The aim of this paper is to show that Philip Kitcher’s ideal of well-ordered science has limitations when trying to implement it as a guide for current commercialized science. After describing Kitcher’s ideal, I give a critical assessment of the approach, uncovering three limitations. I then explain some of the major changes that science organization has undergone in recent decades and how they affect scientific research. I argue that the limitations of Kitcher’s approach follow from not taking these changes sufficiently into account.
In recent decades, philosophers of science have become increasingly concerned with the social dimensions of scientific knowledge. Philosophers such as Helen Longino (1990, 2002a), Philip Kitcher (2001, 2011), Miriam Solomon (2002), Heather Douglas (2009), and Janet Kourany (2010) have sought to incorporate the social aspects of science, while retaining the normative commitments of philosophy of science. Some of the major theoretical approaches in social epistemology of science, however, tend to ignore or underestimate the role that the current state of science organization plays in the production of scientific knowledge.1 The world has changed significantly in recent decades, and science has changed with it. The present regime of global commercialization and privatization of research has changed the way scientific knowledge is produced, raising new epistemological problems for the philosophy of science.
Within social epistemology of science, Philip Kitcher’s work on the relation between science and democracy has attracted a great deal of attention. Both in Science, Truth and Democracy (2001) and the more recent Science in a Democratic Society (2011), Kitcher examines the role of science in democratic societies, and builds an ideal account of the organization of science that he calls the ideal of well-ordered science. Kitcher is one of the few philosophers who have addressed directly the task of building a positive program in philosophy of science in view of the challenges posed by its social organization. Kitcher is thus a key representative of the type of socially concerned philosophy of science that I want to address.
In the paper, I examine Kitcher’s ideal of well-ordered science. My aim is to show that the ideal has limitations when trying to implement it as a guide for current research, given the changes that the commercialization and privatization of science have brought about. The paper is divided in three sections. First, I briefly present Kitcher’s ideal of well-ordered science, emphasizing its procedural account of deliberation and the system of public knowledge in which it is embedded. Second, I give a critical assessment of the approach, uncovering three limitations. Third, I argue that the limitations follow from not taking sufficiently into account the current social organization of science. To do so, I also explain some of the major changes that science organization has undergone in recent decades. I conclude with some suggestions regarding the direction that well-ordered science, and social epistemology of science more generally, could take to overcome such limitations.
2. The Ideal of Well-Ordered Science
In his book The Advancement of Science (1993), Kitcher faced the challenge of providing a social epistemology of science able to identify “the properties of epistemically well-designed social systems, that is, to specify the conditions under which a group of individuals, operating according to various rules for modifying their individual practices, succeed, through their interactions, in generating a progressive sequence of consensus practices” (1993, p. 303). Without compromising the traditional goals of philosophy of science (e.g., the aim for truth, justification, and objectivity), his book became an important example of how one could think about the social dimensions of science in normative terms. In particular, Kitcher introduced resources from economic theory, i.e., rational choice models, to try to capture epistemological efficiency in terms similar to economic efficiency.2 In very broad terms, Kitcher tried to show that many of the motives that one could consider detrimental for decision making in science (e.g., desire for credit) can actually play a constructive role in the process of knowledge production.
In Science, Truth, and Democracy (2001), Kitcher moves away from questions pertaining to the internal organization of the scientific community—i.e., what he calls the division of cognitive labor—to questions pertaining to the broad organization of science within a democratic society—i.e., the division of epistemic labor. The book poses one fundamental question: “what is the collective good that we want inquiry to promote?” (2001, p. 145), which Kitcher tries to answer with his ideal of well-ordered science. In general, the ideal is a procedural account of deliberation in which a representative group of tutored citizens discusses and decides which lines of research scientists ought to pursue given what is most convenient for society at large.
