Tuesday 24 March 2015

Implicit Bias, Confabulation, and Epistemic Innocence

This is the last in our series of posts on the papers published in a special issue of Consciousness and Cognition on the Costs and Benefits of Imperfect Cognitions. Here I summarise my paper ‘Implicit Bias, Confabulation, and Epistemic Innocence’.

I explore the nature of confabulatory explanations of action guided by implicit bias by focusing on two imaginary cases: 


The case of Roger: ‘Roger is on a hiring panel deciding from a stack of CVs which candidates to invite to interview. Roger thinks of himself as an egalitarian, and not as somebody who is sexist. The CVs are not anonymous with respect to gender. Roger chooses not to invite any female applicants to interview. Katie is one of the female candidates who Roger chooses not to invite to interview. Katie’s CV is of equal or better quality than at least some of her male competitors who did get invited to interview, and had Katie’s CV been headed with a male name, Katie would have been invited to interview’ (p. 3).

The case of Sylvia: ‘Sylvia is walking down the road on her way to work. Sylvia thinks of herself as an egalitarian, and not as somebody who is racist. She sees a black man walking towards her. Sylvia crosses the road. The man is not acting in a threatening manner. Had the man not been black, Sylvia would not have crossed the road’ (p. 3).


I outline five common features of confabulatory explanations, they (1) are false or ill-grounded; (2) are offered as the answer to a question; (3) have a motivational component; (4) fill a gap, and (5) are reported without any intention to deceive. I discuss the cases of Roger and Sylvia in light of these features, suggesting that the explanations given by these subjects of their decisions or actions are cases of confabulation. 

Next I introduce epistemic innocence. I say that ‘an epistemically faulty cognition is epistemically innocent if, at a given time, it endows some significant epistemic benefit (Epistemic Benefit) onto the subject, which could not be otherwise had, because alternative, less epistemically faulty cognitions are in some sense unavailable to her at that time (No Alternatives)’ (p. 6–7).

If a cognition is epistemically innocent that is not to say that it is free from faults or even epistemically good—epistemic innocence does not track epistemic goodness. The claim is only that the cognition confers some epistemic benefit which is otherwise unavailable to the subject at that time. I suggest that confabulatory explanations of decisions or actions guided by implicit bias meet the conditions on epistemic innocence. I identify two epistemic benefits which might be had by such explanations: 


(1) maximizing the future acquisition of true beliefs by filling explanatory gaps, and 
(2) maintaining consistency among a subject’s beliefs about herself.

With respect to (1) I suggest that in having an explanation and endorsing a position one can receive feedback and the problematic things that are believed (I did not invite Katie to interview because her CV was of poorer quality or I crossed the road because the black man was behaving in a threatening way) become available to introspective reflection, and open to feedback and revision. With respect to (2) I argue that confabulatory explanations allow one to maintain a coherent self-concept. The explanation maintains consistency between beliefs about the values one is committed to, and the belief that one made some decision or performed some action.

I argue that these benefits are only available via the confabulatory explanation because less epistemically faulty cognitions which deliver the same benefits are in some sense unavailable to the subject who confabulates. I focus on motivational unavailability, an alternative explanation is unavailable in this sense ‘if it is inhibited or not accessed due to motivational factors’ (p. 7). Confabulatory explanations of actions driven by implicit bias are sometimes such that less epistemically faulty alternative cognitions are motivationally unavailable to the subject insofar as they cite implicit biases against certain groups, biases which she is motivated not to recognise in herself or have others recognise in her.

My conclusion is that, at least in some cases, confabulatory explanations of decisions or actions guided by implicit bias can be epistemically innocent, and the epistemic evaluation of confabulatory explanations ought to take into account the context in which the cognition occurs. If we do this we are able to provide a richer account of the epistemic status of confabulatory explanations, and we can resist the view that pragmatic benefits come at the expense of epistemic ones. If we focus on the context in which confabulatory explanations occur, and their potential epistemic benefits, we can give a richer epistemic evaluation of them.

No comments:

Post a Comment

Comments are moderated.