"c-4 0_�19"5

ARISTOTLE
POETICS

INTRODUCTION, COMMENTARY AND

APPENDIXEs· BY
D. W. LUCAS .
ERCEVAL MAITLAND LAURENCE READER

. IN CLASSICS IN THE

UNIVERSITY OF,CA·MB RIDGE

FELLOW OF KIN·G
•
Iic
oLLEGE

OXFORD
AT THE CLARE�DON PRESS
 INTRODUCTION

I. ARISTOTLE'S WORKS

CICERO was acquainted with two types. of Aristotelian

writing: 'De summo autem bono quia duq genera librorum
sunt, unum populariter scriptum quod £gwTEptKov appella­
bant, alterum limatius quod in comrnentariis reliquerunt . . ' .

(De Finibus 5· 12). 'Commentarii' , which translates imo­

p,vqp,a-ra, can stand for anything from rough notes to such
sophisticated works as Caesar•s records of his campaigns.
The exoteric works, presumably the same as those referred
to as £K8£8of.L€vot Aoyot in the Poetics 54hrS, must be the class
whose fluent style is elsewhere praised by Cicero.1 These
'published' works are all lost, unless the l4 01Jvatwv IlolttTEta
recovered from a papyrus is to be reckoned among them.
Probably they were for the most part early works, many of
them dialogues, though less dramatic than those of Plato.2
Our Corpus Aristotelicum .consists of -works of the type
called by the early commentators, though not by Aristotle
himself,J aKpoap,anKa 'workS for listening to'. Jt is not
known for certain how these often jerky and discontinuous
discourses were actually used. They have beeri thought of
as lecture notes, either used by the lecturer or taken down by
the pupil, as sketches for proposed works, or as summaries
of works already completed, but it is· pretty generally
agreed that they formed part of· a course of oral instruction
and were not intended for wide circulation outside the school. 4 .

I 'Flumen orationis aureum fundens' (A cad. 2. II9); 'dicendi incredi­

bili quadam cum copia tum etiam ' suavitate' (Top. 1. 3).
a Ad Att. 13. 19..
3 The Letter to Alexander in which the word occurs (fr. 662) is not
authentic.
4 See de Montmollin, p. 343; W. D. Ross, Aristotles (London, 1949),
pp. 16 and 316; H. Jackson, ]Phil. 35 (192o), 191-200; Bonitz 104b44.

IX
 I N T RODUC T I O N
They vary much in degree of finish, and the Poetics is
among the least finished, being in parts little more than a'
series of j ottings.
It is something of a mystery why the more elaborate
works were driven out of circulation by the less finished
during the early centuries of the Roman Empire, and there- ·
after lost. A story is told in Strabo (668-9) and in Plutarch
(Sulla 26) that the worl<s used in Aristotle's school passeq
through Theophrastus into the possession of Neleus, who
__

hid them in a cave to keep them out of the hands of the

book-collecting kings of Pergamum. They were recovered
early in the first century B.c. and taken to Rome by Sulla, ·
where they eventqally received s.cholarly publication froin
Andronicus of Rhodes ; it is suggested that the effect of this
was to turn attention away from the philosophically inferior
exoteric works. This would imply that these were the only
copies of the esoteric works and that the essential Aristotle
was lost to the world for two �centuries. Scholars vary in·
their ability to believe this. 1
The characteristics of the surviving works have an im­
portant consequence. The Poetics, more than most, is dis­
jointed, full of interruptions, of digressions, and of failures
in connexion . . It is in the nature of notes to be disj ointed.
·
It is also in their nature that they should be revised,
supplemented, and supplied with alternatives, and if they
are the property of a school, they may be worked over by
different hands.2 Accordingly the interpreter of the Poetics

I There is no agreement whether or not the influence of the sup­

posedly lost works is to be found in the scanty. philosophic remains of
the period. Zeller, Phil. der Griech.3 (1879), ii. 2, ch. 3, maintained that
the Physics was known to Poseidopius and that traces of most of the
major works can be found. K. von Fritz in Entretiens Fond. Hardt, iv,
p. 86, asserts that Polybius did not know the Politics nor Euclid the
Analytics, which implies thatthey were not available. For the history
of the Peripatetics between Theophrastus and-Andronicus see C. 0.
Brink in RE Suppl. B. vii, especially 923 ff.
2 The extreme position is taken by F. Grayeff (Phronesis i. 105 ff.),

X
 I N T R O DUC T I O N

is perpetually confronted with an awkward choice. He can

explain an apparent failure of cohesion by saying that the
writer put down enough to indicate for his own use a certain
sequence of ideas, and that the connexion would be made
clear in a · spoken version embodying the necessary transi­
tional passages. On the other hand, by removing a phrase
or a sentence it is often possible to make a confused passage
logical and coherent, and the assumption that a marginal
addition has got into the text, or that alternative versions
have been c�mbined is not, given the apparent nature of
the work, implausible. Again, Aristotle appears at times
blatantly to contradict what he has said elsewhere. Should
we go to all lengths to resolve such inconsistencies, or allow
that two views may appear in notes which, not being in­
tended for posterity, were never finally adjusted? There is
no lack of sentences which can be made to appear intrusive,
and editors have made the discovery that, if much of the
book is )eft out, the rest becomes easier to explain. But
attempts to recover an original-Poetics by stripping off later
additions rest on the assumption, which may not be true,
that the original Poetics is still there. If what we have was
assembled from a larger collection of notes, parts of the
original can have been lost when alternative drafts were
combined.
The right course would seem to be to warn the reader
of the suspicions which may reasonably be entertained as to
the continuity of the existing text, and then to make eveiy
endeavour to find a meaning for it, resorting only as�a final
expedient to excisions or to the assumption that there is
a lacuna.

who says that all \ve have is 1} f3t{J)uo8�K7J .:4.p,a-ro-rE>.ovs Ka� 8t:ocf>p&.a-rov
Ka� -rwv p..:,.· av-rovs- 'in which as it stands there may not be a single
chapter . of purely Aristotelian origin'.

XI
 INT RODUCTION

II. THE LITER ARY WORKS AND THE SECOND

.
BOOK OF THE POETICS
I

Aristotle, unlike his master Plato, did not regard the

materia� world and ordinary, unphilosophic activities as
trivial. He tried to give a rational account not ortly of
rhetoric, but of poetry and music ; quotations in ancient
writers and other testimony show that he Wr-ote a number
of works on these subjects, though he mentioned painting
and sculpture only incidentally, while architecture was not
counted among the fine arts· in the ancient world. Among
the exoteric works was the dialogue named Gryllus after
Xenophon's .son who fell at the battle of Mantinea in 362
B.c. In this Aristotle raised the question, in opposition to
!socrates,· whether rhetoric was an art at all.1 Here the
· influence of. Plato was still strong. More important was the
dialogue in three books On Poets.2 It is impossible to re­
construct the work, but fragments show that it touched on
some of the same topics as the Poetics, no doubt in a more
leisurely manner, and it is likely that it is one of the sources
from which Aristotelian ideas on literature passed to a wider
audience.
The remaining literary works seem to have been intended
for use within the school. In addition to the Rhetoric in
three books, which is extant, there was a Texvwv Evvaywyr/,
a summary of rhetorical theories in two books, and a sum­
mary of the [echne, the Handbook, of Theodectes.. Nothing
i� known of the Ilepi MovutKfj�. On Music1 which may have

1 Quint. 2. 17. 14. 'The young A. scorned judicial oratory, the old A.

analyzed it' ; see G. Kennedy, The Art of Persuasion in Greece (London,

. 1¢)3), pp. 83-87
z See Rostagni, '11 dialogo a. llE:p' lloLTJTiAJV', Riv. Fil. N.S. iv. 433
·

and v. 145 (1926, 7), = Scritti Minori 1. 263. He assumes a closer re­
semblance to the Poetics than there is warrant for.
It is noteworthy that the dialogue form was used also by the Peri­
patetic Satyrus for his work on the lives of the tragic poets, POxy. ix.
1176, ed. G. Arrighetti (Pisa, 1964).
Xll
 INTRODUCTION

dealt with the mathematical aspects of the subj ect. There ·

were six books of 14:rrop�JLa-ra 'OJLTJP"'((l, Homeric Problems, of
which the Poetics provides a sample in Ch. 25 ; it may have
developed from the edition of th_e_ lliad.-which Aristotle is
said to have made for the young Alexander. The single
book of llot."'}Tt.Ka appears from its position· in the listi of
Aristotle's works to have been concerned with similar 'prob­
lems' in other poets. Finally there is a group of works based
on researches in records and archives. A few years before
331. B.c. Aristotle compiled, with the help of his nephew
Callisthenes, a list of victors at the Pythian· festivals which
is the subject of an existing inscription (Dittenberger, Syl­
loge, i. 275) .. He compiled siruilar lists of the victors at the
Dionysiac festivals at Athens and of the plays which were
produced on each occasion (Didascaliae) . It is usually sup­
posed that they were earlier than the Poetics, and that the
knowledge of fragedy there displayed was based in part on
researches carried out in connexion with tbese lis.ts (see
Pickard-Cambridge, Festivals, pp. 103-26).
Our Poetics is almost certainly the work listed as flpayJLa­
-rEta T€xirrJs flotTJTLKfjs, Treatment of the A rt of Poetry,· 2
books. 2 It follows directly after works on rhetoric in the
I The lists of A.'s works, of which the most important is that given
by Diogenes Laertius (5. 22-27}, are printed at the beginning of Rose's
ed. of the Fragments, and they are discussed by P. Moraux, LesListes
anciennes des oufirages d' A. (Louva:in, 1951): see the same author's A. et
son ecole (Paris, 1962), pp. 279--So. The source of Diogenes' list, which
begins with the exoteric works and . groups the rest by subjects,
is generally believed, to be the Peripatetic Hermippus of Smyrna,
who wa� associated with the Alexandrian Library, and presumably
composed his list before the disappearance, if they did disappear, of
Aristotle's esoteric works. Moraux himself attributes it to a later
Peripatetic, Ariston of Chios.
z This is the only work in the list described as a 1rpayp.aula, a word
which A. often used with reference to his inquiries, e.g. Soph. El. 183b4.
Gudeman and de Montmollin consider our Poetics too unfinished for
even such limited circulation as was intended for the esoteric works.
The existence of'a second book was denied by A. P. McMahon, HSCP
28 (1917}, 1-:-46.
Xlll
 I N T R O DUCT I O N
ancient lists. Even without this external evidence there
would be reason to believe that the Poetics consisted of two
books or, since the division into books need not be Aris­
totle's, that a substantial portion of the work is missing.
The scheme clearly implies a section on comedy to balance
that on tragedy, and it is specifically promised at 49h21.
Further, it is pro!llised. at Politics I34Ib38 that a full account
of katharsis will be given €v 'T'ois 1rep� 7TOt7JnKfjs 'in the work
on poetry', and this could appropriately be part of the
comedy section. 1
The .existence of a . second book is supported also by the
subscriptio of William of Moerbeke's Latin version. com­
pleted in 1278 'Primus Aristotelis de arte poetka liber
explicit'. 2 Similarly the reference to the Margites at 48h3o is
cited by Eustratius (c. A.D. noo) on Ethics II4Iai4 as oc­
curring in -r� 1rpwnp 1rept 7TOt7J7'tl<fjs, the first book. Even if
those are right who think our Poetics too rough and in­
coherent to be the finished version of Aristotle's work,
what we have must still be the draft of a part only of the
original (see p. xiii, n. 2).

Ill. ARISTOTLE AND HIS PREDECESSORS

When Aristotle wrote the Rhetoric he was only doing more

comprehensively and scientifically what others had done
before him, but in writing the Poetics it can be said with fair
· I Vahlen argued that the words of Proclus on Pl. Rep. (r, p. 49
Kroll, see p. 52) refer to a discussion of katharsis applying both to
tragedy and comedy, and that it is to be assumed that this was part
of the missing book, Gesammelte philolog. Sc.hrijten, i. 233.
Other refs. to the Poetics, Tois 11'Ep� 71'0'TJT'Kfjs, in A. are all . from the
Rhetoric, 1372a�, 1404a38, 1404b7, I405a5, 14i9b5; the first and last of
these refer to a definition of the laughable and its various kinds (Ei�TJ),
which must have been more extensive than our 49a32-37, and some of
· the ot her passages referred to are suspiciously brief. Rhet. 1404b28
cites Tois 11'Ep� 71'o,�aEws. The tenses used imply an order P.Q_litics,
Poetics, Rhetoric, but such evidence is not conclusive.
z Aristoteles Latinus xxxiii {Bruges/Paris, 1953).
XlV
 I NTRODlfC T I O N
confidence that he broke new ground even though there
was some overlap with his more popular dialogue On Poets.1
There is no trace of any previous attempt to lay down the
principles and to guide the practice of poetic composition ·

in the way that rhetorical handbooks indicated the prin­

ciples of that art.2 Yet given the immense · importance of
poetry in Greek life, both in education and in public festi­
vals, it must inevitably have been the subject of frequent
discussion. This the Poetics would show even if there were
not other evidence.J References to arguments about the
plate where tragedy and comedy originated (48aJo), to con­
flicting views on the primacy of plot or character (soaiS-38),
'
on the merits of single and double plots (SJaiJ), on the
proper way of eliding a tragedy (53a24), and the mention of
the severity shown by contemporary critics (56a5) suggest
a plentiful expression of opinion on literary subjects, though

we have no means of telling how much was spoken and how

I Soph. El. 18489 : 7T£pl p.�v -rwv p'rJ-ropucwv V7T�PX£ 1ToAAa �eal 1raAa'a
-ra A£yop.£va, · 'On rhetoric (�s opposed to logic) much had been said,
some of it long ago'.
On the fl£pi flo,7J-rwv see p. xii.
:z Poetry had long been recognized as a techne; cf. Aristoph., Ran.

939 (of tragedy) : d;\,\' ws 1rapl.\af3ov -r�v -rlxvTJv 1rapa aov. But as with
wo�epuns-, the art of delivery, though it was recognized, ov1rw �£ avy- ·
K£mi' TEXVTJ 1T£pl av-rwv (R. I403b35)·
3 Gudeman, in an attempt to correct what he conceives to be the
over-emphasis on Plato as a source for the Poetics, gives on p. 10 of his
Introduction (and in English in Class. Studies in Honour of]. C. Rolfe,
pp. 75-100) a long list of works with literary titles, not all of which need
have existed when A. wrote. W. Kranz, Stasimon (Berlin, 1933), pp.
4-7, lays emphasis on Hellanicus among his predecessors, but there is
nothing to suggest that there was a serious critical literature before A.
Cf. L. G. Breitholtz, Die dorische Farce (Goteborg, 196o), pp. 35-40.
On the critica1 notions that can be · extracted from earlier Greek
literature see E. E. Sikes, The Greek View of Poetry (London, 1931),
chs. I and 2; J. W. H. Atkins, Literary Criticism in Antiquity (Cam­
bridge, 1934, ch. z; G. M. A. Grube, The Greek and Roman Critics (Lon­
don, 1965), chs. 1-3. The chief texts are collected in G. Lanata, Poetica
Preplatonica (Florence, 1963). '

XV
 INTRODUCTION

much written. ·The sections· on style and grammar (Chs.

19-22) must owe something to the handbooks on rhetoric
_and to the linguistic speculations of the sophists, among
whom Protagoras is mentioned by name at 56h15. Chapter
25, on the solution of the problems presented by poetry,
clearly has a long ancestry in the difficulties, both literary
and ethical, presented above � by Homer's works. And in
the discussion whether epic or tragedy is superior Aristotle
obviously has in mind the expression of a view the opposite
of his own,· perhaps .Plato's. Plato indeed may have pro­
vided him with the �tarting point for his treatment of
several important topics.
In any society where poetry and the arts are important
some ideas about their nature and function begin sooner or
later to circulate. In the world which Homer describes the
subject-matter of the poet is heroic deeds and tales of the
gods; by telling them he casts upon his listeners a spell of
delight. 1 His poetic gift is divine and comes to him from the
Muses, but it is a craft that he possesses, not a fitful in­
spiration.2 Hesiod too received his powers from the Muses,
the daughters of Memory ;J and he was the first to raise the
question of the truth and falsehood in the poet's message.4
Archilochus knew of the fierce inspiration of those in the ·
grip of a power outside themselves, and Pindar proclaimed
the inferiority of acquired skill to native genius.s The
spread of education and the ability to read, which is pre­
supposed by the demand for written laws, caused men to
reflect· on the educational value of what they read-mainly
Homer. Though his. tone was pure, and the aristocratic
ideal continued to be based on the heroic standards which
he enshrined, there was room for much offence in his gods.

1 IC7JA7J8p.&s, Od. n. 334; Ol>..e,s, Od. 17. 518-2I.

·

z a'aov '�
�I� a�· 7Janav � , Od 8 64; 8£OS
ao'a7JV1 \ a£
� I fLO' £V .,p£cnv
• ..l.. -
\ o'p.as 1TaV1'o,as
I .

3 Theog. 29-32.
� '
• •

£vlt/>vaEv, Od. 22. 347· 4 Theog. 27.

s Archil. fr. 77, see p. So ; Pindar, Ol. 2. 86, 9· 100. See Sir Maurice
Bowra, Pindar (Oxford, 1¢4), ch. I.

XVI
 INTRODUCTION

Some-Xenophanes, Pythagoras, Heracleitus-denounced

the poet who so portrayed the divine, others took refuge in
explanations which assumed a hidden and deeper meaning.
The first . was probably Theagenes of Rhegium about the
end of the sixth century.1
A no -less lively stimulus to discussion must have come
from the institution of literary contests at Greek festivals, ·
especially the tragic contests at Athens. Contests are de-
. cided by comparison of one work witn another, and such
. comparisons lead naturally to the development of a critical
VQcabulary and to the establishment, even if unconscious,
of critical standards. As the verdicts of the .judges were of
interest to all, the growth of a . critical attitude was rapid, as ·

is · clearly shown by the frequency and quality of literary

allusions in comedy.2 The contest of Aeschylus and Euri­
pidesin the Frogs of Aristophanes is a remarkable piece of
impressionistic criticism which achieves all that is possible
in a medium which forbids sustained seriousness. But in
treating a question of principle, the . poet's purpose in writ­
ing, Aristophanes is less happy. In a sense it is true that any
poet who takes his own work seriously hopes that he will
'make men better citizens'.J If he believes that he has some­
thing to say, he believes that men will be the better for
hearing it, even .if he does not. aim at inculcating specific
virtues. Probably Sophocles, for instance, was conscious of ..
working within a framework of values which many, to the

I In Pl. Ion 530 c, D Metrodorus of Lampsacus and Stesimbrotus of

Thasos are mentioned as the leading expounders of Homer. Gods
might be identified with the elements or with human· faculties, the
method by which Theagenes explained away the battle of t�e gods in ·
Il. 20, or a hidden meaning might be found in a frivolous story; as
Socrates jestingly interprets the story of Circe (Xenophon, Memorab.
I; 3· 7). Plato did not approve of such interpretation by wovo£a (Rep.
378D).
2 In addition to Acharnians, Thesmophoriazusae, and Frogs there
were many comedies, now lost, in which literary themes were· pro-
minent. 3 Ran. 1009.

XVll
 INTRODUCTION

detriment of the city, did not. accept. But when Aristo­

.phanes makes Aeschylus justify the poets on the ground ·
that they convey useful information on curing diseases or
drawing up an army, he puts him in the same ridiculous
position as the Ion of Plato's dia�ogue who claimed to have.
acquired from his familiarity with. Homer a knowledge of
generalship. 1 Whether or not Aristophanes is wholly serious
here, Plato's reductio ad absurdum would lack point unless
such claims were actually made on behalf of poets. Those
who use poets for education may easily come to assume that
. poets write to educate, and centuries la.ter Plutarch, in his
De Audiendis Poetis (M. 14 E ff.), often argues as though
this were the case.
Some have believed that the literary contest in the Frogs
presupposes a society in which literary critiCism was widely
practised. It is clear that there was lively inter�st and
debate, but there is no evidence that it did more than touch
the fringe of the subject.z The growth of rhetoric as a con-. ·

scious art in the second half of the fifth century directed

attention to words and to the formal structure of sentences.
The sophists Prodicus, Protagoras, and Hippias, and, among
philosophers, Democritus are known to have been interested
in this kind of investigation. An awareness of the impor�
tance in rhetoric of arrangement and transitions might ·

well awaken interest in the way poets handled similar prob­

lems.3 Above all Gorgias, who defined rhetoric as 'the art of
1 Ran. io3o-6, Ion 541 A.
z M. Pohlenz in an influential article on 'The Beginnings of Greek
Poetic', Naih. Gott. G. (1920), 142 ff. asserted the need to assume the
existence of a body of critical theory behind the Frogs. The founder of ·

this he discovered, following Suss, Ethos (Leipzig, 1910), pp. 49 ff., in

Gorgias. ·His reconstruction is of what could have happened rather than
of what there is reason to suppose did happen; cf. 0. Immisch ed. of
Gorgias' Helen (Kleine Texte, Berlin, 1927), pp. 28-3o. Refs. to further
criticisms of the article in Radermacher, Aristoph. Froschez (Vienna,
1954), p. 368. ·
3 Protagoras seems to have suggested that the battle between
Achilles and the river Xanthus, Jl. 21. 211-384, provided a transition

XVlll
 INTRODUCTION

persuasion', asserted, the' power of words, with or without

metre, to stir the emotions and control the mind, producing
apate, deceit. It could be that Paris' eloquence deceived
Helen, in ·which case she deserves no blame for not resisting
him. Tragedy too is a source of apate, but this is a justifiable
deceit and those who succumb to it are wiser than those
who do not, meaning, perhaps, that · the audience must
co-opera,te by accepting the conventions of drama if it is to
enjoy it.1 · There is no necessity to link this apate through
word and persuasion with the .apate in the sense of ·'illusion'
produced by the artist with his pigments, which amounts to
deceit only in quite exception:;tl (:ases when a viewer is
.tempted to take a picture for the reality. But .apate was
used in connexion with the visual arts as early as Em..:
pedocles, fr. 23. 9, and the two were brought together round
aboutAoo B:C. in the sophistic treatise on The :Two Argu­
ments2 where it is said that the best poet and the best artist
is the one who most deceives by producing things like the
truth. Here, as in the epigram of Gorgias, the word is
· used with a conscious aim at paradox. The idea that poet
and artist or sculptor are doing essentially the same thing
was perhaps first expressed by Simonides in his celebrated
·comparison (see p. 269). The idea could be developed with
reference to the process in terms of mimesis, or with reference
to the effects in terms of apate,J but it is impossible to trace
this development, as the few relevant statements cannot be
dated with any precision.

between the battle of mortals, Greeks and Trojans, and the battle of
gods,'385-513, see fr. of schol. on 21. 240, POxy. ii. 68=Protag. fr. A 3o�
I GorgiaS, Hel 8 : €L' OE . 1.
. \,
"'' 1\0')'0S \ '
• 0• 1TEtOaS
, Kat\ TT}V "t'VXTJV I
a1TaTT}OaS
\
• • • •

I
Fr. B. 23 (Plut. M. 348 c) : a1TaTT}V, cfJs r. 1/JTJalv, �v 0 TE a1Ta,:�aas
� """ ' �
otKatOTEpOS TOV f'T}\ a1TaTT}OaVTOS I Kat 0 a1TaTTJ 8'' ,J.. I
E&S OO"t'WTEpOS TOV""' f'TJ\
t 1

a1TaTT}8,VTOS. .
,. .J.I
2 A
Lltaaot\ A oyot I. 3· 10 : EV ' \ . !.' ,
Kat\ .,wypa"t'tq.
\ ,.. '1:
I
yap Tpaytpoo1Touq. oaTtS <Kat')
•

... .... _!'\ ... A ..

1TI\E&OTa Esa1TaTq. Op.Ota TOtS WI.T}8WOtS 1TOtWV1 OVTOS aptOTOS,
• •

3 This term, common iii Plato, is· conspicuously absent from the
Poetics. There is some approach to it at 6o813, 61b11.

XIX
 I N T R ODUC T I ON
It would be more possible to decide whether Gorgias was
·
productive only of _epigrams or of a serious theory · if we
knew more of two works, both probably of _the late fifth
century, the 'On Poetry' of Hippias and of Democritus. We
know from Plato (Hippias �fa. 285 B-E, Hippias Mi. 368 B­
n) that Hippias concerned himself with ·words and rhythms,
but we are told nothing more. ' Titles of worRs by Dem6-
critus preserved in Diogenes Laertius (9. 48) show similar
interests, but he is famous principally for his insistence that
poetry is the result of inspiration, 'excludit sanos Helicone
poetas' (Hor. Ars Poet. 295), a surprising belief to be held by
a materialist (but see commentary on 55332). This, how­
ever, tells us nothing about Democritus' view on the nature
of poetry as mimesis or apate, though one would suppose
that a work of this period on poetry would deal with the
problem.
Finally there was one author of the late fifth century who
seems to have been an historian of literature in something
like the modern sense of the term, Glaucus of Rhegium,
whom Aristotle may well have used. Fragments of his
work On the Ancient Poets the title need not be his own­
-

deal with the early development of Greek lyric and in­

clude a valuable scrap of information about Aeschylus'
Persae.1
Whatever Aristotle inay have owed to fifth-century specu­
lations, there can be no doubt about the influence exercised
· on him by Plato, whose pupil and follower he was from the
time when he came to Ath_ens at the age of seventeen until
Plato's death twenty years later. Plato has much to say
about poetry. In the Ion and, more eloquently, in the
Phaedrus2 he describes poetry as divinely inspired, 'the
madness of the Muses' ; but the compliment is two-edged,
I 11£pl Twv J1pxalwv 11o£'T}Twv ",cal MovatKwv. Fragments in F.Hist.G.
ii. 23. Damastes of Sigeum, pupil of Hellanicus, wrote 11£pl IIot'T}Twv
Kal Eot/JtaTwv.
z Ion 534 B; Phaedr. 245 A.
XX
 INTRODUCTION

since the theory of inspiration is used to explain the in­

ability of poets to give a rational account of what they have
said. Only in the Symposium (209 A-D) is there a faint hint
that po�ts may have a glimpse of ideal truth. In the early
part of the Republic (376 c-402 A) he: examines the value of
the poets, espeCially Homer, for primary education, and
finds the moral standards expressed and implied generally
unacceptable. He had already in the Protagoras (347 c) re­
j ected the use, popular with the sophists, of poems as the
starting-point for discussions on moral questions among
adults. It is in the last book of the Republic that Plato
delivers his main attack on the arts,1 using his Theory of
Ideas to show that artists and poets are gmlty of the most
dangerous of ·an deceptions,· representing appearance as
reality. For if ideas alone are real and the world known to
the senses is only a shadow of the ideas, then the arts yield
the shadow of a shadow at the third remove from truth.
A subsidiary argument (6o5 B-Do7 A.) shows that the emo­
tions aroused by poetry are as deleterious as its . moral
standards, and encourage weakness rather than self-control.
Accordingly poetry is rejected as neither revealing truth nor
helping the temperance of the emotions. To the second-best
state described in the Laws, the work of Plato's old age,
poets are indeed admitted, though only under the super­
vision of those who have knowledge of good and evil
(658 E--66x D). Almost all existing poetry is condemned, and
it is unlikely that any of the poets banished from the
Republic would have cared to accept the terms offered for
aqmission to the city of the Laws.
In private life Plato's attitude to literature seems to
. have been more genial ; he knew the poems of Homer prac­
tically by heart, and in . his works never tired of quoting
them ; he admired Sophron to whose art he probably owed ·
much,2 and if it is true that he sent a disciple to Colophon to
1 595 A-6or B. See Appendix · I.
2 Diog. Laert. 3· r8. See commentary on 47bro .

XXl
.
 INTRODUCTION

collect the poems of Antimadms,1 his interest in poetry was

not unduly narrow: ·
Aristotle, in the Poetics, has little to say of genius or
inspiration,2 nor is he concerned about the religious or moral
implications of the myths,J since he clearly does not expect
them much to affect educated a,dults. As he did not accept
the Theory of Ideas and the status attributed in it to ·the
material world, he was under no necessity to controvert
Plato's account of the artist as a mere imitator ; but in
showing how the poet can reveal significance by generalizing
and universalizing he may have meant to show up· the
inadequacy of Plato's view. It is more certain that the
theory .of emotional purgation is an answer to Plato's com­
plaint that drama encouraged the dominance of the emo­
tions.• Finally, though Aristotle was not interested in . the
educational effects of drama; he may have been influenced
by Plato's strictures on myths which show virtue defeated
and vice triumphant when he laid down his requirement
that the tragic sufferer should not b� a character of un­
blemished excellence.

IV. THE TEXT AND ITS TRANSMISSIONs

Neither: before nor after the alleged loss of Aristotle's esoteric

writings does the Poetics seem · to ·have been widely tead.
Throughout the last three centudes B.c. there was a con­
siderable output of critical literature from the PeripatetiC

1 Proclus, In Timaeum i. 90· ..

z Mentioned only inCh. 17, where it is probably to be understood in

terms of the physiology of the four humours. See commentary arid
Appendix IL
J Cf. 6ob36 where Xenophanes' objections are dismiss�d with a shrug
of the shoulders.
4 See the passage of Proclus' Commentary on the Republic printed
on p. 52. ·

s For a fuller account of the sources of the text see the Latin Intro­
duction toR. Kassel's Oxford ClassiCal Text.
..

XXll
 INTRODUCTION

school based on the works of Aristotle and Theophrastus.

Quotations which we meet as examples first in Poetics or
Rhetoric frequently recur in later literature,1 but there is
no_ passage earlier than the fourth century A.D. of which it
. can be asserted with confidence that it is derived directly
from the Poetics.2
The Poetics seems never to have been the sv:bject of
a Commentary. But it was certainly known in Byzantium,
and it was translated into Syriac probably at the end of the
ninth century A.D. The Syriac version is lost except for
part of Ch. 6, but a few years later the Syriac was done into
Arabic by Abu Bisr (d. 94o), and this translation, which has
survived almost entire, is the earliest witness to the Greek
text, though a halting one. For not only is it at two removes
from the Greek, but it is accessible to most of us only in
a Latin translation. Further, both Syriac and Arabic trans­
lators were at ·the disadvantage· of scarcely knowing what
a tragedy was.J The first complete Latin rendering of the
Arabic was given by D . S. Margoliouth, Professor of Arabic
at Oxford, in his edition of the Poetics (Oxford, 19n); but
this has been superseded by the version of]. Tkatsch pub­
lished posthumously at Vienna, vol. i, 1928, vol. ii; 1932; the
translation is accompanied by a rambling commentary
usable only with the aid of the index.
· Probably within a generation or. two of the translation
I See G. Else, The Origin and Early Form oJGk. Trag. (Harvard and
London, r¢6), p. 113, n. 52; F. Solmsen in Hermes, 66 (1931), 241-67.
A striking ·example is the passage from the Rainer Papyrus given'
on p. 159.
2 The earliest are: Themistius Or. 27, 337 B. froin 49b6, which is in

part a later insertion ; the story of Mitys (52a8) appears in De Mirab.

Auscultat� 846a; sentences from ch. 20 on grammar are quoted in the
Commentaries of Ammonius and of Boethius on the De Interpretatione
(see Bywater on 56b2o).
3 Else gives · as an example of the errors . to which such a work is
prone the Latin version of 51b21 qui ponit, where the Syriac translator
read .l4ya8w�os .l4v8E; as l4.ya8wv 8s av 8jj� which the Arabic necessarily
followed.

