A Moral Account of First-person Authority

This post is by Fleur Jongepier (picture below). Starting Autumn 2017 she will be based in Cambridge (UK), working on the role and value of self-knowledge in contemporary liberalism.

In the previous blogpost, I introduced some examples that I suggested provide a challenge to what I referred to as the ‘traditional’ approach to the notion of first-person authority, namely, the view according to which first-person authority and self-knowledge always come and go together. I ended the post by mentioning the following three views about the relation between first-person authority and self-knowledge:

1. The Decoupling View. They have first-person authority despite not having self-knowledge.
2. The Negative Traditional ViewThey do not have self-knowledge, therefore do not have first-person authority.
3. The Positive Traditional View. They have self-knowledge and therefore have first-person authority.

One reason for thinking the second, Negative Traditional view (‘no self-knowledge, no authority’) is mistaken is because it appears to yield way too rigorous consclusions. It fails to reflect the way that we treat people’s sincere self-ascriptions. Even if someone says “I want to kill myself” and is clearly self-deceived somehow, it does not seem to be the case that we would be inclined or indeed justified to simply overrule or correct her self-ascription, and so it seems we should not deny her first-person authority.

Even if someone is wrong when s/he says “I feel X” or “I want Y”, it is still in an important sense out of place or inappropriate to e.g. say “No you’re not” – even if you’re entitled, epistemically speaking. So even when a person issues a self-ascription that is a very poor guide to how s/he will act in the future, this does not license us to correct, challenge or overrule the self-ascription. Indeed, we need to take their self-ascriptions seriously if we want to try to make them see their mistake or change their minds. Granting someone first-person authority is a condition for e.g. engaging in psychotherapy at all (see Strijbos and Jongepier forthcoming).

What about the third option? On the assumption the traditional view is right, granting that subjects have first-person authority in the given examples might also mean they therefore must have self-knowledge, given that on the traditional view, first-person authority and self-knowledge are a package deal. Those defending this option will thus want to say that e.g. the person saying he wants to kill himself in some sense does have self-knowledge, even if he does not (luckily) take any steps to act according to his own self-ascription. The defender of the third view will thus have to argue that it’s not evident that one lacks self-knowledge if one does not act in accordance with one’s own self-ascription.

And indeed such a view has been defended (see e.g. Ferrero 2003 and Bortolotti 2009). Luca Ferrero instance claims that “First-person authority is characteristic of self-ascriptions of present attitudes” and that “The distinctive first-person authority of [someone’s] self-ascriptions concerns ... whether she takes responsibility for them, not whether the self-ascribed attitude is both correct and a reliable guide to future conduct” (2003, 570 emphases in original). Lisa Bortolotti in a similar vein claims that subjects can have knowledge of their attitudes “no matter how representative of their future behaviour those attitudes would [turn] out to be” (Bortolotti 2009, 639). The subject, Ferrero and Bortolotti emphasize, has self-knowledge and responsibility with respect to what s/he is currently thinking or judging, and it’s this type of self-knowledge that lies at the basis of his self-ascription being authoritative.

I think this third view is much more plausible than the second, but is still somewhat unsatisfying. The person who self-ascribes the intention or desire to commit suicide indeed knows what she is currently thinking, and perhaps we can assume she can ‘take responsibility’ for her self-ascription in some sense (notice that this is much harder for the example of the person suffering from OCD). But the question is: is first-person authority due to this type of self-knowledge? That is, to the sort of synchronic self-knowledge that we have ‘here and now’ as it were?

For example in the case of the person having suicidal thoughts, the (lack of) first-person authority is, according to the third, Positive Traditional view, not connected to the fact that she is right about the fact that she wants to die, but about whether she is right about here and now having occurrent thoughts to that effect. How plausible is this? If nothing that she says reflects having the actual desire but only reflects her occurrent thoughts or judgments, then I don’t see how her self-ascription of her desire or intention, could be true.

In other words, it’s just not plausible that the person has self-knowledge in the relevant sense, by which I mean knowledge of her mental states, that is, even though she may well have knowledge of what’s going on in inner speech, say, or knowledge of her judgments. This is also evident in the second example where the subject explicitly rejects her own self-ascription (“I want to kill my daughter”) by denying it reflects the sort of person that she is. If first-person authority and self-knowledge are to be intimately connected, then first-person authority had better be connected to a type of self-knowledge that is more substantial and less easily obtained.

I think this leaves the first, ‘decoupling’, view. I think what we should say about the sort of cases I’ve been discussing – and, indeed, about the grounds of first-person authority in general – is that such authority is not connected so intimately to people’s self-knowledge as is sometimes assumed. It’s perfectly possible for someone to lack self-knowledge – to fail to know or be radically deceived about one’s own intentions or desires – whilst having first-person authority all the same. That is, in spite of self-deception or self-ignorance, we still have strong reasons not to correct, challenge, ignore or otherwise disrespect ‘epistemically bad’ avowals. The grounds for first-person authority are thus, perhaps, not epistemic after all.

But then what are they? I don’t have the space to defend a full alternative view here, but here’s a rough sketch: the grounds of first-person authority, in ‘good’ and ‘bad’ cases, are not epistemic but moral. Our not being inclined to, and more importantly not being justified to, challenge, correct or ignore people’s self-ascriptions, has to do primarily with a certain sort of respect that we owe to persons; a sort of respect that is due to them even if they are deeply mistaken about their own minds. I think the idea is simple enough: granting and being granted first-person authority is just part of what it means to treat someone as a person.

As for instance Philip Pettit puts it: “To be treated properly as a person ... is to be treated as a voice that cannot be dismissed without independent reason: to be taken as someone worth listening to.” (1997, 91). In all fairness, I think the third view can take these moral considerations regarding first-person authority on board as well, if, at least, it grounds first-person authority not in the sort of synchronic self-knowledge we have of our occurent judgments, but rather in our capacity to judge at all. In that case, ‘respecting someone as a person’ is just short for ‘respecting another’s capacity for judgment’.

This is no proper defence of the Decoupling view of course. But in any case, I think a ‘moral’ approach towards first-person authority might be worth exploring, and more generally that (some) discussions about self-knowledge might be able to benefit from discussions in moral and political philosophy.