WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

The Peripatetic Interpretation 
of Plato's Tripartite Psychology 

P. A. Vander Waerdt 

T HE PERIPATETIC author of the Magna Maralia begins his ac­
count of moral virtue with a doxography in which he criticizes 
Pythagoras for referring virtue to number, and Socrates for 

identifying the virtues with E1nUTfjl-Wt and ignoring the alogical part 
of the soul, thus doing away with 7Ta8o~ and ~8o~ (l182alO-30).1 He 
then attributes to Plato bipartition of the soul, of which he approves 
(l182a26), and proceeds to criticize him for confusing investigation of 
moral virtue with that of the good. The passage begins as follows: 
IJ-ETa TavTa 8£ TIA£hwJI 8tEtAETO T-ryJl tjJvx-ryJl Er~ TE TO AO'YOJl EXOJl Kat 
EtC:; TO aAO'YOJl op8w~, Kat a7TE8wKEJI EKauTCy [Tac:;] apETac:; Tac:; 7TPOO"Tl­
KOV<Tac:; (l182a24f). This attribution of Aristotelian bipartition to Plato 
is especially puzzling because the terms aAo'YoJl and AO'YOJl EXOJl are 
used nowhere in the dialogues to refer to parts of the soul: they first 
appear in the fragments of Aristotle's Protrepticus (frr.B23f, B59-70).2 

1 On this doxography and its parallels see R. Walzer, Magna Moralia und aristoteli­
schen Ethik (= NPhilUnt 7 [Berlin 1929]) 75-83; F. Dirlmeier, Aristoteles, Magna Mor­
alia (Berlin 1958) 158-67; O. Gigon, "Die Sokratesdoxographie bei Aristoteles," 
MusHelv 16 (959) 174-212. 

The fragments of Aristotle's Protrepticus are cited from I. DUring's edition (Goteborg 
1961); those of Xenocrates from M. Isnardi Parente, Senocrate-Ermodoro, Frammenti 
(Naples 1982); of Theophrastus from W. W. Fortenbaugh, Quellen zur Ethik Theophrasts 
(Amsterdam 1984); of Posidonius from the collection of L. Edelstein and I. G. Kidd 
(Cambridge 1972); and of Plutarch from F. H. Sandbach's Loeb Moralia XV (London 
1969). Peripatetic writers are cited, where possible, from the relevant volumes of F. 
Wehrli's Die Schute des Arisroteles (Basel 1944 and later), and the Divisiones Arisroteleae 
from H. Mutschmann's edition (Leipzig 1906), which must be studied in conjunction 
with the important additional textual evidence adduced by P. Moraux, AntCl 46 (977) 
100-27. Arius Didymus' epitome of Stoic and Peripatetic ethics, preserved in Stob. Eel. 
2.7, is cited according to the numeration of C. Wachsmuth's edition (Berlin 1884); 
Galen's De Placitis Hippocratis et Platonis according to P. De Lacy, CMG V 4.1.2 (Ber­
lin 1978); and Ps.-Andronicus' nt:pt 1Ta(JWIJ according to A. Glibert-Thirry's edition 
(Leiden 1977). For the sake of convenience the .:1L8aa"KaALKO~ TWl! IIAaTwlJo~ 80y­
~TWIJ is cited according to the numeration of C. Hermann's Platonis Dialogi VI (Leip­
zig 1892), although I have also consulted the improved text of P. Louis, Albinos, Epito­
me (Paris 1945), on which see H. Cherniss, AJP 70 (949) 76-80; a new text by J. 
Whittaker will appear in the Collection Bude. 

2 It is evident from the Protrepticus that the division into aAoYOIJ and AO,),OIJ EXOIJ was 
well established in the Academy during Plato's lifetime, since the terminus post quem of 

283 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

284 PERIPATETIC PSYCHOLOGY 

One might suppose that these bipartite terms simply correspond to 
the author's criticism of Socrates for doing away with TO aAO,),OJl 

IJ-EPO<; rij<; tfJVX7}<; (I 182a2 1); but in this very passage the author also 
uses the term AO,),(.(J"TLKOJI in its tripartite sense to refer generally to 
the rational part of the soul (a19), and elsewhere freely employs 
tripartite terminology. Clearly, the author's ability to represent tripar­
tition is not properly at issue here (c! 1185a2I). 

The author's attribution of bipartition of the soul to Plato has often 
been invoked to support various reconstructions of the genesis 
and development of this doctrine in the early Academy,3 particularly 
by those who postulate a 'development' in Plato's thinking, from the 
tripartite psychology of the Republic to the alleged bipartition of the 
Laws, and who therefore welcome the early and seemingly unambig­
uous testimony of 1182a24f.4 The importance accorded this evidence 

this work is 353 (taking it as a reply to Isocrates' Antidosis: cf B. Einarson, TAPA 67 
[1936] 272-77, and DUring [supra n.1119-24, 33-35). A number of bipartite divisions, 
drawing upon Plato's tripartition and differing from one another in significant ways, 
date from this period: the anonymous version of bipartition preserved in the Topics 
(298f infra); that of Xenocrates (n.5 infra); that advanced in such E~WTEPLKOL A.O')'OL as 
the Protrepticus and De justitia; and later, that of Aristotle's writings on 1ToA.LTLKJ}, which 
explicitly draw upon these E~WTEPLKOi A.O')'OL (Eth.Nic. 1l02a26-1103a3, 1 138b8-13, 
1 138b35-1139a17, 1143b14-17, 1144al-13, 1166al-b29; Eth.Eud. 1219b26-1220a14, 
1221b27-34, 1246a26-b36; Pol. 1254a38-b10, 1260a5-17, 1287alO-33, 1333aI7-30, 
1334b7-28). (The bipartition attributed to TillE'> at De An. 411b5-7 might represent yet 
another version.) 

In his commentary on Iamblichus' Protrepticus 34.5-36.26 (Pistelli), from which DU­
ring prints frr.B23-30, D. J. Allan, ArchGP 57 (975) 246-68, denies the attribution of 
this passage to Aristotle and assigns it instead to Posidonius' Protrepticus. His arguments 
are not persuasive, however, for he fails to identify a single distinctively Posidonian doc­
trine or phrase Gn any event, we know almost nothing about the Protrepticus [frr.l-3]) 
and does not adequately explain the compositional disjunction between 34.5-22 and what 
follows; moreover, there is no reason to apply this thesis to 34.5-22 (= frs.B23-24), 
which even Allan admits must be based upon an early work of Aristotle (cf frr.B59-
70, B 11-21), and which has close parallels in the treatises he overlooks (cf e.g. Pol. 
1333aI6-25, 1334bI4-29). Thus while the details of the moral psychology of the Protrep­
ticus are unrecoverable, it was clearly one in a series of Academic writings that advanced 
bipartition, as is evident from Aristotle's own testimony at Eth.Nic. 1l02a28, where he 
introduces his division into aA.o')'OlI and A.O')'OII exoll by referring to the etwTEpLKOL A.0YOL, 
on which see P. Moraux, Le Dialogue "Sur la Justice" (Paris 1957) 15-22,41-47; I. DU­
ring, Aristotle in the Ancient Biographical Tradition (Goteborg 1957) 426-43; and F. Dirl­
meier, Naturphilosophie bei Aristoteles und Theophrast (Heidelberg 1969) 51-58. 

a On the Academic debate concerning soul-division, see P. Moraux, "From the 
Protrepticus to the Dialogue On Justice," in Aristotle and Plato in the Mid-Fourth Century, 
edd. I. DUring and G. Owen (Goteborg 1960) 113-32; D. A. Rees, "Bipartition of the 
Soul in the Early Academy," JHS 77 (1957) 112-18, and "Theories of the Soul in the 
Early Aristotle," in DUring and Owen 191-200; and, on De An. 432a24-b7, P. A. 
Vander Waerdt, "Aristotle's Criticism of Soul-Division" AJP (forthcoming). 

4 Commentators who have used this passage uncritically include F. A. Trendelen­
burg, Aristotelis De Anima (Berlin 1877) 441; R. Heinze, Xenocrates (Leipzig 1892) 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 285 

is hardly surprising in view of the difficulties surrounding the Aca­
demic debate on soul-division. Not only is the evidence fragmentary,5 
but interpretation of the relevant Platonic texts is complex: even in 
the canonical exposition of tripartition in Republic 4 there are sug­
gestions, which are rejected, of a version of bipartition (see 299f 
infra)~ and in the passage of the Laws (9.8638) that would determine 
whether the soul-division of that dialogue is to be construed as bipar­
tite or tripartite, it is left undetermined whether OvJ.LOC; is a 1TCX(JOC; or 
a JUpoc; of the sou1.6 Yet it must be insisted that the use of 1182a24f 
as historical evidence is uncritical and involves serious misunder­
standing. The most detailed discussion, Dirlmeier's note ad loc.,7 is 
content to assimilate the division into aAo'Yov and AO'YOV EXOV to 
Platonic tripartition, and thus fails to account for either the historical 
background or the philosophical motivation for 1182a24f. Similar lack 
of attention to the doctrinal differences between the bipartition attrib­
uted to Plato at 1182a24f and Platonic passages that suggest biparti­
tion of various kinds has led other commentators to assimilate the 
division aAo'Yov/ AO'YOV EXOV to the (JV"fITOV/(JEtOV dichotomy of the 
Timaeus (42E-44D, 69s-72D; cf Polito 309c), even though both 

142; G. Rodier, Aristote: Traite de fAme II (Paris 1900) 529f; R. D. Hicks, Aristotle: De 
Anima (Cambridge 1907) 550; E. Zeller, Die Philosophie der Griechen 5 ILl (Leipzig 
1922) 843f n.3; F. M. Cornford, "The Division of the Soul," Hiblour 28 (1929-30) 
206-19; F. Solmsen, "Antecedents of Aristotle's Psychology and Scale of Beings," AlP 
76 (955) 150; D. A. Rees (supra n.3) 113f, 118; Dirlmeier (supra n.O 163-65; T. M. 
Robinson, Plato's Psychology (Toronto 1970) 121 n.4; W. Hardie, Aristotle's Ethical 
Theory2 (Oxford 1980) 218; Isnardi Parente (supra n.D 398. 

