Future Poswa

Future Poswa is a hypothetical projection of the Poswa language thousands of years into the future, after its speakers had been overwhelmed by immigration from the Moonshine Empire.

I have had different ideas as to when this should be spoken. It was originally meant to align with Xykhasl, which was spoken in 12850 AD. But other writeups said that it was only 2000 years after classical Poswa. So taking the middle of these puts it at about 11820 AD, just over 3,000 years beyond classical Poswa.

I will never fully create this language in the sense of giving it its own lexicon, .... I will just derive words ad hoc from the Poswa dictionary. I will focus on just this one daughter language even though there would most likely be at least a small group of them.

Naming

I may change the name to something like Sunflower Poswa or Camp Poswa depending on the climate of where the people live. Camp Poswa is a reference to an idea I had in high school for a "bigger" PIE-style language which somewhat resembles what Future Poswa has developed into.

I had earlier considered Lava Poswa, as in the scene where the people are driving through a rocky landscape sweeping hot lava away with their windshield wipers. But this might be better as a retrospective term for Classical Poswa, where everything was hot and uniform with no parts of speech. I have since re-used this metaphor in the Lava Bed languages which are only distantly related.

I could go into physics and call classical Poswa Planck Epoch Poswa and then have something like GUT Poswa for this, but even though I like physics in general I don't think I want to use proper names.

Another possible name is Chrysalis, which really doesnt describe the language at all, but has a sentimental feeling because it was once the name for a new version of Moonshine that was originally going to be the superstratum of this language.

Family

Poswa only sprouted languages that shared its acoustic characteristics. While languages like Leaper developed a few offshoots that came to sound somewhat like Pabappa and Poswa, neither Pabappa nor Poswa ever developed offshoots that sounded like Leaper.

Future Poswa is in most ways more archaic than Pabappa even though it has been evolving for 5,000 years (5547 AD to ~10500 AD) whereas Pabappa only had 3,000 years (5547 AD to 8773 AD). This is because Pabappa began rapidly changing as soon as it broke away whereas the Poswa branch was very conservative during the first three thousand years (and to some extent also during the remaining two thousand).

Scratchpad

PLural verbs

May 30, 2020

Implement /-wošub/, etc as ordinary PIE-style verb inflections so that Future can have a distinct plural conjugation instead of building the plural by applying a formulaic transformation to the singular. This could enable mismatches. It is also possible that this will happen only for SAPs, and that the 3rd person will be the same in singular and plural. THe advantage of this is that the form that would otherwise be the plural 3p could instead be the exclusive 1p plural or perhaps 2p. (Though it would make more sense to build it "properly")

The reflex of /-wošub/ is /-osu/, but it could cause mutation of a preceding consonant.

Imperatives

May 5, 2020

Imperative /-wub/ and /-ub/ merge, so it is no longer possible to distinguish the ordinary 1st person imperative ("let me eat!" from the passive 1st person imperative ("eat me!"). One of the forms will need to be extended. Note that in Classical Poswa, these were often but not always distinguished not just by the /w/ but also by different consonants at the end of the verb stem, so if this process could be generalized, it could save the distinction.

There could be a merger between "I want to ___" and "Let me ____". The two are formed in completely unrelated ways so there would be no chance of ambiguity. This means that the basic form is the passive, and therefore Future Poswa has an atomic form for the 1st person passive imperative ("___ me!")

Apr 27, 2020

Maybe use Poswa's B-__-V affixes as noun classifiers. e.g. Poswa B-mp-V could be used on nouns for artificial and manufactured items. It is canonically inalienable in Poswa and might still be in Future Poswa.

Poswa has alternations like

pumu "blanket"

pubbo "my blanket"

pumompo "my blanket, which i made"

pumompešo "my blanket, which you made for me"

If retained vestigially in FP, there could be a 3x3 matrix like

       1P        2P        3P
1P     -mpo      -mpose    -mposa
2P     -mpeso    -mpe      -mpesa
3P     -mpaso    -mpase    -mpa ~ -mpasa

These could either be /s/ or /š/ depending on whether the sound shifts. However, it is likely that all of the forms with /s/ will simply fall out of use.

