INT J LANG COMMUN DISORD, JANUARY–FEBRUARY 2014,

VOL. 49, NO. 1, 139–147

Research Report
Stuttering inhibition via altered auditory feedback during scripted telephone
conversations
Daniel Hudock† and Joseph Kalinowski‡
†Department of Communication Sciences and Disorders, Division of Health Sciences, Idaho State University,
Pocatello, ID, USA
‡Department of Communication Sciences and Disorders, College of Allied Health Sciences, East Carolina University,
Greenville, NC, USA
(Received November 2012; accepted July 2013)

Abstract
Background: Overt stuttering is inhibited by approximately 80% when people who stutter read aloud as they hear
an altered form of their speech feedback to them. However, levels of stuttering inhibition vary from 60% to 100%
depending on speaking situation and signal presentation. For example, binaural presentations of delayed auditory
feedback (DAF) and frequency-altered feedback (FAF) have been shown to reduce stuttering by approximately
57% during scripted telephone conversations.
Aims: To examine stuttering frequency under monaural auditory feedback with one combination of DAF with FAF
(COMBO-2) and two combinations of DAF with FAF (COMBO-4) during scripted telephone conversations.
Methods & Procedures: Nine adult participants who stutter called 15 local businesses during scripted telephone
conversations; each condition consisted of five randomized telephone calls. Conditions consisted of (1) baseline
(i.e. non-altered feedback), (2) COMBO-2 (i.e. 50-ms delay with a half octave spectral shift up), and (3) COMBO-
4 (i.e. 200-ms delay and a half octave spectral shift down in addition to the COMBO-2). Participants wore a
supra-aural headset with a dynamic condenser microphone while holding a receiver to their contralateral ear when
making telephone calls.
Outcomes & Results: Stuttering was significantly reduced during both altered auditory feedback (AAF) conditions
by approximately 65%. Furthermore, a greater reduction in stuttering was revealed during the COMBO with four
effects (74%) as compared with the COMBO with two effects (63%).
Conclusions & Implications: Results from the current study support prior research reporting decreased stuttering
under AAF during scripted telephone conversations. Findings that stuttering was significantly reduced to a greater
extent under the COMBO with four effects condition suggest that second signals reduce stuttering along a
continuum. Additionally, findings support prior research results of decreased stuttering frequency under AAF
during hierarchically difficult speaking situations. Clinical application of these findings may be that people who
stutter can use specific software or smartphone applications that produce second speech signals to inhibit stuttering
frequency effectively during telephone conversations.

Keywords: stutter, stuttering, fluency, stammer, stammering, telephone.

What this paper adds?

What is already known on the subject?
Previous research has shown that stuttering can be immediately and effectively reduced during the perception of
second speech signals. This reduction in stuttering by hearing a second speaker, or one’s own voice altered then
feedback to them as they speak varies in level of reduction depending on speaking situation and signal presentation.
Perception of second speech signals during hierarchically difficult speaking situations maintains significant reductions
in overt stuttering behaviours. Second signals that are more speech-like with additional gestural information offer
greater reductions in stuttering frequency.

Address correspondence to: Daniel Hudock, Department of Communication Sciences and Disorders, Division of Health Sciences, Idaho State
University, 921 South 8th Avenue, Mail Stop 8116, Pocatello, ID 83209–8116, USA; e-mail: Hudock@isu.edu
International Journal of Language & Communication Disorders
ISSN 1368-2822 print/ISSN 1460-6984 online  C 2013 Royal College of Speech and Language Therapists

DOI: 10.1111/1460-6984.12053
140
What this paper adds?
The current study demonstrated that monaural presentation of AAF significantly reduced stuttering frequency during
hierarchically difficult situations. This is the first time that combinations of altered feedback have been shown to
offer differential effects from single signal presentations. Multiple combinations of AAF exhibited more robust
stuttering reductions than one combination of DAF with FAF. Results indicate that reductions in stuttering during
the perception of second speech signals occur along a continuum that is dependent on situational and signal factors.
Introduction
Stuttering is a speech disorder overtly characterized by
syllable repetitions, phoneme prolongations and postu-
ral fixations. These overt speech disruptions are immedi-
ately and dramatically reduced from 60% to 100% when
people who stutter (PWS) perceive a second speech sig-
nal that is presented in conjunction with their own
speech output (Hargrave et al. 1994, Hudock et al. 2011,
Kalinowski et al. 1993, Macleod et al. 1995). Second
speech signals are speech signals that are presented to the
speaker during their on-going speech productions. For
example, when a second speaker verbally reads or recites
content with a person who stutters (PWS), stuttering
is typically reduced to the extent of being non-existent.
Another form of second speech signals is using one’s
own speech to produce a second altered signal using a
digital signal processor that feeds the generated signal
back to the speaker. Altered auditory feedback (AAF)
uses one’s own speech to produce an altered speech sig-
nal by feeding the speaker’s production into a digital
signal processor that generates a second temporally and
or spectrally altered signal. Two examples of these sig-
nals are delayed auditory feedback (DAF) and frequency
altered feedback (FAF). DAF is a delayed representa-
tion of one’s speech that is fed back to them; FAF is
a spectrally shifted representation with minimal delay.
DAF (Kalinowski et al. 1993, Kalinowski and Stuart
1996) and FAF (Armson et al. 1997, Howell et al. 1987,
Kalinowski et al. 1993, Macleod et al. 1995, Stuart et al.
