Академический Документы
Профессиональный Документы
Культура Документы
Antike Naturwissenschaft
und ihre Rezeption
AKAN-Einzelschriften
Band 10
Victor Gysembergh, Andreas Schwab (Éd.)
Le Travail du Savoir /
Wissensbewältigung
Mirjam E. Kotwick
The text of Aristotle’s Metaphysics came down to us in two versions, called the α- and
the β-version.1 Our direct access to these two versions is limited to the manuscript
tradition that began in the ninth century AD.2 The separation of α and β, however, very
likely took place before the end of the fourth century AD.3 As the frequent and some-
times radical disagreement between the α- and β-version shows, the text of the four-
teen books of the Metaphysics underwent significant alterations after they were
brought together into an edition, presumably in the first century BC.4 Ancient com-
mentaries written before the ninth century serve as witnesses to the state of the text of
the Metaphysics before the beginning of our manuscript tradition. They are therefore
crucial for understanding how the α- and β-version came about, and also for deter-
mining which of the two versions (if either) leads us more reliably back to the original
text of Aristotle’s Metaphysics.
The larger context of this paper is the exploration of the relation between com-
mentaries on Aristotle’s Metaphysics and the transmission of the Metaphysics text.5
Ancient commentaries offer us indirect access to Metaphysics exemplars that are often
much older than the manuscripts that came down to us via direct transmission. In ad-
dition to giving information about otherwise lost ancient exemplars, the commentaries
also give us important information concerning the question as to whether and how far
these commentaries are sources for emendations of or interpolations into the trans-
mitted text.6 For example, we already have evidence that Alexander of Aphrodisias’
* I am most grateful to David Sedley for his illuminating thoughts on this paper. I am
further in debt to Edmond Kotwick, Andreas Anagnostopoulos, Christopher Noble, and
Phillip Horky for their comments on earlier versions of this paper.
1 For a complete stemma codicum see Harlfinger 1979. For a detailed discussion of the
puzzle of the two versions of the Metaphysics text see Primavesi 2012a.
2 The earliest surviving manuscript is the Vindobonensis gr. 100 = J, witnessing to the α-
version.
3 See Christ 1886: VII, Jaeger 1912: 181, Jaeger 1957: ix-x, Alexandru 2000: 13-14,
Primavesi 2012a: 457-458, and Kotwick 2014: 3-7; 278-279.
4 To the question of the first-century BC edition by Andronicus of Rhodes see Moraux
1973: 45-94, Gottschalk 1987: 1083-1097, Barnes 1997: 21-63, and Hatzimichali 2013.
5 In my dissertation “Alexander von Aphrodisias und der Text der Aristotelischen Meta-
physik” (Kotwick 2014), I explored the relation between Alexander of Aphrodisias’
commentary and the transmission of the Metaphysics.
6 For the relation between Alexander’s commentary and book A of the Metaphysics see
Primavesi 2012a; for the relation between Alexander’s commentary and books A-Δ and
parts of book Λ of the Metaphysics see Kotwick 2014.
214 Mirjam E. Kotwick
7 See Primavesi 2012a, where it is shown that Alexander’s commentary was used, among
other sources, for an editorial revision of Book A of the β-version of the Metaphysics. In
Kotwick 2014, it is shown that the influence of Alexander’s commentary on the
Metaphysics text is confined neither to the first book nor to the β-version of the text.
8 Arist. Metaph. A 3, 983a24-b6.
9 For an interpretation of this section of the Metaphysics see Betegh 2012.
10 For the role of Nous in Anaxagoras’ system see Laks 2002 and also Sider 2005: 123-
147 and Curd 2007: 192-205.
11 Arist. Metaph. A 3, 985a10-29.
12 Aristotle discusses other aspects of Anaxagoras’ theory – and often more extensively –
in A 3, 984a11-16 (on this section see Mansfeld 2011; cf. also A 3, 984b18-22; A 7,
988a28) and esp. in A 8, 989a30-b24 (on this section see Primavesi 2012b).
On Aristotle’s Metaphysics A 4, 985a18-21 215
13 For the Greek text see Primavesi 2012a. The English translation given above is based on
Ross’ translation in Barnes 1984, but has been modified.
14 Betegh 2012: 133 also retains the α-reading and translates it thus: “For Anaxagoras uses
nous as a deus ex machina for the making of the world, for whenever he is at a loss to
say thanks to what cause something necessarily is the case, he drags nous in, but in
other contexts he alleges as the cause of things that come to be anything else rather than
nous.” (my underlining). By inserting “something” into the English translation Betegh
(just as Ross and Curd 2007 in their translations) slides over the oddity of a missing
subject in the relative clause introduced by the α-line. Cf. also Cherniss 1935: 234.
