VICTORIA UNIVERSITY LIBRARY
TORONTO, ONTARIO
SOURCE:
/
ARISTOTLE S
RITICISMS OF PLATO
BY THE LATE
J. M. WATSON
GUTHRIE SCHOLAR IN THE UNIVERSITY OF ST. ANDREWS
HONORARY SCHOLAR OF ORIEL COLLEGE OXFORD
HENRY FROWDE
OXFORD UNIVERSITY PRESS
LONDON NEW YORK TORONTO & MELBOURNE
1909
o
OXFORD : HORACE HART
PRINTER TO THE UNIVERSITY
30/10
NOTE
THIS essay is published after much hesitation; for it
is certain that Watson would not have wished it to appear
in print. I discussed it with him shortly before his death
in 1903, and I know that he regarded it as only a sketch,
which he intended to work up during the next year or
two. It must be remembered that he was only twenty-
four when he wrote it. Even so, however, it will be
admitted that, if he has not answered the question with
which he deals, he has asked it in the right way. Some
readers will note stray indications of a solution rather
different from the main position of the essay.
Watson s friends have decided to print his work, in
order that some memorial may remain of a singularly
gifted young man, to whom they were deeply attached.
If he had lived, there can be no doubt that he would have
been one of the first scholars of his day.
JOHN BURNET.
ARISTOTLE S CRITICISMS OF PLATO
FROM the days of the Greek commentators onward, it
has been a-standing charge against Aristotle that he did
not understand his master s philosophy. Syrian, 1 for
example, representing the Neoplatonists in general, says in
grandiloquent language that Aristotle s criticisms * no more
affect the divine doctrines of Plato than the Thracian shafts
reached the gods of heaven . Similar reproaches are to be
found in Simplicius and Philoponos. In modern times
to pass over the controversies before the eighteenth century
it has been repeatedly maintained that Aristotle first
misunderstands his master s teaching and then criticizes
the result of his own misunderstandings. On the other
hand, champions of Aristotle have not been wanting, though
they are perhaps in a minority. Hegel, 2 the founder of all
modern study of Aristotle, treats the supposition that
Aristotle did not understand Plato as an altogether
arbitrary and unfounded assumption * in view of Aristotle s
fine deep thoroughness of mind, perhaps no one knows
him better .
The origin of this diversity of opinion is not far to seek.
On the one hand, as ancient and modern commentators
alike point out, Aristotle is constantly * Platonizing 3 . In
his every work may be found, if not explicit approval or
quotation of his master, at least innumerable reminiscences,
conscious or unconscious, of Plato s doctrine or language.
But, on the other hand, Aristotle seems to criticize Plato
1 Syrian on Met. B. 997 b 5 sqq. (Aristotelis opera Berol. 1870, v, p. 849 a 32).
2 Hegel, Werke, xiii, p. 189.
3 Cf. Aristotelis Fragmenta, Rose, p. 432 (Teubner, 1886) 77877 Se Kal Iv o7s
6 Aristotle s Criticisms of Plato
unfairly and pedantically. He misconceives the mythical
character of the Timaeits ; he treats poetry as though it
were science ; he denies to Plato the credit of investigations
and metaphysical discoveries in which, nevertheless, the
master had at least foreshown the way to the pupil. More
over, in his attack on the Ideal theory especially, he has
been thought to set up a straw man of his own making
before proceeding to demolish it. It would seem then to
be well worth inquiry, (a) how far such charges of mis
understanding and unfair criticism are justified ; and (b) how
far the peculiar nature of Aristotle s criticisms can be
naturally and rationally explained.
In entering on these questions, it would be of great
service to know the exact order in which the works of
Aristotle were written. Thus the chronological accuracy
with which we can now 1 trace the various utterances of
Leibnitz in relation to Spinoza are most illuminating for
the criticisms passed by the former on his great predecessor.
But in the case of the Aristotelian Corpus a historico-
chronological inquiry is complicated by cross-references
and other difficulties, and as yet the few writers who have
undertaken such an inquiry have been able to arrive only
at probabilities and approximations. The application of
stylistic methods could hardly be so important or fruitful
here as it has been in the case of the Platonic dialogues :
still the researches begun by Blass 2 are in the right
direction.
The dialogue Eudemus may be taken as one of Aristotle s
earliest writings. It seems to have been thoroughly
Platonic, defending indeed, in the spirit of the Phaedo, a
doctrine of personal immortality which Aristotle in maturer
1 Since Stein, Leibnitz und Spinoza.
2 F. Blass in Rhein. Mus. 30. He applies to Aristotle the test of avoidance
of hiatus.
Aristotle s Criticisms of Plato 7
years, after his physical studies, did not see his way to
accepting. The Eudemus and the lie/)! tyiXoo-o^ias were
probably written, though not necessarily published, while
Plato still lived, and already in the latter dialogue we find
Aristotle up in arms against the Platonic theory of Ideas.
It is true that he is profoundly conscious of the enormous
advance made in mathematics and philosophy during the
Platonic age ; such progress, he thinks, had been made in
a few years that philosophy in a short time would be
absolutely complete . But even at this early period he
has definitely broken away from the Platonic position ; he
protested in the plainest terms that he could have no
sympathy with this doctrine, even should his opposition be
put down to a contentious spirit of rivalry . 1 Another
passage, quoted by Syrian, shows that Aristotle had also
already made up his mind on the untenability of the theory
of Ideal numbers. 2 Here too he decisively declared the
world to be not only unending, but also without beginning
in time. 3 Obviously the reader , the mind of the school ,
was to be no mere disciple in philosophy.
To the same period must belong the notes which were
taken by Aristotle, as by other pupils, of Plato s lectures
On the Good (ne/n Taya0o). Even Aristotle seems to
have found them obscure 4 (/5r?0eWa euznyjxara)8<Sy) ; so we can
well believe what he used to tell (del Sujyetro) of the utter
perplexity with which an audience, that had come eagerly
expecting to hear about happiness and human good, found
itself listening to a lecture on mathematics, numbers,
geometry, astronomy, and finally that Good was One . 5
1 Rose (Teubner), p. 27.
2 Rose, p. 27. This passage also is from the Second Book, which contained
the criticism of Plato. The remarks on the advances in philosophy probably
came in the First, though Rose gives them under the Protreptikos ; v. By water
in Journ. of Phil. vii.
3 Rose, p. 33. * Rose, p. 41. 5 Rose, p. 24.
8 Aristotle s Criticisms of Plato
Aristotle had little sympathy with the later mathematical
speculations of his master.
The criticisms of Plato s Ideal theory in the Metaphysics
would probably be less perplexing had Aristotle s ITepi ISewi;
come down to us. Syrian, 1 it is true, says Aristotle had
no arguments additional to those set forth in Met. A and M,
but the testimony of such a partisan is worth nothing ; and
Alexander, commenting on Met. A. 9, has a different tale to
tell. 2 Unfortunately little or nothing is known as to the
date of this Critique of Idealism , though probably it too
belongs to the first Athenian period.
Perhaps the first work of the Aristotelian Corpus, as we
now have it, is the Topics. Here, at least in Books II-VI,
we find everywhere Platonic expressions (e.g. piTtyeiv) and a
Platonic standpoint, not merely the Platonic soul-division,
but even the Ideas (Ibtai) employed for the positive purpose
of testing definitions. 3 But, as has appeared above, he is
already the antagonist of the Platonic theory of Ideas, and
we find him in the Topics supplying points (TO TTCH) or
ready arguments against the Idealists J (roVot xp^M 01 irpbs
TOVS TLOcptvovs Ibeas etmt). 4 One of these, which occurs in
the Soph. El.f is the famous argument of the third man 6
(rpiTos avOptoiros) which Aristotle shows has no relevancy
except where (as in the Ideal theory) the common predicate
(rd Kotvfi Kar-riyopov^vov), e. g. man J , is hypostasized into a
particular (ro 6e n). Plato is mentioned by name four times
1 Rose, p. 148.
3 Vide especially on 991 a 8 sqq., where Alexander reproduces from the Second
Book of the IIe/>t ISeuv a number of Aristotle s arguments against the Ideal theory
as held by Eudoxos. Some of these apply equally to the irapovaia of the Ideas
on Plato s theory.
3 Top. 137 b 3, 147 a 5. * Ib. 143 b n sqq., 148 a 14, 154 a 18.
5 c. 22. 178 b 36.
6 That we have here really the familiar third man and not merely a sophistic
quibble against the concept in general has been shown by Baumker, Rhein. Mus.
34? PP. 73 sqq.
Aristotle s Criticisms of Plato 9
in the Topics, but nowhere else in the whole Organon. In
Post. An. 1 there is an explicit attack on the ev irapa ra -n-oAAa,
and the Ideas are once impatiently dismissed as mere
Tperioy/ara, 2 i. e. they have more sound than sense.
It is disputed whether the Organon is followed by the
ethical or by the physical treatises. The former, Rose s
opinion, is more probable than Zeller s, and at all events
Eth. i. 6 reads as if it were early. Plato is referred to
approvingly in the Ethics three times by name, twice
without name, 3 while whatever may be thought of
the criticism in i. 6, its intention obviously is to
be conciliatory. Met. A. 9 is the only passage where
Aristotle, in speaking of the Academy, uses the first
person plural and ranks himself as a Platonist, 4 and this
probably means that he had not yet developed his own
system. Met. A. 9 is known to be a rechauffe of the
arguments of the Ilept I6eo>r, and the latter is at all events
quite early.
There is no need to dwell on the later works. Three
remarks may be made : (a) There are no direct criticisms
whatever in the Rhetoric or Poetics, though in the latter
especially they might be expected. The Rhetoric has an
interesting notice of the exasperation felt by the partisans of
the Idea (ot TH rr\ t6ea, sc. tyikoTLpovpevoi) at attacks on this
favourite doctrine. 5 (b) The relation of Metaphysics A. 9
to its duplicate in M. 4 and 5 is still an unsolved problem.
A. 9 has been thought later and more mature, because (e.g.)
instead of saying that the Ideas are more in number (TrAeuo)
than the particular things of sense, A. 9 contents itself with
1 i. ii. 77 a 5. 2 Ib. i. 22. 83 a 32.
3 A. 4. 1095 a 32. B. 3.11045 12. K. 2. 1172 b 28. Cf. E. i. 1129 a 6sqq.,
K. 9. n8oa 5 sqq.
4 riOefAfv, olofjifOa, ou <f>ajji(v, &c. : in Eth. i. 6 TO. oitcfta dvaipfiv. The first
person plural occurs also twice in Met. B. 997 b 3 and icoa b 14, as if simply by
reminiscence of A.
5 Rhct. ii. 2. 1379 a 34-
10
Aristotle s Criticisms of Plato
the more guarded phrase just as many or at all events no
fewer (Ua r) ov/c eAarrco) ; still, even if in A. 9 we have the
criticism of the Ideal theory in its final form, this does not
exclude the very early date of most of the arguments, (c) It
might be thought that the references to Plato would in all
probability grow sharper and more unsympathetic as Aris
totle s own system took definite shape. Thus the criticism
of Plato in the last chapter of Book VIII of the Politics is
rather more direct, downright, and unceremonious than
usual (e.g. 1316 b 17 TOVTO 5 &m i/r58os), and this chapter
Newman thinks is of a * somewhat later date than the rest
of the book . Nevertheless, even in the Metaphysics,
there is no perceptible change of tone, and Plato is
mentioned by name and with approval no less than four
times. 1 Chronology, in short, seems able on this question
to yield little definite result. 2
A. Aristotle s Metaphysical Criticisms
We pass at once then to the metaphysical criticisms,
which are the most numerous and the most important.
The difficulties here may be resolved into the following
five problems :
(1) In Met. A. 6 Aristotle states as Plato s a doctrine we
should never have extracted from the Platonic dialogues
alone.
(2) The doctrine which Aristotle controverts is sometimes
directly at variance with that of the Dialogues. Thus
Aristotle says Plato made Ideas of natural things (6iroVa
T. 5. loiob 12, A. ii. ioi 9 a 4, E. 2. 1026 b 14, A. 3. 1070 a 18.
irtain methods of statistical inquiry might be useful, in answer e.g. to the
ons : () what is the comparative frequency of Aristotle s criticisms of
to and of the Platonists, and also of the direct and the indirect references to
\ (*) in what parts of Aristotle s philosophy is the criticism
wt , and where, if at all, is it silent? (,) how far are the criticisms in all
m ml the branches of philosophy, dialectical ?
Aristotle s Criticisms of Plato n
* to the exclusion of artificial products ; he states,
moreover, that orthodox Platonism 2 rejected Ideas of
negations and (according to the usual interpretation) also
of relations (TO. vpos n ). 3
(3) He attributes to Plato a doctrine of Ideal numbers,
which (at least in the form stated) critics have found it
hard to ascribe to Plato as a serious philosophical theory.
(4) The centre of Aristotle s attack is the transcendence
of the Ideas (abvvarov etvai x co/us 1 rr)V ovaiav /cat ov f) owa). 4
Now it has been maintained (a) that Plato never held such
a doctrine at all in Aristotle s sense ; or (b) that in a later
stage of his thinking he recognized this defect in his meta-
physic, and himself overcame and rejected the dualistic
severance (TO \<*>plfav Met. M. 9. 1086 b 4) of universal and
particular.
(5) Aristotle denies to Plato the recognition of final and
efficient causes, 5 which nevertheless seem in the Dialogues
to be laid down with as much emphasis as by Aristotle
himself . 6
The fourth problem deserves fuller statement. In the
Parmenides the aged philosopher of that name criticizes
with great earnestness a theory of Ideas which is unmis
takably that of the Republic and Phaedo. The difficulties
urged against it are so serious that the Parmenides has
again and again been declared spurious, 7 on the ground
that it is not given to any philosopher, however great, to
overleap the limits of his own system, and that to ascribe
it to Plato is to make of a single philosopher both Plato
1 Met. A. 3. 1070 a 18. 3 Met. A. 9. 990 bit. 3 990 b 16.
4 991 b i. Cf. De Caelo i. 9. 278 a 16 efre yap tanv fiSrj, Kadairfp (paaii/ rivts
KT\., (ire mat x (t} P tffr ov prjOlv rwv TOIOVTOJV, where the Platonic Idea and self-
subsistence are interchangeable terms.
5 Met. A. 9. 992 a 24. 6 R. G. Bury, Philebus, Introd., p. li.
7 Notably by Ueberweg and Ribbeck, the latter of whom says the Farm.
signifies den Umsturz der gesammten Platonischen Ideenlehre (Phil. Monais-
hefte xxiii. 1887).
12 Aristotle s Criticisms of Plato
and Aristotle at once. But to waive this question for the
moment, two points are all-important to notice for the
present inquiry, (i) All the difficulties urged in the
Parmenides arise from the absolute transcendence of the
Ideas, their complete severance from the world of sense. 1
This, in the first place (a) makes jW0eis impossible ; for,
whether participation takes place by whole or part, in
either case the self-dependent unity of the Idea is sacri
ficed. Moreover, since avro^yeOos e.g. is severed (xpk)
from ra TroAXa /zeyaAa, the latter may be compared with the
former, and, it is asserted, another etSos ptyedovs is needed
to make avropeydos great. 2 Secondly (b) it makes
also impossible ; for, if the Ideas are a second world
aiTa Kad aura, Farm. 129 d) and yet like the particulars, there
must be a third Idea or irapabctyiM to explain this likeness,
and again we get an infinite regress. 3 Thirdly (c) it makes
knowledge impossible. A really noumenal world is ipso
facto unknowable ; i. e. we cannot know God, and moreover
the converse also is true, God cannot know us. 4
(2) The second point to be noted is the striking fact that
Aristotle uses most of these identical arguments of the
Parmenides, and yet never once refers to this dialogue,
either when he reproduces its objections in Met. A and Z, or
in the whole course of his works. He twice employs the
Tyuros avdpuvos argument, 5 he says the same Idea will be
at once copy and type, 6 he points out by arguments similar
to those of the Parmenides the impossibility of /meflefis or
TrapotmV he asserts that the Ideas, being transcendent, do
not explain knowledge. 8 His contention that the Ideas
1 Cf. Farm. 129 d, 130 b, d, 131 b, 133 a. z Farm. 132 a-b.
3 Farm. 132 d-e. * Farm. 133 b sqq.
5 Met. A. 9. 990 b 17, Z. 13. 1039 a 2. 6 Met. A. 9. 991 a 31.
7 Z. 14. 1039 a 26 sqq. ; cf. Farm. 131 a sqq., also Alexander on Met. A. 9.
591 a 8 (Hayduck, p. 97. 27-98. 23) reproducing the Tltpl I
* A. 9. 991 a 12.
Aristotle s Criticisms of Plato 13
contribute nothing whatever as the causes of phenomena l
is merely a summing up of Plato s conclusion that neither
fxefofu nor fjLL{j.iiarLs is possible, if the Idea is x^P^ a ^
Ko.0 avro. In fact, the chief Aristotelian objections are
simply based on the absurdity in all its consequences of a
common predicate which is at the same time substance
(ova-La), the absurdity of a universal thing , a Ka06\ov which
is at the same time yapurTov? We seem forced, then, on
the horns of a dilemma. Either Plato, in spite of the
annihilating assaults (grundsturzende Eimvande 3 ) of the
Parmenides, did not, in his later system of metaphysics,
abandon the transcendence of the Idea, or Aristotle is not
merely guilty of plagiarism, but has grossly and unpardon-
ably misrepresented his master s teaching. It must appear
in the sequel whether this dilemma is simply another
instance of the dichotomous either ... or\ which works
so much havoc in philosophy.
Doubtless the easiest method of solving all the problems
is to assert that Aristotle misunderstood Plato and that
there is no more to be said. But even were this asser
tion admitted, it would at least be necessary, following
his own constant example, to show some plausible
OLTLOV rr?s eKTpoTnj?, some reason for the aberrations of an
Aristotle. 4 The problem is not solved by ignoring it. We
pass on then to consider various theories, which, in different
ways, really attack the difficulty.
First Problem
It is natural to begin with Zeller s Platonische Studien,
which, though published in 1839, still remains the best
essay on this subject as a whole. Zeller is most helpful
on the first of the problems above propounded. No one,
1 991 a 9. * Vide especially Met. M. 9. 1086 331 sqq.
3 The phrase is Ueberweg s.
4 Met. N. 2. 1080 a I. Cf. Politics ii. 5. 1263 b 30 nlnov TTJS
1 4 Aristotle s Criticisms of Plato
even after a complete course of the Platonic dialogues,
including the Philebus and Timaeus, can come to Aristotle s
account of Plato s philosophy in Met. A. 6 without ex
periencing a shock of surprise, and it was Zeller s great
service to show that this chapter implied no esoteric
Platonic doctrine, but could be explained partly from
the dialogues themselves, partly from the precise and
logical character of Aristotle s thinking, which constantly
strives after definite and clear connexion.
On one particular point, according to Zeller, Aristotle
has misinterpreted Plato. He has identified the matter of
the world of sense (Space, the Unlimited, the Great and
the Small ) with the multiplicity, the non-being, the other
ness, which forms the material principle of the Idea. That
is, he makes the One and the Great and Small the ele
ments (orotxeta) of the Ideas, and says they are at the same
time the principles of reality (CTTC! 5 atria ra eidr; rots aAAoi?,
TCLKtivvv oToix^ta utorwt tLrfOr] (sc. nA.dVa)z;) r>v OVT&V etvai
a-TOL^a 1 ). This mistake, according to Zeller, is easily
intelligible for two reasons, (i) Plato himself had talked
of the Unlimited or Great and Small in reference to the
Ideas, and had not explained how this Unlimited was
related to corporeal matter. (2) Aristotle s view is meant
to offer a solution of the fundamental difficulty in Plato s
philosophy, viz. that, from Plato s standpoint, there is no
possible way of deriving phenomena from the Ideas.
