Talk:Tiki forum archive/Forum 5/7964
Source: 2013 Tiki forum 5, thread #7964. Original authors, UTC times, post IDs, and reply-parent IDs are preserved. This is an archive, not an active article discussion.
nominative-accusative & ergative-absolutive
nominative-accusative & ergative-absolutive
tijlan <[email protected]> · 2009-02-16 10:32 UTC · Tiki post #7964
nominative-accusative type (e.g. English, Japanese, Esperanto) and the ergative-absolutive type (e.g. Basque, Tibetan, Mayan)? Compare the following Basque-Esperanto-Lojban sentences:
Gizon-a | etorri da. Vir-o | ekalvenis. lo nanmu | co'i/ba'o klama
Gizon-ak | mutil-a | ikusi du. Vir-o | knab-on | vidis. lo nanmu | lo nanla | pu viska
"-a" in Basque is the absolutive suffix, which is translated in Esperanto as the nominative suffix "-o". Then "-ak" is the ergative, translated in Esperanto again as the nominative, and this time the Basque absolutive "-a" becomes the accusative in Esperanto. Esperanto, because it runs on an explicit case system, cannot handle the ergative-absolutive languages without explicitly distorting their original typological quality. In Lojban, on the other hand, cases are not (at least morphologically) specified. It therefore buffers the difference between the nominative-accusative and ergative-absolutive languages. It allows a Basque/Tibetan/Mayan etc. speaker to communicate with an English/Japanese/Esperanto etc. speaker without being distracted at the aforementioned dissimilarity.
pe'ipei mu'o mi'e tijlan
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
nominative-accusative & ergative-absolutive
Adam Raizen <[email protected]> · 2009-02-16 11:01 UTC · Tiki post #7965 · reply to Tiki post #7964
> Gizon-a | etorri da. > Vir-o | ekalvenis. > lo nanmu | co'i/ba'o klama > > Gizon-ak | mutil-a | ikusi du. > Vir-o | knab-on | vidis. > lo nanmu | lo nanla | pu viska > In Lojban, on the other hand, cases are
[...]
> It therefore buffers the
> difference between the nominative-accusative and ergative-absolutive
> languages. It allows a Basque/Tibetan/Mayan etc. speaker to
> communicate with an English/Japanese/Esperanto etc. speaker without
> being distracted at the aforementioned dissimilarity.
>
So then does English also buffer the difference between the two, since it also does not morphologically specify case? I think Lojban is as nominative-accusative as it gets. (Predicate logic was invented by people who spoke and were familiar with nominative-accusative languages.) If the correct Lojban for the first sentence were "klama lo nanmu" (not "klama *fa* lo nanmu), then we could probably say that Lojban is ergative-absolutive.
--
Adam Raizen <[email protected]>
nominative-accusative & ergative-absolutive
tijlan <[email protected]> · 2009-02-16 12:41 UTC · Tiki post #7966 · reply to Tiki post #7964
> So then does English also buffer the difference between the two, since it > also does not morphologically specify case?
Though irregularly, it does specify for at least some of the pronouns. In "I saw him", the accusative case of "him" is morphologically specified (in contrast to "he"). So is "I".
> I think Lojban is as
> nominative-accusative as it gets. (Predicate logic was invented by people
> who spoke and were familiar with nominative-accusative languages.) If the
> correct Lojban for the first sentence were "klama lo nanmu" (not "klama *fa*
> lo nanmu), then we could probably say that Lojban is ergative-absolutive.
Syntactic ergativity is quite rare (http://en.wikipedia.org/wiki/Ergative-absolutive_language). And I don't see why the x2 of "klama" being the agent/subject would be so indicative of an ergative-absolutive syntax. The Basque example already shows that the subject doesn't necessarily follows the verb in an ergative-absolutive language. I've learned that there are also nominative-accusative languages where the verb preceds the subject, such as formal Arabic, Gaelic, Hawaiian, and, to a lesser extent, Romanian, Hungarian and Finnish.
mu'o mi'e tijlan
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
nominative-accusative & ergative-absolutive
Adam Raizen <[email protected]> · 2009-02-16 12:54 UTC · Tiki post #7967 · reply to Tiki post #7964
> > I think Lojban is as > > nominative-accusative as it gets. (Predicate logic was invented by people > > who spoke and were familiar with nominative-accusative languages.) If the > > correct Lojban for the first sentence were "klama lo nanmu" (not "klama > *fa* > > lo nanmu), then we could probably say that Lojban is ergative-absolutive. > > Syntactic ergativity is quite rare > (http://en.wikipedia.org/wiki/Ergative-absolutive_language).
