The Speed of Articulatory Movements Involved in Speech Production in Children With Dyslexia

Abstract

A group of children with dyslexia (mean ages 9 and 14 years) was studied, together with group of children without dyslexia matched for age. Participants were monolingual native speakers of the Bosnian language with transparent orthography. In total, the diagnostic tests were performed with 41 children with dyslexia and 41 nondyslexic children. The participants were asked to produce monosyllables, /pa/, /ta/, and /ka/, and the trisyllable /pataka/, as fast as possible. Analysis was undertaken in four ways: (1) time of occlusion duration for plosives (duration of stop), (2) voice onset time for plosives, (3) diadochokinetic rate—articulators rate measured by pronunciation of monosyllables and the trisyllable, and (4) time of moving articulators from one gesture to another—time of interval length (from the explosion of one plosive to the start of the explosion of another plosive). The results suggest that children with dyslexia have significant problems with the speed of articulatory movements involved in speech production.

Keywords

dyslexia articulation

Two current major hypotheses inform the etiology of developmental dyslexia. One of these assumes that the phonological processing with emphasis on the segmental level identifies the core deficit in developmental dyslexia and that it cannot be reduced to domain-general deficits of temporal information processing. The other hypothesis implies that phonological processing deficits are symptomatic of domain-general dysfunction and that some dyslexia subtypes are related to domain-general deficits of temporal information processing for auditory and visual stimuli (Wolff, 2002).

A developmental model of reading claims that a deficit in motor-articulatory feedback in developmental dyslexia is connected with the development of deficient phonological awareness and nonlexical reading. The articulatory representation includes the oral-motor programs for the movement of the articulators to produce the pseudoword’s phonemes and consists of the somatosensory awareness of the distinctive features within these motor programs (Conway, 2003).

Clinical observations and experimental reports provide evidence about speech problems in poor readers (Kamhi, Catts, & Mauer, 1990). In the literature, articulatory disorders have been associated with dyslexia (Lalain, Joly-Pottuz, Nguyen, & Habib, 2003). Speech production skills include articulation difficulties as well as difficulties with the acquisition of the phonological system (De Bree, 2007). Use of speech coding in an articulatory loop is the same in children with and without dyslexia, but the system operates less efficiently in children with dyslexia (Hulme & Roodenrys, as cited in Conway, 2003).

It is widely considered that speech production difficulties in individuals with dyslexia are a reflection of their deficits in phonological processing (Catts, 1986; De Bree, 2007; Griffiths & Frith, 2002). The inability to associate the position of their articulators with speech sounds may impair the development of phonological awareness and the ability to convert graphemes to phonemes, which may be related to programming or feedback deficits (Heilman, Voeller, & Alexander, 1996).

Several researchers investigated speech production and found that children with dyslexia exhibited residual speech difficulties. Studies have found word-specific rather than phoneme-specific production errors in children with dyslexia, especially with multisyllabic words. These difficulties reflect the children’s poor phonological representations and processing (De Bree, 2007). Gerrits and De Bree (2009) found that speech perception and production performance of children at familial risk of dyslexia was poorer than that of controls. Speaking rate and the proportion of pausing time to speaking time may provide an early indication of reading outcome in children at high risk for reading disability (Smith, Roberts, Smith, Locke, & Bennett, 2006).

Kamhi and colleagues (1990) examined the speech production abilities of young, poor readers. Their results showed that poor readers made more speech errors than good readers when producing both words and phrases. The poor readers took significantly longer to produce three of the words correctly than did their nondisabled peers. The data indicated that encoding limitations, rather than speech production limitations, were primarily responsible for the longer acquisition time.

Articulatory awareness was measured in adults with dyslexia, on a task that is not confounded with orthography, and the group with dyslexia showed deficits on this task in comparison to the control group. Information about articulatory movements for specific phonemes is less accessible to dyslexics because of a deficient phonological processing system. An articulatory awareness deficit can persist into adulthood, even when literacy skills are no longer seriously impaired, which suggests that this may be part of a cognitive deficit that underlies dyslexia (Griffiths & Frith, 2002).

In speech research, a metronome is often used to measure timing precision. Wolff (2002) asked adolescents and adults with dyslexia to repeat the bisyllabic consonant-vowel, or CV, sequences /pa-ta/ and /ta-ka/ and the trisyllabic/ a-ta-ka/ correctly in time with a metronome beat. The findings indicated that during a motor sequencing task, students with dyslexia anticipated the signal of an isochronic pacing metronome by intervals that were two or three times as long as those of age-matched readers or adults without dyslexia. Children with dyslexia deviated more from the prescribed rate, repeated syllables too slowly at all metronome speeds, and made more speech sequencing errors than nondyslexic children with other learning disabilities (Wolff, Michel, & Ovrut, 1990).

