Publications

Abstract

Large auditory-visual (AV) speech corpora are the grist of modern research in speech science, but no such corpus exists for Australian English. This is unfortunate, for speech science is the brains behind speech technology and applications such as text-to-speech (TTS) synthesis, automatic speech recognition (ASR), speaker recognition and forensic identification, talking heads, and hearing prostheses. Advances in these research areas in Australia require a large corpus of Australian English. Here the authors describe a blueprint for building the Big Australian Speech Corpus (the Big ASC), a corpus of over 1,100 speakers from urban and rural Australia, including speakers of non-indigenous, indigenous, ethnocultural, and disordered forms of Australian English, each of whom would be sampled on three occasions in a range of speech tasks designed by the researchers who would be using the corpus.

Abstract

English listeners largely disregard suprasegmental cues to stress in recognizing words. Evidence for this includes the demonstration of Fear et al. [J. Acoust. Soc. Am. 97, 1893–1904 (1995)] that cross-splicings are tolerated between stressed and unstressed full vowels (e.g., au- of autumn, automata). Dutch listeners, however, do exploit suprasegmental stress cues in recognizing native-language words. In this study, Dutch listeners were presented with English materials from the study of Fear et al. Acceptability ratings by these listeners revealed sensitivity to suprasegmental mismatch, in particular, in replacements of unstressed full vowels by higher-stressed vowels, thus evincing greater sensitivity to prosodic goodness than had been shown by the original native listener group.

Abstract

Two phoneme goodness rating experiments addressed the role of orthographic knowledge in the evaluation of speech sounds. Ratings for the best tokens of /s/ were higher in words spelled with S (e.g., bless) than in words where /s/ was spelled with C (e.g., voice). This difference did not appear for analogous nonwords for which every lexical neighbour had either S or C spelling (pless, floice). Models of phonemic processing incorporating obligatory influence of lexical information in phonemic processing cannot explain this dissociation; the data are consistent with models in which phonemic decisions are not subject to necessary top-down lexical influence.

Abstract

Three experiments, in which Japanese listeners detected Japanese words embedded in nonsense sequences, examined the perceptual consequences of vowel devoicing in that language. Since vowelless sequences disrupt speech segmentation [Norris et al. (1997). Cognit. Psychol. 34, 191– 243], devoicing is potentially problematic for perception. Words in initial position in nonsense sequences were detected more easily when followed by a sequence containing a vowel than by a vowelless segment (with or without further context), and vowelless segments that were potential devoicing environments were no easier than those not allowing devoicing. Thus asa, “morning,” was easier in asau or asazu than in all of asap, asapdo, asaf, or asafte, despite the fact that the /f/ in the latter two is a possible realization of fu, with devoiced [u]. Japanese listeners thus do not treat devoicing contexts as if they always contain vowels. Words in final position in nonsense sequences, however, produced a different pattern: here, preceding vowelless contexts allowing devoicing impeded word detection less strongly (so, sake was detected less accurately, but not less rapidly, in nyaksake—possibly arising from nyakusake—than in nyagusake). This is consistent with listeners treating consonant sequences as potential realizations of parts of existing lexical candidates wherever possible.

Abstract

Recognizing word boundaries in continuous speech requires detailed knowledge of the native language. In the first year of life, infants acquire considerable word segmentation abilities. Infants at this early stage in word segmentation rely to a large extent on the metrical pattern of their native language, at least in stress-based languages. In Dutch and English (both languages with a preferred trochaic stress pattern), segmentation of strong-weak words develops rapidly between 7 and 10 months of age. Nevertheless, trochaic languages contain not only strong-weak words but also words with a weak-strong stress pattern. In this article, we present electrophysiological evidence of the beginnings of weak-strong word segmentation in Dutch 10-month-olds. At this age, the ability to combine different cues for efficient word segmentation does not yet seem to be completely developed. We provide evidence that Dutch infants still largely rely on strong syllables, even for the segmentation of weak-strong words.