According to the ideal of well-ordered science, scientific inquiry is divided into three phases. The first phase is dedicated to setting the agenda, i.e., to allocating resources among different possible scientific research projects. During the second phase the chosen research projects are pursued in the most efficient way within the scientific community. And, finally, results are translated into practical consequences during the third phase. For decisions to be made properly during the agenda-setting phase, Kitcher envisages a process of ideal deliberation in which citizens with different initial preferences come together to discuss the available research possibilities and make the relevant decisions. Given that they are not scientific experts, but common citizens, they ought to be tutored to understand the epistemic and practical significance of the proposed research. As deliberation advances, the group establishes its priorities concerning the outcomes to which inquiry might prove relevant, and then the group assesses the possibilities of particular scientific ventures delivering what they collectively want. A disinterested arbitrator uses the information about probabilities together with the collective wish list to draw up possible agendas for inquiry. Finally, the group determines appropriate budgets and research agendas (2001, 118–121).
Initially, Kitcher does not give a detailed account of the research phase. He emphasizes the importance of pursuing scientific projects with an efficient distribution of cognitive labor though, suggesting that he is drawing from the account of the scientific process given in The Advancement of Science (1993). With regard to the application phase, Kitcher suggests a similar deliberation procedure to determine how to best apply scientific results. A general picture of the whole process is summarized in Figure 1.
Through his ideal of well-ordered science, Kitcher favors a view of scientific organization in which decisions regarding research agendas and application of results are made by a group of tutored citizens that is familiar enough with scientific process and diverse enough to be representative of society. Ideal deliberations, however, leave the research process untouched: deliberators may choose pertinent lines of research over others and proper application of scientific results, but they have no say during the research phase, where scientific inquiry takes place (i.e., where hypotheses are drawn, experiments designed, data collected, results interpreted, etc.).
Science, Truth, and Democracy (2001) examined the problem of incorporating socially relevant goals into the process of scientific inquiry. In Science in a Democratic Society (2011), Kitcher expands this analysis in terms of the relation between a system of public knowledge and a democratic society: “We urgently need a theory of the place of Science in a democratic society—or, if you like, of the ways in which a system of public knowledge should be shaped to promote democratic ideals” (2011, p. 25). He divides the general structure of public knowledge into four processes: (i) investigation, or the process in which lines of research are determined and executed; (ii) submission, or the process through which people become qualified to contribute to the “public depository”; (iii) certification, or the process through which the contributions get accepted into the public depository; and (iv) transmission, or the process through which parts of the pubic depository become available to the larger public (2011, p. 91).
At this point, Kitcher’s analysis of the structure of public knowledge connects with his ideal of well-ordered science as developed in Science, Truth, and Democracy. The agenda-setting phase of the previous account corresponds, broadly speaking, with the investigation process; as does the previous application phase with part of the transmission process. Kitcher then proceeds to expand his account to the processes of submission and certification, and also broadens the scope of the transmission process.
Kitcher then argues for an inclusion of ideal deliberations in all four processes. In sum, Science in a Democratic Society expands the account of an ideal division of epistemic labor in society, the ideal of well-ordered science, started in Science, Truth, and Democracy. Figure 2 illustrates the expansion of the ideal to the four processes involved in the structure of public knowledge.
The ideal of well-ordered science attempts to organize only a fraction of scientific research, leaving untouched the organization of practices within the scientific community, i.e., the division of cognitive labor.3 Accordingly, the ideal is a normative account of science organization, but only a partial one. As stated in these two books, the ideal of well-ordered science aims at an appropriate division of epistemic labor within democracies, in which science is guaranteed to be responsive to social needs through processes of ideal deliberations.
3. Some Limitations of Well-Ordered Science
In this section, I present three limitations of the ideal of well-ordered science that emerge when trying to implement the ideal as a guide for current scientific research.
3.1 On the Ideal Character of Well-Ordered Science
Kitcher acknowledges the limitations that an ideal account of scientific organization might have: “Well-ordered science is an ideal. It may seem a utopian fantasy, the sort of thing that may figure in philosophical discussions but that has little place in a realistic account of the science.” However, he clearly thinks that ideals are worth conceiving, as long as they are in principle realizable: “[M]eaningful ideals are those for which we can envisage a path that might lead us toward them, and a philosopher who proposes an ideal should be able to point to the initial steps” (2011, p.125, emphasis his). Even though he attempts to identify some steps to put forward his ideal, as well as some of the difficulties that it might encounter in practice (§20), the question of the plausibility of well-ordered science remains open.