1996) are effective at reducing stuttering frequency by
between 60% and 80%. Signals that are self-generated
and externally altered offer the benefit of not relying on
a second speaker to be present.
Researchers have extensively examined second sig-
nals that reduce stuttering. Some of these signals in-
clude masking noise (Kalinowski et al. 1993), pure-tones
(Saltuklaroglu and Kalinowski 2006), DAF (Kalinowski
et al. 1993, Kalinowski and Stuart 1996), FAF (Armson
et al. 1997, Howell et al. 1987, Kalinowski et al. 1993,
Macleod et al. 1995, Stuart et al. 1996), and combined
DAF and FAF (Hargrave et al. 1994). Stuttering is typ-
ically reduced by approximately 40% when PWS per-
ceive masking noise and pure-tones, and by 80% when
perceiving signals that are more speech-like. Researchers
Daniel Hudock and Joseph Kalinowski
have reported statistically equivalent reductions in stut-
tering frequencies during oral reading conditions when
using DAF delays of 25, 50 and 75 ms (Kalinowski and
Stuart 1996), and 50, 200 and 400 ms (Dayalu 2004).
They have also reported similar results with FAF spec-
tral shifts from one-half octave downward to one-half
octave upward in one-quarter octave increments (Stu-
art et al. 1996) and one full octave shift downward to
one full octave shift upward in half octave increments
(Hargrave et al. 1994). Additionally, reductions in stut-
tering frequencies have been noted in single DAF or FAF
presentations as compared with combined DAF and
FAF (Macleod et al. 1995). During oral reading, stut-
tering frequency is typically reduced by between 60%
and 80% with DAF, FAF or combined DAF and FAF
conditions, across the specific tested delays or frequency
shifts.
Support from previously mentioned studies and ex-
amination of other factors and situations using second
speech signals have lead to clinical application of sec-
ond speech signals for reducing stuttering. Researchers
have developed portable devices that offer binaural and
monaural DAF, FAF or DAF with FAF. It should be
noted that binaural signal presentation is more effec-
tive at reducing stuttering than monaural presentation
(Stuart et al. 1997). To the best of the authors’ knowl-
edge, this is the only study that specifically examined
AAF in binaural and monaural presentations. The au-
thors did report a significant difference between bin-
aural versus monaural presentations, but for practical
purposes of hearing listeners, reducing environmental
distraction and decreasing the conspicuousness of de-
vices, the assistive products are typically implemented
with monaural presentations. The portability of these
devices offers many opportunities for clinical applica-
tion across many settings with increased fluency. Re-
searchers have employed second speech signals dur-
ing monologues, conversations, verbal presentations and
during telephone conversations. For example, Armson
et al. (1997) reported a 74% reduction in stuttering
frequency with FAF using a one-half octave downward
shift during oral reading to audiences of two, four and
15 people, with no effect for audience size. Additionally,
Zimmerman et al. (1997) examined the effect of
DAF and FAF on stuttering frequency during scripted
Stuttering inhibition during scripted telephone conversations
telephone conversations. Stuttering frequency was re-
duced by approximately 57% during AAF conditions
and did not reveal significant difference between DAF
and FAF.
For PWS, the telephone is often described as one
of their most feared and avoided situations (Georgieva
1994, Leith and Timmons 1983). Speaking on the tele-
phone to a naı̈ve listener is typically ranked highest on
their hierarchy of difficult speaking situations (Blood-
stein and Bernstein-Ratner 2007). Stuttering frequency
has even been shown to increase during this difficult
situation relative to conversations held within clinical
settings (Ladouceur et al. 1982). To date, however, only
a limited number of studies have examined stuttering
frequency during telephone conversations. The major-
ity of these studies used clinicians and researchers, and
briefed listeners or family members during timed con-
versational exchanges (Craig et al. 1996, Stager et al.
1995). To evaluate stuttering during telephone conver-
sations, some researchers may call clients or have clients
call family members or staff from clinical settings (Craig
et al. 1996, Stager et al. 1995). However, using these
methods to evaluate true fluency levels during gener-
alization may be problematic. If the PWS speaks to
someone who is familiar with stuttering, the listener
may not portray typical responses during the exchange.
In addition, if participants remain in comfortable envi-
ronments where they may have performed similar tasks
numerous times, their level of fluency may not be a true
representation. Procedures may be further adjudicated
by not having participants use scripted texts, as some of
the most frequently employed compensatory strategies
by PWS are circumlocutions and substitutions to avoid
saying words that they fear stuttering on. Such com-
pensatory strategies often occur to the extent of being
almost reflexively produced to avoid saying words with
anticipated difficultly. By substituting words or altering
word order, PWS may exhibit the illusion of being more
fluent than they actually are.
Only a limited number of studies have examined
stuttering episodes of PWS during telephone con-
versations with unfamiliar listeners (Bray and James
2009, Chambers 2009, Zimmerman et al. 1997).
Zimmerman et al. (1997) had nine adult PWS call busi-
nesses on a binaural headset while reading scripted texts
during baseline, DAF and FAF conditions. DAF was set
at a 50-ms delay; FAF had a downward shift of one-
half octave. They reported 60% and 55% reductions for
DAF and FAF, respectively, with no statistically signifi-
cant differences between DAF and FAF. Bray and James
(2009) examined stuttering frequencies and attitudes of
participants who stuttered during two baseline phases
and one experimental phase of a study. However, there
were some limitations with their study. For example, the
authors claimed that only five participants were exam-
141
ined during the study, but reported participant numbers
ranging from one to eight. The researchers did not re-
port why the final number of participants was five and
not eight.
