#common goods

1. What is (Scientific) Knowledge?

During the past years, many critical studies of the commodification of science have been published (see, e.g. Bok 2003; Krimsky 2003; Resnik 2007; Radder 2010a). These criticisms naturally lead to the question of a more developed account of alternatives to commodified science. This is the central question addressed in this article. The question will be specified in terms of the notion of a common (or public) good: Can and should science, especially scientific knowledge, be seen as a common good?

In the classical economic literature on the public good of scientific knowledge, the idea of a public (or common) good has received much attention. Remarkably enough, in this literature there has not been a similar focus on the notion of scientific knowledge. But an appropriate discussion of our subject also requires a thorough account of scientific knowledge. Therefore, I will first address this issue. I start with a review and assessment of a theory of knowledge, including scientific knowledge, which is quite dominant in current analytic philosophy. This is the view of knowledge as justified true belief (JTB). My main sources are the review article by Steup ([2005] 2014) in the authoritative Stanford Encyclopedia of Philosophy and the contemporary introduction to epistemology by two prominent epistemologists, Goldman and McGrath (2015). I think it is reasonable to assume that these sources are sufficiently representative of the views of many analytical epistemologists. For my purposes in this article, it is not necessary to discuss all kinds of details. A focus on the general characteristics of the JTB theory suffices.

The reasons to start with this theory are, first, that it includes an explicit and coherent account of what knowledge is. Furthermore, although it focuses almost exclusively on everyday knowledge, it does claim to be applicable to scientific knowledge as well. Finally, philosophy of science itself includes far less explicit or coherent discussion of the nature of scientific knowledge. Often, the presupposed concept of knowledge is left implicit or it remains relatively undeveloped. Instead of providing a developed account of scientific knowledge as such, usually philosophers of science immediately proceed to more specific questions, such as those about the differences between empirical and theoretical or between descriptive and explanatory knowledge.

The basic idea of the JTB theory is that a subject S knows that p, if and only if S believes p, p is true and S is justified in believing p. Here, S is an individual person and p is a constative proposition (that describes some fact or arrangement of facts). For example, Jasper knows that it is 4 pm now, if he believes the proposition that it is 4 pm now, if it is in fact 4 pm now, and if he believes that it is 4 pm now because he has just checked his watch. This justification in terms of a ‘because’ can be explained in two ways: it is either based on empirical evidence possessed by S or on the fact that the belief has resulted from a reliable cognitive procedure. Possible sources of the justificatory evidence or procedure may be perception, introspection, memory, reason or testimony.1 An important overall goal of the JTB theory is the refutation of global skepticism, that is, the philosophical view that we cannot have knowledge at all. One form of skeptical argument says that there are important cases of incompatible propositions that may lay equal claims to truth and justification. A much-discussed example (to which I return in Section 1.1) is the incompatibility between the propositions ‘I am an ordinary human being’ and ‘I am a brain-in-a-vat’ (that is, I consist of merely a brain, but one that receives exactly the same input, works in exactly the same way as the brain of me as an ordinary human being, and so leads to exactly the same experiences).

1.1. Seven Basic Problems of the JTB theory

In spite of their currency, the central claims of the traditional JTB theory (Goldman and McGrath 2015, chap. 3) are rather questionable, whatever their further refinements or developments. In this section, I have a critical look at seven central claims of this theory. Based on these criticisms, Section 1.2 sketches an alternative theory of knowledge that deviates from the JTB theory in important respects.

(1) Traditional JTB theory is limited to propositional knowledge, to knowing that. This means that other forms of knowledge have been excluded from the beginning, which seems hard to justify. Especially important in the context of this article is know-how, the knowledge of how to do or make something (for instance, baking a cake, making an experiment work, or designing and manufacturing a technological device).

(2) One of the key notions of the theory is justification. In principle, it is correct that knowledge requires some form of justification (or confirmation, the term usually employed by philosophers of science). The JTB theory includes extensive discussion of what justification means and which sources may justify a belief. I cannot, and need not, review these debates here. One question, however, is relevant to the purposes of this article. Is it sufficient that there is a justification, which is available in the relevant epistemic community? Or should S be actually capable and in the right position to perform the justificatory procedure for the belief in question?

This distinction between a weak and a strong interpretation of S’s justification is particularly (though not exclusively!) pertinent in the case of testimony. Much, or even most, of what we know, we know indirectly, from others through different kinds of testimony. This is how most of us know that people have a liver or that now, in 2017, there is a terrible Civil War going on in Syria. Similarly, practicing science and generating scientific knowledge are highly dependent on the testimony of other scientists. Scientists constantly rely on a lot of knowledge (for instance, the knowledge of background theories or of the working of instruments) with which they themselves are not familiar and which they may not even understand.

Of the many issues prompted by the notion of epistemic testimony (Goldman and McGrath 2015, chap. 9), I will only discuss one. This is an issue that pertains to scientific, or more generally, to any kind of specialist knowledge. As a rule, this kind of knowledge is only understood by the members of a community of skilled practitioners. Consider the claim p that ‘Active medulloblastoma enhancers reveal subgroup-specific cellular origins’. This claim constitutes the title of a research article in a recent issue of the journal Nature. In many cases, individual people will have no understanding of this field of science, not merely in the sense that they lack the skills to judge the truth of this claim but in the stronger sense that they have no clue at all about the meaning of the terminology in question. Yet, they may well know that Nature has a high reputation and that published research articles in this journal have been thoroughly reviewed by experts. If this applies to S, he may rightly assume that p is true and that his belief of p is justified. Therefore, in this and similar cases, it seems that S knows that p without having any understanding of what p means.

In this way, the many cases of testimony concerning specialist knowledge seem to lead to a dilemma. On the one hand, if we only require that ‘somewhere’ a justification is available, the individuals S may possess a lot of knowledge, but this knowledge will be extremely superficial. On the other hand, if we require that the individuals S have to be capable to perform the justificatory procedure themselves, the knowledge of individuals will be substantive but also limited to those few areas with which they are thoroughly acquainted through experience and education.

The traditional JTB theory of knowledge seems to opt for the first approach (or, at least, not to support the second one). Thus, Goldman and McGrath (2015, 56) write:

It is important, here, that we understand justification in terms of reasonableness of belief, and not in some stronger sense, such as being able to justify your belief through convincing verbal argumentation.

As we have seen, this weak interpretation of justification includes cases where S does not even understand the proposition p. Such cases, however, do not allow the appropriate use of this knowledge ‘in reasoning about what to do and what is the case’, which is claimed to be one of the reasons for adding justification to mere true belief (Goldman and McGrath 2015, 56). For this reason, I think that we need both the strong and the weak interpretation of justification, as I will explain in Section 1.2.

(3) JTB theorists strongly rely on made-up and highly fictitious assumptions and thought experiments. An extreme example is the discussion of the above-mentioned brain-in-a-vat argument (Steup [2005] 2014, 20–28; Goldman and McGrath 2015, chap. 4). Its central premise is the skeptical claim: ‘I do not know that I am not a brain-in-a-vat’. The brain-in-a-vat hypothesis is, or can always be, construed in such a way that it states a difference (with the ordinary conception of human beings) which, by definition, does not make any difference. For this reason, it is not relevant to the issues under discussion in this article. Let me just note that the extensive discussion of hypotheses like the brain-in-a-vat shows that the possibility of a disembodied knower is taken very seriously. This might explain that, in this kind of traditional epistemology, the in-depth, recent literature on embodied, embedded cognition is either unknown or ignored.2

A second thought experiment is less extreme but still suffers from serious problems. This is the ‘fake barns’ example, a so-called Gettier case (Steup [2005] 2014, 2–3; Goldman and McGrath 2015, 60–61). Along a road in a rural area there seems to be a row of barns. In driving by, Corry makes a claim concerning a particular item of the row, namely the proposition p that this item (over there) is a barn. Since this is really a barn and Corry has evidence based on her reliable eyesight, she seems to know the proposition that p. Yet, unknown to her, it is the case that this particular item of the row is the only real barn, while all the others are merely barn facades. For this reason, JTB theorists conclude from this example that Corry does not really have the knowledge that p. More generally, such Gettier cases are taken to imply that justification and truth are not sufficient to make a belief into knowledge. A further condition or specification should be added, which excludes such cases of what is called ‘epistemic luck’.

