• 1  The exact meaning and translation of the starting assumptions has actually been a matter of much c (...)
• 2  For a (partial) list of scholars that do not regard this section as a self-standing hypothesis, se (...)

1In the second part of Plato’s Parmenides, the eponymous character formulates two contradictory assumptions, “if the one is” and “if the one is not”,1 and scrutinises what follows from them for both the one and “the others” (τὰ ἄλλα, i.e. the things other than the one). Each combination of a starting assumption and a subject gives rise to two apparently contradictory series of deductions: one of them denies of the subject, while the other attributes to it, a given set of predicates (e.g. if the one is, then the one is neither in motion nor at rest, etc.; if the one is, then the one is both in motion and at rest, etc.). In the case of the positive starting assumption, Parmenides draws a further series of deductions concerning the one; this series (the third one) is mixed of both positive and negative deductions. Since, contrary to most modern commentators, the ancient interpreters generally considered this section to be on an equal footing with the others, there are, on the traditional account, nine series of deductions – which are usually called “hypotheses”.2

• 3  Plotinus, Enneads, V 1 [10], 8.23-26 (line numbering after Henry and Schwyzer’s editio minor, 1964 (...)
• 4  <κατ’> εὐθεῖαν (in Parmenidem VI, 1056.23 Luna-Segonds [= 1056.13 Steel]).
• 5  See the interpretations discussed by Proclus, In Prm. VI, 1056.1-1064.12 (= 1056.1-1064.12 St.). L (...)

2No later than Plotinus, Platonists started to conceive of the hypotheses as descriptions of different levels of reality, arranged according to a descending hierarchical order; in other words, any given hypothesis has a lower-ranked object as compared to the preceding hypothesis. Thus, Plotinus identified the “ones” of the first three hypotheses with the One above being, with intellect-being, and with the hypostasis soul, respectively.3 Unlike Plotinus, the subsequent philosophers provided a systematic interpretation of the entire second part of the dialogue. In this regard, two interpretative paradigms are to be distinguished: the earlier Neoplatonists proceeded, as Proclus puts it, “in a straight line”,4 that is, they simply extended the Plotinian interpretative model to all hypotheses; on the contrary, later Neoplatonists generally conceived of the last four hypotheses, which derive from the negative starting assumption, not (primarily) as descriptions of realities, but as proofs by contradiction of the preceding demonstrations.5

• 6  In Prm. VI, 1052.32-1064.12 (= 1052.25-1064.10 St.). For an overview of this section, see Saffrey- (...)
• 7  Loc. cit. in the preceding note.

3The principal task for a Neoplatonic commentator on the Parmenides consists, therefore, in matching the appropriate subject-matter to every hypothesis. Our main and almost sole source for the pre-Proclean part of this story is Proclus, who, in his commentary on the Parmenides, explains and discusses the previous Neoplatonic interpretations.6 Starting from Saffrey and Westerink’s introduction to Proclus’ Platonic Theology,7 this section has deservedly attracted scholarly attention; most contributions, however, simply expound the content of Proclus’ survey of his predecessors: especially in the case of the “straight-line” interpretations, they do not attempt to penetrate the meaning of the identifications proposed by these authors, when it is not transparent, nor to clarify the rationale behind them. In this paper, I shall address these questions in relation to the first authority discussed by Proclus, i.e. Plotinus’ disciple Amelius.

• 8  See Brisson 1987 (which is by far the most exhaustive presentation of Amelius).

4In particular, in section 2 I present the extant evidence about Amelius’ reading of the Parmenides; in doing this, I address (among others) the question of where he expounded his interpretation: I survey the possible answers suggested by Brisson in his fundamental study on Amelius8 and argue that he probably wrote a commentary proper on this dialogue (although we cannot be certain of this). In section 3, I reconstruct his unusual division of the second part of the dialogue into eight hypotheses (instead of the traditional nine); while this topic has occasionally been touched on in previous literature, it has not been thoroughly studied, nor have the two possible interpretative options been compared and discussed: a closer examination is therefore needed. Subsequently, I try to elucidate some further aspects of Amelius’ exegesis on which Proclus is less explicit than we may wish, and which modern studies have not adequately clarified, either: what exactly is the identity of the various layers of reality figuring in Amelius’ interpretation? In particular, on what basis did he discern no fewer than three material layers (section 4)? How are some anomalies in Amelius’ arrangement of the subject-matters to be explained, and what do they tell us about the governing principles of his exegesis (section 5)? In these two sections, I focus mostly (but not exclusively) on his reading of hypotheses 5-8, which he connected with the material realm, since this is both the most original and the most puzzling part of his interpretation. By answering the questions above, I try to disclose the logic underpinning his exegesis of the Parmenides and show that the latter, while resting on hermeneutic assumptions very different from ours, is not so haphazard as a superficial reading of Proclus’ report could suggest. In the conclusion (section 6), I show how the results of the previous analysis can contribute to a better understanding of the subsequent Parmenides exegesis.

• 10  In Prm. VI, 1053.9-33 (= 1053.8-27 St.).

6After this, Proclus comments on this interpretation: he approves of it for restricting the selection of the subject-matters to principles, but he criticises it for reducing the number of the hypotheses from nine to eight, and for twice disrupting the descending hierarchical ordering.10

• 11  See on this Dillon in Morrow-Dillon 1987, p. XXXV-XXXVI, and (less convincing) Tarrant 1993, p. 14 (...)
• 12  Both here and in the subsequent pages of the commentary, when presenting the opinion of some previ (...)
• 13  In manuscripts F and G, which belong to the so-called Greek branch of the tradition, the scholion (...)
• 14  As is stressed by Luna-Segonds 2017, p. 23, n. 7 (= n. c., p. 204), and p. 147.
• 15  Both Steel et al. (2007-2009, vol. I, p. XXV; cf. Steel 1999, p. 296) and Luna-Segonds (2007, vol. (...)
• 16  Dillon (2002, p. 330, n. 7) suggests another possible explanation of the scholiast’s remarkable kn (...)

