The Lagoon

by Armand Marie Leroi

  The two styles also have a very different relationship between theory and data. In the first, a specific hypothesis is tested. The result is either consistent with the hypothesis or not. In the second, a narrative is constructed. You let the data speak. What they tell you is, of course, strongly influenced by what you think and hope you’ll hear. When Glanvill complained that Aristotle forces ‘Experience to . . . yield countenance to his precarious Propositions’, he identified the danger. We are much more keenly aware than Aristotle was that any given empirical pattern may be explained by several different models, yet even so remain susceptible to the same mistake. That is why both the styles of science that I have spoken of are needed. Data trawls and pattern analyses give you models; targeted experiments tell you whether or not they’re true. Many scientists use them both.

  Besides the lack of strong causal inference, the second – Aristotelian – style has another weakness. Lots of data always means low-quality data, especially when it’s gathered from anywhere and everywhere.* Genbank – the vast database in which biologists deposit their DNA sequences – is notorious for its errors. This doesn’t stop them from using it. They take the data, run what checks they can and hope that the errors cancel out and that truth will emerge in the aggregate. That, too, seems to be very much Aristotle’s style. He makes hundreds of factual claims that he must have known were based on uncertain data. It was probably a deliberate choice. His data were grist for the empirical generalizations and causal theories that were his ultimate goal; and he seems to have felt that the risk of incorporating dubious data was worth the prize of finding them. The data from his soothsayer-derived passages on animal behaviour are feeble. But they do illustrate the different kinds of interactions among animals, some of them competitive and some of them predatory, and suffice for an important generalization about how agonistic interactions among animals increase when food runs short. In The Heavens he suggests that we should still theorize even when the evidence is very slight and the object of our investigations is far away. History judges such scientists bold when they’re right – and rash when they’re wrong.

  CXII

  THERE IS A belief, and I think it is a very widespread one, that something is wrong with Aristotle’s explanations; that they are, in some way, fundamentally unscientific. Sometimes it is said that his appeal to the ‘natures’ of things is circular. In Le Malade imaginaire, 1673, Molière’s Aristotelian quacks explain that opium induces sleep because it possesses a sleep-inducing principle. Ever since, arguments of this kind have been known as virtus dormitiva explanations and rightly treated with scorn. At other times it is said that Aristotelian natures possess a ‘creative impulse’ or else ‘occult forces’. Applied to his biology, these are polite ways of saying that he’s a vitalist – which many have said too. And then there are those who have said that final or formal causes are those creative impulses and occult forces and have no place in modern science.

  All of these charges, endlessly repeated, are echoes of the Scientific Revolution. Often they have been repeated by Aristotle’s foes who knew little of what he said or did. Yet even those who have known Aristotle intimately, and loved him dearly, have sometimes thought his explanations bankrupt. William Ogle did. So too, remarkably, did D’Arcy Thompson. On Growth and Form, the strange and beguilingly beautiful book that he published just seven years after his translation of Historia animalium, is a paean to Democritus.

  Over the last fifty years or so, scholars have explored, uncovered and displayed the explanatory wealth of Aristotle’s biology as never before. I have tried to show something of what they have found. Their discoveries, and our ever-changing understanding of the natural world, demand that we reopen the investigation into these antique accusations.

  The claim that Aristotle’s explanations are not merely wrong but unscientific comes down to the claim – as old as Bacon – that they are unmechanistic. Let us accept its premise: that a scientific account, ancient or modern, of some phenomenon must give a mechanistic explanation for it, or at least permit the possibility of one. Most scientists will find this uncontroversial. The question is: what do we mean by ‘mechanistic’?

  The term is a slippery one. We can surely agree that a mechanistic explanation is, minimally, one couched in terms of a physical theory. Beyond that, views vary. Here are a few definitions that I think are wrong. Some philosophers and historians also demand that the physical theory in question be a correct one, or at least a particular one – Newtonian mechanics or atomism, say. Such restrictions are obviously ahistorical. Why should any particular physical theory be so privileged? Physical theories come and go: the discovery of subatomic particles may have rendered Dalton’s atomic chemistry redundant or even wrong, but not unmechanistic, much less unscientific.