But Aristotle s solution that Idea and phenomenon are
composed of the same elements (crroix eta) really cuts away
the ground from under the whole Ideal theory. It renders
the Ideas a superfluous second world, and makes easy
Aristotle s criticisms of the transcendence of the Ideas
and the Mathematical (ra Crafty. In short, this
single alteration of Plato s doctrine once admitted, we
1 Met. A. 6. 987 b 18.
Aristotle s Criticisms of Plato 15
have the key to unlock all the more important differences
between the metaphysical system of the dialogues and
that of Met. A. 6. 1
Dr. Jackson, in his valuable contributions towards the
understanding of Plato s later doctrine, seeks to disprove
the opinion of Zeller that Aristotle has somewhat mis
apprehended Plato . 2 He comes to the rescue with a new
interpretation of the Philebus? It has long been a problem
of Platonic interpretation where we are to find the Ideas
in the division of all reality (navra ra vvv ovra kv ru> iravrij
Phil. 23 c) given in that dialogue. Dr. Jackson proposes
to find them in the third class of the division the WKTOV
yeW, the same class as that in which the particular pheno
menon is included. This original suggestion is not so
paradoxical as it might at first sight appear. The Philebus
states explicitly that in all being there is present Limit
(Tre pas) 4 and Unlimitedness (aveipia) ; these, therefore, must
appear in the Idea as well as in the sensible particulars,
and the only question is, How is Idea differentiated from
particular? Jackson answers that while the indefinite
matter (TO p.aX\ov KCH ro riTTov) is the same for the Idea and
the particular, the Tre pas or limitant quantity (ro itoa-ov) of
the particular differs from, but at the same time more or
less approximates to, the limitant quantity (ro pirpiov) of the
Idea, and the more nearly the Tre pas of the particular
approximates to the irepas of the Idea, the more closely
the particular resembles the Idea . 5
It will be seen that the special feature in this interpreta
tion is the distinction (in the exposition of Phil. 24 C sqq.)
1 Zeller, Platonische Studien, p. 300, pp. 291 sqq. Also in Plato (E. T.),
pp. 319 sqq.
2 Plato (E.T.), p. 327.
* Jackson s articles are to be found \njourn. of Phil, x-xv, xxv. His treat
ment of the Philebus comes in vol. x, pp. 253-98.
* Phil. 1 6 C. 5 Journ of Phil, x, p. 283.
Aristotle s Criticisms of Plato
between TO wo<nw and TO j^rpicw, the latter being the formal
element of Ideas, and >o-a the various formal elements
of the particulars. Jackson finds this reading of the Pkilebus
confirmed by Met. A. 6. By inventive exegesis and emen
dation of one refractory passage, he makes out (i) that
TO fxfya KCU TO piKpo v are the equivalent of the more and
less of the Philebus : (2) that TO to KCU ol apiB^oi correspond
to Topfrpiov /cat Ta TTOCTCL : (3) that the e o>J/ yfyvcrai of Philebus
(27 A) are the same as the oroixeta of Met. A, and the ele
ments of the Ideas are the elements of all things : (4)
that the two elements are, both in Philebus and Met., the
origin of good and evil respectively. In short, the doctrine
ascribed to Plato in Met. A. 6 is precisely the doctrine of
the Philebus:
It will be admitted that Jackson s interpretation of this,-
one of the most abstract chapters in the whole Metaphysics,
is much more ingenious than convincing. In fact it is
a tour de force, and is at once seen to be so on any inves
tigation of all the relevant passages. 1 Still this applies only
to statement (2) in the above summary, and though for it
little can be said, in his other identifications Jackson is,
with certain reservations, entirely justified. One result he
has certainly brought out with clearness. The Idea, which
is usually thought of as simple and indivisible, undoubtedly
appears in the classification of the Philebus if it is meant
to appear at all as a compound, a result of /mtfis- just as
the concrete particular is. This is precisely how the Idea
appeared to Aristotle, a compound of elements
In 987 b 21 he adds KCU. TOVS apiOpovs after us 8* ovoiav TO ev, bracketing TOVS
d/x0/ious in b 23. His other emendations (Journ. of Phil, x, p. 294) are improve
ments, but the important one in b 21 contradicts the sense and the connexion. The
<TTeu X a are not the Great and Small, the One, and the numbers, but simply the
Great and Small and the One ( . the Idea of Good). He is further quite wrong
in the assertion (x, p. 291 sq.) that the Idea in A. 6 except 9 (988 a 10) is not
the formal cause but the type of the particular.
Aristotle s Criticisms of Plato 17
And further, it seems incontrovertible that the Philebus
favours Aristotle s statement that the elements of the Ideas
are in some sense or other the elements of all reality.
But we must now consider Zeller s theory more directly.
Several objections may be urged against it :
i. Aristotle asserts that the elements of the Ideas were
to Plato the elements of all things. But he nowhere says
these elements are identical for the Ideas and for pheno
mena. Not one of the passages adduced by Zeller can be
said to prove this ; some of them are decisively against any
such supposition. Thus in Phys. A. 2, after showing that
Plato identified space with matter, and remarking that the
matter (v\-n) of the Timaeus is different from that described
in the unwritten doctrines (aypa^a boypara), Aristotle pro
ceeds : Plato however . . . must state why the Ideas, i. e.
the numbers, are not in space. For his teaching is that the
participant and space are interchangeable terms, whether
the participant be the great and small (according to the
aypafya boy^ara) or vArj, as he has written in the Timaeus .
According to Zeller, this reproach presupposes that the
matter of the Ideas is identical with the matter of the
material world, i. e. space. But surely had Aristotle ever
meant that space was the matter of the Ideas, he would
have said so, and not taken the roundabout method of the
above quotation in order to establish his point. He would
not have introduced the objection in the way he does, as
if it were a consideration that might have escaped Plato s
notice, but would simply have said, Space is a O-TO^IOV of
the Ideas : hence the Ideas must be spatial . As it is, he
proceeds to justify his reproach, which on Zeller s view he
certainly would not require to do. His proof is as follows :
Plato identifies TO //eraATjTmKoV with space : now TO /^TaAr}-
TITLKOV participates in the Ideas ; .*. space participates in the
Ideas ; ,*. the Ideas must be spatial. In fact, therefore, this
B
i8 Aristotle s Criticisms of Plato
passage, so strongly relied on by Zeller, really goes against
his view. It expressly distinguishes the space of the
Timaeus from the later material principle, viz. the Great
and the Small , which Plato had laid down in his lectures.
Similarly in Phys. iii. 6, 1 we read : If the Great and
Small is the encompassing principle in the sensible and
intelligible world alike, then it ought to comprehend the
intelligible world . Simplicius 2 explains quite satisfac
torily. According to Aristotle, the infinite ov Trepte xet aAAa
Treptexerot, and qua infinite, it is ayvuo-Tov. Now Plato
admits that the Great and the Small in the sensible
world (i. e. space) wep^x" at<r<Va, and therefore makes
them unknowable. He ought to admit then that the
Great and Small in the intelligible world also Tre/aie xei
(sc. ra rorjra) and therefore makes the intelligible world
1 unknowable . This conclusion is absurd, since it is the
very nature of vo-rjrd to be knowable.
The tentative tone of both of these passages would be
quite unintelligible had Aristotle believed in the identity of
1 the unlimited in sensibles with the unlimited J in Ideas.
Consequently when in Phys. iii. 4, 3 we read that Plato s
faeipov existed both in the world of sense and in the
Ideas , there is no reason to conclude that this frircipov is
for both numerically the same. In Met. A. 6. 988 a 10,
Aristotle states that the Ideas result from two causes:
formal TO Iv, material the Great and the Small. Pheno
mena also arise from two causes : formal the Ideas,
material the Great and the Small. Now, were the
material cause identical for both Idea and phenomenon,
this passage would mean that the Ideas, which determine
the Great-and-Small, are yet themselves partly the result
of that Great-and-Small, a contradiction which there is as
little reason for attributing to Aristotle as to Plato.
1 207 a 29. 2 Sc/iol. 368 a 28. 3 203 a 9.
Aristotle s Criticisms of Plato 19
2. Further, it has not escaped notice that while Aristotle
speaks of the Indeterminate Dyad as the material prin
ciple of numbers, he never applies this phrase to the
material principle either of geometrical magnitudes or of
the physical world. Zeller, indeed, while admitting this,
says the Indeterminate Dyad is simply the Great-and-
Small numerically expressed . But here is the whole
point. Aristotle expressly distinguishes species 1 of the
Great and Small ; one of these species (a) (ro TTO\V K<U TO
oXiyov, Met. N. i. 1088 a 19) is the material principle of the
Ideal, as also of the mathematical numbers, and is other
wise called the Indeterminate Dyad . Another species
(b), the Great-and-Small properly speaking, is the material
element of geometrical magnitudes. As Great and Small
is also the generic name for the material principle, Aristotle
can use the phrase both for (a) the indeterminate dyad, and
for (b) the Great-and-Small of magnitudes ; 2 but he never
conversely uses the phrase the Indeterminate Dyad in
reference to both.
Still another species (c) of the Great-and-Small might be
looked for, viz. the material principle of phenomena, the
empty space (r6 rijs x^pas) of the Timaeus. But the his
torian of the problem of matter in Greek philosophy 3 has
shown that Plato in his later thinking, under Pythagorean
influence, probably subsumed the space of the Timaeus
under the more comprehensive category of TO a-ntipov, or,
as he said in his lectures, the Great and the Small . The
Platonic system advances ever further in the way of
1 Met. M. 9. 1085 a 9-12.
8 The passage (Met. N. 2. 1089 a 35 ov yap 5?) T) Svas 17 dopiaros alria ov 5 rb ^fya
KOI TO niKpov TOV Svo \tvKo. KT\.} would be conclusive that Aristotle was careful to
distinguish these two, were it not for the unfortunate ambiguity by which ov8i
like teal may merely be explicative that is . As it is, therefore, we should
render : It is not the indeterminate dyad (species) nor in short the great-and-
small (genus) that can explain &c.
3 C. Baumker, Das Problem der Materie, p. 196 sqq.
B 2
20
Aristotle s Criticisms of Plato
resolving the physical and the concrete into metaphysical
and mathematical abstractions/ 1 In the striking phrase
of one of the Greek commentators, Plato had completely
mathematicized nature ^are^a^anKewaro rrjz^tW). 2 This
is why Aristotle objects to Plato s great and small that it
is too mathematical a substrate (fw0Tj/xariKa>r^>a #A?/) ; it
may explain mathematical magnitudes but not physical
bodies (vAr/ ao-w/zaro?). 3
Aristotle, then, cannot be charged, in his account of
Plato, with annulling the distinction between the Un
limited in Space and that plurality which is also in the
Ideas . 4
3. Again it should be noted that one of Zeller s main
reasons for rejecting Aristotle s testimony about the de
rivation of all things from the principles of the Ideas, is
simply his own preconceived theory as to the relation
of particular and Idea in the Platonic system. Zeller
thinks the particular is, or was meant to be, absolutely
immanent in the Idea/ the latter being the sole reality.
This, according to Zeller, enables Plato to escape such
difficulties as those raised in the Parmenides? But now
comes the question: Whence the distinction of things
from the Ideas? and to this the Platonic system, as
such, contains no answer . 6 There is an inextricable
contradiction between the absolute reality of the Idea
alone, and the admission, nevertheless, of a kind of
existence that cannot be derived from the Idea . 7
Now this view seems but one result of the radical mis
conception which vitiates Zeller s account of the whole
Platonic philosophy. He attempts, that is, to deal with
1 Ibid. p. 197.
a Quoted by Gomperz, Griechische Denker, vol. ii (on Plato s Matter, p. 606 n.).
3 A. 9. 992 b 2 ; A. 7. 988 a 25 ; cf. N. 2. 1089 a 32-b i.
4 Zeller, Plato, E. T., p. 332. 5 p /a/0> E< T>> p ^3.
c Plato, p. 319. 7 Plato ^ p> 333< Similarly in
Aristotle s Criticisms of Plato 21
the dialogues as one whole, and as furnishing one fixed
and immutable system. He still does not accept a later
date for the great metaphysical dialogues Parmenides,
Sophist, Philebus. Yet in these later dialogues there seem
to be various attempts made at a derivation of the sensible
from the Idea, and one of these is by the method of identity
of elements. We have seen this already in the case of the
Philebus ; in more abstract phraseology a similar doctrine
appears already in Parmenides 142 D. Here Plato shows
that the whole universe contains as aspects (popia) unity
and existence (rd %v KOL TO elvcu), and so likewise does every
smallest part of the universe contain these same two ele
ments, or parts , of ideality and reality. This whole
question belongs strictly to a history of Plato s later
metaphysics; all that need here be insisted on is that
Aristotle has not been proved guilty of any such funda
mental misapprehension as is implied by Zeller s theory.
4. Finally, it should be noted that Xenokrates, 6 y^o-tw-
raro? r&v HXaTuvos &KpoaT<t>v, accepted the doctrine of first
principles which Aristotle ascribes to Plato, and derived
all things from Unity and Indeterminate Duality. Speu-
sippos, indeed, derived merely numbers from Unity and
plurality, and, unlike Plato, for the explanation of everything
else he set up several distinct principles. But it was
precisely for this reason that Aristotle reproached him
with making the Universe like a bad tragedy fragmentary
and episodic (e77tcro8ta>8?; TTJV TOV iravros ovcriav TTOLOVOTLV).
Second Problem.
To turn now to the second main problem. Zeller, in
Platonische Studien, had treated Aristotle s statements as
to the contents of the world of Ideas as merely mistaken.
Similarly Bonitz on Met. A. 9, where Aristotle is thought
to state that orthodox Platonism did not admit Ideas of
22
Aristotle s Criticisms of Plato
relations, is highly indignant with Aristotle for alleged
unfairness in argument. 1 Zeller, by the time he wrote his
History, had come to see that the only satisfactory way of
accounting for Aristotle s words in the Metaphysics was to
suppose Plato had actually made these changes. But
even there Zeller suggests no rationale of them; the
original point of view was in these cases abandoned ;
in other words, they were arbitrary modifications. 2
Now Dr. Jackson seeks to make good this deficiency in
Zeller by showing how Plato, in a radical reconstruction
of his system initiated by the Parmenides, was led naturally
and inevitably in his second theory of Ideas not only to
the doctrine of Met. A. 6, and the transcendency (tSe cu
X<t>piorai) of which Aristotle complains, but also to the
retrenchment and revision of his list of Ideas. According
to Jackson, in Plato s later theory there are no Ideas of
relations (e. g. OIMOIOV &v6fwiov t &c.) nor presumably of ayaQ6v t
KCLKOV .* Accordingly the Timaeus recognizes avra *aO y
avra et8T? of the four elements and of the several species of
animal and vegetable, but of nothing else.
That the Ideal theory of the Phaedo and Republic under
went considerable modification after the Parmenides can
no longer be regarded as doubtful. But as to the parti
cular form of the reconstruction, Jackson is, in some
respects, unfortunate. We must consider briefly his two
central positions (i) the substitution by Plato of pCwo-is
and transcendence for /xe tfefi? and immanence, and (2) the
retrenchment by Plato of the list of etdrj.
As to (i) at least three insuperable difficulties have been
pointed out.
1 Bonitz, Metaph. ii, p. in. He thinks Aristotle is refuting Plato by means
of contemporary Platonism. Really this is one among many passages which
show conclusively that Aristotle is not thinking directly of Plato at all.
Zeller, Plato, E. T., p. 275.
s Jackson, Journ. of Phil, xiii, p. 271.
Aristotle s Criticisms of Plato 23
(a) The metaphor of /xe 0ef is is not altogether dropped in
dialogues admittedly later than the Parmenides. 1 It is
true that Jackson s theory does allow of jjieflef is to a certain
extent, but only because he makes an arbitrary and
untenable distinction between ef8>? and avra KaO avra i6>/. 2
(b) The substitution of the Idea as -napa&tiypa or archetype
does not, as Jackson supposes, avert the objections urged
against the Ideal theory in the Parmenides. The relation
between archetype and copy cannot possibly be any other
than that of resemblance, and hence the attempted solution
by /Liifi?7<ns (6jj,oLovo-6aL y eoi/ceWi, etKao-07Jz;ai) lends itself (equally
with the metaphor of /xeflcfis) to the objection of the third
man . Moreover, for describing the relation of particular
to universal, fu /M?<ns is, as Hegel says, a more figurative,
childish, and untutored expression than jme0cfis.
(c) The new view of the Idea as archetype is not a theory
alternative to that of j^0eis, but is clearly described, in
Parmenides 132 D, as merely a special case of it. Aristotle
also joins them both in a single condemnation. 3
(2) Jackson s theory that Plato restricted Ideas to
natural kinds is (in Aristotelian phrase) still more im
possible . 4 In the first place (a) such a theory is directly
opposed to the natural interpretation of Parmenides i3oB-E.
In this, one of the most striking passages of the dialogue,
Ideas of relations are postulated first in order, even before
Ideas of qualities, and it is precisely with organic types
(e.g. man) and the primary forms of matter (fire, water)
that doubt and difficulty (diropio) first arise. The explicit
testimony of this passage must far outweigh a mere
1 e. g. Soph. 255 A ; Tim. 51 A.
2 It will be found stated by Jackson in Journ. of Phil, xi, p. 322 n. ; cf. xiv,
p. 214.
3 Met. A. 9. 991 a 20 TO 8e Xiyttv ira.pa.5eiy para avra eti/cu cu fjiTtx* lv O.VTWV
TaAAct tctvo\oyfiv kcfrlv KT\.
* en
24 Aristotle s Criticisms of Plato
inference from Jackson s interpretation of the difficult sen
tence with which the Parmenides closes. 1
Morover this, the natural interpretation of the Parmenides,
is alone consonant with the whole course of Plato s Idealism.
As has been pertinently said, 2 the Auto-bug was not of
more importance in Plato s scheme of the universe than
the avroKaXov or the avrodyaOov. The avroKoXoKVVTr) or the
ai>ToXdxavov? which the comic poets or a Stilpo took as ex
amples for the ridicule or the refutation of the Ideal theory,
were not, we may be certain, put by Plato on the same
level as Ideas of relations and qualities.
Secondly (b) the dialogues later than the Parmenides
present various difficulties on Jackson s theory. Thus in
Philebus 15 A, besides Ideas of man and ox, we have also
those of TO KdXov and TO dyatfoV, and in the Timaeus the
words et8os K<Wou vo^rov (51 C) naturally mean * an Idea for
every universal . 4
Thirdly (c) there is absolutely no warrant for refusing to
recognize as Ideas the categories or ytvrj of the Sophist.
Certainly not then by this theory can Plato s later doc
trine be brought into line with the Aristotelian references.
The very antithesis of Jackson s view, in many ways, is
that maintained by the late Professor D. G. Ritchie. 5
According to it also, the Parmenides ushers in a second
1 Journ. of Phil, xi, p. 322.
8 A. E. Taylor in Mind 1896, p. 304.
3 Epikrates, in his amusing description of a Platonic Sialpems.
4 Cf. Farm. 135 B eiSos cvbs eKaffrov, 135 E iSeav TWV OVTUV eKaffrov. It is
mere dogmatism in support of a theory when Archer-Hind says of the words in
the Timaeus we are to understand by litaaTov only every class naturally
determined, TWV uiroaa Qvaei . It is only a natural extension of such subjective
interpretation when he thinks Ideas ought to be confined to classes of living
things, and therefore says of the Idea of fire (Tim. 51 B) we have in this
passage a relic of the older theory which Plato . . . would have eliminated had his
attention been drawn to the subject .
5 Plato in the World s Epoch Makers Series. Also in a paper on the
Parmenides in Bibliotheque du Congres International de Philosophic .
Aristotle s Criticisms of Plato 25
theory of Ideas . But in this second theory the Ideas are
not cut down ; rather they are extended to the whole field
of the knowable, according to the philosophic advice of
Parmenides to despise none of these things (ovbev avr&v
&Tifj.d(iv, Farm. 130 E). Further, the transcendence of the
Ideas is not increased ; it is recognized as the defect of the
earlier theory, and endeavours are made to overcome it.