7In his Parmenides commentary, Proclus leaves most of his predecessors anonymous;11 in keeping with this practice, he does not mention Amelius by name in these lines, but there is hardly any doubt that the generic phrase οἱ μέν (“some [commentators]”) alludes to him. This information comes from some marginal scholia on Proclus’ commentary, which disclose the names of most of the commentators referred to by Proclus in his survey; in particular, the scholion on 1052.32 (= 1052.25 St.) expressly identifies the interpreter at issue with Amelius.12 Since this note is found in both branches of the tradition of the Parmenides commentary,13 it must have been already present in the archetype,14 which probably dated from the end of the fifth or from the sixth century.15 The scholion is, therefore, very ancient; it could even be traced to the very Athenian school, which would explain why the scholiast is so well-informed.16

• 17  To my knowledge, no scholar has ever challenged the scholiast’s identifications, except for Tarran (...)
• 18  Theol. Plat. I 1, p. 6.16-7.8 (Amelius is mentioned on p. 6.21).
• 19  For a comparison of the two lists, see Saffrey-Westerink 1968, p. LXXV-LXXXIX, especially p. LXXXV (...)
• 20  Further evidence of the scholiast’s accuracy (at least as concerns his identification of the third (...)

8Since the scholiast’s identifications result in a perfectly plausible history of the Parmenides exegesis, there is no reason to question their reliability.17 Moreover, in the first chapter of the Platonic Theology, Proclus briefly surveys the “exegetes of the Platonic epopteia” (i.e., substantially, the Neoplatonic interpreters of the Parmenides);18 in this passage, Amelius, Porphyry, and Iamblichus are the first three commentators proper (that is, excluding Plotinus) mentioned by Proclus, exactly as in the list of names provided by the scholiast;19 this confirms, at least in part, the latter’s testimony.20

• 21  See Asclepius of Tralles, in Nicomachi Arithmeticam Introductionem I, § μδ’ [44] ll. 3-5 (p. 32) T (...)
• 22  Brisson 1987, p. 828.
• 23  See Proclus, In Prm. III 832.12-833.18 (= 832.9-833.14 St.), especially 832.13-17, 833.12-14 (= 83 (...)
• 24  D’Hoine (2010, p. 236-240). The identification of these philosophers with Amelius is judged likely (...)
• 25  Amelius’ position on the the extent of the realm of ideas is analysed by d’Hoine (2010), who consi (...)
• 26  See Brisson 1987, p. 826-828.
• 27  See Brisson 1987, p. 860.

9At this point, the question naturally arises as to where Amelius expounded his interpretation of the Parmenides. The most straightforward supposition is that he developed his exegesis in a commentary on this dialogue: for, the observations that Proclus reports in the above-quoted text are not just scattered remarks, but systematically cover the entire second part of the Parmenides. Furthermore, there is some (admittedly not conclusive) evidence to the effect that Amelius dealt with the first part, too: for, he is the only Platonist known to have committed himself to the thesis that there exist ideas of evils;21 and, as L. Brisson suggests,22 it is not unlikely that he defended this view in the context of a commentary on Prm. 130b1-e4, where Plato discusses the extent of the realm of ideas. Indeed, in his commentary on 130c5-d2, Proclus refers to some Platonic predecessors of his that had posited ideas of evils;23 even though he does not disclose their names, he is likely to have Amelius in mind, as is argued by P. d’Hoine:24 now, the fact that Proclus alludes to Amelius precisely in the context of his commentary on Prm. 130c5-d2 corroborates Brisson’s supposition.25 If this is so, the hypothesis that Amelius wrote a commentary on the whole dialogue appears even more likely. Note that, if Amelius did devote an exegetical work to the Parmenides, this was probably not an isolated case in the body of his writings: the content of several exegetical remarks by Amelius on other Platonic dialogues has come down to us;26 in particular, the nearly twenty comments on Timaeus passages that are preserved by Proclus27 can hardly stem from any work other than a commentary on the Timaeus.

• 28  Brisson 1987, p. 826.
• 29  This holds particularly for Amelius’ reflections on the extent of the realm of forms, which might (...)
• 30  This was precisely the fate of an anecdote concerning Amelius’ classroom teaching on the Timaeus t (...)

10On the other hand, it is possible to envisage two alternative hypotheses (also put forward by Brisson)28 concerning the original context of these observations: namely, they could come from some other work(s) by Amelius,29 or even be traces of his oral teaching in the context of Plotinus’ school, which were possibly preserved by Porphyry in his commentaries on these dialogues.30

• 31  Brisson 1987, p. 795.
• 32  In Prm. VI, 1053.34-35 (= 1053.28-29 St.; transl. Morrow-Dillon, modified; the Greek text is quote (...)
• 33  Cf. also Iamblichus’ chapter (κεφάλαιον) against Amelius and Numenius mentioned in Proclus, in Ti. (...)

