• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Temporal and aerodynamic aspects of velopharyngeal coarticulation: Effects of age, gender and vowel height

ProQuest Dissertations and Theses, 2011
Dissertation
Author: Fadwa Ahmad Khwaileh
Abstract:
Previous studies on the normal patterns of velopharyngeal coarticulation did not provide a multidimensional description of the phenomenon. The primary objective of this study was to determine the effects of age, gender and vowel height on the temporal and aerodynamic aspects of nasal airflow segments related to velar coarticulation in the normal speech of children and adults. A secondary objective was to determine the within speaker variability of the segments. Speakers consisted of 20 children between the ages of 5 and 7 years, 20 children between 9 and 11 years and 20 adult speakers 18 years or older. Nasal and oral air flows were collected from the participants using partitioned oro-nasal masks during the production of vowel-nasal-vowel sequences (VNV) including /ini/ and /ana/ embedded in two carrier phrases. Temporal and aerodynamic measurements were obtained for anticipatory and carryover nasal airflow for (VNV) sequences including absolute (in seconds) and proportional duration, as well as the volume of nasal airflow (in milliliters) and the ratio of nasal to oral-plus-nasal airflow volume. A mixed design 3 x 2 x 2 x 2 ANOVA procedure was used to determine the effects of age group, gender, vowel height and production level (type of carrier phrase) on temporal and aerodynamic aspects of anticipatory and carryover nasal airflow. In addition, coefficient of variation (CV) was computed for both temporal and aerodynamic measures as an index to speaker's variability. Group Analysis of Variance 3 x 2 ANOVA procedures were used to determine the effect of age group, gender, or both on within speaker variability for all temporal and aerodynamic measurements. The results of the analysis suggest a significant age effect ( p < .001) on both temporal aspects and on the absolute volume (ml) of anticipatory nasal airflow. Duration, absolutely (sec) and proportionally, and volume of nasal airflow (ml) decreased with increasing age. No significant age effect was found for carryover nasal airflow. However, a significant interaction between gender and vowel height was found. Female speakers produced longer duration than male speakers on high vowel contexts, and women produced greater volume of nasal airflow (ml) and greater ratio of nasal to oral-plus-nasal airflow. A significant production level effect was also found. Generally, all speakers exhibited reduced absolute (sec) and proportional duration as well as reduced nasal airflow volume (ml) when the carrier phrase contained `say' preceding the VNV sequence compared to the one without 'say'. Results of the CVs analysis showed main effect of age as well as age and gender interaction. Results indicate a reduction on variability with increasing age. Older boys and men exhibited greater variability than older girls and women particularly on high vowel context. Results of the study indicate that children and adults produce distinct patterns of temporal and aerodynamic aspect of anticipatory nasal airflow. Findings were consistent with previous studies that reduction in duration of speech segment and reduction in variability is a general pattern of speech development. It is also suggested that subtle gender differences in oral-pharyngeal anatomy as well as vowel-specific production patterns may explain the gender difference on high vowels. Results of the study were discussed in the light of Gestural Phonology view of speech development and velar movement. Clinical implications were suggested for the diagnosis and treatment of patients with velopharyngeal dysfunction.

vii TABLE OF CONTENTS

CHAPTER 1. INTRODUCTION .....................................................................................1

CHAPTER 2. LITERATURE REVIEW .........................................................................5

Theories of Velopharyngeal Coarticulation .....................................................................5

Developmental Patterns of Coarticulation .......................................................................8

Velopharyngeal Function and Gender Differences .......................................................12

Velopharyngeal Coarticulation and Vowel Height ........................................................14

Speech Aerodynamics ....................................................................................................18

Statement of the Problem ...............................................................................................20

CHAPTER 3. METHODS...............................................................................................22

Participants .....................................................................................................................22

Speech Sample ...............................................................................................................23

Instrumentation ..............................................................................................................24

Data Segmentation and Data Analysis ...........................................................................24

