Spontaneous Speech Production Research Articles

Vowel acquisition by prelingually deaf children with cochlear implants

Phonetic transcriptions (study 1) and acoustic analysis (study 2) were used to clarify the nature and rhythm of vowel acquisition following the cochlear implantation of prelingually deaf children. In the first study, seven children were divided according to their degree of hearing loss (DHL): DHL I: 90–100 dB of hearing loss, 1 children; DHL II: 100–110 dB, 3 children; and DHL III: over 110 dB, 3 children. Spontaneous speech productions were recorded and videotaped 6 and 12 months postsurgery and vowel inventories were obtained by listing all vowels that occurred at least twice in the child’s repertoire at the time of recording. Results showed that degree of hearing loss and age at implantation have a significant impact on vowel acquisition. Indeed, DHL I and II children demonstrated more diversified as well as more typical pattern of acquisition. In the second study, the values of the first and second formants were extracted. The results suggest evolving use of the acoustic space, reflecting the use of auditory feedback to produce the three phonological features exploited to contrast French vowels (height, place of articulation, and rounding). The possible influence of visual feedback before cochlear implant is discussed.

Melodic intonation therapy in Persian aphasic patients

Background : Melodic intonation therapy (MIT) is a well-known method of aphasia rehabilitation using prosodically based melodic phrases. The literature includes MIT adapted to many languages and efficiently applied to certain groups of non-fluent aphasic patients. However, there has been no report on the efficacy of the method for Persian-speaking patients. Aims : The objective of this study was primarily to investigate the effects of 15 sessions of MIT treatment (adapted to the Persian language) in selected patients with non-fluent aphasia; primarily on expository speech (phrase length and number of correct content units) and oral expression skills (repetition, responsive naming, and confrontational naming), and secondarily, on auditory comprehension abilities (word discrimination, commands). Methods and Procedures : Participants in the study included seven right-handed Persian-speaking patients afflicted with chronic (>14 months post-onset) non-fluent aphasia. Based on the rules of Persian prosody, MIT was adapted to the Persian language. Using a pre, post treatment design, each outcome measure was tested twice before and twice after MIT treatment. Changes in the variables not treated were also measured as baselines controlling treatment effects. Outcomes and Results : Using the Wilcoxon signed-rank test, improvements in the selected variables were shown to be statistically significant after the treatment phase and not during the treatment-free phases. Non-target variables remained unchanged after the treatment and throughout the non-treatment phases. Conclusions : Our study showed that MIT can be adapted for Persian aphasic patients and administered with measurable positive results after 15 sessions of treatment. MIT improved primarily spontaneous speech production, and as minor effects, selected oral expression and auditory comprehension subtests. Therefore MIT might be considered as a method for the rehabilitation of selected non-fluent Persian aphasic patients. More long-term follow-up studies with randomised controlled clinical trials are needed for stronger conclusions.

Null subject pronouns in Broca's speech production

Background: Pronouns are known to be often missing from the spontaneous speech production of Broca's aphasic patients. Null subject pronouns are found in nonfinite utterances. These structures may result from the underspecification of the Tense head and the unavailability of Nominative case for a subject. It is not clear, though, whether null subject pronouns are restricted only to nonfinite structures.Aims: The paper investigates the occurrence of null subject pronouns in finite sentences of Broca's patients. A detailed study into the syntactic and distributional properties of finite null subjects will reveal their nature and origin in aphasic speech production.Methods & Procedures: The spontaneous speech of 13 Dutch patients was investigated. The following variables were included: finiteness (finite versus nonfinite utterances), sentence position (initial versus non‐initial position), and type of pronoun (personal versus demonstrative pronoun). The speech production of the patients was contrasted to normal spoken Dutch.Outcomes & Results: The results demonstrate that finite null subject pronouns in Broca's speech production can be interpreted as instances of normal omission phenomena found in Germanic languages (Topic Drop and I‐Drop).Conclusion: A unified account is proposed for nonfiniteness and finite null subjects in terms of underspecification of functional heads in the syntactic structure. Underspecification is not due to a linguistic impairment, but rather a means to reduce the processing load of the utterance. It is concluded that nonfiniteness is a core phenomenon in Broca's speech production, while finite null subjects reflect an additional omission phenomenon.

