> 1. <ce> = [ke] if the <e> is due to i-umlaut of a back vowel.
> 2. <ce> = [ke] within a word, if preceeded by a back vowel.
> 3. <ce> = [ke] if the <e> arose before a continuant that was
> originally syllabic (e.g. æcer "acre" < Proto OE *ækr).
>
> 4. Otherwise <ce> did result in the affricate [tS]. Before a
> stressed syllable this became <cie> in Early West Saxon. In late

WS

> the same combination is often spelt <cy> (but also <ce>, <cie> and
> <ci>).

Still bearing in mind all those provisos, I'd just like to modify 4,
thus:

4. Otherwise <ce> did result in the affricate [tS]. Where the <e>
is from Proto OE, before a stressed syllable this became <cie> in
Early West Saxon. (In late WS the same combination is often spelt
<cy> (but also <ce>, <cie> and <ci>).) This still leaves the
combinations <cea> from PrOE *kæ, and <céa> (=Anglian <cé>) from
PrOE *kæu < Gmc. *kau, each of which developed the pronunciation
[tS].

Llama Nom

>
> Okay, these are the rules for when <ce> = [tSe] in OE (not

counting

> the combination <sc>), according to Campbell's Old English

Grammar.

> At least this is how I understand them... They work most of the
> time, but not always. Some exceptions are due to analogy with

parts

> of the paradigm where different rules applied. Another exception

is

> the middle consonant of <cieken> "chicken", as indicated by the
> spelling in the Mercian Rushworth Gospells, a unique scibal

attempt

> at distinguishing the front and back pronunciations of <c>. This
> can't be due to analogy with other parts of the paradigm, but

might

> be accounted for by dissimilation. Anyway, here are the rules.
> (For <ce>, read <ce> or <cce>.)
>
> 1. <ce> = [ke] if the <e> is due to i-umlaut of a back vowel.
> 2. <ce> = [ke] within a word, if preceeded by a back vowel.
> 3. <ce> = [ke] if the <e> arose before a continuant that was
> originally syllabic (e.g. æcer "acre" < Proto OE *ækr).
>
> 4. Otherwise <ce> did result in the affricate [tS]. Before a
> stressed syllable this became <cie> in Early West Saxon. In late

WS

> the same combination is often spelt <cy> (but also <ce>, <cie> and
> <ci>).
>
> Some OE textbooks print a dot above <c> in words like 'bæc'
> and 'æcer'. This indicates that they are thought to have been
> palatal stops in early OE, something like [c], as in
> Icelandic 'kenna'. According to Campbell, in such positions, <c>
> never evolved into [tS], but instead reverted to [k]. Regarding
> Rule One, don't forget that /æ/ counts as a front vowel, so if <e>
> is due to i-umlaut of /æ/, this would result in affrication.