Abstract

Two artificial-language learning experiments directly compared English, French, and Dutch listeners’ use of suprasegmental cues for continuous-speech segmentation. In both experiments, listeners heard unbroken sequences of consonant-vowel syllables, composed of recurring three- and four-syllable “words.” These words were demarcated by(a) no cue other than transitional probabilities induced by their recurrence, (b) a consistent left-edge cue, or (c) a consistent right-edge cue. Experiment 1 examined a vowel lengthening cue. All three listener groups benefited from this cue in right-edge position; none benefited from it in left-edge position. Experiment 2 examined a pitch-movement cue. English listeners used this cue in left-edge position, French listeners used it in right-edge position, and Dutch listeners used it in both positions. These findings are interpreted as evidence of both language-universal and language-specific effects. Final lengthening is a language-universal effect expressing a more general (non-linguistic) mechanism. Pitch movement expresses prominence which has characteristically different placements across languages: typically at right edges in French, but at left edges in English and Dutch. Finally, stress realization in English versus Dutch encourages greater attention to suprasegmental variation by Dutch than by English listeners, allowing Dutch listeners to benefit from an informative pitch-movement cue even in an uncharacteristic position.

Abstract

In three experiments, the processing of lexical tone in Cantonese was examined. Cantonese listeners more often accepted a nonword as a word when the only difference between the nonword and the word was in tone, especially when the F0 onset difference between correct and erroneous tone was small. Same–different judgments by these listeners were also slower and less accurate when the only difference between two syllables was in tone, and this was true whether the F0 onset difference between the two tones was large or small. Listeners with no knowledge of Cantonese produced essentially the same same-different judgment pattern as that produced by the native listeners, suggesting that the results display the effects of simple perceptual processing rather than of linguistic knowledge. It is argued that the processing of lexical tone distinctions may be slowed, relative to the processing of segmental distinctions, and that, in speeded-response tasks, tone is thus more likely to be misprocessed than is segmental structure.

Abstract

Research on the exploitation of prosodic information in the recognition of spoken language is reviewed. The research falls into three main areas: the use of prosody in the recognition of spoken words, in which most attention has been paid to the question of whether the prosodic structure of a word plays a role in initial contact with stored lexical representations; the use of prosody in the computation of syntactic structure, in which the resolution of global and local ambiguities has formed the central focus; and the role of prosody in the processing of discourse structure, in which there has been a preponderance of work on the contribution of accentuation and deaccentuation to integration of concepts with an existing discourse model. The review reveals that in each area progress has been made towards new conceptions of prosody's role in processing, and in particular this has involved abandonment of previously held deterministic views of the relationship between prosodic structure and other aspects of linguistic structure

Abstract

Psycholinguists strive to construct a model of human language processing in general. But this does not imply that they should confine their research to universal aspects of linguistic structure, and avoid research on language-specific phenomena. First, even universal characteristics of language structure can only be accurately observed cross-linguistically. This point is illustrated here by research on the role of the syllable in spoken-word recognition, on the perceptual processing of vowels versus consonants, and on the contribution of phonetic assimilation phonemena to phoneme identification. In each case, it is only by looking at the pattern of effects across languages that it is possible to understand the general principle. Second, language-specific processing can certainly shed light on the universal model of language comprehension. This second point is illustrated by studies of the exploitation of vowel harmony in the lexical segmentation of Finnish, of the recognition of Dutch words with and without vowel epenthesis, and of the contribution of different kinds of lexical prosodic structure (tone, pitch accent, stress) to the initial activation of candidate words in lexical access. In each case, aspects of the universal processing model are revealed by analysis of these language-specific effects. In short, the study of spoken-language processing by human listeners requires cross-linguistic comparison.

Abstract

Words can be distinguished by segmental differences or by suprasegmental differences or both. Studies from English suggest that suprasegmentals play little role in human spoken-word recognition; English stress, however, is nearly always unambiguously coded in segmental structure (vowel quality); this relationship is less close in Dutch. The present study directly compared the effects of segmental and suprasegmental mispronunciation on word recognition in Dutch. There was a strong effect of suprasegmental mispronunciation, suggesting that Dutch listeners do exploit suprasegmental information in word recognition. Previous findings indicating the effects of mis-stressing for Dutch differ with stress position were replicated only when segmental change was involved, suggesting that this is an effect of segmental rather than suprasegmental processing.

