The Nature and Functioning of Repetition in Stuttering

The Nature and Functioning of Repetition in Stuttering
American Speech-LanguageHearing Association Convention,
San Diego, CA,
November 17-19, 2011
Ai Leen Choo, Julie A. Hengst and Nicoline G. Ambrose
Department of Speech and Hearing Science, University of Illinois at Urbana-Champaign
For example, measurements of repetition particularly the number of
repetition units and types of repetition including part-word, single
syllable and phrase repetitions are measures that have been used to
distinguish abnormal disfluency, that is stuttering (Ambrose & Yairi,
1995, 1999). Repetition in stuttering has also been utilized as a
measurement for the assessment of severity (Ambrose & Yairi, 1995,
1999; Riley, 1980). An increase in these types of repetition would
indicate greater severity of stuttering.
Other disfluencies (OD)
Other conversational
repetition (OCR)
Conversational repetition
Repetition is a pattern that is ubiquitous in everyday behavior and
communication (Johnstone, 1994; Schegloff, 1979; Tannen, 2007). In
everyday conversation, repetition is pervasive and a major resource in
communication (Hengst, Duff & Dettmer, 2010; Johnstone, 1994).
Specifically, repetition functions in production, comprehension,
connection and interaction to maintain fluency and connection among
interlocutors and coherence within discourse (Norrick, 1987; Tannen,
2007). Repeating an utterance can also stall or slow a conversation to
an appropriate speed, allowing the speaker to hold the floor while
formulating the next utterance (Tannen, 2007). It also supports
meaning and poetics (Agha, 2007; Lakoff & Johnson, 1980).
Repetition can come in many forms: self vs. other, verbatim vs.
paraphrase, and delayed vs. immediate (Johnstone, 1994; Norrick,
1987; Tannen, 2007).
Goals of the Current Study
This study examines stuttering disfluencies and conversational
repetitions in 10-minute conversational interviews obtained on 4
adult males with mild to moderate-severe stuttering and 4 age-sexmatched fluent peers.
Preliminary Research Questions:
1) What are the overall and relative amounts of repetition
(stuttering and conversational types of repetition) used by pairs
who are and are not managing stuttering?
2) Do pairs managing stuttering show similar patterns in the types
of repetition (e.g., who was being repeated –self or other; time
frame of repetition – immediate-within-turn, immediate-across turn
or delayed; similarity to previous utterance – verbatim or
paraphrase) compared to those who are not?
Method
This current study presents a new discourse analysis of data
previously collected for a broader study of stuttering (conducted at
the Illinois International Stuttering Research Program-funded by the
RO1-DC01529 grant). As part of the original study, clinicians
interviewed all participants in a clinical setting prior to their
participation in various experiments.
•
•
•
•
Sub codes
Part word
Single syllable
Disrhythmic phonation
Phrase
Revision
Example
Bu-bu-but
And-an-and
Pa—pathologist
I like to – I like to
Hearing
impaired…impairment
Interjection
I am also taking uhm
neurology and uhm…
Other instances of repetition
that could be self or other,
immediate or delayed,
verbatim or paraphrase.
I remember this room now
actually…but I remember it
being a lot larger than it is.
SLD: are identified as any immediate self repetition of part-word,
single-syllable word and disrhythmic phonation.
OD: are identified as any immediate self repetition of phrase,
revision and interjection that could be verbatim or paraphrased.
OCR: are identified as self or other repetition of words or phrases
that could be verbatim or paraphrase, immediate or delayed.
Moments of repetitions can be double coded. The repetition that
was most often double coded was as OD and OCR. For example,
phrase repetitions (OD) may also be coded as verbatim
repetitions (OCR).
Subcoding of moments of repetition
• All instances of repetition could further be coded as self- or other-;
immediate-within-turn, immediate-across turn or delayed; and
verbatim or paraphrase according to the preceding utterance.
• Self-repetition occurred when the speaker repeated himself while
other-repetition was repeating what someone else has said.
• Delayed repetitions were separated from the preceding utterance
by some intervening material while immediate repetitions were not
separated by any intervening material.
• Verbatim repetitions were repetitions that were exact to the
preceding utterance but paraphrased repetition deviated from the
preceding material in some manner.
Percent repetition
Percent repetition
Results
40%
IM-AT
IM-AT
IM-WT
Amount of Repetition:
• 5713 instances of repetition identified across whole dataset
• 2459 in 4 AWS sessions & 3254 in 4 FC sessions
Delayed
Delayed
IM-AT
IM-AT
IM-WT
IM-WT
FC
Clinician
60%
40%
20%
0%
0%
AWS
Clinician
Delayed repetitions occurred more frequently than immediate-withinturn or immediate-across-turn repetitions in AWS, FC and their
clinicians.
Percent of Verbatim and Paraphrase repetitions
by AWS and their clinicians
Identifying moments of repetition
• Drawing on the stuttering literature (Yairi & Ambrose, 1999) and
Tannen (2007) we defined 3 broad categories of repetition and
used the types of repetition to obtain a total count of repetition in
the sample.
