Global context effects on processing lexically ambiguous words: Evidence from eye fixations

Readers' eye movements were monitored as they read biased ambiguous target words in the context of a short paragraph. Two aspects of context were manipulated. The global context was presented in the topic sentence of the paragraph and instantiated either the dominant or the subordinate meaning of biased ambiguous target words (those with highly dominant meanings). Local contextual information either preceded or followed the target word and was always consistent with the subordinate interpretation. Consistent with prior research, we obtained a subordinate bias effect wherein readers looked longer at the ambiguous words than control words when the preceding context instantiated the subordinate meaning. More importantly, the magnitude of the subordinate bias effect was the same when global context alone, local context alone, or local and global context combined were consistent with the subordinate meaning of the ambiguous word. The results of this study indicate that global contextual information (1) has an immediate impact on lexical ambiguity resolution when no local disambiguating information is available, (2) has no additional effect when it is consistent with local information, but (3) does have a slightly delayed effect when inconsistent with local information.     

Practice makes perfect: Training the interpretation of emotional ambiguity

The interpretation of emotionally ambiguous words, sentences, or scenarios can be altered through training procedures that are collectively called cognitive bias modification for interpretation (CBM-I). In three experiments, we systematically manipulated the nature of the training in order to discriminate between emotional priming and ambiguity resolution accounts of training effects. In Experiment 1 participants completed word fragments that were consistently related to either a negative or benign interpretation of an ambiguous sentence. In a subsequent semantic priming task they demonstrated an interpretation bias, in that they were faster to identify relatedness of targets that were associated with the training-congruent meaning of an emotionally ambiguous homograph. We then manipulated the training sentences to show that interpretation bias was eliminated when participants simply completed valenced word fragments following unrelated sentences (Experiment 2), or completed fragments that were related to emotional but unambiguous sentences (Experiment 3). Only when participants were required to actively resolve emotionally ambiguous sentences during training did changes in interpretation emerge at test. Findings suggest that CBM-I achieves its effects by altering a production rule that aids the selection of meaning from emotionally ambiguous alternatives, in line with an ambiguity resolution account.     

The Development of Strategic Processing of Ambiguous Words: Riddles Versus Neutral Context

Children in Grades 4 and 6 read isolated target words following the auditory presentation of an ambiguous word (e.g., deck). The ambiguous word occurred as the last word in either a neutral sentence (e.g., “There is the deck”) or as the last word in a riddle (“Why couldn't anybody play poker on the ark? Because Noah sat on the deck). Related target words were consistent with either the dominant meaning (e.g., porch) or the subordinate meaning (cards) of the ambiguous word. In the neutral context, Grade 4 children showed equivalent facilitation for dominant and subordinate targets relative to unrelated target words; Grade 6 children showed facilitation only for the dominant targets. In the riddle context, both groups of children showed equivalent facilitation for the two types of related targets. The neutral context results supported the hypothesis that the two bases of context effects—automatic lexical access and selective access—develop at different rates (Simpson &; Foster, 1986). The results obtained from the riddles suggest that the selective process is strategic (Ceci, 1989).     

Resolving syntactic category ambiguity: An eye-movement analysis

In two experiments, participants' eye movements were monitored as they read sentences containing biased syntactic category ambiguous words with either distinct (e.g., duck) or related (e.g., burn) meanings or unambiguous control words. In Experiment 1, prior context was consistent with either the dominant or subordinate interpretation of the ambiguous word. The subordinate bias effect was absent for the ambiguous words in gaze duration measures. However, effects of ambiguity did emerge in other measures for the ambiguous words preceded by context supporting the subordinate interpretation. In Experiment 2, context preceding the target words was neutral. Ambiguity effects only arose when posttarget context was consistent with the subordinate interpretation of the ambiguous words, indicating that readers initially selected the dominant interpretation. Results support immediate theories of syntactic category ambiguity resolution, but also suggest that recovery from misanalysis of syntactic category ambiguity is more difficult than for lexical-semantic ambiguity in which alternate interpretations do not cross syntactic category.     

Contextual strength does not modulate the subordinate bias effect: Evidence from eye fixations and self-paced reading

Eye movements were recorded in order to examine how different sources of information—namely, meaning dominance and strength of biasing context—influence the processing of biased ambiguous words. Gaze durations were longer on ambiguous target words when the preceding context instantiated the subordinate interpretation, even with strongly biasing contexts. Identical results were obtained with a self-paced reading study. Thus, contrary to recent findings (Kellas, Martin, Yehling, Herman, & Vu, 1995), the subordinate interpretation of a biased ambiguous word was not selectively accessed even when the preceding context strongly biased that interpretation. Discrepancies between the present experiments and the Kellas et al. experiment are discussed.     