Statistical Analysis .........................................................................................................26

CHAPTER 4. RESULTS .................................................................................................27

Descriptive Statistics ......................................................................................................27

Statistical Results ...........................................................................................................40

Temporal Measurements ............................................................................................40

Aerodynamic Measurements......................................................................................49

Speakers’ Variability .....................................................................................................65

CHAPTER 5. DISCUSSION ..........................................................................................71

Age Effect ......................................................................................................................71

Gender and Vowel Height Effects .................................................................................74

Effect of Production Level .............................................................................................77

viii Speakers’ Variability .....................................................................................................77

Clinical and Research Implications ...............................................................................78

Conclusions ....................................................................................................................80

LIST OF REFERENCES ................................................................................................81

APPENDIX A. CONSENT FORM ................................................................................90

APPENDIX B. CHILD ASSENT....................................................................................96

APPENDIX C. SUBJECT BACKGROUND FORM ...................................................97

APPENDIX D. LANGUAGE ITEMS ............................................................................99

VITA................................................................................................................................101

ix LIST OF TABLES

Table 1.

Age (years;months) and gender distribution of the speakers. ......................22

Table 2.

Descriptive statistics for the absolute duration of anticipatory nasal airflow in seconds as a function of age group, gender, and vowel height and production level. ....................................................................................28

Table 3.

Descriptive statistics for the absolute duration of carryover nasal airflow in seconds as a function of age group, gender, vowel height and production level. ...........................................................................................29

Table 4.

Descriptive statistics for the proportional duration of anticipatory nasal airflow as a function of age group, gender, vowel height and production level. .............................................................................................................31

Table 5.

Descriptive statistics for the proportional duration of carryover nasal airflow as a function of age group, gender, vowel height and production level. .............................................................................................................32

Table 6.

Descriptive statistics for the absolute volume of anticipatory nasal airflow in milliliters as a function of age group, gender, vowel height and production level. ....................................................................................34

Table 7.

Descriptive statistics for the absolute volume of carryover nasal airflow in milliliters as a function of age group, gender, vowel height and production level. ...........................................................................................36

Table 8.

Descriptive statistics for the ratio of nasal to oral-plus-nasal airflow volume related to anticipatory nasal flow segment as a function of age, gender, vowel type and production level. ....................................................37

Table 9.

Descriptive statistics for the ratio of nasal to oral-plus-nasal airflow volume related to carryover nasal flow segment as a function of age, gender, vowel type and production level. ....................................................39

Table 10.

Results of 2 x 2 x 2 x 3 ANOVA for the absolute duration of anticipatory nasal airflow. ............................................................................41

Table 11.

Results of 2 x 2 x 2 x 3 ANOVA for the absolute duration of carryover nasal airflow. ................................................................................................45

x Table 12.

Results of 2 x 2 x 2 x 3 ANOVA for the proportional duration of anticipatory nasal airflow. ............................................................................48

Table 13.

Results of 2 x 2 x 2 x 3 ANOVA for the proportional duration of carryover nasal airflow. ................................................................................52

Table 14.

Results of 2 x 2 x 2 x 3 ANOVA for the absolute volume (ml) of anticipatory nasal airflow. ............................................................................55

Table 15.

Results of 2 x 2 x 2 x 3 ANOVA for the absolute volume (ml) of carryover nasal airflow. ................................................................................59

Table 16.

Results of 2 x 2 x 2 x 3 ANOVA for the ratio of nasal to oral-plus-nasal airflow related to anticipatory nasal airflow. ...............................................61

Table 17.

Results of 2 x 2 x 2 x 3 ANOVA for the ratio of nasal to oral-plus-nasal airflow related to carryover nasal airflow. ...................................................64

Table 18.

CVs of the absolute duration of anticipatory and carryover nasal airflow as a function of age group and gender. ........................................................68

Table 19.

CVs of the absolute volume of anticipatory and carryover nasal airflow as a function of age group and gender. ........................................................69

xi LIST OF FIGURES

Figure 1.

