Closing The Gap

Meeting of the Aristotelian Society held at Senate House, University of
London, on 22 February 2016 at 5:30 p.m.
IX—CLOSING THE GAP: A NEW ANSWER TO AN OLD

OBJECTION AGAINST KANT’S ARGUMENT FOR
TRANSCENDENTAL IDEALISM
TOBIAS ROSEFELDT
Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

In this paper I present a new solution to the so-called ‘neglected alterna-
tive’ objection against Kant’s argument for transcendental idealism.
According to this objection, Kant does not give sufficient justification for
his claim that not only are space and time forms of our intuition but they
also fail to be things in themselves or properties thereof. I first discuss a
proposal by Willaschek and Allais, who try to defend Kant against this
charge by building on his account of a priori intuition, and argue that it is
insufficient to meet the objection in its full force. I then present my own so-
lution to the problem. It is based on a reconstruction of Kant’s account of
properties of appearances, and tries to show that this account implies that
spatio-temporal properties could in principle not pertain both to appear-
ances and to things in themselves.
The Gap. Kant’s transcendental idealism consists in the claim that

space and time are not things in themselves and that spatio-temporal
properties and relations do not pertain to objects independently of
the subjective conditions under which these objects appear to us.
Kant’s most important attempt to establish this claim starts with ar-
guing for the assumption that we have a priori intuitions of space
and time (CPR, a22ff./b37ff. and a30ff./b46ff.),1 and then continues
with an attempt to show that such a priori intuitions would not be
possible of features that objects have in themselves, but can only be
explained if we assume that space and time are nothing but ‘forms of
our intuition’ and that the spatio-temporal order of the appearing
1 Kant’s Critique of Pure Reason is cited as ‘CPR’, with page references to the 1781 and
1787 editions, given as a and b respectively. Other works by Kant are cited by volume and
page number in the Akademie edition (‘Ak.’). See the References section for details.
C 2016 The Aristotelian Society

V
Proceedings of the Aristotelian Society, Vol. cxvi, Part 2
doi: 10.1093/arisoc/aow011
182 TOBIAS ROSEFELDT
world is imposed on it by the structure of our sensible faculty (a26/

b42 and a32f./ b49f.). Kant’s line of thought is controversial in sev-
eral respects: One might certainly question the claim that we have an
a priori intuition of space and time, and find his arguments for this
claim inconclusive; and one might also doubt that attributing the
spatio-temporality of the contents of our experience to the structure
of our own mind really explains that we can have an a priori cogni-
tion of this structure. However, the most obvious objection to
Kant’s way of reasoning seems to be that even if Kant succeeded in
proving that space and time are ‘forms of our intuition’, this still

seems insufficient for establishing the claim that space and time can-
not also be features of the mind-independent world. In this case the
structure we impose on the contents of our experience would corre-
spond to the structure that the world has independently of us, be it
as a matter of pure coincidence or as the result of the actions of a be-
nevolent creator who has adjusted our forms of intuition to what re-
ality is like in itself. Hence, Kant seems to have neglected the
following alternative to what he claims to have shown:
(na) (i) Space and time are forms of our intuition; (ii) space and
time are also things in themselves and spatio-temporal prop-
erties and relations pertain to objects as they are in
themselves.
It might be that the fact that space and time are forms of our intui-
tion excludes the possibility that we could ever know that (na) holds
because we would represent the world as having a spatio-temporal
structure no matter whether it has it or not. However, it seems that
we should then also not be able to know that the world does not
have this structure. But this is clearly what Kant assumes when he
claims that things in themselves do not exist in space and time.
In the literature the aforementioned objection is not only known
as the ‘problem of the neglected alternative’,2 but is also discussed
under the label ‘Trendelenburg’s gap’ (‘Trendelenburgsche Lücke’),
which relates it back to a heated debate between Adolf
Trendelenburg and Kuno Fischer between 1860 and 1870.3 In fact
the objection dates back much further and already came up in very
early reactions to Kant’s Critique of Pure Reason, for example, in
2 For an overview of the literature on the topic, see Specht (2014a).

3 See Bird (2006) and Specht (2014b).

V
CLOSING THE GAP 183
Ludwig Heinrich von Jakob’s Prüfung der Mendelsohnschen

Morgenstunden oder aller spekulativen Beweise für das Daseyn
Gottes (1786). Jakob was a Kantian and wanted to defend Kant’s re-
striction of knowledge against Mendelsohn in his book. He lets a fic-
titious interlocutor raise the following worry about Kant’s doctrine
of space and time:
I concede that since we have to think everything necessarily in space
and time, these representations have to be grounded in the nature of
our soul, and that they hence precede all our empirical cognition; how-

ever, could not the nature of things nevertheless be such that these
predicates pertain to them also in themselves and that our soul was
given this constitution in order to prevent errors in our cognition, and
to guarantee that our representations correspond to the things them-
selves?’ (Jakob 1786, pp. 26f.; my translation)
Jakob dismisses this worry quickly and on rather questionable

grounds. He first says that it can be ignored because it only concerns
a ‘hypothesis’, a remark that seems beside the point at issue; and he
then asserts that ‘all philosophers’ would deny that things in them-
selves exist in space and time anyway, a claim that seems blatantly
false (Jakob 1786, pp. 27f.). A third, more promising, reply consists
in the remark that ‘the sensible world is only a phenomenon for us,
and that, since all our predicates are only valid of phenomena, we
cannot find any one of them that was suited for the things them-
selves’ (ibid., p. 28). However, as long as we are not told why a pred-
icate that is valid for phenomena is in principle unsuited to be
applied to things in themselves, this reply simply seems to beg the
question against the interlocutor.
If Kant took notice of the discussion of the ‘neglected alternative’
objection in Jakob’s book, which seems rather likely,4 he must have
been rather unimpressed by it, for he did not feel the need to address
it explicitly in the second edition of the Critique. However, he added
a passage to this edition that, in my opinion, in fact contains all the
material to defend him against the charge of not ruling out the possi-
bility that space and time are forms of intuition and features
of things in themselves at the same time. In this passage, which oc-
curs in b69f. of the Critique, Kant deals with another objection
4 Kant had sent Jakob his own short reply to Mendelsohn’s Morgenstunden (Ak. viii:152–
5); this text was then included in Jakob’s book under the title ‘Einige Bemerkungen von
Herrn Professor Kant’ (Jakob 1786, pp. xlix–lx).

