In Parts of Animals I 1 (and elsewhere) Aristotle makes it clear that his goal in the study of nature is a

Comments on Mariska Leunissen s Aristotle s Syllogistic Model of Knowledge and the Biological Sciences: Demonstrating Natural Processes Allan Gotthelf Introduction In Parts of Animals I 1 (and elsewhere) Aristotle makes it clear that his goal in the study of nature is a body of demonstrative science, an episteme. 1 It has been a longstanding question to what extent either the PA I 1 account, or the actual explanatory practice in the biological treatises, accords with the general theory of demonstrative science that Aristotle offers in the Posterior Analytics. 2 In my two papers on this topic, First Principles in Aristotle s Parts of Animals and The Elephant s Nose: Further Reflections on the Axiomatic Structure of Biological Explanation in Aristotle, I have argued that the accord is much greater than had been recently thought. 3 My focus in those papers, though, was largely on the structure in and across the actual explanations in Parts of Animals. I said little in either paper about the precise content of the Posterior Analytics theory, approaching as I did the problem of the accord between it and the biology from the side of the latter, so to speak, rather than from the side of the theory of demonstration. And I said little to make the case that the Posterior Analytics theory itself was explicitly meant to subsume explanations in natural philosophy, or how it could do that, given the theory s evident focus on timeless matters of fact. Mariska Leunissen, in the paper before us, approaches the matter head on, and to my mind makes a significant advance. I applaud its spirit and many of its conclusions, and delight in the way Leunissen 1 Cf. Gotthelf 1987, 170-71. 2 See, e.g., notes 5 and 6 in Leunissen s paper above. 3 Gotthelf 1987 and Gotthelf 1997. Both papers are being reprinted in a volume of my collected Aristotle papers, to be published by Oxford University Press in its Oxford Aristotle Studies series. 1

takes an obscure chapter of APo. II, illuminates its structure and dark corners, shows its value for understanding how Aristotle may have come to apply the basic Posterior Analytics model of explanation to generated natural things, and then uses that understanding to help illuminate certain key aspects of Aristotle s explanatory methodology in the Generation of Animals. I will raise some questions about particular claims in the paper, but they are meant entirely in a constructive spirit. Leunissen states her project as follows: In this paper I have sought to show how Aristotle in both his theory and practice of the biological sciences is influenced by the two models of demonstration of processes outlined in APo. II 12. (above, 00) The grounds on which claims that a philosopher s work on topic A influenced his work on topic B (or another philosopher s work on topic A) often in fact only establish parallels between those bodies of work or their conclusions. In our case Leunissen has certainly established parallels; but I m inclined to think that she may also have picked up on a real progression in Aristotle s thought, from APo. II 11 to II 12 to PA I 1 and the practice in PA II-IV and GA. I m going to take issue with her reading of the relevant part of APo. II.11, but the issues here are very complicated, and even should I be right in my criticism there, the rest of what Leunissen says still stands, and that rest is the heart of her thesis in this important paper. I have two other questions I d like briefly to raise, centered in the biological texts she discusses; one is about the PA I model (and the PA II-IV practice) and the other about Leunissen s really insightful account of a central feature of A. s methodology in GA (especially in his account of embryogenesis in II 6). Both questions also are meant to underscore and clarify theses of Leunissen s that I think are essentially or largely right. Leunissen s argument in this paper is rich and complex. I have found it useful in reflecting both on the full argument and on its parts to have before me an outline of the paper, and will provide that in my first section here for others who may find it useful. The outline is based in part on an advance copy Leunissen kindly provided me of her PowerPoint slides, and in part on my own analysis of the 2

paper. I ll then go on to organize my comments around that outline, moving for the most part step by step through it. Overview of Leunissen s paper I present my outline as Figure 1. Outline of Mariska Leunissen Aristotle s Syllogistic Model of Knowledge and the Biological Sciences: Demonstrating Processes I. Introduction: the APo. model of demonstration and natural processes Models of demonstration in Aristotle 1) Demonstration of being (in APo.) 2) Demonstration of coming to be (in PA I) 3) Cornucopia of patterns of explanation in biology II Demonstrating processes: some first sketches in APo. Book I Geometrical model strongest, but: Examples of processes: APo. I.24, I.29, I.31 Methodological adjustments for processes: I.4, I.8 Demonstration of processes that occur for the most part (I.30) Book II Syllogistic framework incorporating the four causes ( II 11) Two examples involving change: Figure 1 3

