verb classes - properties, lexical decomposition, nat semantic metalanguage Flashcards
Lexical meaning of verbs:
Inchoative + Causative verbs
Relationship between adjective describing a state and verb describing beginning or change of state, e.g.: 1. Wide (Adj.) 2. The road widened. 3. The council widened the road. 4. Open (Adj) 5. The gates opened. 6. The porters opened the gates. (Saeed, 1997:71-72)
- (1) & (4) describe a state (doesn’t involve change)
- (2) & (5) describe a change of state; they are INCHOATIVE
- (3) & (6) are transitive and their meaning CAUSATIVE
CAUSATIVE-INCHOATIVE ALTERNATION:
Some verbs polysemous between a transitive & intransitive sense
> Timmy broke your favourite cup
> Your favourite cup broke.
Transitive Vs are causative, intransitive Vs are inchoative
» Alternation applies to a number of change of state verbs but not to all of
them
EG DISAPPEAR
>The rabbit disappeared.
>*The magician disappeared the rabbit.
EG ‘cut’ only transitive/causative
> *The snake cut in half
> Common FLA error in children
A number of distinct classes identified in Levin’s (1993) and Rappaport
Hovav & Levin (1998) work, following Fillmore’s (1970) seminal paper
Conative alteration?
- Shared meaning components define class membership
- Class defining components are also grammatically relevant
• Tests show different behaviour of e.g. ‘break’ type Vs and ‘hit’ type Vs
CONATIVE ALTERNATION: (from Latin conor/conari, “to try or attempt”)
> Modifies the interpretation of the V and nature of arguments
»Max kicked the ball/at the ball.
»Max broke the window/*at the window
Verbs: lexical decomposition
Most interesting semantic properties of Verbs are those that define class membership (not idiosyncratic punch vs slap)
LEAD TO: LEXICAL DECOMPOSITION
Use of primitive predicates (Dowty, 1979, Jackendoff 1983, 1991, Rappaport
Hovav & Levin, 1998)
Primitives or primitive terms
theoretical terms that cannot be further
analysed in terms of simpler notions
• Thematic roles are not primitives
• Can be shown by breaking down meaning of verb further
Lexical decomposition example
• Brutus killed Caesar.
Brutus = AGENT & Caesar = THEME
Alternative = break down KILL into “subevents” showing relations between arguments in more detail
[DO*(b)] CAUSE[BECOME[NOT ALIVE)(c]]]
A Doing by Brutus CAUSES it to BECOME the case that Caesar
is NOT ALIVE.
(where ‘not alive’ is a state)
Predicates In CAPS - not same as english words - CAUSE only expresses Direct Causation
ARGUMENTS represented by parts of formula (SCHEMATA)
eg. DO/ACT(b) represents the fact that Brutus is the agent (doer)
• BECOME[STATE(c)] shows that Caesar is the theme of a changeof-state
verb
GENERALISED SCHEMATA in lexical decomposition
More generally, classes defined in terms of generalized schemata
EG:
• Inchoatives: BECOME[STATE(y)] (change-of-state theme)
> The ice melted.
BECOME[MELTED(the ice)]
• Agentives:
DO(x)
> Jane patted Fido.
DO PAT(j, f)
Agentive causatives: [DO(x)] CAUSE[BECOME [STATE(y)] >The hairdresser straightened my hair. [DO (the hairdresser, my hair)] CAUSE [BECOME [STRAIGHT(my hair)]]
Non-agentive causatives: CAUSE(x,[BECOME[STATE(y)]]
> The frost hardened the washing on the line.
CAUSE(the frost,[BECOME[HARD(the washing…)]]
Statives: STATE(x)
> This dish is hot.
HOT(this dish)
ANOTHER APPROACH TO LEXICAL DECOMPOSITION: Wierzbicka’s semantic
primitives
Natural Semantic Metalanguage (NSM) aims to describe all meanings in all
languages; see also work of Cliff Goddard
• Strictly limited set of SEMANTIC PRIMES
• Primes are UNDEFINABLE
• They are UNIVERSAL, i.e. lexicalised in all natural languages
• A number of different languages from different families examined so far
Semantic Primes
I, YOU, SOMEONE (PERSON), SOMETHING (THING), PEOPLE; WORLD
THIS, THE SAME, OTHER, PART (OF), KIND (OF)
ONE, TWO, MANY (MUCH), MORE, VERY, ALL, SOME (OF)
THINK, KNOW, WANT, FEEL, HEAR, SAY
GOOD, BAD; BIG, SMALL
DO, HAPPEN; MOVE, THERE IS, (BE) ALIVE
WHEN, BEFORE, AFTER; A LONG TIME, A SHORT TIME, NOW
WHERE, UNDER, ABOVE; FAR, NEAR; SIDE; INSIDE; HERE
NOT, CAN; IF, BECAUSE, LIKE, IF…WOULD, MAYBE
NSM examples
X feels envy: =
Sometimes a person thinks something like this:
Something good happened to this other person
It did not happen to me
I want things like this to happen to me
Because of this, this person feels something bad
X feels something like this
Sky something very big people can see it people can think like this about this something it is a place It is above all other places It is far from people
NSM positives and negatives
POSITIVES:
• Unlike other approaches to decomposition, not restricted to certain word classes
• Can capture hyponymy (KIND OF) meronymy (PART OF)
• Allows description of meaning in other languages
PROBLEMATIC ASPECTS:
• Primitives very general, and explications are typically imprecise/vague
• Definition of envy would also apply to e.g. self-pity/feeling neglected
• Does not explain meaning relations between primitives (e.g. antonymy with BIG
and SMALL)