Title Page
 Table of Contents

Group Title: Effects of experimental palatal Fistulas on speech and resonance
Title: Effects of experimental palatal Fistulas on speech and resonance /
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00102777/00001
 Material Information
Title: Effects of experimental palatal Fistulas on speech and resonance /
Physical Description: Book
Language: English
Creator: Richtner, Ulla E. M., 1941-
Copyright Date: 1986
 Record Information
Bibliographic ID: UF00102777
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltuf - AEK2515
oclc - 15300479

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
        Page v
    Table of Contents
        Page vi
        Page vii
        Page viii
        Page ix
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
Full Text







Copyright 1986


Ulla E.M. Richtner

To Maj and Nils


My appreciation and gratitude are expressed to my

committee chairman, Dr. William N. Williams, for suggest-

ing the topic and for kindly making available to me needed

laboratory facilities and the resources of a Biomedical

Research Support grant.

Gratitude is also extended to the other committee

members, all of whom showed trust in me and the project.

Dr. Thomas B. Abbott also has been my adviser and his

encouragement and assistance have helped me overcome many

hurdles, not only in my doctoral research but throughout

my graduate career. Dr. Kenneth R. Bzoch has shared

generously his vast knowledge in the area of cleft palate.

Dr. G. Paul Moore has provided me with invaluable clinical

insights in the area of voice disorders. Dr. Nikzad S.

Javid offered his professional skill and cooperation in

construction of the research prosthesis. Special thanks

go to Dr. Paul W. Wharton who volunteered his time and

effort and without whom this work could not have been


I owe special gratitude to Dr. Ronald G. Thomas for

his generous and invaluable help in guiding me through the

labyrinth of statistics and computer use. Special thanks

also are due the six judges who donated their time and

expertise for the perceptual listening task. I extend

especially warm thanks to Marcia Buchanan not only for her

editorial assistance but for her personal interest and con-

cern. I would also like to take the opportunity to thank

everyone from the Departments of Speech and Oral Biology

who shared their professional skills, positive attitudes,

and research equipment.

Rotary International and Ahldn-stiftelsen of Sweden

contributed to the completion of my graduate studies by

their generous stipend contribution.



ACKNOWLEDGMENTS....................................... iv

ABSTRACT.............................................. viii


I INTRODUCTION AND PURPOSE....................... 1

Introduction.. .................................. 1
Surgical Treatment of Oronasal Openings........ 5
Prosthetic Treatment of Oronasal Openings...... 9
Cleft Palate and Speech......................... 10
Velopharyngeal Insufficiency and Speech......... 10
Palatal Fistulas and Speech.................... 13
Statement of Purpose ........................... 16


Introduction ... ........................... ..... 20
Palatal Embryology and Anatomy................. 21
The Effects of Velopharyngeal Insufficiency
on Speech..................................... 31
Perceptual Judgment ............................ 34
Instrumentation .......................... 39
The Effects of Controlled Velopharyngeal
Insufficiency on Speech....................... 51
The Effects of Palatal Fistulas on Speech....... 58
Research Questions.............................. 64

III METHODS.... .................................... 68

Introduction.. .................................. 68
Materials and Procedures ............. .......... 69

IV RESULTS........................................ 88

Perceptual Measurements........................ 89
Instrumental Measurements...................... 116
Reliability. .................................. 133


Discussion ...................................
Conclusions..... ...............................


A AUDIOMETRIC EVALUATION.........................



D ARTICULATION RATING TEST.......................

E HYPERNASALITY RATING TEST......................


PERFORMANCE .................................. ..

BIBLIOGRAPHY..... .......... .......... .................

BIOGRAPHICAL SKETCH....................................














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of
the Requirements for the Degree of Doctor of Philosophy



Ulla E.M. Richtner

August, 1986

Chairman: William N. Williams
Major Department: Speech

Palatal fistulas are frequently the result of congenital

anomalies, trauma, or tissue breakdown following cleft

palate repair. It has been observed clinically that speech

articulation and voice resonance are impaired as the result

of an opening between the oral-nasal cavities. Extensive

documentation of velopharyngeal insufficiency and its
2 2
influence on speech suggest that a 10 mm to 20 mm opening

through the velar port results in significant distortion of

speech. However, little quantitative information is

available concerning similar effects of palatal fistulas.

The purpose of this study was experimentally to manip-

ulate size and location of openings through a subject's

prosthetic palatal prosthesis and to measure the effect of

these openings on the individual's speech articulation,

resonance quality, nasal airflow, intraoral air pressure,

and differential pressure. Four experimental openings
2 2 2
(fistulas) of the respective sizes 5 mm 10 mm 20 mm ,



and 30 mm2 were drilled in a removable palatal appliance

covering a subject's complete cleft of the hard and the soft

palates. The four locations of these openings were in the

anterior, middle, and posterior part of the hard palate and

one velar opening, all along the midline. The effect that

each hole size and location had on the individual's speech

and resonance was evaluated perceptually by six judges.

Nasal airflow, intraoral air pressure, and differential

pressure were measured instrumentally.

Analyses of the data revealed that both hole size and

location of a fistula determine the degree of speech impair-

ment. The findings indicate that the greatest distortion

of articulation and resonance quality was perceived by the

judges when the fistula was located in the velar port and

at the most anterior position of the hard palate (approxi-

mately at the incisive foramen). The size of a fistula

was also found to be a significant variable with greater

distortions occurring with increased openings. Even the

smallest fistula of 5 mm2 caused mild distortion to the

speech signal which was statistically significant in the

three instrumental measurements. The five perceptual tests

indicated that speech, in terms of articulation proficiency

and nasality, was significantly distorted when the fistula
2 2
size was between 10 mm and 20 mm .



The production of oral speech sounds is dependent upon

a complete separation between the nasal and oral cavities,

thus directing the vocal stream from the larynx through the

oral cavity alone. This separation between the oral and

nasal cavities is primarily anatomical. The structure

separating the two is the palate. It consists of bone,

muscle, and mucosal tissue which forms the roof of the

mouth and the floor of the nasal cavity. Complete closure

depends on the action of the soft palate or velum. The

soft palate is elevated through muscular action to make

complete contact with the posterior pharyngeal wall. When

the soft palate is in relaxed position, as during nasal

breathing or production of nasal sounds, air is directed

into the nasal cavity. Thus, an opening in this separation,

either in the stationary hard palate or in the movable soft

palate, will be detrimental to normal oral speech sound

production, since it will allow undesirable airflow into

the nasal cavity. This airflow results in a reduction of

intraoral air pressure which is an important physiological

component for the production of different speech sounds

(Brown & McGlone, 1969). Leakage of air into the nasal

cavity also causes audible, inappropriate nasal air emission

during articulation of consonant sounds. In addition, a

shift in resonance quality, due to acoustic damping related

to the coupling of the nasal cavity with the oropharyngeal

cavities, may occur which is perceived as hypernasality.

Clinical observation has revealed that the magnitude of

the effect of these two factors (hypernasality and nasal

air emission) on speech intelligibility is negatively influ-

enced by the increasing size of a palatal opening. This

means, then, that the larger the palatal opening, the

greater the degree of speech impairment. Articulation

problems likely to occur concomitant with palatal openings

are misarticulations, ranging from indistinct production

due to slight air emission to complete omission of the

sound. Distorted sound production, simple sound substitu-

tion, gross sound substitution, and complete omission of

the sound are influenced by nasal air emission and hyper-

nasality (Bzoch, 1979).

There are three conditions that could result in a

physical breakdown of the palatal structure. The first

condition is a cleft occurring in the hard and/or the soft

palate. The type of cleft is dependent upon the location

and size of the cleft itself. The most common cause of

cleft palate is when atypical embryological development

prevents normal fusion along the midline of the palate.

Congenital overt clefts of the lip and/or palate occur at

an incidence as frequently as one in every 700 live births

(Bzoch & Williams, 1979). Two other causes known to cause

cleft palate would be trauma or ablative surgery. Head

trauma, or accidents that cause orofacial injuries, such

as fractures of skull bones, may include fractures of the

palatal bones, or a cleft along the palatal midline.

Ablative surgery, or removal of tissue, is usually due

to disease typically following removal of tumors.

The second condition of palatal malfunction is velo-

pharyngeal insufficiency (VPI) which is defined as the

inability of the soft palate, or velum, to make adequate

closure with the posterior pharyngeal wall for normal speech

production. Velopharyngeal insufficiency can be a result

of congenital or acquired anomaly and the speech disorders

that typically occur can be compared to those impairments

associated with cleft palate and palatal fistulas. Velo-

pharyngeal insufficiency can also result from neurological

disorders. It is important to remember the physiological

differences between the hard and the soft palates. The

function of velopharynx involves intricate movements and

interactions of a number of muscles active in velopharyngeal

closure. The hard palate, on the other hand, does not

display any myofunctional movement since it consists mainly

of bone and tissue lining and serves as a static border

between the two cavities. It is difficult to determine

the actual incidence of VPI, as it varies with the con-

comitant disorder. The incidence of VPI, as it occurs

after primary cleft palate repair, is reported to be between

20 percent and 40 percent (Bradley, 1979).

The third condition responsible for a palatal break-

down is a fistula in the hard and/or the soft palate. An

oronasal fistula is defined, in this study, as a hole in

the palate connecting the oronasal cavities. It may develop

subsequent to surgical repair of a cleft palate. It may

also occur from traumatic perforation of ablative surgery.

Fistulas also may result from pathologic abnormalities such

as noma (gangrenious sore), syphillitic gummas, leprosy,

and leishmaniasis (infection due to the parasite Leishmania)

(Gordon & Brown, 1980).

It is clear that these three different palatal openings

or insufficiencies are almost exclusively caused by a

morphological abnormality. Only VPI can be diagnosed as

resulting from a functional disorder. Patients who suffer

from any of these abnormalities are referred for either

surgical closure of the palatal openings, reconstruction of

muscular tissue to restore velopharyngeal function, or for

prosthetic treatment. When the differential diagnosis

signals a function disorder, speech therapy should be the

treatment of choice. Sometimes a combination of treat-

ments is needed, depending on the extent of the physical

abnormality. Therefore, a need exists for evaluation of

the impact of palatal disorders on speech intelligibility,

as well as for the quantification of the effect of these

palatal insufficiencies.


Surgical Treatment of Oronasal Openings

Cleft Palate

A variety of surgical procedures is available for

palatal closure and correction of VPI. An excellent in-

depth description of several techniques in cleft palate

surgery is given by Millard (1980), and in several chapters

of Cleft Lip and Palate (Grabb, Rosenstein, & Bzoch, 1971).

Selection of the appropriate surgical technique is depen-

dent upon the pathological extent of the condition, indi-

vidual need, and optimal goal. Important information

regarding morphology, tissue reaction, and optimal timing

for surgery are taken into consideration. There are several

reasons why obtaining an adequate closure between the

oronasal cavities is important. The primary goal in cleft

palate surgery, and perhaps the task that is most diffi-

cult to achieve, is to establish normal speech production.

As a direct result of atypical airflow into the nasal

cavity, the necessary intraoral air pressure cannot be

maintained and speech is affected in the form of hyper-

nasality and nasal air emission. Normal hearing is

important for the individual who is learning how to speak.

Hearing loss that is caused primarily by otitis media

(middle ear inflammation) occurs in almost all cleft

palate individuals (Millard, 1980). The frequency of

otitis media in these patients is caused by impaired palatal


muscle activity involved in the opening and closing of the

auditory tube (Eustachian tube). This is due to upper

respiratory disease which is easily transmitted to the ear

via the Eustachian tube. The tube, being the single dynamic

connection between nasopharynx and the middle ear, serves

as an equalizer for air pressure across the tympanic mem-

brane (ear drum) and loses its function when infected

(Davis, 1970). Speech habilitation is, therefore, compli-

cated by inadequate auditory feedback. Another important

reason for surgical closure of a cleft palate is to allow

the individual normal feeding. Correction of dentition

and establishing of normal swallowing habits are, there-

fore, additional important goals in palatal surgery.

Velopharyngeal Insufficiency

Velopharyngeal insufficiency occurs when the soft

palate is unable to adequately separate nasopharynx from

oropharynx during speech. Regardless of the cause,

resulting from either morphologic or functional disorder,

VPI is an opening between the oral and the nasal cavities

which allows airflow into the nasal cavity during the

production of oral speech sounds.

Several different surgical techniques are available

for the establishment of normal velopharyngeal function.

Some examples of these procedures are palatal pushback

(a palal lengthening procedure), pharyngeal flap (a skin


graft), teflon pharyngoplasty (augmentation by injection).

