You are on page 1of 7

The Classical Quarterly

http://journals.cambridge.org/CAQ

Additional services for The Classical Quarterly:

Email alerts: Click here


Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here

The Manuscript Tradition of Simplicius'


Commentary on Aristotle's Physics i-iv

A. H. Coxon

The Classical Quarterly / Volume 18 / Issue 01 / May 1968, pp 70 - 75


DOI: 10.1017/S0009838800029116, Published online: 11 February 2009

Link to this article: http://journals.cambridge.org/abstract_S0009838800029116

How to cite this article:


A. H. Coxon (1968). The Manuscript Tradition of Simplicius' Commentary on
Aristotle's Physics i-iv. The Classical Quarterly, 18, pp 70-75 doi:10.1017/
S0009838800029116

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/CAQ, IP address: 138.251.14.35 on 31 Mar 2015


THE MANUSCRIPT TRADITION OF
SIMPLICIUS' COMMENTARY ON
ARISTOTLE'S PHYSICS i-iv
T H E following discussion1 of the manuscript tradition of Simplicius' com-
mentary on Aristotle's Physics i-iv originated in an examination of the tradition
of the fragments of Parmenides. It is therefore illustrated not only from Sim-
plicius but particularly from the texts of Parmenides quoted by him. This
will not be misleading, since, though many of these texts are quoted by
Simplicius more than once, there is little or no sign in any manuscript of
interpolation from one passage to another and it is not likely that any scribe
could have interpolated the text from an independent manuscript of Par-
menides.
The critical text of the first four books of Simplicius' commentary on the
Physics, which was published by Diels in Berlin in 1882 and is the foundation
for the text of many fragments of the Presocratics, was based on collations by
Vitelli of three manuscripts DEF and of a fragment of Book i in a copy made
by the scribe of E at the end of his text, which is called by Diels E a . Besides
these Diels lists a considerable number of later manuscripts which are (so far
as I have examined them, justifiably) ignored in his critical apparatus. The
total number of manuscripts listed by Diels of some part of Books i-viii is 44;
a further 25 not mentioned by Diels are listed in A. Wartelle's Inventaire des
manuscrits grecs d'Aristote et de ses commentateurs (Belles Lettres, 1963). I shall
argue that Diels seriously underrated both the value of F and the probability
of contamination between his manuscripts, and that in consequence his text
of some fragments of the Presocratics rests on a false foundation. It should be
said, however, at the outset that Diels's understanding of Presocratic thought
prevented him from going far wrong in the readings he adopted and printed.
The manuscripts on which Diels relies are probably all thirteenth century;
they are:
D ( = Laur. 85. 2.) This is described by Diels as 'madore et situ valde corru-
ptus ut difficilis sit lectu'; in fact the damp stain affects only the upper edge
of the manuscript, which is otherwise easy to read. Diels regards this as the
best manuscript, i.e. as the more careful representative of the better textual
tradition, of which the other main representative is E (Praefatio, p. vi). D
contains Book i and two-thirds of Book ii only. Diels notes that its character
changes after p. 304 of his printed text and it seems clear from his critical
apparatus that from this point to p. 347, where it gives out, it is copied from
a different source related not to E but to F.
E ( = Marc. 229.) This is rated by Diels as only a little less careful than D.
It contains Books i-iv but lacks the prologue and some pages at the end;
pages 52. 17 to 72. 11 (Diels) are omitted but the hiatus is supplied (apparently
from a different source) at the end of the main text.
E a . This fragment copied at the end of E from a separate source covers only
1
