On Tuesday 26 November 2002 08:34, Dick Hudson wrote:
> Dylan:
> >A) I could of done it
>
> ## I've been thinking about this pattern for decades now. The more I think
> about it, the odder it seems.
>
> >My position is basically that the phonological equivalence between the
>
> reduced
>
> >forms of *of* and *have*, together with the (semi-)productive conversion
>
> of P
>
> >to V in English facilitates (leads to, even) the reanalysis of (B) to (C):
> >
> >B) AUX've
> >C) AUX of
>
> ## But phonological equivalence doesn't push us to respell other words -
> e.g. "an" and "and" are identical in phonology but nobody is tempted to
> write "John an Mary". And why is it always "have" that changes to "of", not
> the other way round? And anyway, aren't these people aware of the
> morpho-syntactic pattern. Writing a reduced form of "have" as "of" strikes
> me as the ultimate in perversity.

Sorry, but that strikes me as proscriptive dogma. The development is clearly 
real - I googled for it last night and got 170,000 hits. About 50% were 
genuine (the others were usage guides and 'non-constituent co-locations' like 
"... could, of course, ...")

>
> >ISTR *to* as infinitive marker arose in similar manner. Can anyone give a
> >(brief) description?
>
> ## Now that's an interesting thought, isn't it? If infinitival "to" really
> is a non-finite auxiliary (as I claim), maybe its similarity to the
> preposition "to" acts as a model for non-finite auxiliary "have" and "of".
> I don't remember seeing that suggestion. Pretty tortuous but very
> interesting if true. How can we ever know the truth in this kind of case?
>
> >How about any other arguments or supportive evidence? What age range are
> > the auxiliaries generally acquired by?
>
> ## About 3 I should think. But age of acquiring the spoken forms isn't
> relevant, is it? The reduced form of "have" in speech is easy to
> understand; what's hard is this spelling.

I think it is relevant in that if a child has acquired /@v/ as the relevant 
AUX form, and also that /@v/ + STRESSED > /Qv/ such that the spoken form is 
no longer /h&v/, then the orthographic convention becomes obscure. To some 
extent, _have_ may be on a par with _shall_ insofar as being acquired (or 
learnt, maybe) as a stylistic/registal/rhetorical variant. By the time the 
orthography makes it clear to the child that the 'underlying' form is /h&v/, 
it's too late, parental corrections notwithstanding. The point being that the 
child acquires an auxiliary /@v/ / /Qv/, which happens to coincide with OF. 
I'm NOT suggesting that the preposition (in this case) is converted to an 
aux, I AM suggesting that HAVE could dissimilate into HAVE and OF(aux-v) as a 
result of the phonological properties. This is independent of the existance 
of the preposition: it could happen anyway. The existance of OF(p) leads us 
to view it as an error rather than some other process of language change. 
Thus, it isn't that they don't know the morpho-syntactic pattern, but that 
they analyse a different one. To the extent that orthography is parasitic on 
phonology/phonetics, the spelling is an obscuring epi-phenomenon.

This line of thought (and, after all, that's all it is) is only valid if the 
auxiliary system is well established before reading and writing.

wrt an/and, I do see _an_ for _and_ in email (which is just about as close to 
spontaneous speech as writing gets, so far as I can see) but that could 
easily be dismissed as typos. Secondly, determiners and connectives are (are 
they not) more robust - at least for the present in English - than 
auxiliaries (c.f. wanna, gonna, ...)

Dylan