Adjectival cases vs. adverbial cases

[pharazon]

Trask saith:

A central characteristic of Basque syntax is the use of -ko phrases. A -ko phrase may be constructed from virtually any adverbial, regardless of its internal structure, by suffixing -ko to it; this suffix induces certain phonological changes, notably the loss of the locative case-suffix -n. The resulting phrase is a preposed adjectival modifier.

...
mendira `to the mountain'; mendirako bidea `the road to the mountain'

mendira being the form used in something like mendira joan naiz "I've been to the mountain"; -ra is the allative.

Also, for people knowledgeable about Basque, how does Trask figure that etxeko is derived from etxean and not just etxe? Are there examples elsewhere of morphophonemic changes that would take *etxeanko to etxeko?

[chris_notts]

Also, for people knowledgeable about Basque, how does Trask figure that etxeko is derived from etxean and not just etxe? Are there examples elsewhere of morphophonemic changes that would take *etxeanko to etxeko?

I don't know about the sound changes, but perhaps part of the argument is to do with the fact that in the plural -ko takes the -eta- plural that's pretty much restricted to the local cases? Also you can form adjectival cases by adding -ko to both allative and perlative cases:

mendirako bidea
mendi-ra-ko bide-a
mountain-ALL-REL road-def
the road to the mountain

menditiko bidea
mendi-tik-ko bide-a
mountain-PER-REL road-def
the road from the mountain

but you can't with the normal locative ending:

*mendianko bidea

instead, as you said, you have to say:

mendiko bidea
the mountain road

I've been meaning to buy Trask's "The History of Basque" for some time now, but I haven't gotten round to it. Also, it's very expensive and I'm not sure I can spare the money right now.

EDIT: Okay, so that didn't contribute much. I don't know why I wrote it now... I guess I just like the sound of my own voice.

[hwhatting]

Kazakh uses -ghy/-gi:

Qala-da tura-j-myn
city-loc. live-pres.-1sg.
I live in the city.

qala-da-ghy ?j
city-loc-(attr. marker) house
the/a house in the city

Same is the case for Uzbek.

Russian and German handle this basically the same way as English - the attributive prepositional phrase follows after the non it determines, but there's no other formal distinction:

Das Haus in der Stadt. - Ich lebe in der Stadt.
Dom v gorode. - Ja zhivu v gorode.
The house in the city. - I live in the city.

Best regards,
Hans-Werner

[Echobeats]

Trask saith:

A central characteristic of Basque syntax is the use of -ko phrases. A -ko phrase may be constructed from virtually any adverbial, regardless of its internal structure, by suffixing -ko to it; this suffix induces certain phonological changes, notably the loss of the locative case-suffix -n. The resulting phrase is a preposed adjectival modifier.

...
mendira `to the mountain'; mendirako bidea `the road to the mountain'

Turkish does a similar thing with -ki. İstanbul'daki kişiler = "The people in Istanbul".

Tim.

[TomHChappell]

Firstly: what are the usual terms for cases that result in something that modifies a noun and cases that result in something that acts adverbially? Is adjectival case/adverbial case okay?

Case(s) which make the so-cased noun or NP modify another noun or NP are called "adnominal cases", to distinguish them from "adverbial cases".

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Cases which indicate the relationship the so-cased noun or NP has to the verb or VP or clause as a whole are called "adverbial cases", to distinguish them from "adnominal cases". These mostly split into "syntactic cases" vs "semantic cases"; "syntactic" cases apply to mark the so-cased noun or -cased NP as being in a certain grammatical relation, whereas "semantic" cases apply to mark the so-cased noun or -cased NP as having a certain thematic role within the clause.

The "syntactic cases" are, by Blake, listed as usually being "Nominative, Accusative, Genitive, and Dative" for Accusative languages, "Nominative, Ergative, Genitive, and Dative" for Ergative languages (Blake uses "Nominative" where most people would use "Absolutive"); and I imagine "Nominative, Accusative, Ergative, Genitive, and Dative" for tri-partite languages.

Most "Semantic cases" that I'm aware of are all "adverbial"; the biggest group is probably locative/adessive, allative, ablative, and perlative.

But ISTR (IIRC) there may be a language, or some languages, with a case, or some cases, specially for what we think of as "predicate nominatives".

At least one language has at least one case that applies only to adjectives, not to substantive nouns.

Now, the real question. How do various langs distinguish between these two usages? (Using "case" in a general sense, including affixes and adpositions.)