I am greatly indebted throughout my argument to criticism by my colleague Dr. W.
S. M. Nicoll.
MSS. OF SIMPLICIUS ON PHTSICS i-iv 71
pages 20. 1 to 30. 16, and 35. 30 to 44. 19 of Diels's printed text. According to
Diels these few pages 'ex uno omnium optimo ducta sunt exemplo, quod arche-
typo haud dubie antiquiorem sequebatur memoriam' (Praefatio, p. vii).
F ( = Marc. 227.) This manuscript contains all eight books of Simplicius'
commentary, save for the last half of Book viii, in place of which is sub-
stituted an extract from Philoponus' commentary, which is followed by scholia
derived from both Philoponus and Simplicius. In Diels's view F is 'deterioris
exemplaris propago' and, while contemporary with DE, of greatly inferior
value (ibid.).
In addition to these manuscript sources Diels quotes readings from the
Aldine edition of 1526, which was based on a manuscript (or manuscripts)
akin to F.
According to Diels the archetype of our manuscripts of Books i-iv contained
only those books, and the text of Books v—viii in F and a few other manuscripts
of the complete commentary is derived from elsewhere. This hypothesis would
be difficult to disprove but there is no evidence for it. Diels supports it only
with the remark {Praefatio, p. v) that the early division of the whole work
owing to its length into two parts containing Books i-iv and v-viii respectively
is preserved 'in optimis libris'. But two of Diels's 'best mss.' contain only two
books or less, viz. D with Book i and two-thirds of Book ii and G with Books
v—vi. And it is clear that, even if F is inferior to other existing manuscripts
containing Books i—iv and Books v—viii respectively, this would afford no
evidence that its text derives from separate sources.
The text of our manuscripts of Books i-iv in any case descends from a single
archetype, as is shown by transpositions and lacunae common to them all
(Diels, p. vi). Diels attributes the archetype to a date not much later than
Simplicius ('aliquanto post Simplicium') but he regards the exemplar of the
fragment E a as 'following a text which was certainly older than the archetype'.
E a , however, shares with all our other texts a number of errors which show that
it derives from a common post-Simplician source (e.g. the transposition of the
two sentences in 24. 9-12; Siaipovfievov (for apaiovfievov) 24. 29; the insertion
of ravnrjv yap 28. io; re fi.iv (for TO fxiv) 30. 8), and since, though fragmentary,
it is an important element in the tradition, it seems best to reserve the term
'archetype' for the source common to E a and DEF.
Of the three principal manuscripts which Diels uses in his edition he regards
F as 'much less valuable' than DE on the ground that 'its exemplar was care-
lessly copied and replete not only with lacunae but with interpolations both
conjectural and taken from authors quoted by Simplicius including Plato and
Aristotle' (Praefatio, p. vii). He observes also that the writer of F itself, a monk
called George (whose name he thrice misquotes as Gregory), had access to
more than one manuscript (cf. especially his marginal note at p. 556. 8 (Diels)
ravra ev iroXXoTs ov <f>epeTai). The only merit which Diels allows to F in com-
parison with D and E is that it is more accurate than E in certain minor
orthographical matters {Praefatio, p. viii). It can, however, be shown from
Diels's own edition that the tradition of which F is the chief representative is
at least as reliable as that surviving in DE. The text represented by F was,
as Diels notes, the main source of the manuscript tradition in the fourteenth
and fifteenth centuries, on which the Aldine edition was based, and Diels seems
to have been to some extent biased against F by his own discovery of the separate
tradition preserved primarily by DE.
7a A. H. COXON
Diels's account of the relation of his manuscript sources of Books i-iv may be
represented as follows:

He justifies his assignment of a unique position to E a on the ground that it


preserves the true reading against all other texts in a number of passages, of
which the most important are p . 21. 21 (Diels) avvairoSeUvvTai (crvva.TTo8et.K-
vvvai DEF) and 4 1 . 19 ovvwv [i.e. ovpavcov] (dvUJv [i.e. avBpamajv] D E F ) . There
is, however, at least one conspicuous passage which suggests a different view
of the relation of E a to the other manuscripts; in p. 22. 25, where Diels reads
TIap[J.€VL8rjs IIvprjTOS 'EXedrris, D E have IJapfievlSrjs wpiTrjs eXedrrjs, E a F
IJapfj.evi8rjg irvp rjroi. oeXas n. It should follow from the Dielsian stemma that
irvp TJTOI o-eXas TI was in the archetypal source of E a and F and accordingly
Diels is forced (p. vi, 1. 12) to treat the better reading of /?, the original of DE,
as a Byzantine emendation of these words. The improbability of this is plain;
it seems clear that -nvp TJT<H aeXas TI is a scribe's attempt to convert irvpiTT)-
aeXea-rrjs into familiar words. If this is so, the agreement in error between F
and E a suggests either that both represent a common tradition distinct from
that of DE (in which case the archetype already had TTvpirrjs for IlvprjTos)
or that there has been contamination between F and E a or their sources. A
number of other less striking errors shared by E a and F against DE point in the
same direction, e.g. 22. 32 ofxoicos DE, dvoyioLcos Ea, dvop,ola F ; 26. 24 aid DE,
Plato, del E a F ; 27. 17 eyevvrjaav D E , eyevvrjaav in eyiwqae corr. E a , eyiwqae F J
28. 14 <j>epeadai oVep D E , <f>epea6ai. Kal orrep E a F ; 39. 25 o ju.ei' D , 6 fiov E, fiev
om. E a F; 42. 12 </>r)olv rj DE, <f>r)ol Kal E a F; 21 eV TOVTOV DE, €K TOV E a F. One
case, where Diels seems to be mistaken in accepting the reading of EaF against
DE, is of particular interest for the text of Parmenides, viz. the second quotation
from him on page 30, which in Diels's text concludes
ovveKev OVK dreXevTr/TOV TO tov Oefiis elvar
eon ydp OVK emSetie'y" f [J.rj eov Se dv TTOVTOS e8elro.