15 Primavesi 2012a: 450-451.
16 Primavesi 2012a: 451.
17 Primavesi 2012a: 451.
18 I am in debt to David Sedley for pointing this out to me.
216 Mirjam E. Kotwick
τε γὰρ in line a18, which must be taken in the sense of καὶ γὰρ19 and hence here means
‘both’, makes good sense only in the β-text. For, according to the β-version, Aristotle
says that Anaxagoras uses Nous both for the specific case of kosmopoiia and generally
whenever he is at a loss. On the α-reading, by contrast, the two ways in which Anaxa-
goras employs Nous are no longer two parallel, though distinct ways, but instead are
merged into each other and consequently appear to be basically the same: Anaxagoras
uses Nous both in the kosmopoiia and when he is at a loss in stating the cause of the
cosmos or the kosmopoiia. In the α-version, Anaxagoras’ two applications of Nous are
nearly indistinguishable and so they conflict with the purpose of the τε γὰρ … καὶ-
construction.
This textual detail leads to the second observation that once more reveals the
awkwardness of the additional α-line. In light of the broader context of the passage,
and given that Aristotle’s assessment of Anaxagoras is short and rather general here,
the specificity of the α-line appears as an alien element: Aristotle’s concern in this
chapter is not with the cause of the world (nor for that matter why it exists of neces-
sity), but with cause simply, and Aristotle argues that Anaxagoras does not employ
Nous well as a cause. The specifics of Anaxagoras’ failure given in the α-reading do
not connect with any of the questions discussed in this chapter of the Metaphysics. In
light of these considerations we may take it as most likely that the α-line was not in the
original text of the Metaphysics, but was a later addition to the α-text, or what
Primavesi 2012a calls an “α-supplement”.20
On the assumption that the α-reading is false and the β-reading should be ac-
cepted, the question arises: whence and how did these words come into the α-version
of the Metaphysics text? It is obvious that this line is not the result of a mere copying
error; clearly some thought-process is responsible for it. There are three extant ancient
Greek commentaries on this passage, and so it is reasonable to look there for insight
into what this process might have been. I start with the earliest extant commentary,
that by Alexander of Aphrodisias, written around AD 200. The following is what
Alexander says to our passage:
Alexander, In Metaph. 35,1-4 Hayduck
μηχανῇ δὲ τῷ [2] νῷ Ἀναξαγόραν εἶπε He [Aristotle] says that Anaxagoras em-
χρῆσθαι, ὡς ἐν ταῖς τραγῳδίαις οἱ θεοὶ ployed Nous as a deus ex machina, just as
παράγονται [3] ἀπὸ μηχανῆς ἐν τοῖς in tragedies the gods are carried in by a
ἀπόροις· ὃ καὶ αὐτὸς ἐξηγήσατο. stage device in impossible situations; this
Aristotle himself has explained.
λέγει δὲ τὰ αὐτὰ [4] περὶ Ἀναξαγόρου ἃ He says the same about Anaxagoras as
καὶ Πλάτων ἐν Φαίδωνι. did Plato in the Phaedo.21
19 See Denniston 1954: 536. Here are some parallel cases of τε γὰρ … καὶ in the sense of
“both…and”: EE 1219b1; Cat. 6b22-3; Cael. 268b25; de An. 417a10; APr. 71a3.
20 Primavesi 2012a: 439-458.
21 The translation of Alexander’s commentary is based on Dooley 1989, but has been
modified.
On Aristotle’s Metaphysics A 4, 985a18-21 217
22 Alexander’s paraphrase (παράγονται, 35,2) suggests that he read in his copy of the
Metaphysics παρέλκει (β) instead of ἕλκει (α) in line 985a20. This seems to hint at the
fact that the word ἕλκει (α) is part of the α-supplement, i.e., is a later modification of the
text, and that παρέλκει (β) is the original reading.
23 See Primavesi 2012a: 450-451.
24 Plato, Phd. 96a8-10: ὑπερήφανος γάρ μοι ἐδόκει εἶναι, εἰδέναι τὰς αἰτίας ἑκάστου, διὰ
τί γίγνεται ἕκαστον καὶ διὰ τί ἀπόλλυται καὶ διὰ τί ἔστι. / “I thought it splendid to know
the causes of everything, why it comes to be, why it perishes and why it exists.” (Trans-
lation by Gruber in Cooper 1997).
