PREDICATESAND ARGUMENTS

Oleh Lartini (2040107114)
In a sentence the verb is often best seen as a relational feature and, indeed, that active and passive sentences could be handled as if they were relational opposites. Analysis in relational terms seems to offer a far more satisfactory solution to the problem of a sentence meaning than componential analysis. In an essence such analysis will have much in common with predicates calculus.

The predicate appears to structure multiple levels of meaning. The predicate with its arguments is a proposition and the proposition may become an argument of a superior predicate. This kind of recursion in the semantic structure makes it important to characterize each predicate by reference to its signature — the number and respective types of its arguments. It should later become clear that syntactically each level of meaning must be represented by a differently ordered structure. The first four examples are sentences where the predicate seems to have the simplest semantic structure.









Proposition



(1)


They waited.


WAIT(a)


"waited"

(2)


They waited long.


LONG(WAIT(a))


"waited long"

(3)


They waited in the next room.


IN(WAIT(a),b)


"waited in next room"

(4)


They were in the next room.


IN(a,b)


"were in next room"



Since we are not concerned here with entailment or any other logical relation between sentences, we do not need formulae that express prepositions, but can use what logicians call ‘open sentences’. Thus we can characterize walk, love and give in terms of one-two and three-place predicates. It will be often be convenient to spell out the predicate in full whit the relevant English word; when this is done it will be placed in square brackets [walk], [love], [give]. A major advantage of this approach is that it can ‘handle’ ‘atomics’ components as well as relational ones. For we may regard such as components as a relation involving just one argument.

Predicate calculus provides a simple method of dealing with what is known in grammar as subordination by allowing preposition to function as an argument. Thus we may to analyze Bayu thinks that Rafy loves Luna. By saying that the predicate [think] has two arguments, Bayu and the preposition Rafy loves Luna. We need to is one of the argument indicate that the whole preposition Rafy loves Luna is one of the arguments of [think]. This illustrate that sentence can be given as [think], [love], where the round brackets show that [love] is a single element. This illustrate that preposition with is own predicate and argument can also be an argument of another ‘higher’ preposition.

In this example semantic interpretation has not been very different from that suggested by the syntax of the sentence. But it is possible to break prepositions down into far more basic elements than those indicated by the actual words of the sentence. For instances, we might think of treating Surya gave Toni a book in terms of s three-places of predicate [give]-[give]. But we could instead, interpret the sentence as ‘Surya caused Tony to have a book’. The formula then becomes [cause],[have], where the arguments of [cause] are (Surya) and [have],(‘Toni have the book’). Similarly, we might treat kill as ‘cause to die’ or ‘cause to become not alive’. The latter is more favored, but it also involves the use of the logical operator ~ ‘not’. The formula ‘John killed Mary would then be [cause](x), [become](y), [alive]. i.e. ‘John cause Mary become Mary not alive’; notice that both [cause] and [become] have a preposition as their second argument.

This kind of analysis is often written out in ‘tree diagrams’ which are used for syntax. The generative semantics argued that a representation of this kind did not merely relate to the semantics of kill, but was rather its deep structure. The argument was largely based upon the triple ambiguity of I almost killed him, where it is argued, almost may qualify cause, become, or not alive (ef. Morgan 1969). The first sense applies if I shoot at him but missed ( I almost caused the subsequence events, bur did not). The second applies if I shoot at him and he recovered after narrowly avoiding death ( he become almost dead). The third applies if I shoot him in he was in a state of near death (he become almost dead). On the basis of this it is argued that kill must be interpreted in terms of three sentences in deep structure , for this will make it possible to place almost in each of these three sentences and thus show whether it qualities cause, become, or not other.



BIBLIOGRAPHY



Lionz j, Semantic 2, press indicated of the university of Cambridge, 1977.

Palmer F.R.,Semantic, edition, Cambridge unity press London, new York, 2001.

Pateda, mansoer, semantic leksikal, rineka cipta, Jakarta, 2001.