosmology on that
basis. It is the goddess who uses light and darkness as principles
of cosmology. There is no need, therefore, to hunt for any parti-
cular historical figures who have subscribed to these two as their
principles. Yet the adoption of light and darkness as principles
may still qualify as mortal opinions, since these are opposite attri-
butes that mortals recognize as real.
The goddess says that they are mistaken in this assumption. We
can see why from signpost (g) on the way of what is: what truly is is
one and unique, while mortals posit two contrary forms, separately
from each other (B . –). It is not possible simply on linguistic
grounds to determine whether the goddess thinks the mistake is
positing any of the two, as Cornford took it, or a particular one
of the two, say, darkness rather than light, as Popper argued, or
the other way around, as Aristotle read Parmenides (Metaph. Α ,
b–a). What seems to clinch the case in favour of Corn-
ford is the parallel ways in which the goddess goes on to describe
the two. The mistake, then, relative to the way of Truth, lies in
positing two forms (and after all the goddess does not tell us what
they are until after she has pointed to the mistake). Either form can
be viewed as being in its own right but as not-being in relation to
the other. The significant aspect is that they are opposites (antia,
, repeated ), and from this point of view there is no reason to
single out one rather than the other as the odd one out.
What is striking, however, is that having indicated the mistake,
the goddess proceeds to describe each of the two forms in ways
which suggest that they in some manner satisfy some of the signs
along the way of being. So she says of fire that it is ‘the same as it-
Contrast Curd, Legacy, –.
F. M. Cornford, ‘Parmenides’ Two Ways’ [‘Two Ways’], Classical Quarterly,
(), – at –.
K. R. Popper, ‘How the Moon Might Throw Some of her Light upon the Two
Ways of Parmenides’, Classical Quarterly, (), – at .
A further option, suggested by Coxon, Fragments, , is the mistake of taking
just one of these as the cosmic principle.
It is also compatible with Coxon’s reading, it should be added.
Even G. Vlastos, ‘Parmenides’ Theory of Knowledge’, Transactions and Pro-
ceedings of the American Philological Association, (), – at , concedes
that ‘[Parmenides’] grounding in Ionian physics got the better of his contempt for
Parmenides’ Likely Story
self in all directions [ἑωυτῶι πάντοσε τωὐτόν], but not the same as the
other’, while the opposite, night, is also ‘by itself’ (κατ᾿ αὐτό). The
corresponding sign (h) on the way of being earlier in B had been
‘equal to itself in every direction’ (πάντοθεν ἶσον). In fragment she
says that ‘all is full at once of light and dark night’, but again qua-
lifies this by saying that ‘both are equal, since neither has any share
in what is not’. Here ‘all is full at once’ recalls sign (d), ‘but it is all
full of what is’ (πᾶν δ᾿ ἔμπλεόν ἐστιν ἐόντος) at B . . While there
are two principles, they are present equally in the entire universe.
It is not clear whether this means that they are present equally all
through the universe or whether they are equally represented in the
universe. The alternating circles of light and dark described in A
may favour the second option.
However we read B here, it seems clear that the two principles
individually resemble what is on signs (h) and (d). The mortals go
wrong in posting signs indicating a plurality of beings, when we
know from the way of what is that what is is one and unique. We
have according to the mortals two beings, light and night, though
their manner of being, that of opposites, also implicates them in a
particularly conspicuous way in not-being. One being is the same as
itself—fire is the same as itself—yet it is also different from another
being, night, which itself is the same as itself. By being the opposite
of the other, there is a particular conceptual relation between being
fire and not being night, or vice versa. As opposites, they are ne-
cessarily mutually exclusive, so being qualified by one opposite ne-
cessarily means being different from the other. Put more formally,
where O and O are opposites, for all X, if X is O, then, neces-
sarily, X is different from O. We see, then, how it was appropriate
for the goddess to charge the mortals with holding that ‘being and
not being are the same and not the same’ (B . –).
It is not just individually but also in their combination with each
the mock-world of the senses, and he gave to his doctrine of Being a physical appli-
cation, attributing the self-identity of Being to each component of the “deceitful”
duality of Becoming.’
Adopting the translation of Kirk, Raven, and Schofield, Presocratic. Cf. Finkel-
berg, ‘Being’, : ‘In the two last lines of the fragment Parmenides conspicuously
resorts to the terminology of the Aletheia. The phrase πᾶν πλέον ἐστιν ὁμοῦ φάεος καὶ
νυκτός, echoing as it does πᾶν δ᾿ ἔμπλεόν ἐστιν ἐόντος at fr. . , describes the totality
of the two “forms” as “fullness,” while the statement of their equal “filling ability”
portrays it as uniform in its degree.’
On the opposites as ‘enantiomorphic’ see Curd, Legacy, –.
Thomas Kjeller Johansen
other that the two principles bear similarities to what is. So the
sphericity of the cosmos recalls the comparison of what is to a well-
rounded sphere, sign (h). In both cases, being and the cosmos,
the spherical shape is determined by a notion of Necessity. So in
B it is the surrounding heavens whose growth and limits are de-
termined by Necessity, while Necessity in B . , kept what is
within bounds. It is tempting to think that these must be different
notions of Necessity: one in some sense logical (corresponding per-
haps to krinai logōi at B . ), the other in some sense physical or
material. But however exactly we conceptualize these necessities,
we are surely supposed to think of the cosmic order as resembling
the Necessity that governs what is, according to sign (f). Within
the cosmos, we find circularity again at the centre of the world and
in the alternating circles of light, darkness, and their mixture (A ),
as well as in the shape of individual planets such as the ‘round-eyed
moon’ (B ). Another possible reminiscence of being is the posi-
tion of the earth. The earth, like being itself, is unmoving because
it has no reason to incline one way rather than another (A ). The
principle of sufficient reason was used to justify sign (a): as there
was no need for what is to have come into being later or earlier from
what is not, it has not come into being at all. The principle thus
seems equally applicable to being and the cosmos. This suggests
again an attempt to maximize the intelligibility of the cosmos by
the standards that also apply to being. All seem clear indications
that Parmenides models the cosmos, as a whole and in its parts, on
being.
The observation of D. N. Sedley, ‘Parmenides and Melissus’, in A. A. Long
(ed.), The Cambridge Companion to Early Greek Philosophy (Cambridge, ), –
at , that the words μεσσόθεν ἰσοπαλὲς πάντῃ fall outside the scope of the com-
parison (which is in the dative), is significant evidence in favour of a literal spatial
interpretation. On the other hand, if Parmenides literally meant that being was a
sphere, why say that it is like a sphere (cf. Eudemus fr. Wehrli)?
See Palmer, Parmenides, . While I am sympathetic to Palmer’s modal read-
ing, I think it remains an issue for him to explain just how the Necessity that rules
the cosmos is compatible with the contingent status of the cosmos.
As observed by Cornford, ‘Two Ways’, , ‘The geometrical Sphere fettered
by Necessity in the bonds of its mathematical limit (circumference) has now become
the Sphere of the visible Heaven, that will be fettered by Necessity to hold the limits
of the visible fiery stars.’
B . –: τί δ᾿ ἄν μιν καὶ χρέος ὦρσεν | ὕστερον ἢ πρόσθεν, τοῦ μηδενὸς ἀρξάμενον,
φῦν;
Parmenides’ Likely Story
Tim. –.
Shape: Tim. –; time: –. For this interpretation see T. K. Johansen,
Plato’s Natural Philosophy: A Study of the Timaeus–Critias (Cambridge, ),
–.
Parmenides’ Likely Story
in the mixture of several ingredients, changelessness lies in the Ne-
cessity that regulates the movements of the heavens. In this way the
cosmos emerges as a likeness of being.
If this is right, it is tempting to understand eoikota in the god-
dess’s phrase diakosmon eiokota positively in the sense of ‘likely’.
One might even, as was suggested by Wilamowitz, take the term
as anticipating Plato’s application of eikōs in the Timaeus to cosmo-
logy. Timaeus applies the term to an account that shows the cosmos
to be a likeness of true reality, the eternal Forms. Likelihood is then
a positive value for cosmology to aim at: the better the account re-
presents the world as a likeness, the more likely it is. Plato may well
have taken a leaf from Parmenides’ use of eoikōs if he understood
Parmenides’ being as the reality which the cosmos is shown to re-
semble in the way of doxa. I shall return to this comparison at the
end of the paper.
. Deception
ἅτε οὐδὲν εἰδότες ἀκριβὲς περὶ τῶν τοιούτων, οὔτε ἐξετάζομεν οὔτε ἐλέγχομεν τὰ
γεγραμμένα, σκιαγραφίᾳ δὲ ἀσαφεῖ καὶ ἀπατηλῷ χρώμεθα περὶ αὐτά.
Cf. Lesher, ‘Early’, , and Palmer, Parmenides, : ‘Mortal notions, by con-
trast, are not trustworthy and the goddess’s account of their objects is “deceptive”
(fr. . ) because apprehension of these objects varies just as they themselves do.
Her account is also deceptive because the constant and ready availability to per-
ception of the things it describes has misled ordinary mortals to suppose that such
things are in fact all there is and thus inevitably that they are the most worthy ob-
jects of their attention. Calling her cosmology “deceptive” is another way of warning
Parmenides that ultimately he must fix his thoughts upon what must be.’
e.g. Curd, Legacy, .
Parmenides’ Likely Story
to being. The cosmology is both likely and deceptive, eoikōs in
both its aspects, exactly by its likeness to the way of Truth. Rather
like Gorgias’ intelligent theatre audience, the young man will have
understood the cosmology well to the extent that he is tempted to
be deceived.
. Perception
One might think that this reading sits badly with the role of percep-
tion in deceiving us. Parmenides offers a general account of where
mortals go wrong in B . –:
ἀλλὰ σὺ τῆσδ ᾿ ἀφ ᾿ ὁδοῦ διζήσιος εἶργε νόημα
μηδέ σ ᾿ ἔθος πολύπειρον ὁδὸν κατὰ τήνδε βιάσθω,
νωμᾶν ἄσκοπον ὄμμα καὶ ἠχήεσσαν ἀκουήν
καὶ γλῶσσαν, κρῖναι δὲ λόγωι πολύδηριν ἔλεγχον
ἐξ ἐμέθεν ῥηθέντα.
But you, withhold your thought from this way of enquiry,
nor let habit born of long experience force you along this way,
to wield an aimless eye and echoing ear
and tongue. But judge by reasoning the very contentious examination
uttered by me. (trans. after Graham)
Nor is there any reason to think that Aristotle is ascribing to Parmenides the
same reason for being ‘forced to follow appearances’ that he would himself subscribe
to, e.g. providing the evidence that our scientific theories need to explain.
Like others (cf. H. Granger, ‘The Cosmology of Mortals’, in V. Caston and
D. W. Graham (eds.), Presocratic Philosophy: Essays in Honour of Alexander Moure-
latos (Aldershot, ), – at –), I am sceptical of Aristotle’s association
of specifically earth with not-being. What Parmenides may simply have meant to
convey was that the two opposites as such could both be characterized as being and
not being, viz. being whatever it itself is and not being the other.
Parmenides’ Likely Story
des’ own account of thinking in B , which, if we may trust Theo-
phrastus (A ), is also his account of perceiving:
ὡς γὰρ ἕκαστοτ᾿ ἔχει κρῆσιν μελέων πολυπλάγκτων,
τὼς νόος ἀνθρώποισι παρέστηκεν· τὸ γὰρ αὐτό
ἔστιν ὅπερ φρονέει μελέων φύσις ἀνθρώποισιν
καὶ πᾶσιν καὶ παντί· τὸ γὰρ πλέον ἐστὶ νόημα.
For as on each occasion, he says, is the blending of
the much-wandering limbs,
so is thought present to humans. For the same thing
is that which thinks, the nature of the limbs, in humans
in both each and every one: for the greater is thought.
(trans. Palmer, Parmenides, , slightly revised)
I have suggested that there is a likeness between being and the phe-
nomena. But is this likeness a mere similarity or are the phenomena
See further A. Laks, ‘“The More” and “the Full”: On the Reconstruction of
Parmenides’ Theory of Sensation in Theophrastus, De sensibus, –’, Oxford Studies
in Ancient Philosophy, (), –.
Text from Palmer, Parmenides, .
For an illuminating analysis of this idea see E. Hussey, ‘Parmenides on Think-
ing’, in R. King (ed.), Common to Body and Soul: Philosophical Approaches to Ex-
plaining Living Behaviour in Greco-Roman Antiquity (Berlin and New York, ),
–.
Thomas Kjeller Johansen
as they are because they thereby resemble being? It has been sug-
gested that being is related to the perceptual phenomena rather in
the manner of an Aristotelian substance to its accidents. In this
way the phenomena could be the way they are because they are caus-
ally determined by an underlying substance. The problem with this
suggestion in Parmenides is that it is hard to see any causal relation-
ship between what is and the phenomena. There is nothing, after
all, about being itself to cause things to come into being or pass
away.
In the Timaeus the phenomena resemble being because of the ef-
forts of a creator god, who made the world resemble being. One
could think of the relationship between phenomena and being in
the same way: the cosmos has been made to resemble what is by
a cause external to what is. But there is no direct mention of this
kind of cause in Parmenides. There is, perhaps, a textual path which
opens up for such an interpretation. So B . – refers to a goddess
(daimōn) in the middle of the cosmic rings, ‘who governs [kubernai]
all things: for everywhere she rules over hateful birth and mixture’.
The language is clearly consistent with an organizing god. Again,
in B the goddess is said to have ‘devised’ Erōs, Parmenides us-
ing a word (mētisato) which might equally be translated ‘crafted’.
The relationship between this goddess and the goddess who is nar-
rating the cosmology is obscure. My own suggestion would be that
the cosmic goddess as herself internal and specific to the cosmos,
and so subject to change and diversity, is not the same as the god-
dess of B . Yet one might view the one as a likeness of the other: the
cosmic goddess in her ordering and necessitating of the phenomena
imitates the goddess presiding over the way of Truth. The cosmic
goddess would then be causally responsible for the order of the cos-
mos, and so as a cause also for the similarity between the cosmos
and what is. It would be the cosmic goddess’s intellectual efforts
(her steering and her crafting) that were the cause of the similarity
Discussed and rightly rejected by Palmer, Parmenides, –.
As Cosgrove, ‘Fresh Look’, , observes, ‘Parmenides nowhere suggests that
to eon is responsible for or has somehow produced the phenomenal world of change
that mortals have named.’
Nor are the phenomena themselves characterized as accidental if this is meant
to contrast with their being necessary: the goddess takes them to be regulated by a
kind of necessity, as we saw (B ).
The choice of a specifically female cosmic goddess might strengthen the impres-
sion of a deliberate similarity to the goddess of the proem. Plato’s Demiurge returns
to the Xenophanean (B –) male template.
Parmenides’ Likely Story
between phenomena and being. In so far as she is a god, not a hu-
man, we might think that she has access to the same knowledge of
being as the goddess of the proem. In that case she would have the
tools to arrange the world so as to be a likeness of being. However,
again, there is no evidence in the text that the internal goddess de-
ploys her knowledge in ordering the cosmos. So while there may be
evidence that the cosmic goddess works so as to create a particular
order and that being so ordered makes the cosmos similar to being,
there is no evidence that she has worked the cosmos in this way in
order to make it similar to being.
. Conclusion
It seems prudent, then, to refrain from the claim that the cosmos
resembles being because it has been so made. If we require from a
likeness that it has come about in order to resemble its model, we
should not say that the cosmos is a likeness of being. However, if
we lift this causal requirement, we may say that the cosmos is a like-
ness of being in so far as there is a formal similarity, on some of the
points of the signs, between the cosmos and being. Moreover, this
similarity is clearly not a merely accidental similarity from the point
of view of the poem, and so from the point of view of explaining why
Parmenides constructs the cosmos the way he does. We may distin-
guish, after all, a causal claim (a) about the cosmos from one (b) about
the cosmology, i.e.
(a) The cosmos has been made by the maker of the cosmos as a
likeness of being,
for which there is insufficient evidence in the text of Parmenides,
and
(b) The cosmos has been described as a likeness of being,
which we can, I have argued, justifiably attribute to Parmenides.
(a), then, seems to be the distinctly Platonic innovation: the cause of
the cosmos is a Demiurge who looks to being as a paradigm and cre-
ates the cosmos as a maximal likeness of it. (b) for Plato is grounded
in (a): the world is to be described as a likeness of being because it
was made to resemble it. However, Parmenides proffers us only
See, in particular, Tim. –, where Timaeus explains that we should aim at
a likely account of the cosmos because it has been made as a likeness of being.
Thomas Kjeller Johansen
(b). We may press the similarity somewhat further, since both Plato
and Parmenides maintain
(c) The cosmos is intelligible to the extent that it resembles being
and so also
(d) Cosmology is successful to the extent that it shows the cos-
mos as like being.
But Parmenides again has a different basis for maintaining (d) from
Plato’s: not (a), that the cosmos has come into being as a likeness,
but rather
(e) Any subject is intelligible to the extent it satisfies the signs on
the way of Truth.
Since only being satisfies all of these signs, any subject that is intel-
ligible will, to that extent, resemble being. For Parmenides, then,
the interest in showing the likeness of the cosmos to being comes,
as far as we can tell, out of general conditions of intelligibility rather
than a specific view about the genesis of the cosmos.
The absence of a causal story about how the cosmos comes to
resemble being may give heart to those who would take the cosmo-
logy to be a fiction: if there is no account of how the cosmos derives
being from what really is, perhaps there is no reason to think that
the cosmos has any degree or manner of being. Instead, one might
think that Parmenides has described a cosmos that resembles being
and accommodates our appearances, and so offers a degree of intel-
ligibility, but there is no reason to ascribe any reality to the object
of this story. However, as I argued earlier, the signs on the way
of Truth are not just criteria of intelligibility. Since what is there
for thinking is also there for being, and what is there for thinking is
measured by the extent to which it matches the signs, it is reason-
able to say that the cosmos in so far as it matches these signs also
possesses a degree of being. The fictional reading seems to stumble,
then, on the match between intelligibility and reality on which Par-
menides insists.
Proclus chose to quote Parmenides as saying the same thing as
Plato at a particular point in his commentary, the point ( –)
where Timaeus presents the world as a generated likeness of an
I am grateful to an anonymous reader for stressing this option.
Parmenides’ Likely Story
eternal model, and where he explains the sort of cosmology we may
accordingly aim for. He does not quote Parmenides B at the earlier
stage ( – ) at which Timaeus simply contrasted being with
becoming tout court, assigning knowledge to the first and mere be-
lief with irrational perception to the second. If the argument of this
paper is right, Proclus chose the opportune moment for the quota-
tion: for Parmenides too saw the cosmos as intelligible in the mea-
sure that it resembles being, and he set his cosmology the task of
showing just this likeness. That Proclus overstated the point—Plato
went important steps further in explaining how the likeness came
about—should not distract us from the insight that Parmenides laid
the foundations, not just for Plato’s metaphysics, but also for his
cosmology.
Brasenose College, Oxford
B I B L I O G R A P HY
Betegh, G., ‘What Makes a Myth eikôs? Remarks Inspired by Myles
Burnyeat’s Eikôs Mythos’, in R. D. Mohr and B. M. Sattler (eds.), One
Book, the Whole Universe: Plato’s Timaeus Today (Las Vegas, ),
–.
Brown, L., ‘The Verb “To Be” in Greek Philosophy’, in S. Everson (ed.),
Language, Companions to Ancient Thought, (Cambridge, ),
–.
Burnyeat, M. F., ‘Εἰκὼς μῦθος’, Rhizai, (), –.
Bryan, J., Likeness and Likelihood in the Presocratics and Plato [Likeness]
(Cambridge, ).
Cherubin, R., ‘Light, Night, and the Opinions of Mortals: Parmenides B .
– and B ’, Ancient Philosophy, (), –.
Cornford, F. M., ‘Parmenides’ Two Ways’ [‘Two Ways’], Classical
Quarterly, (), –.
Cosgrove, M. R., ‘What Are “True” Doxai Worth to Parmenides? Essay-
ing a Fresh Look at his Cosmology’ [‘Fresh Look’], Oxford Studies in
Ancient Philosophy, (), –.
Coxon, A. H., The Fragments of Parmenides [Fragments], rev. edn. (Las
Vegas, ).
Curd, P., The Legacy of Parmenides: Eleatic Monism and Later Presocratic
Thought [Legacy] (; repr. Las Vegas, ).
Diels, H., and Kranz, W. (eds.), Die Fragmente der Vorsokratiker, th edn.,
vols. (Berlin, –).
Thomas Kjeller Johansen
Finkelberg, A., ‘Being, Truth and Opinion in Parmenides’ [‘Being’], Ar-
chiv für Geschichte der Philosophie, (), –.
Graham, D. W., Explaining the Cosmos: The Ionian Tradition in Scientific
Philosophy [Explaining] (Princeton, ).
Graham, D. W., The Texts of Early Greek Philosophy: The Complete Frag-
ments and Selected Testimonies of the Major Presocratics, pt. (Cam-
bridge, ).
Granger, H., ‘The Cosmology of Mortals’, in V. Caston and D. W. Graham
(eds.), Presocratic Philosophy: Essays in Honour of Alexander Mourelatos
(Aldershot, ), –.
Guthrie, W. K. C., A History of Greek Philosophy, ii. The Presocratic Tra-
dition from Parmenides to Democritus (Cambridge, ).
Hussey, E., ‘Parmenides on Thinking’, in R. King (ed.), Common to Body
and Soul: Philosophical Approaches to Explaining Living Behaviour in
Greco-Roman Antiquity (Berlin and New York, ), –.
Hussey, E., The Presocratics (New York, ).
Johansen, T. K., Plato’s Natural Philosophy: A Study of the Timaeus–
Critias (Cambridge, ).
Kahn, C. H., ‘A Return to the Theory of the Verb Be and the Concept
of Being’, Ancient Philosophy, (), –; repr. in id., Essays on
Being (Oxford, ), – (cited from the reprint).
Kirk, G. S., Raven, J. E., and Schofield, M., The Presocratic Philosophers
[Presocratic], nd edn. (Cambridge, ).
Laks, A., ‘ “The More” and “the Full”: On the Reconstruction of Parme-
nides’ Theory of Sensation in Theophrastus, De sensibus, –’, Oxford
Studies in Ancient Philosophy, (), –.
Lesher, J. H., ‘Early Interest in Knowledge’ [‘Early’], in A. A. Long (ed.),
The Cambridge Companion to Early Greek Philosophy (Cambridge, ),
–.
Lesher, J., ‘Parmenides’ Critique of Thinking: The poludēris elenchos of
Fragment ’, Oxford Studies in Ancient Philosophy, (), –.
McKirahan, R., ‘Signs and Arguments in Parmenides B’ [‘Signs’], in P.
Curd and D. W. Graham (eds.), The Oxford Handbook of Presocratic
Philosophy (Oxford, ), –.
Mourelatos, A. P. D., ‘The Conception of eoikōs/eikōs as Epistemic Stan-
dard in Xenophanes, Parmenides, and in Plato’s Timaeus’ [‘Concep-
tion’], Ancient Philosophy, (), –.
Mourelatos, A. P. D., ‘The Epistemological Section (b–d) of the Proem
in Timaeus’ Speech: M. F. Burnyeat on eikôs mythos, and Comparison
with Xenophanes B and B’, in R. D. Mohr and B. M. Sattler (eds.),
One Book, the Whole Universe: Plato’s Timaeus Today (Las Vegas, ),
–.
Parmenides’ Likely Story
Mourelatos, A. P. D., The Route of Parmenides: A Study of Word, Image,
and Argument in the Fragments [Route], rev. edn. (Las Vegas, ).
Owen, G. E. L., ‘Eleatic Questions’, Classical Quarterly, (),
–.
Palmer, J. A., Parmenides and Presocratic Philosophy [Parmenides] (Oxford,
).
Palmer, J. A., Plato’s Reception of Parmenides [Reception] (Oxford, ).
Popper, K. R., ‘How the Moon Might Throw Some of her Light upon the
Two Ways of Parmenides’, Classical Quarterly, (), –.
Reinhardt, K., Parmenides und die Geschichte der griechischen Philosophie
[Parmenides] (Bonn, ).
Sedley, D. N., ‘Parmenides and Melissus’, in A. A. Long (ed.), The Cam-
bridge Companion to Early Greek Philosophy (Cambridge, ), –.
Vlastos, G., ‘Parmenides’ Theory of Knowledge’, Transactions and Pro-
ceedings of the American Philological Association, (), –.
Wilamowitz-Möllendorf, U. von, ‘Lesefrüchte’, Hermes, (), –.
TH E M O R A L A N D LIT ERARY
CHA R A C T E R O F H IPPIAS IN
PL A T O ’ S H I P P I A S MA JOR
FRANCO V. TRIVIG N O
[him] in anything’ ( –) and proclaims himself ‘wisest in the greatest number
of crafts’ ( –).
Though there is much dispute about the significance of Socrates’ use of irony,
my aims here are rather narrowly focused on the figure of the ironist as he relates
to an impostor, and I will not discuss the larger questions. Regarding the irony
of Socrates’ claims or statements, I think the safest, most conservative position on
the extent of Socrates’ use of irony would acknowledge that his expressed praise
of his interlocutors as already wise is ironic, at least in the sense of ‘not seriously
meant’. See G. Vlastos, ‘Socratic Irony’, Classical Quarterly, (), –;
id., Socrates, Ironist and Moral Philosopher (Ithaca, NY, ), –; I. Vasiliou,
‘Conditional Irony in the Socratic Dialogues’, Classical Quarterly, (),
–.
For an excellent and very useful account of Socrates’ use of ironic praise see
A. Nightingale, ‘The Folly of Praise: Plato’s Critique of Encomiastic Discourse in
Lysis and Symposium’, Classical Quarterly, (), –, esp. –.
Since the vast majority of our information about the historical Hippias comes
from Plato’s two dialogues, I am bracketing questions about the historical Hippias
The Character of Hippias in the Hippias Major
It seems clear that Plato thinks that sophists form a kind or type and
that they all share the features of the professorial impostor; and yet
he characterizes them differently (cf. the differences between Pro-
tagoras, Hippias, and Prodicus in the Protagoras). It seems impor-
tant in this dialogue that Hippias’ alleged wisdom is highly broad
and general and that Hippias is visually impressive. These are spe-
cific features of Hippias that find a place in the larger construct
of the professorial impostor. The dialogue focuses on one espe-
cially crucial and philosophically significant aspect of Hippias’ self-
ignorance: he thinks he knows what the fine is, but he does not; and,
on this basis, he thinks that he is himself fine, though he is not. If
this is right, then Hippias is a particular kind of fool, a self-ignorant
one, and, for all his technical knowledge—and there is no reason to
deny him this—he is still a laughable figure fit for ridicule.
Of course, if the dialogue’s entire aim were merely to mock and
humiliate Hippias—if this were only comedy—it would make for a
philosophically uninteresting dialogue. Rather, through the char-
acter of Hippias, Plato aims to expose for his readers two par-
ticularly pernicious and common mistakes or confusions about the
fine that Hippias both makes and embodies. This is one crucial
way in which Plato’s comedy goes beyond Old Comedy: whereas
the latter’s use of the impostor device may be, in certain prominent
cases, seriously concerned with exposing public figures as frauds
and the historicity of Plato’s portrayal. On the historical Hippias see W. K. C. Guth-
rie, A History of Greek Philosophy, iii. The Fifth Century Enlightenment [Fifth Cen-
tury] (Cambridge, ), –; Woodruff, Hippias, –; D. Nails, The People of
Plato: A Prosopography of Plato and Other Socratics [People] (Indianapolis, ),
s.n. ‘Hippias of Elis’. Xenophon refers to Hippias in Xen. Symp. . , and a frag-
ment from Proclus (B DK) attributes the discovery of a curve for trisecting an
angle to Hippias. This latter piece of evidence is less certain: Nails takes it to refer to
our Hippias, but OCD, s.n. ‘Hippias ()’, claims that it is a mistake to understand
the fragment as referring to Hippias of Elis.
Kahn, ‘Beautiful’, , cannot understand what purpose Plato could have in
portraying Hippias with such animosity, since he would be ‘safely dead, or at least
extremely aged’ by the time of composition. Surely Plato’s portrayal of sophists and
other opponents of philosophy is not primarily to undermine those th-cent. con-
temporaries of Socrates, but rather to reveal for his th-cent. audience the errors of
those th-cent. figures.
Cf. Woodruff, Hippias, –; Kahn, ‘Beautiful’, . I hope, in what follows,
to provide a more specific ‘philosophical motivation’ for the attack on Hippias than
what Woodruff provides. Though my account is consistent with the five general rea-
sons he gives for an attack on Hippias, I focus on two importantly mistaken claims
that Hippias endorses in the dialogue.
I have in mind here Aristophanes’ targeting of, for example, Creon and Soc-
Franco V. Trivigno
and warning the audience about them, it is not so concerned with
diagnosing the source of the figure’s foolishness or in general with
the moral character of the fool. In short, Plato puts a device from
Old Comedy to properly philosophical use, and in Section I will
highlight two interesting ways in which Plato adapts Old Comedy
to his purposes.
Given that Hippias understands the definition of the fine as the appropriate in
terms of appearance here, it is hardly right to say, as commentators often do, that
he himself suggested it earlier in the dialogue ( –). In that case, he clearly
implicitly intended the appearance-based conception he openly endorses here, and
this is obviously not what Socrates has in mind when he mentions the ‘nature of the
appropriate’ ( –). See H. Olson, ‘Socrates Talks to Himself in Plato’s Hippias
Major’ [‘Talks to Himself’], Ancient Philosophy, (), – at , and my
discussion in sect. below.
The Character of Hippias in the Hippias Major
For example, what all large things are large by is the exceeding. For by that
all large things are large, and even if they do not appear so, if they exceed,
they are necessarily large. Similarly, we say the fine is what all fine things
are fine by whether or not they appear fine. ( –)
Faced with this response, Hippias attempts to close the gap between
appearing and being fine by positing the appropriate as the common
cause of both: ‘the appropriate makes things both be fine and appear
fine, when it is present’ ( –). It is not immediately clear what
Hippias thinks this clarification means for the relationship between
being and appearing fine or how it solves the problem of deceptive-
ness. Perhaps Hippias takes his claim to imply that something is
fine if it appears fine, that is, to license inferences from the appear-
ance of fineness to its reality. Though this inference is not logically
entailed by Hippias’ claim, it would obviate the problem of decep-
tiveness, and indeed would make the identification of fine people a
relatively straightforward matter—whoever appears fine is fine.
Socrates does not pursue the problem of deceptiveness and rather
focuses on whether Hippias’ assertion licenses the inference in the
other direction, i.e. implies that something appears fine if it is fine.
Nor is this inference entailed by Hippias’ initial claim, but Hippias
nevertheless agrees that ‘it [is] impossible [ἀδύνατον] that what really
is fine should not appear to be fine when what makes it appear [fine]
is present’ ( –). The assertion is not obviously wrong, and
this can easily be seen by translating kalos as ‘beautiful’—what is
appropriate makes things beautiful, and if something is beautiful,
it appears beautiful. This is an instructive mistake, since it precisely
calls attention to an ambiguity in kalos and an ensuing problem for
someone who takes it in a visual or aesthetic manner: she cannot see
the gap between being and appearing fine. Indeed, Hippias may be
seen to overlook the gap when, earlier in the dialogue, he is to give an
account of the fine ‘that will never appear [φανεῖται] foul anywhere
to anyone’ ( ) and what he offers is an account of ‘what is [εἶ-
ναι] always finest everywhere for everyone’ ( ). Bracketing
the move from foul to fine, the shift in formulation from appearing
to being suggests that Hippias does not clearly mark the difference
between such expressions. In order to correct for this, Socrates
If I am right, then it seems that Hippias wrongly understands the datives in the
two phrases—‘to anyone’ (μηδενί) and ‘for everyone’ (παντί)—to play identical refer-
ential roles in their respective sentences. However, it would be more natural to take
the second dative as one of advantage, and this grammatical point reflects the phi-
Franco V. Trivigno
adduces the counter-examples of ‘fine customs and practices’ (καλὰ
καὶ νόμιμα καὶ ἐπιτηδεύματα), which do not inevitably appear to be
fine always and to everybody ( – ). This moves the level
of discussion to examples of moral improvement that resist an aes-
thetic interpretation. Such an example is clearly fair game since
Hippias has recently given a speech about the ‘fine practices’ (ἐπι-
τηδευμάτων καλῶν, –; cf. ) and ‘very fine customs’
(νόμιμα . . . πάγκαλα, –) that Nestor recommended to Neo-
ptolemus, though it is certainly revealing that Hippias praised that
speech mainly for its aesthetic qualities ( –).
Further, Socrates brings out more clearly that appearing fine is al-
ways appearing fine to someone. There is always a relevant audience,
and this is congenial to the views of an impressive performer such
as Hippias. Again, translating kalos as ‘beautiful’ reveals why Hip-
pias’ emphasis on appearance, along with the attendant need for an
audience, is not obviously wrong: one might think that beauty re-
quires an audience. One might even think, along these lines, that the
perception of beauty is what grounds knowledge of beauty. How-
ever, as Socrates shows, what causes something to appear fine and
what causes something to be fine are different ( – ). This is
at least in part because appearing fine is always indexed to an audi-
ence’s perception, while being fine is grounded in the intrinsic qua-
lities of the object and independent of anyone’s perception. This is
not to deny that the fine may be relational in the way that appropri-
ateness is relational: it may be fine for a virtuous person to receive
praise and fine for a vicious one to receive punishment. Rather, it
is to underscore that, like largeness, fineness may be relational and
indeed depend on some extrinsic factors, and yet not be subjective
or perceiver-relative. Of course, it may be that Socrates thinks that
being fine causes things to appear fine only to those with knowledge,
such that being fine may play a causal role in some cases of appear-
ing fine and certain informed perceivers may set the standard for
losophical point that the first formulation places humans in the role of determiners
of what is fine, whereas the second make the fine something independently benefi-
cial for humans. Hippias thus misses one of the core implications of the difference
between expressions about what seems fine and what is fine.
Cf. Sym. –, where ‘laws and practices’ that aim at moral improvement
are the next step in the ladder of love. Diotima says that, at this stage, one thinks that
‘the beauty of bodies is a thing of little importance’ ( –). The range of things
that are described as fine is also an issue in Phaedo –, where participation in
the Form is invoked to explain all cases.
The Character of Hippias in the Hippias Major
the rest of us. This is to ground the perception of fineness in the
knowledge of fineness.
In sum, thinking of to kalon in an aesthetic manner leads one to
overlook the gap between appearing and being fine and to make
fineness itself audience- or perceiver-relative. But Hippias has
trouble seeing his mistakes. Even after Socrates’ not so subtle hint
to re-examine his initial response, Hippias still clings to apparent
fineness as decisive for the appropriate, and is puzzled that this
wrecks the argument ( ). Hippias still wants to answer a dif-
ferent question, the one he really cares about, what it is to appear
fine. He holds on, I suggest, because the aesthetic picture and its
attendant claims express his core commitments, and indeed are a
source for his self-esteem. If Hippias is right in not seeing a gap
between looking and sounding, i.e. appearing, fine and being fine,
then he is also right to think that he is a fine human being. On
Hippias’ view, it seems that, if you want to know whether someone
is fine or not, you can tell by looking at them to see if they have the
outward marks of fineness—good looks, nice clothes, gold jewel-
lery, etc. If you want to know whether what someone says is fine,
you have to listen to the rhetorical flourish of the composition. In
sum, the visual and auditory fineness of Hippias—not the content
of his character or of the speeches—is what marks him as fine. He
knows this, or so he thinks, because the audience tells him he is
fine and they are, in the end, the standard for measuring fineness.
Of course, Hippias is completely wrong about the fine and thus
about himself: indeed, it is not hard to see Hippias’ example of the
ridiculous man in fine clothes as unintentionally self-referential.
One might wonder how single-minded concern for the truth is compatible with
ironic praise and self-deprecation. A short response might be to see the latter pair
as strategies for attaining the truth and for enabling an arrogant and self-ignorant
interlocutor to attain truth. Cf. Olson, ‘Talks to Himself’, –.
Franco V. Trivigno
better ( – ; cf. – ); it is in these latter terms that
Socrates describes his own ‘political art’ ( –). One might
be further reminded of the Platonic theory of punishment, which
makes ‘pain and suffering’ necessary for moral reform (Gorg.
–).
R. Blondell, The Play of Character in Plato’s Dialogues (Cambridge, ),
–.
Kahn admits that ‘this faithful and intelligent imitation’ could only have been
written by an ‘intelligent student of Plato’s work’—someone like Vlastos—but never-
theless thinks that someone who takes the Hippias Major to be genuine ‘has no sense
for the art of the greatest prose writer of antiquity’ (‘Beautiful’, ). My claim, a
much more modest one, is that those who take the dialogue to be spurious have no
sense for the art of Plato’s comedy.
See Trivigno, ‘Technē’.
Franco V. Trivigno
it comes to ethical and political matters. One core difference
between them is that, while Ion’s only basis for taking himself
to have practical expertise is his knowledge of Homer, Hippias
seems to have some real independent knowledge. Never in the
Hippias Major is Hippias’ knowledge of mathematics, harmonics,
and astronomy questioned either explicitly or implicitly. On
the basis of this fact alone, one may conclude that Hippias is not
stupid, since he possesses knowledge, and in particular knowledge
of fields that would seem to require some level of intelligence.
But one cannot conclude from this that he is not a fool, since
even a knowledgeable and intelligent person may also be self-
ignorant. One might concede that Hippias is not a complete fool,
since his lack of self-knowledge is restricted to ethical and political
matters, unlike Ion, who thinks himself an expert in prophecy
and generalship, for example, on the basis of having memorized
Homer. Nevertheless, Hippias is a fool in the only way that really
matters, since, like the craftsman in the Apology, he has real know-
ledge but makes the mistake of thinking that his knowledge in
one area extends to knowledge of virtue; Ion’s case is arguably
more like that of the poets, and if that is right, the differences
between Ion and Hippias are less important than what they have
in common:
Both share ‘the same fault’: that is, they think that they possess,
but actually lack, moral knowledge; in the end, moral knowledge
is the only kind of knowledge that matters ( –). Plato por-
trays both figures as self-ignorant impostors using techniques of
characterization borrowed from Old Comedy. This core of moral
self-ignorance is what constitutes their foolishness and makes them
properly laughable and appropriate targets of Plato’s comedy.
Hippias’ speech on fine activities clearly relies on material from the epic poetic
tradition, and if this speech, which he is to deliver again in Athens ( –), is the
speech that Socrates has just heard at the opening of the Hippias Minor, then Homer
is the poet in question: see Woodruff, Hippias, ad loc.
Nor, indeed, is Hippias’ knowledge of such matters and other more practical
skills, such as plaiting, shoemaking, and metalworking, questioned in the Hippias
Minor ( –).
The Character of Hippias in the Hippias Major
. Conclusion
B I B L I O G R A P HY
CATHERINE ROWE TT
The Noble Lie comes in two parts. The first is about autochthony
( –): it claims that people are gestated under the earth, and
that the earth is their mother. The second is the Myth of the Metals
( –): it claims that god infuses a metal deposit into each soul
during the gestation under the earth, different metals for different
people.
The first point to note about the autochthony part is that it is evi-
dently not about what we call birth. The citizens are not required
to believe that they were earth-born as infants, but rather that they
were born into adult citizenhood, at the end of their school edu-
cation. This school-leaving event, presumably at the ephebic age
of something like or ( – ), was a kind of ‘birth from
the earth’. The gestation period preceding this ‘citizen birth’ is the
= [T]. Several previous scholars have observed (in a footnote) that this
event must be or may be an event at the ephebic age (e.g. G. F. Hourani, ‘The Edu-
cation of the Third Class in Plato’s Republic’ [‘Education’], Classical Quarterly,
Philosopher Kings and the Noble Lie
period during which the child is reared by the state education sys-
tem, which Socrates has just finished describing.
Here is what Socrates says:
[T] I’ll try first to convince the rulers themselves and the military, and
then the rest of the community, that all the nurture and education
that we gave them seem like dreams [ὀνείρατα] that they experienced,
or happened round them, when in truth at that time they were being
moulded and nurtured deep under the ground . . . (Rep. –)
The text continues, at –, by explaining that the young people
are ‘born’ fully equipped with armour and other paraphernalia:
[T] . . . when in truth at that time they were being moulded and nur-
tured deep under the ground, both themselves and their armour and
the rest of their manufactured equipment, and then when they were
fully formed, the earth, who was their mother, brought them forth,
so now . . . (Rep. – )
. Dreaming
[T] I’ll try first to convince the rulers themselves and the military, and
then the rest of the community, that all the nurture and education
that we gave them seem like dreams [ὀνείρατα] that they experienced,
or happened round them, when in truth at that time they were being
moulded and nurtured deep under the ground . . . (Rep. –)
I think that many readers take this to mean that people are to be
hoodwinked about the true nature of their upbringing, and per-
suaded that it was illusory. Instead, they will be made to think
‘Offspring’ is the usual translation. I use ‘progeny’ in [T].
See further below, sect. ..
Philosopher Kings and the Noble Lie
something that is literally untrue—namely that, instead of the edu-
cation that they actually had, they were in fact underground being
moulded and gestated (πλαττόμενοι καὶ τρεφόμενοι). Since this is un-
true, and surely they must know what kind of education they really
had, this seems like deception or self-deception. Why would they
believe it? They must be brainwashed, it seems, and having been
brainwashed, they will no longer be lucidly aware of who they are
or how they were educated.
That reading of the passage seems to me to be a total confusion.
Here is a preferable alternative: Socrates explains that the young
adults, emerging from a period of intense education for citizenship,
now become lucidly aware of the true nature of their upbringing. So
far from deceiving themselves into thinking that they were under-
ground when they know full well that they were not, the best of
them will come to realize—to discern in a fully rational way—that
during their education they were in truth underground, and were in
a dream.
The difference between awareness of reality and living in a dream
is a recurrent theme throughout the Republic, not just here. In Re-
public book the lovers of sights and sounds are said to be like those
who dream because they think that the ‘many beautifuls’ are what
the Beautiful is:
[T] . A person who recognizes beautiful objects, but does not recog-
nize beauty itself and can’t follow if someone tries to lead him to
knowledge of it, does he seem to you to be living in a dream [ὄναρ] or
in a lucid state [ὕπαρ]? Consider: isn’t the following what dreaming
[ὀνειρώττειν] is, namely taking what is merely like something else to
be, in reality, the very thing itself, and not just something that is
. Cf. , where the same verb of moulding is used while the god is
adding the metals to the stuff out of which he is making them, and cf. – ,
where the verb is used of the formation of the young child in the nursery. For τρέφω
of prenatal gestation see e.g. Aesch. Eum. ff.
R. Wardy, ‘The Platonic Manufacture of Ideology; or, How to Assemble Awk-
ward Truth and Wholesome Falsehood’ [‘Ideology’], in V. Harte and M. Lane
(eds.), Politeia in Greek and Roman Philosophy (Cambridge, ), – at ,
illustrates such a reading, despite observing the cross-reference to other dreaming–
waking motifs.
‘When they were at that time in truth underground’ (ἦσαν δὲ τότε τῇ ἀληθείᾳ ὑπὸ
γῆς, ) is usually taken to say that they are deceived about how things are in
truth. But the particle δέ is not answering to any μέν clause, so there need be no im-
plied contrast between what they think and what is real. We can read the sentence to
mean that when they come to think that they were down under the earth in a dream,
so indeed in truth they were actually down under the earth in a dream.
Catherine Rowett
like it—no matter whether they are awake [ἐγρηγορώς] or asleep [ἐν
ὕπνῳ] at the time?
. Certainly I would say that such a person is dreaming.
(Rep. –)
Are these resemblances between the two images just random, or did
Plato see them and mean us to see them too? Here are some hints
that he meant us to see them.
First, both texts are explicitly about education. The Noble Lie
completes Socrates’ description of the education of the young
( – ), and of the need to select the ones suited to serving
as Guardians and Auxiliaries ( – ). As we have seen, the
story they must believe is about their education and nurture (the
education just described in Republic ). Birth from this womb
comes at the end of all that.