How then does this theory explain the hostile criticism
of Aristotle ? The answer is : (a) It was probably owing
to the objections of his brilliant pupil (who had come to
the Academy in 367, and to whom there is perhaps a kindly
allusion in the Parmenides itself 1 ) that Plato was led to
reconsider his earlier theory. The criticisms in the
Parmenides were those of Aristotle to start with ; hence
he can dispense with referring to that dialogue, while using
its arguments.
(b) It is not Plato himself that is attacked, but disciples
of Plato, who had not advanced along with him after his
self-criticism in the Parmenides.
(c) The criticism of the Ideal numbers is directed against
Speusippos, to whose Pythagorizing tendencies Aristotle
makes express allusion.
(d) It is in the main not the later but the earlier form of
the Ideal theory that is attacked. As for the remark about
Ideas of relations, Aristotle has been misinterpreted.
SKeTjreoy 8e TtaXiv rt rovrooy Aeyerat KaXws KOL TL ov /caAa>?. 2 Of
the theory as a whole it may be said, as by Aristotle on
the community of goods .in the Republic, that it wears
a plausible look and the student welcomes it with delight
a7ro6e xerai). Nevertheless, though it may not, in
1 v. Farm. 135 D, 137 B-C (6 i/eo/Taros) on the other hand, while Aristotle is
still alluded to, the words rJKiara yap at/ iro\vnpayfj.ovoi may be regarded as a fine
stroke of irony on Plato s part.
3 De Coelo i. 9. 278 a 23.
26 Aristotle s Criticisms of Plato
Aristotle s phrase, be itapnav a5iWo?, it must be admitted
to leave as many difficulties as it solves. Though at the
risk of considerable digression, its main propositions have
here been stated together.
The first of these (a) does not admit of definite proof or
disproof. Aristotle s complete silence on the Parmenides
certainly demands explanation ; nor is it adequate to say
either (like Apelt) that Aristotle did not attach to that
dialogue the same exaggerated importance as the Neo-
platonists and the moderns, or even (with Zeller l ) that the
writings of Plato had not the same significance, as sources
of his doctrine, for Aristotle as for us . 2 Zeller s remark,
as we shall see, is perfectly correct, and must always be
borne in mind. But surely it is more than a mere coinci
dence that the only important dialogue indeed almost the
only dialogue of Plato to which no reference can be found
in Aristotle, should be precisely the work which contains
several of Aristotle s own arguments against that Ideal
theory of which he was the life-long opponent.
In any case, however, whatever solution of Aristotle s
silence be accepted, he can at once be acquitted of any
charge of plagiarism. All the anoptai against the Ideas
are perfectly natural, once phenomenon and Idea are set
over against each other as two independent things . The
T/H TOS az>0/>o>7roy, which is the one distinctive argument common
to both Parmenides and Metaphysics, would arise inevitably
among Greek thinkers, who had a horror of the infinite
process and a passion for refutation by means of it. More
over, the honour of excogitating the third man seems to
1 Plato, E. T., p. 77.
The criticisms in the Parmenides may be regarded as suggested by Aristotle,
but it may be held that Plato was so far from being convinced by them that he
occupies himself in this and later dialogues with criticizing his critic, v. Siebeck,
Platon als Kritiker aristotelischer Ansichten, in Zeitschrift fur Philosophic etc.,
vol. cvii, cviii (1896 et sqq.}.
Aristotle* s Criticisms of Plato 27
belong neither to Plato nor to Aristotle. 1 Alexander,
commenting on Met. A. 9, tells us that Polyxenos the
Sophist* introduced this argument, and he proceeds to
state it in Polyxenos own words. Now Baumker 2 has
shown that it is just the argument of the Parmenides, and
that the reason why, according to Polyxenos, a third man
must be assumed is exactly the ground which induced
Plato himself to set up a second or Ideal man. Polyxenos
was a contemporary of Plato; the latter takes up his
argument in the Parmenides, and shows it is valid as
against one form of the Ideal theory ; and the very method
of allusion to it in Aristotle shows it had long been common
property and a familiar argument of the schools. 3
The second contention of the theory (b) is in part a
familiar one. Already Lotze had said : we are justified
. . . in assuming that Aristotle s attack is in part directed
against certain misunderstandings of the Platonic doctrine
which had gained hold in the Academy at an early period . 4
It has, however, the advantage over Lotze s view that it
does not force us to ascribe to the Platonists a doctrine
which their master had never held at all. 5 It is a theory
which certainly represents a part of the truth. But as a
complete explanation it is open to the insuperable objection
that Aristotle himself is totally unaware of any such
divergence between the master and his school . Had he
1 In Rep. x and Tim* 31 A it is proved that there can be only one Ideal bed
and one avr6^yov because a second would involve a third, and so on. But in
the Parmenides (rpiros &v6panros) it is not Ideas themselves that are spoken of but
Ideas are compared with things .
2 Rhein. Mus. xxxiv, p. 73 sqq. (1879).
3 Moreover Aristotle nowhere claims any of the objections as especially his
own, and it is of the very essence of dnopiai to be avyx\v5fs t v. infra, pp. 121-2.
4 Logic, E. T., p. 444 (ed. 1884).
5 Jackson finds an appeal from the Platonists to Plato in A. 9. 990 b 15
of aKpifSeffTcpoi TWV Xo^cav KT\. But he does not explain (a) why the Republic,
Phaedo, and Parmenides should be honoured with the description of dpi-
i, nor (6) how \6yot in the context can mean expositions .
28 Aristotle s Criticisms of Plato
known of such, it is incredible that he could have missed
the opportunity of appealing from the Platonists to Plato
himself, from the &&v *r\oi to the author of the Parmenides
and the Sophist. This is precisely what he does do on
the question of the Ideal numbers; he commends the
doctrine of the master as against those who denied the Ideas
and retained only the Mathematicals (ra fxaflT^aruca). 1
The third proposition (c) must be rejected in toto. How
ever difficult this problem of the Ideal numbers, there is
no doubt whatever that Aristotle assigns the theory to
Plato. It is true that in Metaphysics M. 4, Aristotle proposes
first to examine the doctrine of Ideas by itself, without the
Ideal numbers, in the form it assumed originally (&s
i>TTt\af3ov tfapxys) with those who first asserted the existence
of the Ideas . But this only proves that the theory belongs
to Plato s later development ; and from De An. A. 2 (where
ra TTC/H $i\oo-o$ias foyoptva 2 have no reference to any work of
Aristotle, but are simply notes of Plato s lectures, of the
same nature as the aypa^a So yjxara) we see that Aristotle, as
usual, is speaking from personal reminiscence of Plato s
teaching. Not to insist on Met. A. 6, where Plato is
compared with the Pythagoreans for making the numbers
(rovs apiOiJiovs) causes of the existence of other things , or
on the similar passage at the end of A. 8, the locus classicus
in Met. M. 8. 1083 a 3 2 sqQ- is Q u i te conclusive. Here
Plato is mentioned by name, the Ideal numbers (ov o-v^XrjToi)
are ascribed to him, and his opinion expressly distinguished
from that of erepot r^es (perhaps Xenokrates) who maintained
the existence simply of the mathematical numbers. Plato
is named also in Pkys. iii. 6, where it is said that though
he made his Infinite (a-neipov) a dyad, he does not employ
it as such : * for in his numbers there is neither the infinite
of diminution, the number one being the smallest, nor the
1 Met. M. 8. 1083 a 22. - 404 b 19.
Aristotle s Criticisms of Plato 29
infinite of increase, since he makes number go only as
far as ten*. The reference here must be to the Ideal
numbers. The evidence, therefore, that Plato held such
a view is ample, even though there be no trace of the
Ideal numbers in the dialogues.
The fourth position (d) as a whole falls to be examined
later. Here we are only concerned strictly with Aristotle s
statements about the contents of the world of Ideas.
Obviously if Aristotle says his antagonists do not recognize
Ideas of relations, negations or arte fada, it can hardly be
the earlier theory of Ideas he is attacking, and Professor
Ritchie s contention would fall to the ground.
(i) As to Aristotle s supposed statement, however, about
Ideas of relations, the theory is justified in suggesting a
new interpretation. The more this alleged dictum of
Aristotle (Met. A. 9. 990 b 15, 16) is considered in the light
not merely of the Platonic dialogues, but even more in
reference to other passages of Aristotle himself, the more
strange it will appear.
(a) The Platonic Ideal theory, after the vision of avrd
TO Ka\6v in the Symposium, had been extended, in the
Phaedo, to Ideas of relation. They at all periods form
Plato s favourite type of example to illustrate his theory
(Phaedo, Republic, Theaetetus, Sophist], and in the all-impor
tant passage of the Parmenides 1 they are selected by
Socrates as examples of the first class of eiSr?, those in
which he has the most implicit confidence. Moreover, since
Aristotle, with his table of categories, does not avoid
confusing relations with qualities, 2 it is certain that Plato
would not escape this confusion, and this is confirmed by
1 130 B-E.
2 Modern logic tends to see in qualities nothing but disguised relations ; to
Aristotle relations are a special kind of qualities. But he does not keep them
apart, v. Zeller, Aristotle, E. T., i, p. 287.
30 Aristotle s Criticisms of Plato
the dialogues. Consequently, once reject the CIVTQKTOV and
the avroKaXov will hardly escape the same condemnation.
Is it then credible that Plato or even the Platonists should
ever have rejected Ideas of relation ?
But (b) even greater difficulties are suggested by
Aristotle s own writings. In Categories 7, we find as
examples of ra npos TL such concepts as TO ftey a
bm\d(riov, TO larov, apen;, eTrtaTTJ/xr;, Seo-Tro r?;?, bovXos. To pass
over the fact that Ideas of every one of these concepts are
to be found in the Platonic dialogues, is it not more than
strange, on the ordinary interpretation of the passage Met.
A. 9. 990 b 16, that Aristotle after stating that the Platonists
reject Ideas of relations should, only a few lines further on,
take as an example of the Ideas he is combating, no other
than the avTobnrXdo-Lov 1 ? Further, the object of the whole
discussion from A. 9. 990 b 22 to 991 a 8 is to show that, on
the basis of what the Platonists say about /xe tfef is, there can
be Ideas only of owtat. Had the Platonists repudiated
Ideas of relations, Aristotle, as has been indicated above,
would scarce have needed all this elaborate argument to
show that Ideas of qualities ought to be likewise dis
carded.
In an interesting passage of the Physics (B. 2. 193 b 34
sqq.) Aristotle, discussing how the mathematician differs
from the physicist, says the former uses abstractions
but is justified in so doing (rite yiWat \}/tvbos
The advocates of the Ideas (ol ras Mas Afyowes), Aristotle
continues, fail to see that they too are guilty of abstraction,
only without the excuse of the mathematician. They
abstract, that is, the objects of Physics. 2 Now odd and
even, straight and curved, number, line, &c., can be
abstracted from motion and sense perception, but this
1 990 b 32.
a TO. <l>va lK * xupifrvnv TJTTOV ovra
Aristotle s Criticisms of Plato 31
ceases to be possible in dealing with bone, flesh, man.
This passage makes it almost unthinkable that the con
temporary Academy had given up Ideas of relations.
Moreover, it can be parallelled by at least two other
passages in the Metaphysics. In 0. 8 Aristotle says the
Platonic dialecticians (ot \v rots Ao yois) are easily convicted of
philosophic ineptitude by the very fact of their positing
Ideas of KIZ^O-U and emor?^. To crown all, in Met. N. i.
1088 a 21 sqq., the Platonists are sharply taken to task for
turning relations into substances. ... It is absurd,
nay rather it is impossible, to make the non-substantial a
principle of, and prior to, the substantial; for all other
categories are posterior to substance/
These passages seem to show that in Met. A. 9, where
Aristotle says Some of the more precise arguments to
prove the existence of Ideas result in the setting up of
Ideas of ra w/oos rt, &v ov (j>a^v ei^ai KaO avro yeVos , these last
words cannot be translated (as by Jackson) relations,
whereof we Platonists do not recognize Ideas . The
authority of Alexander 1 cannot be appealed to on this
passage, as his commentary here is not only obscure and
extremely doubtful otherwise, but also self-contradictory.
He asserts that the Platonists denied Ideas of relations,
because, whereas the Ideas were owwu and self-subsistent,
relations had their being only in ^ irpbs aAA?;/\a a-^la-is. This,
however, is after reproducing an argument (presumably
Platonic) which establishes Ideas of relations , an argument,
in fact, which proves the existence of an CLVTOLO-OV, just as
Plato himself might have done. The Platonists (it would
seem to follow from Alexander s explanation) took no little
pains to establish the existence of Ideas of relations by an
Ao yo?, and at the same time extruded all such
1 p. 82. 1 1-83. 33 (Hayduck).
32 Aristotle s Criticisms of Plato
Ideas from their system. Obviously a new interpretation
is demanded. 1
The clue seems to be supplied by a comparison of our
passage with Eth. Nic. i. 6, taken in connexion with the
fact known about Xenokrates that he admitted only two
categories, the absolute and the relative. 2 In Eth. i. 6.
1096 b 8 Aristotle describes a possible objection to his
previous criticisms. The objection may be represented
thus : You overlook the fact (the Platonists retort on
Aristotle) that we do not acknowledge Ideas of relative
goods (e. g. fire, clothing, wine) but only Ideas of absolute
goods .
Now with this passage in mind, Aristotle s argument in
Met. A. 9, may be paraphrased thus : Some of the more
unimpeachable and rigorous arguments (a/cpi/3ecrre/>oi Ao yot)
of the Platonists to prove the existence of Ideas are forced
to include, among the Ideas thus established, Ideas of
things that belong to the Academic category of the relative
(row irpos ri), and therefore, though these arguments may be
perfectly correct and have at least the merit of consistency,
they are in contradiction with the opinion of the main body
of the * school . In a dialectical argument, such as we shall
see most of Aristotle s refutations are, this revelation of a
discrepancy within the school is all that is required. The
passage is an argumentum ad Platonicos, and has no refer
ence whatever either to Plato or to Ideas of relations. 3
1 It has been seen above that Bonitz is unsatisfactory on the passage. The
interpretation here given is suggested by Professor Ritchie in his Plato.
Like Plato, v. Zeller, Plato, E. T., p. 242 n. ; cf. Philebus 53 D.
3 No doubt it will at first seem conclusive against the above view that Aristotle
s here nevertheless held to be right in what he says of &v6aa Qfoei. But if
even Aristotle s own use of the phrase includes < geometrical magnitudes
0*7*0*, e. g. lines, triangles, &c., v. De Coelo i. i. 268 a 4) might not Plato s use
9f *v, especially in later life when the idea of Nature grew more and more
important to him, have included also qualities and relations? Moreover,
Aristotle in A. 3 does not say that Plato admitted Ideas only of ov6<ra fvaei, but
Aristotle s Criticisms of Plato 33
(2) But Aristotle s remark about oiroo-a ^vo-ei cannot be
explained on the theory that Aristotle is attacking the
earlier Platonism of the Republic or Phaedo. It is said
that this remark (Met. A. 3) does not necessarily imply
any real divergence from the position of Rep. x, where
there is postulated an Ideal bed . There is no science of
beds or houses in the same sense as there is of man or of
the good, and consequently Plato cannot have placed Ideas
of artefacta on the same level as other Ideas. But he need
not have rejected them. We can think a house scientifi
cally by thinking of the end attained by it were it perfect.
Now, in Aristotle s phrase, ^ $wi? re Aos KCU ov eW/ca, and
therefore as soon as a house attains its real end it can be
included among oiroa-a $wet.
This interpretation, which can appeal to ^ h rfj </wm ova-a
xXivri made by the <$>vrovpy6s in Rep. x (597 B, D), overlooks
two points : (a) in the passage of Met. A natural things (ra
4>uo-i, oTrocra (j)vo-ti 1070 a 18, 19) are expressly distinguished
from arte facta, e. g. house (a 14, 15) ; (b) there is evidence
independent of Aristotle that the Academy rejected Ideas of
artificial products. Xenokrates, e. g., seems to have defined
the Idea as archetypal cause of the eternal existences of
nature (alriav 7rapabiyiJiaTiKi]v T&V Kara <$>vv\.v act a-weorcoTooi;).
This view, if it was ever held by Plato, must be later
than that of the Republic, and therefore Aristotle s remark
applies not to an earlier theory which Plato had rejected,
but to a later view represented in his lectures (e$r/ A.
1070 a 18).
(3) As to Ideas of negations, the theory we are consider
ing suffers from an internal inconsistency ; for it admits that
when Aristotle, in a redudio ad absurdum argument against
the Platonists, implies that the latter reject Ideas of
only that natural things to Plato did have Ideas, whereas artificial products
did not, v. infra, p. 34.
C
34 Aristotle s Criticisms of Plato
aTToc^cm? (Met. A. 9. 990 b 12), this can apply only to the
1 final theory of Plato . In the Republic we find Ideas of
the bad and the unjust, in the Theaetetus of KOKOZ;, alvxpw,
and jSios atfeo?, in the Parmenides of <WoVqs, in the Sophist
of ^ " ( ? frP<w)- If he finall y re J ected them > it: was
because the perfect and the beautiful, having more of Wpa?,
can be known more completely than the imperfect and
the ugly. The conception of evil as deviation from a
type appears clearly in the Philebus.
It will now be possible to sum up the positive results
of the discussion on the content of the world of Ideas.
(a) There is some Platonic warrant for the rejection of the
Ideas of negations, and no reason for doubting that, as
Aristotle implies, Plato s followers at least discarded them.
(b) That Plato dropped Ideas of arte facia is supported by
the silence of all the later dialogues, (c) Aristotle is further
right in saying that Plato s Ideas extended to all natural
things (OTTO (pvaei). These words, however, must not be
interpreted more strictly than the context warrants ; thus
they do not exclude concepts like health, triangle, line. 1
(d) The statement that Plato banished from his system Ideas
of relations would be very difficult of acceptance, but
Aristotle does not make such a statement.
Third Problem.
In passing to the third and fourth of our problems, we
must take account of the recent work by M. Milhaud,
Les Philosophes Geometres de la Grece, the second book
of which, dealing with Plato, is, at least in its fifth chapter,
one of the most original contributions of recent years to
the literature of the Platonic question. The theory of
Ideal numbers has long been a mystery to students of
1 In A. 3 Aristotle speaks of i7teta as an example of things that come to be by
art (irav TO Kara Tt\vir)v), y et it also of course exists <pv<r(i, and Aristotle himself
gives avroiiyieia as an example of a Platonic Idea (v. Bonitz, Index, s. v.
Aristotle s Criticisms of Plato 35
ancient Greek philosophy. Aristotle s statements about
these numbers may be reduced to the following : (i) The
Ideas, according to Plato, are numbers. This is stated
without qualification. 1 (2) As to the nature of these numbers,
they are heterogeneous and cannot be added together
(ao-u^A?]roi, bidfopot, 2 qualitatively different). (3) As to their
function, they are causes of things (amot Met. A. 9. 991 b 9,
T&V ovTcav amat TrpcSrat M. 6. Io8o a I4). 3 Critics have not,
as a rule, been ready to accept Aristotle s testimony ; they
regard the numbers as intended by Plato to be at most
symbols of Ideas. 4 Zeller doubts whether Plato ever
actually identified the Ideas with numbers ; he thinks
Aristotle has here allowed himself an inversion \Umstellung)
of the true Platonic doctrine. Plato regarded the numbers
as fallen Ideas (depotenzirte Ideen] ; Aristotle regards the
Ideas as sublimated numbers . Zeller modifies but does
not give up this idea in his History f and he would still
agree with Bonitz in considering the Ideal-number theory
in the light of a mere appendix J 6 to the Platonic system.
Very different are the conclusions reached by Milhaud
regarding the Ideal numbers. He shows 7 how Plato in
his later philosophy came more and more, like Kant, to
a synthetic way of thinking. That is, in seeking to solve
the paradox of ptOe&s propounded in the Parmenides } Plato
gives up all material analogies of whole and part, and after
transferring the question to the world of Ideas, and show-
1 Met. A. 9. 991 b 9 and passim, esp. 1081 a 12. In the difficult sentence
A. 6. 987 b 22 ( out of the great and small by participation of these in the one
come TO ftSij roiis dpiOpovs ) there is no reason to dispute Alexander s inter
pretation, that TO. eft)?; and rovs apiOpovs are put simply side by side in apposition.