A combination of these techniques may be used in order to

obtain optimal results (Millard, 1980). In addition to

surgical repair, treatment may involve prosthetics and/or

speech therapy.

Palatal Fistula

Palatal fistulas, defined holes in the palate con-

necting the oronasal cavities, may appear secondary to

cleft palate surgery or can occur as the result of trauma

or ablative surgery. The frequency of occurrence of fis-

tulas following palatal repair has been reported to range

from 9 percent to 34 percent (Henningsson, 1983; Ross &

Johnston, 1972). Abyholm, Borchgrevink, and Eskeland (1979)

report that frequency of occurrence seems to be dependent

upon surgical technique. They report an overall incidence

of 18 percent. Other frequency occurrences of fistulas

resulting from trauma or ablative surgery are not readily

available in the literature.

Surgical repair of fistulas is undertaken only after

comprehensive evaluation of the anatomical structures. If

speech is distorted and a fistula is present, it is

necessary to differentially evaluate the cause of the

distortion. By temporary obturation of the fistula, the

significance of its relationship to speech and the presence


of any velopharyngeal insufficiency can be diagnosed

(Bless, Ewanowski, & Dibbell, 1980; Bloch, 1979; O'Neal,

1971; Skolnick, Glaser, & McWilliams, 1980). If the speech

sounds normal while the fistula is obturated, it can be

concluded that the fistula is the cause of the speech dis-

tortion. However, if speech continues to be impaired, this

would be an indication of VPI. If it were determined that

fistula closure was necessary, a concomitant decision also

would be made with regard to surgical or prosthetic treat-

ment. O'Neal (1971) has argued that consideration should

be given to delaying or even eliminating the need for

surgical repair of an existing fistula in growing children

with ongoing development of the orofacial and orthodontic


A number of surgical techniques are available for

fistula closure. The preferred techniques appear to be

dependent upon fistula size and location. James (1980)

reported on buccal flap surgery to repair small fistulas.

Additionally, he also reported on the use of large buccal

flaps in conjunction with bone grafts for reinforcement

in large fistulas. He reported large fistulas to be of

0.5 centimeter and 2.5 centimeters in size. Harris (1980)

reported that pharyngoplasty was preferred when closing a

posteriorly located fistula. For anteriorly located

fistulas, Harris reports the use of bone grafts. The

bone graft is subsequently placed between flaps that are


taken from the nasal and the palatal soft tissues. Further,

a bone graft always would be indicated when the fistula

occurs in the alveolar ridge.

Prosthetic Treatment of Oronasal Openings

Individuals with incomplete palatal closure who are

not candidates for or who do not elect surgical treatment

generally are treated with prosthetics (Adisman, 1971).

Different kinds of prosthetic appliances are available,

depending upon the physical defect being alleviated by

obturation. A removable prosthetic appliance is used to

cover the complete cleft of the hard and the soft palate,

with the posterior part serving as the speech bulb.

Closure between the oral and the nasal cavities is estab-

lished as the posterior pharyngeal wall makes contact with

the speech bulb. A simpler appliance, such as a speech

bulb, is used for patients with a cleft of the soft palate

only. Velopharyngeal insufficiency can be treated

prosthetically in a similar manner, resulting in appro-

priate velopharyngeal closure. Reconstructive surgical

repair of fistulas is always the treatment of choice

(Drane, 1973), but prosthetic rehabilitation may sometimes

be indicated. A prosthetic obturation may be recommended

especially with recurring fistulas, soft palate inadequacy,

or when serving as an interim obturator while awaiting

surgery or orthodontic treatment. The traditional


prosthodontic treatment includes either a fixed or a

removable obturator (Reisberg, Gold, & Dorf, 1985). Small

fistulas that do not cause a significant speech disorder

may be temporarily obturated to prevent food and liquids

from entering the nasal cavity during eating and drinking.

These materials may include dental wax, acrylic resin, or

chewing gum (Reisberg et al., 1985).

Cleft Palate and Speech

A cleft palate does not provide the separation between

the oronasal cavities necessary for normal speech produc-

tion. Therefore, it will cause hypernasal voice quality

and excessive nasal air emission. These two factors,

hypernasality and excessive nasal air emission, negatively

affect speech intelligibility. The severity of the speech

disorder resulting from this pathology depends upon the

size of the cleft and the extent of involvement of the

cleft with tissues of the hard and/or soft palates. The

speech intelligibility of cleft palate patients is impaired

by a wide range of errors, including indistinct produc-

tion, simple and gross substitution, and complete omission

of sounds.

Velopharyngeal Insufficiency and Speech

Velopharyngeal insufficiency has been proven to cause

disorders of both nasality and intelligibility of speech,

depending on its size. Thorough investigations of the size

of such openings between the oral and nasal cavities and

their influence on speech has been published. The objec-

tive has been to quantify the magnitude of openings in the

velar port and the influence of these openings on hyper-

nasality and nasal air emission.

The earliest attempts at this quantification were

performed using direct observational techniques, such as

lateral still roentgenography or cineradiography (Benson,

1972; Bj8rk, 1961; Bzoch, 1968; Graber, Bzoch, & Aoba,

1959; Moll, 1960, 1962, 1964; Shelton, Brooks, & Youngstrom,

1964; Subtelny, Koepp-Baker, & Subtelny, 1961). Much

information, therefore, is available regarding the pattern

of velopharyngeal kinesiology and configuration during

speech. This has contributed greatly to the knowledge

base about velopharyngeal function and closure (Aram &

Subtelny, 1959; Bj6rk, 1961; McKerns, 1968/1969; Moll, 1965;

Nylen, 1961; Warren & Hoffmann, 1961). The radiologic

measures obtained were compared to indirect measures, such

as perceptual judgments of speech intelligibility and

voice quality, to acoustical measures (sound spectrography)

(Andrews & Rutherford, 1972; Bj6rk, 1961), or to measure-

ments found in the pneumatic pressure-flow (aerodynamic)

technique (Warren, 1964a,b; Warren & DuBois, 1964).

Further quantification attempts regarding velo-

pharyngeal openings led to research where the velar port


openings were experimentally controlled in normal subjects.

Different sized polyvinyl tubing was inserted through the

nose into the velopharynx to create artificial velopharyn-

geal openings (Isshiki, Honjow, & Morimoto, 1968; Bernthal

& Beukelman, 1977). Similarly, variable aperture speech

appliances with inserts were employed by Andrews and

Rutherford (1972), Liebman (1964), and Watterson and

Emanuel (1981a,b). By manipulating the velopharyngeal port

size in this manner, differing degrees of hypernasality

could be created and computations of the critical size of

openings in the velopharyngeal area performed. Generally,

these reported studies on VPI agree that distortion of

speech quality is introduced when there is an opening
2 2
between 10 mm and 20 mm in the velar port. This is

perceived as hypernasality, with audible nasal air emission

causing distortion of oral consonants.

Based on the literature, it can be concluded that the

larger the velopharyngeal opening, the more distortions

occur in speech characteristics of intelligibility and

resonance quality. However, a complete linear relation-

ship is not always present (Subtelny et al., 1961; Isshiki

et al., 1968; Andrews & Rutherford, 1972; Dalston, 1982).

Liebman (1964) found that posteriorly positioned velar

openings caused fewer speech problems than those anteriorly

positioned. The two factors of size and position have

proven important with velar insufficiencies and are


thought to play an important role with other oral-nasal

openings such as palatal fistulas.

Palatal Fistulas and Speech

Fistulas, like cleft palate and VPI, cause inadequate

closure between the oronasal cavities. Therefore, patients

having fistulas exhibit the same speech impairments as do

cleft palate and VPI patients.

Fistula Location

It is unclear as to the influence of fistula location

on speech. Ross and Johnston (1972) state that fistulas

occurring posteriorly in the hard palate or in the soft

palate cause a cleft palate-like speech, i.e., hypernasal

voice quality and audible, inappropriate nasal air emis-

sion. No clear description is offered regarding the

influence that anteriorly positioned fistulas may have

on speech. It is stated that speech may be affected

because of air escaping into the nasal cavity, influencing

both articulation and voice quality (Ross & Johnston,

1972). Cosman and Falk (1980) reported that defects of

speech production were significant in the presence of

anterior fistulas. These palatal defects located anterior

to the place of articulation of pressure consonants caused

sound substitution. For example, a /t/ sound, an unvoiced


anterior plosive, was substituted for by /k/, an unvoiced

posterior plosive, and the /d/ sound, a voiced anterior

plosive, was substituted for by /g/, a voiced positive


Fistula Size

The importance of fistula size and its influence on

speech is evident in the literature. It has been proven

clinically that speech impairment increases with fistula

size. Morley reported in 1962 that even a small fistula

may affect the development of normal speech. This occurs

because of a prevention of the build-up of intraoral air

pressure necessary for the production of oral consonants.

In the literature published twenty years later, it is

still stated that speech is increasingly and adversely

affected by palatal fistulas of increasing size (Henningsson,

1983; Shelton & Blank, 1984). This information, however,

only describes the fistulas as being either small, medium,

or large without any quantitative definitions of size.

In a review of the literature reporting on palatal

fistulas, no study was found that dealt with quantifica-

tion of palatal fistulas and their influence on speech.

There were only two different studies which defined large

versus small fistulas. Proctor (1969) defined large

oronasal fistulas as ranging from 0.5 centimeter to


2.5 centimeters, while Henderson (1982) defined small

fistulas as ranging from 0.5 centimeter to 1.5 centimeters

in width and 0.5 centimeter to 2.0 centimeters in length.

One pertinent study (Shelton & Blank, 1984) deals with

palatal fistulas and their influence on speech. Shelton

and Blank reported on six patients with oronasal fistulas.

The sizes of the fistulas were classified as small,

moderate, or large, without further specification.

Patients with a small or moderate fistula maintained

sufficient intraoral air pressure for adequate oral sound

production. Patients with large oronasal fistulas demon-

strated a reduction in intraoral air pressure. Nasal air

flow, however, was detected with all fistula sizes and

was found to generate some noise.

With this meager information and inadequate guide-

lines on palatal fistulas, there seems to be a demonstrated

need for more quantitative information regarding the effect

of fistula size and location on speech production. Know-

ledge about how differently sized and positioned fistulas

would influence speech would enhance patient advisement

regarding surgery. In addition, such knowledge would aid

in the selection of prosthetic appliances, including those

for temporary obturation of the opening while eating and

drinking. In this present study, the first steps have

been taken to quantitatively define the impact of palatal

fistula size and location on speech, as it was measured


perceptually (evaluation of articulation and nasality),

and instrumentally (nasal airflow and intraoral and dif-

ferential air pressure). It appears that this new quanti-

tative data, together with earlier well-established

qualitative information, will contribute to better and

necessary guidelines in the clinical management of patients

with palatal fistulas and speech disorders.

Statement of Purpose

The primary objective of this study was to quanti-

tatively and qualitatively analyze speech in terms of

nasality and articulation, as well as some of the associated

aerodynamic correlates in the presence of artificially

introduced openings through the hard palate leading directly

into the nasal cavity. Of particular interest is the

dependence of the speech characteristics on the size and

position of such openings. Definition of the requisite

critical size of VPI is an ongoing endeavor and is of much

interest in the literature. However, there is little

information available on defects in the hard or soft

palate anterior to the velopharyngeal port and the result-

ing impacts on speech. Although clinical observations

suggest that oronasal fistulas contribute to hypernasality

and nasal emission, there is little quantitative informa-

tion describing the influence that fistula size and location

have on speech.


This present study is an attempt to close the gap

regarding this lack of knowledge and the purpose is to

begin defining the influence of size and location of an

opening through the hard palate on speech in terms of

intelligibility and resonance quality. In order to perform

this quantification, the size and location of openings

anterior to the velopharyngeal port were experimentally

manipulated. Openings were introduced in the hard palate

part of a removable palatal prosthetic appliance (research

appliance). A 28-year-old Caucasian male with a cleft of

the hard and soft palate and who wears a removable pros-

thetic palatal appliance was the only subject in this


A combination of three locations (anterior, middle,
2 2 2
and posterior) and four sizes (5 mm 10 mm, 20 mm, and

30 mm') of these oronasal openings were tested for a series

of clinical and physiological speech tests. The clinical

tests incorporated tests of nasal air emission, articula-

tion, hypernasality, an articulation rating, and a

hypernasality rating. The nasal air emission test was

evaluated by the examiner taking a repeated measurement of

two readings. The other clinical tests were presented to

six speech-language pathologists specially trained in

evaluating articulation and hypernasality in patients with

palatal dysfunction. These tests were evaluated by

the six judges who listened to and subjectively scored


the remaining four clinical tests. The speech physiology

testing was conducted by employing an aerodynamic instrument

that computed intraoral air pressure and differential


Since previous research has been concerned basically

with velar incompetencies and with controlled velar port

openings, a replication study was conducted to assess the

influence of velopharyngeal inadequacy of speech. Openings

of the same sizes as in the experimental study were drilled

in the velar part of the speech appliance which would imply

a resemblance to VPI. The same test battery of clinical

and physiological speech tests was administered. Data

obtained from this replication study were compared to the

experimental results, as well as to previous research

reports. Such comparisons were conducted in order to

evaluate the results from the present experimental study

on oronasal openings in the velar region and previously

published results.

It is known from previous research efforts that speech

is negatively influenced by an increased size of VPI.

Furthermore, there are clinical implications stating that

increasing fistula size is detrimental to speech intel-

ligibility. Fistula location and its influence on speech

have been discussed only vaguely but it is believed that

posteriorly positioned fistulas cause more speech problems

than anteriorly positioned fistulas. It was anticipated,


therefore, that the results from this present study would

be negatively influenced also by the magnitude of the hole

size, resulting in imprecise articulation, excessive

nasality and nasal emission, and a decrease in intraoral

air pressure and differential pressure as the oronasal

openings were increased in size. It was anticipated also

that these speech variables, likewise, would be influenced

as the oronasal opening was moved posteriorly.



Maintaining appropriate intraoral air pressure during

the production of oral speech sounds is an important

physiological component of normal speech. Normal speech

production will be adversely affected when an opening

occurs between the oronasal cavities.

Oronasal openings, such as a cleft palate, VPI, or

palatal fistulas permit air to escape from the oral cavity

into the nasal cavity. This air escape results in trans-

mission of the air and sound energy into the nasal cavity,

changing the physiological event of normal speech (Daniloff,

Schuckers, & Feth, 1980; Spriestersbach & Powers, 1959).

This process results in a reduction in intraoral air

pressure and increased nasal air emission. These factors,

reduced intraoral air pressure and increased nasal air

emission, cause abnormal speech production by (a) giving

speech a hypernasal voice quality and (b) reducing articu-

lation proficiency. In addition, there are other palatal

abnormalities that often are undiagnosed at birth, but

which are discovered later, which impair normal palatal



anatomy and physiology. Such abnormalities interfere with

speech proficiency. These later diagnosed abnormalities

occur in the form of submucous clefts, cleft uvula, unusually

high and/or narrow palate, and paresis of the soft palate

muscles. Occurrence of VPI in the non-cleft population

may be due to several conditions, including a neurological

disorder, a function disorder, or tissue ablation, or

tissue reduction following a surgical procedure, such as

a tonsillectomy or adenoidectomy (Bradley, 1979). In order

to more fully understand oronasal openings and their effect

on speech, a brief description of relevant embryology and

anatomy will be addressed. A discussion of the speech

characteristics that occur with various oronasal openings

and how they are measured will follow. A review of

experimental studies on VPI will be considered, followed

by a discussion of the research questions which are the

subject of the present research.

Palatal Embryology and Anatomy

For an in-depth study of embryology and anatomy of

the palatal structures and the oronasal cavities the reader

is urged to consider the works by Bateman (1977), Brescia

(1971), Bzoch and Williams (1979), DuBruhl (1980), and

Zemlin (1968). The following material is excerpted from

the above references.


Palatal Embryology

The initial stage of formation of the embryonic face

including the primary and secondary palates can be seen as

early as in the second month of gestation. The face and

mouth including the lips and the hard palate have com-

pleted their basic development process at the end of the

third month.

The development of the orofacial structures is depen-

dent on the merging and fusion of five prominences, the

single frontal prominence, the paired maxillary prominence,

and the paired mandibular prominence. These are present

in the third week of pregnancy. The mandibular processes

merge at the end of the fourth week and give rise to the

lower lip and the chin.

The frontal prominence grows downward and forward and

becomes the frontonasal process. On either side of this

process the two olfactory pits are formed and as the

frontonasal process grows further downward these pits will

divide it into the median nasal and the two lateral nasal

processes (Bzoch & Williams, 1979). This occurs by the

end of the sixth week (Figure 1). It is the nasomedial

process that forms the midportion of the upper lip, the

medial portion of the alveolar ridge, and the primary


The primary palate is developed as the nasomedial

process grows down and back to form the upper anterior roof


Frontonasal prominence (process)

Nasomedial process

Nasolateral process
Olfactory pits

Maxillary process

Mandibular arch

6 wk (12 mm)

Figure 1. Embryo at 6 weeks (12 mm).

(a) (b)

Figure 2. Palatal structure. (a) Before fusion (7 to 8
weeks), (b) Fused palate (8 to 10 weeks).


of the mouth and it extends vertically to the floor of

the nasal cavity. This plate projects into the oral

cavity as far as to where the incisive foramen is later

located. The nasomedial process then extends laterally

to the alveolar ridge and encompasses the four incisor

teeth (Bateman, 1977). Formation of the primary and

secondary palates in the human embryo has been observed

as early as the second month of gestation (Crelin, 1976).

The secondary palate is the main part of the palate

and includes both the hard and the soft palates. It is

formed from paired shelf-like outgrowths that arise from

the maxillary processes. Further contributors are the

lateral palatine processes or the lateral palatal shelves

that grow medially and inferiorly during the eighth week.

When these palatal shelves first start to develop, the

mandible lies high in the mouth and the tongue must rise

between them. During this time the lateral processes

grow downward on each side of the tongue. In the ninth

week of development a sudden growth spurt of the mandible

increases the vertical distance between it and the

maxillary arch and the tongue therefore drops down. This

allows the palatine processes to rise up and start to

fuse. Fusion of the palates is complete between the end

of the 10th and the beginning of the 12th week (Figure 2)

(Crelin, 1976).


Palatal Anatomy and Physiology

The palatal structures, or the roof of the mouth,

consist of the hard and the soft palates. The hard palate

forms a shelf between the oral and the nasal cavities.

These cavities can be completely separated by the normal

function of the soft muscular palate. Adequate anatomy of

both palates is necessary for normal function.

The hard palate. The hard skeletal palate consists

of two sets of paired bones, the palatine processes and the

palatal bones (Figure 3). The palatine processes are

formed by horizontal, medially directed projections from

the body of the maxilla (upper jaw bones) at the boundary

between the body and the alveolar process. The alveolar

process forms the thick spongy part of the maxilla where

the teeth are positioned. The palatine processes (in

anteroposterior direction) are shorter than the maxilla

and form only about three-fourth of the hard palate. They

terminate posteriorly with a rough edge, to which the

horizontal plates of the palatal bones articulate in the

transverse palatine suture. The posterior borders of these

bones are free and form the posterior nasal spine at

midline. The midline suture terminates anteriorly at the

incisive foramen. In young children, an irregular suture

line can be seen extending bilaterally from this foramen

to the alveolar process. The small triangular-shaped

structure anterior to this suture line is called the


Alveolar process
Incisive foramen

Median palatine
Palatine process
of maxilla
_/ palatine suture
-Palatine bone
Posterior nasal spine

Figure 3. The bones of the hard palate.



Soft palate


Figure 4. The hard and the soft palate.


premaxilla. The premaxilla represents the two parts of

the maxillary bone that unite early in embryonic life. The

suture line indicating the existence of the premaxilla is

almost erased on the adult skull (DuBruhl, 1980).

The hard palate is covered with mucosa of a grayish

pink color and various areas of the hard palate can be

identified due to the varying structure of the submucous

layer. Anteriorly, there is a series of transverse ridges

called rugae, posterior to which we find the midline raphe.

The gingiva and the glandular zone can also be clearly

identified (Figure 4). The hard palate is thick anteriorly

and laterally but becomes progressively thinner toward its

midline. This results in an arching of the palate both

transversely and anteroposteriorly. This arching differs

from one individual to the next and contributes to the

acoustic properties of the oral cavity and probably also

to individual voice characteristics (Zemlin, 1968).

The soft palate. The soft palate, or velum (Figure 5),

consists of muscular and fibrous tissues, nerves, and blood

vessels. It is positioned as a continuum from the posterior

border of the palatine bones of the hard palate. It is

attached bilaterally to the pharyngeal wall and extends

posteriorly into the oropharynx where it ends in a free

margin. The anterior part of the velum, slightly posterior

to the articulation with the hard palate, is the palatal

aponeurosis, a point where the palatal muscles are


- -

HP 4 r

V1 j

HP = Hard palate
V1 = Velum at rest
V2 = Velum at closure
PPW = Posterior pharyngeal wall

Figure 5. Lateral view of the hard and the soft palates.



attached. The posterior middle of the free margin of the

velum consists mainly of the uvula. Bilaterally to the

uvula, two curved folds of mucous membrane arch downward

and outward. These are the anterior and the posterior

faucal pillars (DuBruhl, 1980; Zemlin, 1968; Morley, 1962).

Normal function of the hard and the soft palates. The

function of the hard palate is to serve as the boney wall

separating the nasal cavity from the oral cavity. It thus

builds the floor of the nasal cavity and the roof of the

mouth. A review of the literature finds a prevalence of

the conclusion that the primary muscle active in velo-

pharyngeal closure is the elevator veli palatini muscle.

The contraction of this muscle elevates and pulls the soft

palate posteriorly. It has also been suggested that this

muscle exerts posteromedial movement of the lateral

pharyngeal walls, aiding velopharyngeal closure pro-

ficiency (Dickson, 1972).

The middle third of the velum demonstrates the greatest

range of movement. It achieves closure with the posterior

pharyngeal wall during speech, thus serving as a functional

valve during velopharyngeal closure. This elevation also

is thought to be substantially assisted by the uvular

muscle (Dickson, 1972).


Abnormal Palatal Function and Speech

The palatal anomalies affecting speech intelligibility,

which are the bases in this work, are limited to palatal

fistulas and VPI. The severity of a speech impairment is

dependent upon the size/location of a fistula, or the size

of a velopharyngeal opening. In other words, the severity

is dependent upon the magnitude of the coupling of the

oronasal cavities.

Extensive research has determined that a velopharyngeal
2 2
opening between 10 mm and 20 mm in size will result in

hypernasality with audible nasal air emission, thus dis-

torting the production of oral consonants (Isshiki et al.,

1968; Shprintzen, Croft, Levin, & Rakoff, 1977; Warren,

1964a,b). With the critical size of VPI being defined

there is, however, little information and no quantifiable

data available on the effect of palatal fistulas. It

appears that the severity of a speech disorder--in the

presence of a palatal fistula--will depend on the size and

location of the fistula. However, it has been concluded

that the same components of speech that are present with

VPI (hypernasality with audible nasal air emission) also

occur with palatal fistulas (Lindsay, 1971; Morley, 1962;

Palmer, Hamlen, Ross, & Lindsay, 1969; Reid, 1962; Ross &

Johnston, 1972; Skolnick, Glaser, & McWilliams, 1980).


The Effects of Velopharyngeal Insufficiency
on Speech

The importance of adequate velopharyngeal valving and

its influence on speech have long been recognized. Although

it is difficult to determine the exact frequency of VPI

in the non-cleft population, it has been suggested that

between 20 percent and 40 percent of patients who undergo

cleft palate repair suffer a concomitant VPI (Bradley,

1979). The major anatomical defects which occur are

related to velopharyngeal disproportion, muscular defi-

ciency, or surgical removal of tissue (Bradley, 1979).

Several authors have established the existence of a

close relationship between the degree of velopharyngeal

opening and the severity of speech impairments in normal

and in cleft palate patients (Bzoch, 1968; Moll, 1962; Shprint-

zen et al., 1975; Subtelny et al., 1961). The major speech

distortions associated with VPI can be measured by employ-

ing standardized tests and instruments or by the subjective

judgment rating of resonance quality and articulation. The

characteristics of these speech errors help us to differ-

entially diagnose VPI. Secondary components of speech

behavior, like nasal and facial grimacing, sometimes

accompany VPI (Bzoch, 1979).

Bradley (1979) lists specific speech sound errors as

being weak production of plosive sounds (together with

nasal emission of air), accompanied by a nasal non-speech


sound, called a nasal snort. She also reports that the

high vowels /u/ and /i/ can be expected to sound hypernasal

with accompanied nasal air emission. Bj6rk (1961) has

observed that the velar port is not completely closed when

nasal consonants and their neighboring vowels are produced

in normal speech. Findings by Moll (1962) showed the high

vowels to exhibit more velopharyngeal closure than their

low counterparts, /ae/ and /a/. Furthermore, when the vowels

were uttered in connection with a nasal sound such as /n/,

he observed incomplete velopharyngeal closure, especially

if the vowel preceded the nasal sound. This was confirmed

by Benson (1972), who found open velopharyngeal ports in

63 percent of the subjects phonating /a/, in 23 percent

when phonating /i/, and in 9 percent phonating /u/.

