Omzinesý wrote:
eldin raigmore wrote:
I guess an "effected patient" would be one that is created during the clause?
Yes, I know what it is. .... It is something created in the action.But
I didn't know. Apparently I guessed right; thanks for clarifying!
Omzinesý wrote:
... I'm just interested in where it is. ...
I can't find anything about it online; did you look in Barry J. Blake's book "Case"? If so, did that help?
Lodhas wrote:
Or, if you're asking about grammatical voices, I'm certain there are/were languages with at least 3. I can't remember any off the top of my head though.
Classical Greek, Classical Latin, and IIANM Classical Sanskrit, among other I-E languages of that age.
Also, the languages that
M.H.Klaiman in his book "Grammatical Voice" typologizes* as having a "Basic Voice System"; Fula and Tamil are included.
*He mentions Basic Voice Systems (typically these are nearly the same as systems with a Middle Voice); Derived Voice Systems (where various participants are promoted into and/or demoted out of various morphosyntactically-assigned argument positions, as passivization demotes the Agent out of the Subject position and promotes the Patient into the Subject position); Inverse Voice Systems (typically those with a Hierarchical morphosyntactic alignment, also typically those with an obviative, but the overlap is not complete); and Information-Salience Voice Systems (where, rather than "what's the subject?", the question is either "what's the topic?" or "what's the focus?" or, for at least one language, both of those); and leaves open the possibility that there are other types, or languages that don't belong to a type, or, possibly (for all I can remember), languages that don't have Grammatical Voice.
Zifre wrote:
Is there any language which has no true ditransitive verbs
Yes. Or, at least, I'm sure there are some with no ditransitive verb-
roots; and I think there may be some with no ditransitive verbs at all.
Zifre wrote:
but instead forms them in some periphrastic manner?
I'm not sure. See
http://wals.info/chapter/105 and his references.
Joan Bybee (formerly Joan B. Hill) wrote a book on verb morphology, which may be the book I remember reading something like the following out of (but I could be wrong);
Every language has monovalent (one core argument) verb-roots, and every language has bivalent (two core arguments) verb-roots.
While most languages have many of both of those valencies, many have many more monovalent than bivalent verbs and many others have many more bivalent than monovalent verbs.
Many languages also have several trivalent (three core arguments) verb-roots. And a few have 0-valent or 4-valent verb-roots. But in all languages that have them, trivalent verb-roots are a small(ish) minority of the verb-roots; also 0-valent verb-roots if there are any, and 4-valent verb-roots if there are any, are always smallish minorities.
The most common morphology of verbs is valency-changing.
Some languages without trivalent verb-roots can nevertheless form trivalent verbs from bivalent roots by applying a valency-raising operation, such as causativization or "dative applicativization".
But IIRC there are some languages in which valency can't be raised beyond 2.
Anyway;
If you apply a valency-decreasing operation to a monovalent verb, the result is a 0-valent verb.
In some languages some valency-decreasing operations can be applied twice. For instance, both Turkish and Hindi allow "double passivization". If you do that to a bivalent verb you wind up with a 0-valent verb.
In some languages you can apply a valency-raising operation to a trivalent verb. If you do that, you wind up with a 4-valent verb.
In some languages, you can apply two different valency-raising operations (such as causativization and benefactive applicativization) to the same verb one after another; in some languages, you can apply the same valency-raising operation (such as causativization) twice to the same verb. (For instance, Hindi has "double causativization".)
If you do either of those to a bivalent verb, you wind up with a 4-valent verb.
I don't know of any language that allows you to raise valency higher than 4; or at least, none that has been so analyzed by any linguist who wasn't trying to debunk the whole idea of "valency".
I'm pretty sure some languages don't allow valency to be raised higher than 3; I imagine English might be one of them, depending on how you define your terms.
I'm only slightly less sure that some languages don't allow valency to be raised higher than 2.
Milyamd wrote:
Is there a correlation between noun classes and 3rd-person pronouns (should there be a personal pronoun corresopnding to each of the classes)?
Do they tend to be related in any way (the same roots/morphemes)?
My first impulse was to look at
http://wals.info/feature/combined/44A/43A. But, having done so, I don't know for certain that the answer is there; in fact I realized I'm not certain of the precise meaning of your question.
Most languages (254 out of 378 in their database) don't have any gender distinctions in independent personal pronouns.
Most that do have it only in 3rd person singular (61 out of 124 in their database) or only in 3rd person but in both 3rd singular and 3rd non-singular (42 out of 124).
Another 18 out of the 124 have gender distinctions in 3rd person and also in some other person.
http://wals.info/feature/combined/44A/30A shows that most languages don't have any gender at all; which, for at least 109 of those 254 languages that don't have gendered pronouns, explains that. But there are still at least 15 languages in their database that have genders but not gendered pronouns.
Weirdly, they list 5 languages that don't have genders but do have gendered pronouns. I don't know what the deal is with that. Maybe you should write to them and ask them.