2 Met. M. 6-7.
3 As to how they are causes, v. Met. A, 9. 991 b 9, N. c. 6 ; De An. i. a. 404 b
19 sqq. ; Eth. Eud. i. 8. 1218 a 18 sqq.
4 < quasi symbolanotionum, Bonitz, ii, p. 544 ; Zeller, Plat. Stud., pp. 298,263.
6 v. for Plato s later theory p. 517 ; contrast p. 255 (Plato, E. T.).
6 Bonitz, ii, p. 540. 7 Milhaud, pp. 327 sqq.
C 2
3 6 Aristotle s Criticisms of Plato
ing that there some union of specifically different kinds is
absolutely essential, he finally solves his problem by the
union in every Idea of the heterogeneous elements, being
and non-being. The Idea is a meeting point of the finite
and the infinite, the one and the dyad of Great and Small ;
i. e. the principle of fixity, equality, determination (&), and
the principle of variation, of indeterminate multiplicity
(ao/oioros bvds). But now, corresponding with this spirit of
synthesis, and helping to promote it, a great development
had taken place in the conception of quantity. 1 Incommen-
surables cannot be explained by the old conception of number
as a mere putting together of homogeneous units. In the
case of two incommensurable magnitudes there is no longer
identity of quantitative composition ; one is not part of the
other. Yet there is a relation between them ; quantity can
still fix their mode of dependence, though they are not only
not identical but are in a sense irreducible, one to the other.
In short, what has taken place is a radical transformation of
the idea of number ; its significance has now been enlarged
by the introduction of quality, the heterogeneous. It can
still continue to be called number , no longer, however,
in the sense of o-vW^a povdfiuv, but as fixing the mode of
dependence of the most heterogeneous elements. And of
this new number the only principles that can be assigned
are the principle of variation and the principle of fixity;
hence at once the identity of Idea and Ideal number.
Now here is the central point of Milhaud s theory. 2 The
later Platonic doctrine of Ideas was expressed solely in a
mathematical form ; the Ideas had become Ideal numbers,
intimate unions of quantity and quality/ quantities deter
mining unique and specifically different qualities. Aristotle
1 had not in the same degree come under the influence of the
new geometry ; 3 he saw in number nothing but a total of
pp. 179 sqq. 2 Cf. Taylor in Mind, 1903, pp. i sqq. 3 Milhaud, p. 358.
Aristotle s Criticisms of Plato 37
units in juxtaposition. As a natural consequence he mis
understood the Ideal numbers, and in misunderstanding
them has misunderstood the whole Platonic theory. For the
Idea is related to the particular in a peculiar way which can
only be grasped by bearing in mind its character as an Ideal
number. Once we see that Plato was thinking through
out of mathematics and mathematical analogies, the relation
of Idea to particular no longer presents any difficulty.
Such, in brief, is the theory of M. Milhaud. Before
criticism it will be necessary to look at the nature of Aris
totle s objections to the Ideal numbers.
In Met. M. 6 Aristotle takes up the word * numbers
and, treating number as a whole of units, asks in how many
possible ways these units can be conceived. He answers,
they maybe thought of in three different ways, (i) Every unit
may be combinable with every other, as in the mathematical
number. (2) Every unit may be incombinable with and
qualitatively distinct from every other. Aristotle admits
in the next chapter that no thinker had actually put forward
a theory of units thus incapable of all combination, 1 but he
says that impossible though it may be, it is the theory which
the Platonists in consistency ought to hold. (3) The units
in any one number may be combinable with each other,
but not combinable with the units in any other number.
Thus the Ideal number two, the auto-dyad, is not reached
by adding a unit to the primal one ; instead of this, there
are at once two fresh units produced ; similarly the auto-
triad is formed without the aid of the auto-dyad, the units
in the former being quite different from those in the latter.
This is the opinion Aristotle ascribes to Plato and the
Platonists. 2
Now obviously if the Platonists did not admit that their
Ideal numbers were made up of units (|uoya8es) at all, the
1 M. 7. 1081 a 35 sqq. a M. 7. 1081 a 23-5. 6. 1080 a 23.
3 8 Aristotle s Criticisms of Plato
whole of this elaborate subdivision of Aristotle is entirely
beside the mark. Similarly, when he asks how it is possible
that the dyad should be a single essence (Qfow nm) exist
ing independently of its two units, or the triad indepen
dently of its three units, and proceeds to show exhaustively
that it cannot be the independent unity formed either by
subject and attribute, 1 or by genus and difference, or by
contact or chemical combination or position, again one is
impatiently tempted to demur. If the Platonists made each
number a closed concept different from every other, is it
likely they would have granted that such numbers were
mere wholes of units ?
This is the first difficulty that suggests itself. Aristotle
assumes that every number is made up of juoraSe? and
remains fettered in this orthodoxy 2 throughout his whole
exposition. He brings to bear the whole artillery of
dialectic against the absurdities which attend the postulate
of qualitative differences in the unit. We see that a unit
differs from another unit neither in quantity nor in quality ; 3
units have no difference in kind. But would not Plato have
admitted all this at once, merely adding that as regards
the Ideal numbers such objections were entirely irrelevant ?
Still graver misgivings arise on the perusal of M. 7.
1081 b i, 12 sqq. Whether the units are indistinguishable
or differ each from each, number must of necessity be num
bered by way of addition, e. g. the dyad by the addition of
another one to the unit, and the triad by the addition of
another one to the two, and similarly with the tetrad. This
1 1082 a 15 sqq.
3 So it appears to M. Milhaud. But the case of dVo/tot ypa^ai discussed
below (pp. 48 sqq.) suggests the probability that here too Aristotle is really for
the first time dogmatically establishing the subsequent (Euclidean) view
(cf. M. 7. 1082 b 15) which was already used by mathematicians in practice
(,1080 a 30). Plato, if he did disclaim all notion of novdSes (infra, p. 41), must have
been arguing against the perceptual unit of the Pythagoreans.
3 1082 b 4.
Aristotle s Criticisms of Plato 39
being so, it is impossible that the genesis of numbers should
be as they describe, when they generate them out of the
dyad and the one. Really when a dyad is produced it is a
part (fjiopiov) of the number three, and this in turn a part of
the number four, and so on with the following numbers/
In other words, since all number is Kara Trpoo-Oeo-w, and
the Ideal numbers are not, therefore the Ideal numbers are
impossible.
Aristotle, it is true, proceeds to note an objection which
might be made by the Platonists to the above argument. 1
It may be said (and this actually was their doctrine) that
the Ideal numbers can be produced in a manner that does
not involve addition ; e. g. four is a product of the Ideal
dyad and the indeterminate dyad, and not simply 3 + 1.
Aristotle answers that, if so, the Platonists will have to
admit the existence of three Ideal dyads instead of one,
since there will be not only the original Ideal dyad but
also the two dyads in the tetrad.
Even here Aristotle s commonplace notion of number
seems to obtrude. He first makes as an objection against
the Platonists exactly the dogma which they must have
made a merit of repudiating, viz. that one number is a
part of another; and then, in refutation of their own
doctrine that the indeterminate dyad Mays hold of the
determinate dyad and produces the tetrad (TOV yap Xrj^OtvTos
i}v 6uo7rotos), he seems to think of the tetrad as simply the
dyad repeated two times, i.e. 2 + 2.
In short, to prove there are no Ideal numbers, Aristotle
shows that the Ideal numbers are not arithmetical numbers ;
and to prove that the Ideal numbers do not come from the
one and the indeterminate dyad, he reiterates that the
arithmetical numbers come from addition. It is a plain
case of ignoratio elenchi and of the futility of argument
1 1081 b 21.
4 o Aristotle s Criticisms of Plato
where there is no common ground. All Aristotle can be
said to show is that Plato ought not to have called his i5e a
apse s by the name of number at all. He admits that for
what the Platonists wanted to prove, the vTro fleo-t?, 1 namely,
that the Ideas are numbers, their substitution for addition
of multiplication and derivation from first principles is
sound enough. 2 Where there is no addition one Idea
will not be contained in another Idea as a part. But this
difficulty is avoided only at the cost of a demolition of the
nature of number (noXXa avaipovo-iv, Met. M. 7. 1082 b 33).
The following sentence quoted by Syrian from Aristotle s
early work on Philosophy puts the whole question in a
nutshell : If it is any number other than the mathematical
that the Ideal numbers are, we could have no apprehension
of it. Not one man of us in a thousand understands any
other number than the mathematical (rts yap r&v ye TrAeurra^
f)fjiG>v avviricnv aXXov api6fj,6v ;) 3
The novelty, then, of Milhaud s theory of the Ideal
numbers lies not in pointing out the inadequacy of Aristotle s
criticism. Bonitz 4 had shown already how unsatisfactory
was the method of refutation adopted. Aristotle, according
to Bonitz, ought to have pointed out at once that apiQ^ol aa-v^
P\T]TOI is a plain contradiction in terms ; as it is, he has only
darkened obscurity. Nor again was it a new suggestion
to trace the identity of Idea and number to the participation
by the former in unity and plurality. 5 What Milhaud has
shown, however, is that Plato might just be the one man in
a thousand who could understand a number different from
the mathematical number . No other, it is true, seems to
be recognized even by modern mathematics, but it is
acknowledged that quantities like TT, Vz cannot be ex
pressed numerically by any combination of units, and it is
1082 b 33. > > < )
v. Zeller, p. 51? Plato, E. T.). Cf. Met. M. 7. 1081 a ia-x 4 .
Aristotle s Criticisms of Plato 41
therefore only natural if a mathematician like Plato, who
was at the same time equally great as a metaphysician,
should not merely have been dissatisfied with the ordinary
account of number as avvOeo-is ^waScoi/, 1 but have made an
attempt to replace it by another.
Milhaud s theory, however, is suggestive rather than
final. Three points may be noted in connexion with it.
(i) Aristotle expressly attributes to his antagonists often
using the words tocnrep </>ao-i the view that number is made
up of juomSes, 2 though these are not the same as in the
mathematical number. Thus in M. 7. 1081 a 23-5 he says
of the units of the Ideal dyad that on the theory of Plato
(6 Trpoiros CITTWZ;) their production is due to the equalization
of the great and small by the one J . He states explicitly that
all theorists, with the sole exception of the Pythagoreans,
based their number on the unit (^ora5iKoi>? . . .
The acceptance of Milhaud s theory therefore involves
acknowledgement of a very serious misunderstanding on
the part of Aristotle. Such total misrepresentation is not
altogether unintelligible in view of (a) the sentence above
quoted from the ITepl ^nAoo-oc/ua?, which shows Aristotle s
perfect conviction that the only possible number was based
on the unit, and (b) the probability or rather certainty that
Plato s later mathematical speculations were mixed up with
a great deal of Pythagorean fancy and symbolism. 4 Still
it is very hard indeed to suppose that had the Platonists
rejected all notion of /^oz-aSej they would not have made
this clear. And this objection has especial weight if we
1 Cf. Euclid, Book VII, def. a number is TO *K fj.ova.8oav avyKciftevov ir\^dos.
3 So too Aristotle frequently asks : Whence, on Platonic principles, comes
the Unit ? How do they derive it from the One and the Indeterminate Dyad ?
3 1080 b 30.
* This is admitted even by Blilhaud, pp. 309, 320, 326. It would account for
Aristotle s failing to distinguish the wheat from the chaff.
42 Aristotle s Criticisms of Plato
are to assign so important a place to the Ideal numbers as
Milhaud would have us do. Milhaud s view, indeed, seems
to come perilously near to the old esoteric theory of
Platonism, unless more definite allusions to the Ideal
numbers be discovered in the dialogues. 1
(2) On the other hand, the important passage Met. H.
3. 1043 b 32 seems to lend support on the whole to Milhaud s
hypothesis. Aristotle here asks in what sense substances
can be compared with numbers, for points of comparison
there undoubtedly are. His answer is, that if Ideas are in
any sense numbers, they must be so as closed concepts
(OVTUS 1043 b 33), and not, as some philosophers say, as each
a number of units. . . . Every substance must be an
actuality and a definite thing (eyreAe xeta K<H <iW rts), not,
as some say, in the sense that it is a kind of unit or point.
Now this passage shows clearly enough that Aristotle
objects to the Platonic identification of substance and
number simply because (as he thought) this was equivalent
to making substance like a unit or point. Since cmy^ai or
/xoyades are all qualitatively alike, whence on such a theory
(Aristotle asks) comes the uniqueness of things ? If number
can have a qualitative aspect, can be in any sense ao-vja/SA^ros,
Aristotle s query is answered. The Idea of the Good, as
described in the Philebus, is a unity of multiplicity, a one of
heterogeneous elements; it cannot be compared (as Aristotle
correctly enough points out) with the ordinary arithmetical
number, but why not with an a/nfyx6s acr^/SArjros ? Aristotle,
in his strenuous opposition to the Pythagoreanism in Plato,
certainly seems to have ignored that synthetic aspect of
number which his master had endeavoured to elucidate.
But (3) even if Milhaud s theory be accepted, Aristotle,
Cf. Zeller, Plat. Stud., there is almost no trace of the Ideal numbers in
the dialogues ; History (Plato, E. T., p. 254), the Ideal number theory < has no
place in Plato s writings . Ideas of numbers are common enough ; cf. f) TWV
>vais (Rep. 525 C).
Aristotle s Criticisms of Plato 43
though wrong in what he denies, is right in what he
affirms. With his insistence on definite and clear cut
conceptions, he will have nothing to do with any qualitative
aspect of number ; and it will be granted that on trying to
work out Milhaud s conception of a union of quantity and
specific quality many perplexities are involved. On
the other hand, however, (i) Aristotle is quite sound in his
own view of number, and (2) with his interest in biology and
development, he is really in all his attacks on the number
philosophy of the Academy where philosophy, as he says, 1
had been reduced to mathematics implicitly asserting that
there are aspects and departments of the universe, e. g. life
and mind, in which TJ /uerpTjrtK??, Plato s sovereign science of
measurement, is, if applicable at all, altogether inadequate
to reality. For even if we go to the opposite extreme from
Aristotle, and instead of ignoring the truth of Plato s theory
read into it the fullest possible significance, it is a theory
which reduces all the sciences to one that of quantity. 2
Besides pointing out that mathematics and numbers can give
no account of causality, 3 Aristotle insists on their abstract
nature, and holds that whereas the animate is prior to the
inanimate 4 the Platonists reverse this order. At one time he
seems to have been carried away by the mathematical ideal
of exactness (aKpi /3eta), but by the time he writes the De
Anima and the Metaphysics 5 he sees that after all Psycho
logy, as a concrete study, has really more claim to be
called an exact science than mathematics.
Fourth Problem.
The investigation of the rest of Milhaud s theory leads
straight to the problem of the Transcendence of the Idea
in the Platonic system. We have seen above that the
1 Met. A. 9. 992 a 32. 2 v. A. E. Taylor in Mind, 1903.
3 Met. A. 9. 991 b 9. * Met. M. 2. 1077 a 20.
6 Contrast Post. An. i. 27 with DC An. i. i. 402 a 2 and Met. E. i. 1025 b 7.
44 Aristotle s Criticisms of Plato
weight of Aristotle s critique is directed against the
Kaflo Aou, a universal predicate that is at the same time a
particular. Aristotle could not understand how the general
Idea could at the same time have all kinds of other
properties individuality, completeness, perfection. Now,
according to Milhaud, 1 he would have understood, had he
seen what Plato was thinking of in his Ideal theory, viz.
the analogies of mathematics. Plato s Ideas are not muti
lated and abstract universals, but, in one word, the pure
essences of the mathematician J . The Ideal circle, e. g., is
the circle as defined by its equation in the general form ;
it is at once ev ical TroAAd, since it synthesizes in accord with
one definite law a great multiplicity of positions. It is
participated in by particular circles, but this mode of
participation cannot be represented by any metaphor
borrowed from addition. Further, it is in a sense \upts,
outside the world of sense, for it is never adequately
realized even in the particular circles obtained by giving
numerical values to the terms of the general equation,
much less in the material circles of nature, which are but
feeble and imperfect adumbrations of the Idea. As for the
ovtrta of the Idea, of which Aristotle makes so much, it is
simply the being J of all eternal and immutable truths ; it
is a priori objectivity. Milhaud further tries to show, in
support of his identification of the Ideas with the essences
of geometry, that Aristotle is wrong in placing ra /xafl^ariKa
intermediate between the Ideas and the world of sense,
and that the Platonic dialogues afford no real justification
for his doing so. 2
It will be seen that this theory is not altogether new.
Lotze, as is well known, was convinced that by reality
Plato meant validity , and that when he spoke of the Ideas
1 Cf. A. E. Taylor in Mind, 1903.
3 The opposite view is maintained by Adam, Republic ii, pp. 159-62.
Aristotle s Criticisms of Plato 45
as X^/H S- he meant their eternally self-identical significance .
The ei6o? was valid before we thought about it, and will
continue so without regard to any existence of whatever
kind, of things or of us, whether or not it ever finds
manifestation in the reality of existence, or a place as an
object of knowledge in the reality of a thought . Plato s
transcendence, in short, means nothing but independent
validity*. The advantage of Milhaud s theory is that it
explains the blunder of Aristotle in a much more plausible
way than as the result of a mere ambiguity of the Greek
language. Xenokrates told an intending pupil who had no
mathematics that he could not enter the portals of philo
sophy \aftas yap OVK ex l? <jJ>iAo(ro$ias. The only question
is: Can Milhaud s supposition be admitted here? Has
Aristotle s supposed failure to follow the mathematical
thinking of Plato really led him on this question of
transcendence to a caricature of his master s philosophy ?
(i) The answer must be, in the first place, that such a
supposition is refuted by the testimony of Plato himself.
An unprejudiced reading of the Phaedo or Republic or
Phaedrus will unquestionably confirm Aristotle in that
interpretation of Idea and particular which, with his usual
terseness, he sums up in a word or two in the early part of
A. 6 of the Metaphysics. The particulars of sense are
outside of the Ideas , though receiving their common name
because of them (TO. alo-OrjTa napa TCLVTO. Kal Kara TCLVTCL Ae yecr&u
TI&VTO). The Ideas are definite natures and substances
separate from other things - 1
It may be granted to Lotze that even in the first draft of
his theory the ovo-ta which Plato aimed at expressing was
being in the sense of universal and eternal validity , and
that if (in the Aristotelian phrase) 2 we look to his intention
1 Met. I. 2. 1053 b 21 (Jivaeis rivls Kal ovaiai
8 Cf. Met. A. 3. 985 a 5 ; 8. 989 b 5.
46 Aristotle s Criticisms of Plato
rather than to his words we shall not quarrel with any
such conclusion. But, as Lotze himself really admits,
Plato does not succeed in distinguishing Reality (Sein, ovaia)
from Validity (Geltung\ and what was meant to be simply
independent of individual thought becomes (notably in the
Republic] a reality independent of all thought whatever.
When Plato, therefore, talks of the Ideas as kv rovy
vncpovpaviuor as eorwra Iv rfj Qvrei, he means precisely what
Aristotle expresses in more prosaic language by ova-ia
K e\ W/H 07x0^7 r&v alo-OrjT&v. 1
It need only be noted in a sentence that the natural
interpretation of the Parmenides is directly opposed to any
such theory as that of Lotze or Milhaud. The unre-
generate Socrates of that dialogue, i. e. Plato himself, had
previously, it is indicated, held a doctrine in which the
Ideas were (a) avra KO.O avra, which can only mean transcen
dent and self-subsistent ; and (b) x^P^t which describes
them in a negative way but means the same thing.