If this happens, it will come from the B-stem, even though all nouns probably lose their B-stems. This is remedied by having the B-stem take a dummy morpheme that causes it to be reinterpreted as a D-stem. Thus these nouns will appear regular, rather than alternating between, for example, /pubb-/ and /pumomp-/.

Private verbs

Could redevelop private verbs, but with animals sometimes higher than humans. This is because of their lifestyle.

PROBLEM

Mar 22, 2020

Because of changes detailed below, all nouns in Future Poswa will end in either /-a/ or a suffix that begins with /-a/. This vowel cannot be syncopated, so its presence is a peppapopaba problem.

Pabappa never encountered this problem because even when it simplified its grammar, it still retained the A-stems of most words as the basic form. The /-a/ problem may still at least go away when inalienable nouns are discussed, since the three person suffixes /-o -e -a/ would remain in use. (It is also possible that the anciently lost /-y/ for 1st person may somehow return through analogy, and become either /-i/ or /-Ø/.)

POSSIBLE SOLUTION

See history for other ideas.

May 8, 2020

One possible way to push out of this is to use B-stems for some words, and then add a mandatory /-v-/ that then disappeares. This is still unsatisfactory however as it implies that the words will always have a vowel sequence.

Or perhaps instead, D-stems ending in /-v -ž/ could lose those consonants, but regain them when the noun is used with a person marker. This means that these nouns will in fact distinguish between their free form and their 3p form, meaning that they are alienable. Ideally this category should grow to encompass most of the vocabulary, so perhaps there will need to be innovations that take in more than just /-v -ž/. /-mp/ is one more cluster that could be analogized away; see the word for :"house" for example.

It could be that all nouns that remain with consonants at the end of the stem are inalienable regardless of their meaning, and that analogy will progress to the point that almost all consonant stems are words that are in fact semantically appropriate for this category. For example, /papp-/ "heart" could remain with its final consonants, but /papp-/ "river dam" will need some sort of extender that makes it come to end in a vowel. Longer stems will delete the final consonants instead of adding to them, but only if the final consonants can be analogized to some sequence that in C Poswa was an augmented possessive marker.

The two meanings of /-v-/, which would typically not overlap in C Poswa, could merge here, meaning that nouns for vegetables etc could gain a /v/ and then immediately lose it to analogy, and thus survive in their B stem form instead of their D stem form.

Sound changes

shift /ia ie i io iu/ to /ʲa ʲe ʲi ʲo ʲu/ and then shift /y/ to /i/. On the other hand, this shift is so minor that it would not even need to be represented in the orthography, and it would not create any vowel harmony. It could thus be said that there were no vowel shifts at all in Future Poswa, or at least no unconditional ones.

/ul/ > /ȳ/ is real, though, since it is a long /i/ but did not gain palatalization.

/tw nw/ > /pw mw/ at some point. This might occur after coda delabialization (so that in theory coda /t n/ can exist), and before the later shift of /wu wo/ > /u o/.

The voiced stop /ġ/ changes to /d/, probably taking /dž/ with it.

Probably /ž/ > /z/; may or may not be accompanied by /š/ > /s/.

It is possible that all "solid" consonants further back than alveolar are shifted to alveolar; that is, /š ž tš dž k ġ/ will all shift to something like /s z t d t d/, or even /s d t d t d/, and palatalization will be allophonic before /ia ie i io iu/ if it occurs at all. /r/ will survive, but its pronunciation will become much like /w/. Labialized consonants like /kʷ ġʷ/ might survive as true velars, but might also become so statistically rare that they disappear from the language without any single diachronic shift that removes them.

There will be linguolabial consonants as well. In classical Poswa, the sequences /plw blw lw/ were already linguolabial for most speakers, so Future Poswa might add to these. However it is difficult to generate a nasal except by shifting palatalized labials to linguolabials, which would leave the distribution very unbalanced. If this happens it must be highly conditional.