The ideal deliberation that Kitcher envisages for all four processes of well-ordered science is at the core of Kitcher’s normative account of the social organization of science. For instance, the ideal scenario of deliberations including representatives from society at large, or the ideal assumption that citizens can be tutored on the epistemic and practical significance of scientific research without bias, make well-ordered science a clear attempt to think about what science organization should look like, without material constrains. Kitcher seems to acknowledge this point when he claims that “any actual conversation of this type is impossible” (2011, p. 115, emphasis his).4 He complicates the picture even further when he later acknowledges that “nobody can predict how the ideal conversation would come to conclusion,” (2011, p. 124) making it impossible not only to set the initial conditions for ideal deliberation, but also to conceive its result.5 Given that we cannot suggest what the outcome of ideal deliberations would be (not even in ideal terms) and such a conversation in real life is impossible, what can we really learn from the ideal? By claiming that the conclusion of ideal deliberations are impossible to predict, Kitcher fails to point at “the initial steps” that he considers crucial for connecting the ideal to reality.6 I consider this the first limitation of well-ordered science.
3.2. On the Public Character of Scientific Knowledge
A second limitation of the ideal is related to the framework it presupposes. Kitcher argues that the ideal will organize our system of public knowledge better so that it serves our democratic goals. The ideal, however, assumes that knowledge should be part of a public system. As Kitcher acknowledges “the chief aim of this book is to offer an account of how this broad system of public knowledge, with Science as prominent part, functions to promote our [democratic] values […].” (2011, p. 86)
Some of the main characteristics of this system of public knowledge can be found scattered throughout the book. Perhaps a central one is the idea that science, or knowledge in general, is a public good: “Well-ordered science treats knowledge as public property” (2011, p. 241). When confronted with issues regarding privately owned knowledge, Kitcher favors some kind of regulation or, more extremely, simply “not to allow private property in those instances” (2011, p. 241). Accordingly, he criticizes the strong intellectual property rights that guide science today: “We live in a world with a ramshackle system of intellectual property […]. It urgently needs rethinking” (2011, p. 242), and instead argues in favor of basic research (2011, p. 124). In sum, he is critical of the commercialization of science: “Privatization of scientific research will probably make matters worse” (2011, p. 126, emphasis his). The ideal of well-ordered science presupposes a system of public funding for scientific research in which ideal deliberators have the power of deciding which research is significant enough to be pursued. The idea of private funding of science is not even considered.7
Some of the main aspects of the system of public knowledge that the ideal of well-ordered science envisages are in strong opposition to the current system of science organization. As it will become clear in the next section, nowadays knowledge is increasingly privately owned, support for basic research has decreased and is being isolated to universities, corporations are slowly becoming the main supporters of R&D, and the state’s power to control research is decreasing as corporations move overseas. The privatization scheme that moves scientific research today has little resemblance with the system of public knowledge that Kitcher envisages, in fact running in the opposite direction. Thus, there is an important gap between Kitcher’s ideal and the regime of science organization we live in. For the ideal of well-ordered science to serve as a guide for the organization of current science, this gap needs to be acknowledged and bridged. To do so, one would need to acknowledge the broad social circumstances (i.e., the historical, economic, and political changes of recent decades) that led science to its current state, and then identify plausible ways in which current social arrangements could be changed to achieve the ideal. Kitcher, however, does not provide such path towards the system of public knowledge.
3.3. On “Black-Boxing” the Internal Organization of Scientific Research
Both in Science, Truth and Democracy (2001) and Science in a Democratic Society (2011), Kitcher is concerned with questions regarding the distribution of epistemic knowledge and the proper place of science in democracy. These are in fact crucial questions for a philosopher of science with social concerns. However, one should notice that his ideal of well-ordered science also presupposes, without really discussing, a view regarding the division of cognitive labor, or what is normally called the internal organization of science. Kitcher’s latest books “black-box” this internal organization, focusing on questions about the appropriate goals for research in democracy, the democratic application of scientific knowledge, and the distribution of knowledge. I consider Kitcher’s strategy of “black-boxing” the internal organization of scientific research a third limitation of his ideal.
