The Control Potential of Verbs and Verb Complement Types: An Empirical Study

Comprehending sentences with verb complements very often involves interpretation of an ‘understood subject’, or actualizer of the denoted event. Interpretations of this kind are frequently found obligatorily-restricted in basic infinitive constructions (NP + MATRIX V + infinitive), while sometimes less explicit for basic gerundive constructions (NP + MATRIX V + gerund). This problem, known as ‘control’, is intensively studied by means of syntactic configurations and the thematic roles designated by the matrix predicate. However, less attention has been paid to lexical and usage-based dimensions— distribution of matrix verbs and respective readings invoked by the same lexical items, which are of importance for learner-friendly purposes: when acquiring new verbs, one would naturally expect to find out not only facts about grammatical combinations, but also desirable constructions for carrying out their intended meanings.

This study, therefore, integrates the traditional stance of binding theory with an empirical approach to verbs and the control phenomenon, using examples extracted from the International Corpus of English – the British Component. In addition to the above-mentioned basic constructions, many verbs are found compatible with other complement types, where the understood subject of the second event is overt (e.g. NP + MATRIX V + NP + infinitive, NP + V + ‘that’ clause). Through contrasting these distributions, we have discovered that when a verb is valid exclusively in the two basic types of constructions, the actualizer of the subordinate predicate verb generally cannot be external, i.e. other than the matrix subject itself. Preliminary results have also shown that among all matrix verbs accepted in the basic gerundive construction, above 60% are able to introduce an additional event with an external actualizer in other constructions, while this proportion is only around 40% for those verbs accepted in the basic infinitive construction.