Bill hit John).Grammatical studies, both traditional and contemporary, have recognized that formal distinctions, involving the case or syntactic position of the arguments of a verb or other part of speech, correlated significantly with intuitive semantic distinctions, involving the relations of those arguments to the action or state indicated by the verb. In simple English sentences with two or three nominal arguments, for example, where only syntactic position visibly distinguishes their relation to the verb, we can see that the first, subject position regularly belongs to the initiator of the action, or agent; that the direct object position regularly belongs to the recipient of the action, the patient or more generally the theme; and that the indirect-object position, if there is one, regularly belongs to the person or thing for the sake of which the action is done, or the goal of the action:
Bill hit John).the memory chip controls the circuit). the boy gave the girl the dog, the girl
gave the dog the boy, etc.).In languages with rich case systems, the cases will give information about the semantic relations of agent, patient, goal, and some others, including location and instrument. English, typically among languages with impoverished case systems, signals these relations through prepositions. The cases themselves also correlate with syntactic positions (e.g., with nominative as the case of the subject and accusative as the case of the direct object), so that the formal side comprises both syntactic and morphological elements; the relations between these elements are a subject of intense crosslinguistic study within linguistic theory (see GRAMMATICAL RELATIONS).
The semantic relations between a verbal form and its arguments, as in (1) - (3), were dubbed thematic roles Gruber (1965); the term theta-roles is also in common use, with the same meaning. The notion of a thematic role extends also to the adjuncts of the verbal form, elements that, like the goal, instrumental, and locative phrases in (4), are not obligatory but optional (as indicated by parentheses).
Following standard terminology, say that a verb selects an array of arguments and permits an array of adjuncts, each of which bears a given thematic role. The thematic roles of arguments must be syntactically realized: hence we do not have sentences like *John hit, meaning that John hit someone. Conversely, and with greater generality, no thematic role can be distributed over more than one element, whether it be argument or adjunct; thus, for example, instrumental phrases cannot be repeated, as in (5).
There is, then, for each simple clause a one-to-one correspondence between a certain set of theta-roles and their realization in certain syntactic positions, perhaps in construction with specific cases and/or prepositions (or postpositions). That human languages obey this correspondence principle is the theta-criterion in the sense of Chomsky 1981.
Assuming the theta-criterion and a preliminary grasp of the content of typical thematic roles, theoretical studies have endeavored to discover a full inventory of the thematic roles in principle admissible in human languages, and the syntactic or morphological conditions on the appearance of elements with those roles. Languages generally exhibit a hierarchy among thematic roles, realizing agents as subjects and themes as direct objects; for important surveys, see Rappaport and Levin 1988 and Jackendoff 1987. These studies have also shed light on the scope and limits of certain syntactic alternations. For instance, as observed Fillmore (1968), English and other languages have pairs such as (6) - (7).
That such pairs should exist would follow if the verb swarm were specified for locative (in the garden) and instrumental (with bees) thematic roles, with either thematic element free to occupy the subject position. On the other hand, there are alternations that are not fully productive, as in (10) - (11) versus (8) - (9):
The syntactic distribution shown may be predicted from the meaning of the verbs, in that stuff, but not fill, can take the direct object as its theme.
The subject position is crucially distinguished from others, in that it alone can be occupied by an expletive, as in (12) - (13):
Furthermore, there are verbs whose sole argument may be argued to originate, not in the subject position but in the position of the direct object, the so-called unaccusative verbs. For these reasons, among others, it has been useful to distinguish the elements in close construction with the verb -- the internal arguments -- from the single element that will appear as the subject in a simple clause -- the external argument (the terminology and much of the discussion is due to Williams 1980).
Besides the above and other syntactic studies, semantic research has considered the question how thematic roles contribute to meaning. On one view of the matter, discussed by Carlson (1984), Higginbotham (1989), and Parsons (1990), among others, thematic roles are relations between events or states and the things that participate in them. The semantics of (1), for instance, would then be expressed as (14).
Alternatively, it has been suggested that the content of thematic roles is to be located in semantic postulates governing the verbs that assign them (see Dowty 1988).
Carlson, G. (1984). Thematic roles and their role in semantic interpretation. Linguistics 22:259-279.
Chomsky, N. (1981). Lectures on Government and Binding. Dordrecht: Foris.
Dowty, D. (1988). On the semantic content of the notion 'thematic role.' In G. Chierchia, B. Partee, and R. Turner, Eds., Properties, Types and Meaning, vol. 2, Semantic Issues. Dordrecht: Kluwer, pp. 69-130.
Fillmore, C. (1968). The case for Case. In E. Bach and R. T. Harms, Eds., Universals in Linguistic Theory. New York: Holt, Rinehart and Winston, pp.1-90.
Gruber, J. S. (1965). Studies in Lexical Relations. Ph.D. diss., MIT. Reprinted in Lexical Structure in Syntax and Semantics (1976). Amsterdam: North-Holland.
Higginbotham, J. (1989). Elucidations of meaning. Linguistics and Philosophy 12:123-142. Reprinted in P. Ludlow, Ed., Readings in the Philosophy of Language. Cambridge, MA: MIT Press, pp. 157 - 178.
Jackendoff, R. (1987). The status of thematic relations in linguistic theory. Linguistic Inquiry 18:369-411.
Rappaport, M., and B. Levin. (1988). What to do with theta-roles. In W. Wilkins, Ed., Syntax and Semantics 21: Thematic Relations. New York: Academic Press, pp. 7-36.
Parsons, T. (1990). Events in the Semantics of English. Cambridge, MA: MIT Press.
Williams, E. (1980). Predication. Linguistic Inquiry 11:203-238.
Grimshaw, J. (1990). Argument Structure. Cambridge, MA: MIT Press.
Jackendoff, R. (1990). Semantic Structures. Cambridge, MA: MIT Press.
Levin, B. (1993). English Verb Classes and their Alternations. Chicago: University of Chicago Press.
Levin, B., and M. Rappaport Hovav. (1995). Unaccusativity: At the Syntax-Lexical Semantics Interface. Cambridge, MA: MIT Press.
Wilkins, W., Ed., Syntax and Semantics 21: Thematic Relations. New York: Academic Press.
Copyright © 1999 Massachusetts Institute of Technology