In line with my earlier remark concerning the brain-in-a-vat hypothesis, also this example fails to acknowledge the crucial role of the body in perceptual processes. It is based on an inadequate account of the justificatory practices of perceptual experience, in which perception is reduced to a passive process of noticing something. The active component, that is, the concomitant movement of (parts of) the body, is missing. A full perceptual process of a three-dimensional object would include moving (parts of) our body to (slightly or considerably) different places and directions in such a way that we obtain different profiles of the object, which are then (unconsciously) integrated into the final perception of a three-dimensional object (Heelan 1983, chap. 7; Radder 2006, chap. 6). Corry’s claim is based on a quick and superficial noticing, and not on properly embodied experience. Rather than epistemically lucky, she is epistemically disabled because, in this (constructed) example, the quality of her visual procedure is rather poor. This conclusion is reinforced if we do away with the artificiality of the example. Suppose Corry lives in a world where the fake-objects situation is real and common, that is, a world where people are regularly confronted with adjacent two-dimensional and (similar looking) three-dimensional objects. In such a world, it would be well known that the identification of such objects needs to be based on careful, embodied perceptions. After all, a real three-dimensional barn can provide shelter in the case of bad weather, whereas a mere barn facade cannot do this at all or to a far lesser extent.

(4) A related characteristic of the JTB theory is its meta-universalism. The central methodological role of made-up and fictitious counterexamples shows that JTB theorists still presuppose the possibility and desirability of a universally valid philosophical theory of knowledge. In view of the fact that such universality is already highly unrealistic for science,3 assuming its possibility and desirability for philosophical inquiry demonstrates a serious lack of reflexivity on what philosophy can, and cannot, achieve.

(5) Traditional JTB theory is basically individualistic. The primary locus of knowledge is the individual knower, who appears to be an isolated Robinson Crusoe. To acquire some knowledge, each knower faces the world individually. His or her (material and social) life world is no more than a given ‘externality’. While Steup ([2005] 2014, 29) dubs ‘social epistemology’ an ‘additional issue’, to which he devotes a few lines, Goldman and McGrath (2015, chaps 9–10) discuss this topic in more detail. Yet, their focus is on epistemic aspects of groups and institutions, for instance the different ways of interaction among scientists. That is to say, they assume that the social comes into play only when more than one knower is involved. They do not, or hardly, address the more basic issue that even the knowledge of individual people may possess a constitutively social dimension. In fact, there are many good arguments supporting this view. A convincing criticism of JTB’s individualism has recently been presented by Boaz Miller. His main point is this:

Individualism fails because the justifying elements of many of an individual’s expert-obtained beliefs are distributed among members of her epistemic community. For a wide class of beliefs, only the justification collectively possessed by different members of a subject’s epistemic community amounts to knowledge-level justification. Only a communalist conception of knowledge adequately characterizes such true beliefs as knowledge.4

By way of illustration, consider again Jasper’s assertions about the current time but now transformed into a Gettier example. Jasper looks at his watch and claims that it is 4 pm now, a claim that is in fact true. So he seems to know the proposition p that it is 4 pm now. However, unknown to him, his watch is broken and has been stuck at 4 pm for some time. JTB theorists will conclude from this example that Jasper does not have the knowledge that p. Again, the reason for this is that the justification of his belief in this proposition p is merely a matter of epistemic luck.

In contrast, a non-individualist theory of knowledge takes into account other people and the social setting as constitutive aspects of the epistemic situation of individuals.5 On such a theory, Jasper’s knowledge claim is not properly embedded in the justificatory practices of time keeping. He does not have evidence (since one cannot acquire evidence from a malfunctioning device) and his reliance on the cognitive procedure he has followed is unfounded (since his watch does not work properly). Therefore, just like the fake barn case, this is a case of epistemic misfortune rather than epistemic luck.

Two more specific interpretations of the JTB theory might include a constitutively social dimension: contextualism and externalism. The basic claim of contextualism is that the meaning of the phrase ‘S knows that p’ varies across different contexts of speech (Goldman and McGrath 2015, 107–110; Steup ([2005] 2014, 23–24). Might this kind of contextualism imply a full recognition of the social dimension of knowledge? Perhaps, but not necessarily. First, its JTB versions seem to be limited to explicit, propositional knowledge, thus excluding know-how. Furthermore, as we will see in the next section, the case of skilled knowledge requires that S is ‘actually capable and in the right position to perform the justificatory procedure for the knowledge claim in question’. This entails a stronger form of contextualism than mere variation in the meaning of words. Finally, it is difficult to tell how many proponents of the traditional JTB theory actually endorse contextualism, even if limited to propositional knowledge.

Secondly, there is externalism. Externalists assume that some justifiers may be primarily located outside the knower S. In theory, this might make room for a constitutively social dimension. Yet, this does not seem to be the case in the currently dominant interpretations. They see truth and reliability as paradigm cases of external justifiers (Goldman and McGrath 2015, 42–44). For instance, an externalist justification will be reliable if ‘S’s belief has a high objective probability of truth’ (Steup [2005] 2014, 2). Thus interpreted, this kind of externalism does not include any reference to the social.6

(6) Related to its individualism is the JTB assumption that knowledge is a (specific kind of) belief, that is, a psychological attitude toward a proposition held by a specific individual. This is why Karl Popper calls it a subjective theory of knowledge.7 However, if other people matter epistemically, for example if reliable justification depends on the knowledge of other people, the idea of knowledge as primarily a belief becomes questionable. The reason is that it is often difficult to know what other people believe, unless the belief is expressed, made public, in some way. What is more, in certain cases it may also be difficult to know what we ourselves ‘really’ believe. For instance, many philosophers (and other authors) will be familiar with the experience that only after writing down one’s thoughts, one really knows what one thinks and believes. I do not mean to imply that we should not talk about private beliefs at all, as philosophers holding a behaviorist or a purely linguistic perspective might do. The point here is a pragmatic one: for the purpose of developing an optimally clear account of knowledge, it is preferable to start with publicly expressed beliefs rather than unexpressed, private beliefs.8

(7) Further important questions concern the notion of truth. The JTB theory sharply distinguishes truth from epistemic justification.

Truth, it may be said, is a purely metaphysical concept rather than an epistemological one. Given a proposition, what makes it true or false is simply the state of the world. Its truth-value is not affected by cognitive relations people have toward the relevant state of affairs. (Goldman and McGrath 2015, 5)

The theory implies that false statements cannot represent knowledge. Because justified beliefs may later turn out to be false, JTB theorists emphasize that the proposition p should be true in the absolute, metaphysical sense of the term. Although this may look rather evident, the situation is not as clear-cut as it seems. Take the example of Jasper’s knowledge of the proposition that it is 4 pm now on the basis of a properly working watch that only displays (in a discrete way) the minutes and not the seconds. But suppose it is in fact 5 (or 49) seconds past 4 pm. In almost all daily-life situations these facts will not impair Jasper’s knowledge, even if they do imply that p is false. However, in certain other contexts, where precise timing strongly matters, the falsity of p may invalidate Jasper’s knowledge.

Concerning scientific knowledge further problems arise. In science, it is often the case that older scientific theories are still highly valued and remain to be widely used, even if they are now seen as not being true. Is it reasonable to claim, on the JTB theory, that Newton had no knowledge of mechanics (since his theory of mechanics is now considered to be false)? More generally, quite a few philosophers of science argue that scientific truth functions primarily as a regulative idea that guides scientific development but cannot actually be achieved (either in principle or in many important cases, for instance because of the ineliminable role of idealizations and abstractions). My conclusion is that the claim by the proponents of the JTB theory (and by a critic such as Miller) that p should be true needs, at least, substantial qualification.

1.2. A Multi-dimensional Theory of (Scientific) Knowledge

The JTB theory is meant to be a general theory of knowledge. Therefore, it should also apply to scientific knowledge (even if the theory is almost always illustrated with the help of examples of everyday knowledge). Thus, Goldman and McGrath (2015, 241) see science as ‘a paradigm of epistemic activity’. However, given the fundamental problems explained in the previous section, we need an alternative account of both ordinary and scientific knowledge that goes far beyond the traditional JTB theory. In this section, I sketch such an alternative, which avoids the problems of the JTB theory and includes the suggested solutions.9 I distinguish three types of knowledge.

S knowsi that p, if and only if S’s claim that p is justifiably true.

S knowss that p, if and only if S’s claim that p is justifiably true and S is actually capable and in the right position to perform the justificatory procedure for this claim.

S knowsh how to do or make a, if and only if S is actually capable to realize a (under normal conditions) in a stable and reproducible way.

The first type is based on the weak notion of justification. This type of knowledge is very close to what we usually call (correct) information; hence the ‘i’ of informational knowledge. Put differently, informational knowledge is exclusively based on justificatory testimony. Because it seems incorrect to say that S knows that p, if (s)he has no understanding of the content of p at all, informational knowledge does require at least a minimal understanding of the meaning of p. In contrast, the second type of knowledge is based on a stronger form of justification: it requires the skills to perform the procedures that support the claim (which of course presupposes a thorough understanding of its meaning); hence the ‘s’ of skilled knowledge. This ‘performance’ may be verbal, as in the case of stating the justificatory reasons, or it may be non-verbal, as in the case of correctly realizing an observational process. Both informational and skilled knowledge are essentially embedded in a broader epistemic context. The two types of knowledge involve, primarily, a claim or assertion (that is, a publicly expressed belief) made by one or more people. Both types have to be included in a theory of knowledge, because the wider dissemination of informational knowledge claims and the skilled performance of justificatory procedures constitute two important dimensions of human epistemic practices.10 Finally, by also including a characterization of embodied know-how (to perform an action or make an artifact a), the JTB limitation to (constative) propositional knowledge has been removed. Moreover, other forms of knowing (for instance, to know a person) should be legitimate parts of a comprehensive theory of knowledge that accounts for the multiple dimensions of our epistemic notions and practices. That is to say, the three types of knowledge are not meant to be exhaustive. In contrast, Goldman and McGrath seem to assume that only one concept of knowledge can be the correct one. This assumption is, for instance, implied in their discussion of the results of experimental philosophy (Goldman and McGrath 2015, 191–194).