Only at the end of Science in a Democratic Society (2011, §35) does Kitcher provide a brief account of the internal organization of scientific research, defending a threefold strategy that includes cultivating team players, instituting norms, and taking advantage of market forces (all coherent with his previous account ). Even though this is only a sketch, we can notice a couple of things. First, the internal organization of scientific research is not subject to the same kind of democratic process as the other phases. Second, Kitcher’s remarks on the internal organization of scientific labor are clearly not central to his project in Science in a Democratic Society. Scientific research at its core, which includes the process of framing hypotheses, designing and conducting experiments, sampling and interpreting data, sharing results with peers and submitting them to an evaluation process, all seems to disappear in the background or taken as given.
Kitcher has dealt with the problems arising from the division of cognitive labor in the past, especially in The Advancement of Science (1993), where he presents a positive conception of the scientific process, showing confidence in scientists’ abilities to attain good results. In his latest work he maintains this view, privileging science as a social institution, claiming it should serve as a model for the broad system of public knowledge: “Science has become embedded in our public knowledge system, not only an important segment of it but, through the prestige acquired by its most spectacular successes, a part that sets standards to which other forms of inquiry should aspire” (2011, p. 100).
Kitcher assumes that the scientific process works appropriately, locating the problems for science organization, not in the way scientific knowledge is produced, but in the way lines of research are defined and scientific knowledge is applied, certified, and transmitted. The picture presupposes a distinction between the internal and external aspects of science organization. The internal aspects are constrained to the space traditionally assigned to knowledge production, and presuppose a “narrow” view of scientific research, which only takes into account the social aspects of the scientific community. The external aspects of science organization are located in the space of interaction between science and society, in which problems of agenda-setting, certification, transmission, etc. are negotiated. These external aspects may be counted as part of scientific research under a “broad” conception of the term.8Figure 3 attempts to capture this picture.
If this is an accurate picture, Kitcher assumes one can examine the internal or the external aspects of science organization separately. However, this assumption closes the possibility of examining the relation between these aspects, which in turn obscures the epistemological problems emerging from these interconnections. As I argue in the next section, the current commercialization and privatization of science affects the scientific process “narrowly” conceived in ways that the ideal of well-ordered science is unable to address.
4. The Commercialization and Privatization of Scientific Research
In the previous sections, I showed how Kitcher’s ideal of well-ordered science attempts to provide a normative framework for the organization of epistemic labor in democratic societies, and I identified some of the limitations that the ideal encounters when trying to use it as a guide for current research. In the last section, however, I assumed that the current organization of science has certain characteristics which lead to certain epistemological problems that Kitcher’s ideal might not be able to address. In this section, I provide some evidence for the claim that the organization of scientific research has undergone important changes in the last decades, leading to a new regime of commercialization and privatization of science.9 My aim here is to buttress the claims I made with respect to the limitations of well-ordered science in the previous section by emphasizing the extent to which the commercialization and privatization of science—normally understood as external factors—have affected the internal organization of research, and how far are these from the system of public knowledge that the ideal of well-ordered science presupposes.
4.1. Major Changes in Science Organization
With the end of the Cold War, Americans also witnessed the end of a key institution in scientific research: the in-house corporate lab.10 As a number of scholars have identified, a general concern with a loss of US competitiveness during the late 1970s and early 1980 led to major changes in industrial R&D, which eventually culminated with the destruction of the in-house laboratory (see, e.g., Hunt 1999; Hart 2001; Coriat and Orsi 2002; Mirowski 2011). The retreat of the military from research funding, the rigid hierarchical organization characteristic of US corporations, and an enrollment crisis in higher education, were among the main factors contributing to the perception that American competitiveness was in decline (Mirowski and Sent 2008).11 The in-house corporate lab was then replaced, through a series of legislative and structural transformations, with more flexible research units.
The strengthening of intellectual property (IP) legislation can be considered one of the crucial changes in the US R&D model. Legislative acts, such as the Bayh-Dole Act of 1980, restructured the boundaries between public and private research, allowing commercial exploitation of government funded research, while protecting private R&D (Barben, 2007, 62).12 The Bayh-Dole Act, in particular, granted universities IP rights over federally funded research, opening the door for the university to profit from commercial R&D.
