How abstract is syntax? Evidence from structural priming.

In 1990, Bock and Loebell found that passives (e.g., The 747 was radioed by the airport's control tower) can be primed by intransitive locatives (e.g., The 747 was landing by the airport's control tower). This finding is often taken as strong evidence that structural priming occurs on the basis of a syntactic phrase structure that abstracts across lexical content, including prepositions, and is uninfluenced by the semantic roles of the arguments. However, all of the intransitive locative primes in Bock and Loebell contained the preposition by (by-locatives), just like the passive targets. Therefore, the locative-to-passive priming may have been due to the adjunct headed by by, rather than being a result of purely abstract syntax. The present experiment investigates this possibility. We find that passives and intransitive by-locatives are equivalent primes, but intransitive locatives with other prepositions (e.g., The 747 has landed near the airport control tower) do not prime passives. We conclude that a shared abstract, content-less tree structure is not sufficient for passive priming to occur. We then review the prior results that have been offered in favor of abstract tree priming, and note the range of evidence can be considerably narrowed-and possibly eliminated-once effects of animacy, semantic event structure, shared morphology, information structure, and rhythm are taken into account.

It takes a village: The role of community size in linguistic regularization.

Studies of artificial language learning provide insight into how learning biases and iterated learning may shape natural languages. Prior work has looked at how learners deal with unpredictable variation and how a language changes across multiple generations of learners. The present study combines these features, exploring how word order variation is preserved or regularized over generations. We investigate how these processes are affected by (1) learning biases, (2) the size of the language community, and (3) the amount of input provided. Our results show that when the input comes from a single speaker, adult learners frequency match, reproducing the variability in the input across three generations. However, when the same amount of input is distributed across multiple speakers, frequency matching breaks down. When regularization occurs, there is a strong bias for SOV word order (relative to OSV and VSO). Finally, when the amount of input provided by multiple speakers is increased, learners are able to frequency match. These results demonstrate that both population size and the amount of input per speaker each play a role in language convergence.

Children and adults successfully comprehend subject-only sentences online.

In many offline studies, children show selectively better comprehension of sentences with the focus particle only when it modifies the object argument (Jane only ate an apple) than they do when it modifies the subject argument (Only Jane ate an apple). Here we explore the nature of this asymmetry by examining performance in a different kind of task: the moment-to-moment comprehension of unambiguous sentences. If past errors reflect a fundamental difference in representation or complexity of computation, we would expect the same asymmetry in this task. We observed that adults were able to successfully predict the target referent for both types of only-sentences, as indicated by anticipatory looks, while 6- to 8-year-old children could do so only for subject-modifying only-sentences. These findings suggest that much of the asymmetry in past work may be due to task demands. We discuss the implications of these results for children's syntactic and pragmatic development.

Event Structures Drive Semantic Structural Priming, Not Thematic Roles: Evidence From Idioms and Light Verbs.

What are the semantic representations that underlie language production? We use structural priming to distinguish between two competing theories. Thematic roles define semantic structure in terms of atomic units that specify event participants and are ordered with respect to each other through a hierarchy of roles. Event structures instead instantiate semantic structure as embedded sub-predicates that impose an order on verbal arguments based on their relative positioning in these embeddings. Across two experiments, we found that priming for datives depended on the degree of overlap in event structures. Specifically, while all dative structures showed priming, due to common syntax, there was a boost for compositional datives priming other compositional datives. Here, the two syntactic forms have distinct event structures. In contrast, there was no boost in priming for dative light verbs, where the two forms map onto a single event representation. On the thematic roles hypothesis, we would have expected a similar degree of priming for the two cases. Thus, our results support event structural approaches to semantic representation and not thematic roles.

How broad are thematic roles? Evidence from structural priming.

Verbs that are similar in meaning tend to occur in the same syntactic structures. For example, give and hand, which denote transfer of possession, both appear in the prepositional-object construction: "The child gave/handed the ball to the dog." We can call the child a "giver" in one case and a "hander" in the other, or we can refer to her more generally as the agent, or doer of the action. Similarly, the dog can be called the recipient, and the ball, the theme. These generalized notions of agent, recipient, and theme are known as thematic roles. An important theoretical question for linguists and psycholinguists is what the set of thematic roles is. Are there a small number of very broad roles, perhaps with each one mapping onto a single canonical syntactic position? Or are there many distinct roles, several mapping to the same syntactic position but conveying subtly different meanings? We investigate this question across eleven structural priming experiments on Amazon Mechanical Turk (total N = 2914), asking whether speakers treat the thematic roles recipient and destination (i.e., location or spatial goal) as interchangeable, suggesting the broad role of goal, or distinct, suggesting two separate roles. To do so, we look for priming between dative sentences (e.g., "The man gave the ball to the dog"), which have a recipient role (dog), and locative sentences (e.g., "The man loaded hay onto the wagon"), which instead have a destination role (wagon). Our pattern of findings confirms that thematic role mappings can be primed independent of syntactic structure, lexical content, and animacy. However, we find that this priming does not extend from destinations to recipients (or vice versa), providing evidence that these two roles are distinct.

Priming is swell, but it's far from simple.

Clearly, structural priming is a valuable tool for probing linguistic representation. But we don't think that the existing results provide strong support for Branigan & Pickering's (B&P's) model, largely because the priming effects are more confusing and diverse than their theory would suggest. Fortunately, there are a number of other experimental tools available, and linguists are increasingly making use of them.

Scalar adjectives and the temporal unfolding of semantic composition: An MEG investigation.

A growing body of research implicates the left anterior temporal lobe (LATL) for combinatorial semantic processing. However, magnetoencephalography (MEG) studies have revealed this activity to be timed quite early, at 200-250ms, preceding the most common time window for lexical-semantic effects. What type of semantic composition could the LATL perform at 200-250ms? We hypothesized that the LATL computes an early stage of composition, taking as its input only the most readily available lexical-semantic information. To test this, we varied the context-sensitivity of prenominal adjectives, postulating that only context-insensitive intersective adjectives (e.g., dead, Italian) should compose in an early time window, whereas the composition of context-sensitive scalar adjectives (e.g., fast, large) should be delayed until the interpretation of the subsequent noun is fully determined. Consistent with this, early combinatory effects in left temporal cortex were observed only for intersective adjectives, though in this study the effects were somewhat more posterior than in prior reports. Overall, our results suggest multiple stages of semantic composition, of which the LATL may index the earliest.