The following text is mainly derived from the report Contract Research: Openness, Quality, Accountability (Kaiser et al. 2003).
In 1986, Knut Erik Tranøy defined scientific activity as ‘the systematic and socially organized a) search for, b) appropriation and production of, and c) administration and communication of knowledge and insight’ (Tranøy 1986:59). Later works have followed this tradition of associating systematic approaches with social organization (Kaiser 2000:152).
According to Plato the “classic” definition of "science" (scientia, Latin forknowledge) is:
Knowledge is true belief/opinion combined with an account of the reason to believe it. Or, in brief, as is stated in the standard textbook version: Knowledge is justified true belief.
In essence the scientific revolution can boast of having altered perceptions of science/knowledge in two respects: (i) it added specific perceptions on scientific methodology, and (ii) it anchored knowledge in a socially organised scientific community. Thus a tentative definition may be given as:
Scientific knowledge is true belief/opinion combined with a scientifically justified account of its cause, and anchored in the norms of the scientific community.
Verifiability, public, criticism and universalism
The various phases of scientific institutionalisation can be categorised: amateur science (approx. 1640–1800), professional science (approx. 1800−1870), industrial science (1870–1940), and Big Science (1940–present day). It was the scientific academies that, in this initial and important stage of modern science (so-called amateur science), administered the normative basis for scientific activity. It is not easy to unambiguously characterize the prevailing norms. One important element was the experimental method and its underlying view of nature, something that was somewhat cryptically referred to as ‘Natural Philosophy’ or ‘Natural Knowledge’. Another important element concerned linguistic standards and the role that mathematics can play in attempting to describe the laws and regularities of nature. A universal canon of such methodological rules was not formulated, however. Theorists such as Francis Bacon and René Descartes made influential suggestions regarding methodology and scientific policy, but without achieving the same paradigmatic effect as practical examples of successful science.
How systematically or methodologically grounded this science in fact was has been debated for years, though we may assume, with certain caveats, that the underlying base was the concept of the intersubjective verifiability of results. Other basic norms were, on the other hand, successfully institutionalized. One such norm was the demand for public knowledge, which specifically entails that all scientific insights are to be included in a common arsenal of knowledge. Secrecy is disavowed and the personal possession of knowledge is rejected. The justification for such a norm lay in the assumption that the common benefit is greater when knowledge is publicly disclosed than when knowledge is considered to be private property. This was simultaneously the basis for scientific journals. The second norm that was institutionalized is a direct consequence of the first, and can be described as institutionalized and systematic criticism. When knowledge becomes public, either through lectures or publication, it is the task of the scientific community to critically examine the argumentation in order to detect potential mistakes. Bacon, in particular, had pointed out the danger of becoming the victim of wishful thinking and other ‘Idols of the Marketplace’, as he named them.
In the practice of the scientific academies lay also the seeds of yet another research norm, namely universalism. Bacon had vigorously argued that the affairs of science required the greatest minds of every country. The reality was that the tiny scientific community of the initial phase consisted of a few thousand scholars, who kept in constant touch with one another across national borders.
Merton on the ethos of science (CUDOS)
In 1942 the American sociologist Robert K. Merton formulated what he considered to be the basic ethos of science, i.e. its normative foundation (Merton 1973). His four primary norms are:
- Communalism (the demand for communal possession of scientific knowledge; public knowledge).
- Universalism (the rejection of any preferential rights to science; everyone has an equal opportunity irrespective of social background, nationality, etc).
- Disinterestedness (independence from special interests).
- Organized scepticism (the demand for systematic criticism of scientific claims).
Later on he added a fifth norm:
5. Originality (rewards in the form of special recognition are awarded to those who first bring to light new knowledge).
In light of the preceding comments regarding the scientific academies and the scientific revolution, it is easy to see that Merton’s 1st, 2nd and 4th norms are inspired by this history. It is, however, a slightly different case with the 3rd and 5th norms, and at the outset there is reason to believe that they stem from more recent times. The norm of disinterestedness seems to be clearly inspired by Max Weber’s (1864–1920) postulate of value neutrality. This can be related to the expectation that science should be objective in the sense that it is not steered by subjective values and prejudices, and provides balanced presentations.