Efficient explanation: of war by being first to attack separation in time Teleological explanation: of walking after dinner by health makes use of efficient explanation: walking after dinner (= simul. making food not floating) having food not floating (= suml. def.) health (Demonstrations here involve geneseis: timeline is important II 12 shows why.) II. Three models of demonstration of processes in APo. II 12 Formal structure: demonstrations of coming to be = demonstrations of being with tensed terms Three models of temporal relations between cause (BaC) and effect (A s coming to hold of C): Model 1: cause and effect occur simultaneously Model 2: cause and effect are non-simultaneous, but of same type Model 3: cause and effect are non-simultaneous, but follow each other (eternal/cyclical) [put aside] Model 1: cause and effect occur simultaneously Coming to be of A of C = simul. Coming to be of B of C Tenses of middle and major correspond (examples: [a] in II 12; [b] in GA V): [a] APo. II 12: coming to be of [being] eclipse[d] of moon (or: of earth-in-middle to moon coming to be ice of water/coming to be solid [or: of solidity] of water [b] GA V: coming to be of pathema of part or of change of material constitution of part. This is not the same structure as in the teleological explanations introduced in PA I.1, where there is an earlier [efficient] cause (i.e., materially necessary antecedents) and a later effect = final cause (i.e., a part s being a certain way). Model 2: cause and effect are non-simultaneous but are of the same type (homogonos) Data: in continuous linear sequence we can explain later by earlier via: [i] one-directional inference from chronologically posterior to prior, where [ii] middle and extremes are homogonos [= in same tense], and [iii] premises are immediate Figure 1 (cont d) 4

Three hypotheses of this paper: (a) Aristotle has in APo a modal notion of conditional necessity (b) This model provides the conceptual foundations for the model of demonstration in the natural sciences that Aristotle develops in PA I the signal feature of which is the teleological interpretation of the consecutive causal chain. (000) (c) And it explains Aristotle s methodological preoccupation with determining what process or development comes to be before what in natural causal sequences... a method that surfaces in a couple of passages in PA I-II, but which is particularly prominent in GA II. (000) Elaboration of the three hypotheses (a) Distinguishing causal use of necessity from solely modal use (based on Kupreeva 2010) The former involves a type of causality operative in nature: it refers to the necessity of materials acting according to their own nature dependent on or independent of some preexisting internal potential for form that needs to be realized (i.e., conditional vs. material necessity) (000) (This use is not in focus in II 12; only the modal use is.) The modal use focuses on the question of what kind of causal inferences we are allowed to draw in cases where A is the cause of B, but where B does not occur at the same time as A but instead occurs later. (000) This question is independent of the question of what sort of causality is involved. The discussion of the example of the house and its foundations in II 12, for instance, is prior to, or abstracted from, concern with the teleological character of the cause that appears in, e.g., Physics II.9 and PA passim. (b) Seeing PA I s explanatory model as growing out of this model. PA I claims that in natural case there is (as in our Model 2): (i) one-directional inference from posterior to prior and (ii) a (Model 2-like) conditional necessity (with causes producing effects for the most part ). Figure 1 (cont d) 5