Abstract

We propose that word recognition in continuous speech is subject to constraints on what may constitute a viable word of the language. This Possible-Word Constraint (PWC) reduces activation of candidate words if their recognition would imply word status for adjacent input which could not be a word - for instance, a single consonant. In two word-spotting experiments, listeners found it much harder to detectapple,for example, infapple(where [f] alone would be an impossible word), than invuffapple(wherevuffcould be a word of English). We demonstrate that the PWC can readily be implemented in a competition-based model of continuous speech recognition, as a constraint on the process of competition between candidate words; where a stretch of speech between a candidate word and a (known or likely) word boundary is not a possible word, activation of the candidate word is reduced. This implementation accurately simulates both the present results and data from a range of earlier studies of speech segmentation.

Abstract

Dutch and Spanish differ in how predictable the stress pattern is as a function of the segmental content: it is correlated with syllable weight in Dutch but not in Spanish. In the present study, two experiments were run to compare the abilities of Dutch and Spanish speakers to separately process segmental and stress information. It was predicted that the Spanish speakers would have more difficulty focusing on the segments and ignoring the stress pattern than the Dutch speakers. The task was a speeded classification task on CVCV syllables, with blocks of trials in which the stress pattern could vary versus blocks in which it was fixed. First, we found interference due to stress variability in both languages, suggesting that the processing of segmental information cannot be performed independently of stress. Second, the effect was larger for Spanish than for Dutch, suggesting that that the degree of interference from stress variation may be partially mitigated by the predictability of stress placement in the language.

Abstract

Finnish vowel harmony rules require that if the vowel in the first syllable of a word belongs to one of two vowel sets, then all subsequent vowels in that word must belong either to the same set or to a neutral set. A harmony mismatch between two syllables containing vowels from the opposing sets thus signals a likely word boundary. We report five experiments showing that Finnish listeners can exploit this information in an on-line speech segmentation task. Listeners found it easier to detect words likehymyat the end of the nonsense stringpuhymy(where there is a harmony mismatch between the first two syllables) than in the stringpyhymy(where there is no mismatch). There was no such effect, however, when the target words appeared at the beginning of the nonsense string (e.g.,hymypuvshymypy). Stronger harmony effects were found for targets containing front harmony vowels (e.g.,hymy) than for targets containing back harmony vowels (e.g.,paloinkypaloandkupalo). The same pattern of results appeared whether target position within the string was predictable or unpredictable. Harmony mismatch thus appears to provide a useful segmentation cue for the detection of word onsets in Finnish speech.

Abstract

There are currently a number of psycholinguistic models in which processing at a particular level of representation is characterized by the generation of multiple outputs, with resolution involving the use of information from higher levels of processing. Surprisingly, models with this architecture have been characterized as autonomous within the domain of word recognition and as interactive within the domain of sentence processing. We suggest that the apparent internal confusion is not, as might be assumed, due to fundamental differences between lexical and syntactic processing. Rather, we believe that the labels in each domain were chosen in order to obtain maximal contrast between a new model and the model or models that were currently dominating the field.

Abstract

There are currently a number of psycholinguistic models in which processing at a particular level of representation is characterized by the generation of multiple outputs, with resolution - but not generation - involving the use of information from higher levels of processing. Surprisingly, models with this architecture have been characterized as autonomous within the domain of word recognition but as interactive within the domain of sentence processing. We suggest that the apparent confusion is not, as might be assumed, due to fundamental differences between lexical and syntactic processing. Rather, we believe that the labels in each domain were chosen in order to obtain maximal contrast between a new model and the model or models that were currently dominating the field. The contradiction serves to highlight the inadequacy of a simple autonomy/interaction dichotomy for characterizing the architectures of current processing models.

Abstract

Two lexical decision experiments in Cantonese are described in which the recognition of spoken target words as a function of phonological similarity to a preceding prime is investigated. Phonological similaritv in first syllables produced inhibition, while similarity in second syllables led to facilitation. Differences between syllables in tonal and segmental structure had generally similar effects.