A significant proportion of children with dyslexia show a comorbidity with motor disorders, which—it has been hypothesized—reflects an automaticity deficit or a cerebellar dysfunction. The automaticity/cerebellar deficit framework researchers postulated that the cerebellum was unable to regulate motor control. Therefore, articulation in speech would lead to deficient phonological representations and/or to automate of overlearned tasks in reading and would affect the learning of grapheme and phoneme correspondences (Caylak, 2010).

The automaticity/cerebellar deficit framework provides demonstration that it is possible to explain motor, speed, and phonological deficits within a unified account (Nicolson & Fawcett, 2005). If an infant has a cerebellar impairment, this will first show up as a mild motor difficulty—the infant may be slower to sit up and to walk, and may have greater problems with fine muscular control. If we consider that our most complex motor skill is articulation, the infant might be slower to start babbling and, later, talking. Speech and walking may be less fluent in infants with cerebellar impairment. If articulation is less fluent than normal, it takes up more conscious resources, leaving fewer resources to process sensory feedback. Processing the auditory, phonemic structure of spoken words may be less complete (Fawcett & Nicolson, 2004).

Kasselimis, Margarity, and Vlachos (2008) carried out a study with aims (a) to assess the cerebellar deficit hypothesis by examining children’s performance in cerebellar and cognitive tasks associated with dyslexia and (b) to investigate if there is a differentiation in articulation speed in children with dyslexia. Children with dyslexia showed significant impairment in one cerebellar test and performed significantly worse during the articulation speed test than the control group. The present study supports the cerebellar deficit hypothesis and the relationship between reading impairment and speed of articulation.

There are three current theoretical issues that have not yet been fully resolved in the literature. It is not clear (1) whether the problems arise specifically in speech motor planning or whether they also reflect slower speech production, (2) whether such deficits are found only with more complex stimuli, and (3) whether it is only the occurrence of errors in speeded repetition that results in deficient performance (Fawcett & Nicolson, 2002).

There is growing interest in examining whether the manner in which a particular orthography accurately represents phonology relates to dyslexia (Goulandris, 2003). The aim of the present study is to examine the articulatory movements in children with dyslexia who speak Bosnian, a language with transparent orthography. Transparent orthographies are those where the grapheme–phoneme correspondence is one to one. The mappings from letters to sound are much more consistent, with very few irregular words (Gupta & Jamal, 2006). Consistency in sound–letter associations and primarily one-to-one phoneme–grapheme correspondence makes Bosnian a transparent orthography.

There has been little systematic research on the articulation of children with dyslexia in Bosnia and Herzegovina. In this article, we concentrate on articulatory movements in children with dyslexia with the aim of investigating whether they have slower articulation of speech. Also in this article we report research on whether children with dyslexia simply have slower articulation during pronunciation of a single phoneme or whether the slower articulation problem also occurs when moving articulators from one articulatory gesture to another.

Method

Design

The research presented here was designed to examine the speed of articulatory movement involved in speech production. Articulation of voiceless plosives was measured. Plosives represent the basis for the pronunciation of some other consonants, both in terms of behavior of speech organs during the articulation and in terms of acoustic structure, because their elements can be found in some other consonants. They are very important for development of the sound system in children. It is no wonder that they occur among the first sounds children make. They are simpler, with frequency slightly represented in speech (20.5%), and more resilient than other sounds on a variety of pathological deviations in articulation. The group of plosives consists of six consonants of the Bosnian language: p, b, t, d, k, and g. The common feature of the plosive is airflow stop through partitions (occlusion), then a sudden release of the airstream (explosion). Partitions are created using different parts of speech organs. Lips are the most important for pronunciation of sounds p and b, tongue tip is the most active for pronunciation of sounds t and d, and tongue back is the most engaged for pronunciation of sounds k and g (Vladisavljevic, 1981). So, tasks in this study included lips, tongue tip, and tongue back movements.

Analysis was undertaken in four ways: (1) time of occlusion duration for plosives (duration of stop), (2) time of voice onset time for plosives, (3) diadochokinetic rate—articulators rate measured by pronunciation of monosyllables and trisyllable, and (4) time of moving articulators from one gesture to another—time of interval length (from the explosion of one plosive to the start of the explosion of another plosive).

Participants

A group of children with dyslexia (mean ages 9 and 14 years) was studied, together with group of children without dyslexia, matched for age. Participants were monolingual native speakers of Bosnian. The participants were enrolled in Grades 4 and 9 in two public schools in Tuzla Canton. The children with dyslexia were recruited from the same classes as the typically achieving readers, and they did not attend additional remedial reading classes. According to the teachers’ reports, none of these students had any speech, language, or intellectual problems. None had ever been referred for speech or language treatment. All of them were children of upper-middle-class families. All were first diagnosed with dyslexia during this research.