In addition to strong IP rights, the new model of science organization also procured weaker antitrust laws, transnational trade agreements, outsourcing and off-shoring of research to low-wage countries with weaker regulatory environments, the availability of cheap and reliable communication technologies, and the creation of new research structures, such as Contract Research Organizations (Mirowski 2011, pp. 94–95, 122). The combination of these conditions has led to the current state of science organization, characterized in broad terms by the commercialization of scientific research.
According to the Science and Engineering Indicators 2012, gathered by the National Science Board (NSB) under the auspice of the NSF,13 the corporate sector is by far the largest performer and funder of R&D in the US. In 2009, the corporate sector performed 71% of the US total of R&D support, while funding 62%. The 2012 indicators also show a decreasing trend in federal funding of R&D. The federal government was the largest funding source of R&D in 1967, constituting 67% of the national total. In 1979, however, that percentage had already been cut by half, reaching 25% in 2000 (NSB 2012, pp. 4–13). Looking at the trend in funding sources for R&D since 1953 (Figure 4), it becomes clear that federal funding of scientific research has been replaced by the corporate sector in the past decades, supporting the idea that a new regime of science organization, led by the commercialization and privatization of research, has been emerging since the late 1970s and early 1980s.
Moreover, the commercialization of science is evident not only in the percentage of corporate funding and performance of R&D, but also in the focus on applied research and development, and the move away from basic research. Although the academic sector remains the first performer of basic research (53% of the total in 2009), this type of research constitutes only the 19% of R&D performed in the US. The corporate sector has become the main performer and funder of both applied research, which constitutes 18% of total R&D performance, and development, which is by far the largest component of R&D in the US with 63% of total R&D performance (2012, p. 4-4). As one would expect, the focus on basic research proper of the Cold War has been replaced by applied R&D, as the private sector plays a bigger role as funder and performer of scientific research.14
4.2. How Commercialization Affects Science “Internally”
Granted that there have been important structural changes in science organization during the last decades, I argue that these changes are affecting not only the external organization of research (i.e., lines of research funded, type of applications, etc.), but also the more internal aspects of research (i.e., peer review, experiment design, data collection and interpretation, etc.). Let me give two examples to illustrate this point: the case of publication planning firms and the “funding effect.”
In “Ghosts in the Machine” (2009), Sergio Sismondo carefully describes the central aspects of ‘publication planning’, a practice that has become customary for pharmaceutical companies to ensure that their scientific research gains credibility through publications in top peer reviewed journals. Sismondo summarizes the process as follows:
Most sponsored clinical trial research is handled by contract research organizations (CROs), the data they produce is typically analyzed by pharmaceutical company statisticians, papers are written by medical writers, and the whole process is guided and shepherded through to publication by planners and planning teams […]. To gain the most commercial value from research, the papers publicizing it are written under the names of independent medical researchers […].(Sismondo 2009, p. 172)
Another illustrative example of how the commercialization and privatization of science are affecting science organization “narrowly” conceived is the “funding effect.” David Michaels explains the phenomenon as follows:
When a scientist (or the scientist’s employer) is hired by a firm with a financial interest in the outcome, the likelihood that the result of that study will be favorable to that firm is dramatically increased. This close correlation between the results desired by a study’s funders and those reported by the researchers we now call the “funding effect.”(Michaels 2008, p. 143)
The financial conflict of interest that lies behind the “funding effect” is increasing as more and more scientific research moves to the private sector.15 In this way, the commercialization and privatization of science are affecting the way in which scientists conduct their research. In Michaels’ words, “The separation between academic science and the business world is disappearing. As a result, the whole culture of science is changing rapidly” (2008, p. 243).
4.3. Philosophy of Science and the Challenge of Commercialization
After explaining the main changes that the organization of science has undergone in the past decades, and giving some examples of how these changes affect the internal aspects of scientific research, we can now give further support to some of the claims made in the previous section. First, we can say that, despite the many epistemological and ethical advantages that a system of public knowledge might have, the organization of scientific knowledge is currently moving away from such a system. Without a plausible connection between the ideal of well-ordered science and the current organization of science, i.e., without drawing a path that can take us from where we are today to where the ideal prescribes we ought to be, it is not clear how well-ordered science contributes to addressing the epistemological problems emerging from commercialized science.