Merton’s norm of originality serves to remind us that an adequate understanding of the social dynamics of science must include the established system of scientific credit. We find the norm of originality explicitly expressed in the PhD regulations of most universities.
Justification of the norms:methodology
Merton justifies his norms on the basis of two parallel arguments: firstly the historic argument that the norms can be deduced from the institutional arrangements that have encompassed modern science since the scientific revolution; and secondly a functionalist argument that the norms, in tandem, enable the production and quality assurance of knowledge that society expects science to provide – an argument that many may consider to be problematic.
However, the problem lies precisely in the fact that this demarcation between “scientific” and “non-scientific”is not clear cut, and in many cases this will lead to a debate also within the scientific community and among philosophers of science. By adhering to the pluralistic view, however, it would be possible to state a sort of minimal answer to the problem: “scientific” is that which competently uses the tools that at any given time are to be found in the scientific toolbox. Thus, the choice of (empirical) method does not have to be decisive as long as a) a method is used at all, b) this method is suitable for solving problems of the given type, and c) the method is used competently.
The entire point of using a method is that it serves as a quality assurance ofthe knowledge and insights that research produces. Methods should not only provide results, they should also, based on given preconditions, enable the systematic and intersubjective verifiability/quality control of the results. When knowledge is presented as being scientific or research-based, the conditions must be such that peers can review the fundamental data and based on a certain method, assess whether the conclusions are valid. This is the most important prerequisite for scientific quality assurance.
Traditions for justification
It is one thing to become fascinated by the underlying method of science and the inherent rules and regulations of the scientific community, but it is something quite else to explain to others why scientific activity is valuable. The latter aspect is what is normally called the justification of science. Justification pertains to the external aspects of science, and it has become customary to discern between two fundamental forms of justification. Aristotelian justification takes as its starting point the inherent worth of knowledge as a sort of project of self-realization for mankind; Baconian justification, on the other hand, takes as its starting point the utility of knowledge, where the utility is assessed in regard to people other than those who produce the knowledge (see Tranøy 1986, Kaiser 2000: chap. 2). Tranøy has introduced the general concept of welfare functions and welfare effects to take into account ‘the different ways in which insight, and in particular systematic scientific insight, affect the individual and social welfare of human beings, for good or for bad’ (Tranøy 1986: 78). Seen in relation to the problems of modern research policies, the two above-mentioned traditions of justification provide only a sketchy blueprint, but it is not unreasonable to imagine that also later ideas of justification will tend towards one of these two basic forms.
A more recent model of the justification of research is associated with John Desmond Bernal (1901 – 1971). From the 1930s on, in England in particular, there emerged a group of Marxist-inspired theorists who viewed Soviet science as an ideal. Øyvind Såtvedt summarizes Bernal’s programme of research policy in four points (Skoie and Såtvedt 1998: 33):
- The principle that research must be systematically organized and that priority must be given to applied research that is beneficial to society.
- The principle that all researchers are ethically obligated to fight against any misuse of the knowledge that science has produced. There is no moral boundary between the production of knowledge on the one hand, and the application of this knowledge on the other.
- The principle that science is an instrument for social transformation (emancipation) and is rooted in practical life.
- A set of historiographical theses to be employed when describing the history of science.
There are many noteworthy aspects of this model (sometimes referred to as ‘Bernalism’): (i) all science is justified by its utility, (ii) freedom and self-determination are limited by governmental research policy, (iii) value neutrality and disinterestedness are replaced by a comprehensive ethical obligation towards the production and application of knowledge that is beneficial to society, and (iv) when the ethical basis of science is linked to social transformation and practical activities, then it is natural for science to forge close alliances with varying user groups in society.
Considering the reality of research policy today, it is striking how close it is to Bernalism, and how far removed it is from the ideals of Humboldt and Weber. Referring to Eirikur Baldursson, Såtvedt states that it is ‘a paradox that while Communism – which to a great degree inspired Bernal’s thinking on research policy – has largely been dethroned in recent decades, the ideas of Bernalism seem to be alive and well’ (Skoie and Såtvedt 1998: 35). He goes on to quote Aant Elzinga: ‘Bernal’s ideas have been “taken over by the captains of industry and ministers of government in the postwar period”’ (ibid. 1998: 35).