The diagnosis followed the recommendations in the Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (DSM-IV; American Psychiatric Association, 1994), namely, normal level of general intelligence, and no neurological, sensory, or educational deficit that could justify their reading impairments. None of the participants reported a history of neurological diseases, psychiatric disorders, or hearing problems. All children received an evaluation of nonverbal IQ with Raven’s Progressive Matrices by a psychologist who verified the criteria of normal intelligence.

Children with dyslexia were tested by a speech language therapist, an expert on dyslexia. All participants were given tests of reading and writing. For the evaluation of reading and writing abilities, a reading test for school children was used measuring reading time and reading errors for real words and for nonwords. Writing was measured by 2-min spelling and 1-min writing. Tests titled One-Minute Reading Aloud (Furlan, 1965) and Word List for Checking Reading Skills (Matanovic-Mamuzic, 1982) were used. The 2-min spelling, 1-min writing, and nonsense passage reading subtests were adapted for the Bosnian language from the Dyslexia Screening Test–Junior (Fawcett & Nicolson, 2004).

For all reading and writing measures, to qualify as having dyslexia, participants had to score below two standard deviations of the mean performance of the controls. In total, the diagnostic tests were performed with 41 children with dyslexia and 41 children without dyslexia. (See Table 1 for participant characteristics and group differences.)

Table 1.

Test Scores and Ages for Children With and Without Dyslexia.

	Children With Dyslexia (9 years old)		Children Without Dyslexia (9 years old)			Children With Dyslexia (14 years old)		Children Without Dyslexia (14 years old)
	M	SD	M	SD	p	M	SD	M	SD	p
Age in years and months	9.26	1.07	9.38	.97	.710	14.16	.66	14.30	.71	.500
1-min reading^a	26.10	7.51	67.10	12.15	<.001	65.30	14.53	99.67	11.81	<.001
Errors, real-word reading^b	25.87	11.30	2.6	1.99	<.001	10.05	2.19	.43	.75	<.001
Reading time, real-word reading(s)^b	238.07	115.28	86.13	17.46	<.001	106.00	24.44	38.43	10.78	<.001
2-min spelling^c	7.33	2.19	13.22	2.39	<.001	14.85	2.74	22.33	2.11	<.001
1-min writing^c	5.43	2.14	8.95	2.78	<.001	10.78	5.65	16.35	3.12	<.001
Nonsense passage reading^c	26.58	16.96	62.53	3.878	<.001	41.75	11.34	70.88	3.28	<.001

One-minute Reading Aloud (Furlan, 1965).

Word List for Checking Reading Skills (Matanovic-Mamuzic, 1982).

Subtests adapted for the Bosnian language from the Dyslexia Screening Test–Junior (Fawcett & Nicolson, 2004).

Materials and Procedures

Speaker data collection

The participants were asked to produce monosyllables, /pa/, /ta/, and /ka/, and the trisyllable /pataka/, as fast as possible. The children were instructed as follows: Take a deep breath and repeat pa-pa-pa/ta-ta-ta/ka-ka-ka/pa-ta-ka, as fast as possible.

Participants repeated the stimuli until instructed to stop, after more than three repetitions, because three repetitions were planned to be used for analysis. Three trials of each task were taken. The experimenter previously demonstrated by example how to do that and gave instructions to pronounce it as fast as possible. The fastest and most accurate production among the three trials was selected for analysis.

Recording

The recording was done in a quiet room. The speech productions of the participants were recorded using a Sony (MZ-R91) Portable Mini-Disc Player/Recorder and a Sony (ECM-MS907) Electret Condenser Microphone, with microphone-to-mouth distance of 30 cm.

Editing

Data was stored on a personal computer. Dr. Speech software (Version 4: subprogram, Real Analysis: Tiger Electronics, Seattle, WA) was used in the analysis. Real Analysis software is a real-time speech assessment and training system that helps to analyze, document, teach, reinforce, and report speech waveform in various applications. It is a powerful clinical tool that provides highly versatile information for speech assessment, including among several features, real-time recording and playback (Huang & Lin, 1995).

Analyses

The total duration of the monosyllables and trisyllable were measured in all three repetitions, and the duration of the second syllable of the trisyllable in three repetitions. The results obtained by producing monosyllables were used to analyze the speed of movement of the lips, tongue tip, and tongue back. Duration from the beginning of explosion of the phoneme /p/ to beginning of explosion the next /p/ was used for analysis of the movement of the lips. The same procedure was used with the measurement of /t/ and /k/ for getting information about the movement of the tongue tip and tongue back.

Incomplete closures of stop consonants reflect undershooting of the articulatory movements (Esther, 2001). According to Ackermann and Ziegler (1991; as quoted in Esther, 2001, p. 11) “incomplete closures resulting from reduced extension of the mandibular, labial, or lingual movements or from reduced occlusive force may be expected to result in an increase of sound pressure during stop realization.”

