Friday, 20 July 2007

'Pure' medical science

The future of ‘pure’ medical science: The need for a new specialist professional research system

Bruce G. Charlton and Peter Andras

Medical Hypotheses 2005; 65: 419-425


Over recent decades, medical research has become mostly an ‘applied’ science which implicitly aims at steady progress by an accumulation of small improvements, each increment having a high probability of validity. Applied medical science is, therefore, a social system of communications for generating pre-publication peer-reviewed knowledge that is ready for implementation. However, the need for predictability makes modern medical science risk-averse and this is leading to a decline in major therapeutic breakthroughs where new treatments for new diseases are required. There is need for the evolution of a specialized professional research system of pure medial science, whose role would be to generate and critically evaluate radically novel and potentially important theories, techniques, therapies and technologies.


Pure science ideas typically have a lower probability of being valid, but the possibility of much greater benefit if they turn out to be true. The domination of medical research by applied criteria means that even good ideas from pure medical science are typically ignored or summarily rejected as being too speculative. Of course, radical and potentially important ideas may currently be published, but at present there is no formal mechanism by which pure science publications may be received, critiqued, evaluated and extended to become suitable for ‘application’.

Pure medical science needs to evolve to constitute a typical specialized scientific system of formal communications among a professional community. The members of this putative profession would interact via close research groupings, journals, meetings, electronic and web communications – like any other science. Pure medical science units might arise as elite grouping linked to existing world-class applied medical research institutions.

However, the pure medical science system would have its own separate aims, procedures for scientific evaluation, institutional organization, funding and support arrangements; and a separate higher-professional career path with distinctive selection criteria. For instance, future leaders of pure medical science institutions would need to be selected on the basis of their specialized cognitive aptitudes and their record of having generated science-transforming ideas, as well as their research management skills.

Pure medical science would work most effectively and efficiently if practiced in many independent and competing institutions in several different countries. The main ‘market’ for pure medical science would be the applied medical scientists, who need radical strategies to solve problems which are not yielding to established methods. The stimulus to create such elite pure medical science institutions might come from the leadership of academic ‘entrepreneurs’ (for instance, imaginative patrons in the major funding foundations), or be triggered by a widespread public recognition of the probable exhaustion of existing applied medical science approaches to solving major therapeutic challenges.

Over recent decades, most of medical research has become an ‘applied’ science [1]. Now, it is likely that the great bulk of medical funding is spent on projects which are designed reliably to generate incremental technological advances by modest extrapolation from already-existing theories and techniques. Consequently, while there has been a steady improvement in most aspects of clinical practice, there has been a decline in the frequency of major ‘breakthroughs’ [2] and [3].

Modern medical science, therefore, aims at steady progress through an accumulation of small benefits. Applied medical science is subject to a type of pre-publication peer review which aims at ensuring that the published research is very likely to be valid and ready for application by other applied scientists [1]. In a nutshell, modern medical research operates much like traditional industrial research and development (R&D) – more akin to the ‘trial and error’ process of technological evolution than to the bold, science-transformative ‘conjectures and refutations’ of classical or ‘pure’ Popperian science [4].

But radically new theories are usually wrong and totally new technologies usually do not work (at least in the short-term). There is, therefore, a strong tendency for applied medical scientists to become risk-averse, and to seek progress by accumulation of the most predictably valid advances on current knowledge and practice. This unambitious, safety-first tendency has been both disguised and exacerbated by the greater use of hype and spin in communicating medical science, in order to exaggerate the scientific significance of even very modest increments of knowledge. To the outsider, it may be impossible to know for sure to what extent a scientific specialist is ‘bullshitting’ in this way. The result is an ‘inflation’ of communications, in which more and more claims of importance mean less and less [5]. Nonetheless, behind this ever denser smokescreen of more effective public relations, medical science seems to be achieving fewer qualitative ‘breakthroughs’ [3].