12Of course, this conclusion should not lead one to take it for granted that Proclus based his account directly on Amelius’ Parmenides commentary. In fact, one cannot rule out that he relied on at least one intermediary source: according to Brisson, “plusieurs indices nous amènent à penser que l’École d’Athènes ne connut l’œuvre d’Amélius qu’à travers le témoignage de Porphyre et celui de Jamblique”.31 Indeed, at the end of the section about Amelius, Proclus himself refers to “quite a few other commentators” who “have dealt with these authorities [scil. Amelius] and written against them”;32 these “other commentators” are likely to be (some of) the interpreters subsequently examined by Proclus, such as Plutarch, Iamblichus,33 and Porphyry, who was an important source of information about Amelius and many others of his predecessors. Obviously, this does not exclude that, besides reading Amelius’ critics, Proclus had a first-hand knowledge of his commentary; but it certainly proves that he did not need to peruse his predecessor’s work to be informed about its content.

• 34  These formulae are quoted in Luna-Segonds 2017, p. 1, n. 5 (= n. c., p. 164-168, in part. p. 164).
• 35  See Proclus, In Prm. VI, 1057.6-1058.21 (= 1057.5-1058.16 St.), esp. 1057.8-10 (= 1057.7-8 St.) an (...)

13Time has now come to examine the content of the Proclus testimony translated above. The most striking feature of Amelius’ interpretation is that he divided the second part of the Parmenides into eight hypotheses: while most modern scholars subscribe to this position, the ancient commentators were almost unanimous in distinguishing nine hypotheses, as is indeed suggested by a prima facie reading of the text: the second part of the dialogue is manifestly divided into nine sections by as many explicit transitional formulae.34 Indeed, Amelius and the mysterious “philosopher from Rhodes” (who argued for a ten-fold division)35 are the only Neoplatonic commentators known to have proposed a different divisio textus.

• 36  ἴχνη (In Prm. VI, 1042.3, 5 [= 1042.3, 4 St.]).
• 37  In view of this, Saffrey’s (1984, p. 4) scepticism about the possibility of reconstructing Amelius (...)

14As Proclus does not go into the details of the commentators’ divisiones textus – nor does he usually clarify the reasons for their subdividing the second part the way they did –, it is up to us to determine what chunk of text corresponded to each hypothesis. Luckily, we are considerably helped in this task by the nine transitional formulae – the “traces”, to resume Proclus’ image,36 through which Plato highlights the structure of the text and divides it, as I mentioned, into nine sections: since this division is so explicit, it is only reasonable to assume that the advocates of a nine-fold divisio textus took each hypothesis to coincide with one of these sections (the beginning of each hypothesis being marked by a transitional formula), and that Amelius, like most modern commentators, arrived at the number of eight hypotheses by merging two of them, while leaving the others as they are.37

• 38  Dillon (2002, p. 331-332) takes this to be the only possible explanation of Amelius’ identificatio (...)

161) Amelius possibly holds, like the modern advocates of an eight-fold division, that the third section (Prm. 155e4-157b5) is not a self-standing hypothesis, but, rather, a part of/an appendix to the previous one(s).38

• 39  As is suggested by Corrigan (2010, p. 39-40).
• 40  Οὐκοῦν καὶ συλλήβδην εἰ εἴποιμεν, ἓν εἰ μὴ ἔστιν, οὐδὲν ἔστιν, ὀρθῶς ἂν εἴποιμεν; – Παντάπασι μὲν (...)
• 41  Soph. 237b7-239c9.
• 42  The supposition that Amelius considered the last section devoid of any positive content could also (...)

172) The other possibility is that he denied the title of hypothesis to the last section (165e2-166c2):39 here, Plato briefly argues that, if the one is not, the others have none of the previously discussed properties, and closes the deduction proper (before the general conclusion of the dialogue) by these words: “– In short, if we were to say, ‘if one is not, nothing is’, would we not be speaking correctly? – Absolutely”.40 Since this is the most cursory of the nine sections, and the most radically negative in its conclusions, Amelius possibly thought of it as utterly devoid of any content. Alternatively, he may have maintained that, after exhausting everything that can be said to exist in some way or another (including the non-being of bare matter), Plato last turns to absolute non-being, with which he deals, as far as possible, also in the Sophist;41 this would be in keeping with the descending hierarchical order generally observed by Plato on Amelius’ interpretation. Or he might have attributed to the last section roughly the same function that the philosopher from Rhodes, Plutarch, Syrianus, and Proclus would later assign to the last four hypotheses, and Damascius to the seventh and the ninth one: namely, to prove that, if the one is-not in the sense that it is utterly non-existent, the unacceptable consequence follows that nothing is. Be that as it may, one can surmise that Amelius assigned no positive content to the last section and therefore treated it as a sort of appendix to the previous one.42

• 43  I treat this hypothesis as a positive one, although it contains a minority of negative deductions.

• 44  This undesirable consequence is acknowledged by Dillon himself (2002, p. 338, n. 19).
• 45  Note that the matter endowed with ἐπιτηδειότης, as an intermediate layer, can fit both a positive (...)

19This consequence is, however, highly problematic. For, hypothesis 6_A, which deals with matter as already ordered and as having received the forms in actuality – that is, matter endowed with the whole spectrum of formal determinations – ends up corresponding to hypothesis 7_T, where every possible determination is excluded from the one-that-is-not; and hypothesis 7_A, which discusses prime matter in itself, as bereft of any formal determination and even of the fitness for the participation in forms, coincides with hypothesis 8_T, where all possible properties (or, at least, the appearance of them) are attributed to the others. Analogous considerations hold for enmattered forms (hypothesis 8_A). On the first reading, then, Amelius’ ordering of the various levels of matter becomes completely incomprehensible.44 Conversely, on the alternative reconstruction Amelius’ exegesis is totally consistent with the quality of the hypotheses.45

• 46  On this, see below, section 5.

