Philosophy and Phenomenological Research, 16 (1980), 93–105
INGARDEN vs. MEINONG ON THE LOGIC OF FICTION
At the tum of the present century one problem, the so-called
problem of the existence of the world, lay at the center of philosophical argument. The dominance of those who took the idealist side
in this dispute, both in Germany and in England, was then rapidly
being brought to an end with the works of Brentano, Meinong, Frege,
Moore, and Russell, all of whom can be seen as having expressed, in
their different ways, reactions against one or other form of Hegelian
idealism. Yet this problem is no longer among the vital concerns of
present-day philosophers. This is not because the realists have simply
won the battle, such that the whole idealist phase in the history of
philosophy from Kant to Bradley could justifiably be discounted as an
unfortunate detour from the high road of Aristotle and the
Scholastics, Leibniz, Frege, and 20th century realism. But certai_nly it
is true that philosophers no longer pose any direct question concerning the (mode of) existence of the real world. One might perhaps say
that a form of empiri"cal realism has come to be taken for granted on
all sides, whether or not this. is regarded as consistent with some kind
of idealist position on a higher ("transcendental") plain.
We might expect, therefore, that there is little of value to current philosophy to be found in a work1 devoted to this issue, devoted,
indeed, to the painstaking establishment of the validity of the realist
position through some fifteen hundred pages of argument. The work
in question was written by a Polish phenomenologist and student of
Husserl, Roman Ingarden -at a time when the idealism-realism
dispute had already ceased to occupy the attentions of those
philosophers now recognized as the forbears of the analytic movement
- as part .of Ingarden's response to his teacher's move away from the
realisf!l of the early Logical lnvesti"gatz'ons to the transcendental
1
Der Streit um die Existenz der Welt, in three volumes, the second in two
parts: vol. I, Existenzialontologie, Tiibingen, 1964; vol. II, Formalontologie, part
1, Form und Wesen, part 2, Welt und Bewusstsein, both Tiibingen, 1965, vol. III
(posthumously prepared for publication; incomplete). Uber die kausale Strukter
der realen Welt , Tiibingen, 1974. Original Polish version of vols. I and II, Krakow,
1947-48. Partial English translation of vol. I under the title Time and Modes of Being, Springfield, III., 1964.
93
94
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
idealist position which we find in Husserl's later works.
We are all familiar, howeveT, with important philosophical texts
whose significance resides elsewhere than their authors had supposed.
And Ingarden's work has, I want to suggest, precisely this quality: for
it is only the by-products of Ingarden's work which promise to yield its
true value for present-day philosophy. In particular it has value for
the new, reformed ontology which has begun to flower in recognition
of the possibilities for hitherto unattainable precision in this field,
through the exploitation of the resources of formal logic. 2 For the
book contains a mass of individual ontological insights ordered within
what must be the most detailed - and the most adequate - on·
to logical system in the whole literature of philosophy. Thus the work
not only considers ontological problems raised by the notion of time
(the modes of being of and the relations between past, present, and
future existents). problems of causality. problems of the nature of
relations and properties, states of affairs, events, actions. processes,
and higher-order wholes (sets, heaps, as well as biologically and
socially unified wholes). It deals also with specific ontological issues
raised by the tradition: e.g., the bundle theory of objects (as sets of
properties), the 'fonn'·'matter' relation, the concept of transcendence
(e .g .. of the external world to individual consciousnesses), and the
nature of the ontological discipline itself. And it deals further - when
considered in conjunction with lngarden's other works in this and
related fields - with problems in the theory of reference, the nature of
meaning, the functions of language, and also with the issues raised by
works of fiction and by the ontological correlates (if any) possessed by
such works in the realm of 'purely intentional objects.'3
It is these last·memioned issues which I wish to consider in the
prei;ent paper. I shall sketch an Ingan.lenian approach to the on·
tology or, when appropriate formal machinery has been introduced,
to the logic of fiction, comparing this with the Meinongean ap·
2
, ,
Work in '.ormal ontology goes back to Frege. Russell. and the Wittgenstein of
~e
I ractatus: fhe most 1m~orta
current work includes e.g .. investigations in1t1ated by Davidson on the logic of events, cf. al50 N. B. Cocchiarella'a studies of se·
con~
order logi.c and of the discpl~
of formal ontology as such. For references see
my ~n
Essay Jn F~rmal
Ontology. Graze r Pl11losoph1sche Siudit•11. VI. 1978.
