Soteriou is interested in showing that supposition is a distinctive kind of “incomplete” mental attitude:
we do not capture adequately the attitudinative aspect of a subject’s mental condition when he supposes that p if we allow that supposing that p can be a stand-alone mental act—that is, if we allow that a subject can be said to be supposing that p for the sake of argument without doing anything else. The fact that a subject has adopted a suppositional attitude toward the content that p depends upon the occurrence of acts that count as discharging the supposition. This is why we fail to capture adequately the attitudinative aspect of a subject’s mental condition when he supposes that p if we say that the subject is merely pretending or imagining that he is representing p as true. When one supposes that p and infers q from p, one isn’t imagining or pretending that one is representing those propositions as true; and this is connected with the fact that when one is engaged in supposition, one is engaged in actual reasoning, not pretend or imagined reasoning. (126)
the subject who supposes that | p for the sake of argument represents p as true by reasoning on the assumption that p (where reasoning on the assumption that p is genuine reasoning, not pretend or imagined reasoning). (126-7)
So, unlike imagining, which is not an act that plays a role in argument, and unlike judgment, which is a stand-alone mental act (i.e. a mental act that one can engage in without doing anything else (see 124)), supposition is an act that presumes the engagement of other acts—specifically, of reasoning—and it raises a self-imposed constraint on the way in which one engages in these other acts of reasoning.
When one reasons on the supposition that p, the relevant constraints on one’s reasoning are self-imposed. They are not simply constraints on one’s reasoning that are imposed by facts in the world whose obtaining one acknowledges. And furthermore, when one reasons on the supposition that p one treats the relevant constraints on one’s reasoning as self-imposed. When one reasons on the supposition that p one recognizes that the constraint of treating p as true is a constraint on one’s reasoning that one has imposed on oneself. One manifests this recognition in the way in which one reasons—for example, by discharging the supposition with an outright conditional judgment or assertion. (127)
I take this to be the clearest statement of the view. It is not adequately captured by the shorter statement that “when one supposes that p for the sake of argument one imposes a constraint on one’s reasoning by reasoning in recognition of it.” The problem with the shorter statement is that it doesn’t make clear what the difference is between recognition of constraint and recognition of self-imposed constraint. The latter is seemingly what is important. We thus need to read “it” in the pithy statement as “self-imposed constraint” and not simply “constraint.”
But this discussion brings out the idea that suppositional reasoning requires reasoning in recognition of self-imposed normative constraints:
one can only genuinely be said to have introduced a supposition into one’s reasoning if one does things that count as discharging that supposition, and one can only discharge the supposition if it has been introduced. (127)
Slightly metaphorically, but using the familiar Kantian terms:
in order for an agent to be capable of governing himself, he must be capable of both imposing obligations on himself, as legislator, and he must be capable of recognizing and acting on those obligations, as the one being legislated to. His authority as self-governing legislator depends upon his own recognition of that authority. In fact it is necessary and sufficient for it. (127)
There is a kind of compatibilism that Soteriou ultimately accepts that is not obviously part of his original argument.
Acting as if one has imposed a constraint on oneself, one thereby imposes the constraint on oneself. One treats oneself as a source of constraint on oneself, and thereby governs oneself. (128)
What is not clear to me is why he thinks this is true. It is true that, if my acting is efficacious, then there is no difference between acting as if I am constrained and acting under such constraint (except that I can presumably remove the constraint at will in the former case). But imagine that my conscious states are epiphenomenal. Appearance in this case would not suffice for reality. If one were on rails, the mere treatment of one’s moving right or left would not make it the case that it was one, rather than the rails, that had the relevant authority.
My objection is too obscure here, but the issue I think I’m ultimately worried about is how to situate the supposed freedom here with respect to the determinism debate. Consider Soteriou’s statement that
the mental activity involved [in acting as if one has imposed a constraint on oneself] is self-determined, in the following respect: one treats oneself as a source of constraint over one’s own thinking, and thereby makes oneself a source of constraint over one’s own thinking. (128)
Is one really acting in a self-determined manner if one is merely acting “as if” one were constrained by one’s self-imposed constraints? For example, if determinism is true then the constraints one acts upon are world-imposed and not self-imposed. The fact that one represents this incorrectly doesn’t obviously entail that one is thereby self-governed, in the way that the running on rails example above was supposed to illustrate. Similarly if indeterminism is true then we have a coincidental series of happenings, but not self-governance.
Bracketing
When one attempts to come up with a proof for the truth of p when one already believes (or knows) that p one brackets one’s belief that p. Importantly, to bracket one’s belief that p is not to withdraw assent from p. The bracketing of one’s belief that p is not something that is subject to epistemic evaluation and it is not something that requires epistemic grounds, whereas withdrawing assent from p (or suspending judgment over p) is subject to epistemic evaluation, and does require epistemic grounds. When one brackets one’s belief that p one does not use p as a premise in the reasoning one is engaged in. (130)
Soteriou then connects bracketing with the discussion on self-imposed constraints like so:
We have a case in which a subject is bracketing his belief that p only when the fact that the subject is not using p as a premise in the reasoning he is engaged in is a constraint on that reasoning that the subject has imposed on himself, and one which the subject treats as a constraint that he has imposed on himself. (130)
The discussion here is interesting and inherits much of the good and the bad from the discussion of supposition, which his conception of bracketing presupposes.
The connection of the view to Descartes’s argument in the Second Meditation (see 131-33) is also very interesting, and very nearly related to my own reading. It lacks a grip on the connection between self-activity and substantiality that Descartes is so clearly exploiting. But useful just the same.
Reflection and Agency
Soteriou criticizes Korsgaard’s reflective distance model (and a related ‘immersion’ model by O’Schaughnessy) for failing to explain how reflective access and the “space of reflective distance” connect to agency in self-critical thought. I’ve been worried about this as well so it is great to see it discussed so explicitly.
Korsgaard then claims that this space of reflective distance ensures that “we are, or can be, active, self-directing, with respect to our beliefs.” It “presents us with the possibility and the necessity of exerting a kind of control over our beliefs.” However, again, one might wonder how | this “space of reflective distance” is supposed to allow us to “exert control over our beliefs.” As Moran
> puts it, it is not as though, glancing inwards, we can simply manipulate our
attitudes as so much mental furniture. So how does agency figure in the exercise of self-critical reflection that is made possible by “the space of reflective distance?” (134-5)
The biggest question that I have after the paper is the relation between the view expressed and the question of compatibilism/incompatibilism. I don’t really know how to evaluate the “treating” doctrine for autonomy, if that is itself isn’t just the expression of a version of compatibilism. But if it is a version of compatibilism, then I don’t see Soteriou as having expressed the relevant line of thought first articulated by Descartes, and reaching its apotheosis in German Idealism, according to which the mind’s rational activity is free activity—freedom and reason are the same, and incompatible with a merely natural casual force.