Title of Dissertation: 
Dissertation directed by: 
ABSTRACT 
TONGUE-PALA TE INTERACTION IN DISCRETE 
AND SEQUENTIAL SW ALLOWING 
Gloria Chi-Fishman, Doctor of Philosophy, 1996 
Adjunct Professor Maureen Stone 
Professor Gerald N. McCall 
Department of Hearing and Speech Sciences 
Historically, swallowing motor control was thought to involve a central mechanism 
that generated patterned responses with little use of sensory input. Although increasing 
evidence of peripheral modulation has altered this concept, our knowledge about the 
flexibility in deglutitive motor control and performance is incomplete. This study sought 
to gain a better understanding by examining lingual motor strategies in light of changing 
bolus properties (volume, consistency) and task demands (discrete vs. sequential 
swallowing). Specifically, the timing and patterns of tongue-palate contact and the 
associated changes in tongue shape and action were examined in five normal adults using 
simultaneous electropalatography (EPG) and ultrasound. Tasks for discrete swallowing 
included 5 and 30 cc of water, 5 and 30 cc of gelatin, and saliva. Tasks for sequential 
swallowing involved drinking 200 cc of water at normal and fast rates. Two analysis 
schemes were used to make timing and percent-contact measurements: segmentation of 
the EPG time series into four stages (prepropulsion, propulsion, full contact, withdrawal), 
and compartmentalization of the pseudopalate into six bins (front, central, back, lateral, 
medial, midline). Results showed little variation in contact pattern as a function of bolus 
property or subject, suggesting considerable stereotypy in lingual motor strategies for 
movement sequencing. However, unlike the conventional description, tongue-palate 
contact during propulsion was multidimensional with two distinct degrees of freedom in 
the front-to-back and the lateral-to-midline continua. Significant (Q<. 0 I) timing 
differences were found in that larger and thinner boluses were propelled faster than smaller 
and thicker ones, and dry swallows had longer full contact than water. For sequential 
swallowing during continuous drinking, the tongue used faster movement speed and 
overlapping gestures to meet the task demands, while propulsive contact pattern remained 
invariant. Thus, the change was not in motor strategies per se but in the timing 
coordination of the "drink" and "swallow" sequences. A 3-D model of oral lingual action 
for swallowing was proposed. Clinical implications were discussed. In sum, results of this 
study support the theory that swallowing motor control includes a peripheral mechanism 
capable of modulating centrally generated responses, and that the deglutitive motor 
program has both invariant and variant parameters. 
TONGUE-PALA TE INTERACTION 
IN DISCRETE AND SEQUENTIAL SW ALLOWING 
by 
Gloria Chi-Fishman 
Dissertation submitted to the Faculty of the Graduate School of the 
University of Maryland at College Park in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
1996 
(' I 
~ 11 
I I 
Advisory Committee: 
Professor Gerald N. McCall, Chair/ Advisor 
Adjunct Professor Maureen Stone, Advisor 
Adjunct Professor Joan Mele-McCarthy 
Associate Professor Sally J. Phillips 
Associate Professor Nan B. Ratner 
Mo.;ry !and 
LD 
J ds, 31 
, M 70d 
C,hi - tt.Sh 1W10£11 
G. 
..I 
©Copyright by 
Gloria Chi-Fishman 
1996 
DEDICATION 
To my husband, Dr. Peter H. Fishman, 
to my mother, Mrs. Pa-Chun Chi, 
and to my late father, Mr. Wan-Te Chi. 
11 
ACKNOWLEDGMENTS 
The author wishes to express her gratitude to Dr. Maureen Stone for her support 
and guidance during the dissertation project, and to Dr. Gerald N. McCall for his 
mentoring and encouragement. The technical assistance of Dr. Paul J. Smith, Edward 
Davis, Daniel Denman, Yongqing Zhang, Rich Sico, Michael Mulleady, and John Crump 
is also greatly appreciated. Special thanks are further extended to Andrew J. Lundberg for 
his assistance during the actual experiments, and to the Department of Linguistics, 
University of California, Los Angeles, for providing the percent-contact computation 
software written by Andrew Mueller. 
The research was supported in part by the National Institute on Deafness and 
Communication Disorders and in part by the Kay Elemetrics Corporation. 
A portion of this dissertation has been published in Dysphagia, Vol. 11, pp. 239-
247, 1996. 
11l 
TABLE OF CONTENTS 
LIST OF TABLES ............. . ...... ... ...... ..................... . vu 
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii 
LIST OF APPENDICES ..... . ......................................... xii 
CHAPTER I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 
Research Rationale: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 
Review of Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 
The Tongue: Composition, Motor Diversity, and Innervation ......... 4 
Swallowing: Conceptual Framework for Tongue Kinematics ....... .. 7 
Neural Organization of Swallowing .. . .................... 7 
Neural control theories .. ... ........ ... ....... . .. 7 
Neural control elements . . . . . . . . . . . . . . . . . . . . . . . . . . 9 
Physiological Mechanism of Swallowing . . . . . . . . . . . . . . . . . . 15 
Oral and oropharyngeal transport of bolus ........... 16 
Pharyngeal transport of bolus . . . . . . . . . . . . . . . . . . . . 18 
Issues in Swallowing Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 
Effects of Bolus Properties ............................ 20 
Volume ..................................... 20 
Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 
Effects of Task Demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 
Tongue-Palate Interaction and Electropalatography ............... 29 
A Theory of Swallowing Motor Control . ....................... 32 
Summary ............................ . ...... ........ .. . . 34 
Research Purpose and Questions ................................... 34 
CHAPTER II. METHOD .... ... ............ .. .............. ... ........ 36 
Subjects .... . ...... .... . . ..................... ... . ............ 36 
Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 6 
Electropalatography .......................... . ............ 36 
Ultrasound Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 
EPG-Ultrasound Recording Synchronization . . . . . . . . . . . . . . . . . . . . 44 
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 
Preparatory Activities ...................................... 45 
Subject Preparation .................................. 45 
Pseudopalate Landmark-Marking on Ultrasound ............ 46 
Experiment 1. Discrete Swallowing . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 
Tasks and Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 
IV 
Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 
EPG Data . ...... . ........................... 51 
Overall analysis framework . . . . . . . . . . . . . . . . 51 
Measurements and statistics .... . ........... 55 
Ultrasound Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 
Overall analysis framework . . . . . . . . . . . . . . . . 61 
Observations and measurements . . . . . . . . . . . . . 63 
Experiment 2. Sequential Swallowing in Continuous Drinking ....... 65 
Tasks and Design ................ . ................. . 65 
Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 
EPG Data . .......... . . . .... . . ....... . . .. .... 66 
Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 66 
Ultrasound Data . . ...... . ..................... 67 
Observations and measurements . . . . . . . . . . . . . 67 
CHAPTER III. RES UL TS ..................................... . ...... . 69 
Experiment 1. Discrete Swallowing ........................ . ........ 69 
EPG Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 
Tongue-Palate Contact Timing . . ..................... . . 69 
Total duration: task effect and subject variability ...... 69 
Stage duration: effects of task and stage, and subject 
variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 
Tongue-Palate Contact Patterns ..... . .................. 79 
Contact patterns for individual stages . . . . . . . . . . . . . . 79 
Task effect . .............. . .................. 90 
Subject variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 
Left-right symmetry ............................ 94 
Ultrasound Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 
Swallow-Related Lingual Behaviors on Ultrasound .......... 96 
Ultrasound versus EPG: Descriptive Comparison . . . . . . . . . . . 97 
Task Effect on Ultrasound ............ . ............... 98 
Lingual Velocity Profiles: Single-Subject Data ............ 102 
Correlation of EPG Timing Measurements with Ultrasound 
Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 
Identification of Unusual Lingual Behaviors ....... .. ..... 107 
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 
Experiment 2. Sequential Swallowing in Continuous Drinking ............ 109 
EPG Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 
Tongue-Palate Contact Timing ........................ 109 
Total duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 
Additional timing data . . . . . . . . . . . . . . . . . . . . . . . . . 112 
V 
Tongue-Palate Contact Pattern ............... . ........ 113 
General characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 113 
Sequential versus discrete: statistical analyses . . . . . . . 122 
Ultrasound Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 
Swallow Volume Estimation .............. . ........... 124 
Lingual Movement Characteristics . . . . . . . . . . . . . . . . . . . . . 124 
Lingual Velocity Profiles: Single-Subject Data ............ 128 
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 
CHAPTER IV. DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 
Discrete Swallowing ....................................... .. .. 133 
Tongue-Palate Contact Patterns ............................. 133 
Effect of Bolus Size and Consistency on Contact Timing .......... 137 
Continuous Drinking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 8 
Sequential versus Single Swallows . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 8 
Motor Control Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 9 
Methodological Consideration ........ . ..................... 147 
Ultrasound as a Supplementary Technique for EPG .................... 148 
Toward a Model of Deglutitive Tongue Action . . . . . . . . . . . . . . . . . . . . . . . 15 1 
Discrete Swallowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 
Continuous Drinking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 
Clinical Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 
APPENDIX A. ULTRASOUND TRANSDUCER PLACEMENT DATA . . . . . . . . 163 
APPENDIX B. STATISTICAL TABLES ... . ............................ 164 
APPENDIX C. TONGUE LOWERING DURING BOLUS INT AKE IN DISCRETE 
SW ALLOWING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 
REFERENCES ..................................................... 174 
VI 
LIST OF TABLES 
Table 3.1. Mean total durations of tongue-palate contact during discrete 
swallowing for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 
Table 3.2. Mean positive peak velocities and onset-to-peak intervals for 
the four tongue segments from different discrete swallows of 
Subject 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 
Table 3.3 . Mean total durations of tongue-palate contact for sequential 
swallows during normal- and fast-rate continuous drinking 
for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 O 
Table 3.4. Average total-duration ratios of Sequential:Discrete water 
swallows for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 
Table 3.5. Estimated bolus size of sequential swallows during continuous 
drinking for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 
Table 3.6. Mean positive peak velocities and onset-to-peak intervals for 
the four tongue segments from the continuous drinking series 
of Subject 5 ............................................ 131 
Vil 
Figure 2.1. 
Figure 2.2. 
Figure 2.3. 
Figure 2.4. 
Figure 2.5 . 
Figure 2.6. 
Figure 2.7. 
Figure 2.8. 
Figure 2.9. 
Figure 3.1 . 
Figure 3.2. 
LIST OF FIGURES 
Schematic diagram of instrumental setup .... . ..... . ......... ... 37 
A subject's acrylic pseudopalate, and the Palatometer electrode 
sensitivity adjustment screen ...................... .. . .... . ... 39 
The head and transducer support (HA TS) system and transducer 
placement calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 
Physical landmarks of a subject's pseudopalate on ultrasound and 
corresponding EPG displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 7 
Subject-by-task repeated measures design for the discrete 
swallowing experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 
Compartmentalization of the pseudopalate: front, central, back, 
rnidline, medial, and lateral bins . ... . ........ . . . .............. 52 
Temporal boundaries for the four stages in the raw EPG time-series 
of a representative discrete swallowing task (including the swallow 
and post-swallow mouth opening) ......... . .... . ............ . 54 
Examples of pre-warp and post-warp data conformity in two 
representative discrete swallows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 
Schematic of functional lingual segmentation, and grid system for 
tracking radial displacement of the anterior, mid, dorsal, and 
posterior tongue segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 
Least-squares mean total durations for Subjects 1 to 5 as a function 
of discrete swallowing task and for different discrete swallowing 
tasks as a function of subject . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 
Mean durations of individual discrete swallowing tasks for the four 
stages ........ . ............ ..... . . ... ............... . ... 73 
viii 
Figure 3.3. Average durations of Stage I (prepropulsion) as a function of 
discrete swallowing tasks for Subjects 1 to 5 ... . .. . ............. 75 
Figure 3 .4. Average durations of Stage II (propulsion) as a function of 
discrete swallowing tasks for Subjects 1 to 5 .................... 76 
Figure 3.5. Average durations of Stage III (full contact) as a function of 
discrete swallowing tasks for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . 77 
Figure 3.6. Average durations of Stage IV (withdrawal) as a function of 
discrete swallowing tasks for Subjects 1 to 5 . . . . . . . . . . . . . . . . . . . . 78 
Figure 3. 7 a. Changes in average percent-contact over time in front, central, 
and back bins from different discrete swallowing tasks for S 1 . . . . . . . . 80 
Figure 3. 7b. Changes in average percent-contact over time in lateral, medial, 
and midline bins from different discrete swallowing tasks for S 1 . . . . . . 81 
Figure 3.8a. Changes in average percent-contact over time in front , central, 
and back bins from different discrete swallowing tasks for S2 . . . . . . . . 82 
Figure 3.8b. Changes in average percent-contact over time in lateral, medial, 
and midline bins from different discrete swallowing tasks for S2 . . . . . . 83 
Figure 3. 9a. Changes in average percent-contact over time in front, central, 
and back bins from different discrete swallowing tasks for S3 . . . . . . . . 84 
Figure 3.9b. Changes in average percent-contact over time in lateral, medial, 
and midline bins from different discrete swallowing tasks for S3 . . . . . . 85 
Figure 3. 1 Oa. Changes in average percent-contact over time in front , central, 
and back bins from different discrete swallowing tasks for S4 . . . . . . . . 86 
Figure 3 .1 Ob. Changes in average percent-contact over time in lateral, medial, 
and midline bins from different discrete swallowing tasks for S4 . . . . . . 87 
Figure 3.1 la. Changes in average percent-contact over time in front central 
' ' and back bins from different discrete swallowing tasks for S5 ........ 88 
Figure 3.1 lb. Changes in average percent-contact over time in lateral, medial, 
and midline bins from different discrete swallowing tasks for S5 .. . ... 89 
IX 
Figure 3.12. Raw EPG data from two representative discrete swallows, 
illustrating Stage-II contact patterns ....... .. .............. .. .. 91 
Figure 3 .13 . Frequency distribution of contact patterns 1 and 2 for different 
discrete swallowing tasks and subjects based on raw percent-
contact data in lateral bin at Stage-II onset . . . . . . . . . . . . . . . . . . . . . . 93 
Figure 3.14. Changes in average percent-contact over time in left and right 
bins from representative discrete swallows of selected subjects ....... 95 
Figure 3 .15a. Changes in tongue shape over time as a function of discrete 
swallowing task (part 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 
Figure 3. 15 b. Changes in tongue shape over time as a function of discrete 
swallowing task (part 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 
Figure 3 .16. 
Figure 3.17. 
Figure 3.18. 
Figure 3.19. 
Figure 3.20. 
Figure 3.21. 
Velocity profiles for four tongue segments from representative 
discrete swallows of S5 ............ · · .. · · · · · · . · · ..... ... . . 
Changes in percent-contact over time in front-central-back and 
lateral-medial-midline bins from representative continuous 
drinking series of S 1 . . . . . . . . . . . . . . . . . . . . · · · . . · · · · · · · · · · · . . 
Changes in percent-contact over time in front-central-back and 
lateral-medial-midline bins from representative continuous 
d 'nki . n ng senes of S2 .' ...... . ......... · · · · · · · .. · · · · · · . · · · · · 
Changes in percent-contact over time in front-central-back and 
lateral-medial-midline bins from representative continuous 
drinking series of S3 ... .. ........... ... ..... . ............ . 
Changes in percent-contact over time in front-central-back and 
lateral-medial-midline bins from representative continuous 
d 'nki . n ng senes of S4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 
Changes in percent-contact over time in front-central-back and 
lateral-medial-midline bins from representative continuous 
drinking series of S5 ............. . ... ...... ... ...... . . ... . 
Figure 3.22. Raw EPG data for two sequential swallows in a representative 
104 
114 
115 
116 
117 
118 
continuous drinking series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 
X 
Figure 3 .23 . Raw EPG data for two sequential swallows in the continuous 
drinking (self-proclaimed "chugging") series of S3 ..... ... ... .... 121 
Figure 3.24. Least-squares means for Front:Back and Lateral:Back percent-
contact ratios at time of maximal midline contact for Subjects 1 
to 5 as a function of sequential swallowing task . . . . . . . . . . . . . . . . . 123 
Figure 3.25. Changes in tongue shape in first and second swallows during 
normal- and fast-rate continuous drinking ......... ...... .. ..... 126 
Figure 3.26. Velocity profiles for four tongue segments from the continuous, 
normal-rate drinking series of S5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 
Figure 3.27. Velocity profiles for four tongue segments from the continuous, 
fast-rate drinking series of S5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 
XI 
LIST OF APPENDICES 
A. Ultrasound Transducer Placement Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 
B. STATISTICAL TABLES ....................... . ............... 164 
B-1 . Random effects analysis of variance on EPG total durations for 
the discrete swallowing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 
B-2. Repeated measures analysis of variance on mean durations of 
EPG stages for different discrete swallowing tasks . . . . . . . . . . . . . . . 165 
B-3 . Post hoc repeated measures multivariate analysis of variance of 
pairwise contrasts for EPG Stage-II duration between selected 
discrete swallowing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 
B-4. Post hoc repeated measures multivariate analysis of variance of 
pairwise contrasts for EPG Stage-III duration between selected 
discrete swallowing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 
B-5. Analyses of variance of main effect and pairwise contrasts on mean 
EPG total durations for sequential, 5-cc water, and 30-cc water 
swallows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 
B-6. Analyses of variance of main effect and pairwise contrasts on 
Front:Back percent-contact ratios at time of maximal midline for 
sequential, 5-cc water, and 30-cc water swallows . . . . . . . . . . . . . . . . 170 
B-7. Analyses of variance of main effect and pairwise contrasts on 
Lateral :Back percent-contact ratios at time of maximal midline for 
sequential, 5-cc water, and 30-cc water swallows . . . . . . . . . . . . . . . . 171 
XII 
C. TONGUE LOWERING DURING BOLUS INT AKE IN DISCRETE 
SW ALLOWING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 
C-1. Digitized ultrasound images illustrating progressive tongue 
lowering to receive bolus during a representative single 
swallow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 
C-2. Composite display oflingual surface curves extracted from the 
digitized ultrasound images shown in C- l . . . . . . . . . . . . . . . . . . . . . . 173 
Xlll 
CHAPTER I. INTRODUCTION 
Research Rationale: An Overview 
The tongue has long been recognized as the "principal mobile agent" (Bosma, 
1957, p. 281) or the "primary manipulator" (Kennedy & Kent, 1988, p. 26) for oral and 
oropharyngeal transport of the bolus during swallowing. In recent years, there has been 
empirical evidence that oropharyngeal bolus propulsion is the direct result of tongue 
pumping action supplemented by pharyngoesophageal pressure change (Cerenko, 
McConnel, & Jackson, 1989; McConnel, 1988). The tongue's vital role in enacting 
swallow-related physiological events is further revealed in clinical reports that poor lingual 
control in neurologic and/or postsurgical oral cancer patients is a major cause of 
swallowing dysfunction (Hirano et al., 1992; Logemann, 1983). 
Structurally as well as functionally, the tongue possesses diverse and unique 
properties. It is capable of performing a wide range of movements with multiple degrees 
of freedom. Such capability is based on the remarkable muscular-hydrostatic 
characteristic of its anatomic composition (Kier & Smith, 1985) and gained from 
interacting with other structures of the oral cavity, especially the hard palate (Stone, 1991, 
1995). Tongue-palate interaction during swallowing conceivably serves critical roles in 
bolus formation, containment, and transport, and in the generation of pressure gradients 
I 
and force necessary for proper propulsion. However, our understanding of such 
interaction is by far incomplete. 
2 
The potential degrees of freedom in tongue movement suggests enormous 
potential flexibility in lingual motor control. It is commonly known that flexibility in 
control and performance is characteristic of motor systems for skilled actions. Indeed, the 
tongue's flexibility, as reflected in complex movement variations, is well documented for 
speech production (Stone, 1991 , 1995). In contrast, little information is available about 
the extent of flexibility in deglutitive lingual motor control. Although there are many 
descriptive accounts of how the tongue moves during swallowing (Hamlet, Stone, & 
Shawker, 1988; Shawker, Sonies, Hall, & Baum, 1984; Shawker, Sonies, Stone, & Baum, 
1983; Stone & Shawker, 1986), minimal data exist on deglutitive lingual motor strategies, 
or on how tongue movements adapt to changes in bolus properties and swallowing task 
demands. Only a few investigations, to date, have specifically examined the relation 
between bolus volume and lingual action during oral and pharyngeal transport of food 
(Hamlet, 1989; Kahrilas, Lin, Logemann, Ergun, & Facchini, 1993; Martin, 1991 ). 
One can only speculate that the paucity of data on deglutitive lingual motor 
flexibility is perhaps because swallowing is not a highly skilled motor behavior. Another 
reason could be that historically the theory of swallowing motor program was based on a 
central mechanism thought to make little use of peripheral information (see Miller, 1982). 
The modern concept of swallowing motor programming, however, does acknowledge the 
role of sensory modulation. Above all, a voluntary initial oral phase is as much a part of 
the complex swallowing process as the reflexive pharyngeal and esophageal phases. The 
3 
new evidence that, for example, larger boluses do induce more vigorous propulsive action 
from the tongue (Kahrilas et al., 1993) supports the notion that deglutitive lingual motor 
control has more flexibility/adaptability than previously believed. On the other hand, 
because of the biomechanical constraints of the oral cavity, the unidirectionality of the 
bolus translocation in oral transport, and the stringent requirement for airway protection 
during swallowing, it is theoretically implausible that for deglutition purposes, the tongue 
needs the same high degree of flexibility as the motor systems for expert skills such as 
playing a musical instrument with the hands. An investigation aiming at better 
understanding the extent of deglutitive lingual motor flexibility and at identifying variable 
versus invariant motor program parameters has, therefore, theoretical justification. 
From the clinical perspective, improved knowledge on deglutitive lingual motor 
flexibility would have a positive impact on patient management. For decades, swallowing 
clinicians and researchers have continued the quest for a sound theoretical foundation on 
which to build effective clinical evaluation and treatment protocols. Information on the 
flexibility and/ or invariance of lingual motor performance would contribute to this 
theoretical base, provide a physiological or kinematic framework for deriving explanations 
for dysphagia symptoms related to lingual motility deficits, and offer insights into 
appropriate management strategies. 
The present study examined an important area--linguopalatal interaction during 
swallowing--where little research existed. Specifically, efforts were made to delineate the 
temporal and spatial details of deglutitive tongue-palate contact and to address lingual 
kinematic issues concerning response modulations due to changing bolus properties 
(volume and consistency) and task demands (discrete swallowing vs. sequential 
swallowing at normal and fast rates). The underlying motive was that by exploring the 
effects of boluses and tasks on specific parameters of tongue action during swallowing, a 
better understanding would be gained of the flexibility in deglutitive lingual motor 
performance and of the swallowing motor program in general. 
Review of Literature 
The Tongue: Composition, Motor Diversity, and Innervation 
The tongue is one of the most complex structures in the oral sensorimotor system. 
It is remarkably diverse in function and unique in biomechanical properties. In humans, 
the tongue participates in a great variety of oromotor behaviors, ranging from primitive 
(e.g., suckling), to stereotyped (e.g., swallowing), to highly skilled (e.g., speech). Its 
intrinsic anatomy is characterized by a mesh of vertical, transverse, and longitudinal 
muscle fibers, fat, and connective tissue with no bones or joints. Its extrinsic muscles--
genioglossus, hyoglossus, styloglossus, and palatoglossus--all have bony origins but 
soft-tissue insertions (see Abd-el-Malek, 1939, Hiatt & Gartner, 1987, and Miyawaki, 
1974). An organ composed primarily of muscles is constant in its volume due to the fact 
that a muscle is an incompressible liquid. Thus, a physical change of the tongue in one 
dimension is naturally compensated for by a change in at least one other dimension (Kier 
& Smith, 1985; Smith & Kier, 1989). These physical and biomechanical characteristics 
suggest enormous potential flexibility in movement. 
4 
The intrinsic muscles produce diverse shape changes in the tongue. Through 
contractions of its vertical and transverse muscles, the tongue lengthens, flattens, and 
widens. Through contractions of its superior and inferior longitudinal muscles either 
singularly or together, the tongue shortens and compresses, and its tip elevates or 
depresses. Differentially combined activations of these muscles accomplish still other 
kinematic goals or shape alterations such as bending, curling, stiffening, and leverage 
(Stone, 1990, 1991 ). 
Actions of the extrinsic muscles add further complexity to the tongue's movement 
repertoire. The genioglossus protrudes the tongue, the hyoglossus retracts and depresses 
it, the styloglossus retracts and elevates it, and the palatoglossus elevates the base of the 
tongue (Hiatt & Gartner, 1987). Moreover, because of their insertions into the lateral 
aspects of the tongue and the intermingling of fibers, the styloglossus and hyoglossus can 
cause medial concavity or convexity in tongue shape, respectively. Because of its 
midsagittal insertion into the tongue, when coactivated with the styloglossus, the 
genioglossus can further deepen the central groove (Stone, 1991). 
5 
It must be emphasized that the lingual muscle groups do not function 
independently. The decussations of the intrinsic and extrinsic fibers mandate concerted 
movement efforts. It is only by intricately coordinated activities of several muscles as well 
as interactions with the other structures of the vocal tract that the tongue accomplished 
complex movements with multiple degrees of freedom (Hiatt & Gartner, 1987; Lowe, 
1981; Stone, 1991). 
6 
Like any organ associated with diverse activity patterns, the tongue has an 
intricate neural control system with elements extending from the periphery through the 
brainstem to the cortex. At the periphery, motor innervation to the lingual musculature is 
supplied by fibers of the hypoglossal nerve (cranial nerve XII: for all except palatoglossus) 
and the pharyngeal plexus (derived from cranial nerves IX, X, and XI: for palatoglossus). 
In addition, the tongue is richly endowed with somatosensory and proprioceptive 
innervations (Bowman, 1968; Cooper, 1953; Fitzgerald & Sachithanandan, 1979; 
Landgren & Olsson, 1982; Lowe, 1981 , 1984; Porter, 1966). At the brainstem level, the 
hypoglossal nucleus contains topographical representations for different lingual muscles, 
and receives afferent projections directly from the nucleus tractus solitarius, the trigeminal 
spinal and mesencephalic nuclei, and indirectly from the glossopharyngeal and vagus 
nerves (see Lowe 1981, 1984 for reviews). A recent electrophysiological study has 
confirmed that tactile and gustatory input to the posterior tongue is received in specific 
regions in the nucleus tractus solitarius (Halsell, Travers, & Travers, 1993). In humans 
and nonhuman primates, muscle-spindle afferents from the tongue are also sent to the 
hypoglossal nucleus through the hypoglossal nerve (Bowman & Combs, 1968, 1969; 
Fitzgerald & Sachithanandan, 1979). At the cortical level, as recent neurophysiological 
evidence shows, the tongue region of the primary motor cortex (tongue-MI) contains 
discrete efferent zones, and different changes in tongue shape and position are due to 
recruitment of different tongue-MI sites (Murray & Sessle, 1992a, 1992b, 1992c ). The 
data cited here suggest several potential neural mechanisms through which peripheral 
sensory input of various types may exert influences on the tongue and elicit adaptive 
lingual motor responses at the brainstem and the cortical levels during a variety of 
oromotor behaviors including swallowing. 
Swallowing: Conceptual Framework for Tongue Kinematics 
Neural Organization of Swallowing 
Swallowing, "the transport of a bolus, liquid or solid, from the mouth to the 
stomach" (Hendrix, 1980, p. 1322), is one of the most basic and stereotyped behaviors of 
the animal world . All living organisms must ingest food and absorb nutrients for life 
sustainment purposes. Even in humans, swallowing is performed without effort or 
attention and with a considerable degree of semiautomaticity. On the other hand, it is 
unquestionably a motor sequence of extraordinary complexity, requiring the coordination 
of a great number of muscles, several cranial nerves, and the interaction of the respiratory 
and alimentary tracts. This complexity can be overlooked because of its seeming 
effortlessness and semiautomated processes. To represent swallowing accurately, the 
simplicity and complexity of the behavior must be reconciled. 
Neural control theories. Two theories have been proposed to account for the 
regulation of the neuromuscular sequence of swallowing, namely, the reflex chain theory 
(Dodds, 1988; Dodds, Stewart, & Logemann, 1990; Miller, 1982) and the central pattern 
generator theory (Dodds, 1988; Dodds et al., 1990; Doty, 1968; Dubner et al., 1978; 
Miller, 1982). The reflex chain hypothesis proposes that swallowing is the motor 
7 
manifestation of a chain oflinked reflexes. As the bolus is passed through the oral cavity, 
it stimulates intraoral sensory receptors, evoking the initial oromotor responses. 
Subsequently, as the bolus continues its path through the pharynx and esophagus, 
receptors are stimulated along the way, thereby eliciting series of muscle contractions 
from one phase to the next. Inherent to this theory is the notion that peripheral sensory 
input is crucial in the initiation and modification of the swallowing motor output. 
8 
The central pattern generator (CPG) theory for swallowing is a "branch" of that 
hypothesized in general motor physiology. A CPG is a network of neurons responsible for 
generating the particular pattern of a motor sequence under experimentally induced 
conditions of complete deafferentation (Grillner & Wallen, 1985). Thus, conceptually, a 
hallmark feature of central pattern generators is their independence from peripheral 
sensory influences. In other words, once initiated, the entire sequence of muscle 
contractions associated with swallowing will course through a series of patterned 
neuromuscular responses with little peripheral modulation (Doty, 1968; Dubner et al., 
1978; Sessle & Henry, 1989). Supporting evidence for this theory was derived from 
animal studies in which experimentally induced sensory deprivation did not alter the 
swallowing motor output or the activity patterns of swallow-related neurons (Car & Arnri, 
1987; Doty & Bosma, 1956; Doty, Richmond, & Storey, 1967; Kessler & Jean, 1985; 
Miller, 1972b; Sumi, 1964, 1970b). 
The CPG hypothesis has gained emphasis over the years. However, as that 
occurred in general motor physiology1, the theory also has undergone scrutiny. 
Increasing empirical evidence indicates that many of the swallowing motor components 
can indeed be modified by sensory input (Dubner et al., 1978; Jean, 1984b, 1990; Miller, 
1982, 1986; Sessle & Henry, 1989). As a result, a revised theory proposes that 
swallowing is likely the efferent output of centrally orchestrated motor programs, but the 
generation of such programs is greatly influenced, if not routinely regulated, by peripheral 
feedback . 
The precise mechanisms of interactions between peripheral sensory input and the 
swallowing motor programs remain unknown. It appears that one means of elucidating 
this issue in normal humans where central electrophysiological experimentation is 
impossible is to critically examine specific deglutitive motor elements while varying the 
properties of the material swallowed or the demands of the swallowing tasks performed. 
Neural control elements. Despite the different theories, a number of basic 
elements are known to participate in the swallowing neural circuitry in mammalians. 
These include the peripheral receptors, primary afferents with their sensory ganglia, 
medullary control neurons, motoneurons and their efferent fibers, and specific muscles 
9 
1A large body of empirical evidence exists today, indicating a high degree of 
sensory regulation in the motor systems of locomotion, flight, respiration, and mastication 
in animals (for review, see Pearson, 1987 and Rossignol, Lund, & Drew, 1988). In speech 
research, there is also evidence for sensory influences on the timing and coordination of 
labial, mandibular, lingual, and laryngeal movements (see Abbs & Connor, 1991). As a 
result, the concept of central pattern generator has undergone revision. Today, the 
prevalent view in general motor physiology is that peripheral afferent signals are basic 
elements in the pattern-generating networks under normal operating conditions in intact 
nervous systems. 
10 
(Doty, 1968; Jean, 1984a; Miller, 1982). In addition, there are subcortical and cortical 
regions capable of modulating swallowing (Hockman, Bieger, & Weerasuriya, 1979; 
Martin & Sessle, 1993; Miller, 1982). 
Physiological studies (Miller & Sherrington, 1916; Pommerenke, 1928; Sinclair, 
1970; see Miller, 1982 for review) have identified a variety of oral, oropharyngeal, and 
hypo pharyngeal sites sensitive to peripheral swallowing elicitation. Dorsal surface of the 
tongue is among these sites, the others being the soft palate, uvula, faucial pillars, 
valleculae, epiglottis, posterior pharyngeal wall, and the cricopharyngeal vicinity. The 
responsible receptors in these areas include mechanoreceptors ( especially the slowly 
adapting type, cf. Dubner, Dubner, Sessle, & Storey, 1978), those sensitive to liquid 
stimulation (Shingai & Shimada, 1976; Storey, 1968), taste buds and chemosensory 
endings (see Dubner et al., 1978 and Miller, 1982), and thermosensitive terminals (see 
Chi-Fishman, Capra, & McCall, 1994). Many of the above-cited studies emphasize that 
successful swallow initiation depended on the properties of the stimuli applied and the size 
of the area stimulated. That is, peripheral sensory input must be of a specific pattern or 
strength, and a broad span of mucosa! surfaces encompassing a sufficiently large 
population of sensory receptors must be activated. (For a review of oral receptor 
morphology and physiology, see Capra, 1995.) 
Miller and Sherrington ( 1916), in particular, reported that in order for a bolus of 
meat or moist cotton ( 1 cc in volume), being dragged through the oral cavity with a string 
attached, to even inconsistently trigger a swallow in decerebrated cats, it had to make 
contact with "most of the whole circumference of the faucial opening" (p. 157). If a food 
11 
bolus alone does not necessarily provide all the impetus needed to trigger a response from 
the brainstem neurons controlling deglutition, what other factors must be in operation 
during normal swallowing? The answer may lie in the action of the tongue. Particles of 
the bolus, with their potentially diverse physical and sensory properties (e.g., volume, 
texture, taste, temperature), come in random contact with mucosal surfaces during oral 
preparation as a result of direct tongue action, gravity, or the biomechanical force 
generated by linguodental, linguopalatal, and linguobuccal interactions ( cf. Kennedy & 
Kent, 1988). The tongue itself, in the process of manipulating, forming, and transporting 
the bolus, also makes repeated contact with a wide section of intraoral and oropharyngeal 
mucosa. Therefore, active movements of the tongue and the coupled motion of the bolus 
certainly have the potential of stimulating multiple intraoral and oropharyngeal receptors 
including those located on and in the tongue, thereby contributing to the generation of a 
swallowing response. 
The role of the peripheral lingual elements extends beyond swallow elicitation. 
Through the detection of bolus properties as well as changes in tongue position and 
contact pressure, lingual mechanoreceptors and proprioceptors conceivably also 
participate in fostering modification of the preprogrammed swallowing motor responses 
(Dodds, 1988; Kahrilas, Logemann, Lin, & Ergun, 1992; Porter, 1966). 
Most of the mucosal receptors participating in swallowing are supplied by the 
glossopharyngeal nerve, the trigeminal nerve (e.g. , lingual branch for tactile information 
from the anterior tongue), the superior laryngeal nerve, and branches of the pharyngeal 
plexus. Several studies (Chi-Fishman et al., 1994; Doty, 1951 ; Sinclair, 1971 ; 
Weerasuriya, Bieger, & Hockman, 1979) have shown that electrical stimulation of these 
primary afferents, especially the internal laryngeal branch of the superior laryngeal nerve, 
can elicit swallowing when the requirements for stimulus-pulse parameters are met. 
12 
It is commonly postulated that a "swallowing center" exists bilaterally in the 
medulla, which consists of two parts: dorsal and ventral. The dorsal part is comprised of 
neurons of the nucleus tractus solitarius and the adjacent reticular formation (see Miller, 
1986 for review). Neuroanatomical tracings have confirmed that these medullary regions, 
through the glossopharyngeal and superior laryngeal nerves, have close connections with 
the oral, pharyngeal and laryngeal receptive fields capable of activating swallowing 
(Altschuler, Bao, Bieger, Hopkins, & Miselis, 1989; Astrom, 1953; Beckstead & Norgren, 
1979; Contreras, Beckstead, & Norgren, 1982; Cottle, 1964; Hanamori & Smith, 1986; 
Kerr, 1962; Sweazey & Bradley, 1986). This dorsal network is believed to serve the 
function of sensory integration and transmission of the integrated signals to the ventral 
part (Miller, 1972a). 
The ventral part of the swallowing center is comprised of neurons located in the 
ventrolateral reticular formation adjacent to the nucleus ambiguus (Doty, Richmond, & 
Storey, 1967; Jean, 1984a; Kessler & Jean, 1985; Miller, 1972a, 1982, 1993; Nakayama, 
Neya, Watanabe, & Tsuchiya, 1974; Neya, Watanabe, & Yamasato, 1974; Roman, 1986). 
Nerve cells of the ventral network are thought to act as command or switching 
interneurons and send patterns of outgoing signals to the motoneurons involved. 
Motoneurons participating in swallowing are located in the nucleus ambiguus, the 
trigeminal, facial, vagal and hypoglossal motor nuclei, and the spinal cord segments C 1, 
13 
C2, and C3 (Amri & Car, 1988; Amri, Car, & Roman, 1990; Dodds et al., 1990; Doty, 
1968; Jean, Amri, & Calas, 1983; Kennedy & Kent, 1988; Miller, 1993). Efferent fibers 
from these nuclei activate the peripheral motor elements to produce swallowing. 
The relationship between deglutition and the hypoglossal nucleus is well 
documented. Past studies have shown that during swallowing the entire pool of 
hypoglossal motoneurons is activated (Miller, 1982) with generally strong and prolonged 
discharge spikes (Car & Amri, I 987; Sumi, 1964, I 969b). Electrical stimulation of the 
glossopharyngeal and the superior laryngeal nerves, while eliciting swallowing, can at the 
same time trigger excitatory responses from the hypoglossal motoneurons (Lowe, 1978). 
Although the hypoglossal nucleus is better known for its motor function, its connections 
with brainstem sensory nuclei (described earlier) cannot be overlooked. Direct afferent 
projections from the nucleus tractus solitarius--the dorsal swallowing center--to the 
hypoglossal nucleus represent one of the avenues through which different types of 
peripheral input (proprioceptive, tactile, pressure, and other types) may affect lingual 
motoneuron activities during swallowing and induce adaptive motor responses from the 
tongue. 
Numerous oral, pharyngeal, and laryngeal muscles participate in the swallow 
synergy (see Donner, Bosma, & Robertson, 1985 for review). Electromyography (EMG) 
data from non-human mammals show intrinsic muscles in the posterior tongue and 
extrinsic muscles such as the styloglossus and palatoglossus to be among those that start 
the synergistic swallowing process (Amri, Lamkadem, & Car, 1989; Doty & Bosma, 
1956; Kawasaki, Ogura, & Takenouchi, 1964; Lowe & Sessle, I 973 , 1974; McNamara & 
Moyers, 1973; Miller, 1972b; Miller & Bowman, 197 4 ). Genioglossus, the tongue 
"protruder" , also exhibits distinct discharge patterns during swallowing (Lowe & Sessle, 
1973, 1974; Miller & Bowman, 1974). There is further evidence that styloglossus and 
hyoglossus, the tongue "retractors", coactivate with geniohyoid during swallowing to 
elevate the hyoid bone while the jaw muscles are simultaneously stabilizing the mandible 
(Amri et al. , 1989). EMG data from humans are essentially similar, but with perhaps a 
stronger indication for intersubject variations (Bole, 1965; Cunningham & Basmajian, 
1969; Hrychshyn & Basmajian, 1972; Lowe, Sessle, & Gurza, 1977; Milidonis, Kraus, 
Segal, & Widmer, 1993; Vitti, Basmajian, Ouellette, Mitchell, Eastman, & Seaborn, 
1975). 
14 
Increasing empirical and clinical evidence suggests that suprabulbar mechanisms 
participate in the modulation of swallowing. (For detailed reviews, see Hockman et al., 
1979, Martin and Sessle, 1993, and Miller, 1982). A number of cortical and subcortical 
sites have direct connections with zones of the brainstem swallowing center, as well as 
swallow-related motor nuclei including the hypoglossal (Bowman & Combs, 1969; Jenny 
and Saper, 1987; Jurgens, 1976; Kuypers, 1958a, 1958b; Niimi, Kishi, Miki, & Fujita, 
1963; Porter, 1967; Rossi & Broda!, 1956; Sirisko, & Sessle, 1983; Walberg, 1957), and 
thus have the potential to exert descending excitatory or inhibitory influences. Among the 
documented regions are those in the anterolateral or dorsolateral frontal cortex (Sumi, 
1969a, 1972a, Miller & Bowman, 1977; Penfield & Rasmussen, 1950; see Martin & 
Sessle, 1993 for review of supporting experimental lesion studies and clinical 
investigations). It has been shown that electrical stimulation to these and other 
15 
suprabulbar regions can directly elicit swallowing (Miller, 1920; Miller & Bowman, 1977; 
Sumi, 1969a, 1972b; Walker & Green, 1938), modify elicited deglutitive responses 
. 
(Bieger & Hockman, 1976; Kessler & Jean, 1986; Sumi, 1971 ; Weerasuriya, Bieger, & 
Hockman, 1979), or evoke corresponding activity in brainstem swallow-related neurons 
(Amri, Lamkadem, & Car, 1991; Jean & Car, 1979; Sumi, 1970a). 
Recent recordings of single neuron activity in the tongue region of the primary 
motor cortex (MI) showed swallow-related discharges, suggesting that tongue-MI 
neurons may participate in the initiation and regulation of swallowing (Martin, Murray, & 
Sessle, 1991 ; Murray & Sessle, 1990). Interestingly, a considerable number of tongue-MI 
neurons have been found to have mechanoreceptive fields on the superior surface of the 
tongue (Martin et al., 1991 ; Murray & Sessle, 1992a). Since the cortex has a well-
established, crucial role in the integration of sensorimotor processes involved in complex 
voluntary behaviors, cortical regions including the tongue-MI zone conceivably are also 
important in the mediation of oropharyngeal sensory input for regulating the volitional 
aspects of deglutition, or in the final adjustment of certain kinematic parameters of 
swallowing. 
