There's a bit of linguistic "inside baseball" in this post and this one, and some of those that they reference. For those of you who think that the issues seem interesting, but are puzzled by some of the discussion, here are a few pointers.
The general background includes the logical interpretation of words like "if", "and" and "or" in terms of truth tables -- e.g. "A and B" is true if both A and B are true, and false otherwise -- and the various apparently non-logical intepretations of such connectives, such as the interpretation of "and" in terms of time sequence ("(first) A and (then) B") or causal consequence ("you touch me again and you're toast"). In an influential essay entitled "Logic and Conversation", the philosopher H.P. Grice proposed a general theory about the relationship between meaning as something sentences have ("sentence meaning") and meaning as something that people do ("speaker meaning"), and he used some examples of extended senses of such logical words as cases in point. He argued that the observed uses could be explained as inferences from the context, the logical meanings of the words involved, and some very general considerations about the nature of communication, in a process he called "conversational implicature."
The question at issue in the discussion between Allan Hazlett and me (and indirectly Geoff Pullum) is what to make of a particular use of "if" in the New York Times. Allan (and indirectly Geoff) suggested that it was really a sort of stealth version of "and". I connected it to a sense for "if" recorded in the OED, as well as some other classes of examples, and suggested that maybe all of them could be analyzed in the style of Grice. Or maybe I should say, "waved my hand in the direction of a suggestion that ..." Aside from any interest that these particular examples may have, there are thus some large general issues looming vaguely in the background.
Posted by Mark Liberman at February 2, 2004 11:53 AM