XXlll
 INTRODUCTION

into Arabic was copied the best and oldest surviving Greek
manuscript, Parisinus 1741, called A or Ac by editors. This
manuscript was still in Constantinople in 1427, but reached
Florence before the end of the century and found a final
home in Paris. Its outstanding value was not recognized
till the nineteenth century. J. Vahlen, who gave a full
account of its readings in his editions of 1874 and 1885,
regarded it as the sole authority from which the text of
the Poetics is derived. .
It was from a closely related manuscript that William de
Moerbeke, who translated much of Aristotle, made his
Latin version in 1278; this survives in two manuscripts, but
they lay unrecognized until 1930. The Latin is occasionally
of service in establishing the reading of A.1
Since Vahlen's day it has been recognized that ther� is
one manuscript which preserves a tradition independent of
A; this is Riccardianus 46 (B orR to editors), which, though
of the fourteenth century, is the second oldest manuscript.
Attention was first called to it by F. Susemihl in 1878, and
some of its readings were published by G. Vitelli in Stud.
ital. di fil. in 1894 and by C. Landi in the following year.
They were given more fully in the apparatus to the edition
of Margoliouth, who tised the evidence of Ch. 16, where
Riccardianus alone has the words that fill a previously
unrecognized lacuna, to prove that it is independent of A,
(see commentary on 55a14). Though Riccardianus has no
descendants, a few of its readings found their way· into
Renaissance manuscripts (see apparatus, p. 3).
The numerous manuscripts of the fifteenth and sixteenth
centuries are all dependent on A, and their readings, though
occasionally of interest, have the authority only of anony­
mous emendations. See E. Lobel, 'The Greek Manuscripts
of Aristotle's ·Poetics', Supplement to the Bibliographical
Society Transactions, no. 9, 1933.
1 See Aristoteles Latinus xxxiii, ed. Minio-Paluello (BrugesfParis,
1953)·
xxiv
 INTRODUCTION

The Poetics became available to the western world for

the first time when Giorgio Valla published a Latin transla­
tion made from a copy of A, Esten�is 1oo gr., in 1498; the
older translation was unknown, and a Latin version of
Averroes' Arabic commentary printed at Venice in 1481 cast
but a fitful light on the subject. The first printed text ap­
peared in the Aldine edition not of Aristotle, but of the
Rhetores Graeci, in 1508. The Poetics was edited probabiy by
John Lascaris, who used an inferior copy of A, Par. 2038, or
a closely allied manuscript; This remained the basis of the
�ext for over three centuries until the superiority of A was
recognized by Vahlen. And it is only, comparatively speak­
ing, within the last few years that the other evidence for the
constitution of a text, the medieval Latin version, the
readings of Riccardianus, and the Arabic translation have
been made available in a form as complete as is likely to
be achieved.
The most recent development, based on the study of the
growth of Aristotle's doctrines and on fresh consideration of
the nature and function of Aristotle's esoteric works, is a
growing suspic;ion 'of the homogeneity of the Poetics. Such
suspicions are not new, and early editors transposed pas­
sages freely; but since F. Solmsen published his article on
'The Origins and Method of Aristotle's Poetics' in C.Q. 29
(1935), 192-201, more systematic attempts have been made
to remove incoherencies and inconsistencies by distinguish­
ing different layers of composition. 1 The scope for disagree­
ment here is certainly not less than in more usual forms of
textual criticism.
1 See Introduction t

XXV
API�TOTEAOY�

·. IIEPI IIOIHTIKH�
 SIGLA·

A= cod. Parisinus I 74I, saec. x/xi

Arec= m(anus) rec(entior) quae hunc codicem correxit

B= cod. Riccardianus 46, saec. XIV

Lat= translatio latina a Guilelmo de Moerbeke a. I178 confecta

Lat 0= translationis latinae cod. Etonensis, ca. a. IJOO
Lat T= translationis latinae cod. Toletanus, ca. a. 1180

Ar= translatio arabica ab Abii Bisr saec. x ad syriacum e�e�plar

(saec. IX ?) confecta, e versione latina Jaroslai Tkatsch

fr Syr= fragmentum translationis syriacae, apud Tkatsch I I 55

rec =codices graeci saec. xv-xvi, quorum lectiones propriae

coniecturarum loco habentur

tP= Guilelmi codex graecus deperditus

E= Syri codex graecus deperditus

ll={!
8=
{�
A=
{�
A PI�TOTEA O Y� ll E PI ll O I H TI K H �

ll£pt 1TOL1JTLKijS arhfjs T£ Kat TWV £l�wv avTijs. ijv TLVa 14478
� · " ;, ' A� �I '
ovvap.tv £KaUTOV £X£L, Kat 1TWS O£L UVVLUTaU a� TOVS f.LV OVS
(J '(}
1\ \ \ "t.
£t• f.LEI\1\£1. OE £K 1TOUWV Kat' 1 0
t I ,, • '
Kai\WS E�£LV 1J 1TOL7JULS1 ETL � '

f "\ ·\
A

I I .. �\ ""' fl ""'
1TOtWV £UTL\ p.opt.WV1
I
Of.LOtWS ,
0£ Kat\ 1T£pt\ TWV ai\1\WV oaa T'YJS
avTijS EUTL p.£8o�ov, AEYWf.L£V ap�ap.oot KaTa cpvatv 1Tpw� -

TOV airo TWV 1TpWTWV. E1T07roda �� Kat � TijS Tpaycp�las

1TOl1Jats en �€ Kwp.cp8la Kat � �t8vpap.f1o1rot"f}TLK� Kat Tijs
avA7JTLKijS � 1TA£lUT7J Kat KtfJaptanKfjS 1raaat rvyxavovatv 1 5
o �aat. p.tp.�a£tS TO avvo>.ov· 8tacp€povat 8€ d>.>.�>.wv rptalv.
� yap rip ev £r€pots p.tp.£ia8at � rip €upa � rep lr€-
I tl I
pwS Kat' f.L'YJ\ TOV \ t \
av•rov Tp01TOV . WU1T£p yap \
Kat\ xpwp.aUL
Kat ax�p.aat 1TOAAa p.tp.ovvraf. TLV£S a1T£LK�'OVT£S ( oi f.LEV
�La TEXV7JS ol 8€ �La UVV7]8£lM). £upot �E 8ta rijs cpwvijs. 20
" ) ""' · I I I tr \ ""'
OVTW KaV TatS Etp7jp.£VaLS T£XVaLS a1TaaaL ftEV 1TOLOVVTaL
\ I f �t
0 .
\ f. (J..,.
I
TTJV f.LLf.L1JULV £V)
pv f.LC[J Kat\ 1\0YCfJ Kat\ app.ovtg., I I
TOVTOLS
� xwpts � f.L€p.typ.€vow olov app.ov.'g. f.LEV Kat pvfJp.ip xpw-
1 '\ ' f ll
'YJ Kat '?} Ktr:apLUTLK1J KaV £C. "TLV£S
. ' \ 't\ ,
ft£Vat f.LOVOV 1J T£ aVI\"fJTLK
tl

lupat Tvyxavwatv o�aat r� tavrat T�v � vvap.tv olov � Twv 2 5 ,

avplyywv, avrcp �E Tip pvfJp.0 [p.tp.ovvrat) xwptS app.ovlas �

Twv opx1JaTwv (Kat yap o�rot �La rwv ax1Jp.art,op.lvwv pvfJp.wv
A t�\ [ t I ]
f.LLf.LOVVTaL KaL\ 1J"(}1J KaL\ 1Ta . I (}
1J KaL\ 1rpa�£LS
I t.
) 1J OE· £1T01TOUa
p.ovov TO iS' >.oyotS' rptAoiS" (Kat) � roi� p.lrpotS' Kat rovr�t.S' £LT£
.
Usque ad - 1448a29 deest B, ex quo codice nondum mutilo fortasse
una alteraque bonarum lectionum quas in hac parte praebet Parisinus
2038 (saec. xvfxvi) fluxit, cf. Lobel, The Greek MSS. pp. 27-3 1
1447•12 Mywp.£v tl> : Myop.£v A 17 f.v Forchhammer : ylv£& II, quod
non vertit Ar (cf. 145 1 a 17) 2 1 Kuv Paris. 2038 : Kat II 25 TU)'-
xavwatv A : -OVU&V rec (prob. Vahlen) oJaat TO&aVTa& Paris. 2038 (Ar) :
habetltes talem Lat : oJaa& A 26 fltp.ovv-rat II (l-''l-'£i-rat Paris. 2038) :
del. Spengel (non habuisse videtur I) � Paris. 2038 (Ar) : ol II 28 f..,.o-
'Tt'oda II: del. Ueberwcg (non vertit Ar) 29 Kai � Lobel : � n (Ar). .
idem mendum in codd. 1\Ietaph. 992b32

3
 1447b A P I I: T O T EA O T :E
17b p.t.yviJaa p.E'T' 0.,\,\�,\wv EiO' ivl 'T£J)£ ylvEL xpwp.br'l TWV p.€-
Tpwv avwvvp.ot Tvyxa�·ovat p.lxpt .TOV VVJJ" ovS€v yap av
\ \
10 EXOtJLEV ovop.aaat
t1 1 I
Kotvov rovs £..'t"wI.,{�povos Kat\' .!:!JEvapxov I
p.t-I
' \ '(' \ ,\ 1 ·�\ � \
,_.

El. ns ota rptp.Erpwv,

II I
p.ovs Kat rovs £.. WKpanKovs oyovs ovoE
1} £,\EyElwv 7} TWV aAAwv TLVWV TWV TOt01lTWV 7TOtOLTO T�V
I ' r Jl lJ I I \
p.tJL7JULV. 7T,\7JV OL avupw1Tot YE avva1TTOVTES TCf'""J f'ETPlfJ
I
TO
\ \ � · \ •
1TOLEtv E',\EYEL07TOLQVS' TOVS' oE E7T07TOLOVS' ' , . Y
ovop.a�ovatv,'
I , WS'
ovx
\ \ \ ' . a KOtVrJ KaTa\ TO\ JLETpOV
A

I \ "" I
15 K,ar"a T1}V p.tJL1}UtV 1TOU]TUS' a,\,\ 1TpOU.;.
, \ \ \ J.. � \ -
·

� 1 "' I
ayopEVOJITES'' Kat yap a.JI . taTptKOJI 1J 'fJVUtKOJI TL ota TWJI
. � ' 10 ·�\ �\
I
JLETpWV ' .J.. I
EK�EpWUtJI, OVTW Ka·,\ELV EO.W aUtJI' OVOEV OE KOtJIOJI I
� \ � ' '
ft

A ' '
, •
EUTLJI 0JL'YJIPCfJ KUt 'EJL1TE00Kt\Et ' 1T,\'Y}JI TO' JLETpOV, .I
OLO TOJI JLEV
not1JT�v StKatov KaAEfv, Tov S€ �vatoAo'lov p.a,\,\ov 1} 1TOt7J �
20 �v· OJLOLWS' S€ Kav Ei 'T£5' a1TaVTa Ta p.€rpa p.tyvVwv
1rotofro T�v p.lp.'Y}at.V Kn0a1TEp Xatp�p.wv €1rol7JaE K€v"i"av­
pov · JLLKT�v par/JqJSlav J.g a1TaVTWV TWV p.€rpwv, KaC 1TOt'Y}-
T�V 1TpoaayopEvr€ov. 1TEpt · p.ev o�v TOVTWV StwplaOw
-
TOVTOV TOV ' I
Tp01TOV, t \ OE
EtUt � � d ""'
'T£VES': at 1Taat XPWV'Tat TOtS' E1ip1J- .-.. ,.. t

' I � \ 1' f L) � \ 1\ \ I ft
25 JLEVOLS', ,\EYW oE otov pvup.ip Kat JLE,,EL Kat p.Erpcp, WU1TEp
ij TE. Twv StOvpap.{3tKwv TTol7JaLS' Kat � rwv vop.wv Kat ij
�I �I � ..I. ' �· ., • '
TE Tpaylpota Kat' 1J• KWJL<fJOta· OLa�Epovat oE on at JLEV
- t �\ \ I I \ ' � \
a1La 1TaULV a·t oE KaTa JLEpoS'. TaVTaS' JLEV ovv EYW TaS'
w
,\
Sta4>opas TWV TEXI'WV J.v o[S' 1TotoiJvraL T�v p.lp.'Y}atv.
1 1
'E71'Et. OE JLLJLOVVTat Ot• JLLf'OVJLEVOt 1TpaTTOVTaS, avayKT} 2
• ' ;
't Sa
' �

�· � , .. .J.. �,
�

, ..
o E TOV'TOVS' 7J ar.ovoawvs 7J �aVAOVS' EtVat ra yap 1J 7J axEoov
., c· , , "o
l) \ - \ tiL)
'\

' \ I 't " I I . ' t

aEL TOVTOLS" aKO,\. OVUEt JLOI-'OLS', KaKI.fl- yap Kat apETrJ TU 1JU1J
� .J.. I I I _, L)1 f <0 I
ota�Epovat 1TaVTES' ) , TJTOL {3E,\novaS' 7J Kau 7JJLaS' . 7J XEtpovas
>I
A

1447b9 avwVVJ.IoOL Tvyxavovo'tV oJaat Lobel (sed participio nil opus, cf. De
cael. 2 94 a I I , Lobeck, Phtyn. p. 27 7) : Tvyxa,·ovaa II: (avwvvl'o>) Tvyxavovaa
Bcrnays, cum nondum innotuisset -Arabis ' qttae est (sivc su11t, Tkatsch, ii.
1 55· 1 ) s i1te appellatione' ; cf. Tzetz. De diff. poet. vv. 1 1 et 1 67 (pp. 35 et 39
.
Kaibel) 14 ovx ws- II : fort. ov Xt�pls, opp. d,\,\Q. Kotvfi 1 5, cf. ow(v 8€
�<otvov 17 1 5 Ka"Ta T�J' rec : Tl)v KaTa A 16 cpt•atKo'' I (Heinsius) :
J.I.O!IO'LKOV n 22 Ka� n : nil mu tandum, vid. Denniston, Greek Part.2,
p. 585 24- a� rcc : oi II {Ar) 28 o�v tf> (Ar) : ov A 29 ot!; Vic­
torius : at> n (Ar) 144883 -4 Ka�lrt . . . -;raVTE<; II (Ar) : sed. Cudeman

4
 II E P I TI O I H T I K H l; 1448•
., ' .J. A \ 1 ' '
7J Kat TOLOVTOVS'1 OJ0'7T£p Ot ypa•� £tS'' n01\U)IVWTO S' P,£V yap
... ' I

"' ' A - "' ' '' y

5
I n I I I I f
Kp£tTTOVS'1 aVO'WV O£ .X£LpOVS'1 .t.UOVUO'LOS' 0£ OP,OI.OVS' £LKa�£V.
8fjAov 8€ oTt Kat Twv A£X0£tuwv £KaUTTJ p.tp.�u£wv ;gEt
TavTas Tas 8tacpopas Kat £uTat £repa T0 €TEpa IL'P.£iu0ai
\· t \ t I t\ I \ \
TOUTOV
""
TOV
'
Tp01TOV,
I
Kat yap EV OPXTJO'EL Kat aVI\T)O'££ Kat
I I ' \ I [ '] \
"
I
K£0aptu£t eun y£vEu0a£ Tavras Tas avop.oLOTTJTaS', Kat TO 1 0
·

7T£pl. TOV� A6yous 8 € Kat T�v ifJt,Aop.erplav, otoJ' .,Op.T)pos p.€v

{3£ATLOVS', KAeocpwv 8€ op.olovs. 'If�p.wv 8€ 0 eautOS' (o) T�$'
TTap<fJotaS' "' ' '
1TOLTJCT.as 1rpwTOS' A
·Kat' J.1\T VLKoxapTJS' '
o' TTJV
' A \ ' "'
.u £LI\Laoa

X£lpous· ·op.olws 8€ Kat 1T£pt Tovs · 8t0upap.{3ous Kat 1T£pt Toils

v6p.ovs, wuTT£P tyast KvKAw1ras Ttp.60£os Kat fPtt\6g£Vos 1 5
/ >I t t "' ' A "' .I. A \ f
IL'ILTJO'atro av TtS'. EV auru oE 'TTl oto..,op� Kat TJ Tpaycp-
'
TJ' o."'£'
A

"' ' "' ' ) ,

.·

"' ' . 7rpos

o1.a ' TTJV
,· Kwp.cpol.av otEO'T!JI<£zr TJ' P,£V yap X£tpovs
{3£ATlovs P,LP,£ iu0a£ {3ovA£rat Twv viJv.
·· "En 8€ Tov;wv Tplr7J 8c.acf;op� TO ws €Kau-ra .TOVTWV
, ,, \ ' ,
3
..... ' ..... \ \ , ,
IL'ILTJO'atTO av TtS'. Kat yap EV TOtS' auTOtS' Kat Ta avTa 2o
, C' \
A 0at £0''TLV OTE P,EV a1TayyEI\I\OVTa1 , 1\ \ " ff I I
p.tp.£t0' TJ ET£pov 'TL YLJIVO-
'

p.EVOV W0'1TEp 11 0p.T)pos 71'.ot£i �. ws i"OJ' avTov Kat p.� P,£Ta-

_! \ \
. f3a/\1\0VTa, .. I t I
TJ 1TaVTaS' WS' trparrovTaS' Kat EV£pyovvTaS' t, TOVS'
\ t . A \

p.tp.ovp.evovst . £v Tptui. 8� Tav;cus 8tacpopais Tj p.lp.T)uls £unv,

t "
WS' £L'Tt0P,EV KaT' apxas J I t
1 £V OLS' T£ < Kat a) Kat WS', WO'T£ 'TT/ 2 5
f' \ " \ t1 fl -
· ·

p.€v · 0 aVTOS' av £tT} IL'ILTJT�S' 'Op.�pcp Eocpot<Afjs. P,LP,OVVTat

y�p ap.cpw 0'1TOV8aloVS' 1 Tfj 8e :4ptO'TOcpaV£L 1 1TpaTTOVTaS' y�p .
A \ "' .I. <If} "' \
A f
p.tp.ovv-rat Kal. opwvTaS' ap..,w. ov£v Kal. opap.aTa Ka/\Et-
\

' ' .I. A "' A "' '

... J1

ft \
u0a1., 'TLV£S' avTa .,auLv. on p.tp.ovvra£ opwvras. oto Kat,
aVTL1TOLOVVTa£ TfjS'. 7'£ Tpaycp8{as Ka� TfjS' Kwp.cp8las oi J1w- 30
8 Tip rcc (Lat) : Td A Io Ka� Paris. 2038 : Ka� To A I 2 & Ttt;
Paris. 2038 : TttS' A I 3 8ft�u18a A, TJ super n scr. m . rec . I 5 wff1r£P I
yiia ICVKAw1Tiia A: sic"t kyclopas Lat : war.£p yelp KvK�W':Ta; Vahlen : yiia esse
fragmentum nominis Ol••cfmas sive Olvwvas susp. Holland (cf. Athen. r g f et
638 b) Ttp.o8£os A (Lat 0) : et timothetiS Lat T I 6 avTfj 8� Tfj ll: Tfj
aV-rjj 8� Victorius : TaVT'fl 8� Tfj Casau bonus : fort. ·aUTjj 8t (TaVr?J) Tfj (non
obstat Ar) 23 1TaVTa;· Jl: 1raVTa Casauhonus 25 �ea� Cl Paris. 2038 :
om; II 27-�8 1TpaTroVTa; 8pwVTas A : -£5'· -fs Casaubonus
lipwVTa; AB : -£5' Casaubonus
• • • • • •

29 a voce nv£; incipit B

5
 1 448• A P l l; T O T·E A O l' l;

pt.£is ( Tfjs p.£v yap Kwp.cp8las oi M£yap£is oi T£ JvTavOa ws

1 \ ,.. ' "" 1 � I I 't"T f 1
£11t. TTJS 11ap avTots OTJ!J.OKpaTtas y£vop.EVTJS .Kat.\ ot £K �t-
\\�
� ,
\ ,
, � '
•1v 'E7Tt.xapp.os o 110t.TJTTJS 1101\1\Cfl ,
•
'
K.£1\tas, £K£t8£V yap 11po-
�
T£pos v Xtwvl8ov Kai Ma'YVTJTos· Kai Tfjs Tpaycp8las .lvt.ot
35 TWV £v Il£�0110VV�UcP) 7TOtOVJL€VOI. Ta ovop.aTa UTJJL€iov· aVTOt
p.£v yap . Kwp.as Tas 7T£pt.ot.Kl8as Ka�£i'v <f>aatv, l4 07Jvalovs
�' � ' . 1:.' '
• , ' , ,. ' \ � ' .
o£ OTJJLOVS, WS KWJL<fJOOVS OVK Q7TO TOV KWJLa�;,£W I\£X 0£VTQS
}'
�.\ \ \
U(\1\a TTJ� KaTa\ KWJLaS I \ I
111\CLVTJ aTt.p.a�;,O!J.EVOVS I '
£K TOV� QUT£WS" ' >I

1:.'
1:.' n
' \
8b Kat\ TO\ 110t£tv
· ·• avTot\ p.£v \
opav, I
•A o·TJVatovs o£ 11paTT£tV 11poa-
�
I

ayop£V£tv. · 11£p'i p.Ev ·oov Twv �tarfoopwv Ka'i 11oaa1. Ka'i

I ""' I
1 1 8W TQVTa. �
Tt.V£S TTJS p.tJLT}U£WS ELp'JU
I
'E0I.KQUI. 1:.' ' � ' "\ ' ' , 1-
oE YEVVTJUat p.o OI\WS T'TJV 110t.T}Tt.KTJV atnat 4
� I \ \ ../. 1' I I � \
' 1 ../.
5 OVO TI.VES Kat aVTQI. �VUI.Kat . TO TE yap !J.I.JLEI.U 8 at UV!J.�VTOV
, 0pW7TOI.S
I I� 1 1 ,1.._ I ' I
EK 7TQI.OWV £UTI.\ Kat.\ TOVT<p ot.a�Epovat.
�
TOtS av
- "\ \ }' 1 <t ,
TWV ai\1\WV �;,!pWV OTt p.tjLT}TI.KWTaTOV I I
EUTI. Kat\ TQS \
p.a8.TJI -
"" \ '
0:£1.S 110t.ELTat. � \
ota I
JLI.!J.TJU£WS TaS 7TpWTaS, Kat.\ TO\ . xat.p£tV I

Tois p.tp.�p.acn 7TclVTas. UTJfJ.E'iov 8£ TOVTOV 7'o avp.{3ai'vov

Jl
1 \ TWV Epywv·
10 £111. ""
at\ yap avTa\ I\V7TTJpWS \ \ 1
opwp.£v, TOVTWV . Tas "" f' ,.,. I ·\

' 1 \ 1\ f3 I ' I � 1'

EtKOVaS TaS p.al\t.UTa TJKpt W!J.EVaS xatpop.£V 0EWpOVVT£S, Ot.OV
87Jplwv T£ p.op<f>as TWV anp.oTaTWV KaL . V£Kpwv. aiTt.oV 8E:
I ft ../. \ ttl:,\
Kat\ TOVTOV, OTt p.av0QV£1.V OV' P,OVOV I I
TOtS �1.1\0UO�OtS I../.
TJOtUTOV
�
�

\ \ • I ' \
�\\
a.l\1\a kat TOtS "\\
ai\1\0tS OJLOtWS, al\1\ ' \ \'
£7Tt {3paxv\ KOI.VWVOV- �

1 5 ULV aVTOV. �,a. yap TOVTO xalpovat TUS ElKovas opwvTES. OTt
avp.{3alv£t 0£wpovvTas p.avOavEt.v Kat avMoyl,£a0at. Tl EKa-
... . f!/ "" ' ' "" ' ' '
'
, ,
aTov, ot.ov OTt. ovros . EK£t.vos· E7T£t. Eav JLTJ TVXTJ- 11po£wpaKws,
3 1 yap n: om. B 33-34 EKEt8£V p.a:yvq-ros g (Ar) : sed. Else : 0 • • •

71'0t1JT�S del. Gudeman 34 Xtwvl8ov Robortellus (Ar) : xwvl8ov B

35 aQrot Spengel : o&rot S 3 6 l487Jvalovs anon. in ed. Oxon. a. 1 76o,
Spengd : O.lh]va'iot S 1448b 1-2 Kal. '11'pouayopEVELV AB (et . ·. . • • •

appellant Lat) : del. Gudeman ( 1888), om. Ar ; sunt sane inepta, sed ad
tationem concludendam necessaria, cf. 8to 144-8a29 4 yEVVfjuat p.€v
o>.ws A : o>.ws yEVfjuat fLEV B 5 a&rat rec : avral. B 6 8tac/JlpoVCTL n :
8ta�pEL B 7 ECTTt ll
I AI. ' . : om. B . 10 av-ra n : aQrwv B I J Kal.
I 7 OVTOS
..
TOVTOV 'Y : Kat TOVTO A ! TOVTCAIV B " AB : OTt
OTt .. TO; El Se
� I

EKE'ivos A : -roiho EKE'ivo G:udeman

6
 II E P I II O lH T I KH :E 1448b

. .
j .
18 o�x · V Ellebodius, Hermann : . o�xt B 20 8� 8 : 8� Vahlen
22 ol B : om. A 1rpos B (Ar) : �ea.t II 2 5 Uf:JLVOT€poL n: -ov B
26 Twv1· A : om. B £�E'Ma-r£p0t n: -ov B 27 7r0LOVVT£S -ll:
'lrOLOVVTa.£ B b£poL AB : aT€pOL Sophianus, alii 29 ElVa.£ ll : £l8ba.t
�\\ ·
B apea.p.EVOLS A : apeap.Evos B (incipiens Lat) 30 EV ot; n : EVOS .B
� \ \J
3 1 ICQ.L• TO• B : Om.· n 3 5 a.I\I\Q. B : a.IV\
•
OTL n .J6 TO unp.a.
• • •

B : Ta .• • uxrlp.a.Ta. n
• 37 imlB£�w A : a1r- B 38 & B : TO A
1449• 1 1rpos Tas Tpa.yqJ8la.s II (Ai') : om. B in fine paginae

7
 1449D AP I � T O T E A O ! �

r6' fL�V o�v E1TLO'K07T£LV £l apa £x£t 7f87J � rpa:yc.p8la TOLS'

£t8£utv lKaJ'WS' 1} ov, avro T£ Ka(J� aVTO Kp'ivat Ka� 7rpos
Ta . 8larpa, a>.>.os >.oyos. Y£VOfLEV7J a� oOv ' U7T� apxfjs\ avro-
t- ,., ' ' ' ' r � ' r , '
10 O'X£otaO'TLK7]S-Kat aVT7J Kat '1} KWfLlpOta, Kat '1} fL€V a7TO
- �i: I \ ot t' I f t;l> \ � \ "' \ ,/, \
rwv £5apxovrwv rov (Jvpap. f1ov, 7J o£ aTTo rwv ra -yai\-
AtKd. C1 €n Kat vvv ev TTo>.>.a'iS' Twv TToA€wv 8tap.ev£t vop.t-
r I \ fLtKpov \ 7]Vs'1} i: I> I <f t I
�op.oa-Kara (J7J 7rpoayovrwv ouov £ytyv£ro
�
cf>av£ 'v avrfjs· KaL 7To,\,\d.s fL£Taf3o,\d.s fL£•raf3a>.oooa �-
1 I
�' £7Tavuaro, 1 \ £UX£ T'1JV Jl \ aVT7JS' -yvuw.
t "' � � \ I
.

1 5 rpay . ota £7T£t Kat ro

r.: rwv VTTC'>Kptrwv TT>.ij8os eg l.vos £ls 8vo 7rpwro� Aluxv­
Aos 7jyay£ Kat . rd. TOV xop�iJ ?j>.arTWO'£ KaL TOV . ,\6yov
A I t;l> \ Kat O'K7Jvoypa-ytav
' ,1, 1
7rpwraywvtO'THV 7rap£UK£Va0'£V' rpHs o£
A

EocpoK>.ijs. €n 8� ro piy£8os· eK p.tKpwv p.v8wv Kat ·>.t-

20 g£WS' y£>.olas Std. TO EK aarvptKOV fL€Taf3a,\£'iv ol{l� aTT-
1 t T€TpafL£Tpov t f3 £tOV £Y£V£TO
A t I
£0'£fLVVV (J'1), TOI. T£ fL£TpOV I £K I tap. .
\ \ \ ""'
TO fL€V yap 7rpWTOV 'T£TpafL£Tplp £XPWVTO ota , TO uarvptK7]V I 1 ,... � \ \ \

\ li;
Kat\ opx7JO'TLKWr£pav
t I 1'
£tvat T7JV \ 1T0,7JUW,
I l\£s£ws t;l> \ y£vop.£V'1}S'
o£ I
1\ \
aVT7J '1}f -yVO'tS'
, \ ,1, I
TO, 0'K££0V fL€Tpov
, A I 1'
£Vpe p.al\tO'Ta '
yap I\£KTL-
2 5 Kov 'TWV p.lrpwv ro lap.{3£'i&v eunv· U7]fL£'iov 8� ToVrov,
1T>.£wra yd.p lap.{3£'ia >.lyop.£v ev rfj 8taAeKT'!:J _rfj 1Tpos
_> \ \ . 1 \ f i: I t\ I
t;l> \ 01\tyaKtS' .., \
ai\1\'1}1\0VS', £s ap.£rpa 0£ .
Ka'\ £K t fJawOVT£S'I T7JS' I\£KT£-
- f \ \t f
I >I 0£ t;l> \ £7T£tO'OOLWV
t t;l> 'Trl\\ 7]I(J'1). Ka'\ Ta\ a/\1\
I N
28 K7JS' apfLOVtaS'. £Tt WS'
6 maiora et honorabiliora Lat : l'£i{ov Kai. EVTt/LOT£pa A : l'£l{w Kai. Evrt�T£­
pov B £lva' ·Taiha EK£lvwv A : Taiha EK£lvwv £lJ•m B 7 £l apa EX£'
rec : si habet Lat : apa EX£' B : '1Taplx£, A 8 _ti8£atv 11 (Ar) : �8lut
B Kpivat Forchhammer : Kplv£Tat Tj J•a{ A : Kplv£Tat £lvat B (Lat)
9 ywol'tVTJS AB : corr. recentiores quidam (qui et awoax£3,acnK� pro -ijs,
fa1so) �·
a
.,.
ovv , 11
B : ovv 1 o Kat' aVT'Il
' ' ' ' '
_ Kat' 11 : Kat aVT'Il B Kat' TJ'
p.€v 11: 1j /LEv B 1 1 tf,a».tKa rec : tf,av>{A)tKa B p 8iaplv£t rec
(Lat, Ar) : -nv AB 1 5 aw�� AB · 1 8 1TpwTaywvtCJT£iv Sophianus
(ct. Rbein. Mus. 1 902, 1 1 7 sq.) : 'ITPWTaywvtCJT�V E: 'ITPWTOS aywvtCJTtKJv ex .
Ar eliciebat !mmisch 1 9-20 decurtata ; plura videtur habuisse 1:, sed
omnia turbata in Ar : ply£9os EK IL'Kpwv l'v9wv, Kai. (� Me's eK) Me£ws
Christ 20 aaTvptKOV B : aaTvptaKov 11 2 3 3( ywopiVTJs B :
fort. 3' E'Y'Y£1iOILtVTJs 28 llia ws rec : J.Mws AB : alia Lat
8
 B E P I I I O I II T I K II l: 1449• .
£KaaTa Koap.7J8·�vat Aey£Tat €aTC.l.l '�'JfL'iv £lpwdva· 7roAv yap 3 0
iiv Laws €pyov £L7] Ot£gt€vat kaO' €u:aaTov.
5 ' tH S€I Kwp.cp8£aI EUTLV WU7r€p £L7T'Of.UV Ifi.LfL7JUtS' <f>avAo-
)\\ \
T€pWV fi.€V, OV fL€JITOL KaTa 1TO.UO.V ICO.KtO.JI, 0.1\1\0. TOV-
t \ """'

alaxpofJ · EUTL TO y£Aoiov p.opwv. T� yap '):'€Aoi6v EUTLJI . ap.&.p-

. 'J' I J f'
TTJfLO.1 TL Kat ataxos avwovvov > \
:Kat ov 'f' .J.(J
apTLKov, owv 3 5 I � \

'0 ' ' \ I I ' ' I

€V VS' TO Y£1\0LOJI 1TpOUW71'0V ataxpov TL Kat OL€UTPO.fLfL€JIOV
A • �

' . • 1 'I'
O.V€V OOVVTJS'. at fLEJI OVJI T'TJS' TpaycpotO.S' fl.€ TO.{3O.O'€LS' KO.t'
,, ·� I - f:! l

� > f' I . \ \ 10 J <

ot WV €f'€VOJITO J
OV 1\€1\7] 0.(1LV, 7J O€ KWfL<tJOLO. OLO. 'TO f.L"f} � \ � I � \ \ \

a�ovoa,£a0at eg ci.pxi}s £,\a0£v· Kat yap xopov Kwp.cpawv 144CJ

J.P' 71'07€ 0 apxwv €8wK£V, aAA' e0£AovTat �uav. 7fS7J S€
ux-rJH�Ta nva aVTijS' exoVU7JS' ol A£YOfl.€VOL aVTijs 1TOi7JTa'i
.
II I� "' \ I "'
fLV7Jf.LOJI€VOJITO.L . TtS' O€ 7rpOUW1TO. 0J.1T€OWK€JI 7] 1Tp01\0YOVS' '1J
� \ I

f I
\ '(}
1TI\7} 7J V1TOKpLTWJI Kat\ 0(10. TOLO.VTO., ,..
-
7JYV07JTO.L
J
. TO O€ p.v- 5., \ �· \ I

['E ' ' 'i: ' ...