V
raised by some of his early critics, namely, that his restriction of our
knowledge to appearances commits him to an extreme and self-
refuting form of idealism that turns everything into ‘mere illusion’.5
Kant answers this objection by distancing his own distinction be-
tween appearances and things in themselves from a certain extremely
idealist misunderstanding of it. What I want to show in the following
paper is that, given his own positive account of this distinction, it in-
deed follows that if space and time are properties of appearances,
they cannot also be properties of things in themselves, and that
Jakob’s blunt claim that predicates applicable to phenomena are in

principle unsuited to be applied to things in themselves is well-
justified.
Before I do so, however, I want to discuss an alternative strategy
to defend Kant against the ‘neglected alternative’ objection. This was
first proposed by Marcus Willaschek (1997), and has recently been
renewed in a slightly modified form by Lucy Allais (2010, 2015).
Although I find Willaschek’s and Allais’s defence ingenious, and
think it can answer the objection on some reading of it, I will argue
that it falls short of showing that Kant is justified in claiming what
he claims, namely, that things in themselves do not have spatio-
temporal properties and do not stand in spatio-temporal relations to
each other (§ii). In §iii, I will then give an detailed interpretation of
the mentioned passage in b69f., and reconstruct his remarks about
the distinction between appearances and things in themselves in a
way that makes intelligible that, once we have shown that space and
5 See the review of the Critique of Pure Reason by Feder and Garve from 1782 (Sassen 2000,
pp. 53–8) and Pistorius’s reviews of Schulze’s Erl€ auterungen über des Herrn Professor Kant
Critik der reinen Vernunft, which appeared in 1786 (Gesang 2007, pp. 3–25; Sassen 2000,
pp. 93–105). In this review, as well as in his review of Jakob’s book from 1788 (Gesang 2007,
pp. 39–71), Pistorius also raised a version of the neglected alternative objection. Pistorius ac-
cuses Kant of having neglected the third of the following three possible alternative views one
might have about our representation of space and time, namely, ‘that they are merely subjec-
tive, or merely objective, or both subjective and objective at the same time’ (Gesang 2007,
pp. 42f.). Contrary to Jakob, Pistorius did not assume that the neglected third alternative is
one in which space and time are forms of intuition and properties of things in themselves, but
characterized it as one in which the spatio-temporal structure of our experience is partially
grounded in an analogous structure of the mind-independent world, so that we are able to ex-
perience a ‘real plurality’ and a ‘real alterability of the represented things in themselves’
(Gesang 2007, p. 9). He also explains why his version of the neglected alternative objection is
superior to the one mentioned by Jakob and not refuted by the kind of reply that Jakob gives
(Gesang 2007, pp. 44ff.). It is an interesting question whether Pistorius’s version of the ‘ne-
glected alternative’ objection can be refuted by means of the strategy presented in this paper. I
will leave answering it for some other occasion.

V
CLOSING THE GAP 185
time are forms of our intuition, it does indeed follow that they can-
not also be properties of things in themselves.
II
Closing the Gap 1: Kant’s Account of Intuition. Willaschek and

Allais both base their answer to the ‘neglected alternative’ objection
on the insight that Kant does not start his argument for the ideality
of space and time from the assumption that we have some a priori

cognition of these objects, but rather from the claim that our a pri-
ori representation of them is an intuition.6 They then try to show
that, given the specific characteristics that Kant assigns to intui-
tions, the objects of this a priori intuition could not be things in
themselves. The decisive feature of intuitions is that they are ‘singu-
lar’ and ‘immediate’ representations. According to Willaschek and
Allais, this characterization implies that intuitions of extra-mental
objects do not represent these objects in virtue of containing as
their content a certain general condition that these objects happen
to fulfil, but rather because there exists a certain causal connection
between representation and what is represented: an intuition i pre-
sents a certain extra-mental object o (and not any other) to an epi-
stemic subject S because it is this very object o that affects S in such
a way that i arises in her. From this it follows that in order for our
a priori intuitions to represent space and time as things in them-
selves, these things would have to be the causal source of our intui-
tions of them.
Now, Willaschek and Allais diverge on how they try to show that
space and time conceived of as extra-mental things in themselves are
not the causal sources of our pure intuitions and hence could not be
what is presented in them. According to Allais, this is excluded sim-
ply because pure intuition does not involve any kind of causal affec-
tion at all. Willaschek, on the other, concedes that pure intuition
does involve affection and is the result of two causal factors: on the
one hand, the forms of intuitions, and on the other, the spontaneity
of the understanding, which—according to §§24 and 26 of the b-
Deduction—has to act under the name of a ‘transcendental imagina-
tion’ in order for space and time to be given to us as objects in pure
6 For the following, see Willaschek (1997) and Allais (2015, ch. 8, §v).

V
intuition. However, given that the factors that are in fact responsible
for our pure intuition of space and time are not extra-mental things
in themselves, but rather features of our own mind, the above princi-
ple excludes that pure intuition could represent extra-mental things
in themselves. So, in the end, both Willaschek and Allais conclude
that it follows from Kant’s account of intuition that the objects of
pure intuition cannot be extra-mental things in themselves, and
hence, since space and time are the objects of pure intuition, that
space and time are not things in themselves.
I think that this an ingenious and very attractive attempt to defend

Kant against the charge of the neglected alternative, which can ac-
count for a number of peculiarities about the way Kant presents his
claims about the ideality of space and time. For example, it can ex-
plain why Kant seemed to have thought that he could infer the ideal-
ity of space directly from the fact that things in themselves and their
features cannot be objects of our a priori intuition (see a26/b42 and
Prolegomena, Ak. iv:282). However, I believe that the proposal falls
short of addressing the ‘neglected alternative’ objection in its full
force. The worry is this: what Willaschek and Allais manage to show
is that if space and time were things in themselves or features of
them, then these things would not be what we are presented with in
a priori intuition. However, this seems to be a merely epistemologi-
cal claim, whereas the original objection dealt with a certain onto-
logical scenario: the neglected alternative was simply described as
one in which space and time are forms of intuition and also things in
themselves, period, not as one in which they are things in themselves
of which we have an a priori intuition. Hence, it seems unclear how
Kant can rule out that, on the one hand, there is one spatio-temporal
structure of which we are aware in pure intuition, but on the other
hand, there is another spatio-temporal structure that exists in itself.
But the second scenario seems exactly what Kant does want to rule
out when he claims that things in themselves do not exist in space
and time. So Kant seems at least to have neglected the following
alternative:
(na*) (i) Space and time are forms of our intuition, and the space
and the time of which we have a priori intuitions are not
things in themselves; (ii) there is another space and another
time, of which we do not have any intuition, that do exist
in themselves.

