Identifying the Null Subject: Evidence from Event-Related Brain Potentials

Event-related brain potentials (ERPs) were recorded during spoken language comprehension to study the on-line effects of gender agreement violations in controlled infinitival complements. Spanish sentences were constructed in which the complement clause contained a predicate adjective marked for syntactic gender. By manipulating the gender of the antecedent (i.e., the controller) of the implicit subject while holding constant the gender of the adjective, pairs of grammatical and ungrammatical sentences were created. The detection of such a gender agreement violation would indicate that the parser had established the coreference relation between the null subject and its antecedent. The results showed a complex biphasic ERP (i.e., an early negativity with prominence at anterior and central sites, followed by a centroparietal positivity) in the violating condition as compared to the non-violating conditions. The brain reacts to NP-adjective gender agreement violations within a few hundred milliseconds of their occurrence. The data imply that the parser has properly coindexed the null subject of an infinitive clause with its antecedent.     

Processing controlled PROs in Spanish

Two experiments were carried out to investigate the processing of the empty category PRO and the time-course of this in Spanish. Eye movements were recorded while participants read sentences in which a matrix clause was followed by a subordinate infinitival clause, so that the subject or the object of the main clause could act as controller of PRO, and therefore as implicit grammatical subject of the infinitive. In Experiment 1, verb control information was manipulated: The matrix clause contained either subject-control verbs like prometer ('promise') or object-control verbs like forzar ('force'). In Experiment 2, the preposition that headed adverbial subordinate clauses was manipulated: Two different kinds of infinitival adverbial clauses were used, expressing purpose (preposition para) and reason (preposition por) and in which control information is primarily induced by the prepositions (para tends to trigger subject-control; por object-control). Experiment 1 showed that readers make immediate use of verb control information to recover the antecedent of the empty category PRO in Spanish obligatory control constructions. The data obtained in Experiment 2 suggest that during the processing of the empty category PRO in purpose vs. reason adverbial subordinate infinitival clauses the control information induced by the prepositions por vs. para is not initially used as a constraint to guide the selection of the nominal antecedent of PRO. In addition, both experiments showed that PRO antecedent selection is a very fast process and that, together with verb control information, recency played an important role. The results are discussed in terms of current psycholinguistic hypotheses about empty categories and in relation to formal linguistic hypotheses about PRO control. We argue that obligatory control ties (which are co-referential ties) are processed fast because they are launched from the same lexical platform that launches all fast syntactic connections (complements, as opposed to modifiers). In short, that such lexically-specified co-referential ties are regulated by linguistic form.     

The Repeated Name Penalty,the Overt Pronoun Penalty,and Topic in Japanese

When reading sentences with an anaphoric reference to a subject antecedent, repeated-name anaphors result in slower reading times relative to pronouns (the Repeated Name Penalty: RNP), and overt pronouns are read slower than null pronouns (the Overt Pronoun Penalty: OPP). Because in most languages previously tested, the grammatical subject is typically also the discourse topic it remains unclear whether these effects reflect anaphors’ subject-hood or their topic-hood. To address this question we conducted a self-paced reading experiment in Japanese, a language which morphologically marks both subjects and topics overtly. Our results show that both repeated-name topic-subject anaphors and repeated-name non-topic-subject anaphors exhibit the RNP and that both overt-pronoun topic-subject and overt-pronoun non-topic-subject anaphors show the OPP. However, a detailed examination of performance revealed an interaction between the anaphor topic marking, reference form, and the antecedent’s grammatical status, indicating that the effect of the antecedent’s grammatical status is strongest for null pronoun and repeated name subject anaphors and that the overt form most similar to null pronouns is the repeated name topic anaphor. We discuss the implications of these findings for theories of anaphor processing.     

Event-related responses to pronoun and proper name anaphors in parallel and nonparallel discourse structures.