5 The evidence does not sustain Heinze's claim (supra n.4) 140-43 that Xenocrates 
transformed Platonic tripartition into the division into a.\o')'ov and .\o1'Ov EXOV, and that 
he is the target of Aristotle's criticism at 432a24-b6. Theodoretus attributes a division 
into alalh)nKov and .\O,),LKOV to Xenocrates (fr.206), but since Aristotle says of the 
aia9TjTLKOV that 0 owe W<; aAo1'Ov owe W<; .\o')'ov EXOV geiTj av TL., p~8iw<; (432a31), his 
criticism here can hardly be directed against the division recorded in fr.206, quite apart 
from the possibility of doxographical confusion (c1 Dox.Graec. 615.8-9). Damascius 
states that Xenocrates and Speusippus considered the soul immortal J-LEXPL Tij., a.\o')'ia<; 
(fr.210, but this statement does not even clearly presuppose a division of the soul «(1 
L. Tanin, Speusippus oj Athens [Leiden 1981] 371-74). One can hardly assume, as, for 
example, does J. Burnet, The Ethics oj Arisrotle (London 1900) 63-65, that Aristotle 
simply took over bipartition from Xenocrates, whatever role the latter may have played 
in its development. 

S The case for tripartition of the soul in the Laws and the inclusion of (}v,.w<; as a 
separate soul-part is argued by T. J. Saunders, Eranos 60 (962) 37-55; W. W. Forten­
baugh, Aristotle on Emotion (London 1975) 23-25, has advanced the view that "an 
implicit or unformulated bipartite psychology" is employed, owing its impetus to the 
Academic investigation of emotion. The moral psychology of the Laws neither differs 
in its terminology from tripartition nor unambiguously advances bipartition, and there­
fore does not require discussion here. 

7 Supra n.1: 163-65 and 278f. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

286 PERIP ATETIC PSYCHOLOGY 

the context of 1 1 82a24f and the principles of Aristotelian bipartition 
plainly rule out this interpretation.8 

Clearly a different approach is necessary. My purpose here is to 
demonstrate, by examination of the author's psychological doctrine 
and its antecedents, that his attribution of the division aAo'Yov/AO'Yov 
ex.ov to Plato is based upon an interpretation of tripartition in the 
terms of Aristotle's doctrine of OPEgc8, resulting in a fundamental but 
most influential misrepresentation of Plato's psychology that oblit­
erated the differences between Platonic tripartition and Aristotelian 
bipartition. The most important of these differences concerns the 
role assigned to the soul's desiderative elements. Although Aristotle 
takes over Plato's division of these elements into (Jv,.w~, emOv,.ua, 
and /3oVA-QUl8, he groups them together into a single IJ-EPO~, which 
he designates emOvIJ-YJTLKfW Kat. OA~ OPEKTLKOV (Eth.Nic. l102b30), 
whereas each tripartite soul-part has its own em(Jv,.uaL. Bipartition 
and tripartition therefore presuppose fundamentally different relations 
between the 7Ta(JYJ and AO'YO~.9 But the author of the Magna Moralia 
(or his source: see Appendix), by imposing the Aristotelian analysis 

8 Although 1182a24f is often compared to the Timaeus, its psychology is an ante­
cedent of Aristotle's scala naturae, not of his bipartite division (cf. Solmsen [supra n.4] 
148-64); and the criticism of Platonic moral psychology that follows 1182a24f has 
nothing to do with the Timaeus; cf. nn.35-36 infra. 

9 In tripartition each of the soul-parts has its own E7TtfJvJLiat (Resp. 580D, 583A), and 
hence represents an independent form of motivation, whereas in bipartition all three 
species of 0PEgt<; are grouped together in the 0PEKTtKOII, which, while alogical in itself, 
is open to the persuasion of AO')'O<; (e.g. Eth.Nic. l102b28-1103a3; Pol. 1334b7-28); for 
the evidence, cf. Fortenbaugh (supra n.6) 31-39. This difference entails an important 
revision in the assignment of {3oVAr,ut<;. Plato does not explicitly attribute {3oVAr,ut<; to 
the AO')'UTTtKOII, and in fact associates it with E7TtfJvJLia (cf. Laws 9.863B6-9, 10.904B8-
c4); but some Academic psychologists who systematized Platonic tripartition locate 
'TTClUa {3oVATjut<; Ell 741 AO,),UTTtK4I (Top. 126aI4). Aristotle objects that it is ClTO'TTOII to 
divide the 0PEKTtKOII (De An. 432b6-7; cf. Vander Waerdt [supra n.3]) and, in con­
trast, assigns {3oVAr,Ut<; together with 6vJ.W<; and E'TTtfJvJLia to the 0PEKTtKOII (cf. Eth.Nic. 
l102b30; Pol. 1334b22-23; Rh. 1369al-4; De Mot. An. 700b22; De An. 411a28, 
414b2, 433a23-28). By reorganizing the desiderative elements of tripartition, Aristotle 
lays the foundation for an entirely different conception of moral action. The AO'YUTTtKOII 

no longer possesses an independent form of motivation, but becomes the seat of ~Et­
IIO'TT/<;, the morally neutral ~vllaJ..U<; of the soul that secures the means to the end set 
by ethical virtue, and which, when fused with ethical virtue, is identical with f/>POllr,ut<; 
(I144a23-37). 

A point of terminology: Plato usually calls the parts of the soul Er~r, or l~Eat, whereas 
the Peripatetics usually call them ~pr, or J.Wpta. Post-Platonic writers generally use the 
term fJvJ..UKOII to refer to Plato's fJv/-WEt~E<;. This terminological shift clearly occurred in 
the Academy, for Aristotle consistently uses -tKO<; formations (Top. 129a12; De An. 
432a25, 433b4), which appear as well in a variety of Peripatetic texts: e.g. Mag. Mor. 
1185a21, Arius 117.17, Aetius Dox.Graec. 390.4, and Div.Ar. XII 15a5, 17b6. In the 
Platonic Definitions, both -tKO<; (415ell) and -Et~E<; (413a7) formations appear, suggest­
ing that this terminological usage was not firmly fixed. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 287 

of OPE~t8 upon tripartition, inevitably rejects the special status Plato 
assigned to the (JvJ.LoEtSec;, and in a bipartite dichotomy opposes the 
em(JvWflTtK()lJ and (JvJ.LoEtSec; to the i\O),t<TTtKOZl. This interpretation 
leads the author to represent the tripartite psychology of the Republic 
as a division into ai\o)'oZl and i\O)'OZl EXOZl (l182a24-25) without even 
explaining (as e.g. Aetius does, Dox.Graec. 389.10-390.4) the analy­
sis of tripartition upon which this attribution is based. 

Our investigation will enable us to clarify a variety of issues. While 
the author's attribution of bipartition to Plato provides no historical 
evidence for the Academic debate on soul-division, it does show that 
the issues and terms of that debate had already been obscured by the 
first or second generation of the Peripatetic school through inter­
pretation of earlier thought in the terms of Aristotelian doctrine.lO 
Again, although the psychological doctrine of the Magna Moralia has 
never been thoroughly investigated, it well illustrates how Aristote­
lian thought was transformed and modified in the discussions of the 
early Peripatos, and how the author vacillates between independent 
philosophical speculation and strict fidelity to the tradition in discuss­
ing problems whose original import and terms he no longer under­
stands.ll The author's transformation of Aristotelian doctrine is of 
considerable importance for the subsequent history of Peripatetic 
ethics, over which the Magna Moralia exercised a wide influence.l2 

Indeed, the Peripatetic interpretation of tripartition sketched in the 
preceding paragraph dominated virtually all subsequent discussion of 
soul-division in later antiquity, and in the sequel to this article we 

10 w. Jaeger, "Uber Ursprung und Kreislauf des philosophischen Lebensideals," 
SitzBeri 1928, 402-12, and Walzer (supra n.1; c/ J. L. Stocks, Gnomon 7 [1931] 145-
51) have shown that the Magna Moralia is the work of a Peripatetic of Theophrastus' 
generation. I am not persuaded by the more recent attempt of J. M. Cooper, AlP 94 
(1973) 327-49, following Dirlmeier, to claim it as the report of a stage in the devel­
opment of Aristotle's moral philosophy earlier than the Eth.Eud. or Eth.Nic. This thesis 
is surely mistaken, and Cooper's arguments are sufficiently refuted by C. Rowe, AlP 
96 (1975) 160-72. A. Kenny's hypothesis that the Magna Moralia "is a student's pub­
lished notes of the course which we have in its authentic form as the Eth.Eud." (The 
Aristotelian Ethics [Oxford 1978] 220) ignores the substantial doctrinal differences 
between these two works (some of them to be explored here) that make this hypothe­
sis unlikely. In my view the Magna Moralia is the work of an early Peripatetic who 
generally bases his exposition upon the Eudemian Ethics (see Kenny's table, p.1l) but 
who readily introduces material from the Nicomachean Ethics (c/ D. J. Allan, lHS 77 
[I957] 7-11, and Gnomon 38 [I 966] 142-44), and who often defends the framework of 
Aristotelian doctrine even when he does not understand its philosophical motivation 
(see 292f infra). The author's attempt to assimilate tripartition to bipartition is clearly a 
post-Aristotelian development. 

