I have been watching a lot of Q’s Greenland recently and hearing her speak Kalaallisut has inspired me to finally sit down and work on my Quriil sound changes. (Also, I was sick for more than a week, and while that sucked, it gave me a lot of time to think about my conlangs again.) I’ve also decided to deviate a little bit from my original plans for the phoneme inventory. In particular, I will throw out the aspirate consonants, and only require one “super-short” vowel instead of three. This will allow for a little more flexibility when applying sound changes.

We will start off from Proto-Ryka-Quriil this time. Again, each sound change has a (NAME), and Proto-Ryka-Quriil words are prefixed with * in the later discussion.

Sound changes

(CENT-VOW) The short central vowel is lowered into a schwa. This is going to be Quriil’s “super-short” vowel. Simple, but effective.

• ɨ̹ → ə

(PAL1) All velar consonants (including the velarized lateral) are palatalized in the vicinity of /i(ː)/. This will help to obscure the consonant correspondences to Ryka a bit.

• kʰ, k, ŋ, ɫ → cʰ, c, ɲ, ʎ | i(ː) _, _ i(ː)

(AFFRIC1) The aspirate plosives are turned into affricates. Since Ryka’s aspirates are already directly transformed into fricatives, I wanted to go a slightly different route here, although most of these former aspirates will end up being fricatives in Quriil as well, as we will see later.

• pʰ → p͡ɸ
• tʰ → t͡s
• cʰ → c͡ç
• ʈʰ → ʈ͡ʂ
• kʰ → k͡x
• qʰ → q͡χ

(LAT1) A sequence of obstruent and lateral approximant merges into a lateral affricate. This especially applies to all vocalic laterals. If this combination is directly followed by another consonant, the schwa is inserted to form the nucleus instead. This gets rid of all vocalic laterals, leaving only true vowels as nuclei, while also introducing Quriil’s characteristic lateral fricative and affricate.

• [+obstruent] [+lateral] (C) → t͡ɬ (ə C)
• C [+lateral] (C) → ɬ (ə C)

(CLUST) Consonant clusters are disbanded by deleting the initial consonant and applying compensatory lengthening to the preceding vowel (if applicable).

• V C₁ C₂ → Vː C₂

(DIPH) Quriil does not have any phonemic diphthongs (although diphthongs can be realizations of stressed short vowels, as we will see later). In the process of getting lost, the less prominent vowel of the diphthong alters the place and/or manner of articulation of the adjacent consonant (if existent). The high front vowel /i̯/ palatalizes labial, alveolar, and retroflex consonants as well as the velar nasal, and velarizes uvular consonants. The rounded back vowel /u̯/ labializes alveolar and palatal consonants, and uvularizes retroflex and velar consonants. It also velarizes the palatal lateral. The central vowel /ɨ̯/ turns plain stops into affricates. Finally, all diphthongs are reduced to plain vowels consisting only of the previous prominent vowel.

• p, t, ʈ → c | _ i̯, i̯ _
• p͡ɸ, t͡s, ʈ͡ʂ → c͡ç | _ i̯, i̯ _
• m, n, ɳ, ŋ → ɲ | _ i̯, i̯ _
• ɫ → ʎ | _ i̯, i̯ _
• q → k | _ i̯, i̯ _
• q͡χ → k͡x | _ i̯, i̯ _
• t, c → p | _ u̯, u̯ _
• t͡s, c͡ç → p͡ɸ | _ u̯, u̯ _
• n, ɲ → m | _ u̯, u̯ _
• ʈ, k → q | _ u̯, u̯ _
• ʈ͡ʂ, k͡x → q͡χ | _ u̯, u̯ _
• ɳ → ŋ | _ u̯, u̯ _
• ʎ → ɫ | _ u̯, u̯ _
• p, t, c, ʈ, k, q → p͡ɸ, t͡s, c͡ç, ʈ͡ʂ, k͡x, q͡χ | _ ɨ̯, ɨ̯ _
• i̯, u̯, ɨ̯ → ∅

(LONG) In non-initial syllables, long vowels are shortened. This is compensated by the insertion or “hardening” of consonants. At the end of words, a plain stop is inserted after the now short vowel. Before plosives and affricates, a matching fricative is inserted. Following nasals are turned into their plosive equivalents, and liquids are turned into the lateral fricative. Again, this helps to mask the genetic relationship to Ryka and may cause some confusion among Rasvrisu historical linguists by spawning unexpected consonants into the words and deleting vowel length contrasts.

