Space and Flux in Plato

Image, Flux, and Space in Plato's "Timaeus"
Author(s): Richard D. Mohr

Source: Phoenix, Vol. 34, No. 2 (Summer, 1980), pp. 138-152
Published by: Classical Association of Canada
Stable URL: http://www.jstor.org/stable/1087872
Accessed: 05-09-2015 20:56 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1087872?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Classical Association of Canada is collaborating with JSTOR to digitize, preserve and extend access to Phoenix.
http://www.jstor.org
This content downloaded from 132.248.9.8 on Sat, 05 Sep 2015 20:56:39 UTC
All use subject to JSTOR Terms and Conditions
IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

RICHARD
ERACLITUS
WROTE
D. MOHR
THAT YOU CANNOT
step into the same river twice
forthe waters are ever flowingnew upon you (B49a, B12, B91, A6). We
are told by Aristotle(Metaphysics1010all-14) that Cratylus,a follower
of Heraclitus,revisedthisslogan to read "one cannot step into the same
rivereven once," and thenstopped talkingaltogether.Cratylusbelieved
that nothingwhatsoevermay be predicatedof the phenomenain flux.
We are also told by Aristotlethat Plato was a followerof Cratylus
and adopted Cratylus' views into his metaphysics(987a30-bl). In this
paper, I will discuss to what degree this is true. The paper has two sections. The firstgives a new analysis of the tortuousand much debated'
passage of the Timaeus in whichPlato discussesthe possibilityof making
predicationsof the phenomena in flux and in the course of which he
formallyintroduceshis conception of space. The passage is Timaeus
49b-50b. The second sectionis a generaldiscussionof Plato's conception
of space.
The second part of the Timaeus (47e ff.) where Plato discusses the
nature of the physical world begins by assuming that some problem
confrontsall who tryto discuss the nature and generationof earth,air,
fire,and water. The problem is to say what each of these severally is
(0rp
48b6-7). In this way, the second half

abcrwiv,
irorf oa-TVKaL OtKaTrov
TL
of the
Timaeus beginslike many of the Socratic dialogues by askingthe
what-is-itquestion. The question, which is repeated at the start of the
gold analogy (50bl), is an internalquestion which asks us to identify
various phenomenalkinds (in the firstcase, earth,water,etc., and in the
1See: Cherniss, H. F., "A Much Misread Passage of the Timaeus (Timaeus 49c750b5)," A7P 75 (1954) 113-130. Cherry, R. S., "Timaeus 49c-50b," Apeiron 2, No. 1
(1967) 1-11. Cornford,F. M., Plato's Cosmology(London 1937) 178-185. Gulley, N.,
"The Interpretationof Plato, Timaeus 49d-e," A7P 81 (1960) 53-64. Lee, E. N., "On
the Metaphysics of the Image in Plato's Timaeus," Monist 50 (1966) 341-368; "On
Plato's Timaeus 49d4-e7," A7P 88 (1967) 1-28; "On the 'Gold-Example' in Plato's
Timaeus (50aS-b5)," in Essays in AncientGreekPhilosophy,ed. J. P. Anton and G. L.
Kustas (Albany, N.Y. 1971) 219-235. Mills, K. W., "Some Aspects of Plato's Theory of
Forms, Timaeus 49c ff.,"Phronesis 13 (1968) 145-170. Mohr, R. D., "The Gold Analogy
in Plato's Timaeus (50a4-bS)," Phronesis 23 (1978) 243-252. Zeyl, D. J., "Plato and
Talk of a World in Flux: Timaeus 49a6-50b5," HSCP 79 (1975) 125-148. I allude to
these works by theirauthors' names and in Lee's case by year.
138
PHOENIx,
Vol. 34 (1980) 2.
139
gold analogy, various geometricfigures).2The what-is-itquestion is not

here,or in the early dialogues, an externalquestion whichasks afterthe
ontologicalstatus of an object. Plato is here interestedin how earth can
be distinguishedfrom water, rather than whether it is a substance,
quasi-substance,quality, quantity, a form,a matter, an Idea, etc. In
orderto answerthe questionof identifying
earth,water,etc., we are told
that a thirdsortof entitymust be added to the furniture
of the universe.
For the firsthalfof the dialogue, two sortsof entitiesweresufficient.
The
one is the Ideas, which are spoken of as paradigms or models, and as
being intelligibleand always unchanginglyreal; the second is the phenomena,whichare said to be copies of thesemodels,are in the processof
becomingand are visible (48e3-49al). This "becoming" (cf. 52d3) refers
to the pre-cosmicor acosmic fluxof the phenomena described at 30a
and 58b-c. The third thingwhich must now be added to the ontology
is called the receptacleof all becomingand is later identifiedwithspace
(52a8). Why thisthirdthingis requiredwe are not yet told; that requires
of identifyingphenomenal kinds. The
elaborating the earlier difficulty
firststage of the elaboration is contained in the followingsentence:
"It is difficultto say with respect to each of these kinds (earth, water,
etc.) what type (brotov)we really ought to call water ratherthan fire,
and what type (b6rotov)ought really to be spoken of as a particulartype
of thingratherthan as all types taken togetheror severally,so that we
reallyuse languagein a somewhat(rnLt)reliableand stable way" (49b2-5).
The problem,it seems,is not thatvariousphenomenaltypesshade offinto
each other,in such a way that,say, we mightbe confusedas to whethera
patch of rust should be called orange or brown. Rather, it is that all
phenomenaltypesseem to be interchangeableso that all the phenomena
are equally well called by any and all names of typesof phenomena.
Why the phenomenaare in this strait has not been fullystated. But
with the statementthat we are eventuallyto end up withlanguage that
is somewhatstable and certain (49b5) we are already given a hint of the
solution to the problem. For this phrase harks back to the principles
whichopen Timaeus' discourse at 29b-c. There it is said that language
will be as certainas its objective referentsare stable; language about the
immutable Ideas will be absolutely certain, while language about
becomingwill be certainto the extentthat becomingis a likenessof the
Ideas (29c1-2). The hint,then,is that the phenomenamay be identified
2Throughoutthe paper by "identify" I mean to pick out an individual as the kind or
type of individual it is, as in the sentence, "the chemistrystudent correctlyidentified
the samples as zinc." I do not mean "identify" in the sense of pickingout an individual
as the individual it is, as in the sentence, "the victim identifiedthe culpritin the police
line-up." Zeyl uses "identify" in this latter sense (128, 130, et passim); whereZeyl uses
the term "identification,"I use the term "individuation."
140
PHOENIX
insofaras they are likenessesor images of the Ideas. On the otherhand,

