Table 1.
Examples of ditransitive sentences used in the present study.
Figure 1.
Structures of ditransitive sentences, together with serial presentation of each sentence.
(A) A succinct version of linguistic tree structures representing the syntactic structures of ditransitive sentences. P+ and P– sentences are in columns, while canonical (C, shown in red) and noncanonical (N, shown in blue) word orders, i.e., the canonicity of sentences, are in rows. Dat, dative case marker; Acc, accusative case marker; pro, pronoun, which is a phonetically null subject. For the syntactic structures of noncanonical sentences (lower row), a noun phrase (NP) closest to a verb (V) is moved to the front of another NP (dashed arrow), and merged with the higher V-bar (V’) to form a verb phrase (VP). The moved NP then leaves a trace in its original or canonical position, producing a gap with a longer structural distance between the second NP and V. In our paradigm, each pair of P+ and P– sentences had the same accusative NP (boxed) and phonologically same verb (circled) (see Table 1). We examined the predictive effects of precedent NPs on the verb, which were expected to be larger for the canonical sentences with shorter structural distances (curved arrows) than the noncanonical sentences. Among the four conditions, an animate NP (with a dagger) appeared only as the dative NP of the P+ sentences. (B) A single trial with a ditransitive sentence. A grey square was presented to inform the participant that the trial had begun. Next, a sentence, consisting of two NPs and a verb, was presented in a serial, phrase-by-phrase manner. A grey triangle was shown after a verb to inform participants to initiate a button press. Interstimulus intervals were randomly varied so that the responses to verbs were not confounded with those to precedent NPs. We mainly analyzed the cortical responses to ditransitive verbs.
Table 2.
Examples of ditransitive sentences under the four conditions.
Table 3.
Examples of grammatical modified sentences with either monotransitive or intransitive verbs.
Figure 2.
Significant activation with canonicity effects on ditransitive verbs.
(A) Cortical activation showing a significant main effect of canonicity at 530–550 ms. A significant C > N effect (corrected p<0.05) was observed at a single cluster in the left (L.) IFG (shown in yellow to black), which was superimposed on a sagittal section of the standard brain at the peak [Talairach coordinates, (x, y, z) = (–48, 10, 18)]. (B) The current density in the left IFG cluster for each of the four conditions (mean ± SEM). An asterisk denotes the significant difference (p<0.05, paired t-test) between the two conditions, under which the same NP preceded a verb (see Table 2).
Table 4.
Behavioral data for ditransitive sentences under each condition.
Figure 3.
Significant activation with canonicity effects on monotransitive and intransitive verbs.
We examined any canonicity effects for grammatical modified sentences with monotransitive or intransitive verbs. Each activation cluster was shown for a representative (i.e., with more activation) time bin of 20 ms, superimposed on a sagittal section of the standard brain at the peak. Paired t-tests resulted in a significant N > C effect (corrected p<0.05) in the following activated regions. The current density for canonical and noncanonical conditions is also shown for each cluster (mean ± SEM). (A) The left SMG activation [peak: (–50, –24, 7)] at 480–500 ms. (B) The left SMG activation [peak: (–57, –27, 11)] at 570–590 ms. (C) The left pSTG activation [peak: (–48, –45, 13)] at 610–630 ms. (D) The right (R.) aMTG/ITG activation [peak: (54, –3, –18)] at 650–670 ms.