  Mechanistic explanations are also sometimes held to be those that eschew any reference to final or formal causes. That seems wrong too. Certain kinds of complex phenomena demand final and formal explanations; mechanistic explanations are not excluded by, but rather complement, them. Other philosophers demand that mechanistic explanations contain explicit comparisons to machines – pulleys or clocks, say. This is also too restrictive. Ask a biologist how proteins are made in the cell, and he will tell you about the ‘ribosomal machinery’. Ask him then what kind of artefact a ribosome resembles and he will say, well, it’s a bit like a CD player since they both translate information encoded in one physical form into another, and it’s a bit like a locomotive since they both travel down ‘tracks’ (mRNA). Probe a little further, however, and he will also acknowledge the vacuity of the similes, that humans have never built anything like a ribosome, nor indeed anything so clever, but that even so the physics do make sense.

  I propose, then, that a mechanistic explanation is simply one that explains a phenomenon in terms of the physical theory of the day. Granting this definition, Aristotle biology is replete with them. They are two of the four planks of his explanatory scheme, the moving and material causes. He is always, to be sure, saying that it is the ‘nature’ of some animal to do this or that and, had he left it there, his explanations would indeed be vacuous or occult. He doesn’t. He then explains how and why.

  In this book I have sketched Aristotle’s account of five interlocked biological processes: (i) the nutritional system by which an animal takes up complex matter from its environment, alters its qualities and redistributes it to its various tissues so that it can grow, thrive and reproduce; (ii) the thermoregulatory cycle by which it maintains itself and which, as it ages, falls apart; (iii) the CIOM system by which an animal perceives and responds to its environment; (iv) the epigenetic processes of embryonic development and its related spontaneous-generator version; (v) the inheritance system. All of these processes are underpinned by Aristotle’s physical theory and are, as such, mechanistic. That the physical theory is wrong is irrelevant; in the long run, all physical theories are.

  All these processes explain some part of the workings of the soul. But soul is not something superadded to them: they are, collectively, soul; more precisely, soul is the dynamic structure of these physical processes (or their result). Again, that Aristotelian souls run on an obsolete theory of motion, a defunct chemistry and an oft-erroneous anatomy is beside the point. Descartes, for all his běte machine rhetoric, had his animals move by means of ‘animal spirits’ percolating through their nervous systems – pneuma by another name. If Aristotle’s biology becomes unmechanistic at any point, it’s when he considers higher cognitive functions – phantasia, reasoning, desire. They’re merely black boxes. But we can forgive him this – they are for us too.

  Although mechanical similes are not needed for a theory to be mechanistic, they are often the sign of one. When explaining how animals work Aristotle incessantly invokes them. Bellows, irrigation ditches, porous pottery, cheese-making, toy carts and, of course, those enigmatic automatic puppets, all appear in his biology. For all that, he never draws the Cartesian comparison of a whole creature to a
machine. Doubtless this is because the mechanical devices of Aristotle’s day were so rudimentary.* We can see that his heart–lung cycle is a thermostat but he obviously didn’t – he just said how he thought it all works.

  This, then, is Aristotle’s dilemma. He sees that artefacts and living things are both made of more basic stuff, that they change and that these changes must be explicable in terms of physical principles. Yet, when looking at his world, he also sees that there is no artefact remotely capable of doing what creatures so effortlessly do. His solution is to acknowledge the parallels but keep them firmly apart. The cybernetic properties of living things even cause him to give them the special ontological status of ‘entities’ – ousai – while denying that status to artefacts. He would surely have dismissed Descartes’ talk of beast machines as empty rhetoric. In Descartes’ hands it was. It wouldn’t stay that way.