Participants in Bray and James (2009) were asked to
make three telephone calls (i.e. one to a person they were
comfortable speaking to, one to a person they were not
comfortable speaking to, and one to a retailer asking for
a brochure) over the span of 1 month, and to record the
three telephone calls. Recorded telephone calls during
the first month created the first baseline, and recorded
telephone calls made during the second month were
the second baseline. Two weeks after the second base-
line, researchers visited the participants to train them on
using a telephone feedback device (VA609 TAD) that
offered a choice between two separate combinations of
DAF with FAF. The device was a handheld telephone
with monaural signal presentation. The first combina-
tion presented a 56-ms delay with a 304 Hz upward
shift in frequency, while the second preset had a 90-
ms delay with a 530 Hz upward shift in frequency. If
the researchers mean that the spectrum was shifted up
by the listed frequency amount, it is similar to one-
quarter octave and half octave shift up, respectively. It
was not reported what preset the five (or eight) partic-
ipants used during the telephone conversations. Bray
and James (2009) only reported the descriptive statistics
for average stuttering collapsed across telephone calls;
8.62% for baseline 1, 8.42% for baseline 2, and 4.82%
for the experimental condition. Findings from the study
indicated a 50% reduction in stuttering. However, due
to the concerns listed above, results should be considered
with caution.
The current study was designed to explore differ-
ences between one combination of DAF with FAF, and
two combinations of DAF with FAF. Although signifi-
cant differences between DAF, FAF, and combined DAF
with FAF have not been consistently reported in the
literature, both authors experience self-reported addi-
tional benefits when using multiple combinations of
DAF with FAF as compared with single effects or one
combination of effects. Also, participants were required
to place telephone calls to unfamiliar persons using
scripted conversations. Similar procedures and scripts as
Zimmerman et al. (1997) were used to reduce partic-
ipants’ use of avoidances, circumlocutions and substi-
tutions. It was hypothesized that differences from base-
line to both conditions and between one signal with
two effects to one signal with four effects would be re-
vealed (similar to the authors’ experiences) when using
a difficult speaking situation (scripted text to unfamiliar
communication partners) to elicit increased frequency
of stuttering. Additionally, the use of monaural signal
presentations in the current study enabled researchers
to compare the results with Bray and James (2009)
142
and to a lesser extent with Zimmerman et al. (1997),
who used binaural presentation. The aim of the current
study was to examine the effect of monaural presen-
tations from one combination of DAF with FAF and
two combinations of DAF with FAF on stuttering fre-
quency during scripted telephone conversations. It was
hypothesized that stuttering frequency would signifi-
cantly decrease during both conditions, and that differ-
ential effects would be noted when comparing the two
effect combinations and four effect combinations.
Method
Participants
Nine native English speaking adults who stutter includ-
ing eight males and one female, mean age of 35.1 years
(SE = 5.3, range = 21–72), with no self-reported history
of concomitant speech, language, cognitive, uncorrected
visual or hearing deficits participated in the study. Par-
ticipants met inclusionary criteria of greater than 5%
stuttering during informal assessment of spontaneous
speech. Prior to experimental procedures, informed con-
sent (approved by the University and Medical Center
Institutional Review Board, East Carolina University)
was obtained from all participants.
Instrumentation and stimuli
Scripts from Zimmerman et al. (1997) were utilized.
Some scripts were altered in order to increase so-
cial relevance for 2012 (see appendix A). Condi-
tion sequences and telephone scripts were random-
ized by assigning individual numerical values to each
then randomly generating number sequences using
http://www.randomizer.org/. Participants spoke into a
Sennheiser HMD 281 Pro 64 Ohm monaural supra-
aural headset with a unidirectional dynamic microphone
(frequency ranges = 50–13 500 Hz) positioned 2.5 cm
directly in front of their mouths. Downstream audio was
received by a Radial JS-3 XLR splitter that delivered the
throughput, unamplified, stereo signal via XLR to 3.5
mm cable into an iMac 10.1 Intel Duo Core computer,
which used Audacity 1.3.14 Beta software to capture
the audio signal of the speaker at 44.1 kHz. The Radial
JS-3 delivered two amplified stereo signals into a Mackie
Micro Series 1202 twelve-channel mixer via XLR con-
nections. From there two lineout stereo signals went
from the mixer into two separate left mono RCA inputs
in Yamaha DSP-1 digital signal processors (DSP). The
first DSP generated a second signal with a 50-ms delay
with positive half-octave shift up. The second DSP gen-
erated a second signal with a 200-ms delay with negative
half-octave shift down. Mono-unbalanced signals from
the two DSPs were then fed into a Rolls Mix Max RM81
Daniel Hudock and Joseph Kalinowski
mixer via line inputs. The mixed signal was then sent to
a Rolls RA68b headphone-amplifying mixer where the
line stereo plug for the Sennheiser headset was inserted.
Frequency ranges for the broadcast quality headset were
8–25 000 Hz.
Specific delay and frequency settings for the com-
bination with two effects (COMBO-2) condition were
DAF – 50 ms with FAF – plus one-half octave. Set-
tings for the COMBO with four effects (COMBO-
4) condition were DAF – 200 ms with FAF minus
one-half octave in addition to the COMBO-2 settings.
