Re[6]: [tied] Surname 'Knyvett' etc. and OFr 'c(a)nivet'

At 3:39:23 PM on Friday, October 10, 2008, Arnaud Fournet
wrote:

> From: "Brian M. Scott" <BMScott@...>

>>> I was suggesting that these writings probably reflect
>>> the disapperance of -h-, There is no reason to suppose
>>> that -h- is represented by either -c- or -s- or -i-.
>>> Cnict is *knit the same way French dict stands for *dit
>>> with no real -c-. Cnist is *knit the same way French
>>> beste stands for *bete with no real -s-. etc. Therefore
>>> relevant.

>> But completely wrong, since the word in question isn't
>> French.

> No,
> English spelling owes much to French writing.

Of course I know that. But in this case I'm talking about
forms from a period within a few decades after the Conquest,
which are very different from the later ME forms.

> out with Middle English ou = French ou is an obvious
> example.

But not particularly relevant, since I'm talking about early
ME, when spellings like <vt>, <ut>, <wt>, <hut>, etc. are
still common (as in the byname <le Vtlage> 1230 'the
outlaw', which is considerably later than the earliest
<Cnivet> example).

[...]

>>> These are just awkward spellings of *knit with no -h-.

>> No, they're attempts to represent a sound that wasn't
>> present in OFr at the time and therefore had no standard
>> representation. The word being recorded was late OE and
>> early ME /knixt/; /k/ and /s/ were two of the closer
>> possible approximations.

> No, I disagree.
> There was no -h-.

Loss of /x/ in this position didn't begin until the 14th
century and wasn't complete in the standard language until
the Early Modern period.

[...]

> You have not given any dates for these forms.

Because anyone with the knowledge to answer my questions
would already recognize them as characteristic of early
forms, though they can also be found later.

> They must be rather late, because they reflect a rather
> advanced phonetic stage for both French and English.

On the contrary, it's easy to find such types shortly after
the Conquest. In particular, <Chenicte> and <Chenist[us]>,
both representing <Cniht>, appear in 1086 in Domesday Book.

However, I've just been reminded that the <st> spellings may
actually represent /xt/, not /st/, as I suggested earlier:
Pope (§1178) notes that in the combination /st/, OFr /s/
developed into /ç/ and /x/ in the 11th or 12th century
before being effaced. The 13th century 'Orthographia
Gallica' makes this pretty explicit: 'Et quant _s_ est joynt
a la _t_ ele avera le soun de _h_, come _est, plest_ serront
soner _eght, pleght_'. In Anglo-Norman the change persisted
into the late 13th or 14th century, as evidenced by
spellings with <h> for expected <s>.)

Brian