Understanding words in sentence contexts: the time course of ambiguity resolution

Spoken language comprehension requires rapid integration of information from multiple linguistic sources. In the present study we addressed the temporal aspects of this integration process by focusing on the time course of the selection of the appropriate meaning of lexical ambiguities ("bank") in sentence contexts. Successful selection of the contextually appropriate meaning of the ambiguous word is dependent upon the rapid binding of the contextual information in the sentence to the appropriate meaning of the ambiguity. We used the N400 to identify the time course of this binding process. The N400 was measured to target words that followed three types of context sentences. In the concordant context, the sentence biased the meaning of the sentence-final ambiguous word so that it was related to the target. In the discordant context, the sentence context biased the meaning so that it was not related to the target. In the unrelated control condition, the sentences ended in an unambiguous noun that was unrelated to the target. Half of the concordant sentences biased the dominant meaning, and the other half biased the subordinate meaning of the sentence-final ambiguous words. The ISI between onset of the target word and offset of the sentence-final word of the context sentence was 100 ms in one version of the experiment, and 1250 ms in the second version. We found that (i) the lexically dominant meaning is always partly activated, independent of context, (ii) initially both dominant and subordinate meaning are (partly) activated, which suggests that contextual and lexical factors both contribute to sentence interpretation without context completely overriding lexical information, and (iii) strong lexical influences remain present for a relatively long period of time.     

Syntactic context and lexical access

Recent studies demonstrating that multiple meanings of an ambiguous word are initially accessed even when only one reading is syntactically appropriate with the preceding context can be criticized on at least two grounds. First, many of the syntactic contexts used were not truly restrictive, and, secondly, subjects may not have had time to integrate the context before processing the ambiguous word. In the present study, subjects listened to a sentence ending in an ambiguous word and then made a lexical decision to a target related to either the appropriate or inappropriate reading. Contexts were completely restrictive, and a pause was introduced between the context and the ambiguous word. Multiple access still obtained, providing further support for the claim that lexical access is not guided by syntactic context.     

Semantic constraints and judged preference for interpretations of ambiguous sentences

It is proposed that the degree of sensibleness of sentences is determined by semantic constraints which may be more or less satisfied. Such continuous semantic constraints were examined in two experiments in which subjects judged the likelihood of obtaining each of the interpretations of ambiguous sentences. The sentences were factorially generated by independently varying the degree to which semantic constraints for each interpretation were satisfied. In one experiment, the semantic constraints were manipulated by varying critical words within the ambiguous sentence; in the other experiment, a preceding context sentence was used. The results of both experiments supported the hypotheses that the judged likelihood was a direct function of the relative sensibleness of the interpretations, that semantic constraints determined the degree of sensibleness of each interpretation, and that these semantic constraints are continuous restrictions which are independent of each other and stable from sentence to sentence in which they occur.     

The Effect of Non-sentential Context Prosody on Homographs’ Lexical Activation in Persian

This study examines the effect of non-sentential context prosody pattern on lexical activation in Persian. For this purpose a questionnaire including target and non-target words is used. The target words are homographs with two possible stress patterns belonging to different syntactic categories. Participants are asked to read out the words aloud and note the first word that comes to their mind. The results show that by reading the target words, both meanings of the target words are activated in mind and the prosodic pattern of the non-sentential preceding context does not affect the activation of the other stress pattern meaning. This result suggests that the metrical prosodic pattern of non-sentential context is not a strong constraint to determine which meaning of the target word must be activated. The experiment also illustrates that the stress pattern used to read the target words does not necessarily matches the stress pattern of the target word which relates to the written word. These findings confirm Swinney (Verb Learn Verb Behav 18:645–665, 1979) and Elston-Güttler and Friederici’s (J Mem Lang 52(2):256–283, 2005) finding that both meanings of an ambiguous word are accessed at the first stage. This study shows that in lack of semantic context, Persian natives behave homographs as ambiguous words and there is no bias towards preferring one meaning over another.     

