Introduction

A prominent characteristic of cognitive science is the collaborative nature of its research efforts. Cognitive scientists specializing in diverse areas often collaborate in cross-disciplinary research projects, traversing the boundaries between multiple disciplines. The multidisciplinary character of cognitive science is essentially related to its object of study: cognition. The task of studying cognition covers such a broad spectrum of topics as to require a division of labor accomplished through collaborative research (Von Eckardt, 2001). This multidisciplinarity highlights a social sense of cognitive science research; in particular, it is suggestive of a social sense of the knowledge that proceeds from the outcome of this research. Social knowing, then, appears to play a significant role in the intellectual structure of cognitive science; it facilitates the coordination underlying the aim of cognitive science, to study and gain knowledge about cognition, or to put it in the words of David Rumelhart, to gain “a clearer picture of mind and cognition” (Baumgartner & Payr, 1995, p. 200).

While the multidisciplinary character of cognitive science is beneficial towards seeking this aim, it presents a significant problem when approaches in social epistemology are applied to cognitive science. In social epistemology, approaches to social knowing generally assume that what is socially known in a group depends, in varying degrees, on uniformity of what is individually known or believed[1] among the members of that group (e.g. The Strong Programme: Barnes & Bloor, 1982; Goldman, 1987; Shapin, 1995). Another way of stating this is that what is socially known supervenes on the mental states of individuals. Given the multidisciplinary character of cognitive science, what is individually known among its members greatly varies. This diversity of individual knowledge presents problems for an account of social knowing in cognitive science. If social knowing depends on supervenience, then the conclusion that seems to follow is that cognitive science either knows very little or nothing at all.

What is really at stake here with the problem of social knowing in cognitive science? On one extreme, the intelligibility of research in cognitive science is at stake. If cognitive science does not have the relevant social properties sufficient for social knowledge, then it raises doubts about the epistemological coherence of its research. However, if an account of social knowing is to be found, this would be an important contribution to understanding the intellectual structure of cognitive science. On the other side of this problem, the significance of the concept of ‘social knowing’ is at stake. If cognitive science can be shown to have a coherent social structure while having such diversity in what is known by individual cognitive scientists, then this raises doubts about the relevance of a supervenience approach to social knowing. Perhaps social knowing does not, in fact, supervene on the mental (including epistemic) states of individuals. The question of whether cognitive science knows anything, then, is a question worthy of some attention.

The goal of this paper is to address the problem of social knowing in cognitive science. The focus is to identify the issues underlying the problem of social knowing in cognitive science and to argue for the analogous approach to social knowing, developed by Alexander Bird (2011), as a solution to the problem. The first task is to elucidate the problem of social knowing in cognitive science. This involves (a) clarifying the problem of social knowing in general, (b) a discussion of the collaborative nature research in cognitive science, (c) determining the sense of multidisciplinarity relevant to cognitive science, and (d) to demonstrate the incompatibility of a supervenience approach to social knowing with this sense of multidisciplinarity.

The second task is to provide an overview of the analogous approach to social knowing and to outline the main arguments Bird makes for such an approach. Bird argues that social knowing is not supervenient on, nor reducible to, individual knowing. The relationship he poses is one of structural analogy where social knowing stands to society as individual knowing stands to the individual—social knowing is analogous to individual knowing.

The third task is to apply this approach to cognitive science. Critically, this section of the paper evaluates the intellectual structure of cognitive science grounded in a discussion of the cognitive division of labor. It is argued that the notion of common ground (Clark, 1996) in a field such as cognitive science is crucial to an understanding of the coordination of its research efforts. This has important consequences for the role of evidence in the carrying out of research in cognitive science.

The Problem of Social Knowing

To better understand the problem of social knowing in cognitive science in particular, it will be useful to first motivate the problem of social knowing in general. Providing a comprehensive overview of the major positions in social epistemology is well beyond the scope of interest for the current discussion—especially since Bird (2011) provides such an overview. The treatment of social epistemology in this section primarily addresses the question of what social knowledge is and secondarily to outline the main claims Bird presents about social knowing that are relevant to the essential focus of the paper: the problem of social knowing in cognitive science.

Social epistemology emphasizes the social context of epistemic states (Goldman, 1987). One particular kind of social epistemology, what Bird calls ‘social-social epistemology’, considers groups and social structures as legitimate possessors of epistemic states (Bird, 2011, p. 2). This approach emphasizes the social nature of the epistemic subject. For example, the epistemic subject of interest in the phrase ‘Google knows how to build autonomously-driven cars’ is the group ‘Google’. The problem of social knowing is to account for the relationship between what is socially known in a society or group and what is individually known. In other words, how closely tied are social epistemic states to the mental states of individuals?