20It is to be noted that the perplexing matchings resulting from the first reading cannot be explained by appealing to the hierarchical principle; for, if Amelius were prepared to ignore the quality and content of the hypotheses for the sake of a strict adherence to the hierarchical ordering, he would not place the intermediate material layer (i.e. matter with fitness for participation) before both the highest and the lowest one.46

• 47  Of all the commentators discussed by Proclus in the survey of in Parmenidem VI, Iamblichus was the (...)

21In addition, even though most modern scholars agree in treating hypothesis 3_T as a part of the second hypothesis, there is, in principle, no reason to believe that an ancient interpreter should question its autonomy any more than that of any other hypothesis. In fact, the interpretation of hypothesis 3_T is one of the longest- and best-established: together with the second and the third one, this hypothesis is the only one that Plotinus explicitly discussed, and, contrary to most hypotheses, its interpretation among Neoplatonists was subject to very few variations.47

• 48  Proclus, in Ti. III, 287.16-288.20 (= vol. II, 213.9-214.4 D.).
• 49  Brisson 1987, p. 839.
• 50  Note that this passage excludes vegetative life from the domain of irrational souls (and, incident (...)

23Let us now turn to a more detailed analysis of Amelius’ interpretation. His reading of the first three hypotheses essentially coincides with the traditional Neoplatonic one, which was initiated by Amelius’ teacher, Plotinus: they are associated with the One, intellect, and the soul, respectively. Contrary to Plotinus (and like Iamblichus), however, Amelius stretches the soul over two hypotheses (3 and 4): for, he distinguishes between rational (λογικῶν) souls, which are the subject-matter of the third hypothesis, and irrational (ἀλόγων) souls, which are the subject-matter of the fourth one. The same distinction can be extracted from a testimony of Proclus’ about Amelius’ interpretation of Timaeus 35b:48 the world-soul – Amelius claims – holds together all the (ensouled) encosmic beings, that is, encosmic gods, demons, men, and irrational (ἀλόγων) beings. As Brisson remarks, “le domaine de l’âme se subdivise ici en deux sous-domaines” – that is, the rational part, comprising the first three psychic classes, and the irrational part, coinciding with the last one – “correspondant respectivement à la 3ème et à la 4ème hypothèses dans l’interprétation de la seconde partie du ‘Parménide’, qui aurait été celle d’Amélius”.49 In particular, the class of the irrational beings is further subdivided into domestic and wild animals; in view of this, it is clear that the irrational souls are those of (non human) animals.50

• 51  Plotinus argues for the total indeterminacy of prime matter in II 4 [12], 6-16; II 5 [25]; III 6 [ (...)

24The most intriguing part of Amelius’ interpretation is, however, that which concerns hypotheses 5-7. How are the three layers of matter to be understood? Clearly, the matter deprived of forms and of ἐπιτηδειότης cannot but be prime matter, which is void of any property; its utter indeterminacy is an important tenet of Plotinus’ philosophy.51 The identity and the status of ordered matter (as well as those of enmattered forms) are no less transparent: this is, as Proclus himself clarifies, formed matter, i.e. that which is endowed with formal determinations and has thus given birth to the bodily level. Of course, all distinction is a purely conceptual one, since the genesis of the sensible world as a consequence of matter’s reception of forms does not happen in time.

• 52  Cf. Hauer 2016, which focuses on the role of ἐπιτηδειότης in Simplicius’ treatment of quality, but (...)

25The exact significance of the intermediate level, which corresponds to the fifth hypothesis, is more difficult to grasp. The notion of ἐπιτηδειότης is, in itself, quite elusive and multi-faceted,52 but its fundamental meaning in this context is comparatively plain: Proclus is basically referring to matter’s capacity or suitability to receive the formal determinations. But what, precisely, is matter with ἐπιτηδειότης? What is the point of positing a further layer in addition to prime and ordered matter, and on what ground can it be envisaged as a conceptually self-standing level in the constitution of the sensible world?

• 53  Corrigan 2010, p. 40 (Corrigan’s italics).
• 54  See Enn. II 4 [12], 1-5 and Narbonne 1993, p. 47-134, 309-324.
• 55  Cf. Enn. II 4 [12], 5.28-35; I subscribe to Narbonne’s understanding of these obscure lines (Narbo (...)
• 56  Cf. Enn. V 2 [11], 1.13-18, and O’Brien 1999, p. 51-53.
• 57  Note that Plotinus sometimes describes soul as matter in relation to intellect (albeit in a purely (...)

26Corrigan surmises that this may be “some kind of intelligible or psychic matter”.53 This would explain why the matter characterised by ἐπιτηδειότης precedes ordered matter; if, on the contrary, we were dealing with some sort of prime matter to which an inclination towards participation has been added, it should follow ordered matter, in keeping with the descending hierarchical ordering of the identifications. Corrigan does not elaborate on these superior kinds of matter, but he presumably has in mind some theories of Plotinus’, which his pupil may well have adopted. Intelligible matter is an aspect or “moment” of the second hypostasis:54 it is some primordial indeterminacy out of which, as it turns towards the one, the many distinct ideas come about.55 As for psychic matter, this notion is not found as such in the Enneads; however, soul derives from intellect in much the same way as intellect derives from the one: the unformed product of intellect’s generative power becomes formed (and, therefore, soul proper) as it turns towards its source.56 This means that there are, in soul, both a formal and an unformed and material aspect.57

• 58  Note the presence of time adverbs in Proclus’ account of the fifth and sixth hypotheses: “already” (...)
• 59  In Prm. VI, 1053.19-24 (= 1053.16-20 St.).
• 60  Ibid. 1053.24-33 (= 1053.20-27 St.): see below, section 5.