Fritzell (1973) also concluded that slight nasal airflow

could be measured in the normal population during the

production of /a/ but not with /i/, or with the fricatives

/f/ or /s/. He pointed out that individuals with VPI could

not attain the intraoral air pressure necessary for the

distinct production of plosive and fricative sounds. The

speech of these individuals, therefore, contained hyper-

nasal voice quality and compensatory speech errors, such

as glottal stops and pharyngeal fricatives.

Shelton et al. (1964) added to this information that

velopharyngeal closure was found to be complete when test-

ing isolated vowels in one subject group. However, this


same group of subjects later presented abnormal closure

patterns when tested in connected speech. This would seem

to indicate that syllable or sentence testing should be

employed when evaluating velopharyngeal closure. These

findings were investigated by Bzoch (1968), who found in

the testing of repeated syllables containing the plosive

/p/, combined with either of the high vowels, /i/ and /u/,

that complete velopharyngeal closure was obtained. This

was also the case when the low vowel /a/ was tested in the

same manner. However, a discrepancy of one millimeter in

the configuration of lower palatal elevation for the lower

vowel /a/ was observed.

Articulatory distortions in individuals with palatal

disorders were investigated by Bzoch (1965). He reported

that two-consonant blends, fricatives, and plosives--in

that order--were the most misarticulated sounds. The

error types included substitutions, distortions, and

omission of sounds. Substitutions can be recognized as

simple or gross substitutions. Gross substitutions are

characteristic for cleft palate patients and for the VPI

population. Gross substitutions included such errors as

pharyngeal fricatives, glottal stops, and nasal snorts.

These occurred as substitutions for plosive and fricative

sounds and occurred in the pronunciation of affricate and

glide speech sounds. The nasal snort was found to substi-

tute for the nasal sounds /m/ and /n/.



Perceptual Judgment

In order to evaluate different speech characteristics,

especially nasality, perceptual judgment has been and is

still used extensively both clinically and in research. It

is a subjective method of determining the degree of speech

impairment using different rating scales. There have been,

for example, 3-, 4-, 5-, and 10-point scales in use

(Counihan & Cullinan, 1970; Isshiki et al., 1968; Subtelny

et al., 1961; Warren, 1964b). Scaling procedures with

values from 0 to 100 has been utilized also (Fletcher &

Bishop, 1970). Different studies employed panels of judges

or single judges. Some studies claim higher reliability

using one judge, while other studies claim higher relia-

bility using a panel of judges (Bradford, Brooks, & Shelton,

1964). The number of judges also has varied considerably,

ranging from 3 judges to 9 judges to 48 judges (Counihan

& Cullinan, 1970; Liebman, 1964; Sherman, 1970; Subtelny

et al., 1961).

It appears especially difficult to obtain reliable

perceptual judgment of nasality. This difficulty is

influenced by many factors, including speaker variations

due to phonation, articulatory precision, regional dialect,

and language (Counihan & Cullinan, 1970; Fletcher, Sooudi,

& Frost, 1974; Warren, Hall, & Davis, 1981). Great

variance in sample selection, using normal and cleft

palate subjects,was also demonstrated. The difficulty to


obtain a homogenous sample was another problem, due to

differences in type and severity of the clefts (Bradford et

al., 1964; Coleman, 1963/1964; Andrews & Rutherford, 1972;

Watterson & Emanuel, 1981a,b).

The selection of different speech samples and the

clinician's training and experience also contributed to

the difficulty of obtaining valid and reliable nasality

ratings (Bradford et al., 1964; Bzoch, 1979; Counihan &

Cullinan, 1970). Speech samples included isolated sounds,

syllables, and/or connected speech in forward or backward

play (Counihan & Cullinan, 1970).

It is difficult to compare the results and to compute

accurate reliability because of the disparity in perceptual

testing of nasality. However, it has been documented

(Philips & Bzoch, 1969) that perceptual judgment of

articulation offers high intrarater reliability averaging

85 percent, although interrater reliability is lower with

an average of 74 percent. Furthermore, Sherman (1970)

found articulation and hypernasality to be functionally

related in children with cleft palate.

Different instruments have been employed to measure

nasality, sometimes in combination with perceptual judgment

testing. Up to this point, however, no single objective

test instrument seems to do so adequately and exclusively

(McWilliams, Glaser, Philips, Lawrence, Lavorato, Beery,

& Skolnick, 1981; Van Demark, Bzoch, Daly, Fletcher,


McWilliams, Pannbacker, & Weinberg, 1985). The traditional

measures used to augment perceptual judgment testing have

been radiographics, aerodynamics, and acoustics (Reich &

Redenbaugh, 1985).

In conclusion, it appears that it has been, and con-

tinues to be, a difficult task to obtain a reliable

measure of nasality. Therefore, a combination of methods

needs to be considered to secure the reliability of

nasality judgment.

The influence of examiner expectancy on subjective

evaluations, such as with perceptual judgments, has been

documented (Ramig, 1982). Physical disabilities, orofacial

anomalies, and other visual as well as auditory cues have

been noted to influence the examiner's judgments in face-

to-face examinations. Most studies reviewed here, however,

have employed tape recorded speech samples.

Subtelny et al. (1961) employed perceptual judgment

rating of nasality and intelligibility in combination with

lateral roentgenography measures. They reported on the size

of the velopharyngeal port opening as it was changing

during speech. The nasality ratings were derived by

using a four-point rating scale. The speech of 70 adult

subjects with cleft palate revealed that a distance between

the posterior pharyngeal wall and the soft palate of 3.5 mm

to 7.0 mm (3.5 mm equals 9.62 mm2) resulted in the auditory


perception of hypernasal speech. Smaller distances (0.5 mm

to 3.0 mm) resulted in the perception of only moderately

nasalized speech.

Word-syllable intelligibility evaluations were assessed

by the same judges and compared to aperture size (Subtelny

et al., 1961). Subjects with complete closure scored on

the average a 10 percent loss of intelligibility. Sub-

jects with the 0.5 mm to 3.0 mm distance showed a 14.3

percent loss, and the 3.5 mm to 7.0 mm group had a 27.31

percent loss. Even greater openings than these were not

accompanied by significant errors of nasality or intel-

ligibility. Only when the openings became as large as

11 mm to 18 mm was intelligibility reduced significantly.

This nonlinear relationship, as described here, has been

evidenced when comparing judgments of hypernasality with

velopharyngeal port size that were measured with lateral

still X-ray (Subtelny et al., 1961), with cineradiography

(Andrews & Rutherford, 1972), or with aerodynamic measure-

ments (Isshiki et al., 1968). It appears, from the results

of the study by Subtelny and co-workers (1961), that small

gaps have a greater effect on nasality than on intelli-

gibility. Subjects with a 3 mm diameter velopharyngeal

port opening still produced intelligible speech but were

judged as moderately nasal. This information agrees with

Bjork's study (1961), in which subjects who displayed a

4 mm large gap were assessed as hypernasal.


Hypernasality Testing

Evaluation of hypernasality can be executed easily by

employing a test described by Bzoch (1979). This test is

the Hypernasality Test which is standardized and convenient

for use both clinically and in research. The results from

the test are derived from a pass-fail procedure of 10 trials

where the correct productions are added together and

presented as a single score, where a maximum correct score

would be 10.

As the subject produces 10 one-syllable words (beet,

bit, bait, bet, bat, bought, boat, boot, but, Bert) once

with the nares open and again with them gently pinched

closed, a shift of voice quality will determine if hyper-

nasality is present. If the velopharyngeal port is inap-

propriately open during production of any of these words,

closing of the nares will result in changing this open-

ended resonator to a closed-end resonator and, consequently,

a change in resonance quality.

Articulation Testing

The Bzoch Error Pattern Screening Articulation Test

(BEPSAT) was designed to evaluate individuals with incom-

plete velopharyngeal closure (Bzoch, 1979). This test was

compared to the Iowa Pressure Articulation Test by

Erickson (1984). A correlation study indicated significant


relationship between the two tests. The two tests also

showed high validity as a tool for appropriately identifying

articulatory errors in the speech of cleft palate individuals.

The BETSAT, however, appears to provide more valuable infor-

mation on the total articulatory behavior, because it

includes all of the distinctive feature groups. There are

31 test words in the BEPSAT, with 23 words containing the

individual sound element in medial position and eight con-

sonant blends where it occurs in initial position.


Clinicians have tried, over the years, to correlate

the perceptual judgment of velopharyngeal competence or

incompetence with objective testing employing various

instruments designed to measure velopharyngeal closure.

There has been a concerted effort to find such a tool that

would measure the magnitude of VPI and to relate it to the

changes of voice resonance and articulation (Bjark, 1961;

Bzoch, 1979; Dalston, 1982; Edgerton, Sadove, Compton, Bull,

Blomain, McDonald, & Bralley, 1981; McWilliams et al., 1981;

Warren & DuBois, 1964). One of the reasons for designing

a measurement instrument was the search for objective and

reliable means to test nasality, since it had been found

that both interjudge and intrajudge reliability of percep-

tual judgment testing were low (Bradford et al., 1964;

Counihan & Cullinan, 1970).


Observational techniques, such as lateral still

roentgenography and cineradiography, have been employed in

a variety of research studies (Benson, 1972; Bj6rk, 1961;

Bzoch, 1968; Graber et al., 1959; Moll, 1960, 1962, 1964;

Shelton et al., 1964; Subtelny et al., 1961). These tech-

niques have also been used extensively in clinical


Other direct observational techniques include tomography

(Bj6rk, 1961; Kuehn & Dolan, 1975) and the employment of

fiberoptics in photodetector assessment (Dalston, 1982)

and in nasopharyngoscopy with videopharyngoscopy (Shprintzen

et al., 1977). Some measures of palatal function and

nasality would include intraoral air pressure using oral

manometers (Morris, 1966; McWilliams et al., 1981), acoustic

analysis with spectrography (Andrews & Rutherford, 1972; Bjork,

1S61; Coleman, 1963/1964; House & Stevens, 1956; Watterson

& Emanuel, 1981a), and nasal airflow testing (Bzoch, 1979).

Another instrument for acoustic analysis is called Tonar

(Fletcher, 1970; Fletcher & Bishop, 1970; Fletcher et al.,

1974). Two additional techniques are the accelerometric

technique also called nasal vibration analysis (Edgerton

et al., 1981; Horii, 1980; Reich & Redenbaugh, 1985; Stevens,

Kalikow, & Willemain, 1975), and respirometrics (Shaw &

Gilbert, 1982).

Derivation of the hydraulic principle was used in an

attempt to measure pressure-flow relationships (Gorlin &


Gorlin, 1951). Warren and DuBois (1964) extended the

attempt by employing the resulting hydrokinetic equation

to obtain indirect measures of the velopharyngeal port area.

The changes in size when different speech sounds were pro-

duced could thus be determined (Warren, 1964a,b). Quanti-

tative measurements from the pneumatic pressure-flow

technique (Gorlin & Gorlin, 1951; Warren, 1964a,b; Warren

& DuBois, 1964) have been employed to obtain indirect

measures of the velopharyngeal port area and to determine

its changes in size when different speech sounds are


Spectrographic analysis of the acoustic signal is

another method employed to assess the degree of nasalized

speech (Andrews & Rutherford, 1972; Bjbrk, 1961; House &

Stevens, 1956; Watterson & Emanuel, 1981a). At other

times a standardized test of hypernasality was administered

(Bzoch, 1979, Bzoch et al., 1984). However, a combination

of methods has been employed to ensure reliability of the

tested material. Although several instruments are

available for the measurement of nasality, it has, never-

theless, continued to be recorded by using perceptual

judgment procedures (Isshiki et al., 1968; Subtelny et al.,

1961; Watterson & Emanuel, 1981a,b).

Lateral Still Roentgenography and Cineradiography

Both lateral still roentgenography and cineradiography

have been used to examine visually the function of the


velopharyngeal port and to determine the relationship

between the velar length and the depth of the nasopharynx

(Figure 5). The patterns of the palatopharyngeal move-

ments and their changes, during speech, give much informa-

tion about velopharyngeal function and closure. When

inspecting the still X-ray or the cinefilm during speech

production, the elevated soft palate can be seen attempting

to make closure with the posterior pharyngeal wall. The

least linear distance between these two anatomical struc-

tures, the soft palate and the posterior pharyngeal wall,

is measured in millimeters. Bj6rk (1961) demonstrated a

strong linear relationship between the width of this gap

as measured from the lateral cinefilms and the area measures

of the gap in the velopharyngeal port as determined in the

cross-sections obtained from tomography. With these two

measures, he developed a diagram that can be employed in

order to determine the size of the coupling area by use of

cinefilms only (Bj6rk, 1961).