Physiological Mechanism of Swallowing 
Swallowing is commonly divided into sequential stages or phases based on 
anatomic landmarks and radiographic characteristics of structural or bolus movement 
(Dodds et al., 1990; Hendrix, 1980; Logemann, 1983, 1988; Miller, 1982, 1986). The 
purpose is often for simplicity of description. (See Kennedy and Kent, 1985, 1988 for 
16 
reviews of different proposals of swallowing stages.) It is generally agreed that at least 
three physiological processes exist in a deglutitive sequence, namely, oral, pharyngeal, and 
esophageal (Miller, 1986). 
The oral stage is marked by voluntary division of the ingested material into smaller 
portions, followed by propulsion of the bolus from the oral cavity into the pharynx. 
During the pharyngeal stage, the bolus is propelled from the oropharynx through the 
hypopharynx and upper esophageal sphincter (UES) into the cervical esophagus. The 
ensuing esophageal stage, completely involuntary, involves the transport of bolus through 
the body of the esophagus and the lower esophageal sphincter into the stomach. For 
conciseness, the literature review that follows will focus on the kinematic events occurring 
after mastication and before esophageal transport. (For reviews of mastication and 
mascatory lingual kinematics, see Anderson and Matthews, 1976, Hiiemae and Crompton, 
1985, Thexton, 1992, and Tamura, Ide, and Kamijo, 1981 . For reviews of esophageal 
physiology, see Kennedy & Kent, 1988 and Nelson and Castell, 1988.) 
Oral and oropharyngeal transport of bolus. Most accounts of deglutition describe 
a momentary, preswallow entrapment of the bolus in a central lingual groove (Bosma, 
1957; Dodds et al., 1990; Kennedy & Kent, 1988; Roberts, 1957). This "spoon-like" 
depression is formed by approximation of the tip, blade, and lateral aspects of the tongue 
with the perimeter of the hard palate, and by elevation of the posterior tongue against a 
lowered soft palate. The approximation of the posterior tongue with the velum is said to 
create a high-pressure zone at the glossopalatal sphincter, serving to prevent premature 
bolus spillage (Dantas, Dodds, Massey, Shaker, & Cook, 1990). It should be pointed out 
that the formation of a spoon-like depression on the tongue surface depends on how the 
bolus is held in the mouth. There is radiographic evidence that in some individuals the 
bolus is positioned under rather than on top of the tongue just before a swallow (Dodds, 
Taylor, Stewart, Kem, Logemann, & Cook, 1989). Obviously, the hold position will 
affect how the tongue initiates its transport movement. 
17 
Typically, oral transport of the bolus begins with elevation of the tongue tip 
(Abd-el-Malek, 1955; Cleall, 1965; Dodds et al., 1989, 1990; Roberts, 1957) and 
associated forward movement of the tongue blade (Stone & Shawker, 1986). The 
anterior, middle, and dorsal portions of the tongue then initiate firm, progressively 
expanding linguopalatal contact from front to back, and in doing so propel the bolus in the 
direction of the oropharynx (Dodds et al., 1990; Kennedy & Kent, 1988). As these 
activities occur, the posterior tongue and tongue base move downward and forward, 
opening the valleculae and enlarging the oropharynx (Dodds et al. , 1990; Mcconnel, 
1988). These movements cause the back of the tongue to assume the configuration of "a 
vertically steepening ramp" (Shedd, Scatlitf, & Kirchner, 1960, p. 847). The original 
mid-tongue grooving has also progressed posteriorly to continue the channeling of bolus 
passage (Hamlet, Stone, et al., 1988). From a biomechanical perspective, these lingual 
motions create changes in oral and oropharyngeal volume, and the resulting pressure 
gradients likely also contribute to intraoral and oropharyngeal bolus displacement 
(Kennedy & Kent, 1988; McConnel, 1988). 
Several other kinematic events occur sooner or later during oral transport. These 
include the onset of glottic closure (Shaker, Dodds, Dantas, Hogan, & Arndorfer, 1990), 
18 
velar elongation and elevation, and anterior and superior movement of the hyoid bone and 
larynx (Dodds et al., 1990; Logemann, 1983). 
Pharyngeal transport of bolus. As the bolus reaches the oropharynx, the posterior 
tongue (including the tongue base) rapidly moves backward toward the soft palate and the 
posterior pharyngeal wall in a piston-like motion, exerting force on the bolus, driving it 
through the pharynx with the aid of gravity (Dodds et al. 1990; Ku, Ma, Mcconnel, & 
Cerenko, 1990; McConnel, 1988; Shedd et al., 1960). This rapid and forceful tongue 
motion has also been referred to as a "pumping action"2 that works in concert with 
pharyngoesophageal pressure change to accomplish the bolus transport task (Cerenko et 
al., 1989; McConnel, 1988), or as a "stripping action" that has a close temporal linkage 
with maximal anterior displacement of the hyoid bone (Hamlet, Jones, Mathog, Bolton, & 
Patterson, 1988). 
Successful pharyngeal bolus transport involves a number of other crucial kinematic 
events, some of which serve protective in addition to alimentary functions. Forceful 
contact between the elevated velum and the pharyngeal walls (posterior and lateral) in a 
sphincteric manner is achieved to prevent nasal regurgitation (Dodds et al., 1990; 
Shprintzen, Lencione, McCall, & Skolnick, 1974). Complete vocal-cord adduction, hyoid 
and laryngeal elevation, inversion of the epiglottis, and closure of the vestibule serve the 
purpose of maximal airway protection (Cook, 1991; Ekberg, 1986; Logemann et al., 
1992; Shaker et al., 1990). Relaxation of the upper esophageal sphincter, coupled with 
2This "pumping action" is not to be confused with the abnormal, repeated pumping 
tongue action seen in some patients with dysphagia as an ineffective attempt to transport 
the bolus through the oral cavity (Logemann, 1983). 
shortening of the pharynx as a result of maximal anterosuperior movement of the hyoid 
bone and the larynx, enables entry of the bolus into the esophagus (Cerenko et al., 1989; 
Cook, Dodds, Dantas, Massey, et al., 1989; Jacob, Kahrilas, Logemann, Shah, & Ha, 
1989; McConnel, 1988; McConnel, Guffin, Cerenk:o, & Ko, 1992). 
19 
There is some controversy over the nature and extent of pharyngeal constrictor 
contributions to bolus transport. Traditionally, it was held that the constrictors 
represented an "engulfment" mechanism for the bolus, and that their sequential activation 
in a caudally-directed progression generated the peristaltic wave responsible for bolus 
transport through the pharynx (Bosma, 1957; Cook, 1991 ; Dodds et al., 1990; Ekberg & 
Borgstrom, 1989; Kennedy & Kent, 1988; Negus, 1948; Palmer, Tanaka, & Siebens, 
1988). However, recent manofluorographic evidence has shown that (a) bolus entry into 
the oropharynx corresponded to anterior movement of the tongue base, (b) subsequent 
bolus transit through the oropharynx corresponded to posterior movement of the tongue 
base, ( c) continuing transit of the bolus through the hypopharynx and the cervical 
esophagus corresponded to progressive oropharyngeal pressure increases occurring while 
the tongue base makes firm contact with the pharyngeal walls, ( d) this tongue-base-
generated pressure is preceded as well as accompanied by continuous negative pressure in 
the pharyngoesophageal (PE) segment, and (e) pharyngeal wall contraction and the 
resultant constriction pressure wave begin as the bolus tail is passing through the cervical 
esophagus (Cerenko et al. , 1989; McConnel, 1988). Given such evidence, the current 
view is that the tongue-generated pressure and the negative pressure in the PE segment 
are the principal contributors to pharyngeal bolus transit, and the constrictors contribute 
20 
more to post-transport clearance of residue than to the actual food transport (Kahrilas & 
Logemann, 1993). 
Issues in Swallowing Kinematics 
While decades of research have improved our understanding of deglutitive 
neurophysiology and kinesiology, the information is far from complete. One important 
question that has been only partially addressed is: What peripheral variables affect 
swallowing kinematics (movement duration, amplitude, velocity, etc.) in normal humans? 
Two categories of peripheral variables are of interest to the present investigation, namely, 
bolus properties and task demands. Literature pertaining to these aspects is reviewed 
below. 
Effects of Bolus Properties 
Primary properties of the bolus include volume, consistency, taste, and 
temperature. This study focused on volume and consistency due to the results of pilot 
work and the pseudopalate constraints discussed in Chapter II. The review below, 
therefore, includes only studies that addressed the effects of volume and consistency on 
swallow-related oral, pharyngeal, and laryngeal kinematics. 
Volume. A number of studies, using different experimental methods, have shown 
that bolus volume can influence swallowing kinematics in the pharynx and larynx. For 
example, graded volume increases (2, 5, 10, and 20 cc in most studies) were found to 
induce progressively earlier onset of velar elevation (Dantas, Kern, et al., 1990), later 
onset of pharyngeal contraction (Jacob et al., 1989; Kahrilas et al. , 1992), earlier and 
21 
longer anterosuperior hyoid movement (Cook, Dodds, Dantas, Kem, et al., 1989; Cook, 
Dodds, Dantas, Massey, et al., 1989; Dantas, Kem, et al., 1990; Jacob et al., 1989), longer 
laryngeal elevation and closure (Kahrilas et al., 1988; Logemann et al. , 1992), and longer 
and greater extent of upper esophageal sphincter opening (Castell, Dalton, & Castell, 
1990b; Cook, Dodds, Dantas, Massey, et al., 1989; Cook, Dodds, Dantas, Kern, et al., 
1989; Dantas, Kem, et al., 1990; Jacob et al., 1989; Kahrilas, Dodds, Dent, Logemann, & 
Shaker, 1988). In addition, bolus movement velocity or acceleration, measured in the 
pharynx, was greater for larger- than for smaller-volume swallows ( e.g., from 15 to 50 
cm/sec in velocity for 1-20 cc boluses, or from 460 to 680 cm/s2 in acceleration for 10-20 
cc boluses) (Kahrilas & Logemann, 1993; Ku et al., 1990). Volume variations, however, 
did not affect the velocity and magnitude of propagated pharyngeal contraction, or the 
extent of constrictor contraction pressure (Ekberg, Olsson, & Sundgren-Borgstrom, 1988; 
Kahrilas et al., 1992, 1993; Kahrilas & Logemann, 1993). Furthermore, findings were 
inconsistent on the relationship between bolus volume and the magnitude of hyoid 
excursions (Jacob et al., 1989 vs. Dodds, Man, Cook, Kahrilas, Steward, & Kem, 1988) 
with a suggestion oflarge intersubject variability (Ekberg et al., 1988). 
Past data concerning the effects of bolus volume on lingual kinematics are not 
completely unequivocal. Using X-ray microbeam, Hamlet (1989) found no significant 
differences in the starting tongue posture, lingual movement direction, extent of grooving, 
peak velocity, and movement sequence as a function of varying volumes (5 , 10, and 15 cc 
of water) . Martin' s X-ray microbeam study (1991), on the other hand, reported 
differences in lingual fleshpoint trajectories between 2-cc and 10-cc water swallows. 
22 
Specifically, the smaller volume had a more anterior and superior starting tongue position 
and a smaller range of tongue dorsum displacement, although the overall pattern of tongue 
action during swallowing was essentially the same regardless of the volume ingested. A 
recent study by Kahrilas et al. ( 1993 ), using simultaneous biplane videofluoroscopy and 
pharyngeal manometry, examined tongue contour changes as well as the pattern and 
timing of lingual movement during barium liquid swallows (1 , 5, 10, and 20 cc). They 
found that larger volume swallows induced earlier onset of lingual propulsion, deeper 
central tongue grooving, and greater extent and duration of centripetal movement of the 
posterior tongue and tongue base. In addition, the depth, width, and duration of 
glossopalatal sphincter opening were greater with larger boluses, and bolus expulsion from 
the oropharynx more vigorous. Similarly, Dantas, Kem, et al. ( 1990) demonstrated that 
graded volume increases in both barium liquid and paste induced a progressively earlier 
onset of anterior movement of the tongue base. 
Consistency. The effects of bolus consistency3 have been extensively documented, 
but mostly for esophageal function . With respect to the oral and pharyngeal stages of 
swallowing, some studies showed that viscous or thick barium, in contrast to thin liquid 
barium, increased the duration and magnitude of submental and infrahyoid muscle activity, 
and prolonged hyoid movement, oral and pharyngeal transit, pharyngeal peristalsis, and 
3
"Consistency" in layman's terms refers to texture, thickness, and firmness. 
Technically, the definitions include density and viscosity. Density, the thickness of 
consistency, may be measured in weight per unit volume (i .e. , g/cm3) . Viscosity, 
expressed in centipoise (cp), refers to the degree ofresistance to flow when pressure is 
applied. Water, for example, typically is 0. 89 cp in viscosity, while the viscosity of syrup 
used in different clinical swallowing investigations may range from 48 to 1100 cp, 
depending on the thickness of the preparation. 
23 
upper esophageal sphincter opening (Dantas & Dodds, 1990; Dantas, Dodds, Massey, & 
Kem, 1989; Dantas, Kem, et al., 1990; Gramiak, Kelley, & Gravinia, 1967; Ingervall, 
Bratt, Carlsson, Helkimo, & Lantz, 1972; Ingervall & Lantz, 1973; Robbins, Hamilton, 
Lof, & Kempster, 1992). However, a recent scintigraphic study (Hamlet et al., 1996) 
found shorter "oral discharge time" with a viscous material ( 10 cc of apple juice thickened 
to 1100 cp) than with water ( 10 cc), and no significant difference between the two in 
pharyngeal transit times. Methodological variations may account for the inconsistencies in 
these results (see Hamlet et al., 1996 for explanations). Findings from comparisons of 
solids/semisolids with water also appeared equivocal. While there was evidence that 
solids/semisolids induced longer pharyngoesophageal contractions (Wilson, Pryde, 
Macintyre, & Heading, 1989) and slower opening but faster closing of the upper 
esophageal sphincter (Castell et al. , 1990b) than did water, there was also a report of no 
significant differences in the degree of pharyngeal and laryngeal elevation, pharyngeal 
transit time, and the width and cross-sectional area of the upper esophageal sphincter 
(Ekberg, Liedberg, & bwall, 1986). 
In reference to the effect of bolus viscosity on lingual action, a recent investigation 
reported that barium paste induced a longer period of tongue-base contact to the posterior 
pharyngeal wall in normal subjects than did barium liquid (Lazarus et al. , 1993). In 
addition, a study of tongue force indicated that viscous materials (pudding and mashed 
potato), in contrast to water, increased both the force amplitude during bolus propulsion 
and the pressure amplitude during pharyngeal residue clearance (Pouderoux & Kahrilas, 
1995). 
24 
As part of the consistency issue4, saliva (dry) and liquid (wet) swallows have been 
compared from multiple aspects of deglutition. During the oral and pharyngeal stages of 
swallowing, dry swallows tended to induce: (a) a higher position of velopharyngeal 
contact and a more pronounced levator eminence (Shprintzen, Lencione, McCall, & 
Skolnick, 1974); (b) a longer duration ofhyoid movement (Sonies, Parent, Morrish, & 
Baum, 1988); (c) faster and shorter pharyngeal contractions (Castell, Dalton, & Castell, 
1990b ); and ( d) shorter upper esophageal sphincter relaxation (Castell, Dalton, & Castell, 
1990a). In reference to the tongue, dry swallows were associated with: (a) longer 
genioglossus muscle activities (Cunningham and Basmajian, 1969; Hrychshyn and 
Basmajian, 1972); (b) a higher and more anterior starting tongue position and a shorter 
period of initial tongue tip elevation (Cleall, 1965); and (c) more angular lingual 
movement trajectories, lower peak velocities for tongue blade and dorsum, slower 
sequence of blade-to-dorsum movement, and smaller overall range of motion (Hamlet, 
1989). In most of the above-cited studies, water was the type of wet material used, and 
the amount consumed per swallow ranged from small (unspecified) to 15 cc. Despite 
these reports of dry-wet contrast, a recent study by Shaker et al. ( 1990) found through 
simultaneous videoendoscopy, videofluoroscopy, pharyngeal manometry, and submental 
surface EMG that saliva and 5-cc liquid barium did not differ significantly in a number of 
4The classification of dry swallow is difficult, because saliva is a complex solution 
apparently different in constituents from water or liquid barium (Baum, 1987; Kaplan & 
Baum, 1993). In some studies a dry swallow is considered volume-related, and assigned a 
110 11 or a small, unspecific quantity on a continuum of graded volumes (cf Dodds, Man, 
Cook, Kahrilas, Stewart, & Kern, 1988, and Kahrilas, Dodds, Dent, Logemann, & Shaker, 
1988). The present study includes the dry swallow under consistency with recognition of 
the possibility for consistency-volume interaction in comparisons of dry and wet swallows. 
lingual, submental, pharyngeal, and laryngeal timing measurements (e.g., the interval 
between the onset of vocal cord adduction and that of tongue base movement). 
Effects of Task Demands 
25 
It is generally known that flexible motor systems are capable of adjusting the 
parameters of their activities in accordance with changing task demands. The human 
speech and limb motor systems are among those possessing such flexibility. The act of 
movement parameter adjustment is sometimes referred to as "response scaling" which may 
be accomplished using different kinematic strategies (Bermejo & Zeigler, 1989; Milner, 
1986). When a complex, discrete motor task is repeated sequentially, parameter 
adjustment conceivably occurs due to the additional requirements for sequentialization and 
possible changes in movement coordination. In continuous speech, for example, there are 
ubiquitous instances of coarticulation where the articulatory movements for discrete 
sounds co-occur in space or overlap in time under the principal of motor economy (see 
Borden and Harris, 1984, Browman and Goldstein, 1990, and Lindblom, 1983). 
To characterize gestural interactions during speech, Browman and Goldstein 
( 1990) have proposed a linguistic gestural model in which task-specific vocal tract 
variables are grouped into gestures, and gestures are represented on different articulatory 
tiers . To produce an utterance, articulators on three different tiers of the oral tract--lips, 
tongue tip, tongue body--as well as those on the glottal and the velic tiers execute 
functionally specific labial, lingual, laryngeal, and palatopharyngeal gestures that reflect 
independent vocal tract variables (e.g., lip aperture, tongue tip constriction location and 
26 
degree, glottal aperture). Gestures within the same articulatory tier may blend, and those 
on different tiers may overlap in time, resulting in different types of changes in the 
movement output and the acoustic consequences. An example of within-tiers gestural 
interaction is the changing of Isl to /JI in "this ship" (uttered as /61J'J1p/; cited from 
Catford, 1977 in Browman & Goldstein, 1990), reflecting the smooth blending of an 
alveolar fricative with a palatoalveolar fricative. An example of across-tiers gestural 
interaction is the lowering of the anterior tongue for /a/ during tongue dorsum elevation 
for /k/ in the syllable-utterance /ska/, as observed on cinefluorography (Borden & Gay, 
1979). This particular example, illustrating the temporal overlap oflingual gestures for a 
vowel with those for a consonant in a speech motor sequence, supports Ohman' s 
hypothesis ( 1966) that functional tongue segments exist and can act relatively independent 
of each other during articulation. 
There is yet another set of examples in the speech literature that illustrates the 
differences between discrete and sequential articulatory movements. These examples are 
found in the vocalization of infants. As Kent and Murray ( 1991) pointed out, the 
monosyllabic utterances (e.g., Iba/) of infants represent unidirectional articulatory 
movements in which one or more articulators move from a constricted to an unconstricted 
vocal-tract position. In babbling, the monosyllabic utterances are reduplicated, resulting in 
multiple, sequential, reciprocal articulatory movements that "are motorically similar to 
rhythmical stereotypies" (p. 41 l ). Examples of rhythmical motor stereotypies can be 
found in the repetitive movements of the limbs, fingers, torso, and head in normal infants 
(Thelen, 1981). 
27 
Closely related to the issue of discrete versus sequential motor tasks in both speech 
and limb motor systems is the issue of increased movement speed. Multiple empirical 
examples exist, indicating that when tasks are performed at a faster rate and consequently 
shortened duration, there are corresponding modifications in movement amplitude and 
velocity (Abbs, 1973; Cooke, 1980; Freund & Bildingen, 1978; Gay, 1981 ; Hughes & 
Abbs, 1976; Milner, 1986; Munhall, Ostry, & Parush, 1985; Ostry, Cooke, & Munhall, 
1987; Ostry, Keller, & Parush, 1983; Stone, 1981; Tuller, Harris, & Kelso, 1982; 
Wieneke, Janssen, & Belderbos, 1987). While the common strategy is to modify peak 
velocity and amplitude in a tightly coupled linear manner, findings on the d~tails of the 
kinematic adjustments are not all consistent, especially in speech (see concise review in 
Ostry & Munhall, 1985). Descriptively, a simple yet explicit account of the effect of speed 
on articulation is that from Stetson (1951, p. 71): "At a rapid rate the movements tend 
either to get into step or to drop in order to simplify the coordination" . 
Within the deglutition motor system, the issue of discrete (single) versus sequential 
swallows and the issue of normal versus fast drinking have not been well addressed. A 
few studies examined the effect of rate on jaw movements during mastication and found 
equal acceleration and deceleration durations for fast jaw opening but longer deceleration 
than acceleration for movements at normal speed (Morimoto, Inoue, Nakamura, & 
Kawamura, 1984; Plesh, Bishop, & McCall, 1987; Ostry & Flanagan, 1989). A few other 
studies examined the effect of multiple swallows, but only with respect to esophageal 
responses. They reported that repeated swallows of liquid, separated by short intervals, 
could dramatically change esophageal motility patterns; for example, the second swallow 
28 
could reduce or interrupt the peristaltic wave of the first swallow, and the residual wave of 
the first swallow could also depress or inhibit the subsequent swallow wave (Ham, 
Georges, Froideville, & Piepsz, 1985; Meyer, Gerhardt, & Castell, 1981 ; Vanek & 
Diamant, 1987). 
Swallowing in normal infants during feeding is typically described as a part of the 
suck-swallow cycle which is repeated successively with a brief between-cycle pause for 
breathing (Bosma, Hepburn, Josell, Baker, 1990; Newman, Cleveland, Blickman, Hillman, 
& Jaramillo, 1991 ; Selley, Ellis, Flack, & Brooks, 1990; Selley, Ellis, Flack, Curtis, 
Callon, 1986; Weber, Woolridge, & Baum, 1986). Given the added breathing pauses and 
the variable suck-swallow ratios (from 1: 1 to 3: 1 ), infant swallows do not occur in the 
same "sequential" manner as those performed on one breath during adult continuous 
drinking. In addition, since the tongue must conform to and interact with the nipple for 
sucking, lingual behaviors during infant swallowing are not comparable to those during 
mature, nonstop cup-drinking. Interestingly, the radiographic investigation by Newman et 
al. (1991) showed that the activity of suck-oral transit overlapped with pharyngeal transit 
in all of the infants studied. Oral and pharyngeal transit overlap conceivably can also 
occur during adult continuous drinking, although the extent of the overlap remains 
unknown. 
Little is known about the effects of task demands on the oral and pharyngeal 
kinematics of swallowing. As described earlier, past studies (Kahrilas et al, 1993; Martin, 
1991) have emphasized that the overall pattern of deglutitive tongue action contains 
invariant features . This impression is based solely on the experimental model of single 
29 
swallows. Are these invariant features carried over to the sequential swallows in 
continuous drinking which is a common event in everyday eating? That is, is sequential 
swallowing during continuous drinking performed by merely concatenating single 
swallows? Does fusing or dropping of movement components occur as a function of the 
inherent rate increase? Most models of voluntary human limb movements describe a 
trade-off between movement speed and accuracy, that affects movement control (Hancock 
& Newell, 1985; Meyer, Smith, & Wright, 1982; Schmidt, Sherwood, Zelaznik, & 
Leikind, 1985). Swallowing, on the other hand, is a motor behavior that presumably 
cannot afford much latitude, if any, for movement inaccuracy due to the need for maximal 
airway protection. How, then, does the lingual motor system accommodate the added 
task demands and achieve a balance between continuous motion at an increased rate and 
swallowing safety? All of these questions required answers. Conceivably, when 
additional demands are superimposed onto a basic task, the reaction of a motor system 
would reflect the nature, strategies, and fundamental flexibility of the underlying 
movement control processes. 
Tongue-Palate Interaction and Electropalatography 
However mobile and biomechanically unique, the tongue cannot accomplish most 
of its functions by itself. To perform its rudimentary tasks during swallowing, it must 
interact with other oral structures, most importantly the hard palate. Tongue-palate 
interaction provides not only the primary means for bolus formation but also the anterior 
and lateral seals necessary for bolus containment. Furthermore, it is a dynamic source 
30 
from which the tongue gains stability, expands its freedom of movement, alters its surface 
contours, and derives the force needed for bolus propulsion. Linguopalatal relationship is 
not exclusively biomechanical, as there is also empirical evidence for sensorimotor · 
connections between the two structures (see Hellstrand, 1982b for review). For example, 
mechanical stimulation of the hard palate in cats, such as the application of long-lasting 
pressure, produced intermittent activities in the genioglossus muscle and the motor 
consequence of rhythmic tongue flapping (Hellstrand, 1982a). 
A safe and noninvasive technique, electropalatography (EPG), is available for 
studying the details--location, pattern, and timing--oflinguopalatal contact. The 
usefulness of EPG has been well established in linguistic research and speech therapy. 
Linguistic research has studied phoneme-specific contact characteristics in different 
languages (Fujimura, Tatsumi, & Kagaya, 1972; Hoole, Ziegler, Hartmann, & Hardcastle, 
1989; Mizutani & Hashimoto, 1988; Nihalani, 1974; Palmer, 1973 ), the effect of palatal 
shape on lingual articulation (Hiki & Itoh, 1986), speech motor adaptation and 
compensatory articulatory movements (Hamlet & Stone, 1978), the relationship between 
cross-sectional tongue shape and linguopalatal contact patterns (Stone, 1991; Stone et al., 
1992), and lingual coarticulatory effects in connected speech (Butcher, 1989; Farnetani, 
Vagges, & Magno-Caldognetto, 1985; Marchal, 1988; Miyawaki, Kiritani, Tatsumi, & 
Fujimura, 1974; Recasens, 1991 ; Wright & Kerswill, 1989). In speech therapy, EPG has 
been used as a biofeedback tool to facilitate the remediation of articulatory defects 
(Fletcher, 1985; Fletcher, Dagenais, & Critz-Crosby, 1991 ; Suzuki, 1989). (For detailed 
reviews of EPG applications, see Hardcastle, Jones, Knight, Trudgeon, & Calder, 1989 
31 
and Hardcastle, Gibbon, & Nicolaidis, 1991 ). In recent years, increasingly more methods 
have been proposed for the reduction and quantification of speech-related EPG data, 
among which are the percent contact analysis, the contact index method, the center of 
gravity index, and the coarticulatory index (Byrd, Flemming, Mueller, & Tan, 1995; 
F ontdevila, Pallares, & Recasens, 1994; Gibbon, Hardcastle, & Nicolaidis, 1993; 
Hardcastle, Gibbon, & Nicolaidis, 1991 ). 
In contrast, there has been minimal utilization of EPG in swallowing research. The 
preliminary observations of a recent single-subject study (Jack & Gibbon, 1995) showed 
that EPG had potential usefulness and limitations in the investigation of swallowing and 
eating. As reported, liquid (milk) and semi-liquid (yogurt) were more appropriate than 
bulky or sticky materials Gelly) to use with EPG, since the latter prevented tongue-palate 
contact. While there was some indication that consumption duration varied with bolus 
consistency (longest for jelly), no specific kinematic issue was addressed in that study. 
There remained a need for a detailed investigation on the spatial and temporal 
characteristics of tongue-palate coordination during swallowing. 
A major weakness of EPG is that it provides data on only tongue-hard palate 
contact. Information on the actual movement of the tongue or the bolus is missing. As a 
compensatory measure, the present study combined EPG with ultrasonography for cross-
referencing of contact patterns to lingual actions. Real-time ultrasound imaging is a 
noninvasive technique widely applied in medical diagnostics, speech studies, as well as 
swallowing evaluation and research. For swallowing research, in particular, ultrasound 
has proven to be useful in examining lingual motility, tongue surface contour changes, 
tongue-hyoid interactions, and the effects of different clinical conditions on oral bolus 
management and transport (Hamlet, Stone, et al ., 1988; Shawker et al., 1983, 1984; 
Sonies, 1991 ; Stone & Shawker, 1986; Weber, Woolridge, & Baum, 1986; Wein, 
Beckler, & Klajmans, 1991 ). 
A Theory of Swallowing Motor Control 
32 
Based on the literature reviewed, a plausible theory of swallowing motor control 
would include both central and peripheral mechanisms, and a general deglutitive motor 
program that provided some but not extensive flexibility in performance. Because 
swallowing is a complex motor act with involuntary as well as voluntary components, its 
successful execution is likely to rely on not only a pattern generator in the brainstem but 
also modulation processes at the suprabulbar/cortical levels and sensory input from the 
periphery. That is, the brainstem control mechanism dictates the basic response patterns, 
while the suprabulbar and peripheral mechanisms exert modulatory influences based on 
changes in task nature and demands. However, swallowing, like walking, is not a highly 
skilled motor behavior that requires extensive learning to perfect . Moreover, the 
persistent, stringent requirement for airway safety mandates certain fixed relationships and 
coordinations among many, if not all, of the deglutitive motor participants. These factors 
indicate that the degree of flexibility in swallowing motor control and performance does 
not need to be as extensive as that for expert skills such as playing tennis or piano. 
This research proposes that the master motor program for swallowing must consist 
of both invariant and variant parameters. An example of invariance is the unidirectionality 
33 
of bolus translocation through the mouth and of the deglutitive action sequence of the 
tongue. In contrast, an example of variability is the duration of swallowing events. 
Duration adjustments may be made as a function of the motor goals, tasks, and physical 
conditions at hand so that optimal lingual movement execution may be maintained. In 
general motor physiology, timing and force have been shown to be variable parameters in 
the general motor programs for behaviors such as wrist movements and writing (see 
Gentner, 1987 for concise review). It is reasonable that these parameters would vary with 
task demands in swallowing as well . 
The precise neurophysiological interaction between the central and peripheral 
mechanisms during deglutition remains obscure; however, the invariance and variability in 
swallowing motor control can be explored by investigating movement characteristics. 
Because motor performances are dependent on tasks, by examining the effects of bolus 
properties and task demands on specific peripheral parameters of deglutitive tongue 
action, the fixed and variable aspects of the lingual motor system are likely to be revealed. 
In tum, a better understanding of the degree of flexibility in the general motor program for 
swallowing would be gained. 
Why is this type of research important clinically? First, there is a need for better 
knowledge about the variable aspects of swallowing, especially those under voluntary 
control, to improve the theoretical basis for treatment decisions. Second, information is 
also needed on the actual tongue modifications used by the swallowing motor program in 
adapting to bolus/task changes for more insight into patient management strategies (e.g., 
when to apply thicker consistency, what type of tongue exercises to use) . 
34 
Summa!:)'. 
It has been well established that the tongue plays a vital role in the biomechanical 
events of swallowing, including bolus formation, containment and propulsion, and must 
coordinate perfectly with other structures of the aerodigestive tract to execute a complete 
and safe swallow. Multiple descriptive accounts exist on the radiographic and 
ultrasonographic characteristics of lingual movements during swallowing. In addition, 
there is an abundance of evidence for peripheral modulation of different elements of 
swallowing. Yet, minimal data are available on deglutitive lingual motor strategies or how. 
tongue movements adapt to different swallowing tasks. 
The tongue relies extensively on interacting with the hard palate to fulfill its 
deglutitive functions. The details of swallow-related linguopalatal coordination, however, 
remained obscure. Whether linguopalatal contact activities would reflect the effects of 
peripheral modulation, especially those resulting from changing bolus properties and task 
demands, also remained undetermined. The technique ofEPG offers an avenue to 
examine these issues and this theory. Information on the spatial and temporal 
characteristics of tongue-palate interaction may further our understanding of deglutitive 
lingual motor strategies and provide insights into the underlying control mechanisms for 
swallowing. 
Research Purpose and Questions 
The overall objective of the present study was to document any flexibility that 
might exist in deglutitive lingual motor performance. The direct goal was to examine in 
35 
detail the timing and patterns of tongue-palate contact with associated changes in tongue 
shape as a function of different bolus properties and swallowing task demands. 
Specifically, the following questions were addressed: 
I . Does linguopalatal contact vary in pattern and/or timing as a function of bolus 
properties (volume and consistency)? 
a. Is there evidence for regional specificity in tongue-palate contact during a 
discrete swallow? If yes, does such specificity change when the bolus 
changes in volume and/or consistency? 
b. Are there functionally distinct stages in the EPG series of a discrete 
swallow? If yes, how are the stages affected by changing bolus properties, 
and how do they compare to those well established for radiographic 
swallowing analysis? 
2. Is there evidence for spatial and/or temporal adjustment of movements as a result 
of changing task demands ( discrete swallowing vs. sequential swallowing at 
normal and fast rates), or is continuous drinking performed by simply 
concatenating single swallows? 
3. Are time-varying changes in tongue shape and tongue-palate contact related 
systematically? 
4. Are there bolus- and/or task-induced differences in the patterns of instantaneous 
lingual velocities across time? 
CHAPTER II. METHOD 
The present investigation consisted of two experiments. Experiment 1 focused on 
discrete swallowing (single swallows), and Experiment 2 on sequential swallowing during 
continuous drinking. Simultaneous electropalatography and ultrasound imaging were 
conducted in both. 
Subjects 
Six healthy adults, three males and three females between the ages of 23 and 47 
years, served as subjects. They had no history of swallowing disorders. All passed a 
swallowing screening and an oral motor examination. All were informed of the purpose, 
procedures, risks, and benefits of the study, and signed the required consent forms. 
Instrumentation 
A diagram of the instrumental setup for this study is presented in Figure 2.1 . 
Electropalatography 
EPG data were collected using a Kay Elemetrics Palatometer (Lincoln Park, NJ, 
Model 6300). The Palatometer is a computer-assisted system that senses and displays 
tongue-palate contact in real time during oral motor activities such as speech and 
swallowing. The system consisted of an external input/output (1/0) module, a PC-based 
36 
Figure 2.1. Schematic diagram of instrumental setup. HATS = head and transducer support; PAL = Palatometer; CSL = 
Computerized Speech Lab. 
0 
0 
HATS system 
video for 
camera head 
CSL 
0 
0 
Dli]Oi------------. 
I 'O'O'O 'O 'O TITITI __.---, . 'O'O'O 'O 'nTIT. 
---------, -
Ultrasound 
Imaging 
System 
u CJ 
EPG-Ul!rasound 
synch signals 
EPG display 
with click 
mic 
= = = = 
= 
Q It ] and /a/ 
= "''------>> computer system 
3DDDC 
3DDDE 
digital A-V mixer 
videotaped 
image 
with inserts 
,, 
38 
printed circuit card, connecting cables, and software. A pseudopalate and a ground 
electrode that introduced a high frequency, low amplitude, harmless current to the wrist 
were worn by the subject. Each pseudopalate (Figure 2.2, upper panel) was a 0.5-mm-
thick, acrylic plate, molded according to a custom stone cast to ensure a snug fit. 
Embedded in the pseudopalate were 96 gold-plated contact sensors, each about 1 mm in 
diameter, which were symmetrically placed along the medial surface of the teeth and 
across the plane of the hard palate. Pseudopalate length was maximal for each subject, 
extending 5-10 mm beyond the third molars, but short of the gag reflex mechanism. The 
distance between the electrodes was normalized across palates relative to the subject's 
dentition and palate size. Extending from each electrode was a fine wire (42 gauge), and 
the wires were gathered at the back of the pseudopalate in left and right bundles. The 
bundles were routed around the third molars and anteriorly along the lateral teeth to exit 
the mouth at the front. They were then housed in polyethylene tubing and affixed to the 
Palatometer 1/0 box via connectors. 
With the Palatometer activated, a nondetectible DC current was sent through the 
fine wires to the electrodes which registered tongue-palate contact at a sweep rate of 100 
samples per second. The contact information was relayed to the computer and displayed 
on the monitor in a pattern closely resembling the actual electrode layout on the 
pseudopalate. The uncontacted, individual electrodes appeared on the screen as pale-gray 
dots, while the contacted electrodes appeared as small, red squares. The overall electrode 
threshold was adjusted for the individual subjects to a range determined in multiple pilot 
39 
Figure 2.2. A subject's acrylic pseudopalate (upper panel), and the Palatometer electrode 
sensitivity adjustment screen (lower panel) with the suitable range for swallowing super-
imposed (striped area) . 
Sensitivity Adjustment for Channel 1 
sensitivity index bar (gray) Acceptable 
range for 
swallowing (striped area) 
Acceptable 
range for 
speech 
40 
studies to be sensitive to tongue contact but insensitive to the bolus itself (see Figure 2.2, 
lower panel). For each swallowing event, the contact information was stored at 10-msec 
intervals in a 4-sec record on a hard disk. The stored EPG records were then reviewed 
using specially developed software. 
Ultrasound Imaging 
Real-time, sagittal and coronal ( cross-sectional) images of the tongue during 
swallowing were collected via an Acoustic Imaging Open Technology Ultrasound System 
(Phoenix, AZ, Model AI5200S) at the Ultrasound Division, Department of Radiology, 
Johns Hopkins University Hospital. A 3.5-MHZ convex curved linear array transducer 
was used. Focal depth was set at 8.6, 10.3 or 12 cm, depending on the subject's tongue 
size, thus producing 26-30 scans per second. Scan angle was 70° . 
A specially designed head and transducer support (HA TS) system was used to 
stabilize the subject's head during data collection (see Figure 2.3, upper panel). The 
HATS unit was built at the Department of Electrical and Computer Engineering, Johns 
Hopkins University. (For a technical description of the system and its validation, see 
Stone & Davis, 1995.) As part ofthis system, four well-padded clamps secured the head 
at the front, back, and sides. The transducer was housed in a transducer holder and 
positioned under the subject's chin in fixed relation to the head. An acoustic standoff pad 
(Kitecko, 3M), coated with transmission gel on its superior and inferior surfaces to 
prevent air bubbles, separated the mandible and the transducer. This permitted jaw 
movement without translation of motion to the scanner. 
41 
Figure 2.3. The head and transducer support (HATS) system (upper panel) and transducer 
placement calibrations (lower panel). The subject's head is stabilized by four well-padded 
clamps from the circular ring of the head holder. The head holder is suspended by vertical 
bars from an H-frame at top. The transducer is positioned under the subject's chin in fixed 
relation to the head. The eye-glass frame, the tongue depressor marking the occlusal plane, 
and the long bone of the mandible were used to measure the transducer angle. (From Stone 
and Davis, 1995.) 
42 
Transducer placement calibrations ( details in Stone & Davis, 1995) were 
performed on each subject prior to data recording, after each break, and whenever 
changing imaging plane. The subject's jaw length was measured with calipers from the 
posterior angle to a tongue depressor held flat against the mental symphysis. For sagittal 
scanning, since the emphasis was on acquiring optimal images of the full length of the 
tongue, the transducer was placed 2. 5-3 .3 cm (31-40% of the individual jaw lengths) 
behind the mental symphysis at a 6-8 ° angle posterior to vertical. For the coronal scans, 
the transducer was positioned 2. 7-3 .2 cm (33% of the individual jaw lengths) back from 
the mental symphysis at an 11-14 ° angle. (Information on jaw length and transducer 
placement for the individual subjects is presented in Appendix A.) "Midline" was 
determined by the space between the maxillary central incisors and aligned with the pre-
marked midline of the transducer. The ultrasound beam angle was measured relative to 
the body (long bone) of the mandible, the occlusal plane, and an eyeglass frame with two 
pre-marked crosses on its side bar (see Figure 2.3, lower panel). The occlusal plane was 
determined by having the subject bite down on a tongue depressor with the molars. Worn 
by the subject, the crosses on the eyeglass frame defined an additional plane, consistent 
across subjects, relative to which the transducer angle could be measured. All transducer 
calibrations were done from the subject's right side. 
With the ultrasound imaging system activated, a beam of ultra high-frequency 
sound wave, emitted from the piezoelectric crystals of the transducer, was directed 
through a 1. 9-mm thick section of the lingual soft tissue. The beam was reflected at 
43 
various places within the tongue and especially at its superior surface where the mucosa 
interfaced with air. The reflections were received by the transducer crystals and 
electronically processed. Within the ultrasound instrument a computer converted the 
returning echoes into a video image in which stronger echoes were displayed as brighter 
regions. The sector scanner contained a curvilinear array of 96 crystals that fired 
sequentially, forming a 90° wedge-shaped image. (For technical descriptions of 
ultrasound imaging, see Hedrick, Hykes, & Starchman, 1995; for information on lingual 
ultrasonography, see Stone, 1996). On the video screen, the tongue surface appeared as a 
white curve, while two black cone-shaped shadows--one anterior and one posterior--were 
cast by the jaw and the hyoid bone. Bone refracts the ultrasonic beam. The sharpness of 
the tongue-surface outline varied across subjects and appeared to depend somewhat on 
the subject's lingual tissue (e.g., amount and distribution of fat, oral moisture, tongue 
size). 
The ultrasound data, collected simultaneously with the EPG data, were recorded 
on videotape with voice commentaries. A digital audio-visual mixer (Panasonic, Model 
WJ-MX30) was used to control the input signals. The subject's head was also videotaped 
throughout data collection, and the image was inserted at the lower left corner of the 
screen via a screen splitter (American Dynamics, Orangeburg, NY, DigiSplit Model 
AD 14 79) to enable monitoring of head position and movements. Each frame of the 
videotape was encoded with a digital timing signal in hundredths of a second using a video 
timer (FOR-A, Natick, MA, Model VTG-33). 