Similarly, in Republic Socrates explicitly claims that the Cave
will represent our experiences with respect to education and the ab-
sence of education:
[T] Next, I said, picture our condition [φύσιν] with respect to education
and lack of education [παιδείας τε πέρι καὶ ἀπαιδευσίας], as like the
following kind of condition: take a look at some people in a kind of
underground cavernous dwelling . . . (Rep. –)
See above, sect. , and below, sect. ..
Catherine Rowett
Socrates speaks of ‘our education’ because he is thinking of the con-
dition of people in general, in non-ideal cities. But by –
we find that this describes the situation in the ideal city too. So
both motifs are about the perfect city’s provisions for educating the
young in the womb of the earth.
Secondly, we should note the many birth-related motifs in the
Cave, and in the other parts of Republic that are about educating
the philosopher rulers. At – a prisoner is ‘released’
and dragged out of the Cave into the light, and at – Socra-
tes asks us to consider ‘how someone will lead them up to the light,
as they say some people come up out of Hades to the gods’. In the
Noble Lie, at –, the earth-mother unfastens and spews out
the neonates at the end of their education. Some of the expres-
sions about emerging into the light after a period of gestation are
parallel to the terminology that Plato uses of the delivery of a child
in the Timaeus. Perhaps we should also see a reference to the rota-
tion of the child in the womb, ready for birth, in the idea of studies
that can ‘turn the soul’ to prepare it for birth into the light. We
are reminded (though not explicitly) of Socrates’ self-description of
himself as a midwife.
The trainee philosophers are to be delivered into the light by
means of an education that reveals a whole new world outside the
Cave. So the Cave is not just about politics. It is not just about need-
ing philosophers to return to run the state: they must also return to
educate the young. Who is it that will first force a youngster to turn
and shed his chains? Who will drag him out into a world he had
no desire to see? Surely a philosopher midwife who returns to the
dream world to forge the gold in the soul of those capable of philo-
sophy and bring them out into the light, where they can flourish.
καὶ μὴ ἀνείη πρὶν ἐξελκύσειεν εἰς τὸ τοῦ ἡλίου φῶς, ἆρα οὐχὶ ὀδυνᾶσθαί τε ἂν καὶ
ἀγανακτεῖν ἑλκόμενον, καὶ ἐπειδὴ πρὸς τὸ φῶς ἔλθοι . . .; (‘and not to let go until he has
dragged him out into the light of the sun, wouldn’t he find it painful and distressing
being dragged along, and when he got near to the light . . .?’).
καὶ πῶς τις ἀνάξει αὐτοὺς εἰς φῶς, ὥσπερ ἐξ Ἅιδου λέγονται δή τινες εἰς θεοὺς ἀνελ-
θεῖν.
ἡ γῆ αὐτοὺς μήτηρ οὖσα ἀνῆκεν (‘the earth, being their mother, let them loose’).
e.g. Tim. –.
– (μεταστρεπτικῶν). Cf. also Aristophanes’ speech at Sym. ,
where Zeus rotates the heads of the newly halved humans.
Theaet. and passim. There Socrates is delivering his pupil’s unborn the-
ories, whereas in the Republic the pupil herself needs to be delivered. Midwives are
still needed, however.
Philosopher Kings and the Noble Lie
. Ideological inhibitions
On the basis of these parallels with the Cave, we are now in a posi-
tion to challenge several dominant ideas about Plato’s ideological
aims in recommending the use of the Noble Lie. First we shall con-
sider what makes Socrates embarrassed about his proposal. Is there
something that would be difficult for his listeners to accept? If so,
what?
[T] But still, you should go on and listen to the rest of the myth. ‘For
all of you who live in this community,’ we in our myth-making role
shall say to them, ‘you are all brothers; but the god who moulded
you, those among you, on the one hand, who are capable of govern-
ing, he mixed gold into the birth for them, making them highest in
honour; those who are auxiliaries, secondly, silver; but iron and cop-
per for the agricultural and other manual workers.’ (Rep. –)
[T] Given that you are all related to each other, all of you will for the
most part beget others like yourselves, but there will be times when
silver progeny will be conceived from gold, or gold from silver, and
all the other metals from each other similarly. Hence the god’s first
and most emphatic message to those who govern is that there is
nothing about which they will be better guardians, and nothing that
they will guard more carefully, than the progeny, as to what exactly
has been mixed into their souls. (Rep. – )
The Myth of the Metals is designed to ensure that citizens are not
misclassified according to who their parents were ( –). This is
the most important instruction that the god gives, because the city
will be destroyed if they ever make a mistake ( –).
Given this rationale, Plato must mean that genetics cannot guar-
antee the transfer of metal from parent to offspring. Does he just
mean that inheritance is fallible (and the failures must be picked up
correctly), or that nothing is inherited? Either position is compatible
with what Socrates has in mind, since the important point is that
what matters for who you are is not who your parents are but what
you are suited to. The only interpretation that cannot be right is
Popper’s view that class is based on inherited racial characteristics.
How exactly you become prone to absorb silver rather than
bronze is under-specified in the myth, except that the work is
attributed to ‘the god’—which, as noted above, is a way of denying
that it can be altered by human intervention. Some passages im-
ply that offspring typically resemble their immediate parents (e.g.
– , within [T]), while others warn the Guardians that
. That is, no one gets mixed metals, as noted by Schofield, Plato, .
See further below (this section).
A similar warning is given about the rules of marriage at –. See n. .
See [T].
There are other ways of reading this passage, which could mean that all the
Philosopher Kings and the Noble Lie
nothing stops their children from having something quite different
( – ). These are not, of course, contradictory.
In any case it is clear that Socrates’ main point is the negative one:
that one absolutely must not rely on the mere probability. The
passage at – instructs the rulers not to cut corners: they must
test for the metal, not go by the parentage. They might be tempted
to use parentage as a rule of thumb, or to promote someone they
think was their own son or daughter. In reality, as we later discover
( , ), the society will not keep track of the parentage of
Guardian offspring—a provision which effectively removes the lat-
ter temptation provided the scheme is correctly followed.
Whether Plato thought that abilities were mainly inherited, or
mainly random, the earth-womb is clearly provided to ensure that
everyone has an equal chance to acquire and develop talents, and
manifest them, before being assigned to their life-plan. We need
not determine exactly what is due to nature and what to nurture,
providing that we understand that the nurture is designed to en-
sure that no one is set up to fail owing to unequal chances.
To sum up, then, Plato’s position is that whatever hereditary in-
fluence there may be on a youth’s capacity to absorb this metal or
that, it is unreliable and must be ignored as regards educating her
for future life (for which nothing but her actual abilities count).
Furthermore, we should not confuse inherited ability with inher-
ited privilege, which is where people gain advantages that do not
match their ability (whether inherited or not), simply as a result of
belonging to some privileged family.
The marriage numbers are a case of genuine lying, because they conceal some-
thing that the rulers know but those affected must not know (for pharmacological
reasons). The Noble Lie is not like that. It aims to see justice done by placing people
in appropriate roles: not something to be concealed. No one is trying secretly to put
people in the wrong roles. So it resembles the healthy stories of gods and heroes,
which are not intended to deceive but to convey the truth in palatable form. The
truth in question is not unmentionable on anyone’s story except Popper’s (which
has no support in the text).
Philosopher Kings and the Noble Lie
lost in the solution. What happened was not that we found a way
to deceive the rulers, but that we found that no deception was re-
quired and it was not a lie after all.
That would be one way to explain what has happened. If that
is what has happened, should we then object to Plato’s procedure,
introducing it as a lie and then showing that for his society it is
not a lie after all? There are clearly good dramatic reasons for that
procedure, and also some political reasons in that it subtly accuses
societies that use such myths to privilege certain families of living
by a lie, whether or not their rulers believe the lies they peddle. For
them, the lie defends privilege by birth instead of eliminating it,
and for them it is a lie.
Alternatively, we might say that its falsity has not been removed.
It remains literally false that the education was underground, that it
was all a dream, that it instilled minerals such as gold and copper and
so on. All these are metaphor. The earth is not our birth mother,
and what was provided in the education was provided by the state,
not the earth. Any truth in this myth is not in its mythical motifs,
rich though they are as metaphors, but in what they stand for, as
the rulers will realize.
So, perhaps after all the rulers will see through the myth to its
truth, in a way that other citizens will not. It transpires, then, that
although the rulers do believe it, and do not deceive or lie to the
people in teaching them the myth, it will never be more than a be-
lievable myth for the ordinary people, while it is only the philoso-
phers who will know how true it is and why. So there is, after all, a
difference in their epistemic relation to what they believe, and the
kind of persuasion achieved. The rulers will believe and understand
why, while the people will just believe (and that will be wisdom for
them too).
University of East Anglia
B I B L I O G R A P HY
Adam, J., The Republic of Plato (Cambridge, ).
Bloom, H. (ed.), Rebirth and Renewal (New York, ).
Clay, D., ‘Plato’s First Words’, in F. M. Dunn and T. Cole (eds.), Begin-
nings in Classical Literature (New York, ), –.
See Rep. – on how the society counts as wise.
Catherine Rowett
Hesk, J., Deception and Democracy in Classical Athens (Cambridge, ).
Hourani, G. F., ‘The Education of the Third Class in Plato’s Republic’
[‘Education’], Classical Quarterly, (), –.
Irigaray, L., Speculum of the Other Woman, trans. G. C. Gill (Ithaca, NY,
).
Krumnow, K. L., ‘Womb as Synecdoche: Introduction to Irigaray’s De-
construction of Plato’s Cave’, Intertexts, (), –.
Loraux, N., The Children of Athena: Athenian Ideas about Citizenship and
the Division between the Sexes [Children], trans. C. Levine (Princeton:
Princeton University Press, ).
Loraux, N., The Invention of Athens: The Funeral Oration in the Classical
City, trans. A. Sheridan (Cambridge, Mass., ).
Murray, P., ‘What is a Muthos for Plato?’, in R. Buxton (ed.), From Myth
to Reason? Studies in the Development of Greek Thought (Oxford, ),
–.
Nightingale, A. W., Spectacles of Truth in Classical Greek Philosophy: The-
oria in its Cultural Context (Cambridge, ).
Page, C., ‘The Truth about Lies in Plato’s Republic’, Ancient Philosophy,
(), –.
Popper, K., The Open Society and its Enemies, i. Plato (London, ).
Rosenstock, B., ‘Athena’s Cloak: Plato’s Critique of the Democratic City
in the Republic’, Political Theory, (), –.
Saxonhouse, A., ‘Myths and the Origins of Cities: Reflections on the
Autochthony Theme in Euripides’ Ion’, in P. J. Euben (ed.), Greek
Tragedy and Political Theory (Berkeley, ), –.
Schofield, M., ‘Metaspeleology’, in D. Scott (ed.), Maieusis: Essays on An-
cient Philosophy in Honour of Myles Burnyeat (Oxford, ), –.
Schofield, M., Plato: Political Philosophy [Plato] (Oxford, ).
Ustinova, Y., Caves and the Ancient Greek Mind: Descending Underground
in the Search for Ultimate Truth (Oxford, ).
Wardy, R., ‘The Platonic Manufacture of Ideology; or, How to Assemble
Awkward Truth and Wholesome Falsehood’ [‘Ideology’], in V. Harte
and M. Lane (eds.), Politeia in Greek and Roman Philosophy (Cambridge,
), –.
ARISTOT L E O N T R U TH-BEAR ERS
. Truth-bearers
This section argues for two claims. First, that there is strong
evidence for the view that Aristotle reserves the whole disjunc-
tive phrase ‘true or false’ exclusively for propositional items.
Non-propositional items are never characterized by the entire
disjunction ‘true or false’ but only by one part of it (in isolation):
either as a ‘false’ thing, or as ‘the truth’, or (far more rarely) as a
‘true’ thing. Second, that when ‘true’ or ‘false’ is used to describe
parts of external reality, Aristotle distinguishes this from his use of
the disjunctive phrase ‘true or false’ elsewhere. When he talks of
a false thing or of the truth in characterizing external things (and
not propositional items), he provides signposts to indicate that
these uses are importantly distinct from that involved in the whole
disjunctive phrase ‘true or false’.
For a recent expression of such doubts see Burnyeat et al., Eta, ad loc.
Aristotle on Truth-Bearers
.. In favour of propositional truth-bearers: Metaphysics Δ and
Ε
The best evidence in favour of taking Aristotle’s truth-bearers as
propositional items is found in Metaphysics Δ . Here, the veridical
use of ‘to be’ applies to true or false affirmations or denials, lin-
guistic items (ὁμοίως ἐπὶ καταφάσεως καὶ ἀποφάσεως, a–).
Items such as ⸀Socrates is cultivated⸣ or ⸀Socrates is not pale⸣ are
true, items such as ⸀the square’s diagonal is commensurate⸣ are false
(a–).
Metaphysics Δ is an important text, not easily dismissed. The
‘Philosophical Lexicon’ is precisely where Aristotle should set out
the central cases of what is true or false in exemplifying the veridical
use of ‘to be’. It would be strange had he included anything apart
from the primary cases of what is true or false. Propositional items,
such as affirmations and denials, play this role.
One might, however, approach the question of Aristotle’s truth-
bearers in a different way, taking as one’s starting point those
chapters which treat the true and the false as their main subject-
matter (as opposed to Metaphysics Δ , which deals with ‘is true’
and ‘is false’ as one among the many uses of ‘to be’). For this
approach, Metaphysics E and Θ are crucial. Consider the
introduction to E :
[T] Let us, then, leave to the side the enquiry into that which is accident-
ally (for it has been sufficiently discussed); but that which is as true
[τὸ δὲ ὡς ἀληθὲς ὄν] and which is not as false [καὶ μὴ ὂν ὡς ψεῦδος],
since they depend on combination and division [ἐπειδὴ παρὰ σύνθεσίν
ἐστι καὶ διαίρεσιν], both taken as a whole are about the distribution
At b we follow the manuscripts against Jaeger, who inserts the article τό
before μὴ ὂν ὡς ψεῦδος. In this way there is no suggestion that in introducing being
true or false Aristotle is speaking of two rather than one single way of being. Being
true or false can constitute a unitary way of being, without undermining the distinc-
tion between truth and falsity. One might describe veridical being as a unitary way
of being with a positive and a negative pole: truth and falsity respectively.
Bonitz’s suggestion to take the preposition παρά at b as ‘causal’ seems
attractive. We sought to accommodate this suggestion in our translation using the
notion of dependence: being true or false (causally) depends on thought operations
of combination and division. The idea would be that combining or dividing items in
thought is the cause for there being items which are true or false, i.e. for there being
linguistic items or mental states which are true or false. It would not be plausible
to hold that such thought operations are causes for an item’s being true (or for its
being false): for it would be a short step from this causal dependence to some form
of idealism. Further examples of the causal use of παρά in Aristotle occur at Pr. An.
David Charles and Michail Peramatzis
of contradictory propositions; for the true [τὸ μὲν γὰρ ἀληθές] takes
the affirmation in the case of what is combined, and the denial in the
case of what is divided, while the false [τὸ δὲ ψεῦδος] takes the con-
tradictory of this distribution . . . (Metaph. Ε , b–, our
translation)
. , a, and . , b (see LSJ s.v. παρά, III. ). A comparable usage can be
found at Metaph. Α , b: ‘perceptibles are called παρά forms’, i.e. after and in
virtue or because of them; cf. EE . , a–: ὁ γὰρ θρασὺς παρὰ τὸ θράσος λέγεται
παρωνύμως. It seems plausible that paronymy is a case of causal dependence of the
paronymous item on the item from which its name derives.
In our view there should not be any brackets at b–; rather, the brackets
seem to fit well around b–: πῶς δὲ τὸ ἅμα . . . ἀλλ᾿ ἕν τι γίγνεσθαι. We shall
return to this point in sect. .
We should note that in discussing [T] and the other texts listed, our aim is to
give a well-grounded and plausible reading, not to consider the whole range of al-
ternatives others have suggested.
Aristotle on Truth-Bearers
propositions or thoughts, not to external things (such as material
entities). Further, it is definitionally dependent on the distribution
of propositions according to the principle that either they are true or
their contradictory is. In what follows we shall call this ‘the Principle
of Contradictory Pairs’ (CP).
Second, the way of being in question is, as Aristotle empha-
sizes, one which belongs per se to thoughts and is distinguished
from a further distinctive way of being, which belongs per se to
external, mind-independent objects (a–). The former de-
pends on thinkers’ ability to combine and divide objects in thought
(b–; see b), the latter does not. On these grounds
he distinguishes between a way of being which does not depend on
our thoughts and one (being true or false) which does. His accep-
tance of a distinct way of being (for external objects) which does
not depend on our thoughts reflects his commitment to realism.
What makes propositions true is independent of our thinking them
to be true.
The way of being applicable to external, mind-independent
things is distinct: it is the way in which objects (not propositions)
are, where objects do not have contradictories, still less contra-
dictories governed by CP (or by a suitable counterpart to CP such
as: either an object or its contradictory is true). Nor does the way
of being for external objects depend essentially on our ability to
combine and divide objects in thought. So understood, being true
CP is cognate with what in C. W. A. Whitaker, Aristotle’s De interpretatione:
Contradiction and Dialectic (Oxford, ), , is called ‘Rule of Contradictory
Pairs’ (RCP). As Aristotle holds in both Categories and De interpretatione –:
‘of every contradictory pair, one member is true and the other false’. (We are in-
debted to Walter Cavini for this point.) It is important to note that CP need not hold
of all propositions. There may, for all it says, be propositions (such as concern, for
example, the future) where it is not the case that either a proposition or its contradic-
tory is true (cf. De interpretatione ).
For a characterization of realism of this form see M. Dummett, ‘Truth’, Pro-
ceedings of the Aristotelian Society, (–), – at : ‘The correspondence
theory expresses one important feature of the concept of truth which is not expressed
by the law “It is true that p iff p” and which we have so far left out of account: that a
statement is true only if there is something in the world in virtue of which it is true’
(Dummett’s emphasis); see also pp. – for a characterization of the realism vs.
anti-realism debate as focusing on the question of whether there is some independent
reality (not of ‘our own making’) answering to our statements or not. See also Dum-
mett, ‘Realism and the Theory of Meaning’, in id., The Logical Basis of Metaphysics
(London, ), – at : ‘[Frege] made a classic pronouncement of the realist
faith, saying that the truth of a proposition has nothing to do with its being taken as
true.’
David Charles and Michail Peramatzis
or false (as a whole) exemplifies a way of being for propositions or
thoughts, not objects in the external world.
Truth-bearers are, we shall assume, the entities to which truth
and falsity together (as a whole) belong. On the basis of the texts so
far considered, it seems clear that Aristotle takes them to be propo-
sitional items. He does not mention any other items which exem-
plify being true or false.
[T] About the incomposites, now, what is being or not being, and what is
the true and the false [τί τὸ εἶναι ἢ μὴ εἶναι καὶ τὸ ἀληθὲς καὶ τὸ ψεῦδος]?
For it is not composite, so as to be when it is compounded, and not be
if it is divided, such as the white log or the diagonal’s being incom-
mensurable [ὥσπερ τὸ λευκὸν 〈τὸ〉 ξύλον ἢ τὸ ἀσύμμετρον τὴν διάμετρον];
nor will the true and the false still obtain in a similar fashion in their
case too [οὐδὲ τὸ ἀληθὲς καὶ τὸ ψεῦδος ὁμοίως ἔτι ὑπάρξει καὶ ἐπ᾿ ἐκεί-
νων]. Or rather just as truth is not the same in the case of these [the
incomposites], so too neither is being the same [for the incomposites]
[ἢ ὥσπερ οὐδὲ τὸ ἀληθὲς ἐπὶ τούτων τὸ αὐτό, οὕτως οὐδὲ τὸ εἶναι], but
truth or falsity is as follows, to grasp and to speak of them is truth
(for affirmation and to speak of something are not the same), whereas
to ignore is not to grasp them [ἀλλ᾿ ἔστι τὸ μὲν ἀληθὲς ἢ ψεῦδος, τὸ μὲν
θιγεῖν καὶ φάναι ἀληθές . . ., τὸ δ ᾿ ἀγνοεῖν μὴ θιγγάνειν]. (Metaph. Θ ,
b–, our translation)
It should be noted that the phrase τὸ μὲν ἀληθὲς ἢ ψεῦδος at b need not
imply that truth or falsity as a whole applies to incomposites in the same way in
which it holds of composites. Rather, we think that the phrase ‘the true or the false’
is a natural way for Aristotle to extend his treatment of truth and falsity to the case
of incomposites. It is an added, separate claim (which he makes in the subsequent
lines) that this notion applies differently to the case of incomposites. That is to say,
in the case of incomposites there is only being as being true but there is no not-being
as being false (although there may still be not-being as being not-true).
Aristotle on Truth-Bearers
even an utterance). They are not external things or ways/types of
being. In the case of incomposites the false is missing: we have in
its place ‘to be ignoring or not grasping’ (b: τὸ δ ᾿ ἀγνοεῖν μὴ
θιγγάνειν). Aristotle later maintains that there is no false (or mistake/
deception) about incomposites, only ignorance (a–: ἄγνοια).
It seems clear that this lack of truth characterizes not a thing in the
external world but a mental state: ignorance.
This second section of Metaphysics Θ closes with an impor-
tant, if difficult, passage summarizing this chapter’s results about
the true and the false as they apply both to incomposites and com-
posites. This passage brings together several of the themes in the
chapter and should be treated in some detail. Here is the text fol-
lowed by a neutral translation (we shall discuss possible alternative
construals shortly):
The claim about dependency in the first sentence involves the no-
tion of depending on (παρά) at b. The idea is that being as
being true and non-being as being false, or more precisely, affirma-
tions and denials which are strictly speaking true or false (b–
), are so derivatively from, or dependent on, combination and di-
vision. After noting that the true and the false are not features of
external, mind-independent things but of thoughts (b–), Aris-
totle proceeds to point out that the type of combination and divi-
sion involved is an operation effected by thought (b–: ἡ
the way of being which is being true or false can still be unitary, as the way in which
certain linguistic items or mental states are, even if truth and falsity are sharply dis-
tinguished (as, for instance, the positive vs. the negative poles of the veridical way
of being). Thus, for example, the veridical way of being may possess a generic sort
of unity, with being true and being false being two opposite poles of this unity. The
latter two would not, of course, be distinct species of veridical being, for this would
immediately undercut the unity of being true or false. Rather, being true would bear
(for example) a certain type of relation to external reality, while being false would
presumably bear the contrary relation to this reality. Compare with this case the sort
of unity that being an animal possesses, and the way in which being male or being
female are subsumed under being an animal, without being two species of it.
As noted earlier, we find promising Bonitz’s interpretation of παρά as ‘causal’,
i.e. as expressing a relation of causal dependency of the way of being characteristic of
propositional items (being true or false) on thought operations. The παρά is attested
in all the main manuscripts; the variant περί is found only in a citation by Asclepius.
David Charles and Michail Peramatzis
συμπλοκή ἐστιν καὶ ἡ διαίρεσις ἐν διανοίᾳ, ἀλλ᾿ οὐκ ἐν τοῖς πράγμασι).
What thought does is to grasp external beings—an object’s being
essentially thus-and-so, or its being of a certain quality, or its being
of a certain quantity, etc.—and to keep them either connected or
apart (b–: συνάπτει ἢ ἀφαιρεῖ ἡ διάνοια).
In Aristotle’s picture, the truth-bearers—affirmations or deni-
als, or other linguistic items or mental states—depend, for their
being true or false, on thought’s combining and dividing external
things. A truth-apt (or truth-evaluable) item constrained by CP,
such as a statement or a belief, depends on our thinking of be-
ings as combined or divided: when one (mistakenly) combines in
thought the square’s diagonal with being commensurate, one ar-
rives at the statement or belief that the diagonal is commensur-
ate, a propositional item which is false. Aristotle’s claim made at
b–a seems to formulate precisely this idea: the cause of
there being truth-bearers, true or false mental states, is the thought
operations of combining and dividing beings (τὸ γὰρ αἴτιον . . . τοῦ
δὲ [ἀληθοῦς καὶ τοῦ ψεύδους; sc. οὕτως ὂν at b, referring back
to τὸ ψεῦδος καὶ τὸ ἀληθὲς at b] τῆς διανοίας τι πάθος).
However, this is not the whole story. Thought and its operations
of combining and dividing depend on there being things (πράγ-
ματα), specifically categorial beings, to be thought of as combined
or divided. This additional dependence relation is signposted in the
text first at b–. The phrase ‘being and not being thus’ (οὕ-
τως ὂν καὶ μὴ ὄν) at b and b clearly refers to being-as-true and
non-being-as-false: these are propositional items which are true or
false. They are sharply distinguished from the basic or central cases
of being, categorial beings such as a’s being essentially F, a’s being
of a certain quality, a’s being of a certain quantity, etc. (b:
τὸ δ ᾿ οὕτως ὂν ἕτερον ὂν τῶν κυρίως, i.e. τὸ τί ἐστιν ἢ ὅτι ποιὸν ἢ ὅτι
ποσὸν ἢ εἴ τι ἄλλο at b–). The view that categorial beings are the
It may be objected that the claim that thoughts cause mental states to be or to
exist may exclude thoughts from being themselves true or false. Alternatively, it may
be suggested that thoughts are true or false, but they are so in a self-caused fashion:
for they make themselves truth-bearers in their own right. In our view, neither of
these two possibilities is compelling or plausible. This objection is defused once we
draw properly the distinction between thinking—the thought operation of combin-
ing and dividing—and thought—the end product of thinking. While the former is
a thought process or activity and so could not be characterized as true or false in
any relevant manner (but perhaps as truth-aiming or truth-achieving), the latter is
a mental state (e.g. what is believed, hoped, regretted, etc.: the object or content of
a mental state) and so is true or false.
Aristotle on Truth-Bearers
basic cases of being (κυρίως) is unsurprising given the remark made
in Metaphysics Δ that categorial being is being in its own right
(a–: καθ᾿ αὑτά).
Aristotle does not rest content with this claim. Being-as-true and
non-being-as-false (as well as accidental being) are not only differ-
ent from but also dependent on categorial being. At a– he
writes that both of these ways of being are about the rest of the ways
of being: they depend on the ‘basic’ (κυρίως), categorial way of be-
ing (and perhaps also categorial being as it is further modified by
being potentially or being actually: ἀμφότερα περὶ τὸ λοιπὸν γένος τοῦ
ὄντος). Similarly, the true and the false (as well as accidental being)
do not reveal anything additional to categorial being; nor do they
require that there exist any further nature of being over and above
categorial being (a: οὐκ ἔξω δηλοῦσιν οὖσάν τινα φύσιν τοῦ ὄν-
τος). It appears that here the true and the false are propositional
truth-bearers, and depend on categorial being: they depend on it
for their being true or their being false (as the case may be).
In sum: Aristotle’s overall view suggests the following picture:
Level Truth-bearers, linguistic items or mental states, which are
products of level operations (thought processes or acti-
vities) and are governed by CP, are, properly or strictly
speaking, true or false.
Level Thought processes or activities, grasping external things as
being connected/combined or as being divided/separated,
yield level items.
Level Categorial beings, external things, are objects which are
or are not (essentially, necessarily, or accidentally) thus-
and-so.
In this picture level items depend on level thought operations for
their existence and nature as truth-bearers. Level thought opera-
tions depend on level : they grasp external beings, and take them as
being combined or divided. Assuming that the same dependence re-
lation is used in both cases, it follows by transitivity that level is ul-
timately dependent on categorial beings in the following two ways.
First, propositional items depend for their being truth-bearers on
Why assume that this relation of dependence is the same across all three levels,
and so warrants transitivity? One way in which to address this question is to take this
relation as a type of causing, producing, or yielding graspable items. Thus, level
yields categorial beings, graspable items in the sense that they are to be grasped as
being combined or being divided in level . Level in turn yields propositional items
David Charles and Michail Peramatzis
there being categorial beings for thought processes to grasp as com-
bined or divided. Second, level propositional items (resulting in
this manner from level combination or division in thought) de-
pend for their being true or their being false (as the case may be)
on whether the combination or division they involve or result from
correctly describes, or fails to describe (respectively), the way in
which categorial beings are combined or not combined (or divided).
To this picture we could add a further, even more fundamental,
level, taking our lead from a–:
Level Substances or/and essences are causes and principles of
categorial being as such.
Aristotle urges us to leave aside the true and the false because its
causes are thought operations. Further, it is dependent on categorial
beings in the ways just suggested. What we should examine, by con-
trast, are the causes and principles of categorial being (in so far as it
is a being). If we assume (plausibly) that Metaphysics Ζ takes up
precisely this task, it is reasonable to infer that the causes and prin-
ciples of categorial beings are substances, or more accurately, the
essences of these substances. This would establish a further depen-
dence relation of the true and the false on Aristotle’s metaphysical
bedrock: substances and essences.
The main differences between the α and the β texts are at b–.
In the α text just cited and our translation for these lines we read:
α (EJ) τὸ δὲ κυριώτατα [sc. ὂν or λεγόμενον ὂν] εἰ ἀληθὲς ἢ ψεῦδος, τοῦτο δ᾿ ἐπὶ
τῶν πραγμάτων ἐστὶ τὸ συγκεῖσθαι ἢ διῃρῆσθαι.
and [being is said] in the strictest way if [something] true or false
[is said of/about it], and this on the side of things is for them to be
combined or to be divided.
On our reading, categorial being will include such items as Socrates, the healthy
one, being healthy. These items will be presented under the guise of the propositio-
nal items which describe them: for instance, Socrates+being healthy when seen from
the viewpoint of (e.g.) the predicative statement ‘Socrates is healthy’. Aristotle con-
tinues to maintain that categorial being is the strictest or most basic way of being.
To use the examples provided in Metaphysics Δ : the categorial beings Socrates and
being healthy (or Socrates’ being healthy) will be present when represented or de-
scribed by expressions such as ‘Socrates is healthy’ (a–). It may be objected
that Aristotle cannot be taking categorial being to be combined or divided, as our
view allows him to be doing. This objection may be fair if one focuses exclusively
on categorial being as introduced and discussed in parts of the Categories. In that
context the categories are items which are ‘without combination’ (ἄνευ συμπλοκῆς:
Cat. , a–, –; , b ff.). But in the Metaphysics, and specifically in Θ ,
as well as Δ and Ε –, Aristotle does not use the notion of κατηγορία by itself.
Rather, he frames his claims in terms of the concept of τὰ σχήματα τῶν κατηγοριῶν.
The former way of speaking of categorial beings picks up items of the form (being)
F, whereas the latter also allows for items such as x’s being (or not being) F. The idea
that categorial beings are combined or divided items seems to underlie the examples
offered in Metaph. Δ , a–, and Ε , b–: the healthy human or a
human’s being healthy; the walking human or a human’s walking; what something
is; that it is of such a quality; that it is of such a quantity. At any rate, the external
things about which true or false propositional items obtain are combined or divided
categorial beings of this sort. They are not the referents of just the subject or just
the predicate terms of propositional items.
Aristotle on Truth-Bearers
‘this’ (τοῦτο at b) as picking up not ‘being true or false’ (ἀλη-
θὲς ἢ ψεῦδος at b–) but the whole phrase ‘the strictest case of
being is present, if something true or false is said of/about it’ (τὸ δὲ
κυριώτατα εἰ ἀληθὲς ἢ ψεῦδος at b–). So understood, ‘this’ does
not refer to a true or false item said of/about a thing but to the strict-
est case of being—categorial beings—about which something true
or false is said. These clauses, as a whole, say: ‘and [being is said] in
the strictest way if [something] true or false [is said of/about it], and
this [i.e. the strictest case of being] on the side of things is for them
to be combined or to be divided’. The outcome of this reading is
clear: for anything to be true or false of/about the [relevant] strict-
est case of being, categorial being, there have to be certain combi-
nations or divisions of external things. External things are not the
truth-bearers but the truth-makers for any item which is, strictly
speaking, true or false. The latter are propositional items said of or
about external objects.
The interpretation just sketched has considerable merits. First,
it takes and makes good sense of the α text in its entirety, without
invoking any elements from the problematic β text. All alternatives
‘cherry-pick’ at this juncture and none retains (as we have done) the
α text in its entirety. Second, our reading maintains the Aristotelian
thesis that categorial being is the basic or central way of being: for,
in our interpretation, the strictest case of being (κυριώτατα [ὄν or
λεγόμενον ὄν]) clearly is identified with being ‘in accordance with
the figures of the categories’. In this way, the introductory lines of
Metaphysics Θ agree with the characterization of categorial being
as being ‘in itself’ or ‘as such’ (καθ᾿ αὑτά) in Δ and, more impor-
tantly, as the ‘strict’ or ‘basic’ being (κυρίως) in Ε . Third, in our
account, truth or falsity are introduced using the subordinate con-
ditional clause ‘if anything true or false is, or is said, of/about the
strictest case of being’ (εἰ ἀληθὲς ἢ ψεῦδος at b–). So under-
stood, Aristotle places an appropriate gap between external things
(the strictest way of being, categorial being) and his truth-bearers.
While the latter, which are true or false, are (said) of or about ex-
ternal objects, they are not identical with them.
There is a fourth advantage. The presence of the subordinate
It is important to note that the connective δέ in the phrase τοῦτο δ ᾿ ἐπὶ τῶν πραγ-
μάτων at b makes good sense on our account. If δέ involves an adversative use,
its function will be to signal a contrast with what is said (if something true or false
is said) at b–. Now we are to focus back onto the side of things (where what is
most strictly is found) and away from such items as thoughts or propositions.
David Charles and Michail Peramatzis
conditional clause (on the basis of the εἰ found in the α text) allows
for several ways to secure a reference of the demonstrative pronoun
τοῦτο at b without taking it to latch onto being true or false.
In the simpler possibility already noted, the conditional clause in-
terrupts the main clause ‘being in the strictest case. . .’, and so the
‘this’ refers back to the subject-matter of the main clause, the strict-
est way of being (categorial being). Alternatively, the conditional
clause is indeed the referent of ‘this’ but the phrase ‘for things’
(ἐπὶ τῶν πραγμάτων) serves to shift the reference back to the strict-
est way of being (‘this, being true or false, from the side of things
is . . .’). Whichever alternative is preferred, the result remains the
same: ‘this’ refers back to the strictest way of being, categorial be-
ing, and introduces the innocuous and unsurprising claim that this
way of being is somehow closely linked to things’ being combined or
divided. It does not render external things (primary) truth-bearers.
For something (e.g. ‘being cultivated’) to be true of the categorial
being Socrates is for the two relevant things, Socrates and being
cultivated, to be combined. For something (e.g. ‘being pale’) to be
false of the categorial being Socrates is just for the two things, Soc-
rates and being pale, to be divided. This is why Aristotle returns (at
b–) to being true or false, and describes what it is for an af-
firmation or a denial to be true or false, given that things (categorial
beings) are combined or divided in the ways outlined at b–. Im-
portantly, to allow for the possibility of complex, combined or di-
vided, categorial beings Aristotle uses (in this chapter but also in
Metaphysics Δ and Ε –) the notion of τὰ σχήματα τῶν κατηγο-
ριῶν, which seems to contrast with the simple κατηγορία, a notion
which does not seem to allow for any combination or division (as it
is an item ἄνευ συμπλοκῆς: Cat. , a–, –; , b ff.).
The fifth advantage of our interpretation is that it allows for
two ways (matching those mentioned in the previous paragraph)
in which to understand the relation expressed by the phrase ‘is for
things’ (ἐπὶ τῶν πραγμάτων ἐστί at b). If the reference of τοῦτο
at b is to the strictest way of being, we can readily accom-
modate the identity reading of ‘to be’. So understood, Aristotle
would be identifying external beings—the strictest case of being,
categorial being, when something true or false is said of them—
with things’ being combined or divided. This is unproblematic: if
something true or false is said of categorial beings, the latter are
combined or divided. Further, even if ‘this’ were to refer to being
Aristotle on Truth-Bearers
true or false, we could understand the phrase ‘this is for things to
be. . .’ as itself meaning ‘being true or false (partly) consists in, or
depends on, things’ being combined or divided’. So understood,
this phrase (τοῦτο δ᾿ ἐπὶ τῶν πραγμάτων ἐστί) would serve to em-
phasize the relation of being true or false to things, and the de-
pendence of the former on the latter: ‘this, being true or false, in
respect of what is required of things, is . . .’. We do not need to
choose between these differing construals: whichever is preferred,
the claim made at b– is about the relation which obtains
between external things and true or false items corresponding to
them. It specifies that the latter’s being true or false depends on
external things, their combination or division. This way of under-
standing b–, in terms of a truth-making claim, affords an ele-
gant transition to b and the following lines. Initially, in lines
b
– Aristotle focuses on the relation between linguistic items or
mental states and combinations or divisions of things. Second, in
lines b– he raises a question about the conditions under which
being true or false obtains: this is a question about the underlying
truth-makers responsible for propositional truth-bearers. Finally,
at b– he clarifies that his main focus in the present context is on
the truth-makers in virtue of which our beliefs or statements (the
appropriate truth-bearers) are true or false.
In this interpretation, the dependency of being true or false (as
a characterization of propositional truth-bearers) on things is en-
capsulated in the ‘thus’ (ὥστε) clause at b–. Because true or
false linguistic items or mental states depend on things, it follows
(ὥστε) that a belief, the mental state of a believer (οἰόμενος), is true
or false (ἀληθεύει μὲν . . . ἔψευσται δέ) depending on whether it de-
scribes how things stand, or not (respectively). A belief which takes
combined things to be combined, or divided things to be divided,
is true, whereas a belief that takes things contrary to how they are is
false. If b– is understood in this way, we get a picture which
fits exactly with the account of truth for composites formulated in
Section . on the basis of Metaphysics Θ , b–a:
What the present formulations add to our earlier [TC] and [FC]
is, first, a more precise characterization of what external things in-
volve (things’ being combined or being divided), and what the re-
lation between this external side and the propositional side consists
in (taking combined things as combined, and divided things as di-
vided). This is a significant addition: it brings Aristotle’s present,
more sophisticated account of the true and the false into line with
the common-sense realist account, in which the true says that ‘what
is’ is, or that ‘what is not’ is not, whereas the false says that ‘what is
not’ is, or that ‘what is’ is not. Indeed, this is exactly how Aristotle
himself defines the true and the false in a straightforward manner
at Metaph. Γ , b–.
There is a further, substantive, claim added by [TC′] and [FC′]
in the present section of Metaphysics Θ . This addition is about
the side of the bi-conditionals concerned with external things. In
Section . we suggested that on the basis of Categories the ac-
count of truth is best understood in terms of bi-conditionals such as
[TC] and [FC]. Thus, we understood the position developed in the
opening lines of Metaphysics Θ , too, in a similar fashion, as co-
dified in [TC′] and [FC′]. For this sort of formulation captures the
claim made at Categories , b–, that true statements and ex-
ternal items corresponding to them ‘converse’ or ‘mutually recipro-
cate as to implication of being’. Bare bi-conditionals, however, fail
to capture the further claim made in Categories concerning the
natural priority of external items over true statements correspond-
ing to them: the former, their being, obtaining, or being the case,
Aristotle on Truth-Bearers
are somehow the cause for the latter’s being true, while the converse
does not hold good (b–). This priority claim is mirrored in
the truth-making role that external things, their being combined
or being divided, serve in the present context of Metaphysics Θ .
This first becomes clear from our reading of the sentence ‘this is for
things to be combined or divided’ at b– as set out earlier: be-
ing true or false (said about categorial being) requires or depends on
things’ being combined or divided, while combinations or divisions
of things do not, in any comparable manner, require or depend on
being true or false. The ‘not conversely’ clause is not added in this
sentence but is explicitly present at b–: it is because you are
pale that our belief that you are pale is true, but not the other way
round.
To conclude: a detailed reading of a–b shows that these
lines (in the α manuscript tradition) support the philosophically
plausible view, which Aristotle holds elsewhere: propositional
items, not external, mind-independent objects, are what is true or
false. There is nothing in these lines to make us accept that Aris-
totle here departed from his standard view, as expressed elsewhere
and in the remainder of Metaphysics Θ , and held—for two short
lines—that objects in the external world are to be described by the
whole disjunctive phrase ‘true or false’.
One can see how this text might have arisen from an intelligent
Here we omit the ἔτι of the β text and adopt the ἐπὶ of the α text. A neutral
translation of these lines of the β text was provided in sect. .
David Charles and Michail Peramatzis
scribe’s misgivings about the α text. He [let us call him ‘B’] was
unhappy about the highly elliptical τὸ δὲ κυριώτατα in the α text:
α (EJ) τὸ δὲ κυριώτατα εἰ ἀληθὲς ἢ ψεῦδος, τοῦτο δ᾿ ἐπὶ τῶν πραγμάτων ἐστὶ
τὸ συγκεῖσθαι ἢ διῃρῆσθαι,
not seeing that the first three words should be spelt out as ‘what is
[said] most strictly to be [is present]’. As a result, he thought that
the initial clause had to be supplemented with an explicit subject
as in ‘what is most strictly’ (τὸ δὲ κυριώτατα ὄν) to form a genuine
subject, and replaced ‘if’ (εἰ) with ‘being’ (ὄν) to do so. As a result,
the phrase ‘true or false’ (ἀληθὲς ἢ ψεῦδος) was naturally taken as the
predicate. So interpreted, the whole clause said that what is in the
basic case is to be identified with what is true or false. His conclusion
(although welcome to later philosophical radicals such as Heideg-
ger) may have seemed to B to conflict with Aristotle’s standard doc-
trine, in which propositional items (and not things in the external
world) were true or false. When he spotted this difficulty, he emen-
ded the next clause to maintain Aristotle’s standard doctrine. He
wrote τοῦτο δὴ ἐπὶ τῶν πραγμάτων ἐστὶ τῷ συγκεῖσθαι ἢ διῃρῆσθαι. His
emendation involved two moves: the first was to read δή rather than
δέ, to take the connective as meaning ‘in truth’ or ‘indeed’, which
allowed him to take the whole clause as restating in a more accurate
or preferable form what had been stated (somewhat misleadingly)
in the previous one. The second was to introduce τῷ instead of τὸ at
b, interpreting the whole clause as saying that truth or falsity
depends on things being combined or divided (perhaps inspired by
the prepositional phrase of causal dependency παρὰ σύνθεσίν ἐστι καὶ
διαίρεσιν at Ε , b). In the revised text, Aristotle is not identi-
fying truth and falsity with things in the external world. In fact, he
is explicitly rejecting this view and, by reformulating the claim at
b, making truth and falsity dependent on things in the external
world (in line with his usual doctrine). So understood, the β text
arises partly from grammatical perplexity over the phrase found
in the older, α reading τὸ δὲ κυριώτατα εἰ ἀληθὲς ἢ ψεῦδος (due to
B’s failure to see the elliptical reading suggested above) and partly
from an understandable desire to make this passage consistent with
Aristotle’s views elsewhere. Given this genealogy of the β text, we
should turn to it only if we are unable to read the α text in a way
which is consistent with Aristotle’s claims elsewhere. It is impor-
These lines of the α text were translated in sect. , in our [T].
Aristotle on Truth-Bearers
tant to emphasize that there is no comparable account of the route
from the β text to the more elliptical α text reading: τὸ δὲ κυριώτατα
εἰ ἀληθὲς ἢ ψεῦδος.
While interpreters who wish to find in these lines a view of truth
consistent with Aristotle’s account elsewhere rely on the β text,
they do not defend it when it departs from the α text. Giles Pear-
son exemplifies this approach, accepting the β text at crucial points
without considering (or supporting) its distinctive readings. How-
ever, this strategy is vulnerable to criticism by those who claim that
his style of interpretation can be made to work only on the basis
of the β text, which they see as having probably been devised to
support precisely the type of interpretation Pearson (and we) fa-
vour. This is why, for example, in their view, the β text replaces the
nominative at b with a dative to make ‘truth-as-being’ arise
because of things’ being composed and divided.
In contrast with Pearson’s account, interpreters aware of the ge-
neral strengths of the α manuscript tradition have sought to fol-
low it as closely as possible, even when it leads them to attribute
to Aristotle views about truth which are either in tension with his
account elsewhere or philosophically counter-intuitive (or both).
Paolo Crivelli seeks (as we do) to defend the majority of the α text.