Every language has the same range of choices:
to the best of my memory they are the following:

* lexical suppletion (like "me" and "I")

* stem modification (like "you" and "your" and "yours")

* affixes (almost always suffixes to indicate Case; but prefixes and infixes and circumfixes and superfixes and transfixes could apply as well, especially for other categories than Case)

* agreement -- some other word is modified to indicate the case of this word (for instance in English, verbs agree only with Subjects, never with Objects)

* periphrasis -- adding a word or clitic (usually an adposition -- a preposition or a postposition or an imposition -- for Case)

* pure syntax -- mostly means "word order" -- e.g. in SVO languages the S and the O might not be differently case-marked, so that "Ess Vees Oh" means "Ess" is the Subject and "Oh" is the Object, whereas "Oh Vees Ess" means "Oh" is the Object and "Ess" is the Subject.

In English, it's done with syntax: the prepositional phrase follows the noun it modifies.

Adverbial: I went to Tokyo. ~ To Tokyo I went.
Adjectival: The road to Tokyo

So, what is the name of the case of "to Tokyo"? Since it modifies the noun "road", it's adnominal; does that mean it can't be the same case as the allative (adverbial) "to Tokyo" in "I went to Tokyo" or "I sent John to Tokyo"? Or is it merely that an allative can have both an adverbial and an adnominal use in English? Or, is it, rather, that the phrase "the road to Tokyo" is basically a clause (maybe something like "the road [over which one would have to travel in order to go] to Tokyo") with a lot of material suppressed (a very Chomskyish solution)?

When the verb has an object, it could theoretically be ambiguous between adverbial and adjectival, but it would be pretty hard to find one that actually is ambiguous.

I didn't understand that sentence. What do you mean? Can you clarify and exemplify?

(Interestingly, the syntax for prepositional phrases is very similar to participles.)

In Japanese, it turns an adverbial case into an adjectival by adding the genitive marker.

Adverbial: T?ky? e itta. (Tokyo ALLATIVE went.)
Adjectival: T?ky? e no michi (Tokyo ALLATIVE GENITIVE road)

In other words, a kind of "case-stacking". This seems extremely natural to me, but many languages do not allow case-stacking; I wouldn't know what was natural in such languages. (My L1 is _not_ a cased language, so none of it is really "natural" to me, I guess.)

What do other langs do? One possiblity is to replace adjectival cases with relative clauses, so that the case within the clause is still an adverbial one: "The road which goes to Tokyo". Does anyone know of a language where that's the only way of expressing things like this?

I'd like to know, too.

I answered the parts of your post for which I was unable to find answers among those already posted.

I want to know, when you know!

Thanks.

Tom H.C. in MI

[¡Papapishu!]

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Interesting. Can you give any specific examples?

In English, it's done with syntax: the prepositional phrase follows the noun it modifies.

Adverbial: I went to Tokyo. ~ To Tokyo I went.
Adjectival: The road to Tokyo

So, what is the name of the case of "to Tokyo"? Since it modifies the noun "road", it's adnominal; does that mean it can't be the same case as the allative (adverbial) "to Tokyo" in "I went to Tokyo" or "I sent John to Tokyo"? Or is it merely that an allative can have both an adverbial and an adnominal use in English?

An allative can have both an adverbial and an adnominal use in English. And not just an allative, but many other prepositions can too. Basically, English doesn't really distinguish between adverbial and adnominal cases.

When the verb has an object, it could theoretically be ambiguous between adverbial and adjectival, but it would be pretty hard to find one that actually is ambiguous.

I didn't understand that sentence. What do you mean? Can you clarify and exemplify?

Nevermind, I found an example:

"I hit the guy with the stick."

"with the stick" could be interpretted as adverbial, meaning the stick is the weapon with which the guy was hit. Or it could be interpretted as adnominal, meaning the guy was holding a stick at the time he was hit.

However, in the sentence "A guy with a stick hit me.", "with a stick" is unambiguously adnominal, since adverbials can't occur in that position.

[TomHChappell]

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Interesting. Can you give any specific examples?

For examples:
See
http://www.alphadictionary.com/rusgramm ... excpt.html
and
http://www.geocities.com/CollegePark/Bo ... asics.html
and
http://www.helsinki.fi/jarj/siula/kirja ... L0704.html
and
http://russianmentor.net/gram/mailbag/topics/gen2.htm
and
http://www.degruyter.de/downloads/3110182874wiese.pdf
and
http://vn.vladnews.ru/rff_archive/russi ... gners5.htm
and
http://www.surrey.ac.uk/LIS/MB/Bibliography.htm
and
http://russianmentor.net/gram/mailbag/index.htm

It looks like all but one of the examples is Russian; which has a "partititive genitive" and a "genitive of negation", one or the other or both or neither of which might be the "Genitive II".
One of the examples seems to be Finnish, or at least some Finno-Ugric language; unfortunately I do not read the language it appears to be written in, but that's my interpretation of "Suomalais-ugrilaisen seura"

In English, it's done with syntax: the prepositional phrase follows the noun it modifies.