·

A ' A. I , '
(}
OVS' 1TOL€W 1TtXO.pfLOS' Kat 'POpfLtS'.] TO fL€V £s apXTJS' €K
1:tK£ALas -ijAO£,, Twv S€ )10�v7Jatv KpaTTJS 1rpwTos �pgo
aif>EfL€VOS TijS' tap.{3tKijS' l8eas Ka0oAov 1TOL€tV t\oyovs Kat
I
f.LV'(}OVS' . 7J f.LEJI OVV €7T01iOLLO. TlJ TpaycpoLf!. fi.EXPL fL€V TOV
f' \ 1' 1 I - �I \ ....,

\ I \ 'D
f.L€TO.' fL€TpOV
,
11oycp ,
/-"f.L'T}UtS €LJIO.t 0'1TOVOO.tWV ,
7JKOI\OVV7]U£V" IO ? '
�

I t \ \ 1 ';" ,
Tlp o£ TO p.erpov .a1TI\OVV EX£LV Kat a1Tayy£"tav
- � \ \
' Etvat, TO.VTlJ
- " \

\:'· .J. I III

. f _! \ ,.. ,.. <t
OLO.'f'€pOVULV" €TL 0€ Tlp fi.7JK€t '1J fL£JI OTL fl.atl.tUTa 1T€tpaTO.L � \ \

V1TO\ p.taV I <\I 'i: \ \ I

•

.
< . . ? "' \ f
1T€ptooov 7]1\LOV €LVO.L "f} fLLKpov Es 0./\/VlTTEW, 7J 0€ .
I � � \

E1T01TOtLa aoptUTOS' Tlp xpovcp kat TOVTC[J 8tac/>EpEL, KalTot

\ ...., I r ,- ..... ' I
TO 1TpWTOV OfLOLWS' EV TO.tS TpaycpotatS' TOVTO E1TOtOVV KO.L £V 1 5 ,.. � I \ '

TOtS' E1TEf!LV. fi.EPTJ o . EO'Tt · TO. fL€V TO.V'Ta, Ta OE toLO. T'TJS'

""' , ,, �, , ' ' ' , , ' � (' ,� -

.
Tpaycpota�·· OL01TEp OUTLS' 1T£pt TpaycpotaS OtoE 0'1TOVOO.LO.S'
�, � , ., ' � , "� ll' '

3 1 8teedvaL n: 8ulvat B 3 3 a.A>ta s : ill , y Friedrich 34 TO1

n: om. B 36 8tEO'T,O•lp.,UtVO�·] 8t' taTp- A : tO'Tp- B 1449bi l).a8& n
(Ar) : ;).axe B xopov n (.Ar) : xpovo�· B 3 ).EYOILEVOL 8 : fort. y&OJL&OL
6 E1TlxapP,o� Ka� cf>op,ut� 8 : sed. Susemihl (om. Ar, cuius obscura verba ' ut
relinquatur omnis sermo qtti est per compepdimn ' ad notam criticam quae
fucrit in :E incertissima conirctura refert Else) .9 {.t.tv Tou Tyrwhitt :
Jl.O�·ov B 10 w:Ta ,ulTpov :\oycp scripsi (non obstat Ar) : p.tTpov
p.ETa ).6yov B : p.hpov ,U£y&.\ov A : excidit in Lat 1 1 Tcp B : ;.� A
16 Tatha rec : Tav-ra 8

9
 1449b A P I � T O TE A 0 1 �
' ' , """
\ \ ('\
. , '
Kat c/JavATJ�. ol8£ Ka£ 7r€pt. E7TWV" a /.UV yap €7T01TO££a
Jl C'
.
I """
£X£ ' . v7Tapx£' TTJ Tpaycp8Lf!-, � 8� ailrfj, o� miVTa lv Tfj
'
20 E7T07TO£ta '
.
•

Il£pt p.�v o3v Tfjs lv lgap.lTpot� p.tp.T}nKfjs Kat 1T£pt 6

Kwp.cp8las vaT£pov lpovp.Ev · 7T£pt 8� ;.payt.p8la� Alywp.£v
, \ Q ' ., - , - ' ' ' ' •
ava"at'OVT£S aVTTJS EK Twv Et.pTJJ.L€VWV TOV yt.vop.£Vov opov
A ' f
TT}S ovata�. la'ftv oJv Tpayt.p8la p.lJ.LTJU£� 1TpagEws U7Tov8al(Js
I ' I (� I . \ I \ I f
25 Kat T£A£la� J.L€Y£ (}OS EXOVUTJS. TJOVUJ.LEV(fJ 1\0ytp xwpts . EKa-
- 't> ' TOtS J.LOpt0£�1 opwvTWV A t' I ' t' ' ' '
UTlp TWV €WWJ1 EV Kat OV' 0£ a1Ta')'-
� IR I
A

m £1\£0V _Kat 'f'ot'ov 7T£pawovaa TT}V Twv Tot.ovrwv

\I
y£,uas, t' ' ,, I ' , . • - ,

f
7Ta(}TJJ.LaTWV I
KaI(}. apat.V. 1\EYW \ I t' \
OE �
TJOVUJ.LEVOV I
J.LEV\ \ I
1\0YOV .
TOV\.

"£XOVTa pv ( (}J.LOV\ \ ( I [ \ 1\ ' �\ . '

Kat app.ovt.aV Kat J.L€1\0�] , TO o€ XWP£� TOtS
' (}at Kat' 7Tai\£V '\ . "
A

" t' � ' ' ' , '

30 €£0€U£ TO ot.a J.L<ETpWV €Vta J.LOVOV 1T€patv£U ETEpa
� \ 1\ ' \ �\ f \ f
OLa f-L€1\0V� . E7TE£ OE 7TpaTTOVTE� 7TOtOVVTa£ TTJV J.LtJ.LTJUW1 1Tpw- A �

TOV jL�v Eg avayKT}� av £iT} n J.Lbpt.OV Tpaycp8{a� 0 Tfjs

\ I \ Ii;
"./.
O'f'£WS I
KOUJ.LO�" -
€tTa f-L€1\07TO££a 'I'
Kat\ 1\Es£� 1 EV TOVTO£S 1 ' \
yap
- \ I \ I t' \ \ li; \ ' \" \ -
7TOWVVTaL TT}V J.LtJ.LT}UtV. I\£YW 0€ /\ESW J.LEV aVTTJV TTJV TWV
' \ ' t' \ � . \
35 J.L€Tpwv avv ' (}£ULV, J.L€1l07Tottav oE o" TTJV \ � '
ovvap.w .,av£pav
II A ' �\ \li; ' I I I �\ .
£X££ 7Taaav. E1T£t oE 7rpas £ws £an J.LtJ.LT}UtS, 7rpaTTn·at o£
t ""' I I I .,
V7TO T£VWV 7TpaTTOVTWV
\
1 OV� C\
avayKT}
J
7T0£0VS T£Va� EtVa£ KaTaI
T€ TO �'I(}OS Kat\ TTJV
\
\ � I
ot.aVO£aV ( ot.a
� \
yap\
TOVTWVI
KO.f.\ TaS \

;o• 7rpag£,s £lval c/Jap.ev 7Totas Twas 7Tlc/JvK£v aina 8vo Twv
[
7Tpd.g€WV elva£1 8taVOLa Kat �(}o�] Kat KO.Ta TaVTa� Kat
I
Tvyxavovat Kat. a1ToTvyxavovat J \ I I
7TaVT£S ) , eaTw " �\
oE -
TTJS \
p.ev
7rpd.g£ws o · p.v8os � p.lp.TJats, ..\lyw yap p.v8ov TovTov -n}v
1 9 1TaVTa rr: 1TaVTW) B 2 I p.Ev B : om. rr 2 3 ava.Aafl&VT£) Bernays :
a1To- E 2+ U1TOOOala, n (Ar) : om. B 2 5 £Kacmp Reiz : -ov E, tractatus
Coislinianus in defin. comoediae 26 a1Tayy£Ma> A : E1T- B 28 -lra­
fJ't]p.aTwv B, I (' dolores ' fr Syr, 'passiones et impressiones ' Ar), tract.
Coisl. in def. com. : p.afJ.,p.&.Twv rr 29 Ka� pl'Ao> E : del. Tyrwhitt
3 I post p.€'Aov> lacuna haustam esse Ka8apu£ws .explicationem censent multi
3+ avr�v E : TaUTT/V Bywater (cf. I450a4) 3 5 p.hpwv A : &vop.aTWV
Hermann 36 1rO.uav A : 1rO.uw Madius 14508I -2 1TEt/JvKEV • • •

.q8os E (Ar) : sed. Else, transposuerant alii alio 2 8,&.vo,a B (Lat) :

- av A Tattra) Ka� rr : TaV7a B 4 � A : om. B
10
 II E P I II O I H T I K H � 1 450•
� ' "n
, n
UVVUE<nV A
TWV 1Tpayp.aTWV I
1 Ta' OE TJU1J 1 Kaun• 0" 1TmOVS' I
-rtVaS' 5
'P I .J.. \ I � I � ·I t <I \ I ·
Etvac. -yap.EV -rovs 1rpa-r-rov-ras, otavotav oE, EV oaots 1\Eyov-
•

· I I "" \ ,L I I
,
TES' a1TooEtKvvaatv �
n 1J Kat a1To-yawov-rat ,
yvwp.1Jv-- a' vayK1J
I
, A �1 1 .,i:
., • '
OVV .,.
1TaU1JS' T1JS' TpaycpotaS' ft.Ep1J ELVat E� 1 Kaun• 0" 1TOta, TLS' EUTLV
An
TJ' Tpaycpoc.a �,
• TavTaA
o� · Ean • ' p.vuos
Kat' 1JU1J"n . ' 1\E�
Kat \ 11:
tS' Kat'
8uivoc.a Kat .ot/JLs Kat p.EAo1roda. ols p.�v yap p.c.p.ovvTat, zo
• I "
� ,
OVO p.Ep1J, ·
EUTtV1 WS'' �·
OE A
p.tp.OVVTat, EV, a" OE
�· A
p.tp.OVVTaL1 I
Tpc.a,
. \ A
Kat' 1rapa TaVTa OVOEV. TOVTOtS' p.EV · � I I '
OVV t OVK 01\tyot
'I' '\ .
I •
aVTWV A t '
WS'
El'TTEiv Klxp1JvTat Tois Ei8£ac.v · Kat yap tot/Jc.s £xEt mivt Ka�
?]8os Kat p.vOov Kat Algw Kat . p.lAos Kat 8uivotav waaVTWS'.
p.lyc.aTov 8� ToVTwv JaTtv � Twv 1rpayp.aTwv aVaTaac.s. 1 .5
•
TJ, yap ,
Tpaycpota �� 1
p.r.p.1JUI.S ' · EUTW• •
OVK n ,
avupW1TWV 2 \ \ ' 1Tpa-
al\1\(l ,

eEWV Kat {3lov [Kat £v8atp.ov{a Kat KaKo8atp.ovla Jv

1rpae£c. €aTlv. Kat To 7-l.\os 1rpa.ets Tts JaTlv. ov 1Tot6-
0
, ' � E' ' "
' Ta . "f)()1J 1TOI.OtI TtVES', KaTa' . OE
T'YJS'" E.LUtV KaTa' p.EV �'
TaS'

1rpagEc.s Ev8alp.ov£s � TovvavTlo�] " ovKoi.Jv o1rws Ta 7f81J p.c.- 20

fl.'1JUWVTat 1TpaTTOVUtV1 &.,\,\a Ta 1}81] UVp.1TEptt\ap.{3aVOVCJLV
' , i: ., t
� ,
ota TaS' 1rpa� Ec.s· wuTE Ta, 1rpayp.aTa ·
1
Kat' o p.vuos An 1\
TEl\ OS'
A t' I ' �' 1\ ' • , " "
TTJS' TpaycpotaS' 1 TO OE TEI\OS' P.EYLCJTOV a1TaVTWV. ETC. aVEV
\ • .. / �� >/
� \ ' () A /
p.Ev 1rpa�EWS'li;
ovK av y£voc.To -rpaycpota, avEv oE 1J wv y£-
VOLT av· at yap TW- � VEWV I
, t ' "" \ I . ' ' � I
TWV 1TI\ELUTWV a'T}()EtS' -rpaycpotat 25
. , ' "\ . ' \\ ' - .,.' -
ELUIV1 Kat 01\WS' 1TOt1)TaL 'TTOI\1\0L TOtoVTOt1 OtOV Kat TWV ypa-
cf>lwv Z£vec.s 1rpos IloAvyvwTov 1Tl1Tov8£v· & ,..,.�., yap
IloAvyvwTOS' ayaOos �Ooypacf>os, � 8�- Z£vgt8os ypacf>� ov8�v
, , , • • 1
•.1.. 1: A ()TlA p1JUELS' • n , , Kat, 1\EsEt \ ,i:
EXEt �n •1uos. E-rt Eav TLS' - E<yEs1JS' 'T]UtKas
5 o AB : & rec (Lat) 6 �ufvotav B: -a Reiz 7 a1To-
-
8EtKVvao-c.v A : -vovat B 'Tt � II: nva B 8 Tfj� B : om.
A 1 3 Tof� AB : w� Vahlen (cf. 1452b14) yap II : om. B
I 6 1rpQ.fEwv B : 1rpagEw� II - 1 7-20 Kat Ev�atp.ovla 'ToWaVTlov seclusi • • •

( 1 7-1 8 Kat fJlov 1TOtO'T7]� secluserat Ritter)

• • • 1 7 Ev�atp.ovla Kat
KaH:o�atp.ovla B : Ev�atp.ovla� Kat '!] KaKo�atp.ov{a II : non exstant in Ar
I 9 Elai.v II : E!1'Ttv B 2 1 avp.1TEpt'Aap.fJ&.vovcTtv B : avp.1rapa- rec, Spengel
27 1TOAVyvW'TOV B (Ar) : -!1'Tov II 28 1ro'Atlyvw!1'To� II : om. B (et Ar,
sed cf. Tkatsch, i. 1 78) 29-30 MgE,� Ka2 �tavolas B (Ar) : corr. ·

Vlithlen

II
 1 450• A P I I: T O T E A O "l' I:
'
30 Kat.' atavot.�
� '
EV... 1TE1TOt7Jf'EVa�, ov' 1TOt7JUEt '
o" .!,1v TTJ...� TpayqJ-
• . TOVTOt '
at.aS"
��
Epyov,
"
a/\1\a
>\\\
17'01\V\ \
f'WI.I\OJI
�\\
7J KaTaaE£UT£pOtS" � '
�
KEXP7Jf'€V7J '
TpayqJata, �I
Exovua
,
a£
�'
p.v "'
()ov teat. ' uvuTautv '
7rpay-
1 1 .I.
/-f.aTWV.
,
7TpOS" '
a£
�\
TOVTOI.S' Ta' f'EYtUTa OtS' 'f'vxaywyEt. f' "" r

1
'fJ
,
' .
-rpay'fJata TOV f'VUOV f'Ep7J EUTtJI1 at TE 1TEplTT
�I """ ILJ I t .
ET£tat Kat ava-
I \ ,

' , A fl
35 ')'VWpt.UEtS'. ETt. UTJf'EI.OV OTt. Kat. Ot EYXEtpOVVT£S' ?TOt.EtV 1rpo- .
\ , , "' """ I

TEpov 8vvavTat Tfi A€gn Kal TOtS" Tj8Eutv aKpt{3ofiv � Ta .

...
?Tpayp.aTa
'
UVVt.CTTaU' (}
at, OtOV Kat' Ot. 1TpWTOt ...
1TOL7JTat' UXEaOV � '

"
a1TaVT£S", '
«PX"'J ' f'EV ' OVJI.,. �at.' OWV f' ./.
'f'VXTJ 0r fl.V"""'0OS' TTJS'
'
.... Tpa-
yqJat.aS'
�I
1 aEVTEpOV
\:' I
a£
�\
Ta 7J 07J 1Tapa1TI\'1
\ ( "
]CTLOV yap
\ I I
EUTW
t
Kat,.
50b J1rl rijs- ypa¢>ucfjs• El yap TtS' £vaAEfi/JEtE TOtS" KaAMUTOtS'
... t I t
�
'flapp.aKots- xva7JV 1 ovK av op.otws- Evo.ypav£tEV Kat\ ,\EtiKo-
I �� ,/. I

.1.. I /
t

tf
ypa'f'7JUaS' EtKOVa) ' ECTTI.V TE f''f'7JUI.S' 1Tpao:,EWS' Kat ata
t I ll; \ � \
T«VTTJV
I

" "'
p.wuuTa
_I \
TWV npaTTOVTWV. I
TptTOVI
a£
�\
7Jt atavota·
� I
TOVTO a£ � �

5 J.unv T6 ,\£y£tV 8vvau8at. 'i'a J.vOVTa Kal Ta app.bTTOVTa1

01TEp £1rl T<iv Aoy�v �ijS' 1roAtnKfjs Kal , PTJToptKfjs lpyov
' 1 � ' 1 ..... ' ' r
ECTTW' Ot f'EV '
yap
'
apxawt.
' ""
1TOI\LTI.KWS'
\
E1TOI.OlJV I\EYOVTaS"1 Ot
" � \ ...
a£ vvv PTJTOptKWS'. E(JTI.V a£ 'I OS' f'EV TO TOI.OVTOV 0 a7JI\Ot
�' ... • ... " · � , .!(} ' ' ....

· \ I . t I
'T'I]V 1TpOatpEUW1 01TOta Tt.S" r·EV · Ot.S" 'I' OVK t ECTTt J1 a7JI\OV 7J ?Tpo-
� "' \
'1\

10 atpEfTat. � t/>EVy£t.]-8t&7TEp ovK lxovut.v 1]0o� Twv ,\6ywv J.v

10 J Ot.S' p.7Ja1:'' 01\WS
... "\
£UTt.V 0 Tt. 1TpOatpELTat 7J'1\ 'f'EVyEt.
Jl ., .1.. f
0t ,\Eywv-
• f

'
A

. � \ f' 1 f II "'- f II
ataVOta
� I
a£ EV t
OLS" a?ToaEt.KVVOVCTt.
� .I I
Tt. WS" £(1Tt.JI 1] WS" OVK EUTW
"' (} 1 �\ t· A t 7J
\
'TJ Ka 0,\OV Tt. a?TO't'atvOVTat . TETapTOV at: TWV f'EV 1\0YWV t
,/. I I \ I

\:' 1 "
,

11: ,\EYW ' \ 1/:

J1 ., \
,\£sLS"' aE1 WC11TEp 1TpOTEpOV
I
Et.p7JTat., 1\Estv ELVat TTJV
8t.a rijs- ovol-'aulas lpp.7Jv£lav1 o Ka� £1rl -rwv J.p.p.lTpwv Kal
·
30 oil B (Ar) : om. ll 3 1 � rec (que Lat) : 7} AB 34- lcrrlv A : £lalv
B 35 on Ka� n: O�L B 39-1450b3 wapd.w;\�aLOII £lKOIIa • • •

hie A : post wpayfLaTwv (145oa.B ) transp. Castelvetro 1 TLS lva).£{-

1/Jf:L£ n : Tt. EllaAf:lt/Jn B 5 Ka� Ta n : Ka� B 8 T� A : om. B
9 &wola TLS t/J (cf. q62a12) : &woia TLS AB 9-10 lv ots r/J£uyn 8 : del. • • •

Bekker (om. Ar) 10-IOI 8L01r£P ,f:Vyf:t. n (Ar) : om. B •101 ots
• •

A : quibm quidem Lat IL1J8 ' o;\ws n: fLTJ 871AOS' ex Ar eliciebat Lobel T'
rcc : TLS' n ' I 2 TWII fL�V Aoywv E : om. Ar ( ' vidmtllrJuisse interpolata '
calamo notavit nescio quis in editionis Oxon. a. 1 76o exemplari
Bodleiano)
' 12
 II E P I IT O I H T I K H :E 1450�
• ' \ IA \
€1TL TWJI 1\0YWII E"XE"L T1JII aVT YJII ouvap.tv. TWJI 0"'� /\OL1TWII
" ' ' ' � I �
I5
A

7j p.€A01Totla p.l.yuJTov Twv �8uup.aTwv. ,j 8€ o�ts 1/Jvxayw-

I J -
YLKOV JL€111 aTE"XVOTaTOII OE" Kat 1JKLUTa OLKE"LOII TYJS 1TOL1J-
. \ I �\ \ fl t ...

TL�S' � yap ;fjS Tpay<pf>las f>v!laJLLS Kat aii€V ayw110� Kat

V1TOKpLTW� EUT�II. ETL 8€ KUp-twTI.pa 1TE"pt Tr,l' a1TE"pyaalav
'TWV oi/JE"wll � TOV UK€V01TOLOV T'X"1J Tijs TWII 1TOL1JTWJI €unll . 20
7 L1twptu�l.llwll sJ TOVTWV, . .\l.ywp.€11 p.E"nl TaVTa 1Tolall
\ � ""' \ I .,. 1
Ttlla OE"t T1JII UUUTaULV €Lila£ TWII 1Tpayp.aTWV1 E1TEL01J TOVTO
"""' I �\ ""

Kat 7rpWTOV Kat p./.ytUTOII . TijS Tpaycp8{a� JuT{II. K€iTaL 8�

' \ I ' ., , 'i: 1' ,
'YJJLW T1JII Tpay<potaV TE"/\E"LaS KaL_ 01\1}) 1rpaf:> EWS ELVa£ JLL-
. �,
''
• A

' I t'l\ t1
p.1JULV EX OVU1JS Tt J.l.EYE"0os · E"UTLV yap OI\OII Kat\ JLYJOEV EXOII 25
I \ �\
I It\ 1 )
iJ-EYE(}OS'. 01\0V OE EUTLV TO t:XOV apxlJV Kat JLEUOV KaL TE"-
�I \ \ \ I \
' \ 't
II

\ 1 c\
1\E"UTYJV• I
apx'TJ
J \ �I
OE" E"UTLV 0 aVTO JLEV\-
JL1J\ E"s- allaYK1JS' JL€T
J I )
Jl \ \ ' ' �· "
£UTLJI1 JLE"T E"KELJIO 0 €TEpov 1T€o',I.i' VK£11 £Lila£ 7J"' YLII£0' . 0at·
f ,
al\1\0 1' f

\ �\ t "\ \
A

\ 1' "'
T£/\£VT7J OE TOVVaVTLOV 0 aUTO JLEV JLET al\1\0 1T£'f'VK£V ELVa£ 7J
t I 0\ t \ \ I,/.
'i: I .
, "' , '
• ' \ I ' �' "\ \
A •
€� a11a')'K7JS' 'YJ WS' £1TL TO 7TOI\U1 JL€Ta 0€ TOUTO al\1\0 OV0£11"
�,
J "' \ \ \ '
30
c\ ""' 1 ft
J.l.E"UOV OE 0 Kat aUTO JLET Ul\1\0 KaL JLET EKELVO £T€p011.
I �\ \ ' \
...
OH apa TOVS' (1UII£0'TWTaS
A 1' f Jt
\
EU JLV'(}OVS' JLTJlot OTrOI(}€11 £TVXEII
, I(}' , \ \\ \
A >I

orrou £ VXE T€1\EVTall.

>/ .,· )
apx£u(}at · P·'YJ T a1'"a KEXpflu(}at Tais
A

E"lp7Jp.l.11atS' lSI.ats� ETL () ' ETrEL T<) KO.AOV KaL {0 ov Kat a1TaV
.... t I ....
7rpayp.a 0 UVII£UTTJK01 EK TLIIWV OV fWVOII TaVTa TE"TUYJL£11a 35
- t\ I 1 I

t' >\\ \
OEL E"X££11 Ul\1\a KaL JLE YE OS' VTrapX££11 JL YJ\ TO\ TVXOIV ' TO\ t I
tf \ I (}
-" (;
A

\ \
yap \
Kai\OV £11
t
J.l.EY£I(}€t Kat\ lasH €UTI.V1
t I
OLO
� \
OUT€
II
1TaJLJLLKpov
I
" I " '
KUI\OV _ tpOV ( UV')'XHT aL yap • , • '
l. A '
aV Tt YEIIOtTO A

'YJ 0€Wpta £')'YV�

TOV avatu8�TOV xpovo v ytvop.I.VYJ) OVT€ 1Tap.p.l.yt:8Es (ov yd.p
_

fl t () t \ \t II () A
ap.a £wpta ytv£Tat a"" OLX£Ta£ TOLS .:wpovut TO Ell 145 \ C\
I I
1J
A

Kat TO o.\ov EK TijS B£wplas) olov E"l p.vplwv UTa8lwv E"LTJ

1 5 .\omwv B (Ar) : .\ouT(uJi ;,lvr£ 11 : .\omwv TTEp.'OTTov r<>c (ct v.I. in tP) ·

i 6 ,;. . . . ol/sts A :
al ot/sns B 1 8 " B : ws II
• • . 22 T�V A : om. B

' ., ' ' ' ' ' .

23 87 (By\vatcr : 8€ B 28 y{v£a8at 11: y£vla8at B 29 auro
' I)) : ct 'p sum J.at : avTo A
· ' '
3 I Kat aVTo II : Kat avro Ktlt ( pnus
p.£v
Kal. � u pra aliTo) R J3 07TOV n : OrTOL R 37 et 39 7TUV
p.tKprw m1v p.ly£8os All : corr. rec (omnino par·vum . . . omni11o
• . • •

magmmr J.at) 39 al•ata8?JTOV 11 (Ar) : ala87JTOV B xpovov .1 : del. Bonitz

IJ
 C O M M E NT A R Y
For abbreviations see p. xxvi.
C H A P T E R S 1-5
I N this · introductory section A. distinguishes the forms of poetry,
which is assumed to be a mimesis of human actions, according to the
medium employed (Ch. 1), to the objects imitated (Ch. 2), and to the
manner of imitation (Ch. 3). In Chs. 4 and 5 he discusses the origins of
mimesis and shows how different forms were developed by men of
different dispositions, giving in the process brief histories of the two
basic forms, tragedy and comedy. He is then ready to proceed in Ch. 6
to his main subject, tragedy, together with epic, which is viewed as
a rudimentary form of tragedy. The' corresponding treatment of
comedy (promised at .49b21) is missing.
CHAPTE R 1
47&8-13. The subject.
47&8. '11'0&TJT&tc'ij� : sc� TlXV7J'> as in Pl. Gorg. 502 c, though the word was
perhaps felt by now to b�. a subst. in its own right. As the subject of
a book it would recall the Tlxv7J P7JTop�K�, the Handbook of Rhetoric ;
the purpose of these books, which had been i.n existence for a
century or more (see Introduction III), was to teach the art of
speaking, but in th� P. A.'s object is mainly to define the nature
and functio.n of poetry, though instructions for the poet are
included.
a.uTi]� : emphasizing the contrast with Twv £lowv atiTij<>, the parti­
cular kinds of poetry.
d8wv : the species of th<;! genus poetry. Used also of the various
types of tragedy (cf. 55b32), and at 56a33 apparently for the parts,
p.lp7J, or p.6p,a:· For A. the first step tmyards the comprehension of
a subject is to divide it according to its natural categories. Cf.
Pl. Phaedr. 265 E : KaT' dS7] SvvauOa, s,aTlp.v£w KaT' apOpa ·n '1Tl�VK£V •

. . . The popular division by metre is rejected (47b13).

47a9. 8uva.J.L&V ! here no more than 'effect', epyov ; frequently in A. it
means 'potentiality' as opposed to 'actuality'. Related to the
Svvap.''> is the TlAos which each £loo<; achieves.
'11'WS 8Ei auviaTa.a8a& Tous J.LU8ou� : 'how plots should be con­
structed ' ; p.vOos normally 'means in the P. the story as organized into
the plot, the TTpayp.aTwv uvuTau's ; uvv,uTava' (sometimes uvvluTauOa,)
represents the activity of the poet which results in the TTpayp.aTwv
UVUTaU'S or uvv8£u,<;. B. cites 50b32 and 53b4 as evidence that the

53
 COMMENTARY
verb is pass. here. Sm;netimes, e.g. at 51b24, 53318, 37, p.fi8os retains
its older meaning of)egendary story or myth, on which tragic plots ·

were normally based. By a natural extension it is used of the

invented plot of comedy (srbr3).
A. has singled out plot for emphatic mention right at the begin­
ning of his work, in which it is the dominating theme. It is by the
construction of plots more than by any other single means that the
poet achieves his purpose.
47a1 0. �ea.�ws €gELv : to be good of its sort : cf. Tov KMws· ExovTa p.fi8ov .

53al2.
11'0L'laLs : lit. the 'making', 'composition' of poetry. It sometimes
preserves this literal sense,_ as possibly at I. 14 below and, E. thinks,
here; However, it early became a general word for poetry (first in
Herod. 2. 23, 82), and must be so used here if it is the subject of 'uTl
in the next clause ; if it is to mean 'composition', the subject of EuTl
must be supplied from 'TIWIJTtKfjs or Elawv, which seems less natural.
1rolTJuts and 1rol1Jp.a later acquired narrower technical meanings, first
perhaps in Theophrastus ; see N. A. Greenberg, HSCP 65 (r¢r),
263 ff., and C. 0. Brink, Horace on Poetry, pp. 62 ff.
It is important to remember that in the words 1rol7Juts, 1rolTJp.a,
'TTOtTJT�s the idea of making is completely dissociated from the idea of
'creating' with which it is frequently combined in English. -!} Tfjs
Tpaycpatas 1rolTJu's means the fashioning of tragedies by a poet in
a sense similar to -!} 'Twv Tpa1TE,wv TTolTJuts, the fashioning of tables by
a carpenter. A. began the elevation of the poet through revealing
that poets gave significance to poems by organizing their structure , ·
by making stories into plots. 'The very word 'TTOtEtv • • means "to
•

create" ' (Gomme, Greek Attitude . to Poetry and History, p. 54), is

r
misleading : Aoyo1rotos means, among other things, 'writer of prose'.
1roawv tca.l 11'oiwv : cf. the closing lines of P. 62b17.
47a 1 1 . f'OPLWV : p.optov and p.epos are used indifferently by A.
47a12. f'E8o8ou : 'inquiry' ; cf. Pol. 1317b34 : b -rfi p.E8oacp -rfi 1rpo -ravTTJS.
.....
ICO.TQ 'I'UaLV: Cf Ph I� \
Q,...b31 : EU'Tt yap KaTa\ Y'VUtV
.L , 'Ta\ KOtVa\ 1TpWTOV
�
t I
OV'TW Ta\ 1TEpt EKaUTOV tota ���
• I • • "

EL7TOVTaS' fl \ •
8EWpEtV
•
A
.

47at3-47b29. The forms of poetry distinguished in terms of medium.