"Right angle" female and "acute" male velar configuration with broken lines showing the velum at rest and arrows showing direction of movement. ....................................................................................................13

Figure 2.

Oral and nasal air flow rates and articulatory positioning for the phrase /se nip/: V-H, velar height; V-P, velopharyngeal distance; and T-C tongue constriction. ......................................................................................17

Figure 3.

Diagrammatic representation of the artificial model of the vocal apparatus used to calculate VP orifice size. .................................................19

Figure 4.

Example of nasal flow, oral flow and voice signal for an adult female production of /ini/. ........................................................................................25

Figure 5.

Absolute duration (sec) of anticipatory nasal airflow during the production of /ana/ as a function of age group and gender of the speakers. .......................................................................................................42

Figure 6.

Absolute duration (sec) of anticipatory nasal airflow during the production of /ini/ as a function of age group and gender of the speakers. .......................................................................................................43

Figure 7.

Absolute duration (sec) of anticipatory nasal airflow as a function of age group and production level...........................................................................44

Figure 8.

Absolute duration (sec) of carryover nasal airflow as a function of age group and vowel type. ..................................................................................46

Figure 9.

Absolute duration (sec) of carryover nasal airflow as a function of age group and production level...........................................................................47

Figure 10.

Proportional duration of anticipatory nasal airflow during the production of /ana/ as a function of age group and gender of the speakers. ..................50

Figure 11.

Proportional duration of anticipatory nasal airflow during the production of /ini/ as a function of age group and gender of the speakers. ....................51

Figure 12.

Proportional duration of carryover nasal airflow as a function of vowel type and age group of the speakers. .............................................................53

xii Figure 13.

Proportional duration of carryover nasal airflow as a function of vowel type and production level. ............................................................................54

Figure 14.

Absolute volume (ml) of anticipatory nasal airflow as a function of age group and production level...........................................................................56

Figure 15.

Absolute volume (ml) of anticipatory nasal airflow as a function of gender and production level. ........................................................................58

Figure 16.

Absolute volume (ml) of carryover nasal flow as a function of vowel type and production level. ............................................................................60

Figure 17.

Ratio of anticipatory nasal to oral-plus-nasal airflow as during the production of /ana/ as a function of age and gender of the speakers. ..........62

Figure 18.

Ratio of anticipatory nasal to oral-plus-nasal airflow volume during the production of /ini/ as a function of age group and gender of the speakers. .......................................................................................................63

Figure 19.

Ratio of carryover nasal to oral-plus-nasal airflow during the production of "say /ini/ again” as a function of age and gender of the speakers. ...........66

Figure 20.

Ratio of carryover nasal to oral-plus-nasal airflow during the production of "/ini/ again" as a function of age group and gender of the speakers. .......67

1 CHAPTER 1. INTRODUCTION

As soon as technical tools were introduced to study speech production, speech became recognized as a continuum of overlapping articulatory movement (Kent & Minifie, 1977). This gave rise to the concept of coarticulation which refers to the phenomenon in speech production where the articulatory and the acoustic properties of sounds are affected by other contiguous sounds. Nowadays, there is a general consensus that coarticulation is responsible for the lack of acoustic and articulatory invariants for speech sounds.

Coarticulatory effects are generally described according to the direction in which they occur. Forward or anticipatory coarticulation occurs when the acoustic and articulatory characteristics of a sound are affected by those of a subsequent sound. Benguerel & Cowan (1974) provided an example of anticipatory lip protrusion in French. They showed that the lip protrusion for the second vowel in VCC ... CV sequence starts as early as the first consonant of a cluster of 4-6 consonants. They even observed the lip rounding gesture to begin, in some cases, during the unrounded vowel preceding the cluster.

This observation of anticipatory coarticulation suggests an active neurological strategy in which the system of motor control has information about all of these segments beforehand (MacNeilage & DeClerk, 1969; Daniloff & Hammarberg, 1973).