In Section 3, I argue that we need the differentiation between the three types of knowledge when dealing with the issue of scientific knowledge as a common good. In the rest of this section, I discuss several features and implications of the theory for the purpose of further clarifying its three parts and their mutual relations.

A bonus of the theory is that it nicely captures the distinctions made in Dutch and German between the related notions of weten/wissen (S knowsi that MIT Press is a North-American publisher), kennen/kennen (S knowss the economic concept of a public good) and kunnen/können (S knowsh how to compose a book index).

The three types of knowledge are related in the following ways. First, having skilled knowledge implies the possession of the corresponding informational knowledge, while the reverse is not true. Second, being actually capable and in the right position to perform the justificatory procedure for a claim implies having the know-how to do a, where a is this specific justificatory procedure. But again, the reverse does not hold: know-how does not necessarily include the making of claims concerning relevant justificatory procedures. Thus, the three types of knowledge are related but distinct. In particular, skilled knowledge requires informational knowledge and know-how, but cannot be reduced to either. Furthermore, these three types should be interpreted as ideal-types, in the Weberian sense of conceptually distinguishable forms of knowledge (see Weber [1904] 1949). However, calling them ideal-types does not imply that these types are idealizations, that is, extreme cases that cannot, or hardly, be found in practice. On the contrary, empirically such ‘pure’ types are quite common. But of course, in practice mixed forms occur frequently as well. As explained in the previous paragraph, skilled knowledge presupposes informational knowledge and know-how. Other forms of mixture are cases where S’s justification is partly based on skills and partly on information or know-how.

The term ‘justifiably true’ introduces an epistemic notion of truth. A claim is justifiably true if there are good reasons for believing it to be true. This epistemic notion of justifiable truth constitutes a crucial difference with the JTB theory, in which metaphysical truth and epistemic justification constitute two separate elements. An important implication is a shift of the possibility of failure from truth and justification to justified truth. Because justification is fallible, the same applies to justified truth. Therefore, this notion allows for revision of knowledge claims. If p turns out to be not justifiably true after all, we cannot have (had) knowledge of p. However, such revisions are not based on absolute, metaphysical truths (to which we do not have access), but on other knowledge claims that are accepted as justifiably true.11 Furthermore, by including notions of approximate truth, truth for a particular domain, referential truth, or even plausibility as potentially legitimate parts of justificatory arguments, the resulting theory of knowledge is less austere. It is a theory that also grants Newton knowledge concerning the domain of classical mechanics and Jasper knowledge of telling the time. A further important implication is that justified truth requires the existence of a specific justification for proposition p. This excludes alleged justifications based on procedures that, in spite of being seen as ‘generally’ reliable, still do not hold in the case in question. In this way, the proposed theory of knowledge avoids or solves various problems of the JTB theory by relaxing its truth requirement while strengthening its justification requirement (see the examples given under points 7a and 7b below).

The three types of knowledge possess a constitutively social dimension. S’s claim that p is justified or S’s know-how of a should be correctly embedded in a relevant community. Distinct communities (and sometimes distinct groups within such communities) employ different, or even rival, justificatory procedures. The justificatory procedures in chemistry differ from those in philosophy; qualitative and quantitative social science employ different and often competing criteria of justification; and what counts as ‘normal conditions’ in justifying know-how will also depend on the social context. More generally, there are many good arguments supporting the claim that both ordinary and scientific knowledge possess a constitutively social dimension. Let me just mention the following three. Concerning ordinary knowledge, a convincing argument from the sociology of knowledge is its account of the so-called ‘problem of the next case’ in the classification of empirical items (Barnes, Bloor, and Henry 1996, chap. 3; Kusch 2002, chap. 15). One kind of argument regarding science is based on the availability of different sets of values that may be legitimately applied in assessing scientific knowledge claims (Longino 1990, 1995). Another argument can be based on the fact that scientific reasoning proves to be dependent on distinct, historically emerging styles of knowing (Kwa 2011).

Like the JTB theory, my multi-dimensional theory aspires to be applicable to both ordinary and scientific knowledge. This does not exclude, however, that there may be significant differences of degree between the two. Generally speaking, in the case of science the cognitive division of labor will be much more pronounced than it usually is in ordinary life. Consider the two knowledge claims mentioned in the preceding section: the claim that ‘in 2017, there is a terrible Civil War going on in Syria’ and the claim that ‘active medulloblastoma enhancers reveal subgroup-specific cellular origins’. The understanding needed for having knowledge of these claims is shared much wider in the case of the former than in that of the latter. A similar difference applies to know-how: although ordinary life does include areas of specialized know-how, the degree and impact of specialized know-how in science is much larger. In Section 3, I will use this multi-dimensional theory, first, in developing a novel notion of the common good of scientific knowledge; second, by arguing that only informational and skilled knowledge (but not know-how) may be a common good; and third, by providing a fundamental critique of privatizing such common goods through patenting the products of scientific research.

Of course, I would welcome critical discussion of this multi-part theory of knowledge. Yet, the use of made-up and fictitious counterexamples should be treated in an equally critical way. As I argued in Radder (2009b, 890) for the philosophy of technology, what counts should be the interesting and important cases, not simply made-up counterexamples to demonstrate the non-universality of the theory. And again, there is no universal criterion for what is ‘interesting’ and ‘important’. In part, it will depend on one’s views on the nature and aims of philosophy.1

(a) By way of illustration, let us see which types of knowledge do, or don’t, apply to Jasper’s attempts at telling the time and to Corry’s engagement with fake barns. As we have seen, Jasper’s knowledge claim that it is 4 pm now involved two cases. In the case of the properly working watch, he possesses all three types of knowledge. He has the required know-how of reading the clock, which he acquired in his first years at primary school. He has the relevant informational knowledge, since the example assumes that his assertion ‘it is 4 pm now’ is justifiably true. And he possesses skilled knowledge because he is also capable and in the right position to perform the justificatory procedure for his claim by looking at his properly working watch. In the case of the broken watch, Jasper still has the know-how and the informational knowledge. But he does not have the skilled knowledge: because his watch is broken, he is not in the right position to use it for reading the time. (b)Secondly, there is Corry’s knowledge of the claim p ‘there is a barn over there’. Since this claim is justifiably true, she does possess informational knowledge of p. We may also assume that she has the know-how to make the relevant perceptions. The case of her skilled knowledge is more complex. Even if she has the capability to perform the relevant justificatory procedure by looking at the ‘barns’, one could question whether she is in the right position to correctly realize this procedure in this specific situation. After all, the example only makes sense if the ‘barns’ are at quite a distance from the road and if she does not look too carefully (in driving by, her main focus will be on the road). The conclusion to be drawn from these facts should be differentiated, as suggested in the previous section. If the occurrence of both two- and three-dimensional objects is very exceptional, we may grant Corry the skilled knowledge that there is a barn over there. We could note, however, that her claim on skilled knowledge is not very strongly justified. This qualification points to the fact that a more detailed account of knowledge should see justification as a matter of degree. However, if Corry lives in a world commonly populated with both types of object, we should conclude that she was not in the right position to perform the careful perceptions that, in this world, are normally required for supporting her claim. In this world, the condition that she ‘exercises her normal perceptual skills properly’ (Goldman and McGrath 2015, 72) would not be met.

2. What is a Common, or Public, Good?

The second challenge of this article is to develop an appropriate notion of a common, or public, good. For the moment I use these two terms interchangeably, as is frequently done in the literature. In Section 2.2, I will introduce a more specific terminology. I proceed in two steps. Section 2.1 reviews an important economic account of public goods, while Section 2.2 discusses several broader socio-political ideas concerning the common good.

2.1. The Economic Notion of a Public Good

First consider an influential view of the notion of a public good from the perspective of economics. The basic distinction is between public and private goods, in an economic sense. That is to say, the two kinds of goods are distinguished primarily in terms of the notion of property, or ownership. A private good is a privately owned one.13 For the purpose of explaining the notion of a public good, the idea of private ownership is seen as relatively unproblematic. The challenge is to determine the nature and function of public goods.