However, we diverge on one crucial point: he replaces the ‘if’ (εἰ)
of the α text at b with ‘being’ (ὄν), following the β text. This
seemingly small emendation leads him to a radical conclusion: Aris-
totle here commits himself to external truth-bearers, modifying or
correcting what he said elsewhere. On this view, being true or false
is straightforwardly a characterization of external things, presum-
ably objects which are combined or divided (or which consist of
combinations or divisions). In Crivelli’s view, the relevant external
entities are states of affairs. He understands the difficult phrase at
b in the β text, τὸ δὲ κυριώτατα ὂν ἀληθὲς ἢ ψεῦδος, to speak of
a strict, or the strictest, notion of truth and falsity.
We do not accept Crivelli’s construal of the τοῦτο at b as
referring to ἀληθὲς ἢ ψεῦδος at b–. As we have shown above,
if one retains the α text in its entirety (with the ‘if’ (εἰ) at b),
Pearson, ‘Being-as-Truth’, – (including n. ). Crivelli, Truth.
We shall not discuss at all this last part of Crivelli’s view.
Pearson suggests, without endorsing, a similar understanding of this phrase
(‘Being-as-Truth’, n. ): ‘that which is most properly called truth or falsity’. By
contrast, Heidegger (Grundbegriffe, , –) identifies what is in the strictest
sense with what is true.
David Charles and Michail Peramatzis
‘this’ can be taken (in the simplest reading) to refer back to what is
[said] in the strictest way and identified (in the case of complexes
such as Socrates plus being healthy) with the objects in question
being combined or divided. There is no need, on the basis of this
passage, to attribute to Aristotle a new way of being true and false
which straightforwardly characterizes external objects.
Crivelli’s proposal faces an immediate difficulty: since the
concept of truth or falsity deployed in Metaphysics E applies to
thoughts and not to things, it is unlike the one mentioned here
(on Crivelli’s reading). His response is to suggest that Metaphysics
E is referring not to the strictest notion of truth (introduced at
b–) but to a looser one which applies to thoughts but not to
things. In this way, the clash between Metaphysics E and Θ
is only apparent: for the earlier chapter claims only that a ‘looser’
notion of truth or falsity does not belong to things, whereas the
later chapter holds that the ‘strictest’ notion of truth or falsity does
indeed belong to things.
Crivelli’s ingenious suggestion has several disadvantages. First,
on his view, while the strictest notion of truth or falsity applies to
things, the looser one (which applies to thoughts, statements, men-
tal states, etc.) does not apply to things at all. Hence, he has to main-
tain that in Aristotle’s view x can be strictly speaking true or false,
without x being true or false. This result, however, seems not only
counter-intuitive in its own right but also contrary to Aristotle’s
standard practice: for he normally holds that if x is strictly speak-
ing F, then it is also F (non-strictly speaking), whereas the con-
verse is not the case. Second, it is unclear why Aristotle here (on
this one occasion) would have introduced a new demanding sense
of truth and falsity with no apparent similar uses of this notion (as
a disjunction) elsewhere. It appears to emerge unmotivated in the
present context (see also the fourth concern below).
Crivelli’s proposal is not without its philological difficulties. The
presence of a connective (whether δέ or δή) at b is surprising
There is, however, another possibility (as mentioned above): even if the condi-
tional clause is indeed the referent of ‘this’, the phrase ‘for things’ (ἐπὶ τῶν πραγμά-
των) serves to shift the reference back to the strictest way of being (‘this, being true
or false, from the side of things is . . .’).
See Crivelli, Truth, –, –; also his ‘Truth in Metaphysics Ε ’, Ox-
ford Studies in Ancient Philosophy, (), –. For criticism see U. Coope,
‘Review of Paolo Crivelli, Aristotle on Truth’, Notre Dame Philosophical Reviews,
.. <http://ndpr.nd.edu/news/-aristotle-on-truth/>.
Aristotle on Truth-Bearers
given that Crivelli thinks that what comes next, things’ combination
or division, is identical with the strictest notion of truth and falsity
(i.e. the subject-matter of the preceding clause, τὸ δὲ κυριώτατα ὂν
ἀληθὲς ἢ ψεῦδος). One would not have expected to find a connective
here at all. Further, it is difficult to understand b– as sup-
porting the reading of κυριώτατα ὄν as qualifying ἀληθὲς ἢ ψεῦδος,
and as introducing the ‘strictest’ notion of truth or falsity. For the
previous clauses had been concerned with ‘being’, not with looser
senses of truth. On Crivelli’s view, Aristotle moves directly from
talking of being to talk of truth without any sign of introducing a
new topic.
Our interpretation avoids all these difficulties. In adopting the
α text in its entirety, we do not ‘cherry-pick’ (as Crivelli does) some
and only some parts of that text. Second, our reading of ‘if true or
false’ (εἰ ἀληθὲς ἢ ψεῦδος) allows us to see this text as fully consistent
with Aristotle’s standard doctrine elsewhere: propositional items
are the prime bearers of what is true or false. There is no tension to
resolve with Metaphysics E . Third, our reading (as noted above)
makes good sense of the contrastive δέ in b, contrasting truth
and falsity as properties of propositions with things in the external
world. Finally, there is no startling (and unannounced) change of
subject from talk of being to talk of what is in the strictest way true
or false, cutting these clauses off from what precedes.
To conclude: since the α text (in its entirety) makes good philo-
sophical sense, there is no reason—in terms either of philosophical
gain or of philological conservatism (faithfulness to the best manu-
script tradition etc.)—to move away from it (at this point). Indeed,
all attempts to introduce aspects of the β text (whether wholesale or
in part) give rise to serious philosophical or exegetical problems of
their own. The α text, then, should be retained (at this point) in its
entirety. Further, we have shown how, in this case, the β text may
have arisen through the (misguided) ingenuity of an acute and phi-
losophically informed scribe. The distinctive features of the α text,
by contrast, are clearly not the results of purely mechanical failures
in transmission or enhancement by the addition of marginal notes.
Our remarks suggest a more general methodological principle.
Pearson’s proposal also does not explain (or mention) the presence of a connec-
tive at this point.
A similar problem arises for Pearson, who suggests that we might understand
the relevant phrase as ‘that which is most properly called truth or falsity’. Nor is it
clear what, on this proposal, is less properly called truth or falsity.
David Charles and Michail Peramatzis
It should be the first aim of interpreters of the Metaphysics to
give a philosophically coherent reading of the α text (subject to
the two conditions just mentioned) consistent with Aristotle’s
views elsewhere. If this principle is accepted, our philosophic-
ally consistent and plausible reading of Aristotle’s discussion of
truth-bearers, which is based on the α text, is to be preferred to
all other readings currently on offer. The latter would depart (to a
greater or lesser degree) from the α text, even were they to succeed
in making Aristotle’s views philosophically intriguing (as perhaps
with Heidegger’s ‘unhidden-ness’ of truth) or fully consistent with
his claims elsewhere.
It is only when this primary aim cannot be realized that we need
to adjudicate between the demands of philological conservatism and
philosophical desiderata such as overall consistency, charity, etc.
Since the latter judgements are complex and contestable, the phi-
losophical scholar should avoid them whenever possible.
BI B L I O G R A P HY
Berti, E., ‘I luoghi della verità secondo Aristotele: un confronto con Hei-
degger’, in V. Melchiorre (ed.), I luoghi del comprendere (Milan, ),
–.
Bonitz, H. (ed. and comm.), Aristotelis Metaphysica, vols. (Bonn,
–).
Burnyeat, M. F., et al., Notes on Eta and Theta of Aristotle’s Metaphysics
[Eta] (Oxford, ).
Christ, W., Aristotelis Metaphysica (Leipzig, ).
Coope, U., ‘Review of Paolo Crivelli, Aristotle on Truth’, Notre Dame
Philosophical Reviews, .. <http://ndpr.nd.edu/news/-
aristotle-on-truth/>.
Crivelli, P., Aristotle on Truth [Truth] (Cambridge, ).
Crivelli, P., ‘Truth in Metaphysics Ε ’, Oxford Studies in Ancient Philo-
sophy, (), –.
Dummett, M., ‘Realism and the Theory of Meaning’, in id., The Logical
Basis of Metaphysics (London, ), –.
Dummett, M., ‘Truth’, Proceedings of the Aristotelian Society, (–),
–.
Heidegger, M., Die Grundbegriffe der antiken Philosophie [Grundbegriffe]
(), ed. F.-K. Blust (Frankfurt a.M., ).
Jaeger, W., Aristotelis Metaphysica (Oxford, ).
Pearson, G., ‘Aristotle on Being-as-Truth’ [‘Being-as-Truth’], Oxford
Studies in Ancient Philosophy, (), –.
does not follow that he was a ‘correspondence theorist’, committed to states of affairs
or facts as truth-makers.
Aristotle on Truth-Bearers
Primavesi, O., ‘Introduction: The Transmission of the Text and the Riddle
of the Two Versions’, in C. Steel (ed.), Aristotle’s Metaphysics Alpha
(Oxford, ), –.
Ross, W. D., Aristotle: Metaphysics [Metaphysics], vols. (Oxford, ).
Whitaker, C. W. A., Aristotle’s De interpretatione: Contradiction and Dia-
lectic (Oxford, ).
THE S T O I C A R G UMENT
F R O M OI K E I Ō SI S
JACOB KLEIN
. Introduction
R. Salles (ed.), Metaphysics, Soul, and Ethics in Ancient Thought: Themes from the
Work of Richard Sorabji (Oxford, ), –; C. Gill, The Structured Self in Hel-
lenistic and Roman Thought [Self] (Oxford, ), –; W. Kühn, ‘L’attachement
à soi et aux autres’ [‘L’attachement’], in M.-O. Goulet-Cazé (ed.), Études sur la
théorie stoïcienne de l’action (Paris, ), –. On the doctrine’s argumenta-
tive structure and place in Stoic ethics see G. Striker, ‘The Role of oikeiōsis in Stoic
Ethics’ [‘Role’], in ead., Essays on Hellenistic Epistemology and Ethics (Cambridge,
), –; ‘Following Nature’, ibid. –; M. Frede, ‘On the Stoic Concep-
tion of the Good’ [‘Good’], in K. Ierodiakonou (ed.), Topics in Stoic Philosophy
(Oxford, ), –; W. Kühn, ‘L’attachement à soi face aux fins morales: une
question stoïcienne’, Archiv für Geschichte der Philosophie, (), –. On οἰ-
κείωσις in its Roman context see G. Reydams-Schils, ‘Human Bonding and oikeiōsis
in Roman Stoicism’ [‘Bonding’], Oxford Studies in Ancient Philosophy, (),
–. For the doctrine’s legacy to Roman political thought see M. Schofield, ‘Epi-
curean and Stoic Political Thought’, in C. Rowe and M. Schofield (eds.), The Cam-
bridge History of Greek and Roman Political Thought (New York and Cambridge,
), –, esp. –. Cf. also M. Schofield, ‘Two Stoic Approaches to Justice’
[‘Justice’], in A. Laks and M. Schofield (eds.), Justice and Generosity: Studies in
Hellenistic Social and Political Philosophy (Cambridge, ), –. On vari-
ous interpretative difficulties, including the alleged gap between personal and so-
cial οἰκείωσις discussed below, see M.-A. Zagdoun, ‘Problèmes concernant l’oikeiōsis
stoïcienne’ [‘Problèmes’], in G. Romeyer-Dherbey and J.-B. Gourinat (eds.), Les
Stoïciens (Paris, ), –. For an interesting if speculative account of the doc-
trine’s influence on Lucan’s Bellum civile see D. George, ‘Lucan’s Caesar and Stoic
oikeiōsis Theory: The Stoic Fool’, Transactions of the American Philological Asso-
ciation, (), –. P. Schmitz argues that Cicero’s account in De finibus
contains significant Peripatetic distortions (‘Cato Peripateticus’: Stoische und peripa-
tetische Ethik im Dialog. Cic. ‘fin.’ und der Aristotelismus des ersten Jh. v. Chr. [Cato]
(Berlin, )). A thorough bibliography of the literature on οἰκείωσις is provided in
I. Ramelli, Hierocles the Stoic: Elements of Ethics [Hierocles], trans. D. Konstan
(Atlanta, ), –.
Broadly Socratic, that is, in view of texts such as Crito –, where to live
well (τὸ εὖ ζῆν) is to live finely and justly (καλῶς καὶ δικαίως).
Jacob Klein
form of behaviour: the readily observable tendency of animals to
care for themselves by pursuing what is conducive to their own
survival and avoiding what is not. Though commentators have
offered various suggestions about the relation between the motive
of self-preservation and the other-regarding dimensions of Stoic
ethics, there is no consensus about how (or indeed whether) the
Stoics integrated them within a single account. The self-regarding
focus of the oikeiōsis doctrine seems to cut against the cosmopolitan
tenor of Stoic ethics.
A second difficulty is posed by the survival of a closely parallel
theory—or cluster of theories—associated with the late Academic
tradition and deriving, as most commentators now agree, from An-
tiochus of Ascalon. Antiochus’ adoption and prominent use of the
oikeiōsis theory tend to confirm its centrality to Stoic ethics, but they
also obscure our view of the doctrine’s role in early Stoic theory. In
later Academic versions of the doctrine, a structurally similar ac-
count of moral development that also begins from the motivational
patterns apparent in animal behaviour is made to support a con-
ception of the human telos that differs in crucial respects from the
Stoics’ own. In the Antiochean accounts of De finibus and , for in-
stance, the motivational impulses of children and animals are said
to confirm a conception of the end according to which states and
conditions external to virtue are goods in their own right and con-
tribute, together with virtue itself, to the happiness of human agents
(Cic. Fin. . –, –; . –). This analysis of the human
good differs importantly from the Stoic identification of goodness
with virtue alone; yet it is supposed to follow from some of the same
On Antiochus’ relation to the Academy see especially M. Bonazzi, ‘Antiochus’
Ethics and the Subordination of Stoicism’, in M. Bonazzi and J. Opsomer (eds.),
The Origins of the Platonic System: Platonisms of the Early Empire and their Phi-
losophical Contexts (Leuven, ), –; Gill, Self, –; R. Polito, ‘Antiochus
and the Academy’, in D. Sedley (ed.), The Philosophy of Antiochus (Oxford, ),
–; D. Sedley, ‘Antiochus as Historian of Philosophy’, ibid. –. On the An-
tiochean background to the oikeiōsis accounts of De finibus and see C. Brittain,
‘Antiochus’ Epistemology’, in Sedley (ed.), The Philosophy of Antiochus, – at
–; B. Inwood, ‘Antiochus on Physics’ [‘Antiochus’], ibid. – at –;
Inwood, Ethics after Aristotle (Cambridge, Mass., ), –.
These accounts resemble the theory set out in Arius Didymus’ survey of Peri-
patetic ethics, preserved by Stobaeus (. –). On the latter see especially H.
Görgemanns, ‘Oikeiōsis in Arius Didymus’ [‘Oikeiōsis’], in W. W. Fortenbaugh (ed.),
On Stoic and Peripatetic Ethics: The Work of Arius Didymus (New Brunswick, NJ,
), –; B. Inwood, ‘Comments on Professor Görgemann’s Paper’ [‘Com-
ments’], ibid. –; Ethics after Aristotle, –.
The Stoic Argument from oikeiōsis
motivational tendencies the Stoics affirm and from the primacy of
self-preservation in particular. By contrast, Antiochus represents
the Stoics’ own conclusions as incompatible with the empirical ob-
servations from which the Stoics begin (Cic. Fin. . –).
Despite these challenges, I want to suggest that there is room for
further attention to the doctrine’s purported ethical import. Ap-
peals to the character of neonatal motivation figure in the ethical
arguments of each of the main Hellenistic schools, and in each case
they seem intended to clarify the structure of fully rational moti-
vation in human agents. The Stoic doctrine of oikeiōsis follows this
pattern in two important respects. First, it treats the capacity for
perception, and the capacity for self-perception in particular, as the
psychological basis of the activities appropriate to animate orga-
nisms in virtue of the constitutions given to them by nature. Se-
cond, it treats the perfection of these capacities and the functions
they control as a sufficient condition of an organism’s teleological
success. These claims are continuous with the Stoic analysis of hu-
man agency, and they answer to the central claims of Stoic ethics:
viz., that virtue consists in a cognitive condition that centrally in-
cludes self-knowledge and that, together with the activities to which
it gives rise, this condition is constitutive of the human good. Stoic
sources insist, as Academic and Peripatetic accounts do not, that
animals are born with a capacity to perceive themselves and their
situation in the world, and that this capacity enables them to co-
ordinate their actions in a way that is appropriate both to their sur-
roundings and to their distinctive constitutions.
Strictly speaking, the Stoics identify goodness with virtue and what participates
in virtue, including especially the activities to which virtue gives rise (D.L. . =
SVF iii. ; Stob. . = SVF iii. ).
I use ‘animate organisms’ to distinguish organisms that possess psuchē from
those that do not in the Stoic scheme. Such a phrase would be redundant in a dis-
cussion of Aristotle’s biology, which treats every living organism as ensouled.
Here and throughout, I employ ‘cognitive’ to characterize a mental state whose
functional role is to represent the world. In this usage ‘cognitive’ applies quite gener-
ally to representational states such as perceptions, beliefs, and judgements no less
than to knowledge. This clarification is important since ‘cognitive’ and ‘cognitive
impression’ are sometimes used to translate the Stoic technical terms καταληπτικός
and φαντασία καταληπτική. So employed, ‘cognitive’ carries a further sense I do not
intend, that of warrant. A warranted belief results, in the Stoics’ view, from assent
to impressions that () are true, () precisely represent their object, and () have (on
the interpretation I accept) a phenomenal character distinct from mental representa-
tions that fail conditions () or (). In my usage, a cognitive mental state possesses a
representational ‘direction of fit’ but need not satisfy any of these further conditions.
Jacob Klein
I will argue that these elements of the oikeiōsis doctrine help to
clarify its role in Stoic ethics. Though the characterization of animal
behaviour central to each of the oikeiōsis accounts does not appear to
constitute an argument for the Stoic analysis of the human telos in
its own right, it does constitute such an argument when conjoined
to a normative assumption the Stoics share with other Hellenistic
schools: roughly, that the earliest object of motivation in animals
and human infants corresponds to the object of motivation in fully
rational human agents, thus providing a guide to the basic charac-
ter of the human end. My aim here is not to defend this assump-
tion but rather to reconstruct its role in Stoic theory as part of an
argument for the Stoic account of the human good. If this assump-
tion is a common starting-point of Hellenistic cradle arguments, as
Jacques Brunschwig has argued, then the analysis of animal psycho-
logy that survives in fragmentary discussions of Stoic oikeiōsis can
be seen to motivate the central tenets of Stoic ethical theory in clear
respects. The Stoics explain the complex, goal-directed behaviours
of animals by appealing to the perceptive and proprioceptive capa-
cities with which they are born. This focus on animal perception
supports the cognitive analysis of virtue the Stoics accept in the hu-
man case: what animals do on the basis of non-rational perception,
the Stoics claim, human beings do on the basis of rational, concep-
tually structured perception and cognition. The Stoic understand-
ing of human virtue as a cognitive condition that centrally includes
self-knowledge is thus one instance of a wider analysis that makes
accurate cognition the basis of appropriate action and teleological
success in rational and non-rational animals alike.
If this reconstruction is correct, the Stoics argue for the primacy
of cognition in their explanation of animal behaviour because they
Since talk of motivation and pursuit introduces an opaque context, it is hard
to formulate this assumption in a way that is both precise and which covers its use
by Stoics, Epicureans, and Academics. One might say that each school appears to
assume that the kind (or kinds) of thing pursued—in the referentially transparent
sense—by infants and non-rational animals is also and exclusively the kind (or kinds)
of thing pursued—in the referentially transparent sense—by fully rational human
agents. For the Epicureans this object is pleasure. For the Stoics, as I argue below,
it is the integrity of the organism’s ἡγεμονικόν and the appropriate functions (κα-
θήκοντα) this secures. There may of course be differences in the specific activities
prescribed to rational and non-rational animals under this assumption, as well as dif-
ferences in the structure and sophistication of their motivations. The supposition is
not that these motivations and activities will be alike in the rational and non-rational
cases but that they will have, as their final object, the same kind of thing. Cf. n.
below.
The Stoic Argument from oikeiōsis
wish to argue that appropriate action is also achieved by human
agents through a cognitive grasp of one’s own constitution and
through the perfection of the faculty on which this grasp depends.
The psychological background of the oikeiōsis account is crucial to
making sense of the doctrine’s role in Stoic ethical theory, for it
suggests an effort by the Stoics to extend key elements of Socratic
psychology to a much broader analysis of the mechanisms by which
animate organisms are regulated by nature so as to achieve their
ends. In what follows, I argue for this interpretation in three stages.
I first offer a brief survey of the available evidence for the theory
of oikeiōsis. I then emphasize a number of difficulties raised by
recent interpretations of this evidence. Finally, I suggest a revised
account of the way in which the oikeiōsis theory supports the central
claims of Stoic ethics: that virtue consists in a cognitive grasp of the
natural order, and that this condition, when perfected, is sufficient
for achieving the human telos.
The state of the available evidence for the Stoic doctrine is com-
plicated, to say the least. It is unclear whether the Stoics them-
selves coined the verbal noun oikeiōsis, and there is no direct textual
evidence that Chrysippus himself used it, though his use of cog-
nates is well attested (Plut. Stoic. repugn. = LS E = SVF
iii. , ii. ; Galen, PHP . . – = Posid. fr. = LS M).
The term first appears in a fragment attributed to Theophrastus
and derives from a family of words occasionally put to philosophi-
cal use by Plato and Aristotle. As Pembroke observes, the verb to
which oikeiōsis is directly related is oikeioun, and this is derived in
turn from the adjective oikeion and the noun oikos. A thing or per-
son is said to be oikeion when it belongs to one either by kinship,
as in the case of family, or by possession, as in the case of property.
Some texts employ the middle-passive form oikeiousthai to suggest
that something has been made an object of care and concern for
a creature by the agency of providential nature. Oikeiōsis, on the
Phot. Bibl. , b–, fr. in W. W. Fortenbaugh, P. M. Huby, R. W.
Sharples, and D. Gutas (eds.), Theophrastus of Eresus: Sources for his Life, Writings,
Thought and Influence, vols. (Leiden, ).
Cf. Hier. El. Eth., col. . –; Plut. Stoic. repugn. (= LS E = SVF
Jacob Klein
other hand, describes an orientation that is at once both cognitive
and motivational. Plutarch (Stoic. repugn. = SVF i. ) of-
fers as an explanation of the Stoic concept the claim that oikeiōsis is
a perception (aisthēsis) and grasp (antilēpsis) of what is appropriate
(oikeion). No English rendering of oikeion is wholly satisfactory,
but the Stoics’ technical usage is well captured by Brennan: a fea-
ture of an animal’s environment may be characterized as oikeion just
in case it is a suitable object of concern for the animal.
The terminology of oikeiōsis figures in a range of texts in con-
iii. ); Galen, PHP . . – (= Posid. fr. = LS M); Alex. Aphr. Quaest. .
Bruns (= SVF iii. ); Mant. . (= SVF iii. ) and . Bruns (= SVF
iii. ). Cf. Inwood, Ethics and Human Action, – nn. and .
In Striker’s paraphrase, οἰκείωσις is the ‘recognition and appreciation of some-
thing as belonging to one’ (‘Role’, ). No single English word quite covers the
complex sense the Stoics gave to the term. ‘Appropriation’ and ‘orientation’ come
close but fail to capture the recognition of personal affinity conveyed by the Greek.
‘Appropriation’ has been the predominant translation since Long and Sedley, how-
ever, who observe that it ‘provides a means, through the verb or adjective “appro-
priate”, of rendering grammatically related forms of the Greek root oik-’ (A. A.
Long and D. Sedley (eds.), The Hellenistic Philosophers [LS], vols. (Cambridge,
), i. ). ‘Appropriation’ also preserves the connection between οἰκείωσις and
appropriate (καθῆκον) action, which the Stoics intend (D.L. . = LS C = SVF
iii. ), and I have occasionally rendered καθῆκον by ‘appropriate’ as well. Besides
that of Long and Sedley, discussions of the term and its cognates include Pembroke,
‘Oikeiōsis’, – and –; Kerferd, ‘Search’, –; Inwood, Ethics and Hu-
man Action, –; Ramelli, Hierocles, . For the use of related terminology in Plato
and Aristotle see esp. Pohlenz, Grundfragen, n. ; Pembroke ‘Oikeiōsis’, –;
Kerferd, ‘Search’, –; Brennan, The Stoic Life, –. In Greek ἀλλότριον is
the contrast term for οἰκεῖον, and ἀλλοτρίωσις (alienation, estrangement) is the cor-
responding verbal noun. Cicero typically renders οἰκείωσις with either commendatio
or conciliatio, with conciliatum and alienum answering to the Greek οἰκεῖον and ἀλλό-
τριον.
Inwood doubts Plutarch’s trustworthiness in this passage on the grounds that
οἰκείωσις ‘depends on perception but is not itself a form of perception’ (Ethics and
Human Action, n. ; cf. Striker, ‘Role’, n. ). This is no doubt correct,
but it is not clear to me that Plutarch’s usage in this regard is any looser than the
Stoics’ own. Given the details of Stoic psychology, which is built around a cogni-
tive analysis of motivation, there is a tight connection between the recognition of
something as οἰκεῖον and the consequent motivation to pursue it (cf. n. below).
Porphyry similarly maintains that perception (τὸ αἰσθάνεσθαι) is the principle (ἀρχή)
of οἰκείωσις and ἀλλοτρίωσις (Abst. . = SVF i. ).
See Brennan, The Stoic Life, : ‘I propose that what it means to take some-
thing to be oikeion is that one treats it as an object of concern.’ As Brennan observes
(in personal corespondence), ‘concern’ must here be understood broadly enough to
cover a spectrum of cases ranging from the nurturing and benevolent to the appe-
titive and predatory. Though it sounds objectionable to describe a newly hatched
chick as οἰκεῖον to a bird of prey as well as to the mother hen, the different ways in
which the chick is of interest to each are presumably to be explained by the differing
constitutions of hens and hawks. Cf. D.L. . (= LS C = SVF iii. ).
The Stoic Argument from oikeiōsis
nection with a distinctively Stoic account of human development.
These reports differ importantly in emphasis, and it is possible to
produce different pictures of the oikeiōsis doctrine depending on
which of them are given the most weight. Two of the fullest presen-
tations, which have also received the most scholarly attention, are
those of Diogenes Laertius and Cicero. Though the shorter of the
two, Diogenes’ account is distinctive for its presentation of material
that appears to be taken from Chrysippus’ lost treatise On Ends, and
it probably gives us our most reliable glimpse of the main lines of the
early Stoic doctrine. Cicero’s version (Fin. . –), though more
detailed in some respects, is also more difficult to attribute. Cicero
does not mention any of the older Stoics by name, and his summar-
ies of the theory may follow later versions that appear to be based
(in De officiis) on Panaetius and are perhaps based (in De finibus) on
Diogenes of Babylon. Finally, there is the detailed, apparently or-
thodox but regrettably fragmentary treatise of the Stoic Hierocles,
which defends specific aspects of the Stoic doctrine against later cri-
tics. These texts can be supplemented by a helpful but incomplete
account of oikeiōsis in one of Seneca’s letters, by shorter passages in
Cicero and Aulus Gellius, and by many oblique references in Epic-
tetus and Marcus Aurelius.
Each of these sources merits individual discussion, but there
is enough common ground among them to supply us with a rea-
sonably uncontroversial overview of the Stoic theory. Diogenes,
Cicero, and Hierocles each begin by alluding to a form of self-
perception that precedes and explains an animal’s earliest impulses,
enabling it to orient and co-ordinate its activities so as to ensure
its own survival. No one, the Stoics observe, teaches a newborn
animal what its limbs and appendages are for, nor the sort of
food it needs, nor the predators it must avoid in order to survive.
His Περὶ τελῶν (D.L. . = LS A = SVF iii. ), on which see esp. Inwood,
Ethics and Human Action, –.
The main texts are D.L. . – (= LS A = SVF iii. ); Hierocles, Elements
of Ethics; Cic. Fin. . – (= LS D); Off. . –, –; . ; Sen. Ep. .
Seneca’s report mentions Posidonius and Archidemus by name. This is presum-
ably Archidemus of Tarsus, as Inwood notes (Seneca: Selected Philosophical Letters
[Seneca] (Oxford, ), ). Other useful references include Plut. Stoic. repugn.
–; Alex. Aphr. Mant. . ff. Bruns; Epict. Diss. . . –; . . ; .
. ; . . ; Ench. . . Academic and Peripatetic versions of the doctrine—
all of which are likely to derive from Antiochus and are probably influenced by
Carneades—include Cic. Fin. . ff.; . –, –; Acad. . –; Gell. . .
(= SVF iii. ); Stob. . – and esp. . –.
Jacob Klein
From birth animals display a striking sensitivity, present without
instruction, to the nature of their own faculties and to the threats
and benefits present in their environment. Hierocles argues at
length that animals continuously perceive not only what their own
physical faculties with their various limbs and appendages are for,
but also what the dispositions of other animals are for and the
actions, such as flight or aggression, that constitute an appropriate
response to them (El. Eth., cols. . –. ). The same form
of self-perception is invoked to explain a range of co-ordinated
behaviour animals display in relation to other animals, including
concern for offspring and, in some cases, co-operation with other
species. Thus a central claim of the oikeiōsis theory is that animals
are born with a capacity to perceive, in a teleologically informed
way, their appropriate relation to a range of complex features of
their environment.
The focus on animal behaviour, however, is apparently not the
central import of the Stoic theory. In each of the main accounts,
this initial focus shifts to an analysis of psychological development
in humans and, finally, to conclusions about the character of the
human good. Unfortunately, Hierocles’ otherwise continuous and
detailed account contains a lengthy lacuna at just the point at which
the case of human development and its implications are about to
be described. The summaries of Cicero and Diogenes briefly out-
line, however, what fuller articulations of the doctrine must have
described in greater detail: the way in which the initial perceptions
and attractions of pre-rational children develop in the ideal case
into the systematic, propositionally structured form of cognition in
which virtue consists. This condition is rooted in a developing set
of conceptions (ennoiai) that appear to involve, as part of their con-
tent, an increasingly articulate awareness of the kind of creature one
is and of the modes of behaviour that are appropriate as a result.
The link between self-perception and appropriate behaviour towards other spe-
cies is explicit in Hierocles (see esp. El. Eth., col. . –). A link between self-
perception and concern for offspring is suggested, though not explicit, in Plutarch’s
remarks at Stoic. repugn. –. I develop this connection below.
That an animal’s self-awareness becomes increasingly refined as it develops is
one of the final points made in Hierocles’ treatise on ethics before the manuscript be-
comes unintelligible in cols. –. For discussion of this point see G. Bastianini
and A. A. Long, ‘Hierocles, Elementa moralia’ [‘Hierocles’], in Corpus dei papiri
filosofici, pt. . Commentari (Florence, ), – at –; C. Brittain, ‘Non-
Rational Perception in the Stoics and Augustine’ [‘Perception’], Oxford Studies in
Ancient Philosophy, (), – at –.
The Stoic Argument from oikeiōsis
It is clear that the Stoics regard such conceptions as partly and per-
haps primarily constitutive of the rational faculty that guides and
explains human behaviour quite generally. Thus Diogenes de-
scribes the human transition to rational maturity as the point at
which ‘reason supervenes as the craftsman of impulse’ (D.L. . =
LS A = SVF iii. , trans. Long and Sedley), and Cicero’s ac-
count describes the eventual appreciation by human agents of the
‘order and harmony of our obligations’ (Fin. . = LS D = SVF
iii. , trans. after Rackham). These descriptions suggest that a
central goal of the oikeiōsis theory is to establish reason’s role in
shaping the motivations of adult human beings and to characterize
this role as the distinguishing mark of human agency.
. Interpretative difficulties
τὴν δὲ πρώτην ὁρμήν φασι τὸ ζῷον ἴσχειν ἐπὶ τὸ τηρεῖν ἑαυτό, οἰκειούσης αὐτὸ
〈ἑαυτῷ〉 τῆς φύσεως ἀπ᾿ ἀρχῆς, καθά φησιν ὁ Χρύσιππος ἐν τῷ πρώτῳ Περὶ τε-
λῶν, πρῶτον οἰκεῖον λέγων εἶναι παντὶ ζῴῳ τὴν αὑτοῦ σύστασιν καὶ τὴν ταύτης
συναίσθησιν· οὔτε γὰρ ἀλλοτριῶσαι εἰκὸς ἦν αὑτῷ τὸ ζῷον, οὔτε ποιήσασαν αὐτό,
μήτ᾿ ἀλλοτριῶσαι μήτ ᾿ [οὐκ] οἰκειῶσαι. ἀπολείπεται τοίνυν λέγειν συστησαμένην
αὐτὸ οἰκειῶσαι πρὸς ἑαυτό· οὕτω γὰρ τά τε βλάπτοντα διωθεῖται καὶ τὰ οἰκεῖα
προσίεται. ὃ δὲ λέγουσί τινες, πρὸς ἡδονὴν γίγνεσθαι τὴν πρώτην ὁρμὴν τοῖς ζῴ-
οις, ψεῦδος ἀποφαίνουσιν. ἐπιγέννημα γάρ φασιν, εἰ ἄρα ἔστιν, ἡδονὴν εἶναι ὅταν
αὐτὴ καθ᾿ αὑτὴν ἡ φύσις ἐπιζητήσασα τὰ ἐναρμόζοντα τῇ συστάσει ἀπολάβῃ·
ὃν τρόπον ἀφιλαρύνεται τὰ ζῷα καὶ θάλλει τὰ φυτά. οὐδέν τε, φασί, διήλλαξεν
ἡ φύσις ἐπὶ τῶν φυτῶν καὶ ἐπὶ τῶν ζῴων, ὅτε χωρὶς ὁρμῆς καὶ αἰσθήσεως κἀ-
κεῖνα οἰκονομεῖ καὶ ἐφ᾿ ἡμῶν τινα φυτοειδῶς γίνεται. ἐκ περιττοῦ δὲ τῆς ὁρμῆς
τοῖς ζῴοις ἐπιγενομένης, ᾗ συγχρώμενα πορεύεται πρὸς τὰ οἰκεῖα, τούτοις μὲν
See n. below.
Jacob Klein
τὸ κατὰ φύσιν τῷ κατὰ τὴν ὁρμὴν διοικεῖσθαι· τοῦ δὲ λόγου τοῖς λογικοῖς κατὰ
τελειοτέραν προστασίαν δεδομένου, τὸ κατὰ λόγον ζῆν ὀρθῶς γίνεσθαι 〈τού〉τοις
κατὰ φύσιν· τεχνίτης γὰρ οὗτος ἐπιγίνεται τῆς ὁρμῆς.
They [sc. the Stoics] say that an animal has self-preservation [τὸ τηρεῖν
ἑαυτό] as the object of its primary impulse, since Nature from the beginning
appropriates it 〈to itself〉, as Chrysippus says in his On Ends book . The
first thing appropriate to every animal, he says, is its own constitution and
the awareness of this. For Nature was not likely either to alienate the animal
〈from itself 〉 or to make it and then neither alienate it nor appropriate it. So
it remains to say that in constituting the animal Nature appropriated it to
itself. This is why the animal rejects what is harmful and accepts what is
appropriate. They hold it false to say as some people do that pleasure is the
object of an animal’s first impulse. For pleasure, they say, if it does occur,
is a by-product which arises only when Nature all by itself has searched
out and adopted the proper requirements for a creature’s constitution, just
as animals [then] frolic and plants bloom. Nature, they say, is no different
in regard to plants and animals at the time when it directs animals as well
as plants without impulse and sensation, and in us certain processes of a
vegetative kind take place. But since animals have the additional faculty of
impulse, through the use of which they go in search of what is appropriate
to them, what is natural for them is to be administered in accordance with
their impulse. And since reason by way of a more perfect management has
been bestowed on rational beings, to live correctly in accordance with rea-
son comes to be natural for them. For reason supervenes as the craftsman
of impulse. (D.L. . – = LS A = SVF iii. , trans. Long and Sedley,
with minor changes)
Thus Pohlenz: ‘Das Grundmotiv der Lehre ist, die Normen für die Lebens-
gestaltung aus einem Urtriebe der menschlichen Natur abzuleiten’ (Grundfra-
gen, ). Pembroke speaks of ‘the morality which the Stoics derived from self-
preservation’ (‘Oikeiōsis’, ). Inwood characterizes the motive of self-preservation
as ‘the starting-point for all value’ in Stoic theory (Ethics and Human Action, )
and attributes to Chrysippus the view that ‘Man’s commitment to virtue could
be derived . . . from the basic instinct of self-preservation’ (Ethics and Human
Action, ). According to R. Salles, ‘On [the Stoic] view, our moral evolution is
determined by the development of our concern for self-preservation’ (The Stoics on
Determinism and Compatibilism (Aldershot, ), ).
White, ‘Basis’, –.
M. Frede, ‘The Stoic Conception of Reason’ [‘Reason’], in K. Boudouris (ed.),
Hellenistic Philosophy, vols. (Athens, –), ii. – at .
The Stoic Argument from oikeiōsis
from self-preservation or even self-perfection to an exclusive in-
terest in observing and following nature’. In her view, the sage’s
motivation ‘will not simply be an enlightened form of self-love’.
The upshot, then, is that the literature on oikeiōsis reflects a basic
disagreement about the correct way to understand the rational de-
velopment of the Stoic sage and, more generally, the relationship
between the observations with which the oikeiōsis account begins
and the conclusions the Stoics go on to draw. Each of these inter-
pretations, moreover, is at points difficult to square with the avail-
able evidence. On the one hand, there is good reason to suppose
that the Stoics regard the primary impulse to self-preservation as
primary to rational agents no less than to children and non-rational
animals. This understanding seems to be required by a number of
texts, and it fits closely with the eudaimonist framework of Stoic
ethics as a whole. On the other hand, the tendency of commen-
tators to understand this impulse in narrowly egoistic terms has
prompted the assumption—traceable especially to the interpreta-
tions of Pohlenz, Brink, and Pembroke—that the Stoics must have
recognized two distinct forms of oikeiōsis, variously characterized
by commentators as outward- and inward-looking or as personal
and social forms of oikeiōsis. This distinction has prompted, in
turn, a number of deflationary conclusions about the point and
Striker, ‘Role’, ; cf. ‘Following Nature’, –.
Striker, ‘Following Nature’, . She suggests that according to Cicero’s ac-
count, ‘self-preservation is replaced as a primary goal by the desire for order and
harmony’ (ibid. ).
The emphasis on self-preservation in the rational case is perhaps strongest in
Seneca’s Letter , where Seneca relates the actions appropriate at each stage of life
to a fundamental orientation towards self that is constant through each of them and
to which a creature’s behaviour is referred. It seems clear from Seneca’s account that
this form of motivation is prior in some respect to every other form of motivational
impulse: ‘If I do all things because of concern for myself [propter curam mei], con-
cern for myself is prior to all things’ (Ep. . , my translation). For discussion
of this passage see Inwood, Seneca, . Similarly strong statements of self-interest
appear in Epictetus, who clearly intends them to apply in the case of the sage (e.g.
Diss. . . –; . . –; . . –; . . –; . . –); cf. Cic. Fin. .
(= LS F = SVF iii. ); Marc. Aur. Med. . ; Alex. Aphr. Quaest. .
Bruns (= SVF iii. ). Cf. Kühn, ‘L’attachement’, –.
See e.g. Pohlenz, Grundfragen, ; Brink, ‘Theophrastus and Zeno’, –;
Pembroke, ‘Oikeiōsis’, –; Inwood, ‘Comments’, –; Inwood, Ethics and Hu-
man Action, – and n. . Julia Annas calls oikeiōsis ‘a disjunctive notion’
(Morality, ). Cf. Blundell, ‘Parental Nature’; Schofield ‘Justice’; Lee, Oikeiosis,
–; McCabe, ‘Two Accounts’; Kühn, ‘L’attachement’, –. Doubts about
this distinction are voiced by Brennan (The Stoic Life, ) and Algra (‘Mechanism’,
n. ).
Jacob Klein
coherence of the Stoic doctrine. Thus Inwood, building on the re-
construction proposed by Brink, concludes that the Stoics intro-
duced social oikeiōsis as a ‘later graft’ onto the original theory, which
initially dealt only with appropriation to oneself. The suggestion
that the older Stoics failed to connect these distinct forms of oikeiōsis
in a coherent way is now embedded in one strand of the interpre-
tative literature. Yet it attributes to the Stoics a disjointed and ap-
parently ad hoc account.
The model favoured by Striker and Frede invites other dif-
ficulties. If the motivations of a fully rational agent are charac-
terized by a fundamental shift away from self-concern, as their
interpretations suggest, it is difficult to understand why references
to the self-preserving behaviour of animals figure so prominently
in the oikeiōsis account at all. As Jacques Brunschwig emphasizes,
Stoic and Epicurean appeals to neonatal motivation are evidently
intended to inform us, in the last analysis, of the structure and con-
tent of the human telos. In each case, these arguments offer us a
story about what does occur in infants and non-rational animals and
about what, given these starting-points, ought to occur in rational
agents. Neonatal motivation and fully rational motivation consti-
tute the two poles of the Epicurean argument, and the Epicureans’
appeal to the former is clearly intended to support their account
of the latter (D.L. . ; Cic. Fin. . – = LS A; Fin. .
–). That the Stoic account works on analogous lines is strongly
suggested by a number of sources. It is assumed by Cicero, for
instance, and it is stated quite clearly by Alexander of Aphrodisias,
who notes that disagreement about the primary object of oikeiōsis
corresponds to disagreement about the highest good, so that these
objects are correlative in an important respect. If the impulse to
Inwood, ‘Comments’, . Inwood suggests that ‘Chrysippus’ failure to forge a
firm and plausible link [between personal and social oikeiōsis] can be seen as the cause
for the confusion seen in later discussions’ (ibid. ). Again, ‘if the Stoics them-
selves had only an ad hoc explanation for the relation of the two oikeiōseis, it is less
puzzling that this late Hellenistic text [i.e. that of Arius Didymus] failed to produce a
philosophically coherent doctrine from them’ (ibid. ). Cf. Brink, ‘Theophrastus
and Zeno’, : ‘Later Stoics could use the whole range of Theophrastus’s oikeiotēs
grafted on to their own doctrines.’ For persuasive criticisms of this suggestion see
Long, ‘Theophrastus’, –.
Brunschwig, ‘Cradle’, – and –.
At Mant. . – Bruns Alexander attributes to each of the main schools what
Inwood usefully calls the ‘alignment condition’, viz. that ‘the distinguishing feature
of the primary object of desire (τὸ πρῶτον οἰκεῖον) corresponds to the distinguish-
ing feature of the ultimate object of desire’ (trans. Inwood, in Ethics after Aristotle,
The Stoic Argument from oikeiōsis
self-preservation is wholly abandoned by rational agents, why do
Stoic accounts of oikeiōsis insist on the primacy of this impulse in
the first place? To hold that human motivation undergoes a radical
shift in the ideal course of development seems to undercut the
dialectical point of the oikeiōsis account.
A central difficulty in understanding the ethical import of
oikeiōsis, therefore, is that of fitting the Stoics’ claims about animal
psychology and teleology together with their claims about rational
human agency. If we emphasize the motive of self-preservation
with which the oikeiōsis accounts begin, we seem unable to explain
the single-minded attachment to virtue that characterizes the fully
rational sage. On the other hand, if we point to discontinuities
between the animal and human cases, we seem to undermine the
dialectical point of the argument as a whole. The Stoics appear
to lack a convincing account of the psychology that bridges the
gap between non-rational and rational forms of motivation, or an
explanation of how neonatal concern for one’s self supports the
other-regarding injunctions the Stoics accept in the human case.
On the whole, commentators have not been optimistic about the
Stoics’ success in bridging these gaps or in showing how appeals
to oikeiōsis could be used to support the central tenets of Stoic
ethics. More recently some have concluded, in view of these
difficulties, that the oikeiōsis theory must not have been offered
as a justification for ethical conclusions at all. Instead of offering
grounds for accepting the Stoic analysis of virtue, it merely details
the psychological route by which this condition is acquired.
