Hostname: page-component-848d4c4894-p2v8j Total loading time: 0 Render date: 2024-05-13T10:00:22.412Z Has data issue: false hasContentIssue false
  • English
  • Français

In Defense of Evidential Minimalism: Varieties of Criticizability

Published online by Cambridge University Press:  09 June 2023

Frank Hofmann Open the ORCID record for Frank Hofmann [Opens in a new window]
Frank Hofmann*
Affiliation:
University of Luxembourg, Esch-sur-Alzette, Luxembourg
*
Email: frank.hofmann@uni.lu
Rights & Permissions [Opens in a new window]

Abstract

This paper will critically engage with Daniel Buckley's argument against “evidential minimalism” (EM), i.e., the claim that necessarily, bits of evidence (are or) provide epistemic reasons for belief. Buckley argues that in some cases, a subject has strong evidence that p (and fulfills further minimal conditions), does not believe p, but nevertheless is not epistemically criticizable and has no epistemic reason to believe p. I will defend EM by pointing out that Buckley's argument trades on an ambiguity between a strong and a weak notion of criticizability.

Keywords

Evidenceevidentialismreasonscriticism
Type
Article
Information
Episteme , First View , pp. 1 - 6
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press

1. Introduction

In this paper I will critically engage with Daniel Buckley's argument against what he calls “evidential minimalism” (EM, for short). I will defend EM, which will not come as a surprise since Buckley – correctly – identifies me as a proponent of EM.Footnote 1 Roughly speaking, EM is the following view:

(EM) Necessarily, bits of evidence (are or) provide epistemic reasons, i.e., normative reasons for belief.

Here, the “providing” is meant to allow for certain necessary conditions in addition to the obtaining of evidence, namely, the possession of the evidence, the ability to use the evidence (and thus some grasp of the evidence's support for p), and the activation of the question whether p in the subject's mind. These further conditions are “minimal conditions” since they are very natural, concern only evidence (apart from the activation condition), and do not bring in any totally different element. In contrast, a teleological condition, such as having a desire or intention to find out whether p, would amount to a non-trivial further condition that is not related to evidence at all and would therefore not count as minimal. Proponents of EM think that we have genuine epistemic normativity without substantive, non-minimal conditions.Footnote 2

Daniel Buckley has presented a rich discussion that brings up highly interesting and relevant topics and ideas. He considers excuses, and he reflects on the distinction between requiring reasons and warranting reasons. Yet on center stage he puts criticizability. This original move leads our attention to important and, arguably, underexplored phenomena. And one cannot but conclude that Buckley's argument pushes the debate one step forward.

Due to lack of space, I will not take issue with all the interesting materials that show up along the way. I will just focus on the connection between criticizability, evidence, and epistemic reasons as involved in Buckley's main argument. This paper will present a dilemma for Buckley's strategy to refute EM in the way he intends to. Neither horn of the dilemma allows for a clear argument against EM.

2. Buckley's argument

Buckley's argument comprises two steps. The first step consists of providing a case in which someone is not criticizable, though she does have strong evidence for the proposition in question and fulfills all the minimal conditions allowed by EM. Here is a case that Buckley thinks is showing that having strong evidence does not guarantee being criticizable.

BORED AT HOME: Katlyn is at home with nothing to do. She recently left her job

in order to take up a better position at a new company. There's a one month interval

in between her leaving the old job and beginning training for the new job.

Katlyn doesn't have to move for her new job, and she's tied up all manner of

loose ends, taken care of various errands and housekeeping, etc. She finds herself

at home one Sunday afternoon and is simply bored with nothing to do. (Buckley Reference Buckley2022: 8)

Now, Buckley goes on and points out that Katlyn has strong evidence for many propositions, for example:

Thus, at any arbitrary time t when Katlyn is at home, she will possess, in the “robust” sense, strong evidence E for truth of the following claim: that she has been alive for more than 200,000 hours. However, Katlyn is of course in no way criticizable for not believing this claim. (ibid.)

