In section 7 of the target article, Carruthers considers the proposal that we have two levels of mentality, conscious and unconscious, corresponding to the two reasoning systems posited by many psychologists, and that we have different forms of access to the attitudes at the two levels – merely interpretative access to those at the unconscious level, but introspective access to those at the conscious level. Prima facie, this mixed position is an attractive one, which does justice both to the evidence for psychological self-interpretation cited by Carruthers and to the everyday intuition that we can introspect our conscious thoughts. Carruthers rejects the option, however. Although conceding that we have introspective access to conscious thinking, he denies that we have such access to conscious judgments and decisions. I argue here that this conclusion is too hasty.
Carruthers' argument turns on the claim that judgments and decisions terminate reasoning processes and produce their characteristic effects directly, without further processing. Conscious thinking, on the other hand, involves rehearsing mental imagery, especially inner speech, and this has only an indirect influence on thought and action. The route may be metacognitive: A rehearsed assertion with content p may give rise to an (unconscious) metacognitive belief, to the effect that one believes that p or that one is committed to the truth of p, which, together with suitable desires, will lead one to think and act as if one believes that p. Or the rehearsed assertion may be processed as testimony, leading one to form the first-order belief that p, which will then guide behaviour in the normal way. On either route, Carruthers argues, the conscious event gives rise to the effects of a judgment only through the mediation of further cognitive processing, and so does not count as a judgment itself. Similar considerations apply to decisions, although here Carruthers mentions only the metacognitive route.
I am sympathetic to Carruthers' account of conscious thinking, and I agree that imagistic rehearsals influence thought and action through the mediation of unconscious cognitive processes. But this is not incompatible with the commonsense view that some conscious events are judgments and decisions. To see this, we need to take seriously the suggestion that the conscious mind is a distinct level of mentality. Carruthers has himself developed a version of this view, arguing that the conscious mind (the psychologists' System 2) is not a separate neural structure, but rather, a higher-level “virtual” one, realized in cycles of operation of a more basic unconscious system (System 1), which, among many other tasks, generates and processes the imagery involved in conscious thinking (Carruthers Reference Carruthers2006; Reference Carruthers, Evans and Frankish2009; for a related version, see Frankish Reference Frankish, Carruthers and Boucher1998; Reference Frankish2004; Reference Frankish, Evans and Frankish2009). And from this perspective it is natural to regard appropriate utterances in inner speech as genuine judgments and decisions – at least when they achieve their effects via the metacognitive route. For these events will terminate reasoning processes at the higher level and on the relevant topic. The further processing occurs at the lower level and is devoted to a different topic. When I rehearse the sentence, “Polar bears are endangered” in assertoric mode, this terminates my reasoning about polar bears. The subsequent unconscious reasoning is about how to interpret and respond to this assertion, not about whether the conclusion it expresses is correct. These processes can be thought of as implementing the higher-level attitude, and their existence does not compromise the status of the conscious event as a judgment.
It is true that the lower-level processes may sometimes fail to generate the appropriate effects (for example, if the desire to execute one's commitments is overridden by a stronger desire), but this is irrelevant. On every view there are some implementing processes, at least at a neurological level, and these processes may go awry. And if we have a settled habit of interpreting appropriate utterance rehearsals as expressions of belief or commitment, and a settled desire to act consistently or to discharge our commitments, then the right effects will follow most of the time. Similar considerations apply to decisions.
The only peculiarity of the two-level view is that the processes that implement conscious judgments and decisions are cognitive ones. But why should that matter? Compare the way the judgments and decisions of a company are implemented. The edicts emerging from the boardroom require further processing in order to affect the activities of the organization, and this processing involves reasoning on the part of the staff involved. (Again, this will have a metarepresentational character, involving beliefs about what the directors have concluded.) But we still want to say that the judgments and decisions were made in the boardroom, rather than in the cubicles of the junior staff.
What about cases in which a rehearsed utterance generates its effects via the second route, being processed as testimony and generating a first-order belief? Here I think Carruthers is right. If further processing serves to evaluate the conclusion reached rather than simply to implement it, then this does disqualify the conscious event from judgment status. But note that in such cases, the agents themselves will not think of the conscious events as judgments. For if they did, they would naturally come to believe that they believed, or were committed to, the conclusions expressed, and the subsequent processing would follow the metacognitive route. Thus, there is no reason to regard such events as candidates for judgments in the first place. (We might think of them as hypotheses or self-suggestions.) Again, the same goes for decisions.
I conclude that Carruthers' case against a mixed position is not compelling. It is important to stress that the proposed mixed position does not involve positing additional introspective mechanisms. Carruthers allows that we have introspective access to conscious (System 2) thinking; I am simply claiming that some of the introspectable events can be legitimately classified as judgments and decisions. The proposal is merely a reconceptualization of the processes Carruthers describes. But it is a natural one, given a two-level view of the sort Carruthers endorses, and one that accords with intuition. For these reasons it should be preferred. Of course, it would be ad hoc if a two-level view were not independently motivated, but it is (see aforementioned citations).