47at3. E11'o11'olia.: as p.lp.TJuts is an activity, A. may be stressing the
process of maldng, 1r,olTJu's, rather than the thing made ; but E1To1Totla
Can mean 'epic' just a5 avapLaVT07TOt{a Can mean 'statuary', and Since
Kwp.cpala is used here without 'TTOlTJuts it is simpler to take all as re­
ferring to product rather than process (cf. 47b26). The four kinds of
· poetry mentioned were the most important at this date. Though
Antimachus was the only recent epic poet of note, the dominance of
Homer ensured that epic should not be neglected ; works in the other

54
 COMMENTARY
three forms were still being produced in large numbers. Under
dithyramb, the choral song of Dionysus, is included the nomos, the
song of Apollo, mentioned separately at 47b26. Dithyramb itself,
originally narrative� had become highly dramatic by A.'s time (cf.
6Ib29); Non-choral lyric, such as was written by Sappho and
Alcaeus, was now nearly extinct.
47& 15. a.u�TJ"''lKijs . . . Kl8a.pla1'lKijs : the aulos, something akin to the
clarinet, was used to accompany th� dithyramb and was regarded as
highly emotional ; the more restrained cithara or lyre was associated
with the nomos, which was originally choral, hut in A.'s time ari
astrophic monody ; both were used in dramatic performances. The
problem is the meaning of the qualification 'most'. The only natural
distinction is that between music unaccompanied by words, 1/Jt>..�
p.ovauc� 'bare music' the Greeks called it-and the use of these
-

instruments to accompany lyric poetry. The former was not much

practised (unless accompanied by the dance, in which case it was not
strictly 1/Jt>..TJ) , though a competition with . the unaccompanied aulos
was part of the Pythian Games from 582 B.C. Plato in a famous
passage, Laws 669 D--670 B, denied that such music had any clear
meaning and condemned it. A. seems to have held that rhythm by
itself was meaningful. But 1] 1r>.ElaTTJ probably carries an admission
that not all 'bare music' was mimetic.
I� the rest of the P. A. has little to say about music in general,
an'd nothing about instrumental music. The subject can be intro­
duced here quite naturally because A. takes 1rot7JnK� to cover
p.ovatK� as a whole, in which music and dancing (47327) were mostly
subordinate to poetry. ' The scheme of the media is made as com­
prehensive as possible, though A. will have little to say about some of
them. The term p.ovatK� occurs only at 62316, referring to the musical
element in tragedy.
For the use of music in education and the rejection of the aulos see
Pol. I34I3I7 ff.
Tuyxavoua&v : see 47b9 n.
47a 1 6. f.UJ11\aE&S : that poetry, music, painting, sculpture, and �a�ce
are all forms of mimesis A. takes for granted. p.tp.£ia8at means to
make or do something which has a resemblance to something else.
No one English word is adequate in all contexts, but basically the
idea of imitation will be present. See Appendix I.
1'0 auvo�ov : not 'in general' but 'viewed collectively'.
8&a.+€poua& 8€ ci��1]�wv Tp&alv : the · mimetic arts can be dis­
tinguished in three ways, according to the medium used (colour,
words, etc.), the subject of this chapter ; according to the object
imitated (men, good or bad), the subject of Ch. 2 ; and according to
the manner of imitation (by narrative or direct speech), the subject

55
 C O M M EN T A R Y
of Ch. 3 · p.ovauc� is included here as part of 7TOLT/TLK� (c£: 62816), but
the P. deals only with imitations 1n which wqrds are used. . . ·
47•17. £v : the regular word to indicate medium (cf. 47b29)·
478 1 8. wa"'I'Ep YC\P �tal xPwl'aa' . . : the first of seven passages in ·
•

which the aCtivities of the poet are illustrated from the visual arts,
which presumably represent for A., as for Plato, the simplest form of
mimesis ; see G.'s note ad loc. In Plut. M. 1 7 F poetry is still a
mimetiC art and aVTLUTpotfoos Tfj 'wypat/J{q. .
XPWI'aa' Kat aX''ll'aa': the media used by painters and sculptors ;
the' latter normally applied pigment to their statues. ax�p.a-ra can
mean, in addition to the static shapes of painter or sculptor, the
shapes into which men put themselves, the postures, e.g. of dancers
as in. axwta-rt,op.Evwv pvOp.i;JV 1. 27, below, and of actors at 6283 ; cf.
Plut. lV.l. 747 c-E. It means also the form or structure of a play
(cf. 4986).
4 7• 19. a"'I'ELICGtovTEs : 'making likenesses'. Words derived from EtKWV'
'an image', were used primarily for visual representation, but they
contain the same basic idea as mimesis. Cf. Xen. Mem. -3. Io. I :
awp.a-ra 8ta TWV xpwp.a-rwv a'ITELKcl,OVTES £Kp.Lp.E'ia0e ; Pl. Laws 668 A :
I I ,/... , J I 'I' I
ftOVULKTJV ')'E 7TaUaV 'f' «f'EV ELK«UTLKTJV TE ELV«L K«L f'Lf'7JTLK7JV,'
- \

ot J&Ev s,a. TEXVTJS

• • • : this parenthesis contains an idea here quite
irrelevant but dear both to Plato and to A. To do a thing a,a. TEXVT/S
requires knowledge of the basic principles of what you are doing :
.cf. £yw 8� TEXV7JV ov Ka�w 8 ii.v 'll a�oyov 7Tpiiyp.a, Pl. Gorg.._ 46S A with
Dodds' note ; you may do it very successfully in the light of long
experience £p.7TeLplq. (avv�Oeta and £p.7Te,pla are coupled by A. at EN
ns8815), but you will not be able to give a rational account of your
procedure. According to Plato's Gorgias rhetorical skill is all of this
empirical sort ; A. tried to make a real science of it in his Rhetoric, in
which he treats not only of tricks of style but of the arguments to be
employed and of the psychology of audiences. Similarly poets
learnt which myths provided good plots by working their way ·
through the available myths, so that they discovered by trial and
error which were really suitable (see 5381 7). A . • in the light of his
TE)(VTj, could have directed them immediately · to the best plots ; so
(51824) Homer did the right thing Tf-rot 8ta -rExv17v 1} 8ta tfovatv. He
could hardly have possessed a complete theory of epic, but he fol­
lowed h�s genius. There is a similar contrast between TEXV17 and
-rvx7J at 54!a 1o. A. converts a chance discovery into a part of TEXVTJ
by giving a rational explanation of it. For � general account of the
difference between TEXVTi and ep.7TEtpla see Met. 98ob28 ff., though
common usage made little of the distinction.
47820. +wvfJs : it is a problem who· the 'others' are who use the tfow� as
their medium. The first thought of the modern reader, for whom
 COMM E N TA R Y
poetry consists primarily o f printed words, will be that sound, the
medium of music, should supply the second analogy. t/>wvfJ, 'the
(human) voice'� can be used metaphorically for the sounds of musical
instruments, and it is conceivable that it could here refer to so"und in
general as the medium of the musicia� (it was so understood by S.
and by Tkatsch), especially as the more general word t/Jot/>os includes
mere noise as well. But this will not do, since the Greeks included
virtually all music in poetry, and A. would be illustrating the media
of the eipTJf-1-lJif' Texvat 1. 21 from themselves. (It is true that rhythm
and harmony, the media of poetry 1. 22, are not the same as sound,
but they are inseparable from sound in this sense). The same objec­
tion applies, though rather less strongly, to the rendering 'voice'.
Most music was sung, and accordingly the voice is one of the media
to be illustrated, not itself a proper illustration. But A. did call the
voice 'the most imitative of human faculties', P.'P.TJTtKwmTov Twv
p.oplwv (R. 1404321), presumably because of its emotional suggestive­
ness, and he might have used it, though illogically, as being a helpful
illustration here. Another and less vulnerable suggestion is that the
reference is to those who supplied entertainment of.a somewhat
vulgar kind by giving imitations of sounds'difficult to render with the
human voice, such as cries of animals and the squeaking of pulley
wheels (cf. Pl. Rep. 397 A, Laws 669 D). Plutarch tells of a Parmenon
whose performance of the part of a squealing pig was agreed to be
better than the real thing (M. 18 c, 674 B). An objection is that these
performances are hardly on a level such that they can suitably be
compared to painting -and poetry. An art is here explained by the
analogy of a parlour-trick. But this is the view of V., G., and R., and
it remains the most likely. E. supposes the voice to be that of the
rhapsode reciting epics.
For a technical sense of t/>wvfJ see 56b38.
47822. £v pu8J1� �ea'i. MyctJ �ea'i. apJiov£� : we thus have rhythm, words,
and music as the media of the elpTJp.ivats Texvats, those listed in
11. 13-15, above. The artist represents material or personal objects by
colours on a two-dimensional surface, in fact· patches of colour, as
they appear on the retina, by patches of colour ; the entertainer re­
presents sounds by sounds ; the poet represents human action by
words, rhythm, and music. From one point of view words differ
from the other media in that, unlike colours and sounds, they . are
conventional symbols. But i\oyos also means speech, and A. probably
thought in terms of the performance of a play in which the spoken
word represents the spoken word accompanying the original action.
fw8p.os is essentially a pattern of recurrence imposed on speech or­
on other sounds giving rise to expectations which are more or less
fulfj.lled, so that even in over-rhythmic _ prose 'one waits for the

57
 COMMENTARY
recurrence', woT£ wcfi\,v �e£' (R. HOsb24). I t i s defined by Pl. Laws
665 A as � Tfjs �ew�a�ws Tae,s, the application of order to movement ;
the ethical effects of rhythms are due to a correspondence between
this ordered movement and the movements of the soul ; see Appendix
I. Aristoxenus, a pupil of A., says (Rhythmica p. 4II Mar.) €an Ta
pv8P,,Cop.£Va Tpla, Me,s, p.li\os, �elv11a's awp.aT'"�· On the history and
many senses of pv8p.6s see E. Wolf, Wiener St. 68 .(!955}, -99 ff.
Myos is a meaningful combination of words ; an individual word
is ovop.a, used in later grammar to mean a noun as opposed to
pfjp.a a verb. (We find ewos in the P. only in the plural, meaning
hexameter verses or epic.)
ripp.ovla 'fitting together', as a musical term is a satisfying relation
between notes. I translate it by 'melody', though this does not
exclude the notion of rhythm. The only English word referring to
pitch alone is the highly technical 'melos'. At 47b25 app.ovlq. is re­
place� by p.li\n and �oyo�s by f'7iP'P, wit�ou t si�nificant change of
meanmg. The combmat10n i\oy6s-app.ov,a.-pv8p.os . occurs at Rep.
398 D, but A.'s classification of the arts in ten�s of the media em­
ployed is probably new. Gorgias' definition of poetry as i\oyov
€xovTa p.lTpov, Hel. 9, was intended to minimize the difference
between poetry and rhetoric in the interests of the latter.
47826. aup£yywv : the syrinx was a less sophisticated instrument tha:n
the aulos. The latter was played mainly by professionals (cf. Pol.
I34I8I7 f.). Alcibiades is said to have rejected the aulos when it was
still fashionable (Plut� Ale. 2. 5).
47827. opx1JaTwv : dancing has not been mentioned among the ·dxva,,
though the choral dance was part of both drama and dithyramb. But
A. must here refer to unaccompanied solo dancing which can hardly
have been common. Solo dancing was highly developed and very
- mimetic ; as the pantomime it became a favourite form of entertain­
ment under the early Roman Empire. Even so, whether for enter�
tainment or for ritual, it was normally accompanied by music. The
main sources of information are Athenaeus, I. 25-27, 37-40, and
Lucian, De Saltatione, especially ch. 6o.
47828. KQl. fi9'1 KQL 11'0.91) KQl. 1rpci�ELS : curiously emphatic. A single
medium can cover the whole field of p.lp.f1a,s. �8os, not quite the same
as character : see on 50b8. wa811 Kai wpae£,S, here a related pair, the
things that are done to a man, i.e. the things that happen to him, and
the things that a man does (cf. 5rbn) ; for different senses see on
49b24, 52bro, 55b34· For the general effect of the dance cf. Pl. Laws
655 D : f''f'�f'aTa TpO'II'WV EaTi Ta 'll'(pi Tas xop£las.

47828-47b24. A digression on the deficiencies of nomenclature.

47&28-47b2. 1) 8e .G.vwvuJ.LoL: after mimesis using rhythm and
. •

ss
 C O M M E N T A RY

rhythm with harmony (instrumental music) we have the medium of

words and of words with rhythm, here called p.ETpots . as at 47b25,
and equivalent to .\oycp with ;JVOp.ijJ. .\oyos or .\oyos t/Jt.\os is an ade­
quate expression for prose, but it ,does not distinguish .mimetic
prose, like mime and Socratic dialogue, from history or a speech in
the courts. As A. points out in the course of the digression which
follows this complaint, the nomenclature of poetry too is unsatisfac­
tory in a similar way in that mimetic and non-mimetic verse are not
distinguished. Homer and Empedocles are both E7To7Totol because
both wrote E7T"', hexameters, though Empedocles' poems have nothing
in common with Homer's except the metre. (The Greeks had no
category of didactic poetry.) .With the traditional text t/Jt.\o'is � To'is
f'ETpOtS and avwvvp.os it is not clear whether A. is lamenting a double
deficiency, the lack of a word for the class of mimetic prose writing,
and of another word for mimetic verse writing regardless of the par­
ticular metre used, or whether he wants a word which should cover
mimetic writing both in prose and in verse. The run of the sentence
rather suggests the latter, the natural sense and the evidence of the
examples, the former, which is the interpretation of M., following
Gallavotti, Riv. di fil. 58 (1930), 74, and of E. Lobel's emendation Kal
Tj for 7j with avwvvp.ot (CQ 23 (1929), 76) puts it - beyond doubt that -
there were two forms of mimetic writing .which lacked a name.
t/Jt.\o'is, which is in the predicative position, is probably to be taken
both with .\oyots and with p.ETpots, i.e . .\oyot that are not in verse and
verses that are not accompanied by music ; cf. Tovs ..\oyovs Kal -t�v
t/Jt.\op.ETplav (48aii, but see note ad loc.). p.ovov in 1. 29 stresses the
absence of rhythm and harmony. Good prose would, in a sense,
possess rhythm, but it is not necessary to the form.
_ The recognition that what we should call 'imaginative' writing is
possible in prose is remarkable, and was a stumbling-block to
Renaissance critics.
47829. �eat TOU'I'OLS • •: a superfluous complication introduced by A.
•

either for the sake of completeness or because he wanted to mention

Chairemon's Centaur (1. b21, below). · A mixture of metres would
usually be a mixture of spoken and sung verses, involving p.l.\os.
Iambics, trochaics, and anapaests could be mixed in drama.
47b9-24. The explanation of the 'nameless' forms leads on to a digres­
sion on the naming of forms in terms of metre, a practice which
conceals the difference between mimetic and non-mimetic ,writing.
The return to the main theme of media is marked by "'tEp: ;.t�v o�v
TovTwv • • at 47b23.
•

47b9. civwvul'o' : A. often remarks that there is no available word, e.g.

EN no7b2, Pol. 1275a30.
TuyxuvouaL : they might have been given names but, as it happens,

59
 COMMENTARY
they have not. For this slightly emphatic use see G. Thomson on
Aes. Ag. 125--9 and W. S. Barrett on Eur. Hipp. 388. At 47a25 it
is a synonym for Eluw.
4?blO. IW+povo� : Sophron of Syracuse (late fifth century) and his
· son Xenarchus wrote mimes, realistic sketches from everyday life,
developed, perhaps, from the comedies of Epicharmus. Some idea of
their nature may be got from the urban mimes of Theocritus (2, 14,
and 15) and from Herodas. But these are in verse, while those of
Sophro_n, though he was sometimes referred to as a poet, were
generally considered to be in rhythmic prose: There is a strong
tradition · that Plato greatly admired Sophron ; A then. 504 B, Diog.
Laert. 3· 18. For the only considerable fragment see Page, Gk. Lit.
Pap., p. 328, and A. S. F. Gow, Theocr. ii. 34· A. himself said that
Plato's Dialogues were half-way between prose and verse (fr. 73).
47b l l , Iw�epa.TLJCou� Myou� : the conversational philosophy of Socrates
gave rise to a new literary genre. Plato's dialogues, some in dramatic
form, some told by a narrator, were the most famous example.
Alexamenus of Teos, whose works are lost, was the first to write in
this form. Xenophon wrote a number of conversation pieces in most
of which Socrates is the chief speaker. An important fragment of the
dialogue llEpi llotTJTuJV seems to be making the same point, ovKoilv
•A \ 1:
ova� £p.p.lTpovs TOVS KaAovp.lvovs Ewtf>povos p.lp.ovs �-�-� tf>wp.EV dvat Aoyovs
A
Kat' JLLfl-TJUELS,
'
TJ.. TOVS'
niiEsaf'EVOV TOVA TTJLOV TOV!;
' ' I
1TpWTOVS .1.8EVTaS
ypa'l' '
TWV EwKpaTtKWV ataAO)'WV (Athen. sos c = A. fr. 72). The text is
uncertain, and probably to be printed as a question. Rostagni sug­
gested (see Introduction II, p. xii) that the llEpi llotTJTwv was
. roughly parallel in structure to the P. and contained a full exposition
of the theory of mimesis. There is no evidence that the reference to
Empedocles fr. 70 (cf. 1. 18, below) came from the same part of the
dialogue. R. suggested also that A. may have pointed out the ironic
implications of the mimetic activities of Plato who was so resolute
in condemning mimesis : cf. Athen. sos B : awos (sc. Plato) TOVS .
· 8taAO)'OVS JLLJLTJT.LKWS ypat/Jas,
·

47b13. 1rA�v . . . : if the mimes and Socratic dialogues were put, for
example, into iambics they would have a common name, but it
would be in virtue of their metre, not their content, and so irrelevant
qua mtmests. .
47b14. -1row� : as TPLJLE-Tpo1Totos is not found, Lobel proposed the ex-
· ·

cision of TptP,lTpwv in 1. 11 (CQ 23 (1929), 76).

47b1 6. +uaLKov TL! i.e. a poem 1TEpl tf>vulws as of Xenophanes or
Empedocles.
47b1 8. 'EJ.L1rE8oJCAEi : poet, philosopher, and mystery-monger of Acra­
gas in Sicily, c. 493-433. Considerable fragments survive of his poem
On Nature and of his Katharmoi ; they bear out the opinion of A.,
6o
 COMMENTARY
quoted by Diog. Laert. (8. 57) from the ll£pt no,'T}'TWV (= A. fr. 76),
that he was a master of poetic dic�ion. This opinion is in no way
inconsistent with what is said here, as diction by itself does not
make a poet. On differences between E.'s similes and Homer's see
B. Snell, Discovery of the Mind, pp. 214-16.
4 7b20. OJlOLWS 8i • • : this takes UP47bi : IL'YVVCJa p.€7' aAA�Awv.
•

Tci JlETpa : E., p. 57, seems right in restricting p.lTpa to non-lyric

metres. It is true, though he does not mention it, that at R. 14o6b36
A. uses p.lTpa in reference to Pl. Rep. 6o1 A, B where both verse and
music are in question, but the lyric element is not essential to Plato's
argument, and A. is notably careless about references. So p.lTpa here
need mean no more than dactylic, iambic, and . trochaic (cf. 59b34_
37) ; otherwise the exaggeration would be considerable.
47b2 1. XaLptJJ1Wv : a contemporary of A. who seems to have aroused his
interest or to have been a topic of the moment. He is · mentioned
again at 6oa2 for his use of metres in combination and at R. 1413b13
as author of tragedies meant for reading rather than acting. Five
iambic lines of the Centaur are cited by Athenaeus (6o8 E) as fr01:n
a Spap.a TToAvp.£Tpov. What A. means by IL'KT�v pmpwSlav is unknown.
A rhapsody is normally a portion of epic of a length to be given at
one performance.
47b22. 1<al. 1TQLT)TTJV 1TpoaayopEuT(ov : �f there were any evidence that
the Centaur was non-mimetic, the point would be that Chairemon,
like Einpedocles, has to be called a poet for lack of any other word.
As it is, we must suppose that TTpouayop€vTlov picks up TTpou­
ayopevovTis in 1. 15. C. cannot be called a TTap.p.£TpoTTo,&s (Tyrwhitt)
'writer in all metres', so he is called TTO''IJT�s.
47b24-29. We return to the main subject, poetic media.
47b25. pu811� KQ.t JlEAEL Kal. JlETJMtJ : cf. 47a22. Both p.l>..os and p.lTpov
imply the presence of words, so i\6yos is omitted. Pl. Gorg. 502 c says
that if these two elements are removed from a poem only i\oyo' are
left.
p.l>..o s denotes words sung and therefore includes app.ovla. But the
words themselves are nowhere treated as a significant part of the
p.l>..o s, which appears to depend for its effect on pv8p.6s and app.ovla.
Equally p.lTpov is never used by A. of lyric metres . pv8p.os contri­
.

butes to the effect of the ·words but p.lTpov is metrical >..oyos and >..oyos
is predominant. Cf. Probl. 920a12, in the time of Phrynichus TToAAa­
TT>..auta elva' TOTE Ta p.l>..'IJ l.v Tais TpaycpSlats TWV p.lTpwv.
47b27. 8LaciJ(pouaL : in the dithyramb and in the nomos (here mentioned
specifically for the first time : see 47a13 n.) all three media are used
continuously throughout-words are sung by a choir that dances ; in
tragedy and comedy the dialogue generally uses only words and
61
 COMMENTARY
rhythm, music and dancing being confined to the chorus, though
actors sometimes sing (cf. 52b18).
There is n:Jthing par�icularly striking about this classification by
media except .A.'s perception that it ought to be made to include the
mime and Socratic dialogue. The arrangement is somewhat con­
fusing, partly because two out of the possible combinations of three
things, taken three, tw0, and one at a time, are left blank. Harmony
cannot stand by itself, since all music is associated, if not with words,
with rhythm, and for the same reason words plus harmony cannot
exist without rhythm. A. begins for no obvious reason with rhythm
plus harmony and rhythm alone, then passes to words alone and
words plus rhythm� The digression on nomenclature divides this
· from the final .sentence on the two ways of combining all three media.
Except for a reference in th.e next chapter pure� music _and dance
have no further place in the P., and song, p.i>..os, though a part of
drama, is discussed not at all.

CHAPTER 2
The second differentia;· the forms distinguished according to- the objects
of imitation. Men are superior or inferior; the writer, like the artist, can
represent either sort. Sometimes, as with the dithyramb, both types can be
represented in a single form, but in general each form is concerned with
only one type, hence the importance of type as a differentia. As is empha­
sized in the last sentence of the Ch., this is the difference between tragedy
_ (along with epic) and comedy (along with iambic or lampoon, 48b24 n.).
Our P., the first book, is mainly about the poetry of superior charac­
ters ; the lost second book dealt with comedy and the inferior. The
same distinction is important also in the account of the differentiation
and development of literary forms given in Ch. 4 ; the main division is
between those who by temperament were attracted to superior or to
inferior human beings as their subject.
4Sat ot f.LLJiOUJiEVo&: the poets, who are ultimately responsible for the
. .

play or epic, as at 48a:26 where Homer and Sophocles are specified .

. But the same word can be used for the performers impersonating
characters, as the dancers at 47a28 or the actors at 62a1o. This
ambiguous use is the less unnatural since down to . the time of
Sophocles the poet was himself the principal actor as well as being ,
producer and inventor of the dances for the chorus ; he was 1-'-'P.TJTTJS
·

on several different levels.

'll'paTTovTas : the idea that men in action are the subject of epic and
drama was already lamiliar. Cf. Rep. 6o3 c : 7rparrovras, tfoap.Ev,
av8pcfJ7TOVS p.i.p.Eirat 1} 'p.tp.TJTtK� /ltalovs 71 EKOValas 7TpaeE,S, and simi­
larly 396 c. Since A. is working up to his definition of tragedy as the
62
 C O M M E N T AR Y
p.lp.71cns of a 1rpa.e,s 49h24, it is likely that 1TpaTToVTas contains the
implication of acting purposefully, not merely doing something : see
note ad loc. Though 1TpaTT£w, unlike the English 'act', never of itself
means to perform on the stage, A. uses the verb indifferently of men
acting (pursuing a course of action) who are the object of the poets'
imitation as here, and the men acting (performing on the stage) who
are the medium of his imitation as at 6oa14, and it is not always clear
which meaning is uppermost in his mind-apparently men in action
(48ai, 23, 27 ; soa6 ; 50b4), men performing (49b31 ; soa21 ; 6oai4)·
civciy�e'l 8£ : the M here is connective, carrying on the force of £1r£l,
rather than apodotic, and the .apodosis begins after some paren­
theses at 81j"ov 8£ in l. 7 : so G. and S.
48•2. a11'ou8a{ous '1 cflauXous : cf. Laws 798 D : IJ-OVUtK� -rpo11wv IL'IL�ILa-ra
Q \
I
t'EIIT,ovwv KatI XEtpovwv
I ' 8 '
av pw1rwv.
17. and rfo. indicate the two ends of the ordinary, aristocratically

based, Greek scale of values. ap�:-r�, 'excellence', forms no positive

adjective of its own and u1rov8afos could be used to fill the gap. Cf. ·
A. Categ. IOb7 : a1TO -r1j<; ap£-r1js a · 01Tov8afos. Tip yap apET�V EX£LV
01Tov8afos My£-rat, and Pol. I324ai3. ap£lwv and apt!7TOS are in fact
related to ap£-r�. o. and rfo. form a regular pair of opposites, e.g. EN
1 113a25, EE 1221b33, expanded at EN 1 145b9 to· o. Ka� E1Tatv£-ros • • •

tf,. Ka� 1/JEK-ros. It is the mark of the o. to concern himself with the
pursu"it of ap£-r�, which from Homer onwards is centred on honour.
The tfoav"os is an inferior being, not because he is actually wicked but
because his capabilities and ambitions are mean. Cf. R. 1387h12 :
ol av8pa1To8w8ns Kal t/Jav"ot Kal atfot"OTLIJ-OL. The word can be used
without any suggestion of reproach as at Thuc. 7· 77· 2, when Nicias,
in catastrophic plight before Syracuse, says he is in the same position
as o tfoav"o-raTos 'the humblest' of his soldiers. But they are people
not worth serious attention and no subject for tragedy, which is
about ol £1ritf,av£is (53a12), heroes and 'persons of quality'. Near
synonyms of o. are XP"lu-ros and E1Tt£tK�s, both words applied by A. to
the characters of tragedy and both by him contrasted with rfo. For
further discussion of o. and kindred words see Vahlen, pp. 267-8.
These connotations of the word o1rov8aios are relevant to a de­
ficiency which has been seen ,in A.'s account of tragedy, notably by
Wilamowitz, Herakles1, p. 107, that he disregards the her<?ic element,
the status of human and divine implicit in the myth, which was its ,
all but universal subject. In fact the conception of the o1rov8aios '
av�p covers many of the same values. None the less it is probably
true that for A. and his age the myth had worn a little thin and they
were disinclined to discover in it the pr�mitive profundities revealed
by twentieth-century critics and psychologists. It is on a more super�
ficial level that he speaks with appr9val of the story of Athena's
63
 COMMENTARY [2. 48az-
rejecti�n of the flute as evAoyws p.ep.v8oAoy"1p.evos (Pol. 1341bj).
Theophrastus added something when he defined tragedy as �pwuc7is
TVX71S 7TEplaTacns. Cf. also the later definition by Poseidonius 7Tol71ats
8' EUTt 11"1ftaVTtKOV 7T0{1Jp.a p.{p."111tll 7TEptEXOV 6elwv TE Kat av8pw7Telwv ;
see Diog. Laert. 7· 6o ; Brink, Horace on Poetry, p. 65.
The superiority of the a. a�p is relevant also to the educational
effects of tragedy. The activities of admirable people must reflect
admirable standards of coriduct. Whether we call this 'didactic' is
a question of words.
�81J : 'disposition' is closer than 'character' to the meaning of �6os,
but it is a disposition acquired chiefly by training, not implanted at
birth. A. derives �6os from · £6os, 'habit'. We become just through
acting justly, though the degree of justness we can achiev� depends
on our natural. endowment (cf. EN I I44b4-1o). Once our �8os is
formed, the decisions we take with a view to action will be largely
determined by it (see EN II03a15). Common usage was rather wider.
Cf. Aes. Ag. 727 : �6os To 7Tpos ToKewv (of a lion-cub), Pindar, OZ.
II. 19, and A. him,self on the 7f871 of the aristocrat, R. 1390brs . .
4Sa3. · TOUTO&!; ciKoXou8Ei f'OVO&!; : 'follow, as effect from cause', 'go
with' (cf. EE 1232a31). TovTots : in spite of TovTovs in the line above,
this refers to a7Tov8alovs � cp. not to 7TpaTTovTas as E. argues. It is true
that �6os is revealed by action (sobS), but it is not the point here,
which is that all 7}671 are comprehended in the terms a7Tov8aios and
cpafi>tos. As A. is going to base a vital distinction on these two cate­
gories he emphasizes that they are all-inclusive . .
4Sa4. i} KQ8' TJf'a!; : equivalent to nvv irilv (1. 18), which is perhaps an
echo of Homer's otot vfiv {JpoTol elatv.
4Sa5. i} KQl To&ouTou!; : this third term, 'those like ourselves', in addi- ,.
tion to those who are better or worse, corresponds to nothing outside
this chapter in the literary forms which it is supposed to illustrate,
and is wholly superfluous. Possibly the comparison with the three
painters was originally made elsewhere in a different connexion. The
contrast between Polygnotus and Pauson appears again at Pol.
I340a36. The requirement . that the characters of tragedy should be
op.otot, which is put forward in Ch. rs, has nothing to do with this
· tripartite division. If tragedy is to arouse the proper emotions, the
characters must be like ourselves to the extent that we can feel
sympathy for them (cf. 53a5).
WcnrEp ot ypcui»Eis : sc. p.tp.ofiVTat.
noMyvwTOS : the celebrated fifth-century painter (cf. 50a27).
4Sa6. nQuawv : perhaps the person mentioned by Aristoph. Ach. 854
and elsewhere in Old Comedy, who is said by the schol. to have been
a painter ; as he corresponds to Hegemon the parodist, it is possible
that he painted caricatures.
 C O M M E-N T A R Y
A.LovuaL05 : probably the fifth-century painter mentioned by Pliny,
NH 35 · 1 13.
48a7 .XEx9ELawv : i.e. in Ch. I .
.

48B9. opx�aEL : dances representing those worse than the average man,
among whom satyrs are no doubt included, are commonly shown on
vase-paintings. Cf. also Laws 814 E.
48a t t . TOU5 Myou5 IJIL.Xoi-LETPtav : prose, or verse but without
• • •

music ; this is more likely than G.'s view that Kat is explan�tory,
',\6yovs that is to say t/JtAop.£rpla'. Presumably Aoyot here are mimes
on various levels of seriousness.
"01-LTJP05 : though as author of the Margites (48b3o) he wrote also
about rf>aiJAot.
·

48at2. K.XEocJlwv : a tragic poet of this name is mentioned in the Suda.

The inappropriate diction of a Cleophon is referred to in R. 14o8axs.
•Hy�!-'CaJv : he lived at Athens in the second half of the fifth century. ·
48B13. 11'apct»8ta5 : perhaps a play on pat/Jcpota, burlesque epic. The
Batrachomyomachia of uncertain date is a surviving specimen, and
the Hymn to Hermes is n,ot far removed in spirit. Burlesque . of
myth was a feature both of satyric drama and of comedy. It is
not_ known what innovations H. made. He is said by Athenaeus
(4o6 E, 6<)9 A), to have produced parodies in the - theatre- possibly
there was a parody contest at the Panathenaea-among them a
Gigantomachia. He used the ludicrous tag Kai. ro TTEpOtKos uKlAos in
the same sort of way as Aristophanes his A'Y/K'Jlhov aTTwA£u£v.
NLKoxO.p'J5
• • AEL).L0.8a : may be the comic poet contemporary
•

with Aristophanes. The title, if Deliad, would mean a tale of Delos,

if Deiliad, an epic of cowardice.
48B14. 8L9up0.!-LJ3ou5 v61-Lou5 : it is hard to believe that many poems
• • •

in these classes were of other than exalted type. But a new and more
flexible kind of dithyramb was introduced early in the fourth cen•
tury (see Pickard-Cambridge, Dithyramb, p. 38). Timotheus, the
innovator in musical technique and friend of Euripides, and Philo­
xenus of Cythera were both writers in the new style. Philoxenus,
when in love with the concubine of Dionysius of Syracuse, is said to
have relieved his feelings by representing his rival in the guise of
a grotesque and love-lorn Polyphemus. Something of his spirit is
probably reproduced in the two Cyclops poems of Theocritus, 6 and
I I . Timotheus can suitably represent the . more serious type of
dithyramb . . Those · who believe that there existed dithyrambs on
three different levels can get a third name by taking as as the last
letters of a poet's name, 11pyiis (Castelvetro) or Olvw1ras (Holland).
What little is known of Argas (see Athen; 131 B, 638 c) suggests
that he was a bad poet rather than a writer of dithyrambs on low
characters.
 C O MM E N T A R Y
48816. QuTff 8£ Tfi 8LQ+op�: the difference of objects presented is·
common to several kinds of poetry, but in tragedy an.d comedy
especially the class of object is the distinguishing feature -of the
poetic form. The account of the central character of a tragedy given
later in the P. (Ch. 13) is not altogether easy to reconcile with the
basic requirement that it should be a1rov8aios. It is worth observing
that, although in Ch. 6 tragedy is defined as p.lp:qats 1rpa�EwS' a1Tov-
8alas, it is the characters that are here the differentia. Presumably
\ \ \ '/: \
a 1rpa�,S' a1rov8ala is to be defined as a 1rpa�,S' c17Tov8alwv. Cf. 48b25 :
I
TaS' /Cal\aS' 1Tpa<:,E'S'
'
ICa' TaS'
""'
TWV TO,OVTWV,
As a,at/Jopa has not been mentioned since 1. 8 Casaubon's TaVT'!l. or
Kassel's avrfi 8e TaVT'!l is an improvement.

