Who Decides a Scientist’s Research Agenda? Choices and Constraints

Abstract

Who should decide a scientist’s research agenda? The scientist him- or herself? Superiors, such as science ministries or university management? Sponsors, such as funding organizations or industry? Interest groups, such as churches, political parties, animal rights organizations or patient initiatives? What is the right balance between the scientist’s autonomous decision on the one hand and players’ interests on the other? These questions will not be answered here, but reflected upon with regard to the difficult trade-offs involved and the threats to scientific freedom that can result from them. For this purpose, I propose an analytical framework that conceptualizes the thematic autonomy of science as a gradual interplay of individual choices of research topics made in the context of the respective scientific community, on the one hand, and external constraints, on the other. With regard to them, seven basic levers will be distinguished. This framework will be applied to the current state of German professors’ freedom to choose their research agenda.

One of the essential elements of academic freedom is that a scientist’s research agenda is based on his or her own preferences and decisions. This is no unequivocal yes-or-no question but a matter of degrees − which makes it sometimes rather difficult to judge whether there is an infringement of academic freedom in a particular situation − or not yet. Those who specifically want to restrict scientific freedom can strategically exploit this blurring by denying that this is what they are doing.

I want to propose an analytical framework that helps us in such considerations of, first of all, understanding, and based on that, judging relevant incidents. As a prototypical case, I will discuss a typical German professor and the context in which she or he is situated while doing research. So what is their scope for free choice of their research agenda, and what constraints limit it? Much of what I will show for today’s German science system is basically similar to many other national science systems of contemporary Western countries. There are national differences which can be important, but here I am concerned with the commonalities.

I will start with a look at how individual choices of research topics are made in the context of the respective scientific community. Then I will consider and assess the external constraints of these choices. Finally, I will draw some conclusions about the current state of German professors’ freedom to choose their research agenda. To provide a broad overview and give an overall assessment of the state of academic freedom in Germany, I cannot go into details or look more closely at particular cases. I will refer mainly to recent literature which summarizes certain aspects of the topic.

Individual Choices, Embedded in the Scientific Community

With regard to research topics, freedom of research does not only mean the choice of what is researched in terms of questions. The contouring of a research topic includes theoretical approaches and methods used. Someone’s freedom of research is not only restricted when it is made difficult or even impossible for them to address a particular research question, but also when they are kept away from using their preferred theoretical perspective or the methods they wish to work with.

Conceptualizing individual choices of research topics from which individual research agendas emerge, I follow a simple model by Jochen Gläser (2006). A researcher can be conceived of as an actor who has developed preferences for certain topics − in terms of questions, theory and methods − via a specific trajectory that typically began during their student years. This is accompanied by the acquisition of certain skills in dealing with these topics. This trajectory is characterized by strong path dependencies − which does not rule out path breaks, i.e., completely new preferences that can then lead to the acquisition of completely different skills. However, it is important for someone’s freedom of research that these path breaks are not imposed from the outside but occur voluntarily on the basis of one’s own desire and insight.

Path dependencies and path breaks take place in the context of a topic-centred scientific community − disciplinary or interdisciplinary − such as particle physics or medieval history; and the more one sees oneself as a competent researcher, the more one is under the spell of the community’s research front and the questions raised there. There is a collective agenda of what the next important steps in the pursuit of knowledge are − and these research questions are categorized according to their importance and their degree of difficulty. The more important a question is, the greater the reputation one can earn by contributing to the advancement of knowledge that is assessed as successful or even pioneering. However, importance often correlates with difficulty and therefore with the risk of failure. In this way, individual preferences and abilities as criteria for deciding which research topics a researcher pursues are contextualized, on the one hand, by how popular a topic is among colleagues at the forefront of research and, on the other hand, by how daring it is.

