Was there a /t͡ʃ/ to /k/ sound change from Old English?

Question

I stumbled upon a strange thing while looking up the etymology of words ending in "le". I looked up "kettle" and saw that it was pronounced with /t͡ʃ/ in Old English and also in Middle English.

An entry from Wikitionary:

From Middle English ketel, also chetel, from Old Norse ketill and Old English ċietel

The pronunciation of Old English ċietel is /ˈt͡ʃi͜y.tel/ where /t͡ʃ/ can be seen.
In Middle English, wikitionary gives three pronunciations: /ˈkɛtəl/, /ˈtʃɛtəl/, /ˈtʃitəl/. One of the pronunciations has /k/, the other two have /tʃ/.

I searched many other words starting with "k" in Modern English, that are from Old English but none of them show the same change. Also, I can't find anything on Google.

Can someone explain why it happened? Was it a one-time change or it applied to other words too?

Answer 1

In general, there wasn't a /t͡ʃ/ to /k/ sound change. It's more likely in cases like this that /k/ represents a form where for some reason, the conditioned /k/ to /t͡ʃ/ sound change that is part of the history of English did not take place.

In native English words, /t͡ʃ/ almost always comes from palatalization of Proto-Germanic *k. (The verb fetch is a rare exception where /t͡ʃ/ may be from palatalized *t.) The palatalization of Proto-Germanic *k usually occurred in certain contexts, such as before the palatal glide /j/ or next to a front vowel.

The change of the reconstructed velar stop [k] to the present day palato-alveolar affricate [t͡ʃ] is thought to have proceeded through the intermediate step of a fronted or palatal stop, written [c] in IPA: [k] > [c] > [t͡ʃ]. Some sources argue that in Old English, the palatalized version of *k was still pronounced as the palatal stop [c], and had not yet progressed to the affricate [t͡ʃ].

Palatalization did not occur in Old Norse, so the absence of palatalization in contexts where it would be expected is often a sign that a word was borrowed from Norse. This may be why Wiktionary mentions Old Norse ketill.

It may also be possible that different dialects of Old English had different forms of the sound change palatalizing or affricating Proto-Germanic *k – although this is debated, and I'm not sure if there could be any difference in the case of a word like ċietel. The specific form ċietel belongs to the West Saxon dialect. The Oxford English Dictionary's entry for kettle traces the form to a "Mercian and Northumbrian form" cętel starting with [k], also acknowledging that some sources attribute the [k] in kettle to Scandinavian influence. However, this is an unupdated older entry, and I'm not sure that it reflects modern scholarship. According to "Palatalization of Velars:A Major Link of Old English and Old Frisian", by Stephen Laker, all dialects of Old English, including northern dialects such as Mercian and Northumbrian, should have palatalized and assibilated *k before a front vowel (page 181).

Answer 2

Palatalisation is the technical term for what caused the change from /k/ to /t͡ʃ/.

You might have noticed that the vowel in the word keep (/i/) is produced further forward in the mouth than the vowel in cool (/u/). The KEEP vowel is therefore called a ‘front vowel’ and the COOL vowel is called a ‘back vowel’.

The /k/ in cool is articulated in the back of the mouth, meaning the closure is made further back in the mouth whereas the /k/ in keep is articulated further forward in the mouth in anticipation of the following front vowel—/i/. Because of the following front vowel /i/ in keep, the closure for /k/ is made further forward in the mouth (closer to the palate) in order to ease the transition.

In Proto-Germanic (ancestors of English), most of the /t͡ʃ/-words were pronounced with a /k/, for instance, child, chin, cheese etc., were pronounced with /k/ (they're still pronounced with /k/ in German). The word church was *kirika in Proto-Germanic (compare Scot. kirk, from Old Norse which retained the original /k/). [Trask Historical Linguistics]

So what happened?

Over time this palatalisation of /k/ before a following front vowel went so far that the /k/ became [tʃ]. That is to say, the closure for the /k/ moved to the front part of the mouth, resulting in an affricate [tʃ]. [Trask Historical Linguistics]

An example of 'palatalisation' in Modern English is the pronunciation of what you as whatcha: the /t/ in what is alveolar while the /j/ in you is palatal, the /t/ in this case is articulated near the palatal in anticipation of the following /j/, resulting in an affricate [tʃ].

According to Peter S. Baker (Introduction to Old English), Old English c (which is sometimes printed with a dot: ċ) was pronounced /t͡ʃ/ before the front vowels i and ie and the diphthongs ea and eo.

I don't know what happened to kettle, but here's what Donka Minkova has to say:

The PDE vocabulary shows many such pairs. The presence or absence of palatalisation of an etymological [k] to [tʃ] is responsible for the different initial consonants in cold-chill, kettle-chettel (dial.). Alternative pronunciations in a derivational set or in OE and ON are also behind the [k] ~ [tʃ] alternation in the histories of bench-bank, birch-birk, chest-kist [...]

[A Historical Phonology of English by Donka Minkova]