Abstract

This chapter highlights that most language behavior consists of speaking and listening. The chapter also reveals differences and similarities between speaking and listening. The laboratory study of word production raises formidable problems; ensuring that a particular word is produced may subvert the spontaneous production process. Word production is investigated via slips and tip-of-the-tongue (TOT), primarily via instances of processing failure and via the technique of via the picture-naming task. The methodology of word production is explained in the chapter. The chapter also explains the phenomenon of interaction between various stages of word production and the process of speech recognition. In this context, it explores the difference between sound and meaning and examines whether or not the comparisons are appropriate between the processes of recognition and production of spoken words. It also describes the similarities and differences in the structure of the recognition and production systems. Finally, the chapter highlights the common issues in recognition and production research, which include the nuances of frequency of occurrence, morphological structure, and phonological structure.

Abstract

Strong and weak syllables in English can be distinguished on the basis of vowel quality, of stress, or of both factors. Critical for deciding between these factors are syllables containing unstressed unreduced vowels, such as the first syllable of automata. In this study 12 speakers produced sentences containing matched sets of words with initial vowels ranging from stressed to reduced, at normal and at fast speech rates. Measurements of the duration, intensity, F0, and spectral characteristics of the word-initial vowels showed that unstressed unreduced vowels differed significantly from both stressed and reduced vowels. This result held true across speaker sex and dialect. The vowels produced by one speaker were then cross-spliced across the words within each set, and the resulting words' acceptability was rated by listeners. In general, cross-spliced words were only rated significantly less acceptable than unspliced words when reduced vowels interchanged with any other vowel. Correlations between rated acceptability and acoustic characteristics of the cross-spliced words demonstrated that listeners were attending to duration, intensity, and spectral characteristics. Together these results suggest that unstressed unreduced vowels in English pattern differently from both stressed and reduced vowels, so that no acoustic support for a binary categorical distinction exists; nevertheless, listeners make such a distinction, grouping unstressed unreduced vowels by preference with stressed vowels

Abstract

Several models of spoken word recognition postulate that recognition is achieved via a process of competition between lexical hypotheses. Competition not only provides a mechanism for isolated word recognition, it also assists in continuous speech recognition, since it offers a means of segmenting continuous input into individual words. We present statistics on the pattern of occurrence of words embedded in the polysyllabic words of the English vocabulary, showing that an overwhelming majority (84%) of polysyllables have shorter words embedded within them. Positional analyses show that these embeddings are most common at the onsets of the longer word. Although both phonological and syntactic constraints could rule out some embedded words, they do not remove the problem. Lexical competition provides a means of dealing with lexical embedding. It is also supported by a growing body of experimental evidence. We present results which indicate that competition operates both between word candidates that begin at the same point in the input and candidates that begin at different points (McQueen, Norris, & Cutler, 1994, Noms, McQueen, & Cutler, in press). We conclude that lexical competition is an essential component in models of continuous speech recognition.

Abstract

Spoken utterances contain few reliable cues to word boundaries, but listeners nonetheless experience little difficulty identifying words in continuous speech. The authors present data and simulations that suggest that this ability is best accounted for by a model of spoken-word recognition combining competition between alternative lexical candidates and sensitivity to prosodic structure. In a word-spotting experiment, stress pattern effects emerged most clearly when there were many competing lexical candidates for part of the input. Thus, competition between simultaneously active word candidates can modulate the size of prosodic effects, which suggests that spoken-word recognition must be sensitive both to prosodic structure and to the effects of competition. A version of the Shortlist model ( D. G. Norris, 1994b) incorporating the Metrical Segmentation Strategy ( A. Cutler & D. Norris, 1988) accurately simulates the results using a lexicon of more than 25,000 words.

Abstract

Japanese, British English and American English listeners were presented with spoken words in their native language, and asked to mark on a written transcript of each word the first natural division point in the word. The results showed clear and strong patterns of consensus, indicating that listeners have available to them conscious representations of within-word structure. Orthography did not play a strongly deciding role in the results. The patterns of response were at variance with results from on-line studies of speech segmentation, suggesting that the present task taps not those representations used in on-line listening, but levels of representation which may involve much richer knowledge of word-internal structure.