This is the internal scientific context in which decisions on topic selection take place; and because up to this point only scientific aspects play a role, one can speak of autonomous decisions in the sense of freedom of research. The legal construct of academic freedom attributes this to individual academics. However, from a sociological perspective what is meant is not the freedom of the person, but of the role; and the role of the researcher cannot be thought of without considering the expectations of its reference group. Freedom of research is therefore not a question of individual arbitrariness, but of individual choices embedded in the disciplinary or interdisciplinary community − which implies that the discipline also allows the daring or even the ‘crazy’ if it appears potentially fruitful for the progress of knowledge.

However, the aforementioned path dependency is reinforced here. On many occasions, a researcher is assessed by peers: when applying for professorships, submitting articles to journals and asking for third-party funding of research projects. The peers tend to quickly categorize him or her into a certain preference and ability profile and then no longer trust him or her with anything else. His or her research agenda can therefore be restricted not only by the fact that the daring is not given a chance, but already by the fact that research topics which do not correspond to his or her track record are made difficult or even impossible for him or her to pursue. This is a considerable internal restriction which is usually neglected in discussions of individual thematic freedom of research. Because my focus here is on external constraints to academic freedom, I will not discuss this any further; but it would be worth further consideration because this categorizing generates mainstream research and discourages attempts to try out unconventional new approaches.

Extra-scientific Constraints

Scientific knowledge production is not based solely on internal production factors and conditions. There are also indispensable external ones, i.e., those not created by science itself. They can enable or constrain. With regard to threats to freedom of research, the focus here is primarily on constraints − including enabling factors that do not get a chance because they are suppressed, either deliberately or unintentionally.

To underline the significance of constraints with regard to a decision’s autonomy of choice, it is worth recalling Herbert Simon’s (1964: 262) succinct statement: ‘If you allow me to determine the constraints, I don’t care who selects the optimisation criterion’. In other words: those who set the constraints can grant scientists the freedom to choose their research topics and still limit their autonomy as much as they like if their constraints are set massively and skilfully enough (Gläser 2010: 365).

Richard Whitley’s (2010, 2014) concept of ‘authority relations’ is a well-suited analytical tool as a starting point for analysing such threats (see also: Gläser 2010; Whitley and Gläser 2014; Schimank 2014: 29−36). It views individual researchers as actors whose room to manoeuvre – here, in the choice of research topics − is shaped by the design of governance structures to which they are subject. Whitley (2010: 5) uses the term ‘authority’ to describe regulated legitimate influence. Influence, however, presupposes intentionality; in addition, the transintentional effects of the actions of others on a researcher’s opportunities for action − e.g., unnoticed or disregarded restrictions on time for research as a result of measures to increase teaching loads – must be taken into account. If one considers influences along with these transintentional effects in their overall impact, the residual amount of room to manoeuvre remaining to the researcher is their ‘protected space’ (Whitley and Gläser 2014: 8) − what they enjoy in terms of research freedom. It has, first of all, a substantial dimension: what scope of possible questions, theories and methods does a researcher have? Second, there is a time dimension: for how long can the researcher devote attention to their own choice within this scope without having to defend the work against disturbances from external forces? For example, does a research grant for a specific project cover three or only two years? The shorter this time, the earlier the researcher must look for further financing.

Against this background, seven levers can be identified by means of which other actors may have an impact on a scientist’s research agenda – either intentionally or transintentionally, and either enabling or, as is of interest here, restricting it. This is not an exhaustive list, but these levers seem to me to be the most important ones in the current German science system. I can only address each of them briefly here by using four dimensions in which a lever can be characterized in terms of how great a potential threat it poses to the freedom of German professors to choose topics, both now and in the future:

• Breadth of effects: How widespread is the threat? Does it affect only a small number of professors, particular fields of research or certain universities − or is it more or less ‘systemic’ in the sense of penetrating more or less all kinds of research?

• Effect intensity: How far-reaching is the threat? Does it remain superficial and relate to minor aspects of the research agenda − or does it mean major cuts in the freedom to decide on questions, theories and methods?

• Intentionality of effects: How purposeful is the threat? Does it arise transintentionally as an unnoticed side effect of actions of certain counterparts of the professors − or do the counterparts want to achieve precisely this restriction of the professors’ freedom of choice?

