Dick:
> >But I think it is, given that
> >* Orthography reflects phonology, not syntax. Purely syntactic orthographic
> >rules are a bugger to learn (cf. apostrophes)
> ## English orthography reflects both phonology *and* morphology/lexemes, so
> we tend strongly to reflect lexemic bonding in spelling

Yes -- what I wrote sounds blatantly false (when said about English
spelling, of all things!). But what I meant is that if a sentence
consists of a phonological structure and semantic/syntactic structure,
with "words" being correspondences between portions of the two
structures, then orthography reflects the phonological structure.

Reflecting inaudibilia orthographically is quite difficult for most
people, and (heterographous) homophones are harder to keep apart
the less concrete their meanings are.

> >* HAVE itself is rather flimsily held together as a lexeme by the
> >shared phonology of its varieties
> ## Yes, but the point is that the "have" in (1) is definitely an instance
> of the same lexeme as the one in (2)
> (1)     I should have done it
> (2)     I have done it
> If not, you're left with an odd restriction on which non-modal verbs have
> an infinitive: all but HAVE/perfect, which happens to share its exact
> valency with a verb that exists in nothing but the infinitive form:
> OF/perfect. This valency similarity is quite striking, even down to the
> shared peculiarity with BE/go:
> (3)     I have been to Paris
> (4)     I may of been to Paris
> It's simply perverse (in my view) not to merge these two forms into a
> single lexeme, especially given that OF/perfect is phonologically identical
> to a common variant form of finite HAVE which, so far as I know, is never
> written OF:
> (5)     Have you finished?  (*Of you finished?)
> (6)     The students have (*of) finished
> So if OF/perfect belongs to the lexeme HAVE/perfect, why not celebrate this
> fact in the spelling?

I take it that "OF/perfect" means "a variety of HAVE/perfect that has
the phonology of OF"?

Why not celebrate it in the spelling?
(a) Because basically orthography is built on phonology, and deviations
from that are stressful and hard to learn, especially when the deviations
are based on distinctions that are rather deeply embedded in implicit
knowledge.
(b) Perhaps also because finite _have_ does not so readily reduce to
[log in to unmask]
        Who's done it?
    * *I* /@/

    ? The lads /@/ been here for hours.
    ? /@/ they been here yet?
    ? /@/ you sold it yet?

> >I see no need for the extra step that identifies the word as OF. I'd
> >say it was aux HAVE that has the phonology of OF, where aux HAVE is
> >defined by its meaning and its combinatorial properties
> ## Yes, just as I argue above. But to me it makes no more sense to write
> "must of been" than to write "pound have apples"

It makes no sense because English spelling is not phonemic, but that
is very difficult for people to handle. Our 'natural impulse' is for
purely phonemic script, IMO. (I mean 'phonographic' in general, not
phonemic in the strict sense; I don't believe that the phoneme itself
is intuitively natural.)

> To make matters even harder to understand, I assume that these misspellings
> are independent - not learned from the sub-culture. This is a matter of
> fact, of course, so we need evidence; but I'd expect the evidence to show
> small kids writing "must of" without knowing that other kids are doing the
> same. Of course they may be reinforced in later life by seeing others doing
> the same, but that doesn't explain why learners do it

All deviations from phonemic spelling have to be learnt and at a relatively
explicit level until internalized through habit.

--And.