EPG-Ultrasound Recording Synchronization 
Synchronization of EPG and ultrasound recordings was accomplished by 
44 
introduci . 
ng two tngger signals: a click from a noise maker and a glottalized /a/ uttered by 
the exper· 
imenter. The choice of two signals was for double-checking. These audio 
signals w . . . 
ere transI11J.tted through a microphone to the Kay Elemetncs Computenzed 
Speech L b (C . 
a SL; Lmcoln Park, NJ, Model 4300) and recorded on a second channel 
sirnuit 
aneously with the EPG data. The Palatometer is designed to use the CSL hardware, 
input device, and powered speaker for the acquisition and playback of acoustic signals. 
The comp . . 
uter-stored EPG record for each swallowmg event, therefore, contamed both 
the conta d 
ct ata and the trigger signals. Through the Palatometer software the acoustic 
Wavefonn of the trigger signals was displayed in a time-linked fashion with the contact 
information. 
The audio trigger signals were also sent to and displayed on an oscilloscope. The 
0 ·1 
sci Iosco · d h · pe screen was continuously videotaped with a separate camera, an t e image 
in~~ . 
at the lower right corner of the video screen for ultrasound data via a screen 
Splitter Th . 
· us, the acoustic waveform was displayed on the ultrasound video frames to an 
accuracy of ± 15 
msec. 
Procedure 
Preparatory Activities 
Subject Preparation 
45 
Each subject underwent a pilot session two days to three weeks prior to the 
experiment. Only the Palatometer was used in this session. The primary purpose was to 
have the subject test-wear the pseudopalate for at least an hour while performing selected 
swallowing tasks. A second reason was to determine if the 30-cc bolus size would present 
any problem. All subjects completed the session without difficulty. They were also 
required to wear their pseudopalates for one to two hours the evening before the 
experiment. 
On the day of the experiment, subjects wore their pseudopalates for at least 30 
minutes during setup and calibrations prior to performing experimental swallowing tasks. 
This allowed adaptation to the palate's presence and to the loss of palatal sensory 
feedback. Previous research indicated that thin (I-mm thick) artificial plates did not affect 
normal articulation after only a few minutes of acclimatization (Allen, 1958). No research 
has investigated the effects of sensory-feedback loss due to an acrylic pseudopalate. In the 
present study subjects were queried before and throughout data collection about palatal 
comfort and swallowing ease. No discomfort or difficulties were reported, although it is 
possible that the pseudopalate had a minimal effect on swallowing. 
46 
Pseudopalate Landmark-Marking on Ultrasound 
Subjects performed non-swallow lingual tasks prior to the swallowing experiments 
for the purpose of identifying the general physical landmarks of the pseudopalate in mid-
sagittal view on ultrasound. Specifically, guided by the experimenter who had visual 
feedback from the Palatometer display, the subject positioned the tongue tip at three 
locations on the pseudopalate, namely, the front, mid, and posterior rim at midline. These 
tongue-tip maneuvers, recorded on videotape, allowed the extraction of three midsagittal 
ultrasound images, each with a matching EPG frame, for use as references during data 
analysis. A representative set of the ultrasound images and corresponding EPG frames is 
illustrated in Figure 2.4. The image in the left upper panel shows the tongue surface as a 
bright white line beginning at the arrow (which marks the approximate point of contact 
between the tongue tip and the pseudopalate) and continuing back and down into the 
pharynx. This particular midsagittal tongue shape corresponds to two rows of front 
contact on the EPG (right upper panel). In the left middle and lower images, a 
discontinuity appears in the tongue surface where the tongue tip (again identified by 
arrows) is angled upward and backward to contact the mid portion and the posterior rim 
of the pseudopalate. 
In addition, several speech tasks were performed for general linking of ultrasound 
tongue shapes with EPG activities. For sagittal imaging, the utterances included /atata/, 
/asasa/, and /akaka/; for coronal imaging, /ajaja/ and /asasa/. Each utterance-string 
was repeated three times. 
47 
Figure 2.4. Physical landmarks of a subject's pseudopalate on mid-sagittal ultrasound (left) 
and corresponding EPG displays (right). Arrows in left upper, middle, and lower panels 
mark, in order, the approximate locations of front, mid, and posterior rim of the 
pseudopalate. All landmarks were identified with tongue-tip movements. In left middle 
panel, tongue tip is retroflexed and positioned high in palatal vault; in left lower panel, more 
pronounced retroflexion is used to reach the posterior target sensors. Discontinuity in 
tongue surface contour occurs becasue the tongue blade is parallel to the ultrasound beam. 
Mid-Sagittal Tongue Contour 
on Ultrasound 
Tongue-Palate Contact 
onEPG 
••• 
••••••• . . . 
. . . 
. . 
• 
• • 
• 
• 
•• • 
• • 
• 
Front 
Mid 
Back 
Experiment 1. Discrete Swallowing 
Tasks and Design 
48 
Tasks for the discrete swallowing experiment incorporated two bolus volumes and 
three bolus consistencies. Specifically, they were: (a) 5-cc water, (b) 30-cc water, 
(c) 5-cc gelatin, (d) 30-cc gelatin, and (e) dry (saliva) swallow. Although it has been 
shown that the average size of a thin liquid bolus is 21 cc (Adnerhill, Ekberg, & Grober, 
I 989), the quantity of 30 cc was chosen in this study to maximally contrast volume 
difference. This decision was further supported by Hamlet's X-ray microbeam study of 
swallowing (1989) in which inconsistent effect of water-bolus size (5, 10, 15 cc) on 
tongue position, grooving and velocity was attributed to the small volume differences. In 
the present study 30-cc water was handled successfully by all subjects, but two subjects 
reported during the post-experiment interviews that the 30-cc gelatin, though manageable, 
"felt large" or "kind of clumsy" when swallowed at once. 
Each swallowing task was repeated six times--three each during sagittal and 
coronal ultrasound scanning. Thus, the experiment was of a subject-by-task, repeated 
measures design (see Figure 2.5). Presentation sequences for tasks and repetitions were 
completely randomized, as was the order of sagittal versus coronal imaging. 
Dry swallows were accomplished by swallowing the saliva present in the mouth at 
the time. In the event of two successive dry swallows, a between-task delay of at least 15 
seconds was introduced. Water and gelatin were presented in syringes. Jello-like boluses 
were prepared with unflavored gelatin (Knox), sugar, water, and red food coloring. The 
Figure 2.5. Subject-by-task repeated measures design for the discrete swallowing experiment. SW= 5-cc water; 30W = 30-cc 
water; SG = 5-cc gelatin; 30G = 30-cc gelatin; sag reps= repetitions during sagittal ultrasound imaging; cor reps= repetitions 
during coronal ultrasound imaging. 
~ S2 u ----+-----+---+-----+-------l 
<D S3 :g s 4 i-----+-----+---+-----t------1 
V) 
Ss i-----+-----+---+------t------1 
S6 .___ _ __.__ _ __._ __ ...._ _ _,_ __ __ 
SW 30W SG 30G Dry 
Swallow Tasks 
randomized, repeated 
50 
use of plain gelatin was necessary to avoid acid-induced activation of electrodes
5
. For the 
gelatin swallows, the syringes were pre-treated by removing the protruded tips and 
creating a hole (6 mm in diameter) at the center of the flat, sanded head. This allowed the 
gelatin to be coarsely mashed while being squirted into the mouth, thus eliminating 
mastication. All gelatin-filled syringes were stored in a cooler with a cooling pack and 
estimated to be 6-10 ° C in temperature. All water boluses were at room temperature 
(23-26°C). Pilot studies indicated no temperature effect6. 
For each swallow, the subject was instructed to keep the tongue in the lowest 
possible position, hold the bolus anteriorly in the mouth without movement, clear the 
entire bolus in one swallow upon hearing the trigger click, and open the mouth slightly 
after swallowing. These prerequisites ensured stable, minimal tongue-palate contacts at 
the start and the finish so that activity onset and termination could be identified without 
ambiguity. Thus, by the requirements of the experimental protocol, the discrete 
swallowing task of the present study consisted of two components: the swallow and the 
post-swallow gesture of mouth opening with tongue lowering. 
5Pilot studies were conducted on three subjects to determine if"taste" could be 
included as one of the variables. The material used included sour (lemonade), sweet 
(sugar water), and plain (tap water) liquid in 5-cc volume and at room temperature. 
Results showed spontaneous, random activation ofEPG sensors by the sour bolus. Due 
to this pseudopalate constraint the "taste" variable was excluded from the present study, 
and plain gelatin was used to prepare the jello-like boluses. 
6Pilot studies were conducted on two subjects to determine the effect of 
temperature on tongue-palate contact. The material used included 5 cc and 30 cc of water 
at room temperature and at 4 °C. Results showed no apparent differences in contact 
timing or pattern for each volume tested. 
51 
Data Analysis 
A total of 30 single swallows were collected from each subject. However, the data 
from one male subject had to be discarded due to an experimental mishap (undetected, 
accidental change in electrode sensitivity setting on the Palatometer). Therefore, only the 
data from the remaining five subjects (S 1 to S5) were analyzed. One 30-cc gelatin 
swallow from S5 was also excluded, because the subject cleared the bolus in two 
swallows. 
All statistical analyses were performed using the SAS System for Windows, 
Version 3.95, Release 6.08 (SAS Institute, Inc., 1992). Unless specified otherwise, the a. 
level was set at .05. 
EPG Data 
Overall analysis framework. A system of analysis was devised, which consisted of 
two aspects: palatal bins and swallowing stages. They were the results of multiple pilot 
studies and provided the spatio temporal bases for detailed examination of EPG data. 
Preliminary examination showed systematic, time-varying regional differences in 
tongue-palate contact activity during swallowing. Based on those differences, six primary 
regions or bins were defined across the electrode array of the pseudopalate, so as to 
capture the salient contact features. They were named: front, central, back, lateral, 
medial, and midline (Figure 2.6). These bins formed two continua: front-to-back and 
lateral-to-midline in a horseshoe configuration. For descriptive purposes, the medial and 
the lateral bins were each subdivided into inner and outer columns. In addition, two 
52 
Figure 2.6. Compartmentalization of the pseudopalate into six primary bins: front, central, 
back, midline (shaded squares), medial (black squares), and lateral (white squares). The 
lateral and medial bins are each further divided into outer and inner columns. 
D 
D 
D 
D D IJ D D 
Front 
o•o D D 11 Doo 
,,___D_ D • • mm • • -----0----" 
D • mill • D 
•• • •• o . • mi • . o 
• 
D • 
.. ; .. 
• D 
D • 
D • 
D • 
D 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• D 
• D 
• D 
• D 
D 
D 
D 
D 
Outer Inner Outer Inner Inner Outer Inner Outer 
Lateral Medial Midline Medial Lateral 
Back 
secondary bins were extracted by dividing the pseudopalate into left and right halves, 
excluding the midline. These bins enabled the observation of symmetry. 
53 
Preliminary examination also showed that the EPG time-series for the discrete 
swallowing task, though representing a continuous physiological event, contained four 
distinct stages that could be distinguished by their time-varying contact patterns. 
Boundary frames for each stage were extracted through visual inspection (intrajudge test-
retest reliability= 95.1 %). These boundary frames showed specific, repeatable features, 
and represented a change from the preceding pattern. 
The four successive stages were named, in order, prepropulsion, propulsion, full 
contact, and withdrawal. The first three stages reflected the swallow component of the 
task performed, while the fourth stage presumably reflected the post-swallow mouth-
opening component. Figure 2. 7, a time-series display of the raw EPG data for a 
representative task at 10-20 msec increments, illustrates the delineation of the four stages. 
The criteria used for their identification were: 
1. Stage I - prepropulsion. The boundary frame marking the onset of 
prepropulsion had to show a change in tongue-palate contact and be preceded by a 
steady-state period of stable or no contact. Subsequent frames in this stage must show 
continuous changes in tongue-palate contact location and number. The changes did not 
have to be systematic. 
2. Stage II - propulsion. The boundary frame marking the onset of propulsion 
began a sequence of progressive, front-to-back activation of a steadily increasing number 
of electrodes until maximum contact was reached. The onset frame had to have a 
Figure 2.7. Raw EPG time-series in 10-20 msec increments for a representative discrete swallowin$ task (including the swallow 
and post-swallow mouth opening gesture). Temporal boundaries for the four stages are: prepropuls10n (I) 2. 71-2.90, propulsion 
(II) 2.91-3.02, full contact (III) 3.03-3 .21, and withdrawal (IV) 3.22-3.41 . Black squares are activated electrodes. The contact 
seen in Frame 2.70 had been unchanged for the preceding270 msec. 
2.70 .... :: . . 
. . • . . 
. :·:.::::::.• 
. . . . 
.. . 
2.a:. / ):·::- . 
·• 
.. :·;_:::::: -• 
. ·: .· .· . ·. ·:. 
. . 
.. 
. .. 
2.98 .... ,·.• •• 
·=~·= =·:=· 
.·:· :.•:•:. ··:. 
...... ... • ... 
:·. . . . ·:. 
a I I I 
.. . I 11 
.. 
3.1 1 • .• ·.,· •':.•. 
• ·.~:= :::,•. 
•a a lal I 1• :.· ... : ,. · ..... 
•• • • • • •• 
·a: : a I II I ' 
11 I a I I' 
•• ! • I •• 
3.27 . :.: : 
·,. -::: ..•. 
.·:: ·.;. .. : ..... :·.·. 
. -:. : · .. ~·. 
.. ::. :· ... 
•• • • ••• 
.. I ,· I u 
2.71 
I 
. . 
. . • . . 
......... 
.. . ;•· ... . 
. .. . 
.. . . .. 
.. 
·• 
2.89 .• ·.: :, . 
.. ... .. 
. ·:,:_: ·-:-:::.:::_:.\ 
. .. 
.. 
2 99 •• • 
. • ,:::1:::,, 
.......... 
. :: ,::,;, :· .. 
.... . ·. ·• ···. 
.: : . . ·. · ..
• • 
•• • • •• 
.. .. 
3.13 .:.,·, •• 
• ·=~== :;::,. 
••••••••• 
. ·:.· .. : ,; .. -.·. 
la I a I a• 
-.:·. :· .. . 
.. . . .. . 
•• • I I II 
3.29_.:-;:,. 
-.;· -:: :. -~.-
:: · ~· : ·.· ..... 
• • • • • • 
a a a I a • a a 
• • • • •• ::•1 ,·:: 
2.73 . :.:: . . 
.. • . . 
.::·-:.:~:.:::·.\ 
,I 
.. 
·• 
2.91 ...... • .• 
II . .:: :::: ::~-. 
. ·:·.: :,-':'·: ::·.:·. 
.  
3.00 -~·1··· I·=~:: =:::. I 
.:• ···=··· ... . .. : ... ·..  
•• • • • •• 
.... : : · ... 
.. . . .. 
.. .  
3.1 s_. :.1· :.,. •••••••• 
•.: ·.·=·=·.·:.· 
. ·•·.· .. : .. · ..... 
•• • • • •• 
.. : : . : · ... 
" I I 
.. 
3.31 
• I 
... 
.. 
·,: · -:::-· :.· 
11 I •. • .• I 1• 
:• .. · . · .. ··. 
I I a I I I 
11 I a I I.._ 
::·, ,·:: 
2.75 .:.:: .. 
. .• .. 
. :·-:::- ·:. 
. ·: . ·. ·. ·.·  
.. .. 
.. .. 
·• 
2.92 •• , •.••• 
••. . ·=· .•• / _:·-:-=::.::: ·.·. 
. ·:. 
. • 
.  
3.01 .... ,., •• 
·=~·· =·::· ::• ::;•:. ··:. 
. ... : . •. · ... 
11 I a I I 11 
.. . . : .... 
::• ! ·~: 
3.1 7 .... ,., •• 
.: .. : =·==· 
... ······ ... 
.·.· . ·.· : ... •. ·:. 
11 II I I II~ 11 
.. . . . . ... 
. .. 
.. I • I 
3.33 .:.::,. 
. ... . 
. .. 
.. 
.  :·.:::- ·: .• 
:: : ·: : \" ·.·.·. 
.. . . . .. 
••• • • • ••• 
... . ... 
.. -
2.77 .:.::, . 
:.:~ :,: : 1 :) (:• • 
,I 
.  
·• 
2.93 .... ..  . 
..... :, .. ·. 
[ .:·-:.=:=.::·:,-.\ 
.  
. 
3.02 .• ,,., •• 
····• ..... 
··=. ·=·=··=·· .. .. . .. .. 
:·: .·: ·. ·. ·· . 
"II I I a I Ill 
I I I I 
::•1.,·~= 
3.19_.:.,· ••••• 
·~=~=:.::=:: • 
'I I •1 I 1• I . • 
11 I I I I I I ' 
. .. : · .... 
11 I I a I a • 
::•1.1•:: 
3.35 . :.::,. 
. .. . 
. ······· .. ·· -::· .. ·· .. 
. ·•· : : . ·. ·. ·•· . . . 
... . .. 
... 
•• 
. .. 
-
2.79_. :-;:,. 
. :·-.: ::: ·: · .. 
. . . · ... 
. .. 
2.94 .... i'·· 
.·=~:~:! :~; -• 
. :.. : . · .... 
.. : . 
. . . 
·• 
.. 
·• 
3.03 -~,·1 .. 
Ill ·=~-= :,::• 
• •••••••••• ::,·::•:.•:. 
.. : : : ·. · ... 
•• • • • •••• 
... . .. 
.• I I I .• 
3.21 .... , ..... 
.·=~== ::::. • 
-::•.::·-.·=-·· 
.•,• I I : I • •.·. 
..::. :-. .. 
.. . . ... 
.. ~ • I •· 
3.37 . ;.::,. 
. .• .. 
. ........ . 
,:: ... : · .. : · .. 
. ...... . . 
.. . . . . 
.. : · ... 
' I' 
•· 
2.81,:):"::- . 
.. ... .:·-:.:\:,:::· .. · 
.. 
.. 
·• 
2.95 a,;.il I 
···~:,;,· . 
........ ..... 
. .. . : . · .. ..• 
.. : . . · ... 
.. . .. 
.  
.. 
3.05 . J,,· 1 •• 
••••• ••••• 
a • . • • '••• I 
.:: •::-.-:-.. 
.... I : : 
1
1 I l~I 
..: .. : · ... 
.. . . .. 
u I I ·• 
3.22 -~,· i,, IV•"••• •,••, 
•,: ··=·=··=·· 1 I •1 I 11 I I' 
11 I a I I I 1• 
11 I I a 11 
.. : . : · ... 
.. I I I I I' 
u . • I •· 
3.39 .:.: :,. 
. . . . 
......... 
. •::·.: ::····: :: ..• 
.. .. 
•· 
. .. 
·• 
2.96 .,\ii,. 
• •••••••••• 
..:·.- ::~·:=·· 
... : . · .... 
•·· . . .. • ·· . I ' I I 
.. . . ... 
.. 
·• 
3.07 .:.,·1 •• 
,:-.: =·::· 
. .. ······ ... :: • ... = •••• ·•· . 
11 II I I II II 1• 
•1 I a I I I 1• 
... . ... 
•• • • I •• 
3.23 .• ·. ::,. 
·=~=i:•:!• 
• •••••••••• 
.. . ... . •: ... 
.·: .. : .. ·.·• 
.. : . . : · ... 
... . .. 
.. • I I •· 
3.41 . ;.::, . 
. .• .. 
,;: ·:::·:· . 
.. . . ·: .. 
. ,, 
.. 
. .. 
·• 
2.91-.,,·1 •• 
••••• ••••• ....... , ... 
..·: .. ~:· .. :· .. 
11 I • I • a• 
.. . . .. 
.. ~ . · ... 
. .. . .. 
·• 
3.09 .... ,., •• 
••••• ••••• .. ........ . 
.. : •/:• .. :· .. 
.··. : : · .. ·· . 
.: ..... · ..
I I I I 
'I I I I a 1• 
.. . • I ·• 
3.25.; ::::·· :. 
·~=···~i··=~. 
.··. ·:: ·.·. ·· . 
1 1
1 I I I I 181 1 
11 II I I I II al 
11 a I I I I 1• 
.. I • I .. 
3.42 . :. : : .. 
. ... 
····· : : .... : ··. · .. 
55 
minimum of three adjacent contacts within two successive rows in the front or central 
bins, since the activation of one or two electrodes might not begin the continuous 
propulsive motion. This stage ended with activation of the posterior-most electrodes in 
the midline bin or the medial bin (inner column), whichever occurred later (see Figure 2.6 
for bins). 
3. Stage III - full contact. This stage consisted of a period of maximal, essentially 
stable palatal contact. The full contact stage did not require activation of all electrodes or 
extreme stability7. 
4. Stage IV - withdrawal. The boundary frame marking the onset of withdrawal 
began a sequence of progressive, front-to-back deactivation of a steadily increasing 
number of electrodes. The end of withdrawal was defined as the first frame with a stable, 
minimal number of contacts that lasted at least 60 msec, irrespective of subsequent, 
additional deactivation. These criteria were set to establish a common reference across 
bolus conditions and subjects. 
Measurements and statistics. Two sets of timing measurements were made: 
1. Total duration. Operationally, the total duration of the discrete swallowing 
task was the interval from the onset ofprepropulsion to the end of withdrawal on EPG. 
Analysis of variance (ANOVA) was performed on the total durations of 149 task 
responses, using a random effect approach (where "subject" and "subject-by-task 
interaction" were specified as naturally random effects). Post hoc contrasts were done 
7
_Flickering ( transient on and off) of some electrodes, mostly in the front and the 
lateral bms, was noted occasionally during this stage, probably as an instrumental artifact. 
56 
with the least-squares means, using the Scheffe method. The specific contrasts included: 
pairwise comparisons of all five swallowing tasks, comparisons between the two sizes 
(5 cc vs. 30 cc), and comparisons among the different consistencies (water, gelatin, and 
dry) . 
2. Stage duration. Based on the boundary frames described earlier, the duration 
of each stage was calculated for all task responses, and the raw data were collapsed by 
stage into mean durations for different boluses. The collapsed data were then subjected to 
a multivariate analysis of variance (MANOVA), using a== .05. The multivariate approach 
was applied to be consistent with the repeated measures design of the experiment. 
However, MANOVA of bolus-by-stage interaction could not be performed due to 
insufficient error degrees of freedom; therefore, univariate tests of interaction were 
conducted. Planned comparisons of the task means were carried out for Stages II and III, 
using a == . 0 I . The a level was more stringent because of conducting multiple contrasts 
for each of these stages. Stages I and IV were examined visually but not analyzed 
statistically. This was due to concerns over the potential artificiality added to these stages 
by the required starting and ending tongue postures. That is, normally the tongue may not 
assume a low position at the start of a swallow and may remain elevated at the end of a 
swallow. 
To extract contact patterns, two data reduction processes were used: 
1. Percent-contact computation. The percentage of tongue-palate contact(% of 
electrodes on) in each palatal bin was tallied at I 0-msec intervals over time for each task 
response, using software developed for the Department of Linguistics, University of 
57 
California, Los Angeles (Mueller, 1992). The data for each stage were then transformed 
into a waveform displaying percent-contact per bin as a function of time. 
2. Time-warping and averaging. Because each subject repeated each discrete 
swallowing task six times, the repetitions must be averaged to extract the underlying 
patterns. Since no two swallowing responses ever can be of exactly the same duration 
( same number of frames), one cannot simply average the frames in one sequence with 
those in another. To create repetitions with the same number of frames, time-warping was 
performed. Time-warping, also referred to as time-aligning or time-normalizing, has been 
employed in speech research for similar purposes, and may be accomplished via different 
mathematical approaches. (For examples of some of the approaches and applications, see 
Cordaro, 1991, Fallside and Woods, 1985, Sakoe and Chiba, 1978; and Smith, Goffinan, 
Zelaznik, Ying, and McGillem, 1995.) In the present study, the time-warping procedure 
was as follows: For each stage, mean durations were calculated for the six repetitions of 
each swallowing task. Next, the stage duration for each repetition was time-aligned 
against the subject's mean. That is, each repetition was either stretched or shortened in 
duration to match the mean duration. Piecewise linear interpolation was used to add or 
delete frames from the warped data, and to recompute the percent-contact for each new 
frame. This method was appropriate, because the raw data showed essentially linear 
frame-to-frame changes in contact. 
Comparisons of pre- and post-warp waveforms showed satisfactory retention of 
percent-contact characteristics in virtually all cases. Two examples, presented in Figure 
2.8, illustrate cases where the pre-warp data series were "stretched" (left panels) or 
Figure 2.8. Examples of pre-warp (upper panels) and post-warp (lower panels) data confonnity in two representative 
discrete swallows. All waveforms show percent-contact changes in the front, central, and back bins druing Stage I. 
Left panels illustrate a case of "data stretching" (from 80 to 260 msec) in one swallow; right panels a case of "data 
shrinking" (from 350 to 170 msec) in a different swallow. 
Example 1: a case of "data stretchin " 
-+- Front ---- Central -1r-- Back 
1 00 l Pre-warp 
.. 75 
u 
J! 
a 50 
u 
,!!. 25 
O ffl~~-----+-----+------+-- --+---~ 
msec O 50 100 150 200 250 300 
~ Front ---- Central -1r-- Back 
100 l Post-warp 
I 75 I 
~::~~~ 
0 
msec 0 50 100 150 200 250 300 
Example 2: a case of "data shrinking" 
~ Front ---Central -1r-- Back 
100 Pre-warp 
.. 75 
u 
J! 
C 50 0 
u 
~ 0 25 
0 ---~~~~:_:__--t----t- ----1 
msec 0 50 100 150 200 250 300 350 
~ Front ---Central -Ir- Back 
100 Post-warp 
.. 75 
u 
c,r 
.. 
C 50 0 
u 
~ 0 25 
0 
msec 0 50 100 150 200 250 300 350 
VI 
00 
59 
"shrunk" (right panels) considerably (by>½ of original durations) to achieve time-
alignment . They represented about 15% of all cases, occurring mostly in Stage I. In 78% 
of the cases, the amount of time-warping was small (<1/3 of pre-warp durations). Time-
warping amount for the remaining 7% of cases fell in between. 
Subsequently, the time-warped percent data were averaged by frame for each bin 
and each stage to extract an underlying pattern for the six repetitions of a single 
swallowing task. Frame-by-frame standard deviations were also calculated. Waveforms 
of the averaged percent-contact data were examined in detail and compared across 
subjects and swallowing tasks. No statistical test was performed on the waveforms. 
Two additional measurements were made, including: 
1. Lateral percent-contact at Stage-II onset. Preliminary examination showed that 
some swallows differed in the extent of lateral contact at the end of prepropulsion or the 
start of propulsion. The underlying issue was whether a prominent lateral seal was present 
just before propulsive activity. Thus, measurements were made of the lateral percent-
contact at Stage-II onset. Based on the computed percentages, the individual swallows 
were categorized into Patterns 1 and 2 (1 = <65%, i.e., without prominent lateral seal; 
2 = >65%, with prominent lateral seal) and verified against the raw EPG time-series data. 
It should be noted that while the cutoff percentage was set at 65, the majority (95%) of 
the raw data displaying Pattern 2 had more than 70% of contact at Stage-II onset. 
Subsequently, the distribution of Patterns 1 and 2 was examined as a function of 
swallowing tasks. Chi-square analysis was planned but could not be successfully carried 
out due to exclusive distribution of one pattern or another to particular tasks. 
60 
2. Order of percent-contact peaks during Stage II. An additional measurement 
was made to assess task effects. This involved identifying the order in which the front (F), 
lateral (L), and back (B) bins reached their respective percent-contact peaks during 
propulsion. A percent-contact peak was defined as the first occurrence of the highest 
percentage in contact for a respective bin within the demarcated stage limits. The front, 
lateral, and back bins were selected for this analysis, as jointly they were most revealing of 
the general contact pattern observed during Stage II. All patterns of sequencing were 
tallied. Because the predominant pattern followed an F-L-B order, the individual 
swallows were classified into an "F-L-B" group and an "others" group. Chi-square 
analyses were performed to examine the distribution of these groups as functions of 
subjects and swallowing tasks. 
Ultrasound Data 
The videotaped ultrasound data were digitized in real time via a Macintosh-based 
image acquisition system (Scion Corp. , Frederick, MD, Model 1200). Frame-by-frame 
examination of the digitized images was accomplished using the NIH Image software, 
Version 1. 5 5 (Rasband, 1994 ). The midsagittal ultrasound scans were used for all of the 
observations and analyses described below. Unlike the midsagittal images that showed 
various movements of the tongue across its length, the coronal images showed repetitive 
and strictly upward-downward movements of a small crosswise tongue section. The latter 
offered no global illustration of swallow-related lingual actions and could not be used to 
cross-refer to the EPG data. 
61 
Overall analysis framework. A system of analysis was devised, which consisted of 
two schemes: fu1:1ctional segments and radial grid. These schemes provided the bases for 
describing ultrasound tongue movements and for cross-referring to the EPG data. 
For convenience of descriptive analysis, the tongue was divided into five functional 
segrnents--anterior, mid, dorsal, posterior, and root ( cf Stone, 1990; see Figure 2. 9, upper 
panel). The root of the tongue, not visible on ultrasound, was not included in the present 
study. 
A transparent, radial grid was superimposed on the monitor screen over the 
ultrasound image (see Figure 2.9, lower panel; modeled after Stone, Sonies, Shawker, 
Weiss, & Nadel, 1983 and Morrish, Stone, Shawker, & Sonies, 1985). Once in place, the 
grid was not removed until all analyses were completed to ensure constant positioning 
over all frames . The grid consisted of a baseline plus six radii angled at 55, 70, 85, 100, 
11 5, 125 degrees from left to right. For placement, the baseline where all radii converged 
was aligned with the scan vertex, and the two outside radii aligned with the scan borders. 
The angles of the four inside radii were set so that their intersections with the surface of a 
"neutral" tongue (resting without bolus in mouth) approximated the centers of the 
respective tongue segments. This grid setup permitted systematic observations and 
measurements of instantaneous radial displacement for the tongue segments. More 
specifically, for each tongue segment on each frame, the measured displacement was the 
x,y coordinates of a point where the tongue surface and a particular grid radius 
intersected. 
62 
Figure 2.9. Schematic of functional lingual segmentation (upper panel, modeled after 
Stone, 1990), and grid system for tracking radial displacement of the anterior, mid, dorsal, 
and posterior tongue segments (lower panel). The tongue root, not visible on ultrasound, 
was not measured. In the lower panel each inside radius of the superimposed grid intersects 
the approximate center of a tongue segment. 
55° 
anterior 
70° 
mid 
85° 
dorsal 
100° 
posterior 
115° 
125° 
baseline 
63 
Observations and measurements. 
l . Ultrasound lingual behaviors versus EPG contact activities. With the grid in 
place, swallow-related movements of the anterior, mid, dorsal, and posterior segments 
were visually tracked by advancing the ultrasound frames one at a time, and their 
characteristic patterns noted. Comparisons were then made between EPG contact 
activities and ultrasound lingual behaviors, and the similarities and differences described. 
2 . Task effect on tongue shape. To enable the examination of changes in sagittal 
ultrasound tongue shape as a function of swallowing tasks, criteria were set up to extract 
specific frames from each swallow for within- and between-subject comparisons. The 
frames selected included: the frame just before the onset of any lingual movement, which 
displayed the tongue in its resting posture with the bolus collected in front; the first frame 
showing the entire bolus on (i.e., no longer in front ot) the tongue surface; and four 
successive frames (identified using the radial grid) showing the point of maximal 
displacement of the anterior, mid, dorsal, and posterior tongue segments. No subsequent 
frames were selected, because "task effect on tongue shape" became irrelevant when the 
bolus left the oral cavity. 
3. Lingual velocity profiles. Using the same radial grid, frame-by-frame 
measurements of displacement were made for the four tongue segments from the swallow 
onset through the end of propulsion (i.e., the point of maximal displacement of the 
posterior segment). The displacement data derived from the grid were used to compute 
instantaneous velocities. Subsequently, a velocity profile--a waveform displaying changes 
in velocity over time--was plotted for each tongue segment in each swallow. However, it 
64 
should be pointed out that the accuracy of frame-by-frame displacement measurements 
depended greatly on the sharpness of the tongue surfaces imaged. Because of 
considerable variability in image quality across subjects accurate measurements could only 
be maintained with the images from Subject 5 (S5). Therefore, only the velocity data of 
this subject were included. 
The velocity profiles of S5 were examined as a function of swallowing tasks. 
Means and standard deviations were calculated by tongue segments and swallowing tasks 
for the "positive peak velocity" and for the "interval from onset to positive peak velocity 
(OTP)". Also computed were the differences in mean OTP values between tongue 
segments. No statistical test, however, was performed. 
4. Correlation ofEPG timing measurements with ultrasound activity. To ensure 
data accuracy, steps were taken to determine the general consistency between EPG timing 
measurements and corresponding sagittal ultrasound tongue shapes. They included 
mathematically aligning the measured EPG events to corresponding ultrasound frames, 
and comparing the tongue shapes depicted on those frames against the particular contact 
activities displayed on EPG. The extent of the "match" was noted. 
5. Identification of unusual lingual behaviors. The ultrasound swallowing data of 
all subjects were screened independently of the EPG data for any unusual lingual behaviors 
such as "double swallows" or excess movements. In cases where the EPG data were 
questionable, the ultrasound images were examined in detail for elucidation. Questionable 
EPG data included task responses where the measured durations seemed unusually long or 
_..-:::=---
short for one or more stages, and occasional cases where contact activities were not 
consistent with the typical observations. 
Experiment 2. Sequential Swallowing in Continuous Drinking 
Tasks and Design 
65 
Tasks for the sequential swallowing experiment included drinking 200 cc of water 
at normal and fast rates. No precise control was exercised over the speed of drinking. 
For the normal rate, subjects were instructed to drink normally upon hearing the trigger 
and swallow consecutively with no between-swallow pauses. For the fast rate, subjects 
were simply asked to "drink as fast as you can." Water was measured and presented in a 
400 cc plastic beaker with part of its wall cut off and sanded to enable tilting of the 
container without tilting of the head. 
One trial at each rate was implemented in randomized order for both ultrasound 
imaging planes. In addition to drinking rate, a naturally present variable was swallow 
order. The starting tongue position for the first swallow was likely to be different from 
that for the subsequent swallows. Thus, Experiment 2 was of a two-level, randomized 
block design with respect to sagittal/coronal ultrasound data. With respect to the EPG 
data, the experiment was of a repeated measures design, as there were two repetitions for 
each rate. 
.. 
66 
Data Analysis 
~ 
EPG Data 
Data from Experiment 2 were examined across subjects and as functions of 
drinking rate (normal vs. fast) and swallow order (first vs. subsequent). In addition, they 
were compared with the discrete/single water swallows from Experiment I . The analysis 
schemes and steps used here were essentially the same as those for the discrete swallowing 
experiment. However, time-warping and frame-by-frame data averaging were not 
performed due to limited sample size. For contact patterns, therefore, the raw EPG 
percent-contact waveforms were examined. 
Measurements. For contact timing, the total duration of each component swallow 
or task response (up to three swallows due to instrumental limit) within a continuous 
drinking series was measured, using the same operational definition specified for discrete 
swallowing. That is, for Experiment 2, this measurement also included contact activities 
from prepropulsion through the end of contact retraction. Mean total durations were 
calculated for the first and the subsequent swallows at each rate, and ANOV A was 
subsequently performed on these means for the effects of "order" and "rate". In addition 
to total duration, a measurement was made of the average total-duration ratios of 
sequential versus discrete water swallowing tasks for comparative analysis. 
For contact pattern, in addition to the analyses on "lateral percent-contact at 
Stage-II onset" and "order of percent-contact peaks during Stage II" (see descriptions 
under Experiment 1), a third measurement was made of the "percent-contact ratios at 
maximal midline" to contrast sequential and single water swallows. The idea was derived 
from preliminary examination of the data. The procedure involved: (a) extracting the 
percentages of electrodes-on in the front, back, and lateral bins at the time of maximal 
midline contact (i.e. , end of propulsion); and (b) calculating the front:back and the 
lateral:back ratios in percent-contact. ANOVA was subsequently performed on each set 
of ratio data. 
Ultrasound Data 
Observations and measurements. 
67 
1. Swallow volume estimation. The sagittal ultrasound data of each subject were 
used to estimate the average bolus size for sequential swallows. The basic assumption 
was that the intake amount varied minimally from swallow to swallow within a continuous 
drinking series. The total amount consumed (200 cc) was divided by the number of 
swallows in a series to obtain an estimated swallow volume. 
2. Lingual movement characteristics. Tongue movements during sequential 
swallowing were examined as functions of drinking rate and swallow order, and compared 
with the characteristics observed for the discrete water swallowing task. Similarities and 
differences were described. In addition, specific ultrasound frames were extracted for 
each component swallow (see criteria specified under Experiment 1) for the examination 
of task-induced tongue shape differences during continuous drinking. 
3. Lingual velocity profiles. Procedures for velocity analyses on the sequential 
swallowing data were the same as those described for Experiment 1. However, the data 
set for each drinking series included values extracted from the first four sequential 
swallowing responses. Descriptive comparisons were made between normal and fast 
drinking rate, between first and subsequent swallows, and between sequential and single 
water swallows. No statistical test was performed on the single-subject data. 
68 
CHAPTER III. RESULTS 
Experiment 1. Discrete Swallowing 
EPG Findings 
I._ongue-Palate Contact Timing 
Total duration: task effect and subject variability. Average total durations of the 
5-cc water, 30-cc water, 5-cc gelatin, 30-cc gelatin, and dry swallowing tasks for Subjects 
1 to 5 (S 1 to S5) are presented in Table 3.1 . One of the six repetitions for the 30-cc 
gelatin swallow of S5 had to be discarded due to "double swallowing," otherwise all 
averages were based on six replications. As shown in Table 3.1, intrasubject replication 
variability was small . In 84% of the cases, the coefficients of variation (CV=SD/M) were 
less than 0.20 (standard acceptable limit of CV in ratio= 2.00). In the remaining 16% of 
cases, the CV values ranged from 0.21 to 0.28. 
Analysis of variance (ANOVA) on the total duration data showed a significant 
subject effect (p, = .0007), a significant interaction (p, = .0001), and a significant but 
weaker task effect (p, = .0488). (The ANOVA table is presented in Appendix B, section 
B-1.) Post hoc tests did not isolate. the effect of swallowing task to any pairwise 
comparisons or complex contrasts related to bolus size or consistency. The exact nature 
69 
Table 3. I . Mean total durations (in milliseconds) of tongue-palate contact during discrete swallowing for Subjects 1 to 
5 (S I to SS). Coefficients of variation (CV) were computed using the formula [SD/M]. Except for 1130cc Gelatin" of 
S5, all values are based on six replications. Data for 1130cc Gelatin" of S5 had one missing replication. 
Single Swallow 
5cc Water 30cc Water 5cc Gelatin 30cc Gelatin Dry 
SI M 922.30 816.44 1066.44 1055.18 1161.04 
SD 135.79 138. 77 103.83 129.57 175.07 
CV 0.15 0.17 0.10 0.12 0.15 
S2 M 1647.52 1233.11 2225.23 1388.52 1662.17 
SD 292.01 178.86 233.62 144.84 260.39 
CV 0.18 0.15 0.10 0.10 0.16 
S3 M 1409.91 1208.17 1486.49 1377.25 1679.05 
SD 133.62 110.41 255.26 121.88 149.31 
CV 0.09 0.09 0.17 0.09 0.09 
S4 M 1467.79 1449.50 1803. 72 1848.65 1664.62 
SD 309.11 297.58 174.61 166.65 241.97 
~ 0.21 0.21 0.10 0.09 0.15 
S5 M 1365.09 1378.55 1414. 75 1944.32 1538.63 
SD 266.05 386.48 120.82 115.97 414.34 
~ 0.19 0.28 0.09 0.06 0.27 
-..J 
0 
71 
of the task effect was apparently masked by large intersubject differences. These 
differences as well as the strong subject-task interactions can be seen in Figure 3.1 (upper 
panel; crossed lines indicate interactions). As illustrated, total durations for S 1 were 
considerably shorter than those of all other subjects. The lack of a consistent pattern 
across subjects was apparent. Nevertheless, a closer examination of the swallowing tasks 
as a function of subjects (Figure 3.1, lower panel) did reveal some task-based tendencies 
with respect to bolus consistency. That is, dry swallows were longer than water swallows 
of either size, and gelatin swallows were longer than water swallows of comparable size. 
Stage duration: effects of task and stage, and subject variability. Average 
durations for the different swallowing tasks by stage are illustrated in Figure 3.2. Task-
based differences were apparent for all stages, though less so for Stage IV. Statistics were 
performed for only Stages II and III, since they did not have an arbitrary beginning or end 
point. 
Results of the multivariate analysis of variance (MANOVA) on stage durations 
showed a significant main effect of swallowing task, and a significant task-by-stage 
interaction. Post hoc comparisons of the task means for Stages II (propulsion) and III 
(full contact) revealed the following : For Stage II, with respect to bolus consistency, 
water had a significantly shorter duration than gelatin, and dry swallows were significantly 
shorter than 5-cc gelatin swallows. With respect to bolus size, 30-cc swallows were 
consistently shorter than 5-cc swallows. Thus, thinner consistency and larger volume 
were propelled faster through Stage II. For Stage III, the only significant duration effect 
72 
Figure 3 .1. Least-squares mean total durations for Subjects 1 to 5 (S 1 to S5) as a 
function of discrete swallowing task (upper panel) and for different discrete 
swallowing tasks as a function of subject (lower panel). SW= 5-cc water, 30W 
= 30-cc water, 5G = 5-cc gelatin, 30G = 30-cc gelatin. 
2500 
u i 2000 
-C 
0 
~ 
... d 1500 
Jg 
~ 
C 
m 1000 
~ 
-s1 
-o- S2 
--•-· S3 
--o- S4 
-:/·::::·:::-~ 
.. "=-
.-· ·~~ .. :--- -
t;:.:;, ... : ~-=-: -----:,-,~s.:,.-l'.'. - . 
...... ss 
500...L..----,-----.----...-----.-----.--- -
5W 
2500 
-u i 2000 
-C 
.2 
-l! d 1500 
Jg 
{!. 
