ABSTRACT:

> Our interest is in the possibility of there being a philosophically interesting set of useful false beliefs where the utility in question is specifically epistemic. As we will see, it is hard to delineate plausible candidates in this regard, though several are promising at first blush. We begin with the kind of strictly false claims that are said to be often involved in good scientific practice, such as through the use of idealisations and fictions. The problem is that it is difficult to see that there would be any epistemic utility in believing such claims, as opposed, say, to merely accepting them. Next we turn to the challenge posed by epistemic situationism , which when embedded within a plausible form of virtue epistemology appears to show that sometimes purely situational factors can play a significant explanatory role in one’s cognitive success. But again it is hard to see how the role that these epistemically beneficial situational factors contribute can be cashed out in terms of epistemically useful false beliefs on the part of the subject. Finally, we turn to the Wittgensteinian conception of hinge commitments ,commitments that are held to be epistemically useful even if false. While the epistemic utility of these commitments is defended, it is argued that one cannot make sense of these commitments in terms of belief. Support is thus canvassed, albeit in a piecemeal fashion, for the thesis that the prospects for there being philosophically interesting cases of epistemically useful false belief are poor.

“Porcupines can shoot their quills”

The way they restrict "false belief" makes the phrase almost an oxymoron. If you "merely accept" that the Earth is a sphere instead of "genuinely believing" it, how is that different from responding yes to the question "is Earth a sphere?", doing your calculations as if Earth is a sphere, making mistakes that reveal that you didn't know the Earth is a bit squished, etc.? All models are false, so either (i) can't be satisfied and must be relaxed, or "technically false" true beliefs are natural and commonplace.

Also, here's my example of an epistemically useful false belief: The idea that there is substance to music theory (more than what you get from a high school education, that is). You will learn a lot of useful things before falsifying it.

Here is one.

Let us grant a couple premises. First the belief that a cookie consent banner that provides a large button to accept and a multiple-click, read the policy first on another page opt-out process, complies with a law that says opt out has to be just as easy to opt in. Second, that your developers are more likely to use such a banner if they believe it's legal than if they knew they were doing something illegal. Third, that you are going to use the tracking information gathered via dark patterns implying consent by default not to sell to data brokers, but to actually better target marketing of your own publication. Fourth, that reading your epistemology journal will actually make the readers you target better at epistemology.

I think I might have discovered an epistemically useful false belief?