  Aristotle’s enemies (and some of his friends) have also made formal and final causes far more mysterious than they really are. Aristotle saw that complex objects – and nothing is more complex than a living thing – cannot assemble willy-nilly by chance but must be modelled on a pattern located elsewhere. Long absent from science, molecular biology made form – eidos – respectable again. In What is Life? Schrödinger, quoting Goethe (‘Being is eternal; for there are laws to conserve the treasures of life on which the Universe draws for beauty’), argued that the chromosomes, which he envisioned as aperiodic crystals, contain a ‘code-script’ and are ‘the law-code and executive power – or, to use another simile, they are architect’s plan and builder’s craft – in one’. The last is one of Aristotle’s similes too. It was Max Delbrück at Caltech who made the connection explicit. In his charming essay ‘Aristotle-totle-totle’ he told of how, in the course of a long correspondence with André Lwoff at the Institut Pasteur in Paris, he discovered the Philosopher’s works. After quoting bits from The Generation of Animals he wrote, ‘What all of these quotations say is this: The form principle is the information which is stored in the semen. After fertilization it is read out in a pre-programmed way; the readout alters the matter upon which it acts, but it does not alter the stored information, which is not, properly speaking, part of the finished product.’ And then he suggested that, were Nobels handed out posthumously, Aristotle should get one for discovering the principle (if hardly the substance, much less the structure) of DNA. In 1969 Delbrück got one for his work on mutation.

  Final causes, too, have been demystified. Aristotle saw that they are needed when the phenomenon to be explained appears to have a goal. They arise then as the answers to several related questions which he asked and which modern biologists do too. When we ask why do goal-directed entities exist, we give Darwin’s answer: because evolution by natural selection produced them. That is shorthand for the whole edifice of population genetic theory that renders benevolent creators null and void. When we ask what their goals are, we answer by pointing to all the adaptive devices that allow them to feed, move, mate, defy their predators and, ultimately, survive and reproduce. It is Bacon’s sneers at teleological explanations of this sort, those ‘remoras and hindrances’, that now look quaint. To argue, as he did, that the functional study of eyelashes, skin and bones should be no part of science is to betray a remarkable incuriosity about the point of one’s own body.

  We can also ask how goal-directed things, living or not, work. That is the most difficult kind of final explanation, and its answer lies in the beating heart of the science of complex objects. Cybernetics, General Systems Theory and Control Theory formalize the general principles; systems biology shows those principles at work in living things; synthetic biology how those same principles can be used to reshape them. In 2010 JCVI-syn1.0, the world’s first artificial cellular life form, fired its molecular motors. The distinction between artefact and organism dissolved in a Petri dish.

  Aristotle’s answers to these questions, all of which are embraced by his final cause, are sometimes similar to ours and sometimes, but hardly surprisingly so, very different. That they are scientific questions and that he gave scientific answers to them cannot be denied; or at least that he did so until he looked to his God to give creatures, not least himself, their ultimate purpose in life.

  What, finally, of Bacon’s accusation that Aristotle’s science was useless to man? It’s the eternal cri de coeur of the science bureaucrat. (You scientists want all the money going, but what, exactly, do we get in return?) Neither complaint, Bacon’s nor the bureaucrat’s, is entirely baseless. But, just as few modern scientists are utterly indifferent to the utility of their work, neither was Aristotle. His father was a physician, so it’s no surprise to find two books titled On Medicine listed among his lost works. And, although his books on ageing –Youth & Old Age, Life & Death and The Length and Shortness of Life – do not reveal what we can do to nurture the internal fire whose vitality dictates the length of our days, he does conclude the latter with this:

  Our investigation into life, death and related subjects is almost complete. On health and disease it is to some extent up to natural scientists as well as doctors to consider their causes. But it is important to note the differences between these two groups of investigators in how they treat different problems; for it is clear that to some extent they cover the same ground; doctors who display curiosity and intellectual flexibility have something to say about natural science and declare that their theories arise from it and the best practitioners in natural sciences tend to end up with medical theories [italics mine].