See lngarden ~Das
lter~isch
Kumtwerk, Halle. 1931; English translation,
Evan~.to'
1973;... he JUSt·menuoned aspects of lngarden's thought are dealt with by
G. Kung m his lngarden on Language and Ontology," Analecta Husserliana II
1972, 204·17.
, .
INGARDEN VS. MEINONG ON THE
LOGIC OF FICTION
95
proach to the same group of issues.
The parallels between Ingarden and Meinong are many. Both
fall within the sphere of Brent'ano-influenced philosophers: Meinong
directly and Ingarden through his contacts with Husserl and Twar·
dowski. The two philosophers share many problems, especially in the
field of ontology (Ingarden in fact makes frequent reference ·to
Meinong's works, especially in his earlier writings). Both were, like
Frege (but unlike the later Husserl), realist philosophers, though Ingarden's realism was, as I hope to show, the more subtle of the two.
And most importantly perhaps, both saw the need for a nonreductivist ontology, that is, for an ontology which would not seek to 'explain away' or 'paraphrase out' our talk about entities of certain kinds
in terms of talk about other, more favored kinds of entities (e.g.,
material objects). Thus neither philosopher suffered from the prejudice in favor of real, concretely existing objects _which has been so
prevalent amongst analytic philosophers up until the present decade,
and nor did they share the related prejudice in favor of the natural
sciences at the expense of disciplines of other types (social science,
say, or literary theory). In consequence they both shared the recognition that there are certain kinds of deliberate, critical concern with
literary works within which fictional objects as such play a central role
as referents. They saw-in modem parlance - that works of literary
criticism, of literary history, etc., involve quantification over, e.g.,
fictional characters as individual objects, just as, say, works of electron physics involve quantification over (are ontologically committed
to) electrons. And they saw also that even outside such 'theoretical'
contexts we meet the same kind of ontological commitment to fictional characters as intersubjectively accessible objects of our acts.
For there is no way in which we can achieve a faithful reading ofa fic-
tional work withou,t presupposing from the start that it is correlated
with its own specific fi'eld of fictional characters.
What has not been seen by proponents of reductivist theories of
fiction is that this presupposition on the part of a given appropriately
quali.fied subject cannot be regarded merely as some sor' of working
hypothesis, eventually to be dismissed as something insubstantial
(spurious, illusory) in comparison with the "genuine" product of a
particular reading ·(conceived, at worst, as the reader's own enjoyment). To see why this is so, a further comparison with the case of
physics may perhaps be helpful: It was with the aid of the
hypothetical conception of the atom as an absolutely simple, cor·
puscular entity that specific research programs were conceived, pro·
96
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
grams which led, eventually, to a more adequate, noncorpuscular
atomic theory. Now in this case the purported ontological correlates
of the original 'working hypothesis' (the 'corpuscles') can be dismissed
as having enjoyed a purely heuristic significance. But this is not so in
the case of the ontological correlates of those on-going 'hypotheses'
of fiction. For the latter
which are projected in our readings of wor~
are (in faithful readings) absolutely secure against the kind of ontological overt~nig
characteristic of scientific advances: there is no
way in which we could reasonably claim to have 'discovered' that
characters of a particular work fail to satisfy the canons of ontological
rectitude. Such a claim could rest only on an assumption taken out
prior to any individual commune with actual fictional texts, rest, that
is to say, only upon just that kind of metaphysical presupposition
which it is the philosopher's duty to reject.
For the same reasons both Ingarden and Meinong allow a place
in their ontologies also for numbers and mathematical and abstract
objects in general, for institutional entities such as universities and
nation states (as entities founded upon but distinct from the totality
of individual subjects, buildings, etc., which support them), and for
certain other kinds of nonreal or nonmaterial entities.