• aː → ak | _ #
• iː → it | _ #
• uː → up | _ #
• Vː → Vɸ | _ p
• Vː → Vs | _ t
• Vː → Vç | _ c
• Vː → Vʂ | _ ʈ
• Vː → Vx | _ k
• Vː → Vχ | _ q
• Vː [m, n, ɲ, ɳ, ŋ] → V [p, t, c, ʈ, k]
• Vː [ʎ, ɫ, r] → Vɬ
• Vː → V

(VOW-MOVE) To obscure the vocabulary further, short high vowels move around in front of certain consonants. This rule does not affect long vowels.

• i → a | _ [+retroflex, +velar, +uvular]
• i → u | _ [+labial]
• u → a | _ [+alveolar, +retroflex]
• u → i | _ [+palatal]

(AFFRIC2) Affricates are reduced again in most places. After a vowel, they turn into fricatives. Word-initially, they become plain stops. Only after consonants are they kept; instead, the preceding consonant is dropped. This scenario is especially fed into by (LONG), since there, a fricative may have been inserted before the affricate to compensate for the shortening of a long vowel.

• [p͡ɸ, t͡s, c͡ç, ʈ͡ʂ, k͡x, q͡χ] → [ɸ, s, ç, ʂ, x, χ] | V _
• [p͡ɸ, t͡s, c͡ç, ʈ͡ʂ, k͡x, q͡χ] → [p, t, c, ʈ, k, q] | # _
• C → ∅ | _ [+affricate]

(LAT2) All palatal laterals turn into lateral fricatives. This is the main source for the lateral fricative apart from (LAT1) and (LONG). The set of palatal fricatives was previously altered by (PAL1) and (DIPH), to obscure this correspondence a little bit.

• ʎ → ɬ

(LAB-FRIC) We now want to reduce the number of distinct consonants for some places of articulations to add some healthy asymmetry to the consonant inventory. First are the labials; here the affricate (and the reverse combination) as well as the plain fricative are reduced to a glottal fricative.

• [p͡ɸ, ɸp] → h
• ɸ → h

(UVU-FRIC) Next are the velars and uvulars. The velar fricative merges into the uvular one. In addition, the velar and uvular affricates are also reduced to a uvular fricative.

• x → χ
• [k, q] → ∅ | _ χ, χ _

(PAL2) Finally, we want to get rid of most of the palatal consonants that we introduced in the beginning. They now merge into the alveolars. The alveo-palatal fricative /ɕ/, which was inherited from Proto-Ryka-Quriil and has not been touched on yet, is also merged into the post-alveolar fricative /s/.

• c → t
• ç, ɕ → s
• c͡ç → t͡s

(Y-DEL) The schwa is dropped at the end of words.

• ə → ∅ | _ #

Resulting phoneme inventory

	Lab.	Alv.	Pal.	Lat.	Retr.	Vel.	Uvul.	Glot.
Stop	p p	t t			tt ʈ	k k	q q
Fric.		s s		ll ɬ	ss ʂ		x χ	h h
Affr.		ts t͡s		tl t͡ɬ	tts ʈ͡ʂ
Nas.	m m	n n	nj ɲ		nn ɳ	ng ŋ
Approx.		r ɹ	j j	l ɫ

The resulting consonant inventory does have a lot of members, but not quite as many as my original draft of Quriil (cf. the summary in Part 1). Most notably, the aspirate consonants are gone, but also any voicedness distinction. What remains are the three laterals, the five places of articulation for plosives, and the reduced velar inventory.

Overall, the consonant inventory is nicely asymmetrical: For the two “central” places of articulation, alveolar and retroflex, it has the full set of stop, fricative, affricate, and nasal. Just like many natural languages, it sports a reduced set at the edges, with only a stop and nasal in the labial and velar columns, and only a stop and fricative in the uvular column. It is not uncommon that gaps in the consonant inventory arise in these places of articulation first, so this distribution is quite naturalistic. For other places of articulation, like palatal or glottal, their more common members are strewn in as well. The lateral fricative and affricate are obviously rarer sounds and add some character to the inventory.

	Front	Central	Back
High	i, ii i, iː		u, uu u, uː
Mid		y ə
Low		a, aa a, aː

The vowel inventory offers no surprises. We have the common vowels /a/, /i/, /u/, both short and long, and the central schwa (only short).

Phonological processes

We have now seen the phonemes of Quriil, but there is more! Quriil has some phonological processes that are not phonemic in that they are entirely predictable and thus also not represented in the romanization. Still, they greatly alter the pronunciation of words, and so further obscure the genetic relationship between Quriil and the Rykaic languages, at least in its spoken form.

