The 'Haldane Principle' and other invented traditions in science policy
- Executive summary
- Alternative stories
- The Haldane Principle
- What is the 'Linear Model'?
- Policy implications
- Further reading
- About the author
- Surprising as it may seem particular accounts of history are among the most important evidence bases for government science policymaking and funding.
- Many historical accounts are misleading about the past - for example that Britain has been good at inventing but bad at developing; that policy has been derived from the Haldane Principle (of 1918) and the Linear Model of Innovation (of 1945).
- Both the Haldane Principle and the Linear Model were invented; the former in 1963 (by Lord Hailsham) and the latter in the 1980s (by academics) in both cases as accounts of the past rather than prescriptions for the present.
- It is important for historians and policy-makers to recognise the way in which past policy disputes shape history, and the ways in which these histories shape policy.
- Policy-makers need to pay great attention to what 'science policy' is actually for, recognising that it is usually policy for something quite limited - government funded research largely performed in universities.
- There are good reasons for continuing to support university research in the recession, but the false expectation that it will lead us out of it is not one of them.
In the case of proposing and arguing about policies for research, histories have long been central. John Desmond Bernal's The Social Function of Science (1939) was a pioneering case. It argued that there was a collapse of innovation, and especially the use of innovations, under capitalism. It called for socialism as well as the planning of science. The Two Cultures (1959) by C. P. Snow gave a historical account of supposed failures in British culture, and its education system, to promote scientific and technological education. We might also note Hilary and Steven Rose's even more historical Science and Society of 1969, and the works of J.G. Crowther, all supporting a more interventionist policy for research. The Thatcher period saw a major political shift, represented by Terence Kealey's Economic Laws of Scientific Research (1995), which despite its title, was mainly a historical book.
Professional historians have long been concerned with policy. 1964 saw the publication of two important official histories of science in the Second World War, by M. M. Postan and others, and by Margaret Gowing. Official histories were meant to inform policy in the future, and the civil histories of the Second World War were, according to Eric Hobsbawm, pioneering works of contemporary history. Neither Postan nor Gowing were then historians of science. In the 1970s historians of science, led by Roy MacLeod, opened up the study of state support of science in early twentieth-century Britain. Again there was a connection to policy - much of the work was done alongside the Science Policy Research Unit (SPRU) at the University of Sussex. SPRU's most important book, published in 1980, Technical Innovation and British Economic Performance, had at least as much evidence from the 1870s as the 1970s. In Manchester too, history and policy marched together, as can be seen in Philip Gummett's largely historical survey Scientists in Whitehall of 1980.
In the 1980s particular historical accounts of British science and technology (often focussing on the period 1870 - 1914), were central to the debate about research, universities and education generally. The Thatcher government was influenced by this work, but it was also central to the argument of its opponents. There was agreement that British culture had been hostile to science, especially applied science, with dire consequences for the economy. The disagreement was in what to do about it. The most influential histories were profoundly negative accounts of British science, a story of failure and decline, made worse it seemed by an impressive number of Nobel Prizes or supersonic aircraft.
This was history with systematic and massive forgetting built in. For example, if one had called in the late 1980s or early 1990s for
- the concentration of most research and development (R&D) funding in one ministry,
- the appointment of the President of the Royal Society as scientific adviser to that ministry,
- the raising of the R&D:GDP ratio to nearly 3%,
- British business to spend more on R&D than that of any country other than the USA,
- and, linking government research to realistic industrial priorities,
you would have been regarded as ludicrously ambitious, given the deeply entrenched cultural hostility this policy would provoke. But, had you been reading the emergent post-declinist history of Britain you would have recognised this as a call for a return to the Britain of the late 1960s.
We have got away from stories of past failure, and we have begun to chart a new account of twentieth century British science, in a refashioned historical context in which it is possible to understand the centrality of H.G. Wells as well as C.P. Snow. It is not a history of lacks, but of huge investments in science and technology, a story of scientist ministers and members of parliament, of businesses run by scientists. It is a complex historical story, one we can only tell when we shed naive assumptions about say, the relations of national investment in research and national economic growth.