Where Meinong and Ingarden differ is that Meinong allows no
place for the crucial characteristic of fictional objects that they are
created at determinate points in time (i.e., by the sentence-forming
acts of the author of the appropriate work). Meinong defends instead
a wholly counterintuitive picture of fictional objects as having been
'picked out' from an all-inclusive domain:' Ingarden, in contrast,
defends the much more reasonable position according to which fictional objects have an ontological status exclusively as the correlates
of determinate networks of conscious acts on the part of the author of
the appropriate work and of subsequent :readers. 5 We might say
• Cf. Meinong's statement of his ~ princle
assumption':
.. .
of the unlimited freedom of
In regard to every genuine or, so to speak, ordinary determination of so· being, it is in rny
power, according to the principle of unlimited freedom of assumption, to pick out-by
means of adequate intention-an entity which in fact has that deterrnina.tion of so· being.
Uber Miiglichkeit und Wahrscheinlichkeit, Leipzig, 1915, as trans. by R.
Grossmann, Met"nong, London , 1974, p. 160.
5
Thus a fictional object has ontological status only in the time subsequent to
the laying down of an appropriate 'access route' - i.e. , through the sentences of the
relev.ant work, sentc~
which will determine the acts of the appropriately
qualified reader as havmg an ontological correlate with a precisely determinate
structure.
lNGARDEN VS. MEINONG ON THE
LOGIC OF FICTION
97
therefore that where Meinong defends a 'classical' logic of fiction. in volving commitment to a 'paradise' of superfluous entities having no
more than ornamental value, the logic of fiction inherent in Ingarden's work is sharply constructivist, having affinities with the
ontology underlying the constructivist abstraction theory developed
by thinkers such as Lorenzen and Thiel. 6
A correlate tendency in the Meinongean approach to the ontology of . fiction is that it tends to underestimate the radical
dissimilarity in ontological structure between real and fictional objects. Perhaps the most important such difference turns on the quite
peculiar ontological incompleteness which is enjoyed by the latter.
Certainly it is true that in the case of real objects (e.g., our ·contemporaries) and historical figures our knowledge is always incomplete,
owing to the fact that it rests on only a finite quantity of information
where the objects of our knowledge have (or had) an infinitely complex array of determinations. But the incompleteness involved here is
purely epistemological: if we know only that Harold the Nth lost an
eye in the battle of X, but not wh£ch eye, than we do not suppose that
after the battle Harold himself was ontologically structured in such a
way that the missing eye was indeterminately neither right nor left.
To countenance such an ontological indeterminacy in objects of the
past would involve too radical a warping of our frame of reference to
real objects as this is extended from the field of coexisting objects into
the field of that which no longer exists. With fictional objects however
we have to come to terms - at least in our theory - with just this kind
of ontological incompleteness. 7 But it is important that this
characteristic of incompleteness of fictional objects raises its head exclusively on the level of theory. This is because every experience of objects which we undergo is characterized by its own (usually
epistemological) incompleteness: even the most adequate perception
8
Cf. e .g., Lorenzen's "Equality and Abstraction," Ratio, 4, 1962, pp. 85·90,
and Thiel's "Gottlob Frege: Die Abstraktion," in Grundprobleme der grossen
Philosophen: Philosophie der Gegenwart I, ed. J. Specht, GOttingen, 1972, pp.
9·44, and in Studienzu Frege, M. Schirn, ed., Stuttgart, 1976, vol. I , pp. 243-264.
Note that a thorough-going abstraction theorist would argue that there is only one
kind of "ticket" for the introduction of new objects, namely an appropriate
equivalence relation, and that the objects introduced are not, in any· case, to be
taken ontologically seriously.
7
This is because real objects, e.g., real human beings, and the fictional
characters modellea after them have the possibility of an infinite number of determinations; not all of these determinations can be "filled in'! on the basis of only a
finite number of sentences such as are found in a work of fiction.