Adverbial: I went to Tokyo. ~ To Tokyo I went.
Adjectival: The road to Tokyo

So, what is the name of the case of "to Tokyo"? Since it modifies the noun "road", it's adnominal; does that mean it can't be the same case as the allative (adverbial) "to Tokyo" in "I went to Tokyo" or "I sent John to Tokyo"? Or is it merely that an allative can have both an adverbial and an adnominal use in English?

An allative can have both an adverbial and an adnominal use in English. And not just an allative, but many other prepositions can too. Basically, English doesn't really distinguish between adverbial and adnominal cases.

Yes, I think it's clear you're right about that.
I wonder if I know enough to help you?
I certainly don't know enough to simply answer your question, but I hope you'll let me know the answer when you find it.

When the verb has an object, it could theoretically be ambiguous between adverbial and adjectival, but it would be pretty hard to find one that actually is ambiguous.

I didn't understand that sentence. What do you mean? Can you clarify and exemplify?

Nevermind, I found an example:

"I hit the guy with the stick."

"with the stick" could be interpretted as adverbial, meaning the stick is the weapon with which the guy was hit. Or it could be interpretted as adnominal, meaning the guy was holding a stick at the time he was hit.

However, in the sentence "A guy with a stick hit me.", "with a stick" is unambiguously adnominal, since adverbials can't occur in that position.

Great example; very clear, now.
The adverbial use, "I, with a stick, hit the guy" meaning "I used a stick to hit the guy", is Instrumental, the adnominal case "I hit the guy-with-a-stick" meaning "I hit the stick-holding guy", is Comitative.
I've heard it said that, diachronically, Instrumentals usually (or, at least, often) "descend" from Comitatives.
Of course, English uses one preposition, "with", for both uses; but, actually, other prepositions are sometimes used for these;
"I went to visit the guy by car" vs "I went to visit the guy by the car"
"I went to visit the guy by the railroad"
and I'm running out of time, but I'm pretty sure there is at least one other English preposition with two uses, one of which is Instrumental, or one of which is Comitative, and the other not. Chances are, also, that one use is adverbial and the other adnominal -- that way one could get ambiguous sentences out of them, too.

Tom H.C. in MI

[hwhatting]

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Interesting. Can you give any specific examples?

For examples:
See
http://www.alphadictionary.com/rusgramm ... excpt.html
and
http://www.geocities.com/CollegePark/Bo ... asics.html
and
http://www.helsinki.fi/jarj/siula/kirja ... L0704.html
and
http://russianmentor.net/gram/mailbag/topics/gen2.htm
and
http://www.degruyter.de/downloads/3110182874wiese.pdf
and
http://vn.vladnews.ru/rff_archive/russi ... gners5.htm
and
http://www.surrey.ac.uk/LIS/MB/Bibliography.htm
and
http://russianmentor.net/gram/mailbag/index.htm

It looks like all but one of the examples is Russian; which has a "partititive genitive" and a "genitive of negation", one or the other or both or neither of which might be the "Genitive II".

I didn't look at all of these, but I don't think that it makes sense to refer to the genitive of negation as "genitive II". While the partitive genitive is formally different from the "ordinary" genitive (ending in -u instead of -a), and therefore qualifies as an extra case (although of marginal use, differentiated only in a small group of nouns, and seeming to become obsolete in the contemporary language), the "genitive of negation" shows no such differentiation - it is just one usage of the normal genitive besides others. Whoever calls that one "genitive II" has, by my book, done a sloppy analysis of the Russian case system.
Best regards,

Hans

[TomHChappell]

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Interesting. Can you give any specific examples?

For examples:
See
[snip]
It looks like all but one of the examples is Russian; which has a "partititive genitive" and a "genitive of negation", one or the other or both or neither of which might be the "Genitive II".

I didn't look at all of these, but I don't think that it makes sense to refer to the genitive of negation as "genitive II". While the partitive genitive is formally different from the "ordinary" genitive (ending in -u instead of -a), and therefore qualifies as an extra case (although of marginal use, differentiated only in a small group of nouns, and seeming to become obsolete in the contemporary language), the "genitive of negation" shows no such differentiation - it is just one usage of the normal genitive besides others. Whoever calls that one "genitive II" has, by my book, done a sloppy analysis of the Russian case system.

That would be me; and what I did a hurried job of was reading those references. I think most of the nine that refer to Russian (probably) call the "partitive genitive" Gentive II, and (probably) don't call the "genitive of negation" Genitive II.

Best regards,

Hans

Thanks, Hans.

[TomHChappell]

The temptation to call any adnominal case "genitive" seems to be so strong that some language's adnominal cases are just called "Genitive I", "Genitive II", etc.

Interesting. Can you give any specific examples?

I found another example; it's Tsez.