). This assumption structures the Carneadia divisio, and Alexander here applies
it for his own critical ends, much as Cicero does in De finibus . Influenced as they
are by Carneades, Alexander’s testimony and representation of the Stoic position
must be treated with care. In my view, however, what is distorted in Carneadean
accounts of οἰκείωσις is not the ‘alignment condition’ but the psychological details
that underpin Stoic versions of the argument. In particular, Carneadean accounts
obscure the distinctive analysis of the πρῶτον οἰκεῖον, with its emphasis on cognition
and self-perception, from which Stoic versions of the argument begin. As I argue
below, this distortion lies behind the allegation—implied in De finibus , explicit in
De finibus —that the Stoic account of the human telos is inconsistent with the moti-
vational starting-points the Stoics themselves accept. See further n. below.
Brink’s article seems to have introduced the talk of a ‘gap’ between personal and
social oikeiōsis into the anglophone literature (‘Theophrastus and Zeno’, –). Cf.
Pembroke, ‘Oikeiōsis’ –; B. Inwood, ‘Hierocles: Theory and Argument in the
Second Century ..’ [‘Hierocles]’, Oxford Studies in Ancient Philosophy, (),
– at –; Zagdoun, ‘Problèmes’, .
Striker, ‘Role’, ; S. Menn, ‘Physics as a Virtue’, Proceedings of the Boston
Jacob Klein
The interpretative claim I want to develop is that the Stoic argu-
ment fits the dialectical pattern prominent in other Hellenistic ap-
peals to neonatal motivation. The Stoic theory of oikeiōsis is more
than a description of the process by which rational agents acquire
the beliefs in which virtue consists; it is also offered as a justification
of Stoic claims about the cognitive character of virtue and about the
sufficiency of virtue for happiness. My reconstruction takes as a
starting-point two pieces of evidence for the Stoic view. () Seneca
(Ep. . = LS F = SVF iii. ) and Alexander of Aphrodisias
(Mant. . – Bruns = SVF iii. ) confirm that, according to at
least some of the Stoics, the object of a creature’s primary oikeiōsis
relation is not self-preservation narrowly construed, but the pre-
servation (tērēsis) of its constitution (sustasis/constitutio). () Ga-
len preserves a fragment of Posidonius in which Chrysippus is said
to have restricted the oikeion to the fine (kalon) alone (PHP . .
– = Posidonius fr. = LS M). An animal’s impulse to self-
preservation, I will argue, should not be identified with an impulse
to pursue its physical well-being per se, nor with a form of motiva-
tion that is radically altered or replaced in the fully rational case.
Rather, it should be identified with an impulse to preserve (tērein)
its leading faculty or hēgemonikon in a condition of conformity to
nature. Being directed at the hēgemonikon itself, the primary im-
pulse is prior not in order of time or strength, but in so far as it
has as its object the preservation of the faculty in which each of an
Area Colloquium in Ancient Philosophy, (), – at and n. ; Brennan,
The Stoic Life, –.
This latter claim should be distinguished clearly from a third possibility the
Stoics do not intend: that the oikeiōsis account supplies epistemic grounds for the
beliefs in which virtue consists. It is clear that no one could acquire the demanding
cognitive condition the Stoics identify with virtue merely by accepting basic theore-
tical claims about virtue. As I argue below, the oikeiōsis account offers one kind of
support for Stoic claims about the character of virtue and the objects towards which
virtuous motivation is directed. One does not acquire the set of beliefs in which vir-
tue consists merely by accepting these claims, however.
Alexander reports that those Stoics who ‘are thought to speak more subtly and
to make more distinctions about this say that it is to our own constitution [σύστα-
σιν] and its preservation [τήρησιν] that we have been appropriated’ (Mant. . –
Bruns, trans. Sharples). Pohlenz takes this to refer to Chrysippus (Grundfragen,
–). Here appropriation to one’s σύστασις is contrasted with appropriation to one-
self simpliciter. Seneca’s Letter is devoted to considering whether ‘animals have
an awareness of their own constitution’ (Ep. . , trans. Inwood). Cic. Fin. .
(= SVF iii. ) preserves a similar distinction, noting that an animal is appropriated
to its constitution (‘suum statum’) in particular. See esp. Inwood, Ethics and Human
Action , nn. and , and Seneca, –.
The Stoic Argument from oikeiōsis
animal’s impressions and impulses originates. As Nathan Powers
has recently emphasized, the Stoics ‘ascribe to the hēgemonikon con-
trol over all an animal’s life functions’. An animal’s impulse to
preserve this faculty must then be a higher-order impulse to pre-
serve the integrity of the faculty that comprehensively determines
its first-order motivations. It is primary, that is to say, in that it un-
derlies each of an animal’s activities and explains its inclination to
perform them. It is best thought of, I believe, as an orientation to-
wards appropriate function, an animal’s disposition to carry out the
kinds of activities implicit in its physical constitution. This disposi-
tion is made possible, in turn, by the mechanism of self-perception,
which enables an organism to grasp the ends for which it has been
framed by nature.
This understanding of the prōtē hormē helps to resolve the ten-
sions I have noted. On this account, the Stoics do not view the
primary impulse as a distinct form of oikeiōsis underwriting a dis-
tinction between oikeiōsis towards self and oikeiōsis towards others.
Instead, they view an animal’s constitution—the prōton oikeion that
is the object of its primary oikeiōsis relation—as a template for each
of its first-order impulses: a specification of the patterns of beha-
viour appropriate to a creature of its kind. Thus the vulnerability
of a hen’s constitution determines, under relevant conditions, that
flight is an appropriate response to predators, while the constitution
of the pea crab determines the appropriateness of its co-operation
with the pinna mollusc. So understood, the primary impulse will
be satisfied not simply by those activities that promote an orga-
nism’s physical survival, but by the full range of behaviours the
Stoics regard as appropriate (kathēkon) to the organism. In a fully
rational agent, this impulse is indeed a form of appropriation to
D.L. . (= SVF ii. ): ‘By the ruling part of the soul [ἡγεμονικόν] is meant
that which is most truly soul proper, in which arise presentations [αἱ φαντασίαι] and
impulses [αἱ ὁρμαί] and from which issues rational speech’ (trans. Hicks). Accord-
ing to Arius Didymus, ‘Every soul has some ruling faculty [ἡγεμονικόν τι] in it; and
this is its life and perception and impulse’ (fr. phys. Diels = SVF ii. , trans.
Powers). Cf. S.E. M. . ; . –.
N. Powers, ‘The Stoic Argument for the Rationality of the Cosmos’ [‘Argu-
ment’], Oxford Studies in Ancient Philosophy, (), – at .
Seneca gives the example of a hen (Ep. . ). Cicero, Plutarch, and
Athenaeus independently describe the interest Chrysippus took in the pea crab,
pinnotheres pisum (whose name preserves the Greek verb τηρεῖν), because of its
supposedly commensalistic relation to the pinna mollusc (Cic. Fin. . = LS F =
SVF iii. ; ND . – = SVF ii. ; Plut. Soll. an. = SVF ii. b; Athen.
– = SVF ii. a).
Jacob Klein
what is fine since virtue, as the Stoics conceive it, consists in a per-
fected state of the hēgemonikon and in the activities which flow from
this condition. Under this interpretation, Alexander is to be taken
seriously when he observes that Hellenistic disagreements about
the nature of the prōton oikeion correspond to disagreements about
the structure of the human telos. As a condition of the hēgemonikon,
virtue will itself be the primary object of oikeiōsis for fully rational
agents, and the concern to guard and preserve their own virtue will
underwrite and condition each of their first-order activities and im-
pulses. On this account, an animal’s primary impulse to preserve
its constitution answers in a straightforward way to a fully rational
agent’s concern to preserve her virtue.
This reconstruction fits the broader pattern of Hellenistic argu-
ment emphasized by Brunschwig, and I will argue that it is sup-
ported by a range of considerations drawn from Stoic physics and
psychology. In particular, the Stoic argument can be seen to rest on
a wider analysis of the conditions under which animate organisms
quite generally achieve their telos. The key to its structure is the
supposition that the Stoics apply a single criterion of teleological
success to organisms at each level of the scala naturae. At each level
of the cosmic order, the hēgemonikon is also a mode of pneuma, the
direct vehicle of divine activity in the physical cosmos. In the hu-
man case, the physical perfection of the rational hēgemonikon—i.e.
virtue—is clearly sufficient for the realization of the human end. If
one considers the evidence for Stoicism that is securely free of Aca-
demic influence, it appears quite plausible to suppose that other
animate organisms similarly achieve their ends, in the Stoics’ view,
when they preserve the faculty in which each of their psychic mo-
tions originates, so that their activities are correctly administered
by nature as a result. If that supposition is correct, the explana-
A parallel made largely explicit by Cicero at Tusc. . –.
I borrow some of this language from J. Cooper, ‘Chrysippus on Physical Ele-
ments’ [‘Elements’], in R. Salles (ed.), God and Cosmos in Stoicism (Oxford, ),
– at n. .
The term ‘preservation’ glosses over some complexities, including especially
the fact that the sage appears to achieve virtue only through a lengthy process by
which she frees herself from falsehood and previously accumulated cognitive error.
To speak of preserving the ἡγεμονικόν may thus seem to imply continuity of a sort of
perfection that was never there to begin with. Here it is important to distinguish two
kinds of development that are implicit but not clearly distinguished in the sources
that survive. On the one hand, the activities of animals as the Stoics conceive them
appear to conform to nature more or less regularly from birth to death (cf. nn.
The Stoic Argument from oikeiōsis
tion of animal behaviour developed by the Stoics is part of a general
analysis that supports their claim that virtue is sufficient for hap-
piness. This interpretation helps to explain why Stoic texts appear
to treat the primary impulse as the unified basis of appropriate ac-
tions (kathēkonta), whether self- or other-directed, and also why the
Stoics regard an analysis of animal psychology as instructive for the
human case.
and below). In the human case, however, this developmental process goes off
the rails at an early stage because of human beings’ susceptibility to external influ-
ences (on which see esp. M. Graver, Stoicism and Emotion (Chicago, ), ch. ).
The starting-points of nature, says Diogenes, are never corrupt, but rational agents
are corrupted, ‘sometimes by the persuasiveness of things from without, sometimes
through the teaching of associates’ (D.L. . = SVF iii. , trans. Graver). As a
result, the ideal of sagehood seems usually to be conceived of as a kind of recovery
from errors that universally affect human agents (cf. esp. Galen, PHP . . –).
This picture supposes another ideal in the background of the Stoic view, however:
the theoretical possibility of a human being whose cognitions are never distorted in
the first place, so that her actions conform to nature from start to finish. It is in this
sense that preserving one’s ἡγεμονικόν would presumably guarantee conformity to
nature all along, a realization of the τέλος at every stage of development. The pos-
sibility of this sort of diachronic conformity to nature seems to be envisioned by
Seneca’s discussion of οἰκείωσις in Letter .
Inwood suggests that ‘Hierocles’ decision to focus on self-perception as the
central question is both deliberate and unusual’ (‘Hierocles’, ). See further ibid.
: ‘Why this obsession with one theme [self-perception], to the disadvantage of
hormē and of oikeiōsis itself?’ Contrast Long, ‘Representation’, – and n. .
I agree with Long that perception and self-perception are foundational to the Stoic
theory. As I argue below, these features of the οἰκείωσις doctrine support the Stoic
claim that virtue is a wholly cognitive condition.
Many of the Stoics’ examples of self-preserving animal behaviour belong to a
common stock of received zoological wisdom. It is useful to compare the examples
Jacob Klein
What distinguishes the position of each school is rather the details
of the psychological account offered to explain it. Stoic sources in-
sist, as Antiochean sources do not, that animal activity of any sort is
made possible only by the animal’s perception of itself, and that this
awareness always accompanies and informs its perception of what
is external (Cic. Fin. . = SVF iii. ; Hier. El. Eth., col. .
–). Second, it is clear from late discussions of oikeiōsis that
Stoic claims about self-perception—rather than self-preservation
per se—became a particular point of contention with rival schools,
one that surfaces provocatively in overtly ethical contexts. Thus
Seneca argues, in a letter ostensibly devoted to ethical concerns,
that an animal is able to co-ordinate its movements only through
the perception (sensus) of its own constitution (constitutio) and that
a capacity for self-perception, even if confused or inarticulate, must
therefore be present at birth (Ep. . = LS B). Similarly, the
primary aim of Hierocles’ ethical treatise is to defend the Stoic
theory of oikeiōsis against critics who reject Stoic claims about
perception in particular.
These points suggest that we should not expect the ethical sig-
nificance of the Stoic argument to turn narrowly on the claim that
animals seek to preserve themselves but on the distinctive psycho-
logical explanation the Stoics offer for this mode of behaviour. Here
it is important to distinguish clearly between three elements of the
oikeiōsis accounts: () activities directed towards an animal’s phy-
sical survival, prominently emphasized by Seneca and Hierocles;
() an animal’s primary impulse to preserve and maintain its own
constitution (sustasis), mentioned by Diogenes and confirmed by
other texts; () the phenomenon of self-perception, which the Stoics
take to be a precondition of motivation generally. On the one hand,
it is clear that Hierocles and Seneca present () as evidence of the
psychology associated with () and (), i.e. that the Stoics appeal to
the self-preserving behaviour of animals in order to establish their
claims about self-perception and the primary impulse. This sup-
position is uncontroversial, and it fits the pattern of argument pre-
offered by Hierocles, Seneca, and Cicero (ND . –) with those of Aristotle’s Hi-
storia animalium, Plutarch’s De sollertia animalium, and Philo’s De animalibus. What
is distinctive in Stoic theory is the claim that these behaviours can only be explained
(and in fact are comprehensively explained) by the animal’s perceptual states.
See further Inwood, Seneca, –.
Cf. Inwood, ‘Hierocles’, –; Long, ‘Hierocles’, .
The Stoic Argument from oikeiōsis
served in each of the main sources. Commentators have further as-
sumed, however, that the Stoics narrowly associate the psychology
of () and () with (), i.e. that they treat self-perception and the
primary impulse as the motivational basis of only those patterns of
appropriate behaviour—such as the pursuit of food or flight from
predators—that are narrowly directed towards an animal’s physi-
cal survival. This further assumption explains why commentators
have regarded the primary impulse of Diogenes’ account as an in-
adequate basis for the impulses of the fully rational sage and why
they have treated it, instead, as a form of motivation that must either
be abandoned by fully rational agents or somehow integrated with
concern for and appropriation to others.
I believe this latter assumption—that according to Stoic theory
self-perception and the primary impulse ground narrowly self-
preserving or self-directed behaviours—should be given up. This
is the assumption that motivates the alleged distinction between
personal and social forms of oikeiōsis, but it is not required by the
dialectical structure of the Stoic argument, and it fits poorly with
additional features of the Stoics’ view, including, in particular,
their well-attested interest in the co-operative behaviour of so-
cial animals. It is clear that the Stoics regard the self-preserving
behaviour of animals as evidence of their capacity to perceive
themselves and of their consequent impulse to preserve their na-
tural constitutions. It is not clear, however, that they regard an
animal’s perception and preservation of its constitution as the basis
of narrowly self-preserving behaviour alone. The Stoics might well
appeal to the self-preserving behaviour of animals as evidence for
a particular motivational account without supposing, in addition,
that this account explains only behaviour of that form.
The clearest confirmation of this interpretation is found in the
psychological details that underpin the oikeiōsis theory. These
emerge most clearly in the fragmentary remains of the Hierocles
manuscript, which preserves in greater detail than any other source
the Stoic analysis of phantasia and hormē in non-rational animals
and indicates the broader basis of this analysis in the Stoic (and
largely Chrysippean) doctrine of pneuma. The primary aim of
Hierocles’ treatise is not to show that animals preserve themselves
from birth, but that the co-ordinated behaviour this requires can be
The edited text with Italian commentary appears in Bastianini and Long,
‘Hierocles’.
Jacob Klein
explained only by a mode of cognition that registers the teleological
significance of the animal’s own faculties. The thrust of the treatise
is strikingly similar to that of Seneca’s Letter in that it is direc-
ted against critics who reject Stoic claims about self-perception in
particular. The manuscript begins with a short argument to show
that ‘an animal differs in two respects from what is not an animal:
perception and also impulse’ (col. . –). Hierocles then narrows
his focus to perception in particular (col. . –). In the text that
follows he attacks the critics of self-perception by drawing their
attention, first and foremost, to a range of behavioural phenomena
easily taken for granted. These include the animal’s pursuit of its
own physical survival, but they also extend to more complex forms
of co-operation with other animals. Thus Hierocles emphasizes
that a bull perceives the use of its horns (and Epictetus adds that
it deploys them in defence of the herd). Other animals are said,
strikingly, to perceive their own co-operative arrangement (sum-
basis) with other animals.
The inferences Hierocles draws from this catalogue of observa-
tions are intended to demonstrate the fact of self-perception, but
they are also intended, as his preamble makes clear, to support a
more general explanation of animal motivation in terms of cogni-
tion and, in particular, in terms of cognition whose content is both
evaluative and factive. This account is rooted in the Stoics’ highly
specific analysis of the physical soul in animate organisms, which
Hierocles takes some care to explain at the outset of his treatise (El.
Eth., col. . –). According to this analysis, the organic body of
an animal is itself a compound of pneuma, an admixture of the active
elements of air and fire together with the passive elements of earth
and water with which it is further interblended. This compound is
One class of these critics appears to deny that self-perception occurs at all; the
other denies that it occurs as soon as an animal is born. See Inwood, ‘Hierocles’,
– and –; Long, ‘Representation’, .
‘We do not need to speak of the latter of these for the moment, but it seems
appropriate to say a few things about perception, for this contributes to cognition
of the first appropriate thing [φέρει γὰρ εἰς γνῶσιν τοῦ πρώτου οἰκείου], which is the
subject we said is the best starting-point for the elements of ethics’ (my translation).
El. Eth., col. . –; Epict. Diss. . . –; cf. Marc. Aur. Med. . .
El. Eth., col. . ; cf. Cic. Fin. . = LS F = SVF iii. ; ND . – =
SVF ii. .
The Stoic analysis on which Hierocles draws is well attested in other sources. A
thorough discussion is A. A. Long, ‘Soul and Body in Stoicism’ [‘Soul and Body’],
in id., Stoic Studies, –.
The Stoic Argument from oikeiōsis
responsible for the functions of nutrition and growth, and it is com-
mon to plants and animals alike. In animate organisms, however,
pneuma has a further, higher-order expression the Stoics identify as
soul properly speaking (S.E. M. . = LS F; cf. D.L. . – =
LS O = SVF ii. ). When a plant-like embryo leaves the womb,
a portion of its leading faculty or nature (phusis) is cooled by contact
with air so that its tensional properties begin to sustain the capaci-
ties of phantasia and hormē required for animal motion (Plut. Stoic.
repugn. – ; Hier. El. Eth., col. . –). These capa-
cities, which belong to the animal’s hēgemonikon, are held to be the
defining features of animate organisms, and the Stoics analyse them
as capacities for two distinct types of motion: phantasia being a pas-
sive alteration (kinēsis) that represents its cause, hormē being an ac-
tive movement (kinēsis) of hegemonic pneuma reactively reaching
towards the represented object. Both types of motion are expres-
sions of the divine and active principle working in passive matter to
bring about its providential aims.
Commentators have sometimes emphasized the segmented char-
acter of cognitive development as the Stoics conceive it: the sharp
difference between the cognitive abilities of children and adults or
the starkness of the gap between the conditions of virtue and vice.
For perception and impulse as the defining features of animate organisms see
Hier. El. Eth., col. . –, and Philo, Leg. . – (= LS P = SVF ii. ) and .
(= LS P = SVF ii. ). On the basic κινήσεις of the animal soul see Origen, Princ.
. . – (= LS A = SVF ii. ); Clem. Strom. . . . –. (= SVF ii. );
Cic. Off. . (= LS J). Cf. G. Kerferd, ‘The Origin of Evil in Stoic Thought’
[‘Origin’], Bulletin of the John Rylands University Library of Manchester, (),
– at –; ‘Two Problems concerning Impulses’, in Fortenbaugh (ed.), On
Stoic and Peripatetic Ethics, – at .
As στοιχεῖα, earth and water are also compounds of god and featureless matter.
But as Long puts it, it is in air and fire that god appears in propria persona, as it were
(‘Soul and Body’, ). These active elements of which πνεῦμα is comprised are most
closely analogous to the active organizing features of divine reason at earlier stages
of cosmogenesis and to the original, primal στοιχεῖον characterized by Zeno (per-
haps proleptically, as Cooper suggests) as designing fire (πῦρ τεχνικόν), by Cleanthes
as flame (φλόξ), and by Chrysippus as a flash (αὐγή). See Cooper, ‘Elements’. Cf.
Marc. Aur. Med. . ; Alex. Aphr. Mixt. . – Bruns.
e.g. Pohlenz, Die Stoa, . On the Stoic account of concepts, animal per-
ception is largely if not wholly non-conceptual. For the view that animals have
‘quasi-concepts’ according to the Stoics see Brittain, ‘Perception’. Contemporary
accounts resembling the Stoic view of non-conceptual perception include that of
Gareth Evans: ‘Most organisms have at least a rudimentary capacity to recognize
at least some among other members of their own species, if only their parents
and offspring; and since so much of an organism’s welfare (especially among the
social animals) is dependent upon successful interrelation with other members of its
Jacob Klein
The Stoic account of the animal soul ensures, however, that the
faculties that control and explain the motivations of non-rational
animals are parallel in basic respects to the human case. This iso-
morphism is required, in part, by the fact that the Stoics treat the
division between phantasia and hormē as an exhaustive classifica-
tion of the motions to which the animal soul is subject above the
level of nutrition and growth. Phantasia itself is an exceedingly
broad category in Stoic theory, covering every mode of cognition
to which non-rational and rational animals are subject. In the
rational case, memory (mnēmē), concepts (ennoiai), belief (doxa),
knowledge (katalēpsis), and experience (empeiria) are all defined in
terms of phantasia, and virtue is itself a physical configuration of
the soul built up from assents to propositionally structured impres-
sions (phantasiai). The impressions that belong to animals, on the
other hand, are not fully conceptual or propositional in the Stoics’
sense, and it is unclear to what extent animals may entertain a gen-
eralized piece of content independently of overtly perceptual states,
on the Stoic account. In animals too, however, phantasia involves
an object of which the animal is made aware through its causal im-
pact within the soul—an imprint, in Zeno’s terminology—so that in
veridical cases of perception the phenomenal character of the repre-
sentation is causally linked to the perceived object. The basic ele-
species, we should expect to find an informational system of the kind I have outlined
developing out of this primitive capacity as early as any other’ (The Varieties of
Reference [Reference], ed. J. McDowell (Oxford, ), ). Evans treats this
capacity for recognition in animals as a non-conceptual one (–).
These parallels are especially emphasized by V. Goldschmidt, Le Système stoï-
cien et l’idée de temps [Système stoïcien], th edn. (Paris, ), –. Cf. n. below.
According to the Stoics, φαντασία is the most basic representational faculty, and
every other kind of mental representation is explicated in its terms. In this respect
the Stoic taxonomy of mental states differs importantly from Aristotle’s. For both
Aristotle and the Stoics, desire of the most basic sort (ὄρεξις for Aristotle, ὁρμή for
the Stoics) requires αἴσθησις. On the Stoic account, but not on Aristotle’s, αἴσθησις
also requires φαντασία. Hence on the Stoic account, but not on Aristotle’s, desire of
any form requires φαντασία. Aristotle attributes ὄρεξις and αἴσθησις to some animals
which lack φαντασία (DA b, a–; cf. H. Lorenz, The Brute Within: Appe-
titive Desire in Plato and Aristotle (Oxford, ), ch. ). On the Stoic account of
φαντασία see Long, ‘Representation’, –.
See the discussions in Brittain, ‘Perception’, and id., ‘Common Sense: Con-
cepts, Definition and Meaning in and out of the Stoa’ [‘Common Sense’], in D.
Frede and B. Inwood (eds.), Language and Learning: Philosophy of Language in the
Hellenistic Age (Cambridge, ), –.
According to the basic Stoic definition, a φαντασία is ‘a pathos occurring in the
soul, which reveals itself and its cause’ (Aët. Plac. . . – = LS B = SVF ii. ,
The Stoic Argument from oikeiōsis
ments of this analysis cut across the differences between the highly
conceptualized modes of cognition that belong to rational agents
and the non-rational perceptions of other animals.
One of Hierocles’ central concerns, then, is to show that the be-
haviours he catalogues are rooted in the soul’s capacity to represent
the world through physical interaction with it. From the point of
view of action theory, what is most significant in this account is
not the thesis that motivation always occurs in conjunction with an
appropriately discriminating mental representation. It is rather the
claim that cognition of this form determines impulse, that an or-
ganism’s activities are caused and controlled by mental states that
have, as we would now say, the cognitive direction of fit, whose
function is to grasp and conform to the contours of the world. In
animals and human agents alike, the outward motions of impulse
are engendered and wholly explained by representations arising
in the perceptive pneuma of the animal’s hēgemonikon. In the hu-
man case, such representations may be propositional, susceptible
of verbal articulation, and conditioned by a range of antecedent
cognitive conditions that also have a propositional and conceptual
content. Though an animal’s impressions of this form are not of
comparable structure and complexity, they are nonetheless inchoate
analogues of the impressions that belong to human agents. Even in
animals, such impressions appear to require the possibility of con-
tent that minimally includes an evaluative component, representing
the world in a manner capable of producing and structuring impulse
in ways commensurate with the animal’s faculties and basic needs.
Without exception phantasia is the causal basis of impulse, and its
content explains the force and direction of an animal’s activities.
As Inwood has emphasized, these distinctive psychological
claims are notably absent from parallel accounts of animal psycho-
logy found in Antiochean sources. Though perception is said in
De finibus and to guide and shape an animal’s impulses as it
trans. Long and Sedley). Zeno seems to have characterized an instance of φαντασία
as a τύπωσις (imprint), while Chrysippus preferred the less committal ἑτεροίωσις (al-
teration). See S.E. M. . (= SVF i. ). Representations lacking the appropriate
causal connection to the world are φανταστικά or φαντάσματα, in the Stoic scheme.
See D.L. . (= LS A = SVF ii. ); Aët. . . – (= LS B = SVF ii. ).
Such impressions, that is to say, are informed by ἔννοιαι (conceptions) and have
λεκτά (verbalizable representations) as their content. I do not mean to imply a direct
equivalence between these Stoic categories and contemporary ones.
Inwood, ‘Hierocles’, –.
Jacob Klein
matures, it is not there presented as a sine qua non of motivation
generally, nor is it regarded as a faculty responsible for generating
impulse as soon as the animal is born. By contrast, Stoic versions
of oikeiōsis make it clear that accurate cognition is the sole basis of
teleologically appropriate behaviour in animals and humans alike.
Hierocles’ account of the psychology that underpins animal be-
haviour corresponds, therefore, to an important and controversial
Stoic claim about the structure of rational motivation: namely, that
there are no sources of impulse that are not governed by cognitive
appraisals of some form. Though it supports more complex dis-
cursive and inferential functions than those that occur in animals,
the rational faculty that controls right action in human agents does
so through an accurate representational grip on the world. On the
Stoic account, appropriate action in animate organisms generally is
preceded and structured by a cognitive grasp of the natural order.
. Self-perception
mental state a ὁρμή is supposed to be: for example, whether ὁρμή is the causal re-
sult of a purely representational state or whether it is a single state that performs the
functions of both representation and motivation, thereby resembling the ‘besires’
of some contemporary theories. I am here suggesting the latter sort of account and
thinking, in particular, of Margaret Graver’s characterization of the evaluative con-
tent of a hormetic impression as registering its ‘motivational aspect’. See Graver,
Stoicism and Emotion, .
Galen, PHP . . – (= LS T = SVF iii. ); . . – (= SVF iii. ).
Cf. Cic. Tusc. . . The possibility of a mismatch extends to ὁρμαί in that they are
a product of evaluative representations and may be correct or incorrect as such. On
the relationship between slackness (ἀτονία) of soul and failures of perception and im-
pulse see esp. Kerferd, ‘Origin’, –. As Kerferd puts it, ‘When there is a good
state of tension in the soul—eutonia—a man judges rightly and does well’ (). So
too ‘wrong judgements . . . are themselves kinēseis and hormai and are all cases of the
hēgemonikon disposed in a particular way’ (). On the physical basis of Stoic virtue
see further M. Schofield, ‘Cardinal Virtues: A Contested Socratic Inheritance’, in
A. G. Long (ed.), Plato and the Stoics (Cambridge, ), – at –.
According to Stobaeus, ‘Proper function [τὸ καθῆκον] also extends to the non-
rational animals, for these too display a kind of activity which is consequential upon
their own nature’ (. = LS B = SVF iii. , trans. Long and Sedley). Accord-
ing to Diogenes, ‘Proper function [τὸ καθῆκον] is an activity [ἐνέργημα] appropriate
to constitutions that accord with nature [ταῖς κατὰ φύσιν κατασκευαῖς οἰκεῖον]’ (D.L.
. = LS C = SVF iii. , trans. Long and Sedley).
The Stoic Argument from oikeiōsis
pneuma, and it is through these faculties that nature regulates (oiko-
nomein) their activities. The Stoics may well have thought of the
slackening of hegemonic tension in animals as a kind of falling away
or loosening of pneuma’s controlling grip on the organism, engen-
dering an imprecision in its perceptions and impulses.
One way of understanding the impulse to self-preservation, then,
is as a creature’s inborn tendency, implanted by nature, to sustain
the tensional properties of its hēgemonikon in a way that guaran-
tees the accuracy of its representations and the appropriateness of
its impulses. So understood, self-preservation involves preserving
the very properties that determine the animal’s motions, ensur-
ing the correct relationship between the animal’s psychic pneuma
and the organic matter it permeates and controls. Such a view might
be spelt out in different ways consistently with the physical model
implied by the sources. What is crucial to the oikeiōsis account,
however, is that the primary impulse to self-preservation, as the
Stoics conceive it, appears to be satisfied when each of an animal’s
first-order impulses and activities is commensurate with the kind
of creature it is. In perceiving and preserving its hēgemonikon, an
That is to say, impression and impulse cannot occur without corresponding
changes in the physical configuration of the ἡγεμονικόν itself. This point follows trivi-
ally from the Stoic claim that impressions and impulses are alterations of hegemonic
πνεῦμα (e.g. S.E. M. . –). It should be distinguished from both () the claim
that impressions and impulses supervene on properties or states of the sense organs
with which the ἡγεμονικόν is interblended and () efforts to give a general charac-
terization of the relation between physical states of the ἡγεμονικόν and phenomenal
or representational mental content. Since we have determinate evidence for just one
type of content associated with φαντασίαι—the incorporeal λεκτά that are said to sub-
sist (ὑφίστασθαι) in accordance with rational impressions (D.L. . = LS F; . =
LS D = SVF ii. )—it is not easy to give a general account of the latter relation
as the Stoics conceive it.
It might be, for instance, that the Stoics think of the primary impulse as
an orientation whose satisfaction consists in the appropriate functioning of the
creature, so that accurate perception and appropriate impulse are constitutive
of self-preservation as the Stoics understand it. It is worth remembering in this
connection that Chrysippus identifies action (whether he means capacity or activity
is unclear) with the ἡγεμονικόν itself. It therefore seems possible that he included
καθήκοντα—those actualizations or functions (ἐνεργήματα) of these capacities that
are οἰκεῖον to an organism’s constitution—within the scope of the primary impulse.
See Sen. Ep. . (= LS L = SVF ii. ), on which see B. Inwood, ‘Walking
and Talking: Reflections on Divisions of the Soul in Stoicism’, in K. Corcilius and
D. Perler (eds.), Partitioning the Soul: Debates from Plato to Leibniz (Berlin, ),
– at –. Alternatively, the Stoics might suppose that an appropriate tension
in the soul’s commanding part is causally sufficient to ensure the correctness of an
animal’s perceptions and impulses, so that satisfying the primary impulse is distinct
from but nonetheless secures appropriate function.
Jacob Klein
animal is at the same time guarding the integrity of the represen-
tations and impulses that originate there, bringing them into line
with the constitution given to it by nature. As long as the pneuma
that constitutes it as a living organism is maintained in its proper
condition, each of an animal’s activities will conform, like the con-
tent of its representations, to the patterns of nature itself.
This analysis is reflected in each of the oikeiōsis accounts that can
be reliably associated with the Stoics independently of Antiochean
sources, and it makes a crucial difference to interpreting the Stoic
theory. In particular, it explains why the Stoics themselves did not
distinguish narrowly between personal and social forms of oikeiōsis
or between an animal’s self- and other-regarding activities. That
misleading distinction rests on the supposition, due to Pohlenz,
that the primary impulse is a first-order motivation narrowly as-
sociated with self-preserving activity. By contrast, on the account
I have offered an animal’s impulse to preserve itself is not a first-
order motivation in competition with oikeiōsis towards other things
(such as offspring or other animals) but is rather the psychologi-
cal basis of its recognition and appropriation of them as oikeion. An
animal’s self-perception and consequent appropriation to self ex-
plain its self-preserving behaviour, certainly, but they equally ex-
plain other-regarding behaviours such as care for offspring, defence
of the herd, and co-operation with other species. For an animal to
perceive and preserve its constitution is for it to confirm and sus-
tain itself as the kind of thing nature constituted it to be, with all
the attendant activities this implies.
e.g., Epict. Diss. . . ; . . ; . . –; . . ; . . ; . . –; Marc.
Aur. Med. . ; . ; . ; . ; . ; . ; . ; . ; . ; Cf. D.L. .
(= LS C = SVF iii. ).
At Diss. . . – Epictetus draws an explicit parallel between an animal’s συν-
αίσθησις of its own capacities (παρασκευαί) and the acquisition of moral knowledge
in the human case. Cf. also Diss. . . ; . . ; . . –. According to Marcus
Aurelius (Med. . ), one of the properties of the rational soul is that it sees itself
(ἑαυτὴν ὁρᾷ). He goes on to say that one becomes good through conceptions (θεω-
ρήματα) both of universal nature and of the distinctively human constitution (Med.
. ).
Cf. Schmitz, Cato, –. As a corporeal thing, the ἡγεμονικόν is an object of
αἴσθησις, and the Stoics do not hesitate to apply a range of normative and aesthetic
predicates to the physical structure of πνεῦμα that constitutes virtue in human agents
(Stob. . – = SVF iii. ; Cic. Tusc. . . = SVF iii. ). Appropriate actions
are those that ‘contain all the numbers of virtue’ (‘omnes numeros virtutis conti-
nent’) in so far as they flow from a ἡγεμονικόν whose quantifiable physical properties
preserve a harmony with the cosmos (Cic. Fin. . = LS H = SVF iii. ). Cf.
D.L. . (= SVF iii. ) and A. A. Long, ‘The Harmonics of Stoic Virtue’, in id.,
Stoic Studies, – at .
R. Brouwer, The Stoic Sage: The Early Stoics on Wisdom, Sagehood, and Soc-
rates (Cambridge, ), .
Jacob Klein
of pneuma that represents the world and controls assent to im-
pressions in rational agents. The same use of tērein occurs in
Marcus Aurelius, who characterizes the hēgemonikon as a daimōn,
which is to say, pneuma at the highest level of expression and the
embodiment of divine reason in each human agent. Both authors
appear to identify the preservation of this faculty in the Stoics’
technical sense with virtue and right action. Nor does this appear
to be an innovation of later Stoicism. To preserve one’s daimōn is
identical, as Diogenes makes clear, to living according to human
and cosmic nature, engaging ‘in no activity wont to be forbidden by
the universal law, which is the right reason pervading everything’
(D.L. . = LS C = SVF iii. , trans. Long and Sedley). The
Stoics identify eudaimonia quite literally with the perfection of
one’s daimōn.
It appears, moreover, that the oikeiōsis theory defended by Chry-
sippus was developed within a wider teleological analysis of orga-
nisms generally. Though the Stoics do not extend the categories of
goodness or eudaimonia to animals, they do speak of the ends (telē)
appropriate to animals and of the conditions under which these ends
are realized. Cicero’s criticisms of the Stoic oikeiōsis theory mention
a treatise in which Chrysippus offered a general survey of animal
species, then proceeded to discuss the end appropriate to each (Fin.
. = SVF iii. ). Epictetus similarly maintains that ‘of beings
whose constitutions [kataskeuai] are different, the works [erga] and
ends [telē] are likewise different’ (Diss. . . –). A. A. Long em-
phasizes the following neglected passage from Hierocles’ treatise on
marriage:
[E]very [non-human] animal lives consistently with its own natural consti-
tution [τῇ ἑαυτοῦ φυσικῇ κατασκευῇ]—and every plant indeed too according
Diss. . . ; . . ; . . ; . . –; . . ; . . ; . . ; . . ;
. . ; Ench. .
See esp. Marc. Aur. Med. . ; cf. . ; . ; . ; . ; . .
See esp. Epict. Diss. . . –; . . –; . . ; . . ; Ench. ; Marc. Aur.
Med. . ; . . A kind of contrapositive principle appears in the Discourses, where
inappropriate action is said to destroy the characteristic features of a human being:
e.g., Diss. . . –; . . –; . . –; . . –; . . . I am grateful to
Tad Brennan for pointing out several of these references.
G. Betegh correctly argues, contra Rist, that δαίμων here is an alternative de-
scription of the agent’s ἡγεμονικόν. See Rist, Stoic Philosophy, ff., and Betegh,
‘Cosmological Ethics in the Timaeus and Early Stoicism’, Oxford Studies in Ancient
Philosophy, (), – at ff. Cf. Brouwer, The Stoic Sage, n. .
D.L. . (= LS A = SVF iii. ). Cf. Porph. Abst. . . – (= LS P).
The Stoic Argument from oikeiōsis
to the plants’ so-called life—except that they do not make use of any calcu-
lation or counting or acts of selection that depend on testing things: plants
live on the basis of bare nature, and [non-human] animals on the basis both
of representations that draw them towards things appropriate and of urges
that drive them away. (Stob. . )
It therefore seems clear that the older Stoics gave some account of
the ends of animals and that they supposed, in addition, that an
organism’s constitution determines the character of its end. An or-
ganism’s end answers to its natural impulse to preserve its constitu-
tion, and the content of its end depends on the distinctive features
of the constitution it aims to preserve.
It is important, however, to distinguish two ways of understand-
ing the teleological framework behind the oikeiōsis account. If the
Stoics regard the integrity of the hēgemonikon as a condition that
guarantees appropriate action in non-rational animals, as I have ar-
gued, a crucial question is whether they also regard it as sufficient
to ensure that a non-rational animal achieves its end. Clarity on this
point is important for making sense of Stoic appeals to oikeiōsis as a
basis for ethical theorizing since, as I have noted, Cicero and Alex-
ander characterize the prōton oikeion as something that is correlative
to the account of the end adopted by each school, so that the object
of motivational attachment observable in animals and infants cor-
responds to the end to which a rational agent ought to be attached.
Quoted and translated in A. A. Long, ‘The Logical Basis of Stoic Ethics’, in
id., Stoic Studies, – at –. As Long observes, the passage closely resembles
Diogenes’ brief account of the oikeiōsis doctrine, which refers to Chrysippus’ treatise
On Ends.
The reliability of Alexander’s report, influenced as it is by Carneades and by
Alexander’s own Peripatetic aims, is open to question. Yet Cicero’s consistent agree-
ment on this point (Fin. . ; . –; . ; . ; . –; . = LS G; . –;
. ) and the likelihood that the early Stoic theory of οἰκείωσις was developed as a
counterpoint to the Epicurean cradle argument tend to confirm Alexander’s claim.
Moreover, the basically orthodox Hierocles also says that an account of the πρῶ-
τον οἰκεῖον is an appropriate starting-point for a treatise on ethics (El. Eth., col. . ,
–). This appears to mean not that a treatise on ethics should begin with some
remarks on developmental psychology but that an understanding of the earliest, na-
tural object of motivation should constrain an ethical account in substantive ways,
informing our conception of the human good. If the Stoics do not regard the object
of neonatal motivation as correlative to the rational case, it is hard to understand the
extent to which they were prepared to argue for a highly specific account of animal
psychology in ethical contexts. For the suggestion that the Stoic theory is a response
to Epicurean cradle arguments see Pohlenz, Grundfragen, –; Die Stoa, –.
On the Carneadean arguments in Alexander’s Mantissa see G. Striker, ‘Antipater,
or the Art of Living’, in ead., Essays, – at –; R. W. Sharples (trans.),
Jacob Klein
If animals and infants are characterized from birth by an impulse
to preserve their governing faculty, and if the satisfaction of this
impulse is a sufficient condition of their teleological success, this
analysis would appear to support the Stoics’ claim, in the realm of
ethics, that virtue is sufficient for happiness.
A standard way of understanding Stoic teleology, however, rather
assumes that the Stoics regard external objectives—the prima secun-
dum naturam/prōta kata phusin that figure prominently in sources
associated with Carneades—as essential to the end of non-rational
animals, so that the human end differs, in this fundamental respect,
from that of lower animals. This picture is strongly implied by
Cicero’s account of oikeiōsis in De finibus , which represents human
motivation as shifting in the course of development from a concern
with external objectives to a singular concern with right action:
A human being’s earliest concern is for what is in accordance with nature.
But as soon as one has gained some understanding, or rather ‘conception’
(what the Stoics call ennoia), and sees an order and as it were concordance in
the things which one ought to do, one then values [aestimavit] that concor-
dance much more highly than those first objects of affection. (Fin. . =
LS D = SVF iii. , trans. Woolf)
How and where did you suddenly abandon the body and all those things
that are in accordance with nature but not in our power, finally discarding
appropriate action itself? How is it that so many of the things originally
commended by nature are suddenly forsaken by wisdom? Even if we were
seeking the supreme good not of a human being but of some living creature
who had nothing but a mind, . . . this mind would not accept the end you
are proposing. It would want health and freedom from pain, and would
also desire its own preservation as well as the security of those goods I just
mentioned. It would establish as its end a life in accordance with nature,
and this means, as I said, possession [habere] of things that are in accordance
with nature, either all of them or as many as possible of the most important.
(Fin. . – = LS K, trans. Woolf, emphasis added)
come to see my reason as the thing that is my own or oikeion to me, the thing whose
welfare is my primary concern, such that if my reason is in good shape then I am in
good shape. My final end, my summum bonum, is determined by the kind of thing I
am’ (The Stoic Life, ; cf. ‘Souls’, –). According to Radice, Cicero’s account
suggests that the development of reason in human beings ‘modifica profondamente
il fine’ (Oikeiosis, ). The Stoics suppose that ‘la nascita della ragione crea una
frattura netta all’interno della natura umana sicché l’autoconservazione può essere
solo della ragione’ ().
D.L. . (= LS A = SVF iii. ). Cf. Cic. Tusc. . ; Marc. Aur. Med.
. .
The Stoic Argument from oikeiōsis
pneuma that structures the physical cosmos. Though animals go
looking for what is appropriate and avoid what is not, nothing in
the passage suggests that Chrysippus held that an animal’s telos de-
pends on anything more than the integrity of its hēgemonikon and
the appropriate modes of action this secures. Though the hēge-
monikon is the faculty by which an animal governs its own motions,
it is also the faculty by which cosmic reason governs it. The result-
ing picture is that of pneuma working in and through each kind of
thing, establishing it as the kind of thing it is and bringing about
the activities appropriate to it, moving it towards what is oikeion
and away from what is allotrion. Such an account fits suggestively
with Sextus’ remark that the Stoics believe it can be shown, on the
basis of animal behaviour, that only the fine is good (M. . –
= SVF iii. ).
We might suppose, then, that in older Stoic theory something
analogous to the sufficiency of virtue for human beings applies at
every level of the scala naturae, so that the end of each kind of
organism is achieved through the preservation of its hēgemonikon.
There are, to be sure, significant differences between non-rational
animals and rational human beings in Stoic theory: the sophistica-
tion of their concepts, the capacity for language, and the mecha-
nism of assent being the most salient examples. These differences
determine, for the Stoics, a set of evaluative categories that apply to
rational agents alone, including especially those of happiness, good-
ness, and justice. A central reason why they do so, however, is that
the conceptual sophistication of rational cognition enables human
beings to mirror in the structure of their own cognitions—as ani-
mals cannot—the order and beauty instantiated in the cosmos itself.