(2) Secondly, that Aristotle, who had the benefit of Plato s
own conversation and instruction for twenty years, should
never once have seen what Plato meant (according to
Milhaud) by the transcendence of the Idea and the par
ticular s participation therein, is simply incredible. Even
an utter distaste for mathematics would not explain such
a misunderstanding. Aristotle was the acutest mind of the
school, and where the fundamental problem of /ueflcfis was
concerned his universal curiosity was not such as to be
repelled even by the abstractions of the higher mathematics.
Yet he says in explicit terms that the nature of participation
1 A. 7. 1073 a 4, 5. It is curious that few have been found to dispute
Aristotle s statement that the fj.6pia xtapiara of the Platonic soul-division means
actual and not merely ideal severance (De. An. 4135 28 xupiara KaOdirtp rives
c^cum/), yet this separation is quite as much a hard saying as the self-depen
dent existence of the Idea.
Aristotle s Criticisms of Plato 47
was left by Plato an open question , and this is borne out
by the dialogues themselves. 1
(3) Moreover Milhaud s theory seems (a) unduly to de
preciate the mathematical intelligence of Aristotle, and
(b) conversely to modernize the thought of Plato to the
neglect of the historical development.
(a) Is it so certain, as is often assumed, that Aristotle
was a weakling in mathematics ? The very fact of his
being a member of the Academy already implies that he
could not have neglected the subject. Cantor, who refers
to his fine mathematical intellect (feinen mathematischen
Geist), notes his separation of Geometry from Geodesy, 2
just as Plato had previously distinguished Arithmetic
from Logistic. 3 Though the specially mathematical works
ascribed to him are lost, and though the Mechanics are
spurious and the Problems not to be relied on as evidence,
still even in the authentic works we have ample evidence
that he took the keenest interest in all the problems of
mathematics. Further it is curious that he seems to have
understood the famous Nuptial number , 4 the obscurity
of which has been proverbial from the days of Cicero
onward. In the Metaphysics 5 Aristotle says the uni
versal circle or circle in general (6 Ka96\ov KVK\OS) is
1 Met. A. 6. 987 b 14. The phrase dffxtaav Iv Koivy for en/ is often mistrans
lated. It cannot be rendered (as by Ueberweg) omitted to investigate (cf.
Gomperz, diese Frage haben sie unerledigt gelassen ; Bonitz in media reliquerunt
[Index 400 a 5; differently at 128 b 38]). It means they left over for subse
quent inquiry . Now this actually describes with complete accuracy what we
find in the dialogues. Cf. Farm. 133 a dAXd ri a\Xo Set focii/ & /j.tTa\an0dv(t.
This other way , however, is not to be found, and can at most only be
read into the dialogues. Why indeed may not the above words of Aristotle be
the missing reference to the Parmenides ? Cf. also Plato, Phil. 15 b, where again
the problem of p*0fis is raised but not solved.
a Met. B. 2. 997 b 32 sqq. 8 Cantor, i, p. 239.
* PoL v. 12. 1316 a.
5 Z. 10. 1035 a 33-b 2 (in a 34 we should read rts os with E) ; cf. n. 1037
a 2 sqq. 6 air\ws \ty6fJifvos KVK\OS has no v\ij : individual circles have VOTJTT) I/AT/.
48 Aristotle s Criticisms of Plato
a concept that has no matter*, not even vXrj ro^rr/, and
this would seem to be exactly what Milhaud makes of
Plato s Ideal circle, simply an algebraical equation. It
is a pity M. Milhaud did not think it worth while to con
tinue his mathematical researches as far as Aristotle. 1
(b) Interpretation of the old in the light of the new is
the very life of all philosophical exegesis. But where the
question is an historical one, as to how far one thinker
has understood another who was his contemporary, it is
a primary necessity that interpretation should be as closely
literal and objective as possible. Now Milhaud is not
only less than just to Aristotle in his desire to make the
most of Plato, but also tends to put the latter out of per
spective by crediting him with mathematical concepts that
are essentially modern.
We may illustrate this by means of the theory of indi
visible lines (foopoL ypa/^jutat), which will show that Aristotle
may be a sound critic even of Plato s geometry, and there
fore unlikely to misinterpret his master s philosophy owing
to alleged sciolism in Mathematics. This interesting theory
is usually ascribed to Xenokrates, but Aristotle had often
heard Plato himself state it to his pupils in lecture (TroAAd/a?
eriflet, Met. A. 9. 992 a 22). This genus (that of points) was
one of which Plato disputed the very existence. He said
the point was a geometer s assumption, and though he was
ready to call it the starting point of the line, the real starting
point, as he often used to lay down, consisted of indivisible
lines/ 2 It was a theory that was found very hard of com
prehension by the Greek commentators; thus Simplicius
1 A work by G6rland on Aristotle s Mathematics seems unfortunately, at least
to judge by Gomperz s review in Archiv of 1903, to be useless for purposes of
objective study.
54 The passage is a difficult one to render and difficult in itself. A very
different translation and application of the passage will be found in Milhaud,
PP- 340-3i whose treatment however seems far from satisfactory.
Aristotle s Criticisms of Plato 49
is lost in wonder that it should have been put forward
by such a * mathematical man as Xenokrates. Aristotle
brings an argument against it in the passage from which
we have just quoted, and it is refuted at length in a treatise
(Trepl drojuwif ypa/xjuwi/) written by one of Aristotle s pupils
probably Theophrastos.
Now a modern mathematician coming to this theory
might be able (in Aristotle s phrase) to give it an up-to-date
interpretation V He might say that Aristotle and his pupil
had misconceived and traduced a very important doctrine
no less, in fact, than a rough anticipation of modern
infinitesimals. Just as in modern mathematics zero =
a quantity smaller than any assignable quantity, so if the
line be conceived as diminished till it is smaller than any
assignable line, it becomes an aro^os ypajuijuuj, i.e.- a point;
not, however, an Euclidean point, but one from which, by
taking an indefinite number of them, it will be possible to
construct a line (fyx^ yP a /W ? > A - 9- 99 2 a 22 )- It might be
admitted that the view of Plato and Xenokrates was defec
tive compared with that of the moderns, because while the
modern view, with its phrase smaller than any assignable
quantity , does not deny the Euclidean conception of in
finity but simply dispenses with it, Plato, on the other
hand, by definitely talking of indivisible J (arojuoj) de
liberately puts in the place of Euclid s point without parts
something which actually has parts, but of which the parts
are practically denied. 2
Such a theory might quite conceivably be put forward,
and would not be refuted by an appeal to the authority of
Aristotle. For, it would be said, Aristotle and his pupil
) Met. A. 8. 989 b 6.
2 A very close parallel might be found in Herbart, who, distinguishing starre
Lime and stetige Linie, constructs the former out of points in just this non-
Euclidean way (cf. Marcel Mauxion, La Metaphysique d Herbart, pp. 115-16.
Paris, 1894).
D
5 o Aristotle s Criticisms of Plato
had not to the same degree as Plato come under the
influence of the new geometry . They assumed the
complete validity and sufficiency of the orthodox view
according to which the line is divisible ad infinitum. But
surely Plato knew this as well as Aristotle. The hitter s
whole refutation consists, it would be said, in the appeal
to Euclid ; he says the Platonists do not speak the
language of orthodox mathematics , their views being
quite peculiar to themselves . x
Such a theory might be made very plausible. But it
would undoubtedly be shattered on a careful consideration
of the development of geometrical thought after the time of
Zeno. 2 Zeno had shown once for all that the line was not
made up of an infinite number of points : consequently it
devolved on Plato to make a fresh start. He frankly
accepted Zeno s results. The point was simply a geo
metrical assumption , i. e. the mere mathematician J may
talk of the points of a line, but the philosopher sees that the
line is something quite different from the point and cannot
be explained as made up of them. It may be explained,
however, if it is made up of something homogeneous with
itself, i. e. of lines. Only they must be very small lines
so small, in fact, that they cannot be cut into smaller ; they
must be indivisible lines . Plato s view was partly right,
and marked a clear advance on the Pythagorean view. It
contained, however, a contradiction ; for, though a line can
be made by adding smaller lines, these smaller lines can
always be divided into yet smaller. It only remained for
Aristotle to point out this contradiction, and establish
s, Met. M. 6. 1080 b 29 ; idiat nvfs Sc^cu, Met. N. 3. 1090 b 29.
2 Cf. Hepi dr. 7pa/^., which begins by giving some of the reasons which led to
the doctrine. One of these is connected with the Ideal theory, 968 a 9 sqq. ;
another is the demolition by Zeno of the Pythagorean conception of the line,
968 a 1 8 sqq.
Aristotle s Criticisms of Plato 51
thenceforward the Euclidean view nav <rwexes
ets aet Siatpera. 1
As illustrating Aristotle s method of criticism, however,
one of his refutations of the indivisible line deserves
a little examination. It is in the chapter above quoted
Metaphysics A. 9. Aristotle is pointing out the difficulties
that attend the derivation of lines, surfaces and solids from
the Platonic first principles the one and the great and
small. After showing that their attempted derivation is
inconsistent with their own belief that the line inheres
in the surface, and the surface in the solid, Aristotle comes
next to the point. How, he asks, will the Platonist deri
vation show that the point inheres in the line ? Plato,
it is true, tried to evade the difficulty by saying there is
no such thing as the point. The line, according to Plato,
was not made up of points at all, but of indivisible lines ,
and therefore, if the line is derived from first principles,
nothing more is needed.
Then follows Aristotle s objection. The point must
exist; for lines, even if they are indivisible lines, must
have an end (Wpas) 2 , i.e. a point. Bonitz says this is
a petitio principii. So it would be, were not Aristotle all
through this passage arguing from the Platonic standpoint.
As he is himself careful to add, the same argument as
proves the existence of the line proves also the existence
of the point. 3 In other words, Plato says that surface is
the end of a solid and the line the end of a surface,
therefore, he ought, in consistency, to admit that the point
is the end of the line. Plato had seen that lines were
not made up of points, but unfortunately he had not gone
on to say that similarly planes could not be made out of
lines, nor solids out of planes. Aristotle s argument,
therefore, is dialectical, but perfectly justified.
1 Physics vi. i, v. passim. 2 992 a 23. 3 992 a 24.
D 2
52 Aristotle s Criticisms of Plato
It is unnecessary then to consider the details of Milhaud s
theory. It may be held as incontestable that Plato did at
one stage of his thinking hold a doctrine of transcendent
Ideas, such as we find refuted in Aristotle. 1 But now
comes the problem of the Parmenides. If there is one
thing which that dialogue attacks in every conceivable
and possible way, it is just this transcendence of the Idea.
And we have seen that this is the centre also of Aristotle s
attack. The proposition Substance cannot be separated
(x>pk) from that of which it is the substance summarizes,
according to Zeller, the whole difference between the
Platonic and Aristotelian systems ; it furnishes, according
to Bonitz, the summum ac praecipintm Aristoteleae et Plato-
nicae philosophiae discrimen . Here then we are face to
face with the fundamental dilemma already mentioned
what we may call the Parmenides- Aristotle dilemma.
Of this dilemma it has been usual for historians of philo
sophy to accept the first horn that Plato never abandoned
the self-subsistence of the Idea. This view must commit
itself to unnatural interpretations of the Parmenides 2 ; it
tends to minimize either the force of the arguments there
stated or the importance of the whole dialogue ; or again
an easy solution which is no solution it declares the
dialogue spurious.
Further, the Parmenides does not stand alone. If it did
Plato might be regarded, though unwarrantably, as a meta
physical Ariel , writing the Parmenides in an hour of
insight . But in the Sophist also Plato criticizes the
1 Not of course that he consciously held it in the definite and dogmatic form
to which Aristotle, with his preciser terminology, reduces it. Every philosophy
necessarily suffers injustice in being thus restated.
2 Such e.g. as that of Zeller, Plat. Stud., pp. 159-94. Apelt, again, has
:numphantly vindicated the genuineness of the dialogue, but he does so only at
:he cost of ranking its philosophical importance quite low : he calls it < ein
wahres Arsenal von Erschleichungen und Sophismen a < dialectical witches
babbath , &c.
Aristotle s Criticisms of Plato 53
friends of the Ideas , with their doctrine of transcendence
(ovatav x^pts TTOV bLeXopevoi) and their severance of Being and
Becoming (ovo-fa and yeVeo-ts), and in the declaration of the
same dialogue that to go about to separate off (airoxvpiftiv)
one thing absolutely from every other is the very anti
thesis of true philosophy , we seem to find, though the
immediate reference is logical, the spirit of the later
Platonic metaphysic as a whole. Plato seems to have
got beyond the sharp antithesis of the Republic between
seeing and thinking V and to have come to recognize
that the world of knowledge was not a different world from
that of perception, existing independently of it.
But there are difficulties equally great in the way of
accepting the second horn that Aristotle had not the
ability to understand Plato s later Idealism and attributed
to him the crudest form of the theory as the form most
easy to refute. Such a view might indeed appeal to the
many supposed cases of unfair argument used by Aristotle
in his strictures on the Ideas. It is said that he argues
from his own point of view and thus unfairly attributes to
opponents the result of his own deductions. But even if
this were established, 2 it does not make it any the more
intelligible that Aristotle should, from the very first, have
1 Rep. vi. 507 B ra n\v opaaOat <fxifjicv, vofiaOai 8 ov, ras 5 av Ideas voeia&ai jj.& f
opaaBai 5 ov ; cf. in Bk. vii 77 01 fyfcus (f>aivofj.fvrj <i8pa )( o VOTJTOS roiros.
2 A very clear case might be supposed to be afforded by Met. Z. 6. 1031 b 15,
where Aristotle says that if the Ideas are such as some people assert them to
be, then the substrate in other words the particular cannot be substance
(oucrta) . This is urged by way of objection, though it is obvious that Plato (at
least in the first stage of his thinking) would not have admitted the ovata of the
particular. But even here is it not the case that Aristotle is refuting the
Platonists from their own premises? His argument is directed against that
view of the Ideas which makes them like the gods of the popular religion, only
differing from the men in whose image they are made in being a Stot.
Such a view of the Ideas might well commit itself to the assertion attributed to
the Platonists by Aristotle that the non-sensible substances are more substan
tial than the sensible, because they are eternal" 1 (Met. Z. i).
54 Aristotle s Criticisms of Plato
set himself in opposition to the otherworldness of the
Platonic philosophy, had it really ceased to present that
character. The obscurity of Plato s later teaching drove
many from his lectures, 1 and has left traces of itself in
certain passages of Aristotle 2 ; but surely the latter, if
any one, was qualified to understand him.
Other theories finally have sought to avoid the necessity
of taking either side of the antithesis. Plato did abandon
the self-existence of the Ideas and yet Aristotle has not
misrepresented him. Here the most attractive view is one
already partly discussed that which holds Aristotle to
have been aware of Plato s disavowal of transcendence
and to be attacking consequently only the earlier theory
of Ideas. The criticism, it is noted, takes place within the
school, and attacks a doctrine which has several different
and contradictory forms. 3 The arguments are served up
afresh from the Trept i8e<Sr, because that doctrine of exag
gerated transcendence, which even Plato had found it
necessary to censure in some of his pupils, was still
rampant in the Academy at the time when Aristotle put
together his Metaphysics*
Now it may be perfectly correct to say that Aristotle is
attacking an earlier theory of Ideas , but the great diffi
culty is just that he knows of no later theory. He constantly
mentions Plato s theory of first principles (orotxeta), but so
far is he from the knowledge of any change of front with
regard to the Ideas that, on the one hand, Platonists who
might certainly be described as friends of the Ideas are
represented as holding the later doctrine of the One and the
1 Rose, p. 24. 2 e. g. De An. i. 2. 404 b 19 sqq.
3 A. 9. 990 b 9, b u, b 21, 992 a 32.
* From the Platonic side this theory has to face two difficulties : (a} that of the
Titncuus, 51 C sqq., where the Ideas, regarded from the point of view of the
Parmenides, are everything they should not be (51 C, E, 52 A); (6) the difficulties
of identifying the Ideas in the Philebus, with the class of TO irtpas.
Aristotle s Criticisms of Plato 55
Indeterminate Dyad, 1 and, on the other hand, conversely,
Plato in his later philosophy of first principles is still credited
with a pre-Parmenidean doctrine of Ideas. 2
Moreover, though Plato himself is not once mentioned
in the criticisms of Met. A. 9 and Ethics i. 6, it is impos
sible to suppose he is not included in the refutation. 3
Similarly, though in Met. B and Z Aristotle is clearly
attacking the contemporary Academy and a crude doctrine
of eternalized sensibles (di6ia ato-^ra) which was never
held by the master himself, yet Aristotle nowhere says
anything to indicate that the Platonic view in its logical
consequences would not be open to the same difficulties.
He gives it explicitly as Plato s doctrine that he believed
in three orders of existence (ovo-tat), 4 and nowhere is it
stated that he changed this view. In short, the theory
only acquits Aristotle of direct injustice by exposing him
to the same charge indirectly.
Our fourth problem then has evidently reduced itself to
the problem of the Parmenides, which is a standing enigma
in the Platonic philosophy. The interpretation here adopted
of that dialogue seems the natural one, and if accepted it
is impossible to suppose that Plato ever recanted his own
recantation. But there is as yet no agreement as to how
he modified his doctrine, nor is it certain that he ever
found himself in a position to meet satisfactorily the
difficulties of the Parmenides and the innumerable others
in addition to them \ 5
1 Met. A. 9. 990 b 18. 3 Met. A. 6.
3 Aristotle begins the refutation in A. 9 with the words of 5 rds t5as amas
riOffJifvoty but he uses the past tense fKufiiffav b 2, irporj\0ov b 6. This may of
course refer still to none but the Platonists, but it is forced, especially as it is
the case that Aristotle frequently refers to Plato in the plural. Nevertheless it
may be admitted that a single mention of Plato by name (for his view of the
point) and a reference to a single dialogue (the Phaedo) are not what we should
have expected had Aristotle been really attacking a doctrine of Plato s.
* Z. 2. 1028 b 20. 5 Parwt. 135 A.
56 Aristotle s Criticisms of Plato
So much for the thesis ; the antithesis is that Aristotle s
criticism cannot be adequately explained unless the an
tagonists he is refuting actually held a doctrine of tran
scendent Ideas. 1 It is meaningless except as against the
theory of a noumenal world which is a timeless reproduction
of the phenomenal but does not explain it, seeing that the
two are divorced from each other. There is no difficulty
in attributing such a view to members of the Academy ; for
the doctrines of Speusippos and others on the separate
and independent existence of numbers are obviously
a heritage from, and to be paralleled with, the early
Platonic theory of Ideas. But can it be attributed also
to the Plato who wrote the Parmenides and the Sophist
and the Philebus ?
We have here a case of conflicting evidence, and the
data seem hardly sufficient for a solution. The Aristo
telian method of working through the difficulties 2 has in
this case led to little positive result. The dilemma above
stated has of itself no necessary cogency, 3 but the difficulties
which lead up to it have been neither evaded nor solved.
The problem is still sub iudice*
1 On any other theory not one of his criticisms but would fall lamentably
flat, and Aristotle was too keen a dialectician not to have noticed this at once.
Thus take the amusing chapter (Z. 14) in the Metaphysics in which Aristotle turns
the tables on the Platonists. The latter held the Idea was the sole definable ;
Aristotle, however, after showing that of particulars there can be no definition,
proceeds : Neither then can any Idea be defined. For it is a particular, as they
say (us . . . </>a<rt), and separable. Nothing could be more unlikely than that
Aristotle here attributes to the Platonists a mere unwarranted deduction of his
own. So again in Eth. i. 6 it is the Platonists (as Stewart says) who confound
the true with the spurious eternity atSiov with troXvxpoviov.
3 De Caelo iv. i. 308 a 5 ISovres ovv irpurov rd irapa ruv d\\ojv dpij^va, Kal
Siairoptaavres KT\., ib. i. 10.
