Conlang Random Thread

[Qwynegold]

Thanks again for helping me with the question word, everyone.

The only problem is what do I call this what that is kinda like a relative pronoun but not. So far the rule in my grammar has been that question words are only used in questions, so I should mention something about when that's not actually the case.

It's still just the regular question word, not a relative pronoun at all.

Oh, I see. Well that makes things easier.

[bradrn]

Modern Hebrew doesn't have a case system. You have the genitive preposition של /ʃel/, the use of the construct noun or addition of personal suffixed to the construct form. The first is used in most cases, the last is mainly used formally with many exceptions and the one in the middle is just rare, but it is the closest to the idea of nouns modifying nouns. Modern Hebrew just uses adjectives more for description than Biblical.

I've always wondered how come sometimes you could make a genitive construction without של, and also exactly what the construct state is for. Thanks for explaining this!

[Qwynegold]

I have more to say about relatives. But I made a new thread for it, because this is getting long.

[Xwtek]

Do you think it's reasonable to turn dative to accussative if the language originally have no accussative?

[bradrn]

Do you think it's reasonable to turn dative to accussative if the language originally have no accussative?

I wouldn't be able to answer this, but due to the Case Hierarchy it would seem unlikely for a language to have a dative without an accusative in the first place.

[Kuchigakatai]

I hadn't heard of Blake's case hierarchy before, but as a first impression I'm not that convinced by the concept. Isn't it common to have a direct case that merges nominative and accusative while having other cases? Wikipedia mentions Irish is like this, and I would add Romanian (in its nouns and adjectives) and English (in its nouns, as spoken by some if not many native speakers).

[gach]

I hadn't heard of Blake's case hierarchy before, but as a first impression I'm not that convinced by the concept. Isn't it common to have a direct case that merges nominative and accusative while having other cases?

The hierarchy is certainly not strict and it's easy to cook up and justify exceptions to it through historical change or perhaps taking care of some of the core cases (accusative, ergative, genitive) by the means of head marking instead. A previously nonexistent accusative developing from a dative does indeed sound a lot like a direct-oblique case system turning its oblique case into an accusative.

[Xwtek]

How about this. a language has 4 cases: Direct, Genitive, Locative, and Instrumental. There is also two genders, Animate and Inanimate. All The case assignment is:

1. In an intransitive sentence, this is unambiguous. Direct case is used for the only argument.
2. In a transitive sentence, the case assignment is split into many representations:
   1. In most verbs (i.e. eat, kill, beat, etc), both subject and object receive direct case. However, the subject has to be animate, while the object is inanimatized with inanimate gender. If the inanimate noun has to be the subject, passive construction is used instead. (I-DIR eat a.cake-INAM-DIR)
   2. For the movement verbs (to run into something, to go to something, to sit on something), the subject receives direct case, and object receives locative case. (Book-INAM-DIR lies here-INAM-LOC (There is book)) (Note that book is inanimate noun)
   3. For a verb that doesn't give any effects on the noun, the subject is marked with direct case, and the object is marked with instrumental case. (I-DIR see a.flower-INAM-INS)
3. In a ditransitive sentence, the case assignment is split into many representations:
   1. [*[ In a verb related to giving (i.e. to send (a letter), to give, etc), the subject is marked with direct case, direct subject is marked with locative case, and indirect object marked with instrumental case (I-DIR sent him-LOC a.letter-INAM-INS)
   2. In a verb that is basically the causative of the transitive verb, both the old subject and the new objects is marked with direct case, but again the new subject cannot be inanimate and the old subject is inaminatized. The original object retains the case (Note that verb that have both subject and object as direct case cannot have causative form) (etc: I-DIR read him-INAM-DIR, a.book-INAM-INS) (etc: I-DIR put him-INAM-DIR, on.the.chairs-INAM-LOC)
   3. The rest of them (i.e. to tell, to bet etc), both subject and direct object receive direct case. However, the subject has to be animate, while the object is inanimatized. If the inanimate noun have to be the subject, passive construction is used instead. The indirect object is marked with instrumental case. (etc: I-DIR bet him-INAM-DIR, 5.dollars-INAM-INS)