The present study investigated event-related potential (ERP) effects of pronoun and proper name anaphors in both parallel and nonparallel discourse structures. Thirty-seven students processed 400 semantically different text passages. Each trial consisted of two sentences and a comprehension question. The first sentence introduced a protagonist who was referred to by an anaphoric word in the second sentence. The anaphoric word was either a pronoun or a repetition of the proper name of the protagonist and had either the same or a different syntactic role as its antecedent (subject or object). The sentences were presented word by word as rapid serial visual display. Event-related potentials were recorded from 61 scalp electrodes. In agreement with the parallel function strategy, nonparallel discourse structures required longer decision times and exhibited higher error rates than parallel structures. The ERPs revealed two effects: First, pronoun anaphors evoked a more pronounced negativity than proper name anaphors between 270 and 420 ms latency over the frontal cortex electrodes. Another relative negativity occurred between 510 and 600 ms over the parietal cortex electrodes. Second, anaphors in nonparallel positions were accompanied by a more pronounced negativity over the parietal cortex. These data support the idea that an anaphor in nonparallel position triggers extra processing steps, probably search processes in working memory which integrate currently encountered information with previously activated representations.     

The effect of context on the processing of type-shifting verbs

Recent research in lexical semantics has suggested that verbs such as begin and enjoy semantically select for a complement that denotes an activity or an event. When no such activity or event is specified in the form of a progressive or infinitival complement, as in John began (to read/reading) the book, the verb is said to "coerce" the NP direct object to shift its role to encompass the activity that begin requires as complement (e.g., writing, reading). Empirical support for this view has been provided by McElree, Traxler, Pickering, Seely, & Jackendoff (2001). In the present study, however, in two self-paced reading experiments, type-shifting effects (taken to be longer reading times engendered by the computation of the coercion process) were not obtained with sentences in isolation (Experiment 1) or with sentences embedded in contexts that specified the nature of the activity performed over the complement NP (Experiment 2). It is argued that type-shifting verbs are similar to non-preferred verbs for given contexts and that type-shifting operations are pragmatic inferences computed over underspecified semantic representations.     

Processing temporary syntactic ambiguity: The effect of contextual bias

This paper reports two experiments using sentences with a temporary ambiguity between a direct object and a sentence complement analysis that is resolved toward the normally preferred direct object analysis. Postverbal noun phrases in these sentences could be ambiguously attached as either a direct object or the subject of a sentence complement, whereas in unambiguous versions of the sentences the subcategorization of the verb forced the direct object interpretation. Participants read these sentences in relatively long paragraph contexts, where the context supported the direct object analysis (“preferred”), supported the sentence complement analysis (“unpreferred”), or provided conflicting evidence about both analyses (“conflicting”). Self-paced reading times for ambiguous postverbal noun phrases were almost equivalent to the reading times of their unambiguous counterparts, even in unpreferred and conflicted context conditions. However, time to read a following region, which forced the direct object interpretation, was affected by the interaction of verb subcategorization ambiguity and contextual support. The full pattern of results do not fit well with either an unelaborated single-analysis (“garden path”) model or a competitive constraint-satisfaction model, but are consistent with a race model in which multiple factors affect the speed of constructing a single initial analysis.     

Processing temporary syntactic ambiguity: the effect of contextual bias

This paper reports two experiments using sentences with a temporary ambiguity between a direct object and a sentence complement analysis that is resolved toward the normally preferred direct object analysis. Postverbal noun phrases in these sentences could be ambiguously attached as either a direct object or the subject of a sentence complement, whereas in unambiguous versions of the sentences the subcategorization of the verb forced the direct object interpretation. Participants read these sentences in relatively long paragraph contexts, where the context supported the direct object analysis ("preferred"), supported the sentence complement analysis ("unpreferred"), or provided conflicting evidence about both analyses ("conflicting"). Self-paced reading times for ambiguous postverbal noun phrases were almost equivalent to the reading times of their unambiguous counterparts, even in unpreferred and conflicted context conditions. However, time to read a following region, which forced the direct object interpretation, was affected by the interaction of verb subcategorization ambiguity and contextual support. The full pattern of results do not fit well with either an unelaborated single-analysis ("garden path") model or a competitive constraint-satisfaction model, but are consistent with a race model in which multiple factors affect the speed of constructing a single initial analysis.     