The starting point is the observation that a particular good may be socially beneficial, but also offers the opportunity of free riding. Therefore, contributing to its production would not be in the interest of private people. In particular, its production by private enterprises would not be profitable and hence not viable from an economic perspective. In such cases, ‘it is in the interests of society collectively to support production of that good’ (Nelson [1959] 2002, 153). Often-mentioned examples of such public goods are ‘clean air, public parks, public health service, national security, [and] public scientific research’ (Guala 2005, 18). However, these examples still seem to be quite diverse. Therefore, I will focus the discussion on scientific research and its results. Concerning this example, the claim is that, in a free and competitive economy, the information contained in scientific inventions ‘can be appropriated only to a limited extent’, which makes it difficult to exploit it commercially (Arrow [1962] 2002, 175).

The early analyses by Richard R. Nelson and Kenneth J. Arrow have been summarized by stating that public, in contrast to private, economic goods are non-excludible and non-rival (see, e.g. Romer 1990, S73–S78; Stiglitz 1999, 308–311). Non-excludability says that, in a free-market system, a public good cannot be privately appropriated: other people cannot be effectively excluded from freely using it. Non-rivalry means that the use (or consumption) of a public good by one individual does not reduce its use by other individuals. Furthermore, in economic terms the reproduction costs of producing one additional token of a non-rival good (for instance, the additional use of an existing item of knowledge) are taken to be zero or very low (Nelson [1959] 2002, 163–164; Arrow [1962] 2002, 170–171). Both non-excludability and non-rivalry are based on the nature of public goods as embedded in a given socio-economic context.

Thus far, these definitions have been phrased in an absolute way: a good is either public or not public. In practice, goods may also be approximately, or more-or-less, public. The paradigm of a pure public good is an idea.14 Ideas (if they are expressed) cannot be contained; they ‘float’ freely from one person to another. Furthermore, their use by one person does not at all diminish other people’s opportunities of using it. In contrast, a public park (that is, a park that is publicly owned and freely accessible by the public) is not a pure public good. Although it is non-excludable (by definition), it is only non-rival up to a point. Its ‘use’ by one more person will not impede others to enjoy it. But invasion by a large crowd will probably repel many of its initial users. In economic terms, an initially produced pure public good cannot be privately appropriated and the costs of its additional use or consumption are zero. If it is difficult, though not impossible, to appropriate a good and if its reproduction costs are small, but non-zero, the good will only be approximately public. Finally, for cases in which these reproduction costs are not negligible but require substantial investments, one may define the notion of a local public good. A good is locally public if it is non-excludable and non-rival given that, in the relevant location, the necessary reproduction investments have already been made.

2.2. Socio-political Notions of a Common Good

The economic notion of a public good can be summarized as follows. It is a good that is socially beneficial but, as a consequence of its (approximate) non-excludability and non-rivalry, it cannot be privatized. Given that making profit is the primary aim of private economic enterprises, the defining features of a public good imply that it is not suitable for making a profit. However, a more appropriate notion of public (or common) goods should go beyond this limited economic kind of reasoning. Therefore, I will argue that a common good should be in the public interest and should not be privatized, independently of whether it can or cannot be privatized for economic reasons. To avoid confusion with the economic account, I speak of common instead of public goods and of public interests instead of social benefits.

To illustrate the difference between the economic account of a public good and a broader notion of a common good, consider the case of patenting the results of academic research. Writing in the 1960s, Arrow assumed that patenting the results of scientific research would be impossible, or at least very difficult.

[T]he [scientific, HR] inventor will in any case have considerable difficulty in appropriating the information produced. Patent laws would have to be unimaginably complex and subtle to permit such appropriation on a large scale. (Arrow [1962] 2002, 173)

Yet, in this respect he was mistaken, since privatization through patenting proved to be feasible after all: nowadays many results of academic research are being routinely patented (Sterckx 2010). Therefore, extrapolating the economic view to the present, the problem of the underinvestment in scientific research might just as well be solved by patenting its results (rather than by state funding, as originally suggested). From a strictly economic perspective, in the current situation this kind of privatization of science would be unproblematic.

This conclusion contrasts with the scientific ethos described and advocated by Robert Merton ([1942] 1973). Merton’s value of communism entails that the results of science are not, or should not be, privately owned, even if this would be possible from an economic perspective.15 As demonstrated in Radder (2010c), this claim can be specified and strengthened by basing it on a deflationary neo-Mertonian ethos. The key assumptions underlying both Merton’s and a neo-Mertonian approach concerning the relationship between humans and society are quite different from the ones in the prevailing economic accounts. The basic starting point is not the competition between self-interested individuals on a free market. Although Merton explicitly acknowledges the significance of the competition among individual scientists, the justification of scientific knowledge claims is not determined by self-interested researchers but by the collective activity of groups of scientists sharing common values. Moreover, the aim of competition is not, or not primarily, money or profit but rather recognition and reputation (Van den Belt 2010, 202–205). And the claim is that the latter approach to science is preferable to the purely individualistic economic view both on epistemic and on moral, social or economic grounds.

Merton’s approach can be related to some other views of common goods. One view is based on the notion of a commons (Vermeir 2013). Its basic idea is that, for certain goods, collective ownership is preferable to private ownership. A common good, then, is a socially valuable, collectively owned good. A much-discussed example is the collective ownership of land in certain agrarian societies. But one can also think of the many collectively owned infrastructural systems in modern societies.

A related idea is the notion of the gift economy, which is an economy in the broader sense of a system that regulates the production and exchange of goods. In a gift economy, the value of the goods is not determined on the basis of a market with its quantitative laws of supply and demand. Instead their value is qualitative, and constituted through reciprocal acts of the offering and receiving of gifts. A gift economy includes subtle but implicit norms that entail (or hints that suggest) the appropriateness or inappropriateness of acts of giving and receiving in particular situations. Well-known examples include family life and relations of love, friendship and collegiality. Ontologically and politically, a gift economy shares with a commons the communitarian primacy of the collectivity. Yet, in a gift economy, goods are not necessarily (or not even frequently) commonly owned. However, for the case of scientific research, one may combine the ideas of a commons and a gift economy. In this spirit, Koen Vermeir (2013, 2499) claims that ‘the commons model and the gift economy reinforce the Mertonian norms of science’.

Although the idea of the communitarian primacy of the collectivity involves a valuable correction to an exclusively individualistic, market-oriented economy, it also suffers from its own limitations. One important problem is the tendency to conceive of communities as well-determined and more or less closed and self-sufficient entities. A further problem is that actual instances of a commons often prove to be limited to rather specific communities. In an article entitled ‘Whose good is the common good?’, Offe ([2001] 2012, 673–681) discusses four critical questions concerning the appropriateness of the idea of a common good. (1) Which kinds of current communities may be legitimate bearers of a common good? (2) How do common good claims of current communities relate to the interests of future communities? (3) What exactly should be the content of the common good? (4) And who, which people, possess the competence to choose a specific content for this notion? These basic questions spell doubt on the reality and the (moral, political or epistemological) desirability of a purely communitarian notion of the common good. In this article, my focus is on the third question (several aspects of the other questions have been discussed in my account of what are ‘genuine public interests’ in Radder 2016).

Harry Kunneman addresses similar issues and concerns in terms of values. By combining and developing insights from Charles Taylor and Luce Irigaray, he argues that values (and especially basic values or ‘hypergoods’, such as justice or personal integrity, and perhaps also including some of the Mertonian values) transcend not only our individual inclinations or desires but also any specific community in which they are embedded.

It is precisely the transcendent character of these values and their possible source, which implies that each and any articulation of them within a specific tradition and within a specific (linguistic or esthetic) medium of expression can only point to them and ‘evoke’ them but can never adequately characterize or describe them ‘as they really are’.16

For this reason, debate on such basic values (or, in my terms, common goods) will always and necessarily have to deal with different and possibly conflicting articulations of these values. This account of the meaning of values in terms of their ‘horizontal transcendence’ can be related to the issue of non-rivalry, in its broader sense of non-exhaustibility. In Section 3.2, I introduce the notion of non-exhaustibility and connect Kunneman’s account of transcendent values with the general theory of the nonlocality of meaning developed in Radder (2002, 2006).

3. Scientific Knowledge as a Common Good

Having explained the nature of scientific knowledge and several current notions of a common (or public) good, I can now address the question of the common good of scientific knowledge. In Section 3.1, I provide a critical analysis of the implications of the economic accounts of public or common goods for the interpretation of scientific knowledge. On that basis, I then develop my own view of the common good of science in Section 3.2.

3.1. The Economic Arguments for the Public Good of Science17

The primary aim of the articles by Nelson and Arrow was to provide a rational defense of the public funding of science, especially basic science. Their focus is on scientific knowledge, which is described by Nelson as follows.

