The Lagoon

by Armand Marie Leroi

  XXXIX

  ARISTOTLE’S SECOND METHODOLOGICAL insight is his solution to one of the great problems of biological classification – namely, the vexed tendency of organisms to display a mix of apparently incongruous features. Nature’s hierarchy isn’t neat; in fact, it’s a mess. Classify animals by their mode of reproduction (egg-layers v. live-bearers) and you will place them into two groups; classify them by their appendages (legs v. fins) and you will also place them into two groups – but quite different ones. There are, as taxonomists say, conflicts in the data, and either solution, as Aristotle says, risks tearing kinds apart. This is the problem that Plato’s method does not solve. In Platonic division each feature is considered sequentially with inevitably arbitrary results. Aristotle, however, has a far better feel for the order of nature. Here he is deciding how to divide up some land animals:

  Wingless tetrapods all have blood, but the live-bearing ones have hair, the egg-laying ones have scales, as an equivalent of fish scales. The snakes are a kind of animal that has blood, can move on land, but naturally lacks feet and has horny scales. Snakes are generally egg laying but the ekhidna (Ottoman viper) is exceptional in being live bearing. But not all live-bearing animals have hair, some fishes are also live bearing too.

  The trick, it seems, is to consider several features simultaneously – feet (four v. none), reproduction (live-bearing v. egg-laying) and covering (hair v. scales) – and in combination. Three features with two states each yields eight possible combinations, eight possible kinds of animals. But only four actually exist:

  (1) hairy, live-bearing tetrapods

  (2) scaly, egg-laying tetrapods

  (3) scaly, egg-laying apods

  (4) scaly, live-bearing apods

  The first three of these are greatest kinds: zōotoka tetrapoda, ōiotoka tetrapoda, opheis (snakes). The fourth combination, the exceptional viper, is in all respects a snake except that it gives birth to live young. So how should it be classified? Had a Platonic taxonomist considered the viper he would have defined it as a ‘scaly, live-bearing apod’ and so sundered it from the rest of the snakes. Aristotle is subtler. For him a kind is a group of similar creatures, but one with fuzzy boundaries.* Quite sensibly, then, he’s clear that, although the viper may be viviparous, it is nevertheless a snake. This pragmatism is very Aristotelian. He’s always talking about things that are ‘for the most part’ true, as if the organic world is filled with exceptions that one should note but not fuss too much about.

  This sounds casual, but in fact the Platonic and Aristotelian approaches to division represent two very different ways of carving up the world. In ‘monothetic’ classifications the presence of a feature state (e.g. live-bearing) is necessary and sufficient for that object to be included in a class (of live-bearing animals); in ‘polythetic’ classifications classes are identified by the central tendency of all the features and possession of no single feature state is either necessary or sufficient for class membership.* When delineating genē Aristotle takes an implicitly probabilistic stance, analyses the feature matrix and clusters. He needs no computer to do this. When classifying, humans naturally attend to many features and look for associations among them. It is in this spirit that Aristotle says we should begin with the genē that most people use (birds, fish) – at least we should when they’ve got it right.

  The viper is not the only troublesome creature in Aristotle’s bestiary. Ostriches, apes, bats, seals and dolphins are also hard to classify. Most of these animals have features that point to divergent affinities. The origin of the problem is as plain to us as it was obscure to Aristotle: the vagaries of evolutionary history. Closely related species tend to have many features in common due to their descent from a common ancestor. Distantly related species may, however, also share features due to convergent evolution – birds and bats may both have wings, but that does not mean that they are related. Animals may also be confusing mosaics of ancestral and derived features – witness the egg-laying, hair-coated, milk-secreting, duck-billed platypus. The history of systematics can be written as the search for a solution to such confusions. Aristotle may not have understood the cause but he saw and dealt with the consequences. He used the word epamphoterizein – to dualize – of animals whose bodies pointed two ways.