Second, once we have identified some of the ways in which the commercialization of science is affecting the internal aspects of scientific research (as is the case with publication planning and the “funding effect”), we can understand better why we cannot deal with the external organization of science without dealing with the internal organization as well. Treating them as separate spheres might contribute to obscuring the problems that arise when the broader social framework starts affecting the ways scientists conduct research. For instance, the commercialization of science—which in principle could be understood as a broader structural framework for science organization, affecting only the external aspects of research—turns out to be much more pervasive, with consequences for the most internal aspects of scientific inquiry. Normative accounts of the social organization of science should, hence, take into account the interconnections between the “narrow” and the “broad” conceptions of scientific research.
The question now arises: How can philosophy of science provide a normative account of science organization that recognizes the ongoing commercialization of research? Two possible solutions stem from my previous analysis. First, any normative account of scientific knowledge and its social organization that aims to identify and evaluate the problems of commercialization has to take into account the political, historical, and economic changes that have led to that commercialization. And even though normative accounts present ideal scenarios—stating how science organization ought to be, and not how it actually is—they also vary in degree with respect to their departure from the current state of affairs. Given the limitations that the detachment from current science organization presents for the ideal of well-ordered science, I suggest that an account closer to current science—i.e., an account that in formulating a normative stance takes into account the current state of science organization—would be better-off dealing with the challenges of commercialization.
In connection to the previous point, my analysis also suggests a reevaluation of the distinction between the internal and the external aspects of science organization. The challenges that commercialization poses for the production of scientific knowledge, as in the case of publication planning and the funding effect, emerge in new trading zones between what have been traditionally considered “internal” and “external” aspects of science organization. The idea that philosophers of science can study the internal organization of science without worrying about its external aspects, or vice versa, is no longer tenable. One possibility is to qualify the distinction so that it includes new type of interactions. Another one is, of course, getting rid of the internal/external distinction altogether, in which case we would need a more complex account of the factors closely involved in the production of scientific knowledge. Either way, we need a more sophisticated conceptual tool for understanding the relation between private industry and scientific research than the traditional internal/external distinction allows.
For the past decade, Kitcher has been concerned with one of the most pressing questions for philosophers of science today: what is the role of science in democracy? His ideal of well-ordered science responds to this question, favoring a system of public knowledge in which tutored citizens have the power to make socially relevant decisions with respect to scientific research. The ideal of well-ordered science faces at least three limitations for being used as a guide for current research. First, Kitcher does not provide a clear path for working towards the ideal; second, the ideal is framed in terms of a system of public knowledge which is very distant from the current privatization of research; and third, Kitcher’s strategy of “black-boxing” the internal aspects of research undermines the epistemological problems that might arise from the interaction between the internal and the external aspects.
I have suggested two ways in which philosophy of science could move to overcome the limitations: to encourage normative accounts of the production of scientific knowledge that pay attention to the current state of affairs in scientific research and to reevaluate the distinction between the internal and external aspects of science organization. Moving in these directions, I suggest, would help philosophy of science identify and evaluate the problems that this particular organization of science poses for scientific knowledge.
Not all philosophers of science have been indifferent to the situation. See, for example, Justin Biddle’s work on pharmaceutical research (2007) and intellectual property (2012), Kevin Elliot’s work on conflict of interest (2008), Sheldon Krimsky’s research on the privatization of science (2003), Rebecca Kukla’s work on ghostwriting (2012), and Güral Irzik’s work on the impact of commecialization on academic science (2013).
The reader has to wait until one of the last chapters of Science in a Democratic Society (2011) to find a brief reflection on the internal organization of the community of inquirers. In §35, Kitcher argues for a “polythematic strategy for creating and maintaining valuable features of a scientific community,” (p. 210) which includes three mechanisms: cultivating team players, instituting norms, and taking advantage of market forces. We will come back to this point later.