Abstract

Previous research has shown that, in a phoneme detection task, vowels produce longer reaction times than consonants, suggesting that they are harder to perceive. One possible explanation for this difference is based upon their respective acoustic/articulatory characteristics. Another way of accounting for the findings would be to relate them to the differential functioning of vowels and consonants in the syllabic structure of words. In this experiment, we examined the second possibility. Targets were two pairs of phonemes, each containing a vowel and a consonant with similar phonetic characteristics. Subjects heard lists of English words had to press a response key upon detecting the occurrence of a pre-specified target. This time, the phonemes which functioned as vowels in syllabic structure yielded shorter reaction times than those which functioned as consonants. This rules out an explanation for response time difference between vowels and consonants in terms of function in syllable structure. Instead, we propose that consonantal and vocalic segments differ with respect to variability of tokens, both in the acoustic realisation of targets and in the representation of targets by listeners.

Abstract

English and French listeners performed two tasks - click location and speeded click detection - with both English and French sentences, closely matched for syntactic and phonological structure. Clicks were located more accurately in open- than in closed-class words in both English and French; they were detected more rapidly in open- than in closed-class words in English, but not in French. The two listener groups produced the same pattern of responses, suggesting that higher-level linguistic processing was not involved in these tasks.

Abstract

The performance of automatic speech recognition systems is usually assessed in terms of error rate. Human speech recognition produces few errors, but relative difficulty of processing can be assessed via response time techniques. We report the construction of a measure analogous to response time in a machine recognition system. This measure may be compared directly with human response times. We conducted a trial comparison of this type at the phoneme level, including both tense and lax vowels and a variety of consonant classes. The results suggested similarities between human and machine processing in the case of consonants, but differences in the case of vowels.

Abstract

Segmentation of continuous speech into its component words is a nontrivial task for listeners. Previous work has suggested that listeners develop heuristic segmentation procedures based on experience with the structure of their language; for English, the heuristic is that strong syllables (containing full vowels) are most likely to be the initial syllables of lexical words, whereas weak syllables (containing central, or reduced, vowels) are nonword-initial, or, if word-initial, are grammatical words. This hypothesis is here tested against natural and laboratory-induced missegmentations of continuous speech. Precisely the expected pattern is found: listeners erroneously insert boundaries before strong syllables but delete them before weak syllables; boundaries inserted before strong syllables produce lexical words, while boundaries inserted before weak syllables produce grammatical words.

Abstract

Monolingual French speakers employ a syllable-based procedure in speech segmentation; monolingual English speakers use a stress-based segmentation procedure and do not use the syllable-based procedure. In the present study French-English bilinguals participated in segmentation experiments with English and French materials. Their results as a group did not simply mimic the performance of English monolinguals with English language materials and of French monolinguals with French language materials. Instead, the bilinguals formed two groups, defined by forced choice of a dominant language. Only the French-dominant group showed syllabic segmentation and only with French language materials. The English-dominant group showed no syllabic segmentation in either language. However, the English-dominant group showed stress-based segmentation with English language materials; the French-dominant group did not. We argue that rhythmically based segmentation procedures are mutually exclusive, as a consequence of which speech segmentation by bilinguals is, in one respect at least, functionally monolingual.

Abstract

This paper presents lexical statistics on the pattern of occurrence of words embedded in other words. We report the results of an analysis of 25000 words, varying in length from two to six syllables, extracted from a phonetically-coded English dictionary (The Longman Dictionary of Contemporary English). Each syllable, and each string of syllables within each word was checked against the dictionary. Two analyses are presented: the first used a complete list of polysyllables, with look-up on the entire dictionary; the second used a sublist of content words, counting only embedded words which were themselves content words. The results have important implications for models of human speech recognition. The efficiency of these models depends, in different ways, on the number and location of words within words.

Abstract

We report two experiments in which vowels and steady-state consonants served as targets in a speeded detection task. In the first experiment, two vowels were compared with one voiced and once unvoiced fricative. Response times (RTs) to the vowels were longer than to the fricatives. The error rate was higher for the consonants. Consonants in word-final position produced the shortest RTs, For the vowels, RT correlated negatively with target duration. In the second experiment, the same two vowel targets were compared with two nasals. This time there was no significant difference in RTs, but the error rate was still significantly higher for the consonants. Error rate and length correlated negatively for the vowels only. We conclude that RT differences between phonemes are independent of vocalic or consonantal status. Instead, we argue that the process of phoneme detection reflects more finely grained differences in acoustic/articulatory structure within the phonemic repertoire.