  Aristotle classifies some dualizers as he does the viper, by blurring the boundaries of an existing greatest kind. The strouthos Libykos (ostrich, literally ‘Libyan sparrow’) appears, on balance, to be a bird. He avoids the implications of the Barbary macaque. He says that it has some human features (face, teeth, eyelashes, limbs, hands, chest, female genitals, no tail), some tetrapod features (hair, hips, general proportions, male genitals) and some unique features (hind feet that resemble hands), but not where in his classification it should go. On the dolphin, however, he is decisively radical.

  XL

  IN THE MIDDLE OF relating Greece’s battle-scarred dynastic history, Herodotus, inconsequentially, tells the story of Arion, a musician from Lesbos. The beauty of Arion’s music, Herodotus says, was second to none; he invented the dithyramb, the wild measure of the Dionysiac hymns. Arion had long lived at Corinth. This is in the time of the Tyranny of Periander which puts it in the mid- to late seventh century BC. Then Arion moved to Sicily, where he played his harp and grew rich. But after a while he yearned for rocky Corinth and so hired a ship and crew at Tarentum in Apulia to take him back. The crew were Corinthians, so decent fellows all, except that they weren’t. As Italy disappeared, spotting his cash, they made to pitch Arion overboard. Not so fast! said Arion, let me sing for you first. Why not? said the crew. And so Arion put on his finery, plucked his harp, sang his song and then threw himself over the side where a friendly dolphin picked him up, inquired where he was bound and gave him a lift all the way back to Corinth. Of course no one there believed this tale, but then the crew showed up, were duly shocked at finding Arion alive and so confirmed their guilt. And there is still, concludes Herodotus, at Taenarum a shrine with a small bronze statue of a man perched upon a dolphin’s back.

  That Herodotus has Arion leaving Italy via Tarentum (Taranto) is no coincidence, for a youth riding a dolphin was entangled in the city’s foundation myth and stamped on its coins. Pausanias, Aelian, Pliny, Oppian, Ovid and a dozen other ancient writers besides tell of Arion or other dolphin riders, but Aristotle, seeking the plausible core of the myth, only says: ‘On sea-animals: much evidence attests to the mildness and gentleness of dolphins and the passion of their love for boys in the regions of Taras, Caria and elsewhere.’ If that sounds paedophilic, it does in the Greek too. He then goes on to tell how dolphins protect their own, particularly their young, but he’s mostly interested in their anatomy.

  Dolphins, Aristotle says, are supremely swift swimmers and voracious hunters. He says that they copulate and give birth to one or two live young that they suckle via ventral slits. They have internal testicles near the belly and no gall bladders. They have proper bone. They breathe air, have a windpipe and lungs, and a blowhole through which they spout water. When hunting they will plunge into the deep, calculate how long they can stay down, and then shoot to the surface like arrows, flying out into the air, sometimes clearing the masts of boats. As such, they’re just like divers bolting for the surface. When caught underwater in nets they drown, but conversely survive for a long time on land. If taken from the sea they moan but cannot articulate for their tongues are immobile and they don’t have lips. Sleeping dolphins do, however, snore – or so it’s said. They live in male and female pairs for up to thirty years. We know this because fishermen nick their tails and then release them again – which seems to be an account of history’s first mark–recapture study. Sometimes they strand themselves for no obvious reason at all.

  Most of this is accurate. That dolphins snore is dubious, but perhaps we’ll let it go since they do, apparently, vocalize in their sleep. Some scholars think that Aristotle must have dissected a dolphin. I don’t for he also makes some serious mistakes. He
says – and says twice – that the dolphin’s mouth is slung under its head rather like a shark’s. That’s an error made by someone who never saw a dolphin close up. (Pliny, amplifying Aristotle’s error, says that dolphins have their mouths on their bellies, which makes you think that, where the Greek is sometimes wrong, the Roman is often a fool.)* Aristotle also thinks that the blowhole is connected to the mouth since he says it expels water taken in during feeding, but it isn’t and doesn’t. It’s clear that he got his anatomy from some fisherman who butchered a dolphin on the beach. It’s often said that the Greeks cherished dolphins as sacred animals. Oppian, who loved dolphins, said that to hunt them is immoral, as loathsome as homicide, and described, in terms that would do a Greenpeace activist proud, how the beastly Thracians harpooned them. But dolphin hunts must have been widespread for Aristotle describes another technique. He says that nets are set in utter silence and then, when the dolphins are encircled, the hunters make a racket which stupefies and entraps them. There’s no hint of censure: he’s interested only in the fact that dolphins can evidently hear even though they don’t have ears.

