Aristotle's Posterior Analytics is a dense and difficult work. It is also extremely rewarding and philosophically important. Interpreters have toiled in its exegetical and conceptual controversies for centuries. For this reason a volume dedicated to its commentators in Late Antiquity and the Middle Ages is particularly welcome not only to those who work on Aristotle's or the commentators’ texts but also to those who aim to address seriously the philosophical questions raised by the An. post. about (for instance) scientific knowledge, demonstration and definition. There is ample reason, then, to be grateful to the editors of, and the contributors to, this volume for its invaluable addition to the literature on the An. post.
The volume is prefaced with an illuminating introduction in which the editors helpfully outline the conceptual landscape of the commentary tradition on the An. post. not only in Late Antiquity (between the fourth and sixth centuries a.d.) but also in Byzantium, Arabic thought and the Latin Middle Ages (from the eleventh to thirteenth centuries a.d.). The main body of the volume is divided into three parts each of which deals with a central area of the An. post. The first part discusses the commentators’ understanding of how, and in virtue of what cognitive state, Aristotle thinks that the first principles of a science are acquired (An. post. 2.19). The second presents how the commentators grapple with the question of whether, and if so, in what sense, first philosophy (or metaphysics) is a science. The final part is about the commentators’ discussion of Aristotle's account of demonstration, definition and causation, mainly in the second book of the An. post.
In what follows I shall sketch three general areas of criticism which could be levelled against several arguments advanced by the commentators. These criticisms may also apply to the arguments of the ‘meta-commentators’: the modern interpreters of Aristotle and/or the commentators (some of whom are contributors to the volume we are discussing).
The first drawback I see in the commentators’ view of the An. post. is that their readings are often not contextual. They treat Aristotle as a systematic, rigid theorist whose arguments in one treatise are constantly to be understood on the basis of claims made elsewhere in the corpus. Aristotle's arguments, however, should make fairly good sense in their own right. It should not be necessary to invoke material from the rest of his work to understand the arguments in a specific treatise. It is certainly true that questions about the interrelations between different parts of the corpus are pivotal. But they are separate from the initial task of seeking to understand a particular passage.
An example of this non-contextualist approach of the commentators is their view of An. post. 2.19. They typically interpret this celebrated chapter by invoking Aristotle's views of phantasia, nous and thought-abstraction as set out in his De anima (and elsewhere). R. Sorabji notes this tendency and counter-proposes a deflationary reading of the term ‘nous’ as an ‘intellectual spotting’ (pp. 3–4). It would be incorrect, though, to think that an interpreter has either to opt for the heavy-duty reading of the commentators or simply speak of intellectual spotting. Indeed, it would be interesting to ask what this intellectual spotting consists in. Sorabji's reference to Aristotle's account of acumen (agchinoia; An. post. 1.34) could be helpful in this connection. An inquirer is acute in that he or she does not simply spot some important concept or principle but rather grasps such items as causes and principles of certain phenomena. It would be misleading to suppose that Aristotelian nous is a sort of non-discursive intuition or intellectual ‘seeing’, divorced from explanatory and demonstrative concerns. Grasping an item as a cause or a principle entails grasping its position in the wider or even overall explanatory structure of a scientific field.
Second, the criticism levelled against Aristotle by the commentators is sometimes superficial or unfair. For instance, as C. Helmig argues, Proclus criticises Aristotle's account of concept formation in terms of sense perception, memory and experience by insisting that it cannot work unless one resorts to something like Platonic recollection. But Aristotle's argument as set out in An. post. 2.19 can accommodate the Platonist claim that particulars by themselves are not (for instance) perfectly equal or are not identical with Equality. Similarly, it can concede that equality, or being equal, exists. This, however, does not require the existence of a separate Form of Equality which our souls have already grasped in their disembodied phase. Rather, Aristotle can simply hold that there is a common, shared feature, such as being equal, which is a part of the real world but is not para the many particulars. Further, he does not need to agree that in order to grasp this real feature our souls must possess it in advance and eventually recollect it. Rather, his argument is that in this case the relevant pre-existing knowledge is the faculty of sense perception, which together with memory and experience can grasp pre-scientific concepts (what Aristotle calls ennoêmata tês empeirias at Metaphysics A.1, 981a5–7).
My third point may apply more closely to the ‘meta-commentators’ rather than the Ancient commentators. In studying the Aristotelian Organon there is a danger that we may focus exclusively on the ‘formal’ or ‘syllogistic’ aspect of Aristotle's argument without exercising the appropriate metaphysical caution. This strategy can be misleading. Certain parts of the Organon are perhaps metaphysically more innocuous than others. Generally, however, the Organon is not metaphysically neutral. Especially the An. post. (and the Categories) contain metaphysically substantive themes.
An example in which metaphysical sensitivity, as well as consideration of Aristotelian theses canvassed in the metaphysical treatises, would be useful is Philoponus’ (or ‘spurious’-Philoponus’) treatment of An. post. 2.11, discussed in M. Leunissen's excellent study. Leunissen argues convincingly that Philoponus’ construal of the teleological example in An. post. 2.11 is neither the only nor the most compelling reading of the passage. She maintains that the passage involves only one syllogism, the example actually occurring in the text. She points out, however, that this syllogism's middle term does not pick up any final cause. Rather, the final cause is (obliquely) referred to just by the major term. Indeed, she thinks that no Aristotelian demonstrative explanation's middle term refers to a final cause. For such explanations track the causal structure of the real world in terms of material-efficient causation, whereas final causation is not part of the causal fabric of the world (although it is part of our own explanatory practices).
My worry is that Leunissen offers a construal which is artificially limited to the context of the An. post. (and the Prior Analytics). It seems that she understands causation in a narrow, Humean fashion in which only material-efficient causes can discharge the relevant role. Aristotle, however, thinks that final causes are real world causes which make certain kinds of entity what they essentially are. If this is correct, we would expect that at least in some cases the middle term of a demonstrative explanation, which picks up the basic cause of the relevant explanandum, should refer to a final cause: for in such cases the final cause constitutes the basic explanans.
Moreover, in his Metaphysics Z.17 Aristotle understands substance as a principle and a cause. He goes on to specify that this is the essence of a thing, and identifies it with some or other amongst the four causes. It is clear that in some cases he takes the essence to be identical with the final cause. At the same time, though, he applies the causal-explanatory model of the An. post. to substance-kinds and argues that the middle term of the relevant demonstrative explanations picks up the essence. His overall argument, then, entails that the middle term can, in certain cases, pick up a final cause as in those cases the middle term would pick up an essence which is identical with the final cause (Metaphysics Z.17, 1041a27–1041b9). Leunissen's argument would benefit from taking into consideration these metaphysical theses which are not congenial to the idea of removing telê from the causal centre stage of a demonstrative middle term.
These, at any rate, are minor complaints. This volume is an impressive achievement. It will be indispensable not only for those who work in the commentary tradition but for any serious student of the exegetical or the philosophical aspects of the An. post.