C H A P TE R 3
The third differentia, the manner of mimesis (narrative, dramatic, or
mixecl) 4881 9-28 ; followed by a digression on the Dorian claim to the
invention of tragedy and corned!, 48a28-b3
488 1 9. TOUT«a»v : sc. f''f'�aEwv ; in view of 47a16 this is more likely than
a,atPopwv (Vahlen), in spite of 1. 24, below. In fact division by manner
ap.plies only to those forms which use �oyos.
EteQaTQ : the objects ofimitation (cf. Ch. ·2).
48820-24. This is one of the most difficult passages in the P., though
the general drift of the meaning is clear. Plato (Rep. 392 D-:-394 D)
draws a distinction between those forms in which the poet speaks in
his OWn perSOn (8'�)'1JU'S'' 8,' d7Ta)')'EAlas) and those in which he
. speaks through his characters, i.e. between narrative and dramatic ;
but the dramatic form can be introduced into narrative, as it often
was and is, giving the· 'mixed' manner. In this part of the Republic,
· though not elsewhere, Plato adopts a .particular meaning for the
word p.lp.1Ja,s, the activity of the dramatic poet or of the narrative .
poet speaking through his characters, 'impersonation', a usage which
seerris to have affected A., though he did not generally adopt it (but
see 6oa9) . . As Plato explains himself with great care, actually putting
part of Iliad i. 1 7-42 into narrative by way of illustration, it WO\lld
seem that this distinction was unfamiliar. There is no reason to
think that Plato denied that narrative was imitative in the wider
sense of the word.
The first problem is the grouping of the clauses. As the text has
come down to us we have a oTE p.ev followed not by oTE 8£ but by 7j
three times. If we take the first 7j as equivalent to or€ 8£ (a usage not
exactly paralleled, though we have or€ 8£ following al p.ev at 52b5),
then we have three 'manners' of mimesis as follows :
·

66
 C O M M E N TA R Y
(I) The mixed, the poet at one time narrating at another 'becoming
something else', i.e. assuming the role of a character, as Homer does.
(Apparently Homer used direct speech more, and narrative less,
than other epic poets (cf. 6oa7), which A. regards as one of the
reasons for his manifest superiority.)
(2) With the poet maintaining unchanged the part of narrator.
According to Plato this was characteristic of the dithyramb, but
direct speech was certainly · not excluded. Probably A. has no one
·

form in mind.
(3) With the imitators acting and taking part throughout. (The
syntax of this sentence will be considered later.) This is clearly the
dramatic form, in which the poet never speaks in his own person.
On this interpretation, which is B.'s, we have three possible
manners : mixed, narrative, dramatic. Commas are required after
'1TOt£i and . p.£-ra{J&.>,.>..ovTa. This is very · close to the passage of the
Republic which is generally supposed to have inspired these dis­
tinctions.·
Or it is possible to take the third .;;, the one after p.€-rafJ&..Uov-ra, as
equivalent to oT€ S€. Then we have two main divisions, the first of
which, narrative, has two subdivisions. Thus :
(I) with the poet narrating : either (a) becoming (at times) someone
else, like Homer, or (b) maintaining the part of narrator unchanged ;
(2) the dramatic manner as above.
.Commas or dashes are placed after a1Tayy€,\,\ovTa and p.€Ta{J&..UovTa.
Most editors prefer the second interpretation. oTt p.ev suggests
a single main alternative to follow. Given A.'s manner of writing it
is not a serious objection that with €up6v n ytyv6p.f:vov we have to
understand that the poet sometimes speaks in his own person as
well. Apparently A. does not distinguish between passages in which
poets narrate and those in which they speak personally as in invoking
the Muse or commenting on their story, e.g. Il. 23. I76.
4Sal t . ETEp6v TL: the neuter is curious, but the objects of imitation
Were referred tO in the neut. Ta avTct in 1. 20. 6oai0 €lUct'}'€£ av8pa. 1j
yvva.tKa. 1j /1,\,\o :-r£ .ryOos is perhaps relevant. Vahlen compares Phys.
247bi8.
4Sall. 'ITOLEL : probably 'composes'. Cf. s8b8.
4Sal3. ij 1rcivTa.s • • !IL!IOUjiEvousf : the construction of the whole
•

sentence is complicated by its being ace. and infin. depending on

1\ \
eanv. If we rewrite it with finite verbs, it runs Kai ycip £v -rois a.vTois
'lito fl I
(0t f'£f'OVf'€VOS
' fl \
a1Ta'}''}'f:AI\WV, TJ €T€pOV n
I 'Ill tt f .l' \ \ 'ft I f I
KQ.£\ TQ.\ Q.VTa
t \
p.tp.€tTa£
"""
) OT€ f'€V
'}'£')'VOf'€VOS , , , TJ WS 0 aVTOS Kat. f'1J f'€TapQ !1.1\1\WV, TJ '1TaVT€S WS '1TpaTTOVTES
'

' \ \ \

Kal £v€p')'OVVTES (p.tfoVVTa£ oi p.tp.ovp.€VO£),

 COMMENTARY
The change from sing. to plur. is noticeably awkward ; behind i t ,
as B . suggests, is probably the thought that, whereas a single poet or
rhapsode recites, a number of actors perform. We have here the
same ambiguity as was noted in the discussion of 4831. Far-reaching
implications are discovered by J. Jones in his Aristotle and Greek
Tragedy, p. 59· It is all but inevitable that ws Tov athov should be
parallel to ws TTpaTTovTas both from the build of the sentence and
from the fact that the chapter is about TO ws fLirfL�UatTO av TtS. But so
far as A.'s usage is concerned ws· TTpaTTovTas could well be the object
of fLtfL€iu8at equivalent to �ta Twv TTpaTTovTwv, the medium of the
dramatic poet. This woulcl. then be picked up by the · TTpaTTovTas of
1. 27, below (where A could have written ws TTpaTTovTas). , If with
most editors we -retain Tous fL'fLOVfLEVovs, the obscurity is somewhat
increased. Vahlen at one time proposed its excision and Butcher
brackets it in his text. If it agrees with TTavTas TTpaTTovTas it is worse
than superfluous. If it is the object of fLtfL€iu8at it has to be taken as
an ;i.lmost unique pass. = o T�v fL{fLTJUtV TTap€j(wv 54326 : so Tyrwhitt,
and cf. V. ad loc. The most acceptable expedient, if we wish to .keep
Tovs fL· is to take it as subj. of fLLfL€iu8a�r, and ws TTpaTTovTas as obj.,
·

with Casaubon's TTavTa for TTavTas.

.
48825-28. waTE Tft JlEV • O.JicJlw : we have here two illustrious pairs,
• •

Homer and Sophocles, whose subjects are heroic characters, Sophocles

and Aristophanes, whose medium is drama. Aristophanes is scarcely
mentioned by A., who was no enthusiast for the Old Comedy with its
aiuxpoAoyla, 'obscene abuse' (EN I128322). He probably felt more
admiration for Epicharmus, but the two Athenians make a better
pa1r.

With this illustration of the Tptul fnmpopais at 1. 24 the first section

of the ·work is rounded off (cf. 47316) . Our own way of dividing the
arts is different, and the interest of A.'s system is largely historical.
For us music has a separate existence in its own right, and the arts·
which combine music with words and dance, opera and ballet,
belong to music. The distinction between grand and low personages
as objects of representation has had little relevance to literary forms
since Milton and the French Classical drama, The third distinction,
between narrative a.nd dramatic presentation, owes its significance
to the conditions prevailing in the ancient world, where literature
was something heard rather than read. Plato's Ion 535 c leaves no
room for doubt about the dramatic quality of a recitation from
Homer, and of the difference between Homer and an epic poet who
provided no speeches in which the rhapsode could show his po:wers of
impersonation. Comments in the scholia, such as that on Eur. Ale.
163 : fLETEf3TJ �€ EK Tov €gTJYTJTLKoii €TTi To fLtfLTJTLKov, show that later
68
 C OM M E N T A R Y
scholarship continued, whether or not out of mere pedantry, to
recognize · this distinction ; cf. also Vit. Aesch. I9, and Dian. Hal.,
� \ �
Thuc. 37 on .the Melian Dialogue : Kai Ka'r' apxas �LEV EK TOV zatov
, \ • -I.' £KaT€pWV,
• , � o�· aTTOKp,a€WS
E'1H 1-''aS , �
-

0€VTa
,
•

TTpOUWTTOV OTJI\0' Ta /\£X V't' TOVTO

'
' ,

TO axfil-'a a,aTTJp�aas, TO a'1JY1JI-'anKov, TTPOUW'1T0'1TO,€t TOV �LETa TaVTa

au�.\oyov Kai apa,.,anKOV ; cf. also [Longinus] 27. I with D. A. Russell's
note.
Today a not wholly dissimilar distinction arises in the study of the
novelist's technique. 'Dickens's impulse is always to present, in
dialogue and pantomime ; instead of telling us about, he shows us' :
see Wellek and Warren, .Theory of Literature (London, 1949), ch. xvi
'The Nature and Modes of Narrative Fiction'.
There is a certain artificiality in A.'s structure of forms ; we have
a number of media in each of which, and in certain combinations of
which, three kinds of object can be imitated in two (or three) dif­
ferent manners. But of some of these possibilities there are no in­
stances at all, for example, of manner in those _ media which lack
.\6yos, and others are represented by freak instances such as � di­
thyramb about a Cyclops.
4 Sa26 o QUTOS : a compressed expression for 'the same sort of'.
48B28-b2. A digression introduced, or suggested, by the Dorian
associations of the word apav used synonymously with TTpaTTe'v and
from the same root as apa,_,a, 'drama' ; this . was the basis of a Dorian
claim to have originated tragedy and comedy. A. expresses no
opinion on this claim, but presumably thought it of some impor­
tance. On the ?riginal meaning of apav see B. Snell in Philol. Supp. ·
20. I (1928) and H. Schreckenberg, iJpii,_,a (Wiirzburg, 196o). The word
was freely used in fifth-century Attic by both poets and prose writers,
but some fourth•century orators avoided it.
Themistius, who cites A. as his authority on Thespis, states that
tragedy was invented at Sicyon, Or. 27. 337 B. This too may come
from A., but it is hardly proof that A. accepted the Dorian claim.
48B29. QUTci : tragedy and comedy, as suggested by Sophocles and
Aristophanes, who, rather than .the neut. pl. avTa, are probably the
subject of ,_,,,_,ovVTa' ; possibly ot ,_,,,_,ov,.,evo' is to be understood, as
the names of the forms are older than the poets.
4Sa3 t . �vTa.u8a.: Megara next to Attica as opposed to M. Hyblaea
north of Syracuse. The word suggests that the passage was written
at Athens, and so before 348 B.C. or, more likely, after 335·
48B32. 81JJ.LOKpQTLa.s :following the overthrow of the tyrant Theagenes
at a date which is very uncertain, usually taken to be early in the
sixth century. While Attic Old Comedy could have existed only
under a full democracy, there is no evidence· whether Megarian
comedy was of the kind which only a democracy would find

6g
 COMMENTARY
congenial : and it i� questionable whether there could have been such
a democracy in the period of Theagenes. Susarion, whom a dubious
tradition made inventor of comedy and also claimed as Megarian, is
. not mentioned here. Megarian comedy of a later date is referred to
in Aristoph. Vesp. 57 and in EN 1123a24.
48833. 'E-rrlxapJ.&os • • XtwvL8ou Kal Mla.yv')Tos : the Suda gives the
•

jl'or. of Chionides as eight years before the Persian War, i.e. 488 B.c.,
and _Magnes is known to have won his first ·victory at the Dionysia in
473/2. Both may have competed at the first official comic contest in
486 B.C. For the official list of comic poets see Pickard-Cambridge,
Festivals, p. II4. Epicharmus is thought to have begun producing
comedies (mainly non-choral) at Syracuse towards the end of
the sixth century and to have died at a great age about 467 B.C.
Accordingly 'much earlier' is a stronger expression than would seem
justified. .
It is extraordinary that Epicharmus, one of the most famous . of
anci�nt poets, should be described as o wo,7J-r�s. These words at
least are suspect. L. G. Breitholtz, Die dorische Farce im gr. Mut.ter­
. land (Goteborg, 1 96o), is sceptical about the whole tradition of early
Dorian comedy.
48835. iv neXo-rrovvt1actJ : probably connected with Sicyon where, ac­
cording to Herod. 5· 67, 'tragic choruses' performed in honour of the
hero Adrastus early in the sixth century. These choruses would be
pre-dramatic. Epigenes of Sicyon is mentioned in the Suda as a
tragic;: poet after whom Thespis was either the second or the sixteenth
in the list of tragic poets; Corinth, where Arion developed the
dithyramb, was Dorian but doubtfully within the Peloponnes�.
ovoJ.&aTa: i.e. the words �ewp.7]-Mjp.os and 8pO.ll-11pa-r-re,v.
48836. KWJ.&as : the Dorian equivalent of the Attic demes, country .
districts. . (The urban demes may have been an invention of Clei­
sthenes.) A. shows himself aware of the other (and true) derivation of
comedy from Kwp.os, the procession of revellers. · .
Various stories appear in late commentators c_onnecting comedy
with �ewp.a' : see Kaibel, Com. Gr. Frag., pp. 6, n, 16. The version
which A. has in mind is not extant.
A. reverts to the early history of tragedy and comedy in the course
of his sketch of the development of poetry in the two following
chapters : see especially 49a9-b9, where there is no mention of
a Dorian contribution except for comedy. -ra r/Ja>J.uca would be more
easily associated with a �ewp.os than with a �ewp.7J.
48b2. wepl J'EY o�v • • : the subject of classification is now finally dis­
•

missed. But before the main discussion begins it is necessary to

trace the development of poetry -in time and to show how the dif­
ferent forms emerged. How far this account is to be regarded as

70
 · COMMENTA RY -
a mainly logical scheme of development, and how far it is intended as
a record of historical fact, is a difficult question.
A.'s scheme of mimetic arts is given in diagrammatic form by G., .
p. Io8, and with some modification by Sohnsen in CQ 29 (i935), 196.

C H A P TE R 4
The origins of poetry, the division into forms, and the development of
tragedy.
'
·

48b4-24. The two causes- of poetry.

48b4. €oCKa.aL : B.'s rendering 'it is clear that . . . ' is rather strong for
a hypothesis ; the word is used to avoid dogmatism. Cf. EN 1o¢b5 :
'tTdJavwupov 8' £oltCaaw ol Ilv8ayopt:tot A/yetv 'ITEpl athov, where 'seem'
is the natural English.
yEVVijaa.L: 'produce', used of the. invention of dancing (Pl. Laws
673 D).
o>..ws : either 'in general', allowing for other incidental or subor­
dinate causes such as, G. suggests, divine inspiration, or, more likely,
'as a whole', contrasted with the particular causes leading to the
growth of the various species ; so E.
a.t1'La.L 8Uo : the pleasure in imitating and the pleasure in imitations
performed by others ; or pleasure in imitation and the instinct for
melody and rhythm (see on I. 22, below).
48b5. cJ>uaLKa.( : as an integral part of human nature, repeated in aup.�vTov
(£aTtv). S. sees here a denial of theories of divine inspiration. Any­
way the traditional 'ITpwTos evpeT�s is conspicuously absent from this
account, as is any attempt to enlarge on the relationship between
the artist's mimesis and the poet's.
48b7. J.I.Lf1TJTLKWTa.Tov : the neighbouring {c/Jwv has suggested {cpov with
which p.tp.TJTLICW'Ta'TOV is made to "agree instead of with av8pw'ITOt. Cf.
Prob. 9S6al4 : � O'Tt fUP.TJ'TtKW'Ta'TOV (sc. av8pw'ITos) ; p.av8&.vetv yd.p
8&va'Tat 8td. 'TOV'TO.
48b8. Ka.C : this carries on the canst. of a&p.�vTov.
It is characteristic of man not only to imitate but to take pleasure
in the imitations performed by others. Were it not so, we might have
poets but no readers of poetry;
4Sb1 0. €m T&;v ipywv: 'in practice', cf. EN II3Ibi8, rather than 'in the
case of works of art'. But at 62a18 it means 'when plays are per­
formed'. G. seeks to save consistency by reading £K Twv �pywv there.
48b 1 1. �&G.XwTa. .r.Kp,pwf1Eva.s : in the extreme case of aKplj3t:ta viewers
are persuaded that they see not a representation but the reality, as
with Plato's carpenter (Rep. 598 c). In A. a'IT&.TTJ with reference to
·

the arts is conspicuously absent : cf. Appendix I.

JI
 C O M M E N TARY
48b12. Q'"f&OTiiTwv : cf. 'c{Jots Ka� p.Eya>..ots Ka� p.tKpo'is Kal Ttp.lots Ka�
aTtp.oTipots De An; 404b4, and Part. An. 645a15
On the other hand, if we enjoy a representation because the
p.opq,� of the object as represented is pleasing we shall be pleased by
the object itself (Pol. 1340a25).
· Most ancient painting and sculpture (ElKovas covers both) was of
mythical subjects, among which corpses would appear from time to
time, e.g. the children of Heracles or Niobe ; the lowest animals, one
would have thought, less often, as Circe's swine or a hydra.
The difficulty of recognition in the case of nov apxalwv ypaq,lwv is
mentioned at Top. 140a2I.
48b 13. tcQl. Tolhou : we have had the proof that it is so, aTJp.Eiov TOVTov
1. 9 ; we now have the reason for it.
A. nowhere attempts to analyse the difference between the re­
actions of a man viewing a representation and of the same man viewing
the original, and so ignores a basic problem of criticism-unless,
which is not very likely, it formed-part of his explanation of katharsis.
According to De An. 427b21-24 a representation , arouses a feebler
emotion than the original.
f1Qv8civELV • • • -ij8LaTov : cf. Pl. Rep. 475 D, A. Met. 98oa22. The same
explanation- of the same paradoxical fact, that we enjoy looking at
representations of things in themselves unpleasing, is given at R.
1371b4 : E7Tf� 8e To p.av8av£iv T£ �8v Ka� To 8avp.a,£w, Ka� Ta Tota8£
' I f fi;! I f' f"
TOI Tf. p.tp.7]nKOV,
I II
,1.. �
�
ypa'f'tiCTJ\ Kat\ avoptaVTO·

3_
aVa')IKTJ TJOfa ftVat OtOV WU7TEp
7TOtta Kat 7TOt7}TtK7], Kat ?Tav o av EV p.Ep.tp.7]p.Evov u, Kav u ILTJ TJOV
I I 1' I C� \
'

\
,. .-. l\ 't\ '1\ \
avTo\ To\
t

p.Ep.tp.7]1,tfVOV' ov. yap E'IT� TOVTCP xalpEt, d.>..>..a av>t.>t.oytap.os EUTtV OTt TOVTf!
eKE'ivo, waTE p.av8avEtv n avft{JatvEt. The explanation is inadequate.
When we have learnt what already familiar thing a picture re­
presents we have not learnt much. av>t.>t.oyl,£a8at I. r6 is somewhat
nearer the mark. We have the intellectual pleasure of solving
a puzzle, as in the simple delight of the Chorus in the parodos of Eur.
Ion when they recognize the subjects of the ( ?)reliefs at Delphi;
Plutarch has a highly confused discussion of this problem in his De
Aud. Poet., M. r8 A-D. This same idea is introduced far more
plausibly at R. I4Iobro in connexion with metaphor and simile,
where the moment of illumination which comes from recognition of
a not wholly obvious resemblance is well observed. It has no rele­
vance to the aesthetic enjoyment of a picture.
M. states (pp. 35 and 204) that av>t.>..oyl,£a8at here and at 55a7, ro
means 'infer by syllogistic reasoning', at 6rh2 no more than 'con-
. sider', and that the last occurrence is earlier than the invention ohhe
syllogism. While it is true that we have two meanings 'consider' and
'infer', there is nothing in the latter which goes beyond the common
significance of the word in Plato.

72
 COMMENT A R Y
The principle p.av8ci.vetv �8v has a bearing on the enjoyment of
literature. One of the m any elements present in the impulse to
undergo the painful experience of seeing a tragedy is probably the
desire for knowledge, knowledge of the behaviour of human beings in
extreme conditions. But there is nowhere any indication that A.
means this. In prose fiction this element, though not of the first
importance, is more pervasive. It has commonly been regarded as
a merit, however aesthetically irrelevant, if a novel gives us a vivid
picture of life in other times or places or in an unfamiliar milieu.
Further, if poetry has the universal quality ascribed to it in Ch. 9· it
should reveal to us the significance of particular experiences, so
that we should learn from literature a fuller comprehension of the
nature of life itself. But there is nowhere any hint that A. intends by
the pleasure of learning anything of this sort.
48b1 7. o�To� �KEivo� : the masc. is strange after Tl eKaaTov ; cf. 48321 for
the opposite switch. Both the passages from the R. cited on l. 13
above have TovTo £Keivo (which is G.'s reading here) in the corre­
sponding place, and so Aristoph. Ran. 1342, 'so that is what the
dream meant'. As portraiture had little place in A.'s world, the figure
recognized must in · most cases have been a mythological one, but
Alexander, for instance, might be picked out in a battle piece. The
elKov6ypacpos of 54b9 is a portrait-painter.
E. has a suggestion which would give a more real meaning to
p.av8ci.vetv. The spectator is at a zoology lecture and learns from a pic­
ture or diagram to what genus an animal belongs, 'that is a so-and­
so'. But the mention of philosophers in l. 13 is not enough to conjure up
a lecture�room, and it does not suit the passages from the R. R. G� C.
Levens objected, ]HS 81 (x¢1), 190, that TotovTos would be needed.
�1rd �civ 11-iJ Tuxn
• •: this reads like an afterthought ; if the subject
•

of a picture is not at least partly familiar, there can be no recognition,

and pleasure in the workmanship seems a lame alternative, but the
same pleasure in craftsmanship is recognized at Part. An. 645a1o (cf.
Plut. M. 673 0-4 D), which is as near as A. gets to admitting an
aesthetic pleasure. It _should be remembered that the visual arts
are not A.'s prime concern ; they are brought in merely to illustrate
a basic impulse, which led gradually to the development of poetry a!'
we know it. Pl. Laws 668 o, E speaks of the impossibility of judr; .1g
a picture of a kind of animal one has not seen.
48b20. KUTci cJluaw : cf. cf>vaucal, avp.cpVTOV above. It has already been
implied by the examples of the distribution of media among the arts
that harmony cannot exist without rhythm, so they form a natural
pair. The possession of these instincts for rhythm and harmony
distinguishes men from animals according to Pl. Laws 653 E. They
are manifest in children from birth (Probl. 920b3o).

73
 C O M M E N T AR Y
48b2 1 . JlOpLa. : 'sections'. Lines of dactyls, iambs, or trochees are
thought of as pieces cut off from a continuou� rhythmic strip ; cf. R.
14o8b29 : pv8p.or, o� Kal Tel p.lTpa Tf'Tf'Ta (Tp.�p.aTa, B.). p.lTpov often
means 'metre' in the abstract, and we have had it as Aoyor f'£Tc1
p.lTpov ; either will do here.
48b22. tca.Ta f'Ltcpov ·
. • : th� a?JToax£8taap.aTwv, 'improvisations', came
•

first ; men gradually developed them (TTpo&.yovT£s) until they reached

a stage when they became fit to be described as belonging to an art.
Though there is no reference to it, this is not incompatible with the
view attributed to Democritus that p.ovatK'Jf is the product of super­
fluity (Philodemus, De Mus. 31 ; see Koller, Mimesis (Berne, 1954),
pp. 146, 151). The slowness of development might well be due to
lack of leisure in a struggling community.
·

A major problem of the P. is whether (a) the two causes are the
natural tendency to imitate and the natural pleasure in imitations,
or (b) the tendency to imitation is one cause with two subdivisions,
and the other cause is the instinct for rhythm and melody. (a) is
favoured by the phrasing of the passage into which rhythm and
melody are introduced late and unemphatically as though of subor- ,
dinate importance. It can also be argued that A; treats the plastic
arts and pt>etry as similar forms of imitation, and that rhythm and
melody have no connexion with the plastic arts and are therefore
excluded as a main cause. (It is true that pv6p.or was sometimes
applied to objects devoid of motion or repetition, e.g. a shield (Xen . .
Mem. 3· 10. 1o) which is £vpv6p.os if it fits. As Wolf says (see 47a22 n.),
pv8p.os is here equivalent to axfip.a, as it is at Met. 985b15. Later the
critical use of the term seems to have been extended ; Aristides
Quintilianus ( ? third century A.D.) r . 13, says that pv6p.os is used in
three connexions, one of which is ETTl Taw aKw�Twv awp.aTwv, waTTEP
t/Jap.£V Evpv8p.ov av8ptaVTa.) Again, the essentials of imitation can be
effected through the medium of prose to which rhythm is not in­
dispensable. Alternative (a) is accepted by B., R., and S.
In favour of (b), the more likely, it can be argued that it is in
accord with common sense ; that Eyevv"'aav in I. 23 resumes the
yEVVfjaat of I. 4, giving A.'s final statement of his view, and that avTa
in "I. 22, whether we read 1rpor avTa (cf. Tip £vt/JvEs £lvat 1rpor a?JT�v sc.
laTptK�v Met. 1003b3), or make it the object of 1rpoayovns, must refer
to 'TOV p.tp.£ia8at Kal Tfjr app.ovlas Kal 'TOV pvOp.ov of ll. 20, 21, and that
all three must therefore be concerned with the origin of poetry, the
tendency to imitation being one cause and the instinct for melody
and rhythm the second. This is the view of G., M., and E., and of
Tyrwhitt and Vahlen among the older critics. The delayed mention
. of pv8j.Los becomes easier if we regard 48b12-19 as a parenthesis ; M.
·

and E. agree that it is.

74
 COMMENTARY
It is possible to make a guess about the way in which A . reached
his conclusions on this subject. Literature in its most highly de­
veloped form, tragedy, is the imitation of an action in a form using
rhythm and melody. Both these elements. can plausibly be traced
back to very simple beginnings. Man has perhaps had some sort of
song and dance as long as he }_las been man. A. did not know how
long ago men had begun to paint animals on the sides of their caves
(not that this practice is to be accounted for by purely mimetic
tendencies) but he rightly inferred from the habits of children that
the instinct is fundamental However, he was not altogether happy
• .

in combining the two ideas. It is far from clear that the urge to
expression present in the primitive dance has anything to do with the
visual arts, though both may have roots in magic. One may suspect
that both Plato and A. were ill served by a theory of mimesis �hich
could be applied indifferently to painting and poetry. As used in the
discussion of music in the Politics the conc'eption of mimesis is much
less inadequate ; it means there something very close to 'expression'.
But references to accurate (�Kpt{3wp.Evas) drawings of animals serve
-
only to darken counsel.

48b24-4986. The poetry of u7Tov8aiot and tfoav>..o,, hymn-epic-tragedy and

lampoon-comedy.
48b24. 8&Ea11'cui9TJ 8£ : 3E corresponds to JLfV o>..ws at 48b4· This is the
first stage after the yEv£uts of poetry out of improvisations ; poetry
was split into two streams according to the characters of the poets.
In spite of Finsler, Platon und die A. Poetik (Leipzig, 1900), p. 198,
and E., it is hard to believe that the .qBos is that of the poetry rather
than of its writers (and at 49a4), though it is quite true that we here
. revert to the idea explained in Ch. 2, the classification of poetry
according to the objects of mimesis, and this is, s�rictly speaking,
a classification of poems not of poets ; cf. KaTtl T�v olk£lav tfovutv 49a3.
48b25-26. Ka.Acis : clearly indistinguishable from U7Tov8alovs 48a2.
u£p.vos and £�n��s also mean the same in this context as u1rov8aios
and tfoav>..o s respectively. u£p.vos may have particularly appropriate
associations. Cf. T6 Uf:JL�6v ayav Kal TpaytKOV (R. I400b7), JL�TE 7T£pl
E�Tf:AWV Uf:JLVWS (>..Eyetv) (R. 14o8a13), and ironic � UEJLV� av�T} Kal
Bavp.aaT�, � Tijs Tpaycp8las 1TOl1Jats (Pl. Gorg. 502 B).
1rpcige,s see note on 49b24.
48b27. 1Jtoyous : · songs or poems of abuse balancing vp.vovs Kal lyKwp.ta,
poems in praise of gods and men respectively. A. has nothing to say
of the mature forms of hymn and encomium as written, for instance,
by Pindar. (The only Pindar mentioned in the P. is an actor.)
48b28. 1rpo ·ol'�pou : a number of poems of a religious and ritual
nature were attributed to poets supposed to be older than Homer,

75
 COMME NTARY
e.g. Orpheus, Musaeus, Olen. A . was, a t the least, doubtful of their
authenticity (Hist. An. 563a18, Gen. An. 734a19).
ToLo(lTov : this should refer to the several kinds of poem mentioned
iri the previous sentence, but it is clear from what follows that A. is
thinking only of t/Joyo' ; there is no further mention of the serious
sort of poetry until l. 33· It is astonishing that in a passage of this
sort Homer should first be mentioned in connexion with the Margites
· and comedy. It is possible that t/Joyo' are mentioned here because A.
did not wish to exclude the possibility that the supposedly pre­
Homeric vp.vo' were authentic.
4Sb30. o MupyhT)§ : this was a burlesque epic about the ludicrous ad­
ventures of a 'dumb' hero who 1ro.U' �7TlriTaTo lpya KaKws �· �7TlcriaTo
1ravTa. It is again cited by A. as Homer's at EN 1 141a14, on which
the commentator Eustratius (c. A.D. 1oso-u2o) observes : 1rapayn
. , r 1 ,0 1 1 �· , A
• • •

Ka'1 nva 7TO,"'CTW

1 Mapy'T"'V
1 ovop.a.,op.EV.,.,V 1-'"'POV. ILV"lf'OVEVn a avT"lS

ov p.ovov avTOS .M. EV -rep 'irpcfmp IlEp� IloL"lTLKfjS a.Ua l(a� .MpxD\oxos Ka�
KpaT'ivos Ka� KatUlp.axos • • Ka� p.ap-rvpoficr,v Elva' 'Op.�pov -ro 7Tol"lp.a.
•

As the jloruit of Archilochus was about 65o B.c., which is earlier

than the generally re.ceived date of the Margites, the reference should
probably be to the .MpxD\oxo' of Cratinus : see Radermacher in RE 14.
1707. The metre was hexameter irregularly interspersed with iambics,
as exemplified by POxy. xxii (1959), 2309.
Kat Tci ToLa(lTa : there were other qurlesque epics, one of which, the
Batrachomyomachia, survives in a late form. It has no resemblance
to a t/Joyos.
€v ots : in spite of the nearness of -ra -roLafi-ra the reference is to the
t/Joyo' : cf. 59b12 for the position of the relative. It is to lampoons that
the iambic metre was appropriate, while the Margites was not a true
t/J&yos, being akin to comedy (48b38). R., however, disagrees ; as no
t/Joyo' survived, their metre was unknown. Probably A. assumed that
the metre developed during the period of indefinite dur�tion when·
t/Joyo' were being composed. The oldest surviving iambics would be
· for A. those in the Margites, but it is not suggested that Homer in­
vented the metre. Though not all t/Joyo' need have been in iambics,
iap.{Jo' and t/Joyo' in this passage are not to be distinguished.
4Sb3 1 . 1jA&e : 'turned up' : cf. 7Tapa<foavElCT"lS at 49a2.
J.LETpov : A. derived 'iambic' from lap.fJlCw = Ao,Bopw ; it should be
the other way round. As the natural metre for abuse it came Ki:r.Ta
TO cipp.oT-rov and, though it had long ceased to be restricted to this
use, the name remained. (The metre is always lap.{JE'iov ; lap.{J,Kos
means 'abusive', and iap.{Jos can include abusive lines in trochaic
tetrameters (R. 1418b28) ; see K. J. Dover, Entretiens Fondation
Hardt, 10 (1963), p. 186.) .
4Sb32. lG.J.LPLtov ciAA..jAous : this implies abusive exchanges of a ritual
 COMMENTARY
nature such as we know to have been common in many societies :
cf. Herod. 5· 83. 3, and LSJ s.v. y£cpvpl,w. This piece of information
does not seem much at home in its context. For another derivation,
from Iambe, see Hom., H. Dem. 195-204.
48b33. Twv 11'a.>..cnwv : at 53b27 oi 1ra>.mol are extant poets ; here logical
sequence rather than chronology is the subject. There was a long
period when these were the two forms of poetry practised, epic and
iambic. Archilochus is not mentioned as the iambic poet, but his
name is to be supplied mentally. Transitions from an earlier to
a later form, e.g. from hymn to epic, are not under consideration.
48b34-38. wa11'Ep . . . �pa:f.la.To11'ot�aa.�,: the genius of Homer tran­
scends ordinary limitations, ·and he embraces both c17Tov8alous and
cpav>tous. Actually tragic ·poets also wrote satyr-plays, though never
comedies ; cf. Pl. Symp. 223 D.
Tel. a1!'ou�a.ia. . . . 11'0L'1T�� : cf. Xen. Cyr. 3· 3· 39 : E7Ttc1T�JLovfs Ta
.