Preservatory or carryover coarticulation occurs when an articulatory adjustment for a segment appears to carryover to a later one. For example, in the word ‘boots’, lip protrusion associated with the vowel /u/ is retained in the following segments /t/ and /s/ (Kent & Minifie, 1977). Carryover coarticulation is widely believed to be inevitable and caused by mechano-inertial forces acting upon the articulators in motion (Turnbaugh, Hoffman, & Daniloff, 1985; Sereno, Baum, Marean, & Lieberman, 1987; Katz, Kripke, & Tallal, 1991).

The study of the coarticulatory patterns of speech movements is fundamental to our knowledge of articulatory and acoustic properties of speech. In addition, the study of the temporal and spatial aspects of coarticulation may provide valuable information about the size of the organizational units of speech production (Bell-Berti, Krakow, Gelfer & Boyce, 1995). For these purposes, numerous studies have been undertaken to investigate the coarticulatory patterns of several speech sub-systems including lingual, labial and velopharyngeal.

The velopharyngeal port (VP) is a valve located between the oral and nasal portions of the supralaryngeal tract. The primary function of this valve is to control the degree of oral and nasal coupling for normal production of speech. When the VP valve is tightly closed, it allows for sufficient oral air pressure and airflow for the production of hypernasal free obstruents and oral vowels, and when it is open, air will flow through the nasal chamber resulting in the production of nasal consonants and nasalized vowels. Thus, effective control of the degree of oral-nasal coupling is important for the development of intelligible speech.

2 Velopharyngeal coarticulation (VPC) refers to the influence of a nasal consonant onto the preceding segment(s) (i.e., anticipatory) or the following segment(s) (i.e., carryover) during continuous speech. Several articulatory studies employing different observation techniques have documented the existence of a contextual effect of the nasal sound on the adjacent vowels (e.g., Moll, 1962; Lubker, 1968; Ushijima & Swashima, 1972; Kent, Carney & Severeid, 1974). These studies detected a lower position of the velum during the production of a vowel before and after nasal consonants than for vowels near non-nasal consonants. Specifically, Kent et al. (1974) found that in vowel-nasal-oral consonant (VNC) sequences, lowering the velum began during the tongue movement toward the position of the vowel. This suggests that the velopharyngeal port is already open when the oral tract is constricted by the tongue for /n/ or lips for /m/. They also found that in nasal-vowel-oral consonant (NVC) sequences, velum elevation started during or after the constriction of the oral tract for the nasal consonant. This pattern of velar movement causes the vowels before and after the nasal consonant to be nasalized for a certain duration of time.

In addition, studies observed that the velum position is influenced by the tongue height in both oral and nasal contexts. In oral contexts, high vowels are produced with higher velum position and greater VP closure force than are low vowels (Bell-Berti, 1976; Kuehn & Moon, 1998). In nasal contexts, high vowels are produced with a higher velum position and less VP opening than low vowels (Moll, 1962; Clumeck, 1976; Al- Bamerni, 1983). Therefore, the degree of nasality is not only related to the degree of oral- nasal coupling but it is also highly dependent upon the oral configuration.

Aerodynamic data have attracted a considerable amount of interest in the study of nasalization for many reasons. First, it is a non-invasive technique that allows for the collection of a greater amount of data than other observation methods. Second, variation in pressure and flow patterns can allow inferences about articulatory positions associated with the production of various sounds. For instance, variation in nasal flow was found to be related to the degree of VP opening as well as the tongue position (Lubker & Moll, 1965; Warren, Dalston, Trier & Holder, 1985; Warren, Dalston & Mayo, 1993).

