Proto-Ash-Ish

Proto-Ash-Ish (henceforth referred to as PAI) is the reconstructed protolanguage linking the Ash and Ish languages through their own respective protolanguages, Proto-Ash and Proto-Ish.

The PAI speakers would have lived in a mountainous area during a period of glaciation which forced them to seek new lands. Those who would become speakers of Ish left the mountains altogether and found a temperate climate where the sea was not frozen over and became skilled sailors. Those who would go on to speak Ash delved further into the mountains, finding oases of heat in geologically active areas and came to worship the dual nature of fire. Eventually the ice age subsided and perhaps following volcanic eruptions they followed in the footsteps of their long lost kin and arrived at the ocean which now reached much further inland. This is where the two groups would eventually meet again and new cultures combining the old with the new would develop.

Phonology

Ish appears to be much more conservative than Ash or even Proto-Ash when it comes to the inventory of phonemes as the reconstruction of PAI is quite similar.

Vowels

The following vowels (with conventional romanisation rather than IPA) are posited:

	Front	Back
Short	*e	*a

PAI appears to have had neither the nasal vowels of Ish nor the long vowels of Proto-Ash, but otherwise the same two-way contrast as Ish.

Consonants

These are the basic consonants:

	Coronal	Dorsal	Laryngeal
Oral	*t	*k	*q
Nasal	*n
Approximant	*r	*w

Here a big difference from modern Ish is the presence of a nasal consonant whereas Ish has transferred nasality as a feature onto its vowels, losing the consonant in the process.

• As in Proto-Ash, it is unclear exactly what *q was. It merged with *k in Ish so may have been a uvular plosive but in Ash it developed into a glottal.
• The nasal *n likely assimilated the point of articulation of any other consonant in a cluster.
• The approximant *w was probably */ɰ/ as in Ish.
• The approximant *r might have been */ɹ/. It became /l~ɾ/ in Ish and disappeared in Ash.

Additionally there were important clusters that developed into new consonants in the descendants and may have been single phonemes already in PAI:

	Coronal	Dorsal	Laryngeal
Prenasalised	*nt~*ⁿt	*nk~*ⁿk	*nq~*ⁿq
Velarised	*tw~*tʷ~*tᶭ	*kw~*kʷ~*kᶭ	*qw~*qʷ~*qᶭ

The fact that these may have been single phonemes is supported by the fact that no other clusters seem to have been permitted. The convention is still to write these combinations as clusters.

• The velarised consonants may have been labiovelarised already in PAI.
• Consonants could also be both prenasalised and (labio)velarised at the same time.

The main developments into Ash and Ish were as follows:

PAI	Ash	Ish
*k	→ /k/	→ /kˣ/
*nk	→ /ŋᵍ/	→ /k/
*kw	→ /p/	→ /kˣɰ/
*nkw	→ /mᵇ/	→ /kɰ/

Ash developed poststopped nasals (which merged with the plain nasals, poststopping becoming an allophonic feature) and Ish developed a distinction between plain and aspirated or postfricated plosives.

Grammar and features

Nouns

PAI would have had some nominal morphology in the form of case endings inherited by the daughter branches only to be mostly lost or repurposed in various ways.

Deixis

The proximal and distal determiners are an example of common inheritance from PAI shared between the daughter branches.

PAI	Proto-Ash	Ish
*ewe "this"	→ *i(j), *(i)ja- → *i, *ja-	→ (iy)i /(e)ɰe/
*awa "that"	→ *u(w), *(u)wa- → *u, *wa-	→ (ow)o /(a)ɰa/

Verbs

Verb morphology seems to have been reduced mainly to auxiliaries while the main verb was quite invariant. These auxiliaries were inherited by Ash and Ish but treated somewhat differently. In Ash they became prefixes while in Ish they became nominal prepositions or verbal suffixes depending on analysis.

PAI	Ash	Ish
*qeka (transitive verb, direct object)	→ *qaka → ảh- (direct agency)	→ *ek → e(g), iq- /e(k(ˣ))/ (direct object)
*qate (reflexive/passive verb, indirect object)	→ *qati → ảs- (inverse agency)	→ *at → a(d), at- /a(t(ˢ))/ (indirect object, passive)

While these still serve somewhat similar functions in both languages, they work quite differently on a syntactic level.