11 Similarly, Jaeger (supra n.10) 412. 
1~ See Dirlmeier's survey (supra n.l) 99-110; on Arius' use of the Magna Moralia c/ 

n.32 infra. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

288 PERIPATETIC PSYCHOLOGY 

document the influence of this doctrine and trace the complex history 
of its transmission.13 

Our first task, however, is to set forth the evidence and clarify the 
philosophical motivation for the Peripatetic interpretation of tripar­
tition, together with an examination of the psychological doctrine of 
the Magna Moralia. The following discussion comprises two sections: 
in the first we assemble the evidence for the author's interpretation 
of tripartition; in the second we consider its historical antecedents 
and philosophical motivation. 

I 

The author of the Magna Moralia generally patterns his moral psy­
chology on Aristotle's, accepting bipartition (I185b4-13) and the fur­
ther subdivision of the aAo'Yov into the (JPE1TTLKOV or CPVTLKOV (Eth.Nic. 
1102a32-bI2, 1144a9-10; Eth.Eud. 1219b21-24, 37-40) and the E1n­
fJv/J/''1TLKOV Kat OAW" OPEKTLKOV (Eth.Nic. 1102b13-1103a3; Eth.Eud. 
1219b27-1220aI2), as well as the subdivision of the AO'YOV EXOV into 
the AO'YUTTLKOV or {3ovAEVTLKOV and the E1TLO"T"f'I, .. U)vLKOV (I 139a3-15). 
Although he adheres to this framework, the author's understanding 
of Aristotle's moral psychology is strongly influenced by debate cur­
rent in the Peripatos. Consequently, he often preserves Aristotelian 
doctrine and terminology in a context alien to the original, or formu­
lates problems under discussion in the Peripatos in an Aristotelian 
guise. This kind of doctrinal synthesis goes hand-in-hand with interpre­
tation of earlier thinkers through the lens of Aristotelian doctrine, 
Thus, in investigating the author's understanding of tripartition we 
need to consider both his modifications of Aristotelian doctrine on 
soul-division and his Aristotelian interpretation of Platonic psychology. 

We may begin with the author's conflation of the bipartite and 
tripartite terms for the rational faculty, He uses the term AO'YLO"TLKOV 
in its tripartite sense to refer generally to the rational faculty, rather 
than in its bipartite sense with reference to the faculty of CPPOVTJO"L", 
the lower subdivision of the AO')'OV EXOV, Thus, for example, the 
author argues that man acts according to op(Jo" AO'YO", OTav TO aAO­
'YOV ,..tEpO" T~" $VX~" /-LT, KWAVy, TO AO'YUTTLKOV EVEP'Yliv TT,V aVTov 
EVEP'YEtaV (1208a9-1 1) ,14 At the outset, it appears that the author 

13 See P. A. Vander Waerdt, "Peripatetic Soul-Division, Posidonius and Middle 
Platonic Moral Psychology," GRBS 26 (1985), forthcoming. 

14 On op9o~ A01'O~ c/ K. Barthlein, ArchGP 45 (1963) 213-58. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 289 

here follows Aristotle in maintaining that the standard of op(Jo~ AOYO~ 
is cPPOV'YI(TI8 (cl Eth.Nic. l103b31-34, l106b36-1107a2, 1l38b18-25, 
1144b21-30), in which case he would naturally use AO'Y(.UTtKOV to 
refer to the lower subdivision of the AO'YOV EXOV (1l39a7-15)~ but his 
subsequent explanation makes clear that he rather regards it as a 
general term for the rational faculty in the manner of tripartition, for 
he coordinates the aAoyov and AoytUTtKOV with TO XEtpOV and TO 
{3EATtOV (l208a12-18) and with the 7Tafh) and vov~ (l208a19-2l). 
The context shows these terms to be equivalent: in each case man 
acts according to op(Joc; AO'Yoc; when his alogical part does not prevent 
the rational part from fulfilling its EpyoV.15 One consequence of this 
use of AO'YUTT('KOV in its tripartite sense is that the author modifies 
Aristotle's subdivision of the AO'YOV EXOV. Although he locates uo­
cPia in the E7TtUT'YIJ.LOV('KOV , he refers to cPpov1)lT('c; not with AO'Yt­
lTTtKOV, but with a variety of terms: {30VAEVTtKOV, {30VAEVTtKOV Kat 
7TpOatpETtKOV, and 7TpOmpETtKOV J.Lcipwv T7jC; IjJvx7jc; (1l96b l3-34~ cl 
1197b3-1l) .16 Aristotle himself uses {30VAEVTtKOV in this way (Pol. 
1 260al3, Eth.Eud. 1126b25, Mem. 453al3, De An. 433b3, 434a12), 
and the author's usage accords with Aristotelian doctrine (cl Eth.Nic. 
1139a13-15).l7 At the same time, his use of AO'YtUTtKOV as a general 
term for the rational faculty-unexampled in Aristotle-shows that 
he has conflated tripartition with the bipartite doctrine and termi­
nology he generally follows. In tripartition the AO'YLlTTtKOV is used to 
refer generally to the rational faculty; but in bipartition it is used only 

15 This parallelism is marked by the repetition of the verb of prevention: KWAVn 
(1208alO), KWAVELV (a16), KwAvwm (a19). Note also the author's use of AO"),tUTLKOIJ 

and BtaIJ07}TLKOIJ as equivalent terms at 1182aI8-20. 
16 Cf Arius' subdivision: T01) BE AO")'tK01) TO ~IJ 1TEpt TlX ci:iBta Kat TO: lkia (JEWP7}TL­

KOIJ E1TtUT7Jj.LOlJtKOIJ KaAELU8at· TO BE 1TEpt TU eXv8pcdrrtIJa Kat TU <q,8apTU> 1TpaKTLKOIJ 
{3oVAEVTLKOIJ (117.12-15); later, following the Mag. Mar. , he refers to CPPOIJ7}Ut<; as a 
E~t<; {3ovAEVTtKT, (145.20); (1 Dirlmeier (supra n.1) 340. 

17 As Professor Cherniss well explains (per Iitteras): "In 1139all-15 the reason for 
asserting that TO AO")'tUTtKOIJ is a part of TO AO"),OIJ EXOIJ is explicitly said to be the fact 
that AO"),4,EU8at and {3ovAEvEU8at are the same thing. This justification is a criticism of 
the Academic use of TO AO"),tUTLKOIJ to designate the rational part of the soul as a whole. 
The Academics themselves ascribe {3ovAEvEu()at Kat AO")'4,Eu()at as well as BWIJOEL<T()at 

and all such activities to TO AO")'tUTtKOIJ (e.g. Diog. Laert. III. 90 = Div.Ar. 12, p. 15a5-10 
[Mutschmannl, not in Rose), and Plato himself said that in the tripartite soul TO AO"),t­

UTtKOIJ is analogous to TO {3ovAEVTtKOIJ in the tripartite state (Republic 440E 10-441 A3). 
In that case according to 1139all-15 TO AO"),tUTtKOIJ, being {3ovAEVTLKOIJ, cannot be the 
whole of the rational soul but must be only a part of it, the deliberative distinguished 
from the scientific or contemplative part (TO E1Tt<TT7Jj.LOlJtKOIJ) even as Aristotle in De 
Anima 433b3-4 distinguishes among the BVlJaj.LEt<; of the soul the {3oVAEVTLKOIJ from 
the 1J000TtKOIJ and both from the OPEKTtKOIJ ... [MM 1196b 13-17] in substituting {3ov­
AEVTtKOIJ for AO")'tUTLKOIJ merely makes explicit what according to EN 1139a 13-15 it 
[the subdivision of the rational faculty] really means." 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

290 PERIPATETIC PSYCHOLOGY 

for the lower part of the AO)'OV EXOV, and hence the author was faced 
with somehow reconciling this difference.I8 His resolution was evi­
dently not an easy one, for he employs four apparently equivalent 
terms, both bipartite and tripartite, to refer to the rational faculty.I9 
When he uses AO),LCT'Tl.KOV as a general term for the rational faculty in 
opposition to the aAo),ov (I182a18-22, 1208a9-10), or equates the 
AO),C.(J"'T(.KOV with the AO)'OV EXOV in representing tripartition in the 
terms of Aristotelian bipartition (I 182a24-25), it is clear that the 
fundamental differences between Platonic and Aristotelian soul-divi­
sion are no longer understood. 

Two factors are involved in the author's conflation of these two 
conceptions of the rational faculty. The first is his revision of Aris­
totle's doctrine on the relation between the 7ra6"fJ and AO)'O~, and his 
corresponding revision in the role of cPpOV"fJ(J"c.~ in providing the stan­
dard for right action. To understand this aspect of the author's doc­
trine on soul-division we need to recall the dispute between Theo­
ph rastus and Dicaearchus over the relative rank of the theoretical and 
practical lives-a controversy that brought about decisive modification 
of Aristotle's moral philosophy in the early Peripatos.2o The author 
alludes directly to this dispute in his discussion of whether cPpOV"fJ(J"c.~ 
rules over all the soul's parts W(J"TI'EP OOKE' Kat. CtTl'OpE'Tac. (I198b9) in a 
passage (b8-20) that constitutes a citation from Theophrastus; 21 and 

18 Aristotle never uses AO')'tUTLKOV and AO')'OV EXOV as synonyms. He uses AO')'tUTLKOV 

to refer to the faculty of CPPOVTJUt<; (Eth.Nic. 1139a12); and otherwise to the rational 
faculty of the Academic version of bipartition, twice in the De Anima (432a24, 432b6), 
and twelve times in the Topics (c! 298f infra); Phys. 21Oa30 is an echo of Academic 
bipartition, for in Aristotle's moral psychology E'TI"tUrT,f.£TJ would be associated with the 
E'TI"tUTTJf.£OVtKOV (cf Eth.Nic. 1139a6-12). Professor Cherniss explains the ambiguous 
phrase TO AO')'tUTLKOV Kai <> KaAOlljUvo<; VO~ (De An. 432b26) as follows (per lilteras): 
"The Kai here is explicative so that TO AO')'tCTTLKOV might be explained here as 'what is 
called VO~' by people (scil. Platonists); but Aristotle immediately proceeds to consider 
VO~ in its two parts, 'contemplative' and 'practical', and even to use it to cover cpav­
TaULa W<; VOTJULV TLva. In short he may here be using AO')'tCTTLKOV Kai <> KaAovjUvo<; 
VO~ purposely to cover all possible theories of intellection in his attempt to prove that 
in no sense can this be the cause of locomotion." 