Scientific research may be defined as the human activity directed towards the advancement of knowledge, where knowledge is of two roughly separable sorts: facts or data observed in reproducible experiments (usually, but not always, quantitative data) and theories or relationships between facts (usually, but not always, equations).18

Next, scientific knowledge is claimed to be (approximately) non-excludible and non-rival. For this reason, private enterprises will not be keen to fund scientific research. Yet, scientific knowledge begets important social benefits. Applied research is crucial in solving existing practical problems, while basic research regularly creates unforeseen opportunities for novel practical applications. Because of these valuable social benefits, governments should step in and provide appropriate funding of both basic and applied science.19

How to assess these views? Let me first mention three general points. The first is the limitation of the public good of science to the economic value of the results of scientific research. Possible broader values of both research and education, for instance aiming at personal development or responsible citizenship, are mentioned but not systematically taken into account (Nelson [1959] 2002, 153–154). The second point concerns the relationship between basic science and its technological applications. On this point, the two authors agree with the view of ‘technology as applied science’. This view was the dominant one in the 1950s and 1960s. However, more recent research (reviewed in Radder 2009a, 69–74) has convincingly shown that it is untenable. Finally, as mentioned in Section 2.2, the authors did not anticipate the big increases in patenting the results of academic research. Although they were right that this would require a complex and difficult process, the large-scale patenting of scientific research results did prove to be possible.20

A fundamental criticism of the view of scientific knowledge as a non-excludible and non-rival public good has been launched by Michel Callon. His criticism is based on a view of the nature of science and scientific knowledge that is quite different from the one held by Nelson and Arrow.

Science is a thing, or rather a set of complementary things: it does not exist outside of the diverse materials in which it is inscribed. (Callon 1994, 402)

The same applies to scientific knowledge.

The main point is that scientific knowledge is endowed with a physical nature to the extent that it can circulate, be exchanged, or be engaged in commercial transactions. Being a thing, scientific knowledge can even be stolen. This materialism might seem vulgar and shocking but it is, nevertheless, perfectly defensible. (Callon 1994, 397–398)

Moreover, since things are always localized and exist at a specific place only, a further essential feature of scientific knowledge is its local nature.

Rules, practices, cultural forms and relationships with things all vary from one collective to another. Diversity and the local are at the heart of science. (Callon 1994, 418)

On this basis, Callon claims that scientific knowledge is both excludable and rival, and therefore not a public good in the economic sense of that term. He focuses on what he calls codified knowledge, the knowledge contained in generally accepted public statements, but implies that his arguments will apply a fortiori to tacit or incorporated knowledge. Codified scientific knowledge is excludable because the use of this knowledge (such as the statement that ‘DNA’s structure is a double-helix’) requires extensive investments in having available the required instrumentation, skills, money and people to reproduce this knowledge. Callon’s argument against the non-rivalry of codified scientific knowledge is related to this. It is summarized in ‘the thesis of the intrinsic inutility of statements’, that is,

the impossibility of endowing a statement with any meaning if the work of the duplication of skills and instruments has not been done. (Callon 1994, 403)

Isolated scientific statements lack meaning and utility, which implies that they are not even economic goods. For this reason, they simply cannot qualify as non-rival and, more generally, as public goods.

According to Callon, it is only the statement plus its local, material (re)production context that possesses meaning and utility. Could it be the case, then, that it is the statement plus its reproduction context that is non-rival? Callon does not address this question, but perhaps his answer could be like this. We live in a world of scarcity. The realization of a specific (re)production context by a researcher or research group requires substantial investments. In a world of scarcity, this realization will be at the expense of further realizations. That is to say, the use of a scientific statement by a particular individual or group is rivalry, because it does reduce the opportunities for uses by other researchers.

In its turn, Callon’s arguments have been criticized by Van den Belt (2010, 210–211). He argues, first, that Callon’s criticism of non-excludability is correct but has already been taken into account in the classic economic view by including the notion of a local public good. I do not think this reply is plausible, however, because the notion of a local public good is flawed. Creating and maintaining the instrumentation and the expertise in the local situation has required and keeps requiring substantial investments (see Callon 1994, 403–406). For this reason, the notion of a local public good goes against the underlying idea of the free availability of public goods. Jefferson’s rendering of the non-excludability of a good (‘the moment it is divulged, it forces itself into the possession of every one’, quoted in Van den Belt 2010, 199) clearly does not apply to local public goods. Local public goods are not available ‘to everyone’, but only to the members of those select communities that have made, and have the means to keep making, the required investments.

Steve Fuller also endorses the non-excludability of (scientific) knowledge. Thus, he claims that ‘the fact that you know something does not exclude me from knowing it’ (Fuller 2010, 284), and explains this claim as follows.

But of course, knowledge cannot be captured − at least not permanently − only because at least some of those who do not currently possess a given piece of knowledge have the (intellectual and material) means at their disposal … to create the same knowledge for themselves. (Fuller 2010, 284)

This claim is questionable, for the same reasons as Van den Belt’s defense of non-excludability. First, whether or not other people will be able to acquire the same knowledge is an empirical question: most probably, some will and some won’t. Second, even if some people succeed in reproducing a piece of knowledge, this fact does not at all imply the claim that this knowledge ‘forces itself into the possession of every one’.21

My conclusion from this discussion of the non-excludability of (scientific) knowledge is as follows. The claim at issue is that, in a free-market economy, ‘other people cannot be effectively excluded from freely using a public good’. Callon focuses on the ‘freely’ and concludes that, in the case of scientific knowledge, there is no free availability at all. The theory of knowledge presented in Section 1.2 agrees with this specific part of Callon’s views. After all, this theory entails that the justification of scientific knowledge claims requires that they are embedded in specific, well-developed, and hence costly, socio-material practices. Van den Belt and Fuller, in contrast, focus on the excludability and claim that full or permanent excludability is improbable. I think that Van den Belt and Fuller are right, but only concerning their own specific points. What they have not achieved is a rescue of the claim that scientific knowledge is a genuine public good from Callon’s criticism. The reason is the one I have already stated above: the idea of ‘the public’ in the definition of a local public good is so restricted that it betrays the original meaning and the normative significance of the notion of a public good.22

Secondly, Van den Belt argues that Callon’s criticism of non-rivalry is incorrect. Scientific statements are not simply local, material things. The statement that ‘DNA’s structure is a double-helix’ is not merely a token, but primarily an instance of a type of statements. It is on types, and not on tokens, of statements that the public good character of scientific knowledge claims is based. The application of the double-helix model to a certain organism by a particular scientist does not ‘consume’ this model, in the sense that less of it is available for another scientist in applying the same type of model to a different organism. I think that this criticism of Callon’s claim that scientific knowledge is not non-rival is correct, but there is more to be said.

In Section 2.1, I stated that the basic meaning of non-rivalry, the fact that a non-rival good is not being consumed in additional uses, is interpreted by economists as saying that its reproduction costs are non-zero (or very low). But in fact these two things are not equivalent. A clear counterexample is the transference of know-how from a teacher to a student. This process does not reduce the know-how possessed by the teacher, and yet it usually requires the availability of a costly education system. In his defense of the non-rivalry of scientific knowledge, Van den Belt assumes what I have called its basic meaning. Callon, however, bases his argument on the economic interpretation of non-rivalry (although he briefly mentions its basic meaning).

From an economic point of view, the property of nonrivalry is essential. It means that the good’s production costs are fixed: once the good has been produced there is no need for continuing investment because there are no production costs in replicating it. (Callon 1994, 400)

He then goes on to criticize the idea of non-rivalry by pointing out the extensive reproduction costs of scientific knowledge.23

I agree that Callon is right on this point. Yet, since this criticism does not undermine the basic meaning of non-rivalry, Van den Belt’s reply still stands.24

The overall results of this discussion are that the idea of a local public good is questionable; that the notions of non-excludability and non-rivalry are being used in different ways; and that the idea of the non-rivalry of scientific knowledge requires a more detailed defense. For these reasons, we need a more systematic argument why scientific knowledge is, or is not, a public, or in my terminology, a common good.

3.2. Which Scientific Knowledge is a Common Good?

Finally, I can address the question of whether and, if so, which scientific knowledge is a common good. In answering this question, I will use the differentiation of (scientific) knowledge into three types (set out in Section 1.2), combined with a socio-political notion of common goods in the spirit of the views discussed in Section 2.2. In addition, however, we will need to tackle the issue of the content of scientific knowledge claims.

Let me first discuss in somewhat more detail Callon’s alternative view that scientific knowledge has a physical nature. On the most natural reading, this seems to imply that statements just are material things. Consider the statement that ‘DNA’s structure is a double-helix’. As a material entity, this statement is a configuration of specific physical forms, namely of the letters on my computer screen. Yet, to claim that the meaning of this statement is exhausted by this local, material configuration is a mistake. After all, this claim would make it a total mystery why completely different material configurations, used within the same context, can make the same statement. For instance, a different configuration on my computer screen, such as ‘DNA’s structure is a double-helix’; or a printed version of this statement as a pattern of ink blots on paper; or in the case of a spoken version, a configuration of vibration patterns in the air; and so on.