In a clinical setting, understanding the extent of velar coarticulation is critical for distinguishing between normal and abnormal degrees of nasality. Such distinction is necessary if appropriate intervention decisions for velopharyngeal insufficiency (VPI) are to be made. Pressure-flow data have proven useful in the distinction between adequate and inadequate VP function. For instance, estimates of the size of the VP orifice during the production of /p/ in “hamper” have been suggested as a primary aerodynamic diagnostic procedure to describe the adequacy of the VP valve (Dalston & Warren, 1986). However, temporal aspects of nasal flow events have been shown to more accurately describe varying degrees of hypernasality seen in patients with VPI (Warren, Dalston, Morr, Hairfield & Smith, 1989; Warren, Dalston & Mayo, 1993) .

Consequently, numerous studies have investigated the temporal and aerodynamic aspects of the VP mechanism during speech production in non-cleft individuals (Zajac & Mayo, 1996; Zajac, 1997, 2000; Leeper, Tissington & Munhal, 1998; Zajac & Hacket,

3 2002). The impetus of these studies was to establish normative standards of pressure-flow characteristics that can be useful for the assessment of VP adequacy. However, these studies were only concerned with VP movement and timing patterns in the /mp/ sequence in “hamper”. Studies involving speech samples that allow for the description of aerodynamic and temporal patterns of both anticipatory and carryover VPC are still sparse.

Several acoustic and kinematic studies have investigated the emerging speech patterns of children (e.g., Smith, 1978; Sharkey & Folkins, 1985; Nittrouer, 1993; Smith, Goffman & 1998). These studies concluded that children’s speech is characterized by less stability, decreased rate of speech production and increased variability. As the child grows, these qualities gradually diminish reflecting the improvement occurring with neuromotor maturation.

Many studies have investigated age-related differences in the extent and degree of coarticulation (e.g.,

Repp, 1986; Sereno & Lieberman, 1987; Hodge, 1989; Nittrouer, Studdert-Kennedy & McGowan, 1989; Katz, Kripke, & Tallal, 1991). These studies have demonstrated mixed results concerning whether children show greater, less or equal gestural overlap than adults. The conflicting results can be attributed to several factors including use of different measures, analysis of different articulatory subsystems, varied phonetic composition of syllables, and differences in the length and complexity of utterances (Nittrouer, 1993). However, several of these studies reported increased variability in children’s coarticulation patterns compared to adults.

Most studies conducted on coariculatorry patterns of children’s speech have focused on labial and lingual coarticulation; few studies have investigated the effect of age on the extent of VPC (Thompson & Hixon, 1979; Flege, 1988; Ha & Kuehn, 2006). These few studies have not established consistent developmental patterns. For example, Thompson and Hixon (1979) reported an increase of anticipatory nasal airflow with increasing age. In an acoustic study , Flege

(1988) reported no differences between adults and children age 5-10 in the duration of the acoustic signal related to anticipatory VP opening; whereas, another acoustic study (Ha & Keuhn, 2006) showed that children demonstrated longer durations of the nasal acoustic signal associated with anticipatory VPC than did adults during the production of /pamap/, /pimip/, and /pumup/. It seems that it is still premature to have a general agreement about the developmental aspects of VPC; therefore,

further investigations employing other experimental techniques and speech tasks are needed.

Relative to gender differences, aerodynamic studies again report inconsistent outcomes regarding the differences in the degree of anticipatory nasal airflow between male and female speakers. For instance, Thompson and Hixon (1979) reported that more female speakers demonstrated nasal airflow at the midpoint of i 1 in /ini / than did male speakers. On the other hand, no gender differences were reported in the magnitude of anticipatory nasal airflow (Hoit, Watson, Hixon, McMahon, & Johnson, 1994), or ratios of nasal to oral-plus-nasal flow and sound pressure levels (SPL) (Zajac, Mayo & Kataoka, 1998) at the midpoint of the i 1 in /ini/. These studies have only reported

4 aerodynamic measurements. It is possible that the gender factor may have an effect on the temporal aspects of nasal airflow segments related to VPC.