25 Phd. 97b8-c6: Ἀλλ’ ἀκούσας μέν ποτε ἐκ βιβλίου τινός, ὡς ἔφη, Ἀναξαγόρου
ἀναγιγνώσκοντος, καὶ λέγοντος ὡς ἄρα νοῦς ἐστιν ὁ διακοσμῶν τε καὶ πάντων αἴτιος,
ταύτῃ δὴ τῇ αἰτίᾳ ἥσθην τε καὶ ἔδοξέ μοι τρόπον τινὰ εὖ ἔχειν τὸ τὸν νοῦν εἶναι πάντων
αἴτιον, καὶ ἡγησάμην, εἰ τοῦθ’ οὕτως ἔχει, τόν γε νοῦν κοσμοῦντα πάντα κοσμεῖν καὶ
ἕκαστον τιθέναι ταύτῃ ὅπῃ ἂν βέλτιστα ἔχῃ· / “One day, I heard someone reading, as he
said, from a book of Anaxagoras, and saying that it is Mind that directs and is the cause
of everything. I was delighted with this cause and it seemed to me good, in a way, that
Mind should be the cause of all. I thought that if this were so, the directing Mind would
direct everything and arrange each thing in the way that was best.”
218 Mirjam E. Kotwick
Socrates goes on to describe how his hope was dashed when he actually read Anaxa-
goras’ book, for Anaxagoras, in fact, made no use of Nous in his explanations but
“mentioned as causes” only things like “air and ether and water.”26
Socrates’ hope was that Anaxagoras would use Nous “to explain why the world
is so of necessity” (ἐπεκδιηγήσεσθαι τὴν αἰτίαν καὶ τὴν ἀνάγκην, 97e2). These words
indeed sound similar to the words given only in the α-version of our Metaphysics
passage.27 We read there that Anaxagoras drags in Nous whenever he is at a loss to
explain why it (the world) exists of necessity: διὰ τίν’ αἰτίαν ἐξ ἀνάγκης ἐστί. In terms
of content and expression there seems to be a clear connection between Plato’s formu-
lation and the α-supplement. Could some scribe, following Alexander’s commentary,
have transferred the line from the Phaedo into Aristotle’s text? Not likely. The
differences are so great that this passage from Plato is not alone sufficient to account
for the α-line in the Metaphysics. The formal differences between the two expressions
rule out a copy-paste transfer of the line from Plato into Aristotle. Furthermore,
Aristotle’s and Socrates’ criticisms of Anaxagoras are not precisely the same. For
Aristotle (according to the β-version), Anaxagoras employs Nous only when he is at a
loss and therefore inappropriately. For Socrates, Anaxagoras makes no use of Nous
and therefore ignores the Good as principle.28
After Alexander the next oldest commentary on the Metaphysics passage is that
of Asclepius of Tralles. We will see that Asclepius’ commentary provides crucial
information concerning the origin of the α-line. Asclepius wrote his Metaphysics
commentary ἀπὸ φωνῆς29, that is, on the basis of lectures of his teacher Ammonius
Hermiae.30 Asclepius’ commentary heavily depends on Alexander’s commentary, and
in his commentary on books A-Γ, Asclepius draws several passages from Alexander.31
26 Phd. 98b7-c2: Ἀπὸ δὴ θαυμαστῆς ἐλπίδος, ὦ ἑταῖρε, ᾠχόμην φερόμενος, ἐπειδὴ προϊὼν
καὶ ἀναγιγνώσκων ὁρῶ ἄνδρα τῷ μὲν νῷ οὐδὲν χρώμενον οὐδέ τινας αἰτίας
ἐπαιτιώμενον εἰς τὸ διακοσμεῖν τὰ πράγματα, ἀέρας δὲ καὶ αἰθέρας καὶ ὕδατα
αἰτιώμενον καὶ ἄλλα πολλὰ καὶ ἄτοπα. / “This wonderful hope was dashed as I went on
reading and saw that the man made no use of Mind, nor gave it any responsibility for
the management of things, but mentioned as causes air and ether and water and many
other strange things.”
27 See Cherniss 1935: 234, Primavesi 2012: 451, and Betegh 2012: 133.
28 See Betegh 2012: 133. Cf. also Loredana Cardullo 2012: 308 n. 523.
29 For the term ἀπὸ φωνῆς see Richard 1950. Cf. also Hoffmann 2009: 615-616.
30 See the titles of the commentary books in Ascl. In Metaph. 1,2-3; 113, 1-2; 137,2-3;
222,2-3. There are also several references to the verbal utterances of his teacher through-
out Asclepius’ commentary. See Verrycken 1990: 204-210; Luna 2001: 99-106 and
Luna 2003. Despite the (possibly total) dependence of Asclepius’ commentary on the
work of Ammonius, I use the name ‘Asclepius’ to refer to the author of the content of
this commentary.