the firststage of the elaboration of the problem of identifyingthe
phenomenain saying that a phenomenalobject in and of itselfcan have
any and all predications applied to it, more or less accuses the phenomena of being referentsfor self-contradictory
propositions. Plato
makes the correspondingontologicalclaim in the Phaedo wherehe claims
of the phenomena,just as phenomena,that theyare "never the same as
themselves"(Phaedo 80b4-5).
To assert this logical and ontologicalcontradictoriness,
though,is not
to assert directlythat the phenomena in no way exist, but rather,I
suggest, is to assert only that the phenomena are in flux (cf. Phaedo
80b4-5 with Philebus59bl).3 This is how Aristotleinterpretsthe predicational schemas of the Heracliteans (Metaphysics1010a6-37) and contradictorypropositionsseem to be entailed by Plato's analysis of change
in his discussion of "the suddenly" in the Parmenides.4Note also the
descriptionat Timaeus 57c3-4 of thingsthatare changingas thingswhich
are becomingdifferent
fromthemselves.
That the firststage of the elaborationof the problem of identifying
phenomena suggests that the phenomena are subject to contradictory
predicationsshould leave it no surprise,then, that the finalelaboration
of theproblemis the assertionthat thephenomenaare in flux.The reason
the names of all types fit each phase of the phenomena is that the
phenomena are in a universal flux,in which every type is constantly
transmutinginto all othertypes,as describedat 49b7-c7, wherewateris
traced througha cycle of transformations
whichturnswater into all the
otherthreetypesof primarybodies.
The difficulty
of answeringthe originalwhat-is-itquestion,then,is the
of
or state the kind (7ro-ov)
of phenomenal
difficulty being able to identify
that
in
are
flux
and
transmute
into
each
other.This
objects, given
they
is
at
stated
49c7-d3.
The
sentence
fuses
the
difficulty finallyfully
problem
of identificationwith the assertion that the phenomena are in flux:
"accordinglysince each of these (earth,fire,etc.) neverappears the same
[i.e. is in flux],which of these types [irotov]
could one steadfastlyaffirm
withoutembarrassmentas being thisparticulartype and not another?It
3Some critics wish to claim that for Plato, though the phenomena are in some sense
disparate, they are in no way (self-)contradictory.See forinstance R. E. Allen, "The
Argument fromOpposites in Republic V," Rev. of Metaphysics 15 (1961) reprintedin
Essays in Ancient Greek Philosophy 166-167. Phaedo 80b4-5 and Timaeus 49b2-5,
however,would seem to underminethis claim.
4Parmenides155e-157b, and cf. 156e7-8 for the fullgeneralityof the doctrine. Here
it is claimed that when a thingis changingfromhaving a propertyto having its opposite,
it passes throughwhat is called "the suddenly" in which it neitherhas the propertynor
its opposite (156d). This is logically equivalent to a denial of the law of non-contradiction.
141
is not possible." The rest of the passage which gives the answer to this
problemof identificationruns as follows:
theseis byproceeding
in thefollowing
Rather,thesafestwayby farto speakconcerning
manner.What we see, sinceit alwaysis becomingat different
timesand in different
ways-forexample,fire-weshouldnotsay on each occasionthatthisis fire5but that
ofa certainkind[,r roto^VroY]
is fire,andon eachoccasionsay notthatthisis
something
of a certainsort[r6 rotodrov]
is water,nor shouldwe call
water,but thatsomething
else [say,"air" or "earth"] whichwe
(this),as thoughit had somestability,anything
whenwe pointand use the terms"this" and "that" [as in
supposerevealssomething
"thisis fire"].For it fleesand does not abide the assertionof"this"or "that" or any
assertionthatindictsthemas beingstable.But insteadofcallingthem[earth,fire,etc.]
each of thesethings["this,""that,"etc.],we shouldthuscall these[thefleeting
phases
ofthephenomenal
in
flux]"a certaintype"[rbrotoirov],whichalwaysrecurssimilarly
each and all cases; forinstance,fireshouldbe called"thatwhichis on each occasionofa
certaintype"[,r ... roLtorov],
and so witheverything
thatgoes throughtransformations.But, thatin whicheach of theseappearsand againout of whichtheypass away
[i.e. thereceptacleor space],thisalonebycontrastoughtto be designatedby theuse of
the terms"this" and "that", but whatis of any kindwhatsoever,
hotor whiteor any
ofthecontraries
and all thatconsistsofthese,this[i.e.space]shouldnotbe calledanyof
these["hot,""white,"etc.].6
At 50c4-5thereferents
of"a certainsort"(roto7rov),
whicharesaid to
enterintoand leave thereceptacle,
areexplicitly
equatedwithimagesof
theIdeas. The passage,then,is to be interpreted
as follows.
The passagereconfirms
theviewthatthephenomena
as influxcannot
be identified
as to type.Nevertheless,
the passage maintainsthat the
can be identified
and are subjectto thepredication
phenomena
roLo0rov,
"of a certainsort."' It is as recurringratherthan as being in fluxthat a
type of phenomena may be identified.What recursin the fluxof phe-
5WithCherniss(116) and Mills(154,n. 14) I takethepresenceof7rp (d6) as decisive