Syntax

Thanks to the grammatical cases word order was probably somewhat free in PAI, reflected by the resulting differences in the daughter branches as they solidified their own word order in different ways as the case endings collapsed. Ash settled on auxiliary-noun-verb whereas Ish settled on verb-auxiliary-noun.

The auxiliary verbs appear to have been used with a kind of participle ending in *-er.

A sentence in PAI may thus have looked as follows:

*ntaw qeka te kaw-er
fire.SUBJ AUX.ACT water.OBJ whirl-PTCP
Fire stirs water.

*kaw-er qeka te ntaw
whirl-PTCP AUX.ACT water.OBJ fire.SUBJ
Fire stirs water.

It seems Ash preferred the first syntax while Ish settled on the second. The corresponding sentences in their own protolanguages would have been something like the following:

Proto-Ash	Proto-Ish
*naw qaka ti kwa:	*kaɰe ek te daɰ

The auxiliary *qate was somewhat unusual. It appears to have been reflexive while able to take a benefactor as a direct object. Some of its usages might be compared to the verbal "be-" prefix of Germanic languages. Looking at its modern reflexes, it seems to have denoted a lack of volition.

*taw-er qate ewe ntaw
stand-PTCP AUX.PASS PROX.OBJ fire.SUBJ
Fire was stood for/by this one.

This corresponds quite straightforwardly to modern Ish, besides a change in meaning, and the fact that a verb prefix is now required as well:

attayáti dow
/VC-tˢaɰe=aT-ɰe taɰ/
NDIR-appear=NDIR-PROX fire
I saw the fire.

The order of the referents and the participle were probably free to move around:

*te qate kaw~kaw-er
water.OBJ AUX.PASS whirl~REDUP-PTCP
It is being whirled for/by water.

This instead corresponds to modern Ash syntax although this is no longer the common word for water:

see ảsgoa
/ti hat͡s-kuwa/
water/air INV-LOCV:DYN
They are moved by water/air.

Reduplication

Reduplication seems to have formed frequentatives already in PAI as reflexes show up in Proto-Ash as well as modern Ish. In the latter it forms the continuous aspect whereas in Ash the reduplicated forms seem to simply have become separate words or to have replaced the non-reduplicated forms altogether, although this might not have happened until later when it may have been necessary to differentiate between words that were becoming homophonous due to sound changes.

PAI		Ash		Ish
*kaw~kaw(-ar)	"whirl, swirl, whorl, cycle, wrap"	→ *ku~kwa: → goa	"be/move around, elapse"	→ qacow /kˣa.kaɰ/	"be passing (of time, weather)"

Vocabulary

Some core vocabulary is shared between Ish and Ash but does not always mean the same thing or have a similar sound anymore. This word stock includes elements of nature, common actions, and cultural aspects.

PAI		Ash		Ish
*ntaw-	"fire, burn, glow"	→ *n(t)w-aj- → mee	"fire, heart, core, essence"	→ dow /taɰ/	"fire, light"
*qat-	"uncomfortable, harsh"	→ ảdla	"cold"	→ qat /kˣatˢ/	"bad"
*taw(-er)-	"stand"	→ *twa: → laa	"be, stand"	→ tayi /tˢaɰe/	"appear, see"

Developments

The general changes of individual sounds are fairly straightforward given the small inventory of PAI itself as well as its descendants. However in particular contexts many factors would have come into play at once, yielding more complex changes. Likewise roots were often extended with additional affixes in the daughter branches and so may not always correspond to them perfectly.

Laryngeal and rhotic colouring

The back consonant *q sometimes affected vowels in the daughter languages.

PAI		Ash		Ish
*nteq-	"experience, feel"	→ "suffer" → naa; nahga	"die"; "itch, irritate"	→ nẽɱ /tẽkɰ/	"be, live, make"
		→ nao	"(sentient/mortal) individual, soul"	→ nẽɱ /tẽkɰ/	"(sentient) being, individual"

In this case we see backing of *e to *a in Ash while in Ish rhinoglottophilia served as one of the sources of its nasal vowels. Ish also appears to have extended the root with a suffix of which today remains only /-ɰ/.

The same seems to have happened around *r:

PAI		Ash		Ish
*ner-	"smell"	→ *nana → ıdna͠a	"nose"	→ nẽdẽ /DẽDẽ/	"nose, smell"