I used to think that the main problem with moral psychologists’ use of ‘autonomy’ is that ‘autonomy’ is too ambiguous a word. There is the autonomy that all rational beings supposedly have, vs. the autonomy of which I would have more of if I could drive a car, to mark only one ambiguity. In my first book, I listed 8 different ways in which the term is used and I would not be surprised if someone else finds 10 more. It is very easy to conflate at least 2 “autonomies”, but now I think the ambiguous nature of the word ‘autonomy’ is not the only problem with using it. It’s equally bad that ‘autonomy’ as used by some of us moral psychologists is a term of art that is used as if it were a “natural language” term. Agent autonomy is, correspondingly, a theoretical construct about which we are expected to have pre-theoretical intuitions. The technical nature of the term ‘autonomy’ (and often of related and even fancy-er terms like ‘agential authority’) can easily become invisible to those who use it regularly, much the way I imagine some songwriters no longer notice that “self” does not, in fact, rhyme with “else”.
I will grant that ‘autonomy’ has various uses in natural language: there are autonomous vehicles, after all, and a Basque Autonomous Community. One can also grant that ‘autonomy’ meaning something like “the right to make decisions for oneself, free of coercion, especially paternalistic coercion” is almost natural English – American medical and nursing students take to it very quickly. However, ‘personal autonomy’ as used by moral psychologists is no more ordinary English than,say, ‘internal reason’. Saying that ‘autonomy’ means “self rule” isn’t helpful. ‘Self rule’ is only used in ordinary language with regard to nations, not individuals. ‘Agential authority’ or ‘agential’ for that matter is clearly philosophers’ talk – my spellchecker won’t even let me write “agential”. Even ‘agent’ is a term of art, unless we are talking about the sort of agent who spies or the sort who might help you break into the entertainment industry. Non-philosophers who are plenty educated enough to bandy about such words as ‘irrational’ or ‘bad faith’ never, and I mean never, say “I wonder if, when you scream at me, it’s an autonomous action on your part” or “he is so in love that his self-rule is compromised”, or “I figure her belief in astrology expresses no agential authority”. ‘Self-control’ is the closest natural English term we have.
Why does it matter? Terms of art are legit, of course, and philosophy is not all about natural language nor all about intuitions. However, it is an error to use terms of art, steeped as they are in theory, to elicit intuitions. We do not have pre-theoretical intuitions about which individuals and actions are autonomous.
The word “autonomy” does have natural connotations. It is a suggestive term. If asked who is more autonomous, a master or a slave, any guessing undergrad will notice that the slave sounds less autonomous. But if you raise such a question as “who is more autonomous, a slave with perfect self-control or a master who suffers from chronic, ubiquitous, terrible weakness of will?” – do not expect natural language or pre-theoretical intuition to give you the answer. Are rational beliefs more autonomous? Do they form in a more autonomous way? If I were the guessing undergrad I would say “yes”, because autonomy sounds like a good thing and rationality is presumably a good thing. Thus it sounds more plausible that they go together than that they conflict. Why be pessimistic? Beyond the positive connotation shared by ‘rationality’ and ‘autonomy’, I think here is only one honest pre-theoretical answer to the question whether rational beliefs form in a more autonomous way, and the answer is “I don’t know”. This can be followed by: what exactly do you mean by ‘autonomous’, and how is it different from what you call ‘rational’?
A term of art with rich connotations is a treacherous thing. Consider by contrast the term “internal reason” – a boring, bloodless, un-suggestive, connotation-free technical term. The words “internal” and “external” are massively overused by analytical philosophers. I often wish philosophers called their views more original and imaginative names than “internalism” or “externalism”. For one, there are too many internalisms and externalisms and it’s mighty confusing. I could go on further, but I admit it’s basically a stylistic and pedagogical issue. A person with a better verbal memory than mine and a much greater patience with monotony might find nothing amiss with the way we call things “internalism” and “externalism”.
‘Autonomy’ has a deeper problem, as can be shown by the fact that so many people argue about what makes an action autonomous (or not). People don’t disagree that way about which reasons are internal. If you happen to think that Bernard Williams, when he lists the sort of things that can give rise to internal reasons, includes things that don’t belong together – say, desires and values – you do not as a rule argue that Williams made an error and called some reasons “internal” that in fact aren’t (as we can all intuit!). You say that the distinction needs to be redrawn or that a new distinction needs to be added. On the other hand, two philosophers could easily come to argue as to how autonomous an agent Homer Simpson is, and then it is often understandably hard for them to keep their hands off their intuition pumps.
I have been contemplating the term ‘agential authority’ because I have been asked if I don’t think (“don’t you think?”) that irrational beliefs express less agential authority than rational beliefs. The question was asked as if intuition would be enough, or almost enough, to show me that as well. ‘Agential authority’ is another term of art which is sometimes treated as natural language. It does not have as many meanings and connotations, but it is metaphorically evocative in a potentially misleading way.
Imagine that you are trying hard to grade papers despite feeling urges to do just about anything else – play with the cat, watch Netflix, go for a walk (they tell me some people even clean). In such moments, it is natural for you to feel as if your psyche resembles a country, your deliberating self is like a legitimate government, and whatever it is inside you that doesn’t follow your best judgment (the urges? the fraction of the your inner “nation” that support these urges? not clear, really, but whatever it is) is like an organization that defies the authority of the government. Since grading the papers is usually the rational thing to do, such experiences can lend plausibility to the idea that rationality in general is like good government and irrationality in general like crime or insurrection. Now, some of you might recall that elsewhere I take the analogy between your deliberating self and the government to be a bad one even in the case of akratic action. It strikes m as similar to the analogy we make between people and kettles when we feel that by expressing anger we “blow off steam” and thus save ourselves from bigger anger: very intuitive but, despite years of great minds accepting it, ultimately mistaken (PSA: science shows that when you “let out” the anger you feel you increase it. Nothing “blows off”. Kettles have nothing to do with it). This isn’t the place to get into my arguments against people being like countries, but they are not based on denying the intuitive appeal of the self-government trope when you contemplate akrasia.
However, I think that analogizing irrationality in general, and epistemic irrationality in particular, to failure of government invites doubts of the sort that Hume expressed when he considered the theologians’ analogy between the world and a complex artifact that surely cannot be there by accident. Hume asked, essentially, what non-religious reasons we have to see the world as analogous to a lovely object made by an artisan and not, say, to a lovely plant growing from the ground. In a similar vein, take ordinary irrational belief formation. I do not see a particular (pre-theoretical) reason to liken the person who irrationally comes to believe that Elvis is alive to a country experiencing insurrection or infiltration – (i.e. a failure of authority) – and not to a plant suffering from rot, or a heat-guided missile guided the wrong way, or Starbucks putting Leonard Cohen’s Hallelujah in its cheery Christmas music mix, or any number of other things going wrong. So no, I don’t see any (pre theoretical) reason to take irrational beliefs to express less ‘agential authority’ than others. Why not think that they express more, at least sometimes? Who is more authorial of, or has more authority over, her beliefs, the one who boldly tortures the data until they confess, producing an original conspiracy theory, or the wimp who surrenders to the power of the data? I can see why we might call the latter more rational, but I see no strong pre-theoretical, intuitive pull to calling her more “agential”. Or to calling her less “agential”. Or even to thinking, as my undergrads would put it, that belief showing “agential authority” is a thing.
P.S Is epistemic blame a thing? Here are some thoughts. Is ’identification’ any better than ‘autonomy’ or ‘agential authority’? Here is a link – it’s called Just the Booze Talking. Tim Schroeder and I also have a paper about identification but it isn’t funny, really.