• Trend of effects: Are the incidents temporary or stagnating at the same level − or has the threat increased and does it have the potential to increase further? In other words, can a trend of endangerment be identified?

Lever 1: Regulation

Regulation takes the form of general legal provisions and the resulting directives in specific cases – for example, an animal protection law and its authorization or non-authorization of certain animal experiments. In Germany, such bans on particular lines of enquiry, theories or methods require justification by other legal rights that are classified as equally or even more important than the freedom of research. For example, certain biological and medical research methods can be excluded on the grounds of animal welfare or ethical considerations, such as in stem cell research; or data protection as a prerequisite for personal self-determination prevents certain questions and methods of social research.

The civil clauses which many German universities use to prohibit research that primarily serves military purposes are an interesting case in point. They were introduced in the 1980s and 1990s at many universities to secure the freedom of research against influences from the military. Since the start of the Russian war against Ukraine, a debate has started as to whether this restriction on the choice of topics should be lifted in view of the altered geopolitical security situation. In other words, doing away with these regulations is now seen by many as a facilitation of research which is necessary from a societal perspective – and those researchers who want to serve this wish should have the freedom to do so.

A special case of regulation in Germany is the long-standing and frequently lamented ‘Zerwaltung der Forschung’ (Meusel 1977) − a play on words for a sub-type of bureaucratization that alludes to the fact that general administrative frameworks of public service applied to universities are often inappropriate for science. In recent years, there have been complaints, in particular about the personnel regulations of the German civil service, such as rules on fixed-term employment contracts, the effect of which has been that the most talented employees have repeatedly not been allowed to be hired for research projects. Above a certain level of restrictions imposed in this way, one can certainly speak of a threat to academic freedom − albeit an unintended collateral damage of rules that may work well in other parts of the public service.

The above examples show that in the German science system − as in many other Western countries − there are no formal authorities who can order a professor to devote him- or herself to certain research topics or to apply certain theoretical perspectives or methods. Professors cannot be formally forced to pursue a specific research agenda; instead, voluntariness is formally the principle. They can, however, be forced not to do certain kinds of research. Still, if university management or science ministries formally interfere with the freedom to choose research questions, theoretical perspectives or methods in a way that is deemed illegal, the scientists concerned have legal recourse. In this respect, a precaution against threats to the freedom of research is institutionalized.

It is therefore true that regulation can certainly mean far-reaching and very targeted threats to thematic freedom of research. However, regulation is usually not comprehensive, but only affects individual professors or sub-areas of research fields; and there is no recognizable trend towards an increase in Germany, although in the 1990s representatives of the life sciences painted a grim picture of the restrictions placed on the freedom of embryonic and genetic research by political regulations in response to civil protests, objections from churches and other ethical objections (DFG 1996). These temporary interferences were soon overcome.

Lever 2: Discourse Controls

Bans on speech and thought − often referred to as ‘cancel culture’ − are the ‘soft power’ counterpart to regulation (for an overview, see Ash 2002). Something is not formally forbidden − but is attempted to be stopped by those who have a similar or even greater power of enforcement through moral ostracism. These can be non-scientific actors: political groups, religious communities or other advocates of certain values such as animal-rights activists. However, actors within the science system, such as colleagues or students who take moral offense at someone’s research topics and results, can also play an important role here.

Discourse controls can relate not only to research questions, but also to theories and methods. For example, in the social sciences it can be the case that − in order to comply with the requirement that scientific statements be free of value judgements − the theories used do not adopt a certain moral assessment of the topic under consideration, even if it is considered the only legitimate one. If, for instance, a researcher of colonialism does not condemn the atrocities committed by the colonial rulers, but merely describes and explains them, this may lead contemporary critics of colonialism to the conclusion that he or she accepts or even approves of these acts. To these critics, this research is considered acceptable only if the researcher takes on the ‘guilt trip’ (Collins 1997) imposed on him or her, for which a moral statement is not enough: a re-framing of the research question and the theoretical perspectives and methods used is demanded. Otherwise, the researcher must be forbidden to write and speak on that topic.

