Theme 21. The case grammar

Plan

1. The problem of the grammatical subject and object in the grammatical category of voice.

2. The problem of voice and transitivity of verbs.

3. The case theory by C.I. Fillmore as an attempt to deal with the semantics of the subject and the object:

a) the difference between the grammatical and the semantic subject;

b) C.I. Fillmore’s proposal of cases as ‘a set of universal, presumably innate, concepts’;

c) F.R. Palmer about the explanatory force of the so-called ‘case’ theory.

We discussed the problem of subject and object before. One attempt to deal with the semantics of such categories is to be found in ‘case theory’ associated with C. J. Fillmore.

We begin by considering a trio of sentences such as John opened the door with а key, The key opened the door and The door opened. There is the same verb, open, in all three, and in all three it is active. Yet the grammatical subjects are John, the key and the door respectively. Fillmore suggests that these facts can be accounted for if we handle John, the key and the door in terms of ‘case relations’ that are not directly related to grammatical subject and object, the case of each noun being the same in all three sentences. Thus John is AGENTIVE (=‘actor’) throughout, the key is INSТRUМЕNТАL and the door is OBJECTIVE. We can now state that nouns in any of these three cases, agentive, instrumental or objective may be the grammatical subject. With an active verb, however, which of them is to function as subject is determined by a simple rule of precedence that sets the cases in the order agentive, instrumental, objective. That is to say, if the agentive is present (John), it will always be the subject - John opened the door with а key, but not *The key opened the door by John or *The door opened with а key by John. Similarly if there is an instrumental but no agentive, the instrumental will be the subject - The key opened the door, but not *The door opened with а key. Only if the objective is alone can it be the subject - The door opened.

As a means of relating sentences such as these case grammar works well. We could easily produce a comparable set with ring - John rang the bell with а hammer, The hammer rang the bell, The bell rang. In identifying agentive, instrumental and objective we should be merely developing the notions of ‘actor’ and ‘goal’ (or ‘deep’ subject and object) that were suggested before. What is new is that we are not concerned solely with active and passive sentences, but in addition with trios such as those we have illustrated, and that we need a further category, instrumental (for agentive and objective can be identified with actor and goal). But the categories would still be formal - based only on relations of a transformational kind between sentences.

Fillmore’s proposals go far beyond this. He suggests rather that his case notions are ‘a set of universal, presumably innate, concepts’ and proceeds to define them in semantic terms. To begin with he suggests six cases, Agentive (‘typically animate perceived instigator’), Instrumental (‘inanimate force or object causally involved’), Dative (‘animate being affected’), Factitive (‘object or being resulting from the action or state’), Locative (‘location or spatial orientation’), Objective (‘the semantically most neutral case’).

On the basis of such definitions (I have quoted only the most relevant parts), Fillmore proceeds to suggest what cases are required for other verbs. Thus kill must have a dative and either an agent or an instrumental or both. (The person killed is indicated by the dative, since he is the animate affected, and the killing is obviously done by an animate (Agentive) or inanimate (Instrumental) or both - The man killed him, The rock killed him, The man killed him with a rock). Die involves only a dative. Similarly show needs agentive, dative and objective (one shows something to somebody), while see requires dative and objective. There is a contrast between see and look (at), in that the latter requires agentive and objective (with look the person takes an active part, with see he is merely affected).

Although case theory has some initial plausibility, there are grave difficulties with it. To begin with, if the cases are defined semantically, we shall have all the usual problems of the vagueness of semantic definition. Let us consider first the distinction between agentive and instrumental. Fillmore suggests that the agentive is ‘typically animate’, but we can always invent examples where there are two inanimate objects involved, e. g. The storm broke the glass with the hailstones. We must assume, therefore, that although agents are usually animate, animateness is not an essential part of the definition. But if so, we soon find borderline cases. For, if John may be agentive in John broke the window and the rock instrumental in The rock broke the window, it is not easy to see why the wind should be instrumental rather than agentive in The wind broke the window, The wind rustled the leaves or why John should be agentive if he broke the window accidentally or as a result of someone pushing him. Similarly, there are problems with other cases. Part of Fillmore’s argument, it will be recalled, was that the choice of subject of a sentence was determined by certain ‘formal’ grammatical rules rather than the more ‘notional’ case. He argues that the same is true of the direct objects and illustrates this with comparison of present something to someone and present someone with something. Presumably he wishes to say that, although the direct object is something in the first sentence and someone in the second, this is unrelated to case for something is objective and someone is dative in both examples. Thus this analysis probably results from taking the first example as basic – where objective and dative are identical with object and indirect object, and then proceeding to the argument that the same cases hold for the second example because of its similar meaning and in spite of the difference in its form. But why should we not, instead, argue that the second example is to be treated in terms of objective (someone) and instrumental (something) as the surface form suggests? (Or, perhaps, dative and instrumental, since someone is animate?) For if we want to establish the semantics, we might ask whether present someone with something is like reward sоmeопe with something or punish someone with something. And are these not close to beat someone with something? We are faced once again with a problem that is typical of semantics – where to draw a line between what at first seemed to be quite distinct categories.

There are also many distinctions that case grammar cannot handle. Again we may take Fillmore’s own examples: John smeared paint on the wall / John smeared the wall with paint. Bees are swarming in the garden / The garden is swarming with bees. In each pair the cases would have to be the same for John, paint, wall, bees, and garden, but the meanings of the sentences are not the same. Similarly, the verbs buy and sell seem to involve the same cases – and indeed part of the attraction of case grammar is that they do – yet John bought a book from Mary is not the same as Mary sold John a book. These examples suggest quite strongly that there is some function of the grammatical subject that makes these paired sentences different. Fillmore admits this, but argues that we can only investigate the function of the subject, if we are clear about the case-functions. But this is an odd conclusion. Was not the argument in favour of case the fact that grammatical subject has no consistent function?

References:

1. Иванова, И.П. Теоретическая грамматика современного английского языка: учебник / И.П. Иванова, В.В. Бурлакова, Г.Г. Почепцов. – М.: Высш. шк., 1981. – С. 241-252.

2. Пальмер, Ф.Р. Семантика. Очерк = Palmer, F.R. Semantics. A New Outline / F.R. Palmer; предисл. и коммент. М.В. Никитина. – М.: Высш. шк., 1982. – С. 82-87.

3. Филмор, Ч. Дело о падеже / Ч. Филмор // Новое в зарубежной лингвистике. – Вып. 10. – М.: Прогресс, 1981. – С. 369-495.

4. Филмор, Ч. Дело о падеже открывается вновь / Ч. Филмор // Новое в зарубежной лингвистике. – Вып. 10. – М.: Прогресс, 1981. – С. 495-530.