In addition to common sense, the equivalence of these material expressions constitutes a basic premise underlying copyright laws. In a standard phrasing, a copyright owner has the exclusive right of ‘translation, reprinting, recitation, broadcasting, reproduction on microfilms or in any other physical way, or by similar or dissimilar methodology not known or hereafter developed’. Thus, if a statement would be no more than a single local and material thing, one could not even make sense of the notion of copyright.

It might be, though, that Callon’s claim that scientific knowledge is material by nature is merely unfortunately phrased and is not really what he means. First, he might reply to the argument above by admitting that the different material configurations are indeed equivalent. In this case, however, he owes us a reason why they are equivalent, in spite of their obvious material differences, and he has not provided such a reason. Second, he might respond that he did not mean to reduce the statements to their material features. This response is suggested by his view that scientific knowledge claims are inscribed in, and thus not identical to, material things. But again, this leaves open the question what it is that is being inscribed, a question that has not been answered by Callon.

The challenge, therefore, is to explain why the same proposition can be expressed by a variety of different material forms and hence what is meant by the content of a knowledge claim and its relation to its material form. In Section 1.2, I defined the notions of know-how and of informational and skilled (scientific) knowledge. However, these definitions did not include, at least not explicitly, a reference to the content of these knowledge claims. Therefore, the account given so far is not sufficient for answering the question of whether or not these types of scientific knowledge constitute a common good. In addition to the issue of the justification of this knowledge, we have to address its content. For this purpose, I apply my theory of the abstract nature and the non-local meaning of (scientific) concepts (see Radder 1993, 2002, 2006). This theory is complex and although I cannot explain and defend it in detail here, a brief summary is needed for the purpose of this article.

The relevant issue (also mentioned by Van den Belt 2010, 225, note 36) concerns the meaning of the concepts employed in making informational and skilled knowledge claims. These concepts possess both a structuring and an abstracting meaning component. As abstract entities they have a non-local meaning. That is to say, their meaning transcends the meaning they possess as interpretations of all the local processes in which they have been employed thus far. Put differently, their meaning is not exhausted by the sum total of all their past and present uses. For instance, the meaning of the concepts of DNA and of a double helix is not exhausted by their past and present uses. For this reason, these concepts are extensible: they might be successfully extended to (possibly radically) different organisms, which may either be existing but as yet unknown or which may be completely novel, technologically realized organisms. In this specific sense, informational and skilled knowledge is non-exhaustible. Scientific concepts have the potential to be useful in an indefinite number of new situations. Moreover, the potential novel uses of informational and skilled scientific knowledge may be radically different from their original uses. This ontological notion of non-exhaustibility is related to but also qualifies the (basic meaning of the) notion of non-rivalry: these types of knowledge are potentially useful in novel situations. They will be actually useful if they can be successfully realized in those novel situations. Yet, their potential is not diminished by any of their actual uses. In the first part of this section, I criticized Callon’s claim of the material nature of scientific knowledge. The non-exhaustibility of informational and skilled knowledge implies that his claim that scientific knowledge is essentially local is mistaken as well.25

The notion of non-exhaustibility is akin to Kunneman’s idea of the horizontal transcendence of values (discussed in Section 2.2). Their similarity derives from the fact that the meaning of both values and concepts is non-local: it is not, and cannot be, fully determined by past and current practices. There are also some differences, however. One is that non-exhaustibility applies not just to value concepts but to any concept that has a non-local meaning; furthermore, it applies equally to mundane and to profound or ‘hyper’ concepts.

Thus, non-exhaustibility is a consequence of the conceptual interpretations employed in expressing informational and skilled knowledge claims. As abstract entities, these concepts do not coincide with the socio-material processes in which they have been realized thus far.26 For know-how, the situation is different, since know-how is performed and revealed through specific actions of specific people in specific, normal situations. Of course, we can talk about our know-how. Yet, its ideal-typical core is how to do things, not how to make verbal claims about propositions that are composed of concepts. Furthermore, I do not want to deny that know-how can be transferred to new situations, for instance from a master to an apprentice. The point is, however, that the know-how in this new situation will be basically similar to the know-how in the original situation (and certainly not radically different, as may be the case in extending a concept to a completely novel context). In this sense, a specific instance of know-how remains bound to a specific set of ‘normal’ reproduction conditions.

My first conclusion is that an item of scientific knowledge is a common good, only if it is non-exhaustible.27 Thus phrased, this is a necessary condition. It holds for informational and skilled knowledge but it does not apply to know-how. Yet, for the purpose of explaining the notion of a common good non-exhaustibility alone does not suffice. An appropriate account of scientific knowledge as a common good should include a second necessary condition, saying that (acquiring) the knowledge in question serves a public interest. Together, the two necessary conditions are sufficient: an item of scientific knowledge is a common good, if and only if it is non-exhaustible and in the public interest. As we have seen, know-how is not non-exhaustible and therefore cannot qualify as a common good. This does not mean, however, that know-how is not important and can be ignored. First, as explained in Section 1.2, although skilled knowledge, because of its non-exhaustibility, cannot be reduced to know-how, it does require this type of knowledge: being actually capable and in the right position to perform the justificatory procedure for a claim requires having the know-how to carry out this procedure. Second, know-how is epistemically and socially important because it may be in the public interest.

If non-exhaustibility were the only (necessary) condition, the account would merely allow negative conclusions, concerning what is not a common good. A further reason for including the public-interest condition is that non-exhaustibility as such lacks a specific evaluative content. If the non-exhaustibility condition would be sufficient, the informational and skilled knowledge employed in the construction of gas chambers or nuclear weapons would be a straightforward example of a common good. Of course, adding this second condition requires a further explanation of the complex notion of a public interest and of the senses in which science can be in the public interest.28 Moreover, adding the public interest condition implies that more aspects of science need to be taken into account than just its epistemic dimension. For this reason, it makes sense to speak of the common good of science, and not merely of scientific knowledge.

In concluding this section, let me summarize the road I have taken in terms of agreements and disagreements with the views discussed and assessed in this article. Like the JTB theory, I have emphasized the importance of justification. Furthermore, although the social is a constitutive dimension of knowledge, we should not replace the primacy of individuals by the other extreme, the primacy of the social. In this respect, my view differs from Martin Kusch’s communitarian theory, according to which knowledge is primarily social.29 Moreover, by assigning knowledge to a claim made by one or more individuals, I do not concur with Popper’s theory of knowledge as consisting of (approximately) true propositions in an abstract world-three. In my view, what is abstract and transcends our individual and social contexts is not (justified) truth but non-local meaning.30 Yet, in several other respects the multi-dimensional theory of knowledge presented in Section 1.2 differs significantly from the JTB theory.

Furthermore, the analysis of the economic notion of public goods resulted in the following conclusions. First, the claim of the non-excludability of scientific knowledge, also endorsed by Van den Belt and Fuller, has been shown to be ambiguous. Even if it is true that scientific knowledge cannot be fully or permanently contained in its original context, this fact alone is not enough to rescue the general notion of a public good. Second, the idea of non-rivalry (in its basic meaning) proved to be helpful and is related (albeit in a significantly qualified sense) to the notion of non-exhaustibility. The first conclusion agrees with Callon’s claim that the fruits of scientific knowledge are not simply freely available to everyone. It affirms the significance of the embodiment and embeddedness of all human cognition mentioned in my criticism of the JTB theory. However, by taking into account the content of scientific knowledge claims, Callon’s criticism of the non-rivalry of codified knowledge proved to be untenable.

Yet, the shift from non-rivalry to non-exhaustibility and the inclusion of the public-interest condition also moves my final account of the common good of scientific knowledge further away from the economic terminology and position. That (informational and skilled) scientific knowledge is non-exhaustible is a factual claim. However, as I have argued in Radder (2016), claiming that something is in the public interest requires a normative stance. Therefore, stating that a particular item of scientific knowledge is a common good is a normative claim. This is an important further respect in which this view differs from the economists’ theory of a public good. In the economic approach, the notion of the social benefit of a good merely functions as an intuitive, external starting point of the analysis. It is not explicitly analyzed and does not constitute a separate component of the theory. In contrast, adding the condition of being in the public interest to the condition of non-exhaustibility brings my account of the common good of scientific knowledge more in line with the explicitly normative socio-political notions of a common good discussed in Section 2.2. In sum, the question of which scientific knowledge constitutes a common good has been answered as follows: all items of informational and skilled knowledge that meet the public-interest condition.

4. Implications for the Politics of Science

In concluding the article, I will demonstrate that the proposed account of the common good of scientific knowledge is not only of theoretical significance. It does have concrete, practical implications for an assessment of the commodification of academic research, and hence for the politics of science. In discussing this subject we should, at least, address two issues: secrecy and privatization.31 In as far as it is a common good, scientific knowledge should, first, not be kept secret but made public and, second, it should be freely available and not be privatized. Both issues are complex. Here, I briefly discuss one aspect of the commodification of science, to wit the patenting of the results of scientific research. In particular, I focus on a specific kind of patenting, namely product patenting.

