CC BY
21
Эпистемология и философия науки 2019. Т. 56. № 2. С. 60-76 УДК 167.7
Epistemology & Philosophy of Science 2019, vol. 56, no. 2, pp. 60-76 DOI: 10.5840/eps201956230
Beyond the 'null setting':
the method of cases
in the epistemology of testimony
Axel Gelfert - PhD
in Philosophy, professor. Technische Universität Berlin. Straße des 17. Juni 135, 10623 Berlin, Germany. e-mail: a.gelfert@tu-berlin.de
Epistemologists of testimony have tended to construct highly stylized (so-called "null setting") examples in support of their respective philosophical positions, the paradigmatic case being the casual request for directions from a random stranger. The present paper analyzes the use of such examples in the early controversy between reductionists and anti-reductionists about testimonial justification. The controversy concerned, on the one hand, the source of whatever epistemic justification our testimony-based beliefs might have, and, on the other hand, the phenomenology of testimonial acceptance and rejection. As it turns out, appeal to "null setting" cases did not resolve, but instead deepened, the theoretical disputes between reductionists and anti-reductionists. This, it is suggested, is because interpreters 'fill in' missing details in ways that reflect their own peculiarities in perspective, experience, upbringing, and philosophical outlook. In response, two remedial strategies have been pursued in recent years: First, we could invert the usual strategy and turn to formal contexts, rather than informal settings, as the paradigmatic scenarios for any prospective epistemology of testimony. Second, instead of "null setting" scenarios, we can focus on richly described cases that either include, or are embedded into, sufficient contextual information to allow for educated judgments concerning the reliability and trustworthiness of the testimony and testifiers involved. The prospects of both of these approaches are then discussed and evaluated.
Keywords: social epistemology, testimony, reductionism, anti-reduc-tionism, method of cases, social context
л
за пределами «нулевой настройки»: ситуационный метод в эпистемологии свидетельства
Аксель Гелферт - доктор философии, профессор. Технический университет Берлина.
Straße des 17. Juni 135, 10623, Берлин, Германия. e-mail: a.gelfert@tu-berlin.de
Эпистемологи свидетельства склонны к разработке очень изощренных примеров (так называемых нулевых установок) в поддержку соответствующих философских позиций, причем основное направление здесь задает ориентир на «случайного незнакомца». В данной статье анализируется использование таких примеров в полемике об обосновании свидетельства между редукционистами и антиредуктивистами. Противоречия здесь возникают, с одной стороны, по поводу эпистемических оснований верований, основанных на свидетельствах, а с другой стороны, в связи с феноменологией признания и отрицания свидетельств. Автору представляется, что апелляция к случаям с «нулевой установкой» не только не способствует разрешению споров между редукционистами и антиредукционистами, но, напротив, лишь усугубляет их. По-видимому, это связано
60
© Axel Gelfert
с тем, что интерпретация оказывается «дополнена» самим интерпретатором - его мировоззрением, опытом, воспитанием и философскими взглядами. В этой связи в последние годы были разработаны две стратегии коррекции такой интерпретации. Первая предлагает вместо неформальных условий обратиться к формальным контекстам с целью выявления парадиг-мальных сценариев для всякой эпистемологии свидетельства. Вторая требует отказаться от сценария «нулевой установки» и сосредоточиться на хорошо описанных кейсах, которые включают в себя достаточную информацию о контексте, что позволяет сформулировать обоснованные суждения относительно надежности и достоверности свидетельских показаний. Автор анализирует перспективность каждой из названных стратегий. Ключевые слова: социальная эпистемология, свидетельство, редукционизм, антиредукционизм, ситуационный метод, социальный контекст
1. Introduction
Philosophical methodology, in recent years, has undergone something of a revival, in that it has moved beyond the confines of a specialist pursuit and has actively been taken up in various corners of philosophy. The present paper offers one such application - a fairly narrow one, it should be added - to the field of social epistemology. In particular, it discusses the method of cases, and its limitations, in the epistemology of testimony. Much of the early work in contemporary epistemology of testimony was based on the assumption that cases should be sufficiently 'stylized' (i.e., should abstract from empirical detail), so as to allow for generalizations about all, or at least most, of the testimonial information we receive. However, far from eliciting stable intuitions about when it is rational to trust someone's testimony, this emphasis on the so-called "null setting" [Adler, 2006] deepened the divide between (in the early phase of the debate) reductionists and anti-reductionists about testimonial justification, or so I shall argue. In recent years, alternative approaches have emerged that aim to sidestep various methodological issues, by looking beyond the null setting towards, on the one hand, empirically rich descriptions of socially situated practices and, on the other hand, examples from literature and film which offer contextual information and narrative unity.
The remainder of this paper is as follows. In Section 2, the relevant meta-philosophical background is sketched, including the controversy over whether certain philosophical methodologies, such as the method of cases, can be defended by appeal to a specialist notion of 'philosophical expertise'. Section 3 offers a sketch of contemporary epistemology of testimony, paying special attention to the argumentative dialectic between reductionists and anti-reductionists about testimonial justification. On the basis of a much-discussed example due to Tony Coady [1992], it is shown how cases that
conform to the "null setting" fail as arbiters between the two opposing camps. Section 4 looks at how contemporary scholars in the epistemology of testimony have moved beyond the early emphasis on stylized examples, turning instead to an analysis of socially situated practices and to literature and narrative for inspiration. While the "null setting" may still have a role to play for the analysis of certain types of testimony, its oversize influence in early epistemology of testimony has rightly diminished, or so I conclude.
