On being the right size

J.B.S. Haldane taught biology that scale is destiny: a mouse built to the plan of an elephant collapses under its own weight. The UK research system has spent forty years being scaled, concentrated and reorganised – and almost nobody has asked the biologist’s question of it. Not how big. What shape.

In 1926, J.B.S. Haldane published four pages that biology never quite got over. “On Being the Right Size” makes a single observation and follows it without mercy: for every animal there is a most convenient size, and a change in size demands a change in form.¹ The mouse scaled up to elephant proportions does not become an elephant; it breaks its legs. The insect scaled to the size of a dog does not menace the neighbourhood; it suffocates, because the tubes that breathe for it stop working beyond a few millimetres. Nothing in biology is size-neutral, which is why nature, given a new size, always draws a new animal.

British research policy has never had its Haldane essay, which is an irony, because it has had a Haldane principle for a century. Since the first Research Selectivity Exercise in 1986, through every RAE and REF that followed, the working assumption has been that the system’s problems are problems of quantity – more funding or less, more concentration or (occasionally, faintly) less. Quality-related funding flows where assessed strength already is; assessed strength accumulates where QR flows. The result is a system that has been made steadily bigger in some places and steadily thinner in others, by an instrument that has never once been asked what form the new size demands.²

“We have run a forty-year experiment in concentration without a control group”

The concentration argument is a scale argument in disguise, and it is remarkable how rarely it is examined as one. Critical mass is invoked as though it were a law rather than a metaphor. The honest literature on returns to scale in research units is old, mixed and mostly unread: beyond a moderate size, per-researcher productivity in many fields is flat, and in some it declines – the meetings grow faster than the papers.³ We have run a forty-year experiment in concentration without a control group, and the closest thing we have to a result is that the places where the money pooled are excellent, which is what you would expect whether the policy worked or not.

Haldane’s animals fail at their surfaces. The mouse’s problem is heat: too much skin for its volume, so it must eat constantly or die. The elephant’s problem is the opposite. Scale changes the ratio of inside to outside, and it is at the outside – the surfaces, the interfaces – that the trouble shows up first. A research system has surfaces too: where discovery touches training, where universities touch firms and hospitals and government departments, where knowledge is supposed to leave the building. A system optimised for internal mass and starved at its surfaces has entirely predictable pathologies, and we can recite ours from memory: absorption problems, diffusion problems, a productivity puzzle that sits precisely at the interface between what the science base knows and what the economy does with it.

Ask the functional question – what is each part of the system asked to carry? – and the shape of a better answer appears. Discovery science plausibly does need depth, and depth may need pooling. Doctoral training needs community above a minimum – a cohort, not a corridor – which argues for floors rather than peaks. And absorption, the movement of knowledge into the bloodstream of places and industries, needs the opposite of concentration: it needs presence, nearness, surface area. Some functions scale like bone. Others scale like skin. Funding them all through one instrument, tuned to one virtue, is how you build the mouse on the elephant’s plan.

None of this requires a revolution, which is fortunate, because the system could not survive another one. It requires the next spending review to ask a different first question. Not the sector’s question – how much? Not the Treasury’s – how little? The biologist’s: what form does this size demand? A research system, like a mouse or an elephant, is not right or wrong in itself. It is the right size for what it has been asked to carry – or it is not. ●

Spotted a politician meddling in the research system? Or a blurry line between academia and government? Email the man himself: haldane [at] researchagenda [dot] news