V
CLOSING THE GAP 187
Both Allais and Willaschek are aware of this kind of worry. Here is
how Allais responds to it:
[I]t is important to notice that the claim that things in themselves are
not spatio-temporal is not exactly what Kant asserts in the Aesthetic
(although he does provide argument for it in the Antinomies). Rather,
he says that space and time represent no property of things in them-
selves or relations between things in themselves (‘Der Raum stellet gar
keine Eigenschaft irgend einiger Dinge an sich, oder sie in ihrem
Verh€altnis auf einander vor’ a26/b42). He says that we have represen-
tations of space and time that play a fundamental role in our experi-

ence of objects, and that these representations do not present us with
mind-independent features of reality. This is not a positive claim about
the nature of things as they are in themselves (that they are not spatio-
temporal), but instead a claim about what is presented to us: our repre-
sentations of space and time do not present us with mind-independent
features of reality. It is compatible with this that there should be some
structure in things as they are in themselves; the point is that this struc-
ture is not what it is that is present to us when we represent space and
time. Here again, the idea that intuitions immediately present their ob-
jects makes sense of Kant’s conclusion: it enables us to say that even if
there were something like space and time in mind-independent reality,
this something would not be that of which our representation of space
is a representation, since it would not be that which is present to us in
a priori intuition. (Allais 2015, p. 198)
There seem to be two possible interpretations of what Allais is saying

here. According to one reading, which is the one supported by most
parts of the passage, and hence probably the one that correctly rep-
resents what Allais has in mind, Kant simply does not want to rule
out the alternative (na*), or at least he does not want to do so in the
Transcendental Aesthetic, but only later by means of the indirect ar-
guments presented in the Antinomies chapter. In the Aesthetic, Kant
would not want to deny that things in themselves exist in space and
time, but only that a mind-independent space and time could be
what we are presented with in pure intuition. I do not have any com-
plaints against this strategy as an attempt to clarify what Kant
should have said in order to avoid the objection. However, I doubt
that it correctly captures what Kant actually asserts in the
Transcendental Aesthetic. For example, in the passage about time
that corresponds to the claim about space that Allais quotes, Kant
not only says that our representation of time does not represent us

V
with a thing in itself, but he claims that time is nothing that exists in
itself or pertains to objects as an objective determination (a31/ b49).
And there are also other passages where he makes the stronger claim
about space (see a39f./ b56f.).
Now, some parts of what Allais writes also allow for a stronger
reading. When she says that ‘if there were something like space and
time in mind-independent reality, this something would not be that
of which our representation of space is a representation’, then this
could be understood as an objection to the effect that I have some-
how misdescribed the neglected scenario (na*) as one in which

‘there is another space and another time, of which we do not have
any intuition’. For whatever structure there is in the world as it is
in itself, so the objection goes, it could in principle not be a spatio-
temporal structure, but only something ‘like’ or ‘analogous to’ the
structure of space and time. If this is what Allais wants to claim,
then she would be able to deal with the ‘neglected alternative’ objec-
tion in its (na*)-variant, for then nothing could be correctly charac-
terized by us as space and time that is not the object of our pure
intuition.
In order for this strategy to work, we need an explanation of why
Kant is allowed to claim that no structure of reality that is not the
object of our intuition could correctly be described as space and
time. It is important to note that the required argumentational re-
sources cannot come from Kant’s account of intuition alone. What
we would need in addition is an argument to the effect that we lack
any representational resources, be they intuitional or conceptual,
that allow us to characterize a possible scenario in which there is a
spatio-temporal structure that is not identical to that which is pre-
sent to us in pure intuition. For only then could we assume that the
above description of (na*) is in fact empty. Allais does not make
any attempt to provide such an argument, but the basic idea of it can
be found in a passage by Willaschek. Willaschek claims that there is
an analogy between what Kant says about space as the object of
pure intuition and what semantic externalists say about the meaning
of fictional names such as ‘Hamlet’, namely, that these names could
in principle not refer to any real person even if there happened to be
a person in the real world that has all the features associated with
the bearer of the name in the respective story. He writes:

V
CLOSING THE GAP 189
Genau das gilt nach Kant nun auch für den Raum: Selbst wenn es
unabh€angig von unserem ‘Gemüt’ etwas geben sollte, auf das alle allge-
meinen Merkmale des Raumes zutreffen (Dreidimensionalit€at,
Homegeneit€at, Unendlichkeit, etc.), es w€are nicht der Raum, sofern wir
darunter den Gegenstand unserer Raumvorstellung verstehen. Damit
es sich bei einem Gegenstand . . . tats€achlich um den Raum handelt,
muß er dem externalistischen Anschauungsbegriff zufolge die Ursache
unserer Raumvorstellung sein. Wenn diese also eine Anschauuug a pri-
ori ist, kann ihre Ursache kein von uns unabh€angiger Gegenstand . . .
sein. (Willaschek 1997, p. 555)

As with the quote from Allais, this passage also allows for a weak
reading according to which it simply amounts to the claim that a
mind-independent structure could never be identical to ‘our space’,
that is, the space that is the object of our intuition, even if it was dis-
playing all the characteristics that the latter has. And again this read-
ing would not suffice to rule out (na*) for it would leave open the
possibility that besides the space that we intuit there is another space
in itself. However, Willaschek can also be understood as drawing the
analogy between Kant’s conception of space and the semantics of fic-
tional names further, and ascribing to Kant not only an externalist ac-
count of the intuition of space, but also an externalist account of the
semantics of the term ‘space’. According to such an account, the mean-
ing of the term ‘space’ would tie its referent to its actual causal source,
and hence rule out that this term could refer to anything that is not the
object of our pure intuition. This would suffice to rule out scenario
(na*), for the description of a scenario in which there is a space as the
object of pure intuition as well as another space as a thing in itself
would be just as empty as the description of a scenario where Hamlet
is not only a character in the story but also a person in the real world.
Whether or not a reply to the ‘neglected alternative’ objection
along these lines makes for a plausible interpretation obviously de-
pends on whether it is legitimate to assume that Kant would assent
to the relevant semantic externalism. Now, it seems not entirely
anachronistic to assume that Kant would have subscribed to a cer-
tain form of semantic externalism, at least with respect to genuine
singular terms. Kant thought that the only genuine singular mental
representations of which we are capable are intuitions, not concepts
(see Logik Pölitz, Ak. xxiv:567; and Wiener Logik, Ak. xxiv:908).
So he might have agreed that genuine singular terms could also only
be given a meaning by associating them with a particular intuition,