againsttakingrov^roand r7 rotoLroyas thesecondaryobjectsof irpocrayopebew
(d6).
as Cornford
IIVpat d6, by position,is obviouslynotresumptive,
(179) and Zeyl (132)
seemto takeit.
'The gold analogyimmediately
ensuesand runsas follows:
But we musttryagainto speakstillmoreclearlyconcerning
this.For,
IA: ifa man,whileforming
out ofgoldeverytypeofshape,neverstoppedremolding
each into all of the restand if someoneindicatedone of the shapesand asked"what
is it?"
IB: then,by farthesafestanswerwithregardto truthwouldbe "it is gold,"
and all theothershapeswhichcometo be in (thegold),
IIAI: but,thetriangle
2: theseshouldneverbe said to be real,
3: since,indeed,theyundergochangeeven whilethisis beingasserted,
IIB: butif,then,he is willingalso to acceptwithsomecertainty
(the answer"it is)
of a certainkind,"we shouldbe glad.
something
For a commentary
to theselinessee myarticlecitedin note1.
arerotoirov("of a certainsort")answersthequestion
7Theclaimthatthephenomena
pronouncorrela7rotor;("of whatsort?")(49dl). Totoirovis simplythedemonstrative
tivewiththeinterrogative
It has beenraisedagainstmethatmytakingrotoIrov
7rotov.
142
PHOENIX
nomena are images (50c4-5). We can tell that what recursis the same
recurringimage by referringit to the originalof which it is an image.
The phenomena,then,have a double aspect. On the one hand, theyare
in flux;on the otherhand, they are images of Ideas. Insofaras the phenomenaare in flux,nothingwhatsoevermay be said of them.But insofar
as theyare imagesof Ideas, theymay be identifiedaccordingto kind.
The introductionof the phenomena as images solves the problemof
the phenomena.This solutiongives the formalreason forthe
identifying
introductionof the receptacle.The thirdthingis needed as a mediumin
which the phenomena,as images, may appear. This need was not felt,
though,until the problemof identifyingthe phenomenahad been clarified(49b-c). That the phenomenaare in fluxrequiredsome means other
than direct appeal to the phenomena themselvesin order to identify
them.The thirdthinghaving the nature and power (49a4-5) of a receptacle and mediumis a requisiteforfulfilling
this need, since it makes it
possible forthe phenomena to appear as images and so makes possible
the identification
of the phenomenainsofaras theyare images of Ideas.
Furtherby drawingintodoubt theontologicalstatus ofthe phenomena
as being in flux(esp. 50b3, IIA2 in note 6) and by contrastingthe phenomena as in fluxwith the phenomenaas images, our passage leads the
way to the assertionthat the shiftingphenomenaas imagesin the receptacle "cling in some way to existenceon pain of being nothingat all"
(52c3-5). Thus is resolved the problem,leftover fromthe Republicand
not resolved in the Sophist,of how Becoming holds a middle ground
between Being and Non-Being: the fact that the phenomenaare reflections of Ideas saves them fromthe non-existenceand completeunintelligibilitywhich threatensthem as a resultof theirbeing in flux.8
Though my interpretationof the passage is simple, it has eluded
Platonic scholarship. Some critics (notably Zeyl) have construed the
problemwithwhichthepassage deals as a problemofhow thephenomena
areindividuatedratherthanidentified
and claimthatin thepassagesensible
phenomenaare notin themselvessubstantial,but are made substantialby
theirrelationto the receptacle,whichindividuatesthem and serves as a
substratefortheirchanges. For thesecritics,rotoirovdoes notdistinguish
an aspect of the phenomenaas images fromtheiraspect as in flux,but
ratherapplies even of the phenomenaas in fluxand denotes a form-like
to denotetypeorkindin ourpassageoverlooks
itsdemonstrative
("pickingout") force.
But denotingtype or kind is a standardPlatonicuse forrotovrov:at Symposium
or as we wouldsay "a branch
210d7, rtcrrdljiYv
ro0arrlvmeansa "kindofknowledge"
ofknowledge."
TFortherelationofTimaeus49b-50bto theCratylus
and Theaetetus
see Mohr249-250.
See also N. H. Reed, "Plato on Flux, Perceptionand Language,"ProcCambPhilSoc
18 (1972) 65-77. Reed followsLee's interpretation
of Timaeus49-50.
IMAGE, FLUX, AND SPACE
IN PLATO'S
TIMAEUS
143
quality of the phenomena(in Aristotle'ssense of form).For these critics

