Binding
This morning I realized that I've been using the term ‘bind’ inconsistently when describing the association of an identifier and a value: Usually I have written of “binding the identifier to the value,” but about a third of the time I've swapped the direct object with the object of the preposition, “binding the value to the identifier.” I'm obsessive enough that this really bothers me -- binding isn't a symmetrical relation, so it makes no sense to waffle back and forth between these two ways of construing it.
After some thought, I concluded that the nominalization ‘parameter binding’ and the frequency and importance of the technical term ‘bound variable’ strongly favor the former construction, so I'm going through and changing the passages in which I used the latter one instead.
I'm not quite sure why this issue bothers me so much, but the recent history of ‘substitute’ may have heightened my sensibilities. Technical writers now frequently use ‘substitute’ as if its syntax were the same as the syntax of ‘replace’, writing “substituting A with B” when they should be writing “substituting B for A.” It can be difficult for a reader to be sure about whether A or B ends up in the result of the substitution. I don't want something similar to happen to ‘bind’.