There are no vowel deletions at all, and probably no consonant deletions either. Thus every word retains the same superficial syllabic structure as it had had in classical Poswa, and the stark changes in surface forms are due to massive reorganization of the grammar.

If there are consonant deletions, it may be /f v/ > /h Ø/, but this would be conditional.

Long vowels and new diphthongs appear from vocalization of coda liquids:

/el il ol ul yl/ > /ē ʲī oi ui ʲī/ <--- but /ui/ > /ȳ/ at least conditionally
/ar er or ur/ > /ā ea oa ua/
/alʷ elʷ ilʷ olʷ ulʷ ylʷ/ > ??????
/arʷ erʷ irʷ orʷ urʷ/ > /au eu ʲiu ō ū/

All gaps on the left side of the list are deliberate; classical Poswa never had /al/, /ir/, etc.

These changes were actually complete in classical Poswa, but were never represented in the orthography because the surface diphthongs here behaved differently from the inherited diphthongs. In classical Poswa, syllables with /ae ia ie io iu/ could have an additional consonant in the coda, but syllables with the other diphthongs could not. In Future Poswa, this situation may still be true or it may have collapsed such that all nuclei behave the same way.

deletion of longs

It is possible that all long vowels in unstressed syllables could shorten and falling diphthongs could shrink to monophthongs. Then, if /v/ > /Ø/, they would reappear again after that shift. Likewise, there could be deletion of geminate consonants overlaying unstressed syllables, and reduction of any other unstressed clusters besides /mp mb nt nd/.

Even the longs that appear from loss of /v/ would probably still shorten in closed syllables. e.g. /tuvumb-/ "almost". This could also be used to get /-ovos/ > /os/.

Other possible sound changes

/mb/ > /m/ when overlaying two unstressed syllables, or perhaps unconditionally. Because in classical Poswa intervocalic /mb/ was much more common than intervocalic /m/.

/wu wo/ > /u o/. (Strictly speaking it is the other way around, but the orthography prefers to use the simplest possible spellings and Poswa is treated as if it was its own phonetic alphabet.)

Possible target phonology

 CONSONANTS

 Rounded labials:    pʷ   bʷ   mʷ   w    hʷ   lʷ^?
 Plain labials:      p    b    m   (v)  (f)
 Linguolabials:                                  (to be filled in later)
 Alveolars:          t    d    n   (z)   s    l
 Dorsals:                                h    r

 VOWELS

  a   e   i   o   u
 ʲa  ʲe  ʲi  ʲo  ʲu
 
  ā   ea  ia  oa  ua
  ae  ē       oi  ui
  au  eu      ō   ū
         ʲī 
         ʲiu

The plain, un-iotated /i/ will be spelled y.

/lʷ/ is probably not real.

Because /ʷ/ and /ʲ/ cannot occur in the same syllable, even through vowel collision, it is possible that they can be analyzed as the same feature. And since there are no labialized alveolars, it could be said that both features belong to the consonants and that the vowel system is just /a e i o u/ with a range of permissible sequences.

A third possibility is to get rid of coarticulation altogether and analyze the maximal syllable structure as CC(w|y)VC, and just say that sequences like /tw/ do not exist.

Grammatical changes

The grammar of nouns and verbs becomes extremely difficult, because despite the general trend towards the reduction of complexity in the grammar, the simplifications lead to a vast increase in exceptions to the rules. For example, even if Poswa had 400 noun declensions, the speakers learned them all because they all followed patterns predictable from the phonemes towards the end of the word. Whereas Future Poswa may have only 15 declensions, but they are based on what the word used to look like rather than what it looks like in Future Poswa.

Likewise, the classical Poswa verb system with its many infixes was very complicated, but speakers only had to learn three stems for each verb (normal, conjugated, and oblique) whereas in Future Poswa the infixes become fusional and some verbs may have more than a dozen different mutated stems.

pop "two; twins; a natural pair" and popu "some of, part of" could merge, likely in favor of the latter, and take on a merged meaning similar to English "couple (of)".