(According to Bernard Comrie's "Valency-Changing Derivations in Tsez" (Chapter 11 of Dixon & Aikhenvald's "Changing Valency"),
"Tsez, also known as Dido, one of the Tsezic languages spoken in western Daghestan in the Caucasus, and in lowland areas of Daghestan. Traditionally" [Tsez is part of Tsezic which is part of Avar-Andi-Tsez which is part of Daghestanian which is part of Nakh-Daghestanian=N.E.Caucasian=E.Caucasian].)

"... A fruther point relative to noun declension is that Tsez has two genitive cases, genitive-1 in "-s" which is used when the genitive's head is a noun in the absolutive case, genitive-2 in "-z" when the gentive's head is a noun in any other case."

I think I didn't find it the first time because I "Googled" on Genitive-II rather than on Genitive-2.

Doing so I find suggestions that modern French has two genitives; a suggestion that English's "of" genitive and English's "Saxon" genitive (I think that's the "-'s" genitive) should be treated as Genitive-1 and Genitive-2; and a suggestion that modern English pronouns have two genitives. Genitive-1 = {my,our,your,her,their}; Genitive-2 = {mine,ours,yours,hers,theirs}.
See for example
http://www.sil.org/linguistics/glossary ... ryid=11451
http://www.uoregon.edu/~tpayne/engram/E ... tion03.pdf

The following URL
http://courses.nus.edu.sg/course/ellmd/ ... Saweru.pdf
says that the Oceanic language Saweru has two genitives; Genitive-1 is used for inalienable possession among other uses, and Genitive-2 is used for alienable possession among other uses, if I understood correctly.

http://www.bohemica.com/czechonline/ref ... nitive.htm
may be suggesting that Czech has two genitives -- one basically possessive, the other basically partitive.

http://lists.ibiblio.org/pipermail/b-gr ... 13240.html
suggests Biblical Greek "had four genitives"; but I don't think it really means four different cases, I think it means four different readings of the single genitive case.
To quote the authors,
"1) an objective genitive,
(2) a subjective genitive,
(3) an attributive genitive, or
(4) as a genitive of apposition.
(1) objective genitive: "for obedience to faith" -- but is faith a person to whom one offers obedience? hardly.
(2) subjective genitive: "faith's obedience" -- but is faith a person who can obey?
(3) attributive genitive: "obedience of faith" = "obedience associated with/dependent upon faith" = "faithful obedience" -- that's what strikes me as most appropriate in this instance.
(4) genitive of apposition: "obedience, i.e. faith" -- this doesn't make a lot of sense to me.
"

There's also more about Russian's two genitives.

Tom H.C. in MI

[Travis B.]

Doing so I find suggestions that modern French has two genitives; a suggestion that English's "of" genitive and English's "Saxon" genitive (I think that's the "-'s" genitive) should be treated as Genitive-1 and Genitive-2; and a suggestion that modern English pronouns have two genitives. Genitive-1 = {my,our,your,her,their}; Genitive-2 = {mine,ours,yours,hers,theirs}.
See for example
http://www.sil.org/linguistics/glossary ... ryid=11451
http://www.uoregon.edu/~tpayne/engram/E ... tion03.pdf

This is a rather interesting mapping, in that cases which involve the use of of with personal pronouns would actually be using said hypothetical "genitive-1" and "genitive-2" simultaneously, rather than just "genitive-1" alone. This is shown by the phrases:

• That is a house of mine
  *That is a house of my
  *That is a house of me

As shown here, when using of with personal pronouns, which would be treated as being "genitive-1", one must use genitive pronouns (not possessive determiners) of the form mine, yours, theirs, etc., which are denoted above as being "genitive-2". Even though I haven't spent the time to delve through the linked information above, it would be interesting to see how these cases are handled in this scheme then.

[Glossotechnites]

(4) genitive of apposition: "obedience, i.e. faith" -- this doesn't make a lot of sense to me.

Genitive apposition, as in English: the city of London = the city, i.e./that is to say London.

I found a (Dutch) website about the genitive's functions in Latin.
Latin has 12 Genitives, i.e. 12 genitive functions; apart from the genitive possessive and the four Greek genitives you mentioned, it has:
Clarifying Genitive: modifies and clarifies a general term (expressed by the noun that is used with the genitive) by explaining what the term means, such as: a reward in money -- praemium pecuniae
Genitive Descriptive: expresses the physical or mental quality with the name of the person or thing that has this quality (always used with an adjective), such as: a man of great courage -- vir magnae virtutis
Genitive of Measure: indicating the distance, the lenght, the quantity, the age etc. of something, such as: a month of thirty days -- mensis triginta dierum
Genitive of Indefinite Value: I sold my country estate for a high price -- Magni pretii villam vendidi
Genitive Adjective Object: I have pot full of gold -- Ollam auri plenam habeo.
Genitive Part of a Greater Unit: the bravest of them all -- horum omnium fortissimi
Genitive Adjunct of Place: The consuls are located in Rome -- Consules Romae versabantur

[Echobeats]

Genitive Adjunct of Place: The consuls are located in Rome -- Consules Romae versabantur

Actually, that one is a locative, which happens to be identical to the genitive/dative in the first declension. If that were a masculine singular town, say, the ending would be -o, not -i, and if it were a grammatically plural town, it would be -is or -ibus, not -orum, -arum or -um.