the whole passage looks forwardto Aristotle'sform-matter
distinction
(see Zeyl esp. 146-148). The formalreason forthe introductionof space,
on Zeyl's account, is to serve as a principleof individuationand as a
substrateforchange.9
Three other critics (Cherniss, Lee, and Mills) form a group which
claims the passage has nothingto do with form-likeproperties;each of
these critics, though, takes rotoTrov
as picking out a differentsort of
entity.Chernisstakes roitoroV
correctlyas referringto images of Ideas,
but wishes to separate the images from the phenomena in flux, by
equating imagesratherwithwhat he calls "self-identicalcharacteristics,"
which,he claims, cause the phenomena to come into existenceby their
entrancesinto space (128-130). The chiefobjection to this view is that
there are at least four places in the dialogue, three in the immediate
context,where the phenomenaas images and the phenomenaas in flux
are taken as extensionally equivalent (29c1-3, 48e7-49al, 50d1-2,
52a4-7).'0
Lee, thoughcorrectlyseeing that images and that whichis in fluxare

denotesselfco-extensive,neverthelessclaims,like Cherniss,that rotorrov
identicalcharacteristics,
whichhoverin an ontologicaltwilightsomewhere
9Zeyl's understandingof the problemwithwhich the passage deals and its solution are
the result of (1) his translating,like Cornford(178-182), droZov (49b2 and 3) and
7rotov
(49dl) quantitatively rather than qualitatively and so failing also to see that ravb'r
rovro (49c1-2),
'v (50a7) referto types
(49c7),
(49dl), and
ro'-wv
r7, abr,v
abr&,v
and not to individuals,
raised by the fluxof the phe(2) failingto see that the problem
nomena is a problem of identifyingthem rather than findinga permanentsubstrate for
them, i.e., failing to see that 49a6-b2 refersback to 48b5-8, and (3) failing,like Lee
(1967, 22), to see (a) that the stability of discourse made possible by our passage is a
qualified one and (b) that 49b5 refersback to 29b-c.
Zeyl, however,is right,as against Cornford,in seeing that rotoirov can denote properties in any category (substance, quality, quantity, relation, etc.) and is not restricted
just to the categoryof quality, as Cornfordwould have it (Zeyl 147). Cornfordconfuses
two senses of substance in his analysis of our passage. Cornfordwishes to claim that our
passage in calling phenomena rotovrov treatsphenomena as non-substantial,wheresubstance is understood as a composite of formand matter, but he then unwarrantedly
assumes from this claim the furtherclaim that
rotov^rovcannot denote substances,
where substances are understood as essences or natural kinds as opposed to accidents
(Cornford 178-181, esp. 181).
10Despite this overwhelmingevidence, Cherniss claims that 50c3-4 supports his view
by its implyingthat the transientphenomena are apparent alterations of the receptacle
induced by copies of Forms enteringit (129). But ratherthis sentence, which in context
is offeredas a reason (-yip, c2) whythe receptacleis characterlessand unchanging,simply
warns against inferringany diversityof the receptacle itself from the diversityof the
phenomena in it. It is the diversityof the receptacle that is apparent, and, as we are told
elsewhere,altogetherimpossible (50b8-c2, 50d-51a). The firstb8of 50c4 is adversative;
4alverat (c3-4) here means "appears to be x, but indeed is not x" ratherthan "is visibly
representedby x" or "shines forthvisibly as x."
144
PHOENIX
between the Ideas and the phenomena, but which are not images of
Ideas (1966, 366-368; 1967, 22-25). This interpretation,however,collides with the explicitequation of that whichis denoted by roWtDrov
and
the images of Ideas at 50c4-5.
Mills goes so far as to suggest that the referentsof row70rov
are the
Ideas (154), despite the fact that the Ideas are explicitlysaid not to
enterinto anythingwhatsoever(52a3) and in particularnot to enterinto
the receptacle(52c6-dl). Indeed, this propertyof the Forms is the very
reason given at 52a-b fordifferentiating
Formsfromtheirimages.
The convolutionsof the Platonic metaphysicswroughtby Cherniss,
Lee, and Mills are motivated by an attempt to get around asserting
whichtheytake to be
Tro7oLro ("of a certainkind") of thephenomena,
denied in the sentenceof my translationwhichstarts"what we see ..."
(49d4-6). Here I take the participialconstructioncausally ratherthan
purelydescriptively,
renderingit "since it always is becoming. . ." and
it
to
mean
taking
"insofaras it always is becoming." This construction
and interpretation
is justifiedby the factthat the phraseis a summaryof
the precedingdescriptionof the phenomenalflux(49b7-c7), whichraised
the difficulty
of identifyingthe phenomenaas being of a certainkind. I
then
that my simpleinterpretation
of the passage holds.
suggest,
The problems(insurmountablein my view) which beset the interpretationsof Zeyl, Cherniss,Lee, and Mills are dissolvedifwe read 49b-50b
along the lines I have suggested.The passage distinguishesbut does not
separate the phenomena viewed as being in universal flux from the
phenomenaviewed as images of Ideas. As in flux,the phenomenacannot
be identifiedaccordingto kind. The mutabilityof the phenomenadraws
into doubt theirintelligibility
and even theirveryexistence.On the other
hand, as images of Ideas, the phenomenaare subject to the predication
Troovrov,are saved from utter non-existence,and can be identified
accordingto kind. As images of stable reality,the phenomenaare subjects for discourse which is stable to the degree that its objects are
likenessesof the Ideas.
I now wish to appraise the consequencesour newlyinterpretedpassage

has forPlato's overall conceptionof the thirdthingor space. I will do
this in two parts, firstby showingwhat lightour passage throwson the
various metaphorswhichPlato uses to describethe thirdthingor space
and secondlyby discussinghow the passage helps us determinewhether
or not Plato's conception of space operates philosophicallylike Aristotle's conceptionof matter.
Plato uses two groups of metaphorsto describe the thirdthing.One
145
group of metaphorstreats the thirdthingas a containerand suggestsa