Smooshing together of huge affix chains

Possibly evolve a new accusative. Instead of "the girl acted on the boy", say "the girl, the boy she acted on". A few verbs might agglomerate onto the noun stem so that e.g. there would be alongside the accusative a new case for "the ___ she helped", etc. And dozens of "cases" like this could exist side by side with the canonical ones. This is another idea that was stable in classical Poswa because there was no distinction between parts of speech, but in Future Poswa it would be unstable.

Note that in classical Poswa, the "the ____ she _____ed" forms are all based on B-stems, which likely disappear in FP.

Other grammatical change ideas

Full or partial loss of A- and B-stems, with their functions being taken over by suffixed forms of D-stems. Thus the D-stem comes to be the most basic form of the word, and sound changes look mysteriously haphazard, such as tepe "thorn" showing up as tepped-. Most speakers were not troubled by this because the A-stems appeared only in fossilized compounds if at all, but when these compounds did appear, their meanings were more difficult to comprehend than they had been in classical Poswa. Many stems would be completely unrecognizable, such as wutu "champion, winner" appearing as wubb-, and plamba "rain cloud" appearing as plannob-.

Defective stems like pappipe "sunflower", which had no D-stems, might be exempt from this change, or might form new D-stems analogically which then would be used to create the new forms that replace the A- and B-stems. For example, the word for sunflower could be reanalyzed as a verb with the stem pappip-, and then a new form such as /pappipa/ or /pappipana/ would be coined to replace the noun.
While it might seem more efficient to have the B-stems be the new base forms instead of D-stems, this could not happen because there was never a stage in Poswa in which the B-stems could occur without an immediately following consonant suffix. This in fact goes back thousands of years before Poswa. Pabappa sometimes captured B-stems and made them the basic form of words, but only because Pabappa lost final /-s/ and therefore created doublets of many words. Future Poswa loses none of its final consonants and so cannot do this.

Since D-stems in classical Poswa always ended in consonants or consonant + /i/, any suffixes attached to them would have begun with vowels. The vowel /u/ could become a Future Poswa suffix denoting a noun formed from a D-stem, but the general sound of the language would be better preserved if this vowel was /a/ even though an /a/ would be ambiguous with the third person present form of the D-stem verb. In some cases, a suffix might need to be added to this suffix, so that nouns would become obligate compounds, and, for example, the word for rain cloud might be /plannobapa/, using the inherited suffix -pa "in the sky", with no shorter form of the word available except as a verb. Other available noun-forming suffixes, each of which would attach to the /-a/ suffix, would be -na "residue, result(ative)", -ta "agent", and -p "verbal noun (archaic)". This /-p/ was already archaic even in classical Poswa but could reappear by extension of use from other homophonous suffixes such as the subjunctive /-p/ and the reflexive /-p/.

Use of bare /-a/ as the minimal suffix would be helped by the fact that many such words already existed in classical Poswa, such as pižupa ~ pižup- "school", meaning that the ambiguity was already present in the language and any potential problems had been worked around already. Whereas there were few such words where the A-stem ended in /-u/.

Preexisting sound changes would carry over, but syncretism may set in, again making the sound changes look haphazard. For example, the suffixes /-ta -pa -p/ might all share the same /-f/ alternant while /-na/ gets /-v/. This would obey a longstanding Poswa grammatical operation, but in fact the classical Poswa suffix /-pa/ "in the sky" evolved from a form with an initial cluster and did not participate in the /p/ > /f/ sound change. This could cause /-ta/ to be reinterpreted as an animate noun marker and /-pa/ as an inanimate, or at least non-human, one.

It is quite possible that a new rule that all nouns must end in vowels would appear, since they would all derive from D-stems with a vocalic suffix, and because the inherited case endings would at least include /-p -m -s/, all of which would require a stem ending in a vowel. This assumes B-stems disappear entirely, as B-stems were the means that classical Poswa used to attach /-p -m -s/ to stems whose surface forms ended in consonants. If A-stems disappear entirely, B-stems must disappear entirely as well, and it is also possible that B-stems disappear entirely even if some A-stems remain.