·
98
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
of a real object, for example, is always one-sided. Thus we can do no
other, in our pretheoretical commune with objects of any kind, than
ascribe all inadequacies in our knowledge to the side of evr-p~nt
epistemological incompleteness. As we shall see, Meinong's ontology
too allows a place for 'incompleteness' on the side of the fictional objects themsefves. But for Meinongeans the completeness/incompleteness dichotomy is not recognized as a radical opposition; it is seen
rather as a spectrum, and a spectrum, moreover, which is such that
every position is filled, atemporally, by one or other potential 'object
of assumption:
·
Since Meinong fails to take seriously the fact that fictional objects, like institutional objects of other kinds, are created in time, this
implies that he is committeed to an ontology of objects which is clumsily over-generous. Indeed it may not be too much to say that the
massive over-generosity of Meinong's own Gegenstandslehre has had
serious negative consequences for the subsequent history of
philosophy, for the excesses of Meinong's own nonreductivist theory
of entities became associated in the minds of philosophers wit h the
discipline of ontology as such. Thus it came about that reductivist
theories of various kinds came to be accepted as the notm throughout
the philosophical community.
What is worse, perhaps, is that even against this over-rich
theoretical background Meinong is still incapable of coming to grips
with quite central problems in the ontology of fiction. This can be
seen particularly clearly on the basis of an elegant formulation of the
Meinongean approach to entity theory recently developed byTerence
Parsons.8 According to Parsons we are to conceive all objects,
whether real or fictional, possible or impossible, as represented by
clusters of properti'es. Thus
Hannibal
cluster (or set) of Hannibal's properties
Jimmy Carter
cluster of Jimmy Carter's properties
Hamlet
cluster of Hamlet's properties (i.e., as determined by
Shakespeare's text)
Meinong's golden mountain
cluster {golden, mountain}
Meinong's round square
cluster {round, square}
The absolutistic generosity of Meinong's ontology is to be accounted
=
=
=
=
=
a Se·~
"A Prolegomenon to Meinongian Semantics," journal of Philos<>phy,
71. 1974, pp. 561-580 and "A Mein<mgian Analysis of Fictional Objects," Grazer
Philosophische Studien, I, 1975, pp. 73-86.
INGARDEN VS. MEINONG ON THE
LOGIC OF FICTION
99
for by the fact that objects are associated with (identified with?) every
cluster of properties-not merely with those which are somehow actualized in the real world or (e.g.,) in one or other concretely existing
work of fiction. That is to say, the given Jist is to be regarded as having been continued in some systematic way until all combinations of
properties on the right-hand side have been exhausted, such that
each would then correspond to an 'object' on the left. Such 'objects'
then fall into two classes: (i) those which correspond to 'complete'
clusters of properties, 9 some of which would exist ·in the real world,
some of which (say, the cluster correlated with a green-haired j immy
Carter) would fall short of existence of any kind .. (ii) Those which correspond to clusters which are £ncomplete - and the Meinong- Parsons
theory is at least in this one respect adequate as an ontology of fiction,
since it satisfies the requirement mentioned above, that all objects of
fiction should be recognized as suffering from an intrinsic ontological
incompleteness.
There are other respects, however, in which the theory shows
itself to be inadequate to the job in hand. For not only is the underlying ontology committed, from the point of view of a theory of fiction,
to far too many objects - it is also, for particular cases, committed to
far too few: that is to say, the framework is insufficiently delicate to
capture certain nuances characteristic of fictional works. This can be
shown by means of a very simple example: 10 consider the (one
sentence) novel°:
Schaub was a dragon who had ten magic rings.
According to Parsons' rules we have
ring 1 = cluster { r£ng, owned by Schaub, mag£c}
but then also
ring 2 = cluster {ring, owned by Schaub, magz"c}.
That is to say: ring 1 and ring 2 are identical - and similarly of course
for the cases of rings 3, 4, 5, . . . .
This failure to account for the particular diversity involved is not
merely a result of some artificiality in the example chosen: it results,
rather, from the crudity of the notions of property and list/ cluster/set
9
In a sense of 'completeness' parallel to that which was introduced above. For
reasons expressed in note 11 below we are reluctant to assume a perfect identity of
the two notions.
10
.
Adapted by Parsons from an example by David Lewis; see "A Meinongian
Analysis," p. 82.