"bucket" theoryof space. The othergroupof metaphorstreatsthe third
thingas a mirroror more accuratelyas a mediumor fieldforreceiving
images.
The third thing is a containerinsofaras it is called a receptacle or
reservoir (49a6) and space (52a8) and is compared to a winnowing
basket (52e6) and to a place forsitting(f8pav,52bl). Philosophicallysuch
metaphorsof the thirdthingas a containerserve to distinguishPlato's
conceptionof space fromconceptionswhich take space as equivalent to
materialextension.Such conceptionsin general(1) makespace that out of
which phenomenal objects are made and (2) make space ontologically
dependenton theexistenceofphenomenalobjects. By usingthecontainer
metaphorPlato wishes to deny both (1) and (2). On the metaphorof a
container,the thirdthingwith its contentsis ratherlike an urn fullof
potsherds.It makes sense on this metaphor to say that one can place
objects in the receptacle and again take them out keepingeach intact,
unaltered,and numericallyone. That this is so, however,I suggest,is a
non-relevantconsequenceof the containermetaphorand is not intended
to have metaphysicalimplications.
For, on the other hand, Plato speaks of the thirdthingas thoughit
were a mirror(49e7-8 with 50c5). He does not specificallyuse the term
"mirror"of thethirdthing,no doubt because thatwould entailthenotion
of a perceiverin accordance with his earlier analysis of mirrors(46a-c,
cf.Sophist266c) and he does not wish to implythat the relationbetween
the Ideas and the receptacle implies the presence of a perceiver (see
61c-d). Plato may be a phenomenalistwithregardto perception,but he
is not a subjective idealist.
Plato does, however,explicitlyuse metaphorswhichportraythe third
thingas a medium,as opposed to a container.Such are the comparisons
of the thirdthingto a scent base (50e) and the descriptionof the third
thingas an
(a softsubstance capable of receivingimpressions,
EKa-yeov
And viewed functionallyratherthan descriptively,the
50c2, cf. 50e8-9).
thirdthingas comparedto a mother(50d3) shouldbe rankedas a medium
(see Cornford,187), as should perhaps also the metaphorof the third
thingas a nurse (49a6). Now philosophicallythe metaphorof space as a
mirroror medium serves to show the ontological dependence of the
phenomenaon both the Ideas and the thirdthingand suggeststhat only
the thirdthingand not its contentsis stable and so subject to beingcalled
"this." The metaphoralso servesto distinguishthethirdthingfromthose
conceptions,includingAristotle's,which view space as that which surrounds on the outside the extension of phenomenal objects. These
philosophicalfunctionsdo not hold forthe metaphorof the thirdthing
as a container,whereobjects in it are surroundedby it, are stable, and
146
PHOENIX
are ontologically
of it, like marblesin a goldfish
bowl.So
independent
of the thirdthing
Plato whilesettingforththe philosophical
functions
uses metaphorswhich,if read literally,would,whentaken together,
to thecontentsof thereceptacle.Which
assignincompatible
properties
set ofproperties
shouldwe choose?
It is clearfromourpassageTimaeus49b-50bthatit is themetaphors
ofspaceas a mediumforreceiving
imageswhicharemeantto be takenat
facevalue and thatthereadingoftheothermetaphors
mustbe brought
intoaccordwiththeconception
ofspace as a medium.Andindeedthis
can be done.
For Plato is notcommitted
to thecontents'of the thirdthingbeing
like the contentsof a container,since the philosophicalpointsPlato
wishesto makeabout the thirdthingby usingthecontainermetaphor
holdalso of thethirdthingviewedas a mirror,
eventhoughthemirror
doesnotmakethepointsas perspicuously
as doesthecontainer
metaphor
A mirror
is notequivalentto theextension
ofan object,noris
metaphor.
it ontologically
on itscontents,
noris it a materialconstituent
dependent
ofits contents.
We neednotthenaccusePlato ofanyphilosophical
confusionabout thestatusof the thirdthingand theproperties
whichthe
phenomenahave as phenomenain it. WhereasPlato is committedto
seeingthe contentsof the thirdthingas beinglike the contentsof a
mirror(each mentionof imagespresentin somethingmakes such a
to viewingtheconcommitment,
esp. 52c3-5),Plato is not committed
tentsofthethirdthingas thecontentsofa container.
Nor doeshe make
appeals to thecontentsof the thirdthingviewedas contentsof a container.
When,then,Plato says thatthe contentsof thereceptacleappearin
and thenvanishout of thereceptacle(49e7-8),he does notmean (and
cannotconsistently
maintain)thatthecontentsofthe receptaclecan be
placed in and taken fromit, each remainingintact,unaltered,and
one,likemarblesin a goldfish
numerically
bowl,thoughCherniss'sselfidenticalcharacteristics
wouldcommitPlato to thiscontradictory
view.
Ratherthecontentsappearin and vanishfromthereceptacleas images
in a medium."To vanish"here,then,meansnot to go elsewhere,
but
means
that
the
into
simply
image'soriginalstopsbeingprojected
space.
We can clarifyPlato's conception
ofspaceevenfurther
ifwe compare
it to Aristotle's
of matter.Philosophers
and criticsof many
conception
have
from
Aristotle
to
the
to identify
tended
thetwo,
persuasions
present
Plato's
of
as
a
failed
formulation
of
Aristotle's
taking
conception space
ofmatter.Thisviewis maintained
I suggestlargelybytaking
conception
various phrases which Plato uses to describethe receptacle and wedding

these to Aristoteliandescriptionsof matter without seeing what philosophical roles the two entities play in each philosopher's system. By
147
viewing Plato's space as a permanent,characterlesssubstrate which is

knownonly by some non-standardmode of cognition(cf. 51a-b), these
critics make Plato's space look much like Aristotle'smatter, which is
permanent,lacks formal properties,and is known only by analogy.
However, ifwe distinguishthe philosophicalfunctionsof the two sortsof
entitiesin each philosopher'ssystem,we will see that theyhave littlein
common.There are fivebasic functionswhich matterserves (or can be
construedas serving)in Aristotle'smetaphysics:it serves (1) as a material
cause or that out of whichphenomenalobjects are (made), (2) as a principle of individuation,(3) as a subject ofpredication,(4) as a substratefor
change, and (5) by inference,as a principleof existence.I suggestthat
only the last functionis held in commonbetweenAristotle'smatterand
Pla.to's space.
(1) It is largely by pressing the comparisonof space to gold in the
Timaeus that criticshave been able to construespace as a materialconstituentof phenomenalobjects or that out of whichphenomenalobjects
are made" and so claim that Plato, like Aristotle,is viewingphenomenal
objects verymuchlike substancescomposedof formor shape and matter
(see, e.g., Zeyl 142 top, 147). I suggest thoughthat Plato's description
of the gold as that out of whichshapes are formed(50a6) is an exigency
of the metaphorand is not the relevantaspect of the gold that is being
compared to space. Gold is chosen in the Timaeus, I suggest, not to
typifymatter enteringinto a substance as the constituentwhichoffers
stabilityto form,as the bronzeof a sculptureoffersstabilityto the shape
of the sculpture,but exactlybecause of its malleability.Though the gold
in itselfis permanent,it does not transmitor contributeits permanence
to the phenomena.Justlook at the sentencein whichthe gold metaphor
arises (IA, 50a5-bl): the figuresin gold are being ceaselessly remolded
into each otherand it is thisimpermanenceof the phenomenathat raises
the difficultyof identifyingthem. This difficultyis resolved by our
abilityto identifyphenomenaas images. Appeals to the gold do not help
us to identifythe phenomenaor to view the phenomenaas stable. Indeed
the gold itself,as that which is ceaselessly molded, is not even being
treated here as a medium forimages, whetheror not these images are
treatedlike Aristotelianforms.
When space is viewed as a medium forimages (50c), the images are,
to borrow a phrase fromLee, non-substantialimages, images, that is,
whichdepend fortheircontinuedexistenceupon the persistenceof their
originals(Lee, 1966, 353-360). They are like images in mirrors,or like
shadows, and are unlike statues of living individuals,where the image
1"See forexample, G. S. Claghorn, Aristotle'sCriticismof Plato's Timaeus (The Hague
1954), chapter 2, and more recentlyW. K. C. Guthrie, A Historyof GreekPhilosophy5
(Cambridge 1978) 253, and 262-269, esp. 265.
148
PHOENIX
may persist, as the result of its material component,even after its