31 It is likely that Asclepius himself complemented the lecture notes of his teacher with
passages from Alexander’s commentary (see Tarán 1969: 8; Luna 2001: 108 with n. 36;
Loredana Cardullo 2002: 508). Nevertheless, it is safe to assume that Ammonius had an
On Aristotle’s Metaphysics A 4, 985a18-21 219
exemplar of Alexander’s commentary in front of him while lecturing (see Tarán 1969:
12).
32 See Phd. 98b1: ἑκάστῳ οὖν αὐτῶν ἀποδιδόντα τὴν αἰτίαν καὶ κοινῇ πᾶσι τὸ ἑκάστῳ
βέλτιστον ᾤμην καὶ τὸ κοινὸν πᾶσιν ἐπεκδιηγήσεσθαι ἀγαθόν· / “Once he had given the
best for each as the cause for each and the general cause of all, I thought he would go on
to explain the common good for all,…”.
33 In the Metaphysics we find the phrase in Δ 2, 1013a35 and Ζ 16, 1040b30.
34 Phd. 97c3-5: καὶ ἔδοξέ μοι τρόπον τινὰ εὖ ἔχειν τὸ τὸν νοῦν εἶναι πάντων αἴτιον, …· /
“and it seemed to me good, in a way, that Mind should be the cause of all.” See also
Ascl. 31,30: “καὶ πάντων αἴτιον αὐτὸν”.
35 Cf. Loredana Cardullo 2012: “in certi casi”. For the use of the idiom ἔν τισιν (“in some
cases”, “in certain moments”, “in certain states”) cf. Alex. De fato 176,5 Bruns: ἀλλὰ
220 Mirjam E. Kotwick
α-line, however, speaks not of just any moments, but of a crucial and significant
moment: Anaxagoras is at a loss when he has to say why the world exists of necessity.
Three things, then, speak against Asclepius having read the α-line. First, it would be
strange that Asclepius would read this line, only to hide this important information
behind the indefinite ἔν τισιν in his paraphrase. Second, the plural ἔν τισιν does not at
all square with the specificity of the singular moment stated in the α-line.36 Third, the
wording of the α-line is so close to Plato’s own formulation that again it would be
strange that Asclepius would see the similarity between Plato and Aristotle, which
would even strengthen his own understanding of the passage, but then not emphasize it
in his commentary. Asclepius’ indefinite ἔν τισιν therefore indicates that his Meta-
physics exemplar, just as the β-version and Alexander’s text, lacked a clear specifi-
cation of why and when Anaxagoras was at an explanatory loss.
Let us now read further in Asclepius’ text. After briefly discussing another sen-
tence in Aristotle’s text, Asclepius comes back again to the critique of Anaxagoras.
This time Asclepius has Alexander’s comments on the Aristotelian passage precede
his own.
Asclepius, In Metaph. 32,2-7 Hayduck
μηχανῇ δὲ [3] τὸν Ἀναξαγόραν εἶπε τῷ He says that Anaxagoras employed Nous
νῷ χρῆσθαι, ὥσπερ ἐν ταῖς τραγῳδίαις as a deus ex machina, just as in tragedies
οἱ δαί-[4]μονες παράγονται ἀπὸ the gods are carried in by a stage device in
μηχανῆς [καὶ] ἐν τοῖς ἀπόροις, <οἷον> impossible situations; as, for example, in
ἐν τῷ Ἱππολύτῳ [5] ἐφάνη ἡ Ἄρτεμις the Hippolytus Artemis appears to Theseus
ἀποροῦντι τῷ Θησεῖ· οὕτως οὖν καὶ ὁ in despair. The same holds also for Anaxa-
Ἀναξαγόρας ἡνίκα [6] ἀπορεῖ πρὸς τὴν goras: when he is at a loss, in respect to the
κοσμοποιίαν διὰ τίνα αἰτίαν ἐξ ἀνάγκης, making of the world, to say through what
τότε ἕλκει τὸν [7] νοῦν. ἐν δὲ τοῖς κατὰ cause it exists of necessity, then he drags in
μέρος πάντα μᾶλλον αἰτιᾶται ἤπερ τὸν Nous. But in the particular cases, he makes
νοῦν. a cause anything rather than Nous.
4 οἷον ex alt. rec. addidit Hayduck || 6 τίνα ex alt. rec. scripsi : τινα codd.
μὴν τὰ οὕτως ὄντα ἔν τισιν οὐ δι’ αἰτίας τινὰς προκαταβεβλημένας … / “But the things
which are in certain states in this way are not in them on account of certain causes laid
down beforehand…” (translation: Sharples; my underlining). Cf. also Alex. In Top.