Linking facial animation, head motion and speech acoustics

Facial motion during speech is a direct consequence of vocal-tract motion which also shapes the acoustics of speech. This fact suggests that speech acoustics can be used to estimate face motion, and vice versa. Another kinematic–acoustic relation that occurs during production of speech is between head motion and fundamental frequency ( F 0). This paper focuses on the development of a system that takes speech acoustics as input, and gives as output the coefficients necessary to animate natural face and head motion. The results obtained are based on simultaneous measurements of face deformation, head motion and speech acoustics collected for two subjects during production of naturalistic sentences and spontaneous speech. The procedure for estimating face motion from speech acoustics first trains nonlinear estimators whose inputs are line spectrum frequency pair coefficients and whose outputs are marker positions on the face. These estimators are then applied to test data. The estimated marker trajectories are objectively compared with their measured counterparts, yielding correlation coefficients between 0.8 and 0.9. Linear estimators are used to relate F 0 and head motion. As F 0 to head motion is a one-to-many problem, constraints must be added for the estimation of head motion. This is done by computation of the co-dependence among head motion components. Finally, measured and estimated face and head motion data are used to animate a naturalistic talking head.

Acoustic Characteristics of Adolescent Actors’ and Non-Actors’ Voices

The aim of the present study was to investigate differences in the acoustic characteristics of the voice between adolescent actors and non-actors. The experimental sample consisted of 10 actresses and 10 actors while the control sample included 13 girls and 14 boys. Phonation of the vowel /a/, spontaneous speaking and oral reading provided a set of acoustic variables (fundamental frequency, jitter, shimmer, speaking and reading ranges); t test showed statistically significant differences between actresses and non-actresses in speaking range, reading F₀ maximum, and reading range, whereas between actors and non-actors the difference was found in reading range only. The results showed that drama in education without systematic voice training had no effects in terms of acoustic characteristics of the voice.

Using uh and um in spontaneous speaking

The proposal examined here is that speakers use uh and um to announce that they are initiating what they expect to be a minor (uh), or major (um), delay in speaking. Speakers can use these announcements in turn to implicate, for example, that they are searching for a word, are deciding what to say next, want to keep the floor, or want to cede the floor. Evidence for the proposal comes from several large corpora of spontaneous speech. The evidence shows that speakers monitor their speech plans for upcoming delays worthy of comment. When they discover such a delay, they formulate where and how to suspend speaking, which item to produce ( uh or um), whether to attach it as a clitic onto the previous word (as in “and-uh”), and whether to prolong it. The argument is that uh and um are conventional English words, and speakers plan for, formulate, and produce them just as they would any word.

A preliminary investigation of phonetic diversity in adults with motor speech disorder

A preliminary investigation of the phonetic diversity characterizing the speech of non-impaired and speech-impaired adults was performed. Two measures of phonetic diversity, the ratio of consonants-to-vowels (C:V) and the phonetic product (PP), were applied to the spontaneous speech productions of a group of non-impaired adult females and adult females with a motor speech disorder. The C:V ratio evaluates the ratio of supraglottal-to-glottal articulations occurring within a conversational speech sample. The PP evaluates the frequency of occurrence of consonant and vowel sounds within an utterance according to place of articulation categories. Significant differences in C:V ratio were found between the two groups; however, the PP failed to discriminate the groups. Results are discussed with regard to the general articulatory characteristics of non-impaired and speech-impaired adults, as well as the usefulness of examining phonetic diversity in adults with motor speech disorders.