Grammatical gender could evolve, likely from interpretation of words for man and woman (or perhaps girl and boy, or perhaps all four) as variants of the generic agent marker -ta.

Noun plurals

If the plural marker remains interior to the person marker (that is, "my three boys", not "three (of) my boys"), then it is okay to retain the inherited cooperative plural -bum, which would attach to the primordial A-stem and then collapse to just -b- and then take the C-stem endings on top of that. Thus, although A-stems would no longer exist, there would be a separate plural stem for some nouns that retained the original A-stem form even though it was grammatically a C-stem.

Note, this is at odds with the earlier claim that all nouns were based on D-stems except inalienables.

There would be extremely many of these nouns, since probably more than half of all classical Poswa nouns had A-stems that differed from their C-stems in nontrivial ways. Thus, it is likely that massive analogy would occur to reduce the variation. On the other hand, the loss of A-stems in Future Poswa was not a conscious process aimed at simplifying the language; it simply happened because C-stems came to be used as the productive form and A-stems became less prominent over time. Thus it is possible that noun plurals could actually be more difficult to learn in Future Poswa than they had been in classical Poswa.

Some examples of the wild variation between A- and C-stems in classical Poswa that could lead to wild variation in plural formation in Future Poswa are below:

 A-stem       C-stem      meaning
 
 pupop        pisi-       judge
 wanna        wamb-       mat, towel, pad, mattress
 fempis       fempap-     square
 plampum      plampev-    roots of a plant

Inalienable nouns

Loss of A- and B-stems creates a wide category of inalienable nouns. Those that retain their D-stems will not be inalienable except for some that may join the category for semantic reasons.

Possible ideas for inalienable nouns:

Semantically inappropriate inalienable nouns

Loss of A- and B-stems in nouns that already lacked D-stems would leave just the C-stem remaining, and make the noun inalienable regardless of its meaning. Some of these might be repaired by using the third person possessive form and attaching a dummy noun in the genitive afterwards. For example, the stranded form pepop- "library" would conjugate like any other inalienable noun into pepopo, pepope, pepopa for 1st/2nd/3rd person, but might also develop a new bare form such as pepopabas by adding the noun /ba/, in the genitive case, after the noun. But note that inflections would then go inside, creating forms like /pepopambas/ "in the library", never *pepopabašam. So perhaps this does nothing but create a postposed definite article for the noun, and the two words should be spelled separately.

These postposed definite articles could be seen as gender markers, or at least noun classifiers, as they nearly duplicate the original role of the Gold noun classifiers that had emerged over 9,000 years earlier. Some nouns would have no article, and these would be the ones considered semantically appropriate inalienables such as "arm".

The definite article may end up moving to the front of the noun. It may also somehow lose its /-s/, which was originally required by the grammar (could not use /-m/, even when referring to a place, because of the possession marker.)

It is more likely that words like these will be padded with B-stem "linkage morphemes". thus the stem for library is not /pepop-/, but /pepopamp-/, where /mp/ means "made, created". THe bare form of this just deletes the /mp/. it is also possible that all such words generalize the linkage morpheme to something such as /d/ or /Ø/ (from /v/).

Impossessibles

Like Pabappa there may be also a category of nouns that have no possessed forms because they are proper nouns, singular entities (e.g. the sun), or belong to nature. In Pabappa, this is only an informal category because alienability is not grammaticalized, but many such nouns in Pabappa exhibit paradoxical inflection whereby the form of an impossessible noun ends in /-i/, the basic possessive ending, since there is no context in which a further /-i/ would need to be added to it.

Something similar could happen in Future Poswa but through a different process. In Future Poswa, the endings would not simply evolve, but would need to be added deliberately. Thus speakers could start saying /pupo/ "our sun" instead of just /pipi/ "sun". (But note that this word never had a C-stem in classical Poswa, and it would need to be created by analogy; they might choose the etymologically unsound /pipo/ instead.) Grammatically, this would probably fall into the inalienable category, and it would just happen that the 2nd and 3rd person forms are never used.

