No Such Thing as a Fish

In 1952, Gustav Wängsjö published a 612-page monograph on early fossil vertebrates from the Arctic island of Spitsbergen. These fossils were the remains of sluggish, fish-like animals covered in bony plates, on which, more than 380 million years ago, the finest details of nerves and arteries had been as faithfully impressed as fingerprints in clay. Wängsjö studied them with the care that a criminologist might apply to the dabs of a suspect, but the most remarkable feature of his monograph is its postscript. In the space of two pages Wängsjö was forced to retract most of what he had written in the previous six hundred. His interpretations were not correct, he confessed; instead, his findings confirmed the previously published ideas of Erik Stensiö, doyen of Scandinavian palaeontologists. It was reported that Stensiö had said, when Wängsjö presented his dissertation: ‘If you publish this, you will amount to nothing more than a schoolmaster in Norrköping.’ Thus did academic authority cauterise the truth; for time has confirmed most of the younger man’s conclusions. Stensiö lived to more than ninety, working every day in the Natural History Museum in Stockholm. Wängsjö became a schoolmaster in Norrköping. That’s authority for you.

Cladistics was to supply a response to such authority, as a method of classifying animals and plants that does not rely on the opinion of the expert of the day. Rather, it is an explicit and objective method which assesses the relative significance of animal and plant characteristics in order to create a rational classification. A senior professor could no longer dictate the interpretation younger palaeontologists must follow, for this new method proposed rules beyond the reach of his say-so. The reasoning behind cladistics was originally developed by a German entomologist, Willi Hennig, in the late 1950s. Instead of evolutionary trees featuring the ‘ancestors’ of a particular species, organisms were now classified by the features they shared with one another. Maps of descent were replaced by diagrams of relative common ancestry – known today as ‘cladograms’. Estimates of kinship replaced narratives of parentage. The familiar nostrum ‘mankind descended from the great apes’ was replaced by ‘great apes and mankind share a common ancestor.’ This might seem like pretty small beer, but it had profound consequences.

The 1980s were an exciting time for taxonomic biologists, as more and more converts were won over to this new method of understanding biological relationships. Henry Gee was one of those who hung around the Cranley Arms in South Kensington (‘The Cladist’s Arms’), where the leaders of the ‘British school’ supped more pints of Courage Director’s Bitter than was good for them and rubbished the reputations of the unbelievers. ‘Time, gentlemen, please’ was the nearest thing to the words ‘Deep Time’ that I heard in those heroic days. I suspect that Gee wishes he could still go down there. But the Cranley Arms has gone, and with it some of the ideas that were part of the intoxication.

Cladistic methods generated violent arguments. Some of the famous ones are described by Gee, and well described, too. During the Devonian period, more than 375 million years ago, the first primitive creatures took the step from water to land. The question was: which ‘fish’ was closest to the first terrestrial animal with a backbone that walked on four legs? Lifelong enmities were spawned from the resulting disagreements – which often centred on which particular fossil ‘fish’ might have provided the ideal ancestor. The argument from descent was central, but cladistic analysis meant we could take a fresh look at the whole business. One consequence was that the crossopterygian fishes (distantly related to the famous ‘living fossil’ coelacanth Latimeria) were dethroned from their privileged position as ancestors. Those palaeontologists who had used their authority to support one ancestor or another from among these animals were furious. The cladistic analysts then reconsidered the whole question of what actually constituted a fish. The answer came out clearly: there was no such thing as a fish. The term was a vernacular soubriquet for slippery, silvery things (mostly edible) that bore little relationship to any biological classification on cladistic lines. Pooh! came the response: we know a fish when we see one – as well as we know a mammal or a bird. The new mindset took some years to win the day, but it won in the end. Nowadays biologists are routinely able to cope with the notion that humans are more closely related to a lungfish than to a salmon, and laugh at the notion of superficial fishiness. And in the intervening years eight-fingered vertebrates close to the common ancestor of all land-lubber animals have been broken out of Devonian rocks in Greenland. The story of the colonisation of land has become both richer and more curious.