In the above, we outlined some of the historical developments and debates regarding the normative foundation of science, with regard to its inner social organization, its methods and its traditions of justification. These are the three areas that are particularly important for our present concern with for instance contract research.
Of course, the normative foundation of science involves more than those areas outlined above. An essential element that we have not dealt with pertains to problems of scientific integrity, and there are in that context several norms that stipulate scientifically correct behaviour.
The scientific community follows a set of internal scientific rules (norms) that allows for the emergence of scientific knowledge in a dynamic network of researchers in which each individual contributes insights, based on the previous insights of others, and submits them to the community to be critically tested and validated. Respect for the contributions of others and giving adequate credit are therefore important in the scientific community. This is expressed, for example, in the rules for crediting others’ contributions, i.e. usually in respect of references, as well as in the rules for authorship and publication. Plagiarism of the work of others damages the trust on which the scientific community is founded. The most damaging form of variance is dishonesty (fraudulent research), for example falsifying a test or test data. Scientific misconduct (dishonesty) is therefore often classified as so-called FFP (“fabrication, falsification, plagiarism”). Following the entry into force of the Norwegian Act on ethics and integrity in research, allegations of such misconduct are scrutinized by a separate commission. Less extreme violations of the internal norms of research are sometimes referred to as QRP (“questionable research practices”).
Another element of the normative foundation of science concerns research subjects, whether humans or animals, for example in medical research. No objective justifies the unethical use of research subjects. Clear standards have been drawn up for this, for example in the Declaration of Helsinki.
Thus if we consider research practice as a whole in the sense of a set of internal research rules as well as actions that are censurable from the point of view of research ethics, we must take into account different norms or sets of norms. An early publication on dishonesty in research published by the research ethics committees (Elgesem, Jåsund, Kaiser 1997) attempted to provide an overview of such internal research considerations in a table (see table 1).
Table 1: Overview of various sets of ethical norms of importance for research practice (from Elgesem, Jåsund, Kaiser 1997).
|Internal research ethics||
Violation of the duty to be truthful
|Authorship and giving adequate credit||Violation of verifiability norms||Violation of general ethical norms|
Selective use of data or methodology to verify the hypothesis
|Plagiarism of text, data or ideas||Deliberate destruction of data in order to prevent others verifying results||
Non-compliance with licensing rules, failure to obtain permission
Exposing animals or humans to risks
Deliberate misrepresentation of someone else’s results
Failure to report fraud when discovered
Failure to recall publications that are based on dishonesty or serious errors
Failure to credit co-authors
Failure to cite all sources
Inadequate storage or distribution of data
Publishing in the media without documented results
|Deliberately misleading information in project applications|
Freedom, openness and independence?
In terms of research policy, scientific research is often characterized by three key values/norms, i.e. freedom, openness and independence. The above-mentioned observations provide a platform for reflecting on these:
The norm of freedom
The norm of freedom in research has been asserted and defended in connection with the emergence of modern research universities in the 19th century. It refers at the outset to the universities’ freedom to determine themselves how the accumulation of scientific knowledge should be managed and renewed. Specifically, the norm has consequences for appointments, internal organization and teaching at the institutions. As a direct consequence of the norm, scientific personnel at the institution are given the freedom to choose their own field of work and their own research activity. The research activity is in other words not only exclusively controlled by researchers, but also initiated by researchers. The norm was ideologically associated with the perception of scientific knowledge as a cultural good, something that was indirectly beneficial for society in connection with formative education and schooling. The norm has been limited and modified in pace with the increase of institutes for applied research, which were exclusively utilitarian. In such institutes, research was still largely controlled by researchers, but not initiated by them – the objectives were defined externally. The development of Big Science after World War II, however, as well as later developments that took place on the outer limits of the universities, limited the researchers’ control over research. Larger and more complex projects, with predetermined objectives and a clear view towards technological application, demanded a form of supervision and organization that was not necessarily a matter of internal interest to science. Remnants of free research (i.e. research that is controlled and initiated by researchers) are still to be found in today’s universities, but when research funds are scarce and large resources are demanded in certain fields, researchers have in reality limited opportunity to initiate projects. The actuality of the norm of academic freedom, such as it is adapted to the modern world of research, is therefore open to debate (Menard et al. 1996).
