Grammaticalization Quickie Thread

[Travis B.]

Typically more frequently used forms are less subject to analogy, hence that very frequently used "irregular" forms of words are likely to survive when less commonly used forms, which are likely to be subject to analogy in the long run.

[StrangerCoug]

OK, I think I'm going to need some help here. Here's what the slots look like for the finite forms of the verbs:

Every combination of the above is theoretically legal—all 768. Only the root slot (obviously) and the person slot have to be filled; if the slots to their right are not filled, it's treated as the first entry under them.

I should explain the number system of this language: A singular/plural distinction is made for count nouns, while a collective/singulative distinction is made for mass nouns. The singulative of a mass noun can be pluralized like a count noun. The plural marking in the verbs is morphological, not semantical—you don't have to mark a noun as plural (and if you don't for the subject, you can't make the verb plural). You do, however, have to mark when a mass noun is singulative. I've given serious thought of having more than one of my daughter languages abandon the singular/plural distinction leaving the collective/singulative distinction, as a matter of fact.

For the nonfinite verb forms, it simplifies to this (here, all the slots behave the same—the infinitive is represented as a blank in the first slot)

Here, NOT every combination is legal even in theory: you can only fill the tense slot when it's a gerund I meant participle. That leaves only 64 finite forms instead of the expected 96.

Right now I'm only able to think of two simplifications besides the above:

1. My conculture going "We know what number the first-person inclusive's supposed to be, you don't have to mark it" (the most obvious to me).
2. Blur the tense/aspect distinction a little (not to typical SAE levels), especially as regards future/prospective.

What other distinctions do you see going away/being analyzed on the way toward a fusional language?

Edit: Fixed the charts; see two posts down.
Edit #2: Fixed another disparity between my post and my file.

[Travis B.]

These are only minor reductions in the complexity of the above, but some things to do is to reduce the complexity of the numbers distinguished by the verb, i.e. treat the singulative singular as singular and the singulative plural as plural, so the verb only has to distinguish two numbers, singular/collective/singulative singular and plural/singulative plural, to not distinguish all combinations of tensel. aspect, and mood, e.g. eliminate the present perfective, and not distinguish tense or aspect in the irrealis, and to not mark the negative on the verb, and instead mark it with an adverbs and other negative words.

[StrangerCoug]

I accidentally reversed the imperfective and perfective in the charts (it's the latter, not the former, that's unmarked)—will get to them—but I like your ideas for the most part.

I thought of a further simplification: Is it really that useful to distinguish aspect in the infinitive? I don't see many cases where I'd want to—but then again, I'm still developing this conlang and I haven't tried writing in it.

[hwhatting]

A few remarks:
1) 1st inclusive / exclusive makes only sense for non-singular, as the distinction implies a group.
2) Can you explain what the prospective / retrospective do? If they mean what the labels say and are distinct from future / past, they're relative tenses, not aspects, and should be either in the tense slot or have their own slot.

Is it really that useful to distinguish aspect in the infinitive?

The Slavic languages or Greek (Aorist stem infinitve versus present stem Infinitive), e.g., distinguish aspects in the Infinitive.

e.g. eliminate the present perfective, and not distinguish tense or aspect in the irrealis,

Again, languages with a pervasive aspect distinction distinguish aspect in all of these (Slavic) or most of these (no "aorist" present in Greek). A better route to go down would be to limit the distinction to the past tense.

On the whole, the number of categories doesn't look excessive to me, although I have questions on some of them (see above). Out of "frequently found" categories, you don't have an imperative (or similar - exhortative, optative) and you don't have a passive voice. As long as most of the forms are transparent, I don't think speakers will have a problem with keeping track.

[mèþru]

I plan to have a language family where each branch develops their own, very different gender system (I use gender in the sense of any noun class):
Can a language develop a system of infixed genders? The way I envisioned it, the infixes are marked on the first and last word of a NP. If the last word of the NP is a verb, adverb, clitic or particle, then the gender is marked on an auxiliary instead (none of those can occur in the beginning of a NP).
If this system is possible, can it be borrowed into another language (in this case, a close relative) with an already existing gender system? The other gender system uses fusional case-gender-number suffixes on nouns, adjectives and pronouns. It has only three genders, while the infixes have much more (although most of these can be identified as subgenders, such as human feminine is subgender of human, which in turn is a subgender of the animate category, which covers all things perceived to be living (other categories include insects, four-legged things and plants)).

[StrangerCoug]

1) 1st inclusive / exclusive makes only sense for non-singular, as the distinction implies a group.

Kinda what I was getting at with the simplification.

2) Can you explain what the prospective / retrospective do? If they mean what the labels say and are distinct from future / past, they're relative tenses, not aspects, and should be either in the tense slot or have their own slot.

Retrospective: Because I didn't want to use the term "perfect" and create confusion with the perfective, and I want it to work similarly to English Prospective would be the opposite of retrospective, as expected.

That said, I do not want to combine retrospective and prospectives together with the absolute tenses, and I want both of them marked. If they get their own slot, what do I call "not retrospective and not prospective"?

[hwhatting]

Retrospective: Because I didn't want to use the term "perfect" and create confusion with the perfective, and I want it to work similarly to English Perfective would be the opposite of retrospective, as expected.