Occasionally, nasalization of speech is noted in spite

of obvious closure visible on the lateral film. Therefore,

it must be assumed that although closure might take place

at one point as the velum stretches toward the posterior

pharyngeal wall, it might do so only at that point which

can be noted on the lateral film. One or more parts of

the velum, laterally or medially to the touching point,

may not attain complete closure. This lack of complete

closure pattern, or uneven closure, would result in an


opening of the velopharyngeal port. This can be observed

with basal view cineradiography (Skolnick, 1970; Williams,

1986) or through tomography (Bjork, 1961). The limitation

of the lateral measurement and the radiation danger has

led to the development of other instrumental methods of

velopharyngeal quantifiable assessment such as naso-

endoscopy (Dalston, 1982).

Aerodynamic Technique

Intraoral air pressure is a physiological component

related to the production of different speech sounds. The

normal production of plosive and fricative sounds is depen-

dent upon adequate oral cavity pressure. Depending on the

size of a velopharyngeal deficiency and airflow into the

nasal cavity, interference with the pressure sounds can

be expected (Shelton et al., 1973). A number of instruments

have been developed to measure airflow and air pressure

with the basic components being flowmeters and pressure

transducers. The flowmeter records the volume rates of

airflow, and pressure transducers record air pressure

within the vocal tract (Lubker, 1970; Warren, 1973).

The pneumotachograph sensing element is the most

common and most reliable type of flowmeter. The principle

of this instrument is that there exists a proportionality

between the airflow rate in a tube and the pressure dif-

ference between its endpoints. This instrument typically


measures the volume of air in cubic centimeter per second

(Counihan, 1979; Hardy, 1965; Lubker, 1970).

It has been established that there is no simple rela-

tionship between the nasal airflow and the size of the

velopharyngeal port. There are factors such as respiratory

effort and nasal airway resistance that influence airflow

in subjects with VPI (Hutters, 1982; Warren, 1973).

Hutters (1982) claims that her findings indicate that

nasal airflow alone will provide reliable information about

velopharyngeal closure and that testing of nasal airflow

would be valuable as a diagnostic tool.

In order to obtain data on intraoral air pressure, it

is necessary to use a sensitive pressure transducer which

is connected to an amplifier and a recording device. Its

connecting tube, or catheter, is placed in the mouth per-

pendicular to the airflow with the requirement that it is

neither too long, nor that it has a too large or too small

a diameter (Lubker, 1970).

One of the aerodynamic techniques, the pressure-flow

technique, has been described by Warren (1964a,b) and by

Warren and DuBois (1964) as a method of calculating the

area size of VPI. It involves recording the amount of

airflow through the velopharyngeal orifice simultaneously

with the differential pressure across it during speech.

Airflow is measured with a pneumotachograph and differen-

tial pressure with a pressure transducer. By using these


measures, the velopharyngeal area can then be calculated

by using a hydrokinetic equation (Warren & DuBois, 1964,

p. 52; Warren & Devereux, 1966, p. 105).

VPA = Vn
.65 2


VPA = velopharyngeal area in mm2

Vn = rate of airflow in cm /sec

AP = differential pressure in cm H20, converted to

D = density of air (0.001), gm/cm3

.65 = correlation coefficient

Warren (1964a) employed this method and equation to

establish the relationship between oropharyngeal pressure,

nasal airflow, and velopharyngeal port size. He examined

the critical size openings in the velopharyngeal port in

10 normal-speaking subjects. The velopharyngeal opening

was seldom found to be larger than 10 mm2 and was usually

smaller than 3 mm2 under these conditions. Only once was

it open more than 20 mm2. In a succeeding experiment with

10 speakers with cleft palate, Warren (1964b) found that

those speakers who were considered to be rehabilitated,

that is, exhibited acceptable voice quality, rarely evidenced

an orifice larger than 10 mm2. With the nasal speakers,


however, he frequently recorded the aperture to be larger

than 20 mm2 and the extremely nasal speakers displayed

openings as large as 100 mm He concluded that the

critical range of closure, and where speech is perceived
as hypernasal, starts at about 20 mm.

Warren (1979) developed an instrument, the Palatal

Efficiency Rating Computed Instantaneously (PERCI). It was

designed to evaluate palatal efficiency (pressure drop

across an oral/nasal orifice) both for clinical use and

for research. This instrument permits instantaneous com-

puting of measures such as intraoral air pressure, nasal

airflow, and the differential pressure between the oral

and the nasal cavities. Other measurements that can be

obtained from its second version, PERCI II, are airway

conductance and the area of the orifice related to the

palatal mechanism (Microtronics Corporation, n.d.). Intra-

oral pressure is detected by a pressure transducer and the

nasal airflow by a pneumotachograph. Input to these instru-

ments is transmitted via catheters inserted into the mouth

and nose, respectively. When complete separation exists

between the oral and the nasal cavities, the nasal pressure

will be zero while the oral pressure will range between 3

to 7 cm H20. With incomplete separation, such as when a

palatal fistula is present, the difference in pressure

would typically vary with the size of that opening (Fig-

ure 6).



(a) Intraoral Air Pressure


(b) Differential Pressure

Figure 6. Palatal efficiency rating computed instantaneously
(PERCI). (a) Intraoral air pressure, (b) Dif-
ferential pressure.


When recording airflow with a pneumotachograph, such

as the one used in the PERCI instrumentation, the use of a

face mask is required to channel oral and nasal airflow.

Such a mask would typically be an anesthesia--or a diver's--

mask with built-in openings for the necessary catheters.

The use of face masks has been criticized. Lubker and Moll

(1965) cautioned that it may restrict articulatory move-

ments. Furthermore, the amount of pressure that is required

for the mask to obtain an airtight fit also may influence

the normal speech behavior. The face mask has been

criticized also in that the air storage in it would provide

a resistance to the expiratory airstream and, therefore,

induce elevated intraoral air pressure and airflow

(Counihan, 1979). However, these considerations have not

been substantially confirmed. With the use of a short,

straight catheter with a wide opening, resistance is not

clearly affecting speech measurement data.

Sound Spectrography

Sound spectrography has been widely employed to measure

nasalized speech. The spectrograph analyzes complex sig-

nals, such as the speech signal, as a function of frequency,

time, and intensity. The different speech sounds create

certain patterns on the visual output, the spectrogram.

The most important characteristics in evaluating presence

of nasality are the formants, which can be seen as dark


areas on the spectrogram. They consist of regions of

accentuated intensity and represent the resonances of the

vocal tract (Hadding & Petersson, 1970). The three first

formants (Fl, F2, and F3) usually can be clearly dis-

tinguished, and sometimes there is even a clear display

of F4 and F5.

Schwartz (1979) described the nasalized feature

characteristics as a reduction in the intensity of the

first formant, antiresonances, extra resonances, and a

change in the center frequencies of the formants. Other

spectral features that can be associated with nasality also

may occur. These have, however, been observed to be too

unpredictable or too sporadic to deserve mentioning. Even

the four characteristics just described may not always

occur. However, whenever a nasalized vowel is produced,

it is likely at least one of them will be seen on the


In an attempt to evaluate velopharyngeal area size

during normal speech and during simulated nasal speech,

simultaneous cineradiography and sound spectrography were

employed by Bjirk (1961). Thirty adults and 26 children,

all normal speakers, served as subjects. The findings

showed that when relatively small changes of nasality were

registered by the spectrograph, the cineradiographic

readings would indicate a velopharyngeal opening of
approximately 10 mm In a second study, these normal


speaking subjects were asked to mimic hypernasal speech.

The coupling areas were then found to have increased to
2 2
between 100 mm and 280 mm

Nasal Airflow

Velopharyngeal and/or palatal openings will permit

airflow into the nasal cavity and nasal air escape can be

recorded. A convenient and simple test, the Nasal Emission

Test, can be employed for evaluation of nasal air emission

(Bzoch, 1979). Bzoch, Kemker, and Dixon-Wood (1984)

stated that clinical experience has proven it to be a valu-

able speech evaluation tool of velopharyngeal function in

cleft palate patients. Bzoch et al. (1984) continued to

report that, in a comparison study between cineradiographic

measures of the velopharyngeal port and an analysis of the

spoken test words, there was a 96 percent agreement of

palatal function when the speech of 40 patients was

tested. The standardized Nasal Emission Test tests 10

two-syllable words (people, paper, puppy, pepper, piper,

baby, Bobby, bubble, B.B., bye-bye). As each word is

repeated by the subject, the examiner observes any airflow

through the nose using a simple indicator, such as a small

paper paddle about the size and shape of a toothbrush held

under the nose (Bzoch, 1979).

The results are recorded from a pass-fail procedure

of 10 trials where the correct productions are added


together and presented as a single score, with a maximum

correct score of 10. This test should always yield a

perfect score of 10 to indicate complete closure of the

velopharyngeal mechanism.

The Effects of Controlled Velopharyngeal
Insufficiency on Speech

Some experimental studies were performed where the

investigators created different size velar port openings

in normal speaking subjects. This involved (a) inserting

of different sized polyvinyl tubing in the velopharynx

(Bernthal & Beukelman, 1977; Isshiki et al., 1968) or

(b) using a variable aperture speech appliance with

inserts (Figure 7) (Andrews & Rutherford, 1972; Liebman,

1964; Watterson & Emanuel, 1981a,b). By manipulating the

velopharyngeal port size, different degrees of hyper-

nasality could be created and computations of the critical

size of openings in the velopharyngeal area were performed.

In order to determine whether there is a critical

size of the velopharyngeal insufficiency at which point

speech becomes unacceptable, Isshiki and co-workers (1968)

conducted an experimental test with 11 young adult normal

speakers. Polyvinyl tubes were inserted into the velo-

pharynx where artificial openings were created at diameters

of 5 mm, 7 mm, 9 mm, and 12 mm. Articulation and hyper-

nasality were perceptually judged by five speech specialists


~------ ZZD


Figure 7. Two examples of tools for introducing velo-
pharyngeal insufficiency in normal speakers.
(a) Polyvinyl tubing (Isshiki et al., 1968),
(b) Variable palatal prosthesis with inserts
(Andrews & Rutherford, 1972).





on a three-point and a five-point rating scale, respec-

tively. The findings indicated that at no specific point

along the range of dimensions did speech suddenly sound

abnormal. The authors claim, instead, that there is a

gradual change in both intelligibility and nasality. Speech

quality was judged as acceptable--although slightly nasal--

up to a critical aperture size of 5 mm in diameter which
equals 19.6 mm The articulatory errors did not exceed

25 percent at this size opening. When the larger size

aperture of 7 mm (38.5 mm2) was tested, nasality was

unquestionable. Articulation errors increased to more

than 60 percent, and overall speech was unacceptable. Both

nasal airflow and intraoral air pressure were measured for

all of the above tube size conditions. The results showed

increasing nasal air excape and decreasing intraoral pres-

sure with increasing hole size.

Andrews and Rutherford (1972) designed a study aimed

at assessing the contribution of the nasally emitted sound

to the perception of nasality. Through the use of special

instrumentation they were able to separate the orally

emitted sound from the combined orally and nasally emitted

sound as they are produced in normal speech. By applying

a variable aperture palatal prosthesis with inserts

(Figure 7) they created the following velopharyngeal
2 2 2
conditions: no opening, 60 mm 120 mm 180 mm and

240 mm2 openings. In order to evaluate hypernasality,
240 mm openings. In order to evaluate hypernasality,


data from perceptual judgment and spectrographic recordings

were collected. Perceptual findings show a clear trend of

more hypernasality ratings for the combined nasal and oral

signal than for the orally emitted alone, especially for

the vowel /u/. Spectrographic data supported these findings

by displaying clear nasal features on the spectrogram, such

as a combination of increased intensity and reduced fre-

quency of the third formant for /u/, and extra resonances

between F1 and F2 for /i/.

The effects of velar perforations on speech were

studied by Liebman (1964). She employed an experimental

prosthetic speech appliance that was constructed to simulate

these perforations in a variety of sizes and locations,

singly and in pairs. Listeners' judgments of nasality

and intelligibility in one speaker served as indicators of

the influence of these artificial velopharyngeal openings

on speech. The general findings indicated that speech

became worse when the apertures were larger and placed in

a more forward position. Furthermore, if the hole was

divided into two openings of equal combined size and

located in the same general area, speech was noted to be

improved. This finding has not been explained satisfac-

torily but was suggested to have clinical implications.