[Kuchigakatai]

The idea of turning dative marking into accusative marking is also reminiscent of the so-called "personal a" of Spanish, where the old dative marking of Late Latin became accusative marking for rational direct objects (humans, gods and pets) in Old Spanish, which might be extended to other types of nouns in the future, and where indirect objects have been reinforced or modified with other elements still staying separate from direct objects.

In most verbs (i.e. eat, kill, beat, etc), both subject and object receive direct case. However, the subject has to be animate, while the object is inanimatized with inanimate gender. If the inanimate noun has to be the subject, passive construction is used instead. (I-DIR eat a.cake-INAM-DIR)

I wouldn't call that a passive. This is what a direct-inverse language is, so the verb would be said to be in inverse voice when it has an inanimate subject.

In a verb that is basically the causative of the transitive verb, both the old subject and the new objects is marked with direct case, but again the new subject cannot be inanimate and the old subject is inaminatized. The original object retains the case (Note that verb that have both subject and object as direct case cannot have causative form) (etc: I-DIR read him-INAM-DIR, a.book-INAM-INS) (etc: I-DIR put him-INAM-DIR, on.the.chairs-INAM-LOC)

I think I understand what you're saying here perfectly, but a lot of people here on the ZBB are going to be confused by this paragraph. The lack of a translation in your examples may suggest a different reading, and I'm not sure if I'd really gloss that "put" verb as "put". I do understand what you mean because Mandarin 放 fang4 is used this way, meaning both "put sth somewhere" and "lie/remain placed somewhere" (with an appropriate aspect morpheme), but in general it's good if you assume your English-speaking readers are not familiar with such semantic ranges and so provide an explanation.

The rest of them (i.e. to tell, to bet etc), both subject and direct object receive direct case. However, the subject has to be animate, while the object is inanimatized. If the inanimate noun have to be the subject, passive construction is used instead. The indirect object is marked with instrumental case. (etc: I-DIR bet him-INAM-DIR, 5.dollars-INAM-INS)

Again, that would not be passive voice, that would be an example of inverse voice.

[Xwtek]

I wouldn't call that a passive. This is what a direct-inverse language is, so the verb would be said to be in inverse voice when it has an inanimate subject.

While it corresponds to direct-inverse system in one of its descendants, it's not yet a direct-inverse now, because:

1. The original subject is deleted or is marked with instrumental case, not direct case as it should if it is really direct-inverse.
2. It IS a valence decreasing voice. The translation of "the building has been destroyed" IS translated by passive voice. It's just that the sentence "the trap kills the prey" cannot be translated in active voice and have to be passivized as "the prey is killed by the prey"
3. The Direct-Inverse marking is also unusual since this language only has subject agreement anyway, and it's incomplete like a typical SAE language.

[mèþru]

FYI Direct case systems are pretty rare.

[Whimemsz]

FYI Direct case systems are pretty rare.

So what?

I wouldn't call that a passive. This is what a direct-inverse language is, so the verb would be said to be in inverse voice when it has an inanimate subject.

That's not what direct/inverse languages are, no. Though this has been, or was, a subject of debate for a while; some people have tried to analyze inverse verbs, at least in some languages, as being obligatory passives -- or analyze obligatory passives as inverse verbs -- but while this analysis might work in some languages it doesn't work in most of them, IMO. I would just say that languages that require passive constructions when the action goes against a prominence hierarchy are most of the way along a continuum toward direct/inverse languages, and are very well positioned to turn into one. As Akangka points out, in most languages there are ways of distinguishing a passive construction from a (theoretical) active inverse construction -- in the case of Akangka's language, evidently, it's that passive constructions can drop the demoted original subject and the valence of the verb is decreased [although, I don't see from the glosses how this is determined in your lang, Akangka], both of which are normal features of passive constructions but are not features of inverse constructions.