Types of repetition
Stuttering like disfluencies
(SLD)
60%
IM-WT
Percent of Verbatim and Paraphrase repetitions
by FC and their clinicians
100%
100%
Paraphrase
80%
80%
Paraphrase
Percent repetition
Percent repetition
The communication disorders perspective provides a very specific
view of repetition which has been instrumental in measuring and
diagnosing the disorder. Repetitions of sounds, words, syllables and
phrases are the hallmark of stuttering and as such are key features
that have been used to identify and describe it.
Delayed
Delayed
20%
60%
40%
Verbatim
Paraphrase
Paraphrase
Verbatim
Verbatim
FC
Clinician
60%
40%
Verbatim
20%
20%
0%
0%
AWS
Clinician
Verbatim repetitions were more common than paraphrase repetitions
in AWS and their clinicians. However, instances of verbatim and
paraphrase repetitions were almost similar in FC.
Discussion
Patterns in the Types of Repetition:
• 389.3 mean number of repetition per thousand words in AWS
• 298.6 mean number of repetition per thousand words in clinicians
interviewing AWS
• 296.4 mean number of repetition per thousand words in FC
• 403.2 mean number of repetition per thousand words in clinicians
interviewing FC
Mean repetition per thousand words
Mean repetition per thousand words
400
400
300
300
# of words
Repetition in stuttering
Transcribing
• All eight conversational samples were fully transcribed using
consensus procedures (see Hengst, 2001, 2003).
80%
80%
200
100
200
100
0
AWS
0
Clinician
FC
Clinician
The pattern of mean repetition per thousand words was reversed in
AWS and FC.
Examples of Coding on Transcripts:
Percent of Self- and Other-repetition featured by
AWS and their clinicians
Percent of Self- and Other-repetition featured by FC
and their clinicians
100%
100%
Other
Coding for other SLD and OD
Other
80%
60%
40%
Self
Other
80%
60%
40%
Self
Self
Self
20%
20%
0%
0%
AWS
The present study drew upon Tannen’s work and approach to the
study of stuttering to interpret the interactions of pairs managing
stuttering. Specifically, the aim of this pilot study was to compare
the overall and relative amounts of repetition, and patterns in the
different types of repetition in functional communication systems
managing and not managing stuttering.
Results of this pilot study suggest that:
• The functional communication system may be strategically
managing different rates and types of repetitions including those
regarded as disfluencies. For example, the higher mean rates of
repetition per thousand words in AWS was complemented by lower
rates of repetition in their clinicians/conversation partners. Similarly,
FC showed a reverse pattern where their lower instances of
repetition were complemented by their clinicians’ higher rates of
repetition.
• The overall and relative amounts of repetition indicate the functional
communication system is able to accommodate for deviations or
variations. Although AWS featured more self-repetition (which
included SLD, OD and some instances of OCR) than FC, both
groups showed higher rates of self-repetition compared to otherrepetition.
Acknowledgments
Other
Percent repetition
By integrating the approaches of interactional sociolinguistics and
communication disorders, the present investigation will be able to
provide insight into how stuttering impacts communication strategies
when the functional system is managing stuttering. Specifically, the
goal of the pilot study is to investigate how stuttering intermingles with
other forms of planned and unplanned conversational repetition.
Discourse Analysis of Repetition
Data transcription and coding was completed by a team of five
undergraduate research assistants and the primary author. Consensus
transcribing and coding involved at least three passes through each
session, with a final consensus pass used to resolve discrepancies.
100%
100%
# of words
Repetition in the flow of conversational speech is a complex
phenomenon that includes a wide range of verbal and nonverbal
behaviors. Traditionally, researchers in the field of communication
sciences and disorders (CSD) have focused on repetitions that signal
disordered systems and that disrupt the smooth flow of speechlanguage production. In contrast, communications researchers outside
of CSD have focused broadly on the routine ways that repetition
supports the smooth production of talk and the coordination of
interlocutors during interaction.
Percent repetition
Background
Coding for other conversational repetition
Data Set and Participants:
• 4 male adults who stutter (AWS) (mean age= 23.2 years,
stuttering severity= mild to moderate-severe).
• 4 male adults with normal fluency served as fluent comparisons
(FC) (mean age= 23.1 years).
• 10-minute conversational elicitation interview during which an SLP
ask the participant about his hobbies, school, travel and work.
Percent of Immediate-within-turn, Immediate-across-turn and
Delayed repetition featured by FC and their clinicians
Percent of Immediate-within-turn, Immediate-across-turn
and Delayed repetition featured by AWS and their clinicians
Clinician
FC
Clinician
Self-repetitions were more abundant than other repetition in AWS, FC
and their clinicians.
The authors would like to thank the participants for their time and
researchers who were involved in the transcription and data
analysis including Elise Brucks, Emily Freeman, Allison Hilger,
Wendy Love and Keren Wasserman. This study was made
possible with the RO1-DC01529 grant, P.I. Drs. Ehud Yairi and
Nicoline Ambrose.