100
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
which form the basis of Parsons' theory. 11 This in turn rests on the
fact that the theory has been developed within a wholly artificial settheoretical framework where the overriding criteria are those of
mathematical elegance and simplicity-not criteria of phenomenological (nor ontological) adequacy to any sphere in which the theory
would eventually be applied. Note, however, that we are not criticizing the u~e
of formal methods as such: an adequate ontological theory
will have a complexity so great as to be expressible only by appeal to
some correspondingly adequate technical language. But this complexity will itself have a certain quasi-empirical nature, to the extent
that ontology comes to resemble certain scientific theories (such as,
say, chemistry or theoretical ·biology) more nearly than those
branches of mathematical logic hitherto developed.
One further point in which Parsons' formalization of the
Meinong.ean approach approximates to an adequate theory turns on
the fact that, as Ingarden was the first to point out, 12 fictional objects
possess a quite peculiar double property-structure. This is reflected in
the fact that we encounter two quite different sets of statements about
fictional characters, with two quite different kinds of truth-behavior.
The first set, containing what we shall call A-st~emn,
consists of
metalevel statements about fictional characters as such, typical of
those which are to be found in critical and theoretical texts. (Examples would be: 'Sherlock Holmes was presented by Conan Doyle in
novel N as a violin-playing detective'; 'Sherlock Holmes is an inadequately developed character'; etc.). B-statements are statements to be
found in the novels themselves or - within certain limits - statements
deducible from the latter by appeal to logical and scientific laws. 13
Now our account of the structure of fictional objects must reflect (i)
the fact that A-statements may be true, unconditionally, and that
such statements satisfy the law of excluded middle; and then also (ii)
that we can acknowledge B-statements as in some way 'correct' or 'incorrect' only if we interpret the predicates involved more or less
11
This is something which Parsons promises to refine in a future paper (cf. Loe.
cit., p. 76), though the sketches he gives of his intended refinements suggest that he
will continue to associate (ontological) properties too closely with (linguistic)
predicates. Note that Meinong's own original theory of properties (or better: of indi~ual
accidents) was much more subtle (cf. Grossmann, op. cit. Ch . I), though
correspondingly difficult to capture within a formal ontological theory.
12
Cf. Das literarische Kunstwerk, p. I 20ff et passim.
13
See]. Woods, The Log£c of Fiction, Paris and the Hague, 1974, §13, for an
indication of the kind of limitations involved.
INGARDEN VS. MEINONG ON THE
LOGIC OF FICTION
101
metaphorically. Thus consider the B-statement: 'Sherlock Holmes
was a detective,' properly acknowledgeable as correct, even though
the fictional object designated by 'Sherlock Holmes' was not a detective in the strict sense of the term-since only flesh-and-blood human
beings may qualify for the status of detectivehood. We must also note
that B-statements fail to satisfy the law of excluded middle, in the
sense that there are incorrect B-statements which are such that their
negations, too, are incorrect. (Consider, e.g., the statement-pair:
'Hamlet was left-handed,' 'Hamlet was not left-handed'). This failure
corresponds to the ontological incompleteness of fictional objects
mentioned above as a factor in both the Ingardenian and the
Meinong/Parsons approach to the ontology of fiction.
(i) and (ii) together imply that whilst we may associate with each
true A-statement concerning a given fictional object some determination intrinsic to the object itself, B-statements-precisely the statements which determine the 'properties' of Parsons' cluster theorymust be dealt with in a quite different way. Ingarden's suggestion is
that we develop a conception of fictional objects as radically distinct,
in their property behavior, from real objects of the material world.
Fictional objects are to be conceived, he argues, as possessing two
distinct ranks of properties, or rather, one rank of propert£es in the
strict sense, corresponding to true A-statements, and a second rank of
merely ascribed characteristz'cs, corresponding to correct B-statements. Any resultant incompatibility is made harmless by the fact
that members of each group are acceded to within distinct contexts: it
is only on the theoretical level that we accede to rank A properties,
where it is exclusively rank B characteristics which ho•d our attention
during any actual reading of the work.