2. Meta-Philosophical Background: Intuitions And Expertise
Metaphilosophical considerations tend to enter philosophical discourse at crucial junctures in the formation, or consolidation, of a subdiscipline or nascent tradition. Examples from twentieth-century Western philosophy might include the rise of logical empiricism, and its subsequent disintegration (and partial absorption) into a variety of subfields and approaches. Similarly, the divide between, say, analytical and phe-nomenological traditions within Western philosophy has, on occasion, given rise to methodological and metaphilosophical reflection, yet with the deepening of the divide over time, occasions for productive exchanges have diminished. In recent years, metaphilosophical discussions have flourished within what may be called 'mainstream analytical philosophy', and this, too, may be considered a side effect of the proliferation of methodologies and approaches, along with the erosion of tacitly shared philosophical commitments.
At the heart of recent metaphilosophical controversies within mainstream analytical philosophy has been the method of cases and conceptual analysis, where this refers to the dual strategy of, on the one hand, conjuring up particular cases (in the spirit of philosophical 'thought experiments') and, on the other hand, attempting to dissolve any disagreement through careful - and competent - attention to the contituent concepts at issue. This overall approach spans the various subdisciplines, ranging from metaphysics and the philosophy of mind all the way to epistemological projects such as demarcating the applicability of the term 'knowledge'. In fact, one of the most spirited defences of the philosophical utility of conceptual analysis, by Frank Jackson, is entitled From Metaphysics to Ethics: A Defence of Conceptual Analysis [1998], thereby signalling the appeal, and near-universal ambition, of conceptual analysis as a philosophical methodology. At the same time, criticisms and challenges abound, much of which has traditionally centred on the tacitly assumed analytic/synthetic distinction (usually considered to have been undermined by Quine [1951]) and on the undue emphasis on language and concepts [Williamson, 2004].
The power of the method of cases resides in its ability to prompt powerful intuitions, for example concerning the extension of certain concepts (e.g.,'knowledge'), which help adjudicate philosophical questions and demarcate conceptual boundaries. As Jennifer Nagel notes, "[a] well-constructed case can elicit a powerful intuitive verdict," even if "the power of intuitive responses is somewhat mysterious" [Nagel, 2012, p. 495]. Part of the mystery is due to the lack of immediate transparency and the lack of a phenomenology of deliberation. Typically, in philosophical argumentation, cases are so constructed as to either illustrate an already accepted point or sway the hearer/reader, by bringing about an alignment between her intuitions and those of the speaker/author. As Nagel rightly notes, this failure of transparency may lead "us, as self-conscious epistemologists, to wonder about the epistemic legitimacy of the method of cases" [ibid.]. Extending this line of thought further, one could harbour the suspicion that, perhaps, the method of cases amounts to nothing but a careful manipulation of an audience's seemingly intuitive responses. Even when intuitive responses appear to be unanimous, this may well turn out to be an artefact of irrelevant influences that have unduly skewed a particular audience's response - or so the objection goes.
Most recently, proponents of so-called experimental philosophy have issued just such a caveat with respect to the probative force of case-based intuitions. Joshua Alexander and Jonathan Weinberg, in a survey article on the relationship between experimental philosophy and mainstream analytic epistemology, characterize as one type of concern the "restric-tionist" worry that "experimental evidence seems to point to the unsuit-ability of intuitions to serve as evidence at all" [Alexander & Weisberg, 2007, p. 63]; on this view, intuitions in general are too unreliable and temporally unstable to serve as any sort of durable foundations of philosophical practice. This contrasts with the "proper foundation" view, which holds that we should empirically ascertain which intuitions (regarding the types of examples and thought experiments philosophers tend to devise) are stable and durable enough to serve as a basis of theorizing. Crudely speaking, restrictionist views are pessimistic, proper foundation views optimistic about putting intuitions on a firm empirical footing, yet both entail a revision of the standard practice of assuming, without further analysis, that philosophical intuitions have probative force. A further issue concerns the differences between what one might call 'raw' vs. 'educated' intuitions. As Nagel notes, it should be "possible to have theory-driven epistemic intuitions, for example, after becoming very well-rehearsed in applying the verdicts of some particular analysis" of a given philosophical concept (e.g. 'knowledge'). On the one hand, this relates to the worry that even seemingly uniform intuitions may still be subject to cultural modulation; on the other hand, it opens up the possibility of insisting on there being expert intuitions, i.e. intuitive - that is, instant
non-deliberative - responses by properly educated experts well-versed in the correct underlying theories.1
Timothy Williamson has forcefully pushed back against the experimentalists' challenge to the 'traditional' way of doing philosophy. First, philosophers need not accept the tacit premise that evaluations of particular cases amount only to the pitting of one set of intuitive seemings against another. In his The Philosophy of Philosophy [2007], he offers an alternative account of philosophical thought experiments "as employing deductively valid arguments with counterfactual premises that we evaluate as we evaluate other counterfactuals, using a mixture of imaginative simulation, background information, and logic" [Williamson, 2011, p. 215-216]. Couched in these terms, philosophical thought experiments no longer appear so radically different from, say, the experimental method in the sciences. When constrained in the right way, by logic and background knowledge, our intuitions may well track objective counter-factual relationships in the world. Furthermore, philosophers may be especially well placed to engage in such reasoning, given that even experimentalists concede that "philosophical training does typically bring a mastery of relevant literatures both contemporary and historical, and even specific technical skills such as argument evaluation and construction" [Weinberg, Gonnerman, Buckner & Alexander, 2010, p. 334]. Hence, Williamson argues, we should "not regard philosophical training as an illegitimate contamination of the data, any more than training natural scientists how to perform experiments properly is a contamination of their data" [Williamson, 2007, p. 191]. To be sure, philosophical inquiry needs to be conducted responsibly and in accordance with disciplinary standards, yet when these are in place, it is entirely defensible on the basis of the well-earned expertise of its practitioners. No expertise is without bounds, of course, which is why Williamson demands that "[p]hiloso-phy students have to learn how to apply general concepts to specific examples with careful attention to the relevant subtleties, just a law students have to learn how to analyze hypothetical cases" [Williamson, 2007, p. 191]. Elsewhere, he insists that the gulf between the thought experiments of philosophy and the real world of lived experience may not be as deep as the experimentalist critics have tended to assume: "Critics of 'armchair philosophy' tend to forget that there are real life analogues of some philosophical thought experiments" [Williamson, 2011, p. 217].
