ArticlePDF Available

Abstract and Figures

Purpose: The purpose of the present study was to extend previous research by analyzing the ability of adults who stutter to use phonological working memory in conjunction with lexical access to perform a word jumble task. Method: Forty English words consisting of 3-, 4-, 5-, and 6-letters (n = 10 per letter length category) were randomly jumbled using a web-based application. During the experimental task, 26 participants were asked to silently manipulate the scrambled letters to form a real word. Each vocal response was coded for accuracy and speech reaction time (SRT). Results: Adults who stutter attempted to solve fewer word jumble stimuli than adults who do not stutter at the 4-letter, 5-letter, and 6-letter lengths. Additionally, adults who stutter were significantly less accurate solving word jumble tasks at the 4-letter, 5-letter, and 6-letter lengths compared to adults who do not stutter. At the longest word length (6-letter), SRT was significantly slower for the adults who stutter than the fluent controls. Conclusion: Results of the current study lend further support to the notion that differences in various aspects of phonological processing, including vision-to-sound conversions, sub-vocal stimulus manipulation, and/or lexical access are compromised in adults who stutter.
Content may be subject to copyright.
RESEARCH ARTICLE
From Grapheme to Phonological Output:
Performance of Adults Who Stutter on a
Word Jumble Task
Megann McGill*, Harvey Sussman, Courtney T. Byrd
Department of Communication Sciences and Disorders, The University of Texas at Austin, Austin, Texas,
United States of America
*megannmcgill@gmail.com
Abstract
Purpose
The purpose of the present study was to extend previous research by analyzing the ability
of adults who stutter to use phonological working memory in conjunction with lexical access
to perform a word jumble task.
Method
Forty English words consisting of 3-, 4-, 5-, and 6-letters (n= 10 per letter length category)
were randomly jumbled using a web-based application. During the experimental task, 26
participants were asked to silently manipulate the scrambled letters to form a real word.
Each vocal response was coded for accuracy and speech reaction time (SRT).
Results
Adults who stutter attempted to solve fewer word jumble stimuli than adults who do not stut-
ter at the 4-letter, 5-letter, and 6-letter lengths. Additionally, adults who stutter were signifi-
cantly less accurate solving word jumble tasks at the 4-letter, 5-letter, and 6-letter lengths
compared to adults who do not stutter. At the longest word length (6-letter), SRT was signifi-
cantly slower for the adults who stutter than the fluent controls.
Conclusion
Results of the current study lend further support to the notion that differences in various
aspects of phonological processing, including vision-to-sound conversions, sub-vocal stim-
ulus manipulation, and/or lexical access are compromised in adults who stutter.
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 1/19
OPEN ACCESS
Citation: McGill M, Sussman H, Byrd CT (2016)
From Grapheme to Phonological Output:
Performance of Adults Who Stutter on a Word
Jumble Task. PLoS ONE 11(3): e0151107.
doi:10.1371/journal.pone.0151107
Editor: Peter Howell, University College London,
UNITED KINGDOM
Received: October 20, 2015
Accepted: February 22, 2016
Published: March 10, 2016
Copyright: © 2016 McGill et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information files.
Funding: These authors have no support or funding
to report.
Competing Interests: The authors have declared
that no competing interests exist.
Introduction
Stuttering is a multifactorial communication disorder that interrupts the forward flow of
speech production (e.g., [1], [2], [3], [4]). There are significant data to suggest deficits in pho-
nological working memory may be one of the factors that contribute to the difficulties persons
who stutter have establishing and/or maintaining fluent speech, particularly when presented
with cognitively demanding tasks (for review, see [5], [6]; cf., [7]). Reduced speed and accuracy
in aspects of phonological working memory, including phonological encoding and sub-vocal
rehearsal, have been observed in children and adults who stutter in vocal (e.g., [8], [9], [10],
[11], [12]) and nonvocal tasks (e.g., [13, [14], [15], [16]). Additionally, neuroimaging results
using diffusion tensor tractography and magnetoencephalography lend further support to the
assumption that the phonological processing of persons who stutter differs from fluent controls
(e.g., [17], [18], [19], [20], [21], [22], [23]).