Now Parsons too, as we said, is committed to a double structure
view of fictional objects. But note how inadequate is his double structure theory in comparison with that which has been developed by Ingarden. Parsons' two ranks of properties are: set-theoretical properties (is included t'n, is nonempty, intersects with, and so on), which
correspop.d to our A-properties, being seen as possessed by the fictional objects themselves; and properties determined by the sentences
of the novel in question ( our characteristics). Note first of all that
this double structure is possessed also by real objects in Parsons'
theory, for real objects too are identified as particular sets. The
radical heterogeneity between real and fictional objects is therefore
masked. In the Ingardenian framework , in contrast, A-proet~s
consist precisely of those formal and intentional ('nonexistence entail-
=
102
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
ing') properties14 characteristic of fictional objects as such: is a fic1883, is presented in
t£onal object, was created by Conan Doyle
work Was short and fat, etc., and the radical difference between the
manner in which material properties are associated with fictional objects and· the manner in which such properties are possessed by real
material objects proper is thereby brought into light, something
which is quite excluded within the Meinong/Parsons framework.
Underlying this tendency of the Meino-ngean approach to mask
the ontological differences between real and fictional objects is a
conception of fictional objects as possible entities. Thus Doyle is seen
as describing, in his Holmes novels, a world which is a· possible alternative to this, the actual world in which we live. Holmes differs ontologically from e.g., Jimmy Carter only in that Holmes happens not to
exist in the actual world: he is a pure possibile. This is not to
suggest that Parsons logic of fiction is cons~te
with a possibilia ontology for fictional characters. For it seems clear- on the assumption
that objects can be identified as sets of properties at all- that only
those sets which are complete, in the sense determined above, can
qualify as either possible or actual realia. A Meinongean who wanted
to develop a possibilia ontology of fiction against the background of
Parsons' theory would therefore find himself committed to a view of
fictional objects as intrinsically complete, determined in every respect
down to the lowest possible differences, on the basis of a finite
number of sentences in a literary workl15
The possibilia vie~
· suggests a conception of fictional objects as
cows in a distant field, works of fiction being regarded as telescopes
which enable us to gain access to and distinguish features of these
peculiar groups of neighbors. Ingarden's work however, which has
been supported by recent arguments of Kripke and Woods, points to
the absolute untenability of this conception of fictional objects. The
arguments involved can be most easily expressed, perhaps, if we appeal to the notion of admiss£ble trans£tz'on from one object, state, or
event to another related object, state, or event. The following present
themselves. on the basis of our present knowledge of the laws.of (e.g.,)
zn
14
Cf. Cocchiarella, "Existence entailing attributes, modes of copulation and
modes of being in second order logic," Noils, 3, 1969 , pp. 33·53.
·
16
See Woods, "Fictionality and the Logic of Relations," Southwestern journal
of Philosophy, 7, 1976 , pp. 51-63, and ch. 2 of The Logic of Fictz'on.
INGARDEN VS. MEINONG ON THE
LOGIC OF. FICTION
103
physics, as valid examples of admissible transitions:
Jimmy Carter (alive)
__.. Jimmy Carter as corpse
Jimmy Carter as youth
_. Jimmy Carter as trappist monk
this drop .of water
__.. a totality of free oxygen and
hydrogen molecules, etc.
The following, on the other hand, present themselves as transitions
which are inadmissible:
Prime Minister
Gladstone
....,.. a transcendental number
this table here
_. that table there
this rabbit here
~
a tiger, a bar of gold, ... , etc.
Note that these transitions are divided into admissible and inadmissible ontologically, and not as a reflection of our knowledge at any
given stage; for even though appeal must be made to that knowledge
in drawing up an exemplary list of cases of either kind, the line between the two kinds of cases is something to be discovered by our experiments, and not created. Thus it can be that a transition which
seems, at one state, clearly inadmissible, say
John's heart beating
__.. John's heart beating
inside John's body
inside Tom's body,
may reveal itself as-in itself-admissible as a result of advances in
medical technique (but only in those cases where both John and Tom
are alive subsequent to the putting into practice of those advances).