C 
:i 1000 
30W 5G 
Swallow 
30G Dry 
- .. ·SW 
-o ·30W 
...... SG 
--<>-· 30G 
-Dry 
500J_---,------,,----,-- --,----S5::,:::-----
S1 S2 S3 S4 
Subject 
-CJ 
(1) 
V, 
E 
-C 
0 
:;:; 
cu 
~ 
::::, 
C 
Figure 3.2. Mean durations (with error bars) of individual discrete swallowing tasks for the four stages. SW= 
5-cc water; 30W = 30-cc water; SG = 30-cc gelatin; JOG= 30-cc gelatin. 
1200 CJSW 
~30W 
(=::::::::::::::::::j5G 
T ITI!Il30G 900 
T 1-Dry 
600 
300 
oll_J3ffiillilllli1L 
I. Prepropulsion II. Propulsion Ill. Full Contact IV. Withdrawal 
Stage 
74 
at the a.= .01 level was that the dry swallow was longer than either 5-cc or 30-cc water 
swallows. (Tables for the MANO VA tests of main effects and contrasts are presented in 
A ppendix B, sections B-2 to B-4.) 
The average duration (standard deviation) for each of the four stages, collapsed 
across swallowing tasks, was 251.33 (±208.62), 320.03 (±158.60), 585 .13 (±257 .67), and 
2 88 .77 (±150.90) msec, respectively. These means were not significantly different due to a 
large degree of variability in the measurement units, as reflected in their respective 
standard deviation values. 
Averaged timing data from the different swallowing tasks for the individual 
subjects in Stages I to IV are illustrated in Figures 3.3 to 3.6, respectively. Intrasubject 
timing variability was greater on the whole for Stage I ( overall coefficient of variation = 
.56) than for Stages II to IV (overall coefficients of variation =.20, .23, .31 , in order). 
Intersubject timing variability for the individual swallowing tasks was expected in all 
stages. Specifically, in Stage I (prepropulsion, Figure 3.3), both S2 and S5 were 
noticeably different from the other subjects for the 5-cc gelatin swallow in that the former 
had the longest and the latter had the shortest duration. In addition, S2 was the only 
subject whose 5-cc gelatin swallows averaged considerably longer than the larger-volume 
swallows of the same consistency, whereas for all other subjects the opposite was true. In 
Stage II (propulsion, Figure 3.4), the subjects without exception took longer time to 
propel the 5-cc gelatin bolus than they did the other swallowing tasks. In Stage III (full 
contact, Figure 3.5), the dry swallow was the longest in three subjects (Sl , S2, S3) and 
the second longest in the other subjects. Interestingly, S 1 had considerably shorter 
Figure 3.3. Average durations of Stage I (prepropulsion) as a function of discrete swallowing tasks for Subjects 1 to 5 
(SI to SS). SW= 5-cc water; 30W = 30-cc waster; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
1200 
900 
-(.) 
Cl) 
ti) 
E 
-C: 600 
0 
.:; 
ca 
... 
:::::, 
C 
300 
Prepropulsion Stage 
5G 
Swallow 
30G 
CJS1 
E2:ZJS2 
n,,,,,,,,,,ns3 
E=3S4 
-ss 
Dry 
Figure 3.4. Average durations of Stage II (propulsion) as a function of discrete swallowing tasks for Subjects l to 5 
(Sl to SS). SW= 5-cc water; 30W = 30-cc waster; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
1200 
-<.> 900 Q) 
Cl) 
E 
-C: 
0 
:.;::. 600 e 
:::::s 
C 
Q) 
C, 
,S 300 
en 
Propulsion Stage 
30W 5G 
Swallow 
30G Dry 
c=i S1 
i;;-: ,/1 S2 
GED S3 
E=:3S4 
- ss 
-...J 
0\ 
Figure 3.5. Average durations of Stage III (full contact) as a function of discrete swallowing tasks for Subjects 1 to 5 
(S 1 to SS). SW = 5-cc water; 30W = 30-cc waster; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
1200 
-u 900 Cl) 
en 
E 
-C: 
0 
~ 600 
::::, 
C 
Cl) 
a, 
!!! 300 u, 
Full Contact Stage 
30W SG 
Swallow 
30G 
T 
Dry 
c:::J S1 
fZ2:J S2 
EE]S3 
E=3S4 
-ss 
Figure 3.6. Average durations of Stage IV (withdrawal) as a function of discrete swallowing tasks for Subjects 1 to 5 
(S 1 to SS). SW= 5-cc water; 30W = 30-cc waster; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
1200 
-CJ 900 Cl) 
"' E 
-C: 
0 
.:; 600 C'CS 
i.. 
:::::s 
C 
Cl) 
C> 
C'CS 300 
-
"' 
Withdrawal Stage 
30W 5G 
Swallow 
30G Dry 
CJS1 
BJS2 
EiliTI S3 
~S4 
- ss 
-..J 
00 
79 
Stage-III durations for all swallows than the other subjects, making this stage the primary 
source of the intersubject variability for the total swallow duration described in th.e 
prev ious section. In Stage IV (withdrawal, Figure 3.6), S3 had the longest mean 
durations for all swallowing tasks except 30-cc gelatin, and S2 had the shortest mean 
durations for all tasks except the dry swallow. 
Tongue-Palate Contact Patterns 
Contact patterns for individual stages. Average percent-contact profiles (time-
series waveforms) from the different swallowing tasks for Subjects 1 to 5 are presented in 
F igures 3. 7 to 3. 1 18, respectively. Each figure consists of two parts with Part (a) 
illustrating data for the front, central, and back bins and Part (b) illustrating data for the 
lateral, medial, and midline bins. Contact characteristics specific to each stage are as 
follows: 
1. Prepropulsion stage. The prominent percent-contact pattern, Pattern 1, for this 
stage was characterized by small, scattered changes in contact over time, mostly in the 
lateral/back bins, resulting in lateral contacts of <30% by the end of the stage. This 
pattern occurred in 80% of the data. A variant of this pattern, Pattern 2, was observed in 
20% of the data. It was characterized by activation of a greater number of lateral 
electrodes (>65%) as well as some outer-medial electrodes, resulting in an apparent lateral 
seal (see Figures 3.7b--5W and Dry, 3.9b--5W, and 3.10b--30W and Dry). 
8In order to achieve optimal illustration of contact patterns, the X-axes in these 
figu_res are sc~led uniformly within subjects (i.e., to the longest averaged swallow for each 
subJect) but differently across subjects. 
80 
Figure 3.7a. C~an~es in average percent-~ontac~ over tim~ in front_ (thick solid li~e), 
ntral (thin sohd hne), and back (dashed hne) bms from different discrete swallowing 
ce ks for S 1 (female, age 47). Vertical lines demarcate the four stages. SW = 5-cc 
taster 30W = 30-cc water; SG = 5-cc gelatin; 30G = 30-cc gelatin. wa , 
100 
t, 75 
"' C: 50 0 
0 
.,. 25 
0 
msec 0 
100 
t, 75 
~ 50 0 
0 
.,. 25 
0 
msec 0 
100 
t, 75 
"' C: 50 0 
0 
.,. 25 
0 
msec 0 
100 
t, 75 
nl 
C: 50 -0 
0 
.,. 25 
0 
msec 0 
100 
ti 75 
nl 
- 50 C 0 
0 
.,. 25 
0 
msec 0 
., 
/ .,,-' 
-
,,, 
250 
250 
-----
500 
500 
' \ 
' 
750 
750 
~
~---
\ 
., : 
.,-
., _. 
; 
., 
250 
,I __ .,, 
250 
500 
I ,, 
,I 
500 
.,. 
,, 
750 
750 
...., ________ _ 
\ 
' 
', 
' 
,_ 
1000 
1000 
' 
\ 
' 
' \ 
\ 
1000 
1000 
' 
' \ 
5W 
1250 
30W 
1250 
5G 
1250 
JOG 
1250 
Dry 
81 
Figure 3.7b . Changes in average percent-contact over time in lateral (black solid line), 
medial (dashed line), and midline (gray solid line) bins from different discrete 
swallowing tasks for Sl (female, age 47). Vertical lines demarcate the four stages. SW 
= 5-cc water; 30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
100 5W 
-
75 I.) 
nl 
c 50 0 
(.) 
';te 25 
msec 0 250 500 750 1000 1250 
100 30W 
ts 75 
~ 50 0 (.) 
';te 25 
0 
msec 0 250 500 750 1000 1250 
100 5G 
-
75 I.) 
.l!! 
C 50 0 
(.) 
'#- 25 
0 
msec 0 250 500 750 1000 1250 
100 30G 
ti 75 
.l!! 
C 50 0 
(.) 
,;;e. 25 
msec 250 500 750 1000 1250 
100 "1_ ...... . ....... - .... . . ........ .. ... .. ... .. Dry 
u 75 
nl 
i: 50 0 
(.) 
,;;e. 25 
0 
msec 0 250 750 1000 1250 
82 
Figure 3 .8a. Changes in average percent-contact over time in front (thick solid line), 
central (thin solid line), and back (dashed line) bins from different discrete swallowing 
tasks for S2 (female, age 46) . Vertical lines demarcate the four stages. SW = 5-cc 
water; 30W = 30-cc water; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
-u 
"' C: 
0 
(.) 
'#-
100 
75 
50 
25 
SW 
0 +--""=----__;___--+-- --;..---+-- ----+-- ----;J---t ..... - ---~- - - --j 
2400 
msec o 400 800 1200 1600 
2000 
30W 
100 
'ti 75 
~ 0 50 
(.) 
'#- 25 
.,-
o -1-- - -i---+-- ---- +-------;J....,i--- ---+-- - ---t----------1 
msec 0 800 1200 1600 2000 2400 
100 
~ 75 
g 50 
(.) 
'#- 25 
msec 
100 
u 75 
.!!! a 50 
(.) 
'#- 25 
SG 
400 aoo: 1200 1600 2000 2400 
30G 
0 ~~-~-~_l--+-----r-l~'~-t--- --t-------1 
msec O 400 800 1200 \ 1600 2000 2400 
100 
Dry 
400 800 1200 1600 2000 2400 
83 
Figure 3.8b. Changes in average percent-contact over time in lateral (black solid line), 
medial (dashed line), and midline (gray solid line) bins from different discrete 
swallowing tasks for S2 (female, age 46). Vertical lines demarcate the four stages. SW 
= 5-cc water; 30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
100 
- 75 ~ 0 50 
0 
"ii! 25 
SW 
0 '!=-''-'..&.:..--+--+------i---+------+-----;---...... -----+------j 
msec o 400 800 1200 1600 2000 2400 
100 30W 
u 75 
~ 0 50 
0 
'if!. 25 ...... ~--...., 
Q~'!!t:ilo,~~-4-----+---~-'-----1----t-------l 
msec 
100 
u 75 
!S 5 50 
0 
"ii! 25 
100 
u 75 
<G 
°g 50 
0 
'.f!. 25 
msec 
0 <ioo 
400 
800 ~200 1600 2000 2400 
5G 
30G 
Dry 
800 1200 1600 2000 2400 
84 
Figure 3.9a. Changes in average percent-contact over time in front (thick solid line), 
central (thin solid line), and back (dashed line) bins from different discrete swallowing 
tasks for S3 (male, age 23). Vertical lines demarcate the four stages. SW= 5-cc water; 
30W = 30-cc water; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
5W 
100 
u 75 
res 
c 50 0 (.) 
~ 25 
0 
msec 0 250 500 750 : 1000 1250 1500 1750 
30W 
100 
--
______ .... 
u 75 ' .. 
i 
... 
50 ' 0 ... ... 
-(.) 
' 
"#- 25 ... .. 
... 
0 
msec 0 250 500 750 1000 1250 1500 1750 
5G 
100 
u 75 
.,.---' res 
c 50 0 (.) 
"#- 25 
0 
msec 0 250 500 750 1000 1250 1500 1750 
30G 
100 
u 75 
res 
c 50 0 (.) 
~ 25 
0 
msec 0 250 500 750 1000 1250 1500 1750 
100 
Dry 
u 75 .1i'-, -~1 -
.., 
' : '~ ., i: 
. - .,. 
50 ' : _ ... 0 I ; / 0 ., ... 
"#, 25 
./' -~ 0 
msec 0 250 500 750 1000 1250 1500 1750 
85 
Figure 3.9b. Changes in average percent-contact over time in lateral (black solid line), 
medial (dashed line), and midline (gray solid line) bins from different discrete 
swallowing tasks for S3 (male, age 23). Vertical lines demarcate the four stages. SW= 
5-cc water; 30W = 30-cc water; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
100 
·-· ···--······ · ··-···---1--···· 
5W 
ti 75 
"' § 50 
0 
'#- 25 
o -P-=-~--+----+-+-----t----i--+-- ---+'.'.!!!lli....,. _ _,_ ___ --l 
msec O 250 500 750 j 1000 1250 
100 
ti 75 
"' i: 50 0 
0 
-;;e 25 
msec 
100 
ti 75 
"' i: 50 0 
0 
-;;e 25 
250 500 1000 1250 
-·······•· ························r-, 
. ..... 
1500 1750 
30W 
1500 1750 
5G 
0 'F-''--i----+-----+----t------+-----;--f---__::~--------i 
msec o 
100 
~ 75 
§ 50 
0 
250 
'#- 25 ~ .......... _..., 
500 
, 
)so 1000 1250 1500 1750 
] i 30G 
,--~-···- ..... . . - ................... ·- t · 
0 p~~~~:..:..:_--f---+---+-----t--+-~~..::-:=-+-----l 
msec o 500 750 1000 1250 1500 1750 
100 Dry 
ti 75 
"' § 50 
0 
~ 25 
o ........... 11!'!::.----+-t-----+-+---+----+----+.---~~ .... --~ 
msec O 250 500 750 1000 12$o 1500 1750 
86 
Figure 3. 1 Oa. Changes in average percent-contact over time in front ( thick solid line), 
central (thin solid line), and back (dashed line) bins from different discrete swallowing 
tasks for S4 (male, age 37). Vertical lines demarcate the four stages. SW= 5-cc water; 
30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
100 -
u 75 
!9 
C: 50 0 
(.) 
';I! 25 
.... _ 
0 
msec 0 
100 
u 75 
!9 
C: 50 0 
(.) ;-----' ~ 25 ., 
0 
msec 0 
msec 0 
100 
-
75 .., 
i 50 0 (.) 
~ 25 
0 
msec 0 
100 -
u 75 I <a 
-
I C 50 0 J (.) 
.I 
';it. 25 
" 
., 
0 
msec 0 
.. 
., '. - --- ' '\ 
' 
400 800 1200 
\ 
\ 
\ 
\ 
'-
"" .... 
)------------------~, \ 
\ 
\ 
, : 
., 
' \ 
' ...... 
400 800 12od 
400 800 
400 800 1200 
400 800 1200 
5W 
1600 2000 
30W 
1600 2000 
5G 
2000 
1600 2000 
Dry 
1600 2000 
I 
: f 
I 
87 
Figure 3.10b. Changes in average percent-contact over time in lateral (black solid line), 
medial (dashed line), and midline (gray solid line) bins from different discrete 
swallowing tasks for S4 (male, age 37). Vertical lines demarcate the four stages. 5W = 
5-cc water; 30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
100 
ti 75 
~ 0 50 
0 
'#- 25 
SW 
.... j• .................... . 
0 -.....;1--~---+---i-----+------++---·""'••--+---------i 
msec o 400 800 12oq 
100 
ti 75 
.!!! 5 50 
0 
"#- 25 
. : ................................ 
1600 2000 
30W 
0 ~ ......... _,_ ..... ~:__--i-------1-----++~ ..... ---t--------, 
msec o 40b 800 12otj 
100 
... 75 ~ 0 50 
0 
"#- 25 
. . . 1 ........... . . .. .. .... . 
1600 2000 
5G 
0 .............. ~~---+-----'---+------+-----4....:!9-,,_.~---i 
msec o 400 800 1200 ~ 600 
100 
ti 75 
s 5 50 
0 
'#- 25 
2000 
30G 
oC~~;;,;.;.;.~~~~=--ti--- ---1-------+~...:.:.:~ ~ 
~ 0 400 1200 1* 2000 
100 
Dry 
...................... ":, 
ti 75 
"' § 50 
0 
'*- 25 
0 .,.......,,,,=--'-------4-------+---------++-~- ... - ~ -------i 
msec o 800 1:zqo 1600 2000 
Figure 3.1 la. Changes in average percent-contact over time in front (thick solid line), 
central (thin solid line), and back (dashed line) bins from different discrete swallowing 
tasks for S5 (female, age 23). Vertical lines demarcate the four stages. 5W = 5-cc 
water; 30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc gelatin. 
100 
u 75 
J! § 50 
0 
~ 25 
5W 
88 
0 -1-""""=::::..__:_ _ -t _ _ __j__ __ +-__ __;,.___~-_,;;;._,_' _ __ -+---------i 
1200 1600 2000 msec o 
100 
u 75 
J! § 50 
0 
'$. 25 
400 800 
30W 
0 +-- - - - - --+:--=..:::..=__ __ ___._...,;__ ___ -'-+--',.~~-----------< 
msec O 400 800 1200 2000 
100 
u 75 
J! § 50 
0 
#- 25 
., ., 
_., 
., ., 
·-----.,., 
.I . 
, 
1600 
5G 
O~'--------+--_i_--f------+-~~~--t-------l 
1600 2000 msec o 
100 
u 75 
J! § 50 
0 
#- 25 
msec 
400 
400 
800 
30G 
----
800 1200 1600 2000 
89 
Figure 3.1 lb. Changes in average percent-contact over time in lateral (black solid line) 
medial (dashed line), and midline (gray solid line) bins from different discret; 
swallowing tasks for S5 (female, age 23). Vertical lines demarcate the four stages. SW 
= 5-cc water; 30W = 30-cc water; SG = 5-cc gelatin; 30G = 30-cc gelatin. 
100 SW 
ti 75 
.! 
C 50 0 
0 
-a'- 25 
0 
msec 0 400 800 1200 1600 2000 
100 30W 
-
75 u 
"' c 50 0 
0 
-a'- 25 
msec 400 1200 1600 2000 
100 SG 
ti 75 
"' c 50 0 
0 
-a'- 25 
0 
msec 0 400 800 1600 2000 
100 30G 
-
75 u 
"' c 50 0 
0 
-.,e. 25 
0 
msec 0 400 800 1200 1600 2000 
Dry 
msec o 400 800 1200 1600 2000 
90 
2 . Propulsion stage. The predominant propulsion pattern, which followed the 
prepropulsion Pattern 1, began with an abrupt onset of front contact, followed by a steady 
increase in contact in both a front-to-back and a lateral-to-midline progression, ending 
with maximal contact for all bins ( see Figure 3 .12, top). This pattern, occurring in 80% of 
the data, created a lateral seal simultaneous with the backward propulsion of the bolus. 
Pattern 2, which followed Pattern 2 ofprepropulsion, was seen in the remaining 20% of 
the data. This pattern was characterized by increasing contacts in a primarily 
fro nt-to-back progression, occurring in the medial and midline bins (see Figure 3.12, 
bottom.) 
3. Full contact stage. This stage was characterized by stable, maximal contact. 
Such characteristic can be seen in all panels of Figures 3. 7 through 3 .11 . 
4 . Withdrawal stage. The characteristic pattern for this stage was a progressive 
front-to-back and simultaneous midline-to-lateral withdrawal of contact (Figures 3.7-
3 .11 ). The pattern of percent-contact retraction was remarkably consistent within subjects 
and across swallowing tasks . 
Task effect . No striking effects of swallowing tasks on overall contact patterns 
were observed. Occasionally, however, there appeared to be slight patterning differences 
among the tasks. In particular, for Stages II and Ill, the 30-cc gelatin had an average of 
18% less contact (range across subjects = 2-40%) than the other swallowing tasks. In 
addition, for Stage I, dry swallows had a slightly greater contact (2-17%, M = 9%), 
usually in the back bin, than the other bolus conditions. 
Figure 3 .12. Raw EPG data from two representative discrete swallows, illustrating Stage-II (propulsion) contact patterns. Pattern I 
(upper two rows) shows simultaneous front-to-back and lateral-to-midline increase in contact. Pattern 2 (lower two rows) shows 
prominent lateral seal at Stage-II onset and primarily front-to-back progression in contact increase. Each series begins with the last 
frame of Stage I (prepropulsion ). The vertical line demarcates the onset of Stage II. 
Pattern l: 
.. 
2.1s ••• , ...... 
. ·····~. ·-~··!:~ .. :., 
.. . ·: . ·: .. . 
:·· . : ... .. 
I I • • I 
.. . . .. 
.. . . .. 
•• • • 
2.ss.:.11 •• 
• ·=~·:::·~=·. • • o I O O O 11 
:: : ·.· : ·: ·. ·.·• 
.. 
.. 
2.7~ ......... . 
1 •,::1,1::,•, 
.. ·· .; :· .. -~. 
.. : : ·. · .... 
... . . 
.. . . .. 
.. . 
.. 
... 
.. 
Pattern 2: 
2.33.;.:: .. 
·=-·::: ··=· ... ·: .:. ·:,•· 
... : . · ... . 
..
.. . . . .. .. 
• •• .. . ... 
.. . . ... 
•• • ii •• 
2.45 •• ,..,, .. 
·~·= =··=· ... ........ . :.•,': :·.·.-.-. 
.. : . . · ... 
.. . . .. 
~=· • ·:: 
2.34 ...... . 
. .. :.: .. , .. 
•,:·.<: :.•:,• 
11 a . • . I I ,I 
.. 
11 I : •• 11~ 
• •• 
•• • • •• 
~= • i • :: 
2.46 •• ,.,., .. 
·=~·= : .. :. ....., . . . 
11 I •,1 • -: I•• 
.. . . ... 
11 I • I 11 
.. . . .. 
.. . . .. 
II II 
2.59 •• ,. ..... 
·····=····· ··~·-:: :- •:,• 
. ·:: . : : ·: ·. •.·. 
.. . . .. 
.. 
.. 
2.77,,\11 .. 
·=-·==· .. =· 
·····~·-··· 
I, I I, • , I I 11 
.··· : . · .. ··•  . . .
•• • • •• 
.. . . .. 
•· .. 
235 •• •.11 •• 
-=~::: : :::• 
. ·:: .... : · ... :·.·. 
.:·. . : .....  
• • 
•1 a I a• 
... 
II 
••• 
.. 
2.47 •• ,. .,, •• 
..... ..... 
··= ··=·=··=·· ........... 
.. . : . ·. ···. .. . .. 
... . .. 
... . .. 
•• •• 
2.61 .... 11 •• 
. •:::li•i::•. .. . . . .. :.·: ·: : ·: •. ~ .. 
.. .. 
.. 
.. 
2.1a •• , ..... 
.-:::=•====·. 
.. ·· .. •:• ...... 
.. : : ·. ·. ·· . 
.. . . . 
.. . . .. 
. .. 
.. 
... 
.. 
236 ......... . 
·=~ : : : ~~==· 
.·:· . : : · .. •·.· . a I . . I a 
..... · ...... 
•1 I I 11 
:: I • I:: 
2.48 .... 1'·· 
••••• ••••• .......... 
,:: ···= .. • :· .. 
.... • .... ··· . 
.. . . . . 
... . .. 
... 
.. 
I•• 
-
2.63.,\11 .. 
• •:::1i1.;:~-. 
I .·~ • .I: 1, I :· .. 
:: . : · .. •·. 
• • I I, 
.. 
2.79 .......... . 
• •:::=•==··. I.-: •./: •..: :•,. 
11 I • I•• 
.. . . .. 
I I , I 
I I a a• 
• •• 
II 
. .. 
.. 
2.3~ .......... . 
. -.~~:'!=··· 
. :•.:: · .. ·· .. 
:• •· .· ·. ·. ··. I a I I 
•a a I I' 
:: · .. :: 
2.4~·=::)!::•. 
··= ··=·=··=·· :.• I 'I: 1,1 I • 'a 
.. . . . . · ..
. .. · . . · ... 
'I I 
II 
... 
-
2.65 ..... ,., •• 
•••• • ••• • ••• 1 ••••• .. . .... :.· . ·: : ·: . ··. 
.. .· . . . · ..
.. 
.. 
2.~ ......... . 
• :::::1:::::, 
I, I 11 • 11 I 11 
11 I I • I I 1• 
.. . . . ... 
.. . . .. 
... . .. 
II -
2.38 ............ . 
.·~~ .. !~··· 
.:• . : : · .. ·· .. I I • • I I 
..·: . ·. ·.· .. 
.. . . .. 
:: · . ·:: 
2.so •• ,,, •• 
.·~:= ::::•. 
,:: .,. ..... :· .. 
. ·: ., . : . . ·.·. 
.. :~ . :· ... 
... . .. 
aa '!I II 
267 • .,,., .••• 
.1, •••••• 
I 'I~ •111 I .. . . . .. 
:.·: ·: : ·: •. ·.·. 
.. . .. 
.. . · ... 
.. 
2a2.:.11 •• 
·=··=·=·:=· 
.. 
.. 
......... ... 
•• I ... I,.. I ol 
.·: ..... ·:. 
.. : . . · ... 
... . .. 
.. -
2.3?.::;I:::·. 
··=·.-.:=··=·· 
.··. : . ·.· .. ·• 
•1·: . ·. •• '1• 
"I I I I ' 
... 
II 
2.s1 • .-.1, •• ••••• ••••• 
··=. ·=·=··=·· 11 .•, a,: I• ' 
la I I I 11 
. .. : . · .... 
'I I I I I I' 
' la I I' 
•· I .. 
2.69 ..... ,, .. 
·=··= ... ~ • ...... : .. . 
. ' ....... . 
.. : ..... . 
.. . . . .. . 
.. . . .. 
.. . .. 
.. 
2.~ ........... . 
.:,:::,:::•: . 
··=·,:,.•: •. 
,9:1 I : a··:. 
.. : : . · ... 
•a I I .. 
•• •• 
241 .... , ... . 
····: .... . 
·-:·: .:: ··=·· 
. .. : . ·.• •.. 
•1 I • I a• 
11 a • • ·. I 11 
.. . . .. 
... 
.. . ·~ 
2s~.:.1, ••• 
.·.~:= ===··· 
.. ·: ···=··· :· .. • •••••••• 
•e I I I ~ a•' 
.. . . . ... 
... 
.. ' 
. .. 
I ·• 
2.11 •• ,, ..... 
• ... I •.• I 
,•,:•:: : ·:.·i·~ 
.. ·:-:· ·.· .. 
.. . . . .. 
.. . · ... 
.. .  
•· .. 
2.84 ........ . 
·=··=·= .. :• ........ ... 
....... ,.. ... 
,•: I I : I I•:. 
·,::. :· ... 
::•, ,•:: 
2~ •• ,., ••••• 
.·.~:= :::,•. 
-:: ._:: · ..  :· .. 
11 I • I 1• 
... : . · .... 
"I I I I" 
"I I I I" II I • II 
2.53.:·1'·· 
••••• ••••• I 1,11 •111 I 
.:: ·:·=··: :· .. 
.•: .. : .. •:. 
·a:: I:· ... 
• 'I I 
.. ' 
. .. 
.. 
273 ..... ,, •• 
• ·=:~ui::• • 
.. '.. .. 
:.· .· ·: : ·: •. ·.·. 
.. . . . .. 
.. : · ... 
.. . .. 
.. .  
2.87 ........... . 
.•.::=1===··· I~:•; I•,.•:-.. 
,',• I I : I I 1111 
.. : : : ; · ... 
::•1 1•:: 
2.4:;.:::1··::•. 
. .... : .... 
..·: ·-·:· .. :· .. 
.. ... : . ·. ··· . 8 • I I 
... . .. 
•1 I 
.. . •:;. 
254,:.1, •• ••••• ••••• 
•.: · ·=·=··=·· 
.····-·:•:,··. 
"II I • I I • lal 
. .. · : : : · ... 
~~ I ~ • I I=~ 
92 
Examination of contact patterns 1 and 2 ( described above for the prepropulsion 
and the propulsion stages) as a function of swallowing task revealed that Pattern 1 
occurred across all swallowing tasks, and Pattern 2 occurred in all except the 30-cc gelatin 
boluses. The dry swallow had a relatively greater frequency of Pattern 2 (15/32 cases) in 
contrast to the other tasks (6/32 for 5-cc water, 7/32 for 30-cc water, 4/32 for 5-cc 
gelatin), particularly for S 1 and S4 (see Figure 3.13). 
Analysis on the order in which maximal contact was achieved in the front (F), 
lateral (L), and back (B) bins during Stage II showed a variety of sequences. However, 
the predominant pattern, observed across all tasks in 102 of 149 individual swallows, 
followed a front-lateral-back progression of"peaking". In 17 of 149 swallows, the front 
and the lateral bins reached maximal contact at the same time, followed by the back bin. 
This was considered a variant of and therefore grouped with the front-lateral-back pattern. 
Since the remaining patterns (e.g., L-F-B, F-B-L, B-F-L, L-B-F, B-L-F) were few in 
frequency (most with only 2 or 3 cases), they were grouped as "others." Subsequent Chi-
Square test of the two groups showed a significant subject effect (discussed below) but no 
significant effect of swallowing tasks. 
Subject variability. The post-warp, frame-by-frame descriptive statistics (i.e., M, 
SD, and CV) were examined by bin, stage, and task for intrasubject variability in overall 
percent-contact patterns. Data from S 1 and S4 were on the whole somewhat less variable 
than those from the other subjects, although all subjects showed greater replication 
variability in Stage I than they did in the other stages. The CV data revealed that 
I : . 
Figure 3 .13 . Frequency distribution of contact patterns I and 2 for different discrete 
swallowing tasks and subjects based on raw percent-contact data in lateral bin at 
Stage-II onset. Pattern I = <65%, Pattern 2 = >65%; SW= 5-cc water, 30W = 30-
cc water, 5G = 5-cc gelatin, 30G = 30-cc gelatin. Inserted percentages are the 
means for the respective patterns, swallowing tasks, and subjects. 
Cl Pattern 1 D Pattern 2 
51 
52 
53 
54 
55 
0 2 3 4 5 6 
93 
94 
replication variability was high (CV > 2) in 19% of all cases, moderate ( 1 < CV < 2) in 
another 19% of cases, and low (mean CV = .20) in the remaining 62% of cases. In all 
subjects, trial-to-trial variations for Stages II and III were mostly small. 
Intersubject variations in overall percent-contact patterns appeared trivial. 
Occasionally, some subjects showed less front and/or lateral contact in selected stages, or 
used extra preparatory tongue movement during Stage I. Substantial subject differences 
were seen in the use of Patterns 1 versus 2. S 1 and S4 used Pattern 2 more often than the 
other subjects (see Figure 3.13). Subject variations were also apparent in the ordering of 
percent-contact maxima during Stage II. S3 and S4 (both male) were the primary sources 
of the "other" sequence patterns (13/30 swallows in S3, 10/30 in S4), while the remaining 
subjects showed predominantly the front-lateral-back sequence. Of all the single-swallow 
analyses and observations, this was the only occasion where some differences attributable 
to subject gender appeared to exist. However, due to the small sample size, gender effect 
was not specifically evaluated, and no conclusions could or should be drawn. 
Left-right symmetry. Remarkable left-right symmetry (see Figure 3.14, upper 
panel) was observed in the majority of the data. On the occasions where asymmetry was 
seen, the left and the right sides differed minimally in the amount of percent-contact and 
not at all in overall contact pattern (see Figure 3. 14, middle panel). Most of the 
asymmetry tended to occur near the beginning of Stage I (prepropulsion) and/or near the 
end of Stage IV (withdrawal) (see Figure 3.14, lower panel). 
' I 
., 
'I 
95 
Figure 3.14. Changes in average percent-contact over time in left (gray solid line) and 
right ( dashed line) bins from representative discrete swallows of selected subjects. Upper 
panel shows the remarkable symmetry seen in most cases. Middle panel demonstrates 
left-right differences in amount but not pattern of percent-contact. Lower panel shows 
typical locations of assymetry: near Stage-I start and near Stage-IV end. Vertical lines 
demarcate the four stages. 
100 
75 
-u 
"' i: 50 0 
0 
'#, 
25 
0 ' 
msec 0 
100 
75 
-u 
"' i: 50 0 
0 
'#, 
25 
0 
msec 0 
100 
75 
25 
, .. 
., 
I 
I 
.. 
., 
' ' 
,-~·-·-·---·~--·~·~ . ' ' . 
" 
\ 
\ 
~ 
I 
'\ 
• 
400 800 1200 1600 
: : 
i II ~~ ... .. .,. . ..,. ...... __ ., ____ ' 
.. i .fl 
r , r. 
• :, \. 
' 
• ,,i 
•11. • I ,. , ,_....... :
.. ' 
\ 
4. 
400 ado 1200 16QO 
: --,. ... "F9"~--· :\ 
• i : I 
I ' I 
' 
' 
I 
' 
2000 
' 
.. 
-1 
2000 
0 ..---------+--+------+---,-----+----+----!--- -----1 
msec 0 250 500 750 1000 1250 
96 
Ultrasound Findings 
The sagittal ultrasound data on discrete swallowing for all subjects were examined. 
However, due to considerable variations in image quality, only the data from SS were used 
in illustrations and for displacement measurements. 
Swallow-Related Lingual Behaviors on Ultrasound 
The ultrasound visualization of the swallowing task for this study contained four 
stages, i.e. , prepropulsion, propulsion, full contact, and withdrawal, just as distinctly 
identifiable as those seen on EPG. However, the ultrasound and the EPG stages differed 
in scope and behavioral details because of the innate capabilities as well as drawbacks of 
each instrumentation. 
Movement-tracking for the four functional tongue segments (anterior, mid, dorsal, 
and posterior) on ultrasound revealed that a swallow or task response typically began with 
anterior and superior movement of all tongue segments from the resting position, loading 
the bolus onto the tongue surface (prepropulsion). Next, the tip of the anterior segment 
was anchored against the tip of the hard palate, followed by serial elevation of the mid, 
dorsal and posterior segments to maximal displacement. With these movements, the bolus 
was progressively propelled backward into the pharynx (propulsion). The entire tongue 
remained maximally elevated against the palate for a period of time (full contact). Finally, 
the tongue began to pull away from the palate and return to the resting position 
(withdrawal). The withdrawing action, as part of the post-swallow mouth-opening 
gesture, often began with the anterior segment and progressed through the mid or dorsal 
segment in a "peel-away" fashion, followed by further lowering of the entire tongue as a 
whole. However, it was also not uncommon to see a slight descent of the posterior 
tongue just before the anterior segment lowered. The basic pattern oflingual actions 
described above was observed across subjects and swallowing tasks. 
Ultrasound versus EPG: Descriptive Comparison 
97 
Because the sagittal ultrasound images depicted movements of only a 2-mm thick, 
midline section of the tongue, precise correlations ofEPG contact details and ultrasound 
lingual behaviors were not possible. For example, any EPG activities in the lateral and the 
medial bins were not visualized on ultrasound. However, with the inclusion of 
synchronized trigger signals ( click and glottalized /a/) on both EPG and ultrasound data, 
gross comparisons could be made of EPG contact activities and ultrasound tongue 
movements. 
Movements of the anterior, mid, and dorsal (partial) tongue segments during 
propulsion and withdrawal parallelled the contact activities observed on EPG during 
Stages II and IV. However, since EPG registers only tongue-hard palate contact, lingual 
actions before and after contact as well as during steady-state maximal contact, though 
very much a part of the swallowing maneuver, are not shown. Specifically, four sets of 
ultrasound lingual behaviors were not reflected on EPG in the present study. These 
included: (a) some of the initial prepropulsive movements (occurring before any tongue-
palate contact), (b) the propulsive motion of the posterior segment as well as part of the 
dorsal segment, (c) the occasional early withdrawal of the posterior segment, and (d) the 
98 
final descent of the entire tongue to the resting position (occurring after EPG withdrawal) . 
As a result, the EPG prepropulsion and withdrawal stages were usually shorter than the 
equivalent ultrasound periods, and the ultrasound full contact period was usually shorter 
than the EPG full contact stage. In other words, while full contact was maintained on 
EPG, propulsive lingual movements were continued in the posterior tongue segment on 
ultrasound. 
The EPG data provided considerably more detailed information than ultrasound on 
the interaction of the tongue and the palate, adding a new and unique dimension to the 
understanding of lingual motor function during swallowing. Details such as the front-to-
back and simultaneous lateral-to-midline contact increase during propulsion, and the front-
to-back and simultaneous midline-to-lateral withdrawal were not visible on ultrasound. 
Task Effect on Ultrasound 
The finding of some significant bolus effects on EPG provided an incentive to 
examine tongue shapes as a function of swallowing tasks on ultrasound . Figures 3 .15 
(Parts a and b) illustrate, in series, temporally comparable, "significant" frames from five 
representative swallows--one for each bolus type--ofthe same subject. As discussed 
below, several bolus-induced differences in tongue shape can be seen in the images, 
though by the time the dorsal segment reached maximal displacement, tongue shapes 
became essentially the same. 
Pre-swallow (Figure 3.15a, Row 1). While in a rest state, the tongue was 
compressed to varying extent, depending on the bolus volume. For the small water and 
3 
Figure 3. I Sa. Changes in tongue shape over time as a function of discrete swallowing task. All frames were selected based on specific 
criteria from representative swallows of the same subject. Frames in Rows 1, 2, and 3, in order, illustrate tongue shapes at rest (tip to 
back from left to right), when entire bolus was first loaded on tongue surface, and at maximal displacement of anterior tongue segment. 
Arrows in each image mark points on tongue surface. SW= 5-cc water, 5G = 5-cc gelatin, 30W = 30-cc water, 30G = 30-cc gelatin. 
5W 5G 30W 30G Dry 
Figure 3 .15b. Changes in tongue shape over time as a function of discrete swallowing task. All frames were selected based on specific 
criteria from representative swallows of the same subject. Frames in Rows 4, 5, and 6 illustrate tongue shapes (tip to back from left to 
right) at maximal displacement of mid, dorsal, and posterior tongue segments, respectively. Arrows in each image mark points on 
tongue surface. 5W = 5-cc water, 5G = 5-cc gelatin, 30W = 30-cc water, 30G = 30-cc gelatin. 
4 
5 
6 
5W 5G 30W 30G Dry 
-0 
0 
101 
gelatin boluses, the tongue was arched and slightly retracted. Large-volume boluses were 
associated with a lower and more compressed tongue. In dry swallows where bolus size 
was minimal, the tongue showed no compression. Differences due to bolus consistency 
were more apparent in the case of large boluses. Gelatin tended to remain as a cohesive 
mass on the tongue with more collected anteriorly than dorsally; water, naturally less 
cohesive, tended to be contained in front of as well as on the anterior tongue. As a result 
and for apparent safety reasons, the anterior tongue was somewhat more retracted and the 
dorsal and posterior segments slightly more hunched for water than for gelatin (see Figure 
3.15a, Row 1, 30W vs. 30G). 
First appearance of entire bolus on tongue surface (Figure 3. l 5a, Row 2). At the 
moment when the entire bolus was loaded onto the lingual surface, the tongue assumed a 
more level, horizontal posture for dry and 5-cc boluses than for large boluses. The large 
volumes continued to induce a depression in the anterior and mid tongue segments. 
Maximal displacement of anterior segment (Figure 3. l 5a, Row 3 ). With the 
anterior segment anchored to the front of the hard palate and with propulsion 
commencing, the tongue was seen to assume a slight backward slope regardless of bolus 
type. Again, the large boluses tended to depress the tongue slightly more. No striking 
consistency-based differences were observed at this point. 
Maximal displacement of mid segment (Figure 3. l 5b, Row 4 ). Backward tongue 
slope was greatest at this moment. The dorsal and the posterior segments were slightly 
higher for 5-cc water and dry swallows than for the other bolus types. 
102 
Maximal displacement of dorsal and posterior segments (Figure 3 .15b, Rows 5 and 
6). The bolus had essentially left the oral cavity at this point. Therefore, little variation in 
tongue shape was observed as a function of swallowing task. 
In short, volume-based differences in tongue shape were observed during the pre-
swallow state, prepropulsion, and most of propulsion, while consistency-based differences 
were more apparent during pre-swallow. It should be emphasized that regardless of the 
task at hand, progressive anchoring of functional segments from front to back appeared to 
be an important and consistent strategy utilized by the tongue to achieve successful 
transport of the bolus through the oral cavity. 
Lingual Velocity Profiles: Single-Subject Data 
Recall that the EPG data showed significant differences in contact timing for Stage 
II (propulsion) as a function of bolus consistency (water faster than gelatin) and size 
(30 cc faster than 5 cc). Comparable differences in lingual movement velocity were 
expected. Frame-by-frame, radial displacement data were extracted for the four tongue 
segments during prepropulsion and propulsion. Velocities were then calculated, and a 
velocity profile--a waveform of changes in instantaneous velocity over time--was plotted 
for each swallow. It should be pointed out that the accuracy of frame-by-frame 
displacement tracking depended greatly on the sharpness of the tongue surfaces imaged. 
In the present study, due to lingual tissue variations in the individual subjects and resultant 
variations in image quality, adequate measurement accuracy could only be maintained with 
the ima e 
g s from Subiect 5 (SS). T 
J herefore, only the velocity data of this subject are 
103 
reported. 
Figure 3 16 ·11 
· 
1 ustrates the velocity profiles of the four tongue segments from 
representa . . 
hve discrete swallows of SS . Intervals equivalent to EPG Stages I - III are 
lllarked. The veloc't fi . . 1 Y pro les captured the time progress10n of bolus transport from front 
to back d 
' emonstrating the stereotypical, spatiotemporal concatenation of the four 
functio 1 na segments. Also visible in the profiles were the upward and downward velocity 
Shifts Whj h 
c corresponded to the ascent and descent of each tongue segment during bolus 
transport 
The succession in which each segment achieved its positive peak velocity, 
unvaryin . . 
g witbin and between tasks, resembled that of a "whip" effect, as if the segments 
Were link . 
s 10 a ch · · 1 am. The dorsal and the posterior segments had relative y greater 
neg · attve ve1 . 
ocity peaks than the other segments (EPG Stage II), indicating tongue 
lowering. 
