2009/6/27 Garth Wallace <[log in to unmask]>

> I think it's that they're blends of morphemes derived from full word
> forms that are still in use on their own and are recognized as still
> having a connection to the "original". For example, the "-n't" suffix
> to modal verbs in English is still seen as a shortened form of "not",
> and it's believed that people using e.g. "can't" are taking "can" and
> "not" and squishing them together (even though it's actually most
> likely that it was learned like any other inflection). If "not" fell
> out of use for some reason, though, leaving only the suffix, it'd
> probably cease to be thought of as a contraction, because there would
> be nothing to contract.

I was going to say the exact same thing, but Garth beat me to it. This in
part constitutes an answer to the "what makes the Spanish future tense
inflections and not contractions" question—simply put, there is no expanded
"infinitive + haber" form with which to contrast the would-be contractions.

Probably there rests a point of difference between Garth's and my opinions,
though, which is that IMO "not" doesn't have to fall out of use as a word
for "don't" to cease being seen as a contraction and instead come to be
viewed as a new word. Rather, speakers just have to stop thinking about "do
not" as equivalent to "don't". That may come about from semantic shift, or
whatever other myriad factors there may be.

(I can't remember the exact example offhand, but IIRC there was one such
word that branched off into two exactly identical forms but each of which
adopted different meanings and came to be spelt differently and seen as
homophones despite their common heritage. Not very helpful example here,
apologies. But if anyone recognises this example and can enlighten the