The relationship between gender and learning styles amongst A level modern languages students

This article describes research conducted with 72 MFL A level students regarding possible links between gender and learning styles. In terms of confirming conventional wisdom, findings only coincided in four areas: ‘male’ risk-taking, spontaneous speaking and self-confidence; and ‘female’ organisation of written work. Gender-linked differences in other areas were not confirmed.

Nasal airflow in French spontaneous speech

The goal of this paper is to compare the degree of anticipatory and carryover phenomena in the behaviour of the velum during the production of French spontaneous speech, and to compare the data with the same speech material, but read. Airflows through the nose and the mouth were taken as an indirect indication of the velum behaviour. French is a language of special interest because there is a phonological contrast between oral and nasal vowels. The results show a tendency for nasal airflow to start before the nasal and a strong propensity to spread after the nasal. No differences could be established between speaking styles (spontaneous vs. read speech) regarding nasal airflow anticipation and carryover.

Aphasia manifestations in Basque

This paper presents a description of the main symptoms of agrammatism found in a corpus of spontaneous speech production of 59 Basque-speaking aphasic patients. The main symptoms relate to deficits in the production of inflectional morphology. From the standpoint of grammatical structure, the deficits involve the production of the syntactic category Infl, that heads the sentence (IP). Basque has a morphologically complex Inflection, containing up to three agreement markers, plus tense and modality specifications. The most outstanding symptom of agrammatism is the absence of inflected verbs, or its extremely low frequency, correlated with a high frequency of uninflected participial forms. The study of the inflected forms produced in aphasic speech reveals a significant number of errors in the selection of appropriate morphemes, particularly number and dative agreement markers, and a considerable number of newly coined impossible forms, resulting form the combination of morphemes from conflicting paradigms. However, the relative order of the morphemes in the auxiliary is correct, which could indicate that the basic structure of Inflection is not affected, at least in the case of patients that produce inflected forms. Errors in case-marking also appear, mostly involving ergative and absolutive markers.

Investigating Unscripted Speech: Implications for Phonetics and Phonology

This paper looks at patterns of reduction and elaboration in speech production, taking the phenomenon of plosive-related glottalization in German spontaneous speech, on the basis of the ‘Kiel Corpus’, as its point of departure, and proposes general principles of human speech to explain them. This is followed by an enquiry into the nature of a production-perception link, based on complementary data from perceptual experiments. A hypothesis is put forward as to how listeners cope with the enormous phonetic variability of spoken language and how this ability may be acquired. Finally, the need for a new paradigm of phonetic analysis and phonological systematization is stressed, as a prerequisite to dealing adequately and in an insightful way with the production and perception of spontaneous speech.

Temporal cycles in speech production

The term macroplanning has been used to refer to the construction of semantic-conceptual structures that enable speakers to accomplish communicative goals during spontaneous speech production. One form of evidence provided as support for the existence of macroplanning are temporal cycles in speech fluency. The assumption is that speakers alternate between phases of low fluency, during which they prepare macroplans, and high fluency, during which macroplans are executed in speech. Existing models of macroplanning suggest that temporal cycles therefore reflect a scheduling problem between processes competing for the capacity of shared resources. Time series analyses confirmed the existence of temporal cycles in spontaneous spoken discourse. In addition, evidence that topic shifts generally precede periods of increasing fluency in the temporal cycle was thought to be consistent with the assumption that the cycle reflects macroplanning activity. Existing models of macroplanning are reviewed in light of these results, and problems associated with the underlying assumptions of these models are discussed.