In the second line DE have eViSee? against emSeves E a F and Diels correctly
notes that ^ eov, which he obelizes, was (as Simplicius' paraphrase shows)
read and written by Simplicius himself. Now it is not impossible that Simplicius
quoted Parmenides unmetrically (cf. fr. 8. 57); but in the first of the two
other places in which he quotes this line DE again read iniSees against en&eves
E a F (40. 6), while in the second (146. 6), though E agrees with F in reading
eTTiSeves, D still keeps eniSees.1 There is little doubt that imSees, which scans
by synizesis <~> w — and allows /AT/ eov to be retained, is the true reading and that
em8eves is an error due to the familiarity of the form in Homer (there is a
similar confusion in the manuscripts between the two forms in Hdt. 4. 130).
But Diels's stemma codicum prevented him from admitting this; for he would
then have had to believe either that the correct reading in DE was the result of
1
There is some inaccuracy in Vitelli's reporting of both these passages.
MSS. OF SIMPLICIUS ON PHYSICS i-iv 73
a
conjecture or that the agreement between E and F was due to accident
or contamination. Either view would in fact have been much easier to believe
than the view that the nearly true reading of DE in 22. 25 irvpiT-qs iXedrrjs
was conjectural; but in that passage he had no possibility of accepting the
text of E a F, whereas here it was just possible for him to do so. The result is
that the reading eVtSeue? has disfigured the text of Parmenides ever since.
A second different view of the textual tradition from that of Diels (again
with the alternative, if Diels's view is right, that there has been horizontal
as well as vertical transmission) is suggested by passages which can be used to
argue that F preserves a tradition independent of that of Diels's other manu-
scripts. Thus in the relatively brief passages which Simplicius quotes from
Parmenides there are two places in which F alone gives the true text against
D E E a , viz. 39. 3 8' ixplvavro F , S' eKplvovro D E , 8e KpLvov. TO E a ; 5 TJTTIOV ov
fiey' F, rjiTiov DEE a . Of three prose passages in which F is right against DEE*
in agreement two are quotations by Simplicius from Plato and here Diels
does not fail to take advantage of the opportunity of suggesting that the scribe
of F has interpolated his text of Simplicius from Plato: 26. 16 rjVTwa F Plato,
•rjv DEE a ; 21 iroiou (superscr. lov) F, TTOLOVV DEE a , 770116v Plato; 28. 32 TOVS F,
TTJS DEE a . In view of F's preservation of true readings in Parmenides, which
can hardly be interpolated, this suggestion is gratuitous and may possibly
account for Diels's refusal to accept the reading of F in 39. 5 as right, until he
came to edit Parmenides.
Against both sets of evidence quoted above stands that adduced by Diels
(p. vii) in favour of his own view of the unique value of E a . Some of this is
plainly worthless; e.g. in 38. 21 tf>rjal E a , <f>aal DE, Diels fails to note that the
word is omitted by F, so that his exemplar may have agreed with the text of
E a ; similarly in 38. 29 raml Ea, ravrrj DE, ravrd F, F may well have copied
carelessly the same text as appears in E a ; and in 40. 25 rrpos E a , Kal DE,
om. F, there is again nothing to show that F was not copying a text identical
with that of Ea (in any case the true text here, as Diels himself notes, is uncer-
tain) . The case for the uniqueness of Ea can only be supported by passages in
which Ea is right against DEF in agreement. Apart from the two major passages
adduced by Diels himself (21. 21 o-vvairoSeiKvvrai E a , o-waTroSeiKvvvai D E F ;
41. 19 ovvuxv E% avcSV DEF) I can find only the following: 26. 24 aT
Plato, dvoXXvofievov E a , aTroAzAvfxevov DE, dnoXeXvfiivo F ; 2J. 15
E a , imapxovTCov DEF; 28. 16 A^Brjplrrjs E a , avSrjplT^s D E F ; 30. 6 TavTov T'
iv Ea, ravrov re ov iv DF, ravrov re ov Kal iv E ; 9, TO iov E a , T' iov D, re iov E,
T€ov F ; 37. 13 reXtiov E a , Kal reAet'ou DEF; 21, iXiygti E a , e'Ae'yx" D E F ; 40. 8
81' o E a , Sto DEF; 41. 12 aneipov Ea, Kal aTreipov D E F ; 30 ovv E a , om. DEF.
These DEF errors are hardly more impressive than those of E a F listed above
and they are fewer in number than the E a F and DEE a errors taken together.
If we take into account also the half-dozen places in these pages where DF or
EE a or EF agree in error against EE a , DF, and DE a respectively, and if we
remember also that both D and E appear to have had alternative sources to
copy where their main source had major omissions, the conclusion seems in-
evitable that we have to do with an 'open' recension, in which, however, the
two manuscripts D and E are, as Diels thought, for the most part closely re-
lated and preserve a tradition distinguishable from that of F and E a . Whether
F and Ea represent a single tradition or whether Diels was right in regard-
ing E a as copied from a source independent of that of DEF, can hardly be
74 A. H. COXON
established. It is unlikely, however, that he was right in supposing that the
'archetype' in his sense, i.e. his hypothetical common source of DEF, was
