Semantic Bias Research Articles

Acquisition of tense-aspect morphology and the regular-irregular debate

This paper reviews research on English past-tense acquisition to test the validity of the single mechanism model and the dual mechanism model, focusing on regular-irregular dissociation and semantic bias. Based on the review, it is suggested that in L1 acquisition, both regular and irregular verbs are governed by semantics; that is, early use of past tense forms are restricted to achievement verbs—regular or irregular. In contrast, some L2 acquisition studies show stronger semantic bias for regular past tense forms (e.g., Housen, 2002, Rohde, 1996). It is argued that L1 acquisition of the past-tense morphology can be accounted for more adequately by the single-mechanism model.

Semantic bias, homograph comprehension, and event-related potentials in schizophrenia

Objectives: It is controversial whether a semantic processing bias for strong associates is present in schizophrenia, and unknown whether the language abnormalities observed in schizophrenia can be attributed to dysfunctions early or late in cognitive processing. Combined behavioral and event-related potential (ERP) data can indicate the nature and timing of such abnormalities. Methods: Sensibility judgements of dominant and subordinate homograph sentences were measured in 12 schizophrenia patients and 13 normal controls. ERPs were recorded to the disambiguating sentence-ending word. Results: All subjects showed greatest misinterpretation of subordinate homograph sentences, but schizophrenia patients particularly misinterpreted these sentence types. For control subjects, subordinate homograph sentences that were classified as nonsensical showed greater N400 than those classified as sensible. By contrast, the N400 of patients was large, regardless of the sensibility judgement – patients’ brains initially responded to all subordinate sentences as if nonsensical. These data are consonant with a semantic bias. However, the patients’ N400 to dominant homograph sentence endings was also larger than that of controls, a finding not consonant with a semantic bias. Conclusions: The behavioral results indicate a selective comprehension abnormality in schizophrenia dependent on the content of verbal memory. The ERP results suggest a pervasive contextual memory failure. A semantic activation decay model is proposed to explain these results.

Contextual Influences on Phonetic Identification in Aphasia: The Effects of Speaking Rate and Semantic Bias

Two experiments examined the influence of context on stop-consonant voicing identification in fluent and nonfluent aphasic patients and normal controls. Listeners were required to label the initial stop in a target word varying along a voice onset time (VOT) continuum as either voiced or voiceless ([b]/[p] or [d]/[t]). Target stimuli were presented in sentence contexts in which the rate of speech of the sentence context (Experiment 1) or the semantic bias of the context (Experiment 2) was manipulated. The results revealed that all subject groups were sensitive to the contextual influences, although the extent of the context effects varied somewhat across groups and across experiments. In addition, a number of patients in both the fluent and nonfluent aphasic groups could not consistently identify even endpoint stimuli, confirming phonetic categorization impairments previously shown in such individuals. Results are discussed with respect to the potential reliance by aphasic patients on higher level context to compensate for phonetic perception deficits.

Masked Semantic Priming of Emotional Information in Subclinical Depression

Most cognitive theories of depression (e.g., Beck, 1976; Ingram, 1984; Teasdale, 1988; Bower, 1981) assume that depressed individuals have an automatic processing bias for negative information (e.g., activation of negative material in memory). In contrast, Williams, Watts, MacLeod, and Mathews (1988) proposed that depression is associated with a negative bias in controlled, rather than automatic, memory processes. Two experiments investigated whether there is an emotion-congruent bias in automatic (implicit) memory in subclinical depression. The first used a primed lexical decision task with briefly presented, masked primes (prime-target stimulus onset asynchrony, SOA, of 56 msec), including both repetition priming and semantic priming conditions. A depression-congruent priming bias was demonstrated in the semantic condition only. The second experiment examined the time course of the depression-congruent semantic priming bias using 56- and 2000-msec SOAs, and confirmed its occurrence in the 56-msec SOA condition. Results of both experiments are interpreted as consistent with a depression-congruent bias in automatic memory processes. Theoretical and clinical implications are discussed.