V
not with a particular concept. If we accept the externalist account of

Kantian intuitions, this would then imply that genuine singular
terms could not possibly denote any objects other than the ones that
are in fact represented in the intuitions with which they are associ-
ated. Now, although Kant sometimes uses the expressions ‘space’
and ‘time’ as general terms, which allow for the plural (e.g. a25/
b39), he certainly also has a use for them as singularia tanta (for ex-
ample, when he says that ‘space is represented as an infinite given
magnitude’; a25/b39f.). And hence, given all we have said so far,
Kant would after all be justified in asserting that space and time are

not things in themselves, and that a structure of mind-independent
reality would not be identical to space even if it displayed all the fea-
tures that we attribute to space. He could do so without adding the
qualification ‘our space’ or ‘space as the object of our intuition’,
since the term ‘space’ itself essentially could not denote any object
other than the space that we intuit in pure intuition.
Does this close the gap in Kant’s argument and eliminate the alleg-
edly neglected alternative? I do not think so. In order to see why, we
have to notice that what I have been negligently referring to as the
neglected alternative is in fact a sum of two such alternatives. Clause
(ii) in the above formulation of (na) reads: (a) ‘space and time are
also things in themselves’, and (b) ‘spatio-temporal properties and
relations pertain to objects as they are in themselves’. (a) and (b)
each correspond to one of the two views about the nature of space
and time that Kant criticizes and contrasts with his own account. (a)
is the Newtonian container view, according to which space and time
are mind-independent substances that exist independently of the ob-
jects that they contain. (b) is the relational conception of space and
time, according to which there are really only things that have
spatio-temporal properties and stand in spatio-temporal relations,
and maybe also the totality of these properties and relations.7 Now,
a reconstruction of Kant’s argument along the lines of Allais’s and
Willaschek’s interpretation in its strong reading might be able to
deal with the (a)-variant of the neglected alternative that could be
constructed on the basis of these two views. For, as we have seen, it
might indeed be false to say that it is possible that space and time are
forms of our intuition and that nevertheless space and time are also
7 Kant distinguishes between these two scenarios in his ‘Schlüsse aus obigen Begriffen’ more
carefully with respect to time (a32f./b49) than with respect to space (a26/b42).

V
CLOSING THE GAP 191
things in themselves, as the Newtonian conception has it.8 However,

this does not rule out the (b)-variant of the neglected alternative, ac-
cording to which space and time are forms of our intuition and there
are nevertheless things in themselves that have spatio-temporal prop-
erties and stand in spatio-temporal relations, although space and
time themselves do not also exist beyond their status as forms of our
intuition. That is to say that the interpretation does not rule out the
following variant of the neglected alternative:
(na**) (i) Space and time are forms of our intuition and are not

things in themselves; (ii) mind-independent things have
spatio-temporal properties in themselves and stand in
spatio-temporal relations in themselves.
In order to rule out (na**), it is not enough to show that the singu-
lar terms ‘space’ and ‘time’ have an externalist semantics and exclu-
sively refer to the objects of pure intuition. One would rather have
to assume that also all general terms that express spatio-temporal
concepts have an externalist semantics and cannot refer to properties
and relations that things have in themselves. However, assuming this
wider form of semantic externalism within the context of Kant’s the-
ory would be implausible for at least two reasons. First, there seem
to be no positive reasons to ascribe to Kant a semantic externalism
also with respect to general terms. Allais and Willaschek assumed an
externalist account of intuitions and singular terms on the basis of
the singularity and immediacy of intuition. But it is exactly these two
features by which Kant distinguishes intuitions from concepts, and
which he holds the latter not to have. General terms express con-
cepts, and Kantian concepts represent certain objects not because
these objects are their causal sources but rather because those objects
happen to have the properties that serve as the content of the con-
cepts. Secondly, there is a positive reason to deny that Kant held a se-
mantic externalism about general terms, namely, that this would
lead to unacceptable consequences within his system. Assume that
the use of one of our empirical concepts C is in fact always the result
of the fact that the relevant affecting things in themselves have a cer-
tain property P, and hence C tracks P, even if we cannot know that
it does. If semantic externalism were true, then in such a scenario, C
8 Independently of this argument, Kant seems to have thought that the Newtonian concep-
tion of space and time is metaphysically so absurd that it is not a serious contender anyway
(see a39/b56 and b70f.).

V
would represent P. This is absurd, since empirical concepts are sup-

posed to be essentially unsuited to represent properties that things in
themselves have.
These objections point to a general weakness of Allais’s and
Willaschek’s reconstruction. Their argument only shows that the ob-
ject of pure intuition cannot be a thing in itself. However, we also
have empirical, affection-based intuitions of spatio-temporal objects,
and it is far from clear how their considerations translate to this
case. Assume that the scenario in (na**) holds, and things in them-
selves stand in spatial relations and have spatial properties.

Perceiving an object as being cubical, say, would then be the causal
effect of the cubicality of that object. Allais and Willaschek would
have to assume that this cannot be the case, because our form of in-
tuition is involved as one of the causal factors that make the percep-
tion possible. But it is not at all clear that this is true. The opponent
who confronts Kant with (na**) could conceive of our form of intu-
ition as a standing causal condition that makes it possible for us to
represent determinate spatial properties such as the cubicality of the
affecting object. Representational externalism alone does not rule
this out.
One might object that Kant has already shown in the first two ar-
guments of the metaphysical exposition of the concept of space that
the relational theory of space and time is wrong. But this is not true.
What Kant has in fact shown is that the relational theory of space
and time, and its assumption of empirical intuitions of particular
spatial and temporal properties and relations, is incomplete, for it
cannot account for the fact that we have a priori intuitions of space
and time that precede all cognitions of particular spatio-temporal re-
lations and are independent of things existing in space and time
(CPR, a23f./b38f., a39f./b56f.). However, in the considered sce-
nario (na**) there would not only be spatio-temporal relations
among things in themselves, but also space and time as forms of in-
tuition that explain these a priori intuitions. And hence, this scenario
seems still a genuine possibility that is not excluded by the Kantian
considerations we have dealt with so far.
What we need in order to rule out (na**) is a different argument
to the effect that it is not only the singular terms ‘space’ and ‘time’
that are unsuited to refer to things in themselves, but also that all the
predicates and concepts that we use in order to speak and think
about the spatio-temporal properties and relations of appearances