Then there was the dispute over whether the closest relatives to birds were, or were not, dinosaurs. There still are researchers who think of birds as special enough to require an equally special series of evolutionary events to allow them to take wing. The cladistic argument is far simpler: birds are little more than feathered dinosaurs, related to a graceful group of miniatures that flourished alongside the giants celebrated ad nauseam in Jurassic Park and Walking with Dinosaurs. Hard though it may seem to associate avian elegance with cumbersome and ferocious behemoths this was the story told by the cladograms, revealed by interpreting the bones of the famous old bird Archaeopteryx unfettered by preconceived notions. That theory has been triumphantly vindicated by the discovery of a whole aviary of new fossils from China, which are best described as dinosaurs with feathers: the most delicate part of a bird’s anatomy evidently anticipated its capacity to fly. These reptilian chickens scuttled along die ground. This means, of course, that the dinosaurs never really became extinct: they simply took to the air.

Cladistics has renewed interest in the descent of our own species, too. The obsession with finding ‘the missing link’ – a single, unique ancestor bridging the gap with the great apes – was replaced by the altogether more sensible idea of analysing newly discovered fossil hominids according to their progressive similarities to Homo sapiens without fingering any particular one as the ultimate grandfather. When the same techniques proved to be useful in processing the huge amount of information obtained from genetic sequencing (a gene embodies a history of its past in the same way that a bone does) the triumph of the cladistic method seemed complete. Here at last was a single technique which could order Nature’s variety, regardless of whether the evidence came from bones or DNA. It was the key to a general reference system for biology.

Henry Gee’s description of how the method works is clear, and even entertaining; all biology teachers will be grateful to him for that. It is a little too simple; you won’t find in the index any of the basic terms that an undergraduate learns straight away: ‘synapomorphy’ and ‘plesiomorphy’ are examples, inelegant terms perhaps, but it wouldn’t overwhelm the general reader to know what they mean. It’s a bit like having a book on evolution that leaves out the word ‘mutation’. But the didactic purpose is served admirably. So, one might ask, why did Gee go on to spoil things?

I have said nothing so far about the Deep Time of the tide. (When the phrase appears in the text it is always capitalised, as if to emphasise that what Gee has to say about it is Very Important.) To give a clear account of the cladistic method would have been service enough, but would not have generated much controversy. Hence the introduction of Deep Time. The idea seems to be to conflate Francis Fukuyama’s ‘End of History’ with Henry Ford’s ‘History is bunk.’ Deep Time, Gee says, makes the whole endeavour of reconstructing evolutionary history nugatory. Scenarios of descent are ‘just-so stories’ which should be despised by true sons and daughters of science. ‘We want to think of the history of life as a story,’ he says, ‘but that is precisely what we cannot do ... What we need is ... a kind of “anti-history” that recognises the special properties of Deep Time.’

This is not profound: it is profoundly misguided, on both a factual and a theoretical level. First, mere antiquity – ‘Deepness’ – is not a measure of what we can or cannot know. We know more about dynastic Egypt than we do about the Dark Ages, even though the former is three times further away from us in time than the latter: it is a matter of record. In the same way, thanks largely to microscopic fossils, we can reconstruct narratives of Cretaceous time, more than a hundred million years ago, with a precision undreamed of twenty years ago. In Gee’s book there is nothing about invertebrate fossils – humble things like ammonites and sea urchins. This omission is important because such fossils are a hundred thousand times more common than the dinosaurs, hominoids and even fishes which provide the examples in his case histories. It is these ordinary animals that have helped science cope with the problems of Deep Time (after so many of these references I am tempted to refer to the DTs), and it is astonishing that the biostratigraphy based on them doesn’t merit a mention.

The construction of the ‘biostratigraphic column’ is one of the scientific triumphs of the last century and a half. It doesn’t depend on assumptions about ancestors, and is now firmly linked to dates from radioactive ‘clocks’, which tell us how long ago a species might have lived. While it is perfectly true that narratives based on the discovery of randomly sampled rarities are not going to carry conviction in the long run (Gee’s selected examples show that well), this is not the same as jettisoning the whole time dimension of the fossil record. That is the equivalent of throwing out coinage as unreliable archaeological evidence in favour of analysing only royal burials.