454,5 and 479,23 Wallies. The meaning of ἔν τισιν in our passage of Asclepius’ com-
mentary becomes crystal clear in light of a comment Asclepius gives a few lines later.
In contrast to the meager treatment Aristotle gave to Anaxagoras’ use of the ‘efficient
cause,’ Aristotle treats Empedocles’ attempts more extensively. Asclepius quotes
Aristotle’s criticism of Empedocles’ principles Love and Strife as well as Empedocles’
unclear distinction between the two (Ascl. 32,16-17 = Alex. 35,9-10; on this excerpt of
Alexander in Asclepius’ commentary see Luna 2001: 111-112). Asclepius then says καὶ
λέγει ἐν τίσιν (Ascl. 32,17 = Alex. 35,10), that is ‘in the case of Empedocles – in con-
trast to the case of Anaxagoras – Aristotle says precisely in which cases he (E.) goes
wrong with his efficient cause’.
36 I will come back to this point when I analyze the second passage from Asclepius’
commentary.
On Aristotle’s Metaphysics A 4, 985a18-21 221
41 For the different functions of the moods in Greek see Kühner/Gerth II: § 567; p. 446-
453.
42 The troubles that the α-line causes in the α-version of the Metaphysics (see above) dis-
appear when the α-line is understood as part of Asclepius’ commentary.
43 Cf. 97d7-e2: καί μοι φράσειν πρῶτον μὲν πότερον ἡ γῆ πλατεῖά ἐστιν ἢ στρογγύλη,
ἐπειδὴ δὲ φράσειεν, ἐπεκδιηγήσεσθαι τὴν αἰτίαν καὶ τὴν ἀνάγκην…, quoted above.
44 Cf. Ascl. In Metaph. 31,28, quoted above.
45 Furthermore, it seems now likely that in line 985a20 the reading ἕλκει (α) instead of
παρέλκει (β) is due to Asclepius’ reformulation of the original wording in the Meta-
physics text (παρέλκει). The form ἕλκει might very well be part of the α-supplement
On Aristotle’s Metaphysics A 4, 985a18-21 223
In light of these changes that Asclepius makes in the other parts of Aristotle’s
sentence, it seems unreasonable to suppose that he suddenly would have quoted verba-
tim the words διὰ τίνα αἰτίαν ἐξ ἀνάγκης, τότε ἕλκει.46 Rather, it seems that he bor-
rows, again, Plato’s words, which must have been fresh in his mind, as he just mo-
ments ago had referenced the Phaedo (31,28). Since it can also be shown that the α-
line fits perfectly naturally in the syntactical context of the passage in Asclepius’
commentary but is very odd within the Metaphysics context, it appears that the phrase
has its original place in Asclepius’ commentary. All signs indicate that Asclepius is
responsible for the wording of this commentary passage, and hence it comes to light as
rather probable that the words of the α-supplement, which we find here in this com-
mentary passage, are Asclepius’ own ‘Platonic’ creation. Therefore it is reasonable to
conclude that the phrase entered Aristotle’s text from Asclepius’, not Asclepius’ from
Aristotle’s.
Scholars have observed a general correspondence between Asclepius’ commen-
tary and the α-version of the Metaphysics text, and have quite justifiably conjectured
that Asclepius worked from the α-version of the text.47 Yet we must be careful, for the
correspondence might be due not solely to the fact that Asclepius used an exemplar
from the α-version, but might also be due to an influence that Asclepius’ commentary
worked on the α-version. Further study must be done on the whole of Asclepius’
commentary before we have sufficient evidence to settle the issue. What can be said at
this point, however, is the following: The α-text that we can reconstruct from our
manuscripts displays the state of the α-version in the ninth century, and the state of the
α-version in the ninth century was very likely different from the state of the α-version
in the fifth century. As we do not at present know when the α-version acquired the so-
called α-supplements,48 the α-line could very well have come into the text after
Asclepius wrote his commentary, and indeed the evidence suggests that it did.
When we compile the evidence gathered up so far – the β-version of the Meta-
physics, Alexander’s copy and also Asclepius’ copy did not contain the words of the α-
supplement; we find the α-supplement in Asclepius’ discussion of the Aristotelian text
and we can explain them as recapitulation of Socrates’ critique in the Phaedo – it
seems reasonable to conclude that the words of the α-line are Asclepius’ own
‘Platonic’ creation. At some point in the transmission process, therefore, they must
rather than, as Primavesi 2012: 451 suggests, due to an accidental loss of the παρ after
the similar looking γὰρ (α, line 985a19). Asclepius uses the verbum simplex ἕλκει
(rather than παρέλκει) in the sense of “pulling in” also in 19,21: διὰ τί τὸ ἤλεκτρον
ἕλκει τὰ ἀχυρώδη / “why amber pulls (in) chaff”. (This passage is again taken over from
Alexander’s commentary: Alex. In Metaph. 16,5).
