4.4. Adopting less reliable (but cheaper) reasoning strategies

К оглавлению1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 

Let’s now acknowledge the mirror image of the point just made: the costbenefit

approach to epistemic excellence suggests that it is possible to

become a better reasoner by adopting new reasoning strategies that are less

reliable than current strategies (lower left quadrant of Figure 3.5). How

can this be? The answer lies in our discussion of opportunity costs. Consider

again Figure 3.6. Suppose that Test Taker is expending c1 resources

on strategy E for solving quantitative problems on the test. For these

problems, there is no other reasoning strategy available that would be

more reliable. Nevertheless, Test Taker is expending a lot of cognitive

effort (c1) on the quantitative problems. By expending much less effort

(c), he can be almost as reliable on these problems. Further, he frees up

resources to tackle new problems or to tackle other problems more effectively.

So by adopting a new reasoning strategy that leads to decreased local

reliability, he can reallocate resources so as to increase global reliability.

The possibility that epistemic excellence might be served by replacing

a reasoning strategy with another that is less reliable and easier to use

might seem paradoxical. But a cost-benefit perspective on epistemology

leads us naturally to recognize this apparently paradoxical possibility. One

might suspect that this insight is unlikely to have much practical applicability.

But the practical import of this prospect becomes evident when

we consider (again) that excellent reasoners reason reliably about significant

matters. Suppose a reasoner employs a consider-the-opposite strategy

for a very wide range of reasoning problems. This strategy involves explicitly

considering reasons for why one’s judgment might be wrong; it has

been shown to decrease overconfidence (Plous 1993, 228). Suppose further

that although the consider-the-opposite strategy makes the reasoner more

reliable, he employs it so often that it has turned him into an unhappy,

neurotic nebbish, impossible to get along with. Even the most trivial judgments

get the full consider-the-opposite treatment. Such a person could

well become a better reasoner (and a happier person) by restricting the use

of the consider-the-opposite strategy to just those problems for which it is

important not to be overconfident. (Don’t apply consider-the-opposite to

the issue of whether you changed the roll of toilet paper; do apply it to the

Extracting Epistemic Lessons from Ameliorative Psychology 69

issue of whether you turned off the safety switch at the nuclear power

plant.)

Let’s now acknowledge the mirror image of the point just made: the costbenefit

approach to epistemic excellence suggests that it is possible to

become a better reasoner by adopting new reasoning strategies that are less

reliable than current strategies (lower left quadrant of Figure 3.5). How

can this be? The answer lies in our discussion of opportunity costs. Consider

again Figure 3.6. Suppose that Test Taker is expending c1 resources

on strategy E for solving quantitative problems on the test. For these

problems, there is no other reasoning strategy available that would be

more reliable. Nevertheless, Test Taker is expending a lot of cognitive

effort (c1) on the quantitative problems. By expending much less effort

(c), he can be almost as reliable on these problems. Further, he frees up

resources to tackle new problems or to tackle other problems more effectively.

So by adopting a new reasoning strategy that leads to decreased local

reliability, he can reallocate resources so as to increase global reliability.

The possibility that epistemic excellence might be served by replacing

a reasoning strategy with another that is less reliable and easier to use

might seem paradoxical. But a cost-benefit perspective on epistemology

leads us naturally to recognize this apparently paradoxical possibility. One

might suspect that this insight is unlikely to have much practical applicability.

But the practical import of this prospect becomes evident when

we consider (again) that excellent reasoners reason reliably about significant

matters. Suppose a reasoner employs a consider-the-opposite strategy

for a very wide range of reasoning problems. This strategy involves explicitly

considering reasons for why one’s judgment might be wrong; it has

been shown to decrease overconfidence (Plous 1993, 228). Suppose further

that although the consider-the-opposite strategy makes the reasoner more

reliable, he employs it so often that it has turned him into an unhappy,

neurotic nebbish, impossible to get along with. Even the most trivial judgments

get the full consider-the-opposite treatment. Such a person could

well become a better reasoner (and a happier person) by restricting the use

of the consider-the-opposite strategy to just those problems for which it is

important not to be overconfident. (Don’t apply consider-the-opposite to

the issue of whether you changed the roll of toilet paper; do apply it to the

Extracting Epistemic Lessons from Ameliorative Psychology 69

issue of whether you turned off the safety switch at the nuclear power

plant.)