Detail #364: Inverse Case Marking

Something that looks sort of vaguely like this if you squint really hard is attested in real languages, but let's make a more distilled form of it. Consider the most core case distinction in many languages - nominative vs. everything else. Some nouns are more likely to be subjects whenever both a subject and some other NP is present, some are more likely to be objects. We basically get a subjecthood hierarchy, which quite probably is not very unlike an animacy hierarchy.

I imagine that this language has but two cases, and the case of the direct object is also used with adpositions and for any oblique uses. Possession could go any which way, maybe with the subject case being the genitive case as well.

For nouns high up in the hierarchy, the nominative is unmarked, and the accusative is marked. Lower down in the hierarchy, the situation is reversed.

Let us use the same case marker for both and call this the 'INV' case.

I see you.INV
man sees stone
knight fought windmill
windmill.INV killed knight.INV

In intransitive clauses, this offers several options. First a nominative-like approach and an ergative-like approach exist. The nominative approach would have subjects marked the same as in transitive clauses, thus

windmill.INV burns
man falls

 The ergative approach would instead produce this outcome:

windmill burns
man.INV falls

We now need to provide an analysis of what's going on under the hood of this system. The nouns will be given two 'classes' – ones high up in the hierarchy are 'I', and lower ones 'II'.

		I	II
nominative		direct	inverse
accusative		inverse	direct

The ergative situation is similarly

		I	II
ergative		direct	inverse
absolutive		inverse	direct

The final situation we can consider is an underlyingly tripartite system

		I	II
transitive subject		direct	inverse
intransitive subject		?	?
transitive object		inverse	direct

This provides us with many options! We can now introduce a split-S type of thing, and we can even do that in multiple ways. We can let 'inverse' signal lack of volition (or whatever, but that's a good go-to example), or we can let the opposite of the transitive subject marking signal lack of volition. Thus, we may have a situation where:

I.INV fall
signals lack of volition, but
fish jump
signals volition, OR

I.INV fall
signals lack of volition BUT
fish.INV jump
signals volition

Fluid-S is afaict typologically less common than a lexically-conditioned split-S,  but fluid-S is easier to make a toy system out of in just a couple of words. For a lexically-conditioned system, way more options exist, of course: some verbs maybe require direct from all nouns, some require inverse, and some require object marking (thus direct from type II nouns and inverse from type I nouns). Basically, the types of language you can imagine form the following set:

Take as possible verb types the set V, consisting of :
D = verbs requiring direct subjects
I = verbs requiring inverse subjects
O = verbs requiring object-forms out of their subjects
S = verbs requiring subject-forms out of their subjects

The languages we can imagine then forms the set P(V), i.e. the power set of {D,I,O,S}. This is {{}, {D}, {I}, {O}, {S}, {DI}, {DO}, {DS}, {IO}, {IS}, {OS}, {DIO}, {DIS}, {DOS}, {IOS}, {DIOS}}.

A language described by the set {IOS} would then be one that has verbs requiring inverse case subjects, verbs requiring object forms and verbs requiring subject forms.

We can also have a situation where all subjects of intransitive verbs just are direct - i.e. no marking whatsoever.

		I	II
transitive subject		direct	inverse
intransitive subject		direct	direct
transitive object		inverse	direct

A situation where both I and II are inverse for the intransitive subject is conceivable, but unlikely and not significantly different from this situation - it would merely be a 'cosmetic' difference.

We can also have a situation where the cut-off between I and II is different for the intransitive subject than for the other two (heck, one could experiment with having that cut-off  in different spots for subjects and objects too!) 

And of course, we can always add in complications with regards to topics - maybe all topics are unmarked regardless of syntactical role?

Detail #363: A (Double-)Quartet of Impossible Cases

These cases probably do not exist in any language

The cases form a two by two system as far as their features go. Consider the result of subject gaps over coordination in languages with syntactic ergativity:

she hit him and ___ ran
run takes an absolutive argument, him is the absolutive of hit, thus it is implicitly him doing the running.

We could overrule this by a case marking either on the "normally ergative" or "normally absolutive" noun - a form of differential case marking. Thus we get two sorts: the puissant ergative  or the compliant absolutive. "Puissant" because it overpowers the absolutive, "compliant" because it cedes to the ergative. Either one should by its existence make the other one superfluous. However, what other uses could they have?

The puissant ergative could maybe appear with causatives? The compliant absolutive could appear with non-volitional intransitives?

I actually see a potentially possible grammaticalization path for the compliant absolutive: a merger with an absolutive pronoun referring to the ergative subject. If the language has a really weird word order, a similar solution for the puissant ergative could be imaginable. The reason I suspect these to be impossible is that they are nouns telling us information about the next clause's verb, and even more so the compliant absolutive is telling us a non-thing: it's telling us which noun isn't the subject, and isn't even necessarily involved.

The next pair of cases would be the analogous forms for nominative-accusative languages: the compliant nominative and the puissant accusative.

These could have similar origins as those in the ergative scenario. However, one might expect these too to have additional uses, such as the compliant nominative maybe always being the subject of passives.

Another thing that often is influenced by alignment is pronoun binding, esp. wrt reflexives and such. We can imagine puissant ergatives and accusatives to control the reflexive pronouns (out of which possessive reflexive pronouns might be the most interesting), or compliant absolutives and nominatives to cede their control of the reflexive pronouns.

Further thoughts in imaginary typology: would there be a (vague) correlation between puissantness and definiteness? How about compliantness and indefiniteness?

EDIT: four additional cases appear if you consider the situation of erg/abs languages with nom/acc syntax (giving compliant ergatives or puissant absolutives), or the case of nom/acc languages with erg/abs syntax (WHICH PROBABLY DON'T EXIST SO DON'T TRY THIS AT HOME FOLKS) (giving  compliant accusatives or puissant nominatives).

A final, late EDIT:
Obviously, these should probably rather be seen as functions of some certain cases than as cases unto themselves, i.e. you'd expect some other  cases to fill these roles. That of course is also sort of up to how they came about - I imagine a compliant nominative could morphologically be the same as the accusative, for instance.

Detail #362: Differential Case and Inverse Systems

While thinking about the sound changes leading up to Ćwarmin and Ŋʒädär, an uninvited idea entered my mind and turned into this post: what can we do to combine differential subject/object marking and inverse alignment?

The idea might seem somewhat preposterous. Differential subject marking obviously acts on the subject, differential object marking on the object, and inverse alignment doesn't formally distinguish the two in any case-like way.

However, maybe there are ways out, and maybe these can have interesting restrictions on them.

Since I have not mentioned differential case in quite a while, differential case is where subjects (or objects) can take different cases depending on some syntactical or semantic factors.

A few options:

1. mutual marking

Both nouns take the same marker (or allomorphs thereof). This could fit for some Finnish-style differential object marking, i.e. conveying aspectual and/or polarity-related information, but is not suitable for Turkish-style differential marking (where the accusative marks the definiteness of a direct object, whereas the nominative implies indefiniteness).

Jazzing it up: cancel the use of the inverse (or maybe the direct) whenever the Differential Subject&Object Marking is in place. Only let a certain assignment of roles to the two nouns be permissible whenever the marking is there. Here, we can get verb-specific things going: maybe one verb forces inverseness whenever differential marking is in place, maybe another forces directness.

2. marking determined by the hierarchy

One could imagine that the marker goes on whichever noun is lower (or higher) on the hierarchy. Maybe several markers can coexist, and two different ones are permissible, as long as they do not try to go on the same noun. Maybe some markers can go on either noun, and some are restricted to the higher or the lower one. An additional complication could be markers that only exist on some level of the hierarchy - i.e. markers only present with first and second person pronouns, or markers only present for inanimate nouns.

Jazzing it up: have some markers go on 'actual subject' or 'actual object' and some markers follow the above rule. Have a hierarchy among the markers as to which beats which.

3. marking determined by topicality or focus

Whichever element is topic, or potentially focus, could be differentially marked, while the other one goes unmarked.

Jazzing it up: again, change how the verb interacts with things. Make the verb intransitive, and the 'subject' an oblique agent or the object and oblique patient, depending on which one is the topic. This would functionally make the intransitive be both a passive and an antipassive at once. Heck, for double topics, you could have a simultaneous passive-antipassive interpretation.

 

Detail #361: Auxiliaries for Voices

Consider a language where even pretty 'simple' voices like reflexives and reciprocals are expressed using auxiliaries. This provides an interesting space for introducing the distinction between reciprocals and reflexives - number congruence!

If there is some form of person marking possible on infinitives as well, one could imagine having the infinitive take singular marking for reciprocals (or maybe vice versa); however, one could also imagine using just plural verb forms for plural reciprocal subjects, and singular verb forms for plural reflexive subjects.

What if we want the reflexivity or reciprocality not to be with regards to a direct object? Maybe the auxiliaries for voice can stack, and some of them give applicatives and circumstantial voices. Maybe the relevant case-marker (adposition, whatever) can be incorporated in the auxiliary, or turned into an argument of the verb?

Also: UPDATE:

I've been doing really gruelling work on the historical sound changes of Ćwarmin and Ŋʒädär, and will soon post some updates on those. Expect pretty detailed stuff.

Detail #360: Fun With Complementizers

Complementizers appear as heads of all clauses in some theories of syntax. In most theories of syntax, they are also at least the heads of subclauses. The idea in some theories, is that something similar to that in "I knew that she likes Victorian-era comedy" even appears as a null morpheme in the onset (or somewhere else) even of main clauses.

Now, in some languages similar things genuinely appear in some clauses, and I wouldn't be surprised if such a thing even appears in all clauses in some language out there. One common such 'main clause complementizer' is the question marker.

Here appears a thing I've seldom seen conlangers do: force complementizers to appear in certain situations with main clauses, but not in others. Maybe negative clauses require a complementizer, maybe certain kinds of statements require them.

As for 'certain kinds of statements', in Swedish, 'att' (similar to English 'that' as a complementizer) sometimes introduces a clause (whose word order then is like that of subclauses), without any main clause, where the statement expresses disdain, admiration or agreement for a fact thus stated:

att han törs!(that) he dares!
how dare he?

att hon gör!(that) she does!
she sure does!

Using complementizers occasionally or regularly for main clauses can be an interesting way of enriching one's syntax as well. One hypothesis regarding verb fronting as a way of marking questions is that the verb actually moves to the zero morpheme question complementizer, and thus is a sort of realization of that complementizer. This of course changes details in the word order. One could have the language sometimes force the subject or object into the C position, and this would change other word order details. Maybe moving the subj to the C position breaks reflexive binding? Maybe it breaks verb congruence? Maybe moving the object breaks transitivity, making a transitive subject marked absolutive (if the language is ergative).

Of course, the presence of an explicit, non-zero complementizer could, as in Swedish, force subclause word order, if there is a difference between these in the language. Thus maybe all negative clauses have subclause word order?

Detail #359: An Ergative Language with Phrasal Verbs

In English, there are a bunch of object-like nouns (that under some analyses are objects), that are marked by prepositions. These occur with a variety of verbs, a handful of these could be, for instance

wait for
hope for
look at
listen to

Now, we could imagine a similar thing in a syntactically ergative language. In a syntactically ergative language, the syntactical subject is the absolutive argument, though, and we could imagine a situation whereby the ergative argument sometimes would be marked by some other case or some adposition. I am not sure whether this should even be considered quirky case 'subjects' (or 'quirky case ergatives', since the ergative is unlike both 'subjects' and 'objects' in this style of analysis.)

Here's a bit of a challenge though: think up some "ergatively phrasal verbs" that seem as natural as 'wait for' or 'look at' or 'think about'.

Detail #358: A Congruence System with some Quirks

For certain kinds of congruence, there are two main types we can consider:

• morphosyntactic congruence
• semantic congruence

The former agrees in case, gender, etc, the latter agrees in some way with the meaning. We find English having a split on this with regards to examples like 'the family is' vs. 'the family are'. Different speaker communities do not agree on which one of these are right, and some may even use both with some subtle meaning differences. However, let's make up some more interesting thing here, such as a rule that tells us when the congruence is semantic and when it's morphosyntactical.

One idea could be that NP-internally, agreement always is morphosyntactical. (I will go and revise this later with regards to participles!) We could also go and say that VP-internally, the agreement is always semantic. However, the subject not being part of the VP, subject agreement on the verb might be exceptional - I'll go with morphosyntactical here. (Here, I am rather agnostic as to which way is most likely in a natural human language; heck, I find myself conflicted on whichever way VP-internal or NP-internal is more likely to go). For the language I am envisioning, the verb also has object congruence.

So, now we have a system where

the family sold-subj:3sg-obj:3pl the flock

We may of course have some gender congruence adhering to this pattern. Now, we may also have a complication with regards to quirky subjects and objects, or oblique ones: the subject might get a third person ('neutral') marking regardless of the subject's person, number or gender, while a primary 'semantic' object that is marked obliquely in some sense still might get some form of object marking on the verb.

Another complication we can introduce is with regards to left-dislocated objects - regardless if it's due to focus fronting or topicalization or whatever other thing, the object may then be considered outside of the VP, and the gap left behind now might not cause congruence.

Participles obviously have features both of adjectives and of verbs. There, passive participles could take semantic agreement, active participles morphosyntactical agreement.

Here, however, we get a lot of possibilities for sliding scales of marking, and this whole notion could be a nice testing ground for looking at probabilistic approaches for grammars: maybe, just maybe, we could build a system where the probability in some context for one kind of congruence is P, and for the other it's 1-P, where P is a real number in the range [0,1]. In different circumstances, the probabilities differ: subject marking on finite verbs has probability P_subj, object marking has P_obj, a left-dislocated object has P_l-obj, adjectives in attributive position have P_attr, and so on. These may further have a hierarchy where a change in the probability of one might force the probability of another to change.