Searching for the Trace: The Influence of Age, Lexical Activation and Working Memory on Sentence Processing

To investigate the stability of trace reactivation in healthy older adults, 22 older volunteers with no significant neurological history participated in a cross-modal priming task. Whilst both object relative center embedded (ORC) and object relative right branching (ORR) sentences were employed, working memory load was reduced by limiting the number of words separating the antecedent from the gap for both sentence types. Analysis of the results did not reveal any significant trace reactivation for the ORC or ORR sentences. The results did reveal, however, a positive correlation between age and semantic priming at the pre-gap position and a negative correlation between age and semantic priming at the gap position for ORC sentences. In contrast, there was no correlation between age and priming effects for the ORR sentences. These results indicated that trace reactivation may be sensitive to a variety of age related factors, including lexical activation and working memory. The implications of these results for sentence processing in the older population are discussed.     

Competence and processing in children's grammar of relative clauses

We report two toy-manipulation experiments investigating 4–5-year-old children's interpretations of relative clauses. In the first experiment we show that the frequency with which relatives modifying the matrix direct object are interpreted as referring to the matrix subject is sensitive to the nature of the material in the relative. Animacy of the relative direct object leads to an increase in subject coreference errors for relatives; a similar effect of animacy was not found for infinitival complements. We propose that such errors arise when the child's sentence processor is taxed. In our second experiment we compare children's interpretations of relatives to their interpretations of temporal participial complements. Children eschewed coreference between the complement and the object of a passive prepositional phrase in the case of temporal participials but not in the case of relatives. We interpret this result as evidence that children are capable of analyzing relatives as constituents of the NP node, and hence that this node is recursive in their grammars. We discuss the results of both experiments in the context of a two-stage parsing model, arguing that our data fit a picture of children's linguistic abilities in which the 4-year-old has a sophisticated competence grammar and a parsing mechanism of essentially the same structure as the adult's. We compare our approach to a number of other approaches to children's relative clause interpretations in the literature, arguing that these alternatives are too structure-specific and can lead to underestimations of the child's competence grammar. We conclude that our analysis is compatible with a picture of acquisition in which the child's competence grammar of relatives is not qualitatively different from the adult's.     

Ambiguous pronoun resolution: contrasting the first-mention and subject-preference accounts

A visual-world eye-tracking experiment investigated the influence of order of mention and grammatical role on resolution of ambiguous pronouns in Finnish. According to the first-mention account, general cognitive structure-building processes make the first-mentioned noun phrase the preferred antecedent of an ambiguous pronoun. According to the subject-preference account, the preferred antecedent is the grammatical subject of the preceding clause or sentence. Participants listened to sentences in either subject-verb-object or object-verb-subject order; each was followed by a sentence containing an ambiguous pronoun that referred to either the subject or the object. Participants' eye movements were monitored while they looked at pictures representing the two possible antecedents of each pronoun. Analyses of the fixations on the pictures showed that listeners used both order-of-mention and grammatical-role information to resolve ambiguous pronouns.     

Exploring the Repeated Name Penalty and the Overt Pronoun Penalty in Spanish

Anaphoric expressions such as repeated names, overt pronouns, and null pronouns serve a major role in the creation and maintenance of discourse coherence. The felicitous use of an anaphoric expression is highly dependent on the discourse salience of the entity introduced by the antecedent. Gordon et al. (Cogn Sci 17:311–347, 1993) showed that, in English, sentences containing repeated names were read more slowly than corresponding sentences containing pronouns when the antecedent of the anaphoric expression was the subject of the previous sentence. This effect was dubbed the Repeated Name Penalty (RNP), and it was further found that this processing delay is eliminated if the antecedent of the anaphoric expression is the object of the previous sentence. The RNP was later extended to Mandarin Chinese (Yang et al. in Lang Cogn Process 14:715–743, 1999) and to Spanish (Gelormini-Lezama and Almor in Lang Cogn Process 26(3):437–454, 2011), which suggests that this might be a universal phenomenon. Moreover, the Spanish results showed an additional effect: sentences containing overt pronouns were read more slowly than corresponding sentences containing null pronouns when the antecedent of the anaphoric expression was the subject of the previous sentence. This effect was dubbed the Overt Pronoun Penalty (OPP) and, like the original RNP, the effect is also eliminated if the antecedent is in object position (Gelormini-Lezama and Almor 2011; Gordon et al. 1993). The similarity of the RNP and the OPP in Spanish suggests that these two processing phenomena might be caused by the same underlying principles. This paper is a critical review of the literature on these processing delays in Spanish and an attempt to integrate the data in a unified framework. Specifically, and following pragmatic explanations like the Informational Load Hypothesis (Almor in Psychol Rev 106:748–765, 1999), the RNP and OPP in Spanish can be understood as superficial manifestations of an imbalance between processing cost and discourse function.     