  Whether first hand or not, Aristotle put his dolphin anatomy to good use. Although it is in many ways like a fish, he recognizes that its moans and snores, lungs and bones, internal testicles and live-born-and-milk-suckling offspring are typically tetrapod features. It also has a feature of its own, the blowhole. In The Parts of Animals he seems unsure what to do with the dolphin, but in Historia animalium, parts of which were probably written later, he assigns it, along with the porpoise and the whale, to a new greatest kind, the kētōdeis from whence derives our Cetacea. He was probably led to erect a new taxon by the fact that several kinds of animals shared this distinctive combination of features; he is a taxonomic pragmatist. Aristotle did not label the cetaceans as mammals since ‘mammal’ was a concept that he did not understand. For him the cetaceans were just one of the great kinds of blooded animals of equal rank to the birds, fish and viviparous tetrapods. Still, he did a lot better than his successors who for two thousand years just called them ‘fish’.

  I have never seen dolphins in Kalloni, but they are sometimes there. A fisherman told me that, in the summer of 2011, a large pod of bottlenoses entered the Lagoon to hunt. Some fishermen, other fishermen he implied, not him, though this was not completely clear, rounded them up and killed them. He explained that the youngsters damage the nets that cost three thousand euros each and that, of the fifty bottlenoses, three got away.

  XLI

  IN CONGRATULATING ARISTOTLE ON his success in ordering the animals I have, however, elided the problem that I began with – namely, whether or not his project was, at heart, a taxonomic one. The zoologists of the eighteenth and nineteenth centuries thought it was. We should not take them at their word. They sought an illustrious predecessor. There are reasons to doubt whether he really was anything of the sort.

  Contra Cuvier, Aristotle never produces anything resembling a coherent, comprehensive classification in which every animal has its place. Further, though he may have perceived nature’s hierarchy, he does not name its levels: from Race to Kingdom, genos suffices for all. He also never tells us how to distinguish one kind from another, and he’s terribly casual about names. In Aristotle’s day salted Sardina pilchardus and Sprattus sprattus were Aegean staples, but he mentions neither the ‘sardella’ nor the ‘papallina’ for both are Roman names. Instead he speaks of the membras, the khalkis, the trikhis, the trikhias and the thritta, all of which seem to be clupeids, but whether sprats, sardines, shads or pilchards (to introduce the equally underdetermined English names) is hard to say, for he gives us few clues to their identities. His higher taxa are feeble. He sees that snakes and lizards are somehow kin but doesn’t bother to give them a family name. He forgets to tell us whether bats are birds or tetrapods or something else again. He never gives the diagnostic features that might define a kind in a usable way; never says ‘a fish is an animal that has gills + scales + fins . . . etc.’, but instead just says ‘fish are a kind’ and assumes that everyone knows what one is. Compare the relentless lists and tables of names and definitions in Systema naturae with the narrative discursions of Historia animalium and it’s plain that very different scientific agendas are at work.

  Aristotle seems to name and classify only when some other purpose demands it. He even says as much. When describing animals we could, he says, talk about a sparrow or crane individually, but ‘insofar as this will result in speaking many times about the same property because it belongs in common to many things, in this respect speaking separately about each one is somewhat silly and tedious’. It’s simply much easier to discuss larger groups composed of animals that have a lot in common.

  But if Aristotle’s descriptive biology is neither Pliniesque natural history nor Linnaean taxonomy what, exactly, is it all about? A hint comes from the structure of Historia animalium itself. At the start of the book he considers how to reduce his data to some semblance of order. The problem that he faces is that faced by any zoologist: should he order it by taxon (e.g. reptile, fish, bird) or by feature (e.g. reproductive system, digestive system, behaviour, ecology)? His solution, a sensible one, is to compromise: ‘Animals differ from each other in their mode of subsistence, in their habits, and in their parts. Concerning these differences we shall first speak in broad and general terms and subsequently we shall treat them with close reference to each particular kind.’