46 Concerning the missing copula ἐστί in Asclepius’ ‘version’ of the α-line see note 38.
47 See Ross 1924: clxi-clxiii; Rashed 2005: CCXVI-CCXVII; Primavesi 2012a: 408 and
458.
48 Primavesi 2012a: 458: “The supplements may have accumulated over an extended
period of time, so that one cannot date the α-version as a whole on the basis of indi-
vidual supplements.”
224 Mirjam E. Kotwick
have found their way from Asclepius’ commentary into the α-version of the Meta-
physics. It is not unheard of that the phrases of commentators found their way into
philosophical texts.49 Such an intrusion into the Metaphysics text could go back to a
gloss containing Asclepius’ explanatory reformulation. The gloss could have been
placed in the margin of the text by a scholar, who wished to explain what Aristotle
meant when he casually said that Anaxagoras was at a loss. This gloss could then have
been accidentally incorporated into the Metaphysics text at a later stage of the trans-
mission. Also possible is this: There are several α-supplements in the first books of the
Metaphysics that seem to go back to scholarly notes or teacher’s remarks on the
Aristotelian text. It could also be the case that someone regarded Asclepius’ reformu-
lation as so helpful that he deliberately put it into the Metaphysics text.
Before I move on to the third and final extant ancient Greek commentator on our
Metaphysics passage I would like to raise and discuss a possible objection to the inter-
pretation of the evidence I have so far offered. In the case of Alexander’s commentary
on the Metaphysics, it can be shown that he used but one Metaphysics exemplar when
composing his commentary.50 The situation is not quite so clear in the case of Ascle-
pius’ commentary: as I mentioned above, Asclepius’ commentary is based on notes
taken from a lecture given by Ammonius.51 Whether Asclepius himself contributed
something original, and if so how much, is a matter of some debate.52 Granting that
Asclepius carried out some sort of revision, adding his own thoughts now and then, one
may wish to entertain the following possible interpretation of the evidence: the first part
of Asclepius’ comments (31,26-28), which showed no evidence of the α-line having
been read, were based on the lecture of Ammonius, whose Metaphysics exemplar did
not contain the α-line; but the second part of Asclepius’ comments (32,5-7), in which
we read the α-line, were the product of Asclepius’ independent research, and his Meta-
physics exemplar did contain the α-line. This objection is no mere idle curiosity, but
indeed points to an area of research that must be undertaken53 when we want to serious-
ly investigate Asclepius’ commentary as textual witness of the Metaphysics text.54
In answering to this objection three points are to be made. First, as I argued
above, the second part of Asclepius’ comments (32,5-7) does not show signs of his
having read the α-line in his Metaphysics exemplar. Rather, the evidence strongly
speaks in favor of the fact that we are dealing with Asclepius’ own reformulation of
Aristotle’s argument, in which Asclepius writes the α-line on the basis of his reading
of the Phaedo. Second, it is more reasonable and economical to suppose that the
student Asclepius worked with the same version of the Metaphysics text as his teacher
Ammonius than it is to suppose that they, although they worked together, worked with
different versions. Further, if their texts were not in fact identical, the two might very
well have been two copies stemming from the same school exemplar. Thus it is un-
likely that Ammonius’ copy did not contain the α-line whereas the text of his student
Asclepius did. Third, we do not know (yet) how Asclepius worked into his com-
mentary the notes from Ammonius’ lectures, nor have all those parts of his com-
mentary that are based on those lectures been distinguished from all parts that are
based on Asclepius’ own independent research.55 So given the current state of scholar-
ship on the compositional principles of Asclepius’ commentary, Ockham’s razor obliges
us to conjecture that Asclepius and Ammonius worked from the same Metaphysics
exemplar, which, as the above arguments show, did not contain the α-line. And so, given
the arguments developed above, Asclepius is most likely the author of our α-line.