1TpOC1�KOV'Ta, .
48b35. �pa.f.lG.TLKo� : the distinction between narrative and impersona­
tion was made in Ch. 3, and Homer was praised for his use of im­
personation, but in view of 8paJLaTo1Tot�CTas below it must mean more
than this, as at 59a19, both more unified in structure and more
generalized in significance. Thus Homer developed the forms in the
direction of tragedy and comedy, but epic and iambic long continued
to be the main forms. That VJLVOt and other lyric poetry continued to
be composed in the post-Homeric period is probably implied by
49a1o-I4.
48b36. axfif.la. . . . U11'E8EL�Ev : Homer indicated the outlines of the
emerging form of comedy. CTX�JLa (cf. 49a6 and 49b3) implies the
structure, the 'set-up', of comedy, hardJy to be distinguished from
£l8os. But it is excessively difficult to believe that the Margites ap­
proached comic form in any respect other than the use of direct
speech. Conceivably the episodes in which the hero was involved
may have had something in common with incidents in the Dorian
· comedy of Epicharmus, who is rio doubt the next in the line of
developtp.ent. In the terms of Ch. 9 Margites would mark an advance
as being a more universal, i.e. typical, figure than the object of
lampoon.
48b37. lftoyov . . . yE>..oiov : the difference is more fully indicated at
49a34-37. Whereas t/J6yos is essentially vituperative, comedy dis­
plays the ridiculous without malevolence, though this is not always ·

true of the Attic Old Comedy. · .

48b38. o ycl.p Ma.pyh'l� . . . : the kinship between Homer's epics and

tragedy was widely recognized. Cf. Pl. Theaet. 152 E : oi aKpot Tijs
,
1TOt7JC1€WS •
fKaT€pas, 11 '
KWJLtpaLaS 11 '
"JLfV ;E7TLXapJLOS, TpaytpaLaS 11 ' ·oJLTJP
Of
, , ,
OS
and the passages cited by G., p. 109. The suggestion of a similar

77
 COMMENTARY
relationship between the Margite$ and comedy was probably novel.
Homer must have marked out the axfip.a. of tragedy too, but in this
case the subject remained unchanged, TO a1rov�a.iov.
49a2. 1ra.pa.cl»a.vdC711 � 8€ T'ij� Tpa.y'tl8la.� : at one level the realization. of
the potential already present in poetry from the beginning is spoken
of as a natural process, at another (49a15 ff.) innovations made on the
initiative of individuals are regarde9 as an important part of this
process.
49a4. otKda.v : see on 48b24.
K(I)J1'tl8o'IToLoL
• • Tpay'tJ8o8L80.aKa)(o,: there seems no reason for
•

this elegant variation. ��Mmwv means to produce a play, and A.

uses �ewp.<pl>o��a&.a�ea.>..o � (EE 1230bi9), without any difference in
·

meaning. The transition avT� Twv e1rwv Tpa.y<p�o�,a&.a�ea>..o, is consis­

tent with 49a19-25 c;mly if epic poets took over tragedy ready-made.
49a6. JJELt(l) Ka.l. EVTLJlOTEpa : 'grander and more estimable'. The
superiority of tragedy to epic as a form is explained in Ch. 26. But it
was only at 'a late stage', 1. 20� that tragedy became aep.vos . and ol
c11TovSa.io' then abandoned epic for tragedy.
This concludes the scheme of the development of the various poetic
forms, which is followed by a fuller discussion of tragedy and comedy.
The confused impression which is left by this middle section of Ch. 4
is due partly to compression, partly to the introduction of Homer,
an historical figure, into what seems to be an account of logical
rather than of historical development, and to the attempt to connect
him with comedy as well as tragedy. Arising out of improvisations
we have the two sequences, hymns and encomia-epic-tragedy, and
lampoons-comedy ; symmetry could be improved by inserting
iambic poetry between lampoons · and comedy, making three of
each ; but though the original lampoons .may not have been in iam­
bics, t/Joyo� and ta.p.{Jo' are essentially the same. The effect of Homer's
genius was to bring both existing forms nearer to the ultimate goal.
This is easily miderstood as · regards epic, but it is hard to believe
that the Margites had mJich in common with a t/Joyos. · No doubt, as
being more humane than a 1/Joyos and less concerned with the foibles
of a particular individual, it bore some resemblance to later comedy.
After Homer, epic (which would include Hesiod) and iambic poetry
continued to be written much as before, but meantime new forms
were developing to the point where Epicharmus could turn to comedy
and Thespi� followed by Aeschylus to tragedy. It is not suggested
that tragedy developed from epic, only that it was its s.pirit1,1al suc­
cessor. According to what follows tragedy in fact developed from
performances which were mainly ludicrous, 49a19. .
We now turn to tragedy and comedy, which are considered
together in their most rudimentary stage, 49a9-14 ; thereafter tragedy

18
 C O M M E NT A R Y

is the subject for the remainder ofCh. 4, 49314-31, and comedy for the
first half of Ch. 5, 49332-b9·
4987-9. A. first asks whether the development is complete. ci.Uos
.\oyos implies that he will return to the subject, but we do not kriow
that he ever did. He appears in fact to answer his question in the
affirmative a few lines later, 49315, but see note ad loc.
49as. d8Ecnv : usually taken as 'constituent elements' = pipTJ (cf.
50313 and 56333), but in neither case is the text beyond doubt. B.
declares this use unexampled except in Plat,o. G. says that it means
the same as ax�fLaTa. It could also refer to 'types' of tragedy (cf.
55b32, 59b8). The question is whether further development is possible.
Further development of the fLEpTJ seems excluded by 49315. E. makes
E(8eatv depend on iKavws ; cf. Pol. 1318b 25 : iKavws £xn To is 1TOAAois,
·

meaning 'is adequate to the basic form'.

1rpos Tel. 8£aTpa: the natural realization by tragedy of its poten­
tialities might be impeded by accidental fact<:>rs like the require­
ments of dramatic festivals-as it might be by dinner and licensing
hours ; cf. 51a6, the length of plays is governed by festival arrange­
ments. The quality of the audience too might be relevant ; cf.. Plato
on their 8eaTpoKpaTla (Laws 659 B, 701 A).
4989-3 1 . Origin and development of tragedy . Every phrase in this pas­
sage has been the subject of controversy. For a general account of
the subject the reader is referred to A. W Pickard-Cambridge's
• .

Dithyramb, Tragedy, and Comedy. Recent work on the subject is

summarized in Fifty years of Classical Scholarship, ed. Platnauer
(Oxford, 1 955), (that on tragedy by Professor Webster, on comed y
by Professor Dover), and more recently by Professor A._ Lesky,
Die 'frag. Dichtung der Hellenen2 (Gottingen, 1964). A full his-
. torical surv:ey of the controversy is given by C. del Grande in
TPATD..:HA2 (Naples, 1962). The latest studies are Die Anfange der
gr. Trag. by H. Patzer (Wiesbaden, 1962), and The Origin and Early
Form of Gk. Trag. by G. F. Else (Harvard and London, 1¢6).
The main point at issue is whether A. is writing from knowledge
of a genuine tradition or propounding a scheme based on his own
guesses and inferences. A. lived some two centuries after the estab�
lishment of the tragic contests. The official records, which he used for
his Didascali�e, began about 501 B.c., and it is unlikely that any
earlier evidence of comparable quality was available. No doubt
some information based on tradition was preserved by fifth-century
writers, information like that given by Herodotus about the 'tragic'
choruses at Sicyon in the early sixth century. A Glaucus of Rhegium,
who is quoted in the Argument to the Persae, wrote Ilep2 TWV apxalwv
. IlotTJTWv Ka2 MovatKwv probably a:bout 415 B.c., and some facts of
literary history were preserved by the Atthidographers, of whom the

79
 COMMENTARY
earliest was Hellanicus of about the same date. It is extremely
doubtful whether A. possessed any detailed information about the
development of tragedy before its introduction at the Dionysia in
or around 534 B.C.
On the other hand, and this is sometimes forgotten, A. had access ·
to an immense amount of literature which is lost to us ; he knew
vh-tually all Aeschylus and some of Phrynichus, though probably
nothing of Thespis ·; also he was doubtless much better supplied
than we are with early satyr-plays. Thus he was in a good position
to extrapolate backwards from mature through early tragedy to
whatever preceded tragedy ; and if it appeared to him that the
answer was some form of choral lyric, this must cai:ry great weight.
It is uncertain how much of early dithyramb survived ; Arion was
the key figure here, but as no quotation from his works remains, it is ·

likely that they were early lost.

Finally, it is worth observing that ·A.'s account of the origin of
·tragedy from a basically ludicrous form fits so badly with the scheme
of development presented in the first part of the chapter that he
would not have been likely to offer it unless he had been reasonably
confident that it was true. . Here again it is important that he knew
more than we do about the early satyr-play. But that he was ready .
to state inferences as facts is proved by 59b3r on the metre of epic.
49a9, l1.1r' cipxi)s QU1'0aXE8,aa1''KTJS : this resumes from aVTOUX£S£aap.aTWV
48b23, the stage of vp.vo£ and 1/Joyo£, Whether A. thought the primi­
tive dithyramb was pte-Homeric we cannot tell.
49a 1 1 . ,.c;,v €gapxov1'CaJV ,.bv 8,8upaJ.1J3ov : one would suppose that the
dithyramb is a form of vp.vos and belongs, in contrast to the phallic
songs, to the poetry of the a11ouSaro, ; but there are difficulties. The
dithyramb is the original song of Dionysus, for its development
see below. Originally, or at an early stage in its development, the
dithyrambic singers would be led by ah exarchon ; cf. the chorus of
mourners in Il. 24. 720. Archilochus (primarily an iambic poet) claims
(fr. 77) 'I know how to lead off the fair song of the lord Dionysus, the
dithyramb, when my wits are blasted with wine'.
< A ' -\ ' 'i: ' l: 1\
WS LI£WVUUO£ •
aVaiCTOS ICW\OV
"
E<:,ap.,a£ f'€AOS
.
olSa SdJVpap.f3ov' otvcp eur�eEpauvw8E'ls tf>pEvas.
This must mean more than that he knew the words, rather that he
knew how to improvise, or to sing new words which appeared to be
improvised on the spot, while the chorus answered with a traditional
refrain. There is no doubt that this is how the word was used by A.'s
contemporary, Heracleides of Cumae, (ap. Athen. I4SD) : �ea2 ifl&.>.>.oua£v
a[ 11aAAa�eal ,avT<p (sc. {3aa£AEf) �eal p.la p.ev £g&.pxn, ai S' a>.>.a, a8pows
t18oua£. The exarchon is separate from the chorus, though still
So
 COM M ENTARY

attached to it ; thus he is the first stage in the development towards

an independent actor. .
ci11'o TC,v Tn cJ!uAAlKci (E�upxovTc.»v) : the phallus was conspicuous in
the cult of Dionysus. At the komos (48a3j), which was · part of the
Dionysia, phalloi were contributed by Athenian settlers overseas,
JG iz 46, cf. Plut. M. 527 D. The accompanying songs were full of the
aloxpo>..oyta appropriate to a fertility ritual. In Aristoph. Ach. 24!-79
Dicaeopolis celebrates his own Dionysia including a phallic proces­
sion and song. Sicyon had its ,Pa>..>..o,P6pot and other towns had similar
bodies ; cf. A then. 621 D-622 D, and 445 A where he speaks of an
Antheus of Lindos who composed songs a J�'9PX£ Tors p.€8' avTov
,Pa>..>.o. ,Popovotv. The word t/J6yos would cover a good deal of the opera­
tions of such bodies, but A. does not seem to have been aware of the
connexion of cult and ritual with early art. It is to be remembered
that the chorus of Attic comedy, unlike the actors, did not wear the
phallus.
o ;g&.pxwv is properly a pr. part. and was so used in the ancient
world. The noun leapxos is attested in Homer, Il. 24. 721, for the
leader of a chorus of mourners, and in Eur. Bacc. 140 and Demos. 18.
260 for one of 'enthusiastic' celebrants not far removed from the
. early singers of the dithyramb.
49a.12. ETl tcul. vuv : as opposed to the dithyramb which was completely
changed from its original form ; cf. Prob. 918b19.
49814. +uvEpov : cf. 7rapa,Pav£lof1S 4982. The development of rhetoric is
described almost in the sam_e terms, Soph. El. 183b17-32.
· I'ETuJJoAcis : the only hint as to the nature of the changes which
filled the long gap between primitive dithyramb and the invention
by Aeschylus of the second actor, 1. 16 below, is given in ll. 19-28.
To this we can add Thespis' introduction of the first actor, see l. r6 n.
Between Archilochus and Thespis a real contribution. was made at
Corinth c . 6oo by Arion, who gave the dithyramb its literary' form.
This is implied by Herod. r . 23, and A. appears to accept this (fr.
677). The further statement in the Suda that Arion invented · the
Tpayuc/Js Tp61ros and introduced speaking satyrs may be derived
ultimately from A. himself, in the latter case from a misunderstand­
ing of his words, since speech cannot be earlier than Thespis on A.'s
hypothesi!). The Tpay,Kos Tp61ros probably refers to the style of
music. At all events the important and perplexing passage,
Herod. 5 · 67, which tells of the Tpay,Kol x6po' at Sicyon, originally in
honour of Adrastus and transferred by Cleisthenes to Dionysus in .
the early sixth century, shows that a form described as Tpay,K6s had
long been in existence. For this reason the reference to tragedy as the
invention of Arion, said to have been contained in Solon's elegies
(loannes Diaconus, Comm. in Hermogenem, Rabe in Rh. Mus. 63
81
 COMMENTARY .

(19o8), rso), need not refer to anything dramatic. The whole problem
is complicated by the uncertain relation of satyrs to dithyramb : see
below.
498 1 5. E11'a.uaa.To,•• +uaw : the tragic form, like an organic growth,
•

develops until it reaches its TEAos, when its potentiality is fully

realized ; cf. Ph. 193a36. This seems to imply that there was no
important change later than the early plays of Sophocles, and that
no further development is to be looked for. Vahlen and G. argue
that 49a7 £l apa EX€' Tjl)'T/ � -rpaycp8{a 'TO iS £i8£UtV iKaVWS is SUfficient tO
prove the contrary. On the other side, and more convincingly, S.

I .,L A_ I
cites Pol. 1252b32-34 : otov yd.p £Kaa-rov £anv -rfjs y£11Ea£ws T£A£a8dUT�s,
\ \ t' I
TaVT'T/V 't'ap,£V 'T'T/V 't'VUW £tVat
t
£Kll.UTOV,
4981 6. u11'oKpLTwv : 'actors' ; it has long been disputed whether the
actor was called a tJ7ToKptrr]s because he answered, or because he
interpreted and expounded. The first is plausible because it may
have been an original function of the actor to answer the questions
of the chortis·leader about what was happening off-stage. But one of
the actor's main tasks was to speak the prologue (see next note), and
then he was answering no one. Hence G. renders vTToKpm]s 'speaker',
a meaning for which there is no warrant elsewhere. E. keeps the
sense of answerer, but denies the title to the first actor, who was
originally the poet himself. The second actor might reasonably be
called the answerer of the first, but then we encounter difficulties
about the number of actors. B. adopts the meaning 'interpreter' in
the sense that the actor is the poet's spokesman, but the poet needed
no spokesman so long as he was himself the. actor, and the word
probably goes back to this period ; cf. Pindar fr. 14ob (Snell), 125
(Bowra), and Page in CR N.S. 6 (1956), 191. While certainty is im­
possible the most satisfactory suggestion is that the actor expounds
the situation to chorus and audience, especially in the prologue, the
speaking of which was one of his earliest functions ; see A. Lesky in
Studi in onore U.E. Paoli (Florence, 1955), p. 469.
1r.Mj8os : as at 28 below it means no more than apt8p,&s. It is strange
that the invention of the first actor, or the transformation of the
exarchon into an actor, is not mentioned. It appears that A. ac­
cepted the tradition that this was due to· Thespis ; cf. Themistius,
0r. 261 31 6 D ! OV '1Tp0U£XOP,£V n.. on TO P,£V 'ITpWTOV 0 xopos £tUtWV ua£V
' ' •
A

'<I: A
" ' ' A ' ' ' ' .t �
' ' £II ' '\ A
' 1\
£tS -rovs 8£ovs, u£U7rtS a£ wpoAoyov 'T£ Kat. P'T/Utv £s£Vp£v, A tUXVAOS a£
�'
' � ' ' ' •

Tpl-rov vwoKptT�v • (this frag. is rejected, without much reason, by

• •

Rose, A. Pseudepigraphus, p. 79). A. may well have thought the

introduction of the second actor even more important. The chorus
is only rarely so far a corporate person that it can engage in genuine
conflict, as does the Chorus of the Supplices with Pelasgus and
the Egyptian Herald. For the confrontation of Clytemnestra by

82
 COMMENTARY
Agamemnon or Orestes actors are essential. It was the introduction
of the· second actor that opened the way for true drama.
4981 7. Ta Tou xopou 1]XuTTc.Jae: no example of a �ingle-actor play has
survived, but the chorus would obviously be dominant, as it is in the
Supplices, the play in which it comes nearest to being an actor, even
more than in the Eumenides.
Tov Myov : the part spoken and not sung . .
Did Aeschylus, in addition to making the chorus less important,
reduce its size ? The dithyrambic chorus numbered so, and according
to Pollux, 4· no the tragic chorus was composed of the same number
down to the Eumenides (4S8 B.c.). That Aeschylus used a chorus of
so, at least in his early period, was believed by many when it was
still accepted that the Supplices, with its two sets of so cousins, was
an early play ; but with the dating of the Supplices about 463 B.C. the
. theory has lost favour, though still maintained by A. Fitton Brown ,
(CR N.S. 7 (19s7), 1). The st:itement in the 'Life' of Sophocles, 4, that
he raised the number of choreutae from 12 to rs receives some sup­
port from the twelve couplets spoken by the ChoiVS at Ag. 1348-71 .
There is no external evidence, apart from Pollux's assertion, that the
· tragic chorus originally. consisted of more than 12.
49•1 8 . 'II'Pc.JTayc.JvtaTEiv : the leading actor is called an aywvta"r�s as
taking part in the ciywv. Pickard-Cambridge, Festivals, pp. 133-6, is
probably right in taking it as a general term for 'playing first fiddle'
sometimes applied metaphorically to an actor ; this is not a normal
use until late (cf, Pol. 1338b3o). With the reading 1TpwTaywvtaT�v the
· metaphor becomes more violent and the constr. dubious : see R.
. Kassel in Rh ; Mus. ros (1962), II7-I9 . .
Tpeis 8£ . . : these are jottings rather than continuous prose ; there
.

is no suitable verb to supply from the previous sentence.

The introduction of the t�ird act9r was attributed by some to
Aeschylus. The ancient 'Life' claims it for him but adds that
Dicaearchus of Messene (a 'pupil' of A.) made Sophocles responsible.
Aeschylus used three actors in the Oresteia of 4S8 B.C. ; Sophocles
first competed in 468. The confusion probably arose because the
innovation took place during the period when both poets were active.
Did A. mention both versions ? Cf. Themistius quoted on l. r6, above.
aKT)Voypacjl£av : 'scene-painting', which made use of the newly dis­
covered knowledge of perspective. According to Vitruvius 7, praef.
n , this was due to the painter Agatharchus, and the initiative came
from Aeschylus. The 'Life' speaks of his attention to stage decor.
On the whole the Greeks, like the Elizabethans, seem to have
achieved their spectacular effects by splendid costumes rather than
by elaborate scenery : cf. Webster, Greek Theatre Production, pp.
13 ff. See, on oifJts, 49b33 n.
 COMMENTARY (4. 4 9!' I g-
4981 9-31 . This passage contains some of the most indigestible matter
in the P. The general intention �eems to be as follows : tragedy as it
first develope� from the dithyramb was rather trivial and in the style of
the satyr-play. Only at a late stage did it acquire dignity, and one of the
accompanying changes was that of metre from . trochaic tetrameter to
iambic trimeter; also the number of episodes., and so the length .of the
plays, was increased. Tliere are a number of incidental obscurities in
this account and there are grave difficulties in reconciling the ac­
count with the few known facts. So long as it was widely accepted
that Tpaycp8la meant 'song of goats' and that satyr-plays had goat­
choruses the theory seemed to receive some confirmation, but it is
now generally believed that the satyrs of the theatre were horse­
rather than goat�men.
498 1 9. tJEyE8os : like 'greatness' in English this can refer both
to physical length and to grandeur of content ; the same ambiguity
extends to p.u<pwv p.tJOwv, 'short', or .'trivial'. yE>.olas and aaTvptKov
(I. 20) imply lack of dignity, £7TEtao8lwv '7T>.�87J (1. 28}, that the origi­
nal plays were short. ·
};.igEws : covers both choice and use of words�
49820. �K aa:rupLKou : this may mean that tragedy, in A.'s view, de-
. veloped from the satyr-play. A connexion between the two forms is
suggested by the fact that satyr-plays were produced in conjunction
with tragedies. This is the only reference to the form by A., if re­
ference it is. But in view of aaTvptK�v • • • .,j.ol7Jaw at I. 22 it is more
likely that a form akin to the satyr-play is intended, which could
well use a 'ludicrous diction' and be associated with lively dancing,
dpx7JanKwT£pav (1. 23). The addiction of satyrs to the dance is easily
illustrated from vase-paintings (see end of next note). ·
The major difficulty is in combining this satyric stage of tragedy
with its origin in the dithyramb and its subsequent development.
The account here given implies that the dithyramb was a ludicrous
form with a chorus of satyrs. The contrast with the phallic komos at
49a1o implies the opposite, and there is no evidence of anything
satyric in what we know of the early dithyramb-which is not much.
However, Webster thinks he has found evidence for a dithyramb
danced by satyrs at the Panathenaea in a fragmentary vase-painting :
see his revision of Pickard-Cambridge, Dithyramb, pp. 34 and ¢,
and Plate I (a). ·
ovi : 'late', with reference to what ? Presumably to the beginning
of tragedy, which cannot be placed much before the middle of the
sixth century, if it was Thespis who introduced the first actor. The
form must have been established some time before Peisistratus gave
· it official recognition at the Dionysia within a few years of 534
B.C. At R. 1403b23 difl£ means early in the career of Sophocles (when
 COMMENTARY
the poet ceased to act in his own plays), at 1403b36, not long before
the time of writing ; at 49b2 oi/JE 1roTf: means 486 B.c., when comedy
was officially recognized some fifty years later than tragedy. Tragedy
appears to have received an access of grandeur, a1reaep.vvv81J (cf. oi
aep.vo-rt:;pot 48b25), when, or soon after, Aeschylus began to compete. ·
He is hailed in the Frogs 1004 as the creator of tragedy :

. &..\.\' cL 1rpw-ros -rwv 'EA.\�vwv 1rvpywaas Mp.a-ra aep.va­

Kai Koap.�aas -rpaytKov Mjpov.
and the lines are quoted at the begmning of the 'Life', which im­
plies that they had become part of the raw material for the history of
tragedy. Cf. also Philostratus Vit. Soph. 9· No one suggested that
he was literally the inventor, but he could be regarded by those with
better evidence than we have as its second founder� Tragedy must
have ceased to be satyric at latest by 492 B.C. (?) when Phrynichus'
play on the capture of Miletus reduced the theatre to tears, seven
or eight years after the first production of Aeschylus. This, if not
obviously 'late', does not seem entirely out of scale with A.'s other
uses of the word, but the time allowed for the changes· is short.
According to a widely believed tradition the satyr-play proper
with its chorus of men with horses' tails was introduced from Phlius
near Corinth by Pratinas about 490 B.C. If A. accepted this, his use
of aa-rvptKos would distinguish the primitive drama from the im­
ported satyr-play. On the other hand, the abundance of satyrs on
vases of the late sixth century suggests a date nearer to 515 B.C. for
the establishment of satyric drama in: Attica, whether due to Pratinas
or not. On satyrs see F. Brommer, Satyroi (Wiirzburg, 1937),
Satyrspiele (Berlin, 1959), E. Buschor, 'Satyrtanze und friihes
Drama', Sitzb. Munch. Phil.-hist. 1943/5.
49•: u . TETp«f1ETpou : that is to say the trochaic tetrameter, a metre .
which is prominent in the Persae and used at the close of the Agamem­
flon (later it was increasingly employed by Euripides in the HF and
subsequent plays). · The use in the Persae may be connected with t)le
employment of the metre by Phrynichus, described in the Suda as
evpETTJS Tov TETpap.E-rpov, with whose Phoenissae the Persae was said
to stand in a close relationship ; cf. Argument to Persae. For the
transition from tetrameter to iambic see R. 1404a31. Iambic tetra-
·

meters occur in Soph. Ichn., a satyr-play.

49•23. opxTJcn&tcWTEp«v : we should expect this to mean _'suggestive of
the dance', as at R. 14o8b36, where in a discussion of prose rhythms
the trochaic is called Kop8aKtKWTepos ; it can hardly mean 'suitable
for dancing to', since tetrameters were normally spoken or intoned,
and were not tised in the strophic systems to which the chorus sang
and danced both dithyramb and tragedy. The general meaning

ss
 CO M M E N T A R Y
must be that, when spoken verse first came in with . Thespis, they
used the trochaic metre because it was in keeping with the tone of
the old tragedy, which was light and close to the satyr-play. But we
hear of dancing to tetrameters in comedy (schol. Aristoph. Nub.
1352) and of recitation to the flute (Xen. Symp. 6. 3). See P. Maa�,
Greek Metre, trans. Lloyd-Jones (Oxford, 1962), 73-77.
).£�Ews 8£ YEVOJlEY'lS : 'When dialogue had come in', (Kassel's ly-
. yEvop.EV'TJS would be even clearer) ; MeEws, speaking as opposed to
singing, is hardly distinguishable from .\6yov in 1. 1 7, above. The
beginning of the change described in yEvop.EvTJs is not subsequent to
-ro p.ev yap Trpwrov. There we�e no (tragic) tetrameters till there was
Me,s, but as Me's became more important the natural tendency to
use iambic rhythm in speech ensured that the trochaic tetrameter ·

was gradually ousted by the iambic trimeter.

49824. TO OLKELOY JlETpov : cf. S9al l, R. 14o8b32.
E�pE : one may agree that this is not historical, see A. M. Dale
CQ l:'l.S. 13 (1963), 48, n. 2, without allowing that EVpE must mean
'invent'. If A. believed that the Margiteswas by Homer, it is unlikely
that he was so · careless as to suggest that the iambic was invented
in the sixth cent.
49828. O.pJ1ovLo.s : is said to refer to the pitch of the voice used by the
Greeks in conversation (cf. R. 1403b31). · We should have expected
rather a reference to rhythm.
E11'ELao8Lwv : probably in the sense defined at 52b2o, equivalent to
an act. If plays grew longer, the acts would tend to be more
numerous. In fact, in extant tragedy the tendency is for the episodia
to become not more numerous but longer. The word appears else­
where in a different sense, 'incident' (see 55b1 n.). This wquld give
the same meaning here ; a longer play would use more incidents. It is
convenient to keep the Greek form epeisodion for the technical sense,
a section of a play.
·

.
·
.

1rA1}9t) : sc. lylvE-ro V., or perhaps a general sense of augmentation

from TJJ�1]8T} • • a'TTEUEp.VvV8TJ
• •

.,a, li.AA' : A. could have given fuller information on the use of

trilogies, on the development of the theatre and stage equipment,
and on the features mentioned in connexion with comedy at 49b4.

CHAPTER 5
Comedy, and the relation of epic to tragedy.
49•32-49b9. Comedy.
49832. wawEp Ef11'0J1EY: this has not been said in so many words.
References back may be to particular statements or to the gist of

86
 C O MMENTARY

a longer passage with no preCise point of reference. An example of

the former iS 48a25, Where WS £i1TOf'EV referS tO the three media
-rptalv of 47ar6. Here the reference is to the general distinction
between tfoav>..ot and a1rov8aiot as objects of imitation. That comedy is
,_dp:qats tfoav>..w v is indicated by Ch. 2, especially 48ar6, and the p1en­
tion of Aristophanes (48a27), and by 48b26. E. discusses the references
in the P. in his note on 6oa3, p. 615 ; cf. Vahlen, p. 259· His distinc-
-tion between WS referring back to a particular statement and W(11T£p,
· o1T£p, etc. meaning Ka-rtl -rov £ip1Jp.Evov -rpo1rov is too clear-cut ; o1r£p
54a9 is a precise reference to 53a19 ; wa1r£p eMxOTJ 54a18 to sobS, 9 ; cL
EN 1104a2 with reference to I094b13. The important point is that
many references are of the second type and it is vain to search for an
exactly equivalent statement. ·
· Objection has been taken to the occurrence here of a premature
definition of comedy. The definition� of tragedy comes after the
history of its development and at the beginning of the section of
which it is the subject ; there must have been a further definition of
comedy in the corresponding position. The reason for the pres�nce
of the five lines 32-37 is perhaps that A., referring to the classifica­
tion by objects in Ch. 2, felt it desirable to point out that, while
tragedy imitated actions of a1rov8aiot in general, comedy was
concerned only with a particular group of t/Jav>..o t. At the same time
he added an explanation of the difference between 1/Joyos and -ro
· yE>..o rov with reference to 48b37, unnecessary perhaps, but not un­
natural, since his subject is still the development of poetry from
· improvisations, not of one form alone. Vahlen, who suggested that
49a32-37 should be moved to after 49b2o, thought that A. was in­
dicating why he chose to treat as a pair tragedy and epic, which are
both serious, rather than tragedy and comedy, which are both
dramatic-an interesting point, but there is nothing to suggest that
A. had it in mind here.
4911:33. Ku�eluv : the condition of the t/Jav>..os, who has not been bred to
aspire to d.pE-r�, but it covers wickedne�s as well as insensitivitY, or
tastelessness. In R. 1383b19-"84a4 subheadings of KaKla are 8nAla, .

a8£K{a, aKo>..aa{a, alaxpoKep8na1 avE>._Ev(J£p{a, Ko>..aKEla, and p.a>..aK{a,

6.).).6.: the sequence p.� • ov p.ev-rot • • • &.A>..a is very elliptical ; the
• •

ci.Ua clause should explain the emphasis on 1raaav, not all' KaKla but
one sort. Friedrich's a.,u� v or the -rofi alaxpofi ov of one of the re-
centiores would make things easier. · .
1'ou ulaxpou : like KMos in both moral and aesthetic senses ; it
means 'ugly' at 6ra13. KaKos·too can have the aesthetic r.eference,
but not, apparently, KaKla. Ugliness is as incompatible with con­
ventional apET� as is baseneSS.
49•34. ,.c, yE).oiov : the -re>..os of comedy, as pity and faar are of tragedy.
 G O M M E N TARY
GJ.LUP'"ll'a : the- ludicrous error might be anything from confusion
between identical twins to falling' into a well,. and in particular un­
awareness of one's own weaknesses. For the distinction between
ap.apTT/I.I.a and ap.apTla see Appendix IV, P· 3dO. It is natural to
assume some correspondence between comic hamartema and tragic
hamartia, though it should be remembered that there is no sign that
A. regarded hamartia as, what we have made it, a sort of technical
term. But the explanation is to be foun:d rather in the Platonic
theory put forward in the Philebus 47 B-so A. Plato, in discussing
mixed pleasures and pains, accepts tragic pleasure as an obvious
fact oTav ap.a xalpovTES Ki\awut 48 A ; comedy offers malicious enjoy­
ment through the spectacle of those deficient in self-knowledge

,/,
(ayvota 48 c) and the ridiculous consequences which follow from
\ �

� � /: I y
0 �· oOsa!.OVTES:
exaggerated -seIf-estee.m 'ITI\E£UTot ')'E 7TEpt
I \
To\ Twv
� t �

Q \ I
Ev Tats 'f'vxats�

I
E,taVTOVS1
I 0 "
• • •

O£f/p.apT7]KaO'£V1 apET'l/
·
t'EI\TWVS OVI( OVTES.
49a35, Kai. atvxos : Kat is explanatory, i.e. it is not the sort of hamartetna
in which a superior character would find himself involved.
civw8uvov Kai. ou cl»8apT&Kov : &8Vvv and rf>8apn�<6s are strong words
which imply violent suffering and danger to life ; cf. tragic 1ra8os
(52bn), 1rpa.�,s rf>8apnK-ft 1j &8vV7]pa which contains deaths, woundings,
and sce.nes ()f physical agony. Plato required (Phileb. 49 E) that the
comic ayvota should be &.{J>..af31]s. That which in tragedy makes a
direct appeal to the emotions is the opposite of what is appropriate
in comedy. Further examples of the &8vv11pa Kal rf>8apn�<a are to be
found in R. 1386a7. There is a limit to the amount of suffering that
cap. be portrayed if a comedy is not to leave a bad taste in the
mouth. A play in which no one suffers at all is unlikely to be
dramatic; The amount of suffering which an audience wi�l take
depends on t�e degree of realism or fantasy with w�ich misfortune is
presented and on the strength of its stomach. Some today find the
- humiliation of a Malvolio offensive, though it does not seem to
have troubled the Elizabethans. It is hard to say what degree of
affliction for unsympathetic characters was acceptable to the Greeks,
but we have it from A. himself that it was in a�cord with the spirit
of comedy when a?To8vz]uKE£ ov8Els V'IT0 ov8Evos 53a38.
49a36. Ell8us : 'immediately', in the sense that the example is instantly
available and does not , have to be searched for. With a slightly
· different application at 52a14 ; cf. the Attic use of avTlKa 'for ex-
· -

ample'.
'11'pOaW1TOV ! 'mask'. Many comic masks were grotesque, but though
the face was twisted out of the normal to give the desired grimace, it
- did not suggest pain. The masks themselves have perished but
numerous specimens in clay and representations in works of art have
survived ; see T. B. L. Webster, Greek Theatre Production, and

88
 COMMENTARY
M . Bieber, History of the Greek and Roman Theatre (Princeton,
1961). A large number of types of mask are listed in Pollux's
Lexicon, 4, 143-54.
49837. J&ETa.PucrEtS : 'changes' undergone by tragedy = p.ETa{jo�ai 49•I4.
This sense of the word is to be distinguished from the changes of
fortune experienced by the characters .of a play, the meaning else­
where in the P., 52•16, 18, 55b27, 29.
49838. ou "E"'I\8a.crw : the sort of developments of tragedy which have
not been forgotten are the first use of masks, prologue, etc. men­
tionea at 49b4, below. Prologues at , least were not earlier than
Thespis, and these words should not be quoted to show that A.
claimed to possess fuller knowledge about the origins of tragedy
than of comedy.
49b l . cr11'ou8utEcr8a.t: in act. the opposite of wai,ELv. Here 'to be taken
seriously' as at R. 1380326 : 0'1TOV8&.,eu8at &).).' ov KO,'Tatf>povEiu8aL,
obviously in contrast to tragedy, which was taken seriously from
the beginning. Comedy too was taken seriously and the facts about
it recorded after it was recognized, but it was recognized at a later
stage of its development than tragedy.
Ka.l yap . . . E8E"oVTa.l �cra.v : the poet who wished to compete at
a dramatic festival asked the competent archon for a chorus, which
would be trained and equipped at the expense of the officially ap­
pointed choregus. How the archon decided between poets when there
were more poets than places we are not told. The earliest mention of
the practice, Cratinus, fr. IS : OS ov8' E8wK' al'TOVV'TL Iotf,oKM�t xop&v,
shows that the problem could be real. Until comedy was recognized
and competition invited, it was no good asking the archon for a
comic chorus, and performances of comedy must have been organ­
ized by private, not necessarily individual, initiative. Such is the
force of £8E�oVTal, 'volunteers', which was also the name of those who
gave comic performances at Thebes, corresponding to the tf>�otf>&pot
.
at Sicyon, etc. ; see note on 49•II.
According to the most natural use of words, which is not alw!\ys
that ofthe P., the £8E�oVTal should be the same persons as the KwfUP8wv.
So far as usage goes, it seems that KwfUPa&s can mean a comic poet,
a comic actor, or a singer in a comic chorus or Kwp.os, the primary
meaning of the word. Thus the Kwp.q18wv of the MSS., 'a chorus of
comic singers', gives adequate sense, though G. objects to it as being
a tautology. Some support for the view that it was the poet or poets
(reading KwfUP8cp B., or Kw�ois Bernhardy) who volunteered the
requisite effort and expenditure may be found in Eustathius on Il.

