Basic classes - ARGUMENTS Flashcards
what is S?
S = Subject-like
Only core argument of a monovalent predicate
Often coincides with a syntactic Subject.
> But not always.
> NOT a syntactic Subject.
What is A?
A = Agent-like
The most agent-like argument of a divalent predicate.
Most frequently express agentive participants.
> When there is one.
> But not always.
What is O?
O = Patient-like
(also called P)
The most patient-like argument
of a divalent predicate.
Most frequently express patientive participants.
> When there is one.
# Patient semantic role.
# Syntactic Object.
what are Universal syntactic-semantic primitives
S = only argument of monovalent predicate
A = most agent-like argument of divalent predicate
O = most patient-like argument of divalent predicate
> Sometimes called P
Misleading labels S is NOT a syntactic Subject. A is NOT a semantic Agent. O is NOT a syntactic Object. (Nor a semantic Patient (P).)
Semantic-syntactic primitives are ‘nothing’.
Which makes them useful in typology!
Universal primitives are NOT core syntactic relations
Subject and Object are ‘grammatical relations’.
> On the basis of syntactic properties (‘deep’).
E.g. Subjects control gapping.
- The man saw the child and Ø laughed.
NOT formal marking (morphology, word order etc.).
S are often Subjects but not always.
Universal primitives NOT Semantic roles
A and P allude to semantic roles.
> Agent is a common label for ‘who does the action’.
> As are often Agents, but many As are not.
Paul (A-Patient) predeceased John (O-patient) by two years.
What is alignment?
Two of these arguments (SAO) align with eachother
> contrasting with the third one
> diff langs do it in different ways
asking: Which of the primitives align?
In ENGLISH S is encoded like A (nominitive-accusative)
Word order
- S and A both precede the verb
> children sleep a lot/children like lollies
> ?lollies like children
Pronouns
- S and A encoded by the same set
> he arrived / he saw him
> *him saw he
not all langs align S with A. FIVE logical possibilities/ FOUR actual possibilities
- [S] [A, O] Accusative focus NOT PLAUSIBLE
- [A] [S] [O] Tripartite
- [A, S, O] Neutral
- [A, S] [O] Nominative-Accusative
- [A] [S, O] Ergative-Absolutive
tripartite alignment
[A] [S] [O]
S, A and O all encoded differently. [A] [S] [O] > Exceptional. > Attested in Burmese (Sino-Tibetan, Myanmar) > Uneconomical?
neutral alignment
[A, S, O]
S, A and O all encoded in the same way
> Rare but attested.
- EG Thai (Kra-Dai, Thailand) has no formal marking of arguments.
- relies entirely on context.
> Problem for effective communication?
Alignment of two of the arguments is a lot more common
accusative alignment
[A, S] [O]
S is encoded like A
> S + A = nominative case
> O alone = accusative case
Nom/acc alignment - like ENGLISH “accusative languages”
> DOMINANT ALIGNMENT
- C. 75% of world langs
- all langs have some form of accusative alignment
Ergative alignment
[A] [S,O]
S is encoded like O
> S + O = absolutive case
> A alone = ergative case.
erg/abs alignment
> SIGNIFICANT MINORITY
- 25% of the world languages
Australia, South America
E.g. Dyirbal (PN, Dyirbalic, Australia)
The historical example (Dixon 1972).
South-East Asia
Papua New-Guinea
rarely fully ergative - lots of LIMITATIONS
DEEP ERGATIVITY
How arguments behave syntactically in clause chaining.
> I.e. proper ergative syntactic subjects.
> The notion of syntactic ‘pivot’
In English the syntactic pivot is S/A. The man saw the child and Ø laughed. > Gapped argument A combines with S. > If A is omitted , it is the same as S in previous clause. >It’s a nominative S/A pivot
Deeply ergative languages have an S/O pivot.
> As if we could say in English
The man saw the child and Ø laughed
> When it is the child who laughed.
RARE
handful of langs in australia
some cases disputed
>Lezgian (north caucasian)
Split ergativity
Some elements in the language align ergatively.
> others accusatively
(Warlpiri, PN, Ngumpin-Yapa, Aust.)
- Independent pronouns and nouns align ergatively.
- Person suffixes align accusatively