If a close relationship between social knowing and individual knowing is assumed, then what is socially known will depend strongly on what is individually known. Bird presents this claim as:

The implication of (K1) is that what is socially known in a group is also universally known by the members of the group. Something would indeed be socially known if it were universally known, however, (K1) also implies that something is not socially known until it is universally known. This is certainly not the case for the example given previously. Google knows how to build an autonomous car even though it is not universally known by individuals at Google.

A less extreme claim can be stated as:

The advantage of (K2) is that it allows for a more moderate dependence of social knowing on individual knowing to be assumed. The implication is that something is socially known once most people know it. Even though (K2) posits a more moderate dependence, it can still be rejected using the example from Google. The individuals who are working on the autonomous-car project are only a small portion of the individuals at Google.

Neither the strong dependence of (K1) nor moderate dependence of (K2) is able to account for social knowing. Bird also considers a claim from the opposite spectrum:

The weak dependence in (K3) seems, at the very least, in accord with intuition. If p is socially known by a group, then it seems intuitive that at least someone individually knows p. A problem with (K3) is that it does not support an explanation of how something becomes socially known. If at least some individual at Google knew underwater basket weaving, is it then necessary that we attribute this knowledge as something socially known by Google? (K3) has the odd implication that whatever is individually known is also socially known.

The intuitive appeal of (K3) can be preserved while avoiding its odd implication with the following claim:

As Bird notes, commitment to (SUP) is very wide. While (SUP) does not directly imply the odd implication of (K3), it is also ambiguous as to how the mental states of individuals actually add to what is socially known. There still seems to be a dependence, though not as strong as (K1), on some uniformity among the individuals of a group. Out of the claims presented so far, (SUP) is the most likely candidate for an account of the problem of social knowing in cognitive science. Two related facts about cognitive science are problematic with regard to (SUP): the collaborative nature of research in cognitive science and the multidisciplinary character of cognitive science.

Collaborative Research in Cognitive Science

The inter-dependence of collaborative research has been instrumental in cognitive science even from the field’s nascent stages. Stephen Palmer, a major figure in the development of cognitive science, in reference to his early collaborations with computer scientists as a graduate student in psychology said, “We were doing two pieces of cognitive science, in the sense that we were worried not only about human psychology but also about how to do computer simulation models” (Baumgartner & Payr, 1995, p. 157). By the time of the founding of the Cognitive Science Society in 1977, there was already an increasing amount of multidisciplinary cognitive research going on—a trend that would only continue to grow. Even early on, there was skepticism about whether this large amount of research represented a coherent intellectual enterprise (Von Eckardt, 1996). Such skepticism seems to stem from a lack of uniformity in what individual cognitive scientists know and believe.

In light of this skepticism, should cognitive science be striving towards more uniformity? This would seem to make cognitive science more compatible with the claims presented in the previous section. Such uniformity, however, would be contrary to what many consider to be an essential characteristic of cognitive science: its interdisciplinary division of labor (van Gelder, 1998; Von Eckardt, 1996, 2001). Hillary Putnam, another instrumental figure in the development of cognitive science, was asked if he regarded the cross-disciplinary work in cognitive science as exciting, he replied, “That is exciting, but as soon as someone tries to say what all cognitive scientists believe, or that we all have to agree with Chomsky or Minsky, then it is going to become a millstone around our necks rather than something good” (Baumgartner & Payr, 1995, p.183).

Conforming to the implications of (SUP) does not seem to be the best way to resolve the problem of social knowing in cognitive science. In fact, were such uniformity achieved, there would be no cognitive science at all; it would end up being subsumed by another discipline altogether. What is it, then, about the multidisciplinary character of cognitive science that makes it so incompatible with (SUP)? There are multiple senses of multidisciplinarity, not all of which are incompatible with (SUP). In the next section, the issue of multidisciplinarity in cognitive science is addressed with a critical eye towards its compatibility with (SUP).

