Brief Candle in the Dark

by Richard Dawkins

  I loved watching the building behaviour of these wasps, partly because I was writing The Extended Phenotype at the time, and animal artefacts played a starring role in my argument. I was especially charmed by the wasps’ habit of exploiting what looked to me like the physical phenomenon known as thixotropy, as a kind of ‘welding’ technique. When a wasp returned to her tube with a ball of mud in her mouth, she would apply the ball to the lip of her tube. She then buzzed her wings noisily and the vibration transmitted through her jaws visibly caused the mud to ‘melt’, like quicksand: not just the new ball of mud but (I suspect, although I couldn’t see for sure) the mud of the lip of the tube, so they melted into each other. Melting, melding, welding – it did indeed look like welding.1 In fact, it looked to me as though the vibration was acting on the mud in the same way as the heat of a welder’s acetylene flame acts on metal, temporarily liquefying the lip so that new mud bonded securely with old. Jane doesn’t think anyone has published this ‘thixotropy’ suggestion before, so I offer it here for what it’s worth.

  Jane and I conducted a graduate seminar in evolution and behaviour at Gainesville, in which we were joined by two other professors. My chief memory of these weekly gatherings was the way in which they became increasingly dominated by the intellectual power of Alan Grafen. On the face of it he was just a graduate student among many (and one of the youngest), but it was remarkable how we all, students and professors alike, fell into the habit of turning to Alan to resolve our difficulties and tell us, in his sharp Scottish tones, how to think clearly about them and reach the correct conclusion.

  My Florida sabbatical wasn’t all wasps and work. Jane, Alan and I were joined by Donna Gillis, a friend of Jane’s from the Zoology Department, and the four of us set off to discover more of Florida. We went on drives to Disney World (Alan insisted on taking all the most hair-raising rides) and Sea World (Alan was the first to volunteer to be pushed into the pool by a performing seal). We went to the university’s Seahorse Key marine biology research station on the Gulf of Mexico coast, where we self-catered and slept in dormitory bunks. We saw Limulus (‘horseshoe crabs’, although they aren’t crabs at all but distant relatives of spiders; Jane later did research on these ‘living fossils’). We saw thousands of ghost crabs (they really are crabs) scuttling into their burrows as we approached, leaving their easily traced footprints. Most memorable was the phosphorescence in the sea, caused by the disturbance of microscopic organisms in the plankton. We played ‘ducks and drakes’, skimming flat stones over the water and watching the glowing ripples as the stones struck the surface. Donna danced on the night beach, her toes in the wet sand inscribing patterns of glowing and then fading phosphorescent blue, charmingly singing of herself in the third person: ‘She’s dancing.’

  On another beach, Alan and I swam naked, which alarmed Jane and Donna because – to Alan’s and my surprise – apparently it was, and is, illegal, even at night. Now that I think about it, an incident many years later gives me reason to believe the illegality really is taken seriously in the United States. The anthropologist Helen Fisher and I were skinny-dipping in Lake Michigan one warm summer night after a hot day of conference speeches at Northwestern University. A police car drew up on the road a hundred yards away. It was dark, so I don’t know how they had spotted us, but they trained a searchlight on us and bellowed through a bull-horn: ‘You are subject to arrest. You are subject to arrest. You are subject to arrest.’ Panic-stricken, without waiting to get dry, Helen and I tugged our clothes on as we ran. No such incident marred Alan’s and my rather swift and brief dip in the moonlit Florida waves. I suspect, with hindsight, that we did it for reasons of bravado more than enjoyment. Jane now tells me she discourages her students from swimming off that particular beach because one often sees sharks there.

  Back at Gainesville, much of my time on this sabbatical was spent writing chapters of The Extended Phenotype, taking advantage of the library, and consulting Alan almost daily on matters of evolutionary theory and how to think straight about it. But I also spent time collaborating with Jane (again relying heavily on advice from Alan) on a new paper called ‘Do digger wasps commit the Concorde fallacy?’

