Let me emerge briefly from my holiday absence (I found myself near a
computer and couldn't resist having a look at Cybalist).
There are two theories (rather than none at all) about the origin of
<she>. It must have developed either from the OE personal pronoun
<he:o> or the feminine demonstrative <se:o ~ si:o>, with the initial
consonant palatalised in a manner that is slightly irregular but by
no means impossible, as a falling diphthong changed into a rising one
([jo:]).
There is a lot of dialectal variation in Middle English with respect
to "she". The Southern and West Midland dialects have conservative h-
initial forms like <heo, ho, he> [hø: ~ he:], weak <ha> [h&] or
characteristically Kentish <hi> [hi:] (the regular development of
<hi:o>, the Old Kentish variant of the pronoun). Northern dialects
have really funny things like <scho, cho> or <3ho> (with the so-
called "yogh" letter). The last form, found in the Ormulum, is often
considered to represent the "missing link" between OE [he:&] and
Northern ME [So:], i.e. [hjo:] or [Co:].
The direct forerunner of the modern standard form, <sche, she> [Se:]
is found in the East Midland dialect. I don't think <s(c)he> is a
development of <scho>. Rather than that, <he:o> or <se:o> changed
into [hjø:] or [sjø:] at an early date in that area, with the
subsequent unrounding of the vowel, the normal treatment in the East
Midlands and the North.
Piotr
--- In cybalist@..., "Glen Gordon" <glengordon01@...> wrote:
>
> Mark O:
> >I recall from something David Crystal wrote. Essentially, he said
no
> >completely convincing etymology for 'she' can be presented.
>
> Ah, and this comes back to whether it is more logical to accept
> the simplest theory to answer a question, or whether to throw
> one's hands up in surrender and accept nothing at all. The latter
> isn't striving to answer the question and is therefore illogical.
> Clearly, "she" is not a loanword from some extinct language unless
> we wish to make one up in our head.
>
> As far as my little brain understands, /heo/ > /sho/ > /she/ makes
> a hell of load of sense. The /-eo/ in /heo/ was pronounced [e&],
> yes? So, [he&] > [hyo] > [Co] (C = IPA's "c cedille"). Simple
> palatalisation.
>
> Eventually, /sho/ becomes the norm. However, by analogy with /he/,
> /sho/ is revised to /she/. What on earth is wrong with this view,
> I'd like to know?