Even more far-reaching discourse controls insist that those who are personally affected by a particular research topic should have the last word on the results of this research. Extreme proponents of so-called ‘transdisciplinary research’ (Lawrence et al. 2022) argue in this way. Only those who have suffered, for example, gender discrimination or a certain illness, often extended to include those who may have suffered both, or who appoint themselves as spokespersons for those who have suffered, are allowed to judge what research results on these topics are valid and reliable and what findings, in contrast, represent ‘epistemic violence’ (Spivak 1988; Brunner 2020) that must be prevented. This amounts to a deliberate relativization or even abolition of the pursuit of objectivity, which is the hallmark of scientific knowledge in contrast to all other forms of knowledge (Schimank 2024: 453−458).

As with regulation, discourse controls pose profound and targeted, but not widespread, threats. Only a limited number of often-narrow subject areas, most of them in cultural studies but a few also in natural sciences – in particular, wherever sex/gender issues are at stake – are affected, and in Germany the confrontation between the ‘old white males’ and their challengers from various movements of identity politics is still much tamer than in the United States. However, a ‘Netzwerk Wissenschaftsfreiheit’ (Network Academic Freedom) was founded in 2021 which understands itself as a defence alliance for academic freedom.

Whereas discourse controls have so far largely emerged from the left-wing, ‘progressive’ political spectrum, corresponding attacks on academic freedom could currently also come from the far right. At the moment, in Germany the right-wing populist party ‘Alternative für Deutschland’ (AfD) has increasing voter support and is by now the second-largest party on the national level, the largest one in two of 16 federal states. Developments such as those in Hungary since Viktor Orban was elected (Magyar and the Republikon Institute 2023) would become possible if the AfD becomes part of the government. The party has repeatedly announced that it wants to abolish, among others, gender studies at German universities and to reorient teaching and research in German history of the twentieth century with regard to the era of Nazi Germany.

Lever 3: Time

Research topics differ in terms of the time they require; and if the time required is greater than the time available, the topic cannot be dealt with adequately or at all. The amount of time a German professor has for research is mainly regulated by the size of his or her teaching load − currently nine semester hours per week. Within this legal framework, the number of students that professors and their teaching assistants have to supervise differs considerably across disciplines and determines the time needed for teaching, including examinations. Student numbers have increased over a long period of time in most subjects in Germany − without this being sufficiently compensated for by additional teaching staff − and are currently stagnating at a very high level. Furthermore, the time needed for administrative tasks is deducted from the time available for research. There are many complaints about more and more bureaucratic ‘red tape’ at German universities – in particular as a consequence of New Public Management (De Boer et al. 2007; Schimank and Lange 2009; Enders et al. 2015). This has made universities part of the ‘audit society’ (Power 1997), which also contributes to reduced time for research; but we lack reliable empirical data about it. Finally – as will be discussed next – research has become increasingly dependent on external funding; but the chances of success have decreased due to increased competition, so a growing proportion of research time is being spent on writing grant proposals, more and more of which are proving unsuccessful.

As a result, time for research has become increasingly scarce in the German science system as a whole; and accordingly, very time-consuming research projects can often no longer be undertaken, or have to be carried out ‘quick and dirty’, which means that the quality of the results is compromised. In particular, research topics that cannot be divided into feasible timeframes suffer under these conditions. The issue here is the ‘near decomposability’ (Simon 1962) of the required work tasks: are interim results already publishable, and how long can the interruption be between two research steps? To be sure, the prevention of those topics where only final results are of interest, as in research on Bose-Einstein condensates, is rarely intentional, but usually an unnoticed or accepted side effect of cost-cutting measures such as serving more students with the same teaching capacity.

This lever is more pervasive than regulation and discourse controls. However, there are also subjects with a low number of students in relation to the number of academic staff, so that the teaching is less time-consuming. As a restriction on research, time is usually less far-reaching than regulation or discourse controls because researchers can compensate for the increased time spent on other activities up to a certain point by extending their working hours – which many of them do to the detriment of their families and leisure activities. This lever is not thematically focused, but, as just mentioned, it affects certain types of research regardless of the topic.

