Wednesday, March 20, 2024

Detail #439: Quasi-parts of speech and quasi-constituents and the Locus of the Abessive

Different languages are analyzed with different sets of parts of speech and constituents/parts of sentences. To some extent these boil down to grammatical traditions, but to a great extent they also boil down to actual grammatical phenomena. However, if we were to compare, say, Swedish and English, the differences are largely superficial: both descriptive traditions work with barely any modifications to describe either language. A trivial example: Swedish grammar tradition has "subjunctions", which in the English tradition are subsumed under "conjunctions". Subjunctions are words that introduce subclauses, i.e. subordinating conjunctions. A description of English would be marginally different if this concept was introduced.

In some languages, however, distinguishing adjectives from nouns - or adjectives from verbs in some other languages - makes way less sense. Applying such a distinction would be looking at it through a decidedly foreign lens. Sometimes, which part of speech a word belongs to is hard to pinpoint: a word may be both a verb and an adverb, or a noun and an adverb, or a noun and an adjective, etc.

A thing that interests me, however, are a variety of ways in which constituents and parts of speech may behave in ways that justify considering them some kind of quasi-PoS or quasi-constituent, constituents that show some kind of uniformity, but cut across other constituents.

One such example I have been sketching over recent months is what I chose to call "the locus of the abessive". This locus is marked by a certain case (which however also is used for some other constituents), but can appear as subjects, objects, indirect objects, possessors, possessums and other constituents. The abessive itself could also be considered a type of quasi-constituent.

The syntax of this entity gets complicated. First of all, the locus can be any of the following:

  • topic
  • subject: I miss her
  • object: he deprived them of shelter
  • indirect object: they gave him no food
  • location: there is no joy in Aylesbury
  • possessor: the orphan's mom
  • possessum: the man's widow
  • infinitives of various functions: there's no reason to hate him
What if the locus of the abessive could be coordinated over gaps, even when it isn't the same role? 'They deprived him of shelter and gave no food' would then have 'him' as the indirect object of 'gave'. This could get really tricky once possessors and possessums and infinitives start getting involved.

 


Tuesday, March 19, 2024

About Bryatesle and its relatives

About Bryatesle and its relatives

Bryatesle is a dialect continuum encompassing about 30 million native speakers. It is the lingua franca of about 40 million additional speakers, has a rich literature - fictional, scientific, philosophic, religious and instructional, in both poetic and prose forms - liturgies in multiple religions, songs, humor, word games, and a variety of other linguistic devices. Dialectal differences are sufficient that several ends of the dialect continuum are not mutually intelligible, but the standard forms of the language provide common ground for scholars, businessmen, clergymen, government officials and regular people.

Short history

The pre-historic connections between Bryatesle and its kin languages indicate a rather sudden expansion from the Dairwueh-Bryatesle urheimat about 3800 years before the present year (bpy), after a previous split from Sargalk.

Bryatesle tribes started forming city states sometimes roughly 1800 bpy, at which time also writing systems were adopted from Tatediem cultures to the south. Maritime and fluvial trade networks led to about 40 city states forming around the sea of Sadgal and the sea of Gudnyt as well as the great lake Pajik over the next 800 years. Expansion both east- and northwards included assimilation of some Cwarminoid and Tatediem populations. Westwards, Dairwueh tribes partially resisted Bryatesle expansion, partially stood as equal partners in trade and industry, partially stood in political unity, partially expanded onto Bryatesle areas. Several westwards city states in fact were bilingual, and conflicts were not necessarily as much between Dairwueh and Bryatesle, but rather between distinct alliances of dairwobryatesle states.

About 1200 years ago, a stronger political unity over the dairwobryatesle world emerged, with  Ykred emerging as a capital of sorts. This unity lasted 300 years, but after 200 years of dissolution and strife, the last 700 years have seen a somewhat less centralized, but still united dairwobryatesle world, now under the domination of the city-state Sţesar. Although some consolidation of the various standard Bryatesle dialects has occurred due to improved communications, significant differences persist.

Geoculturopolitically, there are occasional confrontations with the Ćwarmin civilization to the east. In the far east, the Ŋʒädär have some trade relations with the Dairwuobryatesle. Southwards, we find the Tatediem engaging in trade, diplomacy, proselytization and sometimes war.