The need for professionalization of pure medical science

Medical Hypotheses tries to counterbalance this risk-averse trend by providing a forum for the dissemination of radical and speculative ideas [6]. Typically, these ‘pure medical science’ ideas have a lower probability of being valid, but a possibility of much greater benefit if or when they turn out to be true. In other words, papers in Medical Hypotheses conform to the model of Popperian science in that they are not intended to be validated at the point of publication, rather the ideas are published in order to be validated by future peer usage which checks their coherence and implications [1].

But the policy of a single journal does not have much influence on the social system of science and this matter of validation of pure medical science is precisely where current medical science is deficient. Published pure science papers may be of potentially great significance; but because their arguments (almost by definition) make more radical assumptions than is acceptable in the risk-averse world of applied science, these papers are typically (and rightly) ignored by those working in the more cautious but more probably true world of applied medical science.

In other words, at present pure medical science is usually evaluated by the leading applied medical scientists, who judge it according to their own criteria of short-term, predictable validity – and naturally find that most pure medical science communications are unlikely to be ready for immediate application. But it is inappropriate for pure medical science to be evaluated by researchers who have a very different type of competence and a distinct set of priorities. What should be happening is that communications in pure medical science should primarily be evaluated by expert specialist pure medical scientists and their teams, according to the pure medical science criteria such as their potential to make a transformative difference to the practice of existing science.

Although there are numerous gifted individuals working in pure medical science, there is no professional community of pure medical scientists who might perform such evaluations, no specialized system of communications (journals, meetings, etc.) and no distinct source and mechanism of funding pure medical science. Consequently, many of the best ideas published in Medical Hypotheses and elsewhere fail to be subjected to the rigorous scientific evaluation which such a research system can provide. This means that these ideas are not fully checked and tested, nor developed to the point where they become potentially useable by applied medical scientists. All too often even brilliant, interesting and potentially important theories fail to make any discernable difference to the functioning of medical science.

What is required is the development of a specialized, professional system of pure medical science within which radical theories – and also radical new techniques, technologies and therapies – can be proposed, critiqued and evaluated.

Scientific specialization as a mechanism of progress

One of the most powerful and sustained trends in human history is the differentiation and specialization of social function [7] and [8]. Science has progressed firstly by the division of labour inherent in having professional scientists, then by improvements in the systems of communication between networks of scientists (with the evolution of first letters, journals and scientific meetings; then later telephone, fax, e-mail, internet, etc.). Accompanying this has been the division of science into ever more specific disciplines, within which communications are especially concentrated [9].

So far, medical science has specialized almost exclusively in terms of subject matter – with sub-systems composed of specialists in body systems such as brains, lungs or hormones; or pathological processes such as cancer, genetic mutations or immunity. We are proposing that this specialization should be taken further to spawn new sub-sub-specialties of pure science.

For example, medical research in oncology might continue to include a majority of applied scientists working on incremental improvements to existing therapies and technologies. But there would also be a smaller group of pure medical oncologists whose role would be to develop, evaluate and refine ‘radical’ new ideas. Such pure medical scientists would publish their ideas in original papers and reviews, meet in conferences, conduct experiments etc., – just as the applied scientists do – but with the important difference that the pure medical scientist would be evaluated on their ability to produce qualitatively novel approaches to understand and treat cancer and would not be constrained by the necessity to provide knowledge ‘ready for implementation.’

If traditional scientific specialities are considered vertical separations, between subject based disciplines, then what is being proposed is a horizontal division into a small elite system of pure science above a much larger system of applied science. In the cancer sciences, the great bulk of the specialty would run on the applied science model using relatively minor variations on existing ideas. But within each cancer science sub-specialty, there should also be a separate and smaller sub-specialty of ‘pure’ cancer science, situated ‘above’ the bulk of cancer scientists, which is concerned with proposing and evaluating qualitatively different ideas.

The applied scientist could turn to this pure system for suggestions when they were ‘stuck’ and needed novel theories or techniques to get them out of a scientific dead-end, or past an obstacle which is not yielding to the more usual strategy of minor-variation, trial and selection.

Pure medical science as a profession

The assumption is that – since it is required for improved functioning – a specialized social system of pure medical science will ‘spontaneously’ emerge at some point in the future – indeed it is probable that this is already happening in an inexplicit fashion. But what might be the professional characteristics of such a system?