At this point, however, one further issue needs to be addressed briefly. Whether the activation condition, which arguably is part of the minimal conditions, is fulfilled has not yet been clarified by the case description. So we need to add that the question whether p is activated in Katlyn's mind.Footnote 3 (Perhaps, someone comes up and prompts Katlyn with the question whether p. Or … the question becomes activated in some other way.) Of course, we have to be careful not to put in a desire or intention to find out whether p, since that would clearly go beyond the minimal conditions.Footnote 4 But one can have a question in one's mind without having such an epistemic desire. Suppose we add to the case description that the question whether p is activated for Katlyn (without positing anything that goes beyond the minimal conditions). Then we get the result that Katlyn has strong evidence for p, fulfills all the minimal conditions, yet is not criticizable, according to Buckley.Footnote 5 This makes (S1) come out true:

(S1) There are cases in which someone, S, has strong evidence for p, fulfills all the minimal conditions, does not believe p, and is not criticizable for not believing p. (cf. Buckley Reference Buckley2022: 15)

In a second step, Buckley infers from the lack of criticizability to the non-existence of epistemic reasons to believe p.

(S2) If someone, S, has strong evidence for p, fulfills all the minimal conditions, does not believe p, and is not criticizable for not believing p, then S does not have any epistemic reason for believing p. (Cf. Buckley Reference Buckley2022: 15)

So, roughly speaking, the inference runs from strong evidence to the lack of criticizability to the non-existence of epistemic reasons. And that would refute EM.

3. Criticizability

To begin with, there may be several, and several important, notions of criticizability that target someone's holding or not holding some doxastic attitude. Some of them might not be notions of epistemic criticizability (but rather of practical criticizability). I will put those aside. But even in the epistemic domain, there might be several notions of criticizability. Buckley has described one of them, as follows:

(C1) S is (epistemically) criticizable for (not) holding doxastic attitude A iff it is appropriate to (epistemically) blame S for (not) holding A.

Then, various accounts of epistemic blame can be added in order to derive interesting conclusions. Buckley discusses four accounts, an emotion-based, a desire-based, a relationship-based, and an agency-cultivation account.Footnote 6

Here is an alternative way in which one might think of criticizability:

(C2) S is criticizable for (not) holding doxastic attitude A iff someone can truly judge that S has made an epistemic mistake in (not) holding A.

In effect, this account puts all the money on whether the subject makes an epistemic mistake. This amounts to whether the subject respects or flouts her epistemic reasons. (At least, this is so if being in line with epistemic reasons is all you need to do in order to make no epistemic mistake.) True judgments just mirror the facts, so the semantic ascent to true judgments does not add anything substantial. Criticizability is connected to epistemic mistakes, according to C2. So, in effect, C2 boils down to

(C2′) S is criticizable for (not) holding doxastic attitude A iff S has made an epistemic mistake in (not) holding A.

Of course, one might try to spell out making epistemic mistakes in terms of blameworthiness, but one does not have to.Footnote 7, Footnote 8

4. A significance dilemma

Now, whatever one's favored notion of criticizability is, one aspect has to be clarified. Buckley's discussion strongly suggests that what plays a key role in the kind of cases at stake is whether there is a condition of “significantly mattering” to criticizability. In other words, we have to ask ourselves how to think about cases in which the subject flouts her evidence and where her having or not having an attitude about the relevant proposition p just does not matter significantly. Call this the “significance condition.” (What counts as significant may depend on the context, of course.) On reflection we can state the following dilemma then.

First horn: on a notion of criticizability that includes a significance condition – a strong notion – we get the following result, arguably:

(1) If S has strong evidence for p (and fulfills all the minimalist conditions), does not believe p, and her attitude toward p does not matter, she is not criticizable for not holding the belief that p.

