The Lagoon

by Armand Marie Leroi

  XXXIII

  Forms. Plato outlines his Intelligible Living Creature and its relationship to subordinate Forms at Tim 30C–31A; CORNFORD (1997) pp. 39–42. A. critiques Platonic Forms at Metaph I, 9. A. speaks of many eidē of birds and fishes at HA 486b224. THOMPSON (1910) n. 490b16 was one of the first scholars to point out that A. uses eidos in several different ways, not all of which are consistent with translating it as ‘species’. BALME (1962a) and PELLEGRIN (1986) later developed this line of interpretation further, wielding it against the idea that A. is engaged in a taxonomic project. A. uses the term atomon eidos at PA 643a13, Metaph 1034a5, DA 415b6 and HA 486a16. Even here there is a debate about whether atomon eidos refers to individuals or species or both – A. is by no means clear. Some, e.g. BALME (1987d), HENRY (2006a), HENRY (2006b), have argued that he means individuals, but I find GELBER (2010)’s argument that he usually does mean species, i.e. that two individuals can share the same indivisible form, convincing. This interpretation has consequences for reading A.’s theory of inheritance, for I am thereby compelled to invoke an additional, sub-specific level of heritable variation, which I call ‘informal’ variation; see Chs LXX and LXXIII. A. explains what forms are by means of the carpenter analogy at PA 641a6 and the syllable theory of form at Metaph VII, 17. DELBRüCK (1971) argued for the interpretation of form as information and many have followed him, e.g. FURTH (1988) pp. 11–120, KULLMANN (1998) p. 294 and HENRY (2006a), HENRY (2006b); but see DEPEW (2008) for a different view.

  XXXIV

  The four ways of explaining. A. frequently states his four basic causal explanations, e.g. GA 715a4, from which the quote comes, and Phys II, 3 and PA 642a2. PECK (1943) pp. xxxviii–xliv, LEUNISSEN (2010a) and LEUNISSEN (2010b) give general discussions of A.’s system of causal explanation. LEAR (1988) pp. 29–31 explains how A.’s ‘causes’ differ from Hume’s. The influence of A.’s division of causal explanations on the history of biology is one of the great themes of RUSSELL (1916), a classic. HUXLEY (1942), MAYR (1961) and TINBERGEN (1963) give the different kinds of causal explanations in modern biology; Mayr cites A. explicitly, Tinbergen does not; see also DEWSBURY (1999). The main difference between their list of causal explanations and A.’s is that A.’s does not have an evolutionary dimension where theirs do. Among those who have seen A. as a mere synthesizer or a species of Platonist are POPPER (1945/1962) vol. 2, ch. 11 and SEDLEY (2007) pp. 167–204 – but the tradition is an ancient one and underlay the entire Neoplatonist project.

  XXXV

  The Bird Hall & how to carve up nature. For a history of the Natural History Museum and its exhibits see STEARN (1981). LENNOX (2001b) ch. 2 discusses A.’s options in arranging his data in much the same spirit as I do here.

  XXXVI

  A. as a natural historian. Theodore of Gaza’s preface to his edition of A.’s zoology, GAZA (1476), is given by PERFETTI (2000) p. 16, who also discusses Pliny’s influence on him. BEULLENS and GOTTHELF (2007) discuss the dating and structure of Theodore’s HA. Pliny on the elephant is from Plin VIII, 1, 13, 32 [trans. RACKHAM et al. (1938–62)]. The view that A. was not doing natural history is a commonplace among A. scholars. FRENCH (1994) disagreed, but his view of A. as a natural historian was inconsistent by his own criteria.