The magnitude of the negative peaks, however, showed no consistent task-
specific trend 
s. During EPG Stage III, positive peak velocity for the posterior segment 
Was achj 
eved. This was true for the dorsal segment as well, in all except the 30-cc gelatin 
swallows ( 
see Figure 3.16, 30G). 
Average positive peak velocities from the different swallowing tasks for S5 are 
Presented. 
10 Table 3.2. Also included are the average intervals from onset to positive 
Peak Velo . 
city (OTP). Within-task comparisons showed the dorsal tongue segment to have 
greater Pea 
k velocity than all other segments for all swallow types. Between-task 
c
0111Parison 1 · · h JI 
s revealed that 30-cc water swallows had greater mean peak ve ocities t an a 
Other SWall 
ows for all tongue segments. The most striking contrasts observed were 
Figure 3.16. Velocity profiles for four tongue segments from representative discrete swallows of S5. Values (in mm/sec) are 
instantaneous velocities from first move of tongue to point of maximal posterior displacement. Missing initial data points for 
anterior segment were due to lack of tongue-grid intersections. Vertical lines demarcate corresponding EPG Stages 1-111. 
Ant = Anterior; Dors = Dorsal; Post = Posterior; SW = 5-cc water; 30W = 30-cc water; 5G = 5-cc gelatin; 30G = 30-cc 
90 
60 
30 
II III Ant 
0 -l--1-___.,_~ ..... ------i 
-30 
-60 
90 
60 
30 
0.3 0.6 0.9 1.2 
M id 
0 ...-l:""'"#-~--i--, 
-30 
-60 
90 
60 
30 
09 1.2 
Dors 
O~ -'+<_,~__,--, 
-30 
-60 
90 
60 
30 
09 1.2 
Post 
0 ~-::--:-+---t--1 
-30 0.9 1.2 
-60 
SW 
90 
60 
30 
II III Ant 
0 +--l~----+--, 
-30 
-60 
90 
60 
30 
oj O¥i o.9 1.2 
Mid 
o ~~~-1'---+--, 
-30 
-60 
90 
60 
30 
0.9 12 
Dors 
0 ~~-7#+.-""--t--, 
-30 
-60 
90 
60 
30 
1.2 
Post 
0 ..,Jq~- -lr-~- -, 
-30 0.9 1.2 
-60 
30W 
90 
60 
30 
II III Ant 
O+--l~----i--, 
-30 
-60 
90 
60 
30 
:0.3 !06 0.9 1.2 
Mid 
0 .V--~'--',_.._--, _ _, 
-30 
-60 
90 
60 
30 
ioJ ,o.6 o.9 1.2 
Dors 
0 ~r.,,i,~----, 
-30 
-60 
90 
60 
30 
0.9 1.2 
Post 
0 ~~-..,,_'"-!- -, 
-30 0.9 1.2 
-60 
5G 
90 
60 
30 
II III 
Ant 
0 +--++...._.., ___ _, 
-30 
-60 
90 
60 
30 
3 I 0.6 6.9 1.2 
Mid 
0 ....... ~'=""-t--~-----i 
-30 0.9 1.2 
-60 
90 
60 
30 
Dors 
0 ~,#'Pl&---1-'=--~-----i 
-30 
-60 
90 
60 
30 
1.2 
Post 
0 _ _....__....,.__....__, 
-30 L2 
-60 
JOG 
90 
60 
30 
II III Ant 
0 ~~lo.... ..... __,-----i 
-3o : 0.3 P6 o.9 1.2 
-60 
90 
60 
30 
Mid 
0 ~-+-- ,._---+-----i 
-30 
-60 
90 
60 
30 
0.3 0.9 L2 
Don 
0 ~--r:.+----1---; 
-30 
-60 
90 
60 
30 
0.3 0.9 L2 
Post 
0 1"""+N-_,_,,_~--; 
-30 03 0.9 L2 
-60 
Dry 
-0 
""' 
Table 3.2. Mean positive peak velocities (in mm/sec) and onset-to-peak (OTP) intervals (in sec) for the four tongue segments 
from different discrete swallows of Subject 5. SW= 5-cc water, 30W = 30-cc water, SG = 5-cc gelatin, 30G = 30-cc gelatin. 
Diff = difference in mean OTP values between tongue segments. 
Positive Peak Velocity OTP Interval 
Anterior Mid Dorsal Posterior Anterior Mid Dorsal Posterior 
SW M 21.49 48.72 53.49 30.86 0.42 0.54 0.63 0.73 
SD 8.00 14.76 4.14 8.39 0.05 0.02 0.03 0.03 
Diff 0.12 0.09 0.10 
30W M 53.57 55.44 73.04 67.97 0.32 0.55 0.63 0.73 
SD 5.17 2.32 22.25 7.00 0.02 0.02 0.00 0.00 
Diff 0.23 0.08 0.10 
5G M 33.78 30.30 70.74 61.47 0.27 0.52 0.76 0.86 
SD 2.04 11.49 5.54 19.36 0.09 0.13 0.13 0.14 
Diff 0.26 0.23 0.10 
30G M 27.77 53.30 65.61 53.23 0.35 0.80 0.87 0.98 
SD 5.77 1.56 0.97 2.53 0.12 0.14 0.14 0.12 
Diff 0.45 0.07 0.12 
Dry M 29.13 31.85 70.15 62.71 0.27 0.49 0.61 0.68 
SD 9.93 1.07 9.14 14.61 0.05 0.05 0.04 0.05 
Diff 0.22 0.12 0.07 
-0 
Vl 
106 
between 3 0-cc and 5-cc water tasks for the anterior and the posterior segments, where the 
larger volume consistently induced 38-78% (based on raw data) greater peak velocities. 
The 30-cc gelatin swallows also had 23-67% higher peak velocities than their 5-cc 
parallels, though only for the mid tongue segment. Dry swallows, on the other hand, had 
consistently greater posterior peak velocity than small-sized water swallows. Other 
differences due to bolus consistency were generally inconsequential. The gelatin and dry 
swallows hardly differed in peak velocity for any tongue segment, though most of the 
water swallows had greater peak velocities than dry swallows for the mid tongue. 
Data on the OTP intervals showed no particular size- or consistency-based trends. 
However, the duration between moments of anterior- and mid-tongue peak velocities was 
the shortest for 5-cc water and the longest for 30-cc gelatin. In addition, 5-cc gelatin took 
a longer time than the other tasks to attain dorsal-tongue peak velocity after reaching mid-
tongue peak. 
Correlation of EPG Timing Measurements with Ultrasound Activity 
Guided by the reference images (see Figure 2.4) which depicted the general 
landmarks of the individual pseudo palates, attempts were made to correlate the EPG 
timing measurements and associated contact activities with time-aligned ultrasound tongue 
shapes for general consistency. In 70% of the cases, reasonable parallels between the 
ultrasound and the EPG activities were observed. That is, activation of electrodes in the 
front bin on EPG (propulsion onset) matched maximal displacement of the anterior 
segment on ultrasound; end of propulsion on EPG followed maximal displacement of the 
mid segment but preceded that of the dorsal segment on ultrasound. In 30% of the cases, 
107 
a mismatch by one to two ultrasound frames did occur. Considering the difference in 
sampling rate between the two instruments (1/30 vs. 1/100 sec per frame for ultrasound 
and EPG, respectively), the mismatches appeared relatively trivial. 
Identification of Unusual Lingual Behaviors 
The screening of all swallows on ultrasound revealed only one case of "double 
swallowing" . This occurred on one 30-cc gelatin trial in SS, where the bolus clearly was 
emptied in two successive swallows--80% on the first and 20% on the second ( estimation 
based on the imaged bolus "area" at the onset of each swallow). Because there was 
material remaining in the mouth, the tongue did not lower completely at the end of the 
first swallow before the next wave of propulsive action began. The EPG correlate for this 
particular bolus showed the typical prepropulsion and propulsion activities but 
deactivation of electrodes sporadically on the entire palate near the end of propulsion, 
followed by complete reenactment of the propulsive process, then full contact, and 
withdrawal. That is, on EPG, the first swallow did not have a "full contact" stage or a 
typical "withdrawal" stage, and the second swallow did not have a prepropulsion stage. 
Although most swallows exhibited regular and systematic lingual behaviors on 
ultrasound, extra tongue movements were observed during prepropulsion in 9% of all 
data. These usually occurred as a rise-fall-rerise (in contrast to a simple elevation) of the 
tongue. The corresponding EPG data typically had a longer prepropulsion stage during 
which scattered clusters of electrodes were sequentially activated, deactivated, and 
reactivated. These cases were distributed across all swallowing tasks, but more in S2 than 
in the other subjects. 
108 
Summary 
The EPG data for single swallows showed stage-specific effects of bolus volume 
and consistency in contact timing. Mainly, larger volume and thinner consistency had a 
significantly shorter duration of propulsion (Stage 11), and dry swallows were significantly 
longer than water swallows in full contact (Stage 111). Contact patterns for the individual 
stages were distinctive but varied minimally as a function of swallowing task and subject. 
In the majority of the single swallows examined, tongue-palate contact followed a front-
to-back and simultaneous lateral-to-midline progression during bolus propulsion, and a 
front-to-back and simultaneous midline-to-lateral pattern of withdrawal. These details 
have not been included in conventional descriptions of deglutitive tongue action. 
While the ultrasound and the EPG swallow activities could not be precisely 
matched due to limitations unique to each instrument, the two techniques provided 
supplementary information that made the observation of swallowing more complete. 
Ultrasound imaging supplemented the EPG findings with activities before and after 
tongue-palate contact, while EPG filled in the missing information on lateral and medial 
linguopalatal interactions. The single-subject velocity data provided apparent indications 
of task-induced differences, especially with respect to bolus size. Larger-volume swallows 
did have greater peak velocities than small-volume swallows, at least for some tongue 
segments . This primary finding agreed with the EPG results and supported the usefulness 
of EPG analysis of swallowing. 
109 
Experiment 2. Sequential Swallowing in Continuous Drinking 
EPG Findings 
Tongue-Palate Contact Timing 
Total duration. Average total durations of the component (i.e., first and 
subsequent) swallows during normal- and fast-rate continuous drinking for Subjects I to 5 
are presented in Table 3.3 . It should be pointed out that while the number of"first 
swallows" averaged was consistently 2 (two replications) for each rate, that of the 
"subsequent swallows" ranged from 1 to 4. There were two cases where only one 
subsequent swallow was captured on EPG out of the two trials; as a result, standard 
deviations and coefficients of variation were missing (marked as "NIA" in Table 3.3). In 
general, the data in Table 3 .3 showed small intrasubject replication variability. In 80% of 
the cases, the coefficients of variation were less than 0.20. Although the fast-rate 
sequential swallows averaged 14% shorter than those in normal-rate drinking, the 
differences were not significant, nor was there an order effect. Subjects, however, differed 
significantly (Q = .0001 ), despite an insignificant subject-by-task interaction. The least-
squares means for subjects (see Table 3.3) revealed that SI (female) and S3 (male) on the 
whole drank considerably faster than the other subjects. 
Average total-duration ratios of sequential versus discrete water swallowing tasks 
for S 1 to S5 are presented in Table 3.4. Sequential swallows, in general, averaged 43-
56% faster than single water swallows among the subjects. Subject differences were 
Table 3.3 . Mean total durations (in milliseconds) of tongue-palate contact for sequential swallows during normal- and fast-
rate continuous drinking for Subjects 1 to 5 (S 1 to SS). Coefficients of variation (CV) were computed using the formula 
[SD/M]. LSM = least-squares mean. NI A= not applicable due to single measurement. 
Continuous Drinking 
Normal Rate Fast Rate LSM 
First Swallow Subsequent Swallows First Swallow Subsequent Swallows 
SI M 432.44 569.82 476.35 466.22 483 .11 
SD 19.11 74.42 23.89 66.43 
CV 0.04 0.13 0.05 0.14 
S2 M 739.87 902.03 608.11 655.40 733 .53 
SD 14.33 138.55 9.56 111.64 
CV 0.02 0.15 0.02 0.17 
S3 M 550.68 439.19 324.33 351.35 416.39 
SD 81.22 200.66 105.11 N/A 
CV 0.15 0.46 0.32 N/A 
S4 M 975.68 725.33 950.00 725.34 844.09 
SD 217.87 N/A 199.70 9.08 
CV 0.22 N/A 0.21 0.01 
S5 M 838.18 798.09 721.28 730.97 775.11 
SD 114.18 67.51 21.50 17.48 
CV 0.14 0.08 0.03 0.02 
LSM 707.37 687.20 616.01 591.20 -
-0 
Table 3.4. Average total-duration ratios of Sequential:Discrete water swallows for Subjects 1 to 5 (S 1 to SS). CN = 
continuous normal-rate drinking; CF = continuous fast-rate drinking; subscript F = first swallow; subscript s = subsequent 
swallows; SW = 5-cc water; 30W = 30-cc water; x = mean of means. 
Ratio 
x 
SI 0.47 0.62 0.53 0.70 0.52 0.51 0.58 0.57 0.56 
S2 0.45 0.55 0.60 0.73 0.37 0.40 0.49 0.53 0.51 
S3 0.39 0.31 0.46 0.36 0.23 0.25 0.27 0.29 0.32 
S4 0.66 0.49 0.67 0.50 0.65 0.49 0.66 0.50 0.58 
S5 0.61 0.58 0.61 0.58 0.53 0.54 0.52 0.53 0.56 
x 0.52 0.51 0.57 0.57 0.46 0.44 0.50 0.48 
...... 
...... 
...... 
112 
apparent in that S3 ' . 
s ratios (x = .32) were considerably smaller than those of all other 
subjects (x -
- ·
51
-.58; see Table 3.4). 
Because no significant differences could be attributed to drinking rate or swallow 
Order Withi h . 
n t e sequential swallowing tasks, the data on their total durations were pooled 
to com 
pare with the total durations of 5-cc and 30-cc single water swallowing tasks. The 
ANovA. sh . . 
owed significant effects of task (12 < 0.0001), subject (Q < 0.0105), and task-
bY-subject i . 
nteract1on (Q <0.0021). Post hoc comparisons indicated that sequential 
swallows w . . . . . 
ere significantly shorter than single water swallows of either size m total 
tongue 
·Palate contact duration. (The statistical tables for these tests are presented in 
Append· . 
ix B, section B-5.) 
~ . The same staging criteria set for single swailows were 
Usect to ex . . 
anune sequential swallowing. It was discovered that they could not be stnctly 
appJiect to th 
e majority of the data due to an overlap of propulsion and withdrawal 
activitie . 
s in both contact timing and pattern ( discussed below). As a result, the durations 
of PropuJsio d nJ f 
n and withdrawal were not measured; timing measurements were ma e o y o 
the Preprop 1 . fr h . 
u s1on stage in all cases (i. e., 44 component swallows om t e contmuous-
drinJc.i 
ng series of all subjects) and of the full contact stage in IO swallows where it was 
Present. 
The duration of prepropulsion activity was Jess than 60 msec in 71 % of the 
sequent· 
ta1 swallows pooled from all subjects (M = 61.10, range= 26.18-144.43). No 
St . flking d. 4-l:' 1uerences were observed due to drinking rate or swallow order. On the average, 
prepropulsion was 72% (range= 49-90%) shorter in sequential swallows than it was in 
single water swalJows. 
When present, fuJl-contact activity averaged 112.13 msec in duration (range= 
33. 78-301.69). This was on the average 84% (range= 81-87) shorter than that in the 
single water swalJows. Nearly alJ 1 O cases in which full contact was present came from 
the data of S2 (female) and S4 (male), representing about 55% of their individual total 
number of sequential swaJlows. No apparent relationship was detected between the 
incidence of fuJJ contact and drinking rate or swallow order. 
Tongue-Palate Contact Pattern 
General characteristics. The percent-contact profiles from representative 
continuous drinking series of Subjects I to 5 are illustrated in Figures 3 .17 to 3 .21, 
respectively. Figure 3.22 presents a representative raw EPG time series. In general, 
sequential swalJowing shared some of the basic contact characteristics of discrete 
swalJowing. That is, in each component swallow during continuous drinking, there were 
distinctive, progressive activation of electrodes for propulsive purposes, as well as 
progressive deactivation of electrodes during withdrawal. However, sequential swallows 
differed strikingly from single swallows due to an overlap of propulsion and withdrawal 
activities. Specifically, as seen in 77% of the data, progressive deactivation of electrodes 
(i .e ., onset of withdrawal) began to occur in the front or central bin while back electrodes 
were still being activated as part of propulsion (see Figure 3.22). As a result, the majority 
of the sequential task responses did not have a period of stable, maximal contact. 
113 
... (J 
.s 
8 
~ 
p· 114 
igure 3.17 . . . 
lllediaI- . di". C~anges m percent-contact over time m front-central-back and 1atera1-
47). C~ te ?ms from representative continuous drinking series of SI (female, age 
ontmuous, normal-rate; CF = continuous, fast-rate. 
100 . 
t 75 
.s 8 so 
* 
0 lllsec 
100 
t 75 
.s 8 so 
* 25 
100 
75 
so 
25 
0 
---Front --Central - - - Back 
,,- -
'J- . I,. - ... "" 
/ \ 
\ ', 
\ ' \ \, 
' ' 
"' 
200 400 
... 
600 800 
---Lateral - - - • • • Medial --·Midline 
200 400 600 800 
Front --Central - - - Back 
/.-.,. \. - _,.,'""" 
__ , 
I \~, ' ,, ... 
.... '-' \. 
,. 
- .... , 
\ \ ..._ 
' 
CN 
1000 1200 
CN 
1000 1200 
CF 
..... 
'\ 
... , 0 lllsec 
... 
'--0 800 1000 1200 200 400 600 
Lateral - • • • • • Medial Midline 
10() CF 
ti 
.s 
C: 
8 so 
lilt 25 
800 1000 1200 200 400 600 
115 
Figure 3 .18. Changes in percent-contact over time in front-central-back and lateral-
medial-midline bins from representative continuous drinking series of S2 (female, age 
46). CN = continuous, normal-rate; CF = continuous, fast-rate. 
---Front ---Central - - - Back 
CN 
\ 
... , 
"'\ 
" 
---· - ,,...,.,,,,..,, ..... __ 
600 900 1200 1500 1800 
---Lateral • • • • • · Medial ---Midline 
CN 
100 
t, 75 
"' c 50 0 
0 
~ 
0 
msec 0 300 600 900 1200 1500 1800 
Front ---Central - - - Back 
CF 
100 · ,..-r ___ , 
t, 75 \ / 
" 
\ 
"' 
r \.., .... \ c 50 ~, ... , -\. 0 
0 \ ,,,._ \ \ '#, 25 \ \ .. ",, \ . .-, 
"' 
.... _ 
----'*'""-"'-
~ .... _, 
0 
msec 0 300 600 900 1200 1500 1800 
Lateral • • • • • • Medial Midline 
100 CF 
, , ~ .. 
·-t, 75 . .. ' ' • 
s 
C: 50 0 
0 
~ 25 
0 -
msec 0 300 600 900 1200 1500 1800 
116 
Figure 3 .19. Changes in percent-contact over time in front-central-back and lateral-
medial-midline bins from representative continuous drinking series of S3 (male, age 23). 
CN = continuous, normal-rate; CF = continuous, fast-rate . 
---Front - Central - - - Back 
u 
ra 
c 
0 
u 
100 
75 
50 .., ..... ,, '"""" 
., ' 
'a' 25 - I ,, 
I \ 
----- ' - - - -- -- --0 ..... 
msec 0 250 500 750 1000 
---Lateral····· ·Medial ---Midline 
100 
u 75 
ra § 50 
u 
'a' 
msec 0 250 
c 50 .,. -\ 
~ 1~: ,- ,.,, 
o I \ 
u ~ 
'a' 25 //. \ 
____.// . ..... 
0 .... - , ... , 
msec 0 
100 
ti 75 
~ 
250 
8 50 -
.11-25 - ~ · 
0 , I 
msec 0 250 
500 750 1000 
Front - -- Central - - - Back 
500 
- ---~ 
,.. 
.. ,. \. 
I ' -
750 1000 
Lateral · · • · · · Medial ---Midline 
500 750 1000 
,. 
.. ,, -... 
I .,,.,,. 
" -..... , 
1250 
1250 
1250 
1250 
CN 
1500 
CN 
1500 
CF 
-, 
1500 
CF 
1500 
117 
Figure 3.20. Changes in percent-contact over time in front-central-back and lateral-
medial-midline bins from representative continuous drinking series of S4 (male, age 37). 
CN = continuous, normal-rate; CF = continuous, fast-rate. 
100 
-
75 u 
"' c 50 0 
(.) 
;f!. 
0 
msec 0 
100 
-
75 u 
s 
C: 50 0 
(.) 
;f!. 25 -
0 . 
msec 0 
100 
-
75 u 
"' c 50 -0 
(.) 
;f!. 25 
0 
msec 0 
100 
ti 75 
"' c 50 0 
(.) 
;f!. 25 
0 
msec 0 
---Front ---Central - - - Back 
350 700 1050 1400 
---Lateral • • • • • • Medial --·Midline 
350 
<xt,_· _. --11~-----
100 1050 1400 
Front ---Central - - - Back 
{ ,,. __ ._ , ~, ....... 
l 'v , .. .,_ .. / \ 
I '--·. -, ;\_ \ // \ \ . - \ r \ '" , .,.., \"' ., \ - - -~ .... _, 
\_\ 
- ~ 
350 700 1050 1400 
Lateral • • • • • • Medial • Midline 
.. 
. ,. ·'~ 
' . 
. 
350 700 1050 1400 
CN 
1750 2100 
CN 
1750 2100 
CF 
' \ \ 
"' .. 
1750 2100 
CF 
1750 2100 
'! I 
118 
Figure 3 .21 . 
llledial-midl: C~anges In percent-contact over time in front-central-back and lateral-
23). CN == ;:e ~ms from representative continuous drinking series of S5 (female, age 
100 
ti 75 
.s 8 50 
'* 25 
100 
ti 75 
.s 
8 so 
..41 25 
10() 
t 7 
.s 5 -
8 so 
'* 25 
l'ttsec o 
100 
J 7s 
8 so 
'* 25 
0 
ntmuous, normal-rate; CF= continuous, fast-rate. 
---Front --Central - - - Back 
CN 
300 600 900 1200 1500 
---Lateral · ···· ·Medial --·Mid/ine 
CN 
300 600 900 1200 1500 1800 
---Front --Central - - - Back 
CF 
300 600 900 1200 1500 1800 
--·Lateral .... · • Medial --·Midline 
CF 
300 600 900 1200 
1800 
Figure 3 .22. Raw EPG data for two sequential swallows in a representative continuous drinking series. An overlap of propulsion 
and withdrawal activities is evident. Vertical line demarcates onset of each swallow. 
1.21 · · · 
.•:: i :_) /1.) :•:. 
144 .. ·.:: •.• 
. .... , .... 
..... ···=·· •11 I 11•,1 ,e 
... :: . ~-· 11 I I I I" 
•a I I I I 1• 
•a I I a I 
... 
-
I .a 
II 
1.79.:.:: . . 
: : ·.:.= ~ :_=: : • .-. 
• I 
•, ·,I• 
,•1 I ; : 
.. 
•• ... . . • I 
.. I • I • :: 
2.03 .... : .. 
.. . •.• 
:.·: ·;_: =:: :•: . 
. , . 
2.33.:.::, . 
. ~~:::11::;~, 
I I 11 I al I I 
••••••• .. . . ... 
•1 I I I 1 • 
I I I I 
.  .. 
• • • • ;:·, !•:: 
:~~\l:\:\ ..  
I I I I I I 
.. . . ... 
. . . 
•• • • •• 
II I • I .. 
1.~.:. : .. 
.·.:::·:,: :_:.; :~·. 
1.47 .• ·. : -~· 
. :::::1:::~. 
:• .... : .... ··. 
"II I I I I I•,. 
"I I I I I I ' 
I I I e 
.. . . . ... 
18 • • N 
1.82 .... : .... 
· .. •: :_:_:: ;._ 
., . 
.. . 
... - . 
~= • I 
. .. 
. .. 
. . 
.. 
2.04 .... ~ ·~: . 
,._:j:.)(\:\ .. 
.. . -
2.36 .:.:: .. 
•. :;::I:::: .• 
.. ·: ,::•,. :· .. 
• 1 I I I I I 11 
.. . . . .. 
"I I I I I I 1 • 
I I I a 
~= • I I "=~ 
~D\\t: 
. . . . . 
I I I I I I 
::•1:1•:; 
1.24.,•i•-:. 
:.:i)?:,•\ . 
1.50 . :. ::, . 
. .... , .... . 
...... .... . 
•• : ·:·=··: :· •• 
•• • • ••• 
"I I I I I" 
"I I • I I I 1• 
I I I I 
::•1:1•:: 
1.84.::: :··: 
.·.·:·-:.: =.::: •: . 
•·· ... 
• I • I 
•• • 
. , ' I , I 
I I' 
.. 
.. 
2.06 ····i'·· 
.:/}(:.\· . 
2.4_1. :~~=;; :- . 
. ···=·=··· •. I •a •a I 1 1 I • I I I I • I I 
•1 a I I 11 
.. . . . .. 
.. ,•:I:-. •. 
~= I I : I I~~ 
2.75 .. :. :,. 
:::::,: =.:.:: .... 
.... 
.. . - .. 
• • 
:: I I : . .. H 
1.2~ ::~if==~ . 
. . ... .. .. 
:·: ·.•: ··: :_ ·: . 
1.55. :. : .... 
. ;·,:1:-•:. 
••••••••••••• 
•• • l,i • ••••• 
:.· I II : .. I•:, 
.. : ... · ... 
'• . • • I I . I 
.. I • I N 
'·?:.)1\:-:. 
. . 
I, I • I 
'I I 
·• • I • 
. . 
... 
.. 
2.10 ··"'i'·=· 
. }.~_:::_~: \ . 
2.43. , , . 
. -:::~i~::: . 
.... . ,........ . 
.. : . : ... . 
•a I I I I 11 
I I I I I • I" 
•• : I I I I a• 
.. I • I • ~~ 
?::<-\ 
11 I I I I a a" 
:: • I : I • :: 
1.30 .... ,., ••• 
•••• ••• • ..•.... ..... 
. ·:: ·~·: ··.·: .  
1.6~: :~: :';:,. 
.. ~··=!=•:; •. :.· .... : ~ ..... 
•• a 8 I I I I" 
'• 11 I • • II a• 
11 I I I • I ,I 
II I • I .. 
1.88.;. :, . 
. ::::: :_:_::,;. 
. " I , I '~'. 
2.13 .... ,., •• 
.... ..... 
·~~·-~:! ··~ •. 
:: : .. ·: ·. :. 
2.46_.:-; : .. 
. -.- .. - . 
,,;···::··- · 
. ·..... : ·.·. ·· . 
.. ·: .· . ·. •.· .. 
·a I I I I I 1 · 
~= 1 I : I 1 ;~ 
2.7~.::·: :·::,. 
:::_:··:,: :.:::\. 
.... . 
:: • I • ... .. 
1.33., .. ,., ••• 
••••• ••• • 
I •11~ : ~a•~al 
.. ·~ ·~ . ·: ... 
'•. 
1.63. ·.:; . 
. -:::i~-~: :. . 
. :: ,::· .. :· .. 
•• • • • • ••• 11 I I I - I' 
la 
1
1 I I I 111, 
•• • • • • ••• 
II I I I .. 
1.90 ::':(:':. 
::: :·-:.:: :.-.· :•: . 
,; • I •· 
2.1 ~ .. ":i:"::• • 
.......... . 
···~:·· .. ... . 
:.•: . : . · .. . 
.. . 
.. 
2.49. · .. 
.. :)\:.::,, 
.. . . : .... 
"I I I I I I 
11 
11 
I I I I a• 
11: I I I I•" 
II I • I ·:'.' 
2.8\: : :::·. 
•,:·•: :-·: .. 
:: : :: ·.: ·. · .. 
.. 
. • I 
.. 
1.36 .... 1·1 .. , 
••••• •••• 
••••••••••• 
•• ·~ ·.i. ·,: ~ •• • 
. ·: .. : .. · .. 
1.6~:,':::',:. 
·-~ · .. -:; :,.·;., 
'• I I • .. al' 
.. . . ... 
la I I I I" 
I I I I I I I 
1 I I a I 1 
:: I I : I • ;: 
1.92.;. :,. 
:/.:_: =,:.::~.:-. 
i • • 
2.~:-:i:·:=·· 
···=··!····· ...... .. .... :.·: . : ·. ·. -: . 
.. . 
.. . 
2.52.::':1:':: 
. : ·.:: = · · :,· 
:.· •• ... i ·•· ~ ~· . 
11 I I I I' 
... ·:. =~·· 
la I .. 
.. ' I , I ', . 
2.83_ ..... · .• 
. ... . 
:::_:·-:.: :,:::\. 
• 
i •• 
1.~ .... , ·:··=· 
·=·===.==.·=~·. •,:-:- : ·. ·.• .. 
.. . . .. 
'IO O I •• 
.. 
1.7~. :: : :·,:. 
·.:· . .:::-.·:,· 
~-. .-:· ...... 
I I I I I a I 
I I I a a I I a I 
I I I I 
11 I I • I a .I 
II I I I .. 
1.~.\t !}_::.., 
• ,i 
.. 
.. 
2.2~:··:, ·:::•. 
···=· =··=·· ::: ~·: ·: ..... 
.. . . .. 
'I I • ". I' .. . 
.. 
•· 
2.56_.;.;:, . 
. .. 
.. 
.....  ~·-:.=f =.:: ;~·. 
•1 I I • I I" 
.. : . . . ... 
11 I I I I 81• 
.. I I I ·:'.' 
2.85 .; ..... 
.. /~:,:.: :,:,::\. 
i I • 
1.~···1 ·~~: •• 
··=== ·=·:.· .·.·· ... ~ ..  ·I 111 
a • • I I 
•1 I • 1• 
.. . . .. 
.. . .. 
.. . 
1.7~: .":: :·-: 
:: ;·-:.:: :.:: ;•.·. 
•• I I I 11 
.. . . . ,. 
1 
1 I I I • 1 
• • • • 
1 1 I I I 
I 
I .a 
.. I • I H 
1.96 '•' .. 
.)}:) :_:,::\·. 
'• 
. . 
2.27 .... , =~-.;. 
-~::: ==·~· 
-~·: ·.:: •: ·. ·•· . 
. ·:: . . ·. ·:. 
• 
... . 
... 
.. 
2.~.:'.::, . 
. .. 
.. 
:\·-:.:1:,:::-:. 
11 I I I I •' 
... · : . . ... 
11 I I I • a• 
.. I • I •:'.' 
2.87. ·. : . 
·.· .. · .. 
. , .._:··:,:=.::\·. 
; I 
1.42 ····1=\. 
····: =····· 
··=·:::·.·· :· •• 1 1 I I I • I 1 1 
11 I I I a• 
"I I I 1
1 
I' 
.. . .  
1
1 • II II 
1.76,.::: .. ,:. 
·::·.-.:::.-;: . .-. 
•·· .· .·: ·. ·. ··• 
I I I I I I I~ 
I I I I I 
la I I I ,I 
II I I I II 
1.98 .... :,. 
:,:/:_::_:.\ . 
2.30.,·.:.~. 
.~~:::I:::;;, 
.. :-.:,.•: .. 
"I 8 I I I 11 
'I I I 1· 
·,.·:. :· ... 
• I 
.. 
I 1• 
II 
2.65.::':):'::·. 
:::·.:.=::.:::•:. 
•, I I I .• 
.. . . . .. 
11 II I I I • ••• 
I O I 
:: I I • I • :: 
27?) 1/ ~_;.:., 
"° 
120 
A second characteristic of sequential swallowing was that the withdrawal pattern 
followed a front-to-back progression without a distinct, simultaneous midline-to-lateral 
component (see Figure 3.22). This was observed in 70% of the cases. 
A third characteristic, though somewhat less compelling, was that 89% of all 
sequential swallows showed contact Pattern 1 during prepropulsion (i.e., small, scattered 
changes in contact over a brief period, mostly in the lateral bin, resulting in lateral contacts 
of < 25%). The incidence surpassed that of the same pattern for single swallows by 
almost 10%. 
No striking differences in contact pattern were observed due to drinking rate or 
swallow order. Intrasubject variability in overall contact pattern was small in nearly all 
cases. Intersubject variations, however, were apparent and primarily due to S3 who used 
predominantly back-bin contacts to accomplish propulsion during continuous drinking (see 
Figures 3.19 and 3.23). A post-experiment interview revealed that this subject described 
his manner of drinking as "chugging". Interestingly, even in the self-proclaimed chugging, 
the propulsion-withdrawal overlap was present at times. Additional subject differences 
were seen in prepropulsion and withdrawal patterns. Four of the five cases demonstrating 
prepropulsion Pattern 2 occurred in the data ofS4. Interestingly, SI, S2, and S3 who 
used Pattern 2 in some of their single swallows used strictly Pattern 1 during continuous 
drinking. Cases where the withdrawal pattern was similar to that for the single swallows 
(i .e., in a simultaneous front-to-back and midline-to-lateral progression) occurred mainly 
in the data of S2 and S4. 
Figure 3.23 . Raw EPG data for two sequential swallows in the continuous drinking (self-proclaimed "chugging") series of S3. 
Contact activities occur predominently in the back bin. Vertical line demarcates onset of each swallow. 
1.72. :.::,. 
. ···=···. 
. : :·:.::=:.·: · .. 
·. · ... 
1.73,:j:";:-. 
... • .• , ,•. 
::·.: .... :··-: :_· ... 
•· .. .. • • I• 
1.92. :.':::··: 
. . . . . 
·:·-:.=::.:: :-:. 
I • I • 
.... 
:: • I 
I a•: 
I I I 
I ·~ 
2.7~ : :: ::··: 
-::-.: :::::::-
..... 
·. · ... 
•· 
2.87 .:.:: .. 
.. ?-:-= \-t ...  
•• 
1.94 .:.: :, . 
. ····=·· ··. 
. :·:.===:.·:.• 
. : : .... · .. 
'• • 
~I - ~ 
. .. · I 
I I I I 
••• • I • .. I • I II 
2.74 .;.:: .. 
0 0,0,0,0 I 
.. ..... . 
. :·.: ·:::\ :_···.·. 
•· 
2.8:.:):': :. 
. :·_:·-:.:1=.::·:_. .... 
• 
. . 
I I a 
1.96_ .:.::,. 
. -:-.- ·. 
:,:_: ··:.=:=.:::_',:. 
· .. ·· . 
.. . .. 
... . ... 
II I I " 
2.76 . :.:: .. 
,',,•:•.•', 
.: ·:.=:::::-_. 
.. : ... · ... 
. .. 
.. : ... 
•· 
2.92 .;.:: .. 
. ·,.-:• .•', 
:: _:·-:-==:.::·:_-.... 
a I 
1.78 .:.:: .. 
. . ::i!ii:: ,. 
.. _ .. _: ·.·· ~ ·-: ~:-:. 
. .. 
.. 
1.97_.:.::,. 
. . ,,:•.•. 
.. :_:·-:.= :=.::·:_. .... 
.. -~ 
.. . . . : ~. 
1.~.:.:: .. 
·. :·i=i· :> 
. . .....•. . . 
• • I I I a I • • 
. ·: I ." : •., I·.~ 
a. I • • • . 11 
•· 
1.99. ·,:;. 
. .. .. 
. .... 
,,:_:··:.===..: :_-...  
.. 
I '' 
I •' 
1.82.:.:: . . 1.86 . . :.::,. 
.· .. ·:·.··· .... . 
1.84.:. ::,. 
. : •.: ::- . : .. :: . ·::~· .... ·. ·: . 
:.·. ·: .. ·: . ·•·. . . . . 
I I 1"1 I , .• • I 1
1
1 
••• I • 0 • •• •• • •• 
11 • : I I I 11 11 I I I ,I 
.  • . .1 . • . .• 
.. .. •· • • II I• 
2.01 .:.::,. 
. ... :•.·. 
. : ·-:: = · ·:. 
. . ·:. ·: ·. · .. 
, . 
I• 
2.04_ .:.::,. 
. •.•:·.·. 
2.07 .:.::,. 
:.:_:'/I=.::·::-:· • 
• 
1.88. :.:: .. 
. ·······. 
.. ... ~ ·.:-= ~ =.::; · .... 
11 I I a • I .. 
•. : . • • •• 1• 
I I 
... . .. 
II : I II 
211 .;.::,. 
. ... :•.• . 
:·:·-:.=:=.::·::-:· • 
.. I I .. =~" I : I •· •· ,i • i • ;. •· ii a i I;: a , • ,i I • .. •. i •. 
2.77_.:.:: . . 
···=····, 
······-·· 
. : : ._::·-: :· 
.·:. ; . ·:. 
.. : ... 
2.78 .:.:: . . 
, '•, •: • ,•', 
.. : ·-:-= ==.::·: .. ·· .. ·. 
.·: . . 
.. : ... . 
.. . . .. 
.. • • I• : I 
2.93_.:-::, . 
. .... 
......... 
. .·: -::·-: ::-.. 
I I 
2.80 .:.::, . 2.81 .:.:: .. 
. ··· ····. 
.·•.•:•.•', 
. ....... .
.. .... : :· .. i\ ::·.~ . 
... · . ~ .. 
...... : '/ l :.::· ::·.:·. 
•... . : · ... 
I .. I' 
II : I . . .. I : 
I ,I 
.. 
2.9~··:'·\ .. ::.. 
::)')\=,:::_-., • 
3.00 .... ::,. 
. · •.•:•.•" . 
..::,-.:·-:-=\=.::·:_-.: . 
I • 
2.83. :.:: .. 
.····=····. 
2.85 . :.::,. 
o •,•,•,• I 
. .. ·_:·-:-=1=.::·:.·.:· • :,:_: ··:.=: :.::· :_-.... 
I 
I I • 
•1." a I : .,.11 
•· I : ! • :: •· 
• I I I I • a· 
... 
.. 
3.0~.: :·:1:·:: . 
. .. -_:·-:.=1=.::·:_-.... 
;: . I 
" .. I I• II ! I• II I I I I• •• I I I •. .~. ii ,i 
.. 
•· •· .. •· 
. . 
.. •· •· 
1.90 .:.:: .. 
. ······· . • ........ . 
.: · .. ·:· .. ·· .. 
• • , " • I • '1 ,1 
la I I • 1• 
. .. ·: :· ... 
... I ' 
• I I .. 
2.1~.:.::,. 
···=···· . 
::··:,:=:.::·:_. .... 
,. 
• 
I 
• • 
I 
. .. 
I I• II I II 
•· 
-N 
-
Sequential versus discrete: statistical analyses. Sequential swallows were 
compared with single water swallows for the order in which maximal contact was 
122 
achieved in the front (F), lateral (L), and back (B) bins during propulsion. Results showed 
that single water swallows had a variety of sequences, whereas sequential swallows had 
only a few. The predominant pattern, observed in 46/60 single waier swallows and 35/44 
sequential swallows, followed a front-lateral-back (F-L-B) progression of"peaking". The 
other patterns were few in frequency and thus grouped as "others." Subsequent Chi-
Square analysis of the "F-L-B" and "others" groups showed a significant subject effect but 
no significant effect due to task nature (sequential vs. discrete), drinking rate (normal vs. 
fast), or swallow order (first vs. subsequent). As in the discrete swallowing experiment, 
the sources of the subject effect for this analysis were S3 and S4 (both male). For 
sequential swallowing, while all female subjects showed only the F-L-B sequence, the 
male subjects showed predominantly the "other" sequence patterns (6/7 swallows in S3, 
4/7 in S4). 
Front:back and lateral:back ratios in percent-contact at the time of maximal 
midline contact were calculated for all sequential and discrete water swallows. Results of 
the ANOVA for each comparison showed a significant task effect and a significant task-
by-subject interaction. Subject differences were not significant in either case. (See 
Appendix B, sections B-6 and B-7 for ANOVA tables.) These results are illustrated in 
Figure 3.24, a plot of the least-squares means for the subjects as a function of swallowing 
task. Post hoc contrasts revealed that single water swallows of either size had 
Figure 3.24 
(lower pa · 
1
t east-squares means for Front:Back (upper panel) and Lateral:Back 
Subjects 
1 
~e 
5 
percent-contact ratios at time of maximal midline contact for 
water; 
30
; - (S I to S5) as a function of sequential swallowing task. SW= 5-cc 
continuous£ JO-cc water; CN = eonunuous normal-rate dnnking; CF= 
swallow La
st
-rate drinking; subscrpt F = first swallow; subscript s = subsequent 
same amou t f num ers mean more contact in the back; a ratio of I. 0 means the 
s. ower b 
n ° contact in both bins. 
---s1 
-o- S2 
1.2 
-s3 
1.0 -0- S4 
-.-ss 
.2 
-~ 0.8 
.x 
u 
! 0.6 
.; 
C 
0 
... 
LL 0.4 
0.2 
0 .0 
SW 30W c~ 
CNs CFF 
CFs 
-a-S1 
1.2 -o- S2 
---s3 
1.0 -o- S4 
0 
.: -ss 
1111 0 .8 0:::: 
~ 
u 
~ 0.6 
ra 
... 
s j 
0 .2 
0 .0 
SW 30W Ci'Ji= CNs 
CFF CFs 
123 
I 
1 
I 
1: 
[ 
,, 
1: 
l: 
1: 
,,, 
11 JI 
1: 
II ,, 
I 
I! 
1: 
... 
sig ·r, 124 
l1J lCantly hi h . 
g er ratios than the component swallows of continuous drinking. However, 
there 
Were no signifi . 
. icant differences among the sequential swallows due to drinking rate 
or swallow ord 
er. 
