Académique Documents
Professionnel Documents
Culture Documents
.
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.
Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend
access to Mind.
http://www.jstor.org
AND DIALECTIC
III.-MATHEMATICS
THE REPUBLIC VI.-VII. (I.).
IN
BY F. M. CORNFORD.
THIS paperhas two objects: (1) to definethe mentalexperiences
as noesisand dianoiac;(2) to disentangle
whichPlato distinguishes
certainthreadshe has woven togetherin his schemeof higher
education. The Republicis a long book, and Plato saves space
themeswhich a modernwriterwould
by treatingconcurrently
keep apart. It is hardlytoo much to say that Plato, at the
height of his powers, never wasted a wo'rd; whereas many
readersof Book VII. musthave been troubledby an impression,
do nothingto remove,that dialecticis
whishthe commentaries
describedover and over again' for no clear reason. I hope to
show that each of these descriptionshas a special purposeand
whenthey are assignedto
meaning,whichcan be distinguished
one of education,the otherof research-with
two programmes,
aims and methods.
different
I.
38
F. M. CORNFORD:
IN THE REPUBLIC
Vi.-VII.
39
40
F. M. CORNFORD:
IN THE REPUBLIC
vI.-vii.
41
science must start from necessarytruths taken as indemonstrable. Thus geometrytakes forgranted(Xa,u,4dayet),
besides
the common axioms, (a) the definitions
of its subject-matter
(magnitude)and of certain'essential attributes' of magnitude,
such as 'straight', 'triangular', and (b) the existence of
magnitudeand of certain other primarythingscorresponding
to the definitions,vtz., points and lines. The .existence of
everythingelse (e.g., the various figuresand their properties)
has to be proved by constructionand demonstration. Definitions are not hypotheses: they do not assert 'the existenceof
anything,but only state a meaningthat must be understood.
-Hypothesesare assumptionsof the existenceof thingsdefined.
In speakingof the hypothesesof mathematics(510 C), Plato
primarilymeans the 'absolute' kind of hypothesis,not the
'relative'. He seems, moreover,like Aristotle, to restrict
'hypotheses' to assumptions'of existenceand not to include
definitions;a definitionin 'mathematicsis' itselfan ' account'
of the meaningof a term,and no ' account of it ' can be
(X6oyo9)
demanded. His instancesare 'odd and even (arithmetic),the
figuresand threekindsof angles (geometry),and otherkindred
thingsin each branchof study' (510 C). Accordingto Aristotle,
arithmeticand geometryassume the meaning(definition)
of odd
and even, etc., straightand triangular,etc., but the existenceof
these thingsshould be demonstrated. The only thingswhose
existencethese sciencesare entitledto take forgrantedas ultimate hypothesesare the unit (arithmetic)and magnitude,points,
and lines (geometry). He thus would agree with Plato in condemningmathematicianswho should treat as ultimate hypoof 'odd and even,the figuresand threekinds
thesesthe existence
of angles'.
Probably Plato has in mind also the originalmeaning of
and thinksof the mathematician'sassumptionsas
w7roted7-Oat,
put to the learnerfor acceptancein the processof instruction.
The whole descriptionwould fit the conversationalmethod of
teaching. When he says, 'starting fromthe hypotheses,they
go throughall the rest and reach a conclusiofi6,ctOXOy/OV,ct6'VCO','
this ambiguousword may mean eitherthat the whole demonstrationhangs togetherconsistently
(thoughit hangs,so to say,
in the air so long as the hypothesisis not proved), or that it
rests on agreementbetween teacher and pupil to accept the
hypothesis. The same ambiguityrecurs at 533 C, 'How can
such 0,obXo7yta
(consistencyor agreement)ever become knowledge ? '
Geometrystill awaited its codificationby Euclid. Solid
42
F. M. CORNFORD:
an opDialecticincludes
J7rava8aOjzo7siXp(i/LEVoV,
primaryand common
meaningof 6p,unis 'impulse' or 'effort' or 'impetus'. It is nearerto'
' spring'thanto spring-board
'.
fEravaf3?va&
Kat
E'T7 Ta
aVo)TapC)
IN THE REPUBLIC
vI.-vII.