If the continuity between the imaginary scenarios of philosophical thought experiments and the real world of lived experience is a desideratum for the method of cases in general, it is especially so for the methodology of social epistemology. This can easily be shown via a brief consideration of the dominant approaches to social epistemology, each of which reflects a distinct methodological orientation. Alvin Goldman [2010]
Much, obviously, hinges on what counts as 'correct' and 'proper' in this context.
helpfully gives a taxonomy of such approaches and distinguishes between three global ways of doing social epistemology, which he calls revisionism, preservationism, and expansionism, respectively. The guiding question, which he borrows from William Alston [2005], is that of whether social epistemology is "real epistemology" - where the latter is, quite obviously, itself a contested notion. Those approaches that take an overtly descriptive approach, such as the sociology and psychology of knowledge, may indeed fall outside the scope of "real" epistemology since they do not normally concern themselves with core epistemic notions such as justification and warrant, or so Goldman argues. To the extent that they are sometimes included under the heading of "social epistemology", they reflect revisionism - they quite literally constitute a change of topic. By contrast, preservationism and expansionism actively seek continuity with traditional epistemology, differing mainly on the extent to which they are willing to include new concerns and considerations alongside the familiar notions of (individualist) epistemology. To a first approximation, preservationism acknowledges the existence of "social evidence" (Goldman's term) for individual reasoners, whereas expansionism is open both to the consideration of collective epistemic agents (e.g., groups as know-ers) and to normative questions concerning how epistemic systems can be improved. The latter concern gives rise to an ameliorative conception of social epistemology, which goes well beyond the normative dimension of traditional (individualist) epistemology.
Yet all the major approaches to social epistemology agree on the need for establishing contact with the real world of lived, socially situated experience. Even 'revisionism' does not intend to sever this connection; if anything, it is willing to sacrifice some of the core tenets of traditional episte-mology in order to maintain this continuity. The ameliorative conception, likewise, does not - and cannot legitimately - abstract away from actual existing epistemic practices, even as it seeks to improve them. Social-epis-temological theorizing, thus, inevitably takes place under constraints, not least the requirement to maintain empirical plausibility in the light of facts about human cooperation and sociality.
3. Cases And Intuitions
In The Epistemology Of Testimony
One of the central debates within social epistemology concerns the status of testimony-based beliefs. 'Testimony' here is to be understood as an umbrella term, covering face-to-face conversations, formal declarations (e.g., in court), educational instructions, written notes, books, media, and the like. As such, it is the chief source of knowledge by which we
learn about the empirical world outside of our narrow realm of immediate experience, about the society we live in, and our place within both. Functionally, it is the main means by which people exchange information, it is crucial to the division of epistemic labor, and it is instrumental in securing the continuity of cultural tradition. No wonder, then, that among all the "social operations of the mind" - to use Thomas Reid's phrase - testimony has received perhaps the most sustained attention from (social and mainstream) epistemology. The epistemology of testimony, therefore, provides fertile ground for an exploration of the method of cases in the context of social epistemology - including of its limitations. (For a survey of the field as a whole, see [Gelfert, 2014].)