In fact, he acknowledges that the impossibility of putting the ideal to practice “may incline you to think it absurd to approach scientific significance as I have done,” but then he immediately tries to explain what are his motivations behind it: “Understanding an ideal, however, can sometimes help us to improve our practice, and this is the hope of my proposal” (Kitcher 2011, pp. 115–116). However, if Kitcher cannot draw a clear connection between the ideal and actual practices, the impossibility of actualizing the ideal becomes an important shortcoming of his account.
In his concluding remarks, Kitcher emphasizes again that the ideal of well-ordered science offers “only a blueprint for a conversation, without any ability to predict how that conversation should turn out” (2011, p. 248). Apparently, Kitcher is confident that sketching the steps for ideal deliberators to follow would be enough to guide actual deliberators, despite the impossibility of knowing the results of those deliberations. Although this might be plausible, giving a purely procedural account of research faces further challenges under the current regime of science organization, as we will see in the following sections.
Perhaps a more viable strategy in this case would be starting from “the bottom-up.” As Heather Douglas puts it: “If not even the thought experiment can be conducted, then surely we need to try actual conversation, even if among non-ideal conversationalists” (2013, p. 3). Douglas (2013) raises a similar critique to the one developed here with respect to the ideal character of well-ordered science.
In 2002, Kitcher and Longino published a series of articles in Philosophy of Science discussing each other’s views. Here, Longino recognizes as a fundamental problem of Kitcher’s epistemology its commitment to public funded science and its inappropriateness for current commercialized science: “Now that more and more research, even in universities, is funded by the private sector, it is not clear we have a collective research effort that can be evaluated by reference to well-ordered science, or on what grounds privatized research could be brought under the umbrella of well-orderdness.” (Longino 2002b, p. 567)
One might also attempt an economic interpretation of well-ordered science, distinguishing between the production of scientific knowledge, i.e., the “internal” process of knowledge formation, and the distribution of that knowledge, i.e., the way knowledge is finally allocated. One could then question the separation of the process of production from the process of distribution that might be suggested by Kitcher’s analysis. Although the concept of distribution might be useful to understand Kitcher’s transmission process, I prefer to give an interpretation in terms of the internal/external distinction because it seems to map better on Kitcher’s account of the four processes of public knowledge, which are not limited to transmission, but also include investigation, submission, and certification. I thank one of the anonymous reviewers for suggesting this point.
The use of the concept of “regimes” to describe the major transformations in science organization from the Cold War era to the advent of the “globalized world” was first introduced by a group of European historians of science, who tried to understand this transformation in terms of a change in knowledge regimes (Coriat and Orsi 2002; Coriat, et al., 2003; Pestre 2003a, 2003b; Barben 2007). Drawing upon this literature, Mirowski (2004, 2011) and Mirowski and Sent (2008) have expanded the regime analysis to the case of science organization in the US during the twentieth century.
In this section, I follow Mirowski (2011) in his account of the major changes of science organization in the US during the twentieth century.
You can find an online version of the report at http://www.nsf.gov/statistics/seind12/
The case of the European Union is not very different. According to the European Commission’s 2011 statistics, “The business enterprise sector (BES) was the largest of the four main institutional sectors of R&D performance in 2008, accounting for 63.9% of EU-27 R&D expenditure,” as well as the largest funding sector, financing 55.0 % of R&D, whereas only 33.5 % was financed by the national governments (Eurostat 2011, p. 29).
Other scholars have identified financial conflict of interest as the central problem in commercialized science, e.g., Bekelman, et al.2003; Krimsky 2003; Hardell, et al. 2006. Despite its pervasiveness, we have strong reasons to doubt the viability and effectiveness of the main strategies (disclosure, management or elimination) available to deal with financial conflict of interest [Elliott 2008]).
I am indebted to many for their comments on this paper, especially to Janet Kourany, who is directing my doctoral dissertation from which this paper stems. Thanks also to Phil Mirowski, Don Howard, and Anjan Chakravartty for their valuable comments on earlier chapter drafts. An earlier version of this paper was presented at the 2013 EPSA meeting in Helsinki, Finland, August 28–31. Thanks to the European Philosophy of Science Association and the University of Helsinki for organizing the conference, and to Martin Kusch and Alison Wylie for their useful comments. I would also like to thank Dan Hicks, as well as two anonymous reviewers for their careful reading of the paper and their very useful comments.