Characteristics of plosives are to stop airflow through partition (occlusion), then suddenly release the airstream (explosion). The duration of stopping airflow through partitions, or occlusion, was measured. Vladisavljevic (1981) considered that from an acoustic aspect, there is complete silence during occlusion for voices /p/, /t/, /k/. Suspended airflow, after sudden opening of the partition, appears as a crack of weaker or stronger intensity, which extends almost throughout the frequency spectrum, creating small, characteristic acoustic energy concentrates in certain frequency areas.

Voice onset time was estimated by the measured time from the beginning of the realization of a consonant phoneme–plosive to the occurrence of first laryngeal wave of the vowel that follows. Pronunciation of the trisyllable /pataka/ was used for this purpose.

The speed of making rapid speech movements was also measured by analyzing the trisyllable /pataka/ in all three repetitions, and the duration of the second syllable in three repetitions. Duration of the interval P-T (from beginning of explosion of phoneme P to beginning of explosion of phoneme T) and duration of interval T-K (from beginning of explosion of phoneme T to beginning of explosion of phoneme K) were measured on the same trisyllable. The measurement was carried out on the middle syllable of the trisyllable because it was the most suitable for acoustic measurements. (First and last phonemes were often not sufficiently visible.)

The measurement for all tasks was done in milliseconds. The method used in the present study was a modification of the methods that were used in the research carried out by Hedjever (1996), Duranovic and Zecic (2001/2002), and Duranovic (2003).

Results

Mean for scores of children with dyslexia and scores produced by the control group are shown in Figure 1. It can be seen that the performance of the children with dyslexia was consistently worse than that of the controls, for all tasks. The mean for all 14 tasks is higher for children with dyslexia.

Figure 1.

Mean performance in milliseconds of articulatory movement involved in speech production tasks.

The analysis was performed on the data from the producing monosyllables and the trisyllable to verify that the children with dyslexia and a group of children without dyslexia differ with respect to the speed of simple articulatory gesture production and the speed of moving articulators from one gesture to another, and to verify effect of age on articulatory rates. First, we carried a one-way ANOVA to address effect of age. One-way ANOVA was undertaken separately for each task, involving the two groups. This analysis had one factor: chronological age (9 and 14 years). Age effects were not significant for any task (see Table 2). Because no significant age effects had been found, difference by ages for the children with dyslexia and the control group was not considered in further analysis, and groups were divided only by the presence or absence of dyslexia. Next, ANOVA with the factor of group (dyslexia vs. control) revealed that the region of significant difference between the children with dyslexia and the control group was for all tasks, except for the task where we measured the duration of the second syllable of the trisyllable /pataka/ in three repetitions. It was important to identify whether children with dyslexia perform worse than children of the same age without dyslexia. It may be seen that the children with dyslexia performed significantly worse than the same-age controls on all tasks, except for one of them, which was previously mentioned.

Table 2.

Results of the Analyses of Variance to Show Effects by Age and Group.

	Age		Group
Speech Production Tasks	F	p	F	p
OCCLT	F(1) = .53	NS	F(1) = 68.52	<.001
OCCLK	F(1) = .80	NS	F(1) = 33.17	<.001
VOTT	F(1) = .09	NS	F(1) = 26.29	<.001
VOTK	F(1) = .44	NS	F(1) = 27.24	<.001
PAMIDDLE	F(1) = .08	NS	F(1) = 14.35	<.001
TAMIDDLE	F(1) = .23	NS	F(1) = 37.34	<.001
KAMIDDLE	F(1) = .53	NS	F(1) = 04.22	<.050
PATHREE	F(1) = .34	NS	F(1) = 19.11	<.001
TATHREE	F(1) = .49	NS	F(1) = 39.81	<.001
KATHREE	F(1) = .02	NS	F(1) = 16.27	<.001
PTKMIDDL	F(1) = .44	NS	F(1) = 00.40	NS
PTKTHREE	F(1) = .39	NS	F(1) = 06.21	<.050
PTINTERV	F(1) = .35	NS	F(1) = 49.75	<.001
TKINTERV	F(1) = .07	NS	F(1) = 44.68	<.001

Note: OCCLT = occlusion of voice /t/; OCCLK = occlusion of voice /k/; VOTT = voice onset time of /t/; VOTK = voice onset time of /k/; PAMIDDLE = duration of second syllable /pa/ in three repetitions; TAMIDDLE = duration of second syllable /ta/ in three repetitions; KAMIDDLE = duration of second syllable /ka/ in three repetitions; PTKMIDDL = duration of second syllable /pa-ta-ka/ in three repetitions; PATHREE = duration of monosyllable /pa/ in all three repetitions; TATHREE = duration of monosyllable /ta/ in all three repetitions; KATHREE = duration of monosyllable /ka/ in all three repetitions; ; PTKTHREE = duration of trisyllable /pa-ta-ka/ in all three repetitions; PTINTERV = duration of interval P-T (from beginning of explosion of sound P to beginning of explosion of sound T); TKINTERV = duration of interval T-K (from beginning of explosion of sound T to beginning of explosion of sound K); NS = nonsignificant.