V
CLOSING THE GAP 193
could never be used successfully in order to characterize things inde-

pendently of how they appear to us. Now, I think that Kant does in
fact provide such an argument. This argument does not result from
semantic externalism about general terms, but rather from a clarifi-
cation of the distinction between appearances and things in them-
selves. If we understand this distinction in the correct way, it is
indeed true that once we have shown that spatio-temporal properties
and relations pertain to appearances, we have shown that they can-
not also pertain to things in themselves. And hence Kant is justified
in assuming—to put it in Jakob’s words—that ‘all our predicates are

only valid of phenomena, and we cannot find any one of them that
was suited for the things themselves’. Or so I will argue in the re-
minder of this paper.
III
Closing the Gap 2: Kant’s Distinction Between Appearances and

Things in Themselves. I will base my interpretation on a passage that
Kant added to the second edition of the Critique in order to answer
an objection that was made by some of his critics, namely, that Kant’s
transcendental idealism turns what we normally conceive of as a
mind-independent world into ‘mere illusion’. The charge was that
Kant’s position is not only indistinguishable from that of Berkeley,9
but also leads to an infinite regress, since it takes the human soul to be
an appearance too, whereas in fact the existence of appearances in-
volves a subject to which something appears and which cannot again
be an appearance.10 Kant answers these objections by accusing his
critics of ignoring the distinction between appearance (‘Erscheinung’)
and illusion (‘Schein’), and by clarifying what he means by his claim
that spatio-temporal properties pertain only to appearances and not
to things in themselves. Here is what he writes:
If I say: in space and time intuition of outer objects as well as self-
intuition of the mind represent both of these things as each affects our
senses, i.e. as it appears, that is not to say that these objects were mere
illusion. For in the appearance the objects, indeed even properties that
9 See the review by Feder and Garve (Sassen 2000, pp. 53–8).
10See H. A. Pistorius’s reviews of Schulze’s Erl€
auterungen über des Herrn Professor Kant
Critik der reinen Vernunft (Gesang 2007, pp. 4f.).

V
we attribute to them, are always regarded as something really given,

only that—in so far as this property depends only on the kind of intui-
tion of the subject in the relation of the given object to it—this object
as appearance is to be distinguished from itself as object in itself. Thus
I do not say that bodies merely seem to exist outside me or that my
soul only seems to be given in my self-consciousness if I assert that the
quality of space and time—in accordance with which, as condition of
their existence, I posit both of these—lies in my kind of intuition and
not in these objects in themselves. (CPR, b69)
He then adds a footnote in order to illustrate his remarks:

The predicates of appearance can be attributed to the object itself, in
relation to our sense, e.g. the red colour or fragrance to the rose; but
the illusion can never be attributed to the object as predicate, precisely
that would be to attribute to the object for itself what pertains to it
only in relation to the senses or in general to the subject . . . What is not
to be encountered in the object in itself at all, but is always to be en-
countered in its relation to the subject and is inseparable from the rep-
resentation of the object, is appearance, and thus the predicates of
space and of time are rightly attributed to the objects of the senses as
such, and there is no illusion in this. However, if I attribute the redness
to the rose in itself . . . or extension to all outer objects in themselves,
without noticing a certain relation of these objects to the subject and
limiting my judgment to this, only then illusion arises. (CPR, b69f.
fn.)11
The first thing to note about this passage is that it contains one of the
most explicit and unambiguous statements of the view that the dis-
tinction between appearances and things in themselves is one between
two different aspects of the same object rather than one between two
different kinds of object.12 As Kant makes clear he does not want to
distinguish between appearances on the one hand and numerically
distinct things in themselves on the other, but it is rather the ‘object as
appearance [that] is to be distinguished from itself as object in itself’
(my emphasis). This clarification implies that the claim that material
objects are appearances does not diminish them to ‘mere illusions’, a
11 The passage corresponds to ‘Note iii’ in the section on the Aesthetic in the Prolegomena
(Ak. iv:290ff.).
12 See also the corresponding passage of the Prolegomena (Ak. iv:292–3). For a general de-
fence of the ontological version of the two-aspect reading proposed in the following, see
Rosefeldt (2007), and also Allais (2007, 2015). The following proposal for how to deal
with the ‘neglected alternative’-objection is a modified and enlarged version of the ideas
presented in Rosefeldt (2013).

V
CLOSING THE GAP 195
formulation by which Kant characterizes Berkeley’s position a bit

later (CPR, b71; see also Prolegomena, Ak. iv:374), and by which he
thus describes the view that spatio-temporal objects are mere repre-
sentations or objects that have no being whatsoever outside of the
representations we have of them. Kant stresses that, on the contrary,
his theory assumes that spatio-temporal objects are ‘something really
given’, that is, they not only seem to be distinct from us and given to
us through our senses, as Berkeleian intentional objects would do,
but do in fact have these features. So, the assumption that spatio-
temporal properties are properties of appearances is meant to be

compatible with the fact that they are properties of extra-mental ob-
jects: as Kant puts it in the footnote, they can be correctly ‘ascribed
to the object itself’ (‘dem Objecte selbst beigelegt werden’) as long as
we bear in mind that they are ascribed to this object only ‘in relation
to our sense’. Hence, Kant’s position is not only decisively different
from that of Berkeley, but is also not threatened by the infinite re-
gress objection, for the claim that the human soul is an appearance
does not imply that it does not also exist independently of the repre-
sentations we have of it.
The positive account of the distinction between appearances and
things in themselves that we can extract from the quoted passage in-
terprets this distinction as one between two kinds of properties:
properties that pertain to the object only in relation to epistemic sub-
jects of a certain kind, and which are dependent on the epistemic
constitution of these subjects; and properties that pertain to objects
as they are in themselves, and which are independent of our episte-
mic relation to them.13 What Kant wants to say is that spatio-
temporal properties belong to the first kind of properties, that is,
that they ‘depend on the kind of intuition of the subject in the rela-
tion of the given object to it’. It is clear that by ‘kind of intuition of
the subject’ Kant refers to the assumption that human intuition is
subject to space and time as its a priori forms. In general, to say that
a property P of an object O is dependent on the cognitive constitu-
tion of certain subjects means that O would not have P if the
subjects’ constitution was different in a certain way. Hence, spatio-
temporal properties are dependent on our forms of intuition in that
13 ‘Property’ here is my translation of the German term ‘Beschaffenheit’, which Kant uses in
the passage quoted above. ‘Beschaffenheit’ is a very general and non-terminological expres-
sion to speak about what objects are like, and it is metaphysically less loaded than expres-
sions such as ‘Akzidenz’ or ‘Realit€
at’.