written 'aliquanto post Simplicium'. For the error avUSv for ovvcov, which
is found in all manuscripts but Ea, seems likely to have originated from a
minuscule text, in which a and ov could be easily confused.1
More important for the text as a whole than the relation of E a to DEF is
the question of the relative value of F. I have already quoted Diels's view
(Praefatio, p. vii) that F is 'deterioris exemplaris propago. . . neglegentius
enim ex archetypo descriptum erat illius exemplar neque solum lacunis re-
fertum sed etiam interpolationibus vel ex ingenio petitis vel ex scriptoribus
a Simplicio citatis' etc. Nevertheless in p. 23. 22 (Diels) only F and its deriva-
tives preserve the name of Thales' father (e^afxvov F, c'/c adfj.ov DE, eKodBov E a ).
In the 180 lines (approximately, including repetitions) which Simplicius quotes
from Parmenides there are at least twenty places where the true reading is
given by F alone of Diels's manuscripts ; 2 and in Simplicius' quotations from
Empedocles there are many more places where the right reading is found only
in F. 3 The true reading is often easier to determine in these poetic passages
than in Simplicius' own text, where there is sometimes little to choose between
two variants, and in such passages Diels himself shows no hesitation in follow-
ing F. A glance at Diels's apparatus criticus shows further that F is regularly
right in the company of E a where DE are mistaken. In the scanty passages of
Parmenides (about thirty-three lines) for which E a exists there are at least
eight places where the true reading is given by EaF against DE and none
where DE are certainly right against E a F. And in the nineteen pages of Diels's
text for which E a exists, out of about 121 places where DEE a F offer variant
readings affording a prima-facie basis for analysing the manuscript tradition,
F agrees with E a against D and E (not necessarily in agreement) in about
eighty-four and of these agreements about seventy-three are adopted as true
readings by Diels himself. An analysis of Diels's critical apparatus to these
pages gives the following figures4
Total number of passages where DEE a F offer significant variants 121
Number of passages where Diels follows EaF against D and E 73
Number of passages where Diels follows DE against E a and F 13
Number of passages where Diels regards E a alone as right 16
1 3
I owe this point to Dr. W. S. M. Nicoll. e.g. Simpl. p. 32. 8 opfitoBeioa . . .
2
P. 30. 25 (Diels) avrla F, evavria D E ; 8' \ififveaaiv F, 6p/j,-qo$elaa . . . Xifieveaiv D E ;
CKpivavro F, 8e Kpivavro DE, 27 TJTTIOV OV F, 33. 2 iravTolais ISerjoiv F, TravToLaioiv ISecciv
TJTTIOV TO E, fjinov TO D ; 31. i KOT' CLVTO F, D E ; 34. 2 oi58a/i.d F, ouS* dfui DE.
4
KOTO. TavTo D E ; 2 raiTi'a F, ravavria D E ; The figures are approximate, since there
39. 3 8' eKplvavTo F, 8' (KpivovTo DE, Se are some passages which can be considered
Kplvov. TO E a ; 5 rjmov ov F, rjmov D E E a ; fidy' borderline cases and others where Diels omits
F, om. D E E a ; 78. 8 /J.6VOS F, fiovvos D E ; 12 to give the reading of one of his manuscripts.
ianv F, iv ianv D E ; 87. 1 vdvr' ovofii ICTTOI I have included among the readings of E a F
F, ITOLVT' 6v6/j.aoTai E, TTSM Tovvofia carat D ; accepted by Diels as correct three places in
23 T<S F+jro D E ; 145. 1 (IOVOS F, /iovvos D E ; which he adopted the text of DE in his
18 avorp-ov F, avovqrov D E ; 20 KC F, Kal D E ; edition of Simplicius but changed his mind
22 dirvoros F, airTvaros D E ; 146. 30 oyKco in Poetarum Philosophorum Fragmenta and
F, om. D E ; 180. 1 yvai/ias F 1 , yvdipais DEF 2 ; Fragmente der Vorsokratiker (viz. 23. 11, 13, 19)
6 KOT' (ZVTO F, Kara ravro D E ; 9 ovofiaoTai and among the passages where F alone gives
1 2
F , wvopaoTat. DEF . One of these (30. 25 8' the correct reading, one where Diels takes
iicpivavTo (does not appear in Diels's apparatus. this view in the same works though not in his
Add 86. 27 TOT (i.e.^T'lew) F, TO OV DE. Simplicius (39. 5).
MSS. OF SIMPLICIUS ON PHYSICS i-iv 75
Number of passages where Diels regards F alone as right 6
Number of passages where Diels follows DF against E a and E 4
Number of passages where Diels follows EE a against D and F 3 ^
Number of passages where Diels follows DE a against E and F 2
Number of passages where Diels regards D alone as right 3
Number of passages where Diels regards E alone as right 1
These figures show conclusively that Diels's descriptions of E a as 'uno omnium
optimo ducta . . . exemplo' and of F as 'deterioris exemplaris propago' are
incompatible. The two are of approximately equal worth and, since Diels's
attribution of a high value to E a is indisputable, it follows that for the pages
of Books i-iv for which E a is lacking, i.e. for all but nineteen pages of Diels's
text out of 800, F must be regarded, with due allowance for some manifest
carelessness in copying, as in the present state of our knowledge the main
representative of the better tradition of the text.
University of Edinburgh A. H. COXON

You might also like