There's no morphological inflection in Lojban, so whatever morphosyntactic
alignment it has has to be syntactic and not morpho-.
> And I
> don't see why the x2 of "klama" being the agent/subject would be so
> indicative of an ergative-absolutive syntax.
Because then the argument of an intransitive verb is treated the same as the
object of a transitive verb (treating 'klama' as intransitive, which may be
arguable).
> The Basque example
> already shows that the subject doesn't necessarily follows the verb in
> an ergative-absolutive language. I've learned that there are also
> nominative-accusative languages where the verb preceds the subject,
> such as formal Arabic, Gaelic, Hawaiian, and, to a lesser extent,
> Romanian, Hungarian and Finnish.
Of course. VSO languages are not rare, and most of them are accusative, not
ergative. (You can add Biblical Hebrew to the list.) The question isn't the
order of the sentence constituents, but how the argument of an intransitive
verb is treated compared to the subject and object of a transitive verb.
-- Adam Raizen <[email protected]>
Got sente?nominative-accusative & ergative-absolutive
tijlan <[email protected]> · 2009-02-16 13:53 UTC · Tiki post #7968 · reply to Tiki post #7964
> There's no morphological inflection in Lojban, so whatever morphosyntactic > alignment it has has to be syntactic and not morpho-.
Right, I'm not saying that Lojban is morphologically ergative-absolutive, but that it conveniently neutralizes the differing nominative and ergative perspectives.
>> And I
>> don't see why the x2 of "klama" being the agent/subject would be so
>> indicative of an ergative-absolutive syntax.
>
> Because then the argument of an intransitive verb is treated the same as the
> object of a transitive verb (treating 'klama' as intransitive, which may be
> arguable).
However, the object of a transitive verb in an ergative-absolutive language is not agentive, is it? In "Gizonak mutila ikusi du", "mutila" is the object and has the same absolutive suffix as what "gizon-" as the argument/subject of an intransitive verb would take, but it is not agentive; it's not the boy who is seeing himself. Now if a hypothetical place structure of "viska" is such that the x2 is the agent of "seeing", it remains as the subject, not the object, despite the conversion. Likewise, the hypothetical word order in "klama lo nanmu" simply means that "lo nanmu" is the agentive subject of "klama", which could still be used to correspondently translate sentences of either ergative-absolutive ("Etorri da | gizonak.") or nominative-accusative ("Ekalvenis | viro."). The point is that, in Lojban, speakers of either type can talk without unlearning their native perspective of the linguistic alignment, apart from the word order.
mu'o mi'e tijlan
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
nominative-accusative & ergative-absolutive
Adam Raizen <[email protected]> · 2009-02-16 14:22 UTC · Tiki post #7969 · reply to Tiki post #7964
> Now if > a hypothetical place structure of "viska" is such that the x2 is the > agent of "seeing", it remains as the subject, not the object, despite > the conversion.
You mean a selbri with the place structure of 'selviska', but atomic, not
derived by conversion? I don't think so. Even though you changed the order
of the arguments to something very unconventional, my guess is that the
syntactic structure in the heads of people who actually used that word in
that language would put the agent as the object. One would have to do
linguistic experiments with a native speaker to tell for sure.
> Likewise, the hypothetical word order in "klama lo
> nanmu" simply means that "lo nanmu" is the agentive subject of
> "klama", which could still be used to correspondently translate
> sentences of either ergative-absolutive ("Etorri da | gizonak.") or
> nominative-accusative ("Ekalvenis | viro.").
I don't quite follow you. Is "Etorri da gizonak" good Basque? It looks like
it contradicts your above sentence.