In Process Citation

Pure word deafness (auditory verbal agnosia) is characterized by an impairment of auditory comprehension, repetition of verbal material and writing to dictation whereas spontaneous speech production and reading largely remain unaffected. Sometimes, this syndrome is preceded by complete deafness (cortical deafness) of varying duration. Perception of vowels and suprasegmental features of verbal utterances (e.g., intonation contours) seems to be less disrupted than the processing of consonants and, therefore, might mediate residual auditory functions. Often, lip reading and/or slowing of speaking rate allow within some limits to compensate for speech comprehension deficits. Apart from a few exceptions, the available reports of pure word deafness documented a bilateral temporal lesion. In these instances, as a rule, identification of nonverbal (environmental) sounds, perception of music, temporal resolution of sequential auditory cues and/or spatial localization of acoustic events were compromised as well. The observed variable constellation of auditory signs and symptoms in central hearing disorders following bilateral temporal disorders, most probably, reflects the multitude of functional maps at the level of the auditory cortices subserving, as documented in a variety of non-human species, the encoding of specific stimulus parameters each. Thus, verbal/nonverbal auditory agnosia may be considered a paradigm of distorted "auditory scene analysis" (Bregman 1990) affecting both primitive and schema-based perceptual processes. It cannot be excluded, however, that disconnection of the Wernicke-area from auditory input (Geschwind 1965) and/or an impairment of suggested "phonetic module" (Liberman 1996) contribute to the observed deficits as well. Conceivably, these latter mechanisms underly the rare cases of pure word deafness following a lesion restricted to the dominant hemisphere. Only few instances of a rather isolated disruption of the discrimination/identification of nonverbal sound sources, in the presence of uncompromised speech comprehension, have been reported so far (nonverbal auditory agnosia). As a rule, unilateral right-sided damage has been found to be the relevant lesion.

Speaking fundamental frequency of young adult Arabic men during oral reading and spontaneous speaking in both Arabic and English languages

Currently, there is a paucity of normative data on speech and voice characteristics of different linguistic and ethnic groups. In particular, there is no available published data on the speaking fundamental frequency (SFF) characteristics of the voice of the Arabic population. The purpose of this study was to obtain preliminary data on the SFF of a group of normal speaking, young adult Arabic males. Fifteen, native Arabic, adult men served as subjects and received the identical experimental treatment. Four speech samples were collected from each subject (Arabic reading, Arabic spontaneous speech, English reading, and English spontaneous speech). Results showed the mean and SD of SFF are ([146,9, 15.4], [145.8, 13.8], [149.1, 12.6], and [145.5, 12.0]), respectively. No significant differences were found in the mean SFF between language and type of speech, nor between languages. A significant difference in the mean SFF was found between the type of speech (F1,14=5.51, p=0.03). Reading was significantly higher than speech. Also, Arabic men in this study had higher SFF values than previously reported for young adult males of other ethnic groups [H. Hollien and B. Jackson, J. Phonetics 117–120 (1973); A. Hudson and A. Holbrook, J. Speech Hear. Res. 197–201 (1992)]. For Speech Communication Best Student Paper Award.

Rhythm in speech: What rhythmic organizations reveal about cognitive processes in spontaneous speech production versus reading aloud

The rhythm of speech can be defined from two different standpoints: a metric one and a rhythmic one. Considered as an assimilation tendency involving the regulation of intervals, speech rhythm becomes a metric concept; considered as a dissimilatory tendency involving the production of events, it becomes a rhythmic concept. The illusion of the measurable regularity of speech is supported by the study of reading aloud, erroneously considered to be a ‘neutral’ kind of speech. Because of its improvised nature, spontaneous speech favors a rhythmic organization. Reading aloud, on the other hand, favors a metric organization because of the structured nature of text. Metric analysis is based on the premise that a temporal continuum can be analyzed by quantification, while rhythmic analysis approaches temporal organization through the mechanisms of perception. The metric/rhythumic opposition is linked to the existence/non-existence of units in speech production and perception. The choice of the corpus is highly important: as we have shown, the rhythmic theory depends on that choice.