Ultrasound Findings 
The ultrasound data on sequential swallowing during continuous drinking were 
examined £ 
or all subjects. However, due to considerable variations in image quality, only 
the d 
ata from S5 
were used in illustrations and for displacement measurements. 
Bolus size for the component swallows of continuous drinking was estimated for 
each b' su ~ect · 
'using the sagittal ultrasound data. As shown in Table 3.5, S1 used an equal 
number ofs 
Wallows for both normal- and fast-rate drinking, while all other subjects 
showed sli h . 
g tly fewer swallows (hence slightly greater volume per swallow) while 
drjnl.! 
'U\..lng fast. 
On the average, the two male subjects--S3 and S4--consumed more water 
Per swa11 
ow th d'd . an 1 the female subjects, especially for the fast-rate senes. 
Tongue · · · d 'nki movements for sequential swallowmg dunng contmuous n ng were 
ibnda111 
entally the same as those seen during single swallows (see Figure 3.25; cf single 
swallows in p· . . . 
igure 3.15). That is, the sequence in which the antenor, mid, dorsal, and 
Posterior segm . . . al d' l t 
ents elevated to and withdrew from thelf respective maXIm 1sp acemen 
reI11ained 
unchanged. However, in agreement with the EPG findings, shorter overall 
I 
·[ 
/I 
II• 
/i 
f 
"' 111 
,,, 
,,, 
1/ ;,, 
/I 
f: 
125 
Table 3.5. Estimated bolus size (in cc) of sequential swallows during continuous drinking 
for Subjects 1 to 5 (SI to S5). The total amount of water consumed for each trial was 
200 cc. 
Subject Rate No . ofSwallows Average Amount per Swallow 
SI Normal 7 29 
Fast 7 29 
S2 Normal 12 17 
Fast 8 25 
S3 Normal 5 40 
Fast 4 50 
S4 Normal 7 29 
Fast 5 40 
S5 Normal 16 13 
Fast 13 15 
126 
Figure 3 .25 . panels on left Changes m tongue shape in first and second swallows during normal- (CN, 
illustrate , ) and fast-rate ( CF, panels on right) continuous drinking. Frames in Rows I to 6 
loaded ;n'~ 0rder, lingual surface contours at swallow onset, when entire bolus was frrst Posterior t ongue surface, and at maximal displacement of anterior, mid, dorsal, and 
ongue segments. Arrows in each image delineate tongue surface. 
2206 
- ·· 't i '- \, 
1- - .. _·- ~ ~ 
I 
CN·2 •,1141- .-.: I 
CF - 2nd 
CF -1st 
CN - 1st CN -2nd 
Ii 
I! 
1: 
It 
I' Ii 
I! 
127 
swallow durations and an overlap of propulsion and withdrawal activities characterized the 
ultrasound images of most sequential swallows. The overlap usually became visible near 
the time of maximal displacement of the posterior segment, at which point a slight descent 
of the anterior segment could be seen (see Figure 3.25, Row 6). This was followed by 
further descent of the anterior and mid segments, which coincided with signs of water re-
entry, while the dorsal and posterior segments remained "anchored" (see Figure 3.25, 
Row 1, CN-2nd and CF-2nd swallows). When apparent descent was seen in the dorsal 
and posterior segments, the anterior segment usually had just begun re-elevating (see 
Figure 3.25, Row 2, CN-2nd and CF-2nd swallows) to initiate the propulsive action for 
the subsequent swallow. Thus, it appeared that the dorsal-posterior lowering gesture 
overlapped the anterior re-elevation gesture. As a result of the movement overlaps, most 
sequential swallows during continuous drinking, especially those that were not the first in 
a series, did not display the same extent of prepropulsion and full contact tongue behaviors 
as did the single swallows. 
Because of large intersubject differences in per-swallow volume, the starting 
tongue height for sequential swallows varied considerably from subject to subject. While 
no marked within-subject differences in tongue shape due to drinking rate appeared to 
exist, some variations were observed due to swallow order. Specifically, the anterior 
segment was usually less depressed and the dorsal and posterior segments less humped for 
the first swallow than for the subsequent swallows (see Figure 3.25, Row 1). 
128 
Lingual Velocity Profiles: Single-Subject Data 
Frame-by-frame displacement data were extracted from the four tongue segments 
for the continuous drinking series of SS, using the radial grid system. Instantaneous 
velocities were then calculated for four sequential swallows in each series. The plotted 
velocity profiles (time-series waveforms) are illustrated in Figures 3.26 and 3.27. As in 
the case of single swallows, these velocity profiles clearly captured the anterior-to-
posterior progression of bolus transport and the spatiotemporal concatenation of the four 
functional tongue segments. A comparison of normal- versus fast-rate series and of first 
versus subsequent swallows showed no marked differences in overall waveform 
characteristics. 
Average positive peak velocities for the sequential swallows of SS are presented in 
Table 3 .6. Also included are the average OTP intervals. As in the single-swallow study, 
within-task comparisons showed the dorsal tongue segment to have greater peak velocity 
than all other segments for most sequential swallows. Between-task comparisons revealed 
that "subsequent swallows" in either normal-rate or fast-rate drinking had lower mean 
peak velocities than the "first swallow" for the mid tongue segment but greater mean peak 
velocities for the dorsal and posterior tongue segments. Differences due to drinking rate, 
however, appeared minimal. In contrast to 30-cc single water swallows, sequential 
swallows regardless of rate and order had considerably lower peak velocities for the 
anterior segment, and "first swallows" regardless of rate had considerably lower peak 
velocities for the posterior segment. In contrast to 5-cc single water swallows, 
"subsequent swallows" regardless of rate had greater peak velocities for the dorsal and 
129 
Figure 3.26. Velocity profiles for four tongue segments from the continuous, normal-
rate drinking series of S5 . Values (in mm/sec) are instantaneous velocities from first 
move to end of fourth swallow. Missing data points for anterior tongue were due to 
either lack of or difficulty in discerning tongue-grid intersections. Vertical lines 
separate the four swallows in the series. 
120 T Anterior 
90 
>, 60 
:!:: 
8 30 
"ii 
> 0 
0.5 2 2.si 3 3.5 
-30 I 
-60 
120 
90 
~ 60 
8 30 
"ii 
> 0 
-30 
.5 
-60 
120 
90 
>, 60 
:!:: 
8 30 
"ii 
> 0 
-30 
-60 
120 
90 
>, 60 
:t::: 
u 30 0 
~ 0 
-30 
-60 
130 
Figure 3.27. Velocity profiles for four tongue segments from the continuous, fast-rate 
drinking series of S5 . Values (in mm/sec) are instantaneous velocities from first move 
to end of fourth swallow. Missing data points for anterior tongue were due to either 
lack of or difficulty in discerning tongue-grid intersections. Vetical lines separate the 
four swallows in the series. 
120 Anterior 
90 
>, 60 
:I:: 
8 30 
'ii 
> 0 
-30 0.5 
2 2.5 3 3.5 
-60 
120 Mid 
90 
~ 60 
8 
'ii 
> 0 
3.5 
-60 
': t 
Dorsal 
>, 60 
:I:: 
8 30 -
'ii 
> 0 
3.5 
-30 -
-60 
120 Posterior 
90 
~ 60 
8 30 
'ii 
> 0 
3 3.5 
-30 
-60 
...... 
Table 3.6. Mean positive peak velocities (in mm/sec) and onset-to-peak (OTP) intervals (in sec) for the four tongue segments 
from the continuous drinking series of Subject 5. CN = continuous normal-rate; CF= continuous fast-rate; subscript F = first 
swallow; subscript s = 3 subsequent swallows. CONT ALL = all swallows of continuous drinking. Diff = difference in mean 
OTP values between tongue segments. CNF and CFF lack standard deviation data due to single n. 
Positive Peak Velocitl'. OTP Interval 
Anterior Mid Dorsal Posterior Anterior Mid Dorsal Posterior 
CNF M 20.12 57.91 63.47 37.95 0.17 0.47 0.57 0.97 
Diff 0.30 0.10 0.40 
CNs M 27.20 47.82 97.19 54.14 0.00 0.22 0.32 0.42 
SD 0.00 4.01 11.11 26.32 0.09 0.06 0.09 
Diff 0.23 0.10 0.07 
CFF M 25.98 51.44 56.77 29.03 0.13 0.40 0.47 0.63 
Diff 0.27 0.07 0.17 
CFs M 16.76 24.59 72.77 74.12 0.00 0.24 0.34 0.40 
SD 0.00 1.41 10.28 19.91 0.02 0.02 0.06 
Diff 0.24 0.10 0.06 
CONTALL M 22.52 41.00 77.54 53 .94 0.08 0.29 0.38 0.50 
SD 4.93 14. 13 16.74 25.08 0.09 0.10 0.09 0.21 
Diff 0.25 0.10 0.12 
...... 
w 
...... 
132 
posterior segments (see Tables 3.2 and 3.6). It should be reiterated that this subject's 
average bolus size was estimated to be 13 cc during normal-rate drinking and 15 cc during 
fast-rate drinking (see Table 3.5). 
Data on the OTP intervals revealed no outstanding dissimilarities among the 
"difference scores" (differences in mean OTP values between tongue segments) as a 
function of drinking rate or swallow order. Sequential swallows on the whole were also 
similar in difference scores to the single water swallows (see Tables 3.2 and 3.6). 
Summary 
The EPG data showed that in contrast to discrete water swallows, the sequential 
swallows during continuous drinking had a significantly shorter total task duration, a 
shorter prepropulsion interval, an overlap of propulsion and withdrawal activities, a higher 
incidence of prepropulsion Pattern 1, a predominantly front-to-back withdrawal pattern, 
and significantly lower Front:Back as well as lateral:back percent-contact ratios at the time 
of maximal midline contact. 
The ultrasound data on sequential swallowing during continuous drinking were 
useful in confirming two prominent contact features observed on EPG, namely, the 
shortened overall swallow duration and the propulsion-withdrawal overlap. The single-
subject velocity data showed some differences in peak velocity for selected tongue 
segments due to swallow order but not drinking rate. There was also evidence, based on 
the comparison between sequential and discrete water swallows, that larger boluses had 
greater peak velocities than smaller boluses, at least for some tongue segments. 
CHAPTER IV. DISCUSSION 
The present study represents a first attempt to examine in detail linguopalatal 
contact characteristics with support from associated tongue shape changes during discrete 
and sequential swallowing via simultaneous eletropalatography and ultrasound. This 
chapter discusses findings in relation to the experimental questions raised. In addition, a 
model is proposed for a 3-D representation oflingual action for swallowing. 
Discrete Swallowing 
Tongue-Palate Contact Patterns 
The first question asked by the present study was whether bolus properties--size 
and consistency--would affect EPG contact patterns and timing during single swallows. 
With respect to tongue-palate contact pattern, no striking, bolus-induced differences were 
identified. Instead, systematic, stereotypical contact behaviors persisted across . 
swallowing tasks and subjects. These contact behaviors showed distinct, time-varying 
regional specificity. That is, EPG activities occurred systematically in different palatal 
areas at specific times during a swallow. To delineate the contact characteristics in detail, 
the present study devised three new data analysis schemes: compartmentalization of the 
palate into six bins that encompassed the lengthwise as well as the crosswise spatial 
domains, segmentation of the time course of the swallowing task into four functional 
133 
134 
stages, and the use of time-warping to average by bin and stage across repetitions. The 
division of the palate into bins and the swallowing task into stages does not imply 
interruptions in contact continuity. Rather, these were useful strategies that optimized 
examination of the spatial and temporal details oflinguopalatal interactions unique to 
swallowing. The use of averaged data also added robustness to the interpretation of 
findings . 
The observations of systematic, prescribed tongue-palate contact patterns 
suggested stereotypical lingual movement strategies for discrete swallowing. In her X-ray 
microbeam study (1991) that examined midsagittal movements of the tongue, Martin 
concluded that "during swallowing, the tongue functions as many small units that are 
selectively activated in a highly organized, invariant manner" (p. 283). The finding of this 
investigation supported such a conclusion, providing indirect evidence for concatenated, 
functionally dependent lingual units. However, in contrast to the conventional 
observations of swallowing based on 2-D imaging data, the present EPG findings revealed 
that the linked lingual units possess multidimensionality. Specifically, the swallowing 
movement patterns observed on EPG exhibited two degrees of freedom for tongue-palate 
contact : a front/back continuum and a lateral/midline continuum. In other words, two 
sets of concatenated tongue units were revealed, one in the lengthwise domain and one in 
the crosswise domain. Moreover, during propulsion, the activation of each set oflingual 
units followed two distinct patterns of temporal coordination. In one pattern (80% of 
data), the lengthwise and crosswise units were activated simultaneously in a front-to-back 
and lateral-to-midline progression. In another pattern (20% of data), the lateral and outer-
135 
medial units of the crosswise domain were activated prior to the front-to-back activation 
of the lengthwise units, suggesting the formation of a prominent lateral seal prior to 
propulsion which was subsequently completed by primarily the medial portion of the 
tongue. It was clear that the EPG contact patterns provided important information on 
deglutitive lingual strategies, and indirectly on tongue constraints in stereotypical 
behaviors. Swallowing investigations using midsagittal radiographic or ultrasonographic 
imaging techniques cannot capture the multidimensional components of deglutitive tongue 
behaviors. Even in biplane videofluorography or combined sagittal and coronal 
ultrasound, only a thin slice of the tongue may be visualized at a time, limiting the 
perspective of observable lingual action. EPG, therefore, represents a very useful tool that 
reveals conventionally unavailable information. 
The four EPG stages reflected four specific oral lingual motor goals carried out 
within the experimental design for the discrete swallowing task of this study. It should be 
reiterated that while the motor events for Stages II and III were integral parts of discrete 
swallowing, activities during Stages I and IV were influenced by requirements of the 
experimental protocol and consequently had artificial components. Furthermore, while 
Stage I was not totally unrelated to swallowing, Stage IV might well be, considering the 
specific instructions given to the subjects ( open mouth immediately after the swallow). 
Nevertheless, two interesting observations were borne of Stages I and IV. First, with 
respect to Stage I, it was apparent that normal individuals had the option of 
loading/positioning the bolus in preparation for oral transport with or without a prominent 
lateral seal. Second, with respect to Stage IV, while the retraction of contact followed the 
136 
same front-to-back progression as in Stage II, the medial-to-lateral offset progression was 
the reverse of the lateral-to-medial onset sequence seen in Stage II. This withdrawal 
pattern appeared to reflect the action of the genioglossus muscle. Since this extensive 
muscle inserts midsagittally across the length of the tongue, sequential contraction of its 
anterior, middle, and posterior fibers would produce the front-to-back withdrawal 
progression and at the same time cause an earlier release of contact from the midline than 
from the lateral bins. 
The EPG stages cannot be directly compared to the conventional swallowing 
phases extracted from videofluoroscopy (Dodds et al., 1990) or ultrasonography (Wein et 
al. , 1991). However, some links between the stages and phases can be made. The EPG 
Stages I and II, when combined, reflected the major portion of the videofluoroscopic or 
ultrasonographic oral phase where the main activity was the loading and transport of the 
bolus through the oral cavity. Because EPG registered only tongue-hard palate contact, 
the propulsive elevation of the posterior tongue segment (still part of the radiographic and 
ultrasonographic oral phase) was not shown. It is reasonable to conclude that such action 
occurred during the initial portion ofEPG Stage III. It is likely that the EPG Stage III 
also included the radiographic pharyngeal phase where orally the tongue remained in a 
fully elevated, anchored position while its posterior segment and root (again, not visible on 
EPG) expanded into the oropharynx for pharyngeal bolus transport. Furthermore, it is 
reasonable to infer that the "pharyngeal swallowing response" occurred during EPG Stage 
III . Finally, it is likely that the release of posterior linguopharyngeal contact to reopen the 
respiratory passage, again not visible on EPG, occurred near the end of Stage III while 
137 
oral full contact was still intact. Typically, radiographic staging of deglutitive activities in 
the upper alimentary tract does not consider the activity of the tongue immediately after a 
swallow, as the emphasis is often on the ensuing esophageal responses. However, Wein et 
al. (I 991) described a recovery phase based on ultrasound imaging, in which the tongue 
returned to the resting position. The EPG Stage IV seems comparable to at least part of 
this ultrasonographic recovery phase (up to the point where the tongue has completely 
retracted from the hard palate). As stated previously, the EPG stages are not identical to 
the conventional swallowing phases, because EPG provides information only on the 
interactions between the tongue and the hard palate. Any lingual movement that might 
occur before or after tongue-palate contact, while visible on radiography and 
ultrasonography, is not seen on EPG. 
Effect of Bolus Size and Consistency on Contact Timing 
With respect to tongue-palate contact timing, the EPG evidence showed that 
thinner and larger boluses were propelled significantly faster through Stage II. The slower 
propulsion rate for gelatin than for water supported previous findings that viscosity 
affected deglutitive oral motor behaviors by lengthening the transit time (see Dantas et al. , 
1990, Lazarus et al. , 1993, and Robbins et al., 1992). The faster propulsion rate for 30-cc 
boluses supported the findings ofKahrilas et al. (1993), suggesting that more forceful 
tongue movements were used to complete the all-at-once oral transport of a large 
quantity. Since the 5-cc gelatin bolus used in the present study was both smaller and 
thicker, one could understand why it had the longest propulsion duration in all subjects 
138 
tested. In general, the finding of significant effects of bolus properties on contact timing 
indicated that EPG was a technique sensitive enough to unveil evidence for peripheral 
sensory and biomechanical modulation of the temporal aspect of swallowing movements. 
Another timing effect occurred for dry swallows. In Stage III, dry swallows were 
significantly longer than water swallows. A possible explanation is that the saliva bolus, 
being small, caused small oral pressure gradients and perhaps a small pharyngeal 
gravitational effect. Consequently, a longer time would be needed to generate sufficient 
impetus for pharyngeal transport of the bolus, lengthening Stage III. 
Continuous Drinking 
Sequential versus Single Swallows 
The second question asked by the present study was whether continuous drinking 
would be performed by simply concatenating single swallows. The evidence revealed a 
shared characteristic as well as differences between the discrete and sequential swallowing 
gestures. The basic movement characteristic shared by both swallowing tasks was the 
sequential upward displacement of the four tongue segments from front to back and lateral 
to midline, as the means of backward bolus propulsion. Reflecting a fundamental 
relationship between the tongue and the hard palate, these movement components served 
as the basis for all swallowing tasks, regardless of their properties and demands. 
In two areas, however, the demands of the sequential swallowing task induced 
lingual reactions quite different from those seen for the discrete swallows: increased speed 
of tongue movement, and overlap of gestural components. Specifically, in contact timing, 
,, 
139 
sequential swallows were significantly shorter in total duration than single water swallows, 
indicating a much faster pace oflingual motions. In contact pattern, the majority of the 
sequential swallows were characterized by an overlap of activity between front contact 
retraction and back contact activation with the omission of full contact. On ultrasound, 
there was evidence for the co-occurrence of both tongue tip lowering and tongue body 
elevation, and the reverse, tongue tip elevation and tongue body lowering at different 
times during continuous drinking. This pattern oflingual action and linguopalatal 
interaction was clearly different from that observed for single swallows where a period of 
tongue inactivity consistently followed propulsion and preceded withdrawal. Thus, the 
answer to the second question was that continuous drinking was not carried out by simply 
repeating full and absolute cycles of single swallows (i.e., by successively completing each 
component swallow with preservation of all four stages, as that seen during the discrete 
swallowing task). 
Motor Control Strategies 
The present study proposes that the fundamental motor strategies used for discrete 
swallowing remained unchanged for sequential swallowing. To elaborate on this point, it 
is necessary to first recount the motor elements involved in each behavior. 
Sequential swallowing occurred as part of continuous drinking which contained 
reciprocal water-intake ("drink") and water-clearing ("swallow") cycles. As part of the 
drink-swallow cycles, there were reduplicated opening and closing of the oral tract by 
action of the tongue. The opening event, seen on EPG as contact withdrawal and on 
140 
ultrasound as lingual retraction, reflected the "drink" portion of the cycle, while the 
closing event, seen on both EPG and ultrasound as propulsion, reflected the "swallow" 
portion. Each of these events consisted of a sequence of gestures that involved 
progressive, front-to-back activation of the four tongue segments: front-to-back lowering 
in one case, and front-to-back elevation in the other. 
It has been well established that swallowing is a complex motor process involving 
both voluntary and reflexive physiological components. In this process, responses from 
the oral cavity, pharynx, larynx, and esophagus are integrated to achieve successful 
preparation and transport of food and protection of the airway. It has also been well 
established that the pharyngeal response is governed by the brainstem central pattern 
generator, while the interaction of tongue, jaw and facial movements for food preparation 
and oral transport is under cortical control (Miller, 1982, 1993). The lowering of the 
tongue to receive the bolus is an obvious food-intake behavior. It is uncertain if this 
behavior should be considered as part of food preparation or as a pre-preparation event. 
It is certain, however, that the behavior is voluntary, and that cortical intervention is in 
effect. The elevation of the tongue to propel the bolus is a food transport behavior. This 
behavior, having oral and oropharyngeal components, is both voluntary and reflexive--
voluntary up to the point when the bolus head passes the anterior faucial pillar, and 
reflexive thereafter ( cf. Logemann, 1983). Because swallowing is a continuum of 
integrated motor events, it is difficult to separate one event from the other or to set 
specific event boundaries, especially those occurring in the oral cavity. However, the 
complexity of interaction between different oral motor sequences during continuous 
drinking does suggest that more than one motor strategy is at work. 
141 
The protocol for the discrete swallowing experiment of this study required each 
subject to hold the bolus in the mouth for a short period before being prompted by trigger 
signals to commence task-related oral maneuvers. This requirement, a means of 
experimental control, is certainly not unique to the present investigation. An activity that 
was not considered as part of the task but took place with each discrete swallow in every 
subject was the presentation of the bolus in a syringe by the experimenter. As seen 
consistently on ultrasound, occurring with each bolus presentation was a natural lowering 
of the tongue to receive the material (see Appendix C for illustration of this behavior). 
This lowering gesture without exception followed a front-to-back progression but stopped 
short of the posterior tongue segment, reflecting the mechanism at work for prevention of 
premature spillage and aspiration. The suggestion, thus, is that had the artificial "hold" 
period been removed, or had the subject been asked to self-present the bolus and to 
swallow immediately afterwards (as in the sequential swallowing experiment), the motor 
events seen would probably have consisted of an opening gesture for bolus intake 
followed by a closing gesture for bolus propulsion. 
Conceptually, therefore, the opening and closing motor sequences are inherent 
parts of both continuous drinking and discrete swallowing, but reciprocation of the 
behaviors is characteristic of one and not the other. To efficiently and safely carry out the 
task requirement for reciprocation of motor events, the timing coordination of these 
events was modified. Since the task demanded "sequential swallowing without any 
I 
I 
I 
I 
I 
I 
I 
I 
betw 
een-swa11 142 ow Paus "· 
e, increased t fl" al Option ra e O mgu movement was a necessity and not an 
. As an effc 
ect of the rate i 
overla . ncrease, component gestures of the two motor sequences PPed m ti 
me and with h . . . 
boJus . armony, resultmg m rhythmical, stereotypic reduplication of Intake 
' transport d 
the in:i an swallow. The harmonious merging of the two sequences gave 
Pression th 
at one co f 
swallow n inuous flow of lingual actions took place for each drink-
cycle and fi 
rom cycle to cycle. 
Given th 
e abov · 
drjhl.! e Interpretations, one would then conclude that discrete water 
'AJ\lng w· 
Ithout artifi . 
cycle . cial bolus-holding would be composed of a single drink-swallow 
Inwh· h ic lingual " . 
One w opening" and "closing" events occurred successively. In addition, 0Uld 
Propose th . 
a Strj at continuous water drinking would be perfonned by concatenating Ilg of d" 
Iscrete d 'nk . 
a iiJn . n s with temporal overlaps of the opening and closing sequences as 
Chon of rat . 
ditte e Increase. Recall that for speech Browman and Goldstein (1990) used 
rent a . ' 
rt1culato . 
for th ry tiers to represent linguistic gestures, and proposed two separate ones 
e ton 
8Ue--tip a d b . . . . 
OverJ n ody ( see review in Chapter I). Under a surular prerruse, smce the 
aps dufin 
t g continua d . . . d h 
on8U us nnking were most apparent between the tongue tip an t e 
e bod 
. Y, oral Jin I . . 1nt0 gua gestures for reciprocal drink-swallow cycles might be orgaruzed 
at least th 
ese tw t· 
WhjJ O iers. The model of Browman and Goldstein (1990) proposes that 
e the i . 
' nd1viduaJ 
'ove 1 speech gestures are discrete and invariant across contexts, the rapp· 
Ing activatio f · h " 1 · Conte n ° several invariant gestures" during contmuous speec . resu ts m 
"1-va · l'Ymg art· 8howe iculatory trajectories" (p. 342). Evidence from the present study 
d that h 
t e coo d' · · t 
r mated co-occurrences of two complex yet innately mvanant mo or 
I 
I I 
sequences during continuous drinking also resulted in task-varying lingual movement 
trajectories. 
143 
Lingual gestural rate increases and overlaps brought about the omission or drastic 
shortening of the full contact stage during continuous drinking, indicating that the motor 
correlate of a rapid rate was the "dropping or fusing of movement elements" (cf Stetson, 
1951 ). In single swallows, the consistency of full contact across all subjects suggested 
that a period of stable anchoring of the oral tongue against the palate served a distinct 
function . That function may well have been to maintain oral stabilization so as to attain 
maximal posterior lingual expansion and lingual driving force for pharyngeal bolus 
transport. If this interpretation is accurate, the absence or meager presence of full contact 
in sequential swallowing raises several difficult questions. Is full contact not an innate, 
necessary stage of swallowing, and is oral tongue anchoring replaced by other temporal 
modifications, movement components, and/or biomechanical factors in the aerodigestive 
tract as a function of increased rate? To what extent does "momentum" contribute to 
bolus movement during sequential swallowing, and how is that related to the concurrent 
motor events in the pharynx, larynx, and cervical esophagus? Does linguopharyngeal 
contact pressure differ, for example, in onset/offset timing and magnitude between discrete 
and sequential swallowing? All of these questions remain unanswered. Future 
investigations using combined videofluorography and manometry are likely to offer 
insights. The present study does, however, support the hypothesis that full contact is not 
a dispensable stage in normal discrete swallowing, although it is drastically reduced in 
continuous drinking due to rapid movement rate. An analogy may be found in the 
144 
compression of vowels with rate increase during speech. It is well known that when 
speaking rate is increased, vowels become shortened to a larger extent than consonants. 
Similarly, in sequential swallowing and perhaps even in a fast discrete swallow, Stage III 
has considerably greater latitude for compression than, for example, Stage II. 
Interestingly, despite the considerable increase in lingual movement pace, there 
were no overt indications of the speed-accuracy trade-off that is common for aimed 
movements ( cf. Meyer et al. , 1982 and Newell, 1980). More specifically, no reduction in 
swallowing safety or efficiency (i.e., no choking or spillage ofliquid) was observed as a 
function of rate increase. There are at least three explanations that are not mutually 
exclusive: 
The first explanation lies in the velocity range of deglutitive tongue motion. As 
Meyer et al. ( 1982) put it, "Faster movements can usually occur only at the expense of 
being spatially less accurate on the average, and movements having greater spatial 
accuracy can usually occur only at the expense of being slower on the average" (p. 449). 
The tongue' s action during sequential swallowing never achieved very high velocities even 
on the fast continuous drinking task. Presumably this was necessary in order to maintain a 
high level of movement accuracy, and the reason for high accuracy is the need to protect 
the airway. Although the overlap of gestures allowed greater velocity for sequential than 
for single swallows, there was an apparent limit to the extent of allowable maximal 
velocity for deglutitive tongue movement due to the range of motion involved and the 
issue of airway safety. This does not imply, however, that a threshold for the speed-
145 
accuracy trade-off does not exist in this case, nor does it imply that lingual action during 
swallowing has no finer rate-adjustment potential. 
A second explanation lies in the mechanical properties of the tongue muscles. 
Although the present research did not directly examine lingual muscle mechanics, elements 
such as force, elasticity, and viscosity, as well as other biomechanical factors ( e.g., inertia, 
force inherent in the bolus) undoubtedly were at play and contributed to the task responses 
observed. The factor of muscle contraction force is especially relevant here. It has been 
shown via electrical stimulation to motor nerves that all muscles have an upper limit of 
force increase in response to increasing rate of stimulation, that muscle twitch force curves 
fuse at high stimulation rates, and that marked reduction in the frequency-response curve 
occurs as a result (above 3-4 Hz in speech muscles) (McClean, 1988). Functionally, in 
fact, at very high movement rates (e.g., at maximal diadochokinetic rates) "there is 
invariably a reduction in the structural displacement" as a "result of the mechanical 
properties of muscle" (McClean, 1988, p. 32). Thus, from the perspective of lingual 
muscle mechanics, to preserve adequate displacement of the tongue for effective 
sequential swallowing, lingual movement rate cannot exceed a certain level. Exactly what 
that "certain level" is remains unknown. Also undetermined are: the appropriate amount 
of force needed, the activation levels and relative timing of different tongue muscles, and 
how their intercoordination affects each other's force output as well as the overall patterns 
of lingual displacement. Future studies are needed to directly address these important 
biomechanical issues. It is understood, however, that this type of investigative task is 
challenging because of the complex decussations of lingual muscles and the constant 
-
interaction of all lingual muscles in any movement involving the tongue (Smith & Kier, 
1989). Isolating specific muscles within the tongue for electromyographic or 
biomechanical recording is conceivably difficult. 
146 
A third explanation lies in the probable changes that are occurring elsewhere in the 
aerodigestive tract during conti~uous drinking. With rapid movements of the tongue and 
the successive arrival of water boluses in the pharynx, there are likely to be associated 
adjustments in the movement timing and patterning of laryngeal/pharyngeal structures and 
in their temporal and spatial coordinations with the tongue. The velopharyngeal sphincter, 
the hyoid bone, the epiglottis, the larynx as a whole, and the upper esophageal sphincter 
(USE) are examples of these structures. To execute a complete single swallow, each of 
them engages in "on-and-off' movements. The larynx, for example, elevates as a whole 
against the base of the tongue during the swallow ("on") and lowers to its normal position 
afteiwards ("off'). The upper esophageal sphincter, as another example, relaxes to allow 
bolus passage during the "on" period and recontracts during the "off' period. To preserve 
aiiway safety when liquid arrives rapidly and continuously, these laryngeal/pharyngeal 
structures must simplify their internal as well as external coordinations. Because the "on" 
activities are crucial, one way of accomplishing such simplification is perhaps to adjust the 
range of movement for the "off' period. For example, the larynx may lower but not to the 
extent it does in a single swallow before it re-elevates for the subsequent swallow, and the 
upper esophageal sphincter may not recontract fully or by much until the end of the last 
swallow. On ultrasound, the hyoid shadow exhibited reduced range ofreciprocal forward-
backward displacement from swallow to swallow and did not completely return to its 
147 
resting position until the end of drinking. Since movement efficiency is an important 
consideration of dynamic motor systems, the "partial off' seems a reasonable strategy for 
attaining safe and cost-effective (as in energy and time) movement execution. This theory 
undoubtedly requires verification. Future continuous-swallowing investigations that 
combine EPG with videofluoroscopy or with nasendoscopy should unveil supporting or 
negating evidence. 
Methodological Consideration 
As a first attempt in examining tongue-palate contact for sequential swallowing 
during continuous drinking, this study did not apply stringent controls over the rate 
variable and focused more on the spontaneous management strategies used by the 
individual subjects. The finding of no significant rate difference between "normal" and 
"fast" drinking appeared to reflect this lack of tight control, at least to some extent. When 
left to their discretion, some subjects ( e.g., S 1) used practically the same rate for normal 
and fast drinking while perceiving themselves as using differential speed (based on the 
subjects' statements during post-experiment interviews). A second explanation for the 
similarities across the two rates is that because choking is unpleasant, subjects might have 
a strong disinclination to very fast drinking. Chugging which some but not all individuals 
can perform might be one of the few successful ways to overcome this. Nevertheless, for 
optimal experimental control, future investigations in this area may consider a protocol 
that induces sequential swallowing via mechanical oral infusion of liquid at graded rates 
(e.g., 40, 60, 80, and 100 cc/min; cf Issa & Porostocky, 1994). The use ofan infusion 
148 
pump to systematically control the rate and amount of liquid presentation will increase 
experimental objectivity and permit more accurate calculation of per-swallow consumption 
and more precise data analysis. 
Ultrasound as a Supplementary Technique for EPG 
As stated earlier, the major weakness ofEPG is that it provides information on 
only tongue-hard palate contact. As a compensation for such weakness, the present study 
used ultrasound imaging to gain an avenue for data verification and supplementation. 
Although direct, feature-by-feature matching of the EPG and ultrasound findings was 
impossible due to instrumental differences, gross comparisons could be made in timing as 
well as between tongue shape changes and contact activities. The most important benefit 
from combining EPG and ultrasound was that the two techniques provided entirely 
different perspectives, and together they made the observations of deglutitive tongue 
behaviors more complete. Specifically, EPG supplied lateral and medial contact 
information that was absent on ultrasound, and ultrasound showed posterior tongue action 
as well as pre-contact and post-contact lingual movements that were not seen on EPG. In 
addition, the two sets of data when combined offered insights that would not have been 
found in one alone : 
First, viewing the contact patterns in light of the visualized tongue actions and vice 
versa, one gained a deeper appreciation for how the tongue took advantage of the 
immobile palate during swallowing. Systematic, front-to-back anchoring against the hard 
palate was the apparent and only way through which propulsion could be accomplished. 
149 
As Stone (1995) put it, "The immobility of the hard palate can be a real advantage to the 
tongue because the tongue does not have to deal with palatal variability in the motor 
planning for their contact" (p. 143 ). The rather straight-forward linguopalatal interaction 
could also be an advantage to the overall motor planning for swallowing, considering the 
multiplicity of other neuromuscular coordinations (e.g., linguopharyngeal, linguolaryngeal, 
laryngopharyngeal, etc.) that must occur with precision to maximally ensure the safety of 
the airway. 
Second, tongue shape changes during swallowing, which presumably were the 
results of linguopalatal contact, were noticeably limited in variety. Although some signs of 
bolus-induced tongue shape differences were observed, they tended to be more prominent 
during pre-swallow bolus holding or loading. Thus, unlike speech where fine differences 
in tongue shape are relied on for differential phoneme production, the principal task of oral 
manipulation during swallowing is effective propulsion of the bolus. Shape alterations 
during swallowing, in other words, are secondary results and not part of the primary goal. 
Third, although the tongue possesses multiple potential degrees of freedom in 
movement as well as many spatial options of contact with the palate, only a few are 
required to fulfill the requirements for swallowing. Sequential elevation and bunching in 
conformity with the shape of the palate, lateral arching, lowering, and horizontal/vertical 
expansion and compression appear to be the basic lingual gestures needed. Twisting, 
curling, and other complex local deformations which conceivably are needed in speech and 
in mastication are not necessary for swallowing. This suggests that the recruitment of 
lingual 
rnuscJes . . 1 so 
. , mtnnsic as well . . . . 
In the Oth as extnnsic, is relatively less complex in swallowing than 
ertw . 
o activities. 
Ultrasound . . 
unagmg . 
lllove,,.., provided an additional opportunity to examine recnonal Jino11al 
.. ient ki o· o-
nernatics throu h . 
Profile g the tracking of instantaneous velocities. The velocity 
Pef111itted 
not only th . 
colllpa . e observation of local velocity changes over time but also the flson of 
Peak-veloc 't d' 
the Ve] . I Y ifferences as a function of swallowing task. The shape of 
OCJty r 
p oflles, es e . . . . . 
Provided p cialJy the locations of peak velocit1es along the time continuum, 
at least thr . 
seque ee pieces of information. First, it revealed the invariance of the 
nee of . 
rnaxunaJ d' 
This Isplacement of the four tongue segments across swallowing tasks. 
sup 
Ported th h 
front t e t eory that serial anchoring of each segment against the palate from 
o back is th b . 
Peak e asic motor scheme for bolus propulsion. Second, prominent negative 
S Of the 
dorsal and . . 
Corres postenor segments were consistently shown. Thelf occurrences 
Pondedin . 
lowe . tirne to the EPG Stage II (see Figure 3. 16), reflecting "segmental" 
1lno 
o Of the t 
Peak ongue to allow bolus passage. How the magnitude of these negative 
S Va· fles as a lime . . . . 
add· . tion of bolus properties warrants further mvestigat10n and may offer 
it1ona1 in . 
sight into J' . . . 
Third mgual strategies for bolus accommodation dunng swallowing. 
, for ea h 
Which c tongue segment, detailed information is available on the velocity patterns 
lllay be quant. . . 
kin Itatively analyzed for more extensive examination of task-related, 
elllatic 
Propertie f ·1 bl llol'll]aJ s O deglutitive lingual action. When more data become avai a e on 
Patterns th . d 
describ. ' e velocity profile potentially can provide a basis for companng an 
Ing abnorm 1 J' 
a mgual movements during swallowing. 
151 
Toward a Model ofDeglutitive Tongue Action 
Discrete Swallowing 
In Stone's proposal (1990, 1991) for a three-dimensional model oflingual 
movement during speech, the tongue was divided into functional segments--five sagittal 
(anterior, middle, dorsal, posterior, root) and five coronal (medial, lateral 1, lateral 2). In 
this "functional segment" model, the lengthwise segments operated semi-independently to 
produce local displacement as well as front-to-back rotational effects for articulatory 
purposes. The crosswise segments carried out concerted or opposing movements to effect 
central grooving and left-right asymmetries. Palatal contact or bracing was optional, 
because certain speech sounds are produced without it. However, the tongue does rely 
heavily on contact with the palate to effect shape and positional changes for the majority 
of speech gestures. 
The EPG and sagittal ultrasound data of the present investigation suggest 
considerable applicability of this model for describing tongue movements during discrete 
swallowing. Three conceptual modifications, however, are necessary: 
First, tongue-palate contact is fundamental to swallow-related lingual action under 
normal conditions. Therefore, there are no "unbraced" gestures. The principal goal of 
such contact, as stated earlier, is to achieve positional changes for bolus propulsion and 
not necessarily to effect alterations in tongue shape. 
Second, the lengthwise lingual segments operate as concatenated, functionally 
dependent units during discrete swallowing. In the front-to-back propulsive sweep the 
maximal displacement of each successive unit, though appearing to occur locally, is 
152 
chained to that of the previous unit (cf. Martin, 1991). In addition, the positional stability 
of one unit is a prerequisite for the attainment of maximal displacement of the ensuing 
unit, regardless of the swallowing task at hand. This prerequisite anchoring is due to the 
fact that the tongue has no skeletal support within itself. It uses sequential, sectional 
bracing against the hard palate from front to back to generate and transmit bolus 
propulsive force, and redistribute its own volume from front to back, thereby 
accomplishing the task of food transport . Serial anchoring of the sagittal units, therefore, 
is not only the source of the bolus-transport force needed but also the origin of the 
stereotypic wave-like contours. 
Third, each lengthwise tongue segment is composed of five cross-sectional 
segments which also operate as concatenated, functionally dependent units to contribute 
to the task of swallowing. During propulsion, the crosswise units are responsible for 
symmetrical elevation of the tongue from the sides toward the midline (as if narrowing a 
central channel) to assist in transporting the bolus backward. 
In sum, the functional segment model, in order to be applied to swallowing, must 
be modified to include anchored rather than unanchored gestures. In addition, the model 
must define the tongue units as functionally dependent and concatenated in both the 
lengthwise and the crosswise domains. 
An interesting question that has not been resolved by any current swallowing data 
is how the volume-preserving constraint interacts with the anchoring constraint. The 
anchoring constraint is a given in swallowing, because the tongue must take advantage of 
the palate to propel the bolus. The volume-preserving constraint, originally pointed out by 
153 
I<ier and s · hift d fr lace to place but not 
nuth ( 1985), says that tongue volume can be s e om P 
Jncreas d · ft ngue shapes during 
e or decreased. Recent three-dimensional reconstructwns O 0 
speech confirmed the presence of local displacement-compression tradeoffs on the tongue 
surface (Stone & Lundberg, 1996). For example, during production of the vowel Iii, 
upward displacement of the anterior tongue unit was found associated with inward 
compression of the posterior unit. The authors hypothesized that this tradeoffwas the 
basic me h · · f 1 T d' · al c arusm for tongue positioning in the product10n o vowe s. wo- imension 
radiogra hi d . 11 ·ng P c or ultrasonographic depiction of tongue movements unng swa owi 
cannot provide evidence for such a tradeoff. Granted that the tongue operates differently 
during s all . . . . . . d t b 
w owmg than dunng speech, still the d1fferent1al operations are came ou Y 
the same 1 h h vo ume-preserving muscular hydrostat. At the height of a swallow w en t e 
Posten · · 
or tongue expands to contact the pharyngeal waUs, the rest of the tongue remams m 
firm co t 
n act with the palate. Where is the complementary, local displacement-
compres · 'bl sion tradeoff? How, then, does the tongue preserve its volume? One poss1 e 
explanation . th h . h. . d . . h t 1s at t e tongue 1s t mned, allowing for backwar extens10n wit ou 
shortening or compressing the front. However, based on clinical observation of modified 
b . 
anum swallows, it appears that the posterior tongue reaches its maximal backward 
position in I I . . . h h . . . . fc d c ose tempora proxmuty wit t e hyo1d reaching its maximal upward/ orwar 
Position Th d h . 
· erefore, the upwar yo1d movement, ostensibly compressing the tongue and 
Jaw muse! . . I . d" 1 . es, 1s associated, at east m Irect y, with the backward expansion of the tongue. 
In °ther Words, the hyoid bone compresses not only the jaw opening muscles (mylohyoid, 
154 
geniohyoid, digastric) but also the posterior tongue, allowing it to push outward into the 
pharynx. 