27I find Corrigan’s reconstruction unconvincing. First, the way Proclus structures and phrases his exposition leaves no doubt as to the fact that, in Amelius’ view, hypotheses 5-7 all deal with the same matter, i.e. the matter of bodies, only envisaged in different phases of its formation.58 Moreover, Corrigan’s attempt to harmonise Amelius’ exegesis with a descending hierarchical ordering is misguided: for, Proclus’ criticism testifies that this ordering was not followed in the interpretation of hypotheses 5 and 6,59 and that there had even been attempts to account for the violation of this principle;60 these attempts rest on the premise that the matter of the fifth hypothesis has an intermediate rank between that of the sixth and that of the seventh hypothesis, which would not be the case if we had to do with some kind of supra-sensible matter.

• 61  Ti. 53B2.
• 62  Transl. Zeyl.

28The matter endowed with ἐπιτηδειότης must therefore be one of the layers of the matter of the sensibles. More precisely, I surmise that its nature can be better understood in the light of the Timaeus description of the “material” world. In 48e2-52d1, Timaeus introduces the notion of χώρα, the “space” in which the images of the forms come to be and perish, which is, in itself, utterly indeterminate. He then (52d2-53c1) goes on to explain that, prior to the ordering of the universe by the demiurge, the χώρα is subject to unbalanced forces: the elements proper in the final stage of their development do not exist yet, since the elements only “possess traces (ἴχνη) of themselves”;61 these elements in form of traces or, so to speak, element-traces, disorderly shake the χώρα and are shaken by it in turn. The notion of disorder already emerged in a previous passage (30a3-5), where Timaeus says that “the god [...] took over all that was visible – not at rest but in discordant and disorderly motion (κινούμενον πλημμελῶς καὶ ἀτάκτως) – and brought it from a state of disorder to one of order”;62 the theme of the demiurge’s imposition of order onto disorderly entities, which do not yet correspond to the names we ordinarily use (fire, water, etc.), resurfaces for the last time in 69b2-c3.

• 63  I confine myself to referring the reader to some significant contributions on the topic: Vlastos 1 (...)
• 64  Plotinus’ metaphorical reading of the Timaeus description of the generation of the cosmos has been (...)
• 65  Aristotle, Physics IV 2, 209b11-16; among the many studies on this controversial Aristotelian read (...)
• 66  This identification is particularly evident in in Ti. II, 265.4-270.11, esp. 265.7-13 (= vol. I, 3 (...)
• 67  See Philoponus, Contra Proclum VI 14, p. 164.18-23 (= Porphyry, fr. XLVII, p. 30.9-32.1 Sodano), 1 (...)

29The meaning and the mutual relation of these passages have been controversial since antiquity. I shall leave aside the modern debate63 and confine myself to some quick remarks about the ancient discussions. Generally speaking, the pagan Neoplatonists, starting from Plotinus,64 subscribed to a metaphorical reading of the Timaeus account of the production of the cosmos: they held that the world is generated not in the sense of having a beginning in time, but, rather, in the sense of being eternally dependent on higher causes; and they rejected the idea that the world is brought about by a demiurge acting like a human craftsman – that is, through a process that requires deliberation and consists of several successive steps. Accordingly, these interpreters were of the opinion that, in the passages mentioned above, Plato is not describing a process happening in time, but, rather, the structure of the material world: he is just describing in a mythical fashion some stages that are distinguished only logically. Furthermore, ancient philosophers, starting from Aristotle, identified the χώρα (considered in itself, in abstraction from the element-traces) with prime matter,65 and assumed (less questionably) that the end point, i.e. the universe ordered by the demiurge, is the ordered cosmos in which we live. The reconstruction of what takes place in between was more contentious: some commentators, like Proclus, identified the visible in discordant and disorderly motion of Ti. 30a with the element-traces, thus obtaining only one intermediate phase;66 others, like Porphyry, conceived of them as two distinct layers, thus positing two intermediate levels.67

• 68  In Prm. VI, 1052.39, 1053.23 (= 1052.31, 1053.19 St.).
• 69  κοσμέω: 53b1; διακοσμέω: 53a7, 69c1.
• 70  κοσμέω: in Ti. II, 260.1; 267.10; 270.16-17 (= vol. I, 383.21; 388.27; 390.32 D.; cf. also In Prm.(...)

31My suggestion that Amelius’ interpretation of hypotheses 5-7 may have been connected with his Timaeus exegesis is justified by Proclus’ use of some significant words in his account. First, he employs the verb κοσμέω with reference to the perfected matter of the sixth hypothesis:68 this verb and some of it compounds are used in some of the above-mentioned Timaeus passages to refer to the transition from the second to the third layer (on the three-phase reading);69 furthermore, Proclus, too, often refers to this process by the verbs κοσμέω and διακοσμέω.70

• 71  In Prm. VI, 1053.24 (= 1053.20 St.).
• 72  The equivalence of the ἐμφάσεις τῶν εἰδῶν to the ἴχνη τῶν εἰδῶν unmistakably emerges from Proclus’ (...)
• 73  In Prm. VI, 1052.37-38; 1053.26-29 (= 1052.29-30; 1053.22-24 St.); cf. also (in the negative) 1053 (...)
• 74  In this passage, Proclus argues that it is the “father and maker”’s task to endow matter with ἐπιτ (...)