[W]hat s new [is the] teleological interpretation [of the] causal chain: Inference from posterior to prior becomes inference from end and final cause to conditionally necessary antecedents. (p. 000) In GA order of generation is an ordering principle in discussion of animal reproduction throughout, but especially in discussion of embryogenesis in GA II: what (organ) is source of the other parts, and does it generate them simultaneously (above, Model 1) or in succession (above, Model 2)? Note the complex account in II 6 (742a16-25) of the correlation between causal roles and temporal sequence. IV. Conclusion In this paper I have sought to show how Aristotle in both his theory and practice of the biological sciences is influenced by the two models of demonstrations of processes outlined in APo. II 12. (000) Figure 1 (concluded) Posterior Analytics II 11 on the teleological syllogism As the Outline indicates, Leunissen introduces APo. II 11 as providing evidence, along with other passages in the Posterior Analytics, that Aristotle includes within the scope of that treatise s theory of demonstration explanations of processes and in general change explanations in which temporal relations are central. The II 11 evidence is unequivocal (as is the evidence from book I). Two of Aristotle s three examples of how to syllogize the different sorts of causal explanations are explanations of temporal 6

events or processes: the Athenians coming to be involved in a war and (in certain cases) the taking of a walk after dinner. 4 Leunissen is convincing that these examples in II 11, and especially the teleological one, invited for Aristotle questions about the character of demonstrations of tensed predications, which are precisely the questions to which II 12 is addressed. I find her account of the peculiar way Aristotle here in II syllogizes the teleological case, and thus her full account of the place of II 11 in the sequence from it through II 12 to the biology, less convincing, however. Understanding the teleological example here has proven notoriously difficult; both Ross and Barnes in effect throw up their hands. The problem of course is that the teleological cause (here, health) of the process being explained (here, someone s walking after dinner) is not the middle term of the syllogistic analysis of that explanation which Aristotle appears to offer us. Analyzing the syllogism Aristotle sketches out, Leunissen concludes that the teleological syllogism in the example is (in effect) this: health holds of walking after dinner because being healthy holds (by partial definition, in that context) of the food not floating, and the food s not floating holds of walking after dinner (because walking after dinner leads to the food s not floating). 5 4 Leunissen rightly remarks that [i]t is true that these latter two examples hardly look like scientific demonstrations at all, but presumably Aristotle is less concerned with the details of his examples than with the larger patterns that they are meant to illustrate in for his contemporary audience easy and intuitive ways. (000) See, after all, the opening of the chapter. 5 Leunissen writes in her note 14: Explanandum: Why does he walk? [Why C?] For the sake of health. Why does health hold of walking? [Why A of C?] A = being healthy; B = food not floating (aition = material cause); C = walking after dinner; AaC because of B: being healthy holds of walking after dinner because walking makes the food not floating, and having the food not floating is what being healthy is in this context. 7

This is certainly a natural way of reading the example Aristotle offers; he even says that the cause is not here the middle term. 6 Leunissen refers us for support of this reading to a 2007 paper of hers in which she has given a much fuller account of her way of reading the II 11 examples, and this one in particular. 7 As she puts it in the earlier paper: Demonstration shows how the events to be explained bring about the final cause. 8 As support for this reading of the teleological syllogism, Leunissen attempted to show in that paper that the explanatory practice in PA II-IV too shows how the events to be explained bring about the final cause. I do not think this is right. I am inclined to agree with Leunissen that the explanatory practice in the biology may shed light on the theory embedded in the APo. II 11 teleological example (though we will need to be sure that we aren t finding in APo. II 11 any feature of the later use of teleological explanation that the main developmental thesis of the present paper says should not be present yet in II 11). There isn t time for us to look carefully at the biological passages Leunissen discusses in the 2007 paper, in her attempt to show that there too demonstration [i.e., explanation] shows how the events to be explained bring about the final cause. But from that very formulation I think we can see where that reading goes wrong. For, while it s true that Aristotle often shows in PA how, in the context of the operation of a formal nature (a potential for form ), the materials that are present make 6 APo. II 11 XX. Our question will be whether the syllogism of which this is true is the syllogism that exhibits the teleological cause as the cause. 7 M.E.M.P.J. Leunissen. 2007. The Structure of Teleological Explanations in Aristotle: Theory and Practice, Oxford Studies in Ancient Philosophy 33: 145-78. This paper, she says in the present paper, provides a possible interpretation of the details of Aristotle s account of the syllogistic structure of different types of demonstration in this chapter. (000 and n. 11) 8 Leunissen 2007, 000. 8