* Thus we have shown above that the talk of plagiarism has no relevancy,
s solution will to some extent depend on the possibility or otherwise of
tmg a consistent doctrine from the very difficult chapter A. 6 of the
etaphysics. Two diropuu in connexion with the chapter may be noted : (i) if the
:nng principles of all things (<7To<xa) are the Good and Matter, why the
Aristotle s Criticisms of Plato 57
But whatever the solution of these difficulties, the essence
of Aristotle s criticism will still be justified. There is
a very fundamental difference between master and pupil
in their doctrine of the real. The real had been for Plato
TO. OVTUS fora, the Ideas ; Aristotle surprisingly, inconsis
tently, and yet naturally enough, agrees that this is so
in the case of the highest owux, the Deity. 1 But in the
concrete world the spirit of the observer and student of
nature predominates over the metaphysical tendency to
dualism which he had inherited from his master ; and the
merit of grasping firmly and clearly that the universal
exists in and through the particular, and that the existence
of the particular is in and for the universal , 2 and of carry
ing this doctrine consistently through the whole phenomenal
world, indubitably belongs to Aristotle.
Fifth Problem.
The fifth and last problem brings us to what Aristotle
has to say on the subject of Plato s aetiology.
(i) His main charge in the indictment of Transcendent
Idealism is, that it cannot furnish any explanation of the
world of change and becoming (r&v (fravep&v rd cunor). 3
Thus, after giving his own explanation of yez>e<ns in the
Metaphysics* he proceeds to show that the Ideas (17 T&V
flb&v atria) do not contribute at all to bring about generation
and substances. For (a) t if the form were a self-subsistent
(Platonic) Idea, and existed in that sense, no " this " would
ever have been coming to be. The form signifies the
" such " or the " what ", but it is not a " this " or a " deter-
need of the Ideas as formal causes ? (2) if this be satisfactorily solved, what is
the relation between the One or the Good to the Ideas (Formal Causes) ?
1 Who is pure Form, TO ri -qv dvai TO irpurov (A. 8. 1074 a 35)
2 R. B. Haldane, The Pathway to Reality, p. 52.
3 Met. A. 9. 992 a 24; cf. 991 a 8 -navroiv 5t /zdAtcrra SiatToprjfftifv av TIS KT\.
4 Z. 8.
58 Aristotle s Criticisms of Plato
minate something " V (b) In some cases, viz. the birth of
natural objects, it is matter of plain experience that the
Ideas have nothing to do with the generation. In nature
like is generated by like, man by man, not by the Idea of
man ; and yet, since natural objects are especially Mai, it
is here that the Ideas would be most required. Similarly
it is the doctor, not the Idea of health, that produces health ;
the scientific teacher, not the Idea of knowledge, that pro
duces knowledge. And if Ideas were the causes, why are
they not constantly in operation? Aristotle sums up his case
in Met. A. 10 : The Ideas are not causes at all, but even
granting that they are, at least they are not the causes
of motion (ovn Kiy?i<recos ye). In short, just as Leibnitz
misses final cause in Spinoza, so Aristotle misses efficient
cause in Plato.
Apart from Lotze s remark on the non-efficiency of the
Ideas that neither do our Laws of Nature contain in them
selves a beginning of motion, it might be retorted to
Aristotle by the Platonists that their master had never
said the Ideas could supply an apxn (KU^O-CCOS) yeveWws. In
all Plato s later writings, at all events, the efficient cause
is soul, mind, creator. 2 But (i) as against the Phaedo,
where the Ideas are made the sole efficient causes,
Aristotle s argument is valid, and (2) it is extremely
probable that Plato in his later lectures had made no
mention of efficient causes. He seems to have used no
1 Pseudo-Alexander here remarks that on the Platonic view (a) there might be
avvOtois, as of the bricks that go to build a house, but no yeveais ; (6) just as this
particular wine and this particular honey, if separate existences, may make up
mead but cannot be found in any other mixture, so if avroavOpu-rros is x J P lffT ^ v t
it may in combination with this particular matter produce Socrates, but can
give rise to no other individual till severed from the matter of Socrates
(Hayduck, 496. 20). With Aristotle there is a growth of form into matter
( = formed matter) ; he no longer, like Plato, makes the cause of phenomena
something different from them.
-* Cf. Laws 896 a ^ux 7 ? Atcra^oA^s re ical Kivrjatus anaarjs atria airaaiv.
Aristotle s Criticisms of Plato 59
other causes than his two first principles, the One or the
Ideas, and the Indeterminate Dyad; he probably said
nothing of the Demiurge so often mentioned in the
dialogues, nor even of soul as source of motion. Other
wise Aristotle s objection, that Plato s mathematical matter*
cannot explain motion, would lose all its point. 1
It is no doubt surprising to find that notwithstanding
his attack on Plato, Aristotle himself reduces his four
causes to two, and on the principle of always finding the
ultimate ground should trace back the efficient cause to
the formal. 2 But though the efficient cause of a house to
Aristotle is ultimately the form of the house in the mind of
the builder, still he does not absorb the efficient cause in
the formal ; he recognizes the efficiency of the art of
building or of the builder.
Again Aristotle is justified in the strictures he passes
on Plato s use of the term participation . He says that
Plato cannot tell the cause of the participation ; and if
we answer, with Bonitz, that the cause is the efficient
cause, it must be further asked : In what way is Plato s
efficient cause an atrtoy TTJ? /*e0eea>9 ? Only as a deus ex
machina. Aristotle substitutes for the static conception of
participation and conjunction (//e flefi?, o-wovo-ia, Met.
H. 6) his own idea of growth and development.
(2) After his exposition of Platonism in Met. A. 6,
Aristotle considers it obvious from what he has said
((pavtpov K r&v iprjfj.v(av} that Plato recognizes only two
causes formal and material. From the Platonic dia
logues themselves a very different impression results.
Already Alexander asks the question why Aristotle
refuses to allow to Plato efficient and final causes. But,
1 Met. A. 9. 992 b 7 TTfpt TC Kivfiaecas, cl fjtev . . . tl Se ^17, voOev rj\0w t cf. also
Phys. T. 2. 201 b 20 evioi, erepor^Ta KCU dvifforrjra KOI TO pri ov QaaKovTfs eivat
Phys. ii. 3. 195 b 21.
6o
Aristotle s Criticisms of Plato
to illustrate Plato s recognition of them both, Alexander
might have appealed to much more telling passages than
those he quotes from the Timaeus and the Seventh Epistle.
Thus (a) in the statement at least of universal efficient
cause, no one could be more emphatic than Plato. In
the Sophist the production of animals, vegetables, and
minerals is assigned to God the Artist (0e6s bwiovpy&v).
In the Philebus the cause of the mixture of Limit and
Unlimitedness (rrjs <TVW IWS ^ curia) is thereby the cause
also of genesis, and may be identified with active power
and artist (br^fjuovpyos). Sophist, Timaeus, Philebus, Laws
are in this respect alike. 1
Similarly (b) as to final cause, not to mention the descrip
tion of the Ideas as Archetypes (napa^dy^ara) and of the
Idea of Good in the Republic as not merely highest efficient
but also final cause of the universe, there is to be found in
the Philebus, where Plato completes his theory of causation,
both divine and human, and indicates the four Aristotelian
causes, the very closest parallel to Aristotle s description
of the Deity as the final cause of the universe for which all
the rest of creation yearns and strives. 2 And in Plato s
latest writing, in one and the same passage along with
universal efficient cause (6 roO -navTos e7ri/xeA.ov/xei/os), we have
the following explicit assertion of final cause 3 : Each
part of the universe . . . has the whole in view. This
and every other creation is for the sake of the whole, and
in order that the life of the whole may be blessed. You
are created for the sake of the whole and not the whole
for the sake of you. Every physician and skilled artist
does all things for the sake of the whole, directing his
1 v. Campbell, Sophist, Introd., p. 76.
2 Even here, however, it is noteworthy that the distinctive note of Aristotle s
conception is wholly lacking K i v c T us t
Laws 903 B-C. For explicit assertion of soul as curio ^ra!3o\r]s re noi
KtVT}(rtajs aird<TT)s, v. 896 A-B.
Aristotle s Criticisms of Plato 61
effort toward the common good, executing the part for
the sake of the whole/
Aristotle then does not do justice to Plato s aetiology.
At the same time, if the following considerations be taken
into account, it will be seen that it is in no spirit of grudg
ing depreciation that he finds deficiencies in his master s
doctrine.
(a) As appears from the words (fravepdv ex T&V efynjji^ww,
Aristotle is thinking not of the Platonic dialogues but of
Plato s lectures especially those On the Good . Now
in these the dynamical interest seems to have been entirely
overshadowed by the ontological. 1
(b) Aristotle does not wholly deny Plato s recognition
of final and efficient causes. As to the former, Aristotle
says that in a sense it was postulated by Plato, only not
qua final. That is, Plato identifies it with the formal cause,
and it is only an accident of the formal cause that it
happens at the same time to be good. The Ideas are
final causes, not a7rA<3?, but only Kara (rvpfaprjKos. As to
efficient cause, Plato, like other philosophers, saw it as it
were in dream. 2 In other words, Plato wished indeed to
make his Ideas efficient powers, but seeing that this is
what in Aristotle s opinion they cannot be, Aristotle can
on occasion deny to Plato s system the recognition of any
efficient cause whatever. In a similar vein he says that
no one has clearly assigned even the formal cause, 3 though
the Idea-philosophers (ot ra etSr; rifleVes) come nearest it.
This simply means that Plato s formal cause is not quite
the same as his own. It will be obvious, therefore, that
(i) Aristotle s account of the system presupposes his criti
cism of it, and (2) he refuses to recognize Plato s maker
and father of the universe as any scientific explanation,
1 v. Alexander on A. 6. 988 an (Rose, p. 42).
3 De Gen. Corr. ii. 9. 335 b 8 sqq. 3 Met. A. 7.
62 Aristotle s Criticisms of Plato
and thus eliminates efficient cause from the Platonic
metaphysic. 1
(c) Finally, it is easy for us now to see in the Dialogues,
notably the Philebus, anticipations of Aristotle s doctrine of
the four causes, but only because Aristotle himself has
brought to clear and definite expression the various scat
tered hints of his master s teaching. Nor can it be denied
that the Platonic exposition leaves much to be desired, as
regards both clearness and adequacy. Aristotle feels this
so strongly with reference to Plato s external, as contrasted
with his own immanent, teleology that, forgetting his own
concession elsewhere, he once roundly asserts that the
final cause is not touched by the Ideas . 2 Again, what
is the relation of the Idea of the Good to other ends
(Ideas) or to the special functions (epya) 3 of things?
Efficient causes Plato attributes at one time to Ideas, at
another to soul : which is his real doctrine ? and what is
the relation of Idea to soul ? Aristotle, therefore, while
willing to admit that Plato made stammering efforts
in the direction of efficient and final causes, 4 was per
fectly justified in thinking that he had not fully worked
them out . 5
It is now possible to sum up the positive results arrived
at:
i. The evidence is against the supposition that Aristotle
has misapprehended the Platonic first principles.
1 If 6 0eos is simply popular in Plato for the highest Idea (cf. Zeller, Plato,
E. T., p. 267), then since Aristotle holds there is no efficiency in the Ideas,
efficient cause will naturally in his view disappear from the Platonic system as
a whole.
2 A. 9. 992 a 32. s Cf. Eth. End. i. 8. 1218 a 30.
4 Met. A. 10. 993 a 15. In Aristotle s favourite phrase (cf. A. 4. 1070
10) rpoirov fifv nva itaoat (sc. of amat) irporepov ttpyvTai, rpoirov 8e nva
^Alexander on A. 6. 988 a n (Hayduck, p. 59. 30-60. a), Rose, p. 42 d\\
i(ipya<raTo n ifepl avruv.
Aristotle s Criticisms of Plato 63
2. Aristotle is correct in what he says of the contents of
the Ideal world.
3. On the Ideal numbers Aristotle is at cross purposes
with Plato. Each is right in asserting what the other
denies.
4. Aristotle has exaggerated, but not caricatured, the
transcendent objectivity of the Platonic Idea. The Par-
menides problem is still unsolved.
5. Aristotle is severe on the Platonic aetiology, but not
without justification.
Before completion of the inquiry, by showing how far
the peculiar characteristics of Aristotle s censure of Plato
admit of explanation on general principles, it will be well
to consider very briefly a few of the main criticisms in the
field of Physics, Ethics, and Politics. 1
B. Aristotle s Criticisms of the Timaeus .
As to Physics, a volume might be written on the criticisms
of the Timaeus alone. Aristotle paid particular attention
to this dialogue, not for its metaphysics and its mysticism-
like the Neoplatonists but because it contained all that
Plato had to say on Aristotle s favourite subject the
natural sciences and biology. With its myths and its
mystical mathematics it must have roused all the scientific
spirit of Aristotle into opposition, and that no radical mis
understanding, and certainly no conscious unfairness, can
be proved against him even here is strong proof of the
painstaking consideration 2 which Aristotle gave to all
Plato s opinions, and of the deep respect which he always
paid to the memory of his great master.
1 For Aristotle s criticism of Plato s Logic, especially of the method of
fiiaipfcrts, v. H. Maier, Die Syllogistik des Aristoteles, ii. 2, chapter i, 3 (* Die
Entdeckung des Syllogismus ), pp. 56 sqq.
2 Bacon misconceived this when he compared Aristotle to the Turk (more
Ottomanorum}.
64 Aristotle s Criticisms of Plato
i. Thus it is at first surprising that Aristotle, in pro
ceeding to discuss growth and qualitative change, 1 should
say that Plato s investigations extended only to generation
and destruction, and not even to all generation but only to
the generation of the elements. As to how flesh or bones
or anything of that kind came into being, he has made no
investigation. Now these latter subjects certainly are
considered in the Timaeus* and Plato has also there
treated though very briefly of growth and decay (av&o-ty
and </>0i<ns), but if we look at what Plato says about them
Aristotle s language is easily explained. Aristotle could
have no sympathy with an account which, he might have
said, made marrow out of tiny triangles 3 and imported J 4
the Deity (6 6e6s) 5 into a scientific explanation. In fact it
is clear that Aristotle passes over Plato s account deliber
ately, for he goes on to say, ( Not one of these subjects
(qualitative change and growth) has been treated in any
thing but a superficial way by any one except Demokritos
... no one has said anything about growth which might
not equally well have been said by anybody* (6Yt //r) KO.V
6 n>x<0j; ciireiev). 6 Moreover, in other works, Aristotle does
note Plato s view of respiration and his theory on the
absence of flesh from the cranium, both of which come in
the passage of the Timaeus which is here overlooked.
Aristotle, it is plain, never minces words, but it is only
a very abstract view that can discover detraction or un
fairness in this passage, and in the implied contrast of
Demokritos with Plato and the Pythagoreans.
1 De Gen. Corr. i. 2. 315 a 26. a 73 sqq.
3 Tim. 73 B. As Aristotle had already refuted Plato s derivation of the
elements, he might well in any case think himself able to dispense with special
notice of his theory here (De Gen. Corr. i. 2).
4 Eth. i. 6 ft (ray ay civ ret, 1677. 5 Tim. 73 B, 74 D, &c.
3 I 5 a 34 o\o}s 5J irapcL ra (TmroXfjs irfpi ovSevbs ouSet? (ireffTrjafv fco ArjfjtoKpirov
KT\. The phrase 6 /XT) KO.V o rv\wv tiireitv recurs in Meteor, i. 13. 349 a 16.
Aristotle s Criticisms of Plato 65
2. As is well known, Aristotle takes the Timaeus
literally almost throughout, 1 and an interesting passage
in the De Caelo - shows him to have been perfectly aware
of the reproaches that might be made against him for
doing so. According to Xenokrates and other defenders
of Plato (nves) t Plato s declaration that the world had come
into existence was intended merely for purposes of
exegesis (Si5ao-Ka\ia? y&P LV \ J ust as a geometrical in
structor may represent the gradual coming into exis
tence of a geometrical figure. Aristotle replies that the
parallel will not hold. It is possible to show a geometrical
figure in the making, but there all the parts can exist simul
taneously. In the question at issue, however, when they
say that out of chaos there comes to be a cosmos, these
cannot be simultaneous ; they are prior and posterior, and
to separate off what are prior and posterior there must
necessarily be generation and time/ 3 This objection,
which is perfectly valid as against Xenokrates, only proves,
according to Zeller, 4 that not only Aristotle, but even
Plato s defenders as well, did not recognize the full
extent of the mythical in the Timaeus, the chaos itself
being simply part of the allegory.
Now this illustrates admirably the difficulty of ever
coming to an anchor when once embarked on the sea of
mythical interpretation. Every one will allow it to be
mythical when the Demiurge in the Timaeus 5 mixes
various ingredients in a mixing-bowl. But soon real diffi
culties begin. Aristotle, with his usual acumen, pointed
1 The one exception seems to be the Srjfjuovpyos, on whom Aristotle is silent.
The word in the Platonic sense occurs only once in all his writings in one of
the early dialogues (Rose, p. 29).
2 i. 10. 279 033.
3 Whereas in the case of Siaypa.fjiij.aTa, ovtev ry \p6vcf Kex&pwrai. Cf. on the
whole passage Simplicius (Schol. 468 b 42).
* Plat. Stud., p. 211. 6 41 D.
E
66 Aristotle s Criticisms of Plato
out as a contradiction in the Timaeus that Plato generates
time in time V Xenokrates, to meet Aristotle, puts forward
an attempted solution. Aristotle refutes this and straight
way others, to meet the refutation, declare that the chaos
also is pure allegory 1 . Zeller does not agree with the
Neoplatonists in taking figuratively 2 Plato s derivation
of the elements. Yet, as Simplicius naturally asks, When
so much of the Timaeus must be taken metaphorically, why
not this also ?
In short, even had Aristotle adopted this method of criti
cism with full deliberation, he would still have been justified.
Better the literal interpretation of Aristotle than the
allegorical methods of the Neoplatonists. Whichever
method be adopted, the words are still true which Aristotle
uses of the Timaeus on another question, that what is
written there has no explicitness . 3 The Timaeus, as
Hegel puts it, is the most difficult and most obscure
among the Platonic dialogues , and though the authority
of Aristotle need not establish his way of taking the
Timaeus to be the only one, that he did take it literally is
certainly no proof of his inability to read aright the strictly
philosophic doctrines of Plato. 4
3. Again, in Psychology, Plato s doctrines of the world-
soul meets with no gentle treatment. His probable tale
(which Plato himself had admitted might not be found
1 Physics 0. i. 251 b 17 sqq.
2 ovuPoXiKUK, Simplicius, De Caelo iii. 252 b 23 (v. Baumker, Das Problem der
Materie, p. 169). Why not also the diremption of the soul .
3 ou5fi/a ?xt Siopta^v, De Gen. Corr. B. i. 329 a 13 sqq. Aristotle is saying that
it is impossible to make out from the Timaeus whether Plato s matter can exist
otherwise than in the form of the four elements. He is thinking of the so-called
secondary matter , which certainly does introduce a difficulty into the question
Aristotle is discussing, whether matter can exist x^p^rrj. Archer-Hind miscon
ceives the passage (Timaeus, p. 179).
Cf. Gomperz, Griechische Denker, vol. ii, pp. 483 sqq., on the difficulties of the
Timaeus. He finds Aristotle justified.
Aristotle s Criticisms of Plato 67
1 everywhere and in all respects consistent and accurate ) 1
is taken by Aristotle with complete literalness and criticized
accordingly. In the first place then/ he begins, it is not
correct to say that the soul is a magnitude (fxeyeflo?). 2
This sounds at first extremely unfair, as we know that to
Plato the soul is immaterial. By magnitude, however, it
must be remembered, Aristotle means geometrical magni
tude, quantity qua measurable 3 (e.g. a mathematical
line).
Now the Platonists, as is known from various evidence,
disputed as to whether the soul was arithmetical or
geometrical, a number or a magnitude, but they had no
doubt as to its being one of the two. Zeller thinks Plato
had not expressed himself definitely in favour of one view
or the other, and left the relation of soul to his mathematical
principle (ra ^a^fxartKa) undetermined 4 ; hence the diver
gence on this question between Speusippos and Xeno-
krates, the latter defining soul as a self-moving number .