32Still more telling is the way Proclus characterises the subject-matter of the fifth hypothesis on Amelius’ interpretation. In his criticism of his predecessor’s views, Proclus refers to this matter as that “which possesses bare reflections of the forms” (ἐμφάσεις ἔχουσαν ψιλὰς τῶν εἰδῶν);71 interestingly, Proclus uses the noun ἔμφασις in the context of the Timaeus exegesis, precisely as an equivalent for ἴχνος (Ti. 53b2) – which means that he employs the term to refer to stage ii.72 Moreover, as we saw, Proclus presents the subject-matter of the fifth hypothesis as matter endowed with ἐπιτηδειότης for participation in forms;73 now, in In Prm. IV, 844.5-32 (= 844.4-26 St.), he associates this notion, too, with the second layer of the physical world: for, on his understanding, matter’s possession of the traces of the forms means nothing else than its being suited (its ἐπιτηδειότης) for a complete reception of the forms and the demiurgic order – a suitability which prime matter still lacks.74

• 75  Interestingly, analogous considerations hold for Porphyry’s interpretation of the Parmenides– whic (...)

33Clearly, then, Proclus took it for granted that the referents of the fifth and sixth hypotheses corresponded, in Amelius’ view, to the level of the element-traces and to the ordered cosmos, respectively.75 Naturally, we can fairly expect Proclus to be presenting Amelius’ interpretation in his own lexicon; hence, we cannot be completely sure that Amelius established a connection between the Parmenides and the Timaeus exegesis in so many words, and that these correspondences are not the product of Proclus’ understanding of his interpretation instead. Given, however, that the Timaeus had been regarded, since Aristotle, as the main Platonic authority on the “material” substrate(s), and that the latter figured so prominently in Amelius’ exegesis of the Parmenides, it was only natural that he should draw on the former dialogue in his interpretation of the latter.

• 76  In Prm. VI, 1053.19-24 (= 1053.16-20 St.).

34So much, then, for the content of Amelius’ interpretation and his distinction of three kinds of matter. A further question must now be addressed, regarding the order of the entities he accommodated to the various hypotheses. I already touched on an undisputed principle governing the Neoplatonic exegesis of the Parmenides ever since Plotinus: in the second part of the dialogue, Plato proceeds from the highest to the lowest-ranked entity. Amelius generally abides by this interpretative assumption, but he seems to infringe it twice, as Proclus himself remarks:76 for, he ranks a) the matter endowed with ἐπιτηδειότης above prime matter, and b) enmattered forms below all kinds of matter, despite their being rational principles and, therefore, supposedly closer to the one.

• 77  The addition is not signalled in Cousin’s edition, which is quite inaccurate by modern standards.

372) Cousin inserts an article between αὐτοῖς and ἐπιστήσασιν: παρ’ αὐτοῖς <τοῖς> ἐπιστήσασιν.77

• 78  Cf. Polybius, Historiae XXIX 12.6; Themistius, in Phys., p. 5.27 Schenkl; Proclus, in primum Eucli (...)
• 79  Cf. Zekl’s (2010) rendering as “Begründung” (“Dem ist nun zwar von seiten derer, die das so angeor (...)
• 80  Cf. LSJ s. v. ἐφίστημι, A VI (“give attention”); in this case the verb, being in the sigmatic aori (...)

39First, some remarks about the translation of the sentence, which has been variously rendered by interpreters. The basic meaning of the phrase λόγου τυγχάνειν παρά τινι is simply “to receive attention/consideration/mention by/among” (sometimes in the sense of “being renowned/esteemed among”) and therefore, with reference to a scientific discussion, “to be discussed, treated”.78 Since, however, the content of this λόγος, “discussion”, is precisely the explanation that immediately follows, which is introduced through γάρ – possibly with epexegetic force (πρώτη γάρ, φασίν, κτλ.: “first, they say,” etc.) –, one can also opt for the contextual rendering “receive an explanation from”.79 As for ἐπιστήσασιν (sigmatic aorist participle of ἐφίστημι), ἐπιστῆσαι, used absolutely, means “to pay attention to”, “to deal with or dwell on” a topic (sometimes in contrast to a summary treatment).80

• 81  In Prm. VI, 1052.32-33 (= 1052.25 St.).

40In view of this, the manuscripts’ text (option 1) can be translated as follows: “This has, indeed, received some explanation from them, when they carefully consider [this question]” (or: “who carefully consider this question”). ἐπιστήσασιν is a circumstantial participle expressing time/circumstance, and αὐτοῖς is a pronoun; since it can have no other referent than οἱ μὲν [...] εἰς ὀκτὼ τὰς ὑποθέσεις πάσας διελόντες who are mentioned at the beginning of the above-quoted passage of Proclus81 (“the commentators that divided the hypotheses as a whole into eight”, i.e. Amelius), the explanation that follows must be attributed to Amelius himself.

• 82  I owe this observation to prof. G.W. Most, whom I thank for a helpful exchange on this topic. I se (...)
• 83  In Prm. VI, 1053.34-35 (= 1053.28-29 St.): Ἀλλὰ τούτοις μὲν καὶ ἄλλοι πλείους καὶ ἐπέστησαν καὶ ἀν (...)

41The phrase παρ’ αὐτοῖς ἐπιστήσασιν, however, presents some rhetorical and stylistic/syntactic problems. If the authority that conceived this explanation was Amelius himself, Proclus’ precision that he paid attention to this question is obvious and pleonastic. This redundancy is problematic also from a stylistic point of view: for, the resulting syntagma (in which a preposition governs a pronoun to which a circumstantial participle is attached), although not strictly ungrammatical, sounds very harsh and unnatural,82 especially for such an elegant writer as Proclus is. On top of that, a few lines later, the same verb ἐφίστημι (this time in the root aorist, but with the same meaning) is used for the critics of Amelius.83

• 84  Like Porphyry, Iamblichus, and Plutarch (see above, p. 35). Proclus might be alluding to either on (...)