Indeed I want to take examples from cases well known through historical research by the 1960s: radar, and the British wartime atomic bomb project.
In 1935 there was no radar in Britain; by 1945 it was a vast enterprise. It is an exaggeration to say that radar saved Britain in 1940 but it was a hugely important warfighting instrument. It was used from the first not just for detection of aircraft but to aim guns (on land and at sea), and was soon used for navigation, and targeting. The rate of development was extraordinary. While nearly all night bombers got through in 1940; by 1944 radar-controlled guns firing shells with proximity fuses would have made a second Blitz of the 1940 sort impossible.
It is not sufficiently appreciated that Britain had a very large atomic bomb project at the beginning of the war, probably the largest of any country. Experiments started in the summer of 1939. Unlike in the radar case the work was largely done in universities and industry, with a large contribution by émigrés who could not be directly employed by government. Through the defeat in France, and through the Blitz, this long-term project accelerated. The expectation was that atomic bombs would be cheaper than TNT ones. The British project was to go to North America, with Britain becoming a very junior and often abused partner, in the Manhattan project.
Familiar though these stories are I would like to emphasise some aspects. First neither radar nor the bomb were British inventions - nor did Britain alone work on them. So did nearly every major power. The idea of radar was widely known before 1935. The great and sudden interest in the possibility of the bomb came from experimental work in Continental Europe. In neither case were British academics critical: they had to be inducted into radar; in the case of the bomb it is notable how few of the major British atomic physicists were involved at the beginning. Yet academics were to become essential to development, and as we know both projects needed and used people from many different scientific backgrounds. Universities produced people and knowledge, not inventions.
The second point is the importance of the Oxford-chemistry-industry nexus. Sir Henry Tizard, an Oxford chemist, specialist in aviation fuels, was involved in both as chairman of key committees. The Oxford physicist Frederick Lindemann was a key promoter of the bomb, a project put under the ministerial authority of Sir John Anderson, an Edinburgh chemist. Anderson, who had been a director of ICI, appointed (Sir) Wallace Akers of ICI as head of the bomb project. Akers graduated in chemistry from Christ Church, Oxford, was chairman of ICI's Billingham division until 1937. He then directed ICI's massive rearmament effort, centred on poison gas, explosives and propellants, and later aviation spirit. ICI Billingham was an important place: here was the centre of synthetic ammonia and synthetic petrol production, based on German processes; it recruited the cream of Oxford chemistry and Cambridge engineering, such as Frank Kearton; a young Oxford chemist who worked on aviation spirit and gaseous diffusion of uranium hexafluoride.
Radar too depended on industry, and the British electrical and electronic industry had benefited hugely from Britain having the first large scale television broadcasting, also based on imported techniques. And we might conclude that both stories tell us that Britain was good at developing but not necessarily the best at inventing.
Older stories overemphasised connections between Cambridge, the academy, physics, and the left, and underemphasised those of Oxford, industry, chemistry, and the right. An important reason for this significant disparity in treatment was the genius, fame and public presence of John Desmond Bernal. His Social Function of Science (1939) radically downplayed the significance of industrial research. But just as significant was the work of a Cambridge chemist, C.P. Snow, who merely damned the most extraordinary science-politics partnership of twentieth-century British history - that between the historian Winston Churchill and the physicist Frederick Lindemann (later Lord Cherwell).
The Haldane Principle
The great bulk of British research, just like the radar and the bomb projects, was done outside the research council system. Yet the research councils were long at the centre of accounts of government and research.
The first was the Department of Scientific and Industrial Research (DSIR) established in 1915 with an advisory council under the Privy Council and the Lord President. The idea was extended by the recommendation of Lord Haldane in 1918 that the Medical Research Council be re-established on similar lines. R.B. Haldane was a great Liberal (and later Labour) politician, lawyer and intellectual, who was twice Lord Chancellor.