Consequently Aristotle has not grossly misinterpreted
the mathematical description of the Timaeus, and his
amusing literalness 5 may, after all, be no great injustice,
though we feel that Plato does\oot bear at any time to be
interpreted so literally and dogmatically. 6
Still the chapter in De Anima 7 is by no means open to
the charge of quibbling commonplaceness . 8 It is not a
sympathetic criticism (since it does not allow for possible
development of opinion on Plato s part), but it is nevertheless
perfectly correct to point out that there is a fundamental
1 Tint. 29 C. 2 De An. i. 3. 407 a. 2. 8 Met. A. 13. 1020 a 9.
4 v. Zeller, Plato (E. T.), p. 355 n. 8 Archer-Hind, Timaeus, p. 114.
6 A more indulgent critic than it was Aristotle s nature to be would have
hesitated before ascribing to a great thinker such a patent contradiction as exists
between the Phaedrus (245 E) and the Timaeus (34 B) in regard to eternal motion,
v. Met. A. 1071 b 37 sqq. He would have asked : May not Plato s meaning be
other than the narrative form of the Ttmaeus compels his words to be ?
7 406 b 25-407 b 26. 8 Wallace, De Anitna, Introd., p. 36.
E 2
68 Aristotle s Criticisms of Plato
contradiction between the view of the Timaeus and that of
the earlier Phaedo l in regard to the union of soul and
body. When Aristotle further says on the perpetual
motion of the world soul that this will be violent 2 and that
consequently the soul will enjoy no opportunity for leisure
or rational amusement , but will have the lot of an Ixion
on his wheel ( I&ovos polpav)* there is here no unfairness
whatever. Aristotle is careful to exclude all Matter from
his own conception of the transcendent mind or of Deity,
and simply makes his point here in the most vivid way at
his disposal.
Further, Aristotle is strongly opposed to the Platonic
view that movement is a predicate of soul, or that soul
is the self-movent. 4 Again, his fundamental objection to
all theories of the class to which Plato s belongs is that
they assume it as possible for any soul to clothe itself in
any body after the manner of the stories of the Pytha
goreans . 5 As well expect a carpenter, says Aristotle, to
do his work with a flute. Aristotle s real criticism of Plato
is simply his great conception of soul as the form or
realization of the body, and his real difference from
Plato, here as elsewhere, comes out not so much in his
dialectical criticisms as in the course of his own scientific
exposition. Every one, nevertheless, will acknowledge the
applicability of his criticism of Plato s faulty psychology ,
however Aristotle himself may have failed to maintain the
organic unity of soul. 6
4. As to the nature of Platonic matter, Aristotle s opinion
Is that Plato gives space as its essential definition, i.e.
identifies matter and space. This interpretation, though
often called in question, 7 still holds the field.
1 407 b 1-5. 2 407 b i. s De Caelo B. i. 284 a 27.
4 407 a 32. * 407 b 13 sqq. 6 411 b 5.
7 One of the difficulties is that Plato strenuously rejects the void and so
Aristotle s Criticisms of Plato 69
In one passage, 1 however, Aristotle s method of reading
philosophy backwards results in a considerable variation
from his usual account. He says that Plato identified
Matter with privation , i. e. the direct contrary of Form.
Teichmuller stigmatizes this unheard-of reproach as a
1 crying injustice 2 to Plato. But Aristotle s statement is
very easily explicable, and he has himself (even in this very
passage) supplied us with the means of checking his own
deductions. 3 He is discussing Plato s Matter from the
point of view of his own system, according to which
Matter and privation are differentiated from each other.
Now Aristotle is correct in saying that Plato had not
distinguished these two, and the Platonic Matter, more
over, is certainly not that of Aristotle, whose concep
tion was very different. But to say therefore that
Plato identified his Matter with Aristotle s privation is
while a natural enough conclusion plainly quite un
justifiable.
Connected with this is the question whether Aristotle
means to include Plato among those who said Matter was
the bad . If he did, this would be another injustice to
Plato, arising from the above identification. For if, in
Plato s system, Matter is simply the * privation of the One,
i. e. the Good, plainly Matter is identical with Evil. But
though Aristotle states that Plato makes Matter the ground
of evil and refers to its baneful power on the Platonic
theory, it is almost certainly Xenokrates alone to whom he
alludes as identifying Matter with the evil principle , and
often uses its impossibility to explain certain phenomena that he may be
called the author of the theory of horror vacui ; v. Baumker, pp. 179-80, on
this difficulty.
1 Physics i. 9.
2 eine schreiende Ungerechtigkeit (Studien zur Geschichte der Begniffe).
3 192 a 10 ^XP L /* " 7"P <5/>o nporjXOov cm 5ef nva viroiceiaOai $voiv KT\.,
which means that the Platonic matter after all is more than * non-being .
70 Aristotle s Criticisms of Plato
therefore of this further misconstruction of Plato Aristotle
stands acquitted. 1
4. Still less reason is there for impugning the value of
the authority of Aristotle on the question of Plato s deri
vation of the elements. According to Mr. Archer-Hind,
Plato was presumably as well aware as any one else of the
impossibility of forming solids by an aggregation of mathe
matical planes ... it is entirely preposterous to suppose
that the most accomplished mathematician of his time was
not fully alive to a truth which, as Aristotle himself admits,
e7rt7roA?js eoriv iSeuj. 2 But not only have we the plain
evidence of the Timaeus that in this respect Plato was still
under Pythagorean influence ; the Academy after him, as
we learn from Aristotle, 3 and as we have seen above,
maintained the same doctrine, viz. that solids could be
built up out of planes. As Zeller says, Aristotle here
understands the Platonic doctrines quite correctly. 4 Even
M. Milhaud, who is not disposed to underrate the Platonic
mathematics and on this point suggests a new explanation
by taking Plato s space as full space , admits that Plato s
theory is an extremely curious one . 5 Milhaud is cer
tainly wrong, however, in saying that Aristotle in this
connexion confounds Demokritos with Plato . 6 In a
1 A. 10. 1075 a 35 rb KO.KOV avrb Oarepov rwv aroixfiuv ; cf. 0. 9 ; N. 4. 1091 b 35
T-a avioov = jj TOV Karcov Qvffts. Bonitz (p. 588) thinks Plato alluded to as well as
Xenokrates in this last passage. He refers in proof however merely to
A. 6 fin. (988 a 14), which says that according to Plato evil is caused by v\rj ; cf.
TO KO.KOTTOIQV avTrjs (Phys. i. 9. 192 a 15). It is expressly said to be Pythagorean
to set up KO.KQV and ayaOov as absolute opposites (Met. A. 5. 986 a 26). Baumker
(pp. 205-6) thinks this doctrine of Matter as the bad can be ascribed to the
later Plato, but it has not been shown even that Aristotle does so.
8 Archer-Hind, Timaeus, p. 202 n. This is but one among many instances of
the partisan spirit in which throughout his edition of the Timaeus he champions
ato at the expense of Aristotle. Cf. p. 184, where Aristotle is declared to
have no right to contradict the nineteenth-century hypothesis of Dr. Jackson.
Met. A. 9 . 993 a Io _ 23 with Alexander ad loc.
e M e !l er Plat (E T ) P- 3? 5 n - 5 Milhaud, pp. 299, 3*0.
8 Milhaud, p. 303.
Aristotle s Criticisms of Plato 71
striking passage 1 Aristotle expressly distinguishes the
logical atomism of Plato and Xenokrates from the physical
atomism of Demokritos. The latter, he says, put his trust
in theories that were physical, i. e. appropriate to his
subject ; Plato, on the other hand, had never been at
home in the physical sciences . 2
5. Finally, a very interesting problem is presented by
a passage in Aristotle s De Caelo. 3 Aristotle is discussing
the question Is the earth stationary or not ? and,
according to the reading of Simplicius and the best
manuscripts, writes as follows : Some say that the earth
rests on its centre and is piled up about and revolves
around the axis of the universe, as we read in the Timaeus.
It is now universally admitted that Plato thought of the
earth as stationary, and the only question is, How explain
the remark of Aristotle ? Has he misread the Timaeus
and misrepresented Plato?
Gomperz 4 thinks Aristotle is alluding to Plato s conver
sation or lectures after the date of the Timaeus, and finds
a confirmation of his view in a passage of the Laws 5 where
Plato alludes in a mysterious way to the newly promulgated
doctrine of the youngest Pythagoreans, that the earth
revolves on its axis. The passage, however, does not
support this hypothesis, 6 and had Aristotle heard the
doctrine from Plato personally he would have said so.
Undoubtedly the right explanation is that Aristotle is here
1 De Gen. et Corr. i. a. 315 b 30 sqq. With equal explicitness Plato is con
trasted with Leukippos in i. 8. 325 b 25.
a ocroi fvwKrjttaoi pa\\ov V rots QvaitcoTs KT\., 316 a 6.
3 De Caclo ii. 13. 293 b 30 d\tiffdai KOI KivciaOai Trepi KT\. The above trans
lation would be the literal one (d\(ta9ai, formed into a ball, globed round ) ;
but probably the two words are used synonymously, /tat being explicative.
The Berlin text gives {\\ea0ai irepi, omitting *at KivetaOai.
* Griechische Denker, ii, p. 609 n. 5 vii. 821 sqq.
6 Moreover, Aristotle says nothing about the earth s own axis, but, like the
Timaeus ,4oC), uses the phrase 6 5ia uavros rtra^tVos iro\os, i. e. the axis of the
universe .
?2 Aristotle s Criticisms of Plato
speaking of the interpretation given to the words in the
Timaeus by the later Platonists, who returned to the old
Pythagorean doctrine that the earth with the other heavenly
bodies revolved around the central fire . The Platonists
misinterpreted the semi-obsolete 1 word which had been
used by Plato in the Timaeus ; and Aristotle, whether he
made this mistake himself or not, gives to the passage the
interpretation of contemporary Platonism.
C. Criticisms in the Politics.
Hegel s fine remark, that Plato was not ideal enough ,
applies to his metaphysics when he is compared with
Aristotle, but hardly to his Ethics and Politics. Here
we feel that of the two great philosophers the deeper
mind was Plato s. Hence it is no mean testimony to the
fairness and ability of Aristotle as a critic that his discus
sion of Plato s Republic in the second book of the Politics 2
is generally admitted to be not merely the best of all his
criticisms of his master, but at the same time one of the
most interesting and trenchant passages in the whole of
the Politics. The crispness of the language, the neatness
of the rejoinders, the practical common sense with the philo
sophic penetration that goes beyond it, the judicious sanity
of its estimate of revolutionary schemes, have made it
a model of criticism for all time. It is a thoroughly
gentlemanly criticism, 3 and the odd nature of certain of
1 Semi-obsolete, i.e. in the sense which Plato still gave to it. On the
whole passage, v. Journ. of Phil, v, p. 206 (Campbell). The Platonists natu
rally took the word fl\\onfvr)v to mean rolling ; cf. Arist. Meteor. 356 a 5,
where it is used in this sense ; v. further on the passage, Zeller, Plato (E. T.),
pp. 380-1 n., and Archer-Hind s note on Timaeus 40 B (pp. 132-3).
3 Politics ii. i sqq.
3 Its real philosophic character may be better appreciated if it is compared
with the attitude of others who have taken it in hand to castigate Plato,
whether in the tone of rabid abuse or ridicule which Plato himself anticipated
(Rep. v) or in the narrow, prejudiced and offensive manner of De Quincey
(v. his collected works, Masson, vol. viii).
Aristotle s Criticisms of Plato 73
the objections, coupled with the presence of one or two
at first sight inexplicable misapprehensions, admits, we
shall see, of very easy explanation.
The tone of the chapter on the Laws is different. 1 It is
occupied exclusively apart from the question of over
population with what are, comparatively speaking, details,
and has been excellently called a somewhat grumbling
criticism . 2 The reason is fairly obvious ; the constitution
of the Laws though the mathematics and religion of that
work give it a wholly different appearance from the
Politics is really very close to that of the ideal state of
Aristotle himself. He had reason enough for being dis
satisfied with the Laws 3 and his real criticism is the
Politics itself. But, whereas in the case of the Republic
he could easily point out a sufficient number of airopCat
to justify him in constructing a new ideal state, this is not
so easy with the Laws. Hence the criticisms in general
are trivial and in some cases unjustified. 4
D. Criticisms in the Ethics.
As for the famous criticism in Ethics i. 6 only three
brief remarks may here be made :
(a) This is one of the clearest of the cases in which
Aristotle s arguments, when compared with the exposition
of his own doctrine as a whole, are seen to be mere
Socratic fence. There is a great difference between the
two philosophers, both on the special question of teleology,
and on the connexion of Ethics with Metaphysics, and
morality with religion. But this is not brought out in the
criticism at all.
(b) The contention 5 that the Aristotelian categories
1 Politics ii. 6. 2 Newman, ii, p. 264. 3 Newman, i, pp. 449-54.
* v. Newman s notes, ii, pp. 264-81, especially on 1265 a 39, 1265 b 19 and 22,
1265 b 31, 1266 a i, a 13, a 17.
5 v. Burnet, Ethics, Introd., p. 1.
74 Aristotle s Criticisms of Plato
were accepted by the contemporary Academy would
certainly make the arguments less unreal, and bring the
passage more into accordance with Aristotle s favourite
method of refutation. But the evidence for such a sup
position is of the smallest, and Aristotle constantly
elsewhere uses his logical engine of the Categories for
purposes of overthrow.
(c) It must be admitted at once that, as against the Plato
of the dialogues, the criticism is a failure. The main
point of the chapter seems to come, so to speak, in the
postscript : the universal good is abstract and transcendent,
Xo>/oioToz; avro TL KaO CLVTO. This might apply to the Republic :
it certainly does not to the Philebus. But Aristotle is
probably thinking little of either ; he has in view the Idea
of the Good as it had become in the treatment of the
Platonists, or indeed in the later treatment of Plato him
self, when he reduced the Ideas to Ideal numbers, and
therefore naturally identified the Good with the One.
To this One, Aristotle tells us, as also to the numbers*
Plato attributed an existence independent of real things
(napa. TO. TT pay para). 1
The only other important criticism of Plato in the Ethics
concerns the doctrine of pleasure. Aristotle has here also
been supposed unfair to Plato, but in this case without
reason. For (a) Zeller, 2 who talks of Aristotle s perverse
apprehension of Plato s utterances on this subject, does
not distinguish between Aristotle s criticism in Book X of
the Ethics and that in Book VII. In the latter there is no
reference to Plato whatever ; Aristotle attacks Speusippos
or other theorists who had used the arguments of the
Phaedo or Philebus to support an indictment against
pleasure, (b) In Ethics x. 3 3 Plato s theory of pleasure
as a yWis is attacked, and Aristotle at first sight conveys
1 Met. A. 6. 987 b 29. 2 Plat. Stud., p. 283. 3 1173 a 31 sqq.
Aristotle s Criticisms of Plato 75
the impression that in his account of the painless delights
of knowledge, sight, &c., he is stating an important new
truth. But the explanation is that Plato had certainly
attempted to explain even the pure pleasures as TrAr^co-
oreis, 1 and so had supported the theory of pleasure as
a yeWi? all along the line. The pure pleasures, though
not preceded by pain, certainly are preceded by KCVUHTIS
and ^6eta, so long as these are imperceptible. Odours,
on this theory, would be the food of the nostrils, and
there would be pain felt at the absence of smell did not
the K&OXTIS or depletion of the nostrils happen to be imper
ceptible. Aristotle simply asks if Plato can point out the
ZvbtLd in the pleasures of knowledge, smell, sight, music,
memory or hope. Plato would have to answer that it
could not be shown, it was merely hypothetical, an assump
tion in order to make his theory consistent throughout.
There is consequently nothing at all disingenuous 2 in
Aristotle s criticism. And though the other arguments are
slighter, there is no excuse whatever for the remark that
as usual, Aristotle s objections miss the point J . 3
Conclusion.
Nothing is easier than to cry out against Aristotle s
misunderstandings and perversions of his master s meaning,
but it is much more profitable to try what can be done by
way of explaining them., As this explanation has already
unavoidably formed great part of our inquiry as to how
far Aristotle has actually misrepresented Plato, it only
1 Tim. 65 A ; cf. Phil. 51 B and Rep. 584 C.
2 very disingenuous, Stewart, Ethics ii, p. 417, but his note on 1173 b 13 at
once explains this statement and disproves it.
3 Archer-Hind (Timaeus, p. 236), who mistranslates the passage Eth. x. 3.
1173 b 5 (v. Burnet) and does not say a word of Aristotle s most important
argument. This is one of many cases in which it might be found that Aristotle
is at a much less remove from King and Truth than his critics, and more
correctly apprehends Plato s thought than the lattcr s would-be champions.
76 Aristotle s Criticisms of Plato
remains to sum up under a few general heads some of the
main reasons which lend to the criticisms an appearance
of perversity, captiousness or unfairness, which is really
quite foreign to Aristotle s intention.
Fortunately there is here no question of any of the
motives which actuated either Leibnitz s criticisms of
Spinoza or Schelling s of Hegel. 1 There is here nothing
of that acrimonious hostility which has sometimes dis
graced the philosophy of the moderns; none of the
systematic depreciation by Leibnitz of the arch-heretic
Spinoza, to whom he owed so much ; none of the bitter
rancour with which Schelling pursues Hegel ; none of the
scurrilous abuse lavished on the latter by Schopenhauer.
Of impatience in the criticisms, of causticity, of the pun
gency 2 which is illustrated for us by the surviving specimens
of his wit, there is certainly no lack 3 ; but of acrimony or
personal ill-feeling a review of all the passages reveals no
trace or shadow. Zeller has shown how little weight is
to be attributed to the gossip of the little men of a later
age. Against the tales of an Aelian we have not only
better evidence on the other side, we have the express
testimony of Aristotle himself. In a famous sentence of
the Ethics he tells us that Plato and Plato s friends were
his friends, but not to the prejudice of the sacred claims of
truth. In the Politics* he pays a graceful tribute to his
1 v. Stein, Leibnitz und Spinoza, pp. 229, 252 sqq., &c., and for the relations
of Hegel to Schelling v. Lecture on this subject included in Hutchison Stirling s
What is Thought, &c., pp. 249 sqq.
2 v. the Ka\\iara airo^e^nara in Diog. Laert. Bk. v. II, 17-20.
3 TO. yap 1877 x ai P* TO} > Ktvo\oyftv, OLTOTTOV HOI ctivvarov, Kevuv kan iravrcXus
(De Sensu 437 b 15) : nxdruvi /xcWot \(KTOV (Phys, iv. 2. 209 b 33) : Met. A
29. 1025 a 6 6 j/ T$ iirniq. \6yos TrapoKpovtrai : N. 3. 1091 a 10 : N. 4. 1091 b 26
rro\\r) rts fviropia ayaOwif.
4 Politics ii. 6. 1265 a ii. We may compare one of Spinoza s references to his
father in philosophy, Descartes. In his theory of the Affects , according to
Spinoza, the celebrated Descartes nihil praeler magni sui ingenii acumen
ostendit (Ethics iii, Preface).
Aristotle s Criticisms of Plato 77
master s writings : All the discourses of Socrates alike
are characterized by brilliancy, grace, originality and the
spirit of inquiry/
Aristotle then might at least say that he loved the man
and did worship his memory this side idolatry as much as
any . But not only so, we have actually some evidence
that Aristotle and Eudemos worshipped Plato as a god, 1
whom a bad man could not mention even in praise without
blasphemy, and to whom even a worthy pupil, such as
Aristotle, preferred to allude indirectly, so as not to take
his name in vain . For what other reason does he so
often criticize Plato in the plural number or as Socrates ,
if not to avoid calling attention to the differences between
himself and his revered master ? 2
No explanation, therefore, can be accepted which refers
to personal reasons, the constant sharpness or occasional
unfairness of the criticisms. The theory of deliberate or
purposive misunderstanding can at once be ruled out
of court.
