zompist bboard

Recently, I have been creating conlangs which do not use completely different methods of inflection, encoding meaning in verbs, and then creating sentences. Since they're rather difficult to explain in the company fluency thread with derailing it, I'll explain the systems as best I can here. At the moment there are 3, but I expect that number to go up in the coming weeks. For each, I have either a developed language or a working language to let me test the system. I'll address each system in turn, but don't expect this to be expertly written. Anyway, on with the show.

Vaqnad, an object-less language
This was the first system I thought up, and it disallows the majority of verbs from taking multiple arguments. The copula is the only exception I have so far. In order to express a sentence like 'I see a man', you would have to say 'I see; a man is seen'. This can then be translated as ghazamr kha aghzame vaqnta.
This also allows one to string verbs together indefinitely, if they all have the same subject. An example would be 'I see a man and talk to him', which would be written 'I see talk; a man is seen is talked to'. This can be extended to any number of Verbs without requiring more clauses, while English would need a new clause for each new verb.

Lingua Vocalis, a language where forms of the same verb can be totally different to one another
I created this system in an attempt to violate the universal that languages must have morphemes. It appears that I have not succeeded, but the system I made is interesting anyway. Whilst other language's bound morphemes are essentially fixed in their phonemic structure, Lingua Vocalis' can change drastically depending on what other bound morphemes are present in the same word, and what order they are present in.
No matter what the phonemes present are, each morpheme has an underlying constant structure which can be found by comparing two adjacent vowels. For example, the ergative and absolutive forms of the noun 'crystal' are ieyæ and ïëyœ. In the language's phonology, i and e are one step of height different, and e and æ when separated by a y are two steps of height different. Similarly, ï and ë are one step of height different (both are rounded), and ë and œ when separated by a y are two steps of height different. Using the system of notation I decided to let me write down these differences, crystal-ergative is written iH-HH, and crystal-absolutive is written ïH-HH. Now it can be easily seen that the root meaning crystal is marked by a change in height of one step, then a change in height by two steps.
The entire system also employs changes in roundedness, marked by R, and changed in frontedness, marked by F. These can be applied to any vowel unambiguously, and in ten combinations: R, F, H, RF, RH, FH, HH, RFH, RHH, FHH. There is also a null change, written 0, but it only serves to separate prefices from the root, and is realised as a long vowel. When changes with only a single step are used, vowels are simply concatenated. If there are two changes, either y, w or h is inserted depending on the prior vowel. If there are three changes, a glottal stop (') is inserted. This is because the mid vowels have two types, high and low, which are not distinguished in writing or in casual speech. The mid vowels default to high if they are the initial vowel, so eH is written ei, and eHH is written eyæ.
The upshot of this is that 'I broke it' (IND-TOP-PERF-1.ERG-3.ABS-hit) can be written as either uwi'oöwä:he'ahæ or uwoö'œhä:he'ahæ, among other possibilities. Removing the topic marking from these examples gives u'eëyœ:ho'æha and uwoö'œ:ho'æha.

Lingua Localis, a which conjugates by space, not time
As far as I know, every language encodes some sort of time information on the verb, whether it's when it happened, or if it's going on from the narrative's point of view. This language, however, does not. Rather, it encodes information about where the event happened. For example, it could be near, far, or in the middle. It could be approaching, receding, circling, or staying still. It could be all those things from your point of view, or someone else's. It could be taking place in an indivisible location (or one treatable as such), or moving between separate locations.
As of right now, I have neither a language for it, nor official-sounding terminology, but I can still go into how one goes about translating into this system. Let's start with a simple sentence, say 'I see a book'. To translate this, we need to know how far the book is, which way it's moving, whether it's passing through multiple places or not, and whether all these observations are from your point of view or not. If it's on a table right in front of you, we can say 'I see-NEAR.SELF-STATIC.SELF-POINT book'. To translate that back into English without losing all this spatial information, we'd have to say 'I see a book right here, unmoving'.
This suggests a nominal case unique to this system. As the locative case for natlangs identifies an object's location in space, so this system can have a case which identifies an object's location in time. This is of course more ambiguous, but that can probably be rectified somehow. So, a phase 'book table-TEMPORATIVE' might be translatable as 'the book that was in the location described by the verb at the same time as the table', or more simply, if context allows, 'the book on the table'.

And I think that's just about everything. If you have any questions about this, say, if I was unclear about something, or you think I left something out, feel free to ask. I hope you find this interesting.

The first one sort of works, although it's kind of tortured and if every time you try and express a transitive you just end up phrasing it SUBJ see-ACT OBJ see-PASS you can without too much difficulty analyse that as a system with subjects and objects anyway. Not to mention that you can't really have a passive unless you have objects to promote in the first place - that is, having a passive form of a verb implies that there has to be an active, transitive form.