Intervocalic voicing

This is something I decided on while working on the sound changes and noticed that the consonants lacked any voicedness distinction: All consonants, more precisely all plosives, fricatives and affricates, become voiced between vowels. They remain voiceless at the beginning and end of words.

Vowel realization

Basically the only fixed feature of Quriil’s phonology is its vowel system. While it only has four vowels phonemically, these come in a lot of different flavors depending on the preceding consonant and stress pattern. This table shows the different realizations for the old vowel inventory with the three old “super-short” vowels:

Long vowels are always pronounced as long vowels, but change position and sometimes roundedness according to the preceding consonant. Short vowels are usually diphthongized when stressed and lax when unstressed, and their position is also dependent on the preceding consonant.

The “super-short” or reduced vowel schwa does not change its position, but may vocalize preceding continuants instead of being clearly articulated as a full vowel.

Final vowel reduction

In addition to the different realizations, I have decided that word-final short vowels are usually realized as the reduced vowel unless emphasized, but will resurface in the underlying full form when additional grammatical material (like suffixes) follow.

Generating some Quriil words

Let’s now check out what the Quriil vocabulary will look like! Unlike for Ryka, I do not yet have any fixed vocabulary for Quriil, so we do not need to check whether any words can be reconstructed to a proto form using our sound changes. Instead, we can take the proto words we reconstructed from our Ryka vocabulary and derive Quriil descendants from them! This also allows us to compare Quriil and Ryka cognates to see how visible the genetic relationship actually is. As a reminder: The goal was to obscure it on first sight, but create some notable correspondences under more careful examination.

Without further ado, here’s the word table from Part 2 with the new Quriil cognates:

Meaning	Quriil	Old Ryka	Proto-Ryka-Quriil
eye	nnuu ɳɯː	lu lu	*ṇuu ɳuː
to support	pin pʏɪ̯n	ped pɛd̪	*pin pin
tongue	llasa ˈɬɛɪ̯zɛ~ˈɬɛɪ̯zə	lath laθ	*lhatha ˈʎatʰa
mushroom	put pɔʊ̯t	pot pɔt̪	*put put
fire	sys səs	shyth ɕi̹θ	*shythy ˈɕɨtʰɨ
bone	kaasill ˈkɑːzɪɬ	katkhen ˈkaθxɛn	*katkhilh ˈkatkʰiʎ
salt	taakup ˈtæːɡɔp	tarku ˈtarku	*ṭaarkuu ˈʈaːrkuː
feather	lliipast ˈɬiːbɐst	‘ypput ˈʀipːut̪	*liipuut ˈɫiːpuːt
to open	qiisak ˈqeːzɛk	hyttja ˈhitːi̯a	*qhiitiaa ˈqʰiːti̯aː
sick	nniiq ɳiːq	leuwr lɛu̯r	*ṇiiuṭ ɳiːu̯ʈ
stone	tlyk t͡ɬək~t͡ɬ̩k	qnk ʔŋ̩k	*qlk qɫ̩k

The result, I think, is pretty cool! There are very few audible similarities, and only in the shorter words (mostly qu. put and ry. pot or qu. sys and ry. shyth). The longer words have been obscured pretty well by both language’s sound changes (compare qu. kaasill and ry. katkhen, or qu. qiisak and ry. hyttja!). Where the relationship may be visible in writing, the phonological processes in Quriil hide the similarities in the spoken language (like qu. llasa [ˈɬɛɪ̯zə] and ry. lath [laθ]). Also note that both Quriil and Ryka have native writing systems and the orthography I use here is only a romanization that the speakers themselves won’t use! A linguist studying both languages will be able to see that they are related, while two layman speakers of the languages will not, especially when diefferent grammar rules come into play. This is exactly the effect I wanted to achieve!

Conclusion

After designing the sound changes for Ryka in the last part and seeing 25% of my Ryka vocabulary rendered illegal by them, I was a bit frustrated and questioning whether I really wanted to continue down this path. But seeing Quriil evolve so nicely from Proto-Ryka-Quriil and having the cognate chart in front of me, I feel a lot more optimistic about this! I’m especially curious to see how the Konoic languages will turn out from this… Next up will be Proto-Konoic and Old Nyirvón, since this one is the most fleshed out of the Konoic languages. We can then also check out how to derive Modern Nyirvón and Siidde from Old Nyirvón, and finally, have total freedom to apply fun sound changes for a completely new Asiul! Hurray!