Re: THEORY: Xpositions in Ypositional languages {X,Y}={pre,post}

On Mon, 24 Sep 2007 11:07:34 +0100, R A Brown
<ray@...> wrote:
>Andreas Johansson wrote:
>> Quoting R A Brown <ray@...>:
>[snip]
>>>Which must surely mean that Dryer considers _'s_ in "The guy next door's
>>>wife" to be a postposition; but it's generally considered to be clitic.
>>
>> Dryer's wording surely indicates that he considers it to be BOTH a clitic
>>AND a postposition.
>
>I've just re-read the opening part Dryer's paper again and while he well
>consider _'s_ to be a clitic, he does not treat it as an adposition in
>his paper and, therefore, not as a postposition. I quote:
>"A word is treated here as an adposition (preposition or postposition)
>if it combines with a noun phrase and indicates the grammatical or
>semantic relationship of that noun phrase to the verb in the clause."
>-------------------------------------
I'm not quite certain, from other parts of the paper, that he doesn't regard
case-markers of adnominal cases (such as genitive), that are either free words
or clitics whose position is determined syntactically, as "adpositions"; I get the
feeling that in certain paragraphs he was simply saying that if those were the
only "adpositions" a language had, he would count it as "not having
adpositions" for purposes of this particular count.  Perhaps his decision varied
from one part of the paper to another?  OTOH maybe I'm just wrong.

>Eldin Raigmore wrote:
>>First off: I am still mostly interested in:
>>(1) What is cross-linguistically common among postpositions in
>prepositional  languages?
>
>Personally, I doubt if there is anything common other than the existence
>of a few postpositions in a predominantly prepositional language. Both
>Latin & classical Greek are predominantly prepositional and both have
>postpositions. But I see nothing common to the post positions of both
>languages.
I remember reading something by some professional linguist that said certain
semantics tended to be such postpositions.  I was hoping someone on-list
could either direct me to that paper or to something else like it, since I've lost
track of it.

>>(2) What is cross-linguistically common among prepositions in
>postpositional  languages?
>
>I'm not familiar enough with postpositional languages to comment except
>to say that I strongly suspect we'll find a similar lack of commonality
>of prepositions in those languages.
From the answers I've gotten so far, nobody on-list knows (well, except me,
and I barely "know" of it) of any work supporting such an idea.  I was hoping
someone did, even if they knew it had already been shot down.

>But Dryer's paper simply does not address the issue that you are
>primarily concerned with. He mentions this in the last paragraph on the
>first page, but does not deal with what might cross-linguistically common.
That's true.

>[snip]
>>About inpositions:  As others have pointed out I misunderstood Dryer.
>>As I now understand it (assuming that even now I understand correctly),
>>for him to count something as an adposition _in_that_paper_, it has to be a
>>case-marker which is either a free word or a clitic, whose position (relative
>>to the noun-phrase which is its complement) is _syntactically_ determined
>>(rather than, I assume, _morphologically_ determined).
>
>That is not what I understand. I do not read "has to be" but rather
>"occur or can occur"; I quote:
>"The map also shows a rare type of adposition, what I will call
>inpositions, adpositions which occur or can occur inside the noun phrase
>they accompany."
That "or can occur" causes a problem, as I see it.  If it _must_ occur within its
complement NP, then I guess it would more obviously qualify as an inposition;
if it _can_ within its complement NP, then it's less clear it isn't a preposition or
a postposition; right?  Anyway, you're right, I didn't pay sufficient heed to
Dryer's "or can occur".

>I have already shown that by that definition, classical Latin had these
>so-called inpositions.
If "can" is changed to "must" does Classical Latin still have "these so-called
inpositions"?

>I have also said that I am not convinced by what Dryer writes in this paper
>that inpositions are really a separate category of adposition.
We'd really need a look at his one-step-rawer data, wouldn't we?
It looks like "believing in inpositions" is similar to "believing in Object-Initial
languages".

>[snip]
>>A case-marker which always is inserted into its complement noun-phrase,
>>Dryer calls an inposition for purposes of this paper.  ...
[snip]
>That is certainly not what I understand from Dryer's words and two
>examples.
>
>Firstly, as I have pointed out, he says quite explicitly "adpositions
>which occur _or can occur_ inside the noun phrase they accompany"
>[emphasis is mine}. He clearly does not say that it is always inserted
>in the noun phrase.
Yes, you're right.

