cf NESL (1993) http://www.cs.cmu.edu/~scandal/nesl.html

I personally think mathematical analogies (in particular, strong and [Knuthian] weak components) are more productive ways to think about these issues instead of attempting to drag in physical analogies, but YMMV.