Or cues knowledge of alternatives: Evidence from priming

Recent research has determined that word meanings can instantly influence the meaning and distribution of other words in the sentence. Here, we manipulated basic carrier sentences with the disjunction or linking two nouns that were either filling the same thematic role or not, and were either semantically related or not. Though previous research has shown that one word can prime a semantically related word even in a sentential context, we predicted that if or cues knowledge about contextually‐relevant alternatives, priming for semantic relatives will only obtain when those words also fill the same thematic role. These predictions were confirmed, as self‐paced reading times of the second alternative in the sentence were faster only when the two alternatives shared the same thematic role and semantic category, suggesting that disjunction words like or function similarly to verbs, which cue knowledge about expected argument structure and sense depending on sentential context. The relevance of these findings for basic reasoning phenomena (i.e., the subadditivity effect) is also discussed.     

Expectations from preceding prosody influence segmentation in online sentence processing

Previous work examining prosodic cues in online spoken-word recognition has focused primarily on local cues to word identity. However, recent studies have suggested that utterance-level prosodic patterns can also influence the interpretation of subsequent sequences of lexically ambiguous syllables (Dilley, Mattys, & Vinke, Journal of Memory and Language, 63:274–294, 2010; Dilley & McAuley, Journal of Memory and Language, 59:294–311, 2008). To test the hypothesis that these distal prosody effects are based on expectations about the organization of upcoming material, we conducted a visual-world experiment. We examined fixations to competing alternatives such as pan and panda upon hearing the target word panda in utterances in which the acoustic properties of the preceding sentence material had been manipulated. The proportions of fixations to the monosyllabic competitor were higher beginning 200 ms after target word onset when the preceding prosody supported a prosodic constituent boundary following pan-, rather than following panda. These findings support the hypothesis that expectations based on perceived prosodic patterns in the distal context influence lexical segmentation and word recognition.     

Semantic context effects and priming in word association

Two experiments investigated priming in word association, an implicit memory task. In the study phase of Experiment 1, semantically ambiguous target words were presented in sentences that biased their interpretation. The appropriate interpretation of the target was either congruent or incongruent with the cue presented in a subsequent word association task. Priming (i.e., a higher proportion of target responses relative to a nonstudied baseline) was obtained for the congruent condition, but not for the incongruent condition. In Experiment 2, study sentences emphasized particular meaning aspects of nonambiguous targets. The word association task showed a higher proportion of target responses for targets studied in the more congruent sentence context than for targets studied in the less congruent sentence context. These results indicate that priming in word association depends largely on the storage of information relating the cue and target.     

Remindings influence the interpretation of ambiguous stimuli

Remindings—stimulus-guided retrievals of prior events—may help us interpret ambiguous events by linking the current situation to relevant prior experiences. Evidence suggests that remindings play an important role in interpreting complex ambiguous stimuli (Ross & Bradshaw Memory & Cognition, 22, 591–605, 1994); here, we evaluate whether remindings will influence word interpretation and memory in a new paradigm. Learners studied words on distinct visual backgrounds and generated a sentence for each word. Homographs were preceded by a biasing cue on the same background three items earlier, preceded by a biasing cue on a different background three items earlier, or followed by a biasing cue on the same background three items later. When biasing cues preceded the homographs on the same backgrounds as the homographs, the meanings of the homographs in learner-generated sentences were consistent with the biasing cues more often than in the other two conditions. These results show that remindings can influence word interpretation. In addition, later memory for the homographs and cues was greater when the meaning of the homograph in the sentence was consistent with the earlier biasing cue, suggesting that remindings enhanced mnemonic performance. Remindings play an important role in how we interpret ambiguous stimuli and enhance memory for the involved material.     

Context and lexical access: implications of nonword interference for lexical ambiguity resolution

To eliminate potential "backward" priming effects, Glucksberg, Kreuz, and Rho (1986) introduced a variant of the cross-modal lexical priming task in which subjects made lexical decisions to nonword targets that were modeled on a word related to either the contextually biased or unbiased sense of an ambiguous word. Lexical decisions to nonwords were longer than controls only when the nonword was related to the contextually biased sense of the ambiguous word, leading Glucksberg et al. to conclude that context does constrain lexical access and that the multiple access pattern observed in previous studies was probably an artifact of backward priming. We did not find nonword interference when the nonword targets used by Glucksberg et al. were preceded by semantically related ambiguous or unambiguous word primes. However, we did replicate their sentence context results when the ambiguous words were removed from the sentences. We conclude that the interference obtained by Glucksberg et al. is due to postlexical judgements of the congruence of the sentence context and the target, not to context constraining lexical access.     