The studies that have been discussed up to this point

have focused primarily on critical area size of VPI and

its influence on nasality and intelligibility. The


following studies to be discussed will deal with how

changes of velopharyngeal port size affect special speech

characteristics, such as vowel intensity (Bernthal &

Beukelman, 1977), vowel identification and nasality, and

whispered vowel spectra (Watterson & Emanuel, 1981a,b).

It is observed repeatedly in the clinic that subjects

with VPI speak with a reduction of an overall intensity

level, because of increased coupling between the oral and

nasal cavities. Bernthal and Beukelman (1977) were able

to demonstrate that vowel intensity is a factor of velo-

pharyngeal portal opening. One normal speaking subject

was fitted with a palatal prosthesis in which controlled

openings at the position of the velopharyngeal port could

be manipulated. The result showed that when velopharyngeal

openings of 7 mm2 and 50 mm2 were present vowel intensity
was reduced, especially at the 50 mm aperture.

It was stated earlier that loudness reduction is fre-

quently observed in subjects with VPI. It is, however,

uncertain whether this is a result of the intensity reduc-

tion or if it is caused by a behavior to try and hide a

speech disorder (Morris, 1968). With Bernthal and

Beukelman's results at hand, it appears that there may

exist a close relationship between intensity and loudness

in this population. To compensate for this loss of sound

energy, speakers with VPI often increase their vocal


effort, which might abuse their vocal folds (Bernthal &

Beukelman, 1977).

It is well known that presentation of loud noise alters

auditory feedback and causes changes in speech. The most

noticeable change is the increase in vocal intensity.

Articulation, however, seems to be unaffected by the

introduction of noise, an assumption that was confirmed

by Garber, Speidel, Siegel, Miller, and Glass (1980).

A number of authors studied the acoustics of nasalized

speech by employing analog methods (Watterson & Emanuel,

1981a). It has been suggested also that an increase in

oral-nasal coupling would show reliable and predictable

acoustic effects on the spectrogram. When these theories

were tested empirically, however, the acoustic effects

read from the spectrograms were sometimes much less

reliable and predictable (Curtis, 1970; Schwartz, 1979;

Watterson & Emanuel, 1981a). It was, therefore, reasoned

that one way of avoiding these difficulties would be to

test whispered vowels (Watterson & Emanuel, 1981a). By

employing whispered vowels, continuous spectral energy

would be produced rather than harmonic, which is the result

when using voiced speech. Watterson and Emanuel expected

to observe systematic formant changes in the spectrograms

as the velopharyngeal aperture was manipulated. A variable

speech appliance with inserts allowed for 12.57 mm ,
2 2 2 2
28.27 mm 50.26 mm 73.53 mm and 153.94 mm openings


in the velopharynx. The results indicated that whispered

speech samples did not establish clearly a relationship

between aperture size and degree of hypernasality when it

was measured with spectrography, nor was there a clear dif-

ference between whispered or voiced vowel samples.

Watterson and Emanuel (1981b), therefore, decided to

include a perceptual listening study to measure the effects

of oral-nasal coupling on vowel identification and vowel

nasality. Both voiced and whispered speech were tested

while using the same variable prosthesis as in their pre-

vious study. The results indicate that ratings of nasality

were more reliable with the voiced than with the whispered

test vowels.

The research reviewed above has dealt with determining

velopharyngeal port size, introducing artificial VPI through

a variable prosthetic appliance, and measuring their effect

on speech. They have evaluated speech in terms of per-

ceptual testing of articulation and hypernasality. Measure-

ments on some correlates of speech, such as differential

air pressure and nasal airflow, vocal intensity of vowels,

vowel identification, vowel nasality, and vowel spectra

also were included. All studies concluded that the larger

the velopharyngeal port size the more distortion is

detected in all of the above speech characteristics,

although there is not always a complete linear relation-

ship present.


Subtelny and co-workers (1961) observed that nasalized

speech increased in severity as velopharyngeal openings

increased. When, however, sizes were larger than 7.0 mm

in size,they saw a leveling off in the nasality ratings.

Similar observations were recorded with the intelligibility

ratings. Again, speakers with openings from 7.5 mm to 11 mm

did not score significantly different from the group having

openings from 3.5 mm to 7.0 mm. However, speakers with a

gap exceeding 11.5 mm (11.5 mm to 18 mm) showed an increase

in loss of intelligibility. The major common finding of

all these studies, thus, infersthat an opening between

10 mm2 and 20 mm2 through the velar port results in sig-

nificant distortion of speech intelligibility and

resonance quality.

The Effects of Palatal Fistulas on Speech

Orofacial surgical skills have greatly improved over

the past decade. Occasionally, however, an opening along

the suture line of a repaired palatal cleft will occur

(Figure 8). This opening or fistula between the oral and

nasal cavities results in inappropriate loss of air through

the nose and excessive nasal resonance.

The incidence of fistulas after surgery in cleft

palate patients has varied considerably. Ross and Johnston

(1972) concluded that the incidence range was from


Figure 8. Palatal fistula.


10 percent to 20 percent, depending on different surgical

techniques and cleft type, as well as different sample

selections in the available studies. Postsurgical occur-

rence of fistulas in cleft palate individuals was reported

to be as high as 18 percent by Abyholm and co-workers (1979),

while a range between 9 percent and 34 percent was indicated

by Henningsson (1983). She mentioned no quantitative

indications of fistula size or location, except that these

fistulas usually occur along the midline and that size and

position determine the degree of the symptoms. In addition

to the reported incidence of fistulas following surgery in

cleft palate patients, there are also fistulas resulting

from ablative surgery and from trauma. No information

appears to be available on incidence resulting from con-

genital anomalies, nor has any information been found on

its factual occurrence.

In this context, it should be mentioned, that indi-

viduals with severe clefts, such as bilateral complete

clefts, would develop fistulas secondary to surgical repair

more frequently than patients with a simpler type of cleft.

More residual palatal fistulas were reported to have

occurred when the von Langenbeck palatoplasty had been

performed than after pushback repair (Krause, Tharp, &

Morris, 1976; Lindsay, 1971; Palmer et al., 1979). These

residual fistulas that occurred after the von Langenbeck

procedure were found to be slightly larger than those


resulting from the pushback technique. The two reports by

Lindsay (1971) and Palmer et al. (1969) were in agreement

that the residual fistulas seemed to be occurring mostly

in the anterior and alveolar parts of the hard palate.

In contrast to this information, Reid (1962) described

fistulas as being small, those less than two centimeters

in diameter, while a large hole would be greater than two

centimeters, with its location typically along the midline.

He further commented that such a large residual fistula

would follow pushback surgery. Witzell, Clarke, Lindsay,

and Thomson (1979) also compared results following the von

Langenbeck palatoplasty and the pushback procedure. The

two surgical procedures compared positively when no

residual fistulas occurred.

With the availability of ample information on VPI and

its effect on speech, it might be concluded that similar

data would be available for palatal fistulas and their

effect on speech. However, in a review of the literature,

the only studies found were descriptive and qualitative

reports. These studies mainly dealt with incidence and

descriptions of surgical techniques. Case histories were

also reported, including general descriptions of fistulas.

It was suggested that fistulas alone, or in association

with VPI, contribute to speech-crippling effects similar

to that of an unrepaired cleft palate (Millard, 1980).


As mentioned in Chapter I, it is unclear whether

fistula location is important. It has been suggested that

anteriorly located fistulas exert significant negative

influence on speech production (Cosman & Falk, 1980). Other

authors merely indicated a possible existence of speech

disorders resulting from anteriorly located fistulas (Ross

& Johnston, 1972). Posteriorly located fistulas were,

however, declared to cause cleft palate-like speech (Ross

& Johnston, 1972).

Palmer et al. (1969) and Lindsay (1971) stated in their

respective studies that all postsurgically occurring

fistulas were situated in the anterior portion of the hard

palate. These anteriorly located fistulas were said to be

a source of audible nasal air emission, impairing articu-

lation proficiency. These fistulas also were said to cause

slight nasality (Lindsay, 1971; Palmer et al., 1969). A

report by Musgrave and Bremner (1960) indicated that many

of the complications following the second stage of palatal

repair occurred as midpalatal fistulas. No indication

was given as to how or if any of these fistulas influenced

normal speech production.

Fistula size has been described in a qualitative and

descriptive manner. Clinical experience indicates a

definite increase of speech unintelligibility and nasal

resonance with increasing fistula size. This has been

substantiated in the literature (Gordon & Brown, 1980;


Henningsson, 1983; Morley, 1962; Reid, 1962; Shelton &

Blank, 1984). However, few authors have defined the size

of fistulas in relation to speech disorders. Those authors

who quantitatively defined fistula size were all describing

fistula size in relation to surgical techniques (Henderson,

1982; Proctor, 1969; Reid, 1962).

Palatal fistulas were categorized grossly as small,

moderate-sized, or large. Reid (1962) described small

holes as being less than two centimeters in diameter and

a large hole as being more than two centimeters in diameter.

Speech defects, apparently, were registered only with the

larger fistulas. These large fistulas were reported to

cause nasal air escape and an increased nasal voice

quality. Proctor (1969) defined large oronasal fistulas

to be 0.5 centimeter to 2.5 centimeters in size. No other

sizes were, however, discussed. Henderson (1982), on the

other hand, reported an approximal definition of small

fistulas (0.5 centimeter to 1.5 centimeters in width and

0.5 centimeter to 2.0 centimeters in length), but failed

to offer any information on large fistulas. Gordon and

Brown (1980), who discussed very small and moderate-sized

fistulas, explained the very small hole to be less than a

few millimeters, but declined to define the moderate-size


An article by Shelton and Blank (1984) describes the

single relevant study where palatal fistula size is related


to speech impairments. However, the hole sizes are not

quantified but simply described as small, moderate, and

large. These authors determined that small palatal fistulas

did not exhibit any nasal airflow during speech. The intra-

oral air pressure necessary for normal speech production

could be maintained in the presence of small fistulas.

It was found that the subjects having moderate fistulas

could retain the necessary intraoral air pressure at the

expense of nasal air emission. All subjects with large

palatal fistulas developed nasal air emission and a reduc-

tion of intraoral air pressure. The presence of nasal air

emission and the reduction of intraoral air pressure were

found to be detrimental to the production of adequate

articulation (Shelton & Blank, 1984).

Research Questions

Both clinical observation and much research indicate

that palatal defects resulting in abnormal openings between

the oral-nasal cavities do, in fact, greatly influence

articulation and resonance quality of speech. This, of

course, depends largely on the size, shape, and location

of the opening. Little research is reported which defines

the influence of incomplete separation of the oronasal

cavities anterior to the velopharyngeal port. It is,

however, clear that with a fistula present, leaving all

other structures intact or fully repaired, we can expect


a breakdown of intraoral air pressure, as well as air

excape into the nasal cavity and through the nose. This

affects both resonance quality and articulation of speech.

With the inference from above reviewed research one

must ask the following questions.


Can a fistula in the hard palate cause similar
speech impairments as those resulting from VPI?

It has been shown that an opening in the velopharyngeal

port negatively affects normal speech production. One

would consequently assume that any other abnormal opening,

such as a palatal fistula, connecting the oronasal cavities,

would likewise impair normal speech.


To what extent would a speech impairment vary
with differing sizes of a palatal fistula?

Previous research studies of velopharyngeal closure have

shown that speech becomes progressively more distorted as

the velopharyngeal port size is increased. It would seem

reasonable to predict, and it is clinically known, that

this is also the case with palatal fistulas. However,

since there are no definable quantitative data available,

it would seem pertinent to begin defining fistula size,

measured in square millimeter, and to measure the effect

that different size fistulas exert on speech production.


Is there a critical fistula size at which point
speech can be measured or judged as being dis-


Intraoral air pressure has been found to decrease rapidly

when velopharyngeal openings exceeded 10 mm2. The critical

size of velopharyngeal closure was suggested to occur at

the point which distinguishes velopharyngeal adequacy from

inadequacy. This was found to begin at approximately

20 mm2. From this it was concluded that VPI may begin at

a 10 to 20 mm2 range (Warren, 1964a,b). It seems important

to find out if a similar range exists with palatal fistulas

or if the critical size here would be smaller or larger.


To what extent would a speech impairment vary with
different positions of a palatal fistula?

Anterior positions of velar openings have been demonstrated

to cause fewer speech distortions than posteriorly positioned

ones. Regarding fistulas in the hard palate, there seems

to be no relevant information indicating the influence of

fistula location. One report indicated that a posterior

location of a fistula caused more speech problems than an

anterior location. However, in this context there has

been no description on the exact fistula location and how

it would influence speech. First of all, it is of interest

to examine the significance of the localization between

VPI and hard palate fistulas. Secondly, a definition of

significant differences among fistulas located in the hard

palate would seem pertinent.