It has become standard to introduce the epistemology of testimony via one of its core disputes, between reductionists and anti-reductionists, who have clashed on the issue of what grounds testimonial justification. Anti-reductionists treat testimony as on a par with perception and mem -ory, that is, as a sui generis source of epistemic justification; on this account, a hearer can acquire justification for a belief simply in virtue of accepting the requisite testimony and making the corresponding belief their own. Reductionists reject any such suggestion that testimony is a fundamental source of epistemic justification. Whatever justification a hearer might have for a belief they acquired from someone else's testimony, must ultimately derive from other, more basic epistemic sources: firsthand observation, memory, inference. We cannot expect to be able to verify first-hand every statement of fact we receive, but even when testimony gives us access to no longer directly verifiable states of affairs, our epistemic justification must be based on other, non-testimonial sources -such as inductive evidence of a witness's track record, in conjunction with independent background knowledge. This dispute over the source of justification for our testimony-based beliefs maps on to - but is, strictly speaking, separate from - the doxastic question of how we should react when we encounter a new testimonial claim: should we accept what we are told, or should we reject the testimony (or perhaps suspend judgment for the time being)? Arguably, anti-reductionists should be expected to be more open towards a default stance of acceptance, whereas reductionists will likely hold off on acceptance until such a point as independent (non-testimonial) evidence has been obtained.
It is precisely this need to square general philosophical commitments - to reductionism and anti-reductionism, respectively - with the situational demand of, here and now, having to decide whether to (tentatively) accept or reject a given piece of testimonial information, that renders some particular (imaginary) cases hotly contested. What is also interesting, and perhaps different from other philosophical disputes, is the fact that virtually all contributors to the debate - reductionists and anti-reductionists alike - agree on certain basic constraints. For example, no serious contributor to the debate denies that (some) testimony-based
beliefs should count as knowledge. Even Elizabeth Fricker, one of the most vocal reductionists, calls this the "Commonsense Constraint": "that testimony is, at least on occasion, a source of knowledge" [Fricker, 1995, p. 394]. As we shall see, this means that both reductionists and anti-reductionists use particular cases - sometimes, the same cases -in order to bolster their position; as a result, even where reductionists and anti-reductionists agree on the outcome (e.g., that a given testimony should be deemed acceptable and the corresponding belief justified), they often go out of their way to emphasize that only their own preferred theory of testimonial justification is able to account for this outcome. In order to illustrate this point, I shall discuss one such case in some detail, demonstrating how each camp - represented by Fricker (reductionism) and Tony Coady (anti-reductionism) - highlights (purported) aspects of the case that suit their theoretical position.
In his 1992 book, Coady argues for a form of defaultist 'fundamentalism'2 about testimony. As the starting-point of his argument, Coady adopts a broadly Davidsonian perspective. Where Davidson [1984] tries to refute the possibility of global error at the level of beliefs, by introducing a hypothetical 'omniscient interpreter' into his framework of radical interpretation, Coady extends - problematically, it should be added! - this framework to testimonial utterances, thereby attempting to show that testimony, in general, cannot be radically mistaken. This, Coady supposes, results in a (defeasible) prima facie justification for claims received via testimony. Furthermore, we have a presumptive right to accept testimony unless there are specific reasons not to do so. In the absence of defeaters, we can acquire knowledge directly, simply by accepting what we are told, or so Coady argues. The anti-reductionist story about epistemic justification, thus, translates into a doxastic recommendation in favour of a stance of default acceptance. By contrast, Fricker argues that any presumptive right thesis is radically misguided: "Does not mere logic, plus our common-sense knowledge of what kind of act an assertion is, and what other people are like, entail that we should not just believe whatever we are told, without critically assessing the speaker for trustworthiness?" [Fricker, 1995, p. 400] This ties in with the reductionist demand that testimonial justification must eventually be reducible to non-testimonial evidence, acquired first-hand.
Whereas Coady argues that testimonial knowledge typically is direct, Fricker insists that it must be inferential: "the hearer must always be scrutinising the speaker for telltale signs [of insincerity and incompetence], and she must be alert to the presence of such signs." [Fricker, 1995, p. 405] When contrasted in this way, the two positions could hardly seem more antagonistic. The purity of the contrast, however, comes at the price of plausibility: In its unmodified form, Coady's position would seem to
The term 'fundamentalism' in this context is due to [Kusch, 2002, p. 37].
2
suggest that mere say-so could turn even the most implausible claim into knowledge, whereas, on Fricker's account, even the most innocent claim would have to be subjected to close scrutiny. Yet, as recipients of testimony, we are neither gullible fools nor epistemic detectives; rather, while we often trust what people tell us without close examination, we also frequently reject testimony for all sorts of reasons. In order to maintain empirical plausibility - in the sense discussed towards the end of the previ -ous section - Coady and Fricker must both modify their accounts: Coady must make room for instances of rejection, whereas Fricker must account for the non-inquisitive nature of everyday acceptance of testimony. Coady does so by adding a ceteris paribus clause to his presumptive acceptance thesis, as in the following case (which subsequently came to play a significant role in the unfolding controversy between anti-reductionists and their critics):
I ring up the telephone company on being unable to locate my bill and am told by an anonymous voice that it comes to $165 and is due on 15 June. No thought of determining the veracity and reliability of the witness occurs to me nor, given that the total is within tolerable limits, does the balancing of probabilities figure in my acceptance. [...] There is nothing hesitant or suspicious about the unknown communicant's responses and I entirely believe what he says without adverting to the premisses about reliability etc. [Coady, 1992, p. 143f.; italics added]
As Fricker sees it, the italicized passages in the quotation from Coady point to "precisely the active sub-personal monitoring of the speaker by the hearer for signs of lack of sincerity or competence" [Fricker, 1995, p. 405] that she herself regards as essential to any rational attitude towards testimony.
