Please read the "Epilogue" to Judea Pearl's seminal work, Causality. We've included a PDF here, but the entire book is available for you access on the library VPN at this link: https://doi-org.libproxy.berkeley.edu/10.1017/CBO9780511803161.
As you're reading this epilogue, consider:
- When you think about one thing in your life causing another, what do you mean? Pearl asks if we are satisfied to be as simple an empericist as Hume, who brushes aside questions of one thing causing another:
"Thus we remember to have seen that species of object we call flame, and to have felt that species of sensation we call heat. We likewise call to mind their constant conjunction in all past instances. Without any farther ceremony, we call the one cause and the other effect, and infer the existence of the one from that of the other.” - David Hume.
Thus, causal connections according to Hume are the product of observations. Causation is a learnable habit of the mind, almost as fictional as optical illusions and as transitory as Pavlov’s conditioning.
- Pearl asks for a deeper understanding of one event causing another. Can you survive your world with only the learned association between flame and heat? Or, do you need to know more?
- Michael Streavens argues in his recent book The Knowledge Machine that the thing that has made science, well, science is what he terms "The Iron Rule of Explanation". When there is a disagreement in theory, we produce evidence that would help us evaluate a theory. Does such an empirical view of science leave any room for questions of causality?
For those who are interested, you could (optionally!) read the *first two sections from Chapter 5 to further fill out the problems that are endemic to the cavalier use of regression models (termed SEMs in the reading) to describe causal relationships.
- Section 5.1.1: Causality in Search of a Language
- Section 5.1.2: SEM: How Its Meaning Became Obscured