  XXXVII

  A. as a taxonomist. CUVIER (1841)’s encomium to A.’s skills as a taxonomist is quoted by PELLEGRIN (1986) p. 11. Modern Greek fish names are given in KOUTSOGIANNOPOULOS (2010) – essential for those interested in Greek fishes, but so far available only in Greek. A. speaks of the several breeds of dogs at HA 574a16, the hippos crab at HA 525b7 and the kyanos at HA 617a23. He writes a great deal about cephalopods, notably their anatomy in HA IV, 1 (see this text Ch. XXIII) but also in many other passages besides. On the paper nautilus see HA 622b8; cf. HA 525a19 and Ch. XXVII. The mystery cephalopod ‘that lives in its shell like a snail’ is described – barely – at HA 525a26; see SCHARFENBERG (2001). A. talks about the other kinds of crabs at HA 525b6; cf. PA 683b26. DIAMOND (1966) describes the ability of New Guinea highlanders to discriminate bird species; ATRAN (1993) discusses folk taxonomies in general; he also has a useful chapter on A.’s systematics.

  The greatest kinds. A. gives major statements on greatest kinds: blooded animal, HA II; bloodless animals, HA 490b7, HA 523a31 and other passages in HA IV. He says that names such as ornithes and ikthyes are vernacular at PA 644b5; cf. PA 643b9 Some of his new technical names (e.g. malakostraka) are actually ‘name-like expressions’, i.e. shorthand descriptions that substitute for a noun: cf. APo 93b29–32; PECK (1965) pp. lxvii, 31 and LENNOX (2001a) p. 155. It’s likely that the use of such names began at the Academy; Speusippus, it seems, used malakostraka and was interested in definition; see WILSON (1997). A.’s hierarchy of genē is very incomplete, but he does not allow any given subordinate kind to exist in more than one position in the hierarchy; cf. Top IV, 2. Many genē are, at best, classified into the enhaima or anhaima, e.g. humans come under no genos other than the enhaima, HA 490b18. The following passages support A.’s commitment to hierarchical classification: on blooded animals, HA 505b26; bloodless ones, HA 523a31, HA 523b1; soft-shells, PA 683b26, cf. HA 490b7. He says that in a classification each animal should appear only once at PA 642b30 and PA 643a8. Borges (1942) ‘El idioma analítico de John Wilkins’ [The analytical language of John Wilkins] in BORGES (2000) p. 231, tells of the (apocryphal) Chinese encyclopaedia. A.’s orthogonal classification of polities is given at Pol III, 7. This classification is the result of the method given in Pol IV, 3. Indeed, there A. explicitly compares classifications of states to classifications of animals and recommends that we take all the varieties of organs – of states or animals – and array them orthogonally: ‘There must be as many forms of government as there are arranging the offices.’ But this is exactly what he does not do when classifying animals, for that procedure would necessarily lead to empty classes. For example, suppose you were to classify animals by two kinds of features: oral organs (teeth v. beaks) and dermal organs (hair v. feathers). An orthogonal classification would yield four classes of animals: (i) toothed-hairy, (ii) beaked-hairy, (iii) toothed-feathered, (iv) beaked-feathered. Of these, (i) is the mammals, (iv) is the birds, while (ii) and (iii) do not exist, allowing that the ‘duck-billed’ platypus does not, in fact, have a beak. This shows that orthogonal classifications are inefficient for they do not reflect the real covariance structure among biological entities. I think A. originally considered them, recognized their absurdity and abandoned them, perhaps when he started to do biology. In fact he doesn’t even follow his recommendation for an orthogonal classification in Pol since later he subdivides the genē given at Pol III, 7 further so that his full classification of political constitutions is also, in fact, nested.

  The meanings of genos. Metaph V, 28; see PELLEGRIN (1986) ch. 2.

