![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
In his book Autonomous Knowledge J. Adam Carter argues that the possibility of radical cognitive enhancement shows the need for epistemology to be significantly updated. Reflection on the possibility of such enhancement shows that doxastic autonomy matters. If a belief fails to be autonomous, it cannot qualify as knowledge. Sects. 1-3 of this paper introduce the key components of Carter's autonomy framework and his considerations on the value of knowledge (including his proposed solution to the value problem, i.e. the challenge of accounting for why knowledge has greater value than any of its proper parts). Sects. 4-7 offer a critical appraisal of Carter's framework. We are not convinced that Carter has succeeded in showing that there is an autonomous belief condition on knowledge. In this paper, however, we shall not engage with this issue. Instead, for the sake of our critical discussion, we shall assume that autonomous belief is a condition on knowledge and engage Carter on a cluster of issues concerning axiology.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Musk’s post on X can be accessed here: https://twitter.com/elonmusk/status/1752098683024220632. Musk (Citation2019) is a whitepaper that describes the kind of BCI that Neuralink has been developing.
2 Neuralink’s mission statement page (https://neuralink.com/#mission) was accessed on 22 February 2024. The subsection of the page titled ‘Who It’s For’ specifies people with severe paralysis as the current target group. However, a wider group of individuals is able to register for Neuralink’s patient registry and potentially participate in clinical trials (‘Who can participate in the Patient Registry?’, https://neuralink.com/patient-registry/). Musk’s post on X concerning Telepathy can be accessed here: https://twitter.com/elonmusk/status/1752118131579867417.
3 Watch, e.g., the interview with Musk (https://youtu.be/wsixsRI-Sz4, 56:50 onwards) from the 2016 Code Conference.
4 Watch the entire session on CNET Highlights (https://www.youtube.com/watch?v=sr8hzF3j2fo&t=1990s; the segment pertaining to memory runs from 46:23 onwards).
5 Lynch (Citation2014, 299; Citation2016). See also Pedersen and Bjerring (Citation2014; Citation2021) and Pritchard (Citation2018) for a discussion of the epistemic significance of neuromedia.
6 There are, of course, those who reject the strategy of decomposing knowledge into individually necessary and jointly sufficient conditions – notably knowledge-first epistemologists such as Williamson (Citation2000). Carter (Citation2022, 23–25) offers a brief discussion of knowledge-first epistemology and autonomy. Carter suggests that his rationale for introducing autonomy as being integral to knowledge may well be transposable to the knowledge-first framework.
7 Carter mentions Alston (Citation1988) and Kvanvig (Citation2003) as examples of the good reasons and coherence approaches, respectively, while Greco (Citation2009; Citation2010) and Sosa (Citation2009) are paradigm cases of virtue reliabilists.
8 Carter employs this ‘tinkering’ strategy against the epistemic-condition views mentioned, i.e. ones that appeal to belief based on good reasons, coherence, and virtue reliabilism.
9 Carter (Citation2022, 49) characterizes pre-emption as follows: a subject ‘comes to possess a belief in a way that pre-empts one’s cognitive faculties if and only if the process that issues the belief lacks the opportunity (trivially) to bypass the subject’s cognitive faculties.’
10 Carter (Citation2022) Carter explores the themes mentioned in this section in earlier work (see, in particular, his (Citation2018), (Citation2020), and (Citation2021)). As Carter (Citation2022, 45) notes his proposed account of epistemic doxastic autonomy draws heavily on Mele’s (Citation2001) history-sensitive account of attitudinal autonomy relevant to moral responsibility.
11 The challenge of explaining why knowledge is more valuable than true belief is sometimes called the ‘Meno problem’ because Plato raises it in the Meno. It is also sometimes called the ‘primary value problem’ while the task of accounting for the greater value of knowledge than any proper subset of its parts is sometimes labelled the ‘secondary value problem’ (see, e.g. Pritchard Citation2007). Kvanvig (Citation2003) offers a book-length discussion of the value of knowledge. See Sylvan (Citation2023, Sect. 3) for a critical discussion of Carter’s (Citation2022, Chap. 2) dismissal of internalist accounts of autonomy.