Yours, Tim.

[hwhatting]

I found a (Dutch) website about the genitive's functions in Latin.
Latin has 12 Genitives, i.e. 12 genitive functions; apart from the genitive possessive and the four Greek genitives you mentioned, it has: (snip)

Let's keep clear on this - Latin has one Genitive with several functions! The question of how many funtions a case really has is debatable - like in every discipline, you have lumpers and splitters; additionally, in Latin (and Greek) you have also not only linguists classifying and sorting, but also philologists, grammarians, school teaching traditions... Such enumerations, from my point of view, make sense only in two cases:
a) pedagogical (pupil asking "but why does Cicero use the Genitive here" - teacher anwering "it's the genitivus stupiditatis");
b) for cross-linguistical comparison ("Latin uses the genitive for purpose X, while Finnish uses the partitive").
The discussion about Genitive 1,2,3 etc. above was concerned with languages that have two or more formally different cases that are called genitive I, II, etc. (for lack of imagination of the grammarians, or because both/all these cases have been referred to as genitives in the grammatical tradition).
And, as was said before, Romae in the case you quoted is an archaic locative, not a genitive.
Best regards,

Hans-Werner

[Echobeats]

"it's the genitivus stupiditatis"

Stultorum estis!

[TomHChappell]

I found a (Dutch) website about the genitive's functions in Latin.
Latin has 12 Genitives, i.e. 12 genitive functions; apart from the genitive possessive and the four Greek genitives you mentioned, it has: (snip)

Let's keep clear on this - Latin has one Genitive with several functions! The question of how many funtions a case really has is debatable - like in every discipline, you have lumpers and splitters; additionally, in Latin (and Greek) you have also not only linguists classifying and sorting, but also philologists, grammarians, school teaching traditions... Such enumerations, from my point of view, make sense only in two cases:
a) pedagogical (pupil asking "but why does Cicero use the Genitive here" - teacher anwering "it's the genitivus stupiditatis");
b) for cross-linguistical comparison ("Latin uses the genitive for purpose X, while Finnish uses the partitive").
The discussion about Genitive 1,2,3 etc. above was concerned with languages that have two or more formally different cases that are called genitive I, II, etc. (for lack of imagination of the grammarians, or because both/all these cases have been referred to as genitives in the grammatical tradition).
And, as was said before, Romae in the case you quoted is an archaic locative, not a genitive.
Best regards,

Hans-Werner

Hans is right.

My Tsez, Saweru, Czech, and Russian, (and, possibly, Finnish, English, or French) examples were languages with two different genitive cases.

The Biblical Greek example was four different uses of just one genitive case.

I wouldn't doubt English's genitive(s) also have around a dozen uses -- to give myself a lot of leeway here, between half-a-dozen and two-dozen.

The same would apply probably to Biblical Greek's genitive(s). It's just that in the particular discussion I was quoting, the author proposed just four of them for the passage under discussion.

---

I don't know whether any form of Greek actually had two different "genitive" cases.

I was asked to supply examples of languages with two or more acknowledgedly-different cases, both of which were called "Genitive" by "the experts". I did point to Russian and Czech and Tsez, and showed where some Finnish- and French- and English-discussing sites spoke of possibly more than one genitive.

---

Other contributors to this thread appear to know more about some of the above languages than I do. What do you know of them?

Thanks,

Tom H.C. in MI

[Drydic]

http://courses.nus.edu.sg/course/ellmd/ ... Saweru.pdf
says that the Oceanic language Saweru has two genitives; Genitive-1 is used for inalienable possession among other uses, and Genitive-2 is used for alienable possession among other uses, if I understood correctly.

I would just like to note that the majority of Polynesian languages have this distinction.

And the only reason I say majority is because I don't have a very extensive knowledge of the various Polynesian tongues, as far as I know, they all still have it.

[Zhen Lin]

And the only reason I say majority is because I don't have a very extensive knowledge of the various Polynesian tongues, as far as I know, they all still have it.

The only Malayo-Polynesian language I know is Malay, and it doesn't even have a genitive construct distinct from adjectivial modification, let alone alienable/inalienable possession.

[Drydic]

And the only reason I say majority is because I don't have a very extensive knowledge of the various Polynesian tongues, as far as I know, they all still have it.

The only Malayo-Polynesian language I know is Malay, and it doesn't even have a genitive construct distinct from adjectivial modification, let alone alienable/inalienable possession.

