Plateau languages scratchpad
Posted: Sat Feb 15, 2025 8:03 pm
Decided to write out some ideas that have been bouncing around in my head for a few months now. This is a conlang family set in a region similar to the Interior Plateau area of western North America. It is a relatively dry zone between an extremely large mountain range to the north and a smaller range to the south that separates it from the coast. The languages here are part of a fairly shallow family (~2500-3000 years old) that is divided into two subbranches: the Highland branch spoken in alpine regions and the foothills of the northern moutains, and the Lowland branch spoken in the large river valleys fed by glacier meltwater from the north.
The Plateau people are mostly nomadic hunter-gatherers, although the Lowland peoples are more sedentary and practice some horticulture. The Proto-Plateau speakers were certainly hunter-gatherers, as there are no reconstructable words for domesticated plants/animals other than the word for "dog". There has been a lot of contact across the mountains between Lowland speakers and peoples on the Coast, whose languages belong to several families in a tight knit sprachbund, and there has been linguistic and cultural diffusion going both ways for centuries.
Some of the IRL influences of this family are pretty obvious; the proto-phonology is basically that of Proto-Iroquoian minus /n/ and plus /ɣ/. Some other influences are Inuit-Yupik-Unangan, Salish, and Indo-European.
Proto-Plateau phonology
Consonants
¹: Occurs only in morphophonogical alternation with *y, and in a handful of nominal/adjectival roots, all with diminutive or onomatopoeic associations (eg., *čõkʷəkʷ crow, raven, *tu:čẽh grey (cf. *tu:kẽh black), *ča:ča: little sister (cf. *ka:ɣa: older sister)).
²: The rhotic is reconstructed as an alveolar approximant, since this is the reflex it has in languages where it doesn’t shift to /l/ or disappear.
Vowels
It is likely that the resonants *r *y *ɣ and *w were pronounced as nasals in the vicinity of nasalised vowels, however the exact conditions for this cannot be reconstructed in Proto-Plateau due to differing reflexes in Highland and Lowland languages (i.e., nasalization before a nasal vowel in Highland, contrasted with progressive nasal harmony processes in Lowland).
Syllable and morpheme structure
Proto-Plateau had the syllable structure (s)CV(C). Any consonant could occur in the onset, and any consonant (other than the marginally phonemic *č) could occur in the coda. Roots were always disyllabic, thus the minimal Proto-Plateau root was *CVCV, while the maximal was *CVCCVC. The shape *CVCVC seems to have been preferred.
Suffixes, which were plentiful in the language, favoured simpler shapes, and did not require an onset. Prefixes, of which only around five can be securely reconstructed, all have the shape *CV-, except for the prefix *s-, which seems to have largely functioned as a nominalizer. Thus, onset clusters are limited to *s+C.
Morphophonology
Hardening and Weakening
Proto-Plateau had a process whereby certain suffixes triggered a change in the final consonant of a preceding morpheme. This was likely due to some prosodic feature in pre-Proto-Plateau which was then lost.
Weakening
Certain (but not all) vowel-initial suffixes triggered weakening of a stem-final obstruent. Suffixes that cause weakening are written with a superscript ᵂ before the suffix.
¹: *h weakens to *y following front vowels, *u following rounded vowels, and *∅ elsewhere.
Most often, *s weakened to *r. In some roots however, weakening of *r resulted in *y instead. This may be due to a merger of pre-Proto-Plateau **s and **š into *s, where original **s weakened to *r and **š weakened to *y, resulting in the unpredictable pattern of Proto-Plateau.
Hardening
Certain (but not all) consonant-initial suffixes triggered hardening of a stem-final resonant. Suffixes that cause hardening are written with a superscript ᴴ before the suffix.
Some examples of this process:
PP *kʷiɣis fish *-ᵂa:y collective > *kʷiɣira:y school of fish
PP *ʔisəkʷ house *-ᵂirõ make > *ʔisəwirõ make a house
PP *hiriɣ eat *-ᴴtasə for a long time > *hiriktasə spend a long time eating
PP *tuskay mountain goat *-ᴴkʷi: see > *tuskačkʷi: see a mountain goat
These processes resulted in many intermorphemic clusters in Proto-Plateau, which were then simplified in various ways in daughter languages.
Proto-Plateau Clusters
A velar stop followed by a velar stop results in a geminate of the second, so *k+kʷ and *kʷ+kʷ merge as *kʷkʷ, and *k+k and *kʷ+k merge as *kk. A velar stop followed by *w results in *kʷ, while *ɣ followed by *w results in *ww.
Clusters of two glottals were not permitted, so *ʔ+ʔ and *h+ʔ simplified to *ʔ, while *ʔ+h and *h+h simplified to *h.
The Plateau people are mostly nomadic hunter-gatherers, although the Lowland peoples are more sedentary and practice some horticulture. The Proto-Plateau speakers were certainly hunter-gatherers, as there are no reconstructable words for domesticated plants/animals other than the word for "dog". There has been a lot of contact across the mountains between Lowland speakers and peoples on the Coast, whose languages belong to several families in a tight knit sprachbund, and there has been linguistic and cultural diffusion going both ways for centuries.
