Over at the Modeled Behavior blog, Karl Smith has written a lovely little post on causation.
As he puts it
Causation is a subtle concept. Experiments don't prove causation (whatever that is) except in the simplest case. We are far too confident that we know what we are talking about.
As he puts it
Suppose you had figured out actual causation. How would you know?I was reminded of a (Five year!) old post of mine, quoting a story I read about causation in Dennett's book "Freedom evolves".
Causation is a subtle concept. Experiments don't prove causation (whatever that is) except in the simplest case. We are far too confident that we know what we are talking about.
No comments:
Post a Comment