The capacity for assent that accompanies this conceptual sophisti-
cation moreover allows human beings to be artificers of this order,
together with Zeus, by apprehending and assenting to it. These
As Inwood observes, there is no evidence ‘that the early Stoics said that an ani-
mal is ever oriented to anything but its constitution’ (Ethics and Human Action, ).
Stoic sources sometimes speak of external objectives as οἰκεῖα, but this is presumably
because an animal’s awareness of and dominant impulse to preserve its σύστασις—
the πρῶτον οἰκεῖον or primary object of οἰκείωσις—determines a range of first-order
affiliations and impulses towards what is appropriate (οἰκεῖον) in a derivative sense
(Inwood, Ethics and Human Action, –). Cf. nn. and below.
On this point see esp. Epict. Diss. . . – and . . According to Epictetus,
the fundamental difference between humans and non-rational animals is not the ca-
pacity to use impressions correctly but the capacity to understand this use. Epictetus
says, very strikingly, that if animals understood their own use of impressions the cat-
Jacob Klein
differences all depend on complexities internal to the human hēge-
monikon, however, and there is little reason to suppose that they are
intended to account for basic differences in the criteria of teleolo-
gical success that apply across species. It is consistent with our
sources to suppose that what is transformed by the development of
reason, according to Stoic theory, is not the actual value of external
circumstances nor the actual conditions of teleological success. It
is rather the capacity to appreciate and assent to patterns of action
already present in the rationally organized cosmos.
This way of understanding Stoic teleology assumes that although
the range of activities appropriate to an animal (kathēkonta) will vary
with its constitution, the preservation of its constitution remains, for
each kind of organism, sufficient to achieve its end. On this inter-
pretation, when Epictetus asserts that the works and ends for each
animal are relative to its constitution (kataskeuē), he does not mean
that the basic conditions of teleological success vary widely across
species. He means only that the activities in which an organism’s
end consists are determined by the set of functional capacities it
possesses by nature: on whether, for instance, it was designed by
nature for co-operation with other members of its own kind, as
human agents manifestly are (Cic. Fin. . – = LS F; Off. . ;
Leg. . ; cf. Xen. Mem. . . ). Under this construal, the telos of
a non-rational animal no more depends on external conditions than
does the human telos. It rather consists in the integrity of the ani-
mal’s hēgemonikon and in the performance of appropriate functions
(kathēkonta), which the animal’s capacity for self-perception enables
it to grasp. We might call this the higher-order interpretation of
Stoic teleology: it implies that although the range of activities appro-
priate to an organism may vary from case to case, the telos depends,
for each kind of organism, on the satisfaction of its higher-order
impulse to preserve the constitution given to it by nature.
egories of happiness and goodness would apply to them as well. The donkey would
then be our ‘equal and our peer’ (Diss. . . , trans. Oldfather). On this passage and
on Epictetus’ use of animals as moral exemplars see W. O. Stephens, ‘Epictetus on
Beastly Vices and Animal Virtues’, in D. R. Gordon and D. B. Suits (eds.), Epic-
tetus: His Continuing Influence and Contemporary Relevance (Rochester, NY, ),
– at –.
Pembroke writes that there is ‘a definite emphasis [in Stoic texts] on the con-
tinuity of childhood with maturity, just as there is elsewhere on that of animal
behaviour with human life . . . [T]he line drawn is evaluative rather than descrip-
tive, and ultimately, the continuity is more important than the point of demarcation’
(‘Oikeiōsis’, ).
The Stoic Argument from oikeiōsis
Each of these interpretations of Stoic teleology has a significant
consequence for our understanding of Stoic theory and of the
oikeiōsis account in particular. The first-order construal of Stoic
teleology may seem to spoil the inductive character of Stoic appeals
to animal behaviour, as Cicero points out when he speaks for Anti-
ochus. If Cicero’s report in De finibus is correct, the older Stoics
must have acknowledged that rational human maturity is marked
by an abandonment of objectives that, by the Stoics’ own admis-
sion, matter to the telos of non-rational creatures (and perhaps that
of pre-rational humans). When a human being acquires her rational
nature, she acquires an end that excludes as indifferent resources
and conditions essential to the telos of all other animals (Fin. . ).
If the Stoics were prepared to grant the strongly discontinuous
account of animal and human ends implied by De finibus , their
appeal to animal behaviour seems wrong-headed from the start.
On this account of Stoic teleology, it is difficult to understand why
Stoic arguments for the human good place so much emphasis on
the animal case and why Hierocles and Seneca take such pains to
refute rival analyses of animal behaviour.
The higher-order construal of Stoic teleology has a consequence
of another sort: it requires us to suppose that, at least when he is
speaking for Antiochus, Cicero is sometimes prepared to distort
or conceal older Stoic views in substantive ways. If the Chrysip-
pean theory of oikeiōsis identified the aim of an animal’s basic moti-
vational impulse with the regulation of appropriate activity rather
than the satisfaction of its physical needs, then Cicero’s report in De
finibus significantly mischaracterizes the psychological claims that
are the starting-point of the oikeiōsis theory, replacing the Stoics’
own analysis of non-rational motivation with an account that points
to Antiochean conclusions about the human telos. Since Cicero of-
fers us one of our few windows into Stoic theory, this means we
must treat a substantial portion of our evidence as a misleading
guide to this aspect of older Stoicism. This is a disheartening result.
On balance, however, I believe we should accept the higher-order
framework suggested by Diogenes’ account, for three reasons.
First, there are indications, internal to De finibus, that the asym-
metry between animal and human ends insinuated by Antiochus
is a vestige of dialectical exchanges between the Stoics and the
Academy rather than a feature of older Stoic theory. As Inwood
Cicero understood the doctrinal core of Stoic theory very well, but he is cap-
Jacob Klein
has emphasized, the sharp division between soul and body fore-
grounded throughout book of De finibus suggests that Cicero is
operating with a broadly Academic account of the soul–body rela-
tion in view and that he may simply be ignoring, at this stage in his
dialectical argument, the different physical analysis that underpins
the Stoics’ own account. The psychological starting-points of
the Stoic theory as Cicero represents them in De finibus are not
especially clear, but they are clear enough to register an important
point of difference between the Stoic theory and the Antiochean
versions of De finibus and . In De finibus Cato says clearly
that an animal’s primary attachment is to its own constitution (‘ad
suum statum’) and that it is in virtue of perceiving and caring for
its constitution that an animal possesses impulses towards what is
external. By contrast, though Piso also speaks of the motive of
their own sake. This material sounds very Stoic indeed, and it fits well with Sextus’
observation that the Stoics locate the goods of the soul in the ἡγεμονικόν itself (M.
. –). It fits poorly, however, with the emphasis on primary natural things in
which it is embedded. We may suppose that, for the Stoics, καταλήψεις are indeed
objects of the primary οἰκείωσις relation in fully rational human agents, since they
are physical states of the ἡγεμονικόν (cf. Cic. Acad. . = LS N) and components
of virtue. Externals, though pursued by animals and human agent alike, are pursued
derivatively, as constituents of the καθήκοντα that flow from the ἡγεμονικόν (cf. n.
above). More generally, the cosmic framework so evident in Diogenes Laertius has
all but disappeared from Cicero’s version of the Stoic theory (cf. Inwood, Ethics after
Aristotle, –; Radice, Oikeiosis, ). For disconcerting parallels between Cicero’s
De finibus account and Peripatetic versions of οἰκείωσις see Schmitz, Cato, –.
‘[T]he object of every living creature’s desire . . . is to be found in what is ad-
apted to its nature’ (Fin. . , trans. Woolf).
Structured as it is by the Carneadia divisio, Piso’s account of the Stoic position
does not even acknowledge an organism’s constitution as a possible object of initial
impulse. It is notably and deliberately omitted from the possible starting-points ack-
nowledged by the divisio, and hence from the list of plausible accounts of the good.
At Fin. . pleasure, the absence of pain, and the primary natural things are given
as the only possible objects of an animal’s primary or initial desire. The account pre-
served at D.L. . (= LS A = SVF iii. ) is not even recognized in this scheme.
Cf. Cic. Tusc. . –; Fin. . , ; . (= LS G).
Jacob Klein
Posidonius attributing to Chrysippus the view that we have by
nature an ‘oikieōsis relationship to the fine alone’ (fr. = PHP
. . = LS M). This claim, if Chrysippus made it, is filtered
through Posidonius’ Platonizing account of the human soul and
Galen’s own polemical commentary, but it fits well with the sup-
position that the primary object of appropriation, for animals and
humans alike, is the hēgemonikon itself. Sources closely associated
with Chrysippus regularly align the predicate oikeion with the
condition of virtue and with the activities that flow from a con-
stitution that conforms to nature. By contrast, the focus of the
Carneadean account is not on appropriate activities but on the ex-
ternal objectives or fruits of nature that answer to the Carneadean
conception of the end. The interpretation I have proposed makes
sense of this evidence. If the Stoics identify the prōton oikeion with
the faculty that controls perception and impulse, and if the aim of
the primary impulse is to preserve this faculty as nature intends,
the primary oikēiosis relationship will indeed be directed, in the
case of the sage, at what is fine, at the perfected rational hēgemonikon
and the activities to which it gives rise. This means, in turn, that
the primary object of oikeiōsis will be correlative to the Stoic con-
ception of the telos, according to which happiness consists in virtue
and virtuous activity. On this account, the Stoic argument fits the
broader pattern of Hellenistic appeals to neonatal motivation.
The third and most compelling reason, however, is that the
higher-order construal of Stoic teleology supplies the Stoics with
an argument for the human good that is intelligible against the
background of the cosmological theory developed by Chrysip-
pus. In general Stoic teleology appears to be carefully calibrated
to accommodate the Stoics’ cosmobiology and the interlocking
relation of part to whole. Nature as a whole realizes its purpose
even when the parts do not. But the parts too realize their ends
when there is no impediment to the governing faculty through
which nature works (Cic. ND . ; Tusc. . –; Marc. Aur.
Compare, for instance, the use of οἰκεῖον Plutarch ascribes to Chrysippus at
Stoic. repugn. (= LS E = SVF iii. ) and (= SVF iii. ) with D.L.
. (= SVF i. , iii. ) and Marcus Aurelius’ use of οἰκεῖον at Med. . . In
these passages virtue and its activities—not external objects—are said to be οἰκεῖον.
Cf. also Stob. . (= SVF iii. ); S.E. M. . (= SVF i. ); Epict. Diss. . .
–; and especially Galen, PHP . . – (Posid. fr. = LS M).
On the wide influence of the Carneadean critique see Inwood, Ethics after Aris-
totle, ch. .
The Stoic Argument from oikeiōsis
Med. . ; cf. Nemesius, Nat. hom. . = SVF ii. ). The
account I have proposed attributes to the Stoics the view that the
same conditions of teleological success obtain at every level of the
scala naturae, and it therefore explains the dialectical point of Stoic
appeals to animal self-perception and self-appropriation. Every
organism goes through the motions of fate, but it may do so either
as a segment of pneuma whose representations and impulses are
preserved in a condition of conformity to nature or as one in which
they are not. It appears that Chrysippus is largely responsible for
the role assigned to pneuma as a basic postulate of this account.
The evidence suggests that he took pains to a remarkable degree to
integrate Stoic cosmogony and cosmology with new developments
in the vitalist medical theories of the third century. In Chry-
sippus’ hands, the theory of pneuma provided a unifying basis for
Stoic cosmology and biology, allowing the Stoics to characterize
the motions (kinēseis) of individual organisms both as movements
of distinct souls and as movements of the breath that suffuses and
sustains the cosmos as a whole (e.g. Epict. Diss. . . –).
Though already present in Zeno’s theory, the doctrine of πνεῦμα was reworked
and elaborated by Chrysippus, who appears to have been influenced by the increas-
ingly prominent role assigned to πνεῦμα by the vitalistic medical theories of the
rd cent. and by a need to revise or update Cleanthes’ appeals to heat as an ani-
mating principle. See D. Hahm, The Origins of Stoic Cosmology (Columbus, Ohio,
), –; M. Lapidge, ‘Stoic Cosmology’, in J. Rist (ed.), The Stoics (Berke-
ley, ), – at –; J. M. Rist, ‘On Greek Biology, Greek Cosmology and
Some Sources of Theological pneuma’, in id., Man, Soul and Body: Essays in An-
cient Thought from Plato to Dionysius (Aldershot, ), – at –; D. J. Furley,
‘Cosmology’, in Algra et al. (eds.), The Cambridge History of Hellenistic Philosophy,
– at ; S. Sauvé-Meyer, ‘Chain of Causes: What is Stoic Fate?’, in Salles
(ed.), God and Cosmos in Stoicism, – at –.
Jacob Klein
about the human good. Since accurate perception and appropriate
impulse do not depend on circumstances, the Stoic account implies
that it is possible for each creature to act appropriately regardless
of external conditions, as Seneca’s turtle does when it struggles
to right itself (Ep. . = LS B). This element of the Stoics’
account supports the claim that the human good consists in virtue
alone. At the same time, the Stoics argue that animal behaviour
cannot be explained by the external stimulus of pleasure. Animals
perform the functions appropriate to their make-up even when it is
painful for them to do so, and this behaviour can be explained only
by the mechanisms of self-perception and self-appropriation (Sen.
Ep. . – = LS B; S.E. M. . –). This element of the
Stoics’ account supports the claim that virtue consists in a form of
cognition, a grasp of oneself and of the cosmos as a whole.
Such an interpretation fits closely with the fact that the category
of appropriate (kathēkon) activity applies to organisms quite gener-
ally and with the metaphysical claim that an organism’s hēgemonikon
is the most immediate expression of divine activity in the physi-
cal cosmos, directing its activities in a manner suited to its kind.
An animal of course does not recognize the conditions under which
its own end is achieved, as rational humans may. Yet the patterns
apparent in animal activity evidently provide a model, within the
Stoic scheme, of the psychological coherence and consistency that
is supposed to govern human cognition and action. It is hard to
credit Cicero’s implication, advanced in a polemical context, that
the Stoics treat the conditions of teleological success for human
agents as sharply discontinuous with those that apply elsewhere
in the scala naturae. In particular, it is difficult to believe that the
Stoics treat external conditions indifferent to human happiness as
essential to the telos of non-rational animals, regarding an animal’s
failures to realize these external conditions as teleological failures.
If this reconstruction is correct, it has two consequences for
Contrast Galen, PHP . . –.
Seneca especially compares the form of cognition implanted in animals by
nature to the technical knowledge that constitutes virtue in the human case (Ep.
. –), on which see Inwood, Seneca, –; Origen, Princ. . . – (LS A =
SVF ii. ).
This is not to suggest that non-rational animals conceive of external things as
indifferent or pursue them under this description, on the Stoic account, merely that
external things make no more difference to the τέλος of animals than to the human
τέλος, in the Stoics’ view. The Stoics do emphasize, however, that animals and in-
fants perform the functions appropriate to them with a certain disregard for painful
The Stoic Argument from oikeiōsis
understanding the Stoic doctrine of oikeiōsis. First, it helps to ex-
plain why the Stoics do not appear to offer a derivation of the sort
commentators have sought of other-regarding obligations from an
animal’s appropriation to self. Though self-appropriation gives
rise to behaviours associated with self-preservation in the ordinary
sense, it also explains a broader range of teleologically appropri-
ate activities. These include, most saliently, an animal’s care and
concern for its own offspring and, in some cases, its co-operation
with other animals both within and across species. In the human
case, the leading feature of the rational constitution is its suitability
for social co-operation, evidenced most clearly in the capacity for
language. According to Cicero and Marcus, we are constituted
(constitutum esse/kateskeuasthai) for justice and fellowship, and re-
cognition of this fact grounds our awareness of obligations towards
others (Cic. Leg. . = SVF iii. ; Marc. Aur. Med. . ; . ;
. ; . ). Though the impulse to preserve one’s own con-
stitution is, strictly speaking, a self-directed one, it is doubtful
that the Stoics ever treated it as co-ordinate with the first-order
impulses associated with so-called social oikeiōsis or as one that
could enter into conflict with them. The sources do not support
a tidy association of social oikeiōsis with the presence of reason,
nor do they restrict the primary impulse to self-preservation to
non-rational animals. Self- and other-regarding forms of motiva-
tion cut across the distinction between rational and non-rational
agency. Both are rooted in the mechanisms of self-perception and
self-appropriation that ground the full range of activities appropri-
ate to animate organisms.
Second, this analysis brings out a basic respect in which the Stoic
account of animal psychology is supposed to be normative for the
rational case. In rational agents the character of impulse is con-
ditioned by the discursive and inferential abilities that underwrite
higher-order modes of phantasia and, crucially, by the capacity to
pass judgement on phantasiai the soul consciously entertains. These
or pleasurable results (Sen. Ep. . = LS B; S.E. M. . – = SVF iii. ).
Cf. Goldschmidt, Système stoïcien, –.
Cf. Plut. De amore : ‘But in man, whom she made a rational and poli-
tical being, inclining him to justice, law, religion, building of cities, and friend-
ship, [Nature] has placed the seeds of those things that are excellent, beautiful,
and fruitful—that is, the love of their children—following the first principles which
entered into the very constitution of their bodies [τῶν σωμάτων κατασκευαῖς]’ (trans.
after Brown).
Jacob Klein
cognitive capacities are constitutive of the rational faculty that is a
privileged expression of divine reason, in the Stoics’ view, but they
also bring with them increased scope for representational error,
with the result that rational agents are liable to stray from nature
in ways non-rational animals do not. Nothing is more contrary to
nature than a false assent, Cicero maintains, in that it deforms both
the cognitions of the soul and the impulses to which they give rise
(Fin. . = SVF iii. ). There is a suggestion in several texts
that in respect of representational accuracy, if not conceptual refine-
ment, the pneuma of an animal’s soul adheres more reliably and con-
sistently than that of human agents to the contours of the world.
This point helps to explain why the Stoics regard the precisely con-
sistent activities of non-rational animals—performed on the basis of
factive and largely automatic representations—as a structural model
of the cognitive conformity to nature that is supposed to obtain in
human beings.
Such a reconstruction is consonant with the prominent place
given to the oikeiōsis doctrine in even cursory allusions to Stoic ethi-
cal theory, and it supplies the Stoics with the sort of ethical foun-
dation they need: one that makes the end for each creature consist
in the functions appropriate to it in virtue of the kind of creature
it is. Under this interpretation, the Stoics are no longer saddled
with the burden of explaining how narrowly egoistic motivation is
supplanted in human beings by rational motivation of a fundamen-
tally different character. Instead, they are building an inductive case
for an analysis of teleological functioning that applies throughout
the scala naturae. With this account in view, we can see how Stoic
appeals to oikeiōsis might have figured in a defence of justice, as
Porphyry and the anonymous commentator on the Theaetetus af-
firm, and as passages in Marcus Aurelius and Cicero independently
suggest.
What is most striking about the Stoic theory, perhaps, is the
See esp. Stob. . (quoted above); . . (= LS B = SVF iii. ); Cic.
Tusc. . ; cf. also Plut. De amore –.
According to Goldschmidt, ‘Au sens le plus général du mot, le devoir (καθῆκον)
est le “convenable”, ce qui est en accord avec la nature de l’agent. Cette conformité
s’établit spontanément, comme une conséquence (ἀκόλουθον) naturelle, dans le com-
portement de l’animal; elle existe même chez l’homme, au niveau de la tendance
(ὁρμή). Plus tard, lors du développement de la nature spécifique de l’homme (la rai-
son) cette convenance ne se fait plus spontanément, mais résulte d’un choix réflé-
chi . . . on a vu que le devoir chez l’animal est parfait d’emblée’ (Système stoïcien,
–).
The Stoic Argument from oikeiōsis
degree to which the Stoics are prepared to integrate psychologi-
cal principles within a cosmic framework and interpret their sig-
nificance in cosmic terms. As Pembroke observes, discussions of
oikeiōsis have tended ‘to isolate Stoic ethics from other aspects of
their thinking, but Nature is not a specifically ethical term’. That
the oikeiōsis doctrine is embedded in this way within a cosmic te-
leological scheme is in keeping with the Stoics’ Socratic inheri-
tance through Xenophon, Plato, and Antisthenes. Antecedents to
Stoic conclusions about animal psychology appear in the Cyro-
paedia, and it is known from the reports of Sextus and Diogenes
that Zeno adopted teleological arguments found in the Memorabilia
(Xen. Cyr. . . ; Mem. . . –; S.E. M. . –; D.L. . – =
SVF i. ; cf. Cic. ND . ; . ). ‘I beg you to learn from Chry-
sippus’, says Epictetus, ‘what is the administration [dioikēsis] of
the universe, and what place therein the rational animal has’ (Diss.
. . , trans. Oldfather). On the interpretation I have offered,
Chrysippus argued that the administration of the cosmos is such
that the end for each creature consists in the performance of the ac-
tivities assigned to it and in the preservation of the governing faculty
from which they flow.
Colgate University
B I B L I O G R A P HY
Algra, K., ‘The Mechanism of Social Appropriation and its Role in Hel-
lenistic Ethics’ [‘Mechanism’], Oxford Studies in Ancient Philosophy,
(), –.
Algra, K., Barnes, J., Mansfeld, J., and Schofield, M. (eds.), The Cam-
bridge History of Hellenistic Philosophy (Cambridge, ).
Annas, J., The Morality of Happiness [Morality] (Oxford, ).
Bastianini, G., and Long, A. A., ‘Hierocles, Elementa moralia’ [‘Hierocles’],
in Corpus dei papiri filosofici, pt. . Commentari (Florence, ),
–.
Bees, R., Die Oikeiosislehre der Stoa: Rekonstruktion ihres Inhaltes
(Würzburg, ).
Betegh, G., ‘Cosmological Ethics in the Timaeus and Early Stoicism’, Ox-
ford Studies in Ancient Philosophy, (), –.
Blundell, M. W., ‘Parental Nature and Stoic οἰκείωσις’ [‘Parental Nature’],
Ancient Philosophy, (), –.
‘Oikeiōsis’, . Cf. Radice, Oikeiosis, –.
Jacob Klein
Bobzien, S., Determinism and Freedom in Stoic Philosophy (Oxford, ).
Bonazzi, M., ‘Antiochus’ Ethics and the Subordination of Stoicism’, in
M. Bonazzi and J. Opsomer (eds.), The Origins of the Platonic System:
Platonisms of the Early Empire and their Philosophical Contexts (Leuven,
), –.
Boys-Stones, G., ‘Physiognomy and Ancient Psychological Theory’, in id.
et al., Seeing the Face, Seeing the Soul: Polemon’s Physiognomy from
Classical Antiquity to Medieval Islam (Oxford, ), –.
Brennan, T., ‘Stoic Souls in Stoic Corpses’ [‘Souls’], in D. Frede and B.
Reis (eds.), Body and Soul in Ancient Philosophy (Berlin, ), –.
Brennan, T., The Stoic Life: Emotions, Duties, and Fate (Oxford, ).
Brink, C. O., ‘Οἰκείωσις and οἰκειότης: Theophrastus and Zeno on Nature in
Moral Theory’ [‘Theophrastus and Zeno’], Phronesis, (), –.
Brittain, C., ‘Antiochus’ Epistemology’, in Sedley (ed.), The Philosophy of
Antiochus, –.
Brittain, C., ‘Common Sense: Concepts, Definition and Meaning in and
out of the Stoa’ [‘Common Sense’], in D. Frede and B. Inwood (eds.),
Language and Learning: Philosophy of Language in the Hellenistic Age
(Cambridge, ), –.
Brittain, C., ‘Non-Rational Perception in the Stoics and Augustine’ [‘Per-
ception’], Oxford Studies in Ancient Philosophy, (), –.
Brouwer, R., The Stoic Sage: The Early Stoics on Wisdom, Sagehood, and
Socrates (Cambridge, ).
Brunschwig, J., ‘The Cradle Argument in Epicureanism and Stoicism’
[‘Cradle Argument’], in Schofield and Striker (eds.), The Norms of
Nature, –.
Chalmers, D., The Conscious Mind: In Search of a Fundamental Theory
(Oxford, ).
Cooper, J., ‘Chrysippus on Physical Elements’ [‘Elements’], in Salles (ed.),
God and Cosmos in Stoicism, –.
Cooper, J., ‘Stoic Autonomy’, in id., Knowledge, Nature, and the Good: Es-
says on Ancient Philosophy (Princeton, ), –.
Dirlmeier, F., Die Oikeiosislehre Theophrasts (= Philologus, suppl. .;
Leipzig, ).
Engberg-Pedersen, T., ‘Discovering the Good: oikeiōsis and kathēkonta
in Stoic Ethics’, in Schofield and Striker (eds.), The Norms of Nature,
–.
Engberg-Pedersen, T., The Stoic Theory of Oikeiosis: Moral Development
and Social Interaction in Early Stoic Philosophy (Aarhus, ).
Evans, G., The Varieties of Reference [Reference], ed. J. McDowell (Oxford,
).
Fortenbaugh, W. W. (ed.), On Stoic and Peripatetic Ethics: The Work of
Arius Didymus (New Brunswick, NJ, ).
The Stoic Argument from oikeiōsis
Fortenbaugh, W. W., Huby, P. M., Sharples, R. W., and Gutas, D. (eds.),
Theophrastus of Eresus: Sources for his Life, Writings, Thought and Influ-
ence, vols. (Leiden, ).
Frede, M., ‘On the Stoic Conception of the Good’ [‘Good’], in K. Iero-
diakonou (ed.), Topics in Stoic Philosophy (Oxford, ), –.
Frede, M., ‘The Stoic Conception of Reason’ [‘Reason’], in K. Boudouris
(ed.), Hellenistic Philosophy, vols. (Athens, –), ii. –.
Furley, D. J., ‘Cosmology’, in Algra et al. (eds.), The Cambridge History of
Hellenistic Philosophy, –.
George, D., ‘Lucan’s Caesar and Stoic oikeiōsis Theory: The Stoic Fool’,
Transactions of the American Philological Association, (), –.
Gill, C., The Structured Self in Hellenistic and Roman Thought [Self] (Ox-
ford, ).
Goldschmidt, V., Le Système stoïcien et l’idée de temps [Système stoïcien],
th edn. (Paris, ).
Görgemanns, H., ‘Oikeiōsis in Arius Didymus’ [‘Oikeiōsis’], in Forten-
baugh (ed.), On Stoic and Peripatetic Ethics, –.
Graver, M., Cicero on the Emotions (Chicago, ).
Graver, M., Stoicism and Emotion (Chicago, ).
Hahm, D., The Origins of Stoic Cosmology (Columbus, Ohio, ).
Inwood, B., ‘Antiochus on Physics’ [‘Antiochus’], in Sedley (ed.), The Phi-
losophy of Antiochus, –.
Inwood, B., ‘Comments on Professor Görgemanns’ Paper’ [‘Comments’],
in Fortenbaugh (ed.), On Stoic and Peripatetic Ethics, –.
Inwood, B., Ethics after Aristotle (Cambridge, Mass., ).
Inwood, B., Ethics and Human Action in Early Stoicism [Ethics and Human
Action] (Oxford, ).
Inwood, B., ‘Hierocles: Theory and Argument in the Second Century ..’
[‘Hierocles]’, Oxford Studies in Ancient Philosophy, (), –.
Inwood, B., Seneca: Selected Philosophical Letters [Seneca] (Oxford, ).
Inwood, B., ‘Stoic Ethics’, in Algra et al. (eds.), The Cambridge History of
Hellenistic Philosophy, –.
Inwood, B., ‘Walking and Talking: Reflections on Divisions of the Soul
in Stoicism’, in K. Corcilius and D. Perler (eds.), Partitioning the Soul:
Debates from Plato to Leibniz (Berlin, ), –.
Kerferd, G., ‘The Origin of Evil in Stoic Thought’ [‘Origin’], Bulletin of
the John Rylands University Library of Manchester, (), –.
Kerferd, G., ‘The Search for Personal Identity in Stoic Thought’
[‘Search’], Bulletin of the John Rylands University Library of Manchester,
(), –.
Kerferd, G., ‘Two Problems concerning Impulses’, in Fortenbaugh (ed.),
On Stoic and Peripatetic Ethics, –.
Jacob Klein
Kühn, W., ‘L’attachement à soi et aux autres’ [‘L’attachement’], in M.-O.
Goulet-Cazé (ed.), Études sur la théorie stoïcienne de l’action (Paris, ),
–.
Kühn, W., ‘L’attachement à soi face aux fins morales: une question stoï-
cienne’, Archiv für Geschichte der Philosophie, (), –.
Lapidge, M., ‘Stoic Cosmology’, in J. Rist (ed.), The Stoics (Berkeley,
), –.
Lee, C.-U., Oikeiosis: Stoische Ethik in naturphilosophischer Perspektive
(Freiburg i.Br. and Munich, ).
Long, A. A., ‘Hierocles on oikeiōsis and Self-Perception’ [‘Hierocles’], in
id., Stoic Studies, –.
Long, A. A, ‘Representation and the Self in Stoicism’ [‘Representation’],
in id., Stoic Studies, –.
Long, A. A., ‘Soul and Body in Stoicism’ [‘Soul and Body’], in id., Stoic
Studies, –.
Long, A. A, Stoic Studies (Berkeley, ).
Long, A. A., ‘The Harmonics of Stoic Virtue’, in id., Stoic Studies, –.
Long, A. A., ‘The Logical Basis of Stoic Ethics’, in id., Stoic Studies,
–.
Long, A. A., ‘Theophrastus and the Stoa’ [‘Theophrastus’], in J. M. van
Ophuijsen and M. van Raalte (eds.), Theophrastus: Reappraising the
Sources (New Brunswick, NJ, ), –.
Long, A. A., and Sedley, D. (eds.), The Hellenistic Philosophers [LS],
vols. (Cambridge, ).
Lorenz, H., The Brute Within: Appetitive Desire in Plato and Aristotle (Ox-
ford, ).
McCabe, M. M., ‘Extend or Identify: Two Stoic Accounts of Altruism’
[‘Two Accounts’], in R. Salles (ed.), Metaphysics, Soul, and Ethics in An-
cient Thought: Themes from the Work of Richard Sorabji (Oxford, ),
–.
Menn, S., ‘Physics as a Virtue’, Proceedings of the Boston Area Colloquium
in Ancient Philosophy, (), –.
Peacocke, C., ‘Does Perception Have a Nonconceptual Content?’, Journal
of Philosophy, (), –.
Pembroke, S., ‘Oikeiōsis’, in A. A. Long (ed.), Problems in Stoicism (Lon-
don, ), –.
Pohlenz, M., Die Stoa (Göttingen, ).
Pohlenz, M., Grundfragen der stoischen Philosophie [Grundfragen] (Göttin-
gen, ).
Polito, R., ‘Antiochus and the Academy’, in Sedley (ed.), The Philosophy
of Antiochus, –.
Powers, N., ‘The Stoic Argument for the Rationality of the Cosmos’ [‘Ar-
gument’], Oxford Studies in Ancient Philosophy, (), –.
The Stoic Argument from oikeiōsis
Radice, R., ‘Oikeiosis’: Ricerche sul fondamento del pensiero stoico e sulla sua
genesi [Oikeiosis] (Milan, ).
Ramelli, I., Hierocles the Stoic: Elements of Ethics [Hierocles], trans. D.
Konstan (Atlanta, ).
Reydams-Schils, G., ‘Human Bonding and oikeiōsis in Roman Stoicism’
[‘Bonding’], Oxford Studies in Ancient Philosophy, (), –.
Rist, J. M., ‘On Greek Biology, Greek Cosmology and Some Sources of
Theological pneuma’, in id., Man, Soul and Body: Essays in Ancient
Thought from Plato to Dionysius (Aldershot, ), –.
Rist, J. M., Stoic Philosophy (Cambridge, ).
Rist, J. M., ‘Zeno and Stoic Consistency’, Phronesis, (), –.
Salles, R. (ed.), God and Cosmos in Stoicism (Oxford, ).
Salles, R., The Stoics on Determinism and Compatibilism (Aldershot, ).
Sandbach, F. H., Aristotle and the Stoics (Cambridge, ).
Sauvé-Meyer, S., ‘Chain of Causes: What is Stoic Fate?’, in Salles (ed.),
God and Cosmos in Stoicism, –.
Schmitz, P., ‘Cato Peripateticus’: Stoische und peripatetische Ethik im Dia-
log. Cic. ‘fin.’ und der Aristotelismus des ersten Jh. v. Chr. [Cato] (Berlin,
).
Schofield, M., ‘Cardinal Virtues: A Contested Socratic Inheritance’, in
A. G. Long (ed.), Plato and the Stoics (Cambridge, ), –.
Schofield, M., ‘Epicurean and Stoic Political Thought’, in C. Rowe and
M. Schofield (eds.), The Cambridge History of Greek and Roman Politi-
cal Thought (New York and Cambridge, ), –.
Schofield, M., The Stoic Idea of the City (Cambridge, ).
Schofield, M., ‘Two Stoic Approaches to Justice’ [‘Justice’], in A. Laks and
M. Schofield (eds.), Justice and Generosity: Studies in Hellenistic Social
and Political Philosophy (Cambridge, ), –.
Schofield, M., and Striker, G. (eds.), The Norms of Nature: Studies in Hel-
lenistic Ethics (Cambridge, ).
Sedley, D., ‘Antiochus as Historian of Philosophy’, in id. (ed.), The Philo-
sophy of Antiochus, –.
Sedley, D. (ed.), The Philosophy of Antiochus (Oxford, ).
Sharples, R. W. (trans.), Alexander of Aphrodisias: Supplement to On the
Soul (Ithaca, NY, ).
Sorabji, R., Animal Minds and Human Morals: The Origins of the Western
Debate (Ithaca, NY, ).
Stephens, W. O., ‘Epictetus on Beastly Vices and Animal Virtues’, in D. R.
Gordon and D. B. Suits (eds.), Epictetus: His Continuing Influence and
Contemporary Relevance (Rochester, NY, ), –.
Striker, G., ‘Antipater, or the Art of Living’, in ead., Essays, –.
Striker, G., Essays on Hellenistic Epistemology and Ethics [Essays] (Cam-
bridge, ).
Jacob Klein
Striker, G., ‘Following Nature’, in ead., Essays, –.
Striker, G., ‘The Role of oikeiōsis in Stoic Ethics’ [‘Role’], in ead., Essays,
–.
Von Arnim, H., Arius Didymus’ Abriß der peripatetischen Ethik (Vienna,
).
White, N., ‘The Basis of Stoic Ethics’ [‘Basis’], Harvard Studies in Clas-
sical Philology, (), –.
Woolf, R. (trans.), Cicero: On Moral Ends (Cambridge, ).
Zagdoun, M. A., ‘Problèmes concernant l’oikeiōsis stoïcienne’ [‘Pro-
blèmes’], in G. Romeyer-Dherbey and J.-B. Gourinat (eds.), Les
Stoïciens (Paris, ), –.
ALEX A ND E R A G A I N S T GALEN ON
MOTION: A M E R E L O G ICAL DEBATE?
ORNA HARARI
This argument has two parts. In the first Aristotle assumes that
anything that moves by something comes to rest when something
else ceases moving, and in the second he argues that this assump-
tion holds for anything that moves. The second part opens with a
preparatory stage where Aristotle assumes that anything that moves
Powers and Movements’, Bulletin of the Institute of Classical Studies, (), –
at n. .
Pines had no access to the Escorial manuscript, where Galen’s criticism is
presented. The proponents of the current view do not entirely disregard Galen’s
theoretical commitments: Silvia Fazzo admits that Galen’s criticism derives from
his understanding of the expression ‘moving primarily and in itself’ (‘Alexandre
d’Aphrodise contre Galien’, –) but nevertheless views his criticism as purely
logical. Michael Frede notes that Galen is described in this treatise as an adherent
of Plato (‘Galen’s Theology’, ) but does not examine the implications of this evi-
dence for his understanding of the scope of Galen’s criticism.
All translations from Greek are mine.
Alexander against Galen on Motion
is divisible; posits a magnitude AB and divides it at C into the parts
AC and CB; and presents the thesis that he proves by reductio ad im-
possibile in the next stage: ‘if the part CB does not move the whole
AB does not move’. This argument is based on another assump-
tion that Aristotle presents earlier in Physics . and in Physics .
(a–), ‘AB moves in itself [καθ᾿ αὑτό] and not because one of
its parts is in motion [μὴ 〈τῷ τῶν〉 τούτου τι κινεῖσθαι]’ (b). From
this assumption Aristotle infers that if a part of AB is at rest AB does
not move, for if it moves, it does so when AC is in motion and CB is
at rest, but this conclusion is in conflict with the assumption that AB
moves in itself and primarily (καθ᾿ αὑτὸ καὶ πρῶτον). This inference
leads to the conclusion that AB is moved by something, since in ac-
cordance with Aristotle’s initial assumption it ceases moving when
something (i.e. its part CB) comes to rest. At the final stage, Aris-
totle generalizes this conclusion, arguing that anything that moves
Aristotle proves this assumption at Phys. . , b–a.
In the Refutation of Galen Alexander reports that Galen interpreted this stage of
Aristotle’s argument as a reductio ad absurdum (a– Escorial). Alexander rejects
this interpretation (b, ; a Escorial) because it may support Galen’s conten-
tion that the proposition ‘AC is in motion and CB is at rest’, which is the conclusion
of the reductio step, is impossible.
It is difficult to understand how Aristotle justifies this inference. In Physics .
the distinction between things that move in themselves (καθ᾿ αὑτό) and primarily
(πρῶτον) and things that move accidentally (κατὰ συμβεβηκός) and partially (κατὰ
μέρη) implies that motion should be predicated of its proper subject, not of an ac-
cident or a part of the moving thing. For example, ‘walking’ should be predicated
of man and not of the musical, but ‘playing the piano’ is predicated καθ ᾿ αὑτό of the
musical, not of man. Similarly, ‘convalescing’ is not predicated πρῶτον of the body
when only the health of the eye or the thorax is restored, but ‘growing’ is predicated
primarily of a thirteen-year-old boy, although his eyes no longer grow. Accordingly,
this account does not entail that if a part does not move the whole does not move
primarily and in itself. The abstract magnitude AB admits a similar account. If the
parts AC and CB are distinguished but not spatially separated, nothing prevents the
predicate ‘increasing in size’ from holding primarily and in itself for AB, when the
length of AC increases, while the size of CB remains the same. Aristotle’s inference
may hold for locomotion of abstract magnitudes or homoeomerous bodies. When
AC moves and CB is at rest, the whole moves by virtue of its part because in this
case AC pushes or drags CB. This suggestion may explain why Aristotle uses this
inference in arguing that AB is moved by something. However, Aristotle does not re-
strict his argument to locomotion of abstract magnitudes or homoeomerous bodies.
In view of these considerations, it is difficult to see how Aristotle justifies the in-
ference from the assumption that CB is at rest to the conclusion that AB does not
move primarily and in itself. This inference exposes Aristotle’s argument to Galen’s
criticism. It implies that any whole does not moves primarily and in itself when one
of its parts is at rest, thereby giving rise to the contention that the assumption that
CB is at rest is incompatible with the assumption that AB moves primarily and in
itself.
Orna Harari
is moved by something because anything that moves is divisible and
by the above argument it comes to rest if its part is not in motion.
Galen’s criticism of this argument is concisely described in the
following passage from Simplicius’ commentary on Physics . :
The most scholarly [φιλολογώτατος] Galen criticized this argument, and
others criticize [it] too, on the grounds that it uses an impossible hypothe-
sis that says that CB does not move, although AB moves primarily and in
itself. (. – Diels)
This passage serves as the main evidence in support of the view that
Galen’s criticism is confined to the validity of Aristotle’s argument.
It suggests that Galen directly questions the assumption that a part
of AB is at rest (henceforth the ‘resting part assumption’), but not
the other assumptions or more importantly the thesis that Aris-
totle’s argument establishes. Simplicius does not present Galen’s
argument in detail, but his brief description suggests that Galen
argued that the resting part assumption is impossible because it
contradicts the assumption that AB moves primarily and in itself.
The Refutation of Galen confirms this conclusion. Here Alexan-
der reports that Galen holds that Aristotle’s argument yields no
conclusion because it rests on a contradictory assumption (a–
Escorial) and that the resting part assumption amounts to the
assumption that AB moves and does not move at the same time
(b–, b, a– Escorial).
These considerations clearly show that Galen questioned the vali-
dity of Aristotle’s argument, but they do not entail the conclusion
that his criticism is merely logical. The question whether the resting
part assumption is contradictory cannot be determined on formal
grounds alone but depends on the interpretation of characterization
of AB as moving primarily and in itself. The Refutation of Galen
confirms this conclusion and also informs us that Galen did not base
his argument on Aristotle’s definition of moving primarily and in
As David Ross points out, this argument is problematic because it implies that
the whole’s motion is causally dependent on the part’s motion, whereas Aristotle’s
assumptions entail that these motions are merely logically dependent (W. D. Ross,
Aristotle’s Physics: A Revised Text with Introduction and Commentary (Oxford,
), , and R. Wardy, The Chain of Change: A Study of Aristotle’s Physics VII
(Cambridge, ), –). The debate between Alexander and Galen does not turn
on this problem but, as I show in sect. . below, Alexander indirectly addresses
it through his claim that this argument does not seek to establish a causal relation
between the motions of the whole and the part.
Fazzo, ‘Alexandre d’Aphrodise contre Galien’, .
Alexander against Galen on Motion
itself but, as Alexander says several times, on the assumption that
this expression means ‘having an inner source of motion and not
being moved by any external thing’ (a– Carullah; a–,
b–a, a–, b–, b– Escorial).
Galen’s use of this definition of moving primarily and in itself
casts doubt on the logical interpretation of his argument. On this
interpretation, Galen does not achieve his principal aim; he attempts
to challenge the validity of Aristotle’s argument but fails because he
bases his criticism on his definition of moving primarily and in itself,
and not on Aristotle’s. The logical interpretation is even more prob-
lematic, given that Galen could criticize the argument in Physics .
on the basis of Aristotle’s definition of moving primarily and in it-
self. Galen could argue, as others did in antiquity, that Aristotle begs
the question. That is to say, he could contend that since Aristotle’s
definition implies that a thing which moves in virtue of having a part
that moves does not move primarily and in itself, the assumption
that a part of a whole is at rest amounts to the assumption that the
whole does not move primarily and in itself, and so Aristotle does
not prove but assumes the conclusion that the whole does not move
primarily and in itself (Simpl. In Phys. . – Diels). In a
similar vein, Galen could further argue that Aristotle’s definition is
incompatible with the assumption that AB moves primarily and in
itself. The latter assumption implies that AB moves as a whole and
not in virtue of having a part that moves, whereas the resting part
assumption states that one part of AB, i.e. CB, is at rest, hence these
assumptions amount to the assumption that AB both moves and
does not move primarily and in itself. In contrast, Galen’s definition
of moving primarily and in itself as having an inner source of motion
and not being moved by an external thing does not immediately im-
ply that a thing that moves in this way cannot have one part in motion
In his actual argument Galen makes the stronger claim that the resting part as-
sumption is impossible. This conclusion does not follow from Aristotle’s definition
of moving primarily and in itself because it does not imply that a thing that moves
primarily and in itself cannot stop moving in this way. As I show below, Galen bases
his argument on the assumption that the motion of things that move primarily and
in themselves strongly depends on their substance or nature. By this assumption,
the resting part assumption is impossible because things that move primarily and
in themselves according to his definition cannot stop moving as long as they retain
their substance or nature. For this reason, Galen’s argument leads to the conclusion
that Aristotle’s argument rests on the contradictory assumption ‘AB both moves and
does not move’, whereas the above argument from Aristotle’s definition of moving
primarily and in itself leads to the qualified contradiction ‘AB both moves and does
not move primarily and in itself ’.
Orna Harari
and another at rest, and hence does not straightforwardly lead to
contradiction. These considerations indicate that the scope of Ga-
len’s criticism of Aristotle cannot be understood without a detailed
analysis of Galen’s argument and an examination of the following
questions. () How does Galen derive the contradiction ‘AB moves
and does not move’ from his definition of moving primarily and in
itself? () Does Galen’s definition disprove Aristotle’s thesis that
anything that moves is moved by something, in implying that things
that have an inner source of motion are not moved by something?
And () why does he base his argument on his definition of moving
primarily and in itself and not on Aristotle’s definition?