Phonological working memory
Theoretically, Baddeleys[24] model creates a framework for understanding the contributions
of phonological working memory to fluent speech production. Baddeleys[24] model of work-
ing memory is comprised of the central executive and three supporting systems: (1) phonologi-
cal loop, (2) visuospatial sketchpad, and (3) episodic buffer (Fig 1). The central executive is
thought to support the retrieval and transfer of information from long-term memory to short-
term memory and vice versa. The phonological loop is comprised of the following two critical
components: a phonological store and a sub-vocal rehearsal system. The phonological store
facilitates the ability to hold material to be remembered in a phonological code. This phonolog-
ical code is vulnerable to decay over time (i.e., trace will last approximately 2 seconds), hence
the need for the sub-vocal rehearsal system. The sub-vocal rehearsal system is a silent verbal
repetition process that refreshes the phonologically encoded material, allowing it to be pre-
served in memory for a longer period of time (>2 seconds). The second system, the visuospa-
tial sketchpad, stores and manipulates visual and spatial information for retention. The third
and final component of Baddeleys model, the episodic buffer, passively binds information
from various distinct sources (i.e., phonological, visual, spatial) into chunks or episodes for
transference to the central executive [24].
Nonword repetition, phoneme elision, and nonword reading tasks
A variety of experimental paradigms have been completed with persons who stutter to enhance
our understanding of the contribution of various subsystems of phonological working memory
to stuttered speech. Nonword repetition tasks have been used to explore phonological working
Fig 1. Adapted version of Baddeleys theoretical model of working memory [24].
doi:10.1371/journal.pone.0151107.g001
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 2/19
memory in adults who stutter (e.g., [8], [11], [12], [13], [25]). Nonword repetition tasks tap
into phonological working memory by requiring phonological encoding and, potentially, sub-
vocal rehearsal of a novel phonological string prior to repetition without the influence of
semantic information. Results from several nonword repetition studies have demonstrated that
adults who stutter are less accurate repeating nonwords than their fluent peers (e.g., [8], [11],
[12], [13], [25]). Phoneme elision tasks using nonword stimuli have also recently been imple-
mented to investigate the abilities of persons who stutter to phonologically encode, sub-vocally
rehearse, and manipulate novel phonological strings prior to producing a novel word.
Ludlow et al. [25] completed one of the first investigations of nonword repetition with
adults who stutter. Ten adults who do and do not stutter repeated two, 4-syllable nonwords
multiple times. Adults who stutter demonstrated significantly decreased accuracy compared to
fluent controls, even when provided with the opportunity to repeat the nonwords. The authors
interpreted these findings to indicate that adults who stutter present with phonological encod-
ing deficits as compared to adults who do not stutter. Additional research has also supported
the finding of differences in phonological processing specific to adults who stutter.
Byrd et al. [8] employed a nonword repetition task as well as a phoneme elision task. Four-
teen adults who stutter and 14 adults who do not stutter listened to 48 nonwords, were pro-
vided multiple attempts at production, and then were required to repeat the target nonword
with a sound missing. There was no difference in performance accuracy on the phoneme eli-
sion task between adults who stutter and fluent controls; this task was equally challenging for
both talker groups. However, for the nonword repetition task, results showed repetition accu-
racy was comparable for adults who do and do not stutter for their repetitions of 24 syllable
words, but the adults who stutter required a greater mean number of attempts before accurate
repetition of 7-syllable words. The authors attributed the significant findings for the nonword
repetition task to suggest that there is a deficit in the sub-vocal rehearsal system of adults who
stutter, which is highlighted when the required productions are at lengths that are more chal-
lenging to recall.
Additionally, Sasisekaran [11] tested behavioral and kinematic responses of 18 adults who
do and do not stutter during nonword repetition and nonword reading tasks. Talker groups
did not differ in their nonword repetition accuracy, but adults who stutter demonstrated a
higher percentage of errors in nonword reading than fluent controls. However, the stimuli
used in their nonword repetition task were one to four syllables in length while the stimuli for
their nonword reading task were either six or 11-syllables long. Thus, it is possible that the non-
words used for the repetition task were not sufficiently complex to tax the phonological work-
ing memory of the adults who stutter in a manner that would yield the talker group differences
previously observed by Byrd et al. [8].