originalperishes.The stabilityof the phenomena,as we already know
from29b-c, is dependent on the images' relation to the Ideas, rather
than on the relationof the images to space. Neitherthe gold analogy nor
any other passage in the Timaeus entails viewing space as a material
constituentout of which substances are created. Rather space is a
mediumor fieldin whichphenomenaappear as (non-substantial)images.'2
If Plato had lived into our centuryhe mightverywell have chosen,not
gold, but a movie screen or televisionscreen as his analogue to a field
across whichceaselesslychangingnon-substantialimages may flicker.
(2) For Aristotleone of the main philosophicalfunctionsof matteris
to serve as a principleof individuation.There are many places where
matter is viewed as a principleof individuation(De Caelo 278a26-27;
Physics 190b22-25; Metaphysics1074a33-35) and in MetaphysicsDelta
he goes so faras to say that it is theprincipleof individuation(1016b3233). Socrates and Coriscus are, as men, formallyidentical,but because
they are made out of differentmatter, they are numericallydistinct.
This does not mean that matter forAristotleconsists of pre-packaged
parcels of numericallydistinct stuff,like Democritean atoms or bare
particulars,whichindividuatequalities because the parcels are in and of
themselvesindividuals.'3 Rather Aristotelianmatter individuates,but
is not in and of itselfindividuated; ratherit is like dough to a cookie
cutter: the dough individuatesthe various cookie cut-outs,but does not
consistof a group of individuals independentlyof having been cut into
cookies.Now it is clear thatPlato's space does not consistofpre-packaged
units.He speaks of regionsof space but theseregionsare posteriorto the
images which fillspace (51b4-6): there is a fireregion because fireis
extended throughpart of space. Space does not consist of a grid of
cubbyholes,which images of firemay then fill.But I also suggest that
space for Plato does not serve as a principleof individuationin Aristotle's sense. Plato certainlynever says that space individuates phenomena of the same kind, thoughspace, as a whole,is numericallyone.
'2Guthrie claims that Plato uses the expressions iv W and ri oi indifferently
"in
connectionwithraw material" and cites Philebus 59e and Politicus 288d as evidence (265,
n. 3). In both of these passages, though, Plato is talking of already highly organized
materials which are then, as instruments,put to a finaluse. The referentsof both passages are what in the Timaeus would be called accompanying causes and in any case
referto the receptacle's contents rather than to the receptacle itself. It should be remembered that both the receptacle and its contents, as described in our passage, exist
independentlyof demiurgiccrafting.When the demiurge does start making things,he
makes them out of the receptacle's contents and not out of the receptacle itself.
13Physics1.7 seems to offera single exception to this rule: "Now the subject is one
numerically,thoughit is two in form.For it is the man, the gold-the 'matter' generally
-that is counted, forit is more of the nature of a this" (190b23-26).
IN PLATO'S
TIMAEUS
149
And indeed since space is not that out of whichthe phenomenaare made,
it cannot serve as a principle of individuation at least as Aristotle's
matter individuates. Consider again the relation of original, (nonsubstantial) image, and medium. In this relation it is not the medium
whichindividuatesthe image. There are several ways forthe image to be
multiplied: the medium may be divided, like a shattered mirror;the
originalmay be reproduced(as shadows of two hands cast on a screen),
or the projectinglight may be reproduced(one hand may cast several
shadows, if thereare diverse light sources). But as long as the medium
remains intact and numericallyone, it does not provide the principle
forindividuatingphenomenaof the same kind.
I suggestthat Plato does not have a principleof individuationforthe
phenomena,ratherhe takes the pluralityof instancesof the same Form
as a given (as at 52d5-e2). I suggest that this is also what he intends
when,in discussingthe sortof existencewhichthe physicalworldhas, he
calls phenomenalexistence"partible" as opposed to theimpartibilityof
the Ideas (35a). For this partible existence is later called "dispersed,
scattered, or strewn existence" (37a). Images, then, are strewnacross
space like individualseeds onto a field(aKebaa7iT,37a5). How the projection of images effectsthis we are not told, thoughwe know it cannot be
achieved by reproducingmore originals of the same sort (31a, cf.
Republic597c). The pluralityof thephenomena,then,simplyseems to be
an unanalyzed and possibly unanalyzable given in the Platonic metaphysics.
Cherniss,while wisely avoiding assertingthat space is a principleof
individuationforthe phenomena,has suggestedthat the demiurgeor god
of the Timaeus serves as a principleof individuationforthe phenomena.
This is an elegant variationon the recurrentview in the historyof ideas
that particulars can be individuated by their accidents and relations.
Chernisswrites: "Since the spatial mirroris homogeneousand the Ideas
themselvesare non-spatial,the reflectionsin space would not be locally
distinct; and the Demiurge is conceived as delimitingthem by geometricalconfigurations,
thus representingspatiallythe 'logical' distinctness of theirnon-spatialoriginals."'4This interpretation,if true, would
have the unfortunateconsequence of strappingPlato with a fallacy of
division,but in any case, it runs up against the texts cited above which
suggesta pluralityof the phenomenain and of themselvesprior to any
demiurgicinterventions,especially52d-e, whereit is claimed of the precosmic receptaclethat it receivesshapes of earth and air, is qualifiedby
all the other affectionsthat go with these, has every sort of diverse
appearance, and is filledwith powers that are neitheralike nor evenly
14"The Sources of Evil Accordingto Plato," ProcAmerPhilosSoc98 (1954) reprintedin
G. Vlastos, ed., Plato 2 (Garden City, N.Y. 1971) n. 18.
150
PHOENIX
balanced in any regionof it. It is not possible, then,that thereis in the

pre-cosmosonlyone instanceof each kindofimage spread over the whole
of space, so that the demiurgecan proceed to chop each up into many
instances.
Rather I thinkit is saferto say that Plato does not have a principleof
individuationforthephenomenaand that probablyhe does not need one,
since propositionsall ultimatelyreferto Ideas, which are individuated
by theirunique conceptual content(31a).
(3) One reason I thinkcriticshave tendedto treatPlato's space as a
principleof individuationis because theywish it to be that by whichwe
can pick out numericallyunique subjects forsentences.They want the
relationof visible propertyto space to correspondto and make possible
the logical relation of subject-copula-(adjectival) predicate, and so to
reproduceAristotle'stheoryof predication.Zeyl is quite explicit about
this (146-148). He takes rotovrovas describing"its referent
as an attribute
of somethingelse ... In otherwords,these terms["fire,""air," etc.] are
to be construedas logically (though not grammatically)adjectival.And
this is preciselywhat the rtarbsKaL 6OatosX6yosis: the constructionof
our nominalreferencesto phenomena as adjectival descriptionsof some
basic, permanent subject worthy of that status. This subject is the
receptacle,foronly it can be designatedas ro7ro... Thus it appears that
Plato's use of rovro,r76e,and r6 rotovrov
is the direct ancestor of Aristotle's admittedlytechnicaluse of such locutions" (146-147, Zeyl's emphasis). One problem here is the presupposition(1) that Plato in this
passage is talkingabout the way space determinesthe (logical) formof
propositionsand (2) that the formof "a stable and certainstatement"is
that of subject-copula-adjective.The latter presuppositionis proven
falsesimplyby lookingat the one and only sentencein the dialogue for
whichPlato claims absolute stabilityand certainty(52c6-dl). The sentence reads: "So long as (the) two thingsare different,
neithercan ever
come to be in the other in such a way that the two should become at
once one and the same thingand two." The sentenceis ratheropaque,
but minimallywe have a propositionstatinga relationbetweenat least
fiveterms(sameness,difference,
oneness,twoness,and coming-to-be-in).
The sentencecannot be reduced to a subject-copula-adjectiveform.
In any case, that space is called a "this" is not a consequence of the
term "this"'s picking out a numericallyunique substrate for phenomenalproperties,since the criteriongiven forthe designation"a this"
is stability,not numericaluniqueness (49d7-e4). The stability of the
phenomenathat allows them to be describedby discourse that is somewhatstable (49b5) is a consequencenot of theirrelationto the numerical
uniqueness of space or to an alleged uniqueness whichit is supposed to
provide them severally, but is explicitlythe result of the relation of
IN PLATO'S
TIMAEUS
151
resemblancewhichtheyhold to the Ideas (29c). If the Timaeus suggests