Studies examining the extent of VPC have investigated the phenomenon from a single diamention such as the duration of the acoustic signal related to VPC (Flege, 1988; Ha & Keuhn, 2006) or the aerodynamic aspects of VPC (Thompson & Hixon, 1979; Zajac, Mayo & Kataoka, 1998). According to Bell-Berti (1993), a complete description of velopharyngeal motor control requires examination of segmental, intersegmental, and contextual factors associated with speech; therefore, the present study is designed to obtain simultaneous temporal and aerodynamic characteristic of anticipatory and carryover nasal airflow to investigate age and gender as well as vowel height effects on temporal and aerodynamic aspects of nasal airflow segments related to VPC in vowel- nasal-vowel (VNV) sequences. Another purpose is to determine the within speaker variability for those temporal and aerodynamic parameters. Such efforts would add to our knowledge of normal range and extent of VPC and would be of great value for the clinical assessment of disordered speech resonance.

5 CHAPTER 2. LITERATURE REVIEW

Theories of Velopharyngeal Coarticulation

Findings of studies suggesting that coarticulation is not restricted and may extend to several segments in advance (e.g. Benguerel & Cowan ,1974) indicate it is not merely a result of mechano-inertial forces acting on the articulators but rather involves an active neurological control system that has information about several segments in advance (MacNeilage & DeClerk, 1969; Daniloff & Hammarberg, 1973; Kent & Minifie, 1977); therefore, a large amount of research has been undertaken during the last few decades to develop theories and models to account for the phenomenon of coarticulation.

According to Farnetani and Recasens (1999), the temporal domain of coarticulation refers to how far and to which direction coarticulatory effects can extend in time. It is believed that the temporal and the spatial aspects of coarticulation are cruicial to the testing of coarticulation theories (Farnetani & Recasens, 1999). It is also believed that the extent of the coarticulatory influence provides information about the size and nature of the organizational units of speech production (Bell-Berti et al., 1995).

Data on the temporal extent of anticipatory velar coarticlation as well as labial coarticulation have been interpreted in light of two theoretical frameworks, the featural phonology model (Moll & Daniloff, 1971; Daniloff & Hammarberg, 1973; Hammarberg, 1976) and the gestural phonology or coproduction model ( Fowler, 1980, 1992; Fowler, Rubin, Remez & Turvey, 1980; Bell-Berti and Harris 1981, 1982; Kelso, Tuller, Vatikiotis-Bateson & Fowler, 1986; Browman and Goldstein, 1989; Saltzman and Munhall, 1989; Fowler & Saltzman, 1993).

After the introduction of distinctive features (Jakobson, Fant & Hallé, 1952; Chomsky & Hallé, 1968), the feature spreading theory was proposed to account for the coarticulatory variability of speech production. This theory proposes discrete features associated with phonemes as input units of speech. These features constitute the properties that specify the identity of each phonological segment.

To explain the observation that anticipatory coarticulation is not a result of mechano-inertial constrains of the speech apparatus, proponents of this theory considered coarticulation as a left-to- right array of features that occurs between the segments at a higher (phonological) level before the command is issued to the articulators. According to this theory, the essential properties of a segment are altered and modified due to the influence of neighboring segments. These modifications occur to smooth out transitional vocoids between speech segments (Hammarberg, 1976).

Moll and Shriner (1967) and Moll and Daniloff (1971) adopted a binary feature specification model developed by Henke (1966) to account for the timing of anticipatory coarticulation. This model assumed that a “look ahead” procedure allows the “features” or “goals” of upcoming phonemes to influence those of the current phonemes as long as the anticipated goals are not in conflict with articulatory requirements of the more

6 immediate goals. According to this model, the phonological feature [nasal] has three specification values, [-] for obstruents (i.e., produced with high velar position), [+] for nasal consonant (i.e., produced with low velar position and [0] for all other segments including vowels (i.e., neutral). Anticipatory VPC occur as vowels preceding a nasal consonant equally assume a low velar position. Based on this model, anticipatory VPC in CVnN would extend as a function of the number and duration of the preceding unspecified segments (Daniloff & Hammarberg, 1973).