While it is certainly compelling to argue that, unless caveats of some sort are added to Coady's account, his position is in danger of collapsing into a stance of credulity, he is hardly alone in being forced to modify his account so as to restore plausibility: Fricker, too, must supplement her position of inferentialism with an account that makes sense of the apparent lack of any inferentialist phenomenology in everyday instances of accepting testimony. Fricker's solution is to allow for "automatic and unconscious" monitoring [Fricker, 1995, p. 404], which takes place at the "sub-personal" level: "It is quite inessential that assessment be conscious; it may occur automatically, without the subject's attention being directed to it." [Fricker, 1995, p. 405] This move, however, is problematic for two reasons. First, Fricker motivates her project at least in part as a defence of the critical powers of the epistemically autonomous subject; yet, as I have argued elsewhere, "sub-personal monitoring, strictly speaking, does not amount to any critical assessment at all, since critical judgment requires that the mechanisms and standards by which we judge be open to scrutiny - which, a fortiori, is not the case if they operate
'at an irretrievably sub-personal level'." [Gelfert, 2009, p. 177] Second, Fricker's critique of testimonial fundamentalism is based on the premise that there is a significant difference between perception as a 'direct' epis-temic source and testimony, in that the former, but not the latter, can be relied upon as a default source of knowledge. Yet, once "sub-personal" processes are sufficient to render an account inferentialist, the contrast dissolves, since perception, surely, also depends on all sorts of subconscious inferential processes. Rather than establish an asymmetry between 'direct' pathways to knowledge and 'indirect' testimony, an account in terms of subconscious inferences would seem grist on the mill of those who, like Coady, argue that testimony, in every epistemically relevant respect, is on a par with other epistemic sources.
This is not the place to adjudicate between reductionism and anti-re -ductionism as theories about the nature of testimonial justification, nor between their attendant doxastic recommendations of default acceptance and inferential monitoring, respectively. What is significant for our purposes is the fact that both camps, though overtly opposed to one another on nearly every issue of philosophical significance, are able to recruit the same case - the hypothetical scenario of someone ringing up the phone company to ask about an outstanding bill and accepting what an anonymous voice tells him - to their cause, and confidently assert that it bolsters their preferred position. Where Coady holds the example to show that in our "ordinary dealings with others we gather information" without any "concern for inferring the acceptability of communications from premisses about the honesty, reliability, probability, etc., of our communicants" [Coady, 1992, p. 143], Fricker explicitly claims that the case, as described by Coady, suggests "precisely the active sub-personal monitoring of the speaker by the hearer for signs of lack of sincerity or competence described above." [Fricker, 1995, p. 405] To bring out just how extraordinary the coexistence of these conflicting interpretations is, consider that Fricker demands no modification or amendment of the case as described by Coady - even though Coady had deliberately constructed the scenario with the intention to illustrate, in paradigmatic fashion, the character of non-inferential, default acceptance, in line with his anti-reductionist stance. The method of cases, far from settling the matter by eliciting strong unanimous intuitions, appears to be all but impotent in the present case. Coady's phone bill example, thus, does not serve as a clear-cut thought experiment - in the sense of a philosophical experimentum crucis - but rather as an ambiguous fictional vignette, onto which conflicting theoretical commitments can be projected with astonishing effortlessness.
4. Challenging The 'Null Setting'
What, if anything, has 'gone wrong' in the dispute sketched in the previous section, and why is the method of cases unable, if not to settle the matter, then at least to elicit stable intuitions of the intended kind? Recall that Coady's phone bill example was not meant to reflect a 'messy' real-life situation, where competing factors are always to be expected, but was devised precisely in order to illustrate a paradigmatic case of non-inferential testimonial acceptance. Fricker, in spite of her confidently co-opting Coady's specific case and claiming that it is "precisely" what supports her story, not Coady's, in her general remarks on the issue acknowledges the (special?) volatility of our intuitions concerning testimonial acceptance:
I find my own intuitions about testimony wildly volatile: consider some cases, and it seems obvious that we must have a default position of trust in what others tell us [...]; but consider others, and it seems equally obvious that our attitude to others must be critical and skeptical [Fricker, 1995, p. 406].
While Coady's and Fricker's conflicting interpretations of the phone bill example may be seen as an illustration of this volatility, it is the fact that this volatility persists even for cases that are specifically designed to support one side of the conflict, not the other, which demands an explanation.
On the face of it, there is nothing unusual about Coady's example; it is just the kind of stylized description - an easily imaginable scenario that abstracts somewhat from the messy complexities of the real world, but not so much as to strike one as contrived - that one would expect in this context. After all, it is intended to be illustrative, aimed at lending plausibility to a position, not at providing, say, a far-fetched counterexample. And yet, upon closer inspection, it turns out to be far from trivial. This is because, given the stylized nature of the description, we - as readers of the text - inevitably 'fill in' the details, thereby importing social background knowledge and projecting our own peconceptions and prejudices onto the (imaginary) encounter with the anonymous voice on the other end of the line. Indeed, as Tim Kenyon has noted, Coady's example as described could not serve its intended argumentative purpose, if it weren't for the importation of our deeply ingrained social expectations. When Coady takes our acceptance of the stated answer ("the bill comes to $165") as evidence that, in general, we accept testimony without any thought of "determining the veracity and reliability of the witness" [Coady, 1992, p. 143], he ignores that a significant portion of the facts pertaining to the veracity and reliability of the testimony "have indeed already been substantially determined in the example as written" [Kenyon,
2013, p. 76]. After all, as the case has been constructed, Coady has made the conscious decision to ring up his telephone company - not just anyone - to ask about the outstanding amount.