I knew this, and that's why I said Polynesian, which is quite aways away from Malay.

[TomHChappell]

Doing so I find suggestions that modern French has two genitives; a suggestion that English's "of" genitive and English's "Saxon" genitive (I think that's the "-'s" genitive) should be treated as Genitive-1 and Genitive-2; and a suggestion that modern English pronouns have two genitives. Genitive-1 = {my,our,your,her,their}; Genitive-2 = {mine,ours,yours,hers,theirs}.
See for example
http://www.sil.org/linguistics/glossary ... ryid=11451
http://www.uoregon.edu/~tpayne/engram/E ... tion03.pdf

This is a rather interesting mapping, in that cases which involve the use of of with personal pronouns would actually be using said hypothetical "genitive-1" and "genitive-2" simultaneously, rather than just "genitive-1" alone.

Before I go on to agree with you that, "yes, the examples you give below really are interesting", I want to address the possibility that you've misunderstood me.

As far as I know, no such mapping has been proposed. The writers who think the "of" genitive of nouns is different from the "'s" genitive of nouns, are not the same writers as the writers who think the genitive of the pronouns {my,our,your,her,their} is different from the genitive of the pronouns {mine,ours,yours,hers,theirs}.

Both groups of writers needed some term to call each genitive case.
The writers on the nouns chose (in line with tradition but not very creatively) to call one of them "genitive-1" and the other "genitive-2". The writers on the pronouns also chose (in line with tradition but not very creatively) to call one of them "genitive-1" and the other "genitive-2".
It's probably happenstance that the noun-writers used "genitive-1" for "of" and "genitive-2" for "'s" instead of the other way around. It's probably happenstance that the pronoun-writers used "genitive-1" for {my, our, your, her, their} and "genitive-2" for {mine, ours, yours, hers, theirs}.

So it's surely happenstance that the "of" case of the nouns and the {mine, ours, yours, hers, theirs} case of the pronouns didn't happen to wind up with the same label; either both "genitive-1" or both "genitive-2".

This is shown by the phrases:

• That is a house of mine
  *That is a house of my
  *That is a house of me

As shown here, when using of with personal pronouns, which would be treated as being "genitive-1", one must use genitive pronouns (not possessive determiners) of the form mine, yours, theirs, etc., which are denoted above as being "genitive-2". Even though I haven't spent the time to delve through the linked information above, it would be interesting to see how these cases are handled in this scheme then.

It sure would, IMO.

You know by now, if you read what I've written so far, that you are the first person to propose such a scheme (AFAIK)?

But I think it would be interesting, even if it occurs only in a conlang and never in a natlang.
AFAIK nobody has seriously proposed it as something that happens in English (even you proposed it only accidentally IIUC).
If it does happen in some natlang, does anyone know of an example?
Does anyone think a serious case can be made that it happens in English?
Does anyone know of any professional linguist who has made such a case?

---
Tom H.C. in MI

[zompist]

Every language has the same range of choices:
to the best of my memory they are the following:

* lexical suppletion (like "me" and "I")

* stem modification (like "you" and "your" and "yours")

* affixes (almost always suffixes to indicate Case; but prefixes and infixes and circumfixes and superfixes and transfixes could apply as well, especially for other categories than Case)

* agreement -- some other word is modified to indicate the case of this word (for instance in English, verbs agree only with Subjects, never with Objects)

* periphrasis -- adding a word or clitic (usually an adposition -- a preposition or a postposition or an imposition -- for Case)

* pure syntax -- mostly means "word order" -- e.g. in SVO languages the S and the O might not be differently case-marked, so that "Ess Vees Oh" means "Ess" is the Subject and "Oh" is the Object, whereas "Oh Vees Ess" means "Oh" is the Object and "Ess" is the Subject.

When someone says "These are the only options" it's a challenge to find more. I think I've got one you left out, which we might call case assignment by omission. If an argument is left out, there is probably a default rule that gives the case for the remaining argument(s). In English, if a sentence has only one NP present, it's almost always a nominative subject. If two arguments are present, they're nominative and accusative (rather than, say, nom + dat, or dat + acc).

[Travis B.]

Doing so I find suggestions that modern French has two genitives; a suggestion that English's "of" genitive and English's "Saxon" genitive (I think that's the "-'s" genitive) should be treated as Genitive-1 and Genitive-2; and a suggestion that modern English pronouns have two genitives. Genitive-1 = {my,our,your,her,their}; Genitive-2 = {mine,ours,yours,hers,theirs}.
See for example
http://www.sil.org/linguistics/glossary ... ryid=11451
http://www.uoregon.edu/~tpayne/engram/E ... tion03.pdf

This is a rather interesting mapping, in that cases which involve the use of of with personal pronouns would actually be using said hypothetical "genitive-1" and "genitive-2" simultaneously, rather than just "genitive-1" alone.