43
44
F.
M. CORNFORD:
philosophical
methoddescribedin the passageof the Republic[511B],
which of course does not referto mathematicalanalysis at all;
it may therefore
well be that the idea that Plato discoveredthe
methodof analysisis due to a misapprehension.But analysisand
synthesisfollowingeach otherare relatedin the same way as the
in the dialectician'sintellectual
upwardand downwardprogressions
that Plato's achievement
method. It has been suggested,therefore,
was to observethe importance,fromthe point of view of logical
rigour,of the confirmatory
synthesisfollowinganalysis.'
. .
existing(diagrammata)
are discoveredby beingbroughtintoactuality.
The reason is that the activityis intuition' (or, readingi1 v07jut
employedis an activity').
evepyeta withRoss, 'the intuition
be
IN THE REPUBLIC
VI.-VII.
45
a5TaL O-TOLXELa7t)V
a7ro&L'Ecov
XeyovTaL.
XEyETcIL
OQX(CO . . . To KaTaOLKEVaCov ELTT Tov KaT-acrKEvaCo/iE'vOv LTTOLXELOV,CT TO
TO
vpOTov 7rap' Ev1KXEL?y &ro
VTepoV . . . aXX&s & . . . ElS O l7rXovOEpov
virapxov
Ta
aLT/IaTa
OTOLXELa Tov
OIE(P?7u'rcoav.
2 On Anal. Post. I., 12 (p. 26, 23, Waflies), avaXvELv8& XE'yC VvV To
&
at
V
ras 7rpoTaoTEtvE$EVplLtKElV
TLvos aXr .ov Ov/-rEpado-aTro
TEOEVTOV
0VV?X OiJ
46
F. M. CORNFORD:
XVclV).
EKEt avaXvfavrev
- Kat yaip
so early a
47
willcorrespond,
elusionsoughtwillbe truealso,and thedemonstration
in thereverseorder,to theAnalysis(theSynthesistraversesthe same
seriesof steps,but in deductiveorder); but if (2) we come upon
somethingadmittedthat is false,the conclusionsoughtwill also be
kind, we assume the (construction)
false. (b) In the problematical
propoundedas ifit wereknown; and nextadvance(upwards)through
the sequent steps,taken as true,as far as somethingadmitted(a
construction,possible or impossible,admitted to be a necessary
elementin the desiredconstruction).Then if (1) the admittedthing
sayis possible and obtainable given', as the mathematicians
the construction
propoundedwill be possibletoo, and once morethe
will correspond,
demonstration
in the reverseorder,to the Analysis;
but if (2) we come upon somethingadmittedthat is impossible,the
problem also will be impossible'.'
It is quite possibleto accept the statementthat Plato ' discovered' the methodof Analysis,in the same sense as Aristotle
discoveredthe syllogism; that is to say, he was the firstto
reflectupon the processof thoughtinvolved and to describeit
in contrastwith the process of Syiithesis. And it 'is certain
that in his account of the dialecticalascent Plato is describing
the upward movementof thoughtwhich has been illustrated
fromgeometricalanalysis. Since he assignsthe mental experience called noesis to the corresponding
section of the Line, it
seems justifiableto say that the intuitionused in this upward
1I gatherfromSir T. Heath's discussionof this passage (Thirteen
Booksof Euclid,i., 138) that modernhistoriansofmathematics-'careful
studies' by Haiikel,Duhamel,and Zeuthen,and othersby Ofterdinger
and Cantorarecited-have madenonsenseofmuchofit bymisunderstanding the phrase 'the successionof 'sequentsteps' (T&v r aKoXovOco) as
', as if it were ra' o-v,If3aLvovTa. Some may
meaninglogical 'consequences
have beenmisledby Gerhardt(Pappus,vii.,viii.,Halle, 1871),whorenders
it ' Folgerungen'. They have thenbeen at greatpains to showhow the
can be the consequencesof the conclusion.