  Think of it as the Invitation to Biomedical Science. ‘This our science’, wrote D’Arcy Thompson, ‘is no petty handicraft, no narrow discipline. It was great, and big, in Aristotle’s hands, and it has grown gigantic since his day.’ Aristotle could not have conceived just how vast the science that he founded would become. Yet, as I contemplate the elaborate tapestry of his science, and compare it to ours, I conclude that we can now see his intentions and accomplishments more clearly than any previous age has seen them and that, if this is so, it is because we have caught up with him.

  CXIII

  AND BECAUSE WE know, and know intimately, the one other scientist in history who resembles him more than any other.

  They were so very much alike. Both were the sons of famous physicians, but both preferred to study nature. Both were voracious for facts. Both were ruthlessly, powerfully logical – and not much good at maths. Both were bold and rash in equal measure and, in being so, left us visions of life imbued with – there is only one word – grandeur. If there is a difference, it is only in the scale of their accomplishments. After all, Darwin didn’t invent science itself from scratch; Aristotle did.

  They also shared a scientific style. Seeking facts to support their theories, both cast their nets wide to catch them. Both interrogated farmers, fishermen, hunters and travellers – though Darwin could add pigeon fanciers to the list too.* Both papered over vast inferential cracks in their evidence – Darwin, inter alia, the mechanisms of heredity, gaps in the fossil record and the invisibility of natural selection. Both made voluminous, if often fleeting, observations. And both men occasionally made far too much of the facts they knew, or thought they knew.

  In The Origin of Species, Darwin tells of a small rodent, the Tuco-Tuco, Ctenomys, which infests the Argentinean pampas. It lives in burrows and, so Darwin assures us, is frequently blind; indeed, one that he had kept alive while travelling with the Beagle ‘was certainly in this condition, the cause, as appeared on dissection, having been inflammation of the nictitating membrane’. Such an inflammation, he continues, being injurious to the animal, would tend to select for eyeless Tuco-Tucos and so, eventually, result in something like a mole. It’s a very reasonable argument and an important one too, for it’s the only example that Darwin has of natural selection, the driving force of evolution, in action. Unfortunately, it is almost certainly not true. Some years ago, following in Darwin’s footsteps, I searched for teary-eyed Tuco-Tucos in Argentina and Uruguay
in vain. I interrogated gauchos and scientists, and all denied that the animals have anything wrong with their eyes. A gaucho offered an explanation for Darwin’s observation: ‘Well, you know, when we catch the Tuco-Tuco we hit it with a spade. They move fast and they are fierce! Maybe this is why Carlos Darwin’s Tuco-Tuco bleed from the eyes, eh?’

  The lesson is one that every biologist, every scientist, knows or must learn: that the practice of science demands a peculiar intimacy with the object of your investigations. You must know its form, its foibles, its pretty little ways, for, if you don’t, you’ll make a mistake or else miss some astonishing thing that it does, and that is almost as bad. That is why Darwin spent eight years on barnacles. He sought, in Barbara McClintock’s phrase, a ‘feeling for the organism’. My own postdoctoral supervisor expressed the same sentiment, albeit more restrictively, when, on the first day in his lab, he told me: ‘Know the Worm.’ A Delphic utterance? Not at all. I knew exactly what he meant.*

  Aristotle, I believe, would have too. Intimacy with the natural world shines from his works; it does from Theophrastus’ as well. This intimacy allowed them, the men of the Lyceum, to begin the process of sieving the ocean of natural history folklore and travelogue for grains of truth from which to build a new science. Aristotle even said so:

  Failure to understand what is obvious can be caused by inexperience: those who have spent more time with the natural world are better at suggesting theories of wide explanatory scope. Those who have spent time arguing instead of studying things as they are show all too clearly that they are incapable of seeing much at all.