Evolution of verb endings

The most interesting change is that many of the formulaic verb endings freeze out, so that Future Poswa has, e.g. /-epo/ "I control you", no longer analyzable as /e/ + /p/ + /o/. Many things will need to be reordered, however, to make things like this work properly, since it is almost certain that Future Poswa will retain the simple three-person /o e a/ conjugation pattern.

Possible new type of transitive verb

The -v- and -ž- verbs in Poswa could lose some aspects of their meaning and become new types of transitive verbs, possibly with a passive meaning. The /ž/ in particular could become conflated with the preexisting passives in -ž-, which had been distinct throughout the thousands of years between Play and classical Poswa because they were usually surrounded by different vowels. For example, in classical Poswa, soppobažo meant "you answer me" and soppobežo meant "I feel your answer". However, a few commonly used past tense forms merged: soppobeži meant both "you answered me" and "I felt your answer" in classical Poswa.

Remember that the inner vowel in the primary Poswa transitive verbs was a tense marker, not a person marker. The newly evolved verbs would mark both tense and person on the outside, and mark only person on the inside. It is also possible that the system could become tangled and that the vowels would no longer be segmentable.

/v/

If /v/ shifts to /Ø/ unconditionally, the verbs with /v/ will come to have vowel sequences instead. Assuming this occurs after the shortening of inherited "long" vowels, these would produce true longs, /ō ē ā/, for the forms in which the inner and outer person markers were the same. These could be reinterpreted as intransitives, which would fit with the inherited pattern of intransitive verb affixes having only one vowel.

/ž/

If the new /ž/ verbs come to be seen as passive, then, assuming that the C- and D-stems merge by analogy, there would be two sets of synonymous passive verbs both using the /ž/, which would evolve to /z/ (and perhaps even to /d/). Below is a chart of the two competing verb systems. Note that classical Poswa never used passive verbs with 3rd person patients, and that because there was no inner person marker, the differences between SAP and non-SAP patients were clarified by always requiring a noun in the clause (or by context from a previous clause) when the agent was non-SAP.

Note that the newly evolving verbs forms are descended from constructions that had the patient as the outermost vowel.

The words given in both columns are for the classical Poswa form, not the evolved form. Thus the full endings are spelled out.

               CLASSICAL      NEW
1p>2p PRES     (not used)     soppobože      "You feel my answer"             
2p>1p PRES     soppobažo      soppobežo      "I feel your answer"
1p>3p PRES     (not used)     soppoboža      "They feel my answer"
2p>3p PRES     (not used)     soppobeža      "They feel your answer"
3p>1p PRES     soppobažo      soppobažo      "I feel their answer"
3p>2p PRES     soppobaže      soppobaže      "You feel their answer"

1p>2p PAST     (not used)     soppobožul     "You felt my answer"
2p>1p PAST     soppobeži      soppobeži      "I felt your answer"
1p>3p PAST     (not used)     soppobožel     "They felt my answer"
2p>3p PAST     (not used)     soppobežel     "They felt your answer"
3p>1p PAST     soppobeži      soppobaži      "I felt  their answer"
3p>2p PAST     soppobežul     soppobažul     "You felt their answer"

If it is assumed that in word final position /el/ > /ē/ > /e/ and /ul/ > /ui/ > /i/, the forms on the right will be necessary and the outer vowels will no longer reliably indicate person, since a final /-e/ could be either 2nd person present or 3rd person past, and a final /-i/ could be either 1st person past or 2nd person past (and would still have a few relics in the 2nd person imperative). The above chart is reproduced below with hypothetical merged forms inserted. Forms judged less problematic are in capitals:

               CLASSICAL      NEW                                          RESOLVED    TRADL-EXP
1p>2p PRES     (not used)     soppoboze      "You feel my answer"          SOPPOBABO   soppobabo
2p>1p PRES     soppobažo      soppobezo      "I feel your answer"          SOPPOBABE   soppobabe
1p>3p PRES     (not used)     soppoboza      "They feel my answer"         soppoboza   soppobabo 
2p>3p PRES     (not used)     soppobeza      "They feel your answer"       soppobeza   soppobabe
3p>1p PRES     soppobažo      soppobazo      "I feel their answer"         soppobazo   (not used)
3p>2p PRES     soppobaže      soppobaze      "You feel their answer"       soppobaze   (not used)
3p>3p PRES                                                                 SOPPOBABA   (not used)

1p>2p PAST     (not used)     soppobozi     "You felt my answer"           SOPPOBEBI   soppobebi
2p>1p PAST     soppobeži      soppobezi      "I felt your answer"          SOPPOBEZI   soppobebi
1p>3p PAST     (not used)     soppoboze     "They felt my answer"          soppoboze   soppobebi
2p>3p PAST     (not used)     soppobeze     "They felt your answer"        soppobeze   soppobebi
3p>1p PAST     soppobeži      soppobazi      "I felt  their answer"                    (not used)
3p>2p PAST     soppobežul     soppobazi     "You felt their answer"                    (not used)
3p>3p PAST                                                                 SOPPOBEBE   (not used)

It is possible that a new dummy verb /b/ could be engineered to fill the remaining gaps. Alternatively, there is still ambiguity between 2p and 3p in some cases, and as in classical Poswa, the noun must be repeated for context.

BREAK vers

May 9, 2020

               PRES(A)  PRES(P1)     PRES(P2)  PAST(A)  IMP      FUT9     (RESERVED) 
1p INTRANS     -o                              -i       ----     -opi     ----
1p>2p          -abo     -ade        -ode       -ebi     ----     -abopi   ----   
1p>3p          -abo     ----        -oda       -ebi     ----     -abopi   ---- 
2p INTRANS     -e       ----        ----       -y       -i       -epi     ---- 
2p>1p          -abe     -ado        -edo       -eby     -u       -epu
2p>3p          -abe     ----        -eda       -eby              -abepi   ----
3p INTRANS     -a       ----        ----       -e~am    -ap      -api     ----       
3p>1p          ----     -ada        -ado       ----     ----     -apu
3p>2p          ----     -ada        -ade       ----     ----     ----     ----
3p>3p          -aba     ----        -ada       -ebe     -abap    -abapi   ----

Since FUT3 has only two possible forms, it has merged with FUT2 as FUT9. 3p imperative is actually subjunctive and will be moved there when a SUBJ column is added.

SUBJ is probably just PRES(A) + /p/, and SERIAL is PRES(A) + /s/, both as in Lava Poswa.

BREAK

               ACTIVE   PASSIVE1    PASSIVE2
1p>2p PRES     -abo     -aze        -oze  
1p>3p PRES     -abo     ----        -oza
2p>1p PRES     -abe     -azo        -ezo     (possible future tense -epu for this entry only)
2p>3p PRES     -abe     ----        -eza
3p>1p PRES     ----     -aza        -azo
3p>2p PRES     ----     -aza        -aze
3p>3p PRES     -aba     ----        -aza

               ACTIVE   PASSIVE1    PASSIVE2
1p>2p PAST     -ebi     -ezi        -ozi
1p>3p PAST     -ebi     ----        -oze
2p>1p PAST     -eby     -ezi        -ezi              (status of /y/ vs /i/ is unclear for passives)
2p>3p PAST     -eby     ----        -eze
3p>1p PAST     ----     -eze        -azi
3p>2p PAST     ----     -eze        -azi
3p>3p PAST     -ebe     ----        -aze

"I control you"

Possible way to freeze out the "I control you" type of suffix is to shatter the 3x3 system. For example, one could assume that third person agents cannot control SAP's, and so the outermost morpheme cannot be an /-a/. Likewise, one could find any endings in which both vowels were the same to be redundant, and lose the /opo/ and /epe/ forms as well. This would leave just four:

-apo   "I control them"
-epo   "I control you"
-ape   "you control them"
-ope   "you control me"

If Future Poswa has no vowel shifts, this will still present not much of a simplified system unless analogy takes hold such that the inner vowels are mistaken for something else, perhaps tense markers. One possible source of confusion that could lead to reanalysis of these forms as atomic, however, is that by this point the third person past tense suffix will probably be just a simple -e, homophonous with the 2nd person present. This is from classical Poswa -el (surface pronunciation [ei]) > /ee/ > /ē/ > -e.