These settings were chosen in consideration of prior re-
search (Hargrave et al. 1994, Macleod et al. 1995, Stuart
et al. 1997). Signals were markedly different, yet in-
dividually perceptible via self-report from participants.
Hargrave et al. (1994) reported equivocal reductions
in stuttering frequency between half and one octave
positive and negative frequency shifts during reading
conditions. Similarly, under DAF, researchers reported
equivocal reductions in stuttering frequency between
50 and 400-ms delays (Kalinowski and Stuart 1996).
Macleod et al. (1995) reported equivocal reductions in
stuttering frequency when comparing DAF with a 50-
ms delay, FAF minus one-half octave shift and a combi-
nation of these two signals. Finally, increased reductions
have been noted with binaural compared with monaural
signal presentations (Stuart et al. 1997). The preceding
studies all report approximately between 70% and 80%
reductions in stuttering frequency during oral reading
tasks. Furthermore, general experimental designs for the
current study were similar to Zimmerman et al. (1997)
who used binaural feedback conditions of 50 ms DAF,
one-half octave negative shift down FAF and baseline
non-altered feedback during scripted telephone con-
versations. It should be noted that current procedures
with monaural signal presentations enabled researchers
to capture unaltered downstream audio from the partic-
ipants without recording the speaker on the other end
of the phone.
Procedures
The researcher verbally briefed participants prior to sign-
ing informed consent documents. Prior to beginning the
experiment, participants were requested not to use flu-
ency enhancing strategies and to talk in their natural way
of speaking during all conditions. Volume was adjusted
to the participants’ designated most comfortable listen-
ing level while being able to perceive both combinations
of signals at approximately the same loudness. Partic-
ipants called a total of 15 local businesses requesting
information and items (for a complete list of scripts, see
appendix A). During each condition participants made
five consecutive calls. In order to reduce any possible
carryover effects researchers had participants produce
Stuttering inhibition during scripted telephone conversations 143
Table 1. Raw data with mean and standard errors (SE) for total spoken and total stuttered syllables as a function of AAF condition

NAF COMBO-2 COMBO-4

Spoken Stuttered Spoken Stuttered Spoken Stuttered
Participant syllables syllables syllables syllables syllables syllables
1 233 41 231 10 351 19
2 222 32 221 12 232 4
3 422 125 243 37 279 39
4 218 67 255 33 255 33
5 248 82 236 39 206 20
6 236 54 251 17 246 13
7 264 84 292 36 363 29
8 188 33 155 5 166 1
9 352 104 508 39 407 15
Means 264.7 69.1 265.8 25.3 278.3 19.2
SE 24.8 10.8 32.6 4.7 26.5 4.2
Note: NAF, non-altered feedback; COMBO-2 = DAF (50-ms delay) with FAF (half octave shift up); COMBO-4 = COMBO-2 and DAF (200-ms delay) with FAF (half octave shift
down).

2-min spontaneous conversation samples without audi-
tory feedback between the three conditions.
Speech rate analysis
As longer DAF settings have been demonstrated to
decrease speech rates, researchers calculated partici-
pants’ speech rates (i.e. syllables per second) in order to
account for the fact that a 200-ms delay was only used
during the COMBO-4 condition. To obtain an accurate
speech rate calculation, researchers used only fluently
produced syllables that were at least two syllables distant
from any stuttering episode and contained a minimum
of three contiguous syllables. One recording did not
meet these criteria—in this instance, researchers used
a derived average from the participant’s other four
recordings during the same condition. Mean syllable
rates and their standard errors were 4.6 (SE = 0.5),
4.1 (SE = 0.5), and 4.4 (SE = 0.6) syllables/s for the
non-altered feedback, COMBO-2 and COMBO-4
conditions, respectively. A one-factor analysis of
variance (ANOVA) was utilized to examine differences
in speech rate as a function of condition. There were
no statistically significant differences between syllable
rates [F(2, 16) = 3.29, p = 0.06, η2p = 0.29].
Results
Prior to analysing stuttering frequencies, the re-
searcher orthographically transcribed participants’
audio-recorded telephone conversations to obtain total
syllable production counts. This was done since conver-
sations slightly varied due to the responses of the clerks
and employees. The researcher then analysed conversa-
tions for stuttered syllables and calculated proportional
values for each condition by dividing the number of
stuttered syllables by the total spoken syllables per con-
dition. Stuttering was defined as part-word repetitions,
phoneme prolongations and postural fixations. Intra-
and inter-rater reliabilities were obtained from a ran-
domized 10% of the data. A certified speech–language
pathologist trained in stuttering evaluation performed
the inter-rater reliability using Cohen’s kappa syllable-
by-syllable agreements. Kappa values for intra-rater anal-
ysis were 0.87 and were 0.54 for inter-rater analysis. Val-
ues above 0.41 represent moderate agreement and values
greater than 0.75 represent excellent agreement (Viera
and Garrett 2005).