>It would seem to me that you have extrapolated what he says about the
>Australian language Anindilyakwa to inpositions generally. ...
[snip]

While it seems far from unlikely that I've misunderstood him (since I did so
earlier, maybe more than once), I don't understand exactly what it is you think
I've misunderstood or exactly how you think I misunderstood it.  Can you try
again to explain what I didn't get?

>>... in Dryer's 1047-language sample he found only 7 languages in
>>which "inpositions" were the dominant type of adposition.
>
>Even Dryer admits it's a pretty rare animal. But what I said is that I
>do not find his two examples convincing evidence that we need a third
>category of adposition, namely 'inposition.' One would dearly like to
>know a good deal more about the structure of Tümpisa Shoshone and the
>six Australian languages that are claimed to be 'inposition dominant.'
We'd really need a look at his one-step-rawer data, wouldn't we?  Or even
rawer than that.

>[snip]
>
>>It also makes it extremely unlikely that it would even be possible to define
>>a "superposition" or a "transposition" that would fit Dryer's paper's definition
>>of an adposition.
>
>On this I agree    :)
Yep.  Nevertheless I appreciate David Peterson's and others' attempts.

>[snip]
>>I have seen adpositions which can occur as either pre- or post-
>>called "circumpositions".  I have also seen obligatorily-paired pre- and post-
>> -positions called "circumpositions".  IMO it would be good to have different
>>terms for the two ideas.  I think the obligatorily-paired pre- and post-
>> -position has the better claim on the term "circumposition"; and some
>>other term should be used for the adpositions which can be either
>>prepositions or postpositions.
>
>On this I also agree.
>
>>But I do not know what to call them.
>
>'ambipositions'???
Believe it or not I actually thought of just that after making my last post but
before reading yours.  If need arises I intend to use it; I'll give you credit.

>But my own feeling is that it is an unfortunate chance of history that we've
>finished up with the three terms 'preposition', 'postposition' and 'adposition'.
>It would have been much better IMO if there were one single term (such
>as 'relator').
>After all we just use the single term 'adjective'; we do not say that
>English had 'prejectives' which French favors 'postjectives'! We say
>that in a certain language adjectives normally precede their nouns while
>in another language they normally follow their nouns. It seems to me
>singularly odd that for the group of words we term 'adpositions' we have
>to resort to terms that have nothing to do with their function but only
>with their position.
It's hard to reform terminology just on the basis that it make sense to.  Even
Claudius couldn't change the alphabet.

>-------------------------------------------------------------------------
>>Lastly I wish to point out: I am still mostly interested in:
>>(1) What is cross-linguistically common among postpositions in prepositional
>>languages?
>>(2) What is cross-linguistically common among prepositions in postpositional
>>languages?
>
>If this is so, then it seems to me that Dryer's paper has thrown us off
>the track.
Probably so; but ...
Well, I did ask several other questions, several of which were answered by
Dryer's paper.  In particular (and for example), most languages are
predominantly either prepositional or postpositional, but some languages, even
though they have adpositions, don't have a predominant type.
I consider those questions crucial preliminaries to the remaining unanswered
questions I'm still mainly interested in.  If most languages weren't either mostly
prepositional or mostly postpositional, there wouldn't be enough data to
answer these last two questions.

>But my own feeling is that (1) is like looking for what is
>cross-linguistically common among post-posited adjectives in those
>languages that normally put adjectives before nouns; and that (2) is
>like looking for what is cross-linguistically common among pre-posited
>adjectives in languages where adjectives normally follow their nouns*.
Some languages have adjectives that can occur either side of the noun,
e.g "un grand homme" vs "un homme grand".  Isn't the semantic difference
between these two rather common among such languages that have such
adjectives?  Likewise some languages' numerals can occur either before or
after the noun, with a semantic difference; isn't the semantic difference in
question, more-than-chance similar between such languages?

>*Oh yes, there is of course a commonality among the Romance languages,
>but move along and consider, say, Welsh, and the commonality begins to
>wear a bit thin.
Maybe what I mentioned above is exactly what you were talking about.  How
does it wear thin?  I know no Celtic languages and only two Romance
languages (or maybe only one).

>Ray
Thanks, Ray. (And thanks everybody who replied to this thread.)

-----
eldin