Kent, Carney, and Severeid (1974) tested the prediction of the binary control model of the velar movement. They found that the model failed to account for all patterns of movement and timing of the velum in English. For example, they demonstrated that in a sentence containing vowels and nasals like

many a man knew my meaning, the binary model would predict only one velum position; however, speakers actually demonstrated velar elevation gesture during the production of the vowels. They maintain that even though a binary model would seem attractive and parsimonious, it is not sensitive to fine temporal pattern of articulatory movement.

On the other hand, the gestural phonology line of thought proposes gestures as the fundamental invariant units of production and perception. These gestures are articulatory movements with specified dynamic and temporal structures. The dynamic specification of the gestures determines the kinematics of speech movements and the temporal structure allows them to overlap in time when executed. In other words, coarticulation does not arise from articulatory adjustments between neighboring segments, but from the coproduction, or temporal overlap, of invariant neighboring gestures (Fowler, 1980; Fowler & Saltzman, 1993).

Citing data from anticipatory lip rounding and velar lowering, Bell-Berti and Harris (1979, 1981, and 1982) proposed

a time-locked model or frame model to identify the extent of coarticulation. The model asserts that the articulatory period of a segment is longer than its acoustic period; as a result, the articulatory movements begin before and end after the acoustic period of the segment (Bell-Berti & Harris, 1981). Anticipatory coarticulation is therefore limited and does not extend very far backward in time before the gesture becomes acoustically dominant. Thus, neither the length of the preceding string of phones nor their conflicting or non-conflicting featural specifications are relevant.

Indeed, Bell-Berti and Krakow (1991) attempted to explain findings of Moll and Danillof (1971). They obtained kinematic and acoustic data for the production of CVnN utterances such as “a ansal” and “say ansal”. They observed monophasic and biphasic velum lowering patterns in their test utterances. The monophasic pattern was observed when the vowel sequence was short and the biphasic pattern was observed when the vowel sequence was long. In addition, they included control utterances containing oral consonants (CVnC). They found that the velum lowers after the oral consonant in a pattern similar to the first stage of the biphasic lowering and the initial portion of the monophasic lowering observed in the test utterances. They concluded that this movement of the velum cannot be ascribed to the nasal consonant.

7 The findings of Bell-Berti and Krakow (1991) appear to be in accordance with the earlier articulatory studies on the movement patterns of the velum that showed that in the absence of nasal consonant, vowels and oral consonants are inherently associated with different velum heights (e. g., House & Stevens, 1956; Moll, 1962; Fritzell, 1969; Ushijima & Swashima, 1972; Bell-Berti, 1980; Henderson, 1984).

It seems that Moll and Daniloff (1971) considered the low position of the velum during the production of vowel string in CVnN sequence to be part of the gesture of anticipatory velar lowering and failed to capture the observation that this pattern of velum movement is inherently associated with the articulation of the vowel.

In light of the idea that gestures, the articulatory movements specified in space and time, are the input units of speech production, advocates of gestural phonology proposed that these gestural units are, in fact, the basic ‘constellations’ that constitute the phonological structure of a language (Browman & Goldstein, 1986, 1989). Specifically, individual gestures are produced according to language-specific spatial and temporal rules, and then are combined in precise and consistent ways that the perceiver can recognize and can use to rebuild the phonological representation intended by the speaker (Nittrouer, 1993). This view represents a departure from the linguistic account that phoneme-sized phonetic segments are the primary units of speech production and perception.

Based on the assumption that the phonological structure is represented by units of actions of the vocal tract, gestural phonology provides an attractive view to the central issue of how children’s speech develops into mature adult-like patterns. Browman and Goldstein (1989) established that these gestures are present in the child’s repertoire in a primitive way prior to any linguistic development. As such, phonological development is viewed as harnessing these prelinguistic units of actions to be the basic units of phonological structure. Nittrouer (1996) explained that the main task for a child learning to talk involves learning to coproduce gestures in adult fashion, with the precise degree of spatiotemporal overlap that the language requires.