On this strong notion, however, we cannot infer from the lack of criticizability to the non-existence of an epistemic reason for believing p (generated by the evidence). The reason is simply that not being criticizable might just be due to the lack of significance. On such a notion, cases like the ones Buckley is presenting are not capable of providing an argument against EM. These are all cases where the subject's doxastic attitude toward p does not matter, intuitively. (Note that this is so, no matter what account of epistemic criticism or blame is adopted. For any such account will allow for two notions of criticizability or blameworthiness, one without and one with a significance condition added. Not believing pointless propositions will always pass as not criticizable on a strong notion of criticizability. Thanks to an anonymous referee for bringing up this issue.)Footnote 9

Second horn: on a notion of criticizability that does not include a significance condition as necessary – a weak notion – we get the following result, arguably:

(2) If S has strong evidence for p (and fulfills all the minimalist conditions), does not believe p, and her attitude toward p does not matter, she is criticizable for not holding the belief that p.

On the weak notion, which is blind toward the issue of significance, the cases can intuitively be described as ones in which the subject is criticizable. This is at least my intuition. Admittedly, it is not so easy to tell. But for those who are in favor of EM, it is an open possibility to think that the subject is criticizable after all, on a weak notion of criticizability. Remember: we have put aside how much it matters whether the subject does not believe p. And what remains is a subject who simply fails to form the belief that p, even though she has strong evidence, is able to make use of it, and the question whether p is activated or open in her mind. Isn't this sufficient for having an epistemic reason for believing p? At least, what Buckley has said does not provide any argument against an affirmative answer here. So again, though it is not so clear-cut as for a rich notion of criticizability, we get no argument against EM.Footnote 10

5. Conclusion

Here is a bit of diagnosis. Crudely speaking, what has misled us is ambivalence about the role of significance for criticizability. True beliefs that we would ordinarily classify as pointless, silly, or insignificant (though true or even known) make us hesitate when it comes to raising criticism. What would be the point of engaging in criticism if the attitude does not matter? (After all, criticizing someone likely requires some effort.) So we naturally tend toward a rich, strong notion of criticizability when confronted with such cases. And that then leads us into thinking that the subject has no epistemic reason to hold this attitude, since we simply follow the path from “no criticism” to “no mistake” to “no epistemic reason flouted.” But we can abstract away from the “filter” of significance and, thus, move to a weaker notion of criticizability. (I take it that this is not an easy task. The weaker notion is somewhat technical, not so intuitive, and probably less often involved in ordinary life, if at all.) Then it is far from clear whether there really is no ground for epistemic criticism and, correspondingly, for assigning an epistemic reason. At least, no non-question begging reason for a positive answer has been given by Buckley. Friends of EM are free to stick to a negative answer. But in any case, it will be important to become clearer about the role of significance for epistemic criticism and epistemic normativity in general. Buckley deserves praise for having drawn our attention to this important phenomenon.Footnote 11

Footnotes

1 Cf. Hofmann (Reference Hofmann2021a).

2 In other words, the ground of having an epistemic reason is not the mere existence of evidence, but its existence plus these further minimal conditions concerning the evidence – I take it that the entire discussion is about the nature or ground of epistemic reasons, in the end. For the purposes of the discussion, however, I have framed EM as a modal claim. (Though Buckley does not explicitly put it that way, I believe that this is what he too intends.) In addition, I am putting aside a tricky question about what (EM) would say about blindspot propositions, i.e., propositions that are possibly true but cannot be truly believed (since believing them would make them false, like, e.g., the proposition that it is raining and no one believes that it is raining).

(Thanks to Thomas Raleigh for making me aware of this issue.) I don't think Buckley's argument is related to that question.

3 The notion of a question's being activated in someone's mind is taken from Hofmann (Reference Hofmann2021b: 281), though it is not further spelled out there.