Millard (1980) reported that surgical closure of a

fistula often can be extremely difficult to obtain.


Patients with residual palatal fistulas are fitted instead

with an obturating device to achieve normal speech and to

eliminate food and liquids from passing to the nasal cavity,

as well as preventing nasal fluids from entering the mouth.

It is the intention that, with the results from this study,

quantified definitions of palatal fistulas will facilitate

the choice of treatment for these individuals. The primary

goal of this work, however, is to perform basic research

to obtain quantitative definitions of fistulas in the hard

palate and their influence on speech. In other words, the

purpose of this study is to begin defining the influence

that size and location of an opening through the hard

palate has on speech in terms of articulation and resonance




Measurement of oral speech in the presence of a palatal

defect was discussed in the previous chapter. The basic

speech characteristics related to palatal disorders are

nasality and articulation. Validity and reliability, as

reviewed in Chapter II, were both difficult to obtain in the

severity ratings of hypernasality. Both intra- and inter-

rater reliability were evidenced to be poor. Regarding the

perceptual judgment of articulation, intrarater reliability

was good but interjudge reliability were low.

Due to the ambiguity of perceptual judgment-testing

of nasality, additional tools were developed and tested

in the hope to obtain one single objective measure of

resonance quality. The consensus of the literature

reviewed above was that no single instrument exists that

exclusively documents nasality (Van Demark et al., 1985;

Fletcher et al., 1974). The recommendation has been to

employ a combination of methods to obtain adequate

nasality measurements (Reich & Redenbaugh, 1985). These

combinations usually contained subjective testing, employ-

ing rating scales or standardized tests of articulation



and hypernasality, together with an objective instrumental


Materials and Procedures

To answer the questions posed in Chapter II, palatal

openings mimicing fistulas had to be experimentally

created. To provide a range of sizes and locations of

palatal openings, a research prosthesis for one subject

was constructed. The subject has an unrepaired cleft

palate and wears a removable palatal appliance. To define

and quantify the effects of different palatal openings on

speech, a battery of tests was included. These tests can

be divided into two categories: (a) subjective perceptual

testing and (b) instrumental testing.


One young adult male with complete bilateral cleft

of the hard and the soft palates was employed for this

study. This individual is a 28-year-old Caucasian pro-

fessional. He speaks English with general American

dialect, holds a master's degree in speech pathology and

a Ph.D. degree in education. He wears a removable palatal

appliance and his speech is perceived as normal. The

subject was born with a bilateral cleft of the hard and

the soft palates, the alveolar ridge, premaxilla, and the


upper lip (Wharton, 1978). Surgical closure of the lip

was performed shortly after birth with several secondary

reconstructive repairs at later ages. The hard and the

soft palates have never been closed. It was explained that

the hard palate was never surgically closed because of the

width of the palatal cleft. Instead, the subject was

fitted with a removable palatal prosthesis at age three

in order to obtain separation between the nasal and the

oral cavities. His speech was judged to be extremely good,

only four months after fitting of his palatal appliance,

by the cleft palate team at the University of Florida

teaching hospital. Today the subject still wears a

removable palatal appliance and his speech is perceived

as normal. An audiometric evaluation of the subject's

hearing prior to conducting any testing indicated normal

hearing bilaterally (Appendix A).

Research Prosthesis

A duplicate of the subject's own presently existing

removable palatal prosthesis was constructed by a prostho-

dontist in the Department of Removable Prosthodontics at

the University of Florida J. Hillis Miller Health Center.

The appliance consists of two parts. One is a metal frame

with a special design for retention of the appliance in

the subject's mouth. The other part is an acrylic resin


portion which fits accurately to the cleft area. After

the construction, which required three fitting sessions,

the subject was instructed to wear and adjust to the

appliance, in order to overcome any functional differences

that may have yet been present. After 10 days, it was

decided that the subject had adapted well to this new

appliance. This was independently judged by two speech-

language pathologists. The subject's speech was free from

articulation errors and his resonance quality was normal.

The subject himself expressed satisfaction with his speech

intelligibility, resonance quality, and the ease with

which normal speech could be produced.

Test Conditions

To permit manipulation of the size and location of

openings anterior to the velopharyngeal port, holes were

drilled through the appliance, resulting in an opening

between the oronasal cavities. Speech was then tested,

using one hole at a time. The locations of the openings

were positioned along the midline, in the front one-

third (A), in the middle one-third (B), and in the back

one-third (C) of the hard palate part of the prosthesis.

A fourth location of holes was drilled in the middle

of the speech bulb (D), which is also the velar part

of the prosthesis (Figure 9). The area sizes of the
2 2 2
openings were 5 mm (1), 10 mm (2), 20 mm (3), and

20 mm (4). The area sizes were selected on the


0 5 mm2

0 10 mm2



20 mm

30 mm2
30 mm

A = Anterior of the hard palate
B = Middle of the hard palate
C = Posterior of the hard palate
D = Velar

Figure 9. Sizes and positions of experimental fistulas.


grounds of previous research on critical points of velo-

pharyngeal insufficiency and speech. Warren, Dalston,

Trier, and Holder (1985) concluded that these sizes ranged

from 1 mm to 9 mm2 for adequate closure, from 10 mm2 to

19 mm2 for borderline closure, and inadequate closure was
registered when the VPI was larger than 19 mm

Each opening was drilled one at a time and resealed

after the entire test battery had been completed for that

particular opening. Self-curing resin was used to reseal

the holes to secure an airtight seal. Altogether there

were 16 experimental test conditions, combining the four

area sizes (1, 2, 3, and 4) with the four different posi-

tions (A, B, C, and D). All holes were drilled in

a random order, taking into consideration the physical

limitations of the prosthesis. There was concern due to

the possibility of fracture of the partition between the

largest sizes of positions B and C. This may have

influenced the randomization. The testing order for all

conditions is presented in Appendix B. All tests were

also conducted under three baseline conditions. The

baselines included testing with the subject's normal,

original prosthesis (BN) in place. Baseline testing of

the research prosthesis (BR) was performed before any of

the experimental openings were drilled. A post-test base-

line measurement (PBR) was also included. At the PBR

testing, all openings had been securely plugged with resin.


Human Experimentation Approval

The study was granted Human Experimental Approval by

the Institutional Review Board at the Shands Medical

Center. The subject was informed that his participation

in this study would not include any known discomfort or

anticipated risks. An informed consent to participate in

research form was read and signed by the subject.

Test Environment and Audio Recording

Testing was performed with the subject comfortably

seated in an acoustically sound-proof research unit at the

University of Florida J. Hillis Miller Health Center. A

tape recorder (Sony, TC-158 SD, Dolby system), with its

300 ohm dynamic microphone (Thoro Test) and cassette tapes

(Scotch AVX-50),was used to record all speech samples.

The Bioengineering Service tested the recording unit

through the microphone and found it to present a "MIC"

input constant of 0.005V RMS with a drive level of -3 dB.

As the unit was tested using a sound level meter, it was

found that all values up to a frequency of 6500 Hertz were

normally represented by the tape recording with a tapering

off to 8000 Hertz, whereafter there was a sharp drop in

the quality of tape and recording reproduction.

The microphone was securely fastened to a microphone

stand and placed at a distance of 30 centimeters from the


mouth and at a 90-degree angle of incidence. This position

of the microphone minimized distortions of the acoustic

voice signal by respiratory airflow noise. The VU-meter

or volume indicator on the tape recorder was adjusted and

monitored to control the intensity level (loudness) by the

investigator. No attempt was made to control the loudness

or pitch of the subject's voice. The subject was able to

hold a general level of vocal production (syllables, words,

and sentences) with only a negligible change of decibels

throughout the testing, which was sufficient for this

purpose. It was necessary that all the equipment be

located within the testing booth, together with the sub-

ject and the investigator.

Perceptual Measurements

There was a total of five perceptual tests in the speech

evaluation. These consisted of (1) one standardized test of

articulation, (2) one articulation rating test, (3) one

standardized test of hypernasality without auditory mask-

ing noise, (4) one standardized test of hypernasality with

auditory masking noise, and (5) one hypernasality rating

test. The standardized tests listed above, (1) and (3),

are tests employed at the University of Florida's Cranio-

facial Center and the Department of Communicative Disorders.

The two perceptual rating tests were constructed in an


attempt to evaluate perceptual scaling judgments and to

compare them to the other testing tools.

Articulation. The Bzoch Error Pattern Screening

Articulation Test (BEPSAT) (Bzoch, 1979), previously

described in Chapter II, was employed for the articulation

testing (Appendix C). The BEPSAT was selected because it

is a standardized test and provides valuable information of

the articulatory proficiency of all distinctive feature


The perceptual rating of articulation proficiency in

connected speech was performed by judging four sentences.

These sentences are listed in Appendix D. The sentences

represent a sample of sentences frequently used in the

evaluation of speech disorders related to palatal dis-

orders at the University of Florida Craniofacial Center.

Judgments of these sentences were conducted by employing

a five-point ordinal rating scale (Appendix D).

Hypernasality. For the evaluation of nasality, the

standardized Cul-de-sac Test of Hypernasality was adminis-

tered (Appendix C). This test was described in detail in

Chapter II. This test was repeated a second time while

the subject was exposed to auditory masking noise. The

reason for this was to rule out any effect from compensa-

tion for nasality due to auditory feedback (Coleman, 1963/1964;

Andrews & Rutherford, 1972). To obtain this new test

environment, a second tape recorder, a Sharp Educator


#376895 TLR with headphones (Thoro-Test SHP-2, 8 ohm), was

employed in the experiment. The masking noise consisted

of a prerecorded multitalker tape. Twenty speakers were

recorded simultaneously on this tape. It was stated that

"the recording is unintelligible and the level is stable

over time, having a relatively flat spectrum (-2.5 dB) from

200 to 500 Hz" (Holmes, Frank, & Stoker, 1983, p. 88). The

tape was obtained from a University of Florida, Department

of Speech, audiologist. The multitalker noise tape was

presented to the subject binaurally at a sound pressure

level of 90 decibels as it was adjusted with a sound-level

meter (Davis, 1970).

The second test of hypernasality consisted of the

perceptual rating of eight paired consonant-vowel-syllables

(CVC-syllables). The CVC-syllable "pin" was paired with

another CVC-syllable that contained the plosive consonant

sound /p/, the vowel sound /i/ followed by one each of the

six plosive sounds /p/, /b/, /t/, /d/, /k/, and /g/. There

were, thus, six different syllable combinations. The vowel

/i/ was selected because, as a high vowel, it has been

demonstrated to be the sound where the least velopharyngeal

opening was observed, thus indicating a high level of intra-

oral air pressure (Moll, 1962; Watterson & Emanuel,

1981a). Two of the pairs "pin-pib" and "pin-pip" were

repeated, but not in consecutive order. This repetition

was included to perform a reliability test among and


between the judges. The six judges rated nasality accord-

ing to a five-point ordinal rating scale (Appendix E).

Procedures of Data Collection

The speech samples obtained through the administration

of the articulation and the hypernasality tests were tape-

recorded. There were five different tests altogether--two

articulation tests and two nasality tests, of which one

nasality test was repeated in a masking noise environment.

With the above five tests and the 19 different conditions

under which they were tested, a total of 95 test objectives

was to be perceptually judged. A test tape was then

arranged with the 95 tests occurring in random order.

Due to the repetitious nature and the bulk of this test

battery, it is not included in its entirety in the disser-



The recorded speech samples were subject to perceptual

judgment. Six speech-language pathologists served as

judges. They were selected because of their experience

in working with resonance and articulation disorders

associated with VPI. All but one of the judges were able

to evaluate all 95 tests in one sitting. Each judge spent

an average time of one hour and 30 minutes for this


evaluation. The range was from one hour and 20 minutes

to one hour and 45 minutes. The 95 tests were recorded

on three tapes. Tape one contained tests one through 38,

tape two tests 39 through 78, and tape three the tests 79

through 95. Each judge was presented with a different

sequence of the three tapes, as shown in Appendix F. This

was arranged to avoid a tiring influence on the judges'


Prior to testing, each judge was familiarized with

the four tests and read a set of instructions (Appendix G).