Obscure morphophonology which produces surface structures which are near-impossible to segment is definitely not a new idea and mad stuff appears in natlangs all the time; cf. Navajo. What's more, when it happens in natlangs it a) serves as a viable means of communication and b) is diachronically justified, which in my book gains it extra points (although I appreciate that you are doing artlanging and so presumably care about neither of these things). Your way of doing it is artlangy and vaguely interesting but you certainly haven't eliminated morphemes (as you realise, to your credit).

Not all languages encode time information on the verb - at least not how you imagine they do. I'd imagine all languages have lexical aspect of some kind since aspectual implications are probably inescapably encoded into verbs, but you certainly haven't got rid of that. A lot of languages also encode spatial information on their verbs or by other means.

Whilst I'm sympathetic to attempts to break the mould and come up with interesting new concepts for language, I'd go and read some more about how languages work first.

Unfortunately for me, I am somewhat limited in the read-about-languages department, being on holiday and having a limited amount of money due to being too young to actually get a job I'd like (paper round sucks!). So, I would if I could. However, this is part of why I posted this. I want to know if there are languages which are this different, especially the last two.
The first one doesn't seem tortured to me at all, since verbs don't have to be transitive, and the passive isn't really a passive. Plus, it would get more interesting if I wrote out an entire story, since the two verbs don't even have to be anywhere near each other in the formal dialect, and you can change the verb to give extra meaning to the action, such as this minor change: 'I walk; the building was walked into', which would mean that I am walking along a street, say, then I walk into a building. Plus I haven't even begun to explore how to reduce the system whilst preserving the way it works.
[To be finished once I have looked at Navajo]

KathAveara wrote:Unfortunately for me, I am somewhat limited in the read-about-languages department, being on holiday and having a limited amount of money due to being too young to actually get a job I'd like (paper round sucks!). So, I would if I could. However, this is part of why I posted this. I want to know if there are languages which are this different, especially the last two.
The first one doesn't seem tortured to me at all, since verbs don't have to be transitive, and the passive isn't really a passive. Plus, it would get more interesting if I wrote out an entire story, since the two verbs don't even have to be anywhere near each other in the formal dialect, and you can change the verb to give extra meaning to the action, such as this minor change: 'I walk; the building was walked into', which would mean that I am walking along a street, say, then I walk into a building. Plus I haven't even begun to explore how to reduce the system whilst preserving the way it works.
[To be finished once I have looked at Navajo]

Wikipedia. Google Books. There's also thousands of language grammar and linguistic theory pdfs floating around the web.

*Shrug*

[Continuation of my earlier post]
I've had a look at Navajo, and it doesn't seem to be entirely the same sort of thing. What I thought up looks even more obscure than Navajo, since Navajo at least has some part of the bound morphemes remaining roughly the same, while there are 16 completely different forms for any morpheme with this vowel inventory. If I added central vowels as well, then that would double the number of forms.
I know that not all languages encode tense directly. Mandarin would use adverbs or something, which this system would as well. However, aspect as we have it would not be even implicitly marked on the verb, for they to would be optional adverbs. Thus, a verb like the example I gave could be imperfective, perfective, habitual, or anything else. While it may seem that there is some aspect buried inside the verb, that's probably only because of the translation.

The first idea seems extremely clunky and inelegant and uneconomic and, yes, "tortured." Or rather, torture--to the poor damned souls who are condemned to talk that way. When you talk about your wish to "reduce" the system, I think I see an acknowledgment of its clunkiness.

KathAveara wrote:the passive isn't really a passive

What is it then. When you say 'the dog chases the cat is chased', how do you know who is chasing whom? How are the theta roles being coded?

I can't see how the reduplication of the verb expresses anything that the mere presence of an object doesn't. What is the point of this idea? From a functional perspective, I mean.

Also: do you have to reiterate the verb for oblique arguments as well? Or just the core aguments?

Finally: one of Nort's langs--Proto-Hathic I believe--does something vaguely similar to this, but the 'passive' verb (the one that has the object for its argument) is different from the 'active' (agentive) one, and expresses a result (AIUI).

cromulant wrote:Finally: one of Nort's langs--Proto-Hathic I believe--does something vaguely similar to this, but the 'passive' verb (the one that has the object for its argument) is different from the 'active' (agentive) one, and expresses a result (AIUI).

Huh? Proto-Hathic is the one where you have to serialize most verbs to make them transitive... and there are no morphological distinctions between subject and object anyway. The only lang I have that does anything interesting with the semantics of subject/object is Kannow, where it's about intention and telicity.

Nortaneous wrote:

cromulant wrote:Finally: one of Nort's langs--Proto-Hathic I believe--does something vaguely similar to this, but the 'passive' verb (the one that has the object for its argument) is different from the 'active' (agentive) one, and expresses a result (AIUI).

Huh? Proto-Hathic is the one where you have to serialize most verbs to make them transitive... and there are no morphological distinctions between subject and object anyway.

Nothing I said is inconsistent with any of that.