42Although none of these considerations is conclusive in itself, taken together they do arouse the suspicion that there is something wrong in the phrase παρ’ αὐτοῖς ἐπιστήσασιν, and thus lend support to Cousin’s (2) and Luna-Segonds’s (3) emendations. On both of these reconstructions, ἐπιστήσασιν is preceded by an article and used as a substantival participle; thus, the ἐπιστήσαντες are presented as distinct figures from the author whose interpretation they discuss: other commentators84 examined Amelius’ interpretation and tried to figure out the rationale behind its apparent irregularity.

• 85  ἐφίστημι does sometimes have a polemical force, but only when it is intransitive and, therefore, n (...)

43In particular, 2) Cousin supposes that an article dropped out (by haplography) after αὐτοῖς, which he retains; since it stands in predicative position, αὐτοῖς is being used emphatically. Considering that putting emphasis on ἐπιστήσασιν does not make sense if the latter has the neutral meaning “to consider carefully”, “to deal with”, this integration presupposes that the verb has a more hostile connotation (cf. Morrow-Dillon’s translation as “critics”). Since, however, ἐφίστημι actually has no polemical force (at least, not in the first aorist),85 Luna and Segonds’s conjecture is more plausible: 3) they extract the article from αὐτοῖς, supposing that a confusion occurred between two phrases that differ by one letter only (παρὰ τοῖς > παρ’ αὐτοῖς); this results in the following translation: “This has, indeed, received some explanation from those that have carefully considered [this question]”.

• 86  In Prm. VI, 1053.26-33 (= 1053.21-27 St.).

45On their reading, the matter endowed with ἐπιτηδειότης is placed first because a) it is clearer (τρανεστέρα: that is, presumably, more graspable, being closer to form) than bare matter and obscurer (ἀμυδροτέρα) than the ordered one; and b) it is both more representative of the nature of matter than the latter and more understandable than the former.86

• 87  I take it that this is the gist of the explanation proposed by the ἐπιστήσαντες, even though Procl (...)

46In other words, the ἐπιστήσαντες tried to account for the irregularity on the basis of two criteria, i.e. how representative each layer is of the true nature of matter – which is indeterminateness and dimness –, and how accessible it is to our understanding; both prime and ordered matter should come first according to one of these criteria, but last according to the other, whereas the matter characterised by ἐπιτηδειότης has, in both cases, the intermediate position: placing it first is, therefore, the most balanced solution.87

• 88  See below, section 6.
• 89  Proclus, In Prm. VI, 1055.9-12, 14-16 (= 1055.8-10, 12-14 St. = Iamblichus, in Prm., fr. 2.11-13, (...)

49Indeed, any entity endowed with some formal determination can hardly fit with the nihilistic conclusions of the negative hypotheses; now, the more distant an entity (save the one) is from form and determination, the lower its rank: consequently, the negative quality of a hypothesis is taken as a sign of debasement. This is particularly evident when the same entity, considered in different aspects or stages of its constitution, spans over more than one hypothesis: the highest level (for instance, formed matter) corresponds to an affirmative hypothesis, whereas the lowest one (like prime matter) is matched to a negative one. For the sake of brevity, I shall refer to this exegetic rule, which I have already touched on indirectly in section 3, as the “principle of alternation” (note, however, that the alternation is not an a priori requirement, but a mere consequence of the necessity of conforming to Plato’s text). This criterion is clearly recognisable in Porphyry’s interpretation of hypotheses 4-9,88 as well as in Iamblichus’ interpretation of hypotheses 4-5 and 8-9.89

• 90  This is not, of course, the only possible solution to the second problem: in particular, Amelius c (...)

51When, like in our case, a single entity (i.e. matter) stretches over three hypotheses, the problem is particularly hard to solve: since every hypothesis after the fourth one is followed by a hypothesis of the opposite quality, if Amelius were to stick to the hierarchical principle, he would end up matching the highest and the lowest level to hypotheses of the same quality. But this would clash with the principle of alternation and result in an interpretation at odds with Plato’s text: for, whilst the intermediate level is more or less suitable for either kind of hypothesis, the highest level (ordered matter) and the lowest one (bare matter) can only correspond to a positive and a negative hypothesis, respectively. In addition, the first of the three hypotheses associated with matter (i.e. the fifth hypothesis) has negative deductions, which makes it impossible to treat it as a description of ordered matter. In view of this, Amelius cannot but violate the hierarchical principle and “dislodge” the intermediate layer of matter by placing it first, as the subject-matter of the fifth hypothesis.90

• 91  In Prm. VI, 1053.19-22 (= 1053.16-19 St.).
• 92  In Prm. VI, 1054.7-10 (= 1054.6-9 St.).
• 93  It is interesting to note that some passages of the Enneads suggest that, contrary to what is gene (...)
• 94  Cf. Beutler and Theiler’s notes on 20, 11ff. and 20, 36 in Harder-Beutler-Theiler 1962, vol. b, p. (...)
• 95  More precisely, nature, which is the lowest part of the world soul, possesses rational principles (...)
• 96  ἀλλ᾽ οὐδὲ ὡς εἶδος ἐν ὕλῃ· ἀχώριστον γὰρ τὸ ἐν ὕλῃ εἶδος, καὶ ἤδη ὕλης οὔσης ὕστερον τὸ εἶδος (Enn(...)
• 97  Cf. also Helmig’s (2006) insightful treatment of enmattered forms in Proclus (and the Platonic tra (...)