The third and final extant ancient Greek commentary on this passage of the
Metaphysics is the so-called ‘recensio altera of Alexander’. This commentary covers
the first two books of the Metaphysics and it appears to have been written by an
anonymous teacher sometime around the end of the sixth century AD.56 It heavily
depends on Alexander’s commentary, and also takes over whole passages from
Asclepius’ commentary. As a final step, I would like to look briefly at the evidence
this commentary contains.57 To our passage the anonymous author writes:
Anonymous, Versio altera (vid. Alex. In Metaph. app. 34,10-16 Hayduck)
διὸ καὶ φησὶν ὅτι Ἀναξαγόρας τε And therefore he [Aristotle] says: “And
μηχανῇ χρῆται τῷ νῷ πρὸς τὴν Anaxagoras uses Nous as a deus ex ma-
[11] κοσμοποιίαν, καὶ ὅταν china for the making of the world, and
ἀπορήσῃ παρέλκει αὐτόν. whenever he is at a loss he drags in Nous.”
εἶπε δὲ μηχανῇ χρῆσθαι ὡς ἐν ταῖς He says that he employed it as a deus ex
τραγῳδίαις [12] ἐποίουν οἱ ἐν τῇ σκηνῇ machina as in tragedies they carry in gods
τοὺς θεοὺς παράγοντες ἐν ταῖς ἀπορίαις on the stage in puzzling situations, to
πρὸς ἀντίληψιν τῶν ἀπορουμένων, provide understanding of the matters
puzzled over,
[13] οἷον ἐν τῷ Ἱππολύτῳ ἐφάνη ἡ as, for example, in the Hippolytus Artemis
Ἄρτεμις ἀποροῦντι τῷ Θησεῖ· οὕτως appears to Theseus in despair. The same
οὖν [14] καὶ ὁ Ἀναξαγόρας ἡνίκα holds also for Anaxagoras: whenever he is
ἀπορεῖ πρὸς τὴν κοσμοποιίαν διὰ τίνα at a loss, in respect to the making of the
ποτ’ αἰτίαν ἐξ [15] ἀνάγκης ἐστὶ τότε world, to say through what cause it exists
ἕλκει τὸν νοῦν, λέγων αὐτὸν αἴτιον of necessity, then he drags in Nous. But in
εἶναι· ἐν δὲ τοῖς κατὰ μέρος πάντα the particular cases, he makes a cause
μᾶλλον [16] αἰτιᾶται ἤπερ τὸν νοῦν. anything rather than Nous.
14 αἰτίαν] αἰτίαν ἐστὶ L
55 A starting point for this kind of investigation into Asclepius’ commentary is Luna 2001:
99-106. See also Loredana Cardullo 2012: 75-83.
56 See Golitsis 2014: 214-216.
57 Cf. Golitsis 2014: 210-211.
226 Mirjam E. Kotwick
Lines 10-11, which are clearly a citation from the Aristotelian text, strongly suggest
that the Metaphysics copy of the anonymous commentator did not contain the words of
the α-supplement. We easily recognize lines 11-12 as a paraphrase of Alexander’s
comments on the passage. Lines 13-16 are, besides minor differences, identical to lines
32,4-7 of Asclepius’ commentary, which I suspect to be the origin of the α-supple-
ment. Since Asclepius’ reformulation is also given in this commentary, it could also be
the case that the words of the α-supplement found their way into the α-version of the
Metaphysics via the anonymous commentary.58
By way of conclusion we can say that the line διὰ τίν’ αἰτίαν ἐξ ἀνάγκης ἐστί,
τότε ἕλκει in 985a19-20 of the α-text of the Metaphysics, although part of the
Aristotelian text as printed in the standard Metaphysics editions, was not contained in
the β-version, nor in the copies of Alexander, Asclepius or the anonymous commen-
tator. The line resembles a critical statement that Socrates gives about Anaxagoras in
Plato’s Phaedo. In his commentary on the passage, the Neoplatonist Asclepius com-
bines Aristotle’s critique with Socrates’ critique in the Phaedo and thereby writes a
new version of the Aristotelian passages. In this new formulation, we find the words
διὰ τίν’ αἰτίαν ἐξ ἀνάγκης, τότε ἕλκει, which are (besides the unnecessary ἐστί)
identical with those of the supplement in the α-version. Based on this evidence, it
seems a plausible explanation that at some time after the sixth century AD, but before
the beginning of the manuscript tradition in the ninth century AD, Asclepius’ ex-
planatory addition was incorporated into the α-text.59
58 The minor differences between the α-text and the recensio altera do not help to decide
from which of the two it was more likely taken over: Whereas the wording in the α-text
(διὰ τίν’ αἰτίαν ἐξ ἀνάγκης ἐστί, τότε ἕλκει) shares with the version in the recensio
altera the ἐστὶ (διὰ τίνα ποτ’ αἰτίαν ἐξ ἀνάγκης ἐστὶ τότε ἕλκει), it does not share the
additional ποτ’. Whereas the sentence in Asclepius agrees with the α-version by not
containing the ποτ’, it disagrees with it by not containing the ἐστὶ.