  The method of division. Plato defines kings at States 257–68e. He says, States 266D, that it is amusing that the king has been ‘running a race with the man who is of all men best trained for living an easy life’. In his divisional scheme the sister-group of ‘herders of featherless bipeds’ – kings – are ‘herders of feathered bipeds’ – goose-herds, evidently an undemanding job. In Metaph VII, 12 and APo II, 5, 13, 14 A. follows P.’s method of division, but introduces some technical modifications; his critique becomes more extensive in PA I, 2–3; see BALME (1987b), LENNOX (2001a) pp. 152–472. Some scholars claim that the objective of A.’s division was definition rather than classification, but in PA I, 4 it’s clear that he’s interested in identifying kinds and works with them; see LENNOX (2001a) pp. 167–9 and n. this text Ch. XLI. P. says that ‘we shouldn’t cut across the joints . . .’ at Phaedrus 265E.

  XXXVIII

  Taxonomic methodology. For a list of diaphorai, or differentiating characters, see HA I, 1. On ‘the more and the less’ as distinguishing regular kinds within greatest kinds of birds see PA 692b3, cf. HA II, 12–13, and, in general, HA 486b13, HA 497b4 and PA 644a13; LENNOX (2001b) ch. 7. A. giv
es his animal geometry at IA 4, PA 665a10 and HA 494a20. Man stands alone: HA 490b18, HA 505b31. A. discusses cuttlefish and gastropod geometry at HA 523b22, PA IV, 9, IA 706a34; see also this text Chs XCI and XCVII; for his plant geometry see IA 706b5, LBV 467b2, Phys 199a26 and PA 686b35. A.’s theory of analogues is given at HA 486b18, HA 497b11 and PA 644a22. Some see A.’s usage as being close to OWEN (1843)’s definition of analogy as ‘a part or organ in one animal which has the same function as a part or organ in another animal’. LENNOX (2001a) p. 168 points out correctly that A. often doesn’t make the functional similarity explicit, though he certainly does sometimes (e.g. hearts and heart-analogues). On analogues see LLOYD (1996) ch. 7, LENNOX (2001b) ch. 7 and PELLEGRIN (1986) pp. 88–94 who claims that analogon does not serve a classificatory function, but I find his arguments unconvincing. For the cephalopod ‘brain’ see PA 652b24, HA 494b28 and HA 524b4; LENNOX (2001a) pp. 209–10. RUSSELL (1916) p. 7 and BALME and GOTTHELF (1992) p. 120 agree that A. has an implicit concept of homology; however, it should be noted that parts that are the same ‘without qualification’ has multiple meanings, LENNOX (2001b) ch. 7. A. compares the skeletons of snakes and snake and egg-laying tetrapod at HA 516b20 and PA 655a20, says that snakes are like footless lizards at HA 508a8 and PA 676a25 and speaks of seals at HA 498a32, PA 657a22 and PA 697b5.

  XXXIX

  Polythetic classification. A. considers how to divide up some land animals at HA 490b19. This passage occurs in the middle of a discussion on the greatest kinds so seems to be aimed at delineating them. He says that snakes are a genos at HA 490b23 and HA 505b5. He points to the need to consider many features simultaneously when dividing at PA 643b9. For polythetic classification and its history see BECKNER (1959), and MAYR (1982) pp. 194–5. MAYR (1982) p. 192, LENNOX (2001a) pp. 165–6, 343 and LENNOX (2001b) ch. 7 agree that A. used polythetic classification. On the ostrich see Ch. XXI. A. discusses three kinds of apes at HA 502a34 and PA 689b31 (a possible fourth is mentioned elsewhere). The pithekos is a dualizer not because of convergent evolution but because it falls between two divergent kinds (tetrapods and humans). This is a consequence of A.’s refusal to place humans where they should go, among the zōotoka tetrapoda; he has torn a natural kind apart and done so for no reason other than a belief in the specialness of humans (see Ch. XCVII). On dualizers see LLOYD (1983) ch. I, 4 and LLOYD (1996) ch. 3.

  XL

  Dolphins. Herod I, 24 tells the tale of Arion. For accounts of the dolphin in antiquity, and riders in particular, see THOMPSON (1947) pp. 54–5. A. speaks of paedophilic dolphins at HA 631a8 and the features of cetaceans in general at JSVM 476b12, HA 589a33, PA 655a15, PA 669a8 and PA 697a15. Pliny talks nonsense about dolphins at Plin IX, 7–10.