[A] vast amount of justificatory information is packed into the phrase, 'I ring up the telephone company.' [...] That the person who answers does not provide a name is practically irrelevant. Given the aim of the phone call, the name of the telephone company worker pales in evidential significance compared to the fact that they work at the telephone company. [Kenyon, 2013, p. 76]
If one were to focus merely on what is explicitly stated about the immediate testimonial exchange - an anonymous voice telling Coady to pay a seemingly random amount of money - Fricker's conclusion that we should not take such testimony on blind faith, seems quite plausible. Yet, once we factor in the (entirely implicit) causal history of how the encounter came about - with Coady looking up the phone company's tollfree customer service number, perhaps choosing from a range of options ("Press '1' for inquiries about your bill, '2' for..."), giving the customer service representative his account details, etc. - trusted acceptance of the resulting testimony seems entirely warranted. All the evidence that was utilized by Coady to seek out this particular source of information should legitimately also count as evidence in support of what he is told. Once one recognizes the evidential significance of testimonial contexts and histories, Coady's and Fricker's dispute over what may, or may not, be read into Coady's claim that no thought of questioning his interlocutor occurred to him because there was "nothing hesitant or suspicious about the unknown communicant's responses" - whether it indicates a default stance of trust or whether, on the contrary, it is the outcome of inferential monitoring - almost becomes a side issue.
At the heart of the matter is a more general question concerning which level of detail we should aim for in the description of exemplary cases in the epistemology of testimony. Jonathan Adler has stated that the setting that is "the proper one for investigating the epistemics of testimony" is the so-called "null setting" [Adler, 2006]. As the name suggests, the null setting abstracts from essentially all specifics that, in any actual situation, might be expected to make a difference. Specifically, Adler [ibid.] states five conditions: First, testimony in the null setting must be "limited to brief assertions to avoid internal support due to coherence"; second, "corroboration or convergence of a number of testifiers [...] should be set aside"; third, testimony must be the "sustaining source" of the belief in question; fourth, "we set aside cases of a hearer's attribution of expertise to a speaker on certain topics, as well as a speaker's acting under professional or institutional demands for accurate testimony"; and-fifth, the hearer "has no special knowledge about the speaker" (i.e., must be a stranger in some sense). While it seems plausible that limiting
oneself to cases that conform to the "null setting" may help 'isolate' the specifically testimonial contributions to our epistemic situation, it is equally clear that most real-life scenarios are radically unlike it. Even Coady's phone bill example would, strictly speaking, not qualify since, as discussed, the hearer (that is, Coady) implicitly relies on the institutional contraints imposed upon the (anonymous) speaker, thereby violating condition four. Virtually the only real-world scenario that comes anywhere close to the null setting is the - perhaps for this reason, ubiquitously cited - case of asking local directions from a stranger.
Paula Olmos draws an interesting parallel between this demand for "null setting" scenarios and the equally influential distinction between 'formal' and 'informal' testimony in the epistemology of testimony, where 'formal testimony' typically refers to highly constrained contexts -as in the case of eyewitness testimony before a court of law - while 'informal' testimony is the catch-all term for the more "relaxed exchange" [Olmos, 2008, p. 59] of information, e.g. between casual interlocutors, where formal conditions do not apply (or are not enforced). No wonder, then, that social epistemologists have tended to regard formal testimony as too restrictive, given that only a small portion of our testimonial encounters - though, no doubt, an important one - takes place in formal contexts. Instead, the focus has been on informal testimony, which is generally taken to reflect the "so-called natural, allegedly basic, practices (and settings)" [ibid.] associated with learning from others, and for which the only requirement is "that it be a statement of someone's thoughts or beliefs, which they might direct to the world at large and to no one in particular" [Sosa, 1991, p. 219]. In part, this follows from analytic episte-mology's treatment of testimony as a catch-all for all non-individualist sources of belief and justification, as a result of which opting for the more comprehensive definition - by taking the term 'testimony' to refer primarily to informal testimony (with its greater extension than formal testimony) - seems the way to go. Doing so, however, comes at a price. Precisely because informal testimony is so diverse, it is virtually impossible to come up with meaningful generalizations that would allow one to estimate (in even the most qualitative fashion) the reliability or trustworthiness of the claims and testifiers one is likely to encounter. This leaves only the most bare-bones description of a testimonial encounter - the null setting - as the target for general theorizing in the epistemology of (informal) testimony. Yet, as we have seen, any case that approximates the null setting - if perhaps only imperfectly, like Coady's phone bill example - is hostage to the importation of (tacit) background expectations, which undermine its probative force.