Cladistic method deals in relative common ancestry; it does not have a built-in time-scale. Fossils provide the timescale. Many of the revolutions in our understanding of the Earth have come about through the intersection of Deep Time with evolution; I might cite the great swathes of Precambrian time, during which simple bacterial cells altered the atmosphere to permit the evolution of higher life-forms. Without our knowledge of time to lay alongside fossils and phylogeny we would have had no idea of this early history. Gee is right to say that cladistics is a method of analysis not a narrative, but some of the stories that can be told are both true and momentous, and the methodological purity he seems to advocate smacks of Professor Stensiö. Gee’s cladistics is that of the early 1980s, the days when Colin Patterson – perhaps the leading cladist of his time – and his friends were in the habit of saying that the fossil record was of no use, and that stratigraphy was just a spanner in the evolutionary works. Since then, there has been a rapprochement between time and cladograms of which one gets no hint here. The Gee version of the ‘revolution’ is already old hat. I guess his ideas froze at the time he walked out of the Cladist’s Arms and into the editorial office of Nature, where he still works.

Establishing relationships by the use of cladistics is important but it does not enable us to describe the world. A cladogram tells us nothing about abundance. In the democracy of the cladogram, modern man is the equal of Homo erectus and quite right, too, but there is no doubt whose face we now see from Pole to Equator and whose influence on the biosphere is the greater. Great shifts in the biological world have happened deep in the geological past – the rise of mammals after the demise of the dinosaurs 65 million years ago is just one example – but change in dominance is not necessarily reflected in a family tree. The mismatch between ‘branching time’ and ‘dominance time’ provides an interesting line of research made possible by the interplay of cladistics with Deep you-know-what.

The majority of my colleagues who do cladistic analyses find that they are most useful for identifying problems, and that’s been my experience, too, over twenty years as a practitioner (I don’t believe Gee has ever published a cladogram). Cladograms are not a methodological magic wand that creates order from chaos; successes are often followed by drawn-out and ambiguous campaigns. With each new analysis something on the cladogram changes. The philosophical difference between drawing up a new cladogram from new evidence and reciting a new historical scenario based on new evidence is a slender one: what matters is to accept that the narrative interpretation of the past is as much subject to reinterpretation as is a cladogram. The problem of authority exists only when some ornery individual pulls rank. Some cladistic theoreticians are themselves as bloody-minded as Stensiö was. (Grown men blanch at the prospect of being grilled at a conference by Dr Farris of the Centre for Evolutionary Studies, Stockholm, computer and theoretical pioneer.)

There is more silly stuff in this book. Gee thinks that speculation about the function of organs or the life habits of fossil organisms is not scientific: ‘hypotheses without tests are no more than cocktail party chatter and are without value except as entertainment – they are not science.’ Thus the whole discipline of functional morphology is ditched in a single phrase. Yet a great deal of good science has been done on the body functions of extinct animals: sophisticated engineering experiments, predictions from field occurrence and fossil tracks and so on. What made the TV series Walking with Dinosaurs irritating to the professional was its failure to draw any distinction between reasonable inference and sheer fantasy (all those twee cries). Of course we can never know, but compared with some of the speculation about what goes on in Deep Space (which seems to be scientifically acceptable) our adventures in the past are generally quite cautious.

What puzzled me most in Gee’s book was his glossing over of the way in which cladistics became enmeshed with creationism – an important episode, if only to demonstrate how ingenious ‘scientific’ fundamentalists can be. When Colin Patterson was giving lectures in the American Museum of Natural History in 1981 he was covertly tape-recorded by these people. His disavowal of the need to think in terms of ‘ancestors’ was given much publicity under such headlines as ‘Scientist does not believe in evolution.’ Perhaps Gee thinks that creationists have gone away: well, they haven’t, in fact they recently got the law in Kansas changed to guarantee the teaching there of ‘creation science’. I can imagine them rubbing their hands with glee when they read things like: ‘one can only long for the 18th century, when Linnaeus could classify animals and plants without worrying about evolution.’ I know this quotation is taken out of context, but that’s what creationists do all the time, and it’s a grave mistake to supply them with ammunition.

As for the greatest palaeontologist of the 20th century, Colin Patterson was in no doubt. It was Erik Stensiö, the grand old man of Stockholm. He had the authority that came from greatness; but, like all scientists, he wasn’t always right.