I've recently been reading the bishop John Phillips' translation of the Psalms into Manx from 1610. Here we find some interesting features of early 17th century Manx, showing that many of the phonological changes characteristic of Manx hadn't yet happened Phillips' orthography derives largely from Early Modern English conventions with a few influences of Welsh but is also very innovative in his attempts to represent the numerous vowels that English is unequipped to convey.
Roughly speaking, 17th century Manx, essentially just Phillips, is often referred to as 'Early Manx', the 18th century language of the Bible and various other texts is 'Classical Manx', 19th century material is often 'Late Manx' and the language of the last generation of native speakers is 'Terminal Manx'. I'm generally comparing Phillips' Early Manx here with the later Classical Manx. All the pronunciations I give for Phillips' Manx are my own guesses and should be taken with some salt.
Initial clusters /kn gn tn/ hadn't yet shifted to /kr gr tr/ for Phillips but this change is ubiquitous for later writers. The same shift has occurred in all other Gaelic dialects except Munster where historical /n/ is preserved in clusters. The loss of a nasal consonant is compensated by nasalisation of the vowel which very rarely develops into a following nasal consonant as in cronk. There are some similar developments in Irish in Mayo and just over the sea in East Ulster where áit and oíche (<oidhche) sometimes have nasal vowels written áint oínche and cnoc appears as cnonc as in Manx. Perhaps this is an areal feature of Sruth na Maoile?
(I'm not sure which of /u/ or /i/ is appropriate for knuick/cruink/cnoic. The following palatal/'slender' consonant would raise /o/ > /u/ but fronting again to /i/ is common, I can't think of an example in any Gaelic variety where [uCʲ] and [iCʲ] aren't interchangeable as the Gaelic languages have a vertical vowel system. Also the situation with lenis and fortis /n/ vs /N/ isn't obvious in Manx so I am just using /n/ in these examples.)
On the other hand, Phillips uses <gn> in gniart, gniartoil, gniartaghy for CM niart /Gaelic neart /nʲart/ 'strength', not letting us forget the English origins of his spelling.
The change /sk ʃkʲ/ > /st ʃtʲ/ by assimilation is not present yet, compare Phillips' troske 'fasting' (troscadh) wisky 'water' (uisce), dẃisk 'wake v.' (dúsc, cf Scottish dùisg), with later ushtey, dooisht, trostey /uʃtʲә duːʃtʲ trostә/. This change only occurred in non-initial clusters, so where Phillips has skial, shiuskel 'story, gospel' (scéal, so-scéal) later authors had skeeal, sushtal /skʲiəl soʃtʲal/.
We find vocalisation of bilabial fricatives /f v ṽ fʲ vʲ ṽʲ/ to [w~u] in some words but the consonants are preserved more often than not. This change occurs when the sound is postvocalic or in coda position (ie when it follows a vowel or is at the end of a word/syllable). It seems from Phillips spelling that the broad/slender distinction was maintained between /f v ṽ ~ fʲ vʲ ṽʲ/, however this distinction was later lost in all labial consonants in Manx as in Scottish Gaelic and either replaced with segmental /j/ or just merged into the non-palatal 'broad' sound.
Without vocalisation
* The vowels in terywf, dyfnid and tréyfd are not clear to me.
** Without an <f>, Phillips also has iurin, niẃryin and tréyd are also found.
The diphthongs /iu/ and /eu/ were often in free variation and tended to smooth towards /uː/ in Late Manx. This might explain trooid with /uː/ with a development something like /tʲrʲeṽʲədʲ/ > /treu.ədʲ/ > /truːdʲ/. Or maybe tréyfd/trooid is derived from something else altogether!
These examples also show the effects that palatalisation was having on nearby sounds. The change /a o/ > /e/ and sometimes /i/ before historically palatalised labials is evident as early as Phillips and later words such as goimh /goṽʲ/ > gew /gẽu/ 'pain' or coimirce /komʲərʲkʲə/ > kemmyrk /kemərʲkʲ/ 'refuge'. The nasal quality of vowels is not indicated in the spelling but they are well attested in phonetic descriptions and the same trend of /ṽ ṽʲ/ nasalising nearby vowels happened in all other Gaelic dialects.
With vocalisation
Phillips agrees with later developments and consistently has shiu 'you pl.' /ʃiu/ from sibh /ʃivʲ/. I also notice that some loanwords are put through this change with interesting results. Perhaps this change first affected /f v/ and changed them into phonetically close [w] before dragging other sounds over.
One word that trips me up is kýei, kyêi, kýèi which Phillips uses to translate 'wild'. The graph <yei> often represents /әi/, /әːi/ (or /ɨəi/) along with various other central vowels and diphthongs/triphthongs but in this case a central diphthong is supported by the later spelling keoi(e) which is almost definitely /əːi/. The development of <agh, adh, ogh, iodh> in stressed intervocalic or word final position is /aɣ oɣ/ > /əɣ/ > /əj/ > /əi/ so I'd expect something like *cagh(a), *cadh(a), *cogh(a) or *ciodh(a). The triphthong /ɨəi/ seems to have merged into /əːi/ during the Classical Manx period and the development of long /əːi/ is a bit more complicated … but likewise there's no word I can find that would correspond to /kəːi/ such as *cao(i)dh, *cuaith(e) etc.
17th century Phillips uses kyei to describe wild animals but texts from the 18th and 19th centuries also use keoi(e) to mean 'fierce, mad, raging' and Cregeen has 'wild, mad, in a rage, not tame'. This makes me think that the word represents a palatalised variant of Gaelic caoch 'blind', *caoigh which would yield /kəːi/ in Manx! Semantically the development 'blind' > 'blinded with rage' > 'raving mad' > 'wild (as an animal)' seems plausible, especially with influence from cuthach 'rage, fury, madness' or even a blend of the two words.
Psalm 80:13
13. Ta yn kollagh kýei magh as y ghǽil da rowrey ſhúas: as ta bæin kýæi yn vagher dâ híi ſhuas.
The wilde Bore out of the wood doeth root it up: and the wilde beasts of the field devoure it. (Book of Common Prayer 1605 which Phillips was translating from)
Tá an collach [*caoigh?] amach as an choill dá rómhar suas is tá beathadhaigh [*caoigh] an mhachaire dá hithe suas.