Analysis of variance showed that children with dyslexia differ from the control group in 13 variables in the speed of articulatory movements. Discriminant analysis was employed to examine which of a number of variables best differentiates the groups. The 14 components were used as predictors (all variables of the speed of articulatory movements) in the discriminant analysis to distinguish among the groups (with dyslexia and control). One significant discriminant function was evident, λ = .39, χ²(14, n = 82) = 69.75, p =.00. This function was significant and showed that there was a strong association between groups and predictors. Based on the examination of the standardized discriminant function and structure coefficients (see Table 3), the variable with the highest weight in defining the discriminant function was occlusion of voice T (standardized coefficient = 0.61), followed by duration of interval P-T (standardized coefficient = 0.40) and duration of second syllable /ta/ in three repetitions (standardized coefficient = 0.37). All coefficients were positive. Pronunciation of the second syllable of the trisyllable /pataka/ in three repetitions is clearly not loaded on the discriminant function. It is the weakest predictor, and this fact suggests that it is not associated with dyslexia but, rather, is a function of other unassessed factors. Furthermore, functions at group centroids showed negative values for normally achieving children (−1.249) and positive values for children with dyslexia (1.249). Therefore, normally achieving children would more likely have low scores during the evaluation of the articulatory speed. In contrast, children with dyslexia would more likely have high scores.

Table 3.

Discriminant Analysis: Standardized Discriminant Function and Structure Coefficients.

Speech Production Tasks	Standardized Discriminant Function Coefficient	Structure Coefficient
OCCLT	.613	.732
OCCLK	.216	.509
VOTT	.018	.453
VOTK	−.284	.461
PAMIDDLE	−.142	.335
TAMIDDLE	.365	.540
KAMIDDLE	−.069	.182
PATHREE	.345	.386
TATHREE	.107	.558
KATHREE	−.219	.357
PTKMIDDL	−.065	.056
PTKTHREE	−.492	.220
PTINTERV	.396	.623
TKINTERV	.300	.591

Discussion

The aim of the present study was to compare the speed of articulatory movements of children with dyslexia in transparent Bosnian orthography with a group of children without dyslexia matched for age. The results showed that children with dyslexia have significantly slower articulatory movements than the control group. Children with dyslexia have problems in the speed of articulation, both during pronunciation of single phonemes and during the movement of articulators from one articulatory gesture to another.

To examine the feature of the production of stop consonants in children with dyslexia, the present study compared voice onset time (VOT) in those children with dyslexia and the control group. VOT is the time interval between the burst that marks the release of the stop closure and the onset of quasiperiodicity, which reflects laryngeal vibration (Lisker & Abramson, 1967). The VOT is affected by place of articulation and voicing. Plosives exhibit distinctive acoustic events on a finer time scale. Typically, the closure interval ends in an abrupt increase in acoustic energy across the frequency range. The release interval is measured from this burst onset to the start of periodicity or to the onset of noise or silence. The duration of the release interval is called VOT in cases when periodicity is present (Stouten & Van Hamme, 2009). On a spectrogram, this is seen as the time between a sharp onset of broadband energy (the burst) and the onset of the formant transition (onset of vocal-fold vibration; Gleason & Ratner, 1998). VOT is a temporal acoustic property that specifies voicing in stop consonants. Voiced stop consonants are associated with relatively short VOT values, whereas voiceless stop consonants are associated with relatively long VOT values (Allen, Miller, & DeSteno, 2003). Voiceless stops such as (t) are characterized by VOTs that range between 40 and 100 ms (Gleason & Ratner, 1998). In this research, the mean of VOTs for children with dyslexia was 89 and 88.63 ms. If we consider previous data, we can conclude that they do not have longer VOT than it is expected, but it was significantly longer than in the control group. The performance of the children with dyslexia was inferior to that of their controls on both tasks that were measured—VOT and occlusion. The occlusion of voice T is the function that most discriminates between children with dyslexia and those without dyslexia. Post, Foorman, and Hiscock (1997) examined VOT during reading nonwords. They did not find speech production difficulties of /t/ and /d/.

Vellutino, Fletcher, Snowling, and Scanlon (2004) summarized some of the most important findings from research evaluating the hypothesized causes of dyslexia during the past four decades. They considered different researchers, including Elbro et al. (1998, as cited in Vellutino, Fletcher, Snowling, & Scanlon, 2004, p. 23), who reported that children without dyslexia in dyslexic families had deficits, relative to controls, on tests evaluating articulatory accuracy.

