Double modal

Laurence Horn laurence.horn at YALE.EDU
Sat Dec 30 01:03:15 UTC 2006

At 5:37 PM -0500 12/28/06, Paul Johnston wrote:
>An interesting thing about double modals:  they apparently violate
>certain "inviolable constraints" in GB grammar (I don't know about
>even more modern Chomskyan models), since you're not supposed to be
>able to (or, to can) negate on the second modal, and whether it's
>Tennessee, Tyrone (N Ire) or Tranent (Scotland) that's EXACTLY what
>I might could do that for you  >>  I might couldn't do that for you
>A might cuid dae that for ye >>  A might cuidnae dae that for ye
>My ex-mother-in-law, from Southern  Lanarkshire, had a million of
>these so-called "performance errors", which were, of course, totally
>systematic.  I'm sure there are many more throughout our South as well.

I don't understand what constraints this would violate, especially if
the first (epistemic) modal can be analyzed as an adverb (=
"Perhaps/Maybe I couldn't do that for you").*  The negation on the
modal would indicate that that's the highest predicate.  I'm also
unclear on who, exactly, would (so-)call these performance errors, if
they're systematic.  I'm no Chomskyan, nor do I play one on TV, but
this has always been precisely the point of drawing the
competence/performance distinction, as I understand it:  If it's part
of one's grammar, it's not a performance error.  Am I missing


*At least in the dialects I'm familiar with, the first modal in the
sequence is indeed an external, epistemic one (= it MODAL be
that...), and when it's "may/might" it can be paraphrased as
"perhaps" or "maybe".

The American Dialect Society -

More information about the Ads-l mailing list