  He begins with a general synopsis, paying particular attention to humans, his model. He then treats the gross anatomy of blooded animals: limbs, skin, secondary sexual characteristics, alimentary system, respiratory system, excretory system. Then he considers the bloodless animals system by system, returns to the blooded animals for a look at their sensory systems, the sounds they make and how they sleep. Then come two books on reproductive organs and behaviours again ordered by blooded v. bloodless animals, a book on habits and habitats, one on behaviour and, finally, a book on human reproduction. By the end it’s apparent that he has constructed a comparative zoology – the first.

  He looks at feet and describes how some of the blooded live-bearing tetrapods (mammals) have many digits (man, lion, dog, leopard) while others (sheep, goat, deer, pig) have bifurcate feet with hooves instead of nails and others (horses) have a single solid hoof. Elsewhere he examines fish guts. Besides the usual stomach and intestines many fish have pyloric caecae, appendages that increase the absorptive surface of the intestine. He describes how they vary in number and position. Elsewhere he considers distribution of the sense of smell and so on. All this is a precursor not to the great systematic monographs such as Cuvier’s Poissons but rather to his Anatomie comparée or Owen’s Vertebrate Zoology (1866) in which the animals are cut into bits. You can read Aristotle on tetrapod feet or fish guts, find the cognate sections in Owen and illustrate Aristotle with his plates. We are in the Bird Hall looking at the Cabinet of Parts – he has a section on bird bills and feet too.

  But it isn’t easy to work out Aristotle’s aims. Like all of his extant works, Historia animalium is poor in structure and rich in redundant, inconsistent, misplaced and barely assimilated data. The reader itches to edit it. It was never a polished piece of work, but always a thing in flux; he seems to have composed it piecemeal, adding new information as it came, or else revising in the light of theory worked out elsewhere. It’s also been messed about with – though by whom and how much is hard to say.

  Even so, modern scholars generally agree that it does have a clear purpose. Beneath the disorder, it provides the materials for a data trawl. Aristotle is searching for patterns – patterns of a very subtle sort. He isn’t interested merely in how parts vary, but also in how they covary. This is how he describes the famously intricate four-chambered stomach of a ruminant (modern terms interpolated):

  Live-bearing horned tetrapods which have unequal numbers of teeth in the upper and lower jaw (also called ruminants) have four chambers. The stomakhos [oesophagus], startin
g from the mouth, goes down past the lung from the midriff to the megalē koilia [rumen]. This is rough on the inside and partitioned. And attached to it near the entrance to the oesophagus is the kekryphalos [reticulum], so called because it looks like a stomach outside but inside is like those woven hair-caps. The reticulum is much smaller than the stomach. Connected to this is the ekhinos [omasum], rough and laminated on the inside and similar in size to the reticulum. After this is what is called the enhystron [abomasum], greater in size than the omasum and more elongated in shape. It has many large smooth internal folds. Straight after this comes the gut.

  The description is detailed and true, but its real interest lies in how he introduces this strange stomach as a property of live-bearing tetrapods that are horned and do not have the same number of teeth in both jaws (he’s thinking of the incisors and canines missing in the upper jaws of many ruminants). It is from such associations that Aristotle constructs his greatest kinds, but it’s the associations themselves that he’s after. You can pull together his data and present them as a data matrix that includes, say, six classes of features (tooth number, stomach type, foot type, etc.) and twelve kinds of animals (cattle, pig, horse, lion, etc.).* It shows how the various features go (imperfectly) together. He never constructed such a table – everything is explained, laboriously, in words. But that he had something like this in mind is clear from Historia animalium’s sequel, The Parts of Animals, in which he summarizes the patterns of variation and covariation that he has discovered and explains why they exist. He pulls his data together and weaves a vast causal web that has a single purpose: to discover the true natures of living things.