α-supplement διὰ τίν’ αἰτίαν ἐξ ἀνάγκης ἐστί, τότε ἕλκει, …
Al. rec. alt. διὰ τίνα ποτ’ αἰτίαν ἐξ ἀνάγκης ἐστὶ τότε ἕλκει …
Asclepius διά τινα αἰτίαν ἐξ ἀνάγκης, τότε ἕλκει …
59 The extent of the impact Asclepius’ commentary had on the transmission of the Meta-
physics text has not yet been determined. Further study is also needed to determine with
greater precision which parts of the commentary are Ammonius’ work and which go
back to Asclepius himself. Such information as this is crucial for ascertaining whether
Asclepius and his teacher Ammonius used the same Metaphysics exemplar or had ac-
cess to different ones. Then, we will be able to use Asclepius’ commentary more ex-
tensively and effectively for answering the question of how this commentary relates to
the two versions of the Metaphysics and to what degree it is the source of interpolations
into the text.
On Aristotle’s Metaphysics A 4, 985a18-21 227
Bibliography
Alexandru 2000: Alexandru, Stefan. “A New Manuscript of Pseudo-Philoponus’ Com-
mentary on Aristotle’s ‘Metaphysics’ Containing a Hitherto Unknown Ascription
of the Work.” In Phronesis 44, 1999, 347-352.
Barnes 1984: The Complete Works of Aristotle. The revised Oxford translation, ed. by
Jonathan Barnes, Vol. II. Princeton 1984.
Barnes 1997: Barnes, Jonathan. “Roman Aristotle.” In Philosophia Togata II. Plato
and Aristotle at Rome, edited by Jonathan Barnes and Miriam Griffin, 1-69,
Oxford 1997.
Betegh 2012: Betegh, Gábor. “’The Next Principle’. Metaphysics A 3-4, 984b8-
985b22.” In Aristotle’s Metaphysics Alpha, edited by Carlos Steel, 105-140.
Oxford 2012.
Bonitz 1848: Bonitz, Hermann. Aristotelis Metaphysica, pars prior. Bonn 1848.
Cherniss 1935: Cherniss, Harold. Aristotle’s Criticism of Presocratic Philosophy. Bal-
timore 1935.
Christ 1886: Aristotelis Metaphysica recognovit W. Christ. Leipzig 1886.
Cooper 1997: Cooper, John M. Plato Complete Works. Indianapolis 1997.
Curd 2007: Curd, Patricia. Anaxagoras of Clazomenae. Fragments and Testimonia. A
Text and Translation with Notes and Essays. Toronto/Buffalo/London 2007.
Denniston 1954: Denniston, J. D. The Greek Particles. Second Edition. Oxford 1954.
Golitsis 2014: Golitsis, Pantelis. “La recensio altera du Commentaire d’Alexandre
d’Aphrodise à la Métaphysique d’Aristote et le témoignage des manuscrits
byzantins Laurentianus plut. 87,12 et Ambrosianus F 113 sup.” In Textual
Transmission in Byzantium: between Textual Criticism and Quellenforschung,
edited by V. J. Signes Codoñer and I. Pérez Martín, 201-232. Turnhout 2014.
Gottschalk 1987: Gottschalk, Hans B. “Aristotelian Philosophy in the Roman World.”
In ANRW 36.2, 1079-1174, Berlin/New York 1987.
Harflinger 1979: Harlfinger, Dieter. “Zur Überlieferungsgeschichte der Metaphysik.”
In Études sur la Métaphysique d’Aristote. Actes du VIe Symposium Aristotelicum,
edited by Pierre Aubenque, 7-33. Paris 1979.
Hatzimichali 2013: Hatzimichali, Myrto. “The text of Plato and Aristotle in the first
century BC.” In Aristotle, Plato and Pythagoreanism in the first century BC. New
Directions for Philosophy, edited by Malcolm Schofield, 1-27. Cambridge 2013.
Hayduck 1981: Hayduck, Michael. Alexandri Aphrodisiensis in Aristotelis meta-
physica commentaria. Berlin 1891.
Hoffmann 2009: Hoffmann, Philippe. “What was Commentary in Late Antiquity? The
Example of the Neoplatonic Commentators.” In A Companion to Ancient
Philosophy, edited by Mary Louise Gill and Pierre Pellegrin, 597-622. Chichester
2009.
Jaeger 1912: W. Jaeger. Studien zur Entstehungsgeschichte der Metaphysik des Aristo-
teles. Berlin 1912.
228 Mirjam E. Kotwick