Continuous Drinking 
The model presented above for discrete swallowing is also applicable to 
continuous drinking. For both water intake and clearing during each drink-swallow cycle, 
the tongue uses invariant, progressive front-to-back activation of its oral segments to open 
and close the oral tract and accomplish the motor goals specified. In the sequence of 
closing gestures, these segments continue to function as concatenated, dependent units in 
both the lengthwise and the crosswise domains. In addition, anchored rather than 
unanchored gestures remain rudimentary to the process. In the sequence of opening 
gestures, the linkage and dependency among the tongue segments are again apparent, as at 
no time can the tongue lower itself to replenish water in an order other than front-to-back. 
One may add that this stereotypic front-to-back opening is tightly coupled with and a 
reflection of jaw action, granted that the jaw is constrained in its degree of motion during 
cup-drinking. 
When gestures of the same structure overlap in time, the indication is usually that 
parts of that structure are functionally independent. The temporal overlaps between the 
tongue tip and the tongue body during continuous drinking could, therefore, be interpreted 
that segments of the tongue performed independent gestures across motor sequences 
while simultaneously maintaining within-sequence gestural dependency. While this 
interpretation is supported by multiple analogies in speech (see Browman & Goldstein, 
1990) 
,the highd 155 
egree of inter-s . . 
cannot be equence behavioral correlation during continuous drinking 
overlooked . 
down'' · Specifically, the invariant "tip down, body up" and "tip up, body 
Patterns of r 
p mgual overlaps fl d d' . fu . . . 
rernatu re ecte 1stmct nct1onal s1gmficance: one to prevent 
re bolu . 
s spillage the 
Or opp . . ' other to make room for propulsion. Because these overlaps 
os1tiona1 
movem 
lllotor s ents Were readily predictable, the indication would be that the two 
equences 
' Perfectly · Other ,1 , integrated, were still not totally free from the influence of each 
. vvhat is 
apparent th . 
Organ. s· ' ough, 1s the remarkable flexibility of the tongue as a soft-tissue 
. Ince multiple m 
111 rnult· uscles are in or connected to the tongue, it can theoretically move 
Iple d' 
' 
1fferent 
anoth ways. Slightly more contraction of a set of lingual muscles versus 
er ca 
n result. 
m subtle variations of behaviors. 
Evid 
ence ofth 
differ e Present study supports the notion that swallowing and speech are 
ent beh . 
aVJors ofth 
Colllplex e vocal tract ( cf Martin, 1991 ). Although both are motorically 
' swa11oWin . . 
and s g is highly stereotypic with reflexive as well as voluntary components, 
Peech is i. ~ 
iughJy va . bl . h lllodel na e and completely voluntary. Thus, when applying speec 
s~~ph Ysio/0 · 
to Pr gic or linguistic--to deglutitive behaviors, revisions are necessary in order 
Operly r 
epresent th . 
e characteristics of the motor components being addressed. 
J) Clinical Implications 
Ysphagia m . . . al 
lbn0 ay be associated with a variety ofneurolog1c disorders and structur. l'llla1· · 1hes ( seet 
1
-Wer ogemann, 1983). In adults and children with upper motoneuron, 
llloton 
euron b l · ho have 
'dero ' asa ganglia, and cerebellar dysfunctions, as well as patients w 
oOne 
surgical· t 
intervention for head and neck cancer, deficits in tongue movemen 
156 
and coordination are common and may even be one of the primary causes for dysphagia 
symptoms (Hirano, et al. , 1992; LaBlance, Kraus, & Steckol, 1991 ; Lazarus & Logemann, 
1987; Logemann, 1983; Veis & Logemann, 1985). A general reduction in the range of 
lingual motion and control or a specific impairment such as repetitive tongue 
rolling/pumping (as seen in Parkinson's disease) can lead to delayed oral transit, 
ineffective bolus propulsion, oral and pharyngeal residue, and even premature spillage of 
the bolus which threatens airway safety (Cook, 1991; Dodds, Logemann, & Stewart, 
1990; Logemann, 1983). 
As shown in the present study, stable anchoring of the tongue against the palate is 
crucial for overall swallowing efficiency, because it provides the necessary leverage for 
generating the appropriate oral propulsive force and for achieving maximal base of tongue 
expansion and adequate pharyngeal propulsive force. Thus, results of this investigation 
support not only the conventional linguomotor exercises (e.g., protrusion, retraction, 
lateralization, twist, curl, etc.) but also the practice of deliberately prolonged (for 4-5 
seconds at a time), hard, squeezing lingual action from front to back in association with 
dry swallows to facilitate forceful tongue-palate contact and tongue base retraction. The 
recommendation for using this maneuver with dry swallows is a safety precaution. 
Patients who have adequate laryngeal control or cough reflex may not need to be as 
conservative and may advance from dry to wet swallows using the "lingual hard-squeeze" 
maneuver after a period of systematic training. Moreover, in head and neck cancer 
patients who present abnormal oral and laryngeal signs but benefit from the use of 
supraglottic or Mendelsohn maneuvers (i.e., breath-holding following by swallowing and 
then a deliberate cough, or maximal laryngeal elevation using the suprahyoid muscles in 
the midst of a wallow; see Lazarus, Logemann, & Gibbons, 1993), lingual hard-squeeze 
can be easily incorporated into these exercises for maximal results. 
157 
The modified barium swallow (Logemann, 1983) is probably one of the most 
widely applied protocols for swallowing evaluation. This approach emphasizes the use of 
small bolus size (2 cc of barium liquid and paste, and l/4 of a cookie) as a conservative 
precaution against damaging consequences in case of aspiration. While such rationale is 
appreciated and holds true in many patients, clinicians should not ignore the possibility 
that some patients swallow larger boluses more efficiently than they do smaller ones. The 
radiographic study of Robbins, Sufit, Rosenbek, Levine; and Hyland (1987) showed that 
three of the 64 patients they examined aspirated on only 2-cc but not 30-cc barium 
swallows. These authors speculated that this finding could reflect different muscle 
recruitment patterns for the oral phase of swallowing and the possibly altered threshold for 
the pharyngeal swallowing response. Based on the results of the present study and of 
Kahrilas et al. (I 993), one may add that biomechanically, the benefit of the larger bolus 
could have also been drawn from faster propulsion rate and greater movement vigor. The 
important message is that not all patients will find smaller boluses easier to manage than 
the larger ones. In patients without evidence of aspiration from the modified barium 
swallow, additional radiographic testing using larger boluses may reveal more clues on the 
nature of the breakdown in swallowing efficiency and on clinical management strategies. 
Conceivably, some neurologic patients (e.g., some of those with cerebrovascu1ar accident 
or closed head injury) presenting mild, generalized difficulty in motor coordination may 
158 
show better swallowing performance with larger-and-thicker materials, taking advantage 
of the timing and force factors from bolus size and the increased cohesiveness from bolus 
consistency. 
Continuous drinking, compared with discrete swallowing, is a task that imposes 
even greater demands for motor coordination, because it involves alternating action 
sequencing, gestural overlaps, increased overall movement rate, and a prolonged, 
heightened state of activity in all deglutitive motor participants. These increased demands 
may be further exemplified by the fact that many motor behaviors developmentally follow 
a progression from a discrete to a sequential version with learning and maturity. Indeed, 
children learn to utter single words before coarticulated speech, and to print single 
characters before cursive writing. Developmentally, swallowing is known to first emerge 
in the embryo, and its neural organization is well established by the time of birth (Bosma, 
1973; Humphrey, 1970). However, the motor coordination for cup-drinking and 
sequential swallowing on one breath requires time to develop and mature. 
Severe, generalized motor dysfunction (such as that occurring in the later stage of 
Parkinson' s disease) can cause a patient to lose the coordination necessary for performing 
sequential swallowing. Nonetheless, just as there are patients who manage larger-volume 
swallows better than smaller ones, some dysphagic patients in whom adequate laryngeal 
function is preserved are likely to find continuous drinking easier to achieve than discrete 
swallowing because of movement momentum or other undetermined factors. Clinical 
investigations are needed to evaluate such likelihood in patients presenting different 
diseases and dysphagia symptoms. 
159 
EPG appears to have potential applications in clinical research and management. 
One of the distinct advantages for clinical usage of EPG is that the technique is 
nonhazardous and relatively noninvasive. As such, it may be used repeatedly and for 
prolonged periods, thus permitting extensive observations oflinguopalatal interaction and 
training. As indicated above, abnormal tongue activities of different sorts occur during 
swallowing in several clinical populations. Deviations in tongue-palate contact timing and 
patterns from the normal data are expected in, for example, postsurgical oral cancer 
patients, patients with hypokinesia or hyperkinesia, and some patients after stroke or 
closed head trauma. However, the details of the abnormal contact behaviors have not 
been characterized for different disorders. EPG presents a readily available means for in-
depth quantitative and qualitative characterizations of abnormal linguopalatal interactions, 
and for identification of the specific components of deviant deglutitive tongue action. 
Such detail, typically not accessible via conventional radiography and ultrasonography, 
may provide clues for therapy planning. 
The clinical management procedures for postoperative oral cancer patients with 
dysphagia sometimes include the fitting of a maxillary prosthesis. The purpose is to 
reshape or augment the palate so that the patient achieves better swallowing efficiency 
through more appropriate tongue-palate contact. The effectiveness of such a device has 
been documented (Davis, Lazarus, Logemann, & Hurst, 1987; Logemann, Kahrilas, 
Hurst, Davis, & Krugler, 1989; Robbins, Bowman, & Jacob, 1987; Wheeler, Logemann, 
& Rosen, 1980). In some of the patients the benefits of the prosthesis are immediate. In 
others, however, a period of training is required for adaptation and improvement to take 
160 
place. For these patients, EPG can be conveniently incorporated as a biofeedback tool to 
facilitate the adaptation-improvement process. EPG's applicability in swallowing therapy 
is not limited to the oral cancer patient population. Any patients who present reduced 
lingual control for deglutition may benefit from the visualization of their tongue-palate 
contact and work on increasing lingual control or finding appropriate compensatory 
maneuvers with direct feedback. 
Conclusions 
EPG provided detailed information on linguopalatal contact during swallowing and 
revealed how the tongue used the hard palate to impel the bolus. It also provided inferred 
information on deglutitive tongue movements which were then confirmed using 
ultrasound. Compartmentalization of the palate into six primary bins (front, central, back, 
lateral, medial, midline) extracted deglutitive linguopalatal contact characteristics that 
were regionally specific, three-dimensional, and typically not revealed through 
conventional radiography and ultrasonography. Specifically, tongue-palate contact for 
swallowing, though invariant across bolus types and subjects, was seen to have two 
degrees of freedom: front/back movement and lateral/midline movement. These EPG 
continua reflected a lingual motor system composed of concatenated, functionally 
dependent units in the lengthwise as well as crosswise domains, that produced 
stereotypical tongue movements. 
The discrete swallowing task of the present study was defined as having four 
tongue-palate contact stages-- prepropulsion, propulsion, full contact, and withdrawal--
161 
based on specific, time-varying contact characteristics. The division of the EPG time 
series into these stages was functionally meaningful, as it captured the distinct and unique 
features of each stage, the variations in patterns within a stage, and the relationship 
between patterns across stages. The timing aspect of discrete swallowing can be 
modulated by the sensory features inherent in a bolus, despite the stereotypy in its overall 
motor pattern. In particular, propulsion rate was significantly faster for thinner and larger 
boluses, and full contact was significantly longest for dry swallows. These temporal 
effects indicated that the swallowing central pattern generator was at least partially 
regulated by peripheral input . 
The present study represents an initial attempt to better understand lingual 
behaviors for sequential swallowing. In continuous drinking where there were 
reduplicated drink-swallow sequences, the tongue was seen to use faster movement rate 
and overlapping gestures to maintain safe and efficient swallowing. Gestural overlaps 
were most apparent in the anterior and the posterior tongue, and reflected a change in the 
timing coordination of the motor sequences for "drink" and "swallow" implemented to 
meet changed task requirements . Further research on the responses of other motor 
elements and on oral, laryngeal, and pharyngeal coordinations during continuous drinking 
is warranted and likely to improve our knowledge of swallowing motor control in general. 
Ultrasound provided supplementary data on lingual contour and velocity changes 
over time. The single-subject velocity profiles supported the EPG findings and presented 
additional evidence for bolus- and task-induced differences in peak velocities. The 
velocity profile is a useful tool for the analysis oflingual kinematic properties during 
swallowing. 
162 
There is potential value in using EPG to investigate disordered swallowing, 
particularly the timing and patterning of abnormal tongue movements. In addition to 
possible future applications in swallowing research, EPG might be used for diagnostic 
purposes as a supplement to the more traditional imaging and pressure measurements, and 
for treatment purposes as a biofeedback tool. 
In sum, evidence of the present study supported a theory that swallowing motor 
control included a peripheral mechanism capable of modulating centrally generated 
responses, and that the general deglutitive motor program consisted of both invariant and 
variant parameters. Specifically, movement pattern or action sequence reflected a fixed 
element within the structure of the motor program. Movement timing, on the other hand, 
was variant and could be modified according to the properties of the material ingested and 
the demands of the task at hand. 
APPENDIX A. ULTRASOUND TRANSDUCER PLACEMENT DAT A 
Specific placement of the ultrasound transducer for individual subjects during sagittal and coronal imaging is presented below. 
Angle refers to degree posterior to the vertical. Position refers to distance posterior to the mental symphysis. The transducer 
used was a 3 .5-MHz convex-curved-linear-array type with 96 active elements. 
Sagittal Placement Coronal Placement 
Subject Jaw Length (mm) Angle (0 ) Position (mm) % of Jaw Length Angle (0 ) Position (mm) % of Jaw Length 
Sl 83 .74 8 33.50 40.00% 14 27.63 32.99% 
S2 93 .74 8 31.40 33.50% 11 31.03 33.10% 
S3 82.11 8 25.20 30.69% 11 27.37 33.33% 
S4 87.54 6 29.50 33.70% 14 29.18 33.33% 
S5 87.42 8 33.18 37.95% 14 29.14 33.33% 
164 
APPENDIX B. STATISTICAL TABLES 
B-1 . Random effects analysis of variance on EPG total durations for the discrete 
swallowing tasks, using the general linear models procedure. 
General Linear Models Procedure 
Tests of Hypotheses for Mixed Model Analysis of Variance 
Source: TASK 
Error: 0.9993*MS(TASK*SUBJECT) + 0.0007*MS(Error) 
DF 
4 
Type III MS 
731560.43883 
Source: SUBJECT 
Denominator 
DF 
16.00 
Denominator 
MS 
234021. 98581 
Error: 0.9993*MS(TASK*SUBJECT) + 0.0007*MS(Error) 
Denominator Denominator 
DF Type III MS DF MS 
4 2073593 .8019 16.00 234021.98581 
Source: TASK*SUBJECT 
Error: MS(Error) 
Denominator Denominator 
DF Type III MS DF MS 
16 234145.22582 124 48669.021832 
F Value 
3.1260 
F Value 
8.8607 
F Value 
4.8110 
Pr > F 
0.0444 
Pr > F 
0.0006 
Pr > F 
0.0001 
B-2. Repeated measures analysis of variance on mean durations ofEPG stages for 
different discrete swallowing tasks, using the general linear models procedure. 
Multivariate tests were performed on the variables TASK and ST AGE, but not on 
T ASK*STAGE interaction due to insufficient error degrees of freedom. Instead, the 
interaction was assessed using univariate tests. 
Manova Test Criteria and Exact F Statistics for 
the Hypothesis of no TASK Effect 
H = Type III SS&CP Matrix for TASK E = Error SS&CP Matrix 
S=l M=l N=-0.5 
Statistic Value F NumDF DenDF 
Wilks' Lambda 0.00079757 313.202 4 1 
Pillai's Trace 0.99920243 313.202 4 1 
Hotelling-Lawley Trace 1252.80832 313.202 4 1 
Roy's Greatest Root 1252.80832 313.202 4 
Manova Test Criteria and Exact F Statistics for 
the Hypothesis ofno STAGE Effect 
1 
Pr>F 
0.0424 
0.0424 
0.0424 
0.0424 
H = Type III SS&CP Matrix for PHASE E = Error SS&CP Matrix 
S=l M=0.5 N=O 
Statistic Value F NumDF DenDF Pr > F 
Wilks' Lambda 0.14327308 3.98645 3 2 0.2070 
Pillai's Trace 0.85672692 3.98645 3 2 0.2070 
Hotelling-Lawley Trace 5.97967801 3.98645 3 2 0.2070 
Roy's Greatest Root 5.97967801 3.98645 3 2 0.2070 
Univariate Tests of Hypotheses for Within Subject Effects 
165 
Source 
TASK*STAGE 
Error 
DF 
19 
76 
Type III SS 
2317839.664 
1663900.096 
Mean Square F Value Pr > F G - G H - F 
121991.561 5.57 0.0001 0.0195 0.0004 
21893.422 
Greenhouse-Geisser Epsilon = 0. 13 15 
Huynh-Feldt Epsilon= 0.3682 
166 
B-3 . Post hoc repeated measures multivariate analysis of variance of pairwise contrasts 
for EPG Stage-II duration between selected discrete swallowing tasks. Only cases where 
12<0.05 are included below, although a for these comparisons was set at 0.01. 5W = 5cc 
water; 30W = 30cc water; 5G = 5cc gelatin; 30G = 30cc gelatin. 
Manova Test Criteria and Exact F Statistics for 
the Hypothesis of no Overall INTERCEPT Effect 
on the variables defined by the M Matrix Transformation 
H = Type III SS&CP Matrix for INTERCEPT E = Error SS&CP Matrix 
S=l M=-0.5 N=l 
Contrast variable: 30W - 30G 
Statistic 
Wilks' Lambda 
Pillai's Trace 
Hotelling-Lawley Trace 
Roy's Greatest Root 
Contrast variable: 5W - 5G 
Statistic 
Wilks' Lambda 
Pillai's Trace 
Hotelling-Lawley Trace 
Roy's Greatest Root 
Contrast variable: 30G - 5G 
Statistic 
Wilks' Lambda 
Pillai's Trace 
Hotelling-Lawley Trace 
Roy's Greatest Root 
Contrast variable: 3 OW - 5W 
Statistic 
Wilks' Lambda 
Pillai's Trace 
Hotelling-Lawley Trace 
Roy's Greatest Root 
Value 
0.03998767 
0.96001233 
24.0077058 
24.0077058 
Value 
0.13149271 
0.86850729 
6.60498432 
6.60498432 
Value 
0.21751734 
0.78248266 
3.59733454 
3.59733454 
Value 
0.32967607 
0.67032393 
2.03328048 
2.03328048 
F 
96.0308 
96.0308 
96.0308 
96.0308 
F 
26.4199 
26.4199 
26.4199 
26.4199 
F 
14.3893 
14.3893 
14.3893 
14.3893 
F 
8.13312 
8. 13312 
8.13312 
8. 13312 
NumDF 
1 
1 
1 
1 
DenDF 
4 
4 
4 
4 
NumDF DenDF 
1 4 
1 4 
1 4 
1 4 
NumDF DenDF 
1 4 
1 4 
1 4 
I 4 
Pr > F 
0.0006 
0.0006 
0.0006 
0.0006 
Pr>F 
0.0068 
0.0068 
0.0068 
0.0068 
Pr>F 
0.0192 
0.0192 
0.0192 
0.0192 
Num DF Den DF Pr> F 
I 4 0.0463 
1 4 0.0463 
I 4 0.0463 
I 4 0.0463 
167 
B-3 . (continued) 
Contrast variable: 5G - Dry 
Statistic Value F NumDF DenDF Pr > F 
Wilks' Lambda 0.03570459 108.03 1 4 0.0005 
Pillai's Trace 0.96429541 108.03 1 4 0.0005 
Hotelling-Lawley Trace 27.0076063 108.03 1 4 0.0005 
Roy's Greatest Root 27.0076063 108.03 1 4 0.0005 
Contrast variable: (mean of30W and 30G) - (mean of 5W and 5G) 
Statistic Value F NumDF DenDF Pr>F 
Wilks' Lambda 0.22283645 13.9504 1 4 0.0202 
Pillai's Trace 0.77716355 13 .9504 1 4 0.0202 
Hotelling-Lawley Trace 3.48759618 13.9504 1 4 0.0202 
Roy's Greatest Root 3.48759618 13 .9504 1 4 0.0202 
Contrast variable: (mean of 30W and 5W) - (mean of30G and 5G) 
Statistic Value F NumDF DenDF Pr>F 
Wilks' Lambda 0.03871082 99.3303 1 4 0.0006 
Pillai's Trace 0.96128918 99.3303 1 4 0.0006 
Hotelling-Lawley Trace 24.8325734 99.3303 1 4 0.0006 
Roy's Greatest Root 24.8325734 99.3303 1 4 0.0006 
168 
B-4. Post hoc repeated measures multivariate analysis of variance of pairwise contrasts 
for EPG Stage-III duration between selected discrete swallowing tasks. Only cases where 
12<0.05 are included below, although a for these comparisons was set at 0.01. SW= 5cc 
water; 30W = 30cc water; 5G = 5cc gelatin; 30G = 30cc gelatin. 
Manova Test Criteria and Exact F Statistics for 
the Hypothesis of no Overall INTERCEPT Effect 
on the variables defined by the M Matrix Transformation 
H = Type III SS&CP Matrix for INTERCEPT E = Error SS&CP Matrix 
S=l M=-0.5 N=l 
Contrast variable: 30W - Dry 
Statistic Value F NumDF DenDF Pr>F 
Wilks' Lambda 0.20092697 15.9077 1 4 0.0163 
Pillai's Trace 0.79907303 15.9077 1 4 0.0163 
Hotelling-Lawley Trace 3.97693255 15.9077 1 4 0.0163 
Roy's Greatest Root 3.97693255 15.9077 1 4 0.0163 
Contrast variable: SW - Dry 
Statistic Value F NumDF DenDF Pr>F 
Wilks' Lambda 0.21248622 14.8248 1 4 0.0183 
Pillai's Trace 0.78751378 14.8248 1 4 0.0183 
Hotelling-Lawley Trace 3.70618759 14.8248 I 4 0.0183 
Roy's Greatest Root 3.70618759 14.8248 1 4 0.0183 
Contrast variable: 30W - SW 
Statistic Value F NumDF DenDF Pr>F 
Wilks' Lambda 0.2711003 10.7547 1 4 0.0305 
Pillai's Trace 0.7288997 10.7547 I 4 0.0305 
Hotelling-Lawley Trace 2.68867164 10.7547 1 4 0.0305 
Roy's Greatest Root 2.68867164 10.7547 1 4 0.0305 
(b) 
169 
B-5. Analyses of variance of (a) main effect and (b) pairwise contrasts on mean EPG total 
durations for sequential (pooled), 5-cc water (5W), and 30-cc water (30W) swallows, 
using the general linear models procedure. 
(a) Main Effect 
Tests of Hypotheses for Mixed Model Analysis of Variance 
Source: TASK 
Error: 0.9952*MS(TASK*SUBJECT) + 0.0048*MS(Error) 
DF 
2 
Type III MS 
5202717.8634 
Source: SUBJECT 
Denominator Denominator 
DF MS 
8.02 131054.48319 
F Value 
39.6989 
Error: 0.9851 *MS(TASK*SUBJECT) + 0.0149*MS(Error) 
Denominator Denominator 
DF Type III MS DF MS F Value 
4 889601 .00469 8.07 130121.19376 6.8367 
Source: TASK*SUBJECT 
Error: MS(Error) 
Denominator Denominator 
DF Type III MS DF MS F Value 
8 131497.90186 89 39252.950238 3.3500 
Pairwise Contrasts 
Analysis of Variance of Contrast Variables 
Contrast Variable: (Pooled Sequential) - (SW) 
Source DF Type III SS F Value Pr>F 
MEAN I 2515716.871380 44.84 0.0026 
Error 4 224401 .958920 
Contrast Variable: (Pooled Sequential) - (30W) 
Source DF Type III SS F Value Pr>F 
MEAN I 1590243.128820 58.24 0.0016 
Error 4 109218.359880 
Pr>F 
0.0001 
Pr>F 
0.0105 
Pr>F 
0.0021 
170 
B-6. Analyses of variance of (a) main effect and (b) pairwise contrasts on Front:Back 
percent-contact ratios at time of maximal midline for sequential (pooled), 5-cc water 
(SW), and 30-cc water (30W) swallows, using the general linear models procedure. 
(a) Main Effect 
Tests of Hypotheses for Mixed Model Analysis of Variance 
Source: TASK 
Error: 0.9952*MS(TASK*SUBJECT) + 0.0048*MS(Error) 
Denominator Denominator 
DF Type III MS DF MS F Value Pr > F 
2 6.6546278153 8.01 0.413608646 16.0892 0.0016 
Source: SUBJECT 
Error: 0.9851 *MS(TASK*SUBJECT) + 0.0149*MS(Error) 
Denominator Denominator 
DF Type III MS DF MS F Value Pr > F 
4 0.5999607626 8.02 0.4097039707 1.4644 0.2987 
Source: TASK*SUBJECT 
Error: MS(Error) 
Denominator Denominator 
DF Type III MS DF MS F Value Pr > F 
8 0.415463811 89 0.0295314676 14.0685 0.0001 
(b) Pairwise Contrasts 
Analysis of Variance of Contrast Variables 
Contrast Variable: (Pooled Sequential) - SW 
Source DF Type III SS F Value Pr > F 
MEAN 1 3. 16808000 22.40 0.0091 
Error 4 0.56572000 
Contrast Variable: (Pooled Sequential) - 30W 
Source DF Type III SS F Value Pr > F 
MEAN 1 2.15168000 14.12 0.0198 
Error 4 0.60972000 
(b) 
!e}~ Analyses of variance of (a) main effect and (b) pairwise contrasts on Lateral:Back 
( SW) nt-contact ratios at time of maximal midline for sequential (pooled), 5-cc water 
' and 30-cc water (30W) swallows, using the general linear models procedure. 
( a) Main Effect 
Tests of Hypotheses for Mixed Model Analysis of Variance 
Source: TASK 
Error: 0.9952*MS(TASK*SUBJECT) + 0.0048*MS(Error) 
Denominator Denominator 
DF 
2 
Type III MS 
l.0427676398 
Source: SUBJECT 
DF MS 
8.01 0.0745735983 
F Value 
13.9831 
Error: 0.9851 *MS(TASK*SUBJECT) + 0.0149*MS(Error) 
DF 
4 
Type III MS 
0.0865437127 
Denominator Denominator 
DF MS 
8.03 0.0739163832 
Source: TASK*SUBJECT 
Error: MS(Error) 
Denominator Denominator 
DF Type III MS DF MS 
8 0.0748858503 89 0.0099276721 
Pairwise Contrasts 
Analysis of Variance of Contrast Variables 
Contrast Variable: (Pooled Sequential) - 5W 
Source DF Type III SS F Value 
MEAN l 0.51842000 20.39 
Error 4 0.10168000 
Contrast Variable: (Pooled Sequential) - 30W 
Source DF Type III SS F Value 
MEAN l 0.31250000 11. 75 
Error 4 0.10640000 
F Value 
1.1708 
F Value 
7.5431 
Pr > F 
0.0107 
Pr > F 
0.0266 
Pr > F 
0.0024 
Pr > F 
0.3918 
Pr > F 
0.0001 
171 
APPENDIX C. TONGUE LOWERING DURING BOLUS INTAKE IN DISCRETE SWALLOWING 
C-1. Digitized ultrasound images illustrating progressive lowering of the tongue to receive the bolus during a representative single 
swallow. As shown, displacement of the anterior, mid, and dorsal tongue segments (left to right) is graded in extent, while that of the 
posterior segment is minimal throughout bolus presentation. Arrows in each image mark points on the tongue surface. The time under 
each frame reflects second and 1/100th of a second. 
20:02 20:18 20:32 20:38 
20:48 20:55 20:78 
C-2. Composite display oflingual surface curves extracted from the digitized ultrasound images shown in Appendix C-1 , illustrating 
progressive lowering of the tongue to receive the bolus during a representative single swallow. Curve marked "1st" corresponds to 
frame 20:02~ curve marked "last" corresponds to frame 20:78. Each curve is plotted based on the x- (on X axis) and y-values (on Y 
axis) of the coordinates for 101 surface points. Unit of measurement= pixel (with 320x240 pixels per frame and 18 pixels per cm). 
Edge extraction and curve plotting were accomplished using a Matlab-based image processing package (Sze, 1996). 
60 
80 
-
100 
cu 
X 
s 1st 
>- 120 
last 
140 
160 
front back 
120 140 160 180 200 220 240 
X (pixel) 
REFERENCES 
Abbs, J. H . (1973). The influence of the gamma motor system onjaw movement during 
speech: A theoretical framework and some preliminary observations. Journal of 
Speech and Hearing Research, 1.Q, 175-200. 
Abbs, J. H., & Connor, N. P. (1991). Motorsensory mechanisms of speech motor timing 
and coordination. Journal of Phonetics, 19, 333-342. 
Abbs, J. H. , & Gracco, V. L. (1983). Sensorimotor actions in the control of 
multimovement speech gestures. Trends in Neurology, §., 391-395 . 
Abbs, J. H ., & Gracco, V. L. (1984). Control of complex motor gestures: Orofacial 
muscle responses to load perturbations of the lip during speech. Journal of 
Neurophysiology, g , 705-723. 
Abbs, J. H ., Gracco, V . L., & Cole, K. J. (1984). Control ofmultimovement coordination: 
Sensorimotor mechanisms in speech motor programming. Journal of Motor 
Behavior, 16, 195-232. 
Abd-el-Malek, S. (1939). Observations on the morphology of the human tongue. Journal 
of Anatomy, 73, 201-210. 
Abd-el-Malek, S. (1955). The part played by the tongue in mastication and deglutition. 
Journal of Anatomy, 89, 250-255 . 
Adnerhill, I. , Ekberg, O., & Grober, M. E. (1989). Determining normal bolus size for thin 
liquids. Dysphagia, 1, 1-3. 
Alfonso, P. J. , & Baer, T. (1982). Dynamics of vowel articulation. Language and Speech, 
25, 151-173. 
Allen, L. R. (1958). Improved phonetics in denture construction. Journal of Prosthetic 
Dentistry, ~' 753-763 . 
174 
Altschuler, S. M ., Bao, X. , Bieger, D., Hopkins, D. A., & Miselis, R. R. (I 989). 
Viscerotopic representation of the upper alimentary tract in the rat: Sensory 
ganglia and nuclei of the solitary and spinal trigeminal tracts. Journal of 
Comparative Neurology, 283, 248-268. 
175 
Amri, M. , & Car, A. (I 988). Projections from the medullary swallowing center to the 
hypoglossal motor nucleus: A neuroanatomical and electrophysiological study in 
sheep. Brain Research, 441 , 119-126. 
Amri, M ., Car, A., & Roman, C. (1990). Axonal branching ofmedullary swallowing 
neurons projecting on the trigeminal and hypoglossal motor nuclei : Demonstrated 
by electrophysiological and fluorescent double labeling techniques. Experimental 
Brain Research, fil., 384-390. 
Amri, M ., Lamkadem, M ., & Car, A. (1989). Activity of extrinsic tongue muscles during 
swallowing in sheep. Brain Research, 503, 141-143. 
Amri, M ., Lamkadem, M ., & Car, A. (1991). Effects oflingual nerve and chewing cortex 
stimulation upon activity of the swallowing neurons located in the region of the 
hypoglossal motor nucleus. Brain Research, 548, 149-155. 
Anderson, D . J. , & Matthews, B . (Eds.). (1976). Mastication. Bristol: John Wright. 
Astrom, K . E. (I 953). On the central course of afferent fibers in the trigeminal, facial, 
g lossopharyngeal and vagal nerves and their nuclei in the mouse. Acta 
Physiologica Scandinavica, 29, Supplement 106, 209-320. 
Baum, B . J. (1987). Saliva secretion and composition. Frontiers of Oral Physiology, .Q, 
126-134. 
Beckstead, R. M., & Norgren, R. (I 979). An autoradiographic examination of the central 
distribution of the trigeminal, facial , glossopharyngeal, and vagal nerves in the 
monkey. Journal of Comparative Neurology, 184, 455-472. 
Bieg er, D ., & Hockman, C.H. (1976). Suprabulbar modulation ofreflex swallowing. 
E xperimental Neurology, 52, 311-324. 
Bole, C. T ., II. (1965). Electromyographic kinesiology ofthe genioglossus muscles in 
man. Unpublished master's thesis, Ohio State University, Dayton, OH. 
Borden, G. J. , & Gay, T. (I 979). Temporal aspects of articulatory movements for /s/-stop 
clusters. Phonetica, 36, 21-31 . 
176 
Borden, G. J., & Harris, K. S. (1984). Speech science primer (2nd eel). Baltimore: 
Williams & Wilkins. 
Bosma, J. F . (1957). Deglutition: Pharyngeal stage. Physiological Reviews, 37, 275-300. 
Bosma, J . F . (1973). Prologue to symposium. In J. F. Bosma (Ed.), Fourth symposium on 
oral sensation and perception: Development in the fetus and infant (pp. 1-8). 
Bethesda, MD: National Institute of Health. 
Bosma, J. F ., Hepburn, L. G., Josell, S. D., & Baker, K. (1990). Ultrasound 
demonstration of tongue motions during suckle feeding. Developmental Medicine 
and Child Neurology, 32, 223-229. 
Bowman, J.P. (1968). Muscle spindles in the intrinsic and extrinsic muscles of the rhesus 
monkey's (Macaca mulatta) tongue. Anatomical Record, 161, 483-487. 
Bowman, J. P., & Combs, C. M. (1968). Discharge patterns of lingual spindle afferent 
fibers in the hypoglossal nerve of the rhesus monkey. Experimental Neurology, 
21, 105-119. 
Bowman, J. P. , & Combs, C. M. (1969). The cerebrocortical projections ofhypoglossal 
afferents in the monkey as revealed by the evoked potential method. Experimental 
Neurology, 23, 291-301. 
Browman, C. P. , & Goldstein, L. (1990). Tiers in articulatory phonology, with some 
implications for casual speech. In J. Kingston & M. E. Beckman (Eds.), Papers in 
laboratory phonology I: Between the grammar and the physics of speech (pp. 341-
376). Cambridge: Cambridge University Press. 
Butcher, A (1989). Measuring coarticulation and variability in tongue contact patterns. 
Clinical Linguistics & Phonetics, J., 39-47. 
Byrd, D ., Flemming, E. , Mueller, C. A, & Tan, C. C. (1995). Using regions and indices 
in EPG data reduction. Journal of Speech and Hearing Research, 38, 821-827. 
Capra, N. F . (1995). Mechanisms of oral sensation. Dysphagia, 10, 235-247. 
Car, A, & Amri, M. (1987). Activity of neurons located in the region of the hypoglossal 
motor nucleus during swallowing in sheep. Experimental Brain Research, 69, 
175-182. 
177 
Castell, J. A., Dalton, C. B., & Castell, D . 0. (1990a). Effects of body position and bolus 
consistency on the manometric parameters and coordination of the upper 
esophageal sphincter and pharynx. Dysphagia, ~. 179-186. 
Castell, J. A., Dalton, C. B., & Castell, D. 0. (1990b). Pharyngeal and upper esophageal 
sphincter manometry in humans. American Journal of Physiology, 258, 
Gl 73-Gl 78. 
Catford, J. C. (1977) . Fundamental problems in phonetics. Bloomington, IN: Indiana 
University Press. 
Cerenko, D ., Mcconnel, F. M. S., & Jackson, R. T. (1989). Quantitative assessment of 
pharyngeal bolus driving force. Otolaryngology and Head and Neck Surgery, 100, 
57-63. 
Chi-Fishman, G., Capra, N. F., & McCall, G. N. (1994). Thermomechanical facilitation of 
swallowing evoked by electrical nerve stimulation in cats. Dysphagia, .2, 149-155. 
Cleall, J. F . (1965). Deglutition: A study of form and function. American Journal of 
Orthodontics, il, 566-594. 
Contreras, R. J. , Beckstead, R. M., & Norgren, R. (1982). The central projections of the 
trigeminal, facial, glossopharyngeal and vagus nerves: An autoradiographic study 
in the rat. Journal of the Autonomic Nervous System, 6, 303-322. 
Cook, I. J. S. (1991). Normal and disordered swallowing: New insights. Bailliere's Clinical 
Gastroenterology, ~. 245-267. 
Cook, I. J., Dodds, W . J., Dantas, R. 0., Kem, M. K., Massey, B. T., Shaker, R., & 
Hogan, W . J. (1989). Timing of videofluoroscopic, manometric events, and bolus 
transit during the oral and pharyngeal phases of swallowing. Dysphagia, ~, 8-15 . 
Cook, I. J., Dodds, W . J. , Dantas, R. 0 ., Massey, B., Kem, M. K., Lang, I. M., Brasseur, 
J. G ., & Hogan, W. J. (1989). Opening mechanisms of the human upper 
esophageal sphincter. American Journal of Physiology, 257, G748-G759. 
Cooper, S. (1953). Muscle spindles in the intrinsic muscles of the human tongue. Journal 
of Physiology, 122, 193-202. 
Cordaro, M. A. (1991). A personal computer-based display of three-dimensional tongue 
surface dynamics. Unpublished master's thesis. Johns Hopkins University, 
Baltimore, MD. 
178 
Cottle, M . K. ( 1964 ). Degeneration studies of primary afferents of IXth and Xth cranial 
nerves in the cat. Journal of Comparative Neurology, 122, 321-345. 
Cunningham, D. P., & Basmajian, J. V. (1969). Electromyography of genioglossus and 
geniohyoid muscles during deglutition. Anatomical Record, 165, 401-410. 
Dantas, R. 0 . & Dodds, W. J. (1990). Effect of bolus volume and consistency on 
swallow-induced submental and infrahyoid electromyographic activity. Brazilian 
Journal of Medical and Biological Research, 23, 37-44. 
Dantas, R. 0 ., Dodds, W. J., Massey, B. T., & Kem, M. K. (1989). The effect of high- vs 
low-density barium preparations on the quantitative features of swallowing. 
American Journal ofRoentgenology, 153, 1191-1195. 
Dantas, R. 0., Dodds, W. J. , Massey, B. T., Shaker, R., & Cook, I. J. (1990). 
Manometric characteristics of glossopalatal sphincter. Digestive Diseases and 
Sciences, 35 , 161-166. 
Dantas, R. 0 ., Kem, M . K., Massey, B. T. , Dodds, W. J. , Kahrilas, P. J., Brasseur, J. G. , 
Cook, I. J. , & Lang, I. M. (1990). Effect of swallowed bolus variables on oral and 
pharyngeal phases of swallowing. American Journal of Physiology, 258, 
G675-G681. 
Davis, J. W., Lazarus, C., Logemann, J., & Hurst, P. S. (1987). Effect of a maxillary 
glossectomy prosthesis on articulation and swallowing. Journal of Prosthetic 
Dentistry, 57, 715-719. 
Dodds, W. J . (1988, March). Physiology of swallowing. In W. J. Ravich, M. W. Donner, 
B. Jones, & J. A. Logemann (Chairs), Second symposium on dysphagia. 
Symposium sponsored by the Johns Hopkins Swallowing Center, Baltimore, MD. 
Dodds, W. J. , Logemann, J. A., & Stewart, E. T. (1990). Radiologic assessment of 
abnormal oral and pharyngeal phases of swallowing. American Journal of 
Roentgenology, 154, 965-974. 
Dodds, W. J. , Man, K. M., Cook, I. J. , Kahrilas, P. J. , Stewart, E. T. , & Kem, M. K. 
(1988). Influence ofbolus volume on swallow-induced hyoid movement in normal 
subjects. American Journal ofRoentgenology, 150, 1307-1309. 
Dodds, W . J. , Stewart, E.T., & Logemann, J. A. (1990). Physiology and radiology of the 
normal oral and pharyngeal phases of swallowing. American Journal of 
Roentgenology, 154, 953-963 . 
179 
Dodds, W. J., Taylor, A. J., Stewart, E. T., Kem, M. K., Logemann, J. A., & Cook, I. J. 
(I 989). Tipper and dipper types of oral swallows. American Journal of 
Roentgenology, 153, 1197-1199. 
Donner, M . W. , Bosma, J. F., & Robertson, D. L. (1985). Anatomy and physiology ofthe 
Pharynx. Gastrointestinal Radiology, lQ, 196-212. 
Doty, R. W . (I 951 ). Influence of stimulus pattern on reflex deglutition. American Journal 
of Physiology, 166, 142-158. 
Doty, R. W. (I 968). Neural organization of deglutition. In C. F. Code (Ed.), Handbook of 
physiology: Sec. 6. Alimentary canal: Vol. IV. Motility, Chapter 92 (pp. 
1861-1902). Washington, DC: American Physiological Society. 
Doty, R. W., & Bosma, J. F. (1956). An electromyographic analysis ofreflex deglutition. 
Journal of Neurophysiology, 12., 44-60. 
Doty, R. W., Richmond, W. H., & Storey, A. T. (1967). Effect ofmedullary lesions on 
coordination of deglutition. Experimental Neurology, 11, 91-106. 
Dubner, R., Sessle, B. J., & Storey, A. T. (1978). The neural basis of oral and facial 
function. New York: Plenum Press. 
Ekberg, 0 . (I 986). The normal movements of the hyoid bone during swallow. 
Investigative Radiology, 21, 408-410. 
Ekberg, 0 ., & Borgstrom, P. S. (1989). A graphic representation of pharyngeal wall 
motion during swallow: Technical note. Dysphagia, .1, 43-47. 
Ekberg, 0 ., Liedberg, B., & Owall, B. (1986). Barium and meat. A comparison between 
pharyngeal swallow of fluid and solid boluses. Acta Radiologica. Diagnosis, 27, 
701-704. 
Ekberg, 0 ., Olsson, R., & Sundgren-Borgstrom, P. (I 988). Relation of bolus size and 
pharyngeal swallow. Dysphagia, l, 69-72. 
Fallside, F., & Woods, W. A. (1985). Computer speech processing. Englewood Cliffs, NJ: 
Prentice-Hall. 
