An Experimental study on the processing of two-clause sentences

Résumé

Some temporal conjunctions can express a variety of meanings on a causal-temporal dimension. By inference, a when-clause may indicate a cause, an event prior in time, or even an event occurring simultaneously with the main clause event (Bever & Townsend, 1970). Using a self-paced reading experiment in the program Psyscope, we crossed the independent variables connective (when) and position (first position – Subordinate-Main – and second position – Main-Subordinate) to test if readers focus more on the main clause than on the subordinate clause and if readers use the order-of-mention strategy to interpret two-clause sentences conjoined with when establishing a causal-temporal relationship. Our results showed a significant effect of both structure and position factors. The online results showed that, in general, the main clause receives more reading times than the subordinate clause and that readers spent more time reading the main-subordinate sentences. The offline results indicate that the information about the first event is easier to recall if the when-clause is the first clause. Thus, the subordinate-main condition presented more accuracy rates. The order of mention strategy did not seem to be used during sentence reading processing, however, the accuracy rates were higher in conditions, in which the linguistic order mirrored the occurrence order. The main-subordinate order affected the ease of late comprehension, although it eased the reading processing.