Sorry, I don't understand what you're saying here.

[mèþru]

Retrospective is what they calls "perfect" (They did not invent this usage, it is mentioned in the wikipedia article on the perfect aspect).

[hwhatting]

Retrospective is what they calls "perfect" (They did not invent this usage, it is mentioned in the wikipedia article on the perfect aspect).

I got that, but in that case "Perfective would be the opposite of retrospective, as expected" doesn't make sense.

[mèþru]

I didn't see that part. Now I am also confused.

[Pole, the]

Retrospective is what they calls "perfect" (They did not invent this usage, it is mentioned in the wikipedia article on the perfect aspect).

I got that, but in that case "Perfective would be the opposite of retrospective, as expected" doesn't make sense.

It makes. You just have to realize it's actually a typo / Freudian slip and that should have been "prospective" instead of "perfective".

(Which is obvious if you read it in the context.)

[StrangerCoug]

Retrospective is what they calls "perfect" (They did not invent this usage, it is mentioned in the wikipedia article on the perfect aspect).

I got that, but in that case "Perfective would be the opposite of retrospective, as expected" doesn't make sense.

It makes. You just have to realize it's actually a typo / Freudian slip and that should have been "prospective" instead of "perfective".

(Which is obvious if you read it in the context.)

Damn same first letter... This is what I meant, and I fixed it in my post. (And I solemnly swear to God that the perfective needs another name.)

[hwhatting]

Ok, so what you have in mind here really seems to be a relative tense (retrospective - act in the past with regards to a certain point in time that is not necessary the moment of speaking, prospective - act in the past with regards to a certain point in time that is not necessary the moment of speaking). That means that (a) they can be combined with any tense and (b) that they also can be combined with imperfective / perfective aspects (although some combinations may be disfavoured). So I would put retrospective / prospective into a different slot from both tense and from imperfective / perfective aspect. You'll probably need to think about the combinations - under what circumstances are they used? Are there ways to simplify this? E.g,, Old Slavonic had an aorist (basically, a simple past), an imperfect ("present in the past"), a perfect ("retrospective") and a pluperfect ("past in the past"), but the Northern Slavic languages have given up These distinctions and use the imperfective / perfective aspect to express most of these distinctions.
Another idea for simplification - you rarely find systems where relative tenses are distinguished for all tenses; in generally, the past tense tends to show more granular distinctions than the past or the future tense,

[StrangerCoug]

This is turning to be less of a quickie than I thought xD

in generally, the past tense tends to show more granular distinctions than the past or the future tense,

Now I'M the one thrown for a loop! How can a feature have more granular distinctions than itself?

[hwhatting]

This is turning to be less of a quickie than I thought xD

in generally, the past tense tends to show more granular distinctions than the past or the future tense,

Now I'M the one thrown for a loop! How can a feature have more granular distinctions than itself?

Sorry, that was meant to be "more granular distinctions than the present or the future tense".

[Pole, the]

In western Europe it is common to form a periphrastic past tense by a construction of the verb “to have” and a past/passive participle. (Examples include English, German, French…) In eastern Europe it is common to form a periphrastic possessive construction by using an inflected noun / adpositional phrase and the verb “to be”, so “I have X” becomes “at me is X” &c. (Examples include Russian, Finnish…)

My question is, is it possible to combine these two? I mean, at the beginning you have an inflectional present tense and an inflectional past tense:

siavek en ond
see.PRES-1SG IND dog
I see a dog.

savek en ond
see.PAST-1SG IND dog
I saw a dog.

Then the past tense becomes supplanted with a have-perfect counterpart:

abiek siavà en ond
have.PRES-1SG see.PTCP IND dog
I saw a dog. (lit. I have seen a dog.)

Then “to have” becomes replaced itself:

mi ist siavà en ond
1SG.DAT be.PRES-3SG see.PTCP IND dog
I saw a dog. (lit. At me is seen a dog.)

At this stage, the existential verb can be left out:

mi siavà en ond
1SG.DAT see.PTCP IND dog
I saw a dog. (lit. At me seen a dog.)


Would that be a plausible outcome? Are there languages spoken in Europe doing something similar?

[hwhatting]

Would that be a plausible outcome? Are there languages spoken in Europe doing something similar?

I don't know of any European language doing something similar. It looks a bit like the perfects / past tenses in modern Indic and Iranian languages (which arose via a different pathway, combining the Past Passive Participle with "to be").
One remark - your pathway envisages "have" construction to be replaced by a "be" cosntruction. In the history of the European languages, the development was the other way round - PIE didn't have a word "have" and in the older IE languages you still find constructions with "be" for possession (e.g. Latin mihi est "I have" besides habeo). Russian, probably beacuse it's peripheral in Europe and under the influence of Finno-Ugric and Turkic languages, didn't replace the "be" contructions by "have" (as you probably know, иметь "have" is used much less frequently than the "be" construction). But a language that has "have" replacing it with "be" constructions for possession would go against the trend in Western and Central European languages.

[StrangerCoug]

How common is a combined imperative/cohortative/jussive mood?

[Ser]

How common is a combined imperative/cohortative/jussive mood?

I'd say it'd be unsurprising. Just call it "imperative" though, unless your language is based on Esperanto (in which case "jussive" or "volitive" would be appropriate...).