6 February 2009 A Bleg
(A bleg is a beg performed by a blogger, often in terms of asking for information.)
I can’t help noticing a specific pattern of logic that pops up frequently. I think there must be a name for it, but I can’t find it. Hence the bleg.
Here are three examples: A wife suspects her husband of cheating. Or, I suspect Genbo has been raiding the ice cream while I was giving Zoe a bath. We can ask the suspect if a sin has been committed, but only one of two possible answers will actually contain information. That is, the husband will say “No” whether he cheated or not, so this response essentially conveys no information. Only an answer of “Yes” actually resolves uncertainty. Same thing, of course with Genbo and the ice-cream.
Similarly, assume there is a biomarker for a cancerous tumor. If you have this thing in your blood stream, you definitely have the tumor, as the molecule is created only by the tumor. However, it is possible to have the cancer without the biomarker being detected. So, no information is gained by failure to detect the biomarker. Only through detection is information gained.
Finally, assume that on a given train line there are two colors of trains. The green train could be either local or express, while the red train is always the express. So, if you see a green train coming, it could be either the local or the express. Only sighting the red one resolves information about the coming train.
This pattern pops up often, and I would love to know if there is a name for it. Asymmetrical information would be a great name for it, but that’s already taken by something totally unrelated (as well as by one of my favorite blogs).
Anyone got a clue?
1 comment in “A Bleg”
February 6th, 2009 at 4:31 pm
Not sure if there’s a definite term for what you’ve described, but something similar would be “falsifiability“.
Basically, if something is falsifiable, you can prove it false with an observation. If you can’t prove something false with an observation, then it would be unfalsifiable (well, at least as far as mere observation goes).
Of course, falsifiability does not indicate if something actually is false, just whether it can be shown to be.