Multidisciplinarity in Cognitive Science

Von Eckardt (2001) makes a distinction between at least two conceptions of multidisciplinarity as it applies to cognitive science. The first conception she identifies as the localist conception:

The sense of multidisciplinarity in (M1) is one of a characteristic that applies directly to individual cognitive scientists. Individual research efforts can only contribute to cognitive science if they employ the methods and findings from multiple, distinct disciplines. For a piece of research to count as evidence under (M1), it must be multidisciplinary (i.e. unidisciplinary research is not acceptable as evidence). The locus of multidisciplinarity is in the individual research efforts (Von Eckardt, 2001, p. 455).

The alternative to the localist conception that Von Eckardt offers is that of a holist conception:

Under this conception, multidisciplinarity is a characteristic of cognitive science as a whole rather than a characteristic required of each individual member. This allows for unidisciplinary as well as multidisciplinary research to count as evidence. The locus of multidisciplinarity for (M2) is in the collective research effort (Von Eckardt, 2001, p. 455).

(M1) and (M2) contrast in their compatibility with regard to (SUP). If social knowing supervenes on individual knowing, then a group must consist of individuals of somewhat uniform knowledge and belief in order to possess the relevant social properties sufficient for social knowing. Since the locus of multidisciplinarity is in the individual for (M1), this conception is compatible with (SUP). (M2), on the other hand, implies a division of labor among individual researchers that is not compatible with (SUP)—what is socially known depends on collaboration between individuals, not just on individual mental states.

Von Eckardt (2001) follows up the distinction between (M1) and (M2) with a critical evaluation of the type of research published in the journal Cognitive Science. She concludes that (M1) does not characterize the multidisciplinary nature of cognitive science research. The picture of the solitary scientist that is implied by (M1) is unrealistic to begin with. It is implausible at best to think that a solitary scientist, especially a cognitive scientist, is capable of obtaining all the relevant evidence they need (Polanyi, 1950). Some have even argued that such a view of the autonomous scientist is a flat out myth (e.g. Hardwig, 1991). The collaboration of research efforts implied by (M2) more accurately characterizes the multidisciplinary nature of cognitive science.

Supervenience and Social Knowing

Given (M2), the implication of (SUP) for cognitive science is that very little or nothing is socially known in cognitive science. This is quite odd given the social nature of cognitive science research. Either cognitive science lacks a coherent intellectual structure or (SUP) is false. Bird (2011) argues that (SUP) makes social epistemic states depend too closely on the states of individuals. In this section, one of Bird’s several demonstrations of the failure of (SUP) to account for social knowing will be presented. His arguments against (SUP) support the radical claim he makes that individual knowing is not even necessary for social knowing—social knowing is analogous to individual knowing.

Bird describes two scenarios where a researcher makes a discovery, and at some later point in time, that discovery becomes widely read and discussed.[2]In the first scenario, Dr. N., a researcher in mainstream science, makes a discovery while working in an obscure field and publishes it in the field’s journal through the normal peer review process. After some time, Dr. N. dies and everyone who has read or heard of the discovery has either died or forgotten all about it. Professor O, at an even later time, stumbles upon Dr. N.’s discovery after searching the indexes and cites it in their own work causing Dr. N.’s discovery to be read and cited by many (Bird, 2011, pps. 12-13).

Dr. N.’s discovery functions as a contribution to scientific knowledge. The discovery was certainly socially known by the end of the scenario. The status of the discovery as a contribution to scientific knowledge did not directly depend on the mental states of individuals; being in the relevant public domain, the discovery remained social knowledge even when it was not individually known by anyone. (K3) requires that at least someone has to know something for it to be socially known. (K3) is thus refuted by this counterexample.

The second scenario Bird describes is the case of Dr. Q., a secretive military scientist. Dr. Q makes a classified discovery—known only by himself—which he types up and locks in a drawer. Several years after his death, someone stumbles upon the discovery which turns out to be directly relevant to their work. They arrange for Dr. Q.’s discovery to get published and subsequently cite it in their own research, to which many others begin to read and cite (Bird, 2011, p. 13).

Dr. Q.’s discovery does not become a scientific contribution until it is published. Only then is it able to constitute social knowledge. When these two scenarios are considered together, Bird cleverly demonstrates that social knowing does not supervene on individual knowing. The mental states of the characters in both stories are the same during the intervening period yet the status of either discovery as social knowledge differed. (SUP) fails to account for this difference.