Moreover, Galen’s criticism of Aristotle’s argument is based on
the assumption that the characterization ‘moving primarily and in
itself’ holds only for simple bodies whose parts are similar to each
other and to the whole (b– Carullah; a–, a Escorial).
This assumption may refer to organic tissues or to the natural ele-
ments (i.e. earth, water, air, and fire), but Alexander’s claim that
Galen maintains that only animate bodies move primarily and in
themselves (b– Escorial) suggests that Galen has the former in
mind. This suggestion finds support in Galen’s extant writings,
where the expression ‘simple’ or ‘primary bodies’ refers to homoeo-
merous bodily parts, such as cartilage, bones, nerves, membranes,
ligaments, and the flesh of organs such as heart, liver, brain, kid-
neys, lungs, and spleen. Understood in this light, Galen criticizes
Aristotle’s argument from the standpoint of his own medical the-
ory. His understanding of the expression ‘moving primarily and in
itself’ echoes his definition of one type of motion—activity [ἐνέρ-
γεια]—as an active motion [δραστικὴ κίνησις] that originates from
the thing itself and not from an external thing. And his claim that
this characterization holds only for homoeomerous parts reflects
Pines claims that it is more reasonable to identify the simple bodies with the
elements (‘Omne quod movetur’, ). Alexander’s mention of the elements gives rise
to this interpretation, but it does not accord with Galen’s view as presented in the
Escorial manuscript, which was unknown to Pines.
e.g. Elem. sec. Hipp. , . – De Lacy (= i. – K.); . – De Lacy
(= i. K.); , . – De Lacy (= i. K.); Nat. fac. . , . – Helmreich
(= ii. K.); AA . , ii. K.; . , ii. K.; QAM . , . – Mueller (= iv.
K.); PHP . , . – De Lacy (= v. K.); . , . – De Lacy (= v.
K.); Meth. med. . , x. K. In Galen’s view, homoeomerous parts are simple
in relation to perception, but in effect they are made of the four humours.
PHP . , De Lacy (= v. – K.); De usu part. . , ii. Helmreich
(= iii. – K.); Meth. med. . , x. K.; . , x. K.; Loc. aff. . , viii. K.
Alexander against Galen on Motion
his view that the primary activities of the body are found in these
parts. Considering the centrality of these views to Galen’s medical
theory, it is unlikely that he propounds them in the present context
only for the sake of exposing a logical flaw in Aristotle’s argument
and ignores the thesis that it establishes, although it may cast doubt
on his accounts of activity and homoeomerous parts.
In the light of these considerations, I examine Galen’s argument
against Aristotle in the context of his medical theory. I show first
that his contention that Aristotle’s argument rests on contradictory
assumptions depends on his definition of moving primarily and in
itself and on the assumption that only homoeomerous bodies move
in this way. Specifically, I show that Galen argues that the rest-
ing part assumption contradicts his definition of moving primarily
and in itself because the motion of homoeomerous bodies strongly
depends on their substance and therefore they cannot have one
part in motion and another at rest. In the second part of my dis-
cussion of Galen I argue that his account of the motion of ho-
moeomerous parts entails a rejection of the conclusion of Aristotle’s
argument, since the motion of these parts is internally caused and
while external causes necessitate alterations of this motion, they do
not cause it in their own right. Here I also show that, from Ga-
len’s viewpoint, he effectively challenges the validity of Aristotle’s
argument, since Galen’s definition of moving primarily and in it-
self reflects his interpretation of Aristotle and his followers. Next, I
analyse Alexander’s reply to Galen, showing that it comprises two
independent arguments: one that is confined to the validity of the
inference from the resting part assumption to the conclusion ‘AB
does not move primarily and in itself’, and the other that defends
the thesis that anything that moves is moved by something, by se-
curing the validity and soundness of Aristotle’s whole argument. I
argue that in the first argument Alexander shows that the inference
that Galen criticizes is valid because the resting part assumption can
serve as an antecedent of a true conditional, even if it is impossible as
an assertoric proposition. Regarding the second argument, I claim
that here Alexander shows that the resting part assumption is true
without qualification of all continuous bodies because the parts of
these bodies exist only in potentiality, and so can be at rest so long
as they are continuous with the whole. My analysis of these argu-
ments leads to the conclusion that the debate between Alexander
On this sect. . below.
Orna Harari
and Galen is rooted in a more fundamental disagreement over two
metaphysical issues, hylomorphism and efficient causality, because
it turns on the questions whether inner causal factors are ontolo-
gically distinct from their bearers and whether external causes are
the primary efficient causes.
. Galen’s criticism
.. Galen’s argument
In the Refutation of Galen Alexander presents four slightly different
formulations of Galen’s argument. To clarify the respective roles of
Galen’s definition of moving primarily and in itself and its restric-
tion to homoeomerous bodies, I discuss these formulations sepa-
rately, beginning with the following two:
When some magnitude, in the event that its parts are continuous, moves
essentially and according to first intention, it must be then one of the simple
first bodies. These are the bodies whose parts are similar, since these alone
are the things that move essentially and according to first intention. Since
the things whose natural principle of motion is in them are the first simple
bodies, and since these consist of similar parts, the part in these things is
no other than the whole. Hence, Aristotle was not definitely right with re-
spect to continuous things [in holding] that when one part of them stops,
the whole stops. For the part in these things is no other than the whole.
(a– Escorial)
There is basically no difference between our maintaining that some part
of a thing that moves essentially is at rest and our maintaining that the
whole moves and does not move at one and the same time. For under no
circumstances is there any difference in these things between the part and
the whole. (b– Escorial)
A concise formulation of this inference is found in the Refutation of Galen,
where Alexander reports that Galen claimed that each of Aristotle’s assumptions
entails its contradictory (a– Escorial); that is, the assumption ‘AB moves’ en-
tails that CB cannot be at rest, and the assumption ‘CB is at rest’ entails that AB
does not move.
Orna Harari
.. Internal and external causes of activities
The idea that homoeomerous bodies move primarily and in them-
selves is prevalent in the Galenic corpus, where he often associates
these bodies with a type of motion called ‘activity’ (ἐνέργεια), which
is an active motion (δραστικὴ κίνησις) that arises from the thing itself
(ἐξ ἑαυτοῦ) in contrast to affection (πάθος), which arises from another
thing. This idea finds expression in Galen’s view that there are two
genera of activity: one which is the primary and most principal (κυ-
ριώτατον) activity of the simple homoeomerous parts and the other
which is the accidental or secondary activity of the whole organs
(Const. art. med. , . – Fortuna = i. – K.). In keeping with
this view, Galen classifies diseases into those of the homoeomerous
parts and those of the organic parts, and regards the former as the
primary disease of all. This classification is based on his assumption
that activities such as seeing and walking are not activities of the
whole organs, e.g. of the eye or the leg, but of one of their homoeo-
merous parts, e.g. the crystalline humour and the muscles (Meth.
med. . , x. – K.). Further, Galen assumes that each homoeo-
merous part necessarily has one unique activity. He grounds this
assumption in the natural constitution of these parts, saying that
they were generated by nature for the sake of one activity (Elem. ex
Hipp. , . – De Lacy = i. K.; Morb. diff. , vi. K.),
and associates their activities with their material composition, stat-
ing that the number of activities corresponds to the number of the
mixtures (κράσεις) of the homoeomerous parts (San. tu. . , . –
Koch = vi. K.). In therapeutic contexts, Galen states that the ac-
tivities of homoeomerous parts causally depend on the mixture. In
On the Therapeutic Method . he identifies the active cause (αἰτία
δραστική) of activity with the good mixture (εὐκρασία) of the four
qualities, i.e. hot, cold, dry, and moist (x. K.), and says that
activities come to be by (ὑπό) the homoeomerous parts (x. K.).
However, in the more theoretical context of On the Natural Facul-
ties he calls the cause of activity dunamis (e.g. . , . Helm-
reich = ii. K.), thereby suggesting that the material constitution of
For references see n. above. See also Meth. med. . , x. K., where he
ascribes a similar account of activity to the ‘ancients’. Galen presents another sense
of activity that holds for motion according to nature, regardless of whether it arises
from the thing itself or from another thing (PHP . , . – De Lacy = v. K.).
e.g. Part. hom. diff. , . Strohmaier; Nat. fac. . , . – Helmreich
(= ii. K.); Meth. med. . , x. K.
Alexander against Galen on Motion
the homoeomerous parts does not provide the full causal explana-
tion of their activities. These divergent views can be understood in
the light of Galen’s account of dunamis and his agnosticism regard-
ing the substance of this active cause. The most detailed discussion
of these matters is found in the following passage from Causes of
Pulses . :
From this it is clear that there is a certain cause by which they [sc. the
pulses] are moved for a time, but it is difficult to discover what it is. One
[person] says that it is the innate heat, another that it is the tonos, or the
property of the mixture, or the entire composition of the bodies, or only
the pneuma, or one of these things, or all of them together. And others
presently introduced an incorporeal dunamis that uses all, or some, or one
of the mentioned proper organs of motion. We call this productive [δη-
μιουργοῦσα] cause of the pulses, whatever it may be, dunamis, even if we
are ignorant of its substance, from its being capable of producing pulses,
just as, I believe, we customarily call any other thing dunamis from [the fact
that] it can make the very thing that it can. For dunamis is of something and
we possess an understanding of its conception in [the category of] relative,
and for this reason we call it so when we are ignorant of the substance. (ix.
. –. K.)
Here Galen says that dunamis is a term that he uses for the active
cause of the pulse because he is ignorant of its substance; but he
clarifies that it is not an empty term but a term that expresses the
relation between cause and effect, i.e. the cause’s capacity to bring
about an effect. By using this term, then, Galen can offer a causal
explanation of bodily activities without committing himself to any
definite position regarding the factor that serves as a cause. So in
stating that the active cause is a dunamis, Galen does not offer an al-
ternative to the different accounts that he enumerates above. Being
a relational term, dunamis is not an absolute distinct factor which, to
Cf. Nat. fac. . , where Galen says that each part has its activity because of
(διά) the mixture of the four qualities (. – Helmreich = ii. K.). A weaker
expression of this view is found in On the Preservation of Health . , where he says
that differences of activity follow (ἀκολουθοῦσιν) from differences in mixtures (. –
Koch = vi. K.). For this ambiguity in the related context of the soul see G. E. R.
Lloyd, ‘Scholarship, Authority, and Argument in Galen’s Quod animi mores’, in P.
Manuli and M. Vegetti (eds.), Le opere psicologiche di Galeno: atti del terzo colloquio
Galenico internazionale, Pavia, – settembre (Naples, ), – at –
, P. Donini, ‘Psychology’, in R. J. Hankinson (ed.), The Cambridge Companion to
Galen (Cambridge, ), – at ; P. N. Singer (ed.), Galen: Psychological
Writings (Cambridge, ), n. .
Cf. Nat. fac. . , . – Helmreich (= ii. K.); Prop. plac. . , . –
Nutton; Temp. . , i. K.
Orna Harari
use Galen’s metaphor, inhabits substances as we inhabit our houses
(QAM . , . – Mueller = iv. K.), but expresses only the
relation between cause and effect. Accordingly, within Galen’s the-
ory the ambiguity as to whether activities are caused by the mixture
of the homoeomerous parts or by the dunamis is less consequential
than it may seem at first glance. In Galen’s relational analysis of in-
ner active causes, these two accounts do not refer to distinct causal
factors, i.e. the mixture of the homoeomerous part and a dunamis,
but to the same factor under different descriptions: one that de-
scribes the cause as an absolute entity, and the other, relational de-
scription that expresses only its capacity to cause an effect.
In keeping with this relational account, Galen formulates his
notion of dunamis in exact correspondence to his account of its cor-
relative, i.e. its effect ‘activity’ (ἐνέργεια). He counts dunamis among
the cohesive causes (αἴτια συνεκτικά), which he construes in oppo-
sition to the Stoics as causes of generation, not of existence (Caus.
puls. . , ix. – K.; CC , . – Kollesch–Nickel–Strohmaier),
and stresses that activities are processes, or in his words ‘have
their being in generation’ (Meth. med. . , x. K.). Further,
he states that bodily activities, though they may seem persistent,
are constantly in motion (CC , . – Kollesch–Nickel–
Strohmaier). In On the Utility of the Parts . Galen appeals to
the notion of dunamis in accounting for this constant motion:
Just as Homer makes Hephaestus’ works of art self-moving . . . so you
should observe that in the body of the animal there is nothing inert, nothing
unmoved, but that every part performs its different well-conducted activity
with the aid of a suitable construction because the creator bestowed upon
each certain divine dunameis. (i. . – Helmreich = iii. . – K.)
Of the causes that change pulses, some are causes of their generation while
others are only causes of alteration. Causes of generation are the function
[χρεία] for the sake of which they come to be, the dunamis by which [they
come to be], and the organs through which they are propagated, while the
causes of alteration are the others that are called preceding [προηγούμενα]
and those that are antecedent [προκατάρχοντα] even to them. For the genus
of causes is threefold not only with regard to pulses but with regard to
everything else. One, the primary and most principal, is that which they
call ‘cohesive’ and it derives its name from the fact that it contains their
substance and it is a cause of generation, as said earlier. The other two ge-
nera are not causes of the generation of pulses but they are causes of change
of pulses that have already been generated. (. , ix. . –. K.)
This passage clarifies Galen’s claim that the cohesive causes are the
most principal causes of bodily activities. It shows that this claim
reflects not only his preference for explanations in terms of pro-
ximate causes (Symp. diff. , vii. K.), but his view that bodily
activities are not primarily caused by external causes. According to
this passage, external causes are called causes because of the cohes-
ive causes; they do not change the pulse in virtue of their substance
and nature but because they transmit (διαδίδοσθαι) an alteration to
one of its cohesive causes; and the pulse cannot be altered before the
alteration is transmitted to one of the cohesive causes. Galen does
not clarify these claims here, but his definition of the pulse helps to
elucidate them. In Difference of Pulses . Galen defines the pulse
as follows:
The pulse is a unique activity primarily of the heart and secondarily of the
arteries, which are moved according to dilatation and contraction by vital
dunamis for the sake of the preservation of the symmetry of the innate heat
and the generation of psychic pneuma in the brain. (viii. . – K.)
From this definition we see that the dunamis is not simply an ac-
tive cause but a cause that brings about a certain effect for the sake
of something. This definition explains Galen’s claim that the pulse
cannot be altered before the alteration is transmitted to one of the
cohesive causes. Specifically, it suggests that the process whereby
external cold leads to an alteration of the pulse involves two types
Orna Harari
of causal explanation: a mechanistic explanation that accounts for
the transmission of alteration to one of the cohesive causes and a te-
leological explanation that accounts for the alteration of the pulse.
Closer examination of the final stage that leads to an alteration of
the pulse clarifies this distinction. This stage involves two altera-
tions: in the first fever alters the function of the pulse, i.e. the pre-
servation of the symmetry of the innate heat, and in the second this
effect on the function causes the alteration of the pulse. The first al-
teration is an ordinary qualitative change that fever causes in virtue
of its nature: it alters the symmetry of the innate heat by heating.
By contrast, the second alteration (i.e. the alteration of the pulse)
is teleological: it results in a quicker, stronger, and more frequent
pulsation that counteracts, rather than transmits, the damaging ef-
fect of heating on the symmetry of the innate heat. This alteration
cools the innate heat by drawing in cold air in dilatation and dis-
pelling the smoky residues (καπνώδη περιττώματα) of the innate heat
in contraction. Being teleological, this effect cannot be brought
about by fever or heat by itself, but its occurrence requires the duna-
mis that changes the pulse for the sake of its function, i.e. maintain-
ing the symmetry of the innate heat. Thus, although the alteration
of the pulse is necessitated by external causes, it is an activity that
arises from the thing itself, namely from the cohesive causes.
This conclusion reconciles Galen’s definition of activity as mo-
tion that arises from the thing itself with his claim that alterations
of activities are caused by external causes. It shows that the al-
teration of the pulse is a teleological reaction of the inner dunamis
that externally caused alterations trigger when they affect the func-
tion. Further, it accords with Galen’s On Antecedent Causes, whose
principal aim is to show that external factors cause impairments of
bodily activities although they are necessary but insufficient condi-
tions for these effects. According to the above analysis, the altera-
tion of the pulse cannot occur without the effect of the external cold
and fever on the symmetry of the innate heat, but this effect cannot
cause the alteration of the pulse without the dunamis that changes it
for the sake of its function. Finally, this conclusion holds not only
for the pulse but for all bodily activities, since in the opening para-
This analysis is based on Caus. puls. . , ix. K. A similar analysis is found
in R. J. Hankinson, ‘Philosophy of Nature’, in id. (ed.), The Cambridge Companion
to Galen, – at –. Cf. Hankinson, ‘Causation in Galen’, –.
Cf. R. J. Hankinson (ed. and comm.), Galen: On Antecedent Causes (Cam-
bridge, ), –.
Alexander against Galen on Motion
graph of Causes of Pulses . Galen explicitly states that the dis-
tinction between the three genera of causes (i.e. cohesive, preceding,
and antecedent) holds for all bodily activities, not only for the pulse.
Thus, if cohesive causes are involved in any causal process that leads
to bodily activities, externally caused changes do not suffice to bring
about the teleological effects that the cohesive causes cause. Con-
sequently, altered bodily activities too are internally caused by the
cohesive causes, since they are teleologically aimed processes that
external causes necessitate but do not cause in their own right.
From this analysis we see that in his criticism of Aristotle Ga-
len does not merely challenge the validity of Aristotle’s argument
in Physics . but rejects at the outset the thesis that it establishes.
Galen’s contention that the resting part assumption is impossible
follows from his definition of the expression ‘moving primarily and
in itself’ as having an inner source of motion and not being moved
by any external thing, and from the assumption that only homoeo-
merous parts of animate bodies move in this way. An examination
of this definition and this assumption in the light of Galen’s extant
writings shows that they reflect his view that the natural activities
of the homoeomerous bodily parts are internally caused by their
cohesive causes. Placed in this context, Galen’s argument leads not
only to the conclusion that the thesis that anything that moves is
moved by something is not universally true but to the stronger con-
clusion that this thesis is universally false because homoeomerous
parts are the only things that move primarily and in themselves.
Further, this analysis brings to light two implicit assumptions that
underlie Galen’s criticism: () Aristotle holds that the inner active
cause—the form—is not distinct from the material constitution, and
() external causes are necessary conditions for alterations of acti-
vities, but they neither generate activities nor primarily cause their
alterations. The full significance of these assumptions for the de-
bate between Alexander and Galen becomes clear in the following
discussion of Alexander’s reply.
. Alexander’s reply
.. Hypothetical possibility and the validity of Aristotle’s argument
In addressing Galen’s criticism, Alexander interprets Aristotle’s
argument in the light of his general understanding of Physics .
Orna Harari
as containing more logical proofs (Simpl. In Phys. . –
Diels), and contends that Aristotle’s argument in Physics . is
not demonstrative but logical (a– Carullah; a Escorial).
He claims that Aristotle’s assumption ‘if something comes to rest
because something else ceases moving, it is necessarily moved
by something’ has one argumentative status when the thing that
comes to rest is external and another when this thing is internal. In
the former case the argument is demonstrative because the thing
that comes to rest is the cause of motion, and in the latter case it
is logical because the thing that comes to rest is not the cause of
motion (a–b and b–a Escorial). On the basis of this
interpretation, Alexander argues that the above assumption does
not explain the conclusion that anything that moves is moved by
something but necessarily entails it, since this assumption and the
conclusion are unique properties (ἴδια) of bodies, just as laughing
and being receptive of knowledge are unique properties of every
man (a Escorial). Accordingly, Alexander construes Aristotle’s
argument as a valid and sound syllogism in the first figure (b–
Carullah):
() Anything that moves primarily and in itself does not move
when something ceases moving.
() Anything that does not move when something ceases moving
is moved by something.
Rescher and Marmura translate the words mantiq and mantiqiyya, which usu-
ally mean ‘logic’ and ‘logical’, by ‘dialectic’ and ‘dialectical’. Since the argument
that Alexander describes as mantiqiyya is not dialectical in the technical sense, I fol-
low Pines and use the word ‘logical’ (Pines, ‘Omne quod movetur’, ). According
to Silvia Fazzo, Simplicius’ report serves as one piece of evidence that casts doubt
on the view that Galen’s criticism of Aristotle’s argument in Physics . reflects his
theoretical disagreement with Aristotle (Fazzo, ‘Alexandre d’Aphrodise contre Ga-
lien’, –). However, this conclusion does not follow from Simplicius’ report on
Alexander’s exegetical approach. It only implies that in Physics Aristotle does not
present demonstrative arguments in support of his physical views. In the Refutation
of Galen Alexander concedes that Aristotle’s argument in Physics . is not demon-
strative.
Cf. Simpl. In Phys. . – Diels. In characterizing this argument as de-
monstrative Alexander understands ‘demonstrative’ as equivalent to ‘explanatory’
and disregards the other characteristics of demonstrative premisses, for instance that
they are essential predications. This approach is not uncommon in later antiquity.
For example, in his commentary on the first book of Euclid’s Elements Proclus ex-
amines the question whether Euclid’s proof that the sum of the interior angles of a
triangle is equal to two right angles is demonstrative by focusing solely on the ques-
tion whether it is explanatory (In Eucl. . – Friedlein).
Cf. Simpl. In Phys.. – Diels.
Alexander against Galen on Motion
Therefore:
() Anything that moves primarily and in itself is moved by
something.
The first premiss of this syllogism is a generalization of the con-
clusion of the inference that Galen criticizes on the grounds that
the resting part assumption is contradictory: that is, when CB is
at rest AB does not move primarily and in itself. Therefore, our
understanding of Alexander’s reply to Galen hinges on the answer
to the question how Alexander justifies his contention that this
premiss is true (b Carullah). A superficial examination sug-
gests that Alexander justifies this premiss by arguing that the rest-
ing part assumption is not hypothetically impossible (henceforth,
‘the hypothetical reconstruction’). This suggestion supports the lo-
gical interpretation of the debate between Alexander and Galen, in
implying that this debate is confined to the question whether the
resting part assumption is absolutely impossible (as Galen holds)
or not hypothetically impossible (as Alexander holds). However,
a careful examination of the hypothetical reconstruction casts doubt
on this interpretation, by showing that in his reply to Galen Alexan-
der uses two distinct arguments: the hypothetical reconstruction
that secures only the validity of the inference from the resting part
assumption to the conclusion ‘AB does not move primarily and in
itself’, and the syllogistic reconstruction that secures the validity
and soundness of Aristotle’s whole argument.
In his commentary on Physics . Simplicius describes the hypo-
thetical reconstruction in the following way:
But Alexander, though suspecting the demonstration on many grounds,
chose to say that it is not hypothetically impossible [οὐκ ἔστι πρὸς ὑπόθεσιν
ἀδύνατον] to assume that a part of a thing that moves in itself and primarily
is at rest, ‘for’, he says, ‘only things that are destructive of each other are
hypothetically impossible, for example, sailing through a rock’. (. –
Diels)
Alexander says: ‘This having been said, there is a need to enquire how the
hypothesis that assumes that a certain part in a primary self-mover is at
rest is possible. Someone can say that he [sc. Aristotle] did not make this
claim here but [the claim] that if a part CB were subtracted from the whole
self-mover AB and stood still, the whole would no longer be in motion,
even if the part that was left behind moved, for what was left behind is no
longer a whole, since a part was subtracted from it. But, since it neither is
nor remains a whole, it could not move as whole. The person who says this
would say that he [sc. Aristotle] used [an argument] very (πάνυ) logically
and dialectically but not naturally or demonstratively.’ (. – Diels)
For this reason, the body is not strictly speaking said to be moved by the
soul, for this is said in those cases where the mover and the moved are se-
parate, as in the case of oxen that move a wagon. But since also that which
is moved in virtue of itself is more generally [κοινότερον] said to be moved
by something (for in this way the craftsman [is moved] by art, i.e. in virtue
of it, and in this way fire [is moved] by lightness), the animal is said to be
The elements are not wholes according to Aristotle’s definition in Metaphysics
Δ , since the position of their parts makes no difference. Nevertheless, Alexan-
der applies this argument to their case, since in his view the elements are bodies
or composites of matter and form. On Aristotle’s and Alexander’s divergent views
on this subject see R. W. Sharples, ‘On Being a τόδε τι in Aristotle and Alexander’,
Méthexis, (), –. Further, my claim that Alexander presupposes Aristotle’s
account of self-motion accords with his account of the appetitive part of the soul.
Unlike Aristotle, Alexander holds that this part of the soul is not moved (DA .
– Bruns).
Alexander against Galen on Motion
moved by the soul in the same way too, i.e. in virtue of it. (. –.
Bruns)
Here Alexander clearly states that animate bodies and the elements
are not strictly speaking moved by the soul and the inclinations,
respectively, but in virtue of them. He does not explain the dif-
ference between these descriptions but his discussion offers two
clues. First, his claim that things that are moved in virtue of them-
selves can be more generally described as being moved by them-
selves suggests that, in his view, the soul and the inclinations are
in a sense active causes, otherwise the expression ‘being moved by’
would be altogether inapplicable to them, not merely inaccurate.
Secondly, Alexander’s claim that the expression ‘being moved by’
holds strictly speaking for things that are moved by separate movers
implies that the phrase ‘being moved in virtue of’ is applicable to
cases where the mover and the moved are the same in substrate.
This interpretation finds support in Alexander’s commentary on
Metaphysics Δ . Here he characterizes the ‘dunamis in virtue of
states’ (ἡ κατὰ τὰς ἕξεις δύναμις) as an internal causal factor, saying
that it is ‘similar to active dunamis, differing from it in being ac-
tive [ποιητική] not with respect to another thing but with respect
to the bearer itself’ (. – Hayduck). This view is central to
Alexander’s account of forms. In his De anima he states that forms
not only define bodies and differentiate them from other bodies but
also explain the ways in which they act and are acted upon (. –
Bruns). This account distinguishes, in Alexander’s view, the Aris-
totelian account of activity and passivity from the Stoic and Pla-
tonic accounts. In his commentary on Aristotle’s De sensu he says
that in Aristotle’s view neither bodies nor incorporeals act or are
acted upon, as the Stoics and the Platonists respectively hold, but
they do so in virtue of incorporeals (. – Wendland).
This discussion highlights the point of contention between
Alexander and Galen. It shows that Alexander’s claim that ani-
mate bodies and the elements are moved in virtue of the soul and
For a more detailed discussion of this passage see I. Kupreeva, ‘Qualities
and Bodies: Alexander against the Stoics’, Oxford Studies in Ancient Philosophy,
(), – at . For an analysis of Alexander’s causal understanding
of forms and for more references see Caston, On the Soul, –. In view of this
interpretation, Alexander’s claim in the Refutation of Galen that animate bodies and
the elements are moved by their soul and their inclination is not incompatible with
Alexander’s Greek writings as Silvia Fazzo argues (‘Alexandre d’Aphrodise contre
Galien’, –).
Orna Harari
inclination is similar to Galen’s account of bodily activities, since
in both accounts bodies have internal causal factors. But unlike
Galen, Alexander is not agnostic about the substance of this in-
ternal cause; he views it as an incorporeal form that supervenes
on the material constitution of bodies (De anima . –; . –;
. –; . –; . – Bruns). Through this view, then,
Alexander counters Galen’s criticism by arguing that inner movers
are ontologically distinct entities, therefore apparent self-movers
are moved by something. In so doing, he defends the thesis that
anything that moves is moved by something but also rejects Galen’s
interpretation of Aristotle that takes forms to be identical to the
mixture of the four qualities (QAM . , . – Mueller = iv.
K.). Understood in this light, the debate between Alexander
and Galen is restricted neither to the validity of Aristotle’s argu-
ment in Physics . nor to the thesis that it establishes. Rather,
hinging on whether the whole is similar to or different from its
parts, this debate also concerns the foundations of Aristotle’s phi-
losophy: the ontological distinction between form and matter.
An examination of Alexander’s brief discussion of elemental mo-
tion in the Refutation of Galen brings to light another theoretical
implication of his debate with Galen. Here Alexander explains his
claim that the elements are moved by their inclination through Aris-
totle’s account of elemental motion, claiming that the elements’ mo-
tion is caused also by an external thing, namely the light (i.e. air and
fire) or heavy (i.e. water and earth) elements in actuality, that turns
light or heavy elements in potentiality into heavy or light in actu-
ality (a– Carullah). This explanation reveals a significant
difference between Alexander’s and Galen’s accounts of external
causes. It shows that whereas in Alexander’s view the elements re-
Alexander explicitly rejects the view that the soul is a relative in Quaestio . .
His arguments are reasonably applicable to the forms of inanimate bodies too.
For an analysis of Alexander’s notion of supervenience see V. Caston, ‘Epi-
phenomenalism, Ancient and Modern’, Philosophical Review, . (), –
at , and On the Soul, –.
Richard Sorabji concludes from this passage that Alexander downplays the role
of the external cause because he cites it as evidence that the inclination is the cause
(R. Sorabji, The Philosophy of the Commentators – : A Sourcebook, ii. Phy-
sics (London, ), ; Matter, Space, and Motion: Theories in Antiquity and their
Sequel (Ithaca, NY, ), ). The conclusion does not necessarily follow from
this passage, since it may also imply that the inner cause derives its efficacy from the
external cause. I show below that the latter understanding accords with Alexander’s
general conception of causality.
Alexander against Galen on Motion
ceive their capacity to move downward or upward from the external
causes that make them heavy or light in actuality, in Galen’s view
external causes do not endow homoeomerous bodies with their in-
ner dunamis, and therefore they neither generate activities nor al-
ter them in their own right. This explanation reflects Alexander’s
general account of efficient causality. In his commentary on Meta-
physics Δ he explains Aristotle’s description of one of the senses of
the term ‘cause’ as the first principle of change or rest (a–)
as follows:
He adds ‘first’ because this is principally [μάλιστα] the efficient cause; for
that which is inherent in the generated thing and thus makes the sub-
sequent things has its cause from the thing that generated it. And even
the organs do not have their motion from themselves [ἐξ αὑτῶν], but also in
these the first cause is external to the generated thing. (. – Hayduck)
BI B L I O G R A P HY
Barnes, J., and Jouanna, J. (eds.), Galien et la philosophie: huit exposés suivis
de discussions (Vandœuvres, ).
Caston, V. (trans. and comm.), Alexander of Aphrodisias: On the Soul Part
Alexander against Galen on Motion
. Soul as Form of the Body, Parts of the Soul, Nourishment, and Percep-
tion [On the Soul] (London, ).
Caston, V., ‘Epiphenomenalism, Ancient and Modern’, Philosophical Re-
view, . (), –.
Donini, P., ‘L’anima e gli elementi nel de anima di Alessandro di Afrodi-
sia’, Atti della Accademia delle Scienze di Torino, classe di scienze morali,
storiche e filologiche, (), –.
Donini, P., ‘Psychology’, in Hankinson (ed.), The Cambridge Companion
to Galen, –.
Fazzo, S., ‘Alexandre d’Aphrodise contre Galien: la naissance d’une
légende’ [‘Alexandre d’Aphrodise contre Galien’], Philosophie antique:
problèmes, renaissances, usages, (), –.
Frede, M., ‘Galen’s Theology’, in Barnes and Jouanna (eds.), Galien et la
philosophie, –.
Furley, D., and Wilkie, J. S. (ed., trans., comm.), Galen: On Respiration
and the Arteries (Princeton, ).
Gregoric, P., and Kuhar, M., ‘Aristotle’s Physiology of Animal Motion:
On Neura and Muscles’, Apeiron, . (), –.
Hankinson, R. J., ‘Causation in Galen’, in Barnes and Jouanna (eds.),
Galien et la philosophie, –.
Hankinson, R. J. (ed. and comm.), Galen: On Antecedent Causes (Cam-
bridge, ).
Hankinson, R. J. (trans. and comm.), Galen: On the Therapeutic Method,
Books I and II [Therapeutic Method] (Oxford, ).
Hankinson, R. J., ‘Philosophy of Nature’, in id. (ed.), The Cambridge Com-
panion to Galen, –.
Hankinson, R. J. (ed.), The Cambridge Companion to Galen (Cambridge,
).
Kukkonen, T., ‘Alternatives to Alternatives: Approaches to Aristotle’s Ar-
guments per impossibile’ [Alternatives to Alternatives’], Vivarium, .
(), –.
Kupreeva, I., ‘Aristotelian Dynamics in the nd Century School Debates:
Galen and Alexander of Aphrodisias on Organic Powers and Move-
ments’, Bulletin of the Institute of Classical Studies, (), –.
Kupreeva, I., ‘Qualities and Bodies: Alexander against the Stoics’, Oxford
Studies in Ancient Philosophy, (), –.
Lloyd, G. E. R., ‘Scholarship, Authority, and Argument in Galen’s Quod
animi mores’, in P. Manuli and M. Vegetti (eds.), Le opere psicologiche di
Galeno: atti del terzo colloquio Galenico internazionale, Pavia, – set-
tembre (Naples, ), –.
Mueller, I., and Gould, J. (trans. and comm.), Alexander of Aphrodisias on
Aristotle’s Prior Analytics . – (Ithaca, NY, ).
Orna Harari
Pines, S., ‘Omne quod movetur necesse est ab aliquo moveri: A Refutation of
Galen by Alexander of Aphrodisias and the Theory of Motion’ [‘Omne
quod movetur’], Isis, (), –.
Rashed, M. (ed., trans., comm.), Alexandre d’Aphrodise: Commentaire
perdu à la Physique d’Aristote (Livres IV–VIII). Les Scholies byzantines
(Berlin and Boston, ).
Rescher, N., and Marmura, M. E. (ed., trans., comm.), The Refutation by
Alexander of Aphrodisias of Galen’s Treatise on the Theory of Motion [Re-
futation by Alexander] (Islamabad, ).
Ross, W. D., Aristotle’s Physics: A Revised Text with Introduction and Com-
mentary (Oxford, ).
Schiefsky, M., ‘Galen’s Teleology and Functional Explanation’, Oxford
Studies in Ancient Philosophy, (), –.
Sharples, R. W., ‘On Being a τόδε τι in Aristotle and Alexander’, Méthexis,
(), –.
Singer, P. N. (ed.), Galen: Psychological Writings (Cambridge, ).
Sorabji, R., Matter, Space, and Motion: Theories in Antiquity and their Se-
quel (Ithaca, NY, ).
Sorabji, R., The Philosophy of the Commentators – : A Sourcebook,
ii. Physics (London, ).
Wardy, R., The Chain of Change: A Study of Aristotle’s Physics VII (Cam-
bridge, ).
R A T I O N A L A S SENT
A ND S E L F - R E V E RSION:
A NE OP L A T O N I S T RESPONSE
T O T H E S T O ICS
URSULA COOPE
On the Stoic view, the two important elements in human action and
cognition are assent (sunkatathesis, adsensio) and impression (phan-
tasia, visum). Chrysippus described an impression as an affection
(pathos) in the soul. An impression can be theoretical or practical:
a theoretical impression presents things as being the case; a prac-
tical impression presents things as to be pursued or avoided. An im-
pression can also be either sensory or non-sensory. Though many
of your theoretical impressions arise from the senses, you can also
have non-sensory theoretical impressions, such as the impression
that a certain argument is valid, or the impression that + = .
Sextus Empiricus attributes to the Stoics the view that an impres-
sion is a passive affection (peisis tis), as opposed to assent, which is
an activity (energeia) (M. . ). As passive affections, your im-
pressions are not directly subject to your control. Of course, you
can deliberately act in such a way as to affect the contents of your
impressions. Simple ways to do this include turning your head in
a certain direction, opening your eyes, and perhaps rubbing them
or squinting. Nevertheless, though you may voluntarily turn your
head, open your eyes and rub them, once you have done all this, the
impression you experience is simply a result of your being acted on
in a certain way.
Non-human animals (and young children) are simply led to ac-
tion by their impressions. Thus, if an animal has an impression of
something as to be pursued, it will automatically have an impulse
towards pursuing it and (provided nothing interferes) will act on
The Stoic account of this distinction is helpfully discussed by S. Bobzien, De-
terminism and Freedom in Stoic Philosophy [Determinism] (Oxford, ), ch. , R.
Sorabji, Animal Minds and Human Morals: The Origins of the Western Debate [Ani-
mal] (London, ), ch. , and T. Brennan, The Stoic Life: Emotions, Duties, and
Fate [Life] (Oxford, ), ch. .
Aët. Plac. . . = ps.-Plut. Plac. phil. = LS B = SVF ii. .
Sensory and non-sensory impressions are distinguished at D.L. .
(= LS A = SVF ii. ). See S.E. M. . – (= LS G = SVF ii. ).
As Sextus Empiricus remarks, reporting Stoic views, you can rub your eyes in
an attempt to make your visual impression clearer (M. . – = LS K). Diogenes
Laertius reports that according to the Stoics, you can also alter which impressions
you experience by undergoing a process of training in some craft. When looking at
a table, a skilled carpenter will have different impressions from someone who is un-
skilled. When hearing arguments, someone who is trained in syllogistic will have
different impressions from someone who is not. (D.L. . = LS A = SVF ii. .)
Rational Assent and Self-Reversion
this impulse. Adult humans differ from other animals in that they
do not respond in this automatic way to their impressions. They
have a rational faculty of assent. This is a faculty which allows them
to endorse, or to withhold endorsement from, their impressions.
In assenting to a theoretical impression, a human being makes a
judgement; in assenting to a practical impression, the human has
an impulse (and will act on this impulse unless something inter-
venes).
Assenting, according to the Stoics, is under your direct control.
Thus, we find in Cicero the report that ‘to these things that appear
and are, so to speak, received by the senses, he [Zeno] adds assent
of our minds, which he makes dependent on us and voluntary [in
nobis positam et voluntariam]’ (Cic. Acad. . = LS B = SVF i.
). Similarly, Clement reports that ‘not only the Stoics but also
the Platonists say that assent depends on us’ (is eph’ hēmin: Strom.
. . – = SVF ii. ). The fact that assent differs in this re-
spect from having an impression is explained by Sextus Empiricus
when he reports the Stoic view:
Now cognition [κατάληψις], as one may learn from them, is ‘assent to the
cognitive impression’ and this seems to be a twofold thing, one part of
which is involuntary [ἀκούσιον] and the other part of which is voluntary
and dependent on our judgement [ἑκούσιον καὶ ἐπὶ τῇ ἡμετέρᾳ κρίσει κείμε-
νον]. For the experience of an impression is unwilled [ἀβούλητον], and being
affected in this particular way (as, for instance, with a sense of whiteness
when a colour presents itself, or with a sense of sweetness when something
sweet is offered to his taste) does not depend on the person affected but
rather on the cause of the impression [οὐκ ἐπὶ τῷ πάσχοντι ἔκειτο ἀλλ᾿ ἐπὶ τῷ
φαντασιοῦντι], whereas the act of assenting to this affection lies in the power
of the person who receives the impression [τὸ δὲ συγκαταθέσθαι τούτῳ τῷ
According to some accounts, certain Stoic philosophers attributed assent to ani-
mals too (see Alex. Aphr. Quaest. . , . – Bruns, and De fato . –.
Bruns). Even if this correctly represents the views of some Stoics, it is clear that all
of them agreed that rational assent is impossible for animals and that, since animals
are carried along by their impressions, they do not have the power to withhold as-
sent. If they assent, they do so automatically and invariably. See Sorabji, Animal, ,
and B. Inwood, Ethics and Human Action in Early Stoicism [Action] (Oxford, ),
–.
Exactly what is endorsed is not very clear: is it the impression itself, or instead
a corresponding proposition? This question is helpfully discussed in Brennan, Life,
–.
That is, a ὑπόληψις, δόξα, or κατάληψις. See Bobzien, Determinism, –.
These passages are cited in J. Barnes, ‘“Belief is up to us”’ [‘Up to us’], Pro-
ceedings of the Aristotelian Society, (), –.
Ursula Coope
κινήματι ἔκειτο ἐπὶ τῷ παραδεχομένῳ τὴν φαντασίαν]. (S.E. M. . = SVF
ii. , trans. Bury, modified).
In this paragraph I follow Bobzien, Determinism, esp. – and – (see
also pp. – for her defence of the attribution of this view to Chrysippus).
For example, ‘Dio is walking today’ is possible just in case () being human
(that is, being the kind of thing Dio is) is compatible with walking (in a way that,
say, being a tree or a fish would not be) and () there is nothing external that prevents
Dio from walking today (for example, he is not chained down).
The question of whether this account of possibility provides the basis for a con-
vincing defence of Stoic compatibilism is beyond the scope of this paper. For some
doubts see T. Brennan, ‘Fate and Free Will in Stoicism’, Oxford Studies in Ancient
Philosophy, (), – at –.
See D.L. . (= LS A = SVF ii. , trans. LS): ‘Some impressions are ex-
pert [τεχνικαί] and others not: a work of art is viewed in one way by an expert and
differently by a non-expert.’
Ursula Coope
possible (in such cases) for you to experience different impressions
from those you in fact experience. Why, then, are you responsible
for your assent in a way that you are not responsible even for this
kind of expert impression? The Stoics might perhaps reply that
your character and dispositions are the main cause of your assent
but not the main cause of your impressions. But this reply seems to
depend upon the explanation we are seeking, instead of providing
us with this explanation. To explain the sense in which your char-
acter is the main cause of your assent but not of your impressions,
we would need already to have an explanation of the fact that you
are accountable for your assent in a way that you are not account-
able for your impressions.
The second question left unanswered by Chrysippus’ arguments
about character and modality concerns the difference between hu-
man beings and trainable animals. Some non-rational animals can
be trained. If non-rational animals lack the ability to assent, then
For instance, in a situation in which you, as a carpenter, are having an expert
impression of a table, it is possible for you to be having a non-expert impression. (You
are not prevented by external circumstances from having such an impression, and
having such an impression is clearly compatible with being human.)
There is, of course, an important difference between assenting and having im-
pressions. The nature of the faculty of assent ensures that external circumstances
never force us to give or withhold our assent: the power we exercise in assenting is
a two-way power (a power to assent or not) and whether we assent depends on our
character. (Bobzien makes this point: Determinism, –.) By contrast, our ability
to have impressions is not a two-way power, and we can be externally forced to have
certain impressions. This shows that it will always be possible, on encountering cer-
tain impressions, to assent otherwise than one does, but it will not always be possible,
on encountering external objects, to undergo impressions that are other than those
one in fact undergoes. That would explain why we do not always have power over
our impressions, but it does not explain why we lack such control in those cases in
which our impressions depend on our character and hence are not necessitated by
things that are external to us.
This point is made by Philoponus in his commentary on Aristotle’s De anima
(In DA . –. ; . –; . – Hayduck). (The authorship of this
commentary on De anima has been disputed. For a defence of the claim that it is in
fact written by Philoponus see P. Golitsis, ‘John Philoponus on the Third Book of
Aristotle’s De anima, Wrongly Attributed to Stephanus’, forthcoming in R. Sorabji
(ed.), Aristotle Reinterpreted: New Findings on Seven Hundred Years of the Ancient
Commentators on Aristotle (London, ).) Philoponus says that the horse that has
been trained to respond to the whip has a kind of non-rational assent, and he contrasts
this with rational assent, which (he says) requires being persuaded. Ps.-Simplicius
also takes up the question of how animals can be trained, given that they lack ra-
tionality. He says that some animals have a kind of rationality, in that they can be
habituated by other beings. This is not the same as human rationality: it does not, for
instance, include the ability to deliberate, and it does not enable the animal to train or
habituate itself. But such animals, in virtue of their ability to be trained by humans,
Rational Assent and Self-Reversion
the way in which training affects an animal must be by modifying
its dispositions to experience impressions. But then, just as your as-
sent depends on your character, so also which impressions a trained
animal has when faced with a certain set of external circumstances
will depend on something internal to it: the dispositions it has ac-
quired through training. Moreover, Chrysippus’ account of possi-
bility implies that in cases in which an animal could be trained, it
could have impressions other than those that it in fact has. Why,
then, is a human being responsible for her assenting in a way that
a trainable animal is not responsible for its impressions? Again, the
Stoics have a possible reply. The trained dispositions of an animal
are importantly different from the dispositions that make up a hu-
man character: the former are dispositions for having impressions,
the latter are (or at least include) dispositions to assent. But again
this reply presupposes an answer to our question about the differ-
ence between assenting and having impressions. To explain the sig-
nificance of the difference between these two kinds of disposition,
we need to know why assenting matters for responsibility in a way
that having impressions does not.
