Re: Argument Structures
| From: | The Gray Wizard <dbell@...> |
|---|
| Date: | Wednesday, August 23, 2000, 18:17 |
|---|
> From: jonathan.jones
> Subject: Argument Structures
>
> Apart from the standard "transitive", "intransitive", and
> "ditransitive" verbs found in English, and the "reflexive" verbs in
> Romance languages, I'm not really familiar with many ways to introduce
> various argument structures into languages. So when working on the
> verbs in my (as yet unnamed) conlang, I tried to start from scratch
> and managed to list the following:
I arrived at a very similar distribution of arguments for amman iar, also
after reading Rick's "Lexical Semantics". In amman iar, predicate argument
structure is driven by the valency marker on the lexical verb.
> "transitive": 1 agent, 1 patient [I taught him]
I taught him.
(ner) eleth en alan erorfainaen
\t ner eleth en alan
\m ner -0 el- -eth en al =an -0
\g I -[A] assertive- -past agt.to.pat it =masc -[P]
\p 1per -nom mood- -tense ptp 3per =gnd -abs
\x I did agt.to.pat he
\t erorfainaen
\m er- or= faino -ae -n
\g do- compl= learn -agt/pat -actn/proc
\p agt- pfx= v -val -vc
\x teach
The agt/pat (-ae) inflection indicates this pattern. There are some other
redundant indicators as well. The position of the auxiliary (eleth) and the
selection of the particle (en) as well as the presence of the agentive (er-)
on the lexical verb all contribute.
> "transitive, focused": 1 agent, 1 patient, 1 focus [I taught him
> Linguistics]
I call the third argument here the "theme", but I think we mean the same
thing.
I taught him linguistics
(ner) eleth en alan iardainnon erorfainiaen
\t ner eleth en alan
\m ner -0 el- -eth en al =an -0
\g I -[A] assertive- -past agt.to.pat it =masc -[P]
\p 1per -nom mood- -tense ptp 3per =gnd -abs
\x I did agt.to.pat he
\t iardainnon
\m iar =tain -on
\g language =science -[Obl]
\p n =nsfx -dat
\x linguistics
\t erorfainiaen
\m er- or= faino -iae -n
\g do- compl= learn -agt.pat.thm -actn/proc
\p agt- pfx= v -val -vc
\x teach
Here the valency marker (-iae) indicates the agt/pat/thm argument structure.
Note that this valency inflection is used primarily in older amman iar
texts. The newer texts prefer to indicate the valency of the core arguments
of the predicate only.
> "reflexive" : 1 agent/patient [I taught myself]
The reflexive in amman iar is implemented as a voice inflection since it
deletes a P-function argument of a transitive predicate when the A- and
P-function arguments both have the same referent.
I taught myself
(ner) eleth ervororfainaen en
\t ner eleth
\m ner -0 el- -eth
\g I -[A] assertive- -past
\p 1per -nom mood- -tense
\x I did
\t ervororfainaen en
\m er- vor- faino -ae -n en
\g do- reflexive- learn -agt/pat -actn/proc agt.to.pat
\p agt- voice- v -val -vc ptp
\x teach.self agt.to.pat
Here the valency (-ae) continues to indicate two arguments (agt/pat), but
the reflexive voice conflates them into a single S-function argument that
plays both semantic roles. Note the unusual word order of this form. In
particular note that the auxiliary verb (eleth) remains in place in
expectation of a patientive P-function argument, but the Patientive Particle
moves to sentence-final position. Note also the absence of a reflexive
pronoun.
> "reflexive, focused" 1 agent/patient, 1 focus [I taught myself
> Linguistics]
I taught myself linguistics
(ner) eleth ervororfainiaen iardainnon en
\t ner eleth
\m ner -0 el- -eth
\g I -[A] assertive- -past
\p 1per -nom mood- -tense
\x I did
\t ervororfainiaen
\m er- vor- faino -iae -n
\g do- reflexive- learn -agt.pat.thm -actn/proc
\p agt- voice- v -val -vc
\x teach.self
\t iardainnon en
\m iar =tain -on en
\g language =science -[Obl] agt.to.pat
\p n =nsfx -dat ptp
\x linguistics agt.to.pat
This continues the pattern.