Analogical Approach to Social Knowing

Applying (K1-3) and (SUP) to cognitive science led to the undesirable conclusion that cognitive science either knows very little or nothing at all. The previous section demonstrated the failure of (K1-3) and (SUP) to provide adequate explanations of social knowing in the first place. A solution to the problem of social knowing in cognitive science is found in Bird’s analogous approach to social knowing. This section provides an overview of this approach, outlining the key components of this framework. These include the distinction between established and organic groups, the functional role of knowing, and the properties of a social cognitive structure. It is not within the scope of this paper to detail all of the arguments Bird provides in defense of this view, especially as they are presented in the original paper. The focus is rather on those arguments which are most relevant to the application of this framework to cognitive science.

Group Solidarity

An important consideration of social knowing that has not been addressed yet is an account of group cohesion—what it is that distinguishes the kind of group that can possess social epistemic states from a mere collection of individuals. Bird argues that an account of cohesion should play a significant role in explaining how a group can be said to possesses epistemic states (Bird, 2011, p. 19). His account of cohesion involves a distinction between two kinds of groups: established and organic.

Established groups get their cohesion from joint commitment (Bird, 2011, p. 18). It is by virtue of joint commitment that an established group can possess collective epistemic states. Likewise, the connection between what is socially known and what is individually known will be close one for established groups. Bird writes, “It is because an established group exists in virtue of its joint commitment that for it to have a belief requires joint commitment to that belief” (Bird, 2011, p. 19). These groups can be seen to satisfy (SUP). Bird notes, however, that established groups do not exhaust the social groups that may be possessors of social epistemic states; approaches to social knowing that are committed to (SUP) often overlook the fact that there are non-established groups that possess social epistemic states as well (Bird, 2011, p. 18).

Groups such as scientific communities (e.g. cognitive science) have cohesion through organic solidarity. Organic solidarity involves inter-dependence brought about by the division of labor; the key feature of division of labor is that “individuals and organizations depend on others who have different skills and capacities” (Bird, 2011, p. 20). Groups whose cohesion is rooted in division of labor are referred to as organic groups. While there may exist established groups within a given scientific community, social knowing within a scientific community is rooted in organic solidarity. The very intellectual structure of science, Bird argues, is held together by organic solidarity (Bird, 2011, p. 20). Multidisciplinarity and collaborative research were seen in previous sections to present the most difficulty for an account of social knowing in cognitive science for (SUP). Given the distinction between established and organic groups, these characteristics can be understood as facilitating social knowing as opposed to hindering it.

Functional Role of Knowing

If social knowing is not supervenient on individual knowing, then what exactly is the connection between the two? Bird proposes the relationship to be one of structural analogy: social knowing is analogous to individual knowing. The analogy is rooted in the function of knowledge. Bird begins to develop the analogy through a hypothesis about the function of individual and social knowledge: “(individual) knowing plays a particular role in the structure of one’s ‘mental economy’ and that social knowing plays a structurally parallel role in a social analogue” (Bird, 2011, p. 25).

Bird specifies the role of knowledge in cognition as follows: “the function of the cognitive faculties is to produce knowledge as their outputs and correspondingly the appropriate input into reasoning is also knowledge” (Bird, 2011, p. 25). Cognitive faculties are interdependent in that the outputs of some faculties provide the inputs for others. For example, when a sense organ receives a signal from the environment, its receptors encode and process that signal and pass its output along the neural pathway, depending on which sense organ, progressively encoding inputs, processing, and passing along outputs as it travels to higher faculties. The social analogue is not necessarily found in the hierarchical structure of human cognition but mainly in the division of labor; the output for one faculty is used as the input for another.

Social Cognitive Structures

The social analog to individual cognitive faculties is found in what Bird calls social cognitive structures. These structures are characterized by the following properties:
These three properties support the interdependence brought on by the cognitive division of labor in organic groups.

Cognitive science can be seen to exercise these three properties in the carrying out of its research efforts. The characteristic output of research in cognitive science is propositional. Research is published in journal articles, presented at conferences, and catalogued in textbooks. Contributions to the cognitive science are made available in the public domain propositionally. There are some outputs that are not always propositional. The division of labor in cognitive science can sometimes involve large-scale computational modeling. What is made available to fellow researchers in these efforts may not always be expressed propositionally. Bird also identifies outputs in scientific communities that may not always be propositional. These include experimental and computational techniques, apparatus, and personal skills. The non-propositional outputs in cognitive science—and other scientific communities—facilitate the coordination of research efforts between individual scientists and may be considered as part of their common ground[3] (e.g. Clark, 1996). Bird’s main point is not to claim that all of the outputs in a scientific community will be propositionally, but that the characteristic outputs are.

