In his poem “Dust of Snow,” Robert Frost writes that a snow dusting Has given my heart / A change of mood rather than Has given a change of mood / To my heart. He does so out of attention to rhyme and meter, and probably also to avoid an awkward line like To my heart. But, as I described in last week’s column, even the non-poets among us make similar choices when determining how to phrase a sentence. When asking our neighbor whether she might give us some sugar, as opposed to asking her whether she’ll give some sugar to us, we are all influenced by which parts of our message we want to highlight and what we can reasonably expect our listeners to already know. We also try not to sound too stupid.
But I left out perhaps the most interesting factor of all: we are influenced by the other sentences we’ve recently heard or read.
Now, we certainly do not produce language in a vacuum. Our words are often a response to someone else’s, and if we did not tailor them accordingly, we would appear mad or narcissistic, and have no friends. But we also appear to be swayed by the structure of recently mentioned sentences when forming our own. We’re likelier to produce active constructions in a conversation filled with active constructions, and preposition datives when lots of things have been given or thrown or shown to other things. Even our children are susceptible. Psycholinguists call this phenomenon “syntactic priming.”
Some researchers believe that we store abstract representations of syntactic structures in memory, much the same way that we store representations for words. The act of retrieving the word dog from memory temporarily makes dog easier to access. Similarly, hearing or reading a sentence with a particular syntactic construction (and thus retrieving said construction) temporarily makes the whole construction more accessible. Then, with that construction ready and waiting, so long as we can use it to say what we want to say, we probably will. This is especially the case when our new sentence has a lot of lexical overlap—the same verb, for instance—as the original sentence.
Although syntactic priming most dramatically affects language production, several studies suggest it plays a role in comprehension as well. Let’s say your friend, a teacher, is venting about her day. Whew! she complains. First my car wouldn’t start, and then I noticed I’d left my cell phone at home. When I realized that a student had blown off an assignment, I was so frustrated that I berated the student at the podium. This last sentence can be interpreted in two ways: it could be that your friend did the berating from the podium (linguists call this “high attachment” because the ambiguous modifier at the podium attaches to the earlier of the two possibilities, the verb berated), or it could be that the student was at the podium while being berated (i.e., “low attachment” or connected to the later possibility, the noun student). The interpretation we are likely to go with appears to be influenced by whether we’ve recently encountered a sentence with a “high attachment” modifier, such as The driver ate sandwiches with his friend, or a sentence with a “low attachment” modifier, such as The driver ate sandwiches with peanut butter.
As one might summarize after an evening with Frost’s Collected Poems, should two constructions diverge in a yellow wood, we choose the one last mentioned. That makes a reliable difference.