19 AO')'tCTTLKOV (1182a20, 1185a21, 1208alO); I)WVOTJTLKOV (1182aI8); TO {JEA.TtOV 
o 196a27, a30, 1208a13); AOYOV EXOV occurs eight times. 

20 This dispute is mentioned several times by Cicero: nunc prorsus hoc statui ut, quo­
niam tanta controversia est Dicaearcho, familiari tu~, cum Theophrasto, amico meo, ut iIIe 
fUUS T(lV 7TpaKTLKOV f3Wv longe omnibus anteponat, hie autem TOV (JeWPTJTLKOV, utrique a 
me mos gestus esse videatur (Alt. 2.16, cf 2.2,2.12, 13.30; Fin. 5.4.1 I). On the contro­
versia see Jaeger (supra n.l0) 412-21 and O. Regenbogen, RE Suppl. 7 (1940) 1481, 
1489-91 s.v. "Theophrastos"; and, on Cicero's use of Dicaearchus, S. E. Smethurst, 
TAPA 83 (1952) 224-32. 

21 See G. Heylbut, ArchGP 1 (1888) 194-99, and W. Jaeger, Hermes 64 (1929) 
274-78; Dirlmeier's objections (supra n.l: 354-56) to Jaeger are not cogent and are 
rejected in the most recent discussion, Fortenbaugh (supra n. I) 182-84 on Ll9. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 291 

he illustrates in many ways the weakening of the connexion between 
1TOALTtK7] and the theoretical life, so fundamental for Aristotle, that 
resulted from this debate.22 Such revision of the relation between 
1TOALTtK7] and the theoretical life inevitably entails revision of the 
relation between practical and theoretical reason.23 The author's ac­
count of virtue does not culminate in (]'ol/>La (as in Eth.Nic. 1141a9-
22, 1177a12-1178a8)~ in fact, he even finds it necessary to argue that 
it is a virtue at all 0197b3-11; cf 1198a22-32) and to apologize for 
introducing it in an investigation of 1TOALTtK7] (l197b28-36). This as­
sumption of the autonomy of virtue from the theoretical life leads the 
author to treat as identical standards for opBot;;; AoYOt;;; a variety of 
terms that, as originally formulated by Plato and Aristotle, denoted 
different operations of the rational faculty and presupposed different 
relations between the mxB7'I and AoYOt;;;. The author's conflation of 
bipartite and tripartite doctrine on the rational faculty is one conse­
quence of the early Peripatetic debate over the relation between prac­
tical and theoretical reasoning. 

The second and more general factor involved in the author's con­
flat ion of bipartition and tripartition is his interpretation of the latter 
in the terms of Aristotle's doctrine of opeg"t;;;; consequently, he simply 
equates the AOYUTTtKOV and AOYOV €XOV without attention to the 
different conceptions of the relation between the mxB7'I and AoYOt;;; 
presupposed by each. Before taking up this aspect of the author's 
interpretation, however, it will be helpful to observe the way in which 
he subsumes the E1TLBvJJ,7'ITtKOV and BVJJ,LKOV into a bipartite dicho­
tomy in opposition to the AOYL(]'TtKOV. 

The author departs significantly from Aristotle by replacing the 
OpeKTtKOV with a faculty of 0pJJ,,.q.24 As a result, he does not employ 

~2 The author takes the emancipation of 1TOAlTLKT, from the theoretical life for granted: 
in his introductory doxography he cursorily criticizes Pythagoras, Socrates, and Plato for 
failing to recognize the autonomy of virtue from theoretical considerations (1182alO-
30); and he does not hesitate to oppose the entire previous tradition of Greek ethics: 
a1Tt..~ 8' OVX, Wa-1TEP OLOV'Tat Ot eXt..t..OL, rij .. apErij .. apx~ Kai irYE/.Ulw EU'TLV 0 t..0'Y0", 
at..t..a ,ux.t..t..ov 'Ta 1T(xBT/ (l206b 17-19). 

2:1 Cl the author's ingenious reinterpretation of CPPOVT/UL" as E1TL'TP01TO" 'Ti .. ... 'TTl" 
uocpiar; (1 198b8-20; Lf 1143b33-35, 1145a6-11; Eth.Eud. 1249b9-23) and Jaeger's re­
marks (supra n.10) 408f, which do not depend upon his mistaken view of cppOvT/ULr; in 
Aristotle. 

24 Although the author's account of moral virtue follows Aristotelian lines, his con­
cept of opJ..t-q, which occupies a place in his soul-division analogous to Aristotle's OPEK­
'TLKOV, rests upon entirely different principles, and hence the author does not employ 
the terms E1TLBvJ..tT/'TLKOV or 0PEK'TLKOV. He has no term to replace these, however, and 
accordingly resorts to periphrases: 'TTlr; l/lV)(TI" 'TL J..tfpOC; ~ TjooJ..tdJa (1204b26; c/ b35) 
and EIJ 'Yap -rii l/lvxii EVEU'TLIJ 'Til CPVUEL 'TOLOWOV ~ Op,.,unJ..tEV at..o'YWC; 1TPO" eX alJ Eli 
E)(WJ..tEIJ (I207a38-bI). In one passage of ten lines 0204b26-36) he uses 'TO ,.wPWIJ 'TTl" 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

292 PERIPATETIC PSYCHOLOGY 

either the bipartite or tripartite terms for appetite, and we therefore 
find no explicit evidence, as in the doxographers, that he collapsed 
the fJvi-WEtBE~ and E1T'tfJvl-t"fJTtK()l) into a single aAoyov-a point that 
would assist our argument that he viewed tripartition in the terms of 
Aristotelian bipartition. One passage, however, does betray the au­
thor's familiarity with tripartition; this, together with evidence from 
Arius Didymus, will clarify the situation. 

In introducing the (JPE1T"TtK()l) , which is "apparently" (W~ BOKEt, 
1185a15) a part of the soul, the author finds it necessary to explain 
his choice of terminology: rij~ BE t/lvxij~ T01JTWV I-tEv TWV i-Wpiwv 

'(J' " "" " 'A.. (J '" , 'I. ' ''' fJv ' OV EV atTtOV av Et"fJ TOV TpEo/EU at, OWV TO ",OytUTtKOV "fJ TO /-ttKOV 
" , 'fJv 1,,\ \ §::: , '''' -?'(J'''' " "fJ TO nrc. I-t"fJTtKOV, a"""o uE Tt 1T'apa TaVTa, ~ OV EV EX0I-tEV OtKEW-

TEpOV OVOJUX E1T't(JE'iVat';; (JPE1T"TtKOV (I 185a20-23). The apologetic 
way in which he introduces the (JPE1T"TtKOv ~ETa TOVTO TO iJ.-EAAOV 

, (J " , ~.!i: " ,,,... , " , AE'YEU at OVTE AWV uu~EtEV av OtKEWV Etvat TOVTWV OVTE JUXKpav 

a1T'EXOV, 1185a13-15) suggests that it does not entirely conform to 
the principles of his own psychology; and indeed his subsequent 
denial that it possesses its own apErI] (on the grounds that OVK EOt­
KEV BE ECvat 0PWTI EV T~ I-tOp~ TOVT~, 1185a29) shows that he has 
adapted Aristotle's framework to his own theory of Opl-tr,. Aristotle's 
moral psychology turns upon a bipartite dichotomy between OPEgt~ 
and AOyO~ in which the (JPE1T"TtKOV is not a distinguishing mark of the 
human soul (cj. Eth.Eud. 1219b26-1220a3); accordingly, he is most 
often concerned with the faculties of desire and of practical and 
theoretical reasoning. But Aristotle can extend his division 'down­
ward' to encompass such a soul-function as the (JPE1T"TtKOv, which 
does not contribute to human excellence but which man nonetheless 
possesses as a living animal; and, to incorporate his moral psychology 
into the framework of his scala naturae, he introduces it as his fourth 
soul-part, although usually only to dismiss it as irrelevant to ethics.25 

Our author plainly does not understand why Aristotle included the 
(JPE1T"TtKOV as a separate soul-part; nevertheless, because he takes 
over Aristotle's four-part division, he is faced with explaining its 
inclusion. His W~ BOKEt shows that in defending its relevance he is 

t/Jvx.i/ .. once and rij .. t/Jvx.i/ .. Tt J-LEpo" four times because he does not have a general 
term for this faculty. On 0PILr, in the Magna Moralia see Walzer (supra n.1) 164-70, 
and P. L. Donini, L'Etica dei Magna Moralia (Turin 1965) 179-207. 

25 The 8PE7T'TLKOV is not peculiar to man and therefore has no share in human excel­
lence (Eth.Nic. 1102a32-b12; Eth.Eud. 1219b21-22, 37-40); Aristotle's inclusion of it 
as a soul-part enables him to relate bipartition to the scala naturae of his scientific 
psychology, and thus to contrast man with other forms of life (1 097b24-1098a 18; cf 
Plut. Mor. 4428). 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 293 

following received doctrine, and it is in this context that his enumera­
tion of the tripartite soul-parts should be seen. In claiming that none 
of these parts accounts for growth and nutrition, the author follows 
Aristotle's revision of tripartition.26 Yet the very fact that he is willing 
to think in terms of tripartition shows that, while bipartition had 
become canonical, he and his audience were still sufficiently familiar 
with tripartition to use it as a foil to justify the inclusion of an Aristo­
telian soul-part that did not contribute to his own moral psychology. 
Clearly, bipartition did not entirely eclipse tripartition in the early 
Peripatos,27 and the author's acquaintance with the 8V/-UKC)JJ and €1TL-
8v",,"'rn.KOV compels us to ask whether he considered them separate 

26 In the Timaeus the E1TdJvf.LT/TLKOV, planted midway between the midriff and the 
navel, is said to be tied up at a sort of manger for the feeding of the uW/-UX without 
which no mortal stock could exist, and so, as elsewhere in Plato, it is connected with 
nutrition, plant life, and bodily desires of various kinds (7007-E6; cf 72E3-73A8, 
77A2-c7; Phdr. 247E4-6; Resp. 436AlO-BI, 43702-4, 439AI-08, 580E3-4), as well 
as with a host of diseases of the soul including 8VUKOAia, 8VCT(Jv~, and 8VCT/-UXfJia 
(86B 1-87B8). The E1TLfJVf.LT/TLKOV thus incorporates both nutritive functions that Aris­
totle would assign to the fJPE1TTtKOV, and emotions he would assign to the 0PEKTLKOV; it 
does not distinguish nutritive and emotive functions-both influenced by and indepen­
dent of bodily drives-as Aristotle does; cf Solmsen (supra n.4) 156f. In claiming that 
none of the tripartite soul-parts accounts for nutrition, the author interprets tripartition 
in Aristotelian terms. Yet his reasons for denying apETJ} to the fJPE1TTLKOV show that, 
although he adopts Aristotle's four-part division, he has completely departed from its 
principles. The fJPE1TTtKOV does not, for Aristotle, contribute to Ev&n,.wvia because of 
its status within the scala naturae; for the author, it is because this faculty lacks Opf.LTj 
and hence the capacity for EVEP'YEI.a (I 185a27-35). 

27 Apart from the Magna Moralia and Arius, the most important sources for Peripa­
tetic doctrine on soul-division are the doxographical tradition, derived from Peripatetic 
sources no longer extant (cf supra n.13), and the PS.-Plutarchan De libidine et aegritu­
dine (on authorship and date see Sand bach 's remarks with references to earlier work, 
RevPhil 43 [I969) 211-16), Clearly, Theophrastus conceived the relation between the 
1TaOTj and AO')'O~ along fundamentally Aristotelian lines, whatever his original contribu­
tions to the Peripatetic investigation of 1TafJo<; (as Fortenbaugh recognizes: see his 
"Arius, Theophrastus and the Eudemian Ethics," in On Stoic and Peripatetic Ethics: The 
Work of Arius Didymus [London 1983) 203-23). In several fragments Theophrastus 
refers to fJvJ.Lci<;, E1TLfJv~, and AOYO<; (LI-L4, L88, LI1?); but there is no explicit 
evidence concerning his soul-division. De libid. 6 reports that Diodorus assigned some 
of the 1TaOTj to the aAo'Yov, others to the AO')'tKOV (cf Zeller [supra n.41 11.2 933 n.4). 
The later Peripatetics engaged in extended controversy with the Stoics over the status 
of the 1TaOTj, and accepted a theory of aruOTjCTt~ that caused them to assimilate the 
TJ'YEf.LOVtKOV to Aristotle's bipartite division (cf Dox.Graec. 394.21-25). We are best 
informed about Strato (frr.l07-31), who localized all arulJTjCTt<; in the TJ'YEf.LOVtKOV, but 
it is unclear how he connected his physiological theories with bipartition (cf fr. 74, 
Dox.Graec. 416.10-13). A separate Peripatetic tradition is represented by the Divisiones 
A ristoteleae , the De virtutibus et vitiis, and the nEpi 1TafJwv attributed to Andronicus, 
which coordinate their treatment of the Aristotelian virtues and vices around Plato's 
tripartite division (cf Div.Ar. 12-13 pp.15-17; VV 1249a31-1250a2; De aff. 241.21-
33). On the latter two works and their place in the tradition of Peripatetic moral psy­
chology, see Glibert-Thirry (supra n.l) 1-34. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

294 PERIP ATETIC PSYCHOLOGY 

soul-parts or, like the later doxographers, subdivisions of a single &AO­
'Y0Jl. The evidence from the author's own text is inconclusive, al­
though only shortly after enumerating the tripartite soul-parts he sets 
forth his own division into &A0'Y0Jl and AOYOJl EXOJl (I 185b4-13), with­
out indicating any incompatibility between it and tripartition-a fact 
easily explained if he interpreted tripartition in the terms of bipartition, 
and hence recognized no fundamental difference between them. 

More decisive, however, is evidence from Arius Didymus that 
shows how the Peripatetics incorporated tripartition into Aristotle's 
bipartite framework.28 Arius begins by describing the A0'Yt.KOJl as Kpt.­

Tt.KOJl, and the &AO'YOJl as 0PI-L"I1Tt.KOJl (I17.12); the latter term would 
well describe the faculty of OPI-L"I1 in the Magna Moralia, but it ap­
pears to have originated under Stoic influence, and does not occur 
before Arius.29 Arius then subdivides the &A0'Y0Jl as follows: Kai 

2~ On Arius' sources see, most recently, P. Moraux, Der Aristotelismus bei den Grie­
chen I (Berlin 1973); whether or not Arius knew Aristotle's writings directly, the 
handbooks of Peripatetic doctrine he may have used are likely themselves to have been 
derivative compilations, and hence Arius may be expected to preserve earlier doctrine 
fairly closely even when he is not directly acquainted with the ultimate source. 

29 Arius' use of KPLTLKOV and OPI-''YITLKOV is elaborately paralleled by Numenius, fr.I8 
(in E. des Places, Numenius [Paris 1973]); cf P. Merlan, Philologus 106 (962) 141f. 
The doxographical usage of OPI-''YITLKOV does not permit us to establish its early history 
with any confidence. It is applied to Aristotle (Dox.Graec. 438.12-14, 457.2), to Plato 
(Didasc. 178.32-37, Timaeus Locrus 102E; cf A. E. Taylor, A Commentary on Plato's 
Timaeus [Oxford 1928] 661-63), and (paired with the €7TL9vI-''YITLKOV and AO'Y~KOV) to the 
Stoics and Epicureans (Dox. Graec. 438.15-20). Its appearance in Plutarch's Adversus 
Colotem (Mar. 1122 A - D) shows that it played an important role in the debates of the 
Hellenistic schools, and hence it is hardly surprising that the doxography is unclear. The 
usage of oPI-''YITLKOV as a Stoic soul-part is not well attested (Dox.Graec. 438.15-20), and 
orthodox Stoics employed an eight-part division with five ainffrJrTjpux, the cPwVTjTLKOV, 
the C77TEPI-'UTLKOV, and the 7}'YEIJ.OV~KOV (the evidence is collected in SVF II 823-33; note 
also SVF I 143, II 836, 879 ; cf A. A. Long, Hellenistic Philosophy [London 1974] 
170-78). The Stoics denied the alogical emotions of the soul a status separate from the 
rational faculty (SVFI 209, II 906, III 459-64; Ga\., De plac. 190.12-13,334.1-3; cf A. 
Dihle, The Theory of Will in Classical Antiquity [Berkeley 1982] 61-64), and hence Arius' 
description of the aAoyov as OPI-''YITLKOV contravenes Stoic doctrine and appears to repre­
sent a Peripatetic adaptation of the Stoic theory of OPI-'T,. The status of OPI-'T, in the Stoic 
soul-division is brought out clearly by Plut. Mar. 441c-D (cf 449c); "impulse in 
excess" is in fact the accepted Stoic definition of 7TCi80<; (cf SVF 1205-207; III 377-78, 
391, 462, 479). According to Stoic doctrine, each of the soul's ~P'YI has its own Bvva­
I-'E~<; (SVF II 826), and OPI-'T, was considered a 8vval-"<; AO'Y~KT, (Dox.Graec. 410.28, SVF 
III 463; cf Arius 86.17-87.13 [=SVF III 169]). The Stoic theory of OPI-'T, is set out in 
detail in Galen's polemic against Chrysippus in De Placitis Hippocratis et Platonis IV; 
Chrysippus' definition of OPE~t<; (opi(,ETat yovv av~v OPI-'T,V AO'Y~KT,V €1Ti TL<vo<;> 
ouov XpT, i/80v< TO<; > [238.35fl) suggests how easily OPE~~<; and OPI-'T, might have been 
harmonized by one for whom, after all, in Arius' own words, aAoyov BE AE'YEu8at .pVXTI<; 
~po<; ov Ka8a1Ta~ aAoyov, <lAAll TI) OWV TE 1TEi8Eu(Jat AOYCP, 01TOWV €UTt TO 1TaffrJTtKOV 
(117.7-9; cf Eth.Nic. 1102b13-1103a3). Thus while the use of OPI-''YITLKOV as a soul-part 
contradicts Stoic doctrine, the term was easily employed by Arius, whose understanding 
of Aristotle's OPEKTLKOV (Eth.Nic. 1102b30) was influenced by Stoic OPI-'T,. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 295 

TOl} aAoyov T() /-tEll OPEKTLKOll TWll EcP' i,J.ii,1l E7TdJvWYlTLKOllo TO BE 
1T'por; Tovr; 1T'ATW'LOlJ OtOll a/o.WlJTLKOlJ 8vJ,LLKOlJ (117.l6-18). Wachsmuth 
rightly comments "hanc divisionem animae notum est Platonicam 
esse" (ad loc.); for the subdivision one would expect if the author 
were following Aristotle is that into (JPE1T'TLKOlJ and OPEKTLKOlJ (Eth. 
Nic. l102a33-1103a3). Arius is clearly familiar with this and other 
Aristotelian doctrine on soul-division, perhaps through an interme­
diate source; 30 but the Peripatetic doctrine he here reports has joined 
together the two lower Platonic soul-parts into an aAoyoll, even 
though he characterizes the (JVJ,LLKOll in a way that departs from its 
primary Platonic sense.31 This passage shows how Platonic, Aristote­
lian, and Stoic doctrine on soul-division was harmonized in the Per­
ipatos; and since all the elements present in Arius are present also in 
the Magna Moralia, comparison of these two texts sheds light on the 
history of this process of harmonization in the Peripatetic school.32 As 
one would expect, the doctrine preserved in Arius seems to be more 
fossilized than that in the Magna Moralia: his use of the single term 

80 Arius sometimes bases himself directly on an Aristotelian text (e.g. l39.19-140.6 
with Eth.Eud. 1220b21-33; (f Kenny [supra n.lO] 20r), and incorporates verbatim 
extracts from the Magna Moralia; but it is usually difficult to judge whether his refer­
ences to Aristotelian doctrine derive from the original or from intermediate sources 
(see supra n.28, and n.32 infra). Arius' definition of 17aOo<; as TO OpeKTLKOIJ /-LEP0<; 
T1]<; tjlVX1]<; elOur,."elJolJ V17aKOlJetIJ T~ h.OYLK~ (38.9-10), which clearly recalls Eth. Nic. 
1102b30-1103a3, suggests that he could have represented the OpeKTLKOIJ in terms 
closer to Aristotle's; but, as in the Magna Moralia, the influence of the Stoic doctrine 
of 0P/-L,.q causes him to depart from Aristotle's doctrine on soul-division. 

:11 Aristotle's definition of OPYl1 as an opeg,<; aIJTLh.v~{Tew<; in the De anima (403a31; 
cf Top. 156a31-b4, 127b31-33) was widely quoted in antiquity and probably stands 
behind Arius' usage here; it is translated by Seneca (De ira 1.3) as oil iram esse cupidi­
tatem d%ris reponendi, and is reflected in Plutarch's We; €7TLOv,."w1J TLlJa T()IJ Ov,."o1J olJm 
Kat opeg'lJ aIJTLh.v~(J"ew<; (Mor. 442B; c! PS.-Plut. De libid. O. Arius describes the 
OV/-LLKOIJ in terms of the 17aOo<;, which is regularly called OPYl1 in Aristotle (e.g. Eth.Nic. 
1126al3-blO), but which Plato (Resp. 440A5, c2) and the Academics (Top. ll3a33-
b3, 126alO) locate in the fJV/-LOELOE<;. 

:12 The soul-divisions of Arius and of the Magna Moralia are closely related; although 
not all Arius' doctrine derives from the Magna Moralia, one passage is directly ex­
cerpted, and Arius elsewhere draws upon closely-related Peripatetic sources. Arius' 
catalogue of apemi associated with the ah.0yolJ and h.oyolJ £XOIJ 037.19-23), apart 
from his addition of KClAoKaYClfJia, corresponds verbatim to the enumeration in Mag. 
Mor. 1185b4-13; and both in turn are paralleled by the shorter lists in Eth. Nic. 
1103a3-10 and Elh.Eud. 1220a5-12 «(1 Walzer [supra n.ll 184). Dirlmeier (supra n.O 
2061' has shown that Mag. Mor. 1185b4-13 is excerpted from a Peripatetic catalogue (cf 
Rh. 1362b24) recording doctrine derived from Plato but not integrated into the struc­
ture of the Magna Moralia as a whole (ayxilJoLa, ev,."afJeLa, and /-Lv,.q/-L'rj do not recur). 
On Arius' incorporation of passages from the Magna Moralia see H. von Arnim, Arius 
Didymus' Abriss der peripatetischen Ethik (SitzWien 204.3 [1926]), who wrongly argued 
that Arius derived his knowledge of the Magna MoraNa through Theophrastus (c! 
Dirlmeier [supra n.ll 206f); Allan, JHS (supra n.10); 7-11; D. J. Furley, "A Note on 
Arius and Magna Moralia I.l-2," in Fortenbaugh (supra n.27) 160-64. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

296 PERIPATETIC PSYCHOLOGY 

{3oVAEVTLKOV where the author of the Magna Moralia uses several, 
and his use of 0P#L'YITLKOV where the author-who clearly would use 
this term if he had it-uses long periphrases, show that the termin­
ology of soul-division in the Magna Moralia is more fluid than Arius'. 
It displays inconsistencies that have been systematized away by Arius' 
time.33 Yet the author and Arius agree on every point where we can 
compare them, and hence we are justified in concluding that the 
author, like the later doxographers, collapsed the 8vf.UKOV and E7TL8v­
#L'YITLKOV into a single aAoyov in opposition to the AoytUTLKOV.34 

The mixture of bipartite and tripartite doctrine and terminology in 
the Magna Moralia is the key to the question of the author's attribu­
tion of bipartition to Plato. The author reduces tripartition to a bi­
partite dichotomy, and interprets Plato's tripartite division from the 
perspective of Aristotelian bipartition. He uses AOYOV lxov and Aoyt­
UTLKOV indifferently to refer to the rational faculty, and opposes it to 
the E7TL8v#L'YITLKOV and 8vf.UKOV in a bipartite dichotomy, a usage antici­
pated in Academic discussion (see 298f itifra). In 1182a24f he opposes 
the AOYOV lxov (= AoytUTLKOV) to the aAoyov without mentioning its 
subdivision. The doxographical tradition confirms this interpretation, 
and the author's mixture of bipartite and tripartite terminology shows 
that in attributing the division into aAoyov and AOYOV lxov to Plato he 
records the first instance of this doctrinal harmonization. 

This interpretation not only places the author's doctrine on soul­
division in its proper philosophical and historical setting, but also 
explains the context in which his attribution of bipartition to Plato 
appears. Commentators who refer 1182a24-25 to the Timaeus com­
pletely ignore its context, for the author plainly states that (Plato) 
(l:7Te&JKEv EKauTC~ [TClS"1 aPET£IS" T£1S" 7TpoO"TJKOlJUaS" (I 182a25f), and 
only in Republic 4 are the aPETai correlated with the soul-parts.35 

aa The terminology of 117.11-18 has been coloured by Stoicism but parallels doctrine 
in the Magna Moralia; cf supra n.32. 

34 For the doxographical tradition cf supra n.13. 
35 The context of 1182a24-25 precludes arguing that this passage records esoteric 

doctrine and that late in his life Plato adopted Aristotle's bipartite division. According 
to the author's criticisms, Plato would then have failed to understand the principles of 
the bipartite division for which he abandoned tripartition: that, although he assigned 
the virtues to soul-parts, he failed to recognize that the ethical and intellectual virtues 
correspond to the alogical and logical parts of the soul. But Plato would hardly have 
adopted bipartition without accepting its underlying principles. The suggestion (c./ 
Dirlmeier [supra n.ll 166) that this passage may be referred to Plato's IIEpi TO}, clya8o}, 
is wholly unwarranted: in our earliest testimonium, Aristoxenus (Harm.el. 2.30-31) 
plainly quotes Aristotle as saying that most of those who heard this lecture were disap­
pointed because it did not deal with what was considered the human good, and none of 
our other evidence (collected most recently by K. Gaiser, Phronesis 25 [1980] 5-37) 
suggests that this work dealt with moral psychology. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 297 

The author's subsequent remarks (I182a26-30) appear to incorporate 
stock Peripatetic criticism of Platonic moral psychology, and may 
simply repeat earlier doxography (see Appendix)~ but the only work 
of Plato to which 1182a24-30 can refer is the Republic .36 In repre­
senting tripartition as a division into aAo'Yov and AO'YOV €XOV, there­
fore, the author uses a kind of short-hand to refer to the psychology 
of the Republic.37 It is impossible to know whether it was the author 
or an earlier Peripatetic who first interpreted the psychology of the 
Republic in this way. But even if the author's interpretation derives 
from earlier doxography, his terminology and conception of soul­
division fully accord with it.38 Such harmonization of Platonic tripar­
tition and Aristotelian bipartition must have been commonly accepted 
in the early Peripatos. 

II 

Thus far we have considered the evidence for the author's inter­
pretation of tripartition, leaving aside its Platonic and Academic ante-

36 The author's criticism of Plato for mixing discussion of T(l Oll'Ta and aA:J}8eUl with 
that of apErr1 recalls the ~L<; of Ph/b. 61B-67B, esp. 64E-65A, and may reproduce 
Peripatetic criticism of that work; but it is most unlikely that direct criticism of the 
Phi/ebus is intended here. According to the author, Plato rightly divided the soul and 
assigned the corresponding virtues (J..LEXPL ~JI O~JI 'ToVrov KUAw.,), but then fell into 
error (p.e'Ta ~Jl'TOL 'TOWO OVKETt op8w.,). Only the Republic couples the psychological 
doctrine and investigation of 'TO aya80Jl to which the author here objects, and it pre­
sents them in the order corresponding to the author's criticism. If this interpretation is 
correct, T7JJI yap apen}JI KaTi~EJI Ei<; T7JJI 1TpaYJUl'TELaV T7JJI lnrEP 'Ta-ya8ov refers to 
the exposition of 'TO a-yu8ov in the RepUblic. If the author is reproducing stock Peripa­
tetic criticism, as is likely, doctrine from and criticism of several sources may here be 
conflated. 

