Grammaticalization Quickie Thread

[bradrn]

What's a more succinct way to say "the boundary between two morphemes not bound to each other"? I don't want to use the term "word boundary" because I want to allow for compound words containing more than one free morpheme, and I'd also like some words to be clitics and behave more like bound than free morphemes.

Well, firstly, you’ll have to define exactly what you mean by ‘bound morpheme’, because that’s a pretty ambiguous term. (I’d say most clitics behave more like free morphemes, for instance. Are ‘the’ //ðə=// or ‘a(n)’ //ə=// bound, for instance?) But also, what context will you be using this in? If you’re talking about morphophonology, for instance, I’d call it a ‘juncture’ (e.g. Pawley’s Kalam grammar distinguishes ‘non-final juncture’, ‘final juncture’, ‘question juncture’, ‘exclamation juncture’), but a different term may be better in a different context.

[StrangerCoug]

Well, firstly, you’ll have to define exactly what you mean by ‘bound morpheme’, because that’s a pretty ambiguous term.

In general you can substitute "free morpheme" with "root" and "bound morpheme" by "affix" and still get at what I'm trying to say.

(I’d say most clitics behave more like free morphemes, for instance. Are ‘the’ //ðə=// or ‘a(n)’ //ə=// bound, for instance?)

You do bring up an interesting point. Let's call a clitic a free morpheme with the phonological behavior of a bound morpheme for the purposes of this discussion. (I plan for it to have the orthographical behavior of a bound morpheme, too, but the orthography is not fully developed and I haven't addressed how to deal with clitics in my English transcription, but am leaning toward using a hyphen—it's free since I'm using an interpunct to break up two letters I don't want to be read as a digraph.)

But also, what context will you be using this in? If you’re talking about morphophonology, for instance, I’d call it a ‘juncture’ (e.g. Pawley’s Kalam grammar distinguishes ‘non-final juncture’, ‘final juncture’, ‘question juncture’, ‘exclamation juncture’), but a different term may be better in a different context.

Right now I'm trying to describe what's going on in a morphophonological context. My conlang has sandhi processes that deal with what happens when difficult-to-pronounce consonant sequences occur across syllable breaks, and I want to make clear when the sandhi rules do and don't apply.

[dɮ the phoneme]

could a possessive preposition cliticize onto the preceding noun and eventually be reanalyzed as a sort of construct-state suffix?

[bradrn]

Kwakʼwala has ‘wrong-way’ clitics for definiteness and possession:

kwʼix̣ʔid=ida bəgwanəma=x̣-a qʼasa=s=is tʼəlwagwayu
clubbed=the man=OBJ=the otter=INSTR=his club

The man clubbed the otter with his club

So I’d say that definitely sounds plausible.

[Rounin Ryuuji]

could a possessive preposition cliticize onto the preceding noun and eventually be reanalyzed as a sort of construct-state suffix?

English "of" seems to be in the process of being absorbed into preceding words, often reduced to -a or -o' ("heap of pancakes" might sound, when I'm speaking quickly, like "heapa pancakes" ['çij.pə 'pænˌceiks]); in such a way this reverses the usual order, which would usually be "pancake heap"; similar, with person, "Marta's mother", "The mother of Marta" (though this form sounds stilted, and is more likely in older literature than ordinary speech). From there, you could get "heapa" and "mothera" as possessum forms (however with named animate possessors being rather marked); from there, you get something very close to the possibility of calquing Arabic Malikat Saba as Queena Sheba.

[Vijay]

Doesn't something like this happen in Japanese?

[Rounin Ryuuji]

As far as I know, Japanese の "no" and the archaic genitives が "ga", and つ "tsu", all just sit after the nominal; I can't think of any instances where the order is reversed.

[Vijay]

But it's still cliticizing onto the preceding noun, isn't it?

[Moose-tache]

Mr/Ms The Phoneme specifically asked about prepositions. You're right that Japanese puts no right after the possessor noun, but this is pretty much standard practice for most of the world's head-final languages. The idea of a preposition doing this is a lot weirder. In the examples above it sort of happens in English in partitives and some other circumstances, making a kind of possessee clitic (essentially the reverse of Japanese no). But there is no productive possessee clitic so far as I'm aware in English (I can't say "The importencsa brushing your teeth" to myself without sounding stupid), and unlike clitics like 's, every example is perfectly reversible except for a few lexicalized terms like "man o' war."

[Vijay]

Oh, sorry, yes, they did say prepositions specifically. I don't think that's so weird, though. Romanian does this with the preposition în, e.g. cu noaptea-n cap 'very early in the morning' (cu 'with' + noaptea 'the night' + în 'on' + cap '(my) head'), trandafir cu creanga-n apă 'rose branch touching the water' (trandafir 'rose(bush)' + cu 'with' + creanga 'the [its] branch' + în 'in' + apă 'water'), ies afară-n bătătură 'I go outside the courtyard' (ies 'I go out, exit' + afară 'outside (of) + în (?) + bătătură 'courtyard(?)').

[dɮ the phoneme]

Ok, this does at least seem possible then, but I'm still not sure what kind of pressures would cause this to happen (as opposed to clitisizing the "normal" way).

On a related note, how do construct-state type systems for marking possession tend to develop, anyway?

On an unrelated note: how did German end up with its characteristic syntax, such that underlying order is SOV but the verb or auxiliary in the highest position is raised to a right-branching TP head (or however you want to formalize it)? I've wanted to make a conlang that features an even more extreme version of this for a while, in which the word order is VSO in simple clauses and AuxSOV in the presence of an auxiliary. X-Bar syntax lets you do this (just do what German does but don't raise the subject from Spec-VP to Spec-TP or whatever), but I'm skeptical of syntactic theories in general, so since I've never seen this setup I'm hesitant to use it. I'd presumably be more comfortable if I had a sense of the potential diachronics involved. Anyone know any articles on historical German syntax?

[bradrn]

On an unrelated note: how did German end up with its characteristic syntax, such that underlying order is SOV but the verb or auxiliary in the highest position is raised to a right-branching TP head (or however you want to formalize it)?

You may be interested in this paper: https://www.jstor.org/stable/416416. The argument seems to be that verbs are sort of like second-position clitics, and thus end up in second position themselves.

[Raphael]

Not sure about this - I'd say that in a very simple sentence, such as "Ich esse gerade Kuchen" ("I'm eating cake right now"), German has SVO, and it "only" turns SOV when things get even the slightest bit more complicated.

[dɮ the phoneme]

Not sure about this - I'd say that in a very simple sentence, such as "Ich esse gerade Kuchen" ("I'm eating cake right now"), German has SVO, and it "only" turns SOV when things get even the slightest bit more complicated.

Agreed. When I say "underlying word order", I mean word order prior to movement, in the generative sense. So IIRC a sentence like that brackets as ich [TP esse_i gerade [VP t_i [NP Kuchen ] ] ], where the t_i is a trace representing the location esse has moved from. So German has a head final VP structure but a head-initial TP structure, and there is a rule that moves the structurally highest verb or auxiliary out of its position in the VP and into T-head position. This means that in the absence of an auxiliary, word order is SVO, and in the presence of one it's SAuxOV.

The situation I'm trying to get for my conlang is the same, except instead of SVO I want VSO, and correspondingly instead of SAuxOV I want AuxSOV.

[StrangerCoug]

So I've got a language (which I'm calling Swiftriver in English for at least now) that starts as a split-P language, where the object of monotransitive verbs is declined in the accusative when the verb is dynamic and in the dative when the verb is stative. This was inspired by the Wechselpräpositionen of German, and indeed Swiftriver has that class of adpositions too except that Swiftriver is postpositional instead of prepositional. We'll call Swiftriver's version of Wechselpräpositionen "changing postpositions" in English.

I want Swiftriver to undergo a change (as part of the transition from Old Swiftriver to Classical Swiftriver) where the object of monotransitive verbs is now always in the accusative regardless of whether the verb is dynamic or stative, yet its formal form keeps its class of changing postpositions and there are still clearly distinct accusative and dative forms. What might trigger this change?

Edit: See two posts below and my response in the post below it.

[Rounin Ryuuji]

The easiest one might be alignment with the usage in a substratum or superstratum language, or an areal feature?

[bradrn]

So I've got a language (which I'm calling Swiftriver in English for at least now) that starts as a split-D language, where the object of monotransitive verbs is declined in the accusative when the verb is dynamic and in the dative when the verb is stative. This was inspired by the Wechselpräpositionen of German, and indeed Swiftriver has that class of adpositions too except that Swiftriver is postpositional instead of prepositional. We'll call Swiftriver's version of Wechselpräpositionen "changing postpositions" in English.

I’ve never come across the term ‘split-D’ before. Where did you see it? I know such languages as ‘split intransitive’.

I want Swiftriver to undergo a change (as part of the transition from Old Swiftriver to Classical Swiftriver) where the object of monotransitive verbs is now always in the accusative regardless of whether the verb is dynamic or stative, yet its formal form keeps its class of changing postpositions and there are still clearly distinct accusative and dative forms. What might trigger this change?

I don’t see any reason why it would need a specific trigger — cases expanding their range are very common cross-linguistically.

[StrangerCoug]

So I've got a language (which I'm calling Swiftriver in English for at least now) that starts as a split-D language, where the object of monotransitive verbs is declined in the accusative when the verb is dynamic and in the dative when the verb is stative. This was inspired by the Wechselpräpositionen of German, and indeed Swiftriver has that class of adpositions too except that Swiftriver is postpositional instead of prepositional. We'll call Swiftriver's version of Wechselpräpositionen "changing postpositions" in English.

I’ve never come across the term ‘split-D’ before. Where did you see it? I know such languages as ‘split intransitive’.

It's my miscopy of split-P from here; I've fixed it in my scratchpad. Swiftriver is not split-intransitive—the sole argument of an intransitive verb is always declined in the nominative in this one.

[bradrn]

So I've got a language (which I'm calling Swiftriver in English for at least now) that starts as a split-D language, where the object of monotransitive verbs is declined in the accusative when the verb is dynamic and in the dative when the verb is stative. This was inspired by the Wechselpräpositionen of German, and indeed Swiftriver has that class of adpositions too except that Swiftriver is postpositional instead of prepositional. We'll call Swiftriver's version of Wechselpräpositionen "changing postpositions" in English.

I’ve never come across the term ‘split-D’ before. Where did you see it? I know such languages as ‘split intransitive’.

It's my miscopy of split-P from here; I've fixed it in my scratchpad. Swiftriver is not split-intransitive—the sole argument of an intransitive verb is always declined in the nominative in this one.

Whoops, you’re right — I misread ‘monotransitive’ as ‘intransitive’.

[Richard W]

I typed the example in English because I dont actually know a language that uses person marking inside of a relative clause.

English can. In mine, it generally agrees in person with an antecedent that is a nominative pronoun. From prescriptive statements that the relative pronoun agrees in person with its antecedent, Latin or Greek must also do so.