It appears, then, that the usual method of constructing stylized cases that can serve as 'intuition pumps' faces special challenges when applied to the domain of social epistemology, and to the epistemology of testimony in particular. Precisely because testimony, as a general category,
includes an extremely heterogeneous range of sources, cuts across all sorts of content, can be used both to inform and to obfuscate, and is heavily context-dependent, any stylized cases that omit relevant details are at risk of being 'up for grabs', waiting to be hijacked for specific theoretical agendas. The closer a case description is to the null setting, the more it underdetermines the kinds of intuitions it elicits; as a result, tacit background assumptions and prejudices will guide our interpretation and assessment. One might argue that this is simply to be expected: In the absence of sufficient detail, any 'underdescribed' case will inevitably engage our pre-theoretical intuitions, which in turn will influence how we 'fill in' the missing bits in our subsequent theorizing and analysis. Isn't this precisely how philosophical thought experiments are supposed to function? Yet, in the case of examples drawn from the social sphere, the ease with which we - entirely unreflectively - import peculiarities of our own outlook, experience, upbringing, and even temperament into the assessment and analysis of such cases, means that we cannot legitimately expect them to dissolve disagreement, but merely to illustrate it. And, indeed, this is precisely what we found in the earlier controversy between Coady and Fricker.
Two remedial strategies suggest themselves: First, we could invert the usual strategy and turn to formal contexts, rather than informal settings, as the paradigmatic scenarios for any prospective epistemology of testimony. Second, we could give up on the idea that considering highly stylized "null setting"-type examples offers any sort of specific insight, focusing our attention instead on richly described cases that either include, or are embedded into, sufficient contextual information to allow for educated judgments concerning the reliability and trustworthiness of the testimony and testifiers involved. Both strategies have been pursued in recent years, and I shall close by sketching examples of each. Olmos herself adopts the first strategy, by focusing "on paradigmatically procedural instances" such as "courtroom witnessing, religious rituals, speeches in the public assembly, public or scientific controversies etc., in which the rich variety of social conditions, normative constraints [... ] and concrete configurations of authority, explain many features that tend to become'transparent'" - and hence invisible - "in apparently less rule-governed instances" [Olmos, 2008, p. 59]. In doing so, she adopts Fred Kauffeld's and John Fields's view that formal testimony "makes explicit much that is assumed or goes unremarked upon in ordinary conversational settings" and therefore "can give us guidance as to where we should be looking" [Kauffeld & Fields, 2003, p. 3] in cases of informal testimony, too. Recognizing the empirical richness of the multiplicity of co-existing (and sometimes intersecting) formal contexts is key to the plausibility of this project. For, it is simply not the case that the courtroom is the only formal setting that is highly constrained, with explicit rules on how testimony is to be conducted. Negotiating a contract,
receiving financial advice from one's bank, conducting a parent-teacher consultation, debating an issue in the local council - these are not, or at least not in most cases, casual encounters conducted in the absence of contextual information; on the contrary, they are saturated with background knowledge, procedural rules, and social conventions. Being able to 'place' a testimonial encounter within this spectrum of situated social practices, thus, goes a considerable way towards alleviating any justificatory worries one might have about when, and whom, to trust. This, in turn, reduces the need to impute theoretical significance to the "null setting" in the first place.
Not all testimony, however, will fit neatly into one of the well-established social practices just discussed. How are we to treat such instances of testimony? This is where the second strategy becomes relevant. This strategy holds that, for a well-considered judgment to be possible, each case needs to be imbued with sufficient 'intrinsic' meta-information -concerning the reliability and motivations of the testifier, the causal ancestry of the testimonial encounter etc. - so as to avoid the kind of 'un-derdetermination' that would then have to be filled, in a haphazard way, by each interpreter's particular outlook. On this view, relying on unconstrained intuitions about underdescribed cases would be bound to result in an account of testimony that could not, as a matter of principle, be "revealing with respect to real, complicated testimony". It is only "through examining real and complicated cases of testimony," Lisa Bergin argues, that we can recognize "the fundamentally social nature of knowledge that is not necessarily apparent when examining simplified cases of testimony" [Bergin, 2002, 210-211]. Part of this recognition involves cultivating an awareness of the malfunctions and resistances that often stand in the way of successful actual communication. A similar point has been made in the recent literature on the topic of epistemic injustice, a central tenet of which is the need to think of interlocutors not as isolated individuals "in abstraction from relations of social power (as they are in traditional epistemology, including most social epistemol-ogy) but as operating as social types who stand in relations of power to one another" [Fricker, 2007, p. 3]. One way to achieve the requisite level of empirical richness and social situatedness in one's philosophical examples is to draw on literature and narrative fiction. Often, though of course not always, novels and films are, as Miranda Fricker puts it, "most centrally about how we live our life, how we make sense of our own life, in a narrative way, through highlighting and backgrounding different events" [quoted in Kodsi, 2018]. It is through this integration of (no doubt selective!) background descriptions into a narrative form that literary examples come equipped with contextual information that is absent from the traditional 'bare-bones' descriptions of testimony that approximate the null setting.