The norm of independence
For individual researchers, it has always been a goal that their research should remain unaffected by external, non-scientific interests. At the same time, the social context of science has always required that alliances be struck with other parties and powers-that-be. The greater the need for funding, the closer these alliances became. Alliances always create a mutual dependence and limit the independence of a given party. This entails that the allied partner’s interests will colour the given researcher’s own activity, something that became ever more relevant as the technological and commercial dimensions of research came into focus. Today’s research is characterized by universities and autonomous institutes forging several such formal and informal alliances. Many have noticed a trend towards the increased commercialization of research, something that clearly created new forms of dependency. At the same time, such alliances do not necessarily have to lead to the deterioration of quality specifications in scientific research. In principle it is rather the opposite: it is precisely such quality specifications that make science an attractive alliance partner. It is also because of these quality specifications that the general public is able to have confidence in the results. The challenge lies therefore in the delicate balance between desired connections on the one hand, and a dependent relationship that is detrimental to research quality on the other.
The norm of openness
The norm of openness is not, on the surface, directly evident in traditional formulations of the normative foundation of science. What most resembles the norm of openness is the demand that knowledge be public, and perhaps also the methodological demand for the intersubjective verifiability of research results. Openness with respect to both data and method is a prerequisite for scientific quality assurance. Openness with respect to research results, in the form of publicly accessible publications, is a prerequisite for quality assurance in the form of peer review and for the practical application of research results to the benefit of society. It is well known that complete openness in this respect does not always exist, for example for reasons of national security or to protect industrial secrets. Modern patent practice usually entails a time-limited exclusivity. When discussing Bernalism and post-academic science, however, where special interests to a larger degree influence the contents of research, we asserted that the overriding interests of society linked to the potential welfare benefits of research imply that the norm of openness should include publicly accessible information about a) which research projects are in fact being conducted, b) who is funding this research and which user groups are collaborating on the project, c) how the research is being quality assured, and possibly d) how the results will fall into the public domain, even though this might take place after a certain time span.
Based on our preliminary discussion, we believe that research should also be controlled by two important norms that are not always explicitly mentioned, namely quality and accountability.
The norm of quality
Quality assurance of scientific knowledge has been a central part of scientific activity ever since the scientific revolution. The use of method, combined with intersubjective verifiability in the form of peer review, has always been promoted as key parts of such quality assurance. The system of peer review is relatively sophisticated and well established in today’s situation. As for method, we have emphasized the important distinction between having an awareness of method on the one hand, and the fact that researchers may, on the other hand, have differing estimations and viewpoints on how scientific problems should be tackled. Professional disagreement is therefore not an indicator of quality failure, nor does professional consensus on a given conclusion indicate that the research quality has been assured. Scientific quality assurance refers to processes rather than conclusions.
In research that is out-and-out utilitarian in nature, the required quality assurance will often include other parties than researchers, for example users. In research where common interests are at stake, and where there exists a significant degree of scientific uncertainty (post-normal science), it would be natural to expect that research should be open to input from various interested parties in regard to its preparation, transaction and conclusions. Although adequate quality assurance of scientific research is fundamental to the general public’s confidence in research, it can never guarantee valid, final and commonly accepted research results. Such a guarantee does not exist in cutting edge research. It turns out that the scientific community itself values professional quality highest among the goods that are demanded by ethical guidelines for research (cf. SCRES 2002).
The norm of accountability
Modern science, at least from the Enlightenment onwards, has always considered participation in the affairs of society, be they political, cultural or commercial, as a natural part of its activity. Nor
did Weber’s postulate of value neutrality lead to a lack of social commitment on the part of science. Weber’s postulate did not, however, demand a particular ethical responsibility on behalf of science. As long as the universities had sufficient institutional freedom to control and determine the research agenda themselves, it seemed certain that relevant knowledge would befall the general public. However, when utilitarian ideas exert ever greater influence on what is to be studied, how the research is to be conducted and who shall have access to the results, it seems imperative that research ethics are reinforced in order to compensate on behalf of the overriding interests of society. An empirical study of ethical guidelines for research, conducted by SCRES, shows that social responsibility is the most highly prioritized among the social values that science should seek to attain (SCRES 2002). It is hardly reasonable to claim, as Bernal does, that researchers are categorically responsible for how their results are applied. But it is not unreasonable to contend that researchers are jointly responsible for the use of science in the wider context of society (see Mitcham & Schomberg 2000).