Much has been written on the research councils, mostly by themselves, an important reason for their overrepresentation in accounts of government and research. In 1962 the Secretary of the DSIR, Sir Harry Melville (another chemist), published a very full, though hardly scintillating account of his department. From the point of view of policy debate today there is something interesting about it: there is no mention of something called the 'Haldane Principle' which is supposed to have governed the research councils (and sometimes all British government funding of R&D) since 1918. Here history continues to matter.
Sir Harry Melville was not at fault for omitting it: in 1962 this 'Haldane Principle' did not exist. Searches through JSTOR, Nature and Hansard confirm this. In fact they show there were two different and unrelated Haldane Principles already in the literature. One was the Haldane Principle of functional organisation of government departments (Education rather than Children, Health rather than Old People) which was suggested in his 1918 Machinery of Government Report. The other Principle was due to his nephew, J.B.S. Haldane and was purely scientific.
Hansard contains a trace of the invention of the science policy 'Haldane Principle'. Austen Albu, an Imperial College engineer, industrialist and Labour MP, noted in February 1964 that the Conservative government was attacking the Labour opposition's plans for science, on the grounds that Labour was seeking to control the work of scientists. He said: 'Allied to this political propaganda is the discovery of a new scientific principle-the Haldane principle; the principle that all scientific research in support of Government policy must be conducted in independent research establishments.' That is not much like the principle in play today, but then as now the meaning was far from stable. Later in the year, with the Labour party in office, the Conservative Lord Hailsham (Quintin Hogg, who had been in charge of the research councils from 1957 to 1964) attacked Labour's introduction of a Ministry of Technology, claiming the 'Haldane Principle' as his. Of the Ministry he said:
This is a totally new departure from recent practice and in my opinion at least, is a most retrograde step. Ever since 1915 it has been considered axiomatic that responsibility for industrial research and development is better exercised in conjunction with research in the medical, agricultural and other fields on what I have called the Haldane principle through an independent council of industrialists, scientists and other eminent persons and not directly by a Government Department itself.
Hailsham's Principle is different from Albu's. Hailsham's Haldane principle is that of independence not of science or research establishments, but of research councils of 'industrialists, scientists and other eminent persons'. And he was right to say that large chunks of this arrangement were torn up in 1964, and not merely in relation to industrial research.
Hailsham was, like Haldane, a lawyer; he would later be Lord Chancellor. Hailsham had been Minister for Science, not of science, as he insisted. He wrote an excellent book called Science and Politics (1963), in which he lauded the research council system, defended academic and scientific freedom, and expressed concern that governments were corrupting science. Hailsham was clearly much influenced by the Freedom in Science lobby - a proto-Thatcherite, CIA-funded organisation more focussed on Oxford than Cambridge, among whose luminaries were the Manchester chemist Michael Polanyi and the Oxford biologist J.R. Baker.
For some, like Hailsham, and indeed another great panjandrum of research policy, the chemist Lord Todd, 'The Haldane Principle' had the dignity of antiquity. For others its antiqueness was a measure of its lack of relevance. Many people referring to the 'Haldane principle' did so to say - how could a policy of 1918 be useful in the 1960s! Indeed the Haldane Principle was invented at the moment it was seriously challenged, and would reappear in discussion at just such moments of contention: the Rothschild reforms to some research councils in 1972; in the 1980s, and again more recently. It has been a hot topic for the parliamentary select committee on innovation, universities and skills. The current government says it believes in it. John Denham, then Innovation, Universities and Skills Minister, said that the 'fundamental elements of Haldane' remain valid: that researchers are best placed to determine detailed priorities; that the Government's role is to set the over-arching strategy; and that the research councils are 'guardians of the independence of science' (Speech to Royal Academy of Engineering, 19 February 2009). Whether these principles actually are those of 'Haldane' is a non-question; whether they are good principles is certainly a question. In my view the first would be acceptable to Madame Ceaucescu; the second is highly debatable in relation to university research; the third, if the genuine view of government, would imply there are no independent universities or learned societies or academies- government will decide who the guardians are, and appoint them! Again, 'Haldane' is for some a great truth from 1918, for others an antiquated albatross around the neck of British research policy. It is of course neither of these things.