Lexical structure in parsing long-distance dependencies

We review a series of experiments investigating lexical influences in parsing sentences with long-distance dependencies. We report three primary results. First, gaps are posited and filled immediately following verbs that are typically used transitively, even when the filler is an implausible object of the verb. However, gaps are not posited after verbs that are typically used intransitively. Second, plausibility determines whether or not a filler is treated as the object of a verb when the verb is typically used with both a direct object and an infinitive complement. Finally, verb control information is used immediately in determining which noun phrase will be interpreted as the understood subject of an infinitive complement.     

Singular and Plural Pronominal Reference in Spanish

In two self-paced, sentence-by-sentence reading experiments, we examined the difference in the processing of Spanish discourses containing overt and null pronouns. In both experiments, antecedents appeared in a single phrase (John met Mary) or in a conjoined phrase (John and Mary met). In Experiment 1, we compared reading times of sentences containing singular overt and null pronouns referring to the first or to the second mentioned antecedent. Overt pronouns caused a processing delay relative to null pronouns when they referred to the first antecedent in single but not in conjoined phrases. In Experiment 2, we compared reading times of sentences containing overt and null pronouns referring to singular or plural entities. Plural null pronouns were read faster than their singular counterparts in conjoined conditions. Plural overt pronouns were read more slowly than their null counterparts both in single and conjoined conditions. We explain our findings in a framework based on the notion of balance between processing cost and discourse function in line with the Informational Load Hypothesis.     

Click monitoring revisited: An on-line study of sentence comprehension

Spoken sentence comprehension is based upon rapid and complex psychological processes, yielding a constantly fluctuating cognitive load. The aim of this study was to evaluate on-line click monitoring, a classical but poorly exploited experimental method, which should allow for an easy measurement of processing load at any chosen point of experimental sentences. In Experiments 1 and 2, we obtained longer latencies to clicks located at the boundary of reversible object relative clauses than to clicks identically located in subject relatives and to clicks located earlier within object relatives. Experiment 3 further revealed that this effect of syntactic type was specific to transposed object relatives and did not occur with normal object relatives. In Experiment 4, we observed longer latencies with semantically reversible than with irreversible sentences, but no difference between actives and passives. These results were obtained under strict control of potential lexical and phonological biases, and suggest that on-line click monitoring may be one useful tool in the study of sentence comprehension.     

Differences Between the Morphology of Gender and Number: Evidence from Establishing Coreferences

This work investigates the processing of gender and number features in the selection of a pronoun antecedent in Italian. In Italian there are instances of nouns in which gender and number are treated morphologically on a par, i.e., overtly and regularly marked by a suffix. In these cases, are the two features also treated similarly in processing? The experiments used sentences with two possible antecedents (differing in gender or number) in the main clause and a pronoun in the complement clause. The sentences were visually presented, with a unimodal lexical decision task at the end of the sentence. The results showed a selective reactivation of the antecedent matching the pronoun in either gender or number. The results are discussed in relation to previous Italian experiments, which found that in sentence internal position there is selective reactivation of only the number-matching antecedent. They are taken to support a model of the coreference processor in which gender and number features are used at different processing stages, due to their different syntactic representation.     