V
objects would not be spatio-temporal if we had other forms of intui-

tion. In the footnote, Kant illustrates his claim about the mind-
dependence of spatio-temporal properties by means of traditional
secondary qualities. He obviously expects the reader to agree that
colour and fragrance are mind-dependent in the sense that a rose
would not be red, or smell a certain way, if our sensory system were
not constituted in such as way as to produce certain sensations when
confronted with it. By analogy, he seems to argue, material objects
would not be extended and have a certain shape and size if we were
not constituted in such a way as to structure the contents of our

manifold sensual representations in a spatio-temporal order. As
Kant puts it, spatio-temporal properties are ‘inseparable from the
representation of the object’ and ‘always to be encountered in its re-
lation to the subject’.
Given what Kant says about spatio-temporal properties and rela-
tions in the quoted passage, he seems best interpreted as assigning to
them the ontological status of what we would nowadays call ‘re-
sponse-dependent properties’.14 Response-dependent properties are
properties of objects that essentially have to do with the way we re-
act to them. They can be understood as the higher-order properties
of having some first-order property that elicits a certain effect in us.
Being poisonous is a typical response-dependent property in this
sense, because to be poisonous just is to have some (e.g. chemical)
property that elicits symptoms of intoxication in such-and-such or-
ganisms. Whereas the first-order (chemical) property, which is the
cause of our reaction, is response-independent,15 the higher-order
property of having some such first-order property with such-and-
such effects on us is dependent on our reaction and our constitution.
If our biological make-up was different and we reacted differently to
the respective first-order property of the poisonous object, the object
might no longer be poisonous for us.
To interpret traditional secondary qualities such as redness and
traditional primary qualities such as spatio-temporal extensions and
relations as response-dependent in this sense means that objects have
14 For a more substantive and detailed account of the following, see Rosefeldt
(forthcoming).
15 Or more accurately: it is independent of the particular response by which ‘being poison-
ous’ is defined, that is, independent of the symptoms of intoxication. Understanding proper-
ties that things have in themselves as properties that they have independently of our mental
reaction to them does not commit us to assuming that these properties are intrinsic or non-
relational in every respect.

V
CLOSING THE GAP 197
them because they bring about certain mental effects in us. In the
case of colours, the effect is a certain colour sensation, and colour
properties themselves can be understood as the higher-order proper-
ties of being such as to elicit such-and-such colour sensations in us.
This property is one that an object has only ‘in relation to a subject’,
and is dependent on the mental constitution of that subject, because
objects would not have it if we did not react to them by having the
respective sensations. Now, Kant seems to think that a similar analy-
sis can be given with respect to traditional primary spatio-temporal
properties. Given what he says about space and time as the forms of

our intuition, this seems reasonable. If objects affect our outer sense,
for example, then we not only react to it by having certain sensations
but also by structuring these sensations according to the form of our
outer sense.16 Although the general spatial form of this structuring is
fully determined by us, it is partly due to the affecting object which
particular spatial structuring of sensations occurs in us (e.g. one that
results in the perception of an object as cubical rather than spheri-
cal). Hence, when we ascribe a particular spatio-temporal property
to an extra-mental object, we are right in doing so as long as we in-
terpret it as the higher-order property of being somehow such as to
elicit in us a specific ordering of sensations according to the forms of
our intuitions. We thus avoid what Kant in the footnote describes as
the illusion of transcendental realism, namely the attribution of ‘ex-
tension to all outer objects in themselves, without noticing a certain
relation of these objects to the subject and limiting my judgement to
this’ (CPR, b70).17
16 In a94, Kant calls this structuring of sensations by the a priori forms of intuition the ‘syn-
opsis of the sense’. It is important to note that this synopsis does not yet yield a mental state
that represents a determinate spatial form or a particular temporal determination. As Kant
makes clear here and in many other passages, this only happens when an additional process
of sensible synthesis is performed by the imagination. A characterization of the mental effect
in non-representational terms is important in order to avoid certain circularity worries with
respect to response-dependent properties (see Rosefeldt forthcoming).
17 Vaihinger writes about the footnote, ‘Die Fussnote, vermuthlich erst nachtr€ aglich und
flüchtig hinzugesetzt, . . . bringt nun in die bis jetzt gewonnenen klaren Ergebnisse eine pein-
liche, ja widerw€artige Verwirrung hinein, welche Kant sich und seinen Lesern durch strenge
Gedankenführung wohl h€ atte ersparen können. Anstatt, wie seine Absicht war, den Text zu
erl€autern, hat Kant ihn nur verdunkelt’ (Vaihinger 1892, p. 488). I could not agree less. The
only slightly unfortunate aspect about the footnote is one further example for an illusion
(namely the illusion that Saturn has two handles), which is disanalogous to the other two
examples. Whereas in this case we can distinguish between the property of appearing to
have two handles and the property of having two handles, Kant’s view in the case of colours
and of spatio-temporal properties is precisely not that objects only appear to have these
properties, but rather that having these properties just is a way of appearing to subjects of a

V
Now, the quoted passage also addresses what we have identified

above as the underlying reason for the incompleteness of Allais’s
and Willaschek’s answer to the neglected alternative objection. The
problem was that it is not clear which conclusions we can draw
from the fact that space and time are forms of intuition for cases
where these forms are not the sole source of a representation (as in
the case of pure intuition), but only one of several causal factors, as
in the case of empirical intuition and the concepts that are acquired
through it, where the affection by a mind-independent object causes
our representations. I have argued that assuming semantic external-