A patent is a commercial monopoly, a legally granted exclusive right on the commercial exploitation of a technological invention. The knowledge on which the invention is based cannot be patented. Its role in patent applications is to explain the (industrial) utility of the patentable invention and to provide (a part of) the justification of its reproducibility. Moreover, the patent laws explicitly require that the (scientific) knowledge claims underlying a patentable invention should be made public, and hence not be kept secret. Applicants should disclose this knowledge by providing an explicit description of it. If the patent is granted, this description will be made public by the relevant patent office. Furthermore, patent law allows the free use of this knowledge if applied in inventions that are clearly distinct from the patented invention. Thus, neither informational nor skilled knowledge can be patented.32

In addition, patent laws exempt information, abstract concepts and theories from patentability. For instance, the European Patent Convention excludes from patentability ‘presentations of information’, ‘scientific theories’, and ‘schemes, rules and methods for performing mental acts’.33 A major argument for the exclusion of theories is that they are not technologies and provide no ‘technical teaching’ (Sterckx and Cockbain 2012, 59–60). For these reasons, proponents of the current patenting practices of the results of scientific research see these practices as basically unproblematic and justified.

Yet, this conclusion is far too rash. A closer analysis reveals many questionable aspects of the theory and current practices of patenting (see, e.g. Lever 2012; Radder 2013; Pagano 2014). Here, I will focus on one of these problems, which can be related to the notion of the non-exhaustibility of informational and skilled scientific knowledge. A product patent (for instance on a particular gene) is a patent on the product as such. That is to say, it is valid for any known or unknown process through which the product has been or might be produced. Thus, with the help of the questionable distinction between the invention itself and the patent claims allegedly based on it, the protection acquired through a product patent goes far beyond what has been made available through the actual invention. As I have argued in detail in Radder (2004), a product patent effectively amounts to appropriating a concept. Put differently, the patent claims made through a product patent are, effectively, conceptual or theoretical claims. In this way, the practice of product patenting implies privatizing the non-exhaustibility of scientific concepts and theoretical statements. This kind of privatization appropriates their full, non-exhaustible potential on the basis of a limited scientific achievement; at the same time, it prevents the (wider and possibly improved) realization of this potential by other researchers (see also Van den Belt 2009, 1321–1324). Furthermore, for the subset of products that are, or may become, of public interest, product patenting implies the private appropriation of common goods.

Remarkably enough, what happens in product patenting is a form of privatization, the effective patenting of concepts, that is illegal according to the patent laws themselves. The conclusion is that the practice of product patenting should be discontinued, and certainly in publicly funded institutes of higher education.

Notes

To be sure, different JTB epistemologists may have different views about the actual justificatory force of these sources.

See, e.g. O’Regan and Noë (2001), Anderson (2003), Radder (2006, Part 1), and Noë (2009). Indeed, Goldman and McGrath start their book with positioning epistemology ‘in the sphere of the intellect’ and by contrasting it with moral theory, which addresses ‘the sphere of action’ (2015, 3).

See Thomas Kuhn’s account of the frequent occurrence of anomalies in normal scientific practice (Kuhn [1962] 1970, chap. 6).

Miller (2015, 417–418). Apart from this, he also acknowledges the problems the commodification of science entails for the reliability of testimony as a source of scientific knowledge (426–427). In other respects, the theory of knowledge I develop in this and the following sections, goes beyond Miller’s account, since he does ‘not challenge any orthodox assumption besides individualism’ (420).

See also Hamlyn (1970), which presents the theory of knowledge from a broadly Wittgensteinian point of view. He concludes, for instance, that ‘[i]nterpersonal agreement provides the criterion for the concept of truth’ (142).

See also Martin Kusch’s criticism of attempts to ‘render harmless’ externalist contextualism by appealing to privileged forms of reliability (Kusch 2002, 135–138).

See Popper (1972), but note that in this context ‘subjective’ does not necessarily coincide with ‘relativist’.

In a more principled spirit, I could add as a further reason that normative judgments of beliefs should only count if these beliefs either are public or influence public issues (compare Radder (2009b, 893–894), where I criticized the moralistic or rationalistic presuppositions of philosophical judgments of private beliefs).

A striking feature of the book by Goldman and McGrath is its narrowness. While it claims to be an introduction to contemporary epistemology, one searches in vain for substantive discussion of the many contemporary approaches derived from, say, Immanuel Kant, Ludwig Wittgenstein, Karl Popper, or Jürgen Habermas. The alternative theory proposed in this section includes elements inspired by the views of these thinkers (cf. Radder 2006, [1984/1988] 2012).

In the previous section, I emphasized the primacy of public claims over private beliefs. In order to assign to S the (informational or skilled) knowledge that p, the claim ‘that p’ should be public. In a derivative sense, one may say that S does have the knowledge that p, if (s)he believes that p and if the corresponding public claim would prove to meet the relevant criterion.

See for example the discussion of ‘verifiable truth’ in Radder ([1984/1988] 2012, chap. 4).

For my own characterization of philosophy as a theoretical, normative and reflexive activity, see Radder (1996, chap. 8).

Thus, private goods are not publicly (or collectively) owned. Commodified goods are goods that have, at a certain point in time, been given a monetary value. Such goods can be both privately and collectively owned. The latter is the case when public governments start charging citizens for using a good. Nowadays, drinking water is usually a commodified, collectively owned good. However, drinking water is also available as a private good sold by privately owned companies. In sum, commodification may, but does not need to, go together with privatization (see the account of commodification in Radder 2010b).

Van den Belt (2010, 197–200) credits Thomas Jefferson, one of the eighteenth century founding fathers of the United States, with providing an early, illuminating explanation of the essential public nature of ideas.

See also Ilkka Kauppinen (2014, 402), who emphasizes that ‘the key question is not whether knowledge can be commodified, but whether it should be commodified’.

Kunneman (2010, 331–332); see also the more detailed account in Kunneman (2005, 62–81).

Special thanks go to Henk van den Belt for his extensive and insightful comments, which led to a significant revision of this section.

Nelson ([1959] 2002, 154). Nelson adds that the economic value of scientific knowledge rests on its predictive power. If that were true, it would be very difficult to resist the conclusion that economics, given its widely acknowledged lack of accurate predictions, is not of economic value.

Note that this argument is not necessarily restricted to scientific knowledge. It would also apply to those types of non-scientific knowledge (for instance in the humanities) that can be shown to be non-excludable, non-rival and socially beneficial.

See for instance the prolonged debate on the inclusion of biotechnological inventions in European patent law in the last decades of the twentieth century (Sterckx 2000; Van den Belt 2009, 1319–1330).

In the same spirit, Stiglitz (1999, 310) argues for the moral and political significance of global public goods, ‘the benefits of which accrue to everyone in the world’.

In Radder (2016) I have addressed this point in more detail, in discussing the question of which public should be included in the notion of a public interest.

Note that, on this interpretation, the notion of non-rivalry of a good shares its free availability with the notion of non-excludability. This explains why Callon is able to use basically the same criticism for both notions, and why Van den Belt (2010, 226, note 37) thinks he is confusing non-rivalry with non-excludability.

The same ambiguity between the basic meaning of non-rivalry and its interpretation in terms of zero reproduction costs can be found in Pagano (2014, 1420).

See also the detailed critique of the role of materiality and locality in actor-network theory (a theory co-developed and endorsed by Callon) in Radder (2006, 145–153).

Moreover, as I have argued in Radder (2014, 27–29), they cannot be conceived as mere mental entities either.

Analogous to the economic notion of a public good (see the remark in Section 3.1) this conclusion also applies more generally, to wit to any knowledge claim that employs extensible concepts with a non-local meaning.

Unfortunately, a detailed discussion of these two issues exceeds the space of a single article by far. For an initial account of the former, see Radder (2016); the issue of the public interest of science will be treated in a forthcoming publication.

See also De George (1983, 13–14): ‘Champions of individualism argue the primacy of the individual; champions of collectivism argue the primacy of society. But just as there is no human society without human individuals, there is no human individual without society.’ Translated to the multi-dimensional theory of knowledge presented in this article: just as there is no justified truth without individuals making knowledge claims, there is no justified truth without the epistemic communities in which these claims are, or will come to be, embedded.

For a discussion of Popper’s three-worlds theory, see Radder (2006, 122–125).

‘At least’, because the requirement of being in the public interest involves more than merely making the knowledge public and freely available.

The role of know-how is more complicated, since some scientific methods may be patented while others may not.

European Patent Convention (2016, 108, article 52). Note that the notion of ‘schemes, rules and methods for performing mental acts’ fits well in the theory of scientific concepts as methods of observation, as discussed in Radder (2006, 94–98).