possible and/or necessitate (ceteris paribus) the formation of the part whose presence is being explained, this is only part of the teleological story, and indeed not its central part. 9 The central part is structurally fairly simple. Aristotle seeks to explain why all those animals which possess a certain part possess that part. The explanation makes reference to a functional need of the part, if the animal is to live or to live well. 10 Discovering the part s function may or may not be easy, but that makes no difference to the structure of the explanation, once the function is discovered. Such a demonstration doesn t show how the events to be explained bring about the final cause, it simply shows why those parts, and any events that lead to their formation, must be present. Whatever precisely the role in the teleological story of the materially necessary processes leading (in the context of a potential for form) to it, these processes are not needed to explain the presence of the final cause. What is explained by that part of the account appears to be the connection in the major premise the claim that a certain (version of a) part is needed, given the context of the available materials, if a certain function is to be performed in the life of the animal. Interestingly, it has at least sometimes been claimed that what Aristotle is giving us in the APo. II 11 example is precisely the syllogistic demonstration of the major premise of the teleological syllogism and not the teleological syllogism itself. 11 For, one might think that the teleological syllogism proper has a very simple structure. For, consider the form of the explanation indicated in Physics II 3 when the final cause (the cause hos telos) is introduced: In yet another way we call <something> a cause as <being> the end; this is that for the sake of which e.g., of walking, health. For, why does one walk? We say, In order to be healthy, and speaking so we think we have given the cause. 9 See, e.g., J. G. Lennox, Material and Formal Natures in Aristotle s De Partibus Animalium, in Lennox 2001a. 10 See Gotthelf 1997 for details. 11 E.g., ps.-philoponus, in APo. II, ad loc. 9

What is to be explained is why a certain subject S (however understood, precisely) walks (on a specific occasion). Identifying that the walker has health as his end (and understands that walking contributes to health) provides the heart of the explanation. As one might put it: Walking (in circumstance C) holds of those who have (the maintenance of) health as an end. Having health as an end holds of S (in circumstance C). So walking (in circumstance C) holds of S (in circumstance C). In the APo. II 11 case the circumstance is just having had dinner. Now, I don t have a view as to why Aristotle might have provided in II 11 the outlines of a scientific account of why it would be right for those who have health as an end to walk after dinner, something which the walker might not have an inkling of. I can only imagine that Aristotle had mind the need for this example to reveal the structure of teleological explanation in natural science and not just in practical activity, and that in the scientific case, walking in fact leads to (the maintenance of health) if certain outcomes are materially necessitated. The progression from APo. II 11 to II 12 Interestingly, If Aristotle thought, while writing II 11, that teleological explanations typically (or often) require a coordinate appeal to material-efficient causes to explain one of their premises, then the brief account at the end of the chapter of how these two sorts of causes can be jointly present in one case might be more closely related to the teleological example than many have thought. And, in that case, the movement Leunissen has described from questions regarding the role of temporality in demonstration to the answers to these questions in II 12 would be even more immediate. But however the discussions of teleological explanation in II 11 are connected, it remains the case that II 12 proceeds directly upon an account in II 11 of how the four types of causal account are to be syllogized. And it does so in order to address cases where, as Leunissen well puts it, the explananda and their 10

