headless compound story

Several months ago I posted a brief anecdote about a shrimp salad ... briefly, it was vaguely humourous in that the person who ordered it wanted no shrimp. In a hurry to post it at the time, I gave it the title "headless compound or what" without thinking about it.

Fast forward til last week, when glancing over my site meter stats, I noticed a surge in referrals from Google searches on "headless compounds" (that is, if 3 can be considered a surge, which I think it is in the modest readership of an off-piste linguablog like this one).

It gave me the brief hunch that a student somewhere needed an example of a headless compound for a paper; and if so, I hope they didn't use the notion of a shrimp salad with no shrimp (or at least I hope they didn't cite me). Here's why: it's not a headless compound. Headless compounds are compound structures in which the right element is not really the semantic head of the compound; at least, that's how they're discussed in Pinker's The Language Instinct.

Pinker invokes the notion of headlessness to explain the resistance to irregular pluralization in compounds like low-lifes, Maple Leafs, tenderfoots, and sabretooths. The argument is that a low-life is not a kind of life, and a tenderfoot is not a kind of foot, and so on. Since the basic meaning of each of these is hidden, so is its access to irregular plural morphology (as in lives, leaves, feet, teeth).

Under this understanding of the term, shrimp salad would be a headless compound if it referred to something like the plate one serves shrimp salad on, or if it referred to anything else, as long as it's not a kind of salad. (Wait - did just say plate of shrimp?)

Ironically, not only is shrimp salad not headless, I don't even think it's a compound - in the original anecdote, and in my mind since, it has primary stress on salad, when compounds typically have stress on their first element.

Now if only headless compound were not a kind of compound...