Before I go on to agree with you that, "yes, the examples you give below really are interesting", I want to address the possibility that you've misunderstood me.

As far as I know, no such mapping has been proposed. The writers who think the "of" genitive of nouns is different from the "'s" genitive of nouns, are not the same writers as the writers who think the genitive of the pronouns {my,our,your,her,their} is different from the genitive of the pronouns {mine,ours,yours,hers,theirs}.

Both groups of writers needed some term to call each genitive case.
The writers on the nouns chose (in line with tradition but not very creatively) to call one of them "genitive-1" and the other "genitive-2". The writers on the pronouns also chose (in line with tradition but not very creatively) to call one of them "genitive-1" and the other "genitive-2".
It's probably happenstance that the noun-writers used "genitive-1" for "of" and "genitive-2" for "'s" instead of the other way around. It's probably happenstance that the pronoun-writers used "genitive-1" for {my, our, your, her, their} and "genitive-2" for {mine, ours, yours, hers, theirs}.

So it's surely happenstance that the "of" case of the nouns and the {mine, ours, yours, hers, theirs} case of the pronouns didn't happen to wind up with the same label; either both "genitive-1" or both "genitive-2".

This is shown by the phrases:

• That is a house of mine
  *That is a house of my
  *That is a house of me

As shown here, when using of with personal pronouns, which would be treated as being "genitive-1", one must use genitive pronouns (not possessive determiners) of the form mine, yours, theirs, etc., which are denoted above as being "genitive-2". Even though I haven't spent the time to delve through the linked information above, it would be interesting to see how these cases are handled in this scheme then.

It sure would, IMO.

You know by now, if you read what I've written so far, that you are the first person to propose such a scheme (AFAIK)?

But I think it would be interesting, even if it occurs only in a conlang and never in a natlang.
AFAIK nobody has seriously proposed it as something that happens in English (even you proposed it only accidentally IIUC).
If it does happen in some natlang, does anyone know of an example?
Does anyone think a serious case can be made that it happens in English?
Does anyone know of any professional linguist who has made such a case?

---
Tom H.C. in MI

The main thing is that my statements above are more an analysis of the scheme preposed before than something that I myself have advocated. I myself would prefer a scheme that simply divides the ways of expressing possession into genitive case (note that such is expressed via a postpositional clitic when not used with personal pronouns), possessive determiners, and the preposition of. Note that I do not try to link of with any sort of syntactic case explicitly, but rather treat it in a fashion parallel to German von and Dutch van. Also note that I treat my, your, her, our, and their as solely possessive determiners, mine, yours, hers, ours, and theirs as solely genitive case, and his and its a possessive determiners or genitive case by context.

Yes, you could try to link of with some kind of explicit syntactic genitive case, buy that would require also doing things like linking to with, say, syntactic dative case for the sake of consistency. As I myself do not see any reason to really explicitly identify prepositions with cases, I would prefer to not think of of in terms of cases myself.

[TomHChappell]

Every language has the same range of choices:
to the best of my memory they are the following:
* lexical suppletion (like "me" and "I")
* stem modification (like "you" and "your" and "yours")
* affixes [snip]
* agreement -- some other word is modified to indicate the case [snip]
* periphrasis -- adding a word or clitic [snip]
* pure syntax -- mostly means "word order" [snip]

When someone says "These are the only options" it's a challenge to find more.

Yep.

I think I've got one you left out, which we might call case assignment by omission. If an argument is left out, there is probably a default rule that gives the case for the remaining argument(s).

You may be right, and I guess you probably are right.

This is one of the defining properties of "subjects", according to the "Subject Properties List" which I will re-check out of the library later this afternoon.

So I think this technique works to assign case only (or almost only) when the case so assigned is the case of the Subject (whatever that is -- even in most ergative languages the Agent is the Subject).

Do you know of natlang cases other than "Nominative or whatever the heck we're calling the Subject" that get assigned this way?

In English, if a sentence has only one NP present, it's almost always a nominative subject. If two arguments are present, they're nominative and accusative (rather than, say, nom + dat, or dat + acc).

In English and most other natlangs, if an NP is omitted, it is usually both the Subject of the clause from which it is omitted, and also the Subject of the clause from which it can be "picked up" (for example, it can be omitted from a dependent clause and "picked up" from a main clause; or omitted from a later clause and "picked up" from the preceding clause).

It is almost always the Subject of one clause or the other.

This is true even in languages which are not Subject-Prominent, provided both clauses do have Subjects.

Most languages have Subjects, whether or not most clauses in those languages have Subjects. (Some linguists think all languages have Subjects; but even these do not think all languages are Subject-Prominenent.)

Of course, what's a Subject in one language is not necessarily the same as what's a Subject in the next language.

