= 27d567bc2d48d9f40e9ccc20ea370b15)
Marius comments on Your Evolved Intuitions - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (104)
It is cheating to make predictions that are actually postdictions. To test hypotheses we have to look at questions for which we don't have information and then test those questions by gathering new information. A lot of the "evolutionary psychology" are really "Just So" stories without this kind of hypothesis testing.
A huge amount of science operates on postdictions. That's not 'cheating', it's just not as impressive as Einstein predicting gravitational lensing. Just as in other sciences, including evolutionary biology, evolutionary psychology uses multiple converging lines of evidence to weigh the probabilities of hypotheses (see, e.g., Buss, The Handbook of Evolutionary Psychology, pp. 457-465: PDF).
And in fact, evolutionary psychologists often do make novel predictions and then go out and make the observations that either confirm or disconfirm the hypothesis.
I suspect that those who say evolutionary psychology is nothing but 'just so' stories are not actually keeping up with the science but instead are operating under cached thoughts.
Ok, update my cache: you listed 7 bold postdictions; please list some bold predictions.
Its true that not all science is prospective, but you need to at least avoid looking at the data when making hypotheses you will then test against that data. Often we divide a data set to do this. Its much easier to do this with quantitative hypotheses than qualitative ones.
Here are some additional predictions from evoutionary psychologists.
The first two are excellent examples. Thanks!
knb provides one example here, though I can come back add others later.
Highly relevant is this post from Yudkowsky.
If Buller is to be believed that's another postdiction. The sort of studies I want to see would compare two groups who aren't known to be distinct. An example: Group A and B evolved in different environments and should display some difference in behavior. They have different rates of intron fragments G and H Now look at apparently-homogenous group X, who do not believe they have A or B ancestry. Compare X with intron G to X with intron H and find the predicted behavior difference.
Do we have studies at that level?
Wasn't there a talk that touched on this at the last Singularity Summit? John Tooby IIRC. I don't remember if it was during his talk, the panel discussion, maybe his Q&A... I'm fairly certain at some point he (or someone) named a couple of big predictions. I'll look this up tonight unless someone beats me to it.
Edit: Here is what I was remembering. He talks about his guide to generating good psychological hypotheses through awareness of evolution at 15:00, with a slide with the 5 step plan at 16:40. His primer on evolutionary psychology is here and has an interesting example of a prediction of cheater detection at the end.