  The Concorde fallacy

  Economists recognize the sunk cost fallacy – throwing good money after bad. Before I had heard of it, I had recognized the same error in the context of evolutionary biology, and named it the ‘Concorde fallacy’, first in a 1976 paper in Nature which I wrote jointly with an Oxford undergraduate pupil, Tamsin Carlisle, and then in The Selfish Gene. Here’s the definition of the Concorde fallacy in the Oxford Dictionary of Psychology, edited by Andrew Colman:

  Continuing to invest in a project merely to justify past investment in it, rather than assessing the current rationality of investing, irrespective of what has gone before. Thus gamblers often throw good money after bad in an attempt to escape from escalating debts . . . and the length of time a female great golden digger wasp Sphex ichneumoneus is prepared to fight over a disputed burrow depends not on how much food there is in the burrow but on how much she herself has put there, the wasp that has carried the largest amount of prey into the burrow being generally least willing to give up fighting. The phenomenon was first identified and named in an article in the journal Nature in 1976 by the British ethologist Richard Dawkins (born 1941) and his undergraduate student Tamsin R. Carlisle (born 1954). Also called the sunk cost fallacy, especially in decision theory and economics . . . [Named after the Anglo-French supersonic airliner, the Concorde, the cost of which rose steeply during its development phase in the 1970s so that it soon became uneconomical, but which British and French governments continued to support to justify past investment.]

  Another name I’ve used is the ‘Our-boys-shall-not-have-died-in-vain fallacy’. During the mounting opposition to the Vietnam War, as I remember from my teargas-dodging days in California in the late sixties, one of the arguments against pulling out went something like this: ‘Lots of Americans have already died in Vietnam. If we withdraw now, they will have died in vain. We can’t let all those boys have died in vain, so we have to go on fighting’ (and a whole lot more of our boys will die, but we don’t mention that). Jane and I were mildly disturbed to find, on re-analysing her data, that her Sphex digger wasps appeared to commit their own version of the Concorde fallacy. Here’s how.

  It doesn’t often happen that an entering wasp meets an incumbent at their shared burrow, but when they do they fight, and the loser usually flees for good, leaving the winner in sole control of however many katydids the two wasps have gathered. Presumably the wasps are fighting to determine who gets the burrow, which is valuable to both of them. The value of the burrow is larger when there are lots of katydids in it, and that value should be the same, you might think, for both wasps, regardless of who caught the katydids. So, if the wasps behave like rational rather than Concordian economists, you would expect both to fight harder for a well-provisioned burrow than for a scanty one.

  Sphex ichneumoneus wasps fighting near the entrance to their shared burrow

  Drawing by Jane Brockmann

  But that is not what happens. Instead, Concorde-fashion, each wasp fights as if the value of the nest is determined by how many katydids she herself has caught, regardless of the true future value of the burrow. This showed itself in two ways. First, there was a statistical tendency for the wasp who had contributed the most katydids to end up winning the fight. Second, the duration of each fight was correlated with the number of katydids that the loser had contributed. Here’s the Concordian rationale for that result. Each fight comes to an end when one decides to flee, and this defines her as the loser. A Concordian wasp will give up early if she has contributed few katydids, late if she has contributed many. Hence the correlation between fight duration and number of katydids that the loser had caught.

  Jane and I fretted, half jokingly given my role in naming it, over the fact that digger wasps appeared to commit the Concorde fallacy. Was it just that
, in John Maynard Smith’s jocular phrase, ‘Natural selection has bungled again’? We sought advice, as ever, from Alan, who pointed out a number of things. Animal design, like design by human engineers, is not perfect in some absolute sense. Good design is always constrained. A suspension bridge is not guaranteed to withstand all conditions: the engineer designs it as cheaply as possible within a specified safety margin. What if, for some reason, the sensory and nervous apparatus that a wasp needs to count katydids in a shared burrow is costly, while the apparatus needed to measure her own effort in catching katydids is cheap? The most economical ‘design’ of wasp might then indeed look Concordian, especially if – as is indeed the case – sharing of burrows is not very common.