Event‐related potentials elicited during a context‐free homograph task in normal versus schizophrenic subjects

Thought disorder in schizophrenia may involve abnormal semantic activation or faulty working memory maintenance. Event-related potentials (ERPs) were recorded while sentences reading "THE NOUN WAS ADJECTIVE/VERB" were presented to 34 schizophrenic and 34 control subjects. Some nouns were homographs with dominant and subordinate meanings. Their sentence ending presented information crucial for interpretation (e.g., The bank was [closed, steep]). Greatest N400 activity to subordinate homograph-meaning sentence endings in schizophrenia would reflect a semantic bias to strong associates. N400 to all endings would reflect faulty verbal working memory maintenance. Schizophrenic subjects showed N400 activity to all endings, suggesting problems in contextual maintenance independent of content, but slightly greater N400 activity to subordinate endings that correlated with the severity of psychosis. Future research should help determine whether a semantic activation bias in schizophrenia toward strong associates is reflected in ERP activity or whether this effect is overshadowed by faulty verbal working memory maintenance of context.

Event-related potentials elicited during a context-free homograph task in normal versus schizophrenic subjects

Thought disorder in schizophrenia may involve abnormal semantic activation or faulty working memory maintenance. Event-related potentials (ERPs) were recorded while sentences reading “THE NOUN WAS ADJECTIVE/VERB” were presented to 34 schizophrenic and 34 control subjects. Some nouns were homographs with dominant and subordinate meanings. Their sentence ending presented information crucial for interpretation (e.g., The bank was [closed, steep]). Greatest N400 activity to subordinate homograph-meaning sentence endings in schizophrenia would reflect a semantic bias to strong associates. N400 to all endings would reflect faulty verbal working memory maintenance. Schizophrenic subjects showed N400 activity to all endings, suggesting problems in contextual maintenance independent of content, but slightly greater N400 activity to subordinate endings that correlated with the severity of psychosis. Future research should help determine whether a semantic activation bias in schizophrenia toward strong associates is reflected in ERP activity or whether this effect is overshadowed by faulty verbal working memory maintenance of context.

Effects and limitations of prosodic and semantic biases on syntactic disambiguation.

This paper examined the effects of prosody on the syntactic ambiguity resolution of Japanese sentences, especially with reference to the interaction with semantic bias. Syntactically ambiguous sentences with different types of semantic bias were constructed. The degree of bias in each sentence was evaluated through visual presentation experiments. Three types of sentences were selected based on the results of visual presentation experiments, were recorded with prosody maximally favoring each possible interpretation of the sentences, and were used as the stimuli for the auditory presentation experiments. The results showed that prosodic cues can influence the interpretation of a sentence even when the sentence is strongly semantically biased. The results also showed a limitation to prosodic cues. The prosodic biases alone were not sufficient to fully determine the interpretation of the sentences even when the sentences were neutrally biased semantically.

Constrained relative least general generalization for Inducing Constraint Logic Programs

Relative least general generalization proposed by Plotkin, is widely used for generalizing first-order clauses in Inductive Logic Programming, and this paper describes an extension of Plotkin’s work to allow various computation domains: Herbrand Universe, sets, numerical data, ect. The ϕ-subsumption in Plotkin’s framework is replaced by a more general constraint-based subsumption. Since this replacement is analogous to that of unification by constraint solving in Constraint Logic Programming, the resultant method can be viewed as a Constraint Logic Programming version of relative least general generalization. Constraint-based subsumption, however, leads to a search on an intractably large hypothesis space. We therefore providemeta-level constraints that are used as semantic bias on the hypothesis language. The constraintsfunctional dependency andmonotonicity are introduced by analyzing clausal relationships. Finally, the advantage of the proposed method is demonstrated through a simple layout problem, where geometric constraints used in space planning tasks are produced automatically.

The Processing of Locally Ambiguous Relative Clauses in German

The processing of locally ambiguous German relative clauses was studied with a self-paced reading task. In three experiments, subjects read sentences which contained either a subject relative clause or an object relative clause. The part of the sentence preceding the lexical item disambiguating between subject and object relative clause had a semantic bias either toward a subject relative clause reading or toward an object relative clause reading, or was neutral with respect to the two readings. The results showed a preference for the subject relative clause reading in the neutral condition. This result is in line with the Active Filler Strategy, a parsing principle which has been proposed for the processing of filler-gap constructions (e.g., Frazier and Flores d′Arcais, 1989). Furthermore, semantic plausibility information preceding the disambiguating item did not immediately influence the filler-gap assignment in the present study. The latter result is discussed in relation to studies on the processing of filler-gap constructions in English which do show an immediate influence of semantic plausibility on filler-gap assignment.