The judges were told that they could stop the test tape,

using a stop button on the tape recorder, should they need

to stop for replay or for taking a rest. The judges were

not informed of the masking feature in some of the


Each judge was required to make a total of 589

(31 x 19) articulation test judgments and to rate articu-

latory proficiency from 76 (4 x 19) connected speech

samples (sentences) with a five-point scale. Additionally

each judge was asked to make 380 (2 x 10 x 19) judgments

on the hypernasality test (190 responses from the unmasked

version and 190 responses from the masked version). The

rating of hypernasality included judgments of 171 (8 x 19)

paired CVC-syllables based on a five-point scale. Varia-

tion in the listening environment was minimized and at all


times judging was performed in a closed room with only the

judge and the experimenter present.


Three instrumental tests were employed to further

assess the effect of palatal openings on speech. These

tests represented measurements of nasal airflow, aero-

dynamic measurements of intraoral air pressure, and dif-

ferential pressure. The first test, the Nasal Emission

Test, supplies the airflow data. The second and third

tests were performed by employing the PERCI II instrument

which provides the measurements of intraoral air pressure

and differential pressure.

Nasal airflow. The data on nasal emission of air were

collected by employing the Nasal Emission Test as seen in

Appendix C. This is a standardized test employing simple

and accurate means to measure the presence or absence of

nasal air emission. The test is employed frequently at

the University of Florida's Cranofacial Center and the

Department of Communicative Disorders. The test was

described in detail in Chapter II.

Aerodynamic measurements. Objective instrumental

measurements were next applied to measure the effects of

palatal openings. The aerodynamic entities that were

selected for testing, which are closely related to normal


speech production, were intraoral air pressure and differ-

ential pressure. Measurement of nasal emission of air was

also attempted but the results were inconclusive. They

were, therefore, excluded from the study (see section of

PERCI below, page 82).

The data from the intraoral air pressure and differ-

ential pressure measurements were gathered to increase the

reliability of the perceptual tests. Furthermore, they

were collected to determine whether any of these measures

could be singly employed for nasality detection.

To compute intraoral air pressure and differential

pressure an instrument, PERCI II, was employed. The func-

tion of this instrument was described in Chapter II. To

make PERCI II operative, two separate pressure transducers

and a pneumotachograph had to be connected to the instru-

ment via catheters. These catheters had to be of different

diameters, depending on what part it would serve.

For testing of intraoral air pressure during speech,

an oral catheter, two millimeters in diameter, was attached

to a Stratham PM5ETC pressure transducer, which in turn

was connected to the pressure channel on PERCI II. Before

testing was begun, the pressure channel of the unit was

calibrated with a water manometer, constructed by the


For the differential pressure test, an additional

catheter was added to the pressure transducer. The catheter


size was four millimeters in diameter. It was threaded

onto a nasal olive that was to be placed in one of the

subject's nostrils during testing (Figure 6).

It was planned to include an additional test of nasal

emission with this instrument. It was thought that such

results obtained by instrumental testing would enhance the

results from the perceptual testing of nasal emission. To

measure nasal emission, the flow channel of the instrument

had to be made operative. Nasal flow is measured with a

nasal catheter with an airtight fit into one nostril. The

other end of the catheter is connected to a Fleisch pneumo-

tachograph (pneumotach) Type One. The pneumotach is, in

turn, connected to a heater power supply (filament trans-

former of 6.3 Volt at 1.2 ampere), used to heat the mesh

screen inside the pneumotach cylinder. This is necessary to

prevent condensation from forming on the screen, which would

lead to an erroneous increase in the airflow signal (Micro-

tronics Corporation, n.d.). The small pressure differential

developed across the screen in the pneumotach is then

detected by a Stratham PM15ETC pressure transducer and

the signal is sent to PERCI II for read-out.

Calibration of the flow channel was performed by

connecting the pneumotach to a captured air source. A

flow gauge (Fischer and Porter Company Precision Bore

Flowrator Tube No. FP-1/2-17-G-10/27 sec/sec-air met.

@ 14.7 P.S.I.A. & 700) with a scale range of 500 cubic


centimeters per second was connected to the output side of

the pneumotach. The calibration of the flow channel was

not linear. Deviations were occurring in the lower ranges.

The instrument could be calibrated only at the ranges from

500 cubic centimeters through 300 cubic centimeters per

second. At the level of 250 cubic centimeters the calibra-

tion errors were measured at 10 units on PERCI II. At

200 cubic centimeters, 100 cubic centimeters, and at

50 cubic centimeters the errors were 20, 40, and 50 units

off, respectively. It was expected that the results could

be adjusted to this error in calibration. However, during

testing, it was noted that the errors were too grave and

erratic. Therefore, due to this internal calibration

instability, this measurement of nasal air emission was

excluded from the study.

Procedures of Data Collection

Data from the Nasal Emission Test were collected by

the examiner at the time of the original testing. A small

paper paddle was held to the subject's nose while he pro-

duced each of the 10 words from the test. The test was

repeated twice for reliability. Any movement of the paper

paddle, indicating nasal air flow, was registered. This

test should yield a score of zero movement of the paddle

to record no nasal air leakage. The data from this test


of nasal emission were collected by the examiner and

repeated twice for the three baselines and the 16 experi-

mental test combinations.

The procedures for testing intraoral air pressure and

differential pressure called for employing the instrument

PERCI II. The two tests were measured during speech for

each of the 19 palatal conditions. The test with PERCI II

followed the perceptual testing for each baseline and

palatal opening.

The speech sample employed was one of those suggested in

the PERCI manual (Microtronics Corporation, n.d.). The speech

sample for both intraoral air pressure and differential pres-

sure was the word "pop." Five readings were recorded by the

investigator. It was explained in the PERCI manual that words

and phrases containing the plosive sound /p/ were preferred.

When these sounds are produced, the mouth is closed and a

stagnant column of air is created. This is desirable because

a normal measurement during the baseline tests would rule out

any velopharyngeal incompetence, which was especially

necessary for the accuracy in the experimental tests with

the research prosthesis. These sounds, furthermore,

eliminated any problems that could occur from tongue

placement. For example, if the place of articulation would

occur behind a palatal opening, the true speech effect of

that opening would not be accurately documented. By

employing plosive sounds the effect of respiratory effort


is also controlled in the presence of palatal openings.

In the testing of differential pressure, a ratio of oral

and nasal pressure is measured and the respiratory effort

is cancelled out, because it affects both pressures equally.

It was expected that, with a negligible VPI, the measure-

ments of intraoral air pressure and differential pressure

could be used accurately to measure the influence that

defects in the hard palate, such as fistulas, would have

on these parameters important for normal speech production.

Reliability Testing

The conclusion of the reviewed literature on perceptual

judgment rating of nasality showed low intra- and inter-

reliability (see Chapter II). Evidence from perceptual

rating of articulation had been found to indicate good

intrareliability but less reliable interagreement. Other

perceptual tests, such as the employment of standardized

tests for articulation and hypernasality, had not demon-

strated these reliability problems.

The knowledge of these findings was the prerequisite

for the reliability testing, on the hypernasality rating

test that was performed in this study. As described above,

there were six different paired CVC-syllables in the test.

Two of these CVC-syllable pairs were selected for repeti-

tion for the purpose of reliability testing, thus increasing


the total number of tested syllable pairs to eight. The

repeated syllables were "pin-pip" and "pin-pib." These

CVC-syllables were nestled into the list of the original

test syllables. The judges were not instructed about this

procedure. No one of the judges commented on the fact that

they had identified the repeated syllable-pairs in the test

material. The reliability data were extracted from the

total number of responses on the hypernasality rating

test and were treated separately. Both intrareliability

and interreliability were analyzed with the Spearman rank

order correlation procedure.

Replication Study

The velar opening (D) has been treated as one of the

experimental palatal openings. The same hole sizes were

drilled and the same series of tests were employed for each

of the D-openings. These D-openings were, furthermore,

intermixed in the sampling regarding test order and no

differences in data collection were considered.

Nevertheless, D was originally included for comparison

studies with previous results from research on VPI. It

was intended that if the D-results compared favorably to

previous results on VPI and speech, a common ground could

be established for comparing the effects of VPI on speech

with those from palatal fistulas.


Data Analysis

The data collected from the above-described testing

procedures were subjected to statistical analysis. The data

were organized to permit comparisons between the different

test implementations. The choice of appropriate statistical

tests was discussed with a statistics consultant from the

Department of Biostatistics at the J. Hillis Miller Health

Center. To accurately answer the research questions in

Chapter II, the Statistical Analysis System (SAS) was

employed (Statistical Analysis System Institute, Inc.,

1982a,b). Because of the complicated nature of the data,

each of the perceptual tests was computed independently.

The aerodynamic test scores of intraoral air pressure and

differential pressure were computed together. The program

of statistical analyses included means, standard deviations,

and analysis of variance (Marks, 1982a,b). Duncan's mul-

tiple range test was employed for a posteriori pairwise

comparison among the means (Kirk, 1968). Its purpose is to

further identify which means are different from each other

after the analysis of variance has been performed. If

significant differences can be identified with this a

posteriori test, there is a high degree of confidence in

the conclusions. To determine the degree of relationship

between the variables Spearman's rank order correlation

coefficients were computed.


This study represents an attempt to start defining and

quantifying the influence of palatal fistulas on speech.

Specific sizes and locations were obtained by systematic-

ally manipulating artificial openings in a removable palatal

prosthesis. Four locations along the palatal midline were

utilized and four artificial fistula sizes were used.
2 2
These ranged from 5 mm to 30 mm To address the ques-

tions posed in Chapter II, data were gathered from several

tests of speech and from measurements of the speech cor-

relates of nasal air emission, intraoral air pressure, and

differential pressure. There were seven types of data

obtained from these tests and measures. Five of these data

sets were collected from two perceptual tests of articula-

tion, two perceptual tests of hypernasality, and one set

of data originating from an instrumental pass-fail test of

nasal air emission. The last two data sets were obtained

from instrumental test measurements of intraoral air

pressure and differential pressure.

The purpose of this study was to quantify and define

palatal openings by their influence on speech through

descriptive statistics, rather than to analyze speech as



it was affected by apalatal openings. The objectives were

to define the critical size/location of the different open-

ings, whetherspeech was negatively influenced in any way,

and to quantify each opening size/position by the influence

it exerted on speech behavior. Therefore, each separate

test was analyzed relative to error production on the

perceptual listening tests. The instrumental tests were

analyzed relative to the necessary intraoral air pressure

and differential pressure for maintaining normal speech

production and to quantitatively correlate each size/loca-

tion to its deviation. The presence of any nasal emission

of air was recorded, and the data from two readings were


Perceptual Measurements

There were four perceptual tests in the test battery

judged by six judges. The tests employed were (1) the Bzoch

Error Pattern Screening Articulation Test, (2) the Articu-

lation Rating Test, (3) the Cul-de-sac Hypernasality Test,

and (4) the Hypernasality Rating Test. The Cul-de-sac

Hypernasality Test was used twice, the second time with

auditory masking, thus increasing to five the actual number

of evaluated perceptual tests.

Each of these five tests was assessed with the three

baselines (BN, BR, and PBR) and the 16 experimental openings


2 2
(A, B, C, and D locations and the sizes 5 mm2, 10 mm2
2 2
20 mm and 30 mm2) in random order.

Articulation Test

The response variables from the BEPSAT were determined

by assessing all test sounds contained in each of the 31

words. The number of correct articulations was extracted

and used in the data analysis as the cutoff score that was

to identify complete articulatory proficiency versus non-

proficiency. No differentiation of error type was deter-


The means and standard deviations of correct articulation

among the six judges for all 19 test conditions are shown in

Table 1 and Table 2. Figure 10 and Figure 11 illustrate

these results graphically by size and position.

The analysis of variance indicated significant dif-

ferences between the means (F = 2.12, df = 15, p < 0.01).

Further analysis of the independent variables, position and

size, did not show significant differences between the

positions (F = 0.15, df = 63, p = 0.92). The differences

between the sizes were, however, significant (F = 8.69,

df = 3, p < 0.01). On a size-by-size analysis, the a

posteriori Duncan Multiple Range Test was employed to

identify further differences among the sizes. It indicated
that the 5 mm openings were not significantly different

from the baseline measurements. However, all larger


Table 1. Means of normal articulation scores among the
six judges (Articulation Test; n = 31).

Size (mm )

Position 5 10 20 30

A 28 23 15 23
B 29 26 18 20
C 27 24 21 20
D 26 25 20 21



27 29 30

Table 2. Standard deviations of the normal articulation
scores among the six judges (Articulation Test;
n = 31).

Size (mm )

Position 5 10 20 30

A 3.57 6.83 8.95 6.24
B 2.73 4.11 9.83 7.90
C 4.49 6.25 7.03 6.33
D 3.31 5.89 7.17 9.21



3.88 3.09 0.81

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.9 - mvs