Q \ �· I It • �
'Ta\ '1TaV'T9-
10. 230 : £Ka�oiJvTo 8� Ka� £8E�OV'Ta� 8LMaK�ot, 8pap.aTwv 811�a8�, O'TE 'TLS
IL1J\ .\i\a,.,wv xopov
\
#L1JVE XOP1J'YE'T1JV EXWV EaV'T<p '
'TT«pELXE· At any
�

early date the producer (StMaKa�os) would be the poet. But we do

8g
 COMMENTARY

not know, and cannot expect to· know, how the unofficial production
of plays was managed.
49b2, o+£ 'II'OTE : see note on 49a2o. The date of the introduction of
comedy at the Dionysia, which was unknown when B.'s edition was
publishe9, was 486 B.C. ; see Capps, The lniroduction of Comedy into
· the City Dionysia (Chicago, 1903), and Hesperia, 12 (1943), Io.
49b3. axt\J.LaTa : cf. 48b36 and n. ·
It is curious that A. here writes as if comedy had begun at Athens,
whereas he has already emphasized the earlier development of
comedy at Syracuse. One can only guess that· he had Athens in
mind as the home of the dramatic records and perhaps of literary
history. The archon has no. place outside Athens.
�EYOJ.LEYoL: the meaning 'the so-called comic poets' is in accord
with A.'s usage (cf. Bonitz 424b28-45), but signifies little ; those comic
poets· 'who are spoken of', i.e. whose names have survived, implying
that many had been forgotten because the records began so late, is
. preferred by ·G. and S. Kassel's yev&,..,evot avoids the difficulty.
49b4. 11'poaw1ra: in place of the wine-lees which ":'ere originally part of
the comic disguise, whence TpvyqJ8la ? Tragic masks were said to be
the invention of Thespis, but if we believe the anthropologists that
masks are an original feature of all quasi-dramatic mummeries, the
role of Thespis must be limited to improving them. Later, when
'"p&aanrov was used for a character in a drama, a mask was called
'TTpOUW'TTELOV,
49b5. 11'�TJ9'1 {..11'oKpL1-wv : this A. himself records for tragedy at 49a16.
Mt>st Attic comedy can be performed by three actors, but four or
five are :required in parts of the Lysistrata and of the Frogs ; see
Pickard-Cambridge, Fe!tivals, pp. 148-52. According to the Byzan­
tine Anonymous writer on comedy (Kaibel, CGF, p. 18) it was
Cratinus who introduced regularity by fixing the number at three,
at · a time when there was no accepted limitation ; but since actors
were provided by the state one would have supposed that the num"
ber would be fixed as soon as comedy became a part of the Dionysia,
unless state provision was introduced later. Epicharmus used three
actors (POxy. 2427. I = vol. 25, p. 2).
To 8£ J.LU9ous 'II'OLEiv marks the change from a mere collection of ·

1/J&yo,, i!lvectives against individuals.

The priority of Epicharmus has already been asserted at 4Ba33·
About Phormis, or more probably Phormos, the only information
we have is that he shared with Epicharmus the credit for inventing
. comedy. The alleged titles of his plays. are all mythological, see
Suda s.v. and Themistius, Or. 27. 337 B, which seems to be based on
this passage or on a similar one from the llepl llotfJTwv. The names,
which have no construction, must have been a margin�1 note.

go
 COMMENTARY
49b6, E K lLKEMGs �Me : this implies direct influence of Syracusan on
Attic comedy. Cf. Themistius, loc. cit. : Kwp.cpola T«) ?Ta.\a")v 1fp{aTo
p.£v eK EH<EMas, eKEt8Ev yap TjuTTJ" 'E11lxap,J.os Te Kai .Popp.os, Ka.\.\tov OE
)1fJ�va'E CJVJ11]V{�81]. , ,
•

49b7. KpciT1)S : Crates was producing from about 450 to 430 B.C., when
Cratinus was the most prominent comic poet. That he was younger
than Cratinus does not mean that he could not have made the
innovation credited to him.
'II'PWTOS �pEev : for the tautology cf. ?TPWTOS evlKa from the Didas-
·

calic inscriptions.
49b8, KG86Xou 'II'OLELV Myous KGl f.Lu&ous : this repeats the p.v8ovs of 1. 5
with additional points. Ka8o.\ov : the iap.{3tK� io€a was concerned with
individuals. A properly constructed p.v8os generalizes. This is an
idea by which A. set great store, and it forms a large part of the
.subject of Ch. 9, where Ka8o.\ov is explained. .\oyovs Ka1 p.v8ovs : the
Kal is explanatory; .\oyos being rather more general than p,vfJos. This,
one out of the many senses of .\oyos, suggests a reasonably coherent
story as at 55834, the raw material or argument for a plot ; at 55b17
the ,\oyos, the essential story of the Odyssey, is hardly 9istinguish­
able from the plot : cf. .\oyot Aiuw?TEtot (R. 1393830), Aiuw?Tov p.vOo,
(Meteor. 356bn).
Without this passage (but cf. Kaibel, CGF, pp. 7, 8) no one would
have guessed that Crates had made a particular contribution to Attic
comedy. There is no hint of it in the sketch of comedy before his
own time given by Aristoph. Eq. 507-4o, the last lines of which are
about Crates. Nor is it clear how A.'s standards are to be applied to
extant comedy. Aristophailes is full of invective in his earlier plays.
Is the Clouds still within the province of iap.{3tK� io€a ? S. thinks so.
It is directed against an individual, but it could be argued that
Socrates is a highly generalized sophist as Cleon is a generalized
demagogue, or that both Clouds and Knights have sufficient plot to
remove them from the category of simple i�vective. But one may
suspect that A. preferred the Plutus to either. of them. . It is not
clear how he placed the development of Epicharmus-and Crates in re:
lation to the Homeric Margites. Both were movements away from
the .Poros.
Hereafter in the extant portion of the P. there are only a few
incidental references to comedy. There is ample scope for specula­
tion as to the way in which A. could have developed a doctrine of
comedy, especially with reference to katharsis. Attempts have
been made to achieve this, particularly on the b.asis of the definition
given in the Tractatus Coislinianus (Kaibel, CGF, p. so), see Lane
Cooper, A n Aristotelian Theory of Comedy (New York, 1922, Oxford :
Weimer Press, 1924). The discussion of the ludicrous in Cic� De Orat.

gr
 COMMENTARY
2. 235-47 may be based on A . But it is to be remembered that vir­
tually the whole development of 'the New Comedy was subsequent
to A., and that Peripatetic writers are likely to have modified A.'s
opinions on comedy more freely than those on tragedy, which
changed little after his death. .
49b9-20. There follows a brief statement of the relation between
tragedy and epic with the emphasis rather on the resemblances than
on the differences. Its practical justification is that A. regards epic
as largely contained within tragedy, which is the more fully ·de­
veloped form. In the section on tragedy, . which forms the main
section of the book, he draws his examples freely froin epic, a thing
he could hardly have done had he not clarified the relationship
between epic and tragedy in this transitional passage, of which the
conClusion is 'the man who understands good and bad in 'tragedy
understands it in epic too' 49b17. Vahlen suggests that A. is giving ·

his justification for not treating both forms of drama together.

49b9, 1) f'EY o�v • 1)�eo).ou81)aEV: this sentence, of which the text is
• •

highly disturbed, must give the points of agreement between epic

a,nd tragedy . in terms of the distinctions drawn in the first three
chapters. �KoAov8"1aEv is the opposite of 8,atfJ£pova£v. There is no
compelling reason why �KoAov8"1aEv should be aor. Probably it is
historical ; epic and tragedy long ago reached their final forms. The
most obvious resemblance is that both imitate a1rov8ai'o, (cf. 48a26).
So far as medium is concerned both use Myos and pv8p.os, elsewhere
called p.£Tpov; and, - whatever his exact words may have been, it is
unijkely that A said much more than this. Cf. 7TEp� Tfjs 8'"1Y"Jf'«T£Kfjs
•
.

Ka� €v p.£Tpcp 1-''f'"JT'Kfjs 59a17, with which the section on epic begins.
The p.Eya>.ov of A does not make sense with p.£Tpov .and gives · no
contrast with p.£-rpov a1ri\ovv in the· next line ; it is best dropped. It
does not appear in Arabic and Latin versions.
. 49bl l , T'ti 6E
• • 8&a..Epouaw : the difference between narrative and
•

drama is fundamental, even though Homer is the most dramatic of

narrators. The precise metrical distinction is less obvious. Epic uses
only hexameters, tragedy as we have it uses iambics, tetrameters,
and anapaests (if the last count as p.£-rpov}. We know nothi,lg of
metrical practice in A.'s time. A more notable distinction is that
tragedy has music and song which epic lacks, but that is referred to
in the comment on the p.lpTJ at I. 16, below, and anyway p.£xpi. p.€v • • •

implicitly excludes p.lAos. Probably p.£-rpov itself denotes speaking

verse which could not be set to music (cf. 47b2o n.), but · p.£-rpov
a7TAovv could conceivably imply that epic has the adornment of
metre without melody = p.£-rpov efi,i\&v, which is the differentia speci­
fied in Ch. 1 ; this is E.'s view. But the mostnatural meaning is that
epic uses only one metre. Conversely, in the comparison of tragedy

92
 C O MM E N T A R Y
and epic, 62ai5, tragedy can use the metre of epic, epic cannot use
that of tragedy. .
For f'EXf)L TOU • dval cf. 5IaiO ;...lxp' TOV avvSTJ>tos Elva,,
• •

49b12. 8lacJI€poualv : we should expect the sing.

49b12-1 6. A further distinction between epic and tragedy is in their
p:ij�eos.
The meaning of p:ij�eos here is· one of the major problems of
the P., though its importance is mainly historicaL It has three
possible meanings : (I} physical length of the written work� the num·
ber of feet of papyrus or the number of lines in the epic or tragedy ;
(2) which is closely allied to (I), the time required for the per·
formance of a play or epic. As readers we think naturally of the first,
but for the Greeks, who usually listened, the second is more obvious,
and xp&vcp in I. I4 shows that time in some sense is the meaning here.
But it is manifest nonsense to say that tragedy was ever 'unlimited
in time' and took as long to perform as epic, or that an epic could
continue indefinitely. ·The oldest tragedies are� comparatively short,
and 49ai9 may mean that originally, in A.'s view, they were shorter
still. Accordingly we have to accept (3) the length of time of the
action. This is not to us such an obvious distinction as the physical
length, but it is a perfectly valid one. The Iliad and Odyssey both
extend over several weeks, and the latter is much compressed by the
device of making one of the characters, as part of the action, tell
a story embracing the events of year.s. An epic on Heraeles or
Theseus, unless a similar device was used, must have extended over
a lifetime. It is true also that in older tragedies the time of action
was unrestricted. The events of the Agamemnon could not take
place in a single day, and in the Eumenides the passage of a con·
siderable period is definitely indicated. As . tragedy became more
realistic the continuous presence of the chorus, who in general
neither sleep nor eat on the stage, tended to restrict the action to the
period between dawn and dusk. But pedantry about time, which the
audience was not encouraged to calculate, was not allowed to impede
an otherwise desirable plot. The duration of the journeys required
in, for instance� Trachiniae and Andromache is ignored. Euripides'
Stheneboea seems to have used even freer licence ; cf. Page, Gk. Lit.
Papyri, p. I26 ; and even if the reconstruction of Ziihlke (Philol.
I05 (I961}, I-IS, I98-227) is right, the · extension in time remains
considerable.
Scholars have been reluctant to accept this explanation. It is true
that elsewhere, 55bi6, 59bi7, 62ai8, p:ij�eos means length in the physical
sense. It has been argued, as by B . and R., that the longer the dura­
tion of the action the longer will be the work containing it, so that the
two .meanings are combined. This is not obviously true and, as E.

93
 COMME N T A R Y ·

points out, the longest periods of time in ' the Iliad are accounted'for
·in the fewest words, e.g. ll. 24. 784. The above interpretation alone
is true to the simple facts about epic and tragedy.
I suggest that the reasons why A. expresses length not in the more
obvious way, in terms of the number of lines or number of yards of
papyrus, but in terms of the duration · of the action, is that this
brings out an essential difference between epic and tragedy. Tragedy
is superior to epic in unity because the events are less dispersed in
time. The necessary or probable connexion between events, on
which A. insists, is likely to be closer if one follows directly after
another, and it is remarked that this superiority is due to a develop­
ment within tragedy. Tragedies were always shorter than epics, but
only later did they acquire the cohesion which comes with a shorter
time of action. This idea of superiority is implicit in the conclusion
. reached in the final. comparison of epic _and tragedy, To yap &.8pow­
npov ij8,ov -� 1TOAAcp IC£Kpap.f.vov -rep xpovtp, Myw �· otov £i ns TOV
Ol8l?Tovv 8£LTJ Tov l:otf>o�eMovs ev £1r£aw oao's � l,\u1s (62b1, where,, how­
ever, xpovtp is time of performance). If this is correct, the statement
of the principle of the unity of time, which is based entirely on this
passage, is not so purely a generalization from practice · as has ·

usually been supposed (see below).

t9b13. utro 1-'(nv trEp(o8ov T)Mou : this suggests twenty-four hours
rather than twelve, but twelve, the time the sun is above the
horizon, ·is not impossible and suits the sense better. The Greek
play, like the Greek day's work, began at dawn, as is occasi<:mally
emphasized, e.g: in Soph. Electra, Eur. Jon and Phaethon� and the
action can plausibly be regarded as filling a day. Twelve hoq.rs
suffice for the action in virtually all Sophocles' and Euripides' plays.
J&LKpov ignAAciTTEW implies .neither much exceeding nor falling
short of. Indications of time are rare, much rarer than in Eliza­
bethan plays ; the Choephori ends after dusk (66o-2), the Agamem­
non and I A begin before daybreak, the Rhesus takes place at night.
This is the passage, the only passage; on which was based the law
of the unity of time . which was taken with immense seriousness by
the neo-classic writers and critics of the seventeenth and eighteenth
centuries. The critics were more interested in claiming the support
of the infallible A. for their own views than in finding out what he
said. 'A. was valid in their eyes only to the extent to which he sanc­
tioned a literature already in existence' (Vinaver, Racine (Eng.
trans. Manchester, 1955), I I ff.). Lessing, who was one of the first
critics to �onsider what A. really said rather than what the legislator
of the arts ought to have meant, observed (Hamb. Dram. 46, ed. R.
Rieman, Leipzig, 3· 197) that unity of place and time, in so far as ·
they were observed, were merely a consequence of unity of action,

94
 COMMENTARY
and that in any case the chorus made it natural that playwrights
should tend to observe both. But A. makes it clear that he disap- ·
,proved of the episodic play, 51h33, and no doubt he would have
found fault with the dispersion involved in the plot of A Winter's
Tale, which extends into the second generation, because it must
detract from the unity of the action. And this was no extreme case.
George Whitstone in his Epistle Dedicatory to Promas and Cassandra
speaks of the English dramatist who 'in three houres ronnes throwe
the worlde : marriges, gets children, makes children men, men to
conquer kingdomes, murder monsters, and bringeth Gods from
Heaven and fetcheth Divels from Hel'. (Quoted by L. Hotson, The
Wooden 0 {London, 1959), P·' 188.)
49b1 6. f'EpTJ : the constituer.t elements, six of them, as will be shown in
Ch. 6. p.£>.o� and ot/M a� the two which are lacking in epic : see
59blo.
49b18. &. f&tY yap • • • : repeated in the final comparison of epic and
tragedy 62a14.

CHAPTERS 6-ll. Tragedy.

CHAPTER 6
Definition of tragedy, and its six parts.
Ch. 6 begins with the definition of tragedy in the light of which the
necessary parts are distinguished and proceeds to a discussion of their
relative importance. The whole is closely knit and is not easily divided
into sections. For convenience it can be separated into (1) Definition
49b21-31 ; (2) List of Parts 49b31-so•14 ; (3) Importance of Parts
soa15-b2o, of which the section on Plot extending from the beginning to
soa3g forms the first half.
49bl 1 -31. The definition is derived largely from the conclusions of the
previous five chapters, but some supplementary .explan;1.tions are ·
required where new ideas are introduced.
49bl 1. trEpl. • €polif&EY! not concerning all hexameter verse but only
• •

that which is mimetic. Epic is in fact mentioned frequently in the

present section, but it is not till Chs. 23 and 24 that it is discussed in
its own right. Comedy was presumably treated in the lost Second
Book. Cf. -R. 1419b5 : Eip7J'Tat 1r0aa Ei8TJ yEAolwv EaTlv €v Tot� '?f'Epl
1TO,TJTucijs, about which there is nothing in the P. as we have it.
Iambic verse, the poetry of abuse, may have been subsidiary, like
epic in Bk. I.

95
 C O MM E N TAR Y
49b23. EK Twv ElpYJJ.LEYwv opov : the expected order of the words
would be Tov EIC Twv ELp1Jp.Evwv ywop.lvov : see V. on 55a24. E. explains
• • •

'Y'vop.oov as imperf., repeating the f.yiyvETo of 49"13 and referring to

the way in which tragedy realized its own nature while developing
in time, Twv EipYJp.Evwv being taken as the historical sketch in Ch. 4·
This is not readily intelligible.
49b24. ouaLQS : o vala is frequently used as a technical term in A.'s
philosophy meaning 'essence ', that in virtue of which a thing is what
it is ; see Met. Z, Ch. 4·
49b24-28. That tragedy is a p.lp.'la•s has been stated in Ch. 1, that it
imitates 1rpaTTOVTES in Ch. 2, and the main point of that chapter is
that they are a1rovaa ro, No one English word for a1rovaaros fits
,

both men and action (cf. srb6_n.). p.lyE8os was mentioned in con­
nexion with historical development in Ch. 4, and p.fjiCOS at 49bl2,
though the real significance of p.lyE8os will not emerge until Ch. 1·
. B. was probably wrong in taking TE>.Elas with p.lyE8os f.x�va1Js, ·
'complete as having magnitude' (cf. 50b25). 7}avap.EVCfJ >.oyqJ refers to
the media of poetry defined in Ch. r, as is explained iri the next
Sentence j the force Of apwVTWV ICaL OV a,• a?Ta'Y"/E>.£as haS .been fully
elucidated in Ch. 3, but a,• f.>.Eov Kai �o{Jov is new, and the assertion
that Ka8apats is the end of tragedy is not only new but remains
·

unexplained. .

49b24. 1rp6.gEws : cf. 48ar ; we have had 1rpae's as the complement of

1ra8os at 47a28 and in a more general sense, similar to this passage,
at 48b25. It is necessary to be aware of the connotations which this
word frequently possesses in A.'s literary and ethical works, because
much of what is said in Chs. 7 and 8 about unity of action is implicit
in the word 1rpae'S.· It means, not any random act like opening one's
mouth or crossing the street, but an action initiated with a view to
an end and carried on in pursuit of it ; it can thus include a whole
complex of subordinate actions (cf. srars, 19) ; it is associated with
1TpoalpEULS as EN II39a3I : 7rpaeEwS apx� 1TpoalpEULS ; cf. EE I222bl9 ;
man alone, as. opposed to animals, can initiate action, is apx�
1rpaeEwv (and similarly with the verb EE 1224a28 : o/J �ap.Ev To 1ra,8lov
1TpaTTEtv). Since 1rpae's refers to an action begun for a purpose and
carried on · until it is realized or until the activity thus initiated
terminates, it is implied that it is a complete whole TE>.Elas, the word
here added ; that it is an entity with beginning, middle, and end as
explained in Ch. 7· Cf. E., p. 256, 'The word has the twin implica­
tions of completeness and seriousness'. For a different view, see
Solmsen, CQ 29 (1935), 197. This conception of action has interested
modem writers on tragedy : cf. F. Fergusson, The Idea of a Theatre,
(Princeton, 1949), p. 361 'Thus by action (of the O T) I do not mean ·

the events of the story but the focus or aim of psychic life from

g6
 C O M M E N T A RY
which the . events in that situation result'. In the Appendix on
Plot and Action, p. 229, almost ineffable mysteries are propounded.
Cf. al::;o J. Jones, On. A. and Greek Tragedy, pp. 24-29.
49b25 . .ft8uaJ.LEVctJ MyctJ : a TfSvup.a is something added to food to give it
a pleasant flavour ; the metaphor is maintained at R. 14o6319 where
Alcidamas is said to have used epithets ovx �Svup.an aU' ws i3£up.an,
'not as seasoning but as food'. It is used contemptuously by Pl. Rep.
6o7 A : �3vup.£v1J Movua. Here the tone seems neutral, as at Pol.
1340b16 : � p.ovu£K� tf>vun nov �3vup.£vwv, but . A., like Plato, was
capable of regarding style as an extraneous. addition to matter : cf.
R. 1404324 ff.1 and Pl. Gorg. 502 c. A .;;avup.a was not normally sweet,
though . sugar would count as one but only as making foods more
agJ"eeable.
·

xwpis €�eciaTctJ Twv d8wv : nov �Svup.aTwv is to be supplied after

El3wv. EKauTcp, though only two kinds of.;;avup.a are mentioned, note
ETEpa in I. 30, below. This is a complicated way of saying that
:rhythm alone is used to make more alluring the language of the
dialogue, rhythm with melody is used in the sung parts ; cf. Ch. I,
especially xwpis � J.'EJ.''YP.lvo£s (47323). Since we do not regard style
as a separable ornament of the subject-matter, no translation of
this can sound natural.
49b26. Tois J.Lop{o,s : various parts, of what kind ·is not specified ; at I.
32 p.&p,ov is equivalent to one of the six p.EpfJ. Here something more
like the quantitative parts of Ch. 12 must be intended, epeisodion and
stasimon. . .

8pwvTwv : subj. gen., 'an imitation performed by men acting', not

'of men acting', or it may be gen. abs. with Twv J.''J.'ovp.£vwv under­
stood. This follows from Ch. 3, 48319-28.
49b27. 8,' iXEou �eai ci»6J3ou : see Appendix II on Ka8apu,s.
TftV • • • �eci8apaw : different from Ka8apulv -r£va ; there are various
sorts of Ka8apms ; this is the one appropriate to pity and fear.
KQ.8apU£S appears in a different sense at 55bl5. On the general in­
terpretation see Appendix II.
The vulgate reading on which early commentators worked was
xwpls EICclcT'TOV 'TWV El3wv EV Tois p.oplo£S bpWvTwv, Kal ov a,· E7TayyEMas
O..Ua a,• l>..Eov • • · • The first to punctuate correctly was Giacoinini
(1573) : see Weinberg p. 524. The old punctuation has been per­
versely revived by R. Stark, Aristot. Stud., Zetemata 8, pp. 39-46,
who separates 8pwvTwv from what follows and takes ICal with 1npal�
vovua.
Twv To,ouTwv tra8'1J.LciTwv : attempts have been made, and still are,
to distinguish between 1Ta81Jp.a and 1ra8os. But Bonitz, Aristot. Stud.
v. has convinced most scholars that they are indistinguishable, the
form 7Ta81Jp.aTwv being preferred to 1ra8wv,·which does, however, occur.

97
 C O M M E N T A RY
TC.ov TotovTwv raises a much-discussed problem. It was held by
many of the older commentators and by R. among the moderns
that TOto1hwv is here equivalent to TOVTWJ'. It is a question of some
substance because if TowvTwv = such, then pity and fear are not the
oply tragic emotions, and it is the difficulty of finding other emotions
which may be purged by tragedy that has led to the attempt to
equate TotovTwv with TovTwv. A.'s use of TotoiJTos has been examined
by Beare in Hermathena 18 (1914-19), n6-35 ,and more recently by
C. W. van Boekel, Katharsis (Utrecht, 1957), pp. ;1:46 ff. The fact is
that there seems to be an ambiguity which is common . to Greek
TotoiJT�s and to English 'such•. Both words can mean similar to the
referent as falling. under the same definition but not separately
specified-often virtually synonymous with othos-and similar to
the . referent but falling under a somewhat enlarged definition. In the
first case the point of Twv .-rotoUTwv here would be to include, for
example, olKTos, EK'TI'ATJ�'s, pity and fear in a slightly different form ;
in the second, to include different but kindred emotions such as
opy�. Cf. 's6b1 : olov EAEOV .� c/>o{Jov � opy�v Kal oua TOLaVTa, R. 1378"22 :
opy�, EAEos, t/>o{Jos Kal oua a>.Aa 'totavm, where A. is thinking primarily
of rhetoric, and Pol. 1342a12 : Tovs l>.e�JLovas Kal Tovs c/>ofJTJTLKovs Kal
Tovs o>.ws 1ra67JTLKovs. Since we are so much in the dark as to what
A. meant by Ka6aputs, it is difficult to choose with confidence, but the
first meaning is the - more likely. In any case there is no justification
for introducing under TotovTwv such emotions as ambition ; cf. John­
son in Appendix II, p. 277.
A meaning which, it has been suggested (Pol., ed. Susemihl­
Hicks (London, 1894), p. 652, also by Butcher, p. 240, n. 3)' might be
int�nded by Twv TotovTwv is that the emotions aroused by tragedy are
not identical with the corresponding emotions aroused by events in
real life ; they are e>.eos Kal c/>o{Jos as aroused by imitations ; cf. 53b12 :
' Q'TI'O
TT}V ' \ ' Kat' -L
• ' EI\EOV 'Q
'f'OtJOV � · JLLJLTJUEWS TJUOVTJV
aLa ' ·� ' ; See nOte On 48b 13. . They.
are transformed by 'aesthetic distance'. · If this distinction were .
a basic assumption of the P., it might well be referred to in this sum­
mary fashion, but as A. never clearly makes the distinction it is
·

going far to find it in Twv TotovTwv.

49b29. cipJ.Lov£av [Kat J.LEXos] : the difference between apJLovlav and JLEAos
is that JLEAos implies the presence of words. If Kal JLEAos is in place
here at all, Kal must be explanatory, but as >.6yos + ,Ov8JLoS and
Myos + pvOJLos + apJLovla cover all the media used by tragedy, JLEAos
is best explained as a gloss based on 47b25.
49h31 . After a,a JLEAovs would be a possible place for a similar brief
explanation of the meaning of Ka8aputs. · ·

49b3 t -soat4. The six parts or elements of tragedy are now deduced
from the definition. This passage was discussed at length by Vahlen,
g8
 C OMM E N T A R Y
'A. i.iber die Teile der Tragodie':, Gesammelte philologische Schrijtm
(Leipzig, I9II) i. 235-74, reprinted from Symb. phil. Bonn. in hon.
Friderici Ritschelii (Leipzig, 1864), pp. 158-84.
49b3 1 . "PGTTOVTES : the visible actors. More precisely; the poet makes
the imitation a,.x 1TpaTTovTwv. The play can achieve its Tl>..os without
being performed (623II).
49b33. o�JfE(I)S �eoaJ&os : the actors are visible, and the spectacle they
present is necessarily to some extent an element in the total effect.
Koap.os implies that things are so arranged as to be worth looking at.
The question is whether ot/ns refers only to the appearance of the
actors, who were richly attired, or includes all that we mean by
'spectacle'. There is no doubt that on the Greek stage, as on the
· Elizabethan, the main spectacle was the appearance of the actors,
magnificent or horrific as the occasion might require. At sob2o ot/M
is associated with the aK£Vo1Totos . who is. said to have been concerned
mainly with masks and costumes (Pollux 4· ns ; cf. schol. Aristoph.
Eq. 230). The few spectacular effects of which we hear seem to
· depei.1d mainly on his efforts : the Erinyes in the Eumenides who
caused a panic, perhaps the winged steed of Oceanus in the P V ; cf.
Bellerophon's steed in Euripides' play ; Ion's a<::t with the birds at
Delphi would depend mainly on his own grace, "[Demetrius] Eloc.
195. · But at 53b1-4 ot/Jts seems to refer to the whole content of Tov
- opa.v. Presumably it was not for nothing that Sophocles introduced
scene-painting, and the mechane as a spectacle must have lent excite­
ment to divine epiphanies. From a later date mrs UK1JVtKaL's ot/J£at
Ka>.ov (Argument to Eur. Phoen.), e.g. Antigone on the walls, refers
to more than clothes.
lv TOUTo&s : actors speak and sing, using words, rhythm, and
melody ; these are the media (ev) ; cf. 4 7b29.
49b34. TTJV T&Jv J&ETp(l)v auv8Ea&v: p.hpa are non-lyric metres, 47b2o il.
The whole is short for T�V TWV 01/0f'tlTWV Ell p.E'rpcp avv8£atv�
49b36; 1r&aav : predicative 'in its entirety'.
49b36-50B1 0. €1rEl. 8E • • f.LEAo1ro&{a: this rather cumbrous sentence
•

is best taken as depending on E1TEl down to yvWp.1JV at 5087, with the

apodosis beginning civayK1J oJv. The const. is obscured by the change
from nom. p.vOos at 5034 to the ace. in 7j871 • • a,O.votav which are
•

governed· by Myw. .
49b37. 1rpaTTOVT6Jv: the performers, who have the same 7j871 as the
original characters of the story.
11'o&ous : speakers in the courts argue o1rws Tov Kptn}v 1TOtov Ttva
?Tot'?awatv (R. 1354b2o).
49b38. 1i8os : see on 4832 and sobS.
8&G.vo&av : further explained below and at sob4. We differ from
the Greeks in attributing 8t&vota to the author rather than to his

99
 COMME N T A RY
characters, in whom we do not so sharply separate intellectual power
from the remaining characteristics. A. divides apET� into �8LK� and �La­
vo7JTLK� (EN IIOJ35)· On a man's �Lavota depends his power to assess
a situation, on his .qOos his reactions to it. In drama
- �tavota is mani-
fested mainly in the characters' arguments.
508 1 . 11'0LQS TLVQS : many actions Can be judged on)y in the light of
what we know of the character of the doer and of what he says in
explanation of his actions.
[1r£cJ>uKEV aiTLa •] : probably a marginal explanation of the pre­
• •

vious clause.
5082. TauTas i.e. 1rpagE,s : the end of an action is a further activity and
not a static condition ; this is good Aristotelian doctrine. but not
self-evident, and depends on a particular conception of activity. Cf.
EN I, Chs. 7 and 8.
5083. iaTLv • •: p.fi8os has scarcely appeared since it was numbered
•

among the main subjects of the P. at the beginning of Ch. I . Now it

is introduced with great emphasis ; .we know that tragedy is a p.lp.7JaLs
1rpaeEws, but it is the p.v8os, the 1/Jvx� of tragedy as it is called at 1 . 38
below, which is more particularly the imitation of the action.
soa4. X£yw yG.p f.Lu8ov TouTov : 'for I use the word p.v8os in this sense',
not the most obvious sense, which was simply 'story'. Similarly B.
wrote mVT7JV for avT�v at 49b34.
The poet takes the story; p.v8os in the non-technical sense (see
4739 n.), and reorganizes it in such a way as to bring the parts into
a more logical and significant relation to one another. The story is
a preliminary selection from the stream of events ; in the plot the
story is organized.
soas. auv8EaLY TWV 11'payf.LciTwv : 'structure', developing the hint given
by the word avvlaTaa8at 4739 : cf. avaTaULS in 1 . IS· UVVTL8tvaL is
used of putting together an essentially true story in Eur.
Bacc. 297, Aristoph. Ran. 1052. Cf. Pl. Phaedr. 268 D, T�v To6Twv
(sc. MaEwv) avaTaaLV 7TpE7TOVaav d.M�AOLS TE KQ� Tqi OAqJ CIVVLUTap.lvqv.
soa7. yvwf.L'JV : not 'purpose', though it would make sense here, but as at
50bi21 1Ca8oAOV nd7To�alvoVT!J.L1 'make SOme general Statement'. yvwp.aL
were 'practical maxims' (R. 1394321), but they naturally tended to be
pithy and epigrammatic. The yvwp.o8twKT�s of Cratinus, fr. 307,
being a Evpt7Tt�aptaTo�avl,wv, no doubt appreciated To aTpoyyt5.\ov ;
cf. Aristoph. fr. 47I. Eur. Phoen. is described in the Argument as
yvwp.wv p.EaTov 7ToAAwv TE Kal ICa.\wv.
t L 1' li t / 1 • 1 t IS
" a 1ogtca
" 1
avu.yK') ouv : cf. 49b37 : ovs avayK7J 7Totovs nvas nvat.
consequence of its nature as developed from its definition.
50B8, 11'0LQ TLS : this refers not to the quality Of an individual tragedy
a.s good or bad, but to the nature of tragedy in general. B., who like
most edd. omits Tijs after 1Ttla7Js,.is misleading here.