The future trend in the German science system with regard to this lever remains to be seen. It could amount to either an increase or a decrease in the temporal threat to thematic research freedom. A decrease would occur if the demographically expected − by no means certain − decline in student numbers is not accompanied by a corresponding reduction in teaching staff; an increase in the potential threat would occur, on the other hand, if this reduction in staff due to financial constraints were to be even greater than the decline in student numbers, so that the supervision ratios would increase further despite a decrease of student numbers.

Lever 4: Financial Resources

The sources from which the research of a professor is financed often have a strong impact on his or her research agenda in various ways (Schimank and Hüther 2022; Hüther and Schimank 2023). Two kinds of sources must be distinguished in the German system, as in most other science systems: basic funding by the science ministries and external grants by funding agencies – in Germany, the Deutsche Forschungsgemeinschaft (DFG) as the most important one – or from industry.

In negotiations on the allocation of basic funding to a professorship, the university management can prevent certain research topics by not granting the necessary funds − for a specific laboratory, for example. Cuts to staff positions have the effect of reducing research time; if the staff members in question are specialized in certain methods, for example, research projects that rely on them are undercut − although such cuts are rarely thematically targeted.

In addition to basic funding, and increasingly replacing it, competitive external grant funding can provide incentives to work on specific issues with certain theoretical perspectives and methods. The most obvious case is industry funding, which is often narrowly earmarked for certain research questions and methods. The proportion of external funding that is awarded without thematic, theoretical, methods-related or other specifications, such as particular extra-scientific uses or cooperation in large consortia, has decreased in the German science system – even though demand is increasing at the same time, so that demand is being met less and less. In this way the ‘protected space’ of research is becoming ever narrower and scarcer. External funding is no longer just an incentive that professors can do without, but rather a must, in order to be able to conduct research at all, which is the case in large areas of the natural and technical sciences. This results in an opportunistic adaptation of research agendas to the wishes of funding sources − if someone’s own research agenda does not happen to fit exactly into the advertised funding conditions. This opportunism is the often-desperate attempt to avoid the ultimate elimination of one’s freedom of research: becoming unable to continue pursuing any kind of research at all.

To be sure, the financial situation of university research in Germany is still better than in most other European countries. The chances of being successful with grant applications are higher; and a considerable amount – though not the majority – of funding from the DFG is awarded entirely on the basis of the scientific quality of the applications, without any additional specifications. Nevertheless, being better off than researchers in other countries does not mean that German researchers are well off, and the trend towards an intensification of threats gives cause for concern.

These developments in financial resources are largely universal in the German system; only those who conduct extremely inexpensive armchair research are spared, e.g., philosophers or literary scholars. The resulting restrictions on thematic research freedom are far-reaching, and the long-term trend of their increase is likely to continue in view of the perception that the needs of other government departments, in particular the ministry of defence, are more urgent. The restrictions are not thematically targeted; however, there is a direction of impact such that research that best meets the formalized evaluation criteria of rankings and ratings – another component of NPM − is most likely to be funded.

Lever 5: Publication Opportunities

If research projects are carried out and interesting results are produced, these should be published with good visibility for the relevant scientific community − in a place decided by the researcher, and not by the university management, which has concluded a supposedly financially advantageous contract with a major publisher such as the ‘DEALs’ that a consortium of German research institutions made in recent years with Wiley, Springer and Elsevier. Even though the natural and engineering sciences have long since surrendered to the restrictions that the very few major publishers now impose on them in the form of digital platforms, the cultural and social sciences do not want to go down this path; but in their publishing landscape, the large publishing houses are also on the advance. Highly standardized journal articles are increasingly becoming the only currency that counts in rankings and ratings − a format that does not do justice to many topics and findings in the cultural and social sciences. Still, an opportunistic adaptation of questions, theoretical perspectives and methods to what can be expected to be published in article format is being imposed.