premissesof a demonstration
The whole is clear when 'we see-what :Pappus says-that the same
in Analysis,from
sequenceofstepsis followedin bothprocesses-utpwards
the consequenceto premissesimpliedin that consequence,and downwards
in Synthesis,when the,steps are reversed to framethe theorem or
' in thenatural(logical)order'. You cannot
demonstrate
theconstruction
followthe same seriesof stepsfirstone way, thenthe oppositeway, and
in both directions. And Pappus never
arrive at logical consequences
said you could. He added 66s to indicate that the steps ' followin
' but are not, as aKoXovOa alone would suggest, logically
succession
ofAnalysisand
'consequent' in theupwarddirection. In the definitions
in Euclid XIII. (Heath,ibid.,I., 138)thephrase tLa
Synthesisinterpolated
T7v
aKO?OV'OcV
is
thesequentstepsto some
soughtas admitted(and the passage) through
admittedtruth'. Here again it is translatedby Heiberg(Teubneredit.,
', and by Heath 'throughits consequences'.
III., 365) 'per consequentias
fromPappus'
may have been copied,withabbreviation,
These definitions
statement.
48
F.
M.
CORNFORD:
/LETEXOVOIL
Kara
Ta
IN THE REPUBLIC
vI.-vII.
49
Tovs Xoyovs
EveKa 7ravTas
There are tracesof an Academic
of 'operations'in mathematics,
as to thenatureandpossibility
controversy
whoseobjectsare eternal; but thisquestioncannotbe pursuedhere.
'Thus at Timaeus,30C, the Idea of Animal embraceswithinitself
Te
KcLLav&yKaL(G
*Gon
VrOLOv,uEVOL XEyOVOL
EaVTC
(ev
7rEpLXaO3Ov
TETpayomvLlEtv,
KaL 7rpa'eOS
KTX.
Evovi-asv
L8Eas.
TCO e'rTl
CGTOV..
50
F. M. CORNFORD:
To Proclus the dialectician'scontemplationof the genus proposed for divisionseemed analogous to the geometer'scontemplation of his given figure. The finaldefinitionof the species
consets out in explicitform(Xo6yos~)
the elements(differences)
tained in the essencedefined,as the geometer'scompleteddiagram exhibits the elementsin the demonstration.Each, in
reachinghis results,eliminates,stepby step,irrelevantelements:
the dialectician,as he selectseach differentita,
rejectsits alternative; the geometerretainsonly the elementsthat will figurein
his demonstration
and rejectsothersthat occur to him but are
foundnot to lead to his conclusion.1
(d) States of mind.-It remains to note anotheruse of the
wordsnoesis(or nous)and dianoia. Glauconsays he understands
the account of dialectic as meaningthat ' the fieldof the real
and intelligible
as studiedby the science(6qr1rT
,Uq)
ofdialectic
has a clearertruth(o-aOE'TepoV)
than what'is studied by the
mathematical' arts' as they are called,-whose hypothesesare
taken as starting-points
'; and that ' althoughstudentsare compelled in studyingthese arts to'use abstractreasoning(Stavoota)
and not the senses,yet, because they proceedfromhypotheses
withoutgoing up to a firstprinciple,they do not, you think,
come to have nous (vovovL'Oxev),albeit their objects can be
objects of nous in connectionwith a firstprinciple (KcairoT
I understandyou to describethe
VO0rT6lV 6'VTW /iCTa apXqs,).
state of mind (e4ts) of geometersand othermathematiciansas
not nous but ditanoia,regardingdianoia as somethingbetween
is said to
beliefand nous' (511CD). Later, the mathematician
live in a sort of dream,not seeing realitywith wakingvision
but
(533B); his so-called knowledgeis not science (etT71.kv)
only dianoia (533D).
The contrastbetweenthe two states of mind may be illu-.
stratedfromthe Meno. The slave who is therequestioned2 at
firstproducesfalse beliefsabout the solution of the problem,
then-truebeliefs; but these will become knowledgeonly when
he has been taken throughthe proofrepeatedly'and graspedits
on thereason'(aiTia9 XoctUO'j).
logicalnecessity
'by reflection
51
52
ETC.