Related languages

The Sargalk-Bryatesle-Dairwueh family consists of the following languages and major dialects (italicized)

  • The Hefnarač-Sargaĺk Branch
    • The Sargaĺk branch
      • Sargaĺk
        • northern Sargaĺk
      • Inraj Sargaĺk
      • Geʔamik †
      • Tudiluk †
    • The Hefnarač branch
      • Hefnarač
      • Sindeʔʔet †
      • Bidlahʔa †
  • The Rilgouz branch
    • Simiz †
    • Rilgouz
    • The Adrk languages
      • Adrk
      • Tarts †
      • Vimil
  • The Dairwueh-Bryatesle Branch
    • The Dairwueh Branch
      • Dairwueh
        • Western
        • Central
      • Bundur
      • Vist †
      • Kappeuje †
    • The Nerazg Branch †
    • The Bryatesle Branch
      • Northern
        • Bryatesle
        • Western Tarist
      • Southern
        • Tarist
        • Kurelwai †
      • Trinzlye †

The Hefnarač-Sargaĺk Branch

The Hefnarač-Sargaĺk branch consists of three extant languages, Hefnarač, Sargaĺk and Inraj Sargaĺk. , Inraj Sargaĺk is moribund, with about 500 speakers. Hefnarač and Sargaĺk each have about 20 000 speakers, but language change towards Dairwueh and Cwarmin are weakening them both. More extinct languages are hinted at from old sources. Geʔamik, Tudiluk, Sindeʔʔet and Bidlahʔa have all gone extinct during the last 100 years. Small word lists for about a dozen other languages that probably were also related have been compiled by scholars and missionaries. The time-depth of the relation between Hefnarač and Sargaĺk is probably on the order of 3500 years or more. The most recent common ancestor between Hefnarač-Sargaĺk and Dairwueh-Bryatesle is probably about 4500 years ago or more.

The Rilgouz Branch
The Rilgouz languages are spoken on islands west of the main Dairwueh lands. The total number of Rilgouz speakers probably is about two million, with Adrk and Vimil having about 3000 each. Whether these languages diverged earlier or later than the HS/BD split is unclear, and even then it is a bit unclear which branch they diverge from: there are isoglosses that pair any two of the three branches, exclude the third - both for sound changes, semantic changes and grammar changes. There even are lexemes that single-handedly occupy conflicting isoglosses.

The Dairwueh branch

An overwhelming amount of shared innovations indicate that the Dairwueh and Bryatesle branches are closely related, having diverged at most 3500 years ago. Dairwueh and Bundur separated about 1500 years ago, Vist and Kappeuje were arguably divergent dialects that have since merged into Dairwueh leaving substratal traces.

The Nerazg branch

A few moribund languages with clearly para-Dairwueh/para-Bryatesle features have been assigned into this family. Research is ongoing, but the evidence is unclear.
 
The Bryatesle branch
Bryatesle's nearest living relative is Western Tarist, which diverged under the last 1000 years or so. Trinzlye and the southern branch encompassing Tarist and Kurelwai diverged about 1500 years ago, only to see Kurelwai mostly assimilate back into Tarist (but also to some extent into Bryatesle).

Trinzlye was largely assimilated into Dairwueh, where it has left some traits as well.

Arguably, there is a dialect continuum between Bryatesle and Western Tarist.

Monday, March 18, 2024

Detail #438: Split Marked Nominative

Split Marked Nominative is a phrase that struck me out of the blue, and I felt like it needs a post. However, let's begin by looking at something slightly related.

We'll start, as usual, with Finnish. Because that's where some pretty crazy stuff can be found. Finnish, in some sense, has a "split marked accusative" system. But first, we need to disentangle the Finnish differential object and subject system:
Finnish marks existential subjects (but also objects) by the partitive. Existential verbs tend to be intransitive, so this doesn't affect the object marking much.
Finnish marks atelicity or negativity by having the object in the partitive case. Thus, only telic, positive verb phrases have the object in the accusative.
Now on to the "split marked accusative". First: plural accusatives and pronominal accusatives have no split: -t all the way (for nouns, -t is the nom/acc plural marker, for pronouns, -t is the accusative marker). For singulars, however, if the verb licenses a nominative subject, the noun is in the marked accusative (identical to the genitive). If the verb licenses no nominative subject, however, it is in the unmarked accusative. (Certain auxiliaries require a genitive subject, and e.g. the passive has no subject in Finnish - the object isn't raised to subject. Also, imperatives license no nominative subject.)
 
Now, on to the split marked nominative. In case the atypical constructions require the nominative marking, I think it would be better to describe it by some other term - e.g. quirky case. However, if the nominative marker is present in standard transitive and intransitive clauses, and only get dropped in some contexts, calling it "split marked nominative" makes more sense. If e.g. pronouns keep their nominative in all contexts, and maybe some other markers (optional quantifiers, demonstratives, etc) also signal nominative, this should be good.

So, now, where does the nominative go less marked or unmarked?
  • subclauses
  • with certain auxiliaries
  • (negative) existential statements?
  • with certain types of subjects? (E.g. proper nouns or mass nouns or something?)
  • Certain TAMs?

How would a system like this come about? I guess a simple grammaticalization path would be "degenerate ergativity".