Children’s Knowledge of Ellipsis Constructions in Mandarin Chinese

This study investigated 4-year-old children’s knowledge of the structural constraints on ellipsis constructions in Mandarin Chinese, focusing on the structural difference between verb phrase (VP) ellipsis and null object constructions. In Mandarin Chinese, if the antecedent clause of a VP-ellipsis structure contains an adverbial modifier, this adverbial modifier must be recovered at the elided site in the second clause. However, if the antecedent clause of a null object construction contains an adverbial modifier, the reconstruction of this adverbial modifier at the elided site is not necessary. Two experiments were conducted to examine whether young Mandarin-speaking children are sensitive to this structural difference between the two ellipsis constructions. The results show that children exhibited adult-like performance in both experiments. This is evidence that young Mandarin-speaking children have knowledge of the structural difference between VP-ellipsis and null object constructions with regard to adverbial recovery. We interpreted the findings in conjunction with previous research as evidence that children’s knowledge of the structural constraints on ellipsis constructions is innately specified.     

Effects of Local and Global Context on Processing Sentences with Subject and Object Relative Clauses

An eye tracking study investigated the effects of local and global discourse context on the processing of subject and object relative clauses, whereby the contexts favored either a subject relative clause interpretation or an object relative clause interpretation. The fixation data replicated previous studies showing that object relative clause sentences were more difficult to process than subject relative sentences. Crucially, however, the reading difficulty asymmetry between subject and object relative clause sentences disappeared when the sentences were presented with a local or a global discourse context that favored the objects in the object relative clauses. These findings demonstrate that the evidence for a syntax-based account of sentence processing is found when sentences are presented in isolation. However, if sentences are placed more naturally, in context, discourse factors outweigh the initial structural assignment.     

Case marking in the Parsing of Complement Sentences: Evidence from Eye Movements

An eye-tracking experiment investigated the role of case-marking in parsing. We manipulated the case of pronouns in reduced complement sentences like I recognized you and your family would be unhappy here and I recognized she and her family would be unhappy here , in which the nominative pronoun she immediately disambiguates the sentences, in contrast to the ambiguous you . The nominative pronoun she disambiguates the sentence because I recognised she is ungrammatical, and thus she and her family must be the subject of an embedded sentence and not the NP-object of the preceding verb. Subjects took longer to read she and her family than you and your family during initial processing. The pattern reversed at the disambiguating phrase would be . Unambiguous control sentences containing the complementizer that did not produce case-marking effects. These results demonstrate very rapid effects of case-marking on parsing. Either case information is used immediately, or it is employed after an extremely short delay. We discuss implications for current theories of parsing.     

Processing Italian Relative Clauses: Working Memory Span and Word Order Effects on RTs

In this study, the authors aim to clarify whether the subject–object asymmetry in relative clause comprehension is due to the use of parsing strategies (Active Filler Theory) or to a greater memory load generated by object sentences. Two experiments investigate how individual differences in working memory span may influence the reading times of relative sentences in Italian, a language characterized by a flexible structure. The results of Experiment 1 indicate that object extraction is more complex than subject extraction when sentences have a canonical structure. Furthermore, low-span participants have particular difficulties with object relative sentence comprehension. The results of Experiment 2 show that subject-relative clauses with uncanonical structures are more complex to understand than object-relative clauses, and low-span participants have more difficulties than high-span participants in elaborating both subject and object relative clauses. These data seem to be coherent with the Active Filler Theory.     

Temporal course of activation of the antecedent by the reflexive in syntactically ambiguous sentences in Japanese

Three experiments examined the temporal course of the activation of an antecedent triggered by a reflexive,jibun, in syntactically ambiguous sentences in Japanese. Twenty-two students in Experiment 1 were tested for sentences where a subject word occupied the first position and an indirect object word the second, with the probe given for each word either immediately after the reflexive or at the end of the sentence. Reaction times (RTs) were faster when the probe was for subject (S-probe) than when it was for indirect object (IO-probe) irrespective of the probe position. Experiment 2 tested 22 students to determine the effect of order of mention of participant on the probe recognition time. RTs were faster for the S-probe than for the IO-probe, although the difference between the two probes was smaller when the probe was given immediately after the reflexive. Experiment 3 tested 22 students to examine the effect of context on an activation pattern. RTs were lower for the S-probe than for the IO-probe, though no difference was found between them for the probe given at the end of the sentence. Findings were interpreted as supporting the single-interpretation model for on-line analysis of syntactically ambiguous sentences and also the immediate-activation model in an assignment of the reflexive to its antecedent.