causes are tensed. First Aristotle discusses the causal relations involved and then he discusses how to syllogize them. He doesn t himself say why he turns to these cases, but the most natural explanation is surely Leunissen s, that he s focused on causal relations in which the temporal dimension is central. If that s right, and if it s right that conditional necessity is not present in II.11, then indeed Aristotle is here introducing conditional necessity into the Posterior Analytics theory as a tool to understand these causal relations. From APo. II 12 to PA and GA Leunissen s comparison of the PA I 1 model of the explanation of features of generated natural substances with the second model she finds in APo. II 12 is nicely done and seems almost exactly right. Does the former model grow out of the latter one and, in general do we see an actual progression in Aristotle s thought in the sequence from II 11 to II 12 to PA I.1? If progression refers to a logical progression, the answer strikes me as clearly affirmative; that is a main accomplishment of Leunissen s paper. I don t see that we can, on the basis of the information we have, establish a chronological progression from APo II. 12 to PA I.1; I don t think we yet have unequivocal evidence that PA I.1 solves a pre-existent problem in the systematization of natural science by drawing on the content of APo. II 12. But Leunissen s discussion of the key PA I.1 text does, to my mind, make that plausible. That being so, one gets the whole progression, if one can show that on the correct reading of the teleological example in APo. II 11 which, I ve tried to show, we do not yet have the chapter does not make use of conditional necessity. But whether it does or not, Leunissen has certainly succeeded in showing, perhaps more clearly than anyone before her, that the Posterior Analytics not only recognizes that its theory of demonstrations can be applied to the processes and change characteristic of perishable natural substances, but is also concerned in places with showing how it can be so applied. 11

Leunissen s conclusion, as presented in the Outline above (and in her own text, 000) goes further yet: In this paper I have sought to show how Aristotle in both his theory and practice of the biological sciences is influenced by the two models of demonstrations of processes outlined in APo. II 12. I will conclude, then I conclude, therefore, with two brief comments on her discussion of the biology. - The target of the PA I model. I quote PA I 1, 640a33-b4: Hence, we should if possible say that because this is what it is to be a man, therefore he has these things; for he cannot be without these parts. Failing that, we should get as near as possible to it: we should say altogether that it cannot be otherwise, or that it is at least good thus. And these follow. And because he is such a thing, his coming-to-be necessarily happens so and is such. And that is why this part comes to be first, and then this. And this is the way we should speak of everything that is composed naturally. (tr. Balme 1972; Lennox 2001a is similar) The PA I model is not just about processes, or comings-to-be, and not primarily about these. Leunissen s analysis of the structure of a teleological explanation, discussed above, might allow one to read it that way, but I have raised strong doubts about that analysis. The model here is a model for the explanation of the presence in an organism of the parts it has, in the variation it has, and that s of course the aim of PA II-IV. The explanation of the character and, specifically, the temporal order of the stages in the coming-to-be of that organism come subsequent to that. 12 Nonetheless, the 12 I append here a note (which Leunissen cites in her note 38 above) to my discussion of this matter in my Postscript 1986 to Gotthelf 1976/77: This is not to say the PA is unconcerned with generation. PA is concerned with why animals have the parts they do, and animals have the parts they do partly because some such parts are needed (or are best) for the functioning of the whole and partly because the available material permits those parts: têi men heneka tou... têi de ex anagkês. PA will thus have things to say about the necessary natures of the available materials, and the constraints these natures 12

conditional necessity highlighted at 640a1ff is certainly at work in 640a33ff, and that is all Leunissen really needs. I would like, then to see the PA I 1 s model of the teleological explanation of the presence and character of animal parts incorporated into Leunissen s account of the progression from APo. to PA more fully. But from the standpoint of the main claims of her paper, this is just a detail. - The importance of chronological order in the account of generation in GA. Leunissen s description of the place in GA of attention to the chronological order of generation (00-00) is virtually flawless, I think. The only observation I would make is this: Someone might think that the mere fact that GA is predominantly after the efficient causation of generation 13 is sufficient to explain its focus on chronological order. But that s surely not the whole story, and Leunissen s focus on the overall pattern of explanation including the idea that the end of a generation causes (or, if you like, explains) the stages that lead up to it as well as her focus on Aristotle s explicit concern in II 6 to put the place on what parts could be formed to fulfill a needed function, but it will not focus on the generation process as such, and its timing and stages, and how these are necessitated by the nature of the result to be produced: those are the focus of GA. And GA s initial focus is not so muich the final cause of generation but its efficient cause: what antecedent movements (as it were) produce offspring one in form with their parents and how they do it. (Cf. GA I.1 715a11-18, II.1 733b23-32, etc.) since the movements Aristotle identifies are those borne by the heat transmited by the father, and involve, in addition to low-level material dunameis, a dunamis for form, teleological explanation will be inevitable, the end being what the dunamis that is the efficient cause is for. And teleological explanation does appear in GA, both for generation as such and its general features (II.1, 731b18-732a11, 6 742a16-b17, 743a36-b3) and for particular formation, timings, and sequences (743b3-18, 744a35-b9, etc.). but it is introduced selectively and is not the primary focus of the work. Still, insofar as they are concerned with for-the-sake-or relationships, PA is primarily concerned with being for the sake of, GA with coming to be for the sake of. In her 2007 paper, Leunissen suggests that explanations of the presence of parts called for by the first part of 640a33ff are explanations via the formal cause, and that the teleological part of the overall explanation starts only after that. I do not see the basis for that claim. 13 Cf., e.g., GA I 1, 715a11-18 and II 1, 733b23-32. 13

