My personal opinion on the complex specified information as discussed in “ Intelligent design as a theory of information”, by William A. Dembski

 

 

In this paper Dembski intends to clarify the distinction between his concepts of specified and unspecified information. While referring to his book “The Design Inference” for the details of his approach, he nevertheless intuitively explains  the difference between specified and unspecified information by an example that involves a large blank wall, an archer with his bow, an undefined number of arrows and some painting tool to design figures on the wall.

The archer stands 50 meters from the wall.

 

 

 

Dembski considers 3 possible scenarios.

 

Scenario 1:  the archer simply shoots at the wall.

Scenario 2: the archer first paints a target on the wall, and then shoots at the wall, squarely hitting the target.

Scenario 3: the archer shoots at the wall. And as in the first scenario, the archer shoots at the wall while it is still blank. After having shot the arrow, the archer paints a target around the arrow so that the arrow sticks in the target. In the third case, the final situation is the same of scenario 1, but, underlines Dembski, we have no information about the archer ability.

 

Dembski notices how we do not learn anything about the archer's ability in the first scenario.

In the second scenario we have evidence of the archer's skill. Nevertheless, he states, that it would be wrong to say that specified information has been actualized.

I suggest the same experiment, but let the archer launch many arrows in each case (say 100 of them). Moreover, rather than an undefined blank wall, we know its surface (S)

The scenarios are the same, but the results are:

1)     Scenario 1: almost all arrows are within a surface S’ of the wall.

2)     Scenario 2: almost all arrows are within the target (it has a surface S’, the same of scenario 2.

3)     Scenario 3: almost all arrows are within a surface S’ of the wall.

 

So we have exactly the same physical results in the three scenarios.

Our reference case is an archer launching randomly 100 arrows against a wall of surface S.

As the final result, in all three scenarios is that 100 arrows stroke the wall of surface S within a surface S’. The probability of such configuration is P(100,S’,S). Its distance from a random distribution of arrows arrivals on S can be reasonably estimated by the negative logarithm of P(100,S’,S). So the final result is the same.

The information we get is in scenarios 1 and 2 that of a low probability pattern being actualised.

Only case 2 can be translated into the competence of the archer, but this depend on the narration.

Assume you do not know the story, and see the arrows distribution after the archer has left. The information you get is log₂(S’/S). Of course if S’@S, the information you get is almost zero, even if the archer had painted a target as large as the wall (God knows why!).

The opposite alternative would be to observe the archer during his performance, by getting the whole information on the story. Then you would have asked the archer while he was painting the target after.