The numbers of spoken and stuttered syllables as
a function of feedback condition for each participant
are presented in table 1. The proportions of spoken
and stuttered syllables as a function of feedback condi-
tion for each participant are presented in figure 1 and
table 2. Prior to inferential statistical analysis partici-
pants’ proportion of stuttering values were transformed
into arcsine units. This commonly used transforma-
tion reduces endpoint weighting of proportional data
sets (Viera and Garrett 2005). Two one-factor ANOVA
were utilized to examine differences in the proportion
of stuttered syllables as a function of condition and to
examine differences in speech rate as a function of con-
dition. A significant effect of condition was revealed
[F(2, 16) = 152.35, p < 0.0001, ηp2 = 0.95]. Post hoc
t-tests with Bonferroni corrections revealed differ-
ences between NAF and COMBO-2 (p < 0.05) and
COMBO-2 to COMBO-4 (p < 0.05). Therefore, both
COMBO-2 and COMBO-4 conditions had signifi-
cantly lower proportions of stuttered syllables than the
non-altered feedback condition. The COMBO-4 con-
dition had significantly lower proportions of stuttered
syllables than the COMBO-2 condition.
144 Daniel Hudock and Joseph Kalinowski
significantly more effective at reducing stuttering than
the COMBO-2 condition.
Findings that stuttering was reduced during both
AAF conditions supports prior research results of sig-
nificantly reduced stuttering under AAF during difficult
speaking situations (Armson et al. 1997, Bray and James
2009, Zimmerman et al. 1997). Interestingly, the level
that AAF reduces stuttering appears to vary by speaking
situation, and signal presentation. The monaural presen-
tation of COMBO-4 reduced stuttering to the greatest
extent and COMBO-2 reduced stuttering to greater ex-
tents as compared with Zimmerman et al. (1997) or
Bray and James (2009). Binaural signal presentation
in Zimmerman et al. (1997) reported 60% and 55%
reductions in stuttering frequency for DAF and FAF,
respectively. Bray and James (2009) reported approxi-
mately 50% reductions using monaural presentations
of DAF with FAF. First, a comparison should be made
between the previous two studies. Zimmerman et al.
(1997) used binaural single-effect presentations and re-
Figure 1. Mean proportions stuttering with standard errors as a ported larger reductions than Bray and James (2009)
function of AAF condition. NAF = non-altered feedback, COMBO who used a monaural presentation with combined ef-
with two = DAF (50 ms delay) with FAF (1/2 octave shift up),
COMBO with four = COMBO with two signal, and DAF (200 ms
fects. The current study used a monaural presentation
delay) with FAF (1/2 octave shift down). Error bars represent plus with two and four combinations of effects and reported
one standard error of the mean. the largest reduction in stuttering. Although COMBO-
4 had the largest effect, COMBO-2 still reduced stutter-
ing to greater extents than either other study. Further-
Table 2. Proportion of stuttering as a function of condition
more, the current study and Zimmerman et al. (1997)
Participant NAF COMBO-2 COMBO-4 used scripts, whereas Bray and James (2009) did not.
1 0.18 0.04 0.05 Larger reductions in stuttering from Zimmerman et al.
2 0.14 0.05 0.01 (1997) and the current study as compared with Bray and
3 0.30 0.15 0.14 James (2009) may have occurred due to conversation re-
4 0.31 0.13 0.13 quirements. In other words, participants may have read
5 0.33 0.17 0.10
6 0.23 0.07 0.05
over the scripts prior to the clerk answering the tele-
7 0.32 0.12 0.08 phone or content construction may require additional
8 0.18 0.03 0.01 effort as compared with reading prepared texts.
9 0.30 0.08 0.07 Due to the findings that COMBO-4 reduced stut-
Means 0.254 0.093 0.069 tering to greater extents than COMBO-2, and that all
SE 0.024 0.017 0.016
AAF conditions reduced stuttering to greater extents
Note: NAF, non-altered feedback; COMBO-2 = DAF (50-ms delay) with FAF (half than reported in Zimmerman et al. (1997) or Bray and
octave shift up); COMBO-4 = COMBO-2 and DAF (200-ms delay) with FAF (half
octave shift down). James (2009), it may be interpreted that second sig-
nals effect stuttering along a continuum that depends
on situational and signal factors. For example, Armson
et al. (1997) reported 75% reductions in stuttering fre-
Discussion
quency under FAF during oral reading to audiences of
Generalfindings that AAF reduced stuttering frequency two, four and 15—with no effect for audience size.
during telephone conversations support prior research Similarly, researchers consistently report larger stutter-
results (Bray and James 2009, Zimmerman et al. 1997). ing frequency reductions under AAF during oral reading
In the current study, stuttering frequency was signif- as compared monologues or conversations (Antipova
icantly reduced an average of 72% during both al- et al. 2008, Armson et al. 2006, Armson and Keifte
tered feedback conditions relative to baseline—it was 2008, Armson and Stuart 1998). Under AAF stutter-
reduced 63% in the COMBO-2 and by 74% during ing is typically reduced between 70% and 80% during
the COMBO-4 conditions. This is the first study to oral reading and from 35% to 60% during monologues
reveal differences in stuttering between varied combi- and conversations. Further support for the interpre-
nations of AAF. Lastly, the COMBO-4 condition was tation of second signals reducing stuttering along a
Stuttering inhibition during scripted telephone conversations 145
continuum comes from the fact that white noise and Clinical implications
pure tones reduce stuttering less than signals that are
Results from this study demonstrate that perceiving
more speech like (Kalinowski et al. 1993, Saltuklaroglu
second speech signals with combinations of effects re-
and Kalinowski 2006). Second, non-speech signals typ-
duces stuttering frequency during hierarchically difficult
ically reduce stuttering between 20% and 40% during
communication situations. In addition to reductions in
oral reading—speech signals or signals that are speech-
stuttering under AAF, impact and improvement
like typically reduce stuttering by between 60% and
during difficult speaking situations may be indi-
100%. It may be interpreted that speech and speech-
cated in questionnaire responses from PWS, sug-
like signals possess temporal and spectral characteristics
gesting that they would use and recommend
that are similar to the speech signal and therefore reduce
using AAF during difficult speaking situations
stuttering to a greater extent than non-speech signals.