> "reciprocal": 2 agent/patient sort of:[I taught him and he taught
> me]
Similarly, reciprocals are voice inflections in amman iar. The reciprocal
form of a predicate deletes a P-function argument of a transitive predicate
when the A-function argument consists of more than one entity that share the
predicate relationship with each other.
We taught each other.
eleth en nir ermanorfainaen
\t eleth en nir
\m el- -eth en ner -i -0
\g assertive- -past agt.to.pat I -plu -[A]
\p mood- -tense ptp 1per -num -nom
\x did agt.to.pat nir
\t ervanorfainaen
\m er- man- or= faino -ae -n
\g do- reciprocal- compl= learn -agt/pat -actn/proc
\p agt- voice- pfx= v -val -vc
\x teach.each.other
Note the unusual word order of this form. In particular note that the
auxiliary verb and the Patientive Particle precede the A-function argument
that nevertheless remains in nominative case. This indicates the dual role
of the single surface argument as simultaneously underlying agent
(nominative case) and patient (patientive particle).
> "reciprocal, focused" 2 agent/patient, 1 focus [ we taught each other
> linguistics]
> {actually, I think a better example is trade: [we exchanged
> goods]}
Similarly,
We taught each other linguistics.
eleth en nir iardainnon ermanorfainiaen
\t eleth en nir
\m el- -eth en ner -i -0
\g assertive- -past agt.to.pat I -plu -[A]
\p mood- -tense ptp 1per -num -nom
\x did agt.to.pat nir
\t iardainnon
\m iar =tain -on
\g language =science -[Obl]
\p n =nsfx -dat
\x linguistics
\t ermanorfainiaen
\m er- man- or= faino -iae -n
\g do- reciprocal- compl= learn -agt.pat.thm -actn/proc
\p agt- voice- pfx= v -val -vc
\x teach.each.other
> "intransitive": 1 experiencer [John learnt]
> "intransitive, focused": 1 experiencer, 1 focus [John learnt
> lingustics]
I'm not sure I follow you here. The single argument (S-function) of an
intransitive predicate can play many different roles besides 'experiencer.
How would you categorize these other semantic roles?
>
> Actually, I lie. Some of that was taken from an essay "lexical
> semantics" I found on the net. I forget who wrote it.
Rick Morneau. He used to frequent this list, but I'm not sure he does these
days. Absolute must reading for all conlangers, IMHO.
> Anyway the conlang is VSO, so a verb followed by 3 nouns could mean
> any of those things, with only the vowel sequence of the verb to
> distinguish them. As the language rather hinges on the verbs, I'm a
> little reluctant to press on until I'm sure I've covered all angles.
> So can anyone here please suggest anything I may have missed.
There are a couple of other combinations that I have implemented that may
not be included in your catalog of valences. My complete catalog includes
the following:
1) patient (-a) used in intransitive construction where the argument is
S-function is semantically patientive. (I am sleeping.)
2) agent (-e) used in intransitive construction where the argument is
S-function is semantically agentive. (I ran.)
3) agent/patient (-ae) used in transitive construction where the argument is
A-function is semantically agentive and the argument in P-function is
semantically patientive. (I taught him.)
4) agent/theme (-ie) used in transitive construction where the argument is
A-function is semantically agentive and the argument in P-function is
semantically thematic. (I looked at him.)
5) patient/theme (-ia) used in transitive construction where the argument is
A-function is semantically patientive and the argument in P-function is
semantically thematic. (I saw him.)
6) agent/patient/theme (-iae) used in ditransitive construction where the
argument is A-function is semantically agentive, the argument in P-function
is semantically patientive and the oblique argument is thematic. (I taught
him linguistics)
> (Oh, I can't help feeling that English is a very bad language in which
> to discus this. Now I read back through this message, I begin to
> wonder whether "wonder" used at the top, is acutally NOT transitive,
> after all. Can I "wonder this"? Comments anyone?)
One would "wonder about this."
David
David E. Bell
The Gray Wizard
dbell@graywizard.net
www.graywizard.net
"Wisdom begins in wonder." - Socrates