There are mechanisms in cognitive science that satisfy the truth-filtering property of a social cognitive structure. Bird identifies the peer-review process as well the journal referee as mechanisms put in place to eliminate error. These are direct mechanisms of maintaining the intellectual quality of research. There are also more indirect mechanisms that can be identified. The conventions that are adopted by researchers in a scientific community can include methods for reporting data analyses, specific ways of discussing a given phenomena, and, as Bird mentions, the writing guidelines for publications. These are more indirect mechanisms of quality control.

The multidisciplinary character of cognitive science exemplifies the third property of social cognitive structures. Collaborative research often involves the coordination of evidence from previous research efforts—the outputs function as inputs for further research and the outputs from multiple disciplines become the inputs for cognitive research. The output of research in cognitive science also becomes the input for social action, especially in areas where research on education and technology intersect. Overall, cognitive science can be seen to possess the necessary properties of a social cognitive structure.

How to Do Things with Evidence

An analogous view of social knowing is pertinent to the question of whether cognitive science knows anything as well as to the question of how cognitive science knows anything, especially with the cognitive division of labor needed for the study of cognition. The need to rely on one another's evidence in cognitive science presents a coordination problem (Lewis, 1969) whereby scientists must carry out their research using shared conventions (including experimental techniques, statistical methods, etc.). As mentioned in the discussion of social cognitive structures, the collection of these conventions constitutes the common ground among members of a group. Common ground is crucial for maintaining the intellectual structure of a scientific group. Bird writes, “Scientists will not be able to rely on one another's results, engage in fruitful dialogue, or collaborate on a project unless they share these beliefs, standards, and techniques (Bird, 2011, p.21).”

Social Nature of Evidence

Bird’s account of the functional role of knowing highlights the social nature of evidence. Scientific knowledge is achieved through inference from evidence (Bird, 2010. When an individual in a scientific community puts forth some knowledge as evidence, they are performing two important social acts. These two social acts have to do with Austin’s (1962) notion of an expositive act and a commissive act. In How to Do Things with Words, Austin describes expositives as “involving the expounding of views, the conducting of arguments, and the clarifying of usages and of references” (Austin, 1962, p. 160). Commissives are performed to commit the speaker (for our purposes, a researcher) to a certain course of action (Austin, 1962, p. 156). Though Austin does not put forth these classes as definitive, they are informative for the current purpose of analyzing how members of a scientific group do things with evidence—specifically, how they further the aim of science using evidence.

When an individual member of a scientific group puts forth a proposition p as evidence, on an Austinian view, it could be said that they are performing both a commissive and an expositive act. In general, the expositive is the clarifying reasons for p while the commisive is the commitment to establish p as something that could be socially known by the scientific group. Particularly for cognitive science, the commisive is that p be considered as evidence in multiple disciplines. In other words, the expectation in multidisciplinary research is that the output of that research be relevant across disciplines. This is due to the social cognitive structure of cognitive science; the aim of cognitive science is carried out by the cognitive division of labor among its members.

How to Not Do Things with Evidence

The social nature of evidence is seen here to be contingent upon the organic solidarity within a scientific community. The coordination problem in research is resolved through conventions grounded in the cognitive division of labor—the inter-dependence in organic groups. Problems can arise in a scientific community when evidence is, instead, contingent upon joint commitment within an established group. For example, early experimental psychology was faced with the coordination problem of how to coordinate talk of the mind. Behaviorism emerged as an established group whose cohesion was grounded in the joint commitment to simply not talk about the mind (Robinson, 1995). Evidence became contingent upon this behaviorist convention and compromised the intellectual structure of psychology all together; uniformity in convention overshadowed the cognitive division of labor.

The behaviorist convention is an example of a convention that tries to solve a coordination problem through exclusion. When joint commitment replaces organic solidarity as the cohesion of a scientific community, then the social epistemic states of that scientific community are constrained to (SUP). The problem with this is that (SUP) can be seen to undermine function of evidence in light of the current discussion. Because (SUP) requires that social knowing depend on the epistemic states of individuals, and since joint commitment to a proposition can restrict the individuals that are even able to contribute, cognitive division of labor is drastically limited. In cognitive science, this would be a very undesirable consequence since research efforts depend on collaboration and multidisciplinarity.