(Chambers 2009, Pollard et al. 2009, Stuart
Using Max’s inverse internal model of sensori-motor
et al. 2006, Unger et al. 2012). Similarly, many re-
control the current findings may be adequately explained
searchers note the benefits of AAF during difficult speak-
(Max, Guenther, Gracco, Ghosh & Wallace, 2004). Max
ing situations for PWS (Antipova et al. 2008, van Borsel
et al. (2004) suggest that stuttering may occur due to;
et al. 2003). Other clinical applications can be general-
1) an insufficient preparatory model (efference copy)
ized to PWS using software or smartphone applications
for motor output or that the efference model, sent for
that offer various combinations of second speech signal
comparison to feedback controllers, does not match ex-
effects for them to hear as the speak on the phone. It
ecuted actions, or 2) that PWS have an overreliance
is beneficial to have the altered signal only presented
on afferent, sensory, feedback for online corrections. As
to the PWS, as they would likely not use it if the lis-
noted by Max et al., perception of second speech au-
tener heard the altered signal as well. The altered signal
ditory or visual signals, such as DAF, FAF, or visual
could be presented directly to the earpiece of the de-
speech feedback (visual images of a mouth presented
vice, or transmitted via a Bluetooth capable device (i.e.
to speakers), may decrease stuttering by activating the
Bluetooth headset/ear piece or car interface). Use of this
auditory cortex for efference model comparisons sooner
technology would dramatically increase fluency during
than would occur during speech initiation without feed-
telephone conversations, especially if the technology al-
back. By providing better feedback and improving com-
lowed for various combinations of effects.
parisons, it should also correct deficits in preparatory,
Additionally, to extrapolate the findings with the
forward, internal models. Also noted by the Max et al.
help of previous research, PWS could record and play
(2004), neighboring areas to the auditory cortex, such
back sustained (Dayalu et al. 2011) or intermittent
as the superior temporal sulcus (STS), or the medial
phoneme presentations (Kalinowski et al. 2000), and
portion of the superior temporal gryus (STGm), are
non-linguistically matched content (Guntupalli et al.
activated during perception of visual speech feedback.
2005) while they speak on the telephone to help reduce
It is possible that these sensorimotor integration areas
their stuttering frequency. However, it should be noted
of the STS, STGm and potentially other areas, often
that content which is more speech-like and linguistically
described in mirror neuron system literature (for a re-
matched reduces stuttering to greater extents. Similarly,
view of the mirror system see Oztop, Kawato & Arbib,
perception of vowels reduces stuttering to greater extents
2013; for information on the role of the STS and the MN
than voiceless continuants or non-speech signals. If soft-
during imitation see Molenberghs, Brander, Mattingley
ware or applications would offer these alternatives in a
& Cunnington, 2010), generate the efference model,
portable and Bluetooth pairable device, it would likely
or act as a multisensory/sensorimotor comparator be-
be very beneficial to PWS and potentially for people
tween feedforward and feedback motor control systems.
with other communication disorders.
In the current study, the COMBO-4 condition may
have further increased the effectiveness of the second
signals by presenting additional information for com-
Conclusions
parison and therefore decreased overreliance to a greater
extent. Reductions in stuttering under AAF may also Stuttering frequency was significantly reduced by 72%
be interpreted using the EXPLAN model (Howell and under monaural AAF presentation during scripted tele-
Au-Yeung 2002). The EXPLAN model suggests that phone conversations with naı̈ve listeners. In addition,
internal timekeepers govern motor planning and exe- COMBO-4 reduced stuttering to greater extents than
cution during speech using parallel processes for load COMBO-2. Levels of reduction in stuttering reported
monitoring mechanisms (Howell and Au-Yeung 2002). during this intense hierarchically feared communication
Using this theory, AAF acts as an external timekeeper situation indicate the robust effect of second speech sig-
reducing the load on the system by implementing local nals during speech productions for PWS. Furthermore,
of global rate changes. increased inhibition of stuttering during COMBO-4 as
146
compared with COMBO-2 supports the notion of a
continuum of inhibition during the perception of sec-
ond speech signals. Future studies should examine stut-
tering during monaural and binaural presentations using
the same then different alterations of AAF combinations
compared with single signal presentations during vari-
ous hierarchically difficult speaking situations.
Limitations
Some limitations of the current study include the fact
that monaural headsets and having the speaker hold-
ing the receiver to the contralateral ear are not often
used during typical telephone conversations. Addition-
ally, the use of scripts attempted to account for avoid-
ances, circumlocutions and substitutions while main-
taining consistency between participants. Speakers do
not typically use prepared scripts while making tele-
phone calls.
Acknowledgements
Declaration of interest: The authors report no conflicts of interest.
The authors alone are responsible for the content and writing of the
paper.
References
ANTIPOVA, E. A., PURDY, S. C., BLAKELEY, M. and WILLIAMS, S.,
2008, Effects of altered auditory feedback (AAF) on stuttering
frequency during monologue speech production. Journal of
Fluency Disorders, 33, 274–290.