None of our Stoic sources explicitly answers this question about
why we are responsible for our assent in a way we are not responsible
for our impressions. But an answer can, I think, be inferred from
what these sources say about the nature of assenting and about how
assenting differs from merely having an impression.
On the Stoic view, dispositions to assent are not mere tendencies to
go along with certain impressions. To assent is to make an evaluative
judgement. When we assent (or withhold assent), we use concepts
such as ‘true’ or ‘false’, ‘appropriate’ or ‘inappropriate’ to assess
our impressions, and we reject or accept these impressions in the
light of this assessment. This evaluative judging is what the Stoics
nitive impression, does not have knowledge but only opinion (δόξα), opinion being
a weak assent. See S.E. M. . – (= LS C); Cic. Acad. . – (= LS B =
SVF i. ).
You can, of course, act for a reason in such a way as to bring it about that you
have certain impressions, but in that case it is your action rather than the impression
itself that is based on a reason. What makes it possible for your action to be based
on a reason is just that the assent that prompted it is based on a reason.
This is not to claim that whenever you assent you assent for reasons. The point
is that the strength or weakness of any act of assenting (whether or not it is an act
of assenting for a reason) depends on certain counterfactuals about how the agent
would respond to arguments purporting to present reasons for withholding assent.
Rational Assent and Self-Reversion
refuse to withdraw your assent because you think there is no good
reason to withdraw your assent.
This is, of course, a rather minimal type of control. As we have
seen, for those of us who are not sages this very capacity to assent for
reasons can itself lead to error. One might wonder, then, whether
this capacity can really be what explains our accountability for our
beliefs and our actions. After all, it is possible to have this capacity
but be unable to exercise it well. Consider, for instance, someone
who has been indoctrinated in such a way that she reasons badly:
she assents and withholds assent on the basis of reasons, but she
is able to recognize the force of only a very limited range of rea-
sons. Must we really conclude that such a person is blameworthy?
The Stoics, I suspect, would have been untroubled by this conclu-
sion. But for those modern readers who are squeamish about such a
readiness to assign blame, it is worth noting that the Stoic view that
we are accountable for our assent does not by itself imply that every
mistaken act of assent is blameworthy. It implies only that assent-
ing is the kind of activity that is appropriately praised, blamed, or
excused. Whether any particular mistaken act of assent should be
blamed or excused will depend on further factors.
I have argued that, on the Stoic view, we are responsible for
assenting (and withholding assent) just because assenting (and
withholding assent) can be the exercise of a certain kind of rational
control. This view depends upon the fact that the capacity for
assenting is fundamentally different in kind from the capacity for
having impressions: in exercising our capacity for assent we can
If this is right, then the Stoic view finds an echo in the writings of certain mo-
dern philosophers. For example, Richard Moran claims that we are active in rela-
tion to our judgement-sensitive attitudes (such as beliefs and certain desires), and
he contrasts this with our passivity in relation to our sensations (Authority and Es-
trangement: An Essay on Self-Knowledge (Princeton, ), –). See also Gary
Watson’s claim that having ‘doxastic control’ amounts to having one’s beliefs de-
termined by ‘belief-relevant norms’: ‘our cognitive lives would be out of control to
the extent we were incapable of responding to the norms of coherence and relevant
evidence’ (G. Watson, ‘The Work of the Will’, in id., Agency and Answerability: Se-
lected Essays (Oxford, ) – at ).
Angela Smith makes just this point in defending her account of responsibility,
on which we are responsible for those attitudes and reactions that reflect our evalu-
ative judgements (‘Responsibility for Attitudes: Activity and Passivity in Mental
Life’, Ethics, . (), – at ). This is, she says, giving an account of
‘the conditions under which something can be attributed to a person in the way that
is required in order for it to be a basis for moral appraisal of that person. Merely
claiming that a person is responsible for something . . . does not by itself settle the
question of what appraisal, if any, should be made of the person on the basis of it.’
Ursula Coope
be guided by reasons for assenting, but in exercising our capacity
for having impressions we cannot be guided in this way. However,
if this is the Stoic account, it invites certain further questions.
What exactly does this difference between assenting and having
impressions amount to? What is φ-ing for a reason, and why can ‘φ-
ing’ here stand for ‘assenting’ but not for ‘having an impression’?
You can, after all, have new impressions as a result of listening to
arguments. Why does that not count as being persuaded to have those
impressions or as having those impressions for a reason? Again, what
is it for the sage’s assenting to be guided by knowledge of when one
should assent, and why can his impressions not also be guided by
such knowledge? More generally, what is it about the nature of our
capacities for assenting and for having impressions that explains
these differences between them?
In the remainder of this paper I shall look at a particular way of
explaining these differences between our capacity for having im-
pressions and our capacity for assent. A hint of this explanation can
already be found in the writings of the late Stoic Epictetus, though
I shall argue that this hint is only fully developed in the Neoplato-
nist ps.-Simplicius’ commentary on Aristotle’s De anima. Accord-
ing to ps.-Simplicius’ account, our power of assent is special in that
it can be exercised in a distinctively self-reflexive way. I shall not
attempt a full defence of the view that emerges from this account,
but I shall argue that it does suggest an interesting answer to our
questions about the distinctive nature of our capacity for assent.
Like the earlier Stoics, Epictetus holds that assenting is under our
control in a way that experiencing impressions is not. Impressions
‘are neither voluntary nor subject to one’s control [non voluntatis
sunt neque arbitrariae], but through a certain power of their own
they force their recognition upon men’; whereas the assents by
which these impressions are recognized are ‘voluntary and occur
subject to human control [voluntariae sunt fiuntque hominum arbi-
tratu]’.
However, Epictetus differs from the earlier Stoics in that his
notion of what depends on us (what is eph’ hēmin) is narrower than
From Gell. . . – (Gellius’ report of the fifth book of Epictetus’ Disser-
tationes), trans. Rolfe, modified.
Rational Assent and Self-Reversion
theirs. For Epictetus, something only counts as depending on the
agent if nothing external to the agent could interfere with it. Because
of this, he denies that the physical actions prompted by our assent
depend on us. Instead, what depends on us is just the ‘use of our
impressions’ (that is, assenting, withholding assent, and having the
beliefs and impulses that follow upon assent). This is the only ac-
tivity that cannot be interfered with or prevented. This difference
between Epictetus and earlier Stoics partly arises from a difference
in their motivations for discussing the notion of what ‘depends on
us’. Epictetus’ primary concern in these discussions tends not to be
allocating responsibility for our acts, but rather providing advice
about how to avoid frustration and disappointment. If you only
desire things that depend on you (in Epictetus’ narrow sense), then
your desires will never be frustrated by external circumstances (En-
chiridion, ch. ). As he says, no one can prevent you from assenting to
what is true or force you to assent to what is false (Diss. . . –).
There is, then, an easy answer to the question of why in Epicte-
tus’ sense of ‘depends upon us’ assenting depends upon us whereas
having impressions does not. As we have seen, the nature of assent-
ing ensures that external circumstances cannot force you to assent:
the main cause of assenting or withholding assent is always your
character. By contrast, there is nothing about the nature of having
For a helpful discussion of Epictetus’ views about what is involved in correct
use of impressions see A. A. Long, ‘Representation and the Self in Stoicism’, in S.
Everson (ed.), Psychology, Companions to Ancient Thought, (Cambridge, ),
–.
Thus, at Ench. . – Epictetus gives the following list of things that depend on
us and hence are our deeds (ἔργα): ‘supposition, impulse, desire, inclination, aver-
sion’. At Diss. . . (= LS K) he says that ‘the one thing which the gods have
made dependent on us is the one of supreme importance, the correct use of impres-
sions. The other things they have not made dependent on us’ (trans. after LS). At
Diss. . . – he asks ‘What is your own?’ and answers ‘The use of impressions’.
This is your own, he says, because no one can interfere with it. See also Diss. .
. –. I take passages such as these to support the interpretation I give here (see
also Bobzien, Determinism, –). For a contrary interpretation, on which Epicte-
tus’ view is much closer to that of the older Stoics, see R. Salles, ‘Epictetus and the
Causal Conception of Moral Responsibility and What is eph’ hêmin’ [‘Epictetus’],
in P. Destrée, R. Salles, and M. Zingano (eds.), What is up to us? Studies in Agency
and Responsibility in Ancient Philosophy (Sankt Augustin, ), –. On Salles’s
interpretation, Epictetus would face just the same question about assent that I raised
for the earlier Stoics (since if Salles is right, Epictetus cannot help himself to the
‘easy answer’ I suggest for him below).
This is not to deny that Epictetus was also interested in questions of responsibi-
lity, as Salles emphasizes (‘Epictetus’, –). Salles cites, in particular, Diss. . .
–. See n. above.
Ursula Coope
an impression that rules out the possibility that certain impressions
are forced upon you by external circumstances. If what depends
upon us must be the kind of thing that cannot be interfered with by
external circumstances, then whether or not we assent will depend
upon us, but which impressions we have will not.
If this were all Epictetus had to say about the way in which as-
senting depends on us, then his account would not help to answer
the questions I raised for the earlier Stoics. It would not, for in-
stance, help to explain why assenting, unlike having an impression,
is the kind of thing one can do for a reason or in the light of know-
ledge of whether one should assent. However, Epictetus also has more
to say about the way in which our power for assenting differs from
our power for having impressions. Our power for assent, he says, is
uniquely capable of a certain kind of self-reflexive exercise.
Epictetus’ Dissertationes begin with a chapter on what ‘depends
upon us’. He claims that reason, which he characterizes as ‘the
power of making correct use of impressions’ (that is, of correctly
assenting to or withholding assent from impressions), is the only
faculty the exercise of which wholly depends upon us. What is spe-
cial about reason, he explains, is that it is the only power that can
be exercised self-reflexively: no other faculty can contemplate itself,
and therefore no other faculty can approve or disapprove of itself;
reason ‘alone takes thought both for itself [καὶ αὑτὴν κατανοήσουσα
παρείληπται] (what it is, what it can do, and how valuable it has come
to be) and also for other faculties’ (Diss. . . –). We need to ask
why it is relevant to make this point about reason’s being uniquely
self-reflexive in a chapter about what depends upon us. Does this
fact about reason add anything to Epictetus’ other claims about the
sense in which exercises of reason (in assenting or withholding as-
sent) are uniquely in our power?
Epictetus defends his claim that reason is uniquely self-reflexive
in Dissertationes . . Other powers, he says, are directed at
things that are distinct from (and differ in kind from) themselves.
τῶν ἄλλων δυνάμεων οὐδεμίαν εὑρήσετε αὐτὴν αὑτῆς θεωρητικήν, οὐ τοίνυν οὐδὲ δο-
κιμαστικὴν ἢ ἀποδοκιμαστικήν (Diss. . . ).
For this view see also Marc. Aur. Med. . . : ‘These are the properties of the
rational soul: it sees itself, analyses itself, and makes itself such as it chooses’ (τὰ ἴδια
τῆς λογικῆς ψυχῆς· ἑαυτὴν ὁρᾷ, ἑαυτὴν διαρθροῖ, ἑαυτὴν ὁποίαν ἂν βούληται ποιεῖ; trans.
G. Long).
For discussion of his argument here see A. A. Long, Epictetus: A Stoic and Soc-
ratic Guide to Life [Epictetus] (Oxford, ), –.
Rational Assent and Self-Reversion
For instance, the art of shoemaking has to do with leather, but
the art itself is not composed of leather. The art of grammar has
to do with written speech, but the art itself is not made up of
written speech. Because of this, neither the art of shoemaking nor
the art of grammar can operate upon itself. Reason, on the other
hand, is in a certain sense ‘of the same kind as’ (homoeidēs with) its
subject-matter. The task of reason is the proper use of impressions,
but reason is itself ‘a system from certain impressions’ (σύστημα ἐκ
ποιῶν φαντασιῶν). Because of this, Epictetus says, reason is ‘con-
templative of itself’ (αὑτοῦ θεωρητικός), as well as of other things
(Diss. . . ).
It is not entirely clear what Epictetus means by calling reason ‘a
system from certain impressions’, or how this relates to the older,
Chrysippean claim that reason is a collection of concepts. He
could be using ‘impression’ (phantasia) broadly here, in such a
way that concepts (ennoiai) also count as impressions of a kind.
Alternatively, the ‘from’ (ek) could have a temporal rather than a
compositional sense, and his claim could be that reason is a system
of concepts that is developed out of certain impressions. Either way,
the conclusion he wants to draw from this claim about the relation
between impressions and reason is that any faculty that can assess
the impressions must also be able to assess reason. Thus, the faculty
of reason, since it is what assesses the impressions, must also be
capable of assessing itself. Epictetus goes on to make a similar point
about wisdom (phronēsis). Wisdom contemplates what is good and
bad, but since it itself is good, in contemplating what is good and
bad it must also contemplate itself and its opposite (Diss. . . ).
It might be objected here that the rules of grammar can be applied to the state-
ments that articulate the rules of grammar. For example (as an anonymous reader
pointed out to me), the statement ‘a verb always agrees with its subject’ is itself in
accord with the very rule of grammar it is articulating. When he denies that gram-
mar operates on itself, Epictetus must be distinguishing between grammar itself and
the statements by which the rules of grammar are articulated in language. On his
view, the fact that the rules of grammar can be used to judge such statements does
not show that grammar can ‘operate on itself’.
Galen tells us that Chrysippus described reason (λόγος) as ‘a collection of con-
cepts and preconceptions [ἔννοιαι and προλήψεις]’ (PHP . . , p. M. = SVF
ii. ).
Long suggests this in Epictetus, n. , and cites Plut. Comm. not. –
(= LS F = SVF ii. ), where concepts are classified as ‘a kind of impres-
sion’.
This argument that reason, unlike other arts, is self-reflexive has its roots in
Plato. See Charm. – ; Euthyd. – . Cicero also takes up this ar-
Ursula Coope
We are now in a position to understand the significance of this
claim that reason can evaluate itself. Epictetus’ point seems to be
that you can always reflect upon your own assenting (and withhold-
ing of assent) and judge whether you are right to be assenting (or to
be withholding assent). The claim is that you can evaluate each act
of assent (or of withholding assent), not that you actually do evalu-
ate every such act. If Epictetus held that you must evaluate every
act of assent, he would be committed to an infinite regress of acts of
evaluation (since each act of evaluation itself involves a distinct act
of assenting or of withholding assent). Instead, Epictetus says that
each act of assent can be reassessed and endorsed or rejected.
This is only possible, he argues, because reason (the faculty by
which we assent or withhold assent) is capable of assessing its own
operation. Reason has to be self-reflexive in this way if an infinite
regress of powers is to be avoided. If every act of assent (or with-
holding assent) is evaluable, then either there is one power which
enables us both to assent and to evaluate that act of assent, or our
capacity for evaluating that assent depends on the possession of a
further power. If the capacity for evaluating the original assent de-
pended upon the possession of a further power, that further power
would also have to be a power for assenting (since to evaluate some-
thing is to exercise a power of assenting or withholding assent), so
the operation of that further power would also have to be evaluable
(given the assumption that every act of assenting or withholding as-
sent is evaluable). The operation of this further power would, then,
need to be evaluated either by itself (in which case it would be a
power capable of evaluating its own operation) or by yet a further
power, and so on (Diss. . . –). Epictetus concludes that if
gument, claiming, at Fin. . – (= LS H = SVF iii. ), that ‘wisdom is wholly
directed towards itself, which is not the case with other arts’. In Acad. . Cicero
spells out the sense in which reason ‘judges itself’: reason enables the philosopher to
judge ‘which conjunctions and disjunctions are true, which statements are ambigu-
ous; what follows from something and what is incompatible with it’ (trans. C. Brit-
tain, Cicero: On Academic Scepticism (Indianapolis and Cambridge, )), and in
doing so, reason judges itself.
Of course, this implies that the series of assents is potentially infinite: for any
assent, there can be a further assent that endorses it. But Epictetus takes a potential
infinite of this kind to be unproblematic, so long as it does not imply an actually
infinite series of powers for assent.
Epictetus argues first that what evaluates reason must itself be a kind of reason-
ing power, and then that if no reasoning power were capable of evaluating itself there
would have to be infinitely many such reasoning powers. His argument seems vul-
Rational Assent and Self-Reversion
every act of assent (or withholding assent) is evaluable, there must
be a single power that is capable of evaluating its own use.
These remarks about the self-reflexivity of reason suggest a new
answer to our question about the difference between assenting and
having an impression. Epictetus—I take it—is claiming not merely
that the power assenting happens also to be uniquely self-reflexive,
but rather that there is an essential connection between its being
self-reflexive in this way and its being a power for assenting: an
activity counts as assenting only if it is an exercise of such a self-
reflexive power. If this is right, then assenting differs from having
an impression in the following way. Assenting (or withholding as-
sent) implies having the ability to evaluate one’s act of assent (or of
withholding assent), whereas having an impression does not imply
having the ability to evaluate one’s impression (or indeed, having
the ability to form any kind of meta-attitude towards it).
As we have seen, Epictetus makes this point about the self-
reflexivity of reason right at the start of a chapter on ‘what depends
on us’. We can now see why he thinks the two topics are connected.
The rational power of assenting, because of its self-reflexive nature,
provides you with the ability to re-evaluate any of your acts of as-
senting (or of withholding assent). You can reflect on any of your
beliefs and ask ‘is it really true?’ Or you can reflect on anything
you are about to do and ask ‘is this really an appropriate thing
to be doing?’ Having this ability for re-evaluation is a necessary
condition for exercising a certain kind of control over what you
do or believe: you need to be able to re-evaluate your beliefs and
intentions in order to have the capacity to revise (or reaffirm) those
beliefs and intentions in the light of your thoughts about whether
or not they are justified. A creature that altogether lacked this
capacity for re-evaluation would just be stuck with whatever beliefs
or intentions it had formed, and would thus lack an important
kind of control over what it thought or did. A creature that only
had the power to re-evaluate a certain limited range of acts of
assent would be correspondingly limited in its ability to revise its
beliefs and intentions. In a human being, no act of assenting or
withholding assent is ‘off limits’ for re-evaluation, because reason
nerable to the objection that the regress can be avoided if someone possesses two rea-
soning powers, each of which is used to evaluate the operation of the other. Perhaps
Epictetus thinks that this would really just amount to having a single self-evaluating
reasoning power, made up of those ‘two’ powers.
Ursula Coope
(the capacity for assenting or withholding assent) is a self-reflexive
power.
Can we, then, use Epictetus’ account to answer our earlier ques-
tions about the way in which assenting differs from having impres-
sions? The fact that you can always evaluate any of your assents
does help to explain why your assent depends on you in a way that
an animal’s impressions do not depend on it. Being able to evaluate
your assent is a necessary condition for being able to revise or main-
tain that assent in the light of this evaluation, and this is a necessary
condition that animals lack. On the other hand, as it is a mere ne-
cessary condition, the fact that you can evaluate your assent does
not explain how it is possible for you to revise your assent in the
light of this evaluation. Hence, it does not explain why a human
being can assent (or withdraw her assent) for a reason but cannot
have impressions (or cease to have impressions) for a reason. The
difference between human assent and human impressions cannot
lie in the fact that a human being has the capacity to evaluate each
of her assents; she can evaluate each of her impressions too. The
difference seems rather to lie in the fact that evaluating your assent
can itself be a way of revising that assent, whereas you cannot revise
your impressions simply by evaluating them. When you reflect on
your assent to P and judge that it was mistaken, you thereby cease to
assent to P, but when you judge that your impression that P is inac-
curate, you do not cease to have that impression. An account of the
difference between assenting and having impressions needs to ex-
plain why assent is such as to be responsive to higher-order evaluation,
and why assenting is, in this respect, unlike having an impression.
In what follows I turn to ps.-Simplicius’ commentary on Aris-
totle’s De anima for its presentation of a Neoplatonist view on
which reason is self-reflexive in a more radical way. I shall ask
whether this view can explain the fact that assenting, unlike having
impressions, is such as to be responsive to higher-order evaluation.
There is, of course, a difference in the two kinds of evaluation. For Epictetus,
the power by which you evaluate your assent is the very power you exercise in assent-
ing, whereas the power you exercise in evaluating your impressions is not the power
you exercise in having an impression. But it is not at all obvious why this difference
should imply that assenting is in your control in a way that having an impression
is not.
References are to Hayduck’s edition in the Commentaria in Aristotelem Graeca
series (Berlin, ).
Rational Assent and Self-Reversion
. Ps.-Simplicius’ classification of
soul-power as rational or non-rational
In a later passage he repeats the claim that non-rational cognition cannot ‘ap-
prehend something as true’ and explains that though such cognition can be true, it
cannot ‘judge the very fact that it is true’ (ἀληθὴς μὲν ὑπάρχουσα, οὐκ αὐτὸ δὲ τοῦτο
κρίνουσα ὅτι ἀληθής, . –).
Interestingly, ps.-Simplicius makes an analogous point about the distinction
between rational and non-rational desire. Just as taking something to be true re-
quires a rational cognitive power, so also desiring something as good (as opposed
to as merely pleasant) requires a rational desiderative power. In both cases a kind
of self-reflexivity is required: ‘in the simultaneous perception of something as being
good or true there is necessarily included also the subject’s being benefited or itself
saying what is true’ (. –). Ps.-Simplicius does not explain how to understand
this self-reflexivity in the case of rational desire. If it is to be analogous to what he
says about cognition, then he should say that in desiring something as good, one
endorses the power one exercises in so desiring, as being in this instance rightly
exercised—but that is quite a lot to read into the remark that in desiring something
as good one must grasp oneself ‘as being benefited’.
As should be clear from what I have already said, ps.-Simplicius is here spelling
out what is involved in ‘telling truths’ (ἀληθεύειν), not merely what is involved in a
certain special kind of truth-telling (the kind that involves grasping of oneself that
one is telling truths). His view is that if something is to tell truths, it must grasp of
itself that it is telling truths.
Ursula Coope
in accordance with the reversion of belief to itself, judging as true
its own understanding of things’ (κατὰ τὴν τῆς δόξης πρὸς ἑαυτὴν ἐπι-
τελουμένη ἐπιστροφήν, κρινούσης ὡς ἀληθῆ τὴν ἑαυτῆς περὶ τῶν πραγ-
μάτων σύνεσιν, . –). Of course, judgements can be false, so
one can be mistaken in taking one’s understanding to be true, but
whether or not one is mistaken, ps.-Simplicius holds that one must
take one’s understanding to be true in order to count as assenting.
He says:
ἡ δὲ διάνοια κἂν ἀληθῶς κἂν ψευδῶς προσβάλλῃ τοῖς γνωστοῖς, οὐκ ἀρκεῖται τῇ
προσβολῇ οὐδὲ συγκατατίθεται, εἰ μὴ καὶ ὅτι ἀληθὴς ἡ σύνεσις ἐπικρίνῃ ἢ ὀρθῶς
ἢ οὔ, ποτὲ καὶ περὶ τὴν ἐπίκρισιν ἁμαρτάνουσα. (. –)
Discursive reason [dianoia], whether it reaches out to the objects of cogni-
tion correctly or falsely, is not satisfied with the reaching out and does not
give its assent unless it also judges, correctly or otherwise, that its under-
standing is true (and sometimes it also makes a mistake about this judge-
ment).
Ps.-Simplicius’ claim here is that when one assents (that is, takes
something to be true), one’s belief (doxa) or discursive reason (dia-
noia) must take its own understanding (sunesis) to be true. To work
out what he means by this, we need to answer two distinct ques-
tions: () what is meant by ‘taking understanding to be true’? and
() what is it for belief (or discursive reason) to take its own under-
standing to be true?
Ps.-Simplicius uses the word ‘understanding’ (sunesis) in a very
broad sense. Both rational and non-rational cognition can be an
I translate δοξάζειν as ‘believe’ and κρίνειν as ‘judge’, but I do not think that
much hangs on the distinction here. The Greek word δοξάζειν can also be translated
‘judge’. The English word ‘believes’ tends to be used to attribute a dispositional state
(such that it can be true that x believes P, even while x is asleep), whereas ‘x judges
that P ’ tends to be used of something that one does at a particular time. The relation
between judging and believing is not straightforward. Often, making a judgement is
forming a belief, but judging can also be a kind of activation and reaffirmation of a
belief one already has, and it is possible to come to have a belief that P without there
being any time at which one considered whether P and made a judgement. I shall
argue that according to ps.-Simplicius, one only counts as believing/judging that P
if one endorses the understanding in virtue of which one believes/judges that P. It
does not matter whether we take this claim to be about the disposition of believing or
about the activity of judging, so long as we understand the notion of ‘endorsement’
accordingly: if the claim is about believing that P, then the relevant kind of endorse-
ment is believing that our understanding is correct; if the claim is about judging that
P, then the relevant kind of endorsement is judging that our understanding is cor-
rect. In what follows I shall spell out the claim as a claim about judging that P.
For this reason, ‘understanding’ is a slightly misleading translation, more ap-
Rational Assent and Self-Reversion
exercise of sunesis. For example, at . – he implies that there
is a perceptual kind of sunesis that is exercised in simply perceiv-
ing something (for example, perceiving that there is a man), and he
distinguishes this from a second, rational kind of sunesis that is ex-
ercised in making a judgement on the basis of one’s perception (for
example, judging that there is in fact a man).
Ps.-Simplicius does not spell out what it is for a kind of sunesis to
be true, so here we have to speculate. What does the claim that your
sunesis is true add to the claim that your judgement or perception is
true? His view, I suggest, is as follows. To claim that your judge-
ment is true is to claim that things in fact are as you judge them to
be; similarly, to claim that your perception is true is to claim that
things in fact are as you perceive them to be. To claim that your
(perceptual or rational) sunesis is true is to claim something more
than this: it is to claim that you are achieving truth non-accidentally,
in virtue of correctly exercising your sunesis. Your power of sunesis
is true on a particular occasion if and only if it is achieving truth in
virtue of functioning well as the kind of sunesis it is.
Your perceptual sunesis is functioning correctly just in case your
perceptual powers are functioning well: you are not hallucinating,
your vision is not distorted, and so on. When you make a judgement
on the basis of perceptual sunesis, you judge that your perceptual
sunesis is achieving truth in virtue of its correct functioning. Ps.-
Simplicius gives an example in which you hesitate to make such a
judgement because you are unsure whether what your perceptual
sunesis suggests to you is in fact true. You might see that there is
a man in the distance, but nevertheless hold back from endorsing
your perceptual sunesis and hence from judging that there is a man
until you have confirmed that the thing in question is ‘two-footed
and moving and upright’ (. ). In this case, you initially hold
back from endorsing your perceptual sunesis because you know that
(even when your perception is functioning correctly) the visual ap-
propriate for rational than for perceptual σύνεσις. In what follows I shall mostly just
use the transliterated Greek word sunesis in the main text.
At . – he says that the σύνεσις of a thing is not the same as the σύνεσις of
the fact that that σύνεσις is true (οὐ γὰρ ταὐτὸν ἡ περὶ ὁτουοῦν σύνεσις καὶ ἡ περὶ τοῦ
ὅτι καὶ ἀληθὴς ἡ σύνεσις). In this context he is distinguishing between perceptual and
rational σύνεσις. His point is that the σύνεσις we exercise in simply perceiving a thing
is not the same as the σύνεσις we exercise in judging this perceptual σύνεσις to be true
and hence in making a judgement that something is the case.
As ps.-Simplicius says, truth consists in agreement with the facts (. –).
Ursula Coope
pearance of things that are far away can be misleading: how things
appear in the distance is not always a good guide to how things are.
There is also, I suggest, another way in which you might fail to
endorse your perceptual sunesis. Suppose you suffer from a hallu-
cination of a man, you know you are hallucinating, and yet you have
independent grounds for judging that there is in fact a man there
(perhaps even a man who looks just like what appears to you in your
hallucinatory experience). In this case, what your visual appearance
suggest to you (namely that there is a man) is in fact true, and you
do judge (on independent grounds) that there is a man there, but
you do not judge that your perceptual sunesis is true. You realize
that your visual sunesis is not achieving truth in virtue of function-
ing well as the kind of sunesis it is. Truth here is not an achievement
of your visual sunesis. As a result, though you judge that there is
a man, your judgement is not based on how things visually appear
to you.
To assent, or make a judgement, is to exercise rational sunesis (al-
though, as we have seen, when you make a judgement on the basis of
perception, you also exercise your perceptual sunesis). To take your
rational sunesis to be true is to take yourself to be achieving truth in
virtue of the correct functioning of your capacity for assent.
Ps.-Simplicius does not himself spell out what it is for this capa-
city to be functioning correctly, but we can reconstruct his view on
the basis of certain remarks he makes about the relation between as-
sent and reasoning. In discussing assent, he distinguishes between
two different kinds of cognition. You might fully grasp something
with your intellect, knowing it in an ‘undivided and unitary way’,
or alternatively, you might lack this intellectual grasp, but instead
have the kind of cognition that depends on reasoning (. –).
In the first case your assent need not depend upon reasoning. In
the second case your assent (or withholding assent) should depend
on whether or not there is reason to assent: whether or not you
assent should depend ‘entirely on reason [logos] (whether neces-
sary or convincing), or on the speaker’s being a trustworthy person,
judging this too in accordance with some reason [logos]’ (. –
). These remarks suggest that whether your capacity for assent
A necessary λόγος for a certain conclusion proves that conclusion to be true;
a convincing (πιθανὸς) λόγος gives good reason to believe the conclusion. Here (at
. –) ps.-Simplicius uses the term ‘rational cognition’ (λογικὴ γνῶσις) to refer
narrowly just to this kind of cognition that depends on λόγος (as opposed to the
Rational Assent and Self-Reversion
is functioning correctly depends upon your warrant for assenting. If
your assent is not based on reasoning, then your capacity for assent
is functioning correctly just in case you have the kind of unitary
knowledge that warrants giving this kind of assent. On the other
hand, if your assent is based on reasoning, then your capacity for as-
sent will be functioning correctly just in case the reasons on which
it depends do in fact provide good grounds for assenting. When you
take the sunesis you exercise in assenting to be true, you take yourself
to be achieving truth in virtue of the fact that your sunesis warrants
you in assenting as you do. Once again, then, taking your sunesis
to be true is a matter of taking yourself to be achieving truth non-
accidentally, in virtue of exercising that sunesis. A judgement based
on bad reasoning might still be true, but the sunesis exercised in
making such a judgement would not be true.
This explains what it is to judge one’s sunesis to be true, but it does
not yet explain the peculiar self-reflexivity ps.-Simplicius attributes
to the act of judging. To understand this, we need to answer the se-
cond question I raised for his account. We need to explain what he
means when he says that rational cognition takes its own sunesis to be
true: that is, when he says that belief (doxa) judges as true ‘its own
understanding about things’ (τὴν ἑαυτῆς περὶ τῶν πραγμάτων σύνε-
σιν, . ), or that rational cognition ‘comes about by reverting to
itself and judging its own understanding’ (εἰς ἑαυτὴν ἐπιστρέφουσα
καὶ τὴν ἑαυτῆς κρίνουσα σύνεσιν γίνεται, . –).
intellectual cognition of νοῦς). However, I shall continue (as ps.-Simplicius does
elsewhere) to refer to both this kind of cognition and the kind that is an exercise
of νοῦς as ‘rational’, and to treat both as exercises of different types of rational
σύνεσις.
When you judge that P on the basis of seeing that P, you endorse not only your
perceptual σύνεσις, but also the rational σύνεσις you exercise in endorsing your per-
ceptual σύνεσις, and also the rational σύνεσις you exercise in assenting to P on this
basis. You judge that how things then visually appear to you is a good guide to how
they are (that is, you endorse your perceptual σύνεσις). You take yourself to be war-
ranted in making this judgement about your perceptual σύνεσις (that is, you endorse
the σύνεσις you exercise in assenting to ‘how things now appear to me is a good guide
to how they are’). And you also take yourself to be warranted in assenting to P on
this basis (that is, you endorse the σύνεσις you exercise in assenting to P). An infinite
regress might seem to threaten here. Do you also need to judge that you are warran-
ted in taking yourself to be warranted in endorsing your perceptual σύνεσις, and if so, do
you then also need to judge that you are warranted in making that judgement, and
so on? (Similarly, do you need to judge that you are warranted in taking yourself to
be warranted in assenting to P, and that you are warranted in making that judgement,
and so on?) As we shall see, ps.-Simplicius’ account of the way in which rational
σύνεσις is self-reflexive gives him an answer to this worry.
Ursula Coope
When you judge whether your perceptual sunesis is true, you
use one power (your rational power of assent) to evaluate another
(your perceptual sunesis). When you judge whether a belief you
held ten years ago was warranted, you are exercising your power of
assent to evaluate another exercise of that very power. According
to ps.-Simplicius, the self-reflexivity that is essential to judging is
of a stronger kind: judging is self-reflexive in that it involves one
cognitive act evaluating itself. His view is that any act of judgement
must contain, as part of its content, an endorsement of the sunesis
being exercised in that very act of judgement.
It is perhaps easiest to explain this with an example. Ps.-
Simplicius is claiming that you cannot judge, say, that there is a
blackbird in the garden without, as part of this act of judgement,
judging that you are achieving truth in virtue of using your
understanding (sunesis) correctly in making this judgement. The
content of your judgement will thus be ‘there is a blackbird in
the garden and the understanding I am exercising in making this
judgement is true’. ‘This judgement’ here must refer to the whole
judgement (that is, to the judgement that has as its content the
whole of ‘there is a blackbird in the garden and the understanding
I am exercising in making this judgement is true’). In this sense,
then, your act of judgement is necessarily self-referential: it has as
part of its content something that refers to itself.
As we have already seen, taking this understanding (sunesis) to
be true is taking yourself to be achieving truth in virtue of being
warranted in making the judgement. When you judge ‘there is a
blackbird in the garden and the understanding I am exercising in
making this judgement is true’, you are judging that your under-
standing warrants the whole judgement: ‘there is a blackbird in the
garden and the understanding I am exercising in making this judge-
Interestingly, this view has something in common with a self-referential account
of belief that has been defended by Gilbert Harman. Harman argues that to accept
a conclusion h, as known, one must implicitly accept the self-referential claim ‘h and
there is no actually undermining evidence to the truth of this whole conjunction’ (G.
Harman, ‘Reasoning and Evidence One Does Not Possess’, Midwest Studies in Phi-
losophy, (), – at ). In another paper Harman suggests that the content
of a belief that φ ‘is something like “I am in this mental state because of something
that settles it that φ” ’ (G. Harman, ‘Self-Reflexive Thoughts’, Philosophical Issues,
(), – at ). Harman argues that the fact that certain self-reflexive
thoughts give rise to paradox (as in the famous paradox of the liar) is not a good
ground for rejecting the possibility of any self-reflexive thoughts, and hence is not a
good ground for rejecting this account of belief.
Rational Assent and Self-Reversion
ment is true’. This is the sense in which belief (doxa) judges as true
its own understanding (. ).
According to ps.-Simplicius, part of what it is to make a judge-
ment is to judge that this judgement is warranted. This cannot be
satisfactorily accounted for by supposing that each act of judge-
ment is accompanied by some further act, in which the original act
is judged to be achieving truth in virtue of being warranted. First,
such a view would imply an infinite regress of acts of judging; but
second, and equally importantly, such a view would fail to account
for the essential relation between making a judgement and taking
yourself to be achieving truth because you are warranted in mak-
ing that judgement. Ps.-Simplicius is not claiming merely that in
judging you must also, necessarily, endorse your judging. His point
is that your act of judging only counts as an act of judging in virtue of
your endorsement of it. Your endorsement does not merely accom-
pany your act of judging, it is what constitutes that act as an act of
judging. As such, ps.-Simplicius assumes, this endorsement must
be included in the act itself. This is why the content of the act of
judgement must be self-referential in the way we have described.
We are now in a position to understand why judging must be the
exercise of a strictly rational power. According to ps.-Simplicius, a
power must be rational if it is to be exercised self-reflexively in the
way that is required for judgement. As we saw earlier, the exercise of
human perceptual powers is only self-reflexive in a certain limited
way. You can exercise a perceptual power in being aware of another
distinct exercise of the same power (as when you use your power of
Of course, in reporting our judgements we do not usually articulate this self-
referential content. We say ‘There is a blackbird in the garden’, not ‘There is a black-
bird in the garden and the understanding I’m exercising in this act of judgement is
true’; we say ‘Labour has a good chance of winning the next election’, not ‘Labour
has a good chance of winning the next election and the understanding I’m exercising
in this act of judgement is true’, and so on. On ps.-Simplicius’ view, this is simply a
pragmatic point about how we express our judgements. What we judge will always
include the self-referential content ‘the understanding I’m exercising in this act of
judgement is true’, and it is possible to make this content of the judgement explicit
(as I have done in the examples above), even though in reporting our judgements we
do not generally do so.
At . – ps.-Simplicius says that διάνοια does not give its assent unless it
also (καί ) judges that its σύνεσις is true. But the καί here does not imply two distinct
acts of judgement. His point is that διάνοια must, in one and the same act of judging,
judge both that P and also that the σύνεσις exercised in this very act of judging is true.
It would imply an infinite regress since the act of judging the original judge-
ment to be warranted would itself need to be judged warranted by yet another act of
judgement, that act by yet another one, and so on.
Ursula Coope
sight in being aware of your seeing), but this awareness cannot be
a single self-reflexive act, nor can it be an awareness of the power
itself. The power you exercise in judging must be self-reflexive in
just these ways. Your judging is a single self-reflexive act, and it is
an act in which you grasp (and evaluate) the power by which you
judge, since it is an act in which you endorse the sunesis that you are
exercising in that very judgement.
(c) Are there cases in which you make a judgement without taking
your judgement to be warranted?
It might seem that ps.-Simplicius’ account of judgement is vulner-
able to counter-examples. There are cases in which we are willing
to describe people as ‘judging’ or as ‘believing’, even though they
do not seem to satisfy the conditions laid down by ps.-Simplicius’
account.
Ursula Coope
It is certainly possible to judge that P while judging that your ge-
neral capacity for determining whether P is unreliable. For instance,
you might judge that Labour will win the next election while ac-
knowledging that you are not very good at predicting such things.
This, I think, is not a counter-example. As we have seen, when
ps.-Simplicius says that in making a judgement you must endorse
the understanding you are exercising, he does not mean that you
must take yourself to be exercising a capacity for judgement that
is generally reliable. To endorse your understanding is to take it to
be achieving truth by functioning well in this instance, and hence
to take yourself to be justified in so judging. You cannot judge that
Labour will win the next election while acknowledging that your
judgement is completely unwarranted. That would be guessing, not
judging.
The possibility of having repressed beliefs might also be thought
to provide a potential counter-example. If you can believe that
P without having any conscious access to this belief, then you
can (presumably) unconsciously believe that P while consciously
judging that this belief is unwarranted. Perhaps, however, ps.-
Simplicius could allow for this. He could maintain that when you
unconsciously believe that P you must also unconsciously believe
that you are justified in believing that P. The fact that you can
at the same time consciously judge that you have no justification
for believing P merely shows that your conscious judgements can
contradict your unconscious beliefs.
Certain types of irrationality might seem to provide a different
kind of counter-example. We do sometimes say things like ‘I can’t
help believing this, even though I know I’m not justified in doing
so’. If you are afraid of flying, you might find yourself ‘judging’ that
the aeroplane you are on will crash, even while recognizing that it
is not reasonable so to judge. If you are an optimist, you might find
you cannot help ‘believing’ that ‘something will turn up’, even while
you acknowledge that all the evidence points the other way.
To this, I think, ps.-Simplicius would reply that in spite of the
language we sometimes use to describe such cases, they are not
really examples of genuine judging or believing. They are, rather,
cases in which people behave in certain respects as if they have the
relevant belief, without actually having it. The fearful flier does not
assent to the claim that the aeroplane will crash and hence does not
believe it will crash, but she has the kind of panicked reaction that
Rational Assent and Self-Reversion
would normally arise from such a belief. The optimist does not be-
lieve that something will turn up, but he has the sunny disposition
one would expect such a belief to engender.
The use of the words ‘belief’ and ‘judgement’ in such cases
is merely evidence of a certain looseness in our language. Ps.-
Simplicius is giving an account not of the way in which people use
the word ‘doxa’ (‘belief’), but rather of what it is to hold something
to be true or take something to be the case. He himself makes this
clear when he notes that the word ‘doxa’ is sometimes used (by
Iamblichus, for instance) for a kind of non-rational cognition: ‘the
cognition of what is superficial and seems to be’ (. –). This
use of the word is not, he says, evidence of any real disagreement
between Iamblichus and Aristotle (or between Iamblichus and
ps.-Simplicius): Iamblichus would agree that one must exercise
a rational power in taking something to be the case; when he de-
scribes a kind of non-rational cognition as doxa, he is simply not
using the word ‘doxa’ to mean belief that something is the case.
I have claimed that this account can help to answer our questions
about the reason-responsiveness of assent. This claim might at first
seem surprising. Ps.-Simplicius follows Aristotle in arguing that
belief (doxa) does not depend on us: it is not eph’ hēmin (. –
, commenting on DA . , b–). Prima facie, this looks
like a rejection of the Stoic view that assenting depends on us, and
hence also of the view that assenting is, in a special sense, reason-
responsive. However, the disagreement here is only verbal. Ps.-
Simplicius’ point is simply that we cannot believe at will, which
is something the Stoics never meant to deny. In fact, I shall argue
Ps.-Simplicius could also have cited Plato’s use of ‘δόξα’ at Rep. , – ,
which suggests that someone who is subject to a perceptual illusion has two contra-
dictory δόξαι. In such a case, for example, you might have a non-rational δόξα that
the stick in water is bent and a rational δόξα that the stick is straight. Clearly, the non-
rational δόξα here would not be something that ps.-Simplicius (or Aristotle) would
call a δόξα.
In this, ps.-Simplicius is following Aristotle. Imagining, Aristotle says, de-
pends on us whenever we wish (ἐφ᾿ ἡμῖν ἐστιν, ὅταν βουλώμεθα, b). As Ian
McCready-Flora remarks, Aristotle’s point is that belief is, in this respect, unlike
imagination: ‘we cannot form beliefs simply because we want to’ (‘Aristotle and the
Normativity of Belief’ [‘Normativity’], Oxford Studies in Ancient Philosophy,
Ursula Coope
that the very feature of ps.-Simplicius’ account that enables it to
explain the impossibility of assenting (and hence believing) at will
also provides a basis for explaining the way in which our capacity
for assent is reason-responsive.
Ps.-Simplicius’ defence of the claim that we cannot believe at will
is often thought to be hopelessly confused. McCready-Flora (fol-
lowing Barnes) provides the following reconstruction: ‘Whether a
given belief is true or false is set by the facts. A belief that p (for some
proposition p) is true given that p and false given that not-p. We do
not, in general, control what the facts are, and therefore it is not up
to us whether we have a true belief or a false belief. The truth and
falsehood of any given belief is a semantic necessity, not subject to
our wishes. Belief is, therefore, not up to us.’ As McCready-Flora
goes on to say, ‘This argument fails . . . because imaginings also
have truth-values.’
We can, I think, credit ps.-Simplicius with a better argument if
we consider his remarks against the background of his more general
account of assent. He argues as follows:
πάσης γὰρ κοινὸν τὸ ἢ ἀληθεύειν ἢ ψεύδεσθαι, ἐπειδὴ ἐν συγκαταθέσει πᾶσα
ὑπόληψις. ἡ δὲ συγκατάθεσις οὐ κατὰ μόνην τὴν τῶν προσπιπτόντων σύνεσιν,
ἀλλὰ καὶ κατὰ τὴν τοῦ ἀληθοῦς ἢ ψεύδους διάκρισιν. ἐν δὲ τῇ πρὸς τὰ πράγματα
συμφωνίᾳ καὶ διαφωνίᾳ τὸ ἀληθὲς καὶ τὸ ψεῦδος· τὰ πράγματα δὲ οὐκ ἐφ᾿ ἡμῖν.
