Poet John Ashbery, intervewed by Jillian Tarnald in the NYT Book Review of May 10th:
Are there poets whom you’ve gained greater appreciation for over time?
Certainly Whitman, whose barbaric yawp didn’t impress me at first, but whose silken language did as I began to live with it. And French poets, of whom I published a volume of translations last year. Had I not received a Fulbright and gone to live there for some years, I wouldn’t have been able to appreciate them.
The anaphoric element there is understood here as ‘in France’, but the word France isn’t in the context; instead, we have the word French, which merely evokes France. So that, speaking somewhat metaphorically, the anaphora “goes into” the word French.
The topic is a venerable one, going back 40 years or so, under the name of the Anaphoric Istand Constraint (AIC).
In a Language Log posting on 10/20/07, I summarized some of the history, in which it was suggested that
lexical items are “islands” for anaphora, that parts of lexical items or referents merely evoked by lexical items cannot serve as antecedents for anaphoric elements (of several different kinds).
I gave the invented examples:
ONE: I’m a pianist, but I don’t own one. [one ‘a piano’]
PRO: Flautists can easily take them on planes. [them ‘flutes’]
THERE: I speak Norwegian, but I’ve never been there. [there ‘to Norway’]
The AIC asserts that all three of these are ungrammatical (though comprehensible).
The literature then takes up apparent exceptions and provisos to the AIC.
Later literature takes a quite different view of these phenomena. As I said in a 7/21/12 posting on this blog, in a 1991 article in Language by Ward, Sproat & McKoon, it’s argued that
anaphoric islands aren’t a syntactic/semantic phenomenon at all, but a pragmatic one, with the acceptability of particular examples having to do with processing ease considerations linked to context, discourse organization, and the like.
I’ve posted several times on this blog about particular examples, noting that when AIC violations are viewed in context (rather than as single sentences ripped from their discourses) they usually become innocuous; at the worst, they might require a bit of extra processing time, but most examples (like the Ashbery example above) aren’t even noticed by readers or listeners.
I have a modest collection of AIC violations in real life (one nowhere as big as collections by Gregory Ward and others, in particular Andy Kehler), with just one THERE example:
As my husband is Singaporean, and I’ve been going there for 16 years on-and-off, …
This is easier to process than my Norwegian example above, even out of context, because Singapore is almost entirely untouched within Singaporean. (Contrast Norwegian / Norway and French / France.)
Leave a Reply