Priming vs. Interference in the Use of Optional Infinitival "to"

Fri, 04/28/2017 - 12:30pm
Speaker and Affiliation: 
Tom Wasow, Stanford University

There is a substantial literature on the factors influencing the presence or absence of the word "that" in environments where it is optional.  Another English function word, the infinitival "to", is also optional in certain environments, but there has been little study of the factors conditioning its use.  In a recent corpus study of one such environment (which we dubbed the "do-be construction"), my colleagues and I found (inter alia) a significant effect of priming; that is, if the preceding clause contained an infinitival "to", the probability of an infinitival "to" where it could be omitted was increased.  Another environment in which "to" is optional is in the complement of the verb "help":

Pat helped (Chris) (to) clear the table.

The use of "to" in this environment has not been investigated in experiments or large corpus studies, but it has been addressed anecdotally, particularly in works for English language learners.  There it is claimed:  (i) that there is a subtle meaning difference between the forms with and without "to", and (ii) that the use of "to" is disfavored if "help" itself is infinitival.  This talk will discuss preliminary investigations of both of these claims.  In particular, if (ii) is correct (as it appears to be), it is a case of interference -- the exact opposite of priming.  This raises the question of why a preceding "to" primes the use of "to" in one construction and interferes with it in another.  

Eric Raimy,
Department of Linguistics and Language Institute