(. –)
Telling truths or falsehoods is common to all [sc. hupolēpsis], since all hupo-
lēpsis consists in assent, but assent is not only in accordance with the under-
standing [sunesis] of what impinges, but in accordance with the judgement
of truth or falsehood. Truth and falsehood consist in agreement and dis-
agreement with the facts, but the facts do not depend on us. (my translation)
(), – at ). For an account of what the Stoics meant when they said that
belief depends on us see sect. above.
If this is right, then the explanatory power of the account is, I think, a point
in its favour. In saying this, I am not claiming to give a conclusive argument in its
favour. I shall not, for instance, argue it is the only account that could provide such
explanations. Nor do I mean to claim that ps.-Simplicius took himself to be making
these points about the account’s explanatory power. Although he makes some brief
remarks about the impossibility of believing at will, he does not address my ques-
tions about reason-responsiveness. This last section of my paper is, then, as much
an exercise in drawing out the philosophical consequences of this account as it is an
attempt at strict historical exegesis.
McCready-Flora, ‘Normativity’, –. McCready-Flora is here following
Barnes, ‘Up to us’, –. Barnes attributes this argument to both Aristotle and ps.-
Simplicius. McCready-Flora claims that Aristotle himself had a better argument,
but that ps.-Simplicius misinterpreted him.
Rational Assent and Self-Reversion
Certainly the last two lines of this argument, taken by themselves,
suggest the interpretation given by McCready-Flora and Barnes.
But for a full understanding of what ps.-Simplicius is saying here,
we need to grasp the significance of his remarks about the kind of
sunesis that is exercised in assent and of his appeal to the notions of
telling truths and falsehoods (alētheuein and pseudesthai).
Ps.-Simplicius takes the argument to apply not merely to be-
lief (doxa) but to every hupolēpsis (or ‘rational cognition’, . ).
‘Hupolēpsis’, he says, is a broader term than ‘doxa’, including within
its scope also epistēmē and phronēsis (. –, commenting on
Arist. DA . , b–). What distinguishes hupolēpsis from
both perception and imagination is that it ‘tells truths or falsehoods’
(ἢ ἀληθεύειν ἢ ψεύδεσθαι). Hupolēpsis can tell truths or falsehoods
just because it involves assent (. –). As I remarked earlier,
‘telling truths or falsehoods’ is making true or false assertions about
how things are (not simply being true or false). That is why ima-
gination does not count as ‘telling truths or falsehoods’. When you
imagine things, they appear to you to be a certain way, but imagin-
ing does not involve making a claim about how things are.
As we have seen, ps.-Simplicius holds that making a claim about
how things are is a matter of assenting to something, and that such
assenting must be the exercise of a particular kind of power. It must
be the exercise of the kind of rational sunesis that is capable of self-
As he makes clear at . –: ‘“ὑπόληψις” is a broader term than “δόξα” but
“δόξα” is here used to stand for all “ὑπόληψις”.’ Ps.-Simplicius is presumably influ-
enced here by the remark with which Aristotle introduces this argument: ὅτι δ᾿ οὐκ
ἔστιν ἡ αὐτὴ νόησις καὶ ὑπόληψις, φανερόν (DA b–). Ps.-Simplicius takes ‘νόη-
σις’ here to mean φαντασία (. –), and hence takes Aristotle to be introducing
this argument with the claim that φαντασία is not the same as ὑπόληψις.
McCready-Flora (‘Normativity’, –) makes it a condition on the interpre-
tation of Aristotle’s argument here that the argument should apply only to belief.
As McCready-Flora points out, Aristotle elsewhere says that our exercise of under-
standing does depend on us. This might, then, be thought to be an objection to
ps.-Simplicius, at least in so far as he is attempting to provide an interpretation of
Aristotle. However, there is, I think, a possible reply to McCready-Flora here. The
sense of ‘depends on us’ in which understanding depends on us is rather differ-
ent from the sense in which belief is said not to depend on us. The sense in which
understanding depends on us is that we can exercise understanding or not at will.
But something analogous also seems to be true of belief: I can call to mind a par-
ticular belief at will. On the other hand, the content of understanding is surely not
something we can alter at will any more than we can believe at will. If you take the
belief that P to be unwarranted, you cannot decide to exercise understanding in think-
ing that P any more than you can decide to believe that P.
See above, sect. , esp. n. .
Ursula Coope
reversion. In our passage he says that every hupolēpsis consists in
assent (. ). Such assent, he says, is not simply in accord with
the sunesis of what impinges (that is, the sunesis we have described
as ‘perceptual’). It must also be in accord with one’s judgement of
truth or falsity (truth and falsity being a matter of agreement or
disagreement with the facts) (. –). These last remarks are,
of course, very brief. But his other remarks about assent suggest
the following view. In assenting, one makes a judgement about
how things are. For this, the exercise of perceptual sunesis is not
sufficient. In assenting, one must be exercising rational sunesis and
moreover one must be taking oneself to be achieving truth in virtue
of the correct functioning of that sunesis. Part of what it is to make a
judgement is to take oneself to be achieving truth in virtue of being
warranted in so judging.
If this is ps.-Simplicius’ point, it goes at least some way towards
explaining why believing at will is not possible. If it were possible
to believe that P at will, then you could consciously judge that P
for reasons that you took to have nothing to do with whether or
not ‘P ’ was true. For instance, you could believe that P on the
grounds that this belief would make you happier, while holding that
this was in fact not a good ground for taking ‘P’ to be true. Ps.-
Simplicius’ account (on the interpretation given here) shows why
this is impossible. As we have seen, this account implies that in
the act of judging that P you must take yourself to be achieving truth
(that is, agreement with the facts) in virtue of the well-functioning
of your sunesis. But this implies that in judging that P you must take
yourself to be justified in so judging. You cannot take yourself to be
justified in so judging while at the same time taking your judgement
Ps.-Simplicius himself encourages us to understand his argument here in the
light of his further remarks about assent. For instance, when he later spells out his
view that rational assent is essentially self-reverting (. –. ), he introduces
this discussion by saying ‘The difference between imagination and belief which is
stated now seems to me to be the same as the one that was mentioned before. There
believing was said not to depend on us because it necessarily involves telling false-
hoods or truths, and here it is said that conviction always follows belief’ (. –).
Denying that we can believe at will is, of course, compatible with recognizing
the possibility of wishful thinking: the possibility that our hopes and desires can in-
fluence what we believe without our realizing it.
Showing this is sufficient for showing that we cannot believe at will. However,
a full account of why believing at will is impossible would have to explain why it is
impossible to withhold assent from P while taking the judgement that P to be de-
cisively justified by the evidence. Ps.-Simplicius’ account does not, I think, help to
explain this.
Rational Assent and Self-Reversion
that P to be based on inadequate grounds, and hence to be unjusti-
fied (and of course, by the same token, you cannot make the meta-
judgement that your judgement that P is justified, while taking that
meta-judgement to be unjustified). Thus, the fact that it is im-
possible to believe at will simply follows from the fact that judging
essentially involves taking oneself to be justified in so judging.
In the special case in which your judgement is about some fact
that depends on you, you have a kind of indirect way of bringing
yourself to make that judgement. If you think you would be happier
if you were to judge that you were singing, you can start singing,
and then judge that you are singing. But of course, this does not in-
volve judging that P while recognizing that you are not warranted
in so judging. You can bring yourself to judge that you are singing
just because you can make it the case that you are warranted in so
judging, and you can do this just because you can make it the case
that you are singing. Most of our beliefs are not about things that
depend on us, and so in most cases we cannot, even in this indirect
way, alter our beliefs at will.
If I am right that this is how ps.-Simplicius’ account rules out
the possibility of belief at will, the same considerations also provide
a basis for answering our questions about reason-responsiveness.
Again, it is the essentially self-reflexive nature of assent that is
relevant. As we saw, Epictetus recognized that in order to count
as assenting, you must be able to reflect on whether you should
My argument depends here on the assumption that it is impossible, at the
same time, to make two obviously contradictory judgements. For some discussion
see n. below.
This also gives ps.-Simplicius an answer to a version of Moore’s paradox.
Moore raises a puzzle about assertions such as ‘P, but I am not justified in judging
that P ’. An assertion of this sort is defective in some way, but its content is not in-
consistent (since ‘P’ and ‘I am not justified in judging that P ’ can both be true).
In what way, then, is it defective? Ps.-Simplicius’ answer is that when you express
a judgement by saying ‘There is a blackbird in the garden’, this only partially ex-
presses your judgement. Fully expressed, your judgement is ‘There is a blackbird
in the garden and I am justified in making this judgement’. But this implies ‘I am
justified in judging that there is a blackbird in the garden’ (since if you are justified
in making the whole judgement, you must be justified in making each part of it). So
your judgement, if fully expressed, is in fact inconsistent with ‘I am not justified in
judging that there is a blackbird in the garden’.
Of course, this does not rule out the possibility of altering one’s beliefs in an
even more indirect way, by voluntarily undergoing a process of education, training,
or indoctrination. In denying that belief depends on us, neither Aristotle nor ps.-
Simplicius means to rule out the possibility of bringing oneself to have new beliefs
in this indirect way.
Ursula Coope
be assenting and to reaffirm or revise your assent in the light of
such reflection. His account, however, left unexplained the relation
between re-evaluating your assent and revising or reaffirming that
assent in the light of this re-evaluation. It did not explain why,
when you conclude your assent is not justified, you thereby cease
to assent. Here the account I have attributed to ps.-Simplicius
provides the basis for an answer. As we have seen, on this account
you not merely can, but necessarily do, engage in a certain kind of
reflection on any act of assent. Your endorsement of your assent
is built into the very act of assent itself: your act of judging that P
includes, as part of its content, the judgement that you are justified
in so judging. This explains why, if you re-evaluate your act of
judging that P and conclude that you are not justified in so judging,
you thereby cease to judge that P. Continuing to judge that P
would involve making two obviously contradictory judgements:
the judgement that you are justified in judging that P (made as
part of your act of judging that P) and the judgement that you are
not justified in so judging.
The fact that assent (and hence judgement) is responsive in
this way to higher-order evaluation also explains the possibility of
judging for a reason. You only count as judging that P on the grounds
that Q if your act of judging that P depends in a certain way on
your taking the fact that Q to provide good grounds for so judging.
For instance, you only count as judging on the basis of the weather
forecast that there will be rain if you take the forecast’s prediction
to provide good grounds for judging that there will be rain and you
judge as you do because of this. This implies that were you to
Though one cannot simultaneously make two obviously contradictory judge-
ments, one can simultaneously have an impression that P and an impression that not
P (as, for instance, in the waterfall illusion, when one has the impression both that
the rocks are moving and that they are not moving). A full account of judgement
would also need to explain this difference between judging and having impressions.
The account we have outlined has something to contribute here too, since it shows
there is an important difference between judging that P and not P and having the im-
pression that P and not P. On this account, in judging that P and not P, you would
have to take yourself to be achieving truth in virtue of the well-functioning of your
judgement-making capacity. By contrast, having the impression that P and not P does
not involve taking yourself to be achieving truth in virtue of the well-functioning of
your impression-having capacity.
For the fact that Q to count as your reason for judging that P, your taking the
fact that Q to be a good reason for so judging must have the right kind of explanatory
connection to your judging that P. It is hard to spell out just what this explanatory
connection is. Not just any causal connection will do. Your judgement that P could
Rational Assent and Self-Reversion
become convinced of the unreliability of the forecast, you would
either revise your judgement or find new grounds for it. That is, it
implies that your judging is responsive to a certain kind of higher-
order reflection on whether it is justified by its grounds. Again,
this is just what ps.-Simplicius’ account (as I have interpreted it)
would predict. On that account, making a judgement essentially
involves endorsing the sunesis exercised in that judgement; when
the judgement is based on reasons, endorsing the sunesis exercised
in that judgement just is taking those reasons to warrant making
that judgement.
Your impressions are not, in the same way, responsive to higher-
order evaluation. This is because having an impression does not
essentially involve either endorsing the content of that impression
or taking the impression to be the exercise of a capacity that is func-
tioning correctly. Thus, although you might in fact cease to have a
certain impression as a result of judging that the capacity that led
you to have that impression was not functioning correctly, when
this happens it is simply the result of a causal relation between that
be caused by your taking the fact that Q to be a good reason for judging that P though
your judgement was not based on the grounds that Q. On the other hand, the required
explanatory connection (between your judging that P and your taking the fact that Q
to be a good reason for so judging) cannot be that your reason for judging that P is the
(presumed) fact that the fact that Q is a good reason for so judging. Such a condition
on judging for a reason would give rise to an unacceptable infinite regress: it would
imply that any judgement based on a reason was based on infinitely many reasons. In
order to judge that P on the grounds that Q, you would also have to judge that P on
the grounds that the fact that Q was a good reason for so judging, and you would also
have to judge that P on the grounds that the fact that the fact that Q was a good reason
for so judging was a good reason for so judging, and so on. (For modern discussions
see P. Boghossian, ‘What is Inference?’, Philosophical Studies, (), –, and
R. Neta, ‘What is an Inference?’, Philosophical Issues, . (), –.) On the
account I have attributed to ps.-Simplicius, this condition on judging for a reason
is simply a special case of the more general condition on judging: if you judge that
P, you must take yourself to be warranted in so judging and judge that P because of
this. The ‘because’ here is a kind of essential ‘because’: you only count as judging
that P if, as part of that very act of judging, you take yourself to be warranted in so
judging.
For a full answer to our questions about the difference between assenting and
having impressions it would be necessary also to explain the possibility of withholding
assent for a reason. Ps.-Simplicius does not himself give an account of withholding
assent. If we extrapolate from what he says about assenting, we might expect him to
claim that withholding assent essentially involves judging oneself justified in with-
holding assent. Since the capacity in virtue of which we withhold assent just is our
capacity for assenting (and hence for judging), it is at least possible for withholding
assent to be related in this way to making a certain judgement. However, the attempt
to spell out such a view in detail would take us far beyond anything in ps.-Simplicius.
Ursula Coope
impression and that act of judgement: there is nothing about the
nature of the impression itself that ensures you will cease to have the
impression when you make such a judgement.
. Conclusion
In this paper we have seen how ps.-Simplicius draws upon the Neo-
platonic notion of self-reversion to explain the nature of rational as-
sent. I have argued that this account of assent provides a basis for
explaining a fundamental difference between assenting and having
impressions: the fact that we can assent for a reason but cannot (in
the same sense) have an impression for a reason.
Ps.-Simplicius’ account thus suggests an interesting new view of
the nature of assent, a view that combines elements of Aristote-
lian, Stoic, and Neoplatonist thought. From the Stoics, he inher-
its the view that believing involves assenting. He draws upon the
Neoplatonist notion of self-reversion to explain the essentially self-
reflexive nature of assent. This enables him to defend Aristotle’s
claim that we cannot believe at will. On this account, though we
do not believe at will, we nevertheless have a kind of rational con-
trol over our beliefs: beliefs, by their very nature, are such as to be
revised or maintained for reasons. This account thus provides an
answer to the question we raised for the Stoics: what is it about the
nature of assent that explains why you are responsible for assenting
in a way in which you are not responsible for having impressions?
You are responsible for assenting just because you can assent (or
withhold assent) for reasons, and you can assent for reasons just
because of the essentially self-reflexive nature of the act of assent.
Corpus Christi College, Oxford
BI B L I O G R A PHY
Barnes, J., ‘ “Belief is up to us”’ [‘Up to us’], Proceedings of the Aristotelian
Society, (), –.
Blumenthal, H. J., with intro. and notes by R. Sorabji, ‘Simplicius’: On
Aristotle On the Soul . – [‘Simplicius’ . –] (London, ).
Bobzien, S., Determinism and Freedom in Stoic Philosophy [Determinism]
(Oxford, ).
Rational Assent and Self-Reversion
Boghossian, P., ‘What is Inference?’, Philosophical Studies, (),
–.
Brennan, T., ‘Fate and Free Will in Stoicism’, Oxford Studies in Ancient
Philosophy, (), –.
Brennan, T., The Stoic Life: Emotions, Duties, and Fate [Life] (Oxford,
).
Brittain, C., ‘The Compulsions of Stoic Assent’ [‘Compulsions’], in M.-K.
Lee (ed.), Strategies of Argument: Essays in Ancient Ethics, Epistemology,
and Logic (New York, ), –.
Brittain, C. (trans.), Cicero: On Academic Scepticism (Indianapolis and
Cambridge, ).
Brittain, C., and Brennan, T., Simplicius: On Epictetus’ Handbook –
(Ithaca, NY, ).
Castagnoli, L., ‘How Dialectical was Stoic Dialectic?’, in A. Nightingale
and D. Sedley (eds.), Ancient Models of Mind: Studies in Human and Di-
vine Rationality (Cambridge, ), –.
Caston, V., ‘Aristotle on Consciousness’, Mind, (), –.
Chlup, R., Proclus: An Introduction (Cambridge, ).
Coope, U., ‘Aquinas on Judgment and the Active Power of Reason’, Phi-
losophers’ Imprint, . (), –.
Golitsis, P., ‘John Philoponus on the Third Book of Aristotle’s De anima,
Wrongly Attributed to Stephanus’, forthcoming in R. Sorabji (ed.), Aris-
totle Reinterpreted: New Findings on Seven Hundred Years of the Ancient
Commentators on Aristotle (London, ).
Harman, G., ‘Reasoning and Evidence One Does Not Possess’, Midwest
Studies in Philosophy, (), –.
Harman, G., ‘Self-Reflexive Thoughts’, Philosophical Issues, (),
–.
Huby, P., and Steel, C., in collaboration with J. O. Urmson and with
notes by P. Lautner, Priscian: On Theophrastus On Sense Perception
with ‘Simplicius’: On Aristotle On the Soul . – [‘Simplicius’ . –]
(London, ).
Inwood, B., Ethics and Human Action in Early Stoicism [Action] (Oxford,
).
Johansen, T., ‘In Defense of Inner Sense: Aristotle on Perceiving that One
Sees’, Proceedings of the Boston Area Colloquium in Ancient Philosophy,
(), –.
Long, A. A., ‘Dialectic and the Stoic Sage’ [‘Dialectic’], in J. M. Rist (ed.),
The Stoics (Berkeley, ), –.
Long, A. A., Epictetus: A Stoic and Socratic Guide to Life [Epictetus] (Ox-
ford, ).
Long, A. A., ‘Representation and the Self in Stoicism’, in S. Everson (ed.),
Ursula Coope
Psychology, Companions to Ancient Thought, (Cambridge, ),
–.
Long, A. A., Stoic Studies (New York, ).
Long, A. A., and Sedley, D. N., The Hellenistic Philosophers [LS] (Cam-
bridge, ).
McCready-Flora, I. C., ‘Aristotle and the Normativity of Belief’ [‘Norma-
tivity’], Oxford Studies in Ancient Philosophy, (), –.
Menn, S., ‘Self-Motion and Reflection: Hermias and Proclus on the Har-
mony of Plato and Aristotle on the Soul’, in J. Wilberding and C. Horn
(eds.), Neoplatonism and the Philosophy of Nature (Oxford, ), –.
Moran, R., Authority and Estrangement: An Essay on Self-Knowledge
(Princeton, ).
Nawar, T., ‘The Stoic Account of Apprehension’, Philosophers’ Imprint,
. (), –.
Neta, R., ‘What is an Inference?’, Philosophical Issues, . (),
–.
Salles, R., ‘Epictetus and the Causal Conception of Moral Responsibility
and What is eph’ hêmin’ [‘Epictetus’], in P. Destrée, R. Salles, and M.
Zingano (eds.), What is up to us? Studies in Agency and Responsibility in
Ancient Philosophy (Sankt Augustin, ), –.
Smith, A., ‘Responsibility for Attitudes: Activity and Passivity in Mental
Life’, Ethics, . (), –.
Sorabji, R., Animal Minds and Human Morals: The Origins of the Western
Debate [Animal] (London, ).
Steel, C., ‘Breathing Thought: Proclus on the Innate Knowledge of
the Soul’, in J. Cleary (ed.), The Perennial Tradition of Neoplatonism
(Leuven, ), –.
Steel, C., ‘Simplicius’: On Aristotle On the Soul . – [‘Simplicius’ . –
] (London, ).
Urmson, J. O., with notes by P. Lautner, Simplicius: On Aristotle On the
Soul . –. [Simplicius . –. ] (London, ).
Von Arnim, H., ‘Über einen stoischen Papyrus der Herculanensischen Bi-
bliothek’, Hermes, (), –.
Watson, G., ‘The Work of the Will’, in id., Agency and Answerability: Se-
lected Essays (Oxford, ) –.
Westerink, L. G., and O’Neill, W. (trans.), Proclus: Commentary on the
First Alcibiades of Plato (The Hague, ).
COMMO N SE N S E A N D E X T RA POWERS
KLAUS CORCILIU S
First let us explain dunamis in the strictest sense, which is, however, not
the most useful [χρησιμωτάτη] for our present purpose. For dunamis and
operate on each other’ (), that all properties are causal powers (), and more.
Unfortunately, there is no room to discuss these claims here.
Cf. Metaph. Θ , a–, a ff., a–. Since Michael Frede’s article
‘Potentiality in Aristotle’s Metaphysics Θ’, in T. Scaltsas, D. Charles, and M. Gill
(eds.), Unity, Identity, and Explanation in Aristotle’s Metaphysics (Oxford, ),
–, these passages have received a good deal of attention in recent scholarship,
notably in three monographs: S. Makin, Aristotle’s Metaphysics Θ (Oxford, ),
J. Beere, Doing and Being: An Interpretation of Aristotle’s Metaphysics Theta (Ox-
ford, ), and A. Kosman, The Activity of Being: An Essay on Aristotle’s Ontology
(Cambridge, Mass., ). For the history of the distinction see S. Menn, ‘The Ori-
gins of Aristotle’s Concept of ἐνέργεια: ἐνέργεια and δύναμις’, Ancient Philosophy,
(), –.
The translations are from J. Barnes (ed.), The Complete Works of Aristotle: The
Revised Oxford Translation, vols. (Princeton, ), with slight modifications. I
refrain from translating δύναμις and ἐνέργεια.
The other main family of manuscripts even reads χρησίμη, i.e. ‘useful’.
Klaus Corcilius
energeia extend further than the mere sphere of change. But when we have
spoken of this first kind, we shall in our discussions of energeia explain the
other kinds of dunamis. (Metaph. Θ , b–a)
And:
Although the form always makes the sense organ like it, it is veridically like
it only when the perceptible, the environment, and the sense organ are in
normal conditions. Perception is trustworthy then when it takes place in
‘standard’ or ‘normal’ conditions. ()
In the case of the capacity to make a sound, a subject may have this ca-
pacity without exercising it; it may make an audible sound, in absence of
anyone hearing (first actuality); or it may be sounding (second actuality)
while—and only while—being heard. Thus, sounding is an object’s capa-
city to produce sound, in second actuality. While an object’s capacity to
produce sound can be activated independently of anyone perceiving it, an
object’s sounding depends on the activation of the corresponding percep-
tual power in the perceiver. (; cf. )
tures (the phenomenal features) and their per se accidents (efficient causal features):
A. Silverman, ‘Color and Color-Perception in Aristotle’s De anima’, Ancient Phi-
losophy, . (), – at . Aristotle’s language is compatible with the idea
that the causal and the formal features of perceptual qualities relate in the way in
which formal and efficient causal features relate as different powers that coincide in
the same thing. But for Marmodoro this is not a viable option, as she identifies per-
ceptual qualities with powers (see sect. above).
Common Sense and Extra Powers
Aristotle allowed himself two alternative courses for the activation of causal
powers: mere activation of a power, or activation-cum-change of a power.
The theory of the perceptual medium takes this distinction from Aristotle’s
power ontology to a higher level of complexity and sophistication. We can
outline it as follows. The color of an object causally engages the actually
transparent air (i.e. the medium) without changing it. The air in turn caus-
ally engages the sense organ of sight, the eye, without changing it, but giving
rise to the perceptual experience of the object’s color. So the perceiver per-
ceives the color itself, because her experience of it is the result of a chain
of activations of causal powers, without change, started by the color itself
(similarly with all other senses). (–)
This is a bold move and I think also more than her account re-
quires. To start with, one might think that, on the face of it, the
absence of change thesis contradicts Aristotle’s frequent statements
that the perceptual object inflicts change (kinēsis) on the media and
sense organs. Marmodoro wants to forestall this objection by say-
ing that the only ‘change’ that is going on is precisely the ‘commut-
ing’ of the perceptual form through the medium to the perceiver
(), and that, she argues, is not a change:
The role of the medium then is to be a causal ‘bridge’ between the sensible
and the sense organ detecting it, for the transmission of what I call a ‘dis-
turbance’. The color ‘disturbs’ the actively transparent medium, without
changing it, and this ‘disturbance’ is transferred to the sense organ, without
changing it, but giving rise to the experience of the perceiver. The ‘dis-
turbance’ that color generates cannot be merely a causal engagement with,
for example, the air, since it cannot take place without light. A ‘disturb-
ance’ may be either physical dislocation or qualitative engagement without
change. For example, the bending of a tree’s leaves in the wind, or the re-
flection of an image on a shiny metal surface are qualitative causal engage-
ments of the leaves or the metal surface, without change (in the relevant
Aristotelian sense). Either way, a color engages causally with the actually
transparent medium—air or water—and through this medium it engages
the sense organ of sight, producing the experience of color. (–)
But these two issues can, and I think should, be kept apart. Say-
ing that the changes in the medium do not embody the perceptual
forms they transmit is perfectly compatible with saying that they
are genuine Aristotelian changes (processes). One only has to dis-
tinguish between changes in the medium and the perceptual forms
they carry, as, for example, in the case of the movement of the air
and the audible form it carries. The sounds’ power to cause the
activation of the faculty of hearing may be the potential sound (as
Marmodoro says). But the fact that the air undergoes a change in
virtue of carrying the audible form—audible sounds are transmitted
by portions of air struck in certain ways—need not alter the sound
(provided there are no interfering violent changes), nor does it re-
quire that the sound belongs to the air in the same way in which
it belongs to the audible objects. It also does not threaten to inter-
rupt the direct causal interaction between the active and the pas-
sive powers of perception. In other words, it seems the absence of
change thesis is an unnecessary burden for Marmodoro’s account.
Her thesis is all the more surprising as she elsewhere in the book
adopts a model of how to think of the relation between percep-
tual form and movement in the medium that makes the absence of
change thesis superfluous. This is Theodore Scaltsas’s ‘encoding
model’. In a article he suggests conceiving of the movements
in the air as possessing the audible form only in the very attenuated
sense of being able to cause the corresponding hearing experience in
the perceiver: the audible form is encoded in a certain structure,
sequence, or pattern, with which the air is struck. Scaltsas help-
fully compares this way of transmitting perceptual forms with Aris-
totle’s explanation for the transmission of the human form in sexual
reproduction: the movements of the male semen do not embody
the human form, but are structured according to a certain pattern
that has the power of causing the coming to be of a human foetus
As Aristotle does at DA . , b–.
T. Scaltsas, ‘Biological Matter and Perceptual Powers in Aristotle’s De anima’
[‘Biological’], Topoi, (), –. I was surprised, therefore, to find Marmodoro
saying that her interpretation of the role of the medium ‘thus follows’ Scaltsas ().
Klaus Corcilius
in a suitable recipient. This seems a viable way of reconciling the
idea that there is genuine change in the medium with Marmodoro’s
thesis that there is direct interaction between the power to produce
perception and the perceptual faculty.
I now turn to Marmodoro’s explanation of how the senses ‘take
on’ the perceptual object without matter. She suggests conceiving
of it as a ‘blending interaction’ between two forms, namely the form
of the perceptible quality and the form of the sense organ:
The disturbance caused by a perceptible quality in the sense organ does not
bring about a (persisting) change in the sense organ. Rather, the sense or-
gan, which has its own form [which is a sort of harmonious mean, KC], re-
gisters the disturbance brought about by the perceptible quality by a blend-
ing interaction between its own form and the disturbances suffered. ()
The idea is, I take it, that both the incoming ‘disturbances’ of the
medium and the perceptual organ contain forms. The medium con-
tains the (encoded) form of the perceptual object, and the percep-
tual organ somehow contains the form of the perceptual faculty.
When they come together, both forms blend in some way and con-
stitute the perceptual experience in a way analogous to the way in
which the blend of the form of the aria and the form of the attuned
chords brings about a reverberation of the aria in the lyre. Unfor-
tunately, this is all we are told. Such as it stands, the analogy raises
more questions than it proposes to answer. Most significantly, we
are not told how the metaphysical framework of the first part of
Scaltsas, ‘Biological’, –. His analogy is designed to illustrate how Aristotle
can maintain that perception is irreducible to physical properties without commit-
ting to the physical primitiveness of perceptual experience.
Common Sense and Extra Powers
Marmodoro’s book maps onto the elements of the analogy. Presum-
ably the analogy is designed to illustrate her metaphysical thesis
that perception is literally constituted by the common activity of
the mutual activations of the relevant co-ordinated partner powers.
But how is this supposed to work in this case? Does the ‘blending
interaction’ of the form of the aria and the form of the lyre illustrate
the direct interaction of co-ordinated perceptual powers? How does
that help us? All I can detect in the scenario is how the sounds of the
aria reverberate in a resounding object (the lyre) that is tuned in a
certain way. I can also see how such reverberation may partly con-
stitute an acoustic experience, but only the audible part of it. But I
fail to see how the reverberation of the sounds of the aria in the lyre
illustrates a ‘blending interaction’ of the two forms. As it stands,
Marmodoro’s elaboration of the analogy is opaque.
Let us start with (), the (controversial) claim that the common
sense (aisthēsis koinē) is in charge of the perception of the so-called
‘common sensibles’ (koina aisthēta). These are perceptible features
Common Sense and Extra Powers
shared by more than one sense modality. Aristotle mentions mo-
tion, rest, shape, extension, number, and unity (DA . , a–;
. , a ff.; cf. De sensu , b). With this Marmodoro does
not want to deny that the special senses have perception of the com-
mon sensibles as well; she denies only that they have a full grasp of
them. Very roughly, she argues as follows (–):
(i) The individual senses perceive their corresponding quali-
ties.
(ii) The individual senses also perceive features of their corres-
ponding qualities that occur in other sense modalities as well
(the so-called common sensibles). Sight, for example, per-
ceives not only colours, but also movement, rest, shape of
colours, and so on. However, each individual sense can per-
ceive the common sensibles only as they occur in their own
sense modality.
(iii) To perceive the common sensibles in one sense modality is
to have only a partial perception of the common sensibles
( ff.).
(iv) Each of the common sensibles (besides occurring in differ-
ent sense modalities) is also a single sensible on its own.
(v) Perception has a full grasp of the common sensibles.
(vi) The common sense has a full perceptual grasp of the com-
mon sensibles.
This is meant to establish that there is a special job for the com-
mon sense that none of the other senses can do, namely perceiving
the common sensibles to their full extent (note here that (iii)–(vi)
are without explicit textual support in Aristotle). What is this com-
mon sense? Marmodoro stresses that it is not an extra sense over
and above the five special senses. Rather, it is the five senses ‘act-
ing together as one, that is when acting as integrated parts of the
perceptual system’ (). But, for her, the common sense is more
than just the co-operation of the five senses: it is endowed with ex-
tra powers on its own that enable it to carry out perceptual opera-
tions over and above the five senses. These extra powers, as we will
see, all boil down to bestowing unity on the diverse inputs of the
‘The answer for Aristotle is that no special sense can perceive movement, the
single sensible that is common to a moving-color and a moving-tickle’ ().
‘[A] full grasp of the common sensibles is possible only in the context of all the
perceptions of the common sensibles supplied by the special senses acting as one’
().
Klaus Corcilius
special senses. This unique function, Marmodoro argues, makes
it necessary that the common sense be given a ‘metaphysically ro-
bust’ interpretation (as opposed to the more deflationary views that
deny such extra powers to the common sense and make its unity
and operations depend on other things, such as, for example, the
body). To make her case, she discusses each of the different kinds
of complex perceptual content (starting with types ()–() above),
plus the constraints Aristotle imposes on the workings of the per-
ception of complex content (mainly in DA . , ; De sensu ; and
De somno ). On that basis, finally, we can see how she would an-
swer our initial question, namely how we can hear the bell if the
bell does not causally act on us as a bell (). The answer, in a nut-
shell, is that we do indeed hear the bell, and that we do so directly,
namely via the direct perception of the common sensibles by the
common sense. The perception of 3-D objects, according to her
account, then, is in a way like the full-fledged perception of com-
mon sensibles. Starting with the common sensibles mentioned by
Aristotle explicitly, she writes: ‘the common sensibles are multimodal
sensibles perceived directly by the common sense’ (, emphasis ori-
ginal), and adds that the ‘perceiver perceives directly via the com-
mon sense the ball’s shape, as color and texture unified in a certain
spatio-temporal location’ (, emphasis original).
The idea is that the full-fledged perception of the common sen-
sibles by the common sense renders a coherent multimodal picture
of the common perceptual features of things (shape, unity, num-
ber, etc.). This is then applied to the perception of 3-D objects in,
roughly, the following way:
(i) Common sensibles are ‘ways in which the special sensibles
are clustered together’ (–).
(ii) The common sense has a direct and full-fledged perception
of the common sensibles.
See Gregoric, Common Sense, –, and T. Johansen, The Powers of Aristotle’s
Soul (Oxford, ), –, against both of whom Marmodoro argues at length
(mostly on the basis of her metaphysical distinctions, ff.). It is surprising that
she does not engage with Modrak’s interpretation of the common sense. For, on the
face of it, Modrak’s view has some similarity to her own: ‘The common sense is
not a separate sense; it just is the capacity for the joint exercise of several senses.
By this means Aristotle significantly expands the number of capacities of the per-
ceptual system. Each special sense is limited to a narrow range of objects, but there
are no similar limitations on the objects that can be apprehended through the com-
mon sense’ (D. K. W. Modrak, Aristotle: The Power of Perception [Power] (Chicago,
), ).
Common Sense and Extra Powers
(iii) 3-D objects are (in a way) unified clusters of properties ().
(iv) 3-D objects are perceived directly by the common sense
(–).
Now (iii) needs to be read with a grain of salt. For the com-
mon sense, she adds, does not perceive 3-D objects as mere clusters
of properties, but as structured wholes. How? The answer lies in
Marmodoro’s ‘metaphysically robust’ interpretation of the com-
mon sense. The common sense is the special senses qua one, but
metaphysically speaking it is very different from them. Its ‘robust’
metaphysical status makes it, i.e. common sense, the source of the
unity of the complex input provided by the special senses. To ex-
plain ‘the unification of their [i.e. the special senses’] complex con-
tent’, she writes:
we need to look further to the subsequent models of the common sense and
its perceptual contents. ()
he [viz. Aristotle] derives (metaphysically) the unity of its perceptual con-
tent from the unity of the common sense. ()
Here, two statements are of interest for our present concern. First,
when a plurality of perceivers are exposed to the same percep-
tual object, the special object of perception (the special sensible,
the idion) will be numerically different, and only specifically (ei-
dei) the same for each perceiver. This is well taken care of by Mar-
modoro’s ‘subtle’ realism. Second, all perceivers hear the numer-
ically same 3-D object, which object is also the first mover of the
whole process (τοῦ κινήσαντος πρῶτον; the examples are all primary
substances: a bell, frankincense, a fire). Now, the interpretation of
this is tricky. What seems clear is that Aristotle thinks perception is
veridical with respect to 3-D objects (bodies, the bell, the frankin-
cense, the fire), and that the 3-D objects are also the first moving
causes of perception. This suggests that we perceive 3-D objects via
affection by their perceptual properties, but that the causal origin
of these affections is not the properties themselves but the 3-D ob-
jects whose properties they are. What seems less clear from the pas-
sage is whether Aristotle also thinks that perceptual qualities can be
veridically perceived without 3-D objects. His statement that each
of the affections implies a body (‘is not without body’: οὐδὲ ἄνευ σώ-
ματος, b) can be taken in least two ways: either we say that
the perceptible affection requires a 3-D object for its existence, or
we say that each perceptible affection is already from the beginning
a perceptible affection of a property of a 3-D object. Now the lat-
Common Sense and Extra Powers
ter seems the more attractive option, not only because it fits better
with the passage’s claim that the hearing of the sound of the bell is a
hearing of the bell (because the sound originates in the bell, its first
mover), but also because it does not require an account of the veri-
dicality of the perception of 3-D objects which is radically different
from that of the perception of special sensibles. It allows conceiving
of perception as being immediately of 3-D objects, without taking
a detour over the isolated perception of qualities (which then have
to be put together by an additional synthetic perceptual power).
On that reading the passage suggests that we perceive 3-D objects
holistically even when we happen to engage causally with them via
one sense modality only. Of course, one would have to show how
this can work in the particular case, i.e. how exposure to the con-
tinuous causal influence of 3-D objects on our special senses can
render their perception. Now this, even though it may prove chal-
lenging as well, seems an approach well worth pursuing. Aristotle
on Perceiving Objects, however, does not consider it.
As for the argument that the common sense is in charge of per-
ceiving the common sensibles to their full extent, it has not become
clear to me how we can ‘fully’ perceive common sensibles as single
sensibles (as claimed on page ). Perhaps this is simply a lack of
imagination on my part. I have difficulties in seeing how motion,
rest, shape, extension, number, or unity can be perceived in ways
that transcend the particular ways in which they occur in the dif-
ferent sense modalities.
See the discussion in S. Herzberg, Wahrnehmung und Wissen bei Aristoteles: Zur
epistemologischen Funktion der Wahrnehmung (Berlin, ), –.
It is interesting in this connection that Marmodoro rejects Charles’s proposal
that the common sense is causally assimilated to the external 3-D object, on the
ground that there is no (specific, I take it) sense organ that can be affected by the
multimodal features of the object so as to become like them (). But does ‘causal assi-
milation’ always have to work in this one-to-one correspondence way? Charles does
not think so. He says that the ‘range’ of that model extends also to the perception of
3-D objects (‘cross model object’ in his parlance): ‘Consideration of the case of com-
mon sensibles shows the range of the causal-likening model. It need not be confined
to the type of physiologically based account offered for special perception. Indeed,
it can be applied even in cases where no physiological account of this type is offered.
In the case of perception of common sensibles, several sense organs can together re-
ceive the relevant form of one moving object, when they discriminate one such object
(in favoured cases) under the causal influence of one continuous moving object’(D.
Charles, Aristotle on Meaning and Essence (Oxford, ), –). So, according to
Charles, continuous exposure to the causal influence of the 3-D object can, via the
perception of common sensibles (unity, shape, motion, and so on), generate assimila-
tion to them.
Klaus Corcilius
Regarding the discussion of the metaphysical unity of the com-
mon sense, I will not go through the different stages of Aristotle’s
alleged development here. But it seems to me that, pace Mar-
modoro’s ‘problem-oriented’ approach (), Aristotle is unlikely
to have gone over one and the same philosophical question as many
as six times, or to have gone back and forth between different
chapters within one or even between different treatises with the
same question in mind. To treat Aristotle’s treatises in this way,
I think, misses something very important, namely the systematic
dimension of the order and sequence of his writings, and the subtle
principles of division of explanatory labour by which they are
governed. This holds to some degree for most of his writings, but
the division of explanatory labour is especially conspicuous in the
mutual cross-references in the De anima and the so-called Parva
Naturalia.
. Conclusion
With these claims (each of which would have easily justified a se-
parate book), Marmodoro proposes a radically new picture of Aris-
totle’s metaphysics of powers and theory perception. However, her
ideas, as we have seen, are not always fully worked out, and some-
times they are presented too hastily to be convincing. Furthermore,
the overall argument would have greatly profited if Marmodoro had
pointed out more clearly how her main theses are supposed to con-
nect with each other. This applies especially to () above, where it
did not become clear how the metaphysical framework in () is sup-
Common Sense and Extra Powers
posed to map onto her account of the perception of what she calls
complex perceptual content by the common sense: what exactly are
the active and passive powers involved in the perception of 3-D ob-
jects? How can a perceptual faculty which, unlike the special senses,
is individuated by its cognitive achievement fit into her scheme of
mutually activating active and passive powers? Marmodoro does
not tell us. This, as we have seen, leaves an important lacuna in her
account, mainly because of the repercussions it has on the thesis in
(), that Aristotle has a causal and direct realist theory of percep-
tion. For the way in which Marmodoro establishes that thesis relies
entirely on her co-ordinated causal powers view of perception. As
a result, it is unclear how she wants to preserve the ‘subtle’ realist
picture of perception in the all-important case of the perception of
3-D objects.
As for the argument for (), the thesis that Aristotle is a pro-
ponent of an ontology of pure causal powers, it seems to me that,
such as it stands, and as an interpretation of Aristotle’s texts, it is
ill-conceived. In spite of the impressive amount of ingenuity in-
vested by Marmodoro, readers who are familiar with Aristotle’s
metaphysics are not likely to be convinced of her claim that he
thought of powers as subjects of change (even if it is a ‘transition’).
Marmodoro’s application of this metaphysical framework to mod-
ally specific perception in (), by contrast, should have appeal for
the scholar of Aristotle. Her ‘subtle’ perceptual realism maximally
teases out the explanatory potential of Aristotle’s basic account of
change. As for (), I cannot evaluate here the philosophical merit of
the proposal that the common sense is a ‘metaphysically robust’ ca-
pacity of perceptual synthesis, especially since Marmodoro herself
does not pause to consider its philosophical implications. But there
can be no doubt that her proposal strongly assimilates the common
sense to the synthetic powers of the Aristotelian intellect. Whatever
we may think of it as a thesis, though, Marmodoro’s at times re-
markably ingenious discussions will nourish further discussion in
the field. One does not have to agree with all, or even any, of the
main theses of Aristotle on Perceiving Objects to profit from read-
ing it. I think the book will have a positive effect on the discussion
of Aristotle’s theory of perception, in spite of the methodological
and interpretative reservations I have expressed here. It inspires
readers to think harder about the actual workings of perception in
Aristotle and its philosophical underpinnings. This is always a good
Klaus Corcilius
thing, and especially in the context of Aristotle’s theory of percep-
tion, where the trend of the past few decades was to focus on ques-
tions which drew attention away from the explanatory powers of
Aristotle’s metaphysics of ordinary change. I thus welcome Aris-
totle on Perceiving Objects and recommend it to anyone who wishes
to engage seriously with Aristotle’s philosophy of perception.
University of California, Berkeley
BI B L I O G R A P HY
Barnes, J. (ed.), The Complete Works of Aristotle: The Revised Oxford
Translation, vols. (Princeton, ).
Beere, J., Doing and Being: An Interpretation of Aristotle’s Metaphysics
Theta (Oxford, ).
Caston, V., ‘Phantasia and Thought’, in G. Anagnostopoulos (ed.), A
Companion to Aristotle (Oxford, ), –.
Charles, D., Aristotle on Meaning and Essence (Oxford, ).
Frede, M., ‘Potentiality in Aristotle’s Metaphysics Θ’, in T. Scaltsas, D.
Charles, and M. Gill (eds.), Unity, Identity, and Explanation in Aris-
totle’s Metaphysics (Oxford, ), –.
Furth, M., Substance, Form and Psyche: An Aristotelian Metaphysics (Cam-
bridge, ).
Gregoric, P., Aristotle on the Common Sense [Common Sense] (Oxford,
).
Herzberg, S., Wahrnehmung und Wissen bei Aristoteles: Zur epistemologi-
schen Funktion der Wahrnehmung (Berlin, ).
Johansen, T., The Powers of Aristotle’s Soul (Oxford, ).
Kosman, A., The Activity of Being: An Essay on Aristotle’s Ontology (Cam-
bridge, Mass., ).
Makin, S., Aristotle’s Metaphysics Θ (Oxford, ).
Menn, S., ‘The Origins of Aristotle’s Concept of ἐνέργεια: ἐνέργεια and δύ-
ναμις’, Ancient Philosophy, (), –.
Modrak, D. K. W., Aristotle: The Power of Perception [Power] (Chicago,
).
Scaltsas, T., ‘Biological Matter and Perceptual Powers in Aristotle’s De
anima’ [‘Biological’], Topoi, (), –.
Silverman, A., ‘Color and Color-Perception in Aristotle’s De anima’, An-
cient Philosophy, . (), –.
Sorabji, R., Animal Minds and Human Morals: The Origins of the Western
Debate (London, ).
Wedin, M. V., Mind and Imagination in Aristotle (New Haven, ).
I N D E X L O C O R UM