ism with respect to empirical spatio-temporal predicates does not
have the intended idealist implication that these predicates are not
applicable to things in themselves. On the contrary, we have seen
that a general semantic externalism would force us to assume that
empirical spatio-temporal predicates could represent in-itself
properties in cases where our use of a certain concept happened to
be caused by a particular in-itself property and would in fact track
this property—even if we could not know that this is the case. If
my alternative interpretation is correct, however, and Kant took
spatio-temporal predicates to express higher-order properties that
are defined by means of the mental effects that objects have on us,
then this latter consequence can be avoided. For in order to track
those higher-order properties, it is irrelevant what the first-order
in-itself properties are that cause our representations. What we
track is the property to have some first-order property with such-
and-such effects on us, and this higher-order property is accessible
to us even if we cannot know what the first-order property is that is
present in objects when they have the higher-order property. We do
not even have to know whether it is always the same first-order
property that grounds the higher-order property of which we are
aware. On the other hand, the interpretation makes clear why the
certain kind. For this reason I am also sceptical about an alternative reading of the passage
b69f. which Andrew Huddleston has suggested to me, and according to which it is not the
spatio-temporal properties themselves that are response-dependent, but only their exempli-
fication by empirical objects. The only possible sense I can make of the idea of a response-
dependent form of exemplification would be to assume that for an object to exemplify a
certain property in this sense does not mean that the object really has the property, but
rather that it only appears, or seems, to have it. This kind of ‘exemplification’ would indeed
be response-dependent. However, it would result in exactly the understanding of Kant’s ide-
alism that he himself explicitly rules out when he writes, ‘I do not say that bodies merely
seem to exist outside me . . . if I assert that the quality of space . . . lies in my kind of intuition
and not in these objects in themselves’ (CPR, b69).

V
CLOSING THE GAP 199
properties that we express by empirical spatio-temporal predicates

pertain to objects only in relation to us and are dependent on our
own mental constitution. For if our mind was constituted in a dif-
ferent way, and we had different forms of intuition, then objects
would no longer have the higher-order properties of giving rise to
the particular orderings of sensation that they in fact bring about in
us. And noticing this mind- and response-dependence of spatio-
temporal properties is enough to see that they are not had by things
in themselves. This brings us back to the objection of the neglected
alternative.

The scenario that we found still not ruled out at the end of the last
section was one in which space and time are forms of intuition but,
although space and time do not themselves exist as mind-
independent entities, things in themselves nevertheless have spatio-
temporal properties and stand in spatio-temporal relations. The ob-
jection that Kant neglected this scenario in the Transcendental
Aesthetic is based on a certain understanding of what Kant took to
be his insight about spatio-temporal properties there, namely, that
they only pertain to one kind of entity—appearances—but not to
another one—things in themselves. The objection then reads that al-
though we can only know that they pertain to the one kind of object
and do not know whether they also pertain to the others, this is not
enough in order to rule out that they in fact pertain to both.
Given what we have learned from the passage in b69f., this way of
putting things is mistaken. For, as we have seen, Kant’s distinction be-
tween appearances and things in themselves is not one between two
different kinds of object, but rather one between the way objects ap-
pear to us and the way the very same objects are in themselves, a dis-
tinction that in turn is best understood as one between two kinds of
property, response-dependent and response-independent ones. Hence,
the supposedly neglected alternative should not be described as one in
which appearances as well as things in themselves have spatio-
temporal properties; rather, the related objection would have to read
as follows: ‘Kant claims that space and time are mere forms of intui-
tions, and that spatio-temporal properties and relations are response-
dependent properties and do not pertain to objects in themselves. But
all he is justified in claiming is that we only know that these properties
are response-dependent, which, however, is not enough to rule out
that they are in fact also mind-independent and also pertain to objects

V
in themselves.’ That is to say, the supposedly neglected alternative

would have to be described as (na**B) rather than as (na**A):
(na**A) (i) Space and time are forms of our intuition and are not
things in themselves; (ii) not only appearances, but also
things in themselves have spatio-temporal properties and
stand in spatio-temporal relations.
(na**B) (i) Space and time are forms of our intuition and are not
things in themselves; (ii) spatio-temporal properties and
relations not only pertain to objects as they appear (that

is, are response-dependent), but also to them as they are
in themselves (that is, are also response-independent).
Once we have formulated the objection in this way, we see that it is
no real threat, because the supposed neglected alternative is not a real
possibility. The alternative scenario would have to be one in which a
property that is in fact response-dependent is also response-
independent. However, this is impossible. Response-dependent prop-
erties are essentially response-dependent, and cannot also be
response-independent. If the property of being cubical just is the prop-
erty of being such as to elicit a certain kind of ordering of sensations
in us, then this very property could not be response-independent. By
its nature it is such that an object could not have it ‘in itself’. To ac-
cuse Kant of having neglected an alternative here would be like accus-
ing someone who has claimed that being poisonous is a response-
dependent property that she has not excluded the possibility that be-
ing poisonous might also be a response-independent property and
that things might not only be poisonous for such-and-such beings but
also be poisonous in themselves. This would of course be absurd, for
there simply is no such thing as being poisonous without being poi-
sonous for someone or something. Similarly, if Kant is right about the
nature of spatio-temporal properties, then there simply is no such
thing as having such a property without having it in relation to some-
one with space and time as their forms of intuition.
Of course, denying the response-dependence of spatio-temporal
extension is far less absurd than denying the response-dependence of
a property such as being poisonous. In contrast to the latter case, it
is not a conceptual truth that extension is response-dependent. On
the contrary, philosophers before and after Kant have taken it to be
the paradigm of a property that objects have independently of our

V
CLOSING THE GAP 201
minds, and hence Kant has to present an intricate and controversial

argument in order to show that these philosophers are wrong. If we
look for an analogy from the empirical realm for Kant claims about
extension, we should think of properties of which people have first
overseen, and then later found out, that they depend on certain pa-
rameters. Properties such as taking place at noon, moving upwards,
or happening simultaneously might be good examples. In each of
these cases people first thought that the mentioned properties per-
tained to objects or events in themselves before they later discovered
that they do so only relative to a certain further parameter such as a

time zone, a gravitational field, or, as is shown in the special theory
of relativity, an observer’s reference frame. However, once we have
found out that these properties are dependent on these parameters, it
is no longer an open question whether objects could also have them
in themselves. For there simply is no taking place at noon indepen-
dent of any time zone (on the sun, for example), no upwards motion
in absolute space, and no non-relative simultaneity. Analogously, if
Kant was right at all that the spatio-temporal properties of objects
depend on the constitution of our sensibility, then these objects sim-
ply could not have these properties independently of our sensibility.
That is not to say, of course, that it might not turn out that spatio-
temporal properties are in fact not response-dependent. This sce-
nario simply corresponds to the all too well known epistemic possi-
bility that a philosophical theory is wrong. However, it does not
correspond to the supposedly neglected alternative which Kant
should have addressed in his theory. The neglected alternative objec-
tion deals with a scenario in which Kant is right in what he says
about the status of the spatio-temporality of appearances and only
wrong about the non-spatio-temporality of things in themselves. In
order for this scenario to be actual, spatio-temporal properties
would have to be response-dependent and response-independent at
the same time. Having neglected this ‘alternative’ is not something
he should be blamed for, since it is not an alternative, but an
impossibility.
IV
Conclusion. The neglected alternative on which the objection against