It is precisely this form of ethical co-responsibility that we refer to here as accountability. Such a norm of research ethics essentially includes two aspects: 1) When certain decisions and/or technologies that were in essence made or created on the basis of scientific know-how turn out to have unfortunate, unforeseen side effects, and when there simultaneously is reason to believe that a better or broader preparation of the scientific decision-making basis could have forewarned against such possible side effects, then there is reason to believe that science is ethically co-responsible for the negative consequences. 2) When there exists either specific results or important scientific uncertainty which imply that a planned decision and/or technology could have serious consequences for e.g. society, health or the environment, then it is the ethical duty of science and the individual researcher to ensure that this information is effectively made known to the relevant decision makers and the general public.
We deem that such a norm of accountability overrides all other obligations, for example obligations towards a client or an institution. This norm entails furthermore that science engages itself in public debates where scientific and technological questions are on the agenda or form an essential part of the debate.
We have, in comparison with the norms that were discussed above, downplayed certain other norms, for example the norm of universality. This does not imply that we do not see the value of such a norm; it is rather due to our opinion that the norm is in reality quite complex, and that for example a certain amount of local knowledge also has a part to play in research. We refer here to the bibliography for further discussion (see for example Cole 1992). We have also downplayed Merton’s norm of disinterestedness, which we believe is better maintained by our norms of openness and accountability.
We believe there is reason to contend that openness, quality and accountability form the normative core of research ethics and are applicable to scientific activity in general. The norm of independence has a secondary meaning in the sense that the quality of science should not be unduly influenced by the given researcher’s inevitable relationships of dependency, and as such this norm is important for contract research. On the other hand, it is difficult to uphold freedom and independence as key norms when considered in the general form they were discussed above. Actual research is seldom of the kind these norms seem to imply, and it is difficult to see that they should comprise a minimum standard for research ethics in general – actual research is too complex for that. No researcher is completely independent: we are all dependent on something at all times. On the other hand, both of these norms have certain aspects, pertaining to quality assurance and accountability, that are important to uphold. Subsidiary aspects of these norms will not be included as consequences of our three overarching norms.
This article has been translated from Norwegian by Jennifer Follestad, Akasie språktjenester AS.
Bernal, J.D. 1967. The Social Function of Science. Oxford: Oxford University Press
Cole, S. 1992. Making Science – Between Nature and Society. Harvard: Harvard University Press
Elgesem, Dag; Jåsund Kjetil og Kaiser Matthias: ”Fusk i forskning”. De nasjonale forskningsetiske komiteer. Skriftserie nr. 8. 1997
Funtowicz, S. & Ravetz, J. 1993. “Science for the Post-Normal Age”, Futures 25/7: 735-755.
Kaiser, M. (2000). Hva er vitenskap? Oslo: Universitetsforlaget
Kaiser, M., K. Rønning, K. W. Ruyter, H.W. Nagell og M.E. Grung: Oppdragsforskning – åpenhet, kvalitet, etterrettelighet; De nasjonale forskningsetiske komiteer 2003
Menard, L. et al. 1996. The Future of Academic Freedom. London: The University of Chicago Press
Merton, R. K. 1973. The Sociology of Science – Theoretical and Empirical Investigations. Chicago & London: The University of Chicago Press.
Mitcham, C. & Schomberg, R.v. 2000. The Ethic of Scientist and Engineers: From Occupational Role Responsibility to Public Co-responsibility. Amsterdam: JAI Press.
SCRES 2002. Standards for ethics and responsibility in science – an empirical study. The Standing Committee for Responsibility and Ethics in Science, under The International Council for Science (ICSU), i dag International Science Council: https://council.science/, https://council.science/wp-content/uploads/2017/05/SCRES-Background.pdf
Skoie, H, & Såtvedt, Ø. 1998. Forskning, kultur og autonomi. Et bidrag til debatten om den frie universitetsforskning. Oslo: Norsk institutt for studier av norsk forskning og utdanning: 200.
Tranøy, K.E. Vitenskapen –samfunnsmakt og livsform Universitetsforlaget Oslo 1986, 1991.