In the present study, the performance of the children with dyslexia was inferior to that of their controls on all tests except for the test in which we measured the duration of the second syllable of the trisyllable /pataka/ in three repetitions. We can try to explain this result with the assumption that the speed of moving articulators was not slower in children with dyslexia when it required a smaller number of movements. In the case of three repetitions of the trisyllable /pataka/ and the speed of these movements, differences between children with dyslexia and normally achieving children were significant, when more number of movements were required. Wolff, Cohen, and Drake (1984) found that adolescents with reading problems have greater difficulty rapidly sequencing syllable strings than repeating single syllables. It was argued that the threshold at which movement speed degrades timing precision for coordinated action best characterizes the motor impairment of individuals with reading problems.

Results suggest that children with dyslexia have problems in the speed of articulatory movement involved in speech production. The present findings are in line with earlier reports of a deficit in the repetition of articulatory gestures, problems in speeded articulation in both production and planning of gestures in children with dyslexia (see Fawcett & Nicolson, 2002). Our data extend these previous findings by showing that VOT and time of occlusion are effective acoustic parameters for articulation testing in children with dyslexia. Also, it has been shown that during analysis of the trisyllable /pataka/, duration of intervals of P-T and T-K should be taken into consideration because discriminant analysis showed that these variables best differentiate groups. The results also showed that articulatory speed is important to distinguish between children with dyslexia and control group for both languages with opaque orthography, and in languages with transparent orthography.

Our finding about slower articulation rate in children with dyslexia is in line with a study by Fawcett & Nicolson (1995) who found that for peg placing and articulation rate, the children with dyslexia were significantly slower than their chronological age controls and suggested that children with dyslexia have persistent, and unexpectedly severe, problems in motor skill. Snowling (1981) and Catts (1989) also showed that older children with dyslexia are slower and make more error on complex articulation tasks. This was explained by difficulties in the planning stage of speech production.

Smith, Lambrecht Smith, Locke, and Bennett (2008) examined the development of timing characteristics in early spontaneous speech of children who were later identified as having a reading disability (RD). Early speaking rate was significantly slower in the group with RD, with significantly different patterns of pausing compared with children without RD. Phonetic plans may be shorter and/or less specified in children with RD, surfacing as slow, short speaking turns with increased pausing relative to articulation.

As we previously mentioned, there is no consensus about articulation rate and differences between children with dyslexia and their controls. In the research of Di Filippo and colleagues (2006), children with reading disabilities and control children did not differ in articulation rate, and no clear deficits in articulatory speed were found in the adult group with dyslexia in the research by Pennington, Van Orden, Smith, Green, and Haith (1990). Smith and others (2006) found that children with RD showed a significantly slower speaking rate. Articulation rate did not differ significantly across groups. Results in our research are not the same as these other studies because clear differences in articulation rate can be seen from the results of analysis of variance and discriminant analysis. A possible explanation for these differences in results is that different techniques are used, and some of them are more sensitive than others.

Esther (2001) wrote about diadochokinesis (DDK) and used Lundeen’s (1950) definition as “the maximum speed of movement with which a given reciprocating act can be produced” (Esther, 2001, p. 3). In different literature, DDK analysis is considered as the measurement of rate and accuracy of placement. DDK rate is considered a maximum performance test. It can reflect diminished coordination, range, or rate of articulators in clinical neurology. DDK rate could predict articulatory performance. It might be because DDK rate is sensitive to oromotor deficits (Esther, 2001). Measures of articulatory function are completed via diadochokinetic tasks designed to measure production in various contexts (Robarge, 2009). Diadochokinetic rate is one means of assessing oral motor skills. DDK rate provides information about a person’s ability to make rapid speech movements using different parts of his mouth. In this research, we can talk about diadochokinesis for tasks that have measured the phonemes /pa/, /ta/ and /ka/ where children used the front (the lips), middle (the tip of the tongue), and back (the back of the tongue). On all tasks, children with dyslexia were significantly worse than the control group.

Differences between groups were obtained for the interval P-T and interval K-T. On both tasks, children with dyslexia had significant worse results than the control group, and discriminant analysis showed that these two tasks are two of the most important functions that discriminate between children with dyslexia and children without dyslexia.

These results are in line with previous studies. Lalain and colleagues (2003) used aerodynamic/acoustic data to explore how children with dyslexia produce bilabial stops in French within a sentence, where they occurred in two positions and in three-vowel environments. A study with French dyslexics found that even though these children were able to produce a voicing contrast for /p/ and /b/, their realizations showed different timing of articulatory movement than those of a reading-age and chronological-age matched group.

Several studies connected articulatory and phonological problems in people with dyslexia. Mann and Foy (2007) focused on speech production in relation to the development of phonological awareness. Their results suggest that although the phoneme awareness measures were not apparently related to the articulatory measures, there were consistent associations between rhyme awareness and articulation. This research can be extended with the aim of investigating the connection of current problems in articulation rate with phonological skills in children with dyslexia.