37 The author's criticism of Socrates also derives from interpreting tripartition from 
the perspective of bipartition: the author naturally expects to find moral virtue associ­
ated with the aAO'}lOV; finding that Socrates' moral psychology rests on different prin­
ciples, he criticizes him accordingly. The author's objections to Socrates' identification 
of the virtues with E1TLCT'TiiJUlL (I 182a 15-23; cj. 1183b8-18) are a stock Peripatetic 
criticism (cj. Eth.Nic. 1144b17-30; Eth.Eud. 1216b3-25, 1246b33-36). Aristotle him­
self does not explicitly connect this criticism with bipartition, but the author (in 
1182a17-23) is clearly justified in doing so: Socrates' identification of virtue with E1TL­
urr1f.''TI prevents him from discovering the aAO'}lOv in its own right, and the author is 
historically correct in tracing Socrates' disregard for 1TCx8o<; and 'Jj8o<; to his disregard for 
the alogical part of the soul; see Fortenbaugh (supra n.6) 63-65. 

38 Note that the author takes Plato's soul-division as the antecedent for his own 
bipartition and does not recognize any incompatibility between the two soul-divisions: 
he approves of Plato's division into aAO'}lOJI and AO'}lOJI EXOJl (1182a26); and later, in 
introducing his own bipartite division, he says We;; 4xx~JI (1 185b4), which can only refer 
to his earlier attribution of bipartition to Plato. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

298 PERIPATETIC PSYCHOLOGY 

cedents. Let us now turn to the historical background and philosophi­
cal motivation of the Peripatetic interpretation of tripartition. 

Plato's tripartite psychology was interpreted in various ways by 
members of the Academy during his own lifetime. It is impossible 
now to reconstruct the details of the Academic debate on soul-divi­
sion or the transformation of tripartition into bipartition through 
Aristotle's reorganization of the desiderative elements of tripartition. 
For our present purpose it will suffice to indicate the contrast be­
tween the canonical expositions of Platonic tripartition and Aristo­
telian bipartition, since the author of the Magna Moralia simply 
interpreted the former in terms of the latter. But evidence of a differ­
ent Academic interpretation of tripartition may be found in the psy­
chological examples of the Topics. This evidence must be interpreted 
with caution, for many of Aristotle's examples are anonymous Aca­
demic 8ogat, recorded solely for dialectical purposes (cl IOla30-34); 
one cannot therefore assume that isolated passages on the same 
subject record a related body of doctrine, much less that these 8&gat 
represent Aristotle's own doctrines.39 Indeed, mutually exclusive defi­
nitions of the t/lvXT, are often advanced in the Topics, or a definition 
advanced in one passage is rejected elsewhere, or rejected on differ­
ent grounds.40 But while the passages on soul-division may not record 
a consistent body of doctrine, one passage provides a clear antecedent 
of the author's interpretation of tripartition: 

~ bTt TO 1TOAV SE Kat Ell TOt~ 1TAEia-TOt~ KaOa1TEp TO AOYUTTLKOV 

rSWll 1TPO~ E1TL9vf.L"'1TLKOll Kat 9vf.LLKOll TO TO ~lI 1TpOCTTaTTELlI TO S' 
It 1'"\ ~ , '\. " , '\\.' , , 

lYTTTJPETEtV" OVTE yap TO 1\.0ytCTTLKOll 1TallTO'TE 1TpOCTTaTTEL, al\./\. ElIt-
, , ." \, LL. ' , LL. ' " 

OTE Kat 1TpOCTTaTTETat, OVTE TO E1Ttuuf.L"'1TLKOll Kat uVf.LLKOll aEt 1TPOCT-
, , \. \. " , , ~ ... IJ"".,.' It ,,', , ,.. 

TaTTETat, al\./\.a Kat 1TpOCTTaTTEL 1TOTE, OTall n f.LOx.",pa "'1 'l"VX"'1 TOV 

allOpOmov (129a 10-16). 

Not only does the syntax here make use of the subordination of 
servant to master in opposing the AOYUT'TtKOV to the other two parts, 
but 128b37-39 distinguishes aPE'TT, from E7nCrrT,#LTJ because the for­
mer comes into being in several faculties, the latter only in the Aoyt­
CTTtKOV. Now, it is not clear what conception of the role of the desid­
erative elements in the soul underlies this opposition. We cannot 

39 Cj H. Cherniss, Aristotle's Criticism of Plato and the Academy I (Baltimore 1944) 
80f, and I. DUring, "Aristotle's Use of Examples in the Topics," in Aristotle on Dialec­
tic, ed. G. Owen (Oxford 1968) 202-29. 

40 Consider the definition of soul as OV(Tia E-rTUITT,ILTJ<; 8EKTLK7) (151bl-2; cj 140a35-
38), or Xenocrates' definition of soul as apdJ~<; aVToo; aliTOl' Ktl'Wl' (140b2-7; cf 
120b3-5, 123all-14, 23-26; De An. 404b30, 408b32-409alO, 409b4-18), on which 
see Cherniss (supra n.39) 10-19. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 299 

prove from 129a12-16 that the anonymous Academics simply re­
jected the Platonic status of the Ovj.LOEL8~s, as the author of the Mag­
na Moralia did, although the grouping of the ETn(JvWflTLKOV and 
(JVj.LLKOV in opposition to the AOYUTTLKOV reverses Plato's grouping 
(Resp. 44082, E4-6, 442A4-6~ c! n.44 infra) and may suggest that 
the (JV/ . .LO€LSec; is no longer viewed as the natural ally of the AOYUT'rt­
KOV. Perhaps 129a12-16 may be associated with the systematic assign­
ment of the desiderative elements to each of the tripartite parts at 
126a9-14, where /3ovA'Y'/cne; and alaxvv'Y'/ are located in the AOYUT'TL­
KOV (c! Resp. 571c9, 606c3-6).41 In that case, the version of biparti­
tion advanced at 129a12-16, since it involves a division of the desid­
erative elements among the soul-parts, would have a basis different 
from that of the Peripatetic interpretation of tripartition, which pre­
supposes Aristotle's doctrine of 0pEgLe;. 

In any event, there can be no doubt, from Socrates' arguments 
against Glaucon's tentative assignment of the (Jvj.LoEL8ee; to the E1nOv­
j.L'Y'/TLKOV (Resp. 439E 1-5), that the conception of tripartition as a 
bipartite dichotomy necessarily misrepresents Plato's moral psychol­
ogy. In establishing that the soul's conflicting desires require it to be 
divided into parts, Socrates argues that the phenomenon of thirst, 
when someone is simultaneously thirsty and reluctant to drink, re­
quires at least two parts, TO j.LEv q, Aoyi(,ETaL AoyurTLKOV 'TTpOuayopEv­
OVTEe; Tiie; tjJvxiie;, TO 8E q, €pq. TE Kat 'TTELvii Kat 8LtjJii Kat 'TTEpt Tae; 
aAAae; €mOvj.Lias €'TTTO'Y'/TaL aAOYUTTOV T€ Kat €m(JVj.L'Y'/TLKOV (43905-
8). This dichotomy indeed suggests a bipartite division, and Socrates' 
use of aAoYLuTOV to designate the faculty containing the Em(Jvpi,aL 
(43907) is a clear antecedent to Aristotle's aAoyov.42 But Socrates 
goes on to ask whether TO 8E 8T, TOV (JVj.LOV Kat q, Ovj.LOVj.LE(Ja is a 
third soul-part or has the same nature (oj.LOcpvee;) as either of the 
other two (439E2-4). When Glaucon tentatively associates it with the 
€m(JVj.L'Y'/TLKOV, Socrates objects with the story of Leontius, arguing 
that (Jv~e; is the natural ally of Aoyoe; (€'TTiKOVPOV DV T~ AOYLUTLK~ 

41 If the psychological examples in the Topics are related, it is possible that the four 
cardinal virtues are coordinated with the bipartite version of tripartition at 128b37-39 
and 129aI2-16, as in tripartition; only ¢POVyWV; and O"wcPp0mJVTj are explicitly as­
signed to the AO'YLO"TLKOV and E1TdJvJ.LTjTLKOV respectively (l38b 1-5, 136b 11-14; (1 
134a34, 145a29-32); but the phrase Kat TWV aAAWV apuwv OVT~ iKaO"TTj" AaW 
/3aVOJ.LEVTj" 036b13) suggests that the other virtues are to be assigned to soul-parts 
in the same way, and elsewhere the four cardinal virtues are enumerated (l08al-3, 
150a2-15). 