5. Conclusion
In the previous section, I have discussed two strategies that aim to resolve the indeterminacy of intuitions in the face of "null setting" examples by proposing ways of enriching the descriptions of testimonial cases, yet they do so in different ways. Whereas the first strategy relies on our finegrained background knowledge of the social world and its various distinct situated practices, the second promises more informationally rich case descriptions by tapping into a stock of well-articulated narratives which, often enough, have already been checked for coherence and psychological plausibility by their non-philosophical audiences. By anchoring our intuitions in social experience and narrative coherence, they open up new ways of thinking systematically, and productively, about how we give and receive information in our testimonial encounters with others. The "null setting" that has been so dominant in the epistemology of testimony may still have its place - we all, on occasion, need to rely on a random stranger for local directions! - yet its stranglehold on the philosophical imagination has diminished in recent years, and rightly so.
Список литературы / References
Adler, 2006 - Adler, J. "Epistemological Problems of Testimony", in Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy (Winter 2006 edition). Available at: https://plato.stanford.edu/archives/win2006/entries/testimony-episprob/ [accessed on 24.03.2019]
Alexander, Weinberg, 2007 - Alexander, J., Weinberg, J. M. "Analytic Epistemo-logy and Experimental Philosophy", Philosophy Compass, 2007, vol. 2, no. 1, pp. 56-80.
Alston, 2005 - Alston, W. Beyond "Justification". Ithaca: Cornell University Press, 2005, xiii+256 pp.
Bergin, 2002 - Bergin, L. "Testimony, Epistemic Difference, and Privilege: How Feminist Epistemology Can Improve Our Understanding of the Communication of Knowledge", Social Epistemology, 2002, vol. 16, no. 3, pp. 197-213.
Coady, 1992 - Coady, C.A.J. Testimony: A Philosophical Study. Oxford: Oxford University Press, 1992, 336 pp.
Davidson, 1984 - Davidson, D. Inquiries into Truth and Interpretation. Oxford: Oxford University Press, 1984, xx+292 pp.
Fricker, 1995 - Fricker, E. "Telling and Trusting: Reductionism and Anti-Reductionism in the Epistemology of Testimony (Critical Notice of C.A.J. Coady, Testimony: A Philosophical Study)", Mind (New Series), 1995, vol. 104, no. 414, pp. 393-411.
Fricker, 2007 - Fricker, M. Epistemic Injustice: Power and the Ethics of Knowing. Oxford: Oxford University Press, 2007, 198 pp.
Gelfert, 2009 - Gelfert, A. "Indefensible Middle Ground for Local Reductionism about Testimony", Ratio (New Series), vol. 22, no. 2, pp. 170-190.
Gelfert, 2014 - Gelfert, A. A Critical Introduction to Testimony. London: Bloomsbury, 2014, vi+257 pp.
Goldman, 2010 - Goldman, A. "Why Social Epistemology is Real Epistemo-logy", in: Haddock, A., Millar, A., Pritchard, D. (eds.), Social Epistemology. Oxford: Oxford University Press, 2010, pp. 1-29.
Kauffeld, Fields, 2003 - Kauffeld, F.J. and Fields, J.E. "The Presumption of Veracity in Testimony and Gossip", Proceedings of the 2003 OSSA Conference. Windsor: University of Windsor, 2003. Available at: http://web2.uwindsor.ca/faculty/ arts/philosophy/ILat25/edited_new_kauffeld&fields.doc [accessed on: 21. 03. 2019]
Kenyon, 2013 - Kenyon, T. "The Informational Richness of Testimonial Contexts", The Philosophical Quarterly, 2013, vol. 63, no. 250, pp. 58-80.
Kodsi, 2018 - Kodsi, D. "Flaws and Forgiveness: An Interview with Miranda Fricker", Oxford Review of Books, 15 December 2018. Available at: http://the-orb.org/ 2018/12/15/englands-arch-miser/ [accessed on 24.03.2019].
Kusch, 2002 - Kusch, M. Knowledge by Agreement: The Programme of Communitarian Epistemology. Oxford: Oxford University Press, 2002, 250 pp.
Nagel, 2012 - Nagel, J. "Intuitions and Experiments: A Defense of the Case Method in Epistemology", Philosophy and Phenomenological Research, 2012, vol. 85, no. 3, pp. 495-527.
Olmos, 2008 - Olmos, P. "Situated Practices of Testimony: A Rhetorical Approach", Theoria, 2008, vol. 23, no. 1, pp. 57-68.
Quine, 1951 - Quine, W.V.O. "Two Dogmas of Empiricism", The Philosophical Review, 1951, vol. 60, no. 1, pp. 20-43.
Sosa, 1991 - Sosa, E. "Testimony and Coherence", in id., Knowledge in Perspective: Selected Essays in Epistemology. Cambridge: Cambridge University Press, 1991, pp. 215-222.
Weinberg, Gonnerman, Buckner, Alexander, 2010 - Weinberg, J.M., Gonnerman, C., Buckner, C., Alexander, J. "Are Philosophers Expert Intuiters?", Philosophical Psychology, 2010, vol. 23, no. 3, pp. 331-355.
Williamson, 2004 - Williamson, T. "Past the Linguistic Turn?", in Brian Leiter (ed.), The Future for Philosophy. Oxford: Oxford University Press, 2004, pp. 106-128.
Williamson, 2007 - Williamson, T. The Philosophy of Philosophy. Oxford: Blackwell, 2007, 320 pp.
Williamson, 2011 - Williamson, T. "Philosophical Expertise and the Burden of Proof", Metaphilosophy, 2011, vol. 42, no. 3, pp. 215-229.