There is evidence that children with dyslexia have problems in the speed of articulatory movements involved in speech production, including the speeded production of single articulatory gestures and several articulatory gestures. There is no other visible cause for lower articulation speed except the presence of dyslexia. The results of this study confirm the previous claims about speech problems in children with dyslexia. This study provides useful information about acoustic characteristics of speech, including voice onset time, time of occlusion, and evidence about slower mobility of lips, tongue tip, and tongue back in children with dyslexia. There is also evidence about slower mobility and moving from one gesture to another in children with dyslexia. These results can lead to a conclusion that there are problems in articulatory movements included in the speech production in children with dyslexia.

Slower speed of articulatory movements in children with dyslexia indicates the need for organization in the training for improvement of these skills. Joly-Pottuz, Mercier, Leynaud, and Habib (2008) evaluated the benefit of training sensory-motor aspects of articulatory function in a group of children with dyslexia. Auditory phonological training included daily listening series of exercises, mainly containing triplets of words the dyslexics had to compare phonologically to disclose similarity between 2 of the 3 stimuli. Articulatory awareness exercises put emphasis on phonetic oppositions between voiced and voiceless stop consonants. In the part of the training on the phoneme production with computerized visual feedback, children had to pronounce a chosen phoneme through a microphone, while the program analyzed the acoustic quality of the production; the utterance was then transformed into a visual index. The program was also used for improvement in the accuracy of phoneme production; development of skills in pronouncing a sequence of different phonemes; and by using a specially designed 2- and 4-phoneme contrast to improve the accuracy in contrasting phonemes. Results showed that this training is an efficient way for improving poor representation of phonemes.

It is important for speech therapists in Bosnia and Herzegovina, who provide support for children with dyslexia, to include in their work, among other interventions, remediation tools for the improvement of articulatory function. Further research is needed to explore the effectiveness and usefulness of such training in work with children with dyslexia for Bosnian and other languages with transparent orthography.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Allen

J. S.

Miller

J. L.

DeSteno

(2003). Individual talker differences in voice-onset time. Journal of the Acoustical Society of America, 113, 544–552.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Catts

H. W.

(1986). Speech production/phonological deficits in reading-disordered children. Journal of Learning Disabilities, 19, 504–508.

Catts

H. W.

(1989). Speech production deficits in developmental dyslexia. Journal of Speech and Hearing Disorders, 54, 422–428.

Caylak

(2010). The studies about phonological deficit theory in children with developmental dyslexia: Review. American Journal of Neuroscience, 1(1), 1–12.

Conway

T. W.

(2003). Measuring phonological processing and phonological working memory in adults with developmental dyslexia: A functional magnetic resonance imaging study. Unpublished doctoral dissertation, University of Florida.

De Bree

E. H.

(2007). Dyslexia and phonology: A study of the phonological abilities of Dutch children at-risk of dyslexia. Utrecht, Netherlands: LOT.

Di Filippo

Brizzolara

Chilosi

De Luca

Judica

Pecini

. . . Zoccolotti

(2006). Naming speed and visual search deficits in readers with disabilities: Evidence from an orthographically regular language (Italian). Developmental Neuropsychology, 30, 885–904.

Duranovic

(2003). Osobitosti glasa i govora osoba s Parkinsonovom bolescu i multiplom sklerozom [Characteristics of voice and speech in persons with Parkinson disease and multiple sclerosis]. Magistarski rad, Edukacijsko-rehabilitacijski fakultet, Sveuciliste u Zagrebu.

10.

Duranovic

Zecic

(2001/2002). Brzina pokretljivosti artikulatora kod djece sa cerebralnom paralizom [Speed of articulatory movement in children with cerebral palsy]. Defektologija, 7, 95–100.

11.

Esther

T. S. L.

(2001). The relationship between diadochokinetic rate and accuracy, reading rate, and sentence intelligibility in Cantonese speakers with Parkinsonism. Unpublished manuscript, University of Hong Kong.

12.

Fawcett

Nicolson

(2004). Dyslexia: The role of the cerebellum. Electronic Journal of Research in Educational Psychology, 2(2), 35–58.

13.

Fawcett

Nicolson

(2004). Dyslexia Screening Test–Junior (DST-J). London, England: Psychological Corporation.

14.

Fawcett

A. J.

Nicolson

R. I.

(1995). Persistent deficits in motor skill of children with dyslexia. Journal of Motor Behavior, 27, 235–240.

15.

Fawcett

A. J.

Nicolson

R. I.

(2002). Children with dyslexia are slow to articulate a single speech gesture. Dyslexia, 8, 189–203.

16.

Furlan

(1965). Jednominutni ispit glasnog citanja [One-minute reading aloud]. Zagreb, Croatia: Skolska knjiga.

17.

Gerrits

De Bree

(2009). Early language development of children at familial risk of dyslexia: Speech perception and production. Journal of Communication Disorders, 42, 180–194.

18.

Gleason

J. B.

Ratner

N. B.

(1998). Psycholinguistics. Belmont, CA: Thomson Learning, Inc.

19.