12 Examples of finally value include a rare stamp, the pen used by Abraham Lincoln to sign the Emancipation Proclamation, and a dress worn by Princess Diana. For these examples, see Beardsley (Citation1965), Kagan (Citation1998), and Rabinowicz and Rønnow-Rasmussen (Citation2000), respectively. The distinction between intrinsic and extrinsic value marks a distinction regarding the metaphysical grounds of value. Intrinsic value is value possessed in virtue of intrinsic properties while extrinsic value is value possessed in virtue of extrinsic properties. For a long time philosophers contrasted mere instrumental value and intrinsic value with one another. However, Korsgaard (Citation1983) argues that this is a false contrast, and that the real contrast is between mere instrumental value and final value. Korsgaard’s key move is to take the instrumental/non-instrumental (final) value distinction to concern how things are valued and the intrinsic/extrinsic distinction to concern the (metaphysical) source of value. Korsgaard thinks of final value in terms of something’s being valued for its own sake. Carter characterizes something as having final value in terms of its being valuable for its own sake. We do not wish to take a stance on whether, ultimately, a subjectivist account of final value (in terms of valuing attitudes) or an objectivist account is more plausible. However, for the sake of our discussion we adopt Carter’s formulation.
13 This is an instance of what Sylvan (Citation2018, 391) labels ‘(Stronger Swamping Premiss)’. As Sylvan (Citation2018, 2.1) reports, swamping was initially presented as a challenge specifically to reliabilism. However, as he goes on to note and account for (Citation2019, 2.2) – following Pritchard (Citation2010) (and others) – the scope of the challenge posed by value swamping is broader: it is a challenge for veritism.
14 Alvin Goldman and Erik Olsson have argued that reliabilist veritists can solve the swamping problem by appealing to conditional probabilities. See Goldman and Olsson (Citation2009), Sect. 4.
15 Sylvan (Citation2018, 400–401) endorses a distinction between objective and apparent reasons for actions and attitudes. An objective reason for a belief is a reason that counts in favour of it by bearing on its correctness. Sylvan notes that there is considerable disagreement concerning the nature of apparent reasons. Roughly put, we can think of apparent reasons as reasons given by appearances. Not all apparent reasons are factive – e.g., visual appearances for someone subject to the Müller-Lyer illusion. However, according to Sylvan, other apparent reasons do qualify as factive – and these are the reasons for which he uses the label ‘factively apparent reasons’. This category of reasons include appearances such as seeing that p and remembering that p (Sylvan Citation2018, 400).
16 Carter (Citation2022, 132–134) discusses the prospects of incorporating autonomy into the Sylvanian framework of fitting ways of valuing truth, concluding that they do not look bright. However, see Sylvan (Citation2023, Sect. 4) for an argument to the contrary.
17 It is worth pointing out that Korsgaard (Citation1996) requires something like reflection or self-reflection for self-constituting action. This is indicative of her Kantian sympathies. On Kantian views on autonomy, if one has a given a desire, there is the always the further question as to whether one should act on it. If one decides to do so, then something that is not free (the desire) becomes a free reason (my choosing to act on it). The action, accordingly, can be classified as free or autonomous. This is worth pointing out because on this picture the exercise of the subject’s capacity for reflection plays a central role in explaining the autonomy of action. It is also worth pointing out because there is an epistemic analogue of this in BonJour and Lehrer. Given a perceptual experience, one can ask whether one should be believe on it – whether the experience is a good reason for belief. In their earlier work, Lehrer and BonJour took knowledge to require the subject to identify antecedent mental states as causes of belief and reflectively endorsing them as good reasons. This position would appear to be the knowledge analogue of the Korsgaardian line on autonomous action. Lehrer and BonJour thus appear to have at one point advocated an autonomous knowledge-style view where the requirement of reflective endorsement is what ensures doxastic autonomy. It also appears that a transposition of Korsgaard’s full story concerning autonomous action to the case of knowledge renders an internalist view on autonomous knowledge. By extension, since Carter goes in for an externalist notion of autonomy, the strategy of transposing Korsgaard’s line on action to knowledge is feasibly only if it leaves out certain features or commitments of Korsgaard’s view.
18 Also, for what it is worth, in the discussion of the book symposium Carter classified autonomy and intellectual self-constitution as non-epistemic.
19 For discussion of the question of unity, see Pedersen (Citation2017).
20 To be fair, Carter does not merely stipulate autonomy to be a condition of knowledge. He presents cases in favour of thinking so. Even so, the support offered by these cases in favour of an autonomy-condition on knowledge rests on these cases eliciting the intuition or judgment that autonomy matters to knowledge. This kind intuition or judgment-driven methodology about cases is not uncommon. However, it is a substantial methodological commitment.