100
 COMMEN T A RY
50&9. Taiha 8' • . : the or-der in which the six parts are given bears no
.

relation to the division which follows, but with the transposition of

Jt.E�,s and Su1vow the order would be that in which t,he parts are dis­
cussed, which is more or less the order of importance ; or/M and
p.e�orroda, which come last, are not really a part of the art and are not
discussed. The order A£�,s-SuJ.voLa is common elsewhere.
50&10, 1 1 . ot� JlEV • Tp(a : Mo = AE�LS, p.e�orrola ; Ell = or/JLS (8ia
• •

rrpaTTcwrwv) ; Tpla = p.v8os, 1J8TJ, 8ufvoLa. M�,s applies to the �oyos of

non-lyric parts. Words of lyrics are included in p.£�os. The P. is
almost entirely concerned with the group of Tpla. Robortello (1548) and
Maggi (1546, 1550) explained Tpla as M�,s, �8os, SufvoLa. The passage
was correctly explained by Vettori (156o) : see Weinberg, p. 462.
5011 1 2-14. TouToLs JlEV waauTw� : a passage much discussed, but
• • •

never adequately explained. The apparent meaning is unsatisfac­

tory ; after saying that every tragedy must have all six parts, why
add that 'not a few' poets use them all ? It is vain to cite a�8eLs
Tpay<p8la' (5oa25), because it means deficient in, not devoid of, �8os.
Further, the separation of TolhoLs and eiSeaw is needlessly emphatic,
and ws elrreiv, though not restricted by A. to use with such expres­
sions as 'all' or 'none', has no application here. '1Tiiv sc. Spiip.a is an
odd expression. Some of these difficulties can be avoided by re­
arrangeiJlent or emendation, but as a summing up it is defective,
even apart from the gratuitous change from p.£pTJ to eiSTJ (see56a33 n.).
The only serious alternative is Vahlen's ws eiSeaw, abandoned in his
edn. The meaning of this would be that not a few poets concentrate
on a single p.£pos (cf. 56a3-7) using it, or regarding it (cf. 52b14, Met.
99Bb1o xpija8a, ws y£veaLv) as a special type of tragedy. In the next
sentence he read or/JeLs exELv '1Tav (cf. 62a14), indirect statement of the
poet's thought, that spectacle, or whatever p.£pos it might be, con­
tained all the essentials of drama. This seems an extreme view, but it
makes sense of a sort. It is a difficulty that of. the ei8TJ mentioned in
Ch. 18, though they are connected with the p.EpTJ, only the �8LKov is
common to both (see 55b32 n.). But 53b1 gives some support to the
existence of an el8os connected with c;.p,s ; anyway the ei87] are them-
selves a mystery. _
to+L�: or/Je's is an easy emendation, but why is this p.£pos alone in
the plur.? c;.p,., gives the required sense, but the corruption is harder
to explain. .
5081 5-38. Five reasons why p.v8or; is pre-eminent among the parts.
First reason 50815-23. Drama is concerned with actions, to which
character is incidental.
. : not distinguishable from 7Tpayp.aTwv
508 1 5. 1rpayJlUTWV O'UO'TQO'L�
avv8eaw in I. s, above.
IOI
 COMMENTARY
50a16. l'li'TJaLs .� 1rp6.�Ewv : this follows from the definition and was
.

- assumed at 48ar, at least in so far as 1rpaTTOVTE� are engaged on

a 1rpa�'�' and by Plato before ; it is regarded as self-evident and no
reasons are given. The fact that tragedies are · about actions shows
nothing, for even if the main purpose was to reveal character, charac- ·
ter would have to be revealed largely through action ; cf. what is
said about 1rpoalpecns ·at sobro. A. does no more here than repeat
what is implicit in the definition.
50a17. j3iou : a surprising statement, but apparently it means what it
says, not the lives of individuals ('careers' E.), of which only a minute
proportion can be represented in a play, but the whole complex of
events of which a generalized picture is given, a more philosophic
picture than that given by history. Alcidamas had called the Odyssey
KaAOV av8pw7Tlvov {Jlov KaT07TTpov (R. I4o6bn), and there are other
examples of this wide use : see B., ad loc., and Pl. Laws 858 D. A.
associates {Jlos or 'ru� with Jpya and 1rpa�ets (EN II79b19, EE 1219b2).
In later writers imitation of {)los is associated more particularly with
comedy.
50a17-20. [�at Eu8aLI'ovla . ] : these words can be attached to
• .

the text only by supposing that so�e words, e.g. Kal ev8atp.ov{as,
have fallen out by haplography. But (a) KaKo8alp.wv, -ovla do not
occur elsewhere in A. ; (b) Te>..o s in I. 18 is awkward with TEAos in I. 22,
especially as the first refers to the end of the action which is the sub­
ject of the tragedy, the second to the action itself which is the end of
the tragedy ; (c) A.'s particular views on the end of action are not
very relevant to the importance of action in drama, but they are the
sort of thing that a commentator might be tempted to explain.
The desire for happiness might well be the cause which led to the
initiation of the action which was the subject of a play, but this
action is just as much an action whether the happiness which is it.s
end is regarded as an action or a state. In fact A. was emphatic that
it was an action : cf. EN norars, n76b7, Pol. 1325a32. Even the
{Jlos 8erup1JnKos, which we might regard as the · opposite of action, is
an evlpywi (EN II77ai8).
50a2 1 , 1rpQTTOUaLV , , , aUI'11'EpL)..al'j36.vouaLV ! the subject is o{ p.tp.ov­
ftEVOt, They do not act, i.e. conduct their imitation of the action, in
order to present character, but they present �haracter as an element
in the action which is the main end. On A.'s habit of identifying the
poet with the actors who are his medium see 48a1 n.
50a22. �C71'E , , ! accordingly it follOWS that the plot is the end.
,

It had not been suggested that any other part but .qOos could claim
the primacy. In a different, and more ultimate, sense the .,l>..o s
of tragedy is the emotional effect it produces (and the katharsis) :

102
 COMMENTARY
cf. 6ob24, 62a18, 62b15 ; here i t is the end in the sense that the poet
subordinates all other parts to it.
Ka£ : explanatory of 1rp&:yp.a-ra : 1rpayp.chwv ava-raats which is the
same as p.v8os. Or 1rpayp.a-ra could stand for the materials and p.v8os
for their form when embodied in structure.
Second Reason soa23-29. Tragedy can exist without character, not
without action.
soa23. O.veu • Tpay«tJ8£a : many contemporary plays and some of
• •

Maeteilinck in an earlier generation come nearer to achieving this

than A. would have thought possible.
soa24. QVEU 8£ 1]8wv : B. is right in taking &.�8Eis, 1. 25, below, as
'd efic1ent m character' ; cf. ap.tp.TJ-rws
• • > I " • 1•
Eypa.,Ev, 'drew madequate1y'
· •

(6ob32). A play may be devoid of 'character interest', but it is in­

conceivable that all the participants of a play could be such that the
audience have no expectation as to the kind of decision they will
make. Those who act must be Trawl -rwEs (49b37).
soa25. TWV VE(I]V : see on 53b28. Euripides is less a master of �8os than
Sophocles, but to call him &.�871s would imply a severe standard
indeed. ol 1rpw-rot (l. 37, below) who are contrasted with the vEot are
presumably the earliest of the TTa)t.awl. S. is alone in taking vEwv as
'young' rather than 'r�cent', which contradicts 11. 35-38, below.
50a27. Polygnotus was introduced at 48a5 as an example of a painter
who idealized his figures, see n. ad loc. Zeuxis of Heraclea in southern
Italy worked in the late fifth and early fourth centuries. He is men­
tioned at 61b12 as painting figures more faultlessly beautiful than any
in real life. At Pol. 1340a28-4o it is said that ala8YJ-ra except for
p.ovatK� have little ethical effect, though there Polygnotus is again
mentioned as �O,Kos.

Third Reason 50a29-33. Tragedies which are 1·ich in· character but
poor in plot do not fulfil the funCtion of tragedy.
soa29. £cf>egijs : suggests a mere sequence, 'one after another', as - at
59a27. Sometimes, however, it is a rational or natural order, s�ai.
1]1hKcis : 'expressive of character' B., in this case the true character
of the speaker. >.&yot �O,Kol (R. 1391b22) are speeches which suggest
the speaker has a certain character, not speeches on ethics.
soa30. �v : 'was agreed to be'. A. uses the imperf. to refer to conclu­
sions previously reached ; cf. R. 1363a9 : oo yap TTav-rEs £tf>lEv-rat, -rov-r'
aya8ov �v. -rpaycp�las Epyov recurs at 52b29 with·reference to emotional
effects ; these have not so far been discussed, only mentioned in the
definition 49b27, 28. The majority of commentators-, R., S., E., refer it
to that passage, E. with emphasis on Ka8apats. B. following Vahlen
103
 C O M M E N TA R Y
understands the Epyov to be the production of a 1rp&.��ws p.lp:YJaLs (cf.
49b36), which a mere sequence of speeches could not achieve.
The difference between a. play and a series of rhetorical speeches is
put, with a different p\lrpose, by Pl. Phaedr. 268 c.
50832. �ea.£ : is undoubtedly explanatory : cf. I. 22, above.
50833. Castelvetro transposed the comparison of the painter and the
choice between pleasing colours without form and form without
colours, 50839-b3, to follow directly after 1Tpayp.aTWV. This was ac­
cepted by most editors until Vahlen, who defended the change (Gesam­
melte, pp. 25o-2), altered his mind in favour of the MSS. order in his
edn. ; only E. among recent editors transposes . . Against the chartge
we may argue that there is no easy explanation to account for it ;
that the comparison between a more and a Jess essential dement
stands well at the end of the discussion about plot and character ;
that EUTLV T� p.lp.'f/ULS
• • at sob3 follows a little awkwardly oil apx'f/
•

p.Ev o�v• • • 7]811 at 50838, 39, since both sentences round off the stage of
the argument. On the other hand, while there is an obvious cor­
respondence between an outline drawing and the plot of a play,
a random spread of colours has little connexion with the characters.
But the comparison. would gain immensely in significance if attrac­
tive colours placed at random xvS11v corresponded with eloquent
speeches following one after another NJ�efjs. · Indeed G. goes so far as
to make xva11v refer back to NJ�ffis though he leaves them 12 lines
apart. That the words do. correspond is strongly suggested by the
relevant passage of the Phaedrus. Socrates, criticizing Lysias' speech,
says 'do not the parts . seem to have been thrown down xva.,v at
random ?' and a few lines later 'can you suggest any reason why
Lysias put the parts in this order ovTws N)�efjs ?' and the meaning is
illustrated by the famous Midas epitaph of Cleobulus of Lindus, four
hexameters which can be read in any order, 264 B-D. irfJ�!fjs sug­
gests, not disorder, but absence of any ordering principle.

Fourth Reason 50833-35. The most attractive elements in a play

belong to the plot.
50833. +uxa.ywyEi : cf. sobr6 and Timocles, fr. 6. 6 ; first in Xen. Mem.
3· ro. 6" and Pl. Phaedr. 261 A. This originally striking expression
drawn from necromancy (cf. Aes. Pers. 687) loses a good deal of its
force when it becomes part of the language of criticism, and implies
little more than 'attract' : see G.'s note. It is surprisingly absent
from Gorg. Helen (he uses EKYO'f/T�vw). Eratosthenes stated that the
aim of poetry was 1/Jvxaywyla not StSaaKaAla (Strabo r6).
50834. 1rEpL1rETELa.L �ea.l O.va.yvwpl:aELS : the first mention of these terms,
which are explained in Ch. u . Probably they were already recog­
- nized technical terms.

104
 C O MM E N TA R Y
Fifth Reason and summing up 50835-39. The plot is the most
difficult part of a play to manage.
50a3(». a�epLpoOv: cf. 48bn €lK6vaS' p.&.>.tuTa �Kpt{Jwp.EvaS', to produce
something finished and precise. No . extant tragedy is a youthful
work, unless possibly the Rhesus, which, if it is by Euripides, was ·

written early in his career : see Ritchie, The Authenticity of the Rhesus
of Eur. (Cambridge, 1¢4). The evidence of other periods certainly
suggests that skill in dramatic art comes only with practice, whereas
a youthful rhetorician might be expert at apt speeches.
50837. auv(aTaa8cn : act. and mid. form seem to be used indifferently. ·
ol ,..p&hoL: are these pre-Aeschylean ? See on vEwv I. 25, above. If
this generalization does not strike us, so far as our evidence goes, as
obviously true, it is perhaps a measure of the difference between
character and �8oS'. However, Sophocles is said to have considered
his latest and most mature style to have been �8tKWTaTov (Plut. M.
79 B) ; see Sir Maurice .Bowra, Problems in Gk. Poetry (Oxford, 1953),
PP· Io8 ff.
50838. cipxT) •Kat otov "'ux� : in the light of A.'s ·philosophy these
. .

words carry even more weight thari they ;might seem to. E. quotes
De An. 402a6 : EUn .yap (� .fovx�) otov apx� TWJI {c{Jwv. Soul is the
'form' of man, and plot is of equivalent importance in tragedy.
50&39-50b4, The Second Part �8oS'.
50839. 8EuTEpov : in order of importance. A difficulty of the follow,ing
passage is that �8oS' has already been examined in connexion with
-
p.fi8oS', as being the only rival for primacy, without having been
introduced formally as the second part. It has been defined only
incidentally (49b37) as that in virtue of which men are 1rotol, of one
sort and not another. It is further defined in contrast to a,&.vota
5ob8, below.
50bl, iva�EL"'ELE : trans., the object being 1rlvaKa understood.
50b2, xu8fJv : 'at random'. Ta xva7Jv means prose as opposed to verse
(R. r409b7). At Phaedr. 264 B (see ' 50a32 n.) the parts of Lysias'
speech are said xvaT]JI {J€{J>.Tju8at 'to be thrown down at random'.
Ell+puvELEv : _ give the �aov� which is the TE>.oS' of the arts.
�EuKoypa+�aas : for the formation cf. UKtaypacf>Eiv. There seems
no evidence whether this means to draw in white on a dark ground
or to draw in black on a white ground. That Philostratus mentions
drawing >.£vKjj Tjj ypap.p.jj (Vit. Ap. 2. 22) shows little. The comparison,
if it stands here-cf. 50a33 n:-is between an outline drawing, which
corresponds to the plot, and beautiful · colours grouped without
meaning, corresponding to the characters. While the first pair have
significance in common, the correspondence between characters and
patches of pigment seems incomplete. Gomme, p. 63, following V.
IOS
 C O MM.EN T A R Y
suggests 'not give so much pleasure as if he draws the outline in
black and white first'. For a similar comparison between 7T£ptyparf>�,
'outline', and rf>app.aKa, 'colours' see Pl. Polit. 277 c.
50b3, liLci. TQUTTJY ! i.e. indirectly, secondarily, repeating the point
which was made at soai6 ; tragedy is not an imitation of men, but it
imitates men in the course of presenting human action. The repeti­
tion rounds off the . paragraph.
50b4, 1rpQTTOVTWY here must refer to the original agents, not the per­
formers, as at 48ai .

50b4-12. The Third Part 8tavota.

This is discussed further in Ch. I9, but it is quickly dismissed as
being in the province of rhetoric. .
50b4, To(iTo • • 8UvQa9QL! as �8os is revealed mainly by action, so
•

�tavota is revealed in i\oyot spoken or written, in demonstrations and in

generalizations ; the point is repeated (ll. n, I2) at the end of the
section. · �8os Kai 8tavota are the two aspects of the whole man, cor­
responding, though rather superficially, to character and intellect.
Then as now it was a subject ·of debate with which of the two the
educator ought to be more concerned (cf. Pol. I337a38): In the last
resort they are not completely separable. .
50b5, TeL EYOVTQ ! the man with good a,a.vota can see what there is in
a subject and what there is to be said about it. Cf. Isocr. S· no :
KaTt8wv T� 'fTAfj8os Twv £v6VT-wv £l'fTE'iv, Dion. Hal. Lysias, IS : Evp£nKds
yap £uTw Twv £v Tots 7rpayp.autv £vovTwv ;,\6ywv.
Tci. cipJ.&oTTOVTQ ! cf. S4a22. The two together are much the same as
Ta 8£ovTa, what Thucydides claims to put into the mouths of the
speakers in his History (1. 22. I).
50b6, TWY Mywv : taken by B. and G. to refer to speeches in plays, by
Vahlen� Butcher, R., and S. to speeches in general. M. and E.
bracket the words. At I. IS, below, it has the latter meaning (or just
'prose'), and it is natural to illustrate a,a.vota from rhetoric in general :
cf. s6a34· On the other hand, with ol apxaiot I. 7 we are back again
with the speeches of drama. It is not necessary to specify ;,\6yot as
the part of drama where 8tavota is in place, but A. does sometimes
state the obvious.
Ti]S 11'0"-LTLKi}s KQl P'lTOpLKfj!i : 8tavota is associated with both of
them. The 'fTOAtTtKOS is as old as the 'fT(?AtS ; the P"/TOptKOS depends on
his new art. The meaning of P"'ToptKfjs is plain, but 'fToAmKos is used
in many senses. With this passage may be compared Pol. I274b36 :
TOV 8� 7TOA£TtKOV Kai TOV vop.o8ETOV 'fTfiuav opwp.Ev T�V 7rpayp.aT£lav oJuav
1TEpi 1ra.\w. 'fToAm"� is concerned with the good of the whole 'fTOAts
(EN I094bn). As ethics were conceived as existing within the frame­
work ofthe city, the connexion between Tj8., and 'fToAmK� was close.
ro6
 COMMENTARY

\
Cf. Pol. 1 288hr : 7Tat8tda Kat e8TJ TavTa ax£oov Ta 1rowvvTa a7Tovoaiov
; R' 1356)a26 : TTJ'> 7T£pt Ta TJ"(}TJ 7Tpayp.a-
'
" 11
avopa KatI TaI 7TOtOVVTa 7TOI\LTLKOV
- ,
•
- I I

T£lac;, �v olKaLOV £an 7Tpoaayop£V£LV 1TOALTLK�V ; Satyr. vit. Eur. 39·

iv. 5 (Arrighetti) : 7Tot\Ad. Kat 7Tapa nov KWJ-LLKclJV 7TOLTJ'TWV ap.a avO'TTJPW'>
Uy£-raL Kat 7TOAtnKwc; EvpmlOTJ'> 7Tpoc; aAK�V Kat £vr/Jvxlav 7TapaKaA£t
• • •

Tovc; v€ovc;. There is a touch of irresponsibility about pTJTOptK� : cf.

PTJTOptKOV yap 'TO 'TOLOV'TOV p.U.At\ov 1} 'TtP,wv-roc; at\�8£tav avOpOs (Galen,
Protrept. ro).
50b7. ol •cipxaioL : see note on 53h28.
• •

SObS, {nlTopLKWS : tl�e rhetorical characteristics of Euripides are ob­

vious. In the set debates he seems at times more eager to make his
CharacterS say Ta £vovTa Kat Ta app.O'T'TOV'Ta than to maintain their
consistency. No doubt fourth-century tragedy developed further
along the same lines ; Theodectes was both dramatist and rhetori­
cian. We need not go so far as to take 1roAtTtKws as · equivalent to
tj8tKwc;, but it implies a less ex�lusive interest in persuasiveness· and
point scoring : cf. R. I4I7a23 : p.� W) a7TO Otavolac; My£LV W0'7T€p oi viiv,
d.U' ws am� rrpoatp€a£wc; (for 7Tpoalp£0'LS see below) and 1366�10. But

111 I \
there is ·some overlap between them. G. quotes Dio Chrys. Or. 52
(VO1 111 p. I 60 D"lnd ) : TJ 'T£ 'TOV- Evpt7TWOV
' O'VV£0'L<; 7TOI\L'TLKW'Ta'TTJ Ka_LI
,
• .. • " • • ,

pTJTOptiCwnhTJ ooaa. As an example of a play that was conspicuously

1roAmKoc; E. �uggests the Antig. Eur. Supp., like Thucydides, is
both 7ToAtTtKoc; and PTJTOptKo>. For an extreme of rhetoric we might
instance the agon of the Troades (914-1032).
·

. �9os J.LEV : answered by ·ot�l.vota o€ I. II. oq8o> is introduced here in

order to explain otavota by contrast. According to our oq8oc; we are
1rotol nv£c;, and naturally prefer certain courses of action and make
certain decisions 7Tpoatp€a£ts. The notion of 7Tpoalp£ats makes clear
the connexion between oq8oc; and action, whereas otavota is revealed
mainly in speech. A 1rpoalp£ats is a considered decision made by a
person of mature judgement after due deliberation. Such a choice
cannot be made on the spur of the. moment (EE 1224a4), or by a
child (ibid. 28), or under the influence of violent passion, since 7T. is
f.L£Ta Aoyov Kat Otavolas (lj:N III2ai5 ; cf. schol. Eur. Med. 899) ; Medea
does not kill her children KaTa 1rpoalp£aw. (A. could hardly have
denied that the tempestuous disposition which gave rise to such
violent passions was part of her �8os.) Hence we are told here oq8os
OTJAoi -r�v '1Tpoalp£atv, and only those .\6yot which contain a 7Tpoalp£ats
can reveal oq8os (cf. EN I4I7ai9 : OVK exovaw oi p.a8TJp.a'TLKOt Tf8TJ OT£
o vo€ '1Tpoalp£atv), though otavota can be shown without one. The
connexion between '1Tpoalp£ats and oq8os is a commonplace in A.
(cf. EN I I I Ib4 ff., I I39a3I, R. I395bl4). One would not expect to
find oq8os in a Messenger's speech ; . messengers decide nothing. But
it is obvious that decisions require also the ability to foresee
107
 COMMENTARY
consequences, and sometimes A. makes a place for 8u:fvo£a . too,
·
e.g. Met . ro65a32 : 'Tf'poalpfu's ov xwp�s 8£avolas, and Ph. r¢br8 and 22
where the two words are synonymous : see Ross's note.
In the 'drama of ideas', Sha:w or Pirandello, 8ufvo£a assumes
pre-eminence. Whether A. admitted the existence of such a type
depends on the interpretation of soar2-I4.
50b9, [£v ots o(uc ECM'L 8i]>..ov _ -• .] : either. the 11'poalpfu£s would be
inexplicable but for the �Oos in the speech, or Tl 'Tf'O£TJTEov may be
supplied after SijAov. But the clause is absent in Ar. and . can
be dispensed with. .
There is another serise of�Oos which Vahlen, Gesammelte, pp. 257 ff.,
would introduce here. A speaker's persuasiveness depends partly
on the personal confidence which he inspires ; he may set out to co,n­
vey an impression of a particular sort of personality according to the
nature of his audience and his subject. In rhetorical writings this is
the sense of �Oos most commonly referred to, the character which
the speaker projects. Dion. Hal. Lysias 8, says of Lysias that he
never made a 11'pouw'Tf'ov ci.v718o'TJ'ol71Tov. A. discusses this in Bk. 2 of
his R. This �Oos could be independent of 11'poo:.lpfu,s.
The characters in a play are usually represented as expressing
themselves sincerely, revealing themselves as they are. No doubt the
playwright in representing them uses some of the �arne skills as the
speech-writer putting across a per�onality which will make a favour­
able impression. Occasionally he will go a stage further and re­
present a wolf representing, himself in sheep's clothing, as Creon
in OC 728-6o. · A.'s tendency to identify the dramatist with his
characters makes it difficult to give a precise meaning to �Oos here.
So long as we regard a dramatic character as an independently
existing entity, this sense of �Oos is largely irrelevant, qut it becomes
relevant when we think of the dramatist using his art to present
characters.
For T]Buc� -rpaycp8la see 56ar.
50b12. eca.8o>..ou : such generalizations or yvwp.a' would not express
ultimate truths ; they would be general statements such as are con­
cocted with a view to the particular conclusions that the speaker
wishes to justify.
It is worth noting that there is little room for impassioned speech
within these categories. In later tragedy the expression of emotion
·was left increasingly to the musical parts. A speech like Medea's
(Med. Ior9-8o), which A. probably did not admire (cf. 53b28), con­
tains by these standards little �8os and less s,&.vo£a, though it does
end with a resounding yvwp.TJ, the generalization ou'Tf'f:P p.fylu-rwv ·

at-r,os KaKwv f3poTois ro8o. To many moderns this rather spoils the
e ffect;
ro8
 C O MM E N T A R Y
SObll-1 5. Fourth Part >..£�,,.
[Twv J1EV Mywv] : B.'s emendation T£vv lv .\6yw would mean 'of the
· •

parts in the medium of words' : cf. 47322.

SOb13. >..£�,s can often be rendered 'style' ,' but it covers the whole
process of combining words into an intelligible sequence. Cf. 49b34,
where the phrase T�V TWV p.E-rpwv avv8t:cnv . covers only metrical
language, but in the widest sense the process is the same.
TTJV 8ui. Ti]S ovo11aa£as EPJ1TJVE£av : 'communication by means of
words' ; ovo#Laala = ovof'aTa. From lpf'TJVt:vs, · an interpreter · of
riddling or foreign language, EPf'TJVt:la comes to mean communica­
tion and also, like -"�''' style. See the entry s.v. in the Glossary to
Rhys Roberts's Demetrius of Phalerum 1Tt:pl 'Epf''r)vt:las.
SObl·S . EXEL TTJV auTTJV 8Uvai'Lv : 'has the same effect' and so 'does the
same thing' (cf. 4739). Both in verse (l1Lf'£Tpwv) and in prose (.\oywv) ,
or in this context perhaps in speeches (in real life), .\£�L> is concerned
with communication, with putting the best words in the best place,
but not, of course, the same words in the same places in verse and in
prose. The language of lyric poetry' is probably covered by IL£.\os,
. and neither ep.p.ETpa nor ,\E�,, are relevant to it.

SObl S -lO. The two remaining Parts, p.€.\os and o,Pts.

SOb 1 6. 1)8uaJ16.Twv : cf. 49b28. At EN n7ob29 the point of ij8va1La is that
one does not need much of it.
SOb17. QTIXVOTaTov • • TjKLaTa oiKeiov : the qualities required by the
•

successful producer of visual effects are different from · those re­

quired by a poet, and though the wardrobe-master is allowed to have
a TEXVTJ (1. 20, below), his skill is mainly empirical.
SObl S. 8uvai'LS : A. is emphatic (cf. 53b4, 62312) that the 'effect' of
tragedy does not depend on its being performed. Indeed it appears
that at this date plays were written which were intended only for
reading, avayvwanKol (R. 14I3b12). The earliest ref. to the reading
of stage plays is Aiistoph. Ran. 52 : Dionysus read the Andromeda to
himself.
ciywvos tcai. lnroKpLTwv : '(performance at) a public contest and
actors'. Cf. 53327 C1KTJVWV Kal aywvwv, which is a hendiadys 'scenic
contests'. ay. KaL tJ1T, so constructed ·should mean 'actors' contests'.
There was in fact a prize for acting at Dionysia and Lenaea, hut it
would be pointless to mention it here.
SOblO. cJICEU011'0LOU : schol. to Aristoph. Eq. 230 tells us that none of the
aKEVo1ToLol. dared to make a portrait mask of Cleon for Paphlagon to
wear. Pollux 4 · HS suggests that masks and costumes were the main,
if not the only, concern of the CJKEV01ToLos.
A.'s preference for plot as opposed to character has won little

109
 COMMENTARY ·

approval from most of the critics from the late nineteenth century
onwards. To the generations which were pr�foundly influenced by
Bradley's Shakespearian Studies it was common doctrine that; as
Granville Barker once put it, the purpose of drama was to portray
character. Intetest in the inner life of the individual, which had
been developed by the great novelists of late Victorian times in
England, France, and Russia, caused exaggerated attention to
traits of personality which could be perceived in Shakespeare and
contributed to the spread of the belief that they must be contained,
could one but find them, in all great drama. In fact few Greeks were
interested in the analysis of states of consciousness and the study of
psychological developments,· and the scale of Greek drama allowed
few opportunities for revealing the uniqueness of individuals. The
character w4o appeared in a mask was naturally generalized spiri­
tually in the same way as he was physically and as the style of
masked acting must require. The difference becomes obvious if one
compares the number and variety of the situations in which Hamlet,
for instance, is presented w.ith the restrictions and simplicity of the
Greek stage, from which the background of ordinary life is ahnost
totally excluded. It is impossible to . deny that A. was right in his
priorities so far as concerns Greek tragedy, and it is surprising that
there should have been critics in his own time, as the tone of the
chapter implies there were, who thought character more important
·

than plot.
More recent crit.ics are less out of sympathy with A. 'In drama
•
characterization depends on function ; what a character is depends
on what he has to do in the play. Dramatic function in its turn
depends on the structure of the play ; the character has certain
things to do because the play has such and such a shape.' · Northrop
I•'rye, Anatomy of Criticism (Princeton, 1957), p. 171.

C H A PT E R S 7-14
Chapters 7-14, with the intrusive Ch. 12, deal with plot, its structure,
and its emotional effects. They contain a large proportion of the meat
of the P. Chs. 7-9 form the first part of this section, and in them the
general nature of the plot is discussed. Ch. 7 deals with plot in the
light of the previously agreed definitions, while Ch. 8 throws further
light by showing what it is not and the misapprehensions which have
caused plots to be badly constructed. In Ch. 9 it is shown that a well­
constructed plot represents a more general truth than history can'
usually reveal.
1 10