52As concerns the enmattered forms, they must be placed last, at the level of the eighth hypothesis, since they cannot interrupt the sequence of material layers, nor, as “positive” entities, correspond to the fifth hypothesis. In this case, however, that there is a violation of the hierarchical principle is not so evident as Proclus believes,91 since the status of enmattered forms is somewhat amphibious: true, as the last manifestation of the intelligible, they are closer to the first principles than matter is, and should therefore precede it; on the other hand, insofar as they only exist in matter, one can understand why Amelius (like Porphyry)92 placed them after the material substrates.93 In this regard, it is interesting to quote a passage from Plotinus’ treatise On Difficulties about the Soul (Enn. IV 3 [27], 20.36-38). Plotinus is here refuting the Peripatetic doctrine that the soul is a form in matter;94 in his view, the soul is, rather, the cause of enmattered forms.95 In this connection, he contrasts the ways in which the soul and enmattered forms are present in bodies: “But it [scil. the soul] is not, either, present as the form in matter: for the form in matter is inseparable, and it comes afterwards (ὕστερον) to the matter which is already there”.96 Unlike the soul, then, the enmattered form is inseparable from matter, that is, it does not exist independently of it; and it comes to a material substrate that is thought of as already existing. Thus, although the enmattered form is a “nobler” entity than matter, it is possible to treat the former as something posterior to the latter, since it comes after it and only subsists in it.97

• 98  The problem can be schematised as follows (I indicate by + the hypotheses having positive deductio (...)

53The series of realities to be matched to hypotheses 5-8 can therefore be schematised as follows (the symbols – and + stand for entities corresponding to negative and positive hypotheses, respectively): – + – +. This sequence is consistent with the quality of the hypotheses: the principle of alternation is thus respected; the price to pay for this is a disruption of the hierarchical order.98

• 99  In Prm. VI, 1054.4-10 (= 1054.4-9 St. = Porphyry, fr. 170.8-13 Smith).
• 100  On the posteriority of enmattered forms to matter, see above, p. 51-52.
• 101  In other words, Porphyry divides hypotheses 4-9 into three independent units, each consisting of t (...)
• 102  In Prm. VI, 1054.38-1055.16 (= 1054.31-1055.14 St. = Iamblichus, in Prm., fr. 2 Dillon).
• 103  See n. .75

56By way of conclusion, I shall briefly show that the results of my analysis can be brought to bear on the subsequent exegesis of the hypotheses. In the context of the “straight-line” interpretative model, Porphyry’s exegesis of hypotheses 4-9 can be construed as a successful attempt to reconcile the hierarchical principle with the principle of alternation: he arranges these hypotheses in pairs, each pair being associated with one given reality, considered under either of two aspects of its ontological structure or from two different points of view (hypotheses 4-5: ordered and unordered body, respectively; 6-7: ordered and unordered matter; 8-9: enmattered forms, considered as actually present in matter and in abstraction from it).99 Within each pair of hypotheses, the entity matched to the negative hypothesis is lower-ranked than that matched to the positive one (for example, unordered body is inferior to ordered body), so that Porphyry’s interpretation is consistent with the alternating qualities of the hypotheses. At the same time, the hierarchical principle is followed, too: for, there is nothing to prevent Porphyry from arranging the three main items (namely body, matter, and enmattered form) according to their rank;100 and, of course, the higher-ranked entity of each pair is inferior to the lower-ranked entity of the superior pair (for instance, ordered matter, qua matter, is inferior to unordered body).101 Also Iamblichus carefully avoids the conflict between the two interpretative principles.102 Moreover, as I mentioned,103 also Porphyry, like Amelius, arguably imported some levels of reality from the Timaeus into the Parmenides exegesis.

• 104  In Prm. VI, 1058.22-28 (= 1058.17-22 St. = Plutarch, fr. 62.1-5 Taormina).
• 105  In Prm. VI, 1055.14-16 (= 1055.12-14 St. = Iamblichus, in Prm., fr. 2.15-16 Dillon).
• 106  Damascius, In Prm. IV, 74.1-12 (= Iamblichus, in Prm., fr. 14 Dillon).
• 107  Cf. Dillon 1973, p. 389, 402.
• 108  Proclus, In Prm. VI, 1053.9-15 (= 1053.8-12 St.).
• 109  In Prm. VI, 1054.18-21; 1055.23-24 (= 1054.15-17 = Porphyry, fr. 170.20-22 Smith; 1055.19-20 St.).

59Finally, as a consequence of the reduced number of entities involved, in these later interpretations the “arche-logical” aspect becomes more manifest. It is difficult to determine whether Amelius and the early Neoplatonists in general took the second part of the Parmenides to embrace the entire scala naturae or to be concerned with principles only. On the one hand, no layer of reality seems to be excluded from their interpretations; moreover, in one passage Proclus apparently suggests that the idea that the Parmenides deals with principles is a comparatively recent acquisition, which Plutarch owes to the philosopher from Rhodes.104 On the other hand, it cannot be excluded that, in the opinion of the early-Neoplatonic commentators, all the entities they introduce are somewhat entitled to the name of principle; indeed, a passage of Damascius testifies that Iamblichus (who introduced the celestial and the sublunary body as the subject-matters of the eighth and ninth hypotheses)105 took pains to argue that the composite (such as the four elements, the celestial spheres, and the heaven) can be a principle:106 this suggests that he did hold the dialogue to be a discourse about principles.107 Also, when Proclus praises Amelius for restricting his interpretation to principles,108 or criticises Porphyry and Iamblichus for introducing non-principles into their interpretations,109 he seems to be implicitly presupposing that they shared his assumption about the arche-logical nature of the dialogue (although the last two commentators were not able to develop their interpretation accordingly). If so, his remark on Plutarch and the Rhodian may not mean that the latter was the first commentator to realise that the Parmenides deals exclusively with principles: rather, he was the first whose interpretation was, in Proclus’ view, truly consistent with this assumption. Even on the latter hypothesis, at any rate, the more rigorous selection of entities made by the later commentators allows them to bring this aspect of the dialogue into sharper relief.