Its general characteristics can be inferred from an examination of the typical career pathways of successful scientists of the past – and these have been subjected to close scrutiny by the late Frederic Holmes in a recently published book [10]. He characterizes the specialist education of scientists (and other experts) in terms of apprenticeship, achievement of independence, mastery of a domain – which takes around 10 years – then the achievement of distinction, followed by maturity and ageing.

The undergraduate education is increasingly generic and unspecialized in modern societies [8], so that although individuals interested in a scientific career will probably study mathematics and some natural science at school and during their first degree at college; specifically scientific training begins with the Masters and Ph.D. degrees. Probably, the career paths of pure scientists would diverge from applied science shortly after the Ph.D., by the introduction of a new type of pure medical science postdoctoral training involving apprenticeship to a series of established experts (preferably in different institutional settings in order to facilitate the apprentice’s achievement of independence from the ‘master’ [10]).

The notion of apprenticeship requires amplification. The envisaged specialty of pure medical science is a minority elite activity within any specific branch of science, and an activity which implies the need for distinctive, perhaps unique, personal aptitudes (in contrast with the tendency of most science to restructure itself so as to make specific individual workers replaceable [8]). Therefore, the education of a pure medical scientist would resemble that of a solo concert artist within the classical music system.

A concert soloist requires not only the technical expertise to reproduce music accurately (for instance as a junior orchestral violinist), but extra distinctive individual qualities of musical personality so that music may be ‘re-created’ in performance (e.g., by new ways of phrasing). Such creative soloistic performances may ultimately influence the ways in which junior orchestral musicians are trained. The soloist is expected to have total technical command derived from a prolonged basic training as an essential pre-requisite to higher studies, usually with a recognized expert high-level teacher at a conservatory. The conservatory teacher is typically a successful solo concert artist. (Although the very greatest teachers are not always the very greatest solo performers, the greatest teachers are almost invariably at least very good solo performers.) The selection of candidates for solo performance training is highly competitive and the choice (from among the large number of technically competent candidates) relies upon non-explicit artistic judgment of each student’s performance. This is usually accomplished by comparing rival students in performance-based auditions, examinations or music prize competitions. The precise criteria by which the judges measure musical creativity and personality are not explicitly definable, and are not accessible to the non-expert.

The selection of candidates for apprenticeship in pure medical science should probably be based upon analogous mechanisms to the selection of future potential concert soloists. The basic principles are that the candidate’s performance should be evaluated competitively by recognized pure medical science experts in the same discipline. Without being over-specific, this implies that new evaluation systems be developed, which aim to measure the cognitive aptitude of aspiring pure medical scientists – i.e. their ability to do pure medical science. The candidate would be required to demonstrate, under controlled conditions and in competition with other candidates, their ability to think and communicate in the way which is necessary to pure medical science. And their ability at pure science would be measured by judges who already can think and communicate at a high level in the pure medical science system.

Following graduation from a post-doctoral institution of pure medical science, the next objective would be to spend (approximately) 10 years achieving ‘mastery of a domain’ [10] as an intellectually independent pure medical scientist. In career terms, this might involve appointments to longer term positions at a pure science unit attached to one of the most internationally prestigious medical research institutions. These putative pure science units would presumably be specialized in their scientific subject matter and the pure scientists would interact with the successful applied science research groups already established at these institutions. A precursor of this kind of close interaction between pure and applied scientists might be the way that pure theorists such as Francis Crick have worked-with the more applied and experimental visual researchers at the Salk Institute and other nearby Californian research universities [11].

The leaders, or scientific managers, of pure medical science would be selected from these ‘masters’ according to the evaluation criteria of pure medical science; just as currently the applied medical science leadership are selected from those who are successful as defined by the evaluation criteria of applied science – for example having numerous highly cited publications in high impact journals, large grant incomes and a good record of project management in delivering planned objectives. There would naturally be different evaluation criteria for selecting pure science leaders. The substantive nature of their specifically scientific achievements would be given more weight (i.e. their major work should be both radical and ambitious, and also scientifically important). For pure medical scientists, there should be less emphasis on quantity and more on quality – a record of launching two science-transforming ideas would count for more than publishing 200 mainstream papers.

With respect to publishing within the social system of pure science, the trend in science is for the evolution of specialized communications media – including scientific meetings, journals and other more rapid interactive media such as e-mail groups. In order to render these media efficient, there would almost certainly be restrictions of some kind regarding participants in these communications – for example only ‘certified’ specialist pure medical scientists would be normally eligible to contribute. This would have the disadvantage of excluding uncertified people who nonetheless might potentially make important communications. So, in addition to the mainstream official media of pure medical science, there would also need to be alternative indirect and ‘unofficial’ routes of communication with the pure science system; for instance via the mass media, personal web pages, personal contacts and introductions, and ‘bridging’ journals such as Medical Hypotheses.

Corruption, competition and judgment by achievement

Corruption is a potential problem with social systems that are dependent on subjective evaluation by recognized experts. Since some of the most important evaluation criteria are inexplicit and the process of choosing the elite is intrinsically non-transparent, this can cloak the fact that other kinds of evaluations are in reality being applied. In a corrupt institution, nepotism, friendship, political ideology or sheer bribery rather than cognitive aptitude may become the real, but covert, evaluation criteria determining access.

Historically, there has been a tendency for technical experts to bolster their status by blending it with ‘cultural’ competence and excluding those judged to be uncultured (or having the wrong kind of culture). This applied to elite administrators as different as the Chinese mandarins and the top British civil servants of the 19th and 20th centuries – both of whom received a mainly ‘literary’ education and intense socialization in the required behavior of the ruling class. When cultural considerations become dominant, then social behaviors such as style, accent, manners, personal interests and opinions may become more important evaluation criteria than specific vocational competence. There is a potential danger that this kind of elite snobbery would come to affect pure medical science.

There is no way entirely to prevent a low frequency of cultured-corruption in the short-term, but corruption may be limited and contained over the longer term when the social system tends to punish the results of corruption. The main corrective to institutional corruption is, therefore, performance-based competition between institutions [8]. This operates best when institutional funding and support is based upon achievement. For institutions of pure medical science, relevant achievements would include the output of high level expert graduates who go on to become the next generation of ‘masters’, and the achievement of significant scientific breakthroughs. If the pure medical science institutions were to be supported on the basis of their productivity in these respects, then effective institutions would tend to thrive and corrupt institutions would tend to decline.

To return to the musical analogy, a corrupt musical conservatory who selected their student not on their soloistic aptitude but on the basis of snobbery would damage its output of prestigious solo concert artists and conservatory teachers, and would progressively lose its reputation. Aspiring teachers and soloists would tend to go elsewhere and if institutional support was linked to musical success a corrupt conservatory would decline.

Similarly, the best safeguard of standards in selection for training in pure medical science would be the existence of many competing institutions (preferably in several different and competing countries). Corrupt or incompetent pure science institutions would fail to produce effective pure scientists and the best students would go elsewhere – taking their funding with them. The evolution of useful institutions of pure medical science would, therefore, be assisted by the continued modernization of science in general, including increased competition between individuals and institutions – reforms which are particularly necessary in Europe as contrasted with the USA [12].

Funding elite pure medical science institutions

In the past, major advances in pure medical science were usually the product of a few gifted individuals, often working in small groups [13]. In future, it is likely that pure medical science – like applied medical science – will evolve toward a larger-scale, more explicitly structured form of organization typical of ‘Big Science’ [14] and [15]. Although pure science may not be amenable to the massification of current applied medical science (with laboratories of hundreds) in future it will probably still involve a significant element of research administration.

The probability is that a future pure science system would tend to become dominated by large institutions. What drives this trend is the tendency of systems to replace the un-predictabilities of depending on a few very exceptionally skilled individual humans, with the more stable systems of organization deploying coordinated teams of highly specialized, but more readily available and replaceable individuals [8] and [13]. This fuels the evolutionary replacement of small-scale loosely organized science based on a few accidentally occurring geniuses with large-scale science based on managed teams of technical specialists.

The most probable way that a professional specialism of pure medical science will be ‘launched’ would be the establishment of new pure medical science units at some of the most prestigious world-class science research universities [16] which currently do mainly applied medical science. Presumably, such elite research units would be created and supported in collaboration with some of the major independent medical research funders such as the Wellcome Trust, Howard Hughes Medical Institute, Rockefeller Foundation and the large disease-specific charitable groups. While new-founded centers would initially require substantial government support or private patronage, in the long-term it would be important that there was competition between many different centers in several different countries, to ensure that effort and standards were maintained.

Indeed, the establishment of separate and distinctive funding arrangements for pure medical science is crucial to its evolution. Since applied medical science has a higher probability of success and a much lower risk of failure, pure medical science cannot be evaluated by the same criteria without being driven towards extinction. Credit must be given for scientific ambition and potential importance, as well as for logistical efficiency.

Indeed, it may be that pure medical science is best supported by entirely different mechanisms of funding than the current one of funding research projects. At present funding agencies provide resources to support the process of research in the hope of future achievement. But alternatively, research funders could supply prizes retrospectively to reward actual scientific or technological achievement [17], [18] and [19] and these might be a means of stimulating research to solve specific important problems (e.g., a cure for breast cancer, a way of halting the progress of rheumatoid arthritis). An historical example of a prize successfully stimulating research was devising a mechanism for measuring longitude (although in the event this led to rapid incremental technological improvement rather than the hoped-for major scientific conceptual breakthrough [17]).

In addition, there might be a trend towards the ‘patenting’ of complex scientific theories, especially the kind that derives from long-term research by multi-disciplinary groups. If an institution could develop a theory, technology or therapy that was novel, non-obvious and useful then they might be able to profit financially from this [20]. Income from prizes and patents could potentially support large (and because large, risk-spreading) institutions that aim to win prizes and register profitable patents on a sufficiently regular basis to ensure their own continuation. And some of these putative institutions might operate in the pure medical science arena.

Whatever are the long-term innovations to funding mechanisms, the absolute requirement for initiating and growing a specialized social system of pure medical science is for separate and specified sources of funding dedicated to a pure medical science agenda and using evaluation criteria which are distinctive from applied medical science.

Will it really happen?

A professional, specialized social system of ‘pure’ medical science might be considered just another of those ‘good ideas’ which – as we described above, are published and then disappear, unevaluated. But if there is a real need for pure medical science, then such a system will probably arise – although perhaps not in the precise form we have described.

Where will the effective demand for pure medical science come from? It might come in a top-down way from the leadership of elite academic ‘entrepreneurs’ who intuit a need even before demand is fully expressed: the kind of individuals who have initiated some of the major research institutions and foundations of the 20th century.

Or, because the principal ‘market’ for pure medical science is applied medical science, immediate demand and support for pure medical science may come from applied medical science. This is plausible because the social system of pure science would be aiming to provide knowledge ready for further practical evaluation by the applied science system. But why would the hard-nosed, bottom-line-driven applied scientists be interested in the speculations of pure science? The answer is that most of the time they would not be interested, since the knowledge of pure science is riskier and less dependable. But sometimes the system of applied science cannot solve its problems by the safe and reliable method of incremental extrapolations and then the applied scientists could turn to the system of pure medical science to find a range of non-obvious but potentially useful ideas.

The ultimate demand for pure medical science would come from the public, especially from medical patients and their advocates – who realize that applied medical science is unlikely to deliver new cures for new diseases simply by doing ever more of the same old stuff [3], and who call-for more ambitious (high risk but also high reward) strategies of research.

An analogy can be seen in industrial R&D. Sometimes the demand for radically new strategies comes from the R&D people, sometimes from entrepreneurs who have a hunch about what the public wants (although the public do not yet realize it), sometimes from the public themselves who clamor for progress beyond the reach of established technology (for instance the public demand for ever-bigger, ever-cheaper television screens, which pushed the technology beyond the capabilities and cost of cathode ray tubes). Most of the time, R&D units improve the firm’s products by incremental extrapolation from already existing products. So that the Walkman was an incremental improvement of the cassette player, which was an incremental improvement of the reel-to-reel tape recorder. But this improvement can only go so far until diminishing returns set-in. The constraints of magnetic tape technology eventually became the major limiting factor on progress and then the situation was ripe for a radical new technology such as the CD, which relied on advances in pure science (digitization of sound, miniature lasers, etc.).