There was no Magna Carta, no Bill of Rights, no Human Rights Act, no constitution for government funded research; in 1918 or ever. There were lots of different arrangements and for different aspects. It was the case however, that in the realm of education, and non-departmental research, there was a powerful sense of the limits of appropriate government intervention. In the interwar years schools were partially funded by central government, but not run by them; the same strongly applied to universities, and research councils owed something to this model. Indeed it was often regarded as a great British strength that its school and university teachers were not fonctionnaires. Central funding without central control was seen as a characteristic of a mature and free society. But that did not mean scientists were in charge.
What is the 'Linear Model'?
One of the many oddities of discussion of 'Haldane' is that it does not involve the Second World War. Yet this is eloquent testimony to the fact that research councils were not very significant in the war effort: their spending did not go up very much either during or after the war. University research in nearly all subjects fell severely during the war. Universities became militarised teaching institutions. Students in approved subjects rushed through courses and failure meant call-up. Men in non-essential subjects didn't get to University or didn't finish their degrees; they were called up.
In a view which was rarely expressed, Sir Henry Tizard, of all people, claimed of the Second World War: 'It is a mistake to suppose that science advances rapidly in a war. Certain branches of science may receive a special stimulus, but on the whole the advance of knowledge is slowed.' Bernal and many others had a very different view. Indeed the Second World War is generally regarded as having been hugely important and good for British scientific research.
After the war, the main concern was to increase productivity and exports. Another was to develop the empire, particularly the African colonies, as sources of raw materials to be bought in sterling. A new and very large Colonial Research Council was to be created, but outside the usual structure - it came under the Colonial Office and was soon spending more than the MRC or the Agricultural Research Council, as Bernal was one of the few to notice.
One might assume, from the stories of the transformation of the place of science due to the war, that domestic research would be at the top of the agenda, and so it was in some respects, the atomic bomb being one. Yet in 1948 the man now in effect chief scientific adviser, Sir Henry Tizard, took a different view. He noted that Switzerland and Sweden were doing well, due to focus on technology (a term whose meaning was expanding) rather than scientific research:
I quote them in support of my view that if it is not the general expansion of research in this country that is of the first importance for the restoration of its industrial health, and certainly not the expansion of government research remote from the everyday problems of industry. What is of first importance is to apply what is already known.
It is in fact rather odd for a former rector of Imperial College, London to lament the lack of technological education. Indeed odd for someone who worked so closely with Cambridge engineers. It is an illustration of how difficult sources are to interpret.
The research and development effort of the late 1940s was massively greater than that of 1939, but the great expansion was in military R&D budgets and big civil projects, outside the research council sphere. In the 1950s the expansion in university research and graduate output was on the back of a much larger and hugely expanding effort by industry, and by government departments. As in the war, academic science fed and followed state and industrial projects. By the late 1960s the whole massive R&D effort was being reassessed.
Solly Zuckermann, chief scientific advisor to the government from 1964-71, had an inkling of something well known to economists of technical change of the 1960s, but since often forgotten: that there was no positive correlation between national R&D and national rates of growth. Seeing its significance was a vital breakthrough for historical studies of British science, and for historical reflection on policies for research.
I also cite Tizard and Zuckermann to cast doubt on the view that after the war, policy was dominated by something called 'the linear model of innovation'. This is even more difficult to pin down than Haldane. Following one academic expert, I take it to be the idea that: 'basic' or 'fundamental,' 'pure' or 'undirected,' scientific research is the main source of technical innovation; that the process of innovation is a sequential one; and that the innovation produced is the main source of economic growth. But this view was not held by any serious analyst. Anyone engaged knew that the big money went to big projects to develop existing techniques, driven by military and industrial demands, and that invention had many sources. Tizard and Zuckermann were operating on a different level.