Declarative bias for specific-to-general ILP systems

A comparative study is presented of language biases employed in specific-to-general learning systems within the Inductive Logic Programming (ILP) paradigm. More specifically, we focus on the biases employed in three well known systems: CLINT, GOLEM and ITOU, and evaluate both conceptually and empirically their strengths and weaknesses. The evaluation is carried out within the generic framework of the NINA system, in which bias is a parameter. Two different types of biases are considered: syntactic bias, which defines the set of well-formed clauses, and semantic bias, which imposes restrictions on the behaviour of hypotheses or clauses. NINA is also able to shift its bias (within a predefined series of biases), whenever its current bias is insufficient for finding complete and consistent concept definitions. Furthermore, a new formalism for specifying the syntactic bias of inductive logic programming systems is introduced.

The impact of context and semantic bias on constituent attachment in reading

The impact of context and semantic bias on constituent attachment in reading

Quantifying Prior Determination Knowledge Using the PAC Learning Model

Prior knowledge, or bias, regarding a concept can reduce the number of examples needed to learn it. Probably Approximately Correct (PAC) learning is a mathematical model of concept learning that can be used to quantify the reduction in the number of examples due to different forms of bias. Thus far, PAC learning has mostly been used to analyze syntactic bias, such as limiting concepts to conjunctions of boolean prepositions. This paper demonstrates that PAC learning can also be used to analyze semantic bias, such as a domain theory about the concept being learned. The key idea is to view the hypothesis space in PAC learning as that consistent with all prior knowledge, syntactic and semantic. In particular, the paper presents an analysis of determinations, a type of relevance knowledge. The results of the analysis reveal crisp distinctions and relations among different determinations, and illustrate the usefulness of an analysis based on the PAC learning model.

The Interaction of Lexical and Syntactic Ambiguity

Two experiments investigated comprehension of noun/verb lexical category ambiguities such as trains, in order to determine whether resolution of these ambiguities was similar to other types of ambiguity resolution. Frazier and Rayner (1987, Journal of Memory and Language, 26, 505-526) argued that these ambiguities were resolved with a delay strategy that is not used for other ambiguities. Experiment l′s self-paced reading data replicated Frazier & Rayner′s results but also showed that evidence taken to support delay had other explanations. Experiment 2 investigated the influence of semantic biases on ambiguity resolution and found that three probabilistic factors influenced lexical category ambiguity resolution: (1) the relative frequency of head vs. modifying noun usage of a biasing noun, (2) the frequency of cooccurrence of a biasing noun and category ambiguous word in English, and (3) the combinatorial semantic information in the sentence. The extent to which alternative models account for the use of probabilistic information in ambiguity resolution is discussed.

Acquisition of selectional patterns in sublanguages

When implementing computational lexicons it is important to keep in mind the texts that a NLP system must deal with. Words relate to each other in many different, often odd ways this information is rarely found in dictionaries, and it is quite hard to deduce a priori. In this paper we present a technique for the acquisition of statistically significant selectional restrictions from corpora and discuss the results of an experimental application with reference to two specific sublaguages (legal and commercial). We show that there are important cooccurrence preferences among words which cannot be established a priori as they are determined for each choice of sublanguage. The method for detecting cooccurrences is based on the analysis of word associations augmented with syntactic markers and semantic tags. Word pairs are extracted by a morphosyntactic analyzer and clustered according to their semantic tags. A statistical measure is applied to the data to evaluate the sigificance of any relations detected. Selectional restrictions are acquired by a two-step process. First, statistically prevailing ‘coarse grained’ conceptual patterns are used by a linguist to identify the relevant selectional restrictions in sublanguages. Second, semiautomatically acquired ‘coarse’ selectional restrictions are used as the ‘semantic bias’ of a system, ARIOSTO_LEX, for the automatic acquisition of a case-based semantic lexicon.

The multiple classification of acts and the big five factors of personality

The present study investigated whether Norman's five highly replicable factors of personality could be found in the intercorrelations among prototypicality ratings of acts for traits. Moreover, by factor analyzing cross-correlations among prototypicality ratings, in which the ratings for each trait pair stemmed from different judges, it was checked whether the intercorrelations among prototypicality ratings are influenced by semantic bias. Eight judges rated the prototypicality of 120 varbally described acts for Norman's 40 marker traits. The prototypicality of each act was judged with respect to each trait. When the correlations among the prototypicality ratings were factor analyzed, Norman's Big Five could clearly be identified. This finding was replicated for the cross-correlations, and it was thus demonstrated that semantic bias is unimportant for the correlations among prototypicality ratings. It is argued that the five major factors of personality reflect basic dimensions of meaning that underlie the attribution of traits on the basis of the observation of behavior.

Nine types of transformations: Is linguistics finally becoming meaningful?

Noam Chomsky has led the way for linguists in the development of models which are “elegant” in the scientific sense. Generative transformational grammar is simple, complete, internally consistent, and rule-governed. From its inception, it has had a syntactic bias, however, and Chomsky has retained this syntactic bias through Interpretive Semantics, Extended Standard Theory and into Government and Binding. However, some linguists have taken Chomsky's model in a different direction, developing a semantic bias instead, a bias which Zellig Harris and Noam Chomsky had never originally foreseen or intended. George Lakoff and others have extended the concept to deal with different parts of speech (e.g. “use” vs “with”). Charles Fillmore and others have applied it to converses (e.g. “buy” vs “sell”). The generative semantists have argued that verbs like “kill” can be lexically decomposed into “cause,” “become” and “dead,” and this lexical decomposition is transformational in nature. More recently, the model of componential analysis in particlar and network models in general have extended this lexical decomposition process into all semantic areas, and again the concept of the grammatical transformation has proven very valuable.

Sentence comprehension strategies in specifically language impaired children.

ABSTRACTThis study investigated sentence comprehension and the use of sentence comprehension strategies by specifically language impaired children. Eleven specifically language impaired (SLI) children (aged 4;0 to 6;0), chosen according to strict criteria, were compared with eleven children matched on chronological age and eleven language controls matched on the receptive scores from the Reynell Developmental Language Scale and the British Picture Vocabulary Scale. Subjects acted out active and passive voice sentences that had either a positive or negative semantic bias or were unbiased semantically. The SLI children performed at a signficantly lower level than both the age and language control subjects. They relied more heavily on semantic cues than the language control children and all eleven consistently applied a ‘probable event strategy’. Word order responses that took the first noun to be the actor were used with equal frequency by SLI children and language controls. However, these responses were used less discriminately by the SLI children than the language controls who applied them more frequently to active voice sentences than to passives. The results suggest that the SLI children primarily based their interpretation on semantic expectations or the sequence of content words and relied little on syntactic information.

Can linguistic boundaries change the effectiveness of silence as a phonetic cue?

This study investigated the influence of three kinds of linguistic boundaries—word boundaries, prosodic breaks, and syntactic breaks—on the perception of a silent interval at the boundary site as a cue to the presence of a labial stop consonant. The experimental technique involved cross-splicing of portions of four naturally produced pairs of sentences, as well as presentation of excerpts from these sentences. Although one sentence pair showed a pronounced syntactic boundary effect, the other three (including two that were better controlled for semantic bias) did not, which points to a different, stimulus-specific origin of the effect obtained. Prosodic boundary effects were also generally absent, presumably because the stimuli were constructed such that prosodic variation ceased 78 ms prior to the critical silent interval. Only introduction of a word boundary effected a systematic reduction in stop consonant percepts, although this manipulation was confounded with other contextual factors. On the whole, the data provide little evidence for any direct effects of structural linguistic variables on phonetic segment perception; such effects seem to be restricted to the level of word recognition.

Monozygotic twin cross-correlations as a validation of personality structure: A test of the semantic bias hypothesis.

Monozygotic twin cross-correlations as a validation of personality structure: A test of the semantic bias hypothesis.

Monozygotic twin cross-correlations as a validation of personality structure: A test of the semantic bias hypothesis.

Monozygotic twin cross-correlations as a validation of personality structure: A test of the semantic bias hypothesis.