Goulandris

(2003). Dyslexia in different languages: Cross-linguistic comparisons. London, England: Whurr.

20.

Griffiths

Frith

(2002). Evidence for an articulatory awareness deficit in adult dyslexics. Dyslexia, 8, 14–21.

21.

Gupta

Jamal

(2006). An analysis of reading errors of dyslexic readers in Hindi and English. Asia Pacific Disability Rehabilitation Journal, 17(1), 73–86.

22.

Hedjever

(1996). Akusticka analiza vremenskih segmenata normalnog i poremecenog govora [Acoustic analysis of time segments of normal and disordered speech]. Doktorska disertacija, Edukacijsko-rehabilitacijski fakultet, Zagreb, Croatia.

23.

Heilman

K. M.

Voeller

Alexander

A. W.

(1996). Developmental dyslexia: A motor-articulatory feedback hypothesis. Annals of Neurology, 39, 407–412.

24.

Huang

D. Z.

Lin

(1995). Dr. Speech for Windows (Version 4): User’s guide. Seattle, WA: Tiger Electronics.

25.

Joly-Pottuz

Mercier

Leynaud

Habib

(2008). Combined auditory and articulatory training improves phonological deficit in children with dyslexia. Neuropsychological Rehabilitation, 18, 402–429.

26.

Kamhi

A. G.

Catts

H. W.

Mauer

(1990). Explaining speech production deficits in poor readers. Journal of Learning Disabilities, 23, 631–636.

27.

Kasselimis

D. S.

Margarity

Vlachos

(2008). Cerebellar function, dyslexia and articulation speed. Child Neuropsychology, 14, 303–313.

28.

Lalain

Joly-Pottuz

Nguyen

Habib

(2003). Dyslexia: The articulatory hypothesis revisited. Brain and Cognition, 53, 253–256.

29.

Lisker

Abramson

A. S.

(1967). Some effects of context on voice onset time in English stops. Language and Speech, 10, 1–28.

30.

Mann

V. A.

Foy

J. G.

(2007). Speech development patterns and phonological awareness in preschool children. Annals of Dyslexia, 57, 51–74.

31.

Matanovic-Mamuzic

(1982). Lista sa serijama rijeci za ispitivanje citanja [Word list for checking reading skills]. In: Teškoce u citanju i pisanju [Reading and writing difficulties]. Zagreb, Croatia: Skolska knjiga.

32.

Nicolson

R. I.

Fawcett

A. J.

(2005). Developmental dyslexia, learning and the cerebellum. In Fleishhacker

W. W.

Brooks

D. J.

(Eds.), Neurodevelopmental disorders (pp. 19–36). New York, NY: SpringerWien.

33.

Pennington

B. F.

Van Orden

G. C.

Smith

S. D.

Green

P. A.

Haith

M. M.

(1990). Phonological processing skills and deficits in adult dyslexics. Child Development, 61, 1753–1778.

34.

Post

Y. V.

Foorman

B. R.

Hiscock

(1997). Speech perception and speech production as indicators of reading difficulty. Annals of Dyslexia, 47, 1–27.

35.

Robarge

(2009). Communication management in amyotrophic lateral sclerosis: The role of the speech-language pathologist during disease progression. The Internet Journal of Allied Health Sciences and Practice, 7(2), 1–6.

36.

Smith

A. B.

Lambrecht Smith

Locke

J. L.

Bennett

(2008) A longitudinal study of speech timing in young children later found to have reading disability. Journal of Speech, Language, and Hearing Research, 51, 1300–1314.

37.

Smith

A. B.

Roberts

Smith

S. L.

Locke

J. L.

Bennett

(2006). Reduced speaking rate as an early predictor of reading disability. American Journal of Speech-Language Pathology, 15, 289–297.

38.

Snowling

(1981). Phonemic deficits in developmental dyslexia. Psychological Research, 43, 219–234.

39.

Stouten

Van Hamme

(2009). Automatic voice onset time estimation from reassignment spectra. Speech Communication, 51, 1194–1205.

40.

Vellutino

F. R.

Fletcher

J. M.

Snowling

M. J.

Scanlon

D. M.

(2004). Specific reading disability (dyslexia): What have we learned in the past four decades? Journal of Child Psychology and Psychiatry, 45, 2–40.

41.

Vladisavljevic

(1981). Poremecaji izgovora [Articulation disorders]. Beograd, Serbia: Privredni pregled.

42.

Wolff

P. H.

(2002). Timing precision and rhythm in developmental dyslexia. Reading and Writing, 15, 179–206.

43.

Wolff

P. H.

Cohen

Drake

(1984). Impaired motor timing control in specific reading retardation. Neuropsychologia, 22, 587–600.

44.

Wolff

P. H.

Michel

G. F.

Ovrut

(1990). The timing of syllable repetitions in developmental dyslexia. Journal of Speech and Hearing Research, 33, 281–289.