  As it happens, there is some indirect evidence that the nervous/sensory apparatus a wasp might use to count prey is indeed costly to run. It comes from a related genus of digger wasp, Ammophila, studied in the Netherlands by Gerard Baerends, incidentally the first graduate student of my old maestro Niko Tinbergen. Unlike Jane’s Sphex ichneumoneus, Baerends’s Ammophila campestris is a progressive provisioner. Instead of gathering a full store of food, laying an egg on it and then sealing the burrow and leaving, Ammophila campestris brings food (caterpillars, not katydids in this case) daily to its growing larvae. Moreover, it has two or three burrows on the go at any one time. The ages of its larvae are staggered, and their food needs are correspondingly different. The wasp ‘knows’ that younger larvae need less food than older larvae, and she feeds them unequal quantities accordingly. But now, here’s the surprising fact. She assesses the needs of each of her larvae only during a single early morning inspection round of all her burrows. After the inspection round is over she behaves for the rest of the day as if she is completely blind to the contents of the burrows.

  Baerends showed this in a neat experiment. He systematically switched larvae around from burrow to burrow. No matter how small the larva now in a particular burrow, the wasp would continue to feed it the large prey appropriate to the larger larva who had occupied the burrow during the morning inspection round. And vice versa. It is as though the wasp has an instrument for measuring the contents of a burrow, but this instrument is costly to run so she switches it on only once a day, during the morning inspection round; the rest of the day the apparatus is switched off to conserve running expenses. This would explain the blunders made by the wasps in Baerends’s experiments, blunders compatible with the view that they were now blind to burrow contents. Of course, such ‘blunders’ would not normally matter because, in the absence of a Baerends, larvae don’t hop from burrow to burrow.

  Ammophila campestris really needs its assessment instrument because its normal habit is progressive provisioning of multiple burrows housing larvae of different ages; but even so, it severely limits the time it is switched on. Sphex ichneumoneus, working on only one larva at a time, and sharing burrows rather rarely, doesn’t have such a strong need for this costly apparatus, so it never switches it on, or doesn’t possess it at all. It therefore appears to commit the Concorde fallacy. This, at any rate, was how we rationalized our results. And really, we shouldn’t have allowed ourselves to be disappointed by the wasps’ ‘performance’ – especially if they are as ‘sphexish’ as some philosophers believe. Intelligent humans in positions of power commit the Concorde fallacy, and humans make decisions a lot more silly than that when assessing risks, costs and benefits, as psychologists such as Daniel Kahneman have shown us.

  THE DELEGATE’S TALE

  DAVID Lodge, in his campus novel Small World, compares the academic conference to a Chaucerian pilgrimage:

  The modern conference resembles the pilgrimage of medieval Christendom in that it allows the participants to indulge themselves in all the pleasures and diversions of travel while appearing to be austerely bent on self-improvement.

  I warm to the analogy, having used Chaucer for a different purpose in The Ancestor’s Tale. Here I choose six conferences out of hundreds, as representative stations along my scientific pilgrim’s way.

  Lodge’s cynical view was not contradicted by a memorable conference during the time I was writing The Selfish Gene, sponsored by the Boehringer pharmaceutical company in a splendidly over-the-top German castle. The subject was ‘The creative process in science and medicine’, and it was certainly the poshest conference I have ever attended. The main guest list comprised scientists and philosophers of immense distinction, many of them Nobel Prize-winners. Each of these illustrious figures was allowed to bring along a couple of junior colleagues – squires to their knights, as it were. My old maestro Niko Tinbergen was one of the ‘knights’, and he took Desmond Morris and me as his squires. Other knights (some of them knights in the literal sense) were Sir Peter Medawar (immunologist, essayist and legendary polymath), his ‘philosophical guru’ Sir Karl Popper, Sir Hans Krebs (the world’s most famous biochemist), the great French molecular biologist Jacques Monod and several other household names in science, each with their juniors in attendance. There were only about thirty of us in all. I felt immensely fortunate to be present, and hardly dared say a word.