-----
Tom H.C. in MI

[zompist]

I think I've got one you left out, which we might call case assignment by omission. If an argument is left out, there is probably a default rule that gives the case for the remaining argument(s).

This is one of the defining properties of "subjects", according to the "Subject Properties List" which I will re-check out of the library later this afternoon.

So I think this technique works to assign case only (or almost only) when the case so assigned is the case of the Subject (whatever that is -- even in most ergative languages the Agent is the Subject).

Do you know of natlang cases other than "Nominative or whatever the heck we're calling the Subject" that get assigned this way?

Just to be clear, I was not using "nominative" to mean "subject". You mention ergative languages yourself; if they have this sort of rule it presumably assigns default absolutive case.

I don't know enough Mandarin to use it as an example, but it'd be worth looking at to see if it offers some. I do know that it often organizes sentences by topic/comment and that topics don't have to be primary arguments of the verb.

The other languages I know well tend to have explicit case marking of one sort or another, so they're not much help.

I used this sort of process in Kebreni, with the reverse rule from English: with "NP V", where V is transitive verb, the NP is taken as the accusative.

In English, if a sentence has only one NP present, it's almost always a nominative subject. If two arguments are present, they're nominative and accusative (rather than, say, nom + dat, or dat + acc).

In English and most other natlangs, if an NP is omitted, it is usually both the Subject of the clause from which it is omitted, and also the Subject of the clause from which it can be "picked up" (for example, it can be omitted from a dependent clause and "picked up" from a main clause; or omitted from a later clause and "picked up" from the preceding clause).

It is almost always the Subject of one clause or the other.

Maybe you should give some examples, since I'm not sure if your statement contradicts mine or not.

I was thinking about sentences like "I mailed the book", where the number of arguments alone tells us that neither argument is a dative. We can't say "*I mailed John" (meaning John is the recipient, not the cargo). On the other hand we can say "I told John", so my observation may be quite wrong, or at best it needs refining.

[TomHChappell]

[snip]
Just to be clear, I was not using "nominative" to mean "subject".

Yes, I understood that.

You mention ergative languages yourself; if they have this sort of rule it presumably assigns default absolutive case.

That's what you'd think, but it isn't so in most cases. In most ergative languages the omitted argument is absolutive for intransitive clauses but ergative for transitive clauses.

There are some ergative languages in which the omitted argument is always (or almost always) absolutive. But they are a minority even within the ergative languages.

For most ergative languages the "Subject" is exactly the same as it is for languages with other morphosyntactic alignments.

Weird, huh?

I don't know enough Mandarin to use it as an example, but it'd be worth looking at to see if it offers some. I do know that it often organizes sentences by topic/comment and that topics don't have to be primary arguments of the verb.

It's one of those discussed in the book I just rechecked. I'll see if the book says.
It's true that for most (or at least many) topic-prominent languages the topic doesn't need to be a participant of the comment; but, if it is, it is probably the subject.
It's also true that for (most or many) T.P. languages, the Topic takes over many of the properties that the Subject has in S.P.-non-T.P. languages.

The other languages I know well tend to have explicit case marking of one sort or another, so they're not much help.

I used this sort of process in Kebreni, with the reverse rule from English: with "NP V", where V is transitive verb, the NP is taken as the accusative.

Maybe you should give some examples, since I'm not sure if your statement contradicts mine or not.

In Japanese, the head noun of a relative clause is always its Topic; and it is the noun modified by the RC, regardless of its role in the main clause.

In Philippine languages, on the other hand, the head noun of a relative clause is always its Focus; and it is the noun modified by the RC, regardless of its role in the main clause.

In languages that don't have such big morphosyntactic effects of Topichood or Focushood, the head noun of a RC is usually its Subject. There is a hierarchy of which roles in the RC can be relativized; the Subject is at the top, so that, if anything at all can be relativized, the Subject can.

I was thinking about sentences like "I mailed the book", where the number of arguments alone tells us that neither argument is a dative.

It does?

We can't say "*I mailed John" (meaning John is the recipient, not the cargo). On the other hand we can say "I told John", so my observation may be quite wrong, or at best it needs refining.

Probably only means it doesn't apply to English, or at least not to every English verb.

You mean
"I told John the story"
"I told John"
"I told the story"
?

Or, better,
"I showed the principal the tardy student"
"I showed the principal"
"I showed the tardy student"
?

(It's true none of these are subject-less; but that's because English is a subject-prominent language, whatever S.P. means.)

I still think you are probably right. I just think that the assignment of "whatever case the Subject of this clause would have to be in if this clause's Subject were explicit" is the one that will most often get assigned this way. Note I say most often, not necessarily always.


I wish I could come up with better examples, too. Sorry I haven't yet. But I'm not territorial about the problem; if anyone else comes up with better examples, I'll appreciate it.

-----
Tom H.C. in MI