First off, that language is a proto-language, so I need to me around with the grammar. Secondly, the passive isn't really a passive because all active verbs are intransitive anyway. There is no obligation to express the object of our transitive or ditransitive verbs, which can be used to certain effects, such as suspense if you leave the passive for much later.

cromulant wrote:

Nortaneous wrote:

cromulant wrote:Finally: one of Nort's langs--Proto-Hathic I believe--does something vaguely similar to this, but the 'passive' verb (the one that has the object for its argument) is different from the 'active' (agentive) one, and expresses a result (AIUI).

Huh? Proto-Hathic is the one where you have to serialize most verbs to make them transitive... and there are no morphological distinctions between subject and object anyway.

Nothing I said is inconsistent with any of that.

Where in the thread did I have the result thing? I don't remember there being anything like it, but I'm awful at grammar so I could just be missing an obvious analysis.

Ah. You're right in that you don't say anywhere that the second verb expresses a result--that was my interpretation.

http://www.incatena.org/viewtopic.php?f=12&t=41750#p1025910

Nortaneous wrote:suka feñ !ap mʘet
1S pick_up break 3S
'I broke it' -- lit. 'I picked it up and it broke'

It looks me like the first verb expresses the action and has as its argument the Agent, while the second verb expresses the result of the action, and has as its argument the Patient.

But perhaps I am extrapolating too much about PH from a single example.

EDIT: also, that comment was made just going by memory. As I look at the example again, I'm not sure why I thought the second verb has a different core argument than the first. Probably because you said there were very few transitive verbs, so, if "it" is not the transitive object of "break," then it must be the (patientive) subject.

At any rate, it's an interesting system that got me thinking, and KA's idea reminded me of it a bit.

cromulant wrote:Finally: one of Nort's langs--Proto-Hathic I believe--does something vaguely similar to this, but the 'passive' verb (the one that has the object for its argument) is different from the 'active' (agentive) one, and expresses a result (AIUI).

Regardless of what Nort's languages actually do, if you are going for single argument verbs, it sounds much more natural to mark the objects with a more restricted class of linker verbs than by reduplicating the main verb or some special inflectional form of it. So instead of saying

I hit, man get.hit

for "I hit a man" you'd be saying something like

I hit, man get.affected

The really fun part here is that such linker verbs would very naturally become classificatory so that the choice of which one to use would be determined by the type of the object or the process being described. So for things like eating or being sick you could have

I eat, fruit get.comsumed
"I'm eating a fruit"

and for perception verbs

I see, man cause.experience
"I see a man"

With historical development you could do even more fun stuff. Since there isn't really any good reason why the verbs should avoid multiple arguments, there'd be pressure for the objects to get analysed as arguments of the core verb and the linking verbs to erode into case marking particles. As a result you'd have an array of accusative markers (or perhaps ergatives as well, depending on how the original language worked) where the choice of the proper variant would be determined by the type of action described by the clause.

One further thing: if you are going to code objects using verb serialisation or some sort of clause chaining, it feels natural that similar strategies would be the the preferred way to mark oblique phrases as well. If someone knows any universals relevant for this intuition, please comment.

cromulant wrote:The first idea seems extremely clunky and inelegant and uneconomic and, yes, "tortured." Or rather, torture--to the poor damned souls who are condemned to talk that way. When you talk about your wish to "reduce" the system, I think I see an acknowledgment of its clunkiness.

KathAveara wrote:the passive isn't really a passive

What is it then. When you say 'the dog chases the cat is chased', how do you know who is chasing whom? How are the theta roles being coded?

You can do it without passives:

I see a man. A man is visible.
I talk. A man undergoes/experiences.
The dog chases. The cat tries to escape.

Your examples all use two different verbs. If you've read the thread, you know that this is not a new concept to me. You also know that KathAveara's language uses the same verb twice. Not 'the dog chases, the cat tries to escape' but 'the dog chases, the cat is chased.' Your solution is inapplicable to the language in question.

My questions stand. A gloss of the language's rendering of 'the dog chases the cat' would be helpful.

cromulant wrote:Your examples all use two different verbs. If you've read the thread, you know that this is not a new concept to me. You also know that KathAveara's language uses the same verb twice. Not 'the dog chases, the cat tries to escape' but 'the dog chases, the cat is chased.' Your solution is inapplicable to the language in question.

My questions stand. A gloss of the language's rendering of 'the dog chases the cat' would be helpful.

Hmm, I guess so. A verb inflected for agency of its argument perhaps? I guess a gloss is necessary.

This has been a very interesting discussion to read, and I have made a few modifications and thought more about the most natural solution to the problem. Given that I do not have access to my notes at present, I'll use English to demonstrates my solution.
For 'The dog chases the cat' there are two methods. Firstly, a special verb roughly translatable as 'do' can be used, giving dog chases; cat is done. Alternatively, you can say what the cart does in response, in the active, and link the two phrases with a marker (if you speak a higher register), giving dog chases; cat runs away. The auxiliary verb either refers to the previous verb, or whichever verb it is linked to via markers in the higher registers.
Also, sorry for the long delay. I've had no Internet access for weeks.