Such restrictions can also occur when joint commitment to a particular terminological notion prevents individual researchers from coordinating their research efforts. Berkeley (2008) discusses this issue concerning notions of the term ‘symbol’. Overemphasis on commitment to a particular notion of symbol resulted in little agreement and much confusion among individual researchers. The social function of evidence is undermined when researchers are unable to agree on terminology. The solution that Berkeley argued for was to develop a convention grounded in the collaborative division of labor. Specifically, it was to develop and introduce terminological conventions that resolve the coordination problem through increasing the opportunity for coordination rather than restricting it to adherence to a joint proposition.

Conclusion

Two things were identified at the beginning of this paper as being potentially at stake in a critical analysis of the problem of social knowing in cognitive science: either the epistemological coherence of cognitive science or the supervenience approach to social knowing. Adopting an analogous approach to social knowing provides a framework for better understanding the intellectual structure of cognitive science. Without compromising the diversity of its members (its essential characteristics), cognitive science can be seen to possess social epistemic states by virtue of its organic solidarity. The casualty of the current analysis is the relevance of a supervenience approach to social knowing in organic groups.

Works Cited

Austin, J. L. (1962). How to Do Things with Words. Oxford: Clarendon Press.

Barnes, B., & Bloor, D. (1982). Relativism, Rationalism. In M. Hollis & S. Lukes (Eds.), Rationality and Relativism (Vol. 63, pp. 21-47). Cambridge, MA: MIT Press.

Baumgartner, P., & Payr, S. (1995). Speaking Minds?: Interviews with Twenty Eminent Cognitive Scientists. Princeton N.J.: Princeton University Press.

Berkeley, I. S. N. (2007). What the <0.70, 1.17, 0.99, 1.07> is a symbol? Minds and Machines, 18(1), 93-105.

Bird, A. (2010). The epistemology of science—A Bird’s-eye view. Synthese, 175, 5-16.

Bird, A. (2011). Social knowing: The social sense of “scientific knowledge.” Philosophical Perspectives, 25(1), forthcoming.

Clark, H. (1996). Using language. Cambridge, England: Cambridge University Press.

Feyerabend, P. (1975). Against Method?: Outline of an Anarchistic Theory of Knowledge. London: Humanities Press.

Garrod, S., & Doherty, G. (1994). Conversation, co-ordination and convention: an empirical investigation of how groups establish linguistic conventions. Cognition, 53(3), 181-215.

Goldman, A. I. (1987). Foundations of social epistemics. Synthese, 73(1), 109-144.

Hardwig, J. (1991). The role of trust in knowledge. The Journal of Philosophy, 88, 693-708.

Holyoak, K. J., & Thagard, P. (1997). The analogical mind. American Psychologist, 52, 35-44.

Lewis, D. (1969). Convention?: A Philosophical Study. Cambridge, MA: Harvard University Press.

Polanyi, M. (1950). Scientific beliefs. Ethics, 61(1), pp. 27-37.

Polanyi, M. (1961). Knowing and being. Mind,, 70(280), 458-470.

Shapin, S. (1995). Here and everywhere: Sociology of scientific knowledge. Annual Review of Sociology, 21, 289-321.

van Gelder, T. (1998). The roles of philosophy in cognitive science. Philosophical Psychology, 11, 117-136.

Von Eckardt, B. (2001). Multidisciplinarity and cognitive science. Cognitive Science, 25, 453-470.

Von Eckardt, Barbara. (1995). What is cognitive science? Cambridge, MA: MIT Press.

Appendix

(Case of Dr. N.) Dr. N. is working in mainstream science, but in a field that currently attracts only a little interest. He makes a discovery, writes it up and sends his paper to the Journal of X-ology, which publishes the paper after the normal peer-review process. A few years later, at time t, Dr. N. has died. All the referees of the paper for the journal and its editor have also died or forgotten all about the paper. The same is true of the small handful of people who read the paper when it appeared. A few years later yet, Professor O. is engaged in research that needs to draw on results in Dr. N’s field. She carries out a search in the indexes and comes across Dr. N.’s discovery in the Journal of X-ology. She cites Dr. N.’s work in her own widely-read research and because of its importance to the new field, Dr. N.’s paper is now read and cited by many more scientists (Bird, 2011, pps. 12-13).

(Case of Dr. Q.) Dr. Q. carried out his research in secret. He typed up his results and locked them in a drawer. He died; only ten years later did someone, by chance, come across his typescript in an attic. The results eventually reached the hands of Professor R, who arranged for their publication and cited them in her work, as a result of which they were read and discussed widely (Bird, 2011, p. 13).