Across the board, publication opportunities are becoming increasingly dependent on the behaviour of major publishers. It is difficult to assess how far-reaching and intense the restrictions on thematic research freedom that can be brought about by this lever are and could become. It could be that the humanities and social sciences are more strongly affected than the natural sciences, which, at the moment, are somewhat but not very deeply impacted by these developments – with the exception of exorbitant journal prices now followed by similarly disproportionate article-processing charges for open access publications (for opposing assessments see Ziegler and Dirnagl 2024). More detailed research is required here, too. However, these restrictions are hardly thematically targeted. An increase in restrictions can be assumed as long as the major publishers are able to continue and further develop their successful business model. A recent move is the expansion of scientist tracking by collecting their data traces on the internet so that university management can be offered ‘transparent’ scientists who, as a consequence, have to respond even more opportunistically to management requests when choosing their topics (Gehring 2023).

Lever 6: Career Opportunities

Decisions about the careers of professors − recruitment to certain positions, negotiations about job specifications and personal income − provide university management with opportunities to influence the allocation of resources and time and thus the scope for autonomous choices of research topics and the subsequent work on them. Applicants for professorships are aware of this and base their research agenda on what ‘resonates’ with university management. This is especially true for those who are waiting for their desired first appointment. But if they want to make further progress afterwards, they are well advised to continue to look at how they can help university management to cut a fine figure in the competitive relationships in which they operate − in Germany, above all, by third-party funding successes and, best of all, with large collaborative research alliances.

Since the reform of German professors’ salaries in 2002, career decisions are also more relevant to income than before. In short, the fixed salary of professors appointed after that year was cut by around a third; and this third has since been awarded as a performance bonus − if the performance is appropriate. The university management, who have to decide on this, are primarily orientated towards research performance − in particular the acquisition of third-party funding and publications in peer-reviewed international journals. This is clearly perceived by many as pressure to take these criteria into account in their own research agenda, at least to the extent that they can successively earn a ‘status-appropriate’ salary through performance bonuses. Salary is therefore not only about a higher standard of living, but also about the associated symbolic recognition of someone’s willingness and ability to perform (Janßen et al. 2021).

All in all, this means that career opportunities are also focused on opportunism, which from a certain point onwards creates a tension with the preservation of one’s own thematic freedom of research. Restrictions on thematic academic freedom due to career decisions are widespread and far-reaching as well as thematically unfocused, and they will continue to increase − unless the demand for professorships decreases rapidly because fewer and fewer people want to embark upon the long, precarious path to this position and the uncertain chances of success and, instead, opt for better opportunities for themselves in other occupational fields.

Lever 7: Reputation

A professor’s research activities are ultimately influenced − and this is the generalized symbolic recognition of one’s own scientific achievements − by the acknowledgement of scientific reputation by the scientific community. A scientist gains an impression of his or her own reputation − as the core of personal self-esteem in this role − from many different kinds of communicative signals. The most important ones are successes in the placement of one’s own publications and their subsequent citation, the amount of external funding acquired and calls to professorships.

On the one hand, reputation results from the interaction of the six other levers that contour the room to manoeuvre for research activities and the associated acquisition of reputation; on the other hand, the acquired reputation in turn has an effect on how an individual is able to use and perhaps expand their room to manoeuvre for the continuation of their research agenda.

Reputation can be converted into better career and publication opportunities as well as more financial resources and time. This is the well-known ‘credibility cycle’ (Latour and Woolgar 1979) which has always operated in modern science. But under conditions of increased scarcity of these opportunities and resources, a researcher’s reputation takes on a completely different meaning. In earlier times, reputation primarily was a guide for other academics as readers as to whose publications they should pay attention to in order to be as well informed as possible for their own research; nowadays, reputation is above all else a signal to funding organizations and university management as to whom they should award jobs and money (Schimank 2010). The need of those who want to remain capable of doing research and becoming professors to build up reputations as researchers that − in competition with others − comes as close as possible to the criteria of university management and funding organizations is correspondingly high.