References

Anderson, M. L. 2003. “Embodied Cognition: A Field Guide.” Arti cial Intelligence 149: 91–130.

Arrow, K. J. (1962) 2002. “Economic Welfare and the Allocation of Resources for Invention.” In Science Bought and Sold: Essays in the Economics of Science, edited by P. Mirowski and E.-M. Sent, 165–180. Chicago, IL: University of Chicago Press.

Barnes, B., D. Bloor, and J. Henry. 1996. Scienti c Knowledge: A Sociological Analysis. London: Athlone.

Bok, D. 2003. Universities in the Marketplace. The Commercialization of Higher Education. Princeton, NJ: Princeton University Press.

Callon, M. 1994. “Is Science a Public Good?” Science, Technology, & Human Values 19: 395–424.

De George, R. T. 1983. “Social Reality and Social Relations.” Review of Metaphysics 37: 3–20.

European Patent Convention. 2016. Convention on the Grant of European Patents. 16th ed. München: European Patent Office. Accessed August 8, 2016. http://documents.epo.org/projects/babylon/eponet.nsf/0/F9FD0B02F9D1A6B4C1258003004DF610/$File/EPC_16th_edition_2016_en.pdf

Fuller, S. 2010. “Capitalism and Knowledge: The University Between Commodi cation and Entrepreneurship.” In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 277–306. Pittsburgh, PA: University of Pittsburgh Press.

Goldman, A. I., and M. McGrath. 2015. Epistemology: A Contemporary Introduction. New York: Oxford University Press.

Guala, F. 2005. The Methodology of Experimental Economics. Cambridge: Cambridge University Press.

Hamlyn, D. W. 1970. The Theory of Knowledge. London: Macmillan.

Heelan, P. A. 1983. Space-perception and the Philosophy of Science. Berkeley: University of California Press.

Kauppinen, I. 2014. “Different Meanings of 'Knowledge as Commodity’ in the Context of Higher Education.” Critical Sociology 40: 393–409.

Krimsky, S. 2003. Science in the Private Interest. Has the Lure of Profits Corrupted Biomedical Research? Lanham, MD: Rowman and Littlefield.

Kuhn, T. S. (1962) 1970. The Structure of Scientific Revolutions. 2nd ed. Chicago, IL: University of Chicago Press.

Kunneman, H. 2005. Voorbij het dikke-ik. Bouwstenen voor een kritisch humanisme [Beyond the Self-important Self. Building Blocks for a Critical Humanism]. Amsterdam: Humanistics University Press.

Kunneman, H. 2010. “Viable Alternatives for Commercialized Science: The Case of Humanistics.” In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 307–336. Pittsburgh, PA: University of Pittsburgh Press.

Kusch, M. 2002. Knowledge by Agreement. The Programme of Communitarian Epistemology. Oxford: Oxford University Press.

Kwa, C. 2011. Styles of Knowing: A New History of Science from Ancient Times to the Present. Pittsburgh, PA: University of Pittsburgh Press.

Lever, A., ed. 2012. New Frontiers in the Philosophy of Intellectual Property. Cambridge: Cambridge University Press.

Longino, H. E. 1990. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry. Princeton, NJ: Princeton University Press.

Longino, H. E. 1995. “Gender, Politics, and the Theoretical Virtues.” Synthese 104: 383–397.

Merton, R. K. (1942) 1973. “The Normative Structure of Science.” In The Sociology of Science, R. K. Merton, edited by N.W. Storer, 267–278. Chicago, IL: University of Chicago Press.

Miller, B. 2015. “Why (Some) Knowledge is the Property of a Community and Possibly None of its Members.” The Philosophical Quarterly 65 (260): 417–441.

Nelson, R. R. (1959) 2002. “The Simple Economics of Basic Scienti c Research.” In Science Bought and Sold: Essays in the Economics of Science, edited by P. Mirowski and E.-M. Sent, 151–164. Chicago, IL: University of Chicago Press.

Noë, A. 2009. Out of Our Heads. Why You are Not Your Brain, and Other Lessons from the Biology of Consciousness. New York: Hill and Wang.

Offe, C. (2001) 2012. “Whose Good is the Common Good?” Philosophy & Social Criticism 38: 665-684.

O’Regan, J. K., and A. Noë. 2001. “A Sensorimotor Account of Vision and Visual Consciousness.” Behavioral and Brain Sciences 24: 939–973.

Pagano, U. 2014. “The Crisis of Intellectual Monopoly Capitalism.” Cambridge Journal of Economics 38: 1409–1429.

Popper, K. R. 1972. Objective Knowledge: An Evolutionary Approach. London: Oxford University Press.

Radder, H. 1993. “Science, Realization and Reality: The Fundamental Issues.” Studies in History and Philosophy of Science Part A 24: 327–349.

Radder, H. 1996. In and About the World. Philosophical Studies of Science and Technology. Albany: State University of New York Press.

Radder, H. 2002. “How Concepts Both Structure the World and Abstract from It.” Review of Metaphysics 55: 581–613.

Radder, H. 2004. “Exploiting Abstract Possibilities: A Critique of the Concept and Practice of Product Patenting.” Journal of Agricultural and Environmental Ethics 17: 275–291.

Radder, H. 2006. The World Observed/The World Conceived. Pittsburgh, PA: University of Pittsburgh Press.

Radder, H. 2009a. “Science, Technology and the Science–Technology Relationship.” In Philosophy of Technology and Engineering Sciences, edited by A. Meijers, 65–91. Amsterdam: Elsevier.

Radder, H. 2009b. “Why Technologies are Inherently Normative.” In Philosophy of Technology and Engineering Sciences, edited by A. Meijers, 887–921. Amsterdam: Elsevier.

Radder, H., ed. 2010a. The Commodi cation of Academic Research: Science and the Modern University. Pittsburgh, PA: University of Pittsburgh Press.

Radder, H. 2010b.“The Commodification of Academic Research.”In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 1–23. Pittsburgh, PA: University of Pittsburgh Press.

Radder, H. 2010c. “Mertonian Values, Scientific Norms, and the Commodification of Academic Research.” In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 231–258. Pittsburgh, PA: University of Pittsburgh Press.

Radder, H. (1984/1988) 2012. The Material Realization of Science. From Habermas to Experimentation and Referential Realism, revised edition, with a new postscript. Dordrecht: Springer.

Radder, H. 2013. “Exploring Philosophical Issues in the Patenting of Scientific and Technological Inventions.” Philosophy & Technology 26: 283–300.

Radder, H. 2014. “De cognitieve drie-eenheid: Realiseren, interpreteren en abstraheren [The Cognitive Trinity: Realization, Interpretation, and Abstraction].” Algemeen Nederlands Tijdschrift voor Wijsbegeerte 106: 3–36.

Radder, H. 2016.“Technological Systems and Genuine Public Interests.”In Social Epistemology and Technology: Toward Public Self-awareness Regarding Technological Mediation, edited by F. Scalambrino, 27–36. London: Rowman and Littlefield International.

Resnik, D. B. 2007. The Price of Truth. How Money Affects the Norms of Science. New York: Oxford University Press.

Romer, P. M. 1990. “Endogenous Technological Change.” Journal of Political Economy 98: S71–S102.

Sterckx, S. 2000. “European Patent Law and Biotechnological Inventions.” In Biotechnology, Patents and Morality. 2nd ed., edited by S. Sterckx, 1–112. Aldershot: Ashgate.

Sterckx, S. 2010. “Knowledge Transfer from Academia to Industry Through Patenting and Licensing: Rhetoric and Reality.” In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 44–64. Pittsburgh, PA: University of Pittsburgh Press.

Sterckx, S., and J. Cockbain. 2012. Exclusions from Patentability. How Far Has the European Patent Office Eroded Boundaries? Cambridge: University of Cambridge Press.

Steup, M. (2005) 2014. “Epistemology.” In The Stanford Encyclopedia of Philosophy. Spring 2014 ed., edited by E. N. Zalta. Accessed August 8, 2016. http://plato.stanford.edu/archives/spr2014/entries/epistemology/

Stiglitz, J. E. 1999. “Knowledge as a Global Public Good.” In Global Public Goods: International Cooperation in the 21st Century, edited by I. Kaul, I. Grunberg, and M. A. Stern, 308–325. New York: Oxford University Press.

Van den Belt, H. 2009. “Philosophy of Biotechnology.” In Philosophy of Technology and Engineering Sciences, edited by A. Meijers, 1301–1340. Amsterdam: Elsevier.

Van den Belt, H. 2010. Robert Merton, Intellectual Property, and Open Science. A Sociological History for Our Times. In The Commodification of Academic Research: Science and the Modern University, edited by H. Radder, 187–230. Pittsburgh, PA: University of Pittsburgh Press.

Vermeir, K. 2013. “Scientific Research: Commodities or Commons?” Science & Education 22: 2485–2510.