The 'linear model of innovation' is, like 'Haldane', a term of art in science policy. It is also an invention, in this case an academic rather than a political one. It was invented in the 1980s, though there are a few earlier references. The 'linear model' is also often associated with a foundational document, in this case Vannevar Bush's, Science, the Endless Frontier of 1945. But it is not there, just as the Haldane Principle in not in the Haldane report of 1918. 'Haldane' always had supporters as well as detractors, the linear model by contrast has been universally regarded as a very bad idea by both historians and science policy analysts.
Accounts stressing the linear model serve a purpose: they say that policies for research in the past were wrong; we now need a radically new set. In fact the policies of the past were uncannily like the post-linear model policies proposed. But there is deeper irony: many accounts of the past themselves reflected this invented linear model, giving undue weight to academic research funded by government, according to the Haldane principle. Our historical models have shaped our understanding of history, rather than the other way around.
What then might be the implications for thinking about current policy? First, don't think with Haldane principles and linear models, or two cultures. Think about them and their significance in debates, and how they support each other. Note their rootedness in seemingly solid historical accounts. Note how they limit discussion to particular parts of a much more complex whole.
But that is just the beginning: forget about science and about science policy. For example, is 'science policy' policy for the 'pursuit of knowledge and understanding of the natural and social world'? That is a recent definition of 'science' by the 'Science Council'. It is in fact a definition of 'research'. In fact it is not even a policy for research, it covers a lot less and a lot more. Indeed there has never been even a sensible category of what Lord Zuckermann talked of as 'so-called science policy'.
There is no stable entity called science which is governed, worried about, applauded, condemned, invested in, much less understood. Each small part of 'science' is governed by many institutions, and many policies. Most science is not research. Most research is not academic. Most research is not government-funded. Such research is funded by many bodies. Most research is not nationally organised; its effects are not national, even if nationally organised. The more we think about terms (and I won't go into 'technology') the more fallacies of composition and division mount up. No wonder that linear models and Haldane principles are to the extent they apply at all, hugely misleading.
What then of the economic role of research-council research? In recent years, and especially in the recession it has got increasing emphasis. Writing in the Guardian on 17 May 2009, Lord Drayson, Minister for Science, wrote that 'focusing effort within science is nothing new, whether during war or peacetime, ... there's no U-turn under way in British science; no retreat from excellence; no undermining a diverse base capable of interdisciplinary research; no challenging the independence of scientists, who remain governed by research councils and peer review,' encapsulating the idea that something called science is governed by research councils. He then goes on to make an extraordinary claim: 'What has changed is our sense of urgency to use science as the primary means to achieve a healthy and prosperous Britain' [emphasis added]. As we have seen, the idea that that small fraction of British research funded by the research councils could be the primary means to regenerate Britain is a naive absurdity, one which was not shared by earlier British governments, which were much more committed than New Labour to the urgent modernisation of Britain through science and technology. Those earlier governments were committed to an active national industrial policy, something which New Labour rejects, something without which a nationalist research policy is not merely naive but ridiculous.
This paper is based on the Wilkins-Bernal-Medawar Prize Lecture, given on 20 April 2009 at the Royal Society.
Thanks to Leroy Attz, Chris Bowlby and Geoffrey Lewis.
- David Edgerton, Science, Technology and the British Industrial 'Decline' ca. 1870-1970 (Cambridge: CUP/Economic History Society, 1996) (Contains extensive annotated bibliography)
- David Edgerton and K.S. Hughes, 'The Poverty of Science: A Critical Analysis of Scientific and Industrial Policy Under Mrs Thatcher', Public Administration 67 (Winter 1989), pp. 419-33.
David Edgerton, '"The linear model" did not exist: reflections on the history and historiography of science and research in industry in the twentieth century', in Karl Grandin and Nina Wormbs (eds), The Science-Industry Nexus: History, Policy, Implications (New York: Watson, 2005), pp. 31-57
About the author
David Edgerton is the Hans Rausing Professor in the Centre for the History of Science, Technology and Medicine at Imperial College London. He is the author of Shock of the Old: technology and global history since 1900, Profile Books, 2006. email@example.com
If you have found this paper interesting why not join our email list to receive occasional announcements when new ones are released?