  XLI

  What is the project of HA? MEYER (1855), BALME (1987b) (a revised version of a 1961 paper) and PELLEGRIN (1986) successively attacked the view that A. constructs a classification or even wants to, and this has become, to varying degrees, dogma. But A. clearly does construct a classification, and uses it, even if it is very incomplete and not his primary objective; see LLOYD (1991) ch. 1, LENNOX (2001a) p. 169 and GOTTHELF (2012) ch. 12. A. says why classifications are useful at PA 644a34. The order in which he will arrange his information in HA is given at HA 487a10. My synopsis of HA is based on the order of the books given by BALME (1991) rather than D’Arcy Thompson and earlier authors who used the order imposed by Theodore of Gaza; see Balme’s Introduction for a discussion of authenticity, plan and order of the component books of HA. Balme also makes the case that, among the zoological works, HA was not written first. Indeed, it’s likely that they were all updated and more or less integrated with each other over A.’s lifetime, and perhaps afterwards by his successors, so that it is now very difficult to discern the order of composition. A describes the ruminant stomach at HA 507a32. The view that HA provides the basic material for a demonstrative science is a commonplace among scholars of A.’s zoology.

  XLII

  The need for demonstration. A. alludes to the purpose of HA as material for demonstration at HA 491a12, cf. PA 639a13, PA 640a1 and GA 742b24; see LEUNISSEN (2010a) ch. 3.1.

  XLIII

  Demonstration & the syllogistic. The following works, arranged in order of increasing difficulty, discuss A.’s logic and theory of demonstration: BARNES (1996) chs 7–8, ACKRILL (1981) chs 6–7, ROSS (1995) ch. II, ANAGNOSTOPOULOS (2009c), BYRNE (1997) and BARNES (1993), the last of which really requires a grasp of formal logic. What it takes to have scientific knowledge of something: APo 71b9. The conditions of demonstration are specified as follows: the premises must be true and immediate, APo 71b9; must concern universals, APo 71a8, APo 73b25, cf. DA 417b21, Metaph 1036a2, Metaph 1039a24, Metaph 1086b32; we must have better knowledge of them than the conclusions, APo 72a25. For the story of how the stickleback lost its pelvic girdle, see SHAPIRO et al. (2004) and CHAN et al. (2010). Gasterosteus aculeatus ranges into Greece, but there is no Aristotelian fish that can obviously be identified with it. A. used ‘definition’ in several different ways, see APo II esp. APo 94a11; here I mean ‘conclusion of the demonstration of what something is’ [trans. BARNES (1993)]. For the role that such causal definitions play in his science see Ch. XLVII. Teleological demonstrations, e.g. PA 640a1, are discussed by LLOYD (1996) ch. 1, LEUNISSEN (2010a). A. discusses the need for primary definitions at APo II, 19; see GOTTHELF (2012) ch. 7 for other examples. BYRNE (1997) pp. 207–11 discusses A. v. the sophists.