To come then to verae cansae. (i) Aristotle, some thir
teen years after Plato s death, appeared at last as the head
of a new school. As against the rival Academy he had to
justify himself to the world for doing so, and he is therefore
inevitably concerned to find differences from his master
just where there was most appearance of indebtedness or
similarity. In Leibnitz s criticisms of Spinoza we find
exactly the same thing ; only Leibnitz makes the assertion
1 v. Wilamowitz-Mfillendorf on the well known elegy to Eudemos ( Aristo-
teles und Athen sub fin.).
2 Similarly Aristotle (after the Topics) seems consistently to avoid express
mention of Xenokrates, who was at the head of the contemporary Academy.
We know that Aristotle and Xenokrates were great friends ; yet the latter is
certainly not spared in attack, e. g. in De An. i. 4. 408 b 32 his opinion is, of
all those discussed, TroAv aXoywrarov. Simplicius observes (Schol. 488 b 3)
that it is always simply Plato s 5oa which is the object of Aristotle s attack.
78 Aristotle s Criticisms of Plato
that there is no Spinozism in any part of his teaching 1 :
Aristotle, on the contrary though for the above reason
his direct expressions of agreement with Plato are fewer
than they otherwise might have been has yet, considering
the impersonal nature of all his work, 2 rendered in the
most unequivocal terms his rpo^cia of gratitude for the
master s teaching.
Teichmuller, 3 it is true, holds that if Aristotle had been
quite just to Plato he would have put his own services to
philosophy in the shade, seeing that his own doctrine is
nothing but a systematized Platonism. But neither state
ment is adequate, and certainly not the latter. Aristotle
does advance beyond Plato, and he is not throughout his
works if indeed he is at any time a mere Eristic seek
ing to prove these advances against his predecessor/
Teichmuller exaggerates the element of opposition to
Plato, 4 and takes one single explanation of it as by itself
sufficient.
(2) (a) Aristotle is arguing against contemporaries (ol vvv).
The master had been dead for over fourteen years, but his
more commonplace pupils in the Academy were living and
active, and Aristotle, the founder of the biological sciences,
had little sympathy with their Pythagorizing substitution
of mathematics for concrete philosophy.
(b) It is Plato s lectures rather than his written dialogues
of which Aristotle is mainly thinking in his references.
In the Topics? e. g. he cites three instances of novelty of
1 Stein, Leib. und Spin., p. 230.
Ka6air(p teal 6 yevvatos HXarcav (pijaiv in De Mundo 7. 401 b 24 is just one
of the indications that this work is spurious. It is felt at once that Aristotle
could no more have written like this than Thucydides.
3 Studien zur Geschichte der Begriffe, Berlin 1874.
4 Thus it is nothing but the wish clearly to define his position that leads to
the phrase i)/iefs8e (pafj.fv after the statement or refutation of a theory of the
Platonists or Plato (cf. De Genn. et Corr. 329 a 24, Phys. 192 a 3).
5 vi. 2. 139 b 32.
Aristotle s Criticisms of Plato 79
epithet from Plato, and not one of these is to be found in
the dialogues. 1 Again, it is a very striking fact, that with
all Aristotle s attacks on the Ideal theory, only a single one
of the Dialogues is ever alluded to in connexion with it.
This is the Phaedo, and here he appeals no less than three
times 2 to one identical passage 3 which seems to have
strongly (and unfavourably) impressed itself on his memory.
(c) Some of the misunderstandings are probably simply
due to confused and imperfect recollection of passages
which he did not trouble to refer to. Just as in his fre
quent quotations from Homer he may sometimes be very
wide of the mark, as when he attributes to Calypso words
which are not even those of Circe but are actually spoken
by Odysseus to his pilot, 4 so in quoting Plato he constantly
forgets the connexion. Thus in the sole reference that
can be found in Aristotle to the Politicus* he has not only
carelessly misquoted the passage, but alludes vaguely even
to its author by the very extraordinary phrase Some one
in former time (m . . . rG>v irporepov). Zeller 6 does not do
justice to the strangeness of these words when he says
that here the definite person whom Aristotle is thinking
about is more distinctly and clearly referred to than in
the other anonymous mentions of Plato. v Ei/iot and nvcs
and ot \cyovTcs are regular : n? -r&v irportpov is unique. The
reference remains singular though not unaccountable . 7
Again we are told that Aristotle had made abstracts or
1 Cf. De Gen. Corr. ii. 3. 330 b 16 ftaOdirfp TI\6.TQjv ev rats 8iaiptffrtv, and De Part.
Anim. i. 2 (Zeller, Plato, E. T., pp. 46-7).
2 One of these (Met. M. 5. 1080 a 2) is a duplicate of A. 9. 991 b 3. The other
is De Gen. Corr ii. 9. 335 b 10.
3 Phaedo 100 B sqq. * Ethics ii. 9. 1109 a 31.
5 Politics iv. 2. 1289 b 5 ; cf. Polit. 303 A, B.
6 Plato (E. T.), p. 63 n. As we have seen, Aristotle s mode of anonymous
mention is not the indirectness of disparagement, as it is e. g. in Leibnitz s
Scriptor quidem subtilis at profanus (of Spinoza).
7 Campbell, Introd. to Polit. p. 55.
8o A ristotle s Criticisms of Plato
epitomes of the Republic and Timaeus. 1 If, after doing so,
he thought he might in future consult his memory in
preference to documentary evidence, we have an explana
tion of occasional perversities of allusion. 2 Aristotle leaves
us with the impression that he did not know the Republic
so well as he ought to have done.
(3) We have already seen traces of Aristotle s intense
dislike of the mythical in philosophy. In a passage of the
Meteorologica 3 he says it is ridiculous (ytXolov) to suppose,
like Empedocles, that one has given any explanation by
talking of the sea as the sweat of the earth . For
purposes of poetry, no doubt, this is adequate enough
(metaphor being an adjunct of poetry), but for a scientific
knowledge of nature it is not/ This feeling appears already
in the Topics f where, in the censure of some metaphorical
definitions (all of them seemingly Platonic), it is remarked :
1 Everything said metaphorically is obscure/ Consequently
he has a very real objection to Plato s poetic metaphors \ 5
Of Plato he might have reversed his dictum on Empedo-
kles and said he was a poet rather than a physicist , 6 just
as even his language was half-way between poetry and
prose. 7 Aristotle for the first time introduces a definite
philosophical style ; so too he is for maintaining the
independence and severity of science. He thought it high
time that the mythical should be banished from philosophy.
Its only raison d etre is that the true facts are unknown or
uncertain. And in such a case Aristotle thinks that the
scientific procedure is to say so
1 For the Timaeus v. Simplicius on De Caelo 284 a 27 (the passage on the
world soul), Schol. 491 b ; cf. Zeller, Arist. (E. T.), i. 62.
e. g. Politics ii. 5. 1264 an, 36, b 15. But v. infra, pp. 86 and 87.
" 3- 357 a 24. * vi. 2. 139 b 32.
He missed airovbrj airoSeiKTiitr], A. 8. 1073 a 22.
(f>vffio\6yov fj.d\\ov rj iroirjrrjv of Empedokles (Poetics i. 1447 b 19).
Diog. Laert. iii. 37 (Rose, p. 78).
8 De An. ii. 2. 413 b 25 ; cf. 403 a 8 and Rodier ad loc.
Aristotle s Criticisms of Plato 81
In spite of all this it still no doubt remains unfair to
treat Plato s poetry as though it were science. But if
Aristotle (conformably with his own principles) had refused
to take any notice at all of Plato s fairytale of science , he
would have been thought still more unjust. As it is he
never says of any of Plato s opinions what he does say of
the Pythagorean notion of time, that it is too ridiculous
to investigate its impossibilities .
Parallel with the dislike of the metaphorical and the
mythical is Aristotle s objection to a priori deductions in
the field of Politics. This explains the sharpness of his
criticism 1 on Plato s ideal history of evil in Books VIII
and IX of the Republic. It is not the case that Aristotle
seems to have understood Plato s account as an attempt
to describe the actual facts of Greek history . This would
be incredible in itself (for Aristotle could not suppose Plato
to have been ignorant of the history of his own native
Athens) and is refuted by a careful reading of the passage.
Most of the objections are really on the basis of Plato s own
theory, though Aristotle follows them up at once with
a statement of the actual facts. Aristotle, as he admits
himself, is never an l indulgent critic, 2 and his concrete
mind is not satisfied with Plato s attempt at a philosophy
of history . It is sound, he thinks, neither as the one nor
as the other.
(4) The great philosopher may write a valuable and
excellent history of philosophy, as is proved by the first
Book of Aristotle s Metaphysics, and by its modern parallel,
Hegel s Lectures. But such histories will not be so reliable
objectively as had they been written by lesser men ; con
sequently we are not surprised to find the same charges
made against Aristotle as have also been made against
1 Politics v. 12. * N. 3. 1090 b 14.
F
82 Aristotle s Criticisms of Plato
Hegel. Aristotle, in a word, discusses previous thinkers
from the standpoint of his own system.
An excellent example is furnished by his investigation of
the concept of Space. 1 Plato had nowhere in the Timaeus
expressly discussed the nature of Space as such. But
Aristotle has asked himself as usual: What have my
predecessors taken Space to be? And the answer is
perfectly natural and inevitable : Plato identifies it with
Matter (v\-n). Zeller, therefore, is quite correct in saying
that while Plato asks the question What is Matter? and
answers Space, Aristotle asks the question What is Space ?
and makes Plato answer Matter . 2 Aristotle would himself
have admitted that Plato s problem after all had been differ
ent from his own ; he says before beginning his inquiry,
that he has no previous discussions to go upon. 3
Aristotle more than once in this way discusses under
Physics what had been given by Plato as rather of meta
physical interest. A curious and somewhat different case
is where Aristotle in the Meteor ologica* after discussing
why the sea does not swell in volume with the mass of
river water that flows into it, roundly declares that what
is written in the Phaedo 5 about rivers and the sea is im
possible , and proceeds to *show how. This, as has been
said, is like testing the geography of Dante s Inferno
by the laws and discoveries of physical science . 6 Still
in a sense it is really more of a tribute to his master
than a criticism. Aristotle is aware that Plato has no
scientific theory on the question he is discussing, but
he thinks it worth while giving an exposition and
criticism even of his mythical or probable account in the
Phaedo.
1 Phys. iv. 2. 2 Platonische Studien, p. 212.
3 Phys. iv. i. 208 a 35. 4 355 b 34. 5 in C.
e W. D. Geddes, Phaedo, p. 151.
Aristotle s Criticisms of Plato 83
Still another example may be taken, this time from the
Metaphysics. Aristotle says that Plato in the Sophist
identifies Not-being with falsehood (rd x/^OSo?). 1 Now
Plato in that Dialogue proves that if T& ^ ov is existent,
then such a thing as x^eCSos (ifrcv&qs Sofa, \lfevbrjs Ao yos)
becomes possible. But Aristotle, seeking to find an answer
as to which of the three (Aristotelian) kinds of Not-being
Plato had been thinking of when he used the word, has
naturally but wrongly been led by the words of the Sophist
to identify Plato s Not-being with his own not-being in
the sense of the false (TO w ov o>s \l/vbos).
It is obvious that this accommodating procedure will
sometimes lend an appearance of great caprice to Aristotle s
interpretations of Plato. But even yet whole histories of
philosophy are written under the shadow of the fallacy
that the problems of one age or thinker are present in the
same way to every other.
(5) Aristotle is the analyst par excellence, and, aiming at
definiteness and clearness of doctrine, he is not content till
he has reduced every theory to the special yeW to which
it belongs. This is a natural result of his subdivision and
systematization of all the departments of philosophy. In
Plato s Republic we find together (even in the same book)
Physics, Psychology, Ethics, Politics, Metaphysics; Aris
totle has separate compartments for all of them. The
difference between the two minds comes out very clearly
in a well-known passage of the Politics? where Aristotle
alludes to the * extraneous discourses with which Socrates
has filled the Republic. We here, if anywhere, catch a
glimpse of the real Aristotle from under his mask of
impersonality, and the pupil who compiled the Magna
Moralia reproduces the genuine spirit of his master when
1 N. a. 1089 a 19.
3 Politics ii. 6. 1264 b 39 rots (cuO(v \6yois ireirXrjpojKC TOV Aoyoi/ KT\.
F 2
84 Aristotle s Criticisms of Plato
he says : Plato was wrong in mixing up virtue with his
treatment of the Good ov yap o^/ceio^. 1
This frame of mind will obviously not be the best for
doing complete justice to Plato. Further it goes along
with an attention to details and individual results, which
lends to some of Aristotle s remarks on Plato an appearance
of very carping criticism what Teichmtiller calls Krittelei.
But this only means that in the words of the Parmenides,
philosophy has now taken a firm grip , 2 and the philosophic
thinker no longer fears the falling into some bottomless
pit of absurdity 3 by discussion of the seemingly trivial
and unimportant. Plato in his later dialogues had him
self here shown the way.
Nor again is it any discredit to Aristotle that his anim
adversions should often take the form of a criticism of
language. Himself the creator of a technical philosophic
vocabulary, he could not neglect the terminology of others.
Thus his first few arguments against the Republic of Plato
are footnotes on the ambiguity of the words unity and
all . 4 He was reproached for this tendency even in
antiquity; thus Philoponus 5 says (wrongly) that in re
proaching Plato for identifying space with the participant
and yet not locating the Ideas in space, Aristotle as usual,
attacks the mere word (viz. space). Similarly the modern
critic, speaking of Aristotle s discussion of Plato s theory
of vision, says it is impossible to exonerate it from the
charge of ovo^ar^v Orjpcvo-is } . 6 But if so, the case in point
would prove that philosophy was nothing else than the
kind of word-catching which Aristotle is here accused of.
The passage (De Sensu c. 2) is quite fair. Plato had
attempted to explain why we do not see in the dark. 7 It
1 Mag. Mor. i. i. 1182 a 28. 2 Farm. 130 E. 3 Farm, 130 D.
4 Politics ii. a. 5 Quoted in Baumker, p. i8i 2 .
6 Archer-Hind on Timaeus, p. 157. 7 Timaeus 45 C sqq.
Aristotle s Criticisms of Plato 85
is because the light issuing from the eye is changed and
extinguished when the air it meets has no fire in it.
Aristotle replies that extinction is here a wholly irre
levant concept; it applies to fire or flame, but neither of
these terms can be predicated of light. 1 His own explana
tion makes no use of fire. 2
(6) Lastly, and most important of all, comes the fact we
have so often had occasion to notice, that Aristotle s criti
cisms are dialectical. This means strictly that they are argu
ments based not on true premises, but on premises admitted
by the other side. But the word can be used loosely of
all difficulties (dTro/ncu) 3 that rest on popular premises in
general. The aporetic method proceeds on the principle
that if a sufficient number of shafts be levelled at a target,
some of them at least are bound to hit the mark. In the
Platonic dialogues Plato contrives to let us see when his
arguments are not serious ; in Aristotle, however, the
method has stiffened, the procedure looks more dogmatic
and more of an insult to the reader s intelligence. Yet
Aristotle himself tells us what to expect ; his method is to
register all possible objections (ras evbexwevas aTroptas). 4
And that he is true to this plan is easily proved.
For (i) it is impossible otherwise to explain the frequency
with which objections good, bad, and indifferent are heaped
up together or jotted down in parenthesis with no regard
for order and system, and no link of connexion except his
favourite particle en. One excellent example among many
is afforded in Metaphysics M, where after his main refutation
of the Ideal numbers, the attack is renewed in c. 8, and
a fusillade of varied objections follows, some of them of an
1 437 b 15 sqq. 2 v. De An. ii. 7 ; De Sensu c. 3.
3 Also 8vax ( P ; n> 8voxep eta h ra P a X*l> 8ua*oAtai. Syrian (in Met. 1080 a 9) calls
the arguments against the Ideas ktrixtiprj^ariKol r6iroi.
* Met. A. 7. 988 b 21 ras <i/5ex<ty*Vas airopias 8if\0oj^(v irepi avrwv. Then follows
the criticism of the earlier philosophers (c. 8) and of the Academy (c. 9).
86 Aristotle s Criticisms of Plato
extremely questionable character. So again in the Politics x
Aristotle assumes in one passage that Plato s community
in women and children is to be limited to the guardians ;
in another, 2 after propounding it as an open question
whether, according to Plato, women, children, and property
are to be held in common also by the agriculturists, he
proceeds to set forth the difficulties on either supposition. 3
(2) Not only, for this reason, is it true that many of the
criticisms are weak and do not seem to bite ; others actually
contradict Aristotle s own rulings or remarks elsewhere.
Thus one of the proposals of Plato s Laws that of the
double homestead which Aristotle criticizes in the Politics*
as fatal to domestic economy, is, after all, adopted by
himself. 5 So again he objects to the Platonists that they
make matter the source of multiplicity, for probabilities ,
analogies , and first appearances are against such a
view. 6 At first one wonders if this passage is not a deser
tion of Aristotle s own first principles, till it is remembered
that Aristotle need not himself believe in the validity of
the objections he presents to opponents. One more ex
ample may be cited, from a chapter which is full of
argumenta ad homines as also ad Platonicos. The doctor
does not consider health in general, but the health of man,
or rather of this particular man ; it is the individual that
the doctor cures. 7 Aristotle s own doctrine recognizes
both the particular and the universal side of the art of
medicine, as of all arts 8 ; but it is easy to see which side
will be emphasized when he is making a point against the
Platonists.
1 1262 a 40. a 1262 a 14.
3 There is therefore no unfairness : Plato s position is being surveyed on all
sides. Moreover the Laws shows Plato to have believed in communism as the
true ideal for the whole state, v. Newman, Politics, Introd., p. 159.
4 1265 b 25. * 1330 a 14 sqq. 6 Met. A. 6. 988 a 1-7.
7 Eth. i, 6. 1097 a 1 1. 8 Rhet. i. 1356 b 29, Eih. 1180 b ao, Met. A. 981 a 15-20.
Aristotle s Criticisms of Plato 87
Finally, under this head may be brought certain other
arguments, of which we can only say that they are dictated
by pure eagerness to score a point. We must allow for
the combined pugnacity and pertinacity of Aristotle s
nature ; he was a very militant philosopher, and all is fair
in the war against the Platonists. Thus, in reference to
the Ideal numbers, he asks whence come the units that
make up the Indeterminate Dyad? 1 They must come from
a Dyad also, and, as Alexander adds, it is a strange
doctrine indeed that would make one come from two
instead of vice versa. So again in the Ethics? Aristotle
plays the Philistine in his well-known gibe about the
weaver and the carpenter*. Similarly, in the Politics?
Aristotle need not have been unaware of Plato s real
opinion as to the happiness of the guardians. It was a
point in which his opinion really differed from that of his
master ; and he simply yields to the natural temptation of
quoting Plato in his own support.
It may readily be admitted that Aristotle does not show
to the best advantage in his criticisms of Plato. He is too
full of his own point of view to be a sympathetic critic, and
sometimes too near his master to be an effective one.
Moreover, the thought of Plato refuses to be fettered
within the categories of any system ; the whole is more
than the sum of its parts, the spirit of Platonism is more
than the totality of its doctrines. But nothing could have
been more wisely ordered by the time spirit of Greek
thought than that Plato s work should be continued and
1 Met. A. 9. 991 b 31 with Alexander ad loc. Similarly he is perfectly well aware
of the real nature of Plato s great and small , but at M. 8. 1083 b 23 he treats
them as though they could be separated.
a i. 6. 1097 a 8.
3 Politics ii. 5, 27. It seems incredible that any one who has read the
beginning of Rep. Bk. iv should have so utterly misunderstood it (Campbell
and Jowett, iii, pp. 162-3). I* ts incredible. Aristotle in his dnopiai need no
more be taken always au pied de la lettre than Plato in the dialogues.
88 Aristotle s Criticisms of Plato
extended by one so different in temperament, yet so like
in universality of mind and enthusiasm for philosophy.
It is not proved that Aristotle is guilty towards Plato of
any fundamental misrepresentation; and Plato cannot be
said to be fully known till he is re-read in the light of
Aristotle.
Q
o:
LJ
o
LJ
o