a logical formforpropositionsabout phenomena-Lee has suggestedthe
form"Redness, or whatever-ness,is here"-space does not enter into
thislogical form(ifit does at all) as thesourceofsentences'stability;and
so Plato's space does not,as does Aristotle'smatter,serveas theobjective
correlateto the subjects of predication.
(4) As a resultof not beinga subject forpredication,space forPlato is
not, at least in Aristotle'ssense, a substrateforchange. If by change we
mean a subject's havingone predicateapply to it at time T, and another
at timeT2, thensince space is not a subject forpredication,it is not that
whichabides throughthe change of predicates.It is true that the receptacle, as the gold analogy shows,does abide or remainthe same whilethe
phenomenachange,but it does not, as the gold analogy (properlyinterpreted) also shows, enter into the process of change as a constituentof
that which changes. We cannot point to the receptacle as a constituent
of a phenomenalobject and use it as the referentfordeterminingthat a
phenomenalobject is the same object even though(some of) its properties change. And so too Plato's space, unlike Aristotle'smatter,is not
potentiallythe contraryof any particularphenomenalobject. Space is
not, then, at least in Aristotle's sense, a substrate for change. Space
providesa fieldacross whichimages may flicker,but it does not provide
any continuityto the images as theychange.
(5) Aristotle does not explicitlystate that matter is a principleof
existence for the phenomena, but this seems to be stronglysuggested
by Metaphysics Zeta (1032a20-22, 1039b20-31) and by De Caelo
278b1-2, and it is implied by the doctrine that being and unity are
synonymousin each of the categories (Metaphysics1003b22-37), such
that,since to be one in the substancecategoryis to be one in numberand
so to be individuatedby matter,matteris also the principleof being for
the substance category, namely, it is that by which substances exist
simpliciter.One mightsay that "to be" is "to be enmattered"forAristotle. Plato's view mightbe analogous. Space is a principleof existence
fornon-substantialimages. It is claimed at 52c that since a phenomenon
is a "semblance of somethingelse, it is properthat it should come to be
in somethingelse (namely space) clingingin some sort to existenceon
pain of beingnot at all." It is only as a principleof existencethat Plato's
space and Aristotle'smatterseem to have a similarfunction.
Even the descriptivepropertiesof space and matter are assigned for
different
reasons by each philosopher.That space forPlato is eternal is
the consequence of its being a fieldratherthan of its being that which
abides changes of predicates.'5And Plato's space is characterlessso that
1'50b6-cl; 52a8-bl, noteb', bl; and y)iveoLv,
bl means"comingintoexistence"not
change."
"goingthrough
152
PHOENIX
it offersno interferenceto the receptionof images (50b-c, d-e, 51a7),

while Aristotle's matter is characterless because it is that of which
everythingelse is predicatedwhile it is itselfnot predicatedof anything
else (Metaphysics1028b36-37, cf. 1029a22-23).
As a medium or field which is necessitated by the interpretationof
phenomena as images, Plato's space serves neitheras that out of which
phenomenal objects are made, nor as a principleof individuation,nor
as a subject for predication,nor as a substrate for change. We may
safely conclude with Cherniss then that "in Plato's theory,thereis no
material principleeven forsensibleobjects."'"
THE UNIVERSITYOF ILLINOIS,URBANA
"6Riddleof theEarly Academy(Berkeley 1945) 23.

Space and Flux in Plato

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Space and Flux in Plato

Caricato da

Copyright:

Formati disponibili

Image, Flux, and Space in Plato's "Timaeus"

Author(s): Richard D. Mohr

IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

THAT YOU CANNOT

step into the same river twice

48b6-7). In this way, the second half

IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

gold analogy, various geometricfigures).2The what-is-itquestion is not

insofaras they are likenessesor images of the Ideas. On the otherhand,

IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

5WithCherniss(116) and Mills(154,n. 14) I takethepresenceof7rp (d6) as decisive

IMAGE, FLUX, AND SPACE

quality of the phenomena(in Aristotle'ssense of form).For these critics

Lee, thoughcorrectlyseeing that images and that whichis in fluxare

I now wish to appraise the consequencesour newlyinterpretedpassage

IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

group of metaphorstreats the thirdthingas a containerand suggestsa

various phrases which Plato uses to describethe receptacle and wedding

IMAGE, FLUX, AND SPACE IN PLATO'S TIMAEUS

viewing Plato's space as a permanent,characterlesssubstrate which is

may persist, as the result of its material component,even after its

IMAGE, FLUX, AND SPACE

balanced in any regionof it. It is not possible, then,that thereis in the

IMAGE, FLUX, AND SPACE

resemblancewhichtheyhold to the Ideas (29c). If the Timaeus suggests

it offersno interferenceto the receptionof images (50b-c, d-e, 51a7),

Potrebbero piacerti anche