Epistemic Structure in Non-Summative Social Knowledge

ABSTRACT

How a group G can know that p has been the subject of much investigation in social epistemology in recent years. This paper clarifies and defends a form of non-supervenient, non-summative group knowledge: G can know that p even if none of the members of G knows that p, and whether or not G knows that p does not locally supervene on the mental states of the members of G. Instead, we argue that what is central to G knowing that p is whether G has an epistemic structure that is functioning appropriately in accord with the action-related purposes of the group, and this structure may include non-agential elements such as devices that retain or process information. We argue that recent objections to non-summative group knowledge given by Jennifer Lackey do not in fact succeed in undermining the view, but do help to clarify the nature of non-summative group knowledge. The main upshot of our response to Lackey’s objections is that groups put their knowledge into action in ways that often differ from how individuals do, and social epistemologists should be careful to notice these differences, especially insofar as groups often structure themselves by employing various epistemically-relevant devices.

KEYWORDS:

• Group knowledge
• Alexander Bird
• Jennifer Lackey
• distributed cognition

Acknowledgment

Many thanks to three anonymous referees from this journal for extremely helpful comments that greatly improved the paper.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1. For an influential account of collective (non-summative) group belief, see Margaret Gilbert (1994, 2000, 2002, 2004, 2009, 2013).

2. For (non-exhaustive) major contributions to this literature on non-summativism, see Alexander Bird (2010, 2014), Jeroen de Ridder (2014, 2022), Hakli (2007), List and Pettit (2011), Palermos (2020), Rolin (2008), Deborah Tollefsen (2002, 2004, 2007a), Raimo Tuomela (1992, 2004, 2006, 2011), and K. Brad Wray (2001, 2007).

3. For those interested in a deeper elaboration and defense of non-summativist views of knowledge, other sources may be consulted (see especially Tollefsen 2002).

4. Thanks to an anonymous referee for their helpful suggestions for this section.

5. For an interesting debate about which parts of the scientific community can possess non-summative collective knowledge, see Wray (2007) and Rolin (2008). Wray argues that small research teams are the only parts of a scientific community to which non-summative collective knowledge can be attributed (Wray 2007, 342–3). Conversely, Rolin argues that ‘some assumptions adopted by scientific communities are properly understood as collective knowledge in a plural subject sense’ (Rolin 2008, 119).

6. de Ridder (2022) offers a recent defense of joint-commitment accounts, although his joint-commitment account is a modified version which differs considerably from those that have been traditionally offered in the literature.

7. Additionally, Palermos (2020) focuses on cases of continuous and reciprocal interaction between agents within a group; at least some of the cases of social knowledge that we are concerned with do not have that feature.

8. See Chapter 8 of Hutchins (1995) for an in-depth discussion of the distribution of epistemic labor on the USS Palau and its navigation of San Diego harbor.

9. Palermos and Pritchard (2013) and Palermos (2016) are also explicit in discussing the role that artifacts play in distributed cognition.

10. An emphasis on epistemic communities as the bearers/producers of knowledge (rather than individuals) can be found in the work of Lynn Hankinson Nelson (1990, 1993, 1995), who writes: ‘ … [epistemic] communities are the primary loci – the primary generators, repositories, holders, and acquirers – of knowledge’ (Nelson 1993, 124). In a similar vein, Helen Longino (1990) provides a provocative critique of the supposed ‘objectivity’ of the scientific method, and, like Nelson, advocates for the primacy of groups vs. individuals as the base unit of analysis. Longino writes: ‘Refocussing on science as practice makes possible the second shift, which involves regarding scientific method as something practiced not primarily by individuals but by social groups’ (Nelson 1990, 66–7). Thanks to an anonymous referee for pointing out the historical significance of these sources.

11. There has been some recent debate over whether the scientific community can be said to ‘assert’ any proposition. For more on this, cf. Dang and Kofi Bright (2021).

12. Such policy changes are plausibly held to be ‘non-epistemic’ consequences. For a helpful discussion of non-epistemic consequences of scientific assertion, see Franco (2017).

13. In her recent review of Lackey’s The Epistemology of Groups, Tollefsen offers convincing reasons to consider the GEAA a ‘sophisticated summative account’ (Tollefsen 2021, 4).

14. The term ‘operative member’ is taken from the work of Raimo Tuomela and his non-summative, joint acceptance account of justified group belief (quoted in Lackey 2020, 27).

15. Thanks to an anonymous referee for pushing us to clarify our argument here.

16. This case is adapted very slightly from Lackey (2020, 130). We have added the title ‘MISSING CHILD’ to this example in order to more easily refer to it throughout this section.

17. One way to think of this issue is to appeal to the distinction between propositional and doxastic justification. See Silva and Oliveira (forthcoming) for an overview. The CPD could be said to have propositional justification without doxastic justification (or a proxy thereof).

18. See Cohen (1995) for a classic source on this distinction; also cf. Dang and Kofi Bright (2021).

19. We should point out an inadequacy in Lackey’s critique of Bird’s account of the knowledge of scientific communities (Lackey 2020, §3.3). Lackey argues that, on Bird’s view, the minds of group members cannot contribute positively to group knowledge, but can contribute negatively (as defeaters), and this seems theoretically deficient. However, we cannot see the force of this objection: given that former members of a group did publish the paper in question, minds did positively contribute; if the paper in question merely materialized without anyone writing it (or justifying its contents), then the scientific community should not be said to know its contents. And the reason why current minds don’t positively contribute to knowledge of long-ago published facts is because it is already known by the scientific community. Secondly, Lackey argues that, given that members of the scientific community are constantly questioning previous findings, the scientific community may be said to know much less than one might think. But we believe that this is a welcome conclusion; science involves much less knowledge than one might believe (see Dang and Kofi Bright 2021).

20. In a footnote (Lackey 2020, 115, fn. 6), Lackey notes that Bird does not in the end require accessibility as a condition on group knowledge, but nevertheless Lackey goes on to include an accessibility condition in the account of SK that she attributes to Bird and which she goes on to criticize.

21. Cf. Ritchie (2015, 313) for a similar example and discussion.

22. Bird (2010, §3) writes of the scientific community having a Durkheimian ‘organic solidarity’. On our own view, all that is required for group knowledge is enough structure so that the members of the group can collaborate to perform its function. And Bird himself does not require the kind of dynamical integration such as in Palermos (2016).

23. There is an interesting parallel here with Grice’s (1989) notion that conversations are cooperative endeavors insofar as the participants have shared goals, even if the goals aren’t made fully explicit.

24. We thank two anonymous referees for their suggestions in clarifying this section.

25. Giere accepts that there is distributed cognition, but he denies that there is an irreducible group cognitive agent (2002, 2007) that possesses knowledge (2006, 2007). Thanks to an anonymous referee for suggestions here.

26. See also Hiller (2013) for a structure-based view of group ontology.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article

Notes on contributors

Avram Hiller

Avram Hiller is an Associate Professor of Philosophy at Portland State University. He works in many areas in analytic philosophy.

R. Wolfe Randall

R. Wolfe Randall is a philosophy graduate student at University of California, Santa Barbara. His current research examines social epistemology, social ontology, distributed cognition, and normative questions regarding public advocacy.