beginning, middle, and end in proper causal relation to each other brings that out. See especially her illuminating discussion on p. 000.) I am thus happy to endorse her statement on that page that the analysis of demonstrations of same type processes in APo. II 12 drives at least the mode of inference and the concern for tracking order in generation. This, then, is an exciting paper, and I thank Mariska Leunissen for it. Department of History and Philosophy of Science University of Pittsburgh 1017 Cathedral of Learning Pittsburgh, PA 15260 gotthelf@pitt.edu Bibliography Balme, D. M. 1972. Aristotle s De Partibus Animalium I and De Generatione Animalium I (with passages from II.1-3). Oxford: Oxford University Press. (Rev. edn. with a report on recent work and additional bibliography by Allan Gotthelf, 1992.) Barnes, Jonathan. 1993. Aristotle: Posterior Analytics, 2nd edn. (Oxford: Clarendon Press). Detel, W. 1997. Why all Animals have a Stomach. Demonstration and Axiomatization in Aristotle s Parts of Animals. In W. Kullmann & S. Föllinger, eds., Aristotelische Biologie: Intentionen, Methoden, Ergebnisse, 63-84 (Stuttgart: Franz Steiner Verlag). 14

Gotthelf, Allan. 1976-77. Aristotle s Conception of Final Causality, Review of Metaphysics 30: 226-54. Reprinted with additional notes and a postscript in A. Gotthelf & J.G. Lennox, eds., Philosophical Issues in Aristotle s Biology, 204-42 (New York: Cambridge University Press). Gotthelf, Allan. 1987. First Principles in Aristotle s Parts of Animals. In A. Gotthelf & J.G. eds., Philosophical Issues in Aristotle s Biology, 167-98 (New York: Cambridge University Press). Gotthelf, Allan. 1997. The Elephant s Nose: Further Reflections on the Axiomatic Structure of Biological Explanations in Aristotle. In W. Kullmann & S. Föllinger, eds., Aristotelische Biologie: Intentionen, Methoden, Ergebnisse, 85-95 (Stuttgart: Franz Steiner Verlag). Kupreeva, Inna. forthcoming. Causation and Conditional Necessity in Posterior Analytics II 12. In F.A.J. de Haas & M.E.M.P.J. Leunissen, eds., Interpreting Aristotle s Posterior Analytics in Late Antiquity and the Byzantine Period, (Leiden: Brill). Lennox, James G. 2001a. Aristotle s Philosophy of Biology, Studies in the Origins of Life Science (Cambridge: Cambridge University Press). Lennox, James G. 2001b. Aristotle: On the Parts of Animals I-IV. Translation with notes (Oxford: Oxford University Press). Leunissen, M.E.M.P.J. 2007. The Structure of Teleological Explanations in Aristotle: Theory and Practice, Oxford Studies in Ancient Philosophy 33: 145-78. Leunissen, Mariska & Allan Gotthelf. In preparation. What s Teleology got to do with it? A Reinterpretation of Aristotle s Generation of Animals V. Lloyd, G.E.R. 1996. The Theories and Practices of Demonstration. In G.E.R. Lloyd, Aristotelian Explorations, 7-37 (Cambridge: Cambridge University Press). 15