Abstract

The lexicon, considered as a component of the process of recognizing speech, is a device that accepts a sound image as input and outputs meaning. Lexical access is the process of formulating an appropriate input and mapping it onto an entry in the lexicon's store of sound images matched with their meanings. This chapter addresses the problems of auditory lexical access from continuous speech. The central argument to be proposed is that utterance prosody plays a crucial role in the access process. Continuous listening faces problems that are not present in visual recognition (reading) or in noncontinuous recognition (understanding isolated words). Aspects of utterance prosody offer a solution to these particular problems.

Abstract

The recent claim by Black and Byng (1986) that lexical access in reading is subject to prosodic constraints is examined and found to be unsupported. The evidence from impaired reading which Black and Byng report is based on poorly controlled stimulus materials and is inadequately analysed and reported. An alternative explanation of their findings is proposed, and new data are reported for which this alternative explanation can account but their model cannot. Finally, their proposal is shown to be theoretically unmotivated and in conflict with evidence from normal reading.

Abstract

One of a listener's major tasks in understanding continuous speech is segmenting the speech signal into separate words. When listening conditions are difficult, speakers can help listeners by deliberately speaking more clearly. In three experiments, we examined how word boundaries are produced in deliberately clear speech. We found that speakers do indeed attempt to mark word boundaries; moreover, they differentiate between word boundaries in a way which suggests they are sensitive to listener needs. Application of heuristic segmentation strategies makes word boundaries before strong syllables easiest for listeners to perceive; but under difficult listening conditions speakers pay more attention to marking word boundaries before weak syllables, i.e. they mark those boundaries which are otherwise particularly hard to perceive.

Abstract

The present experiment tested the suggestion that human listeners may exploit durational information in speech to parse continuous utterances into words. Listeners were presented with six-syllable unpredictable utterances under noise-masking, and were required to judge between alternative word strings as to which best matched the rhythm of the masked utterances. For each utterance there were four alternative strings: (a) an exact rhythmic and word boundary match, (b) a rhythmic mismatch, and (c) two utterances with the same rhythm as the masked utterance, but different word boundary locations. Listeners were clearly able to perceive the rhythm of the masked utterances: The rhythmic mismatch was chosen significantly less often than any other alternative. Within the three rhythmically matched alternatives, the exact match was chosen significantly more often than either word boundary mismatch. Thus, listeners both perceived speech rhythm and used durational cues effectively to locate the position of word boundaries.

Abstract

In an earlier study, it was shown that listeners can use prosodic cues that predict where sentence stress will fall; phoneme-monitoring RTs are faster when the preceding prosody indicates that the word bearing the target will be stressed. Two experiments which further investigate this effect are described. In the first, it is shown that the duration of the closure preceding the release of the target stop consonant burst does not affect the RT advantage for stressed words. In the second, it is shown that fundamental frequency variation is not a necessary component of the prosodic variation that produces the predicted-stress effect. It is argued that sentence processing involves a very flexible use of prosodic information.

Abstract

Reaction time to detect a phoneme target in a sentence was found to be faster when the word in which the target occurred formed part of the semantic focus of the sentence. Focus was determined by asking a question before the sentence; that part of the sentence which comprised the answer to the sentence was assumed to be focussed. This procedure made it possible to vary position offocus within the sentence while holding all acoustic aspects of the sentence itself constant. It is argued that sentence understanding is facilitated by rapid identification of focussed information. Since focussed words are usually accented, it is further argued that the active search for accented words demonstrated in previous research should be interpreted as a search for semantic focus.

Abstract

The order of some conjoined words is rigidly fixed (e.g. dribs and drabs/*drabs and dribs). Both phonetic and semantic factors can play a role in determining the fixed order. An experiment was conducted to test whether listerners’ reaction times for monitoring a predetermined phoneme are influenced by phonetic constraints on ordering. Two such constraints were investigated: monosyllable-bissyllable and high-low vowel sequences. In English, conjoined words occur in such sequences with much greater frequency than their converses, other factors being equal. Reaction times were significantly shorter for phoneme monitoring in monosyllable-bisyllable sequences than in bisyllable- monosyllable sequences. However, reaction times were not significantly different for high-low vs. low-high vowel sequences.