42 0: Resp. 441c2, 604D9; Leg. 863B4; Theophr. fr.L1l7; De! 415e7. The Platonic 
antecedents to the division of the soul in Eth.Nic. I 13 are discussed by Dirlmeier, 
Nikomachische Ethik (Berlin 1956) 293. 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

300 PERIPATETIC PSYCHOLOGY 

c/>VCTEI", 441A2-3), that in the divided soul it sets its arms on the side 
of the AO,),W"TtKOV (440A5-E6), and that it is a third part separate 
from the other two (440E 8-441 B 1) .43 The Bv/.WEI,,8et;; always allies 
itself with the Ao,),W"nKov against the E1nBv/-LTlTtKOV, like two parties 
engaged in stasis (8vcUv u7"auw~ov'TC.ov, 440B2) , and never makes 
common cause with the EmBv,."w.1" against the commands of AO,),Ot;; 
(440B4-c5); its natural function is to enforce the deliberations of the 
AO')'W"TtKOV and to ensure that, in point of ruling and being ruled, the 
E'1nBvJJ.TlTtKOV, like the other two parts, acts justly in accordance with 
the principle, 7"£1 aUrol) 7TPClr7"EtV.44 

Clearly, this conception of the Bv/.WEt8et;; as the natural ally of the 
AO,),W"TtKOV, with its own E7TI"(Jv,."w.1,, providing a source of motivation 
independent of that of the other two parts, precludes grouping it with 
the E1nBv/-LTlTtKOV and reducing tripartition to a bipartite dichotomy. 
The Peripatetics who represented the tripartite psychology of the 
Republic as a division into &AO')'OV and AO')'OV EXOV interpreted it in 
terms of Aristotle's doctrine of OPE~I"t;;. Although Aristotle does con­
sider it more disgraceful to yield to bnBv,."w. than to BvJJ.<>t;;, because 
the latter is at least responsive to AO,),Ot;;, he nonetheless deprives 
BvJJ.<>t;; of its status as an independent form of motivation, grouping it 
together with the other ope~Ett;; in the &AO,),OV.45 Attempting to recon­
cile tripartition to the framework of Aristotelian bipartition, the Peri­
patetics rejected the special status of the (Jv/.WEI,,8et;; by assimilating 
BvJJ.<>t;; to the other ope~Ett;; (cf. Mag. Mor. 1187b36-37, 1188a23-26) 
and equated the resulting bipartite dichotomy with Aristotle's division 
into &AO')'OV and AO')'OV EXOV. 

Thus Plato's tripartite psychology differs fundamentally from the 
division into &AO,),OV and AO')'OV EXOV attributed to him by the Peripa­
tetics.46 In order to represent tripartition as a bipartite dichotomy, 
they had to reject the independent status of the (Jv/.WEI,,8et;; together 

43 On the notoriously problematic status of 6v~<; see, most recently, J. M. Cooper, 
"Plato's Theory of Human Motivation," History of Philosophy Quarterly 1 (1984) 3-21; 
cl E. Brann, "The Music of the Republic," ArON 1 (1967) 41-46. 

44 CI 44OE4-6, 441 08-E 1, 442A4-c3, 44208-443B2, 443c9-444A2, 44408-11, 
586E4-587 A5, 589A6-B6; Leg. 64407-645cl; Tim. 70A2-7, 7002-6; Phdr. 246A6-
B4, 253c7-255A I, 255E4-256A6. When Socrates groups the three soul-parts into a 
dichotomy, he opposes the upper two parts to the bn6vIJ-T/nKO/l (cl 440B2, E4-6, 
442A4-6); only in the case of uwq,POuV"TI does he reverse the grouping (442cI0-oI), 
but the agreement by the other two parts that the Ao')'WnKo/l should rule in no way 
alters the natural function of the 6v,."oEWE<; to enforce the {joVhEV8EJI'ra (442B9) of the 
hO')'UTnKO/l (cl e.g. Leg. 645A5-BI). 

45 CI Eth.Nic. 1149a25-b26; Theophr. fr.L88; supra n.9. 
46 CI supra n.9 and the passages cited supra n.44 (add Resp. 602c-605c, Tim. 

4Ic-44c). 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

P. A. VANDER WAERDT 301 

with the distribution of desiderative elements among the three soul­
parts, and to interpret tripartition in the terms of Aristotle's analysis 
of OPE~t.~. Aristotle himself never represents tripartition as a bipartite 
dichotomy, and in the criticism of soul-division at De Anima 432a24-
b7 he clearly distinguishes the two divisions. It was not until a gener­
ation later, when the problems that had motivated the Academic 
debate on soul-division were no longer alive, that the early Peri­
patetics attempted to reconcile Platonic and Aristotelian psychology. 
By interpreting tripartition in the terms of Aristotelian bipartition 
they fundamentally misrepresented Plato's psychology; yet, through a 
complex history of transmission, this Peripatetic interpretation of 
tripartition dominated the doxography and philosophical speculation 
of later antiquity. Virtually all subsequent discussion of soul-division 
represents tripartition in the terms of Aristotelian bipartition. The 
brief doxographical notice at Magna Moralia 1182a24-25 is only the 
first record of a Peripatetic doctrine that shaped the interpretation of 
Plato's psychology for many centuries thereafter. 

ApPENDIX: PLATO IN THE MAGNA MORALIA 

Did the author of the Magna Moralia derive his knowledge of Plato from 
earlier doxography? We have seen (supra nn.36-37) that some of the details 
of I 182a24-30 are well explained by this hypothesis. If the author considered 
his attribution of bipartition to Plato original, he would surely have explained 
his interpretation of tripartition (as e.g. Aetius, Dox.Graec. 389.10-390.4); 
instead, as Dirlmeier (supra n.O 165 recognizes, he presents it as a com­
monplace, requiring no further elaboration. Moreover, if the author based 
his criticism of Socrates and Plato directly on the Republic, why should he 
ascribe bipartition to the latter and argue that the former did not recognize 
the MOYOV? (What source other than Plato's dialogues would the author 
have for Socrates' psychological doctrine?) This inconsistency suggests that 
the author is simply reproducing stock Peripatetic criticism, and indeed his 
manner of presentation suggests that his own usage of doctrine was charac­
teristic of contemporary discussions of soul-division. 

If the doxography of 1182alO-30 does not reproduce a written source 
verbatim, it certainly reflects the understanding of pre-Aristotelian moral 
philosophy current in the Peripatos. It is possible, moreover, that the auth­
or's attribution of bipartition to Plato derived from the Peripatetic work that 
recorded Platonic doctrine and was the source for his catalogue of aPETaL at 
1185b4-13 (supra n.32). In introducing bipartition, the author says W<; cpa~v 
(1185b4), which can only refer to his earlier attribution of bipartition to Plato 
at 1182a24f. The source of the author's catalogue of aPETaL at 1185b4-13 
must also have referred to the division into aAoyov and AOYOV EXOV (it is the 
basis for the classification of the ape-rai). Since 1182al0-30 probably derives 



WAERT, VANDER P. A., The Peripatetic Interpretation of Plato's Tripartite Psychology , Greek, 
Roman and Byzantine Studies, 26:3 (1985:Autumn) p.283 

302 PERIP ATETIC PSYCHOLOGY 

from earlier doxography, and since the Peripatetic catalogue recorded Pla­
tonic doctrine, the author's source for 1185b4-13, might also have contained 
a doxography of soul-division that attributed bipartition to Plato. This evi­
dence is suggestive, but hardly conclusive. 

In his only other reference to Plato, the author explicitly cites the Republic 
(1194a6) for an example of his doctrine of proportionate justice, but his 
discussion (1194a6-29) leaves it uncertain whether he knows the work at 
first hand. Although the author's enumeration of the 'YEWP'YO~, OiKO&)J.LO~, 
Vcfx:i:vTTJ~, and UKVTOTOJ.LO~ does reproduce the membership of the self-suffi­
cient ava'YKaWTaTTJ 1TOA.18 of Resp. 369c-370D, the avaA.o'YLa as an example 
of which the author cites this passage is not mentioned there at all, and his 
citation is marked as an interpretation by the analysis that begins EUTLV 8' -r} 
avaA.o'YLa aVTTJ (1194a 12). As Professor Cherniss has suggested to me, the 
author or his source appears to have taken as implicit in this passage of 
the Republic Aristotle's doctrine of proportionate justice (Eth.Nic. 1132b31-
1133b28). In fact, Aristotle himself uses the same examples as Plato (OlKO­
BOJ.LO~ and CTKVTO'T()J.LO~, 1133a7-10~ 'YEWP'YO~, 1133bl; oiov Kat EV rf11TO)u-

.... ,... ,~\ .... ~ ~ I , Il" , 'I:;' , .... 
TLKT/ T4J CTKVTOT0/-UY aVTL TWV V1TOV'IIUXTWV aJ.LOLiJTJ 'YLVETaL KaT a-;LaV, KaL T4J 
vc/xxvrfj Kat TOL~ A.OL1ToL~, 1163b33-1164al), and he explicitly criticizes 369c-
370D at Pol. 1291alO-22. Hence a Peripatetic might easily read this doctrine 
back into the Republic. (The report of oi IIv8a'YopEwL [1194a30] may be 
explained by the account of TO aVTL1TE1TOv8o~ at Eth.Nic. 1132b20-22, which 
the author follows with modifications') It must remain uncertain whether the 
author himself or some earlier Peripatetic introduced the illustration from 
the Republic. But it may well have been a stock Peripatetic example that the 
author took over from an earlier doxography, and certainly 1194a6-29 may 
not be used as evidence that the author was directly acquainted with the 
Republic. 

In both cases where the author mentions Plato, therefore, he may well be 
drawing upon earlier doxography. Not only is 1182a24f best explained by 
supposing that the author was reporting a Peripatetic commonplace, but the 
criticism of Platonic moral psychology that follows bears all the marks of 
stock Peripatetic doctrine.47 

PRINCETON UNIVERSITY 

September, 1985 

47 I am greatly indebted to Harold Cherniss for discussing this topic with me and for 
very helpful and detailed criticism of an earlier version. For their kind help I also wish 
to thank Albrecht Dihle, A. A. Long, A. E. Raubitschek, and Friedrich Solmsen.