Kant’s central argument for transcendental idealism was built was

V
supposed to consist in a scenario in which space and time are forms

of intuitions, but in which they nevertheless also exist independently
of us and things have spatio-temporal properties and stand in spatio-
temporal relations in themselves. Willaschek’s and Allais’s argument
has shown that, given Kant’s account of intuition, the space and the
time that is the object of our a priori intuition could not be a thing in
itself. We have seen that, with a little help from a rather modest
form of semantic externalism, Kant could also directly conclude that
space and time simpliciter could not exist as things in themselves,
and hence rule out a scenario in which his claims about our forms of

intuition are combined with a Newtonian account of mind-
independent space-time. Moreover, a careful interpretation of Kant’s
distinction between appearances and things in themselves, and of his
account of the nature of spatio-temporal properties as response-
dependent, made clear that, once we have shown that objects exem-
plify spatio-temporal properties and relations as appearances, we
have ipso facto ruled out that these properties could also be exempli-
fied by things in themselves. And hence it is also not possible to com-
bine a view of space and time as forms of intuition with a relational
theory of mind-independent space-time. There is no gap left to be
closed.
Institut für Philosophie

Humboldt-Universit€at zu Berlin
Unter den Linden 6
10099 Berlin
Germany
References
Allais, Lucy 2007: ‘Kant’s Idealism and the Secondary Quality Analogy’.
Journal of the History of Philosophy, 45(3), pp. 459–84.
2010: ‘Kant’s Argument for Transcendental Idealism in the Transcen-
dental Aesthetics’. Proceedings of the Aristotelian Society, 90(1),
pp. 47–75.
2015: Manifest Reality: Kant’s Idealism and His Realism. Oxford: Ox-
ford University Press.

V
CLOSING THE GAP 203
Bird, Graham 2006: ‘The Neglected Alternative: Trendelenburg, Fischer, and

Kant’. In Graham Bird (ed.), A Companion to Kant, pp. 486–99. Oxford:
Blackwell.
Gesang, Bernward (ed.) 2007: Kants vergessener Rezensent: Die Kritik der
theoretischen und praktischen Philosophie Kants in fünf frühen Rezensio-
nen von Herrmann Andreas Pistorius. Hamburg: Meiner.
Jakob, Ludwig Heinrich von 1786: Prüfung der Mendelssohnschen Morgen-
stunden. Leipzig: Johann Samuel Heinsius.
Kant, Immanuel 1900– : Kants Gesammelte Schriften, ed. Koniglich Preu-
ischen, later Deutsche, then Berlin-Brandenburg Akademie der Wissen-
schaften. 29 volumes. Berlin: de Gruyter.

Rosefeldt, Tobias 2007: ‘Dinge an sich und sekund€are Qualit€aten’. In Jürgen
Stolzenberg (ed.), Kant in der Gegenwart, pp. 167–209. Berlin and New
York: de Gruyter.
2013: ‘Subject-Dependence and Trendelenburg’s Gap’. In Stefano Bacin,
Alfredo Ferrarin, Claudio La Rocca, and Margit Ruffing (eds.), Kant und
die Philosophie in weltbürgerlicher Absicht: Akten des XI. Internationalen
Kant-Kongresses, vol. 2, pp. 755–64. Berlin and New York: de Gruyter.
forthcoming: ‘Transcendental Idealism and the A Posteriori Contents of
Experience’. In Karl Schafer and Nicholas Stang (eds.), The Sensible and
Intelligible Worlds: New Essays on Kant’s Metaphysics and Epistemol-
ogy. Oxford: Oxford University Press.
Sassen, Birgit (ed.) 2000: Kant’s Early Critics: The Empiricist Critique of the
Theoretical Philosophy. New York: Cambridge University Press.
Specht, Andrew F. 2014a: Kant and the Neglected Alternative. Ph. D. disser-
tation, Syracuse University.
2014b: ‘F. A. Trendelenburg and the Neglected Alternative’. British
Journal for the History of Philosophy, 22(3), pp. 514–34.
Vaihinger, Hans 1892: Commentar zu Kants Kritik der reinen Vernunft,
Band 2. Stuttgart, Berlin and Leipzig: Union Deutsche Verlagsgesellschaft.
Willaschek, Marcus 1997: ‘Der transzendentale Idealismus und die Idealit€at
von Raum und Zeit. Eine “lückenlose” Interpretation von Kants Beweis in
€
der “Transzendentalen Asthetik”’. Zeitschrift für philosophische For-
schung, 51, pp. 537–64.

V

Closing The Gap

Uploaded by

Copyright:

Available Formats

Closing The Gap

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Closing The Gap

Uploaded by

Copyright:

Available Formats

Meeting of the Aristotelian Society held at Senate House, University of

London, on 22 February 2016 at 5:30 p.m.

IX—CLOSING THE GAP: A NEW ANSWER TO AN OLD

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

The Gap. Kant’s transcendental idealism consists in the claim that

C 2016 The Aristotelian Society

world is imposed on it by the structure of our sensible faculty (a26/

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

2 For an overview of the literature on the topic, see Specht (2014a).

C 2016 The Aristotelian Society

Ludwig Heinrich von Jakob’s Prüfung der Mendelsohnschen

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

Jakob dismisses this worry quickly and on rather questionable

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Closing the Gap 1: Kant’s Account of Intuition. Willaschek and

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

There seem to be two possible interpretations of what Allais is saying

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

not with a particular concept. If we accept the externalist account of

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

things in themselves, as the Newtonian conception has it.8 However,

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

would represent P. This is absurd, since empirical concepts are sup-

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

could never be used successfully in order to characterize things inde-

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

Closing the Gap 2: Kant’s Distinction Between Appearances and

C 2016 The Aristotelian Society

we attribute to them, are always regarded as something really given,

He then adds a footnote in order to illustrate his remarks:

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

formulation by which Kant characterizes Berkeley’s position a bit

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

objects would not be spatio-temporal if we had other forms of intui-

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

Now, the quoted passage also addresses what we have identified

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016

C 2016 The Aristotelian Society

properties that we express by empirical spatio-temporal predicates

Downloaded from http://aristotelian.oxfordjournals.org/ by guest on November 8, 2016