ARMSON, J., FOOTE, S., WITT, C., KALINOWSKI, J. and STUART, A.,
1997, Effect of frequency altered feedback and audience size
on stuttering. European Journal of Disordered Communication,
32, 359–366.
ARMSON, J. and KEIFTE, M., 2008, The effect of SpeechEeasy on stut-
tering frequency, speech rate, and speech naturalness. Journal
of Fluency Disorders, 33, 120–134.
ARMSON, J., KEIFTE, M., MASON, J. and DE CROOS, D., 2006, The
effect of SpeechEasy on stuttering frequency in laboratory
conditions. Journal of Fluency Disorders, 31, 137–152.
ARMSON, J. and STUART, A., 1998, Effect of extended exposure to
frequency-altered feedback on stuttering during reading and
monologue. Journal of Speech, Language and Hearing Research,
41, 479–490.
BLOODSTEIN, O. and BERNSTEIN-RATNER, N., 2007, A Handbook
on Stuttering, 6th edn (New York, NY: Thomson-Delmar).
BRAY, M. and JAMES, S., 2009, An evaluation of a telephone assistive
device (TAD) for people who stutter. International Journal of
Speech–Language Pathology, 11, 54–60.
CHAMBERS, N., 2009, Impact of the telephone assistive device (TAD)
on stuttering severity while speaking on the telephone. South
African Journal of Communication Disorders, 56, 23–34.
CRAIG, A., HANCOCK, K., CHANG, E., MCCREADY, C., SHEPLEY, A.,
MCCAUL, A. and REILLY, K., 1996, A controlled clinical trial
for stuttering in persons aged 9 to 14 years. Journal of Speech,
Language and Hearing Research, 39, 808–826.
DAYALU, V. N., 2004, Inhibition of stuttering from second speech sig-
nals. Unpublished doctoral thesis, East Carolina University,
Greenville, NC.
Daniel Hudock and Joseph Kalinowski
DAYALU, V., GUNTUPALLI, V., KALINOWSKI, J., STUART, A., SALTUK-
LAROGLU, T. and RASTATTER, M., 2011, Effect of continu-
ous speech and non-speech signals on stuttering frequency
in adults who stutter. Logopedics, Phoniatrics, Vocology, 36,
121–127.
GEORGIEVA, D., 1994, Speech situations increasing stuttering in 13–
16-year-old persons. Journal of Fluency Disorders, 19, 175–
175.
GUNTUPALLI, V. K., KALINOWSKI, J., SALTUKLAROGLU, T. and NAN-
JUNDESWARAN, C., 2005, The effects of temporal modifica-
tion of second speech signals on stuttering inhibition at two
speech rates in adults. Neuroscience Letters, 385, 7–12.
HARGRAVE, S., KALINOWSKI, J., STUART, A., ARMSON, J. and JONES,
K., 1994, Effect of frequency-altered feedback on stuttering
frequency at normal and fast speech rates. Journal of Speech
and Hearing Research, 37, 1313–1319.
HOWELL, P. and AU-YEUNG, J., 2002, The EXPLAN theory of flu-
ency control and the diagnosis of stuttering 2005. In Current
Issues in Linguistic Theory Series: Pathology and Therapy of
Speech Disorders (Amsterdam: John Benjamins), pp. 75–94.
HOWELL, P., EL-YANIV, N. and POWELL, D., 1987, Factors affecting
fluency in stutterers when speaking under altered auditory
feedback. In H. Peters and W. Hulstijn (eds), Speech Motor
Dynamics in Stuttering (New York, NY: Springer), ch. 28.
HUDOCK, D., DAYALU, V. N., SALTUKLAROGLU, T., STUART, A.,
ZHANG, J. and KALINOWSKI, J., 2011, Stuttering inhibition via
visual feedback at normal and fast speech rates. International
Journal of Language and Communication Disorders, 46, 169–
178.
KALINOWSKI, J. and STUART, A., 1996, Stuttering amelioration at
various auditory feedback delays and speech rates. European
Journal of Disordered Communication, 31, 259–269.
KALINOWSKI, J., ARMSON, J., ROLAND-MIESZKOWSKI, M., STUART,
A. and GRACCO, V. L., 1993, Effects of alterations in auditory
feedback and speech rate on stuttering frequency. Language
and Speech, 36, 1–16.
KALINOWSKI, J., DAYALU, V. N., STUART, A., RASTATTER, M. P.
and RAMI, M. K., 2000, Stutter-free and stutter-filled speech
signals and their role in stuttering amelioration for English
speaking adults. Neuroscience Letters, 293, 115–118.
KEPPEL, G. and WICKENS, T. D., 2004, Design and Analysis: A Re-
searcher’s Handbook, 4th edn (Upper Saddle River, NJ: Pearson
Education).
LADOUCEUR, R., COTE, C., LEBLOND, G. and BOUCHARD, L.,
1982, Evaluation of regulated-breathing method and aware-
ness training in the treatment of stuttering. Journal of Speech
and Hearing Disorders, 47, 422–426.
LEITH, W. R. and TIMMONS, J. L., 1983, The stutterer’s reaction
to the telephone as a speaking situation. Journal of Fluency
Disorders, 8, 233–243.
MACLEOD, J., KALINOWSKI, J., STUART, A. and ARMSON, J., 1995,
Effect of single and combined altered auditory feedback on
stuttering frequency at two speech rates. Journal of Commu-
nication Disorders, 28, 217–228.