The view of gestural phonology represents an abandoning of the classic assumption that young children first master a repertoire of phonemes and then they establish their lexicon by forming different combinations of these abstract contrasting units (Goodel & Studdert-Kennedy, 1993). Evidence that articulatory gestures are the basic units of contrast was derived from studies of infants babbling and toddlers producing their first words. Lock (1986) noted a similarity between the toddler’s prelinguistic vocal gestures and its first few words. Vihman, Macken, Miller, Simmons & Miller, 1985) observed that consonants produced with high frequency in the child’s babbling are also present in high frequency in his/her first words. Browman and Glodstein (1989) concluded that “The child is recruiting its well-practiced action units for a new task” (p. 204). The idea that pre-linguistic gestures are employed in the service of producing early words was also proposed and supported by other studies (Locke,1983; Studdert- Kennedy, 1987; Vihman, 1991) where 'gestures' are referred to as 'articulatory routines' or ‘word recipes’.

Full document contains 114 pages
Abstract: Previous studies on the normal patterns of velopharyngeal coarticulation did not provide a multidimensional description of the phenomenon. The primary objective of this study was to determine the effects of age, gender and vowel height on the temporal and aerodynamic aspects of nasal airflow segments related to velar coarticulation in the normal speech of children and adults. A secondary objective was to determine the within speaker variability of the segments. Speakers consisted of 20 children between the ages of 5 and 7 years, 20 children between 9 and 11 years and 20 adult speakers 18 years or older. Nasal and oral air flows were collected from the participants using partitioned oro-nasal masks during the production of vowel-nasal-vowel sequences (VNV) including /ini/ and /ana/ embedded in two carrier phrases. Temporal and aerodynamic measurements were obtained for anticipatory and carryover nasal airflow for (VNV) sequences including absolute (in seconds) and proportional duration, as well as the volume of nasal airflow (in milliliters) and the ratio of nasal to oral-plus-nasal airflow volume. A mixed design 3 x 2 x 2 x 2 ANOVA procedure was used to determine the effects of age group, gender, vowel height and production level (type of carrier phrase) on temporal and aerodynamic aspects of anticipatory and carryover nasal airflow. In addition, coefficient of variation (CV) was computed for both temporal and aerodynamic measures as an index to speaker's variability. Group Analysis of Variance 3 x 2 ANOVA procedures were used to determine the effect of age group, gender, or both on within speaker variability for all temporal and aerodynamic measurements. The results of the analysis suggest a significant age effect ( p < .001) on both temporal aspects and on the absolute volume (ml) of anticipatory nasal airflow. Duration, absolutely (sec) and proportionally, and volume of nasal airflow (ml) decreased with increasing age. No significant age effect was found for carryover nasal airflow. However, a significant interaction between gender and vowel height was found. Female speakers produced longer duration than male speakers on high vowel contexts, and women produced greater volume of nasal airflow (ml) and greater ratio of nasal to oral-plus-nasal airflow. A significant production level effect was also found. Generally, all speakers exhibited reduced absolute (sec) and proportional duration as well as reduced nasal airflow volume (ml) when the carrier phrase contained `say' preceding the VNV sequence compared to the one without 'say'. Results of the CVs analysis showed main effect of age as well as age and gender interaction. Results indicate a reduction on variability with increasing age. Older boys and men exhibited greater variability than older girls and women particularly on high vowel context. Results of the study indicate that children and adults produce distinct patterns of temporal and aerodynamic aspect of anticipatory nasal airflow. Findings were consistent with previous studies that reduction in duration of speech segment and reduction in variability is a general pattern of speech development. It is also suggested that subtle gender differences in oral-pharyngeal anatomy as well as vowel-specific production patterns may explain the gender difference on high vowels. Results of the study were discussed in the light of Gestural Phonology view of speech development and velar movement. Clinical implications were suggested for the diagnosis and treatment of patients with velopharyngeal dysfunction.