Some of the IRL influences of this family are pretty obvious; the proto-phonology is basically that of Proto-Iroquoian minus /n/ and plus /ɣ/. Some other influences are Inuit-Yupik-Unangan, Salish, and Indo-European.
Proto-Plateau phonology
Consonants
| Alveolar | Palatal | Velar | Labiovelar | Glottal | |
| Stop | *t | (*č¹) | *k | *kʷ | *ʔ |
| Fricative | *s | *h | |||
| Resonant | *r | *y | *ɣ | *w |
²: The rhotic is reconstructed as an alveolar approximant, since this is the reflex it has in languages where it doesn’t shift to /l/ or disappear.
Vowels
| Front | Central | Back | |
| Close | *i *i: | *u *u: | [|
| Mid | *ẽ *ẽ: | *ə | *õ *õ: |
| Open | *a *a: |
Syllable and morpheme structure
Proto-Plateau had the syllable structure (s)CV(C). Any consonant could occur in the onset, and any consonant (other than the marginally phonemic *č) could occur in the coda. Roots were always disyllabic, thus the minimal Proto-Plateau root was *CVCV, while the maximal was *CVCCVC. The shape *CVCVC seems to have been preferred.
Suffixes, which were plentiful in the language, favoured simpler shapes, and did not require an onset. Prefixes, of which only around five can be securely reconstructed, all have the shape *CV-, except for the prefix *s-, which seems to have largely functioned as a nominalizer. Thus, onset clusters are limited to *s+C.
Morphophonology
Hardening and Weakening
Proto-Plateau had a process whereby certain suffixes triggered a change in the final consonant of a preceding morpheme. This was likely due to some prosodic feature in pre-Proto-Plateau which was then lost.
Weakening
Certain (but not all) vowel-initial suffixes triggered weakening of a stem-final obstruent. Suffixes that cause weakening are written with a superscript ᵂ before the suffix.
| Plain | *t | *s | *k | *kʷ | *h |
| Weakened | *r | *r, *y | *ɣ | *w | *y, *w, *∅¹ |
Most often, *s weakened to *r. In some roots however, weakening of *r resulted in *y instead. This may be due to a merger of pre-Proto-Plateau **s and **š into *s, where original **s weakened to *r and **š weakened to *y, resulting in the unpredictable pattern of Proto-Plateau.
Hardening
Certain (but not all) consonant-initial suffixes triggered hardening of a stem-final resonant. Suffixes that cause hardening are written with a superscript ᴴ before the suffix.
| Plain | *r | *y | *ɣ | *w |
| Hardened | *t | *č | *k | *kʷ |
Some examples of this process:
PP *kʷiɣis fish *-ᵂa:y collective > *kʷiɣira:y school of fish
PP *ʔisəkʷ house *-ᵂirõ make > *ʔisəwirõ make a house
PP *hiriɣ eat *-ᴴtasə for a long time > *hiriktasə spend a long time eating
PP *tuskay mountain goat *-ᴴkʷi: see > *tuskačkʷi: see a mountain goat
These processes resulted in many intermorphemic clusters in Proto-Plateau, which were then simplified in various ways in daughter languages.
Proto-Plateau Clusters
| *t | *s | *k | *kʷ | *ʔ | *h | *r | *y | *ɣ | *w | |
| *t | *tt | *ts | *tk | *tkʷ | *tʔ | *th | *tr | *ty | *tɣ | *tw |
| *s | *st | *ss | *sk | *skʷ | *sʔ | *sh | *sr | *sy | *sɣ | *sw |
| *č | *čt | *čs | *čk | *čkʷ | *čʔ | *čh | *čr | *čy | *čɣ | *čw |
| *k | *kt | *ks | *kk | *kʷkʷ | *kʔ | *kh | *kr | *ky | *kɣ | *kʷ |
| *kʷ | *kʷt | *kʷs | *kk | *kʷkʷ | *kʷʔ | *kʷh | *kʷr | *kʷy | *kʷɣ | *kʷ |
| *ʔ | *ʔt | *ʔs | *ʔk | *ʔkʷ | *ʔ | *h | *ʔr | *ʔy | *ʔɣ | *ʔw |
| *h | *ht | *hs | *hk | *hkʷ | *ʔ | *h | *hr | *hy | *hɣ | *hw |
| *r | *rt | *rs | *rk | *rkʷ | *rʔ | *rh | *rr | *ry | *rɣ | *rw |
| *y | *yt | *ys | *yk | *ykʷ | *yʔ | *yh | *yr | *yy | *yɣ | *yw |
| *ɣ | *ɣt | *ɣs | *ɣk | *ɣkʷ | *ɣʔ | *ɣh | *ɣr | *ɣy | *ɣɣ | *ww |
| *w | *wt | *ws | *wk | *wkʷ | *wʔ | *wh | *wr | *wy | *wɣ | *ww |
A velar stop followed by a velar stop results in a geminate of the second, so *k+kʷ and *kʷ+kʷ merge as *kʷkʷ, and *k+k and *kʷ+k merge as *kk. A velar stop followed by *w results in *kʷ, while *ɣ followed by *w results in *ww.
Clusters of two glottals were not permitted, so *ʔ+ʔ and *h+ʔ simplified to *ʔ, while *ʔ+h and *h+h simplified to *h.