Like finances and career opportunities, this lever has a widespread, far-reaching and increasingly restrictive effect on thematic freedom of research, even if it is not thematically targeted. In other words, the current tendency of German university management to present as many large collaborative research projects as possible – in particular, clusters of excellence − only has unspecific effects on the choice of theories, methods and topics; but this can be just as powerful as direct specifications or exclusions.

Overall Picture

This brief sketch of the external production factors and conditions of scientific knowledge production can be recapitulated in two statements about the potential threat that each of the levers poses to a professor’s freedom to choose his or her own research agenda:

• Four of the seven levers do not currently have a very high potential to jeopardize the freedom of choice of topics for German professors − which of course does not rule out the possibility that individual events could mean a very large reduction in the freedom of research for those directly affected. These four levers are regulation, discourse controls, time, and publication opportunities. However, the future intensity of threats to thematic freedom of research may increase.

• Greater actual – and also potential future − threats to professors’ freedom of choice in research come from the other three levers, especially in interaction with each other − although all three are not thematically targeted: finances, career opportunities, and reputation. In view of the public and media discussions about possible threats to academic freedom in Germany, this means that some parameters − in particular discourse control − tend to be overestimated, while others, such as financial resources and reputation, are underestimated.

The extent to which these levers actually affect the ‘protected space’ of professors is determined by certain intervening variables in each individual case: above all, the discipline or subject area, and the career status of the professor (Janßen et al. 2021). Furthermore, national science systems differ in terms of the exact nature of these external constraints, and constraints can also change considerably over time within a particular science system – as they did in the German case.

Endangered Freedom of Research under Multiple Constraints

A final consideration is the question of whether these assessments are still too simple in one crucial respect. Threats to thematic freedom of research can be very clear in nature: when specific settings of one lever − or combinations of specific settings of several levers − produce specific effects. However, this laboratory situation, in which all other factors are constant, hardly ever exists in reality. On the contrary, a wide variety of constraint setters synchronously and asynchronously turn the various levers in the most diverse directions. Under these circumstances of not one overarching threat but multiple threats, it is extremely difficult to make a sociological assessment of the sum vector of these multiple influences and effects on thematic research freedom; even a subsequent reconstruction of what happened and why is rarely unambiguous. What is much more important, however, is that those who set certain constraints with certain intended effects quickly poke around in their uncertainty to determine whether their plans are working – although they often do not want to acknowledge this. In particular, it is seldom realized that a measure relating to a certain set of levers, which on its own would not have entailed any restrictions on thematic research freedom, can very well have a restrictive effect when it occurs in combination with measures that have a simultaneous effect on other levers.

Such an interplay of effects is not the exception but rather the normal case when − contrary to what Simon assumes − there is not just a single constraint setter at work, but a number of them, some of whom know and try to take into consideration what the others are doing or planning to do, but many of whom do not. It is then quite likely that something I would like to call the Davey Moore effect will occur − after an early topical song by Bob Dylan (1964) about the death of a professional boxer in the ring. ‘Who killed Davey Moore?/Why an’ what’s the reason for?’ is the refrain. In one verse after another, the people involved in the incident have their say: the referee, the spectators, the sports bettors, the manager, the sports journalists and the opponent. They all have plausible reasons for categorically denying any guilt, so that the main question of the song remains just as unanswered at the end as it was at the beginning. The implicit message of the song is nevertheless very clear: It is the constellation of the interplay among the actors that caused the boxer’s death.

It would be worth examining whether the constellation of those who − with regard to the thematic freedom of research of German professors − as ministries, university management, funding organizations, industry, publishers, the public, etc., turn the seven levers without coordination among themselves, resembles the constellation described by Dylan. In any case, what fits are the excuses made by all those involved: ‘What I am doing is not a restriction on the freedom of research’. This apologetic rhetoric is meanwhile suspicious enough to justify a closer look. It may be possible that an even greater danger to individual freedom of research emanates from such a constellation and not from individual constraint setters.