> The point is that, in
> Lojban, speakers of either type can talk without unlearning their
> native perspective of the linguistic alignment, apart from the word
> order.
A Basque speaker sees the argument of an intransitive verb (which is always
x1 in Lojban) as occupying the same syntactic role as the object of a
transitive verb (which for our purposes we'll say is x2 in Lojban. It
certainly isn't x1.) A Lojban speaker, who knows the languages, sees x1 and
x2 as syntactically distinct, which is why Lojban's morphosyntactic
alignment is different from Basque's.
-- Adam Raizen <[email protected]>
Got sente?nominative-accusative & ergative-absolutive
PierreAbbat · 2009-02-16 14:35 UTC · Tiki post #7970 · reply to Tiki post #7964
> 2009/2/16 Adam Raizen <[email protected]>: > > There's no morphological inflection in Lojban, so whatever > > morphosyntactic alignment it has has to be syntactic and not morpho-. > > Right, I'm not saying that Lojban is morphologically > ergative-absolutive, but that it conveniently neutralizes the > differing nominative and ergative perspectives.
I say either Lojban doesn't have a morphosyntactic alignment, or it has one that no natlang has.
If I were explaining morphosyntactic alignment in Lojban terms, I'd say that the subject of a transitive verb in an accusative language is x1, whereas in an ergative language it's x2. I'm not sure how I'd explain split ergativity. The one instance in which I know how it arose, which is modern Indic, the past tense was originally a participle, used as a passive with the instrumental.
Pierre
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
nominative-accusative & ergative-absolutive
Adam Raizen <[email protected]> · 2009-02-16 14:57 UTC · Tiki post #7971 · reply to Tiki post #7964
> I say either Lojban doesn't have a morphosyntactic alignment, or it has one > that no natlang has. >
Formally, it is a bit weird, but I think that the actual cognitive structures that speakers of Lojban use are the same as those used by speakers of accusative languages (and that would be the case even if a monolingual native speaker of an ergative language learned Lojban fluently.)
-- Adam Raizen <[email protected]>
Got sente?nominative-accusative & ergative-absolutive
Jorge Llambías <[email protected]> · 2009-02-16 15:58 UTC · Tiki post #7972 · reply to Tiki post #7964
> > Formally, it is a bit weird, but I think that the actual cognitive > structures that speakers of Lojban use are the same as those used by > speakers of accusative languages (and that would be the case even if a > monolingual native speaker of an ergative language learned Lojban fluently.)
I tend to agree.
If Lojban had lots more gismu pairs like citka/cidja, tirna/savru, bacru/voksa, and assuming those pairs were more or less identical in meaning except for the argument order (which is probably not really the case now), then it could be argued that it was neutral with respect to the accusative/ergative divide.
A speaker with accusative tendencies would go with "lo smacu cu ca'o citka lo cirla", and a speaker with ergative tendencies would say "lo cirla cu ca'o cidja lo mlatu", and they would both mean exactly the same thing.
We do have such pairs for all predicates if we count viska/selviska, catra/selcatra, and so on, but the ergative-like one is marked, so there is a bias towards the accusative alignment.
mu'o mi'e xorxes
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
nominative-accusative & ergative-absolutive
slobin · 2009-02-18 16:21 UTC · Tiki post #7973 · reply to Tiki post #7964
> A Basque speaker sees the argument of an intransitive verb (which is always > x1 in Lojban) as occupying the same syntactic role as the object of a > transitive verb (which for our purposes we'll say is x2 in Lojban. It > certainly isn't x1.) A Lojban speaker, who knows the languages, sees x1 and > x2 as syntactically distinct, which is why Lojban's morphosyntactic > alignment is different from Basque's.
This has remembered me a motto from CONLANG list maybe ten or so years ago, during du'emoi discussion on accusative/ergative topic:
"Murder is just transitive death".
-- http://slobin.pp.ru/ `When I use a word,' Humpty Dumpty said, <[email protected]> `it means just what I choose it to mean'
To unsubscribe from this list, send mail to [email protected]
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