  We sat around a large polished table (I don’t think it was actually round, which is a pity for my ‘knights’ conceit), with our names proud in front of us (incidentally, why, at such events, does the name so often face its owner, who presumably knows who he is, rather than out into the room towards those who might make use of it?). Strewn around the table were notepads and pencils, bottles of mineral water, sweets (ugh) and cigarettes galore. These last were more than usually unfortunate because Karl Popper had a famous distaste for cigarette smoke. On one occasion at a different conference he had risen from the floor to make a special request that nobody should be allowed to smoke. Nowadays such a plea would not be necessary; it would go without saying. But those days were different, and it was a symptom of the regard in which the great philosopher was held that the chairman acceded to his request. Or almost. What he said was: ‘In deference to Sir Karl and out of respect for him, please would any delegate who wishes to smoke leave the hall and smoke outside.’ Sir Karl rose again: ‘No, zat would not be good enough. When zey come back in, I could smell it on zeir bress.’

  So you can imagine the consternation raised by the tobacco largesse scattered over the conference table in our opulent Schloss. Every time the hand of a smoker strayed tablewards, a flunkey would come bustling over to clutch a sleeve and whisper, ‘No please, not to smoke, Sir Karl cannot stand it . . . bitte schön.’ But, as far as I remember, the cigarettes remained on the table in full view, to tempt the unfortunate addicts for the duration.

  The conference was structured loosely around a series of invited presentations, followed by questions from around the table and extended discussion. Every morning at breakfast, with German thoroughness, each of us was handed a massive pile of paper, a complete word-for-word transcript of everything that had been uttered the previous day: every last um and er and infelicitous restarting or recycling of a sentence. I pitied the red-eyed night shift of typists slaving through the small hours to produce this torrent of verbiage. But there was a problem: how to tie each pearl to its oyster – who had said what, in other words. The chairman of each session was briefed to make us preface each intervention with our name. Peter Medawar, who chaired the opening session, also asked the first question and identified himself to the tape recorder with characteristic aplomb: ‘This is Medawar, shamelessly abusing the privileges of the chair.’ But most people, in the cut and thrust of discussion, forgot to say their names, so an alternative solution was deemed necessary. It proved to be even more distracting than the cigarettes. On a rotating stool, perched high on top of the massive, polished table, sat a young woman in a short skirt. Every time one of the delegates started to speak she would swivel, like a gun turret on a battleship, to locate him and write down in a notebook his name and first sentence. These notes were then used by the night typists to attribute each laboriously rendered paragraph.

  It was fa
scinating for a young scientist to eavesdrop as the giants of his profession unmasked their creative processes. Hans Krebs’s recipe for how to win a Nobel Prize was too modest to be credible: it amounted to ‘Go into the lab every day at 9 a.m., work all day until 5 p.m., then go home; and repeat the process for forty years.’ I’ve already quoted Jacques Monod’s engaging revelation that he was in the habit of imagining himself an electron deciding what to do next. I did something similar when I asked myself, following my scientific hero Bill Hamilton, what I would do if I were a gene, trying to get copies of myself passed on to future generations.

  At the very end of the conference, one of the invited guests, a Japanese physicist who had uttered not a word throughout, timidly asked if he might finally say something. He explained that he would lose face if he went back to Japan and confessed that he hadn’t spoken. It would have been technically enough if he had stopped there, but he went on to say something rather interesting. Most physicists, he pointed out, are obsessed with symmetries of various kinds. Japanese aesthetics, on the other hand, favour asymmetry, and perhaps this gives Japanese physics a different perspective. I immediately thought of Pamela Asquith, a young Canadian anthropologist friend doing a study in what might be called meta-primatology – the comparative study of primatologists. Her thesis was that Japanese primatologists brought a different cultural perspective to their monkeys, to complement the western angle. A comparable point has been argued for female primatologists, of whom there are numbers disproportionate to other sciences.