4 This worry is addressed by Buckley (cf. Buckley Reference Buckley2022: 10).

5 We can also suppose that Katlyn does “consciously and vividly ‘see’ the evidential connection” between her evidence and p (Buckley Reference Buckley2022: 11).

6 Buckley refers to Boult (Reference Boult2021) where Cameron Boult describes these four main contenders in the field of epistemic blame.

7 Epistemic mistakes can be understood in terms of violating some constitutive norm of belief. In our case, the relevant norm would be something like the following: “You ought to (if you have strong evidence for p and satisfy the minimal conditions, you believe that p).” This is a proposal that can be found in Hofmann (Reference Hofmann2021b).

8 One might worry that making a mistake could amount to “mere appraisal,” which Buckley wants to exclude from the set of proper kinds of criticism. Buckley holds that proper criticism “must go beyond mere appraisal,” where “mere appraisal” is assessment “vis à vis a standard where this has no real weight or significance in our actual practices” (7). Now, if making an epistemic mistake is understood in terms of violating an evidentialist norm it does have “real weight and significance in our actual practices.” For it is quite common to assess people's beliefs with respect to evidentialist considerations, both in science and in ordinary life. Suitable contrast cases (such as the one of CAREFUL versus HASTY, given by Feldman (Reference Feldman2003: 39)) support the point: subjects whose beliefs are in accordance with their evidence are credited with justified beliefs, those whose beliefs do not fit their evidence are classified as holding unjustified beliefs. In science, the role of evidence is as least as prominent as in ordinary life, as pointed out by Strevens (Reference Strevens2020), for example. In contrast, assessing a subject's beliefs relative to what the tea leaves say (according to some rules of reading tea leaves) or whether these beliefs are in accordance with the subject's desires or … is not generally accepted and does not have “real weight and significance.” So we can conclude that the standard of evidence qualifies as going beyond “mere appraisal,” according to Buckley's own characterization. Many thanks to an anonymous referee for urging me to spell this out in more detail.

9 One might be tempted to interpret Buckley's argument as saying that the minimalist is forced to postulate insignificant epistemic reasons (in cases like Katlyn's) and that if one is not criticizable, one could at best have insignificant epistemic reasons (for having a belief that one actually lacks). Note, however, that there are two notions of an “insignificant epistemic reason”: a reason the flouting of which is insignificant, or a very weak reason. The evidential minimalist can live with saying that Katlyn's epistemic reasons are insignificant in the former sense, but this does not mean that these reasons would be insignificant in the latter sense. Thanks to an anonymous referee for drawing my attention to this point.

10 If we look back to (C1), the notion that connects criticizability to the appropriateness of (epistemic) blame, it is fair to say that it is a strong notion. The appropriateness seems to include an element of significance. After all, blaming someone for something that does not matter is out of place or inappropriate. Since Buckley's discussion proceeds against the backdrop of this strong notion, he would be placed on the first horn. But as I try to argue, if he moved to a weak notion, it would not help much either.

11 I am very grateful for valuable discussions and comments to Yannick Kohl, Thomas Raleigh, and Kieran Salt.

References

Boult, C. (2021). ‘Epistemic Blame.’ Philosophy Compass, 16(8), e12762. doi: 10.1111/phc3.12762.Google Scholar
Buckley, D. (2022). ‘Against Evidential Minimalism.’ Episteme, 1–20. doi: 10.1017/epi.2022.20.CrossRefGoogle Scholar
Feldman, R. (2003). Epistemology. Upper Saddle River: Prentice-Hall.Google Scholar
Hofmann, F. (2021 a). ‘Is Evidence Normative?’ Philosophia 49, 667–84.CrossRefGoogle Scholar
Hofmann, F. (2021 b). ‘Knowledge Norms of Belief and Belief Formation.’ Ratio 34(4), 277–85.CrossRefGoogle Scholar
Strevens, M. (2020). The Knowledge Machine. New York: Liveright.Google Scholar