MAX, L., GUENTHER, F. H., GRACCO, V. L., GOSH, S. S. and WAL-
LACE, M. E., 2004, Internal models and feedback-biased mo-
tor control as sources of dysfluency: a theoretical model of
stuttering. Contemporary Issues in Communication Science and
Disorders, 31, 105–122.
MOLENBERGHS, P., BRANDER, C., MATTINGLEY, J. B. and CUNNING-
TON, R., 2010, The role of the superior temporal sulcus
and the mirror neuron system in imitation. Human Brain
Mapping, 31, 1316–1326.
OZTOP, E., KAWATO, M. and ARBIB, M. A., 2013, Mirror neurons:
function, mechanisms and models. Neuroscience Letters, 540,
43–55.
Stuttering inhibition during scripted telephone conversations
POLLARD, R., ELLIS, J. B., FINAN, D. and RAMIG, P. R., 2009, Ef-
fects of the SpeechEasy on objective and perceived aspects of
stuttering: a 6-month, phase I clinical trial in naturalistic en-
vironments. Journal of Speech, Language and Hearing Research,
52, 516–533.
SALTUKLAROGLU, T. and KALINOWSKI, J., 2006, The inhibition of
stuttering via the presentation of natural speech and sinusoidal
speech analogs. Neuroscience Letters, 404, 196–201.
STAGER, S., LUDLOW, C., GORDON, C., COTELINGAM, M. and
RAPOPORT, J., 1995, Fluency changes in persons who stut-
ter following a double blind trial of clomipramine and de-
sipramine. Journal of Speech, Language and Hearing Research,
38, 516–525.
STUART, A., KALINOWSKI, J., ARMSON, J., STENSTROM, R. and
JONES, K., 1996, Fluency effect of frequency alterations of
plus/minus one-half and one-quarter octave shifts in audi-
tory feedback of people who stutter. Journal of Speech and
Hearing Research, 39, 396–401.
STUART, A., KALINOWSKI, J. and RASTATTER, M. P., 1997, Effect of
monaural and binaural altered auditory feedback on stuttering
frequency. Journal of the Acoustical Society of America, 101,
3806–3809.
STUART, A., KALINOWSKI, J., SALTUKLAROGLU, T. and GUNTUPALLI,
V. K., 2006, Investigations of the impact of altered audi-
tory feedback in-the-ear devices on the speech of people who
stutter: one-year follow-up. Disability and Rehabilitation, 28,
757–765.
UNGER, J., GLUCK, C. W. and CHOLEWA, J., 2012, Immediate effects
of AAF devices on the characteristics of stuttering: a clinical
analysis. Journal of Fluency Disorders, 37, 122–134.
VAN BORSEL, J., REUNES, G. and VAN DEN BERGH, N., 2003, De-
layed auditory feedback in the treatment of stuttering: clients
as consumers. International Journal of Language and Commu-
nication Disorders, 38, 119–129.
VIERA, A. J. and GARRETT, J. M., 2005, Understanding interobserver
agreement: the kappa statistic. Family Medicine, 37, 360–
363.
ZIMMERMAN, S., KALINOWSKI, J., STUART, A. and RASTATTER, M.,
1997, Effect of altered auditory feedback on people who stut-
ter during scripted telephone conversations. Journal of Speech,
Language and Hearing Research, 40, 1130–1134.
Appendix A: Telephone Scripts
Similar to Zimmerman et al. (1997), each partici-
pant initiated the conversation with a ‘Hello’ and ended
the exchange with a ‘Thank you for your time.’
(1) Video store: Can you tell me if you have the movie
The Kings Speech? What are your membership re-
quirements?
147
(2) Museum: What is the special exhibition currently
going on? How much is general admission and
admission to the special exhibition? What are your
weekend hours? Do you have audio guides?
(3) Dry cleaner: How much do you charge to dry
clean a man’s sport jacket? If I drop it off in the
morning, may I pick it up the same day? Are you
open on Sundays?
(4) Office supply: Do you sell external hard-drives?
Where are you located? What time do you close
today? Thank you for your time.
(5) Florist: Can you tell me how much your long-
stem roses are per dozen? How much are a dozen
carnations? Do you take American Express?
(6) Hotel: What are your weekend rates for a double
room? Do you have a fitness center? Do you have
a swimming pool?
(7) Hardware store: Do you have a set of metric socket
wrenches? What time do you close today?
(8) Conference centre: Do you have a conference
room to accommodate a party of ten? Do you
cater? Do you have a lounge in your facility?
(9) Hair salon: How much do you charge for a man’s
shampoo, cut, and blow dry? Do I have to make
an appointment?
(10) Department store: What are your hours of opera-
tion? Do you have any specials going on? Where
is your store located, I’m from out of town and
coming from Greenville?
(11) Restaurant: Do you have a Sunday brunch? What
time do you open on weekdays? Do you have a
children’s menu?
(12) Copy centre: Do you copy homemade DVD
movies? What is your charge per DVD copy? What
are your weekend hours?
(13) Bookstore: I was wondering if you have the book
outliers by Malcolm Gladwell in stock? Is that hard
or soft cover? How much is it?
(14) Sporting goods: Do you carry rollerblades? And
do you also carry snowboards? Are you open on
Sunday nights?
(15) Bakery: Do you make birthday cakes? How many
days’ notice would you require? Do you take
American Express?