The argument concerning the pur~oted
status of fictional objects as possibilia should now be clear. 1 Objects are possz'ble only if
they occur as the second term of some admissible transition, where
the first term is an object in this our actual world. Inspection reveals
that this account is satisfied by all of those objects (states and events)
which we should normally accept as possible existents (for example,
Jimmy Carter's fourteenth child). Now clearly
the flesh-and-blood
a fictional character, created
product of human
_.. as a result of the actions of
sperm and egg
author, printer, publisher, etc.
is an inadmissible object-to-object transition: the most that we can
allow is that some flesh-and-blood human being and some fictional
character could be namesakes, never that they could be identically
18
It rests on intuitions concerning our use of designatory expressions developed
by Kripke in his "Naming and Necessity," in Semantz"cs for Natural Languages,
Davidson and Harman, eds., Dordrecht, 1972, 253-355, 763-69; cf. also Woods,
op. cit. and-from a different, no longer purely linguistic point of view- Ingarden,
Df!r Streit um die Existenz der Welt, vol. I, § 11.
104
PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH
the same object. One consequence of these arguments is that fictional
objects, being excluded from che realm of possibilia (and therefore, a
fortiori, from the realm of actual existents) are necessarily nonexistent - something which must be carefully distinguished from any
claim to the effect that fictional objects necessarily lack any ontological status whatsoever.
It is not our purpose here to present a complete account of In·
garden's own positive views concerning the ontological status of fie·
tional objects. These are presented in great detail in Ch. IX of The
Controversy over the Existence of the World (see note 1 above) and in
his book on The Li"terary of Art: An Investigation on the Borderlines
of Ontology, Logic and the Theory of Uterature (see note 3), a work
which contains, incidentally, a 200 page centerpiece on the theory of
meaning, .many of whose insights have since been rediscovered by
analytical philosophers of language. Let it suffice here to say that a
logical formulation of the Ingardenian ontology of fictional objects
would possess-in contrast to the Meinong/Parsons system discussed
above-the following components:
(i) It would be constructivist. That is to say, fictional objects-and
other kinds of institutional object (e.g., works of art in general, na·
tion states, etc.) - would be indexed temporally and, where relevant,
by associated work. This is in reflection of their being created ex
nihilo, by particular acts of an author (or, in other cases, e.g., by acts
of parliament, acts of war, etc.).17 No sentence involving the use of a
proper name of a fictional character can express a true proposition in
times prior to the temporal index of that character. Note that whilst
the quantificational machinery associated with institutional object
names and variables must, for an Ingardenian, be constructivist, the
logic involved may well be embedded in a classical quantification
theory for real object names and variables. Syntactic 'dualism' would
thus reflect the ontological dualism involved in supposing that whilst
fictional objects are created by acts of consciousness, real existent objects of the material world are ontologically independent of all such
acts.
(ii) An object is distinguished from its stock of properties (i.e., in the
case of a fictional object, from its double stock of properties and
17
Jt is this structure of indices and not. e.g. , the laws of set theoretic identity
or dose vuiants thereof, which will eventually determine the criteria of ident ity for
fictional objects which are to be built into the formal ontology at present under
consideration.
INGARDEN VS. MEINONG ON THE
LOGIC OF FICTION
105
characteristics). Thus syntactically speaking object-names are
distinguished from the names of (e.g.) sets of properties associated
with them. Note that in making this requirement we do not coIQmit
ourselves to the possibility that an object may exist (or have any kind
of ontological status) independently of its actual stock of properties,
nor indeed that properties may exist except as the properties (individual accidents) of determinate objects.
(iii) For the case of fictional objects in particular we have shown
that it is necessary to distinguish properties possessed by the objects in
question, from characterist£cs which are merely associated with those
objects. It seems that we must restrict the normal syntax of attribu·
tion (P(a), R(a,b), etc.) to the former case, introducing some alternative mode of representation for characteristic-association.
Wh~ter
a logic of fiction satisfying these conditions can be
developed which would have a facility equal to those logics already
proposed, e.g., in allowing the expression of shared philosophical in·
sights, is an open question. None of the given conditions need involve
any radical overhaul of our basic logical machinery, indeed it seems
that they would make themselves felt principally in the form of an unwield complexity in the system which resulted, but this, as we have
argued, is no just ground for complaint. For the criteria of
mathematical elegance- and mathematical deviance-for their own
sake are out of place in the realm of formal ontology as a descriptive
philosophical discipline.
BARRY SMITH.
UNIVERSITY OF MANCHESTER.