Sim Aberson wrote me a little while back about a reference on NPR to someone as a retired widow. The clear intention was to refer to this person as someone who was both retired and a widow, as the truth-functional equivalent of widowed retiree. But it’s hard to read retired widow that way; instead, you’re likely to get the interpretation ‘someone who has retired from being a widow’, which is puzzling.
It’s a small point, but the question is: why?
In a very simple world, when an adjective X combines with a noun Y, the semantics of the composite can be computed straightforwardly from the semantics of the parts:
the composite is subsective (or hyponymic): an X Y is a kind of Y; that is, the denotation of X Y is a subset of the denotation of Y; and
the composite is intersective: an X Y has the property denoted by X; that is, the denotation of X Y is also a subset of the denotation of X.
What I’ve presented here is surely the linguistically naive person’s idea about the semantics of adjectival modification (and it has a long tradition in writings about language).
But in the real world, there’s a host of other types of adjectival modification, some of them subtly off this paradigm, others quite strikingly so. On the subtle side, there are cases where X is understood not absolutely, but relatively to Y: a big mouse is not necessarily big, merely big as mice go; a famous cynic is not necessarily famous, merely well-known for cynicism; a wonderful dancer is not necessarily wonderful, merely very talented at dancing. Once you start collecting such cases, you’re likely to conclude that examples like blue circle are not all that common.
Then there are cases where X isn’t restrictive at all, but appositive, as in the pilotless drones ‘drones, which are pilotless’ example discussed here. And cases where the composite isn’t subsective at all, but “resembloid“; an African violet isn’t a violet, though it resembles one. And cases where X is restrictive, but not straightforwardly intersective, as in Canadian border, where Canadian does indeed restrict the interpretation of border, but not by requiring that the border be Canadian (whatever that might mean), but by evoking some referent — Canada, in this case — that can be used to restrict the domain.
(There are other wonders of adjectival modification, and much literature on them. But this will do for now.)
The Canadian border example begins to bring us back to retired widow. Even restricting yourself to borders between geographical or political territories, you can imagine a number of ways in which a border could be related to Canada: it could be a border established by Canada between two other territories, it could be a border established by principles enunciated by Canada, it could be the border of Canada (with some other country) as viewed from within Canada, it could be the border of Canada (with some other country) as viewed from within the other country, and no doubt clever people could imagine more possibilities.
But the fact is that, without considerable contextual support, Z border (where Z is an adjective based on the geographical noun X) is almost always going to be understood as ‘border with X’ (that is, viewed from outside X), which makes things like
We were traveling in Paraguay when the revolution began, and we rushed to the Canadian border.
so baffling. It’s not ungrammatical, but it’s very hard to interpret.
That is, Z border ‘border with X’ is a small-scale idiomatic pattern, something you can get “off the shelf” without having to work out speaker intentions, the contribution of context, and the like. It’s conventionalized, and is particular to the head N border.
Now we’re back to retired widow. This, like Canadian border, is an instance of a conventional pattern, but in this case it’s the second element, rather than the first, that’s the open slot: retired Y ‘someone who has retired from the occupation Y’, as in retired nurse, retired steelworker, retired professor, and so on. Not subsective, and, though restrictive, not straightforwardly intersective.
February 3, 2009 at 7:52 am |
Are you sure that what makes “retired widow” seem odd is grammatical, and not the result of an ingrained cultural bias? We tend to stereotype widowhood as being an alternative to employment, so that it attaches to the adjective “retired” that way, and then the problem comes because widowhood, unlike employment, is not a changeable state. It may sound like I’m reaching, but would you have reacted the same to the phrase “retired widower“? I’m not sure I would.
February 3, 2009 at 11:37 am |
The phrase doesn’t bother me either. “Widowed retiree” raises a temporal question that’s at least as ambiguous, since it suggests that the widowhood occurred after the retirement. The only fully unambiguous description is one that gives all the facts in the correct order, which may take too much time and not be germane to the story: “Ms. Smith, who retired after 47 years as a focus puller, was recently widowed when her husband Philander died in a bottling plant accident.”
In an earlier, misguided phase of my life, I worked as an attorney. My state requires attendance at Continuing Legal Education courses, including a certain number of hours discussing substance abuse and ethical issues. I once attended a lecture by a person who described herself as a “recovering lawyer.” It took me a minute to realize that she meant that she was a lawyer who was a recovering alcoholic. Although I’m not an alcoholic, I’ve occasionally used the phrase “recovering lawyer” to describe myself, and I’ve heard other former practioners use it too.
February 6, 2009 at 7:28 pm |
And then there are composites that are wholesale derived from another composite, like “theoretical physicist” or “computational linguist”. Back in my physics times I once had acquaintances from outside the science circles argue that “applied physicist” must surely be a joke and improper in serious writing.
@hjælmer: “Recovering lawyer” would have been near impenetrable to me
(Note on BrE vs AmE: I’d expect “pensioner” here in the place of “retiree”.)
February 12, 2009 at 11:49 pm |
The phrases “retired woman”, “retired Canadian” or “retired father of three”, in all of which Y is a role similarly difficult if at all to relinquish, don’t create, for me, the same momentary unclarity.
In some phrases, like “retired pigeon fancier” or “retired curmudgeon” my instant reaction is to take Y as an occupation or activity retired into rather than retired from.