5. Conclusion

According to Strategic Reliabilism, the quality of a reasoning strategy is a

function of its expected costs and benefits as well as those of its competitors.

But which of two reasoning strategies has a better cost-benefit ratio for a

particular reasoner is an empirical question. From the perspective of

Strategic Reliabilism, when one assesses positively someone’s reasoning, he

must present evidence that these individuals have availed themselves of the

best strategies of which they are cognitively capable. Of course, conceptual

reject-the-norm arguments don’t do this. They conclude that proponents

of the HB program are mistaken in their negative evaluations of their

subjects’ reasoning (or of the reasoning that leads to the modal answer

given by subjects). But these arguments do not appeal to the contingent,

empirical factors that are relevant to the proper assessment of the quality

of a reasoning strategy. Therefore, conceptual reject-the-norm arguments

are incapable of making the case that our actual reasoning strategies are

epistemically excellent.

It’s useful to note that any argument—whether purely conceptual or

not—that aims to show that our actual reasoning strategies are epistemically

excellent is doomed. There are two reasons for this. First, the

conclusion is known to be false. Ameliorative Psychology offers us a

number of reasoning strategies that are superior to the ones most of us

actually use. Second, suppose the conclusion weren’t known to be false.

Suppose that as far as we know, our actual reasoning strategies are the best

available. It would still be extraordinarily unlikely that there aren’t better strategies out there, just waiting to be discovered. So an argument can’t be

sound if it purports to show that when it comes to our reasoning we don’t

need no stinkin’ badgerers.

There is a way for reject-the-norm arguments to avoid these objections.

They can change the subject. They can employ or define an epistemic

notion such that our actual reasoning strategies are unbeatable

according to that notion. We suspect that something like this is going on

with the conceptual reject-the-norm arguments of Gigerenzer and Cohen.

But as long as these arguments don’t touch upon the quality of subjects’

reasoning—the focus of Strategic Reliabilism—there is no reason for proponents

of the HB program to worry. They can grant that a reject-thenorm

argument shows that there is some Pickwickian sense in which

subjects are ‘‘rational’’ or not making an ‘‘error,’’ and then point out that

on this view, one can be ‘‘rational’’ and ‘‘error-free’’ while regularly engaging

in atrocious reasoning, reasoning that (among other things) predictably

leads to massively suboptimal outcomes. There is, admittedly, an

esthetic drawback to taking this line against reject-the-norm arguments: It

requires the use of an annoying number of scare quotes. The substantive

lesson to draw, however, is that unbridled, a priori optimism about our

cognitive capacities inevitably leads to normative evaluations that either

are meant to be prescriptive and are in practice absurd or are empty of

prescriptive force.

We have argued that two conceptual reject-the-norm arguments fail;

and we have also argued that any conceptual reject-the-norm argument

must fail. It doesn’t follow, however, that every normative assessment

made by proponents of the HB program is correct. When it comes to

assessing the quality of a particular reasoning strategy, there is no substitute

for attending to the costs and benefits of that strategy and its

competitors. What does Strategic Reliabilism say about the reasoning

strategies that proponents of the HB program claim result in biases (e.g.,

base rate neglect)? We turn to this issue in the next chapter.

According to Strategic Reliabilism, the quality of a reasoning strategy is a

function of its expected costs and benefits as well as those of its competitors.

But which of two reasoning strategies has a better cost-benefit ratio for a

particular reasoner is an empirical question. From the perspective of

Strategic Reliabilism, when one assesses positively someone’s reasoning, he

must present evidence that these individuals have availed themselves of the

best strategies of which they are cognitively capable. Of course, conceptual

reject-the-norm arguments don’t do this. They conclude that proponents

of the HB program are mistaken in their negative evaluations of their

subjects’ reasoning (or of the reasoning that leads to the modal answer

given by subjects). But these arguments do not appeal to the contingent,

empirical factors that are relevant to the proper assessment of the quality

of a reasoning strategy. Therefore, conceptual reject-the-norm arguments

are incapable of making the case that our actual reasoning strategies are

epistemically excellent.

It’s useful to note that any argument—whether purely conceptual or

not—that aims to show that our actual reasoning strategies are epistemically

excellent is doomed. There are two reasons for this. First, the

conclusion is known to be false. Ameliorative Psychology offers us a

number of reasoning strategies that are superior to the ones most of us

actually use. Second, suppose the conclusion weren’t known to be false.

Suppose that as far as we know, our actual reasoning strategies are the best

available. It would still be extraordinarily unlikely that there aren’t better strategies out there, just waiting to be discovered. So an argument can’t be

sound if it purports to show that when it comes to our reasoning we don’t

need no stinkin’ badgerers.

There is a way for reject-the-norm arguments to avoid these objections.

They can change the subject. They can employ or define an epistemic

notion such that our actual reasoning strategies are unbeatable

according to that notion. We suspect that something like this is going on

with the conceptual reject-the-norm arguments of Gigerenzer and Cohen.

But as long as these arguments don’t touch upon the quality of subjects’

reasoning—the focus of Strategic Reliabilism—there is no reason for proponents

of the HB program to worry. They can grant that a reject-thenorm

argument shows that there is some Pickwickian sense in which

subjects are ‘‘rational’’ or not making an ‘‘error,’’ and then point out that

on this view, one can be ‘‘rational’’ and ‘‘error-free’’ while regularly engaging

in atrocious reasoning, reasoning that (among other things) predictably

leads to massively suboptimal outcomes. There is, admittedly, an

esthetic drawback to taking this line against reject-the-norm arguments: It

requires the use of an annoying number of scare quotes. The substantive

lesson to draw, however, is that unbridled, a priori optimism about our

cognitive capacities inevitably leads to normative evaluations that either

are meant to be prescriptive and are in practice absurd or are empty of

prescriptive force.

We have argued that two conceptual reject-the-norm arguments fail;

and we have also argued that any conceptual reject-the-norm argument

must fail. It doesn’t follow, however, that every normative assessment

made by proponents of the HB program is correct. When it comes to

assessing the quality of a particular reasoning strategy, there is no substitute

for attending to the costs and benefits of that strategy and its

competitors. What does Strategic Reliabilism say about the reasoning

strategies that proponents of the HB program claim result in biases (e.g.,

base rate neglect)? We turn to this issue in the next chapter.