VERB ENDING MATRIX
     PAST   PRES   IMP
1p   -i     -o     -u* 
2p   -i     -e     -i 
3p   -e     -a     -ob

This chart assumes /ul/ > /ui/ > /i/ at least in final position, that /ub/ > /u/ by analogy, and that /-ae/ is never restored even though it causes conflict with /e/. The /-ob/ will be removed by a change in how the verbs are built.

/ul/>/y/ (that is, unpalatable i) may be more likely.

Note that two /-i/'s were already present in classical Poswa and caused little ambiguity. The use of /-ae/ as the 2nd person present is unlikely; it had vanished in classical Poswa already and would have only been known from poetry. If /-ub/ > /-u/, then /-ob/ might shift to /-o/ by analogy although there is no sound change that could provide this.

Some of the tense markers might be padded with consonants. In particular, /-ob/ could change to /-(ʷ)ap/, which would likely be analogized to just /-ap/. It is even possible that this will merge with the control morphemes up above, making the imperatives and the volition matrix overlap. "Assertive mood" could work for this.

* See above. This is probably a passive.

Imperative vs assertive mood

possible reformed verb ending matrix:

  VERB ENDING MATRIX
       PAST1    PAST2    PRES    IMP    ASSERT    SUBJ   ASSERT2   ASSERT3
1p     -i       -om      -o      -u*    -ope      -op    -opi      ----
2p     -y       -em      -e      -i     -epo      -ep    -epi      -epu
3p     -e       -am      -a      -ap    -ap(V)    -ap    -api      -apu

*Passive.

ASSERT2 may be no more than a generic future tense, but historically it carried the specific meaning of "I will make sure that ...". ASSERT3 is the same, except that it specifies a 1st person patient. Thus /pisitepu/ "you will stick me". Note that this means it is possible to omit the disyllabic transitive verb marker, and in fact that this is descended from the disyllabic marker that coalesced into a single vowel /u/ while others remained in their original form.

The distinction between imperative and assertive could evolve into a politeness distinction, and would share the common trait of the family that the polite forms are actually shorter than the impolite forms. Alternatively it could remain literal, and speakers would use both forms depending on their confidence. The imperative would essentially mean "clean up!" and the assertive would mean "you will clean up (because I can make you do it)".

ASSERT2

It is possible that /-ope -epo/ will shift to /-opi -epi/, under two different analogies; in both cases the speakers change the tense of the outermost ending to the one which causes it to become an /-i/. Then the /-ap(V)/ form would likely be frozen in as /-api/, meaning that it would no longer be possible to distinguish between "I make him clean" and "you make him clean", etc. But it would still be distinct from the imperative "let him clean!" This would effectively make /-pi/ the universal assertive mood marker.

The subjunctive mood is forced onto the chart here because of changes affecting morphemes that it used. There is no longer a distinction between imperative and subjunctive in the third person.

Future tense

Classical Poswa never had a single future tense. There were two methods of expressing it: verbal auxiliaries such as "I plan", "you plan", etc used in conjunction with the imperative, and serial verbs with similar meanings used with the subjunctive, which was unmarked for tense. (That is, "I will walk" was either "let me walk, as I plan" or "I plan that I walk".) A third possibility might have involved verbal infixes.

In Future Poswa, this may continue, but the opening up of a contrast between the imperative and assertive mood may allow the language to express more shades of meaning. Alternatively, a true future tense might arise, but it would very likely have the present tense endings as the outermost markers, not the imperatives or assertives.

Notes