Likewise in clinical medicine, treatment of cancer is still based on long-established methods such as surgical excision, radiotherapy and cytotoxic chemotherapy. These techniques have improved incrementally over several decades to yield more remissions and cures and fewer side effects. But the rate of improvements is leveling-off and there have not been major curative breakthroughs in the treatment of those solid cancers which are the main cause of death in Western societies – breast, gut, prostate, lung, etc., [3]. Only a pure science approach of developing a variety of (competing) radical theories, techniques, technologies and therapies – and subjecting these ideas to specialized professional evaluation – offers genuine hope of progress.

If the public continue to demand new cures for new diseases, then sooner or later this will trigger the growth of a specialist professional social system of pure medical science.


[1] B.G. Charlton, Conflicts of interest in medical science: peer usage, peer review and ‘Col consultancy’, Med Hypotheses 63 (2004), pp. 181–186. SummaryPlus | Full Text + Links | PDF (157 K) | View Record in Scopus | Cited By in Scopus

[2] D.F. Horrobin, Scientific medicine – success or failure?. In: D.J. Weatherall, J.G.G. Ledingham and D.A. Warrell, Editors, Oxford textbook of medicine (2nd ed), Oxford University Press, Oxford (1987), pp. 2.1–2.3.

[3] B.G. Charlton and P. Andras, Medical research funding may have over-expanded and be due for collapse, QJM 98 (2005), pp. 53–55. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus

[4] R.K. Popper, Conjectures and refutations: the growth of scientific knowledge, Routledge and Kegan Paul, London (1972).

[5] P. Andras and B.G. Charlton, Democratic deficit and communication hyper-inflation in health care systems, J Eval Clin Pract 8 (2002), pp. 291–297. View Record in Scopus | Cited By in Scopus

[6] B.G. Charlton, Inaugural editorial, Med Hypotheses 62 (2004), pp. 1–2. SummaryPlus | Full Text + Links | PDF (129 K) | View Record in Scopus | Cited By in Scopus

[7] N. Luhmann, Social systems, Harvard University Press, Cambridge (MA) (1995).

[8] B. Charlton and P. Andras, The modernization imperative, Imprint Academic, Exeter (UK) (2003).

[9] D.L. Hull, Science as a process, Chicago University Press, Chicago (IL) (1988).

[10] F.L. Holmes, Investigative pathways, Yale University Press, New Haven (USA) (2004).

[11] F. Crick, What mad pursuit: a personal view of scientific discovery, Penguin, London (1990).

[12] P. Andras and B.G. Charlton, European science must embrace modernization (correspondence), Nature 429 (2004), p. 699. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus

[13] B.G. Charlton, The last genius? – reflections on the death of Francis Crick, Med Hypotheses 63 (2004), pp. 923–924. SummaryPlus | Full Text + Links | PDF (209 K) | View Record in Scopus | Cited By in Scopus

[14] D.J. de Solla Price, Little science, big science – and beyond, Columbia University Press, New York (USA) (1986).

[15] J. Ziman, Real science, Cambridge University Press, Cambridge (UK) (2000).

[16] Institute of Higher Education, Shanghai Jiao Tong University. Academic ranking of world universities; 2004. [Accessed 02.06.2005].

[17] D.F. Horrobin, Glittering prizes for research support, Nature 324 (1986), p. 221. Full Text via CrossRef

[18] Davis LN. Should we consider alternative incentives for basic research? Patents vs. prizes. Presented at DRUID summer conference; 2002. [Accessed 02.06.2005].

[19] Davis LN. How effective are prizes as incentives to innovation? Evidence from three 20th century contests. Presented at DRUID summer conference; 2004. [Accessed 02.06.2005].

[20] J. Rees, Patents and intellectual propery: a salvation for patient oriented research?, Lancet 356 (2000), p. 849.