21 On this matter, see Pedersen (Citation2017, Sections 2–4).
22 We note that, while Carter’s discussion of Sylvan’s work adopts Sylvan’s own terminology – and is thus put in terms of fundamental/non-fundamental and derivative/non-derivative value – Carter switches terminology when putting forward his own proposal about autonomous belief and intellectual self-constitution. In particular, instead of talking explicitly in terms of derivative value he uses the locution ‘on account of’. Instead of saying that autonomous belief derives its value from bearing a certain kind of relationship to intellectual self-constitution he says that autonomous belief possesses final ‘on account of’ a relationship that it bears to finally valuable intellectual self-constitution. As far as we can tell, the key considerations and arguments presented in this section can be systematically rewritten using ‘on account of’ and still run. For further discussion, see Section 6.
23 We say ‘dangerously close’ because constitution is not identity. Among other things, identity is symmetric while constitution is not. If a is identical to b, then b is identical to a. However, it is not the case that, if a constitutes b, then b constitutes a. Still, when a constitutes b, it makes up b and in this sense there is a very metaphysically intimate relationship between a and b. We note that some philosophers do hold the view that constitution is identity (see, e.g. Noonan (Citation1993) for the view that constitution is a special case of identity).
24 Carter could also endorse an instance of (FDV) put in terms of autonomous beliefs, proper parthood, and either the intellectual self or the self. For these two instances overgeneration would also be an issue – and autonomous beliefs would be a proper part of the self or intellectual self, i.e. that from which they are supposed to derive their final value.
25 We have not considered the following instance of (FDV): autonomous beliefs derive final value from intellectual self-constitution (and so, possess final, derivative value) if (i) intellectual self-constitution is finally valuable, and (ii) autonomous beliefs constitute intellectual self-constitution and derive final value from intellectual self-constitution in virtue of doing so. We have not considered this instance because we find it too heavy on constitution talk. However, even if we left this issue aside, it would face the master concern – specifically, in the form put forward against constitution-involving instances of (FDV).
26 For what it is worth, for the symposium on Autonomous Knowledge, in presenting Carter’s view and commitments one of the authors (Pedersen) explicitly attributed to Carter the idea that autonomous beliefs possess final, derivative. This attribution did not trigger any objections or comments concerning interpretative correctness from Carter.
27 This is Table 5.3 from Carter (Citation2022, 143), with a row added for autonomous and heteronomous actions. This addition simply includes in the table something to which Carter is clearly committed.
28 Earlier we said that we want to remain neutral between subjectivist and objectivist conceptions of final value but that we follow Carter in using terminology suggestive of objectivism for the purposes of our discussion of his view (i.e. ‘valuable for its own sake’ rather than ‘valued for its own sake’). We submit that this part of our discussion is one point where it may be difficult to see how one could sustain an objectivist reading. It seems that an object’s having the relational property of being rare does not by itself suffice for its being finally valuable. Some diseases are rare but that hardly makes them finally valuable. Perhaps here a subjectivist about final value has a relatively straightforward way out: rare stamps are valued for their own sake – i.e. are the objects of a certain attitude – while this is not so for rare diseases.
29 Rabinowicz and Rønnow-Rasmussen say, ‘The final value of a thing [the dress] comes from the importance of another thing [Diana]’ (Citation2000, 42).
30 (IVABa) and (IVABb) are somewhat similar to (FDVABC). (IVABa), (IVABb), and (FDVABC) all attribute derivative value to autonomous beliefs. However, note that there is also a crucial difference. (FDVABC) attributes final, derivative value to autonomous beliefs whereas (IVABa) and (IVABb) only attribute instrumental value to them.
31 Our reasoning assumes that constitution and contribution are closed under proper parthood – i.e. that, if x constitutes y and y is a proper part of z, then x constitutes z; and if x contributes to y and y is a proper part of z, then x contributes to z. On the picture just sketched, one might take the intellectual self and the practical self not to possess final, fundamental value – reserving this status only for the full self and according instrumental value to the two restricted selves in virtue of their status as proper parts of the full self. We leave this issue open here. We note that (IVABa), (IVABb), and the considerations and arguments offered in this section can be rewritten with the intellectual self taking the place of the self and intellectual self-constitution the place of self-constitution. However, we also wish to note that rewriting this section along these lines would appear to give rise to the kinds of issues concerning the metaphysics of value discussed in Section 5.