The maze task: Measuring forced incremental sentence processing time

The maze task is an online measure of sentence processing time that provides an alternative to the standard moving window version of self-paced reading. Rather than each word of the sentence being presented in succession, two words are presented at the same time, and the participant must choose which word is a grammatical continuation of the sentence. This procedure forces the reader into an incremental mode of processing in which each word must be fully integrated with the preceding context before the next word can be considered. Previous research with this technique has not considered whether it is sufficiently sensitive to syntactic complexity effects or to garden path effects. Four experiments are reported demonstrating that reliable differences in processing time for subject relatives and object relatives can be obtained, and that this technique generates garden path effects that correspond closely with the data from eyetracking experiments, but without the spillover effects that are sometimes obtained with eyetracking. It is also shown that the task is sensitive to word frequency effects, producing estimates well in excess of those found with eyetracking.     

Accessing ambiguous words during sentence comprehension

In five experiments visual processing of sentences containing either a more frequent or a less frequent usage of an ambiguous word was examined. When prior or subsequent context was pragmatically related to the relevant sense of the ambiguous word, sentences intending the more frequent sense produced longer ambiguity detection times and shorter immediate comprehension times than sentences intending the less frequent sense. This relative frequency effect was not obtained in comprehension of corresponding unambiguous sentences containing relatively high or low frequency unambiguous synonyms of the senses. These results suggest that access of ambiguous word-senses tends to occur in order of relative frequency, and that multiple access of senses tends only to occur for low frequency usages. When the preceding verb imposed selection constraints on which sense could follow, the frequency effect did not occur consistently; and it was virtually eliminated when biasing context took the form of a previous sentence containing an unambiguous synonym of the relevant sense. Implications for models of access of unambiguous words in sentences, as well as for models of processing ambiguous sentences, are considered.     

Selection mechanisms in reading lexically ambiguous words

Readers' eye movements were monitored as they read sentences containing lexically ambiguous words. The ambiguous words were either biased (one strongly dominant interpretation) or nonbiased. Readers' gaze durations were longer on nonbiased than biased words when the disambiguating information followed the target word. In Experiment 1, reading times on the disambiguating word did not differ whether the disambiguation followed the target word immediately or occurred several words later. In Experiment 2, prior disambiguation eliminated the long gaze durations on nonbiased target words but resulted in long gaze durations on biased target words if the context demanded the subordinate meaning. The results indicate that successful integration of one meaning with prior context terminates the search for alternative meanings of that word. This results in selective (single meaning) access when integration of a dominant meaning is fast (due to a biasing context) and identification of a subordinate meaning is slow (a strongly biased ambiguity with a low-frequency meaning).     

Lossy-Context Surprisal: An Information-Theoretic Model of Memory Effects in Sentence Processing

A key component of research on human sentence processing is to characterize the processing difficulty associated with the comprehension of words in context. Models that explain and predict this difficulty can be broadly divided into two kinds, expectation-based and memory-based. In this work, we present a new model of incremental sentence processing difficulty that unifies and extends key features of both kinds of models. Our model, lossy-context surprisal, holds that the processing difficulty at a word in context is proportional to the surprisal of the word given a lossy memory representation of the context—that is, a memory representation that does not contain complete information about previous words. We show that this model provides an intuitive explanation for an outstanding puzzle involving interactions of memory and expectations: language-dependent structural forgetting, where the effects of memory on sentence processing appear to be moderated by language statistics. Furthermore, we demonstrate that dependency locality effects, a signature prediction of memory-based theories, can be derived from lossy-context surprisal as a special case of a novel, more general principle called information locality.     

Influence of context on the activation and selection of ambiguous word senses

The activation of ambiguous word senses was investigated by measuring the amount of interference in naming the ink color of a word that was either related or unrelated to one of the meanings of a preceding ambiguous word. In agreement with previous results obtained using this procedure (Conrad, 1974), evidence was obtained that both meanings of the ambiguous words are activated even in the presence of biasing context. However, contrary to previous findings, the degree of activation of each word sense depended on its degree of compatibility with the context. These results are consistent with a language processing system in which all interpretations of an ambiguity are accessed and then processed until an accurate determination has been made of which interpretation best satisfies the syntactic and semantic constraints that govern it.