Pronouncing “the” as “thee” to signal problems in speaking

In spontaneous speaking, the is normally pronounced as thuh, with the reduced vowel schwa (rhyming with the first syllable of about). But it is sometimes pronounced as thiy, with a nonreduced vowel (rhyming with see). In a large corpus of spontaneous English conversation, speakers were found to use thiy to signal an immediate suspension of speech to deal with a problem in production. Fully 81% of the instances of thiy in the corpus were followed by a suspension of speech, whereas only 7% of a matched sample of thuhs were followed by such suspensions. The problems people dealt with after thiy were at many levels of production, including articulation, word retrieval, and choice of message, but most were in the following nominal. © 1997 Elsevier Science B.V. All rights reserved.

Speech production and use of sign by young cochlear implant users.

The use of sign and the emergence of speech skills by children who are deaf and use a cochlear implant were examined. Twenty‐five children who had an average of 43 months of experience with a Nucleus cochlear implant and who are educated in a simultaneous communication environment engaged in spontaneous conversation and completed speech production and audiological tests. During spontaneous conversation, the children used voice and sign to express whole words in their conversations 70% of the time, voice only 21%, and sign only 9% of the time, suggesting that use of a cochlear implant does not result in elimination of signing. Children demonstrated a wide range of intelligibility. Pearson correlations indicated that children who had better intelligibility were most likely to use voice only, while children who had poor intelligibility were most likely to use sign only. Analysis of the children’s consonant production over time suggested that fricative production emerges relatively early; but follows emergence of other speech skills, including the ability to produce nasal consonants. Measures of consonant production were significantly correlated with measures of consonant perception. This suggests a close link between children’s ability to produce speech and perceive speech.

Average speaking fundamental frequency in soprano singers with and without symptoms of vocal attrition

Nineteen trained soprano singers aged 18-30 years vocalized tasks designed to assess average speaking fundamental frequency (SFF) during spontaneous speaking and reading. Vocal range and perceptual characteristics while singing with low intensity and high frequency were also assessed, and subjects completed a survey of vocal habits/symptoms. Recorded signals were digitized prior to being analyzed for SFF using the Kay Computerized Speech Lab program. Subjects were assigned to a normal voice or impaired voice group based on ratings of perceptual tasks and survey results. Data analysis showed group differences in mean SFF, no differences in vocal range, higher mean SFF values for reading than speaking, and 58% ability to perceive speaking in low pitch. The role of speaking in too low pitch as causal for vocal symptoms and need for voice classification differentiation in vocal performance studies are discussed.

Comparative vocabulary development in kindergarten classes with a mixed population of monolinguals, simultaneous and successive bilinguals

This article reports on a language‐evaluation study which was made in Brussels kindergartens. More particularly the vocabulary evolution of originally monolingual Dutch‐speaking children, originally simultaneously bilingual children (Dutch/French) and young successive second‐language learners (French at home, Dutch in kindergarten) was charted. A cross‐sectional, as well as a longitudinal investigation was conducted with three‐, four‐, and five‐year‐olds. A special instrument, which evaluates active and passive vocabulary, sentence comprehension, morphology proficiency and spontaneous speech production, was designed for this study. Especially the vocabulary items are discussed in this article as they are representative for all other test parts. The conclusion highlights the possibilities and problems with regard to looking after these children continuously in Dutch‐speaking primary education. The study also seeks an answer as to the relative possibility or impossibility to instruct these groups of children together.

The Pitch of Speech as a Function of Linguistic Community

A recent series of experiments by Deutsch and co-workers has investigated the perception of musical patterns in which the tones are well defined in terms of pitch class, but poorly defined in terms of pitch height. One of these patterns is known as the "tritone paradox." It has been found that listeners' differing perceptions are significantly correlated both with the linguistic community in which the listener grew up and with the pitch range of the listener's spontaneous speaking voice. To explain these findings, Deutsch has hypothesized that listeners acquire an internal representation of pitch classes based on the prevailing pitch range of speech in their linguistic community and that this representation influences both their perception of the tritone paradox and their speech production. The present paper examines this hypothesis in the light of available data about the pitch of speech as a function of linguistic community. It is concluded that these data are surprisingly consistent with Deutsch's hypothesis.