Fitzgerald, M. J. T., & Sachithanandan, S. R. (1979). The structure and source oflingual 
proprioceptors in the monkey. Journal of Anatomy, 128, 523-552. 
180 
Fletcher, S. G. ( 1985). Speech production and oral motor skill in an adult with an 
unrepaired palatal cleft. Journal of Speech and Hearing Disorders, 50, 254-261. 
Fletcher, S. G., Dagenais, P.A., & Critz-Crosby, P. (1991). Teaching consonants to 
profoundly hearing-impaired speakers using palatometry. Journal of Speech and 
Hearing Research, 34, 929-942. 
Folkins, J. W., & Abbs, J. H. (1975). Lip and jaw motor control during speech: Responses 
to resistive loading of the jaw. Journal of Speech and Hearing Research, _lli, 207-
220. 
Fontdevila, ],, Pallares, M. D., &. Recasens. D. (1 994). The contact index method of 
electropalatographic data reduction. Journal of Phonetics.~ l 1-154. 
Freund, H.-J., & Bildingen, H. J. (1978). The relationship between speed and amplitude of 
the fastest voluntary contractions of human arm muscles. Experimental Brain 
Research, 11, 1-12. 
Fujimura, 0., Tatsumi, I. F., & Kagaya, R. (1972). Computational processing of 
palatographic patterns. Journal of Phonetics, l, 47-54. 
Gay, T. (1981). Mechanisms in the control of speech rate. Phonetica, 38, 148-158. 
Gibbon, F., Hardcastle, W., & Nicoladis, K. (1993). Temporal and spatial aspects of 
lingual coarticulation in /kl/ sequences: A cross-linguistic investigation. Language 
and Speech, 36, 261-277. 
Gracco, V. L., & Abbs, J. H. (1985). Dynamic control of the perioral system during 
speech: Kinematic analyses of autogenic and nonautogenic sensorimotor 
processes. Journal ofNeurophysiology, 54, 418-432. 
Gracco, V. L., & Abbs, J. H. (1989). Sensorimotor characteristics of speech motor 
sequences. Experimental Brain Research, 75, 586-598. 
Gramiak, R., Kelley, M. L. Jr., & Gravinia, R. F. (1967). Nasal pressure changes during 
swallowing. American Journal ofRoentgenology, 99, 562-576. 
Grillner, S., & Wallen, P. (1985). Central pattern generators for locomotion, with special 
reference to vertebrates. Annual Review of Neuroscience,~. 233-262. 
Halsell, C. B., Travers, J.B., & Travers, S. P. (1993). Gustatory and tactile stimulation of 
the posterior tongue activate overlapping but distinctive regions in the nucleus of 
the solitary tract. Brain Research, 632, 161-173. 
181 
Ham, H. R., Georges, B., Froideville, J. L., & Piepsz, A. (1985). Oesophageal transit of 
liquid : Effects of single or multiple swallows. Nuclear Medicine Communications, 
Q, 263-267. 
Hamlet, S. L . ( 1989). Dynamic aspects of lingual propulsive activity in swallowing. 
Dysphagia, ~' 136-145. 
Hamlet, S. L ., Bunnell, H. T., & Struntz, B. G. (1986). Articulatory asymmetries. Journal 
of the Acoustical Society of America, 79, 1164-1169. 
Hamlet, S. , Choi, J., Zormeier, M ., Shamsa, F., Staehler, R., Muz, J. , & Jones, L. (1996). 
Normal adult swallowing ofliquid and viscous material : Scintigraphic data on 
bolus transit and oropharyngeal residues. Dysphagia, ll., 41-47. 
Hamlet, S., Jones, L., Mathog, R., Bolton, M., & Patterson, R. (1988). Bolus propulsive 
activity of the tongue in dysphagic cancer patients. Dysphagia, 1, 18-23 . 
Hamlet, S. L., & Stone, M. (1978). Compensatory alveolar consonant production induced 
by wearing a dental prosthesis. Journal of Phonetics,§, 227-248. 
Hamlet, S. L., Stone, M ., & Shawker, T. H. (1988). Posterior tongue grooving in 
deglutition and speech: Preliminary observations. Dysphagia, 1, 65-68 . 
Hanamori, T., & Smith, D. V. (1986). Central projections of the hamster superior 
laryngeal nerve. Brain Research Bulletin, 16, 271-279. 
Hancock, P.A., & Newell, K. M. (1985). The movement speed-accuracy relationship in 
space-time. In H . Heuer, U. Keinbeck, & K.-H. Schmidt (Eds.), Motor behavior: 
Programming, control, and acquisition (pp. 153-188). New York: Springer-
Verlag. 
Hardcastle, W. J ., Gibbon, F., & Nicolaidis, K. (1991). EPG data reduction methods and 
their implications for studies oflingual coarticulation. Journal of Phonetics, 19,. 
251-266. 
Hardcastle, W., Jones, W ., Knight, C., Trudgeon, A., & Calder, G. (1989). New 
developments in electropalatography: A state-of-the-art report. Clinical Linguistics 
& Phonetics, 1, 1-38. 
Hellstrand, E. (1982a) . Reflex control of cat extrinsic and intrinsic tongue muscles exerted 
by intraoral receptors. Acta Physiologica Scandinavica, 115, 245-256. 
182 
Hellstrand, E. (1982b). The neuromuscular system of the tongue. In S. Grillner, B. 
Lindblom, J. Lubker, & A. Persson (Eds.), Speech motor control (pp. 141-157). 
New York: Pergamon Press. 
Hendrix, T . R. (I 980). The motility of the alimentary canal. In V. B. Mountcastle (Ed.), 
Medical physiology. Vol. 2 (pp. 1320-1347). St. Louis: C. V. Mosby. 
Hedrick, W.R. , Hykes, D . L., & Starchman, D. E. (Eds.). (1995). Ultrasound physics and 
instrumentation (3rd ed.). St. Louis: Mosby. 
Hiatt, J . L., & Gartner, L. P. (1987). Textbook of head and neck anatomy. Baltimore, 
MD: Williams & Wilkins. 
Hiiemae, K. M., & Crompton, A. W. (1985). Mastication, food transport, and swallowing. 
In M. Hildebrand, D . M. Bramble, K. F. Liem, & D. B. Wake (Eds.), Functional 
vertebrate morphology (pp. 262-290). Cambridge, MA: Harvard University Press. 
Riki, S., & Itoh, H. (1986) . Influence of palate shape on lingual articulation. Speech 
Communication, .5., 141-158. 
Hirano, M., Kuroiwa, Y., Tanaka, S., Matsuoka, H., Sato, K., & Yoshida, T. (1992). 
Dysphagia following various degrees of surgical resection for oral cancer. Annals 
of Oto logy. Rhinology and Laryngology, 101, 13 8-141. 
Hockman, C . H., Bieger, D., & Weerasuriya, A. (1979). Supranuclear pathways of 
swallowing. Progress in Neurobiology, ll, 15-32. 
Hollshwandner, C.H., Brenman, H. S., & Friedman, M. H.F. (1975) . Role of afferent 
sensors in the initiation of swallowing in man. Journal of Dental Research, 54, 
83-88. 
Hoole, P., Ziegler, W., Hartmann, E., & Hardcastle, W. (1989). Parallel 
electropalatographic and acoustic measures of fricatives. Clinical Linguistics & 
Phonetics, J, 59-69. 
Hrychshyn, A. W. , & Basmajian, J. V. (1972). Electromyography of the oral stage of 
swallowing in man. American Journal of Anatomy, 133, 330-340. 
Hughes, 0 . M ., & Abbs, J. H. (1976). Labial-mandibular coordination in the production of 
speech: Implications fro the operation of motor equivalence. Phonetica, 33, 
199-221 . 
183 
Humphrey, T. (1 970). Reflex activity in the oral and facial area of the human fetus. In J. F. 
Bosma (Ed.), Second symposium on oral sensation and perception (pp. 195-233). 
Springfield, IL: Charles C. Thomas. 
Ingervall, B., Bratt, C. M., Carlsson, G. E., Helkimo, M., & Lantz, B. (1972). Duration of 
swallowing with and without anesthesia of the temporomandibular joints. 
Scandinavian Journal of Dental Research, 80, 189-196. 
Ingervall, B., & Lantz, B. ( 1973 ). Significance of gravity in the passage of bolus through 
the human pharynx. Archives of Oral Biology, 2, 351 -356. 
Issa, F . G . & Porostocky, S. (1994). Effect of continuous swallowing on respiration. 
Respiration Physiology, 95, 181-193 . 
Jack, F . R. & Gibbon, F . (1995). Electropalatography in the study of tongue movement 
during eating and swallowing (a novel procedure for measuring texture-related 
behavior). International Journal of Food Science and Technology, 30, 415-423 . 
Jacob, P. , Kahrilas, P . J. , Logemann, J. A., Shah, V., & Ha, T. (1989). Upper esophageal 
sphincter opening and modulation during swallowing. Gastroenterology, 97, 
1469-1478. 
Jean, A. (1984a). Brainstem organization of the swallowing network. Brain, Behavior and 
Evolution, 25, 109-116. 
Jean, A. (1984b). Control ofthe central swallowing program by inputs from the peripheral 
receptors. A review. Journal of the Autonomic Nervous System, 10, 225-233 . 
Jean, A. ( 1990). Brainstem control of swallowing: Localization and organization of the 
central pattern generator for swallowing. In A. Taylor (Ed.), Neurophysiology of 
the jaw and teeth (pp. 294-321). Houndmills, England: Macmillan Press. 
Jean, A., Amri, M., & Calas, A. (1983). Connections between the ventral medullary 
swallowing area and the trigeminal motor nucleus of the sheep studied by tracing 
techniques. Journal of the Autonomic Nervous System, 1, 87-96. 
Jean, A., & Car, A. (1979). Inputs to the swallowing medullary neurons from the 
peripheral afferent fibers and the swallowing cortical area. Brain Research, 178, 
567-572. 
Jenny, A. B. , & Saper, C. B. (1987). Organization ofthe facial nucleus and corticofacial 
projection in the monkey: A reconsideration of the upper motor neuron facial 
palsy. Neurology, 37, 930-939. 
184 
Jurgens, U. ( 1976). Projections from the cortical larynx area in the squirrel monkey. 
Experimental Brain Research, 25, 401-411. 
Kahrilas, P . J ., Dodds, W . J. , Dent, J. , Logemann, J. A., & Shaker, R. (1988). Upper 
esophageal sphincter function during deglutition. Gastroenterology, 95, 52-62. 
Kahrilas, P . J. , Lin, S., Logemann, J. A., Ergun, G. A., & Facchini, F. (1993). Deglutitive 
tongue action: Volume accommodation and bolus propulsion. Gastroenterology, 
104, 152-162. 
Kahrilas, P . J. , & Logemann, J. A. (1993). Volume accommodation during swallowing. 
Dysphagia, ~, 259-265 . 
Kahrilas, P . J. , Logemann, J. A., Lin, S., & Ergun, G. A. (1992). Pharyngeal clearance 
during swallowing: A combined manometric and videofluoroscopic study. 
Gastroenterology, 103, 128-136. 
Kaplan, M . D ., & Baum, B. J. (1993). The Functions of Saliva. Dysphagia, ~, 225-229. 
Kawasaki, M ., Ogura, J. H., & Takenouchi, S. (1964). Neurophysiological observations of 
normal deglutition. I. Its relationship to the respiratory cycle. Laryngoscope, 74, 
1747-1765. 
Kennedy, D . (1969). The control of output by central neurons. In M.A. B. Brazier (Ed.), 
The intemeuron (pp. 21-36). Berkeley, CA: University of California Press. 
Kennedy, J. G. , III, & Kent, R. D. (1985). Anatomy and physiology of deglutition and 
related functions. Seminars in Speech and Language, Q, 257-273 . 
Kennedy, J . G ., III, & Kent, R. D. (1988). Physiological substrates of normal deglutition. 
Dysphagia, 1, 24-37. 
Kent, R. D. , Martin, R. E ., & Sufit, R. L. (1990). Oral sensation: A review and clinical 
prospective. In H. Winitz (Ed.), Human communication and its disorders : A 
review, 1990 (pp. 135-191). Norwood, NJ: AblexPublishing. 
Kerr, F . W . L. (1962). Facial, vagal and glossopharyngeal nerves in the cat. Archives of 
Neurology, Q, 264-281 . 
Kessler, J.P., & Jean, A. (1985). Identification of the medullary swallowing regions in the 
rat. Experimental Brain Research, 57, 256-263 . 
185 
Kessler, J . P ., & Jean, A. (1986). Inhibitory influence ofmonoamines and brainstem 
monoaminergic regions on the medullary swallowing reflex. Neuroscience Letters, 
65, 41-46. 
Kier, W. M., & Smith, K. K. (1985). Tongues, tentacles and trunks: The biomechanics of 
movement in muscular-hydrostats. Zoological Journal of the Linnean Society, 83, 
307-324. 
Ku, D . N ., Ma, P.-P. , McConnel, F. M . S., & Cerenko, D. · 1990). A kinematic study of 
the oropharyngeal swallowing of a liquid. Annals of Biomedical Engineering,~. 
655-669. 
Kuypers, H . G. J. M. (I 958a). Corticobulbar connexions to the pons and lower brain-stem 
in man: An anatomical study. Brain, fil, 364-388. 
Kuypers, H . G . J.M. (1958b). Some projections from the peri-central cortex to the pons 
and lower brain stem in monkey and chimpanzee. Journal of Comparative 
Neurology, 110, 221-255 . 
LaBlance, G . R., Kraus, K., & Steckel, K. F. (1991). Rehabilitation of swallowing and 
communication following glossectomy. Rehabilitation Nursing, 16, 266-270. 
Landgren, S., & Olsson, K. A. (I 982). Oral mechanoreceptors. In S. Grillner, B. 
Lindblom, J. Lubker, & A. Persson (Eds.), Speech motor control (pp. 129-139). 
New York: Pergamon Press. 
Lazarus, C. , & Logemann, J. A. (1987).Swallowing disorders in closed head trauma 
patients. Archives of Physical Medicine and Rehabilitation, 68, 79-84. 
Lazarus, C., Logemann, J. A., & Gibbons, P. (1993). Effects of maneuvers on swallowing 
function in a dysphagic oral cancer patient. Head & Neck, U, 419-424. 
Lazarus, C. I ., Logemann, J. A., Rademaker, A. W., Kahrilas, P. J. , Pajak, T. , Lazar, R., 
& Halper, A. (1993) . Effects ofbolus volume, viscosity, and repeated swallows in 
nonstroke subjects and stroke patients. Archives of Physical Medicine and 
Rehabilitation, 74, 1066-1070. 
Lindblom, B . (1983) . Economy of speech gestures. In P . F. MacNeilage (Ed.), The 
production of speech (pp. 217-245) . New York: Springer-Verlag. 
Logemann, J. A. (1983). Evaluation and treatment of swallowing disorders. San Diego: 
College-Hill Press. 
186 
Logemann, J. A. (1988). Swallowing physiology and pathophysiology. Otolaryngologic 
Clinics of North America, 21, 613-623. 
Logemann, J. A. (1991). Approaches to management of disordered swallowing. Bailliere's 
Clinical Gastroenterology, ~, 269-280. 
Logemann, J. A., Kahrilas, P . J., Cheng, J., Pauloski, B. R., Gibbons, P. J., Rademaker, A. 
W ., & Lin, S. (1992) . Closure mechanisms oflaryngeal vestibule during swallow. 
American Journal of Physiology, 262, G338-G344. 
Logemann, J. A., Kahrilas, P . J. , Hurst, P., Davis, J., & Krugler, C. (1989). Effects of 
intraoral prosthetics on swallowing in patients with oral cancer. Dysphagia, ~, 118-
120. 
Lowe, A. A. ( 1978). Excitatory and inhibitory inputs to hypoglossal motoneurons and 
adjacent reticular formation neurons in cats. Experimental Neurology, 62, 30-47. 
Lowe, A. A. (1981). The neural regulation of tongue movement. Progress in 
Neurobiology, ]2, 295-344. 
Lowe, A. A. (1984). Tongue movements--brainstem mechanisms and clinical postulates. 
Brain, Behavior and Evolution, 25, 128-137. 
Lowe, A. A., Gurza, S. C., Sessle, B. J. ( 1977). Regulation of genioglossus and masseter 
muscle activity in man. Archives of Oral Biology, 22, 579-584. 
Lowe, A. A., & Sessle, B. J. (1973). Tongue activity during respiration, jaw opening, and 
swallowing in cat. Canadian Journal of Physiology and Pharmacology, ,ll, 
1009-1011. 
Lowe, A. A., & Sessle, B. J. (1974). Genioglossus activity during respiration, jaw opening 
and swallowing in cat and monkey. Journal of Dental Research, 53 (Special Issue), 
201. 
MacNeilage, P. F ., & Sholes, G. N. (1964). An electromyographic study of the tongue 
during vowel production. Journal of Speech and Hearing Research, 1, 209-232. 
Mansson, I., & Sandberg, N. (1974). Effects of surface anesthesia on deglutition. 
Laryngoscope, 84, 427-437. 
Mansson, I ., & Sandberg, N. (1975). Oro-pharyngeal sensitivity and elicitation of 
swallowing in man. Acta Otolaryngologica, 79, 140-145. 
187 
Marchal, A. (1988). Coproduction: Evidence from EPG data. Speech Communication, 1, 
287-295. 
Martin, R. E. (1991). A comparison of lingual movement in swallowing and speech 
production. Dissertation Abstracts International, 52, 1982-B. (University 
Microfilms No. 91-22486) 
Martin, R. E., Murray, G. M., & Sessle, B. J. (1991). Firing patterns and properties of 
neurones in primate tongue motor cortex (MI) in relation to swallowing. 
Neuroscience Abstract, 11, 309. 
Martin, R. E., & Sessle, B. J. (1993). The role of the cerebral cortex in swallowing. 
Dysphagia, ~' 195-202. 
McClean, M. D. (1988). Neuromotor aspects of speech production and dysarthria. In K. 
Y orkston, D. Beukelman, & K. Bell (Eds.), Clinical management of dysarthric 
speakers. San Diego: College-Hill. 
McConnel, F. M. S. (1988). Analysis of pressure generation and bolus transit during 
pharyngeal swallowing. Laryngoscope, 98, 71-78. 
McConnel, F. M. S., Guffin, T. N., Jr., Cerenko, D., & Ko, A. S. (1992). The effects of 
bolus flow on vertical pharyngeal pressure measurement in the 
pharyngoesophageal segment: Clinical significance. Otolaryngology and Head and 
Neck Surgery, 106, 169-174. 
McNamara, J. A., Jr. , & Moyers, R. E. (1973). Electromyography of the oral phase of 
deglutition in the rhesus monkey (Macaca mulatta). Archives of Oral Biology, 18, 
995-1002. 
Mermelstein, p . ( 1973). Articulatory model for the study of speech production. Journal of 
the Acoustical Society of America, 53, 1070-1082. 
Meyer, D. E., Smith, J.E. K., & Wright, C. E. (1982). Models for the speed and accuracy 
of aimed movements. Psychological Review, 89, 449-482. 
Meyer, G. W., Gerhardt, D. C., & Castell, D. 0. (1981). Human esophageal response to 
rapid swallowing: Muscle refractory period or neural inhibition? American Journal 
of Physiology, 241, G129-GI36. 
Milidonis, M. K., Kraus, S. L., Segal, R. L., & Widmer, C. G. (1993). Genioglossi muscle 
activity in response to changes in anterior/neutral head posture. American Journal 
of Orthodontics and Dentofacial Orthopedics, 103, 39-44. 
188 
Miller, A. J. (1972a). Characteristics of the swallowing reflex induced by peripheral nerve 
and brainstem stimulation. Experimental Neurology, 34, 210-222. 
Miller, A. J. (1972b ). Significance of sensory inflow to the swallowing reflex. Brain 
Research, 43, 147-159. 
Miller, A. J. (1982). Deglutition. Physiological Reviews, 62, 129-184. 
Miller, A. J. (1986). Neurophysiological basis of swallowing. Dysphagia, l, 91-100. 
Miller, A. J. (1993). The search for the central swallowing pathway: The quest for clarity. 
Dysphagia, ft, 185-194. 
Miller, A. J., & Bowman, J. P. (1974). Divergent synaptic influences affecting discharge 
patterning of genioglossus motor units. Brain Research, 78, 179-191. 
Miller, A. J., & Bowman, J.P. (1977). Precentral cortical modulation of mastication and 
swallowing. Journal ofDental Research, 56, 1154. 
Miller, F. R. (1920). The cortical paths for mastication and deglutition. Journal of 
Physiology, London, 53, 473-478. 
Miller, F. R., & Sherrington, C. S. (1916). Some observations on the bucco-pharyngeal 
stage of reflex deglutition in the cat. Quarterly Journal of Experimental 
Physiology, 2, 147-186. 
Milner, T. E. (1986). Controlling velocity in rapid movements. Journal of Motor Behavior, 
~. 147-161. 
Miyawaki, K. (1974). A study of the musculature of the human tongue. Annual Bulletin, 
Research Institute ofLogopedics and Phoniatrics (University of Tokyo), ft, 23-49. 
Miyawaki, K., Hirose, H., Ushijima, T., & Sawashima, M. (1975). A preliminary report on 
the electromyographic study of the activity of lingual muscles. Annual Bulletin, 
Research Institute ofLogopedics and Phoniatrics (University of Tokyo), 2, 
91-106. 
Miyawaki, K., Kiritani, S., Tatsumi, I. F., & Fujimura, 0. (1974). Palatographic 
observation of VCV articulations in Japanese. Annual Bulletin, Research Institute 
ofLogopedics and Phoniatrics (University ofTokyo), ft, 51-57. 
189 
Mizutani, T., & Hashimoto, K. (1988, November). Analysis of tongue motion for the 
dental consonants based on the high-speed palatographic data. Paper presented at 
the second joint meeting of the Acoustical Society of America and the Acoustical 
Society of Japan, Honolulu, HI. 
Morimoto, T., Inoue, T., Nakamura, T., & Kawamura, Y. (1984). Frequency-dependent 
modulation of rhythmic human jaw movements. Journal of Dental Research, 63, 
1310-1314. 
Morrish, K. A., Stone, M., Shawker, T. H., & Sonies, B. C. (1985). Distinguishability of 
tongue shape during vowel production. Journal of Phonetics, .Ll., 189-203. 
Mueller, A (1992). Percent. Unpublished computer program written for the Department 
of Linguistics, University of California, Los Angeles. 
Munhall, K. G., Ostry, D. J. , & Parush, A. (1985). Characteristics of velocity pro.files of 
speech movements. Journal of Experimental Psychology: Human Perception and 
Performance, ll, 457-474. 
Murray, G. M ., & Sessle, B. J. (1990). Neurones in primate tongue motor cortex alter 
their firing rates during swallow. Neuroscience Abstract, l,Q, 1221 . 
Murray, G. M ., & Sessle, B. J. (1992a). Functional properties of single neurons in the face 
primary motor cortex of the primate. I. Input and output features of tongue motor 
cortex. Journal of Neurophysiology, 67, 747-758. 
Murray, G. M ., & Sessle, B. J. (1992b ). Functional properties of single neurons in the face 
primary motor cortex of the primate. II. Relations with trained orofacial motor 
behavior. Journal ofNeurophysiology, 67, 759-774. 
Murray, G. M ., & Sessle, B. J. (1992c). Functional properties of single neurons in the face 
primary motor cortex of the primate. III. Relations with different directions of 
trained tongue protrusion. Journal of Neurophysiology, 67, 775-785. 
Nakayama, S., Neya, T., Watanabe, K. , & Tsuchiya, K. (1974). Effects of electrical 
stimulation and local destruction on the medulla oblongata on swallowing 
movements in dogs. Rendiconti di Gastro-enterologia, Q, 6-11 . 
Negus, V. E. (1948). The second stage of swallowing. Acta-Otolaryngologica, 
Supplementum 78, 78-82. 
190 
Neya, T. , Watanabe, K., & Yamasato, T. (1974). Localization of potentials in medullary 
reticular formation relevant to swallowing. Rendiconti di Gastro-enterologia, .Q, 
107-110. 
Nelson, J. B., & Castell, D. 0 . (1988). Esophageal motility disorders. Disease-a-Month, 
34, 299-389. 
Newell, K. M. (1980). The speed-accuracy paradox in movement control: Error of time 
and space. In G. E. Stelmach, & J. Requin (Eds.), Tutorials in Motor Behavior 
(pp. 501-510). Amsterdam: Elsevier-North Holland. 
Newman, L. A., Cleveland, R. H., Blick.man, J. G., Hillman, R. E., & Jaramillo, D. (1991). 
Videofluoroscopic analysis of the infant swallow. Investigative Radiology, 26, 
870-873 . 
Niimi, K. , Kishi, S., Miki, M., & Fujita, S. (1963). An experimental study of the course 
and termination of the projection fibers from cortical areas 4 and 6 in the cat. Folia 
Psychiatrica et Neurologica Japonica, 11, 167-216. 
Ohman, S. E. G. (1966). Coarticulation in VCV utterances: Spectrographic 
measurements. Journal of the Acoustical Society of America, 39, 151-168. 
Ostry, D. J., Cooke, J. D., & Munhall, K. G. (1987). Velocity curves of human arm and 
speech movements. Experimental Brain Research, 68, 37-46. 
Ostry, D. J. , & Flanagan, J. R. (1989). Humanjaw movement in mastication and speech. 
Archives of Oral Biology, 34, 685-693 . 
Ostry, D. J., Keller, E., & Parush, A. (1983). Similarities in the control of the speech 
articulators and the limbs: Kinematics of tongue dorsum movement in speech. 
Journal of Experimental Psychology: Human Perception and Performance, .2., 
622-636. 
Ostry, D. J., & Munhall, K. G. (l 985). Control of rate and duration of speech movements. 
Journal of the Acoustical Society of America, 77, 640-648. 
Palmer, J.B. , Tanaka, E., & Siebens, A. A. (1988). Motions of the posterior pharyngeal 
wall in swallowing. Laryngoscope, 98, 414-417. 
Palmer, J. M. (l 973). Dynamic palatography. Phonetica, 28, 76-85 . 
191 
Pearson, K. G . (1987). Central pattern generation: A concept under scrutiny. In H. 
McLennan, J. R. Ledsome, C.H. S. McIntosh, & D.R. Jones (Eds.), Advances in 
physiological research (pp. 167-185). New York: Plenum Press. 
Pearson, K. G. ( 1991 ). Sensory elements in pattern-generating networks. In N. I. Badler, 
B. A. Barsky, & D. Zeltzer (Eds.), Making them move. Mechanics, control and 
animation of articulated figures (pp. 111-127). San Mateo: Morgan Kaufmann. 
Penfield, W ., & Rasmussen, T. (1950). The cerebral cortex of man: A clinical study of 
localization of function. New York: Macmillan. 
Perkell, J. S. (1969). Physiology of speech production: Results and implications of a 
quantitative cineradiographic study (Research Monograph No. 53). Cambridge, 
MA: M.I.T. Press. 
Plesh, 0 ., Bishop, B., & McCall, W. (1987). Mandibular movements and jaw muscles' 
activity while voluntarily chewing at different rates. Experimental Neurology, 98, 
285-300. 
Pommerenke, W. T. (1928). A study of the sensory areas eliciting the swallowing reflex. 
American Journal of Physiology, 84, 36-41. 
Porter, R. (1966). Lingual mechanoreceptors activated by muscle twitch. Journal of 
Physiology, London, 183, 101-111. 
Porter, R. (1967). Cortical actions on hypoglossal motoneurons in cats: A proposed role 
for a common internuncial cell. Journal of Physiology, London, 193, 295-308. 
Pouderous, P., & Kahrilas, P. J. (1995). Deglutitive tongue force modulation·by volition, 
volume, and viscosity in humans. Gastroenterology, 108, 1418-1426. 
Rasband, W. (1994). NIH Image, Version 1.55 [Computer program]. Bethesda, MD: The 
National Institutes of Health. 
Recasens, D . (1991). On the production characteristics of apicoalveolar taps and trills. 
Journal of Phonetics, 19, 267-280. 
Recasens, D., Farnetani, E. , Fontdevila, J. , & Pallares, M. D. (1993). An 
electropalatographic study of alveolar and palatal consonants in Catalan and 
Italian. Language and Speech, 36, 213-234. 
Roberts, R. I. ( 1957). A cineradiographic investigation of pharyngeal deglutition. British 
Journal of Radiology, 30, 449-460. 
192 
Robbins, J., Hamilton, J. W. , Lof, G. L., & Kempster, G. B. (1992). Oropharyngeal 
swallowing in normal adults of different ages. Gastroenterology, 103, 823-829. 
Robbins, J. A., Sufit, R., Rosenbek, J., Levine, R., & Hyland, J. (1987). A modification of 
the modified barium swallow. Dysphagia, 2., 83-86. 
Robbins, K. T. , Bowman, J. B., & Jacob, R. F. (1987). Postglossectomy deglutitory and 
articulatory rehabilitation with palatal augmentation prostheses. Archives of 
Otolaryngology - Head and Neck Surgery, 113, 1214-1218. 
Roman, C. (1986). Controle nerveux de la deglutition et de la motricite oesophagienne 
chez les Mammiferes. Journal of Physiology, Paris, fil., 118-131 . 
Rossi, G. F., & Brodal, A. (1956). Corticofugal fibers to the brain-stem reticular 
formation. An experimental study in the cat. Journal of Anatomy, 90, 42-62. 
Rossignol, S., Lund, J.P., & Drew, T. (1988). The role of sensory inputs in regulating 
patterns of rhythmical movements in higher vertebrates. In A Cohen, S. 
Rossignol, & S. Grillner (Eds.), Neural control of rhythmic movements in 
vertebrates (pp. 201-283). New York: J. Wiley. 
Sakoe, H., & Chiba, S. (I 978). Dynamic programming algorithm optimization for spoken 
word recognition. IEEE Transactions on Acoustics, Speech, and Signal 
Processing, 26, 43-49. 
SAS Institute, Inc. (I 992). SAS system for Windows, version 3.95, release 6.08 
[Computer program]. Cary, NC: SAS Institute, Inc. 
Schmidt, R. A., Sherwood, D. E., Zelaznik, H. N., & Leikind, B. J. (1985). Speed-
accuracy trade-offs in motor behavior: theories of impulse variability. In H. Heuer, 
U. Keinbeck, & K.-H. Schmidt (Eds.), Motor behavior: Programming, control, 
and acquisition (pp. 153-188). New York: Springer-Verlag. 
Selley, W. G., Ellis, R. E., Flack, F. C., & Brooks, W. A (1990). Coordination of 
sucking, swallowing and breathing in the newborn: Its relationship to infant 
feeding and normal development. British Journal of Disorders of Communication, 
25, 311-327. 
Selley, W. G., Ellis, R. E., Flack, F. C., Curtis, H., & Callon, M. (1986). Ultrasound study 
of sucking and swallowing by newborn infants (letter to the editor). 
Developmental Medicine & Child Neurology, 28, 821-823 . 
193 
Sessle, B. J. , & Henry, J. L. (1989). Neural mechanisms of swallowing: 
Neurophysiological and neurochemical studies on brain stem neurons in the 
solitary tract region. Dysphagia, ~' 61-75. 
Shaiman, S. (1989). Kinematic and electromyographic responses to perturbation of the 
jaw. Journal of the Acoustic Society for America, 86, 78-88. 
Shaker, R., Cook, I. J. S., Dodds, W. J. , & Hogan, W. J. (1988). Pressure-Flow dynamics 
of the oral phase of swallowing. Dysphagia, J,, 79-84. 
Shaker, R., Dodds, W. J. , Dantas, R. 0 ., Hogan, W. J. , & Arndorfer, R. C. (1990). 
Coordination of deglutitive glottic closure with oropharyngeal swallowing. 
Gastroenterology, 98, 14 78-1484. 
Shawker, T. H., Sonies, B., Hall, T. E., & Baum, B. F. (1984). Ultrasound analysis of 
tongue, hyoid, and larynx activity during swallowing. Investigative Radiology, 19, 
82-86. 
Shawker, T. H., Sonies, B., Stone, M., & Baum, B. J. (1983). Real-time ultrasound 
visualization of tongue movement during swallowing. Journal of Clinical 
Ultrasound, 11, 485-490. 
Shedd, D. P., Scatliff, J. H., & Kirchner, J. A. (1960). The buccopharyngeal propulsive 
mechanism in human deglutition. Surgery, 48, 846-853 . 
Shingai, T., & Shimada, K. (1976). Reflex swallowing elicited by water and chemical 
substances applied in the oral cavity, pharynx, and larynx of the rabbit. Japanese 
Journal of Physiology, 26, 455-469. 
Shprintzen, R. J., Lencione, R. M ., McCall, G. N., & Skolnick, M. L. (1974). A three 
dimensional cinefluoroscopic analysis ofvelopharyngeal closure during speech and 
nonspeech activities in normals. Cleft Palate Journal, 11, 412-428 . 
Sinclair, W . J. (1970). Initiation of reflex swallowing from the naso- and oropharynx. 
American Journal of Physiology, 218, 956-960. 
Sinclair, W. J. (1971). Role of the pharyngeal plexus in initiation of swallowing. American 
Journal of Physiology, 221, 1260-1263 . 
Sirisko, M. A. , & Sessle, B. J. (1983). Corticobulbar projections and orofacial and muscle 
afferent inputs of neurons in primate sensorimotor cerebral cortex. Experimental 
Neurology, 82, 716-720. 
194 
Smith, A., Goffman, L., Zelaznik, H. N., Ying, G., & McGillem, C. (I 995). 
Spatiotemporal stability and patterning of speech movement sequences. 
Experimental Brain Research, 104, 493-501. 
Smith, K. K., & Kier, W. M. (I 989). Trunks, tongues, and tentacles: Moving with 
skeletons of muscle. American Scientist, 77, 29-35. 
Smith, T. S. (1971). A phonetic study of the function of the extrinsic tongue muscles 
(UCLA Working Papers in Phonetics No. 18). Los Angeles: University of 
California. 
Sonies, B. C. (1991). Ultrasound imaging and swallowing. In M. W. Donner & B. Jones 
(Eds.), Normal and abnormal swallowing: Imaging in diagnosis and therapy 
(pp. 109-117). New York: Springer. 
Sonies, B. C., Parent, L. J., Morrish, K., & Baum, B. (1988). Durational aspects of the 
oral-pharyngeal phase of swallow in normal adults. Dysphagia, l, 1-10. 
Sonies, B. C., Weiffenbach, J. , Atkinson, J.C., Brahim, J., Macynski, A., & Fox, P. C. 
( 1987). Clinical examination of motor and sensory functions of the adult oral 
cavity. Dysphagia, l, 178-186. 
Statistical Sciences. (I 991). S-Plus. version 3.0 [Computer program]. Seattle, WA: 
Statistical Sciences. 
Stetson, R. H. ( 1951 ). Motor phonetics: A study of speech movements in action. 
Amsterdam: North-Holland. 
Stone, M. ( 1981 ). Evidence for a rhythm pattern in speech production: Observations of 
jaw movement. Journal of Phonetics, .2, 109-120. 
Stone, M. (1990). A three-dimensional model of tongue movement based on ultrasound 
and x-ray microbeam data. Journal of the Acoustical Society of America, 87, 
2207-2217. 
Stone, M. (1991). Toward a model of three-dimensional tongue movement. Journal of 
Phonetics, 19, 309-320. 
Stone, M. (1995). How the tongue takes advantage of the palate during speech. In F. Bell-
Berti & L. J. Raphael (Eds.), Producing speech: Contemporary issues (pp. 143-
153). New York: American Institute of Physics. 
195 
Stone, M. ( 1996). Instrumentation for the study of speech physiology. In N. J. Lass (Ed.), 
Principles of experimental phonetics (pp. 495-524). St. Louis: Mosby. 
Stone, M., Faber, A., Cordaro, M. (1991). Cross-sectional tongue movement and 
tongue-palate movement patterns in [s] and['] syllables. Proceedings of the Xllth 
International Congress of the Phonetic Sciences, 2., 354-357. 
Stone, M., & Lundberg, A (1996). Three-dimensional tongue surface shapes of English 
consonants and vowels. Journal of the Acoustical Society of America, 99, 3728-
3737. 
Stone, M., & Shawker, T. H. (1986). An ultrasound examination of tongue movement 
during swallowing. Dysphagia, l, 78-83. 
Stone, M. , Sonies, B. C., Shawker, T. H., Weiss, G., Nadel, L. (I 983). Analysis ofreal-
time ultrasound images of tongue configuration using a grid-digitizing system. 
Journal of Phonetics, 11, 207-218. 
Storey, A. T. (1968). Laryngeal initiation of swallowing. Experimental Neurology, 20, 
359-365. 
Sumi, T. (1964). Neuronal mechanisms in swallowing. Pfliigers Archiv, 278, 467-477. 
Sumi, T. ( 1969a). Some properties of cortically-evoked swallowing and chewing in 
rabbits. Brain Research, .Ll., 107-120. 
Sumi, T. ( 1969b ). Synaptic potentials of hypoglossal motoneurons and their relation to 
reflex deglutition. Japanese Journal of Physiology, 19, 68-79. 
Sumi, T. ( 1970a). Activity in single hypoglossal fibers during cortically induced 
swallowing and chewing in rabbits. Pfliigers Archiv, 314, 329-346. 
Sumi, T. ( 1970b ). Changes of hypoglossal nerve activity during inhibition of chewing and 
swallowing by lingual nerve stimulation. Pfliigers Archiv, 317, 303-309. 
Sumi, T. ( 1971 ). Modification of cortically evoked rhythmic chewing and swallowing 
from mi db rain and pons. Japanese Journal of Physiology, 2..1, 489-506. 
Sumi, T. (I 972a). Reticular ascending activation of frontal cortical neurons in rabbits, with 
special reference to the regulation of deglutition. Brain Research, 46, 43-54. 
Sumi, T. (1972b). Role of the pontine reticular formation in the neural organization of 
deglutition. Japanese Journal of Physiology, 22, 295-314. 
196 
Suzuki, N . (1989). Application ofEPG to cleft palate and glossectomee cases. Clinical 
Linguistics & Phonetics, J, 127-136. 
Sweazey, R. D ., & Bradley, R. M. (1986) . Central connections of the lingual-tonsillar 
branch of the glossopharyngeal nerve and the superior laryngeal nerve in lamb. 
Journal of Comparative Neurology, 245, 471-482. 
Sze, C .-F. (1996) . Image processing package [Unpublished computer program]. 
Tatton, W. G. , & Bruce, I . C. ( 1981 ). Comment: A schema for the interactions between 
motor programs and sensory input. Canadian Journal of Physiology and 
Pharmacology, 59, 691-699. 
Thexton, A. J . (1992). Mastication and swallowing: An overview. British Dental Journal, 
173, 197-206. 
Tomura, Y., Ide, Y., & Kamijo, Y. (1981 ). Studies on the morphological changes of the 
tongue movements during mastication by x-ray TV cinematography. In Y. 
Kawamura, & R. Dubner (Eds.), Oral-facial sensory and motor functions (pp. 45-
52). Tokyo: Quintessence. 
Tuller, B. , Harris, K. S., & Kelso, J. A. S. (I 982). Stress and rate: Differential 
transformation of articulation. Journal of the Acoustical Society of America, 71, 
1534-1543 . 
Vanek, A. W ., & Diamant, N . E. (1987). Responses of the human esophagus to paired 
swallows. Gastroenterology, 92, 643-650. 
Veis, S . L., & Logemann, J. A. (1985.) Swallowing disorders in persons with 
cerebrovascular accident. Archives of Physical Medicine and Rehabilitation, 66, 
372-375 . 
Vitti, M ., Basmajian, J. V., Ouellette, P. L., Mitchell, D. L., Eastman, W. P., & Seaborn, 
R. D. (1975). Electromyographic investigations of the tongue and circumoral 
muscular sl ing with fine-wire electrodes. Journal of Dental Research, 54, 844-849. 
Walberg, F . (1957) . Do the motor nuclei of the cranial nerves receive corticofugal fibers? 
An experimental study in the cat. Brain, 80, 597-605. 
Walker, A. E., & Green, H. D. (I 938). Electrical excitability of the motor face area: A 
comparative study in primates. Journal ofNeurophysiology, l, 152-165. 
197 
Weber, F., Woolridge, M . w., & Baum, J. D. (1986). An ultrasonographic study of the 
organisation of sucking and swallowing by newborn infants. Developmental 
Medicine & Child Neurology, 28, 19-24. 
Weerasuriya, A., Bieger, D ., & Hockman, c. H. (1979). Basal forebrain facilitation of 
reflex swallowing in the cat. Brain Research, 174, 119-133 . 
Weerasuriya, A., Bieger, D ., & Hockman, c. H. (1980). Interaction between primary 
afferent nerves in the elicitation of reflex swallowing. American Journal of 
Physiology, 239, R407-R414. 
Wein, ~ -, B~ckler, R. , Klajman, s. (1991). Temporal reconstruction of sonographic 
1magmg of disturbed tongue movements. Dysphagia, Q, 135-139. 
Wheeler, R. L., Logemann, J. A. , & Rosen, M. S. (1980). Maxillary reshaping prostheses: 
E~ectiveness in improving speech and swallowing of postsurgical oral cancer 
patients. Journal of Prosthetic Dentistry, 43, 313-319. 
Wieneke, G ., Janssen, P ., & Belderbos, H. (1987). The influence of speaking rate on the 
duration of jaw movements. Journal of Phonetics, .Ll., 111-126. 
Wilson, J . A , Pryde, A., MacIntyre, C. C. A., & Heading, R. C. (1989). Normal 
pharyngoesophageal motility: A study of 50 healthy subjects. Digestive Diseases 
and Sciences, 34, 1590-1599. 
Wright, S., & Kerswill, P . ( 1989). Electropalatography in the analysis of connected speech 
processes. Clinical Linguistics & Phonetics, J., 49-57.