The Boy Who Cried Wolf, and Bayesian priors

Bayesian probability thinking is often applied to the ID controversy, in different directions. Further to: The stories we are allowed to know and tell, here’s an interesting item on Bayesian priors, featuring the Boy Who Cried Wolf. The author wishes to make a political argument, but I want to focus on a specific point of interest:

… the popular conception of the boy who cried wolf is only half complete.

Villagers begin by believing that no one would falsely claim to get attacked by a wolf. But a false alarm – and the possible subsequent false alarms – lower their probabilistic calculations of wolf attacks and the negative consequences assigned to each claim of a wolf attack.

Then, of course, the boy gets eaten by a wolf. More.

He’s certainly right that most people who use the term “crying wolf” as a throwaway putdown do not believe there even is a wolf. Which amounts to missing the point of the story. Actually, if there had really been no wolf at all, that would be a different story called Chicken Little ( the Sky is Falling). Glass notes:

Economist Thomas Bayes pioneered this method of probabilistic thinking, and it’s now universally used in thinking about how peoples’ beliefs should respond to new information.

Basically, it means that if we live in wolf country, we may have to put up with many false sightings of wolves, though not of falling skies. So the first thing we need to do is compile evidence as to whether we live in wolf country. Follow UD News at Twitter!

11 Replies to “The Boy Who Cried Wolf, and Bayesian priors”

1. 1
2. 2
keith s says:

Denyse,

Could you elaborate on how you think this applies to the ID debate?

Is it “The Boy Who Cried Design”?

3. 3
bornagain77 says:

“The Boy Who Cried Design”?”

But keith s, in the story the boy who cried wolf was effectually eaten by a wolf. So are you saying that design is real and people don’t realize it because they have been fooled into thinking design is just an illusion? 🙂

4. 4
bornagain77 says:

correction : “eventually eaten by…”

5. 5
keith s says:

BA,

It’s Denyse’s analogy, not mine. I’m just asking her to make it explicit.

6. 6
bornagain77 says:

but keith s you, not Denyse, are the one who said “The Boy Who Cried Design”?”, I just thought you wanted to know that it did not convey what I presume you, as a dogmatic atheist, intended to convey.

7. 7
Bob O'H says:

Economist Thomas Bayes…

Oh dear, oh dear… Thomas Bayes was a Presbyterian minister. I don’t think he wrote anything about economics.

8. 8
News says:

Surprised keith s has such trouble understanding the point: Bayesian reasoning would take the significance of a wolf seriously so as not to miss one if it is out there, even if some alarms are false.

However, some users of the expression “the boy who cried wolf” clearly misunderstand the point of the story – that there WAS in fact a wolf, who was ignored.

The story those people should be citing is Chicken Little, whose point is that the sky ISN’T falling.

When assessing probabilities, we need to consider the significance of one of our theses being correct. Why is this difficult?

9. 9
keith s says:

Denyse,

I understand the point of Glass’s article.

Denyse,

Could you elaborate on how you think this applies to the ID debate?

Is it “The Boy Who Cried Design”?

10. 10
News says:

keith s at 9: Like the wolf, actual design is one of the possibilities that must be considered when analyzing specified complexity in nature. The same way as it must be considered when analyzing fine tuning in the universe.

Of course, one can just decide it can’t be true and that any other least-implausible hypothesis is better.

11. 11
Mung says:

keiths can defend all his claims.