  XLIV

  Problems with, and critiques of, A.’s theory of demonstration. A. is certainly not oblivious to the problems of (i) falsely inferring causes from associations, (ii) falsely inferring the direction of causation, and (iii) multiple causation. In APo I, 13 he distinguishes between the ‘fact’, by which he seems to mean just the association proved by the syllogism, and the ‘reasoned fact’ by which he seems to mean the association + some other information that will convince us that there is indeed a causal connection, and which way the causal connection runs. In short he seems to argue, reasonably, that some other source of information, external to the syllogism, will show that there is a causal connection and what it is, but his discussion is not very clear; see LENNOX (2001b) ch. 2. The question of why A.’s works aren’t arrayed in syllogistic form has led to much discussion. BARNES (1996) pp. 36–9 lays out one solution, but Kosman, quoted in GOTTHELF (2012) ch. 7, says that the issue is a red herring for nowhere does A. say that science should be presented in this way. I think he doesn’t present his science syllogistically because he can’t. True, APo I, 30 claims that demonstrations can involve relations that are ‘for the most part’, but this seems to violate his universality requirement; BARNES (1993) p. 192 and HANKINSON (1995) discuss the difficulties. A.’s ad hoc explanation for why camels don’t have horns is given at PA 674a30; LENNOX (2001a) pp. 280–1. Demonstration blends into dialectic at EN 1145b2 [trans. modified from NUSSBAUM (1982)]. This leads us to the controversy over the degree to which A.’s official theory of demonstration is found in his biology. Some scholars think that the biology is richly informed by the official theory; others are more ambivalent and point to a diversity of methods of demonstration. Major discussions can be found in BOLTON (1987), LLOYD (1996) ch. 1, LENNOX (2001b) chs 1, 2, LEUNISSEN (2007), LEUNISSEN (2010a), LEUNISSEN (2010b) and GOTTHELF (2012) chs 7–9. A. discusses how to cope with multiple causes – divide and explain – at APo II, 13–18; LENNOX (2001b) ch. 1. He talks about dividing and explaining when prescribing for ocular ailments at APo 97b25 [trans. BARNES (1993)]; for its modern equivalent in cancer research see HARBOUR et al. (2010).

  XLV

  The functional beauty of birds. On the bird winds see Meteor 362a24. A. describes birds and their habits in HA VII, 3 and the tyrannos at HA 592b23. He speaks of the more and the less of bird features at PA 692b4 and the relationship between bird diversity and bios at PA 662a34, PA 674b18, PA 692b20, PA 693a11, PA 694a15, PA 694b12; cf. GA 749a35. See WILSON (1999) for guilds and functional groups in modern ecology. D
ARWIN (1845) p. 380 gives the famous passage about the birds of the Galapagos. A. speaks of how nature makes instruments to fit the function at PA 694b12.

  XLVI

  Teleology in the zoology. A. speaks of the primacy of final causes at PA 639b13 and PA 646a25; see LEUNISSEN (2010a) ch. 7.1, and the reasons that organisms reproduce at GC 338b1, DA 415a25 and GA 731b31. Strictly this argument applies only to (i) sublunary organisms (i.e. it excludes heavenly organisms); (ii) organisms that reproduce (i.e. it excludes spontaneous generators). Here, and elsewhere (Ch. XCVI), I argue that the beneficiaries of reproduction are forms; see LENNOX (2001b) ch. 6 for a somewhat different view.

  XLVII

  Explaining the elephant. A. discusses the elephant’s trunk at HA 497b26, HA 536b20 and HA 630b26; he explains it at PA 658b34 and PA 661a26. There is some inconsistency between HA and PA concerning the lifestyle of elephants. In PA, the aquatic habitats of the elephant are strongly stressed; in HA, however, the elephant is not named among the amphibious animals, and although it clearly lives near rivers, it does not live in them, and is a poor swimmer; see LENNOX (2001a) p. 234, KULLMANN (2007) pp. 469–73 and, especially, GOTTHELF (2012) ch. 8 on A.’s analysis of the elephant. JOHNSON (1980) describes snorkelling elephants, but these days you can also see them on YouTube. At PA 659a25 A. argues that the elephant’s legs are ‘unsuitable for bending’ but elsewhere – HA 498a8, IA 709a10, IA 712a11 – he suggests that they can bend, though the latter passages are, admittedly, unclear. Ctesias is often blamed for telling A. about the elephant’s inflexibility, but no existing text corroborates this, BIGWOOD (1993). For the elephant’s leg in history see TENNANT (1867) pp. 32–42; for the modern kinematics of the elephant’s legs see REN et al. (2008), and for the aquatic ancestry of the elephant see GAETH et al. (1999) and WEST et al. (2003).