More recently, Sasisekaran and Weisberg [12] explored the nonword repetition accuracy of
10 adults who stutter when auditorily presented with nonword stimuli that varied by complex-
ity and phonotactic constraint. Adults who stutter demonstrated decreased accuracy repeating
complex nonwords compared to their fluent peers. Additionally, adults who stutter exhibited
significant practice effects as measured by reduced movement variability for 3-syllable stimuli,
but motoric variability persisted for the longer 4-syllable words. The authors attributed talker
group differences in nonword repetition accuracy to deficits in phonemic encoding and/or
speech motor processes for adults who stutter.
Vocal versus nonvocal tasks
Sasisekaran, DeNil, Smyth, and Johnson [26] explored nonvocal phonological encoding of
adults who stutter using a phoneme monitoring task. Ten adults who stutter and 12 adults who
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 3/19
do not stutter completed phoneme monitoring tasks during a silent naming condition and a
perception condition. During the silent naming condition, participants pressed a button to
indicate whether or not a target sound was present when they silently named a picture. In the
perception condition, participants silently indicated whether or not a target sound was heard
during auditorily presented items. Results revealed adults who stutter were significantly slower
in phoneme monitoring during the silent naming task. The authors suggested that these results
support the notion that adults who stutter exhibit phonological encoding deficits during silent
encoding and segmenting of phonological units as compared to fluent controls.
In an attempt to extend past findings beyond the restriction to vocal performance, Byrd,
McGill, and Usler [13] explored the vocal and nonvocal nonword repetition and phoneme eli-
sion abilities of adults who stutter compared to adults who do not stutter. Nonvocal conditions
were added in this study in order to avoid possible confounds related to errors of motor execu-
tion. In the vocal nonword repetition task, participants listened to and repeated two sets of 4-
and 7-syllable nonwords (n= 12 per set; 24 total). For the nonvocal nonword repetition task,
participants listened to the same set of 24 nonwords and silently identified each target nonword
from a subsequent set of three nonwords. In the vocal phoneme elision task, participants
repeated the 24 nonwords with a designated phoneme eliminated. Similar to the nonvocal non-
word repetition task, in the nonvocal phoneme elision task, participants listened to phoneme
elision instructions and silently identified which subsequently presented nonword accurately
fulfilled the condition. Adults who stutter produced significantly fewer accurate initial vocal
productions of 7-syllable nonwords compared to adults who do not stutter. There were no par-
ticipant group differences for the silent identification of nonwords, but both groups required
significantly more attempts to accurately silently identify 7-syllable nonwords as compared to
4-syllable nonwords. For the vocal phoneme elision condition, adults who stutter were signifi-
cantly less accurate than adults who do not stutter in their initial production and required sig-
nificantly more attempts to accurately produce 7-syllable nonwords with a designated
phoneme eliminated. This group difference was also significant for the nonvocal phoneme eli-
sion condition for both 4- and 7-syllable nonwords. The authors suggested performance differ-
ences on these vocal and nonvocal tasks further support the notion that phonological working
memory may contribute to the difficulty adults who stutter have establishing and/or maintain-
ing fluent speech.
Paradigm Selection
Taken together, results from multiple experimental investigations support the idea that phono-
logical working memory, particularly the subsystems that Baddeley [24] describes as phonolog-
ical encoding and/or sub-vocal rehearsal, may be uniquely compromised in persons who
stutter. To further extend previous investigations of phonological working memory in adults
who stutter that have been restricted to auditory input, the current investigation utilizes visual
stimuli in a word jumble paradigm. This methodological distinction includes the additional
task demands of (1) visual decoding of letter strings, (2) grapheme-to-sound conversions, and
(3) lexical search/retrieval operations.
Baddeley [24] describes a model of the phonological loop that accounts for both visual and
auditory input eventually arriving at the phonological output buffer, which may either lead to
sub-vocal rehearsal or spoken output. Baddeley [24] posits that visually presented material is
decoded at the level of visual analysis then transferred from an orthographic to a phonological
code to be temporarily stored within the phonological output buffer (See Fig 2). While the pho-
nological code is held within the phonological output buffer, the central executive facilitates a
rapid retrieval process from the lexicon. In the current experimental paradigm, participants
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 4/19
were asked to reorder the orthographic-phonological code through a nonvocal manipulation
process. This silent manipulation assists in the trial and error reordering of the visual and pho-
nological information multiple times prior to finally supplying a complete, lexically
Fig 2. Adapted version of Baddeleys[24] proposed structure of the phonological loop with visual
and/or auditory input.
doi:10.1371/journal.pone.0151107.g002
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 5/19
appropriate solution to the phonological output buffer which will then be executed through
spoken word production. Thus, successful performance on this meta-linguistic task requires
neural facility with several phonologically based lexical skill sets, including phonotactic sensi-
tivities, visualization of re-ordered lexical strings, and matching those manipulations with lexi-
cal reality.
Purpose and Related Predictions
The purpose of the present study is to analyze the ability of adults who stutter to use phonolog-
ical working memory in conjunction with lexical access as measured by performance on a
word jumble task. In contrast to previous investigations of phonological working memory that
have utilized auditory input only (e.g., nonword repetition, phoneme elision), the current para-
digm presents visual stimuli that must then be translated into a phonological code to be reor-
dered. Specifically, the word jumble paradigm requires participants to (1) initially encode a
visually presented, randomly ordered orthographic string, and subsequently, (2) mentally
manipulate, in trial and error fashion, various phonologically possible arrangements, that (3)
trigger the emergence and recognition of a real word. We predict that adults who stutter will
demonstrate increased response times and decreased accuracy in rearranging jumbled stimuli
to arrive at real words compared to adults who do not stutter.
Method
Participants
The current study and consent form documentation was approved by the University of Texas
at Austins Institutional Review Board (IRB approval number 2014-03-0068) and written,
informed consent was obtained for each participant. Participants met the following inclusion-
ary criteria: (a) self-reported native English speaker or English speaker with native competency;
(b) 18 years of age or older; (c) no present or prior history of speech, reading, and/or language
disorders (with the exception of stuttering for the adults who stutter); and (d) no neurological,
social, emotional, or psychiatric disturbances. Participants were 26 adults who do (n= 13;
M= 28.15 years; range = 1942; n= 5 female; n= 8 males) and do not stutter (n= 13; M= 28.5
years; range = 1942; n= 5 female; n= 8 males) matched for age (±3 years), gender, handed-
ness, and education-level. All 13 of the adults who stutter had reportedly received prior speech
therapy for stuttering. See Table 1 for participant descriptions.
Classification and Inclusion Criteria
Speech, language, hearing, and vision measures. All participants passed a bilateral pure
tone hearing screening at 20dB for 1000, 2000, and 4000Hz [27]. All participants met the 20/20
criterion on a near visual acuity test [28]. Additionally, all participants were administered the
following pre-experimental standardized language and speech measures: Peabody Picture
Vocabulary TestFourth Edition (PPVT-4; [29]), Expressive Vocabulary TestSecond Edition
(EVT-2; [30]), Comprehensive Test of Phonological Processing subtests Phoneme Elision, Blend-
ing Words, Rapid Digit Naming, and Nonword Repetition (CTOPP; [31]). All participants
reported no past or present diagnosis of a speech, language, and/or hearing impairment with
exception of stuttering for the participants who stutter. All participants reported no present or
prior diagnosis of dyslexia and no present or prior general or specific difficulties with reading.
All participants also reported no significant neurological or cognitive deficits as part of their
medical history.
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 6/19
Talker group classification. A participant was considered an adult who stutters if he or
she self-reported as a person who stutters, if the participant had received a formal diagnosis of
stuttering from a certified speech-language pathologist, if the participant exhibited three or
more stuttering-like disfluencies (e.g., sound/syllable repetitions, sound prolongations, whole-
word repetitions) per 300-word speech sample [32], and if the participant presented with an
overall score of 10 or higher on the Stuttering Severity Instrument for Children and Adults-
Third Edition (SSI-3; [33]). A participant was classified as an adult who does not stutter if he or
she exhibited fewer than two stuttering-like disfluencies [32] per 300-word sample and a score
of less than 10 on the SSI-3 [33], and if the person reported they had no past or present diagno-
sis of stuttering.