[malloc]

Everything about this phonology is frustrating me. It seems remarkably hard to pin down the technical details of the phoneme inventory or justify key phonological rules. In particular, the phonemes /s/ and /l/ are difficult to classify by features and require implausible or complicated sandhi rules to fit into the phonological system. Conventionally /s/ has the features [COR +CONT] since it represents the fricative counterpart to the coronal stop /t/. Yet this featural analysis implies that /s/ should allow the allophone [T] which my language certainly doesn't allow. Instead it seems more specifically [+STRIDENT] than anything. And needless to say, the features [COR +CONT] make little sense in light of the glide hardening rule that promotes /j/ [DOR +HIGH -BACK] to /s/ (made necessary because the language has no posterior sibilants). The phonology fits my phonoaesthetic preferences better than anything I have found thus far, but theoretically it is a horrible mess.

[Xwtek]

Everything about this phonology is frustrating me. It seems remarkably hard to pin down the technical details of the phoneme inventory or justify key phonological rules. In particular, the phonemes /s/ and /l/ are difficult to classify by features and require implausible or complicated sandhi rules to fit into the phonological system. Conventionally /s/ has the features [COR +CONT] since it represents the fricative counterpart to the coronal stop /t/. Yet this featural analysis implies that /s/ should allow the allophone [T] which my language certainly doesn't allow. Instead it seems more specifically [+STRIDENT] than anything. And needless to say, the features [COR +CONT] make little sense in light of the glide hardening rule that promotes /j/ [DOR +HIGH -BACK] to /s/ (made necessary because the language has no posterior sibilants). The phonology fits my phonoaesthetic preferences better than anything I have found thus far, but theoretically it is a horrible mess.

It's pretty common to treat /s/ as palatal instead of alveolar

[mèþru]

Great site for making up a heraldic shield
http://drawshield.net/create/index.html

[bradrn]

Great site for making up a heraldic shield
http://drawshield.net/create/index.html

Impressive! I copied a complex blazon from Wikipedia over and it worked first time!

(Of course, this tool would have limited use if your conworld's heralds differ in any significant way from the traditional European model.)

[Xwtek]

Do you know any language outside Indo-European, Semitic, and Oto-Manguean to have fusional grammar? The best I know is Navajo, but it sounds like if you combine agglutinative part with the fusional one.

[Whimemsz]

.

[Moose-tache]

I kinda want to have a language where most adjectives are actually nouns meaning "ADJ-ness", and the adjectival usage of these words are accomplished by using possessive structures or adpositional phrases.

Lots of languages don't have a distinct category of adjectives. While it's more common for the would-be adjectives to reappear among the verbs, since stative verbs already behave a lot like adjectives, this is not universal. In Japanese you have a class of noun-like adjectives that are only a hop, skip, and a jump away from just being nouns (is na really so different from no?). If you generalized this across the language you could have adjectives basically merge with nouns. For realism, I would recommend still moving some adjectives into the verb camp, and maybe helping to justify the noun-adjectives with lots of borrowings.

[k1234567890y]

I kinda want to have a language where most adjectives are actually nouns meaning "ADJ-ness", and the adjectival usage of these words are accomplished by using possessive structures or adpositional phrases.

Lots of languages don't have a distinct category of adjectives. While it's more common for the would-be adjectives to reappear among the verbs, since stative verbs already behave a lot like adjectives, this is not universal. In Japanese you have a class of noun-like adjectives that are only a hop, skip, and a jump away from just being nouns (is na really so different from no?). If you generalized this across the language you could have adjectives basically merge with nouns. For realism, I would recommend still moving some adjectives into the verb camp, and maybe helping to justify the noun-adjectives with lots of borrowings.

I have done "Adjective = Verb" several times before, but still thanks (: