No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Monitoring sentence comprehension
... These data provide the appropriate background for both a re-analysis of past results obtained with this technique and a discussion of the paradigm's strengths and weaknesses, both of which we undertake throughout the paper. Cutler and Norris (1979) offer a thorough discussion of three kinds of detection tasks employed in psycholinguistics and argue that phoneme-and word-monitoring tasks diverge from a tone-monitoring task in that, inter alia, the former exhibit a general decrease in RTs across a sentence, an effect Cutler and Norris deny to the tone monitoring experiments of Abrams and Bever (1969) and Holmes and Forster (1970). This conclusion, however, is based on a lessthan-careful analysis of the data that Abrams and Bever (1969), in particular, report. ...
... As mentioned in the previous section, these authors established three different click positions in sentences such as since she was free that | day | her | friends asked her to come (before the main clause break, in the clause break, and right after the clause break, all marked with |), and the RTs they obtained certainly exhibit a decrease: 243 ms., 230, and 216. So why do Cutler and Norris (1979) conclude otherwise? Abrams and Bever (1969) exposed their participants to repeated presentations of the same material, and participants' performance progressively improved. ...
... Abrams and Bever (1969) exposed their participants to repeated presentations of the same material, and participants' performance progressively improved. The RTs we have just provided were those of the first presentation, and it is only in the next two presentations wherein the linear decrease in RTs from the first to the third position disappears-the pattern that Cutler and Norris (1979) focus on. Given that these participants were reacting to familiar sentences and click positions in the second and third presentations, those responses are not comparable to those of the other monitoring tasks Cutler and Norris (1979) discuss (or other click-detection studies). ...
Franco, Gaillard, Cleeremans, and Destrebecqz (Behavior Research Methods, 47, 1393-1403, 2015), in a study on statistical learning employing the click-detection paradigm, conclude that more needs to be known about how this paradigm interacts with statistical learning and speech perception. Past results with this monitoring technique have pointed to an end-of-clause effect in parsing-a structural effect-but we here show that the issues are a bit more nuanced. Firstly, we report two Experiments (1a and 1b), which show that reaction times (RTs) are affected by two factors: (a) processing load, resulting in a tendency for RTs to decrease across a sentence, and (b) a perceptual effect which adds to this tendency and moreover helps neutralize differences between sentences with slightly different structures. These two factors are then successfully discriminated by registering event-related brain potentials (ERPs) during a monitoring task, with Experiment 2 establishing that the amplitudes of the N1 and P3 components-the first associated with temporal uncertainty, the second with processing load in dual tasks-correlate with RTs. Finally, Experiment 3 behaviorally segregates the two factors by placing the last tone at the end of sentences, activating a wrap-up operation and thereby both disrupting the decreasing tendency and highlighting structural effects. Our overall results suggest that much care needs to be employed in designing click-detection tasks if structural effects are sought, and some of the now-classic data need to be reconsidered.
... Each occurred as a word-initial singleton in twelve target words. Manner of articulation, word position, and singleton vs. cluster effects on phoneme monitoring RT were thus controlled [16,17,18]. All target words were familiar to children, with a log frequency greater than 3.00 in the CBBC section of the SUBTLEX database [19]. ...
... The target phoneme always occurred in the fifth or sixth syllable of the sentence, to control for sentence-position effects on RT [18]. When recorded, each sentence was produced with the target word prosodically focused, to facilitate phoneme monitoring and control for prosodic effects on RT [16]. Sixteen catch sentences were also created, containing no instances of any target phonemes but matching the test sentences in length and prosodic structure. ...
... Experimental methods can be characterised in many ways, one of which is the division into global and local tasks (see Levelt, 1978;Cutler & Norris, 1979, for similar distinctions). ...
... Except for a few studies (Dell & Newman, 1980;Eimas & Nygaard, 1992 ), the contextual predictability of a critical word was assessed through monitoring responses to the initial phoneme of the word directly following it. With monitoring times to the initial phoneme ot the critical word itself, it is uncertain whether responses are made on the basis of lexical or sublexical levels ot representation (see Cutler & Norris, 1979;Cutler, Mehler, Norris, & Segui, 1987). When responses are made to the first phoneme of the next word, this does not apply. ...
Listeners sometimes have the impression that they know exactly which word a speaker is going to say next. However real such observations are, they are the exception rather than the rule. More than forty years of research has shown that recognizing spoken words is not a guessing game. Normally, word recognition is an extremely fast and efficient process that rarely reaches conscious awareness. Fluent speech is uttered at a rate of two to three words per second, and an adult language user has a mental lexicon in which the knowledge of about 30 000 to 50 000 words is stored (Aitchison, 1994). This implies that a listener has about one third of a second to select one word from this huge mental data base.
... The top-down interpretation of this effect is that when the lexical representation of peel is activated, it correspondingly biases activity to the phonemes within that word. However, Cutler, Norris, and colleagues have argued that these results can be explained without resort to top-down mechanisms (Cutler & Norris, 1979;Norris, McQueen, & Cutler, 2000). Cutler and Norris (1979) suggest that there are two levels of phonemic representation, one that represents the input and one that represents the system's 'guess' based on both the input and the context. ...
... However, Cutler, Norris, and colleagues have argued that these results can be explained without resort to top-down mechanisms (Cutler & Norris, 1979;Norris, McQueen, & Cutler, 2000). Cutler and Norris (1979) suggest that there are two levels of phonemic representation, one that represents the input and one that represents the system's 'guess' based on both the input and the context. Norris, McQueen, and Cutler (2000) suggest that there is a level of representation essentially dedicated to phoneme identification, and that it is activity at this level and not at the phoneme level that receives input from the context. ...
... According to the autonomous race model (Cutler & Norris, 1979), subjects performing the phoneme monitoring task respond on the basis of either a prelexical or a lexical code dependent on which code becomes available first. The prelexical code becomes available from the speech signal, whereas the lexical code is obtained through accessing a particular lexical entry. ...
... The TRACE model (McClelland & Elman, 1986) must assume that there is a fast spreading of activation from a lexical representation to a graphemic representation that can affect the phoneme detection response. The race model (Cutler & Norris, 1979) must adopt the assumption that auditory lexical access also involves the retrieval of orthographic representations, which can contribute to the detection decision in addition to phonological representations. ...
A widely used task in the research on spoken word recognition is phoneme monitoring, in which subjects have to detect phonemes in spoken words. It is generally assumed that this task is performed using phonetic or phonological representations of words only. To test whether an orthographic representation of the words is employed as well, an experiment was conducted in which Dutch subjects monitored for phonemes with either a primary or secondary spelling in phonologicall y matched spoken words and nonwords. Phoneme monitor- ing times were slower when the phoneme had a secondary spelling than when it had a primary spelling. The effect was greater after than before the uniqueness point of the word, and monitoring times were faster for words than for nonwords. These findings indicate that an orthographic representation of words is engaged in phoneme monitoring.
... (The concept of these two levels of processing has, of course, a long history in the field. For previous similar proposals regarding levels of processing in spoken word recognition, see Cutler & Norris, 1979;Foss & Blank, 1980;McClelland & Elman, 1986;Norris, 1994;Radeau, Morais, & Segui, 1995;Slowiazcek & Hamburger, 1992.) In particular, Vitevitch and Luce (1998) suggested that facilitatory effects of probabilistic phonotactics might reflect differences among activation levels of sublexical units, whereas effects of similarity neighborhoods may arise from competition among lexical representations. ...
... The results of a series of experiments using several different tasks and types of stimuli are accounted for by an adaptive resonance framework for spoken word recognition that embodies two levels of representation-a lexical level and a sublexical level. The hypothesis of two levels of representation with dissociable and distinct effects on processing reveals, in part, the complexity of the recognition process: Predicting processing of spoken words involves simultaneous consideration of the nature of the task used to interrogate the recognition process, the level of representation that dominates the response (Cutler & Norris, 1979;Foss & Blank, 1980), and the probabilistic phonotactics and similarity neighborhood structure of the spoken stimulus. ...
Recent work (Vitevitch & Luce, 1998) investigating the role of phonotactic information in spoken word recognition suggests the operation of two levels of representation, each having distinctly different consequences for processing. The lexical level is marked by competitive effects associated with similarity neighborhood activation, whereas increased probabilities of segments and sequences of segments facilitate processing at the sublexical level. We investigated the two proposed levels in six experiments using monosyllabic and specially constructed bisyllabic words and nonwords. The results of these studies provide further support for the hypothesis that the processing of spoken stimuli is a function of both facilitatory effects associated with increased phonotactic probabilities and competitive effects associated with the activation of similarity neighborhoods. We interpret these findings in the context of Grossberg, Boardman, and Cohen's (1997) adaptive resonance theory of speech perception.
... Segmental nodes may then in turn share then-activation advantages with the lexical nodes to which they are connected. Or, alternatively, responses may be driven off of either the lexical or the segmental levels, depending on current activation values at either level (see Cutler & Norris, 1979). If segmental activation dominates early in processing, the advantage afforded by lexical embeddedness may result-at least early on-in faster processing for carrier words with initially embedded items. ...
A large number of multisyllabic words contain syllables that are themselves words. Previous research using cross-modal priming and word-spotting tasks suggests that embedded words may be activated when the carrier word is heard. To determine the effects of an embedded word on processing of the larger word, processing times for matched pairs of bisyllabic words were examined to contrast the effects of the presence or absence of embedded words in both 1st- and 2nd-syllable positions. Results from auditory lexical decision and single-word shadowing demonstrate that the presence of an embedded word in the 1st-syllable position speeds processing times for the carrier word. The presence of an embedded word in the 2nd syllable has no demonstrable effect.
... Two models that are distinguished by whether they allow interaction between the lexical and phonemic level are the TRACE model and the autonomous race model (Cutler, Mehler, Norris, & Segui, 1987;Cutler & Norris, 1979). In the TRACE model, there are three levels: the feature level, the phoneme level, and the lexical level. ...
A series of experiments was conducted to determine whether the effects of lexical status on phonetic categorization were influenced by stimulus naturalness (replicating M. W. Burton, S. R. Baum, & S. E. Blumstein, 1989, who manipulated the intrinsic properties of the stimuli) and by stimulus quality (presenting the stimuli in white noise). The experiments compared continua varying in voice onset time (VOT) only to continua covarying VOT and amplitude of the burst and aspiration noise in no-noise and noise conditions. Results overall showed that the emergence of a lexical effect was influenced by stimulus quality but not by stimulus naturalness. Contrary to previous findings, significant lexical effects failed to emerge in the slower reaction time ranges. These results suggest that stimulus quality contributes to lexical effects on phonetic categorization, whereas stimulus naturalness does not.
... For instance, most abstractionist models do not allow top-down feedback, i.e., Cohort, FLMP (Oden & Massaro, 1978;Massaro & Oden, 1980). Other models like RACE (Cutler & Norris, 1979;Cutler, Mehler, Norris, & Segui, 1987) and Shortlist -both also autonomous-propose parallel and independent phonemic and lexical processing routes, whilst a minority implements top-down feedback directly (TRACE). Still others like Merge (Norris, McQueen, & Cutler, 2000) also have two parallel and independent processing routes like RACE, from prelexical processing units to lexical units, and from prelexical processing units to phoneme decision nodes, but also feedback from the lexical level to phoneme decision nodes (but not to prelexical acoustic processing nodes). ...
Chilean Spanish is special in that it displays particularly high degrees of lenition and elision of [β̞�], [ð̞]� and [ɣ̞] (Pérez, 2007). Interestingly, Chilean Spanish listeners can recover elided units effortlessly, which challenges the assumptions of some lexical access models, such as strong bottom-up abstractionist models (Mitterer & Ernestus, 2006). This proposal reports on a series of perception experiments in which synthetic continua from full approximants to elided variants were presented in several informational conditions. Results showed that increasing the amount of acoustic information and the number of semantic cues had a significant effect on listeners' responses, enabling lexical effects and minimizing phonological recovery. Moreover, these effects were different for the three consonants being tested, probably due to existing links between production and perception. These findings are discussed in light of previous research on lexical effects and recovery, and lexical access models in general.
... A dual-representation model of phonology is also consistent with several strands of thinking in psycholinguistics. For example, Cutler and Norris's (1979) dual-route model of phoneme monitoring (as implemented in Norris 1994) holds that phonemes may be detected by a phonetic route, in a speech mode of listening, or via a lexical route where the presence of the phoneme is deduced from the fact that a word containing the phoneme has just been detected. They identified a number factors that influence which of these two routes will be fastest. ...
... In a recent study with the tone-monitoring technique, Lobina, Demestre, and García-Albea (2018) point out that the data reported in Abrams and Bever (1969) and Bever and Hurtig (1975) may not be as robust as commonly thought, as some important factors were not considered at the time, and as a result, the experimental manipulation then used may have been confounded. Monitoring tasks, in general, exhibit a tendency of RTs to decrease across a sentence (Cutler & Norris, 1979), a factor that was not controlled for by Bever et al. In fact, the tone both Abrams and Bever (1969) and Bever and Hurtig (1975) placed at the end of clauses was typically the first tone in a series, and their data clearly show a decrease in RTs from the first to the last tone position. ...
Monitoring tasks have long been employed in psycholinguistics, and the end-of-clause effect is possibly the better-known result of using this technique in the study of parsing. Recent results with the tone-monitoring task suggest that tone position modulates cognitive load, as reflected in reaction times (RTs): the earlier the tone appears in a sentence, the longer the RTs. In this study, we show that verb position is also an important factor. In particular, changing the time/location at which verb–noun(s) dependencies are computed during the processing of a sentence has a clear effect on cognitive load and, as a result, on the resources that can be devoted to monitoring and responding to a tone. This study is based on two pieces of evidence. We first report the acceptability ratings of six word orders in Spanish and then present monitoring data with three of these different word orders. Our results suggest that RTs tend to be longer if the verb is yet to be processed, pointing to the centrality of a sentence’s main verb in parsing in general.
... Ce que nous percevons consciemment est le résultat d'une construction cognitive, influencée en partie non négligeable par nos connaissances et nos croyances ; l'hypothèse essentielle de la thèse de la modularité est qu'il y a une priorité aux informations sensorielles pour ce qui concerne les premières étapes de traitement, et que celles-ci sont peu affectées par nos connaissances ou croyances. On peut rendre compte des résultats de Warren et de Ganong (cf supra) avec un modèle de traitement non interactif en supposant que le sujet fonde sa réponse, non seulement sur une représentation phonémique extraite du signal, mais aussi sur une représentation post-lexicale après que le mot porteur ait été reconnu (Foss & Blank, 1980;Cutler & Norris, 1979;Cutler, Mehler, Norris, & Segui, 1987). Finalement, la distinction entre interactif et modulaire porte sur la question de savoir si un niveau supérieur (p.ex. ...
Nous présentons une série d'expériences qui montrent que le système perceptif reconstruit automatiquement la structure syllabique lors de la perception de la parole. Nos résultats rejettent la notion selon laquelle les syllabes sont des atomes de la perception de la parole, mais montrent plutôt que la structure syllabique agit comme un cadre d'organisation pour le décodage du signal de parole.
... In questo caso, ai partecipanti viene chiesto di decidere se un determinato stimolo acustico contiene il segmento [b] oppure no premendo sui tasti "b" o "no b" in seguito alla presentazione di ogni stimolo. Questo paradigma risulta superiore rispetto al tradizionale riconoscimento fonemico (phoneme detection; Cutler e Norris, 1979;Cutler et al., 1987), nel quale ai partecipanti viene chiesto di premere un tasto solamente nel caso in cui il suono bersaglio venga riconosciuto. Nel caso in cui il partecipante non prema alcun tasto, non è chiaro se ciò avvenga perché il parlante ritiene il suono assente, oppure perché semplicemente ha avuto un tempo di reazione troppo lungo e non è riuscito a premere il tasto in tempo. ...
Linguistic perception is conditioned by phonology. In this study, this claim is tested on empirical data coming from a set of Emilia dialects (Italy), which are reported to show regressive voicing assimilation (RVA). The hypothesis is that, in the case RVA is part of the phonological competence of the speakers, consonant clusters whose segments display opposite voicing specification are misperceived as showing the same specification. The analysis of the production results show that RVA is systematic, although partial (de)voicing can be found too and sometimes the process of RVA is not applied, presumably under the influence of Standard Italian. Similarly, RVA is shown to variably constrain perception. Taken together, the production and perception data suggests that RVA, rather than a fully systematic phonological process, should be considered a phonetic implementation process applying at the phonetics-phonology interface.
... In autonomous models (e.g., Cutler & Norris, 1979;Norris, McQueen, & Cutler, 2000), on the other hand, phonemic activation itself is uninfluenced by linguistic knowledge. In the Merge model (Norris et al., 2000), the Ganong effect is attributed to a combination (or merging) of the effects of phonemic activation and lexical activation in the service of making the phonemic decisions required in categorization experiments. ...
Listeners tend to categorize an ambiguous speech sound so that it forms a word with its context (Ganong, 1980). This effect could reflect feedback from the lexicon to phonemic activation (McClelland & Elman, 1986), or the operation of a task-specific phonemic decision system (Norris, McQueen, & Cutler, 2000). Because the former account involves feedback between lexical and phonemic levels, it predicts that the lexicon’s influence on phonemic decisions should be delayed and should gradually increase in strength. Previous response time experiments have not delivered a clear verdict as to whether this is the case, however. In 2 experiments, listeners’ eye movements were tracked as they categorized phonemes using visually displayed response options. Lexically relevant information in the signal, the timing of which was confirmed by separate gating experiments, immediately increased eye movements toward the lexically supported response. This effect on eye movements then diminished over the course of the trial rather than continuing to increase. These results challenge the lexical feedback account. The present work also introduces a novel method for analyzing data from ‘visual-world’ type tasks, designed to assess when an experimental manipulation influences the probability of an eye movement toward the target.
... However, another class of models explains lexical effects without lexical feedback. In an autonomous model such as Race (Cutler & Norris, 1979), phoneme identification is performed in parallel along two different routes, phonemic or lexical. Phoneme identification along the phonemic route is based on phonetic information, whereas the lexical route is only available after word recognition since it relies upon phonological descriptions stored in the lexicon. ...
Lexical effects on speech perception are not very reliable and they have been shown to depend on various factors among which word length. In the current models of phonemic decision, lexical effects are conceived as arising from top-down processing, with or without feedback, depending on the model. Lexical effects tend to be stronger in longer words, which can be ascribed to an increase in the amount of lexical evidence. The present study was aimed at collecting further evidence on this point. The existence of lexical effects was confirmed in a series of two experiments on voicing identification in French initial stops. The effects were present for stops in monosyllables and polysyllables whereas they were almost absent in bisyllables. We tentatively explain the U-shaped relationship between lexical evidence and phonemic identification by two different mechanisms which would be both weakly 162 Willy Sernicles, Renaud Beeckmans & Monique Radeau effective with moderate amounts of lexical evidence (in bisyllables). With fairly large amounts of lexical evidence (in polysyllables) the lexical effect would be due to the fairly complex top-down processes postulated in the literature. With low amounts of lexical evidence (in monosyllables), a much simpler mechanism based on a re-analysis of the acoustic input would be at work.
... The second is the theoretical assumption of interaction (lexical-sublexical feedback, which we discuss in detail in Section 4). (Norris, 1994;Norris, McQueen, & Cutler, 1995) is a fundamentalist simulation model that combines aspects of autonomous, feedforward models like Race (Cutler & Norris, 1979) and Cohort II with the competition dynamics of TRACE. A primary motivation in the development of this model was to keep positive characteristics of TRACE (e.g., competition dynamics) while avoiding weaknesses (e.g., the large number of nodes and connections due to reduplication of nodes over time). ...
... Note that the originally posited route from the speech signal to phonemic perception units should be retained so that the production-perception asymmetry would be explained. If the latter route (i.e., from the speech signal to the phonemic perception units) is also part of an 'indirect' route for lexical access, leading from the speech signal to the lexicon through phonemic perception units, the resulting model would have dual routes for lexical access, as in the Race model (Cutler & Norris, 1979, cited by Norris et al., 2000. 20 ...
Phonemic perception exhibits coarticulation sensitivity, phonotactic sensitivity and lexical sensitivity. Three kinds of models of speech perception are found in the literature, which embody different answers to the question of how the three kinds of sensitivity are related to each other: two-step models, one-step models and lexicalist models.
In two-step models (Church, 1987), phonemes are first extracted, and phonotactic repairs are subsequently made on the obtained phoneme string; both phonemic categorization and phonotactic repair are sublexical, and coarticulation sensitivity should only affect initial (pre-phonotactic) phonemic categorization.
In one-step models (Dehaene-Lambertz et al., 2000; Dupoux et al., 2011; Mehler et al., 1990), phonemic categorization and phonotactic repair are sublexical and simultaneous; phonotactic repairs themselves depend on coarticulation cues. Such models can be implemented in two different versions: suprasegmental matching, according to which a speech signal is matched against phonotactics-respecting suprasegmental units (such as syllables), rather than phonemes, and slot filling, according to which a speech signal is matched against phonemes as fillers for slots in phonotactics-respecting suprasegmental units.
In lexicalist models (Cutler et al., 2009; McClelland & Elman, 1986), coarticulation sensitivity and/or phonotactic sensitivity reduce to lexical sensitivity. McClelland & Elman (1986) claim a lexicalist reduction of phonotactic sensitivity; Cutler et al.'s (2009) make a claim implying lexicalist reductions both of phonotactic sensitivity and of coarticulation sensitivity.
This thesis attempts to distinguish among those models. Since different perceptual processes are assumed in these three models (whether sublexical units are perceived, or how many stages are involved in perceptual processing), our understanding of how speech perception works crucially depends on the relative superiority of those three kinds of models.
Based on the results available in the past literature on the one hand, and on the results of perceptual experiments with Japanese listeners testing their coarticulation sensitivity in different settings on the other, this thesis argues for the superiority of the slot filling version of one-step models over the others. According to this conclusion, phonemic parsing (categorization) and phonotactic parsing (repair) are separate but parallel sublexical processes.
... Early models argued that this competition can be implemented without any interaction between words. In these models, as time passes and more information becomes available, previously compatible candidates become incompatible and thus fall out of competition (Cutler & Norris, 1979;Marslen-Wilson, 1987). However, later evidence for some form of interference between active lexical candidates suggested that this kind of race process is not sufficient for describing how words are recognized -there is an additional need for some form of interaction or inhibition among words. ...
Language learning is generally described as a problem of acquiring new information (e.g., new words). However, equally important are changes in how the system processes known information. For example, a wealth of studies has suggested dramatic changes over development in how efficiently children recognize familiar words, but it is unknown what kind of experience-dependent mechanisms of plasticity give rise to such changes in real-time processing. We examined the plasticity of the language processing system by testing whether a fundamental aspect of spoken word recognition, lexical interference, can be altered by experience. Adult participants were trained on a set of familiar words over a series of 4 tasks. In the high-competition (HC) condition, tasks were designed to encourage coactivation of similar words (e.g., net and neck) and to require listeners to resolve this competition. Tasks were similar in the low-competition (LC) condition, but did not enhance this competition. Immediately after training, interlexical interference was tested using a visual world paradigm task. Participants in the HC group resolved interference to a fuller degree than those in the LC group, demonstrating that experience can shape the way competition between words is resolved. TRACE simulations showed that the observed late differences in the pattern of interference resolution can be attributed to differences in the strength of lexical inhibition. These findings inform cognitive models in many domains that involve competition/interference processes, and suggest an experience-dependent mechanism of plasticity that may underlie longer term changes in processing efficiency associated with both typical and atypical development. (PsycINFO Database Record
... Thus, under the autonomous view, the observed biases reflect the integration of multiple information sources, but, crucially, this integration does not affect lower level phonetic processing itself (see e.g., Norris, McQueen, & Cutler, 2000;Mc-Queen, Jesse, & Norris, 2009). A succession of autonomous models has been proposed in the literature, including Race (Cutler & Norris, 1979;Cutler, Mehler, Norris, & Segui, 1987), Shortlist (Norris, 1994), Merge (Norris et al., 2000), and Shortlist's Bayesian implementation (Norris & McQueen, 2008). Although each varies in its details, none allows for higher level modulation of phonetic processing through feedback (McQueen et al., 2006; see also Fuzzy Logical Model of Speech Perception; Massaro, 1989). ...
Although much evidence suggests that the identification of phonetically ambiguous target words can be biased by preceding sentential context, interactive and autonomous models of speech perception disagree as to the mechanism by which higher level information affects subjects' responses. Some have suggested that the time course of context effects is incompatible with interactive models (e.g., TRACE). Two experiments examine this issue. In Experiment 1, subjects heard noun- and verb-biasing sentence contexts (e.g., Valerie hated the . . . vs. Brett hated to . . .), followed by stimuli from 2 voice-onset time continua: bay-pay (noun-verb) versus buy-pie (verb-noun). Consistent with prior research, identification of phonetically ambiguous targets was biased by the preceding context, and the size of this bias diminished in slower compared with faster responses. In Experiment 2, tokens from the same continua were embedded among filler target words beginning with /b/ or /p/ to elicit phonemically driven identification decisions and discourage word-level strategies. Results again revealed contextually biased responding, but this bias was as strong in slow as in fast responses. Together, these results suggest that phoneme identification decisions reflect robust, lasting top-down effects of lexical feedback on prelexical representations, as predicted by interactive models of speech perception. (PsycINFO Database Record
... The absence of interaction between the phonological information and the corresponding orthographic information is inherent to autonomous models such as RACE or MERGE (Cutler & Norris, 1979;Norris, McQueen, & Cutler, 2000). But, strangely, the role of orthography is not mentioned either in a highly interactive model such as TRACE (McClelland & Elman, 1986), although this could easily accommodate the impact of orthography on spoken word recognition by means of interactions between the representations activated at different levels (i.e., features, phonemes, and words). ...
http://dx.doi.org/10.5007/2175-8026.2012n63p161
The levels-of-processing approach to speech processing (cf. Kolinsky, 1998) distinguishes three levels, from bottom to top: perception, recognition (which involves activation of stored knowledge) and formal explicit analysis or comparison (which belongs to metalinguistic ability), and assumes that only the former is immune to literacy-dependent knowledge. in this contribution, we first briefly review the main ideas and evidence supporting the role of learning to read in the alphabetic system in the development of conscious representations of phonemes, and we contrast conscious and unconscious representations of phonemes. Then, we examine in detail recent compelling behavioral and neuroscientific evidence for the involvement of orthographic representation in the recognition of spoken words. We conclude by arguing that there is a strong need of theoretical re-elaboration of the models of speech recognition, which typically have ignored the influence of reading acquisition.
... On this view, if top-down information is integrated directly with sensory information, an organism ipso facto loses the possibility of veridical perception, as there is no distinction between information in the environment and information in the organism. Autonomous models account for lexical effects on sublexical tasks by proposing parallel, competing lexical and sublexical routes (as in the Race model; Cutler & Norris, 1979), or that the locus of sublexical decisions is, counterintuitively, post-lexical. In the Merge model (Norris et al., 2000), for example, there are two banks of phoneme units. ...
Spoken word recognition is a distinct subsystem providing the interface among low-level perception and cognitive processes of retrieval, parsing, and interpretation. The narrowest conception of the process of recognizing a spoken word is that it starts from a string of phonemes, establishes how these phonemes should be grouped to form words, and passes these words onto the next level of processing. Some theories, though, take a broader view and blur the distinctions among speech perception, spoken word recognition, and sentence processing. The broader view of spoken word recognition has empirical and theoretical motivations. One consideration is that by assuming that the input to spoken word recognition is a string of abstract, phonemic category labels, one implicitly assumes that the nonphonemic variability carried on the speech signal is not relevant for spoken word recognition and higher levels of processing. However, if this variability and detail is not random but is lawfully related to linguistic categories, the simplifying assumption that the output of speech perception is a string of phonemes may actually be a complicating assumption.
... However, autonomous models, which do not allow top-down processes (an effect of word-knowledge on phoneme perception is one example of such a process), have had some success in accounting for such findings in other ways. One autonomous model is the race model of Cutler and Norris (1979) and its descendants (e.g., Norris, McQueen, & Cutler, 2000). The model has two routes that operate in parallel. ...
To produce and comprehend words and sentences, people use their knowledge of language structure; their knowledge of the situation they are in, including the previous discourse and the local situation; and their cognitive abilities, including memory, attention, and motor control. In this chapter, we explore how competent adult language users bring such knowledge and abilities to bear on the tasks of comprehending spoken and written language and producing spoken language. We emphasize experimental data collected using the tools of cognitive psychology, touching only briefly on language development, disordered language, and the neural basis of language. We also review some of the major theoretical controversies that have occupied the field of psycholinguistics, including the role that linguistic analyses of language structure should play and the debate between modular and interactive views. We also present some of the theoretical positions that have proven successful in guiding our understanding of language processing. We conclude by discussing the need to integrate studies of language comprehension and language production and pointing to emerging research topics.Keywords:psycholinguistics;auditory word recognition;reading;lexical access;sentence comprehension;word production;sentence production
... However, autonomous models, which do not allow top-down processes (an effect of word-knowledge on phoneme perception is one example of such a process), have had some success in accounting for such findings in other ways. One autonomous model is the race model of Cutler and Norris (1979) and its descendants (e.g., Norris, McQueen, & Cutler, 2000). The model has two routes that operate in parallel. ...
To produce and comprehend words and sentences, people use their knowledge of
language structure, their knowledge of the situation they are in, including the previous discourse
and the local situation, and their cognitive abilities, including memory, attention, and motor
control. In this chapter, we explore how competent adult language users bring such knowledge
and abilities to bear on the tasks of comprehending spoken and written language and producing
spoken language. We emphasize experimental data collected using the tools of cognitive
psychology, touching only briefly on language development, disordered language, and the neural
basis of language. We also review some of the major theoretical controversies that have occupied
the field of psycholinguistics, including the role that linguistic analyses of language structure
should play and the debate between modular and interactive views. We also present some of the
theoretical positions that have proven successful in guiding our understanding of language
processing. We conclude by discussing the need to integrate studies of language comprehension
and language production and pointing to emerging research topics.
... The locus of the observed e ects is an issue of contention between the proponents of autonomous and interactive models of speech processing. The \race" model is an autonomous processing model that was proposed by C u tler and Norris (1979) to account for a diverse set of ndings on phoneme monitoring using words and nonwords. Cutler and Norris argued that a dual-outlet model, in which a prelexical phonetic route competes (i.e., races, hence the model's name) with a lexically derived phonemic route, provided the best account of a host of seemingly contradictory ndings (reviewed therein). ...
... Logan 1985 andStrange 1995) provide evidence for the languagespecificity of the mapping from perceptual information onto phonological form. In psycholinguistics, this mapping is referred to as prelexical speech perception, and complemented by word recognition, where the phonological form is mapped onto a form in the lexicon, see Cutler and Norris (1979), Cutler et al. (1987), and McQueen and Cutler (1997). Boersma (2006a et seq.) employed such a two-staged speech recognition in the BiPhon model (but see already Boersma 1998), where the mapping between forms is formalized with OT constraints and occurs in parallel. ...
In this paper, I argue that a regular diachronic sound change is the result of a different interpretation of the same auditory information, as put forward by Ohala (1981 et seq.). Whereas Ohala describes such an account of sound change as purely phonetic, I show that it involves phonological knowledge, namely the language-specific use of auditory cues and their mapping onto language-specific phonological categories. Two diachronic developments of retroflex segments, namely retroflexion of rhotic plus coronal consonant sequences in Norwegian and retroflexion of labialised coronal obstruents in Minto-Nenana, illustrate these assumptions. For both, the differences across generations are modelled in Optimality Theory with the help of language-specific cue constraints in a perception grammar (following Boersma 1997 et seq.). This approach is shown to be superior to the descriptive approach of cue re-association proposed by Ohala because it provides a formal account that includes differences in cue weighting (especially the disregard of cues that became unreliable) and differences in emergent phonological categories.
... 53 Proponents of interactive models have pointed out that lexical feedback is in line with research showing that lexical knowledge allows listeners to quickly adapt to speakers with unfamiliar pronunciation, 54 but proponents of feed-forward models have countered that feedback for perceptual learning is different from online feedback as is implemented in TRACE. 55,56 Shortlist Shortlist 50 was developed in response to the criticism of duplication and lexical feedback in TRACE, and combines aspects of feed-forward models, such as the phoneme decision model Race 57 and Cohort II, with the competition mechanism of TRACE. The duplication of the entire network for each input feature in TRACE is avoided by implementing Shortlist as a two-stage model in which the generation of lexical candidates and the competition process are separated (see Figure 3). ...
All words of the languages we know are stored in the mental lexicon. Psycholinguistic models describe in which format lexical knowledge is stored and how it is accessed when needed for language use. The present article summarizes key findings in spoken-word recognition by humans and describes how models of spoken-word recognition account for them. Although current models of spoken-word recognition differ considerably in the details of implementation, there is general consensus among them on at least three aspects: multiple word candidates are activated in parallel as a word is being heard, activation of word candidates varies with the degree of match between the speech signal and stored lexical representations, and activated candidate words compete for recognition. No consensus has been reached on other aspects such as the flow of information between different processing levels, and the format of stored prelexical and lexical representations. WIREs Cogn Sci 2012, 3:387-401. doi: 10.1002/wcs.1178 For further resources related to this article, please visit the WIREs website.
Copyright © 2012 John Wiley & Sons, Ltd.
... On this view, if top-down information is integrated directly with sensory information, an organism ipso facto loses the possibility of veridical perception, as there is no distinction between information in the environment and information in the organism. Autonomous models account for lexical effects on sublexical tasks by proposing parallel, competing lexical and sublexical routes (as in the Race model; Cutler & Norris, 1979 ), or that the locus of sublexical decisions is, counterintuitively , post-lexical. In the Merge model (Norris et al., 2000), for example, there are two banks of phoneme units. ...
Spoken‐word recognition is an efficient and generally error‐free process that occurs under a variety of speaking and listening conditions. The talk will focus on the mapping process between the speech signal and access of form and meaning. The nature of the representation that supports spoken‐word recognition will be discussed with a focus on the consequence of ambiguity and mismatching information. Research has been conducted in the past few years suggesting that activation of lexical representations is accomplished via feature mapping. It is argued that this architecture permits lexical activation given incomplete or erroneous input. Phonological variation and some recent work concerning representation and processing of common variants will also be discussed.
... This phoneme-monitoring task is a speeded-response task as processing time is limited (Connine & Titone, 1996). Despite its name, this task reflects lexical processing when meaningful sentences are monitored (Cutler, Mehler, Norris, & Segui, 1987;Cutler & Norris, 1979;Mirman, McClelland, Holt, & Magnuson, 2008). Phoneme monitoring can thus be used to measure the effect of context on the speed of lexical access. ...
Abstract Many older listeners report difficulties in understanding speech in noisy situations. Working memory and other cognitive skills may modulate older listeners' ability to use context information to alleviate the effects of noise on spoken-word recognition. In the present study, we investigated whether verbal working memory predicts older adults' ability to immediately use context information in the recognition of words embedded in sentences, presented in different listening conditions. In a phoneme-monitoring task, older adults were asked to detect as fast and as accurately as possible target phonemes in sentences spoken by a target speaker. Target speech was presented without noise, with fluctuating speech-shaped noise, or with competing speech from a single distractor speaker. The gradient measure of contextual probability (derived from a separate offline rating study) affected the speed of recognition. Contextual facilitation was modulated by older listeners' verbal working memory (measured with a backward digit span task) and age across listening conditions. Working memory and age, as well as hearing loss, were also the most consistent predictors of overall listening performance. Older listeners' immediate benefit from context in spoken-word recognition thus relates to their ability to keep and update a semantic representation of the sentence content in working memory.
... Frauenfelder et al. 's data, likewise, are problematic for an interactive model, but not for an autonomous model, because autonomous theory predicts that the perception of nonwords should be insulated from lexical processing. The autonomous race model (Cutler & Norris, 1979;, as an example, predicts the pattern of RT range results that McQueen (1991) found for wordfinal categorisation. In this model, lexical and pre-lexical phoneme identification procedures are considered to race; the procedure which more rapidly produces an output on a particular trial will be responsible for the phonetic decision on that trial. ...
... Some researchers have argued that some words are decomposed while others are not. Wurm (1997) proposed a dual-route model based on the idea of parallel, competing processes [cf. the Race model of Cutler and Norris (1979)]. In his model, morphologically complex words are processed simultaneously as fullforms and as analyzed constituent morphemes. ...
This study compared models of auditory word recognition as they relate to the processing of polymorphemic pseudowords. Semantic transparency ratings were obtained in a preliminary rating study. The effects of morphological structure, semantic transparency, prefix likelihood, and morphemic frequency measures were examined in a lexical decision experiment. Reaction times and errors were greater for pseudowords carrying a genuine prefix, and this effect was largest for pseudowords that also carried a genuine root. While results were grossly similar for bound and free root types, there were also some important differences. Regression analyses provided additional support for decompositional models: semantic transparency, prefix likelihood, prefix frequency, and root frequency all affected pseudoword rejection times. The results are most compatible with a modification of Taft's (1994) interactive-activation model or a dual-route model.
... By contrast, in order to capture context effects, bottom-up models necessarily allow two routes which can influence phonemic judgments—via phonemic and lexical representations. Various models, both non-connectionist (Cutler & Norris, 1979) and connectionist (Norris, McQueen & Cutler, in press) have been proposed in opposition to TRACE. These models exploit two routes, and hence allow for the possibility of " attentional " switching between them. ...
... Y et, since any phonem e or fragment detection task requires intentional matching between the target and the carrier, it is also possible that at least part of the syllabic eect results from a syllable-struc tured post-lexical phonological representation. For example, an alternative account of the attentional allocation results reported in the previous section is that, as in other phoneme detection tasks, listeners base their responses on either lexical or pre-lexical inf ormation, depending on the experimental conditions (thè`dual code hypothesis' '; Cutler & Norris, 1979) . A s regards the original fragment detection task, both Mehler et al. (1981) and Cutler et al. (1986) assumed that it taps pre-lexical processing. ...
In recognising spoken words, the retrieval of a unique mental representation from the speech input involves the exceedingly difficult task of locating word boundaries in a quasi-continuous stimulus and of finding the single representation that corresponds to highly variable acoustical forms. Many cognitive psycholinguists have proposed that these segmentation and categorisation problems are easier to solve at the sublexical than at the word level: Some sublexical representation would mediate the mapping between the acoustic signal and the mental lexicon. Accordingly, much effort has gone into disclosing a hypothesised universal perceptual building block, for example the syllable. More recent advances in speech processing research indicate, however, that speakers of different languages process speech by relying on units or segmentation strategies that are appropriate to the phonological properties of their maternal tongue. Recent data on this topic will be reviewed, with special emphasis on the stage-processing approach of the experimental situations and phenomena reported. For example, the syllabic effects observed in fragment detection and the phenomenon of blending dichotically presented words are discussed. It will be argued that although there is a strong case for language specificity in listeners' intuitions about the phonological structure of their language as well as in word recognition, less evidence is available regarding the early perceptual stages.
... Ce que nous percevons consciemment est le résultat d'une construction cognitive, influencée en partie non négligeable par nos connaissances et nos croyances ; l'hypothèse essentielle de la thèse de la modularité est qu'il y a une priorité aux informations sensorielles pour ce qui concerne les premières étapes de traitement, et que celles-ci sont peu affectées par nos connaissances ou croyances. On peut rendre compte des résultats de Warren et de Ganong (cf supra) avec un modèle de traitement non interactif en supposant que le sujet fonde sa réponse, non seulement sur une représentation phonémique extraite du signal, mais aussi sur une représentation post-lexicale après que le mot porteur ait été reconnu (Foss & Blank, 1980;Cutler & Norris, 1979;Cutler, Mehler, Norris, & Segui, 1987). Finalement, la distinction entre interactif et modulaire porte sur la question de savoir si un niveau supérieur (p.ex. ...
Psycholinguists define spoken word recognition (SWR) as, roughly, the processes intervening between speech perception and sentence processing, whereby a sequence of speech elements is mapped to a phonological wordform. After reviewing points of consensus and contention in SWR, we turn to the focus of this review: considering the limitations of theoretical views that implicitly assume an idealized (neurotypical, monolingual adult) and static perceiver. In contrast to this assumption, we review evidence that SWR is plastic throughout the life span and changes as a function of cognitive and sensory changes, modulated by the language(s) someone knows. In highlighting instances of plasticity at multiple timescales, we are confronted with the question of whether these effects reflect changes in content or in processes, and we consider the possibility that the two are inseparable. We close with a brief discussion of the challenges that plasticity poses for developing comprehensive theories of spoken language processing.
Expected final online publication date for the Annual Review of Linguistics, Volume 10 is January 2024. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.
Processing speech can be slow and effortful for children, especially in adverse listening conditions, such as the classroom. This can have detrimental effects on children’s academic achievement. We therefore asked whether primary school children’s speech processing could be made faster and less effortful via the presentation of visual speech cues (speaker’s facial movements), and whether any audio-visual benefit would be modulated by the presence of noise or by characteristics of individual participants. A phoneme monitoring task with concurrent pupillometry was used to measure 7- to 11-year-old children’s speech processing speed and effort, with and without visual cues, in both quiet and noise. Results demonstrated that visual cues to speech can facilitate children’s speech processing, but that these benefits may also be subject to variability according to children’s motivation. Children showed faster processing and reduced effort when visual cues were available, regardless of listening condition. However, examination of individual variability revealed that the reduction in effort was driven by the children who performed better on a measure of phoneme isolation (used to quantify how difficult they found the phoneme monitoring task).
The interaction of contextual, high-level linguistic knowledge and the listener’s attention to low-level phonetic details has been the subject of a large body of research in speech perception for several decades. In the current paper, I investigate this interaction by considering the specific phenomenon of word predictability and its role in modulating the listener’s attention to subphonemic details of the acoustic signal. In the first experiment, subjects are presented with a discrimination task in which target words are presented in either predictable or unpredictable sentential context and then repeated in isolation, being either acoustically identical or subtly different. The subjects more accurately discriminate contextually unpredictable words, suggesting more attention to the phonetic details of words in unpredictable contexts. In the second experiment, considering the predictions of exemplar theory, I test whether this perceptual bias could result in changes in production. In this experiment, in which subjects heard and repeated sentences, I find a significant effect of word predictability on how close the subjects’ productions were to the model’s, which suggests a role of predictability on phonetic accommodation. The results of these experiments contribute to our understanding of stored exemplars and suggest the influence of contextual predictability in sound change.
High-variability phonetic training is effective in the acquisition of foreign language sounds. Previous studies have largely focused on small sets of contrasts, and have not controlled for the quantity of prior or simultaneous exposure to new sounds. The current study examined the effectiveness of phonetic training in full-inventory foreign language consonant acquisition by listeners with no previous exposure to the language. Chinese adult listeners underwent an intensive training programme, bracketed by tests that measured both assimilation of foreign sounds to native categories, and foreign category identification rates and confusions. Very rapid learning was evident in the results, with initial misidentification rates halving by the time of the mid-test, and continuing to fall in subsequent training sessions. Changes as a result of training in perceptual assimilation together with improved identifications and reduced response dispersion suggest an expansion of listeners’ native categories to accommodate the foreign sounds and an incipient process of foreign language category formation.
Syllables are identified faster than features or phonemes, at any rate in initial position of the target-item (Savin & Bever, 1970). This demonstration has led many authors to suggest that the syllable is the basic segment in speech perception and that phonemes can only be derived from the analysis of the perceptually primary segment, namely, the syllable. Considerable disagreement remains as to how this observation ought to be interpreted. Savin & Bever considered, that the phoneme had linguistic rather than psychological reality. Their interpretation, however, came up against considerable criticism from several authors including McNeil & Lindig (1973), Healy & Cutting (1976), Foss & Swinney (1973).
We present an auditory presentation technique called segmented binaural presentation. The technique builds on the dichotic listening paradigm (Shankweiler & Studdert-Kennedy, 1967; Studdert-Kennedy & Shankweiler, 1970) and segmented lexical presentation (Libben, 2003; Betram, Kuperman, Baayen, & Hyönä, 2011). The technique allows the first part of a word to be presented to one ear and the second part of the word to be presented to the other ear. The experimenter may thus manipulate whether a stimulus is segmented in this binaural manner and, if it is segmented, the location of the binaural segmentation within the word. We discuss how the technique may be implemented on the Macintosh platform, using PsyScope and freely available software for audio file creation. We also report on a test implementation of the technique using suffixed and compound English words in a lexical decision task. Results suggest that the technique differentiates between segmentation that occurs within and between compound constituents.
The process of mapping acoustic-phonetic level input to a lexical representation is multi-faceted. Models of spoken word recognition provide a variety of processing architectures and make different assumption(s) regarding the unit(s) of representation used in the exchange of information from signal-to-word and the nature of information flow through the system. The current models provide a backdrop for a discussion of some of the advances and debates in the field. Some of the issues considered are: early versus delayed commitment to a lexical hypothesis, consequences of multiple activation, segmentation and lexical access, the processing and representation of phonological variants, and the role of attention in spoken word recognition.
Phoneme monitoring studies from 1969 to 1996 are reviewed and grouped in terms of issues that have been addressed with the task. These issues include the contribution of the lexicon to speech perception, processing complexity, attention, contribution of prosodic information, and the basic unit of speech perception. Within each issue, task demands and artifactual variables have been identified and highlighted.
Older listeners are more affected than younger listeners in their recognition of speech in adverse conditions, such as when they also hear a single-competing speaker. In the present study, we investigated with a speeded response task whether older listeners with various degrees of hearing loss benefit under such conditions from also seeing the speaker they intend to listen to. We also tested, at the same time, whether older adults need postperceptual processing to obtain an audiovisual benefit. When tested in a phoneme-monitoring task with single-talker noise present, older (and younger) listeners detected target phonemes more reliably and more rapidly in meaningful sentences uttered by the target speaker when they also saw the target speaker. This suggests that older adults processed audiovisual speech rapidly and efficiently enough to benefit already during spoken sentence processing. Audiovisual benefits for older adults were similar in size to those observed for younger adults in terms of response latencies, but smaller for detection accuracy. Older adults with more hearing loss showed larger audiovisual benefits. Attentional abilities predicted the size of audiovisual response time benefits in both age groups. Audiovisual benefits were found in both age groups when monitoring for the visually highly distinct phoneme /p/ and when monitoring for the visually less distinct phoneme /k/. Visual speech thus provides segmental information about the target phoneme, but also provides more global contextual information that helps both older and younger adults in this adverse listening situation.
Two experiments compared continuous and discontinuous models of word recognition. Participants heard prefixed words whose full-form and root uniqueness points (UPs) differed, in either a gating or lexical decision paradigm. Identification points and reaction times were analyzed using multiple regression. Full-form UPs predicted performance better than root UPs did. Full-form frequency measures had reliable facilitative relationships with performance while root frequency measures were not consistently significant. Prefix frequency had a reliable, inhibitory effect. Judged prefixedness, semantic transparency, and prefix likelihood were related to performance, alone or in interaction. The results provide evidence for both kinds of word recognition procedures. A model is proposed with two parallel recognition routines: a whole-word routine and a decompositional routine that considers only unbound roots that can combine with the prefix in question. A preliminary rating study provides stimulus values on several dimensions and can be used as a database by other researchers.
An empirical account is offered of some of the constants that infants and adults appear to use in processing speech-like stimuli. From investigations carried out in recent years, it seems that syllable-like sequences act as minimal accessing devices in speech processing. Ss are aware in real time of syllabic structure in words and respond differently to words with the same initial three phonemes if the segmental one is CV/... and the other CVC/.... Likewise, infants seem to be aware that a `good' syllable must have at least one alternation if it is composed of more than one phoneme. When the segment is only one phoneme long, its status is necessarily somewhere between that of the phoneme and the syllable. An important problem that arises with the syllable is that it is an unlikely device for speech acquisition. Indeed, there are a few thousand syllables and the attribution of a given token to a type is far from obvious. Even if physical invariants for syllables in contexts were to be found, the task facing the child still remains one of sorting thousands of types from many more tokens. Issues concerning acquisition versus stable performance will be addressed to further constrain possible models. In addition, I try to show that even though information processing models are useful tools for describing synchronic sections of organisms, the elements that can account for development will have to be uncovered in neighbouring branches.
Using gender decision and shadowing tasks, we compared recognition of French nouns with early or late uniqueness points (UP) that were articulated at three different rates. With gender decision, the medium rate (3.6 syllables (syll)/s), which is close to that used by Radeau, Mousty, and Bertelson (1989), gave rise to a comparable UP location effect. The effect increased at the slower rate (2.2 syll/s), but disappeared at the faster rate (5.6 syll/s). With shadowing, only the slow rate gave rise to a UP effect. A similar pattern of results was found using speech that was linearly compressed or expanded. Because the fast rate is close to that typical of conversational speech, the present results cast doubt on the relevance of the UP in the processing of fluent speech. The implications of rate effects for models of spoken word recognition are discussed.
This article reports two experiments which examined the utility of the phoneme monitoring technique for studying syntactic processing of sentences. In French, by using self-embedded relative clauses, it is possible to isolate and examine the effect of a syntactic cue while controlling the factors known to effect phoneme detection times. Monitoring within and after the relative clause led to significant differences in phoneme detection times for reversible subject and object relatives only after the clause boundary. These results demonstrate the sensitivity of the phoneme monitoring task to syntactic processing and are taken to reflect structural calculations of the underlying grammatical relations for the reversible object relatives. When lexical information was introduced with nonreversible relatives, there was no longer a difference between the detection times for subject and object relatives after the clause boundary. Thus, it appears that lexical information can be used in the attribution of underlying grammatical roles.
From previous research we know that prosodic features are perceptually effective in marking boundaries and that a suitable implementation of these features improves the quality of synthetic speech in terms of acceptability. It can further be assumed that listeners use the perceived prosodic information to compute the meaning of the input speech. This paper, therefore, investigates and determines whether a well-phrased utterance, (that is, an utterance with prosodic boundaries in appropriate positions and with appropriate realizations), is easier to comprehend than a poorly-phrased one. To measure this, we designed a method in which a kind of verification task is combined with a question-answering task ("monitoring for the answer"). The stimulus set consisted of structurally ambiguous sentences. The expectation was that when listeners hear a question followed by an appropriately phrased utterance, they will react more rapidly than when the question is followed by an utterance with neutral phrasing. Also, it was expected that in the latter situation reaction times (RTs) will be shorter than if an inappropriately phrased utterance is presented. The results confirmed the expectations: an appropriately phrased utterance always produced the fastest RTs.
ResearchGate has not been able to resolve any references for this publication.