Wednesday, December 20, 2017

Sargaĺk Cases as Bundles of Features

In this post I try, in retrospect, to analyze the case system of Sargaĺk in terms of features. Unlike the post on the case system of Ŋʒädär, a far share of actual usage examples will also be presented. Warning: this is still somewhat waffly.

The case system consists of the pegative-genitive, the absolutive (with three subcases that differ on syntactical grounds, and one subcase that differs for some pronouns), the comitative-instrumental, the familiar comitative, the locative, the ablative and the lative.

This adds up to a set of maximally eight morphologically distinct cases, maximally ten syntactically distinct cases but minimally six morphologically distinct cases. Each of these sets are worth investigating separately, as is the use of cases with regards to time.

Temporal usages of the locative cases and the absolutive case are the easiest to look into, since the dimensions are rather 'easy' to come up with: punctual vs. time-span, start-point, end-point, location in time vs. amount of time.
The cases that are used in expressions of time are the absolutive, the locative, the ablative, the lative and the comitative-instrumental.

We begin with the ten syntactically distinct cases:
  • ditransitive subject / genitive
  • transitive subject
  • intransitive subject
  • direct object
  • indirect object
  • comitative-instrumental
  • comitative-familiar
  • locative
  • lative
  • ablative
The three additional locative cases that appear on some pronouns will not be dealt with.

An immediate quick partitioning of these might hint at sets along these lines: {{the two transitive subjects, intransitive subject?}, {objects, intransitive subject?}}, {comitatives}, {locatives}.
The number of binary features needed to distinguish ten is 4, or rather 3.3..., while the number of trinary features is 2.1... Since we basically need quite a bit "too much" space no matter what way we do, I will not use them very optimally. Control and direction will both be binary, and I will get back to what they even signify in a bit. The third I'll call 'involvement', and I will give that one three values - active, passive, frame.

dit. subj.++a
tra. subj.+-a
intr. subj.+-a
dir. obj.--p
ind. obj.++p

The lative case is interesting in that in disdainful utterances with verbs of location (rather than verbs of movement), it can be used with a locative meaning, i.e. 'hither I sit' would signify something along the lines of 'bollocks, here I sit again'.
The comitative-instrumental, likewise, takes on a disdainful meaning when used with nouns for which comitative-familiar is expected. These are very limited uses, but they are common.

It turns out approaching the maximal form does not provide any very neat-looking decomposition into features. The eight morphologically distinct cases might be a bit more promising, given that 8 is 23, three features could exactly account for them. We now deal with absolutive, accusative, pegative, comitative-instrumental, comitative-familiar, locative, lative and ablative.
Here, we'll parse the feature 'direction' not as one of physical movement towards, but rather as 'involving' even any sense of metaphorical direction.
"Activity" is a conflated "involvement"-version.


The minimal set of distinct forms occur with inanimate nouns, and most animals (the comitative-familiar is attested with some pets and some anthropomorphized animals in mythology). These nouns distinguish absolutive, pegative, comitative(-instrumental), locative, lative and ablative. This set itself is of some interest, since we also know that the comitative-familiar can be replaced in all positions by the comitative-instrumental for all nouns that distinguish the two, so a more coarse system of features could leave out the comitative-familiar entirely. We now have two options: three features, and two cases get to cover two of the combinations - alternatively one covers three combinations.

We could put the locative ones in one bag, and the three other ones in the other - getting us 'grammatical vs. locational/oblique', and distinguish them by some trinary feature that seems to line up - maybe having the locative and the absolutive, the pegative and the ablative, and the lative and the comitative-instrumental paired up sharing the value of the second feature. This, however, seems somewhat weird. Certainly pegatives and ablatives are slightly similar, both being in some sense 'origins', and locatives and absolutives are sort of similar in 'intrinsicness' to an event. But comitatives and latives do not seem very similar at all, unless we create some kind of wastebin category. I decided to call this wastebin category 'indirect'.


This would seem contrived if it were not for this actually appearing as a pattern in some parts of the language:
  • With passives, where nouns in the three upper cases can be demoted into the three lower cases. This is not very common with the comitative, but nevertheless attested.
  • Adnominally, you get a clear pegative-ablative correlation: animate possessors tend to be pegative, inanimate tend to be ablative.
  • With expressions of time, specific times are given in the absolutive or locative (depending partially on the type of time - named times are absolutive, generic nouns may be locative, partially on whether it's a complement or adjunct, with adjuncts being locative). A similar pattern occurs with time spans but with comitative and lative marking the end of the time-span and the pegative or ablative marking the onset.
  •  A handful of verbs can take  comitative and lative arguments. Among these we find narol, share. One can share 'with' or 'to', where the distinction seems to be one of volition on the part of the subject. 
  • Similarly, a few verbs permit for a similar alternation among direct objects or subjects: ırsal, 'to reach' where the case on the object seems to mainly correlate with some kind of aspectual notion, i.e. absolutive objects indicate arrival, locative objects indicate physical length or habitual arrival, altul 'to embrace, to contain', where absolutive subjects indicate embracing, and locative subjects indicate containing.
However, a different set of patterns fit a different analysis, where a set of binary features appear, but the absolutive and locative take an indeterminate/ambiguous/superimposed/ignored value. Mostly this is a very similar analysis to the previous one, giving similar pairings - pegative-ablative, comitative-lative, and absolutive-locative. The table below does not collapse absolutive and locative onto the same row, however, but uses an analysis where absolutive is 'rather central' and locative is rather 'peripheral'. The terminology here is not very clearly defined, but 'central' vs. 'peripheral' has two independent possible significances: it indicates canonical uses of verbs and adpositions, peripheral signifies less typical such usages; meanwhile, 'central' also signifies topicality, direct physical involvement and so on, whereas 'peripheral' either signifies lesser significance to the topic under discussion or less physical involvement.




The decision to analyse the system like this relates to the fact that we sometimes find patterns where the pegative and comitative operate as a pair of complements, as do the ablative and lative; the absolutive and locative, however, also operate as such a pair as well, so whereas the other cases each have relations to two cases, the absolutive and locative occupy both those spots for each other.

Such examples include certain adpositions (nup, 'under' with the ablative, 'covered by' with the lative; int, 'among' with the pegative, 'surrounded by' with the comitative - generally, the peripheral case for these provides a more 'specific', less general meaning, thus less 'central' to the meaning-space of possible meanings).

The description given here cover the dialects closest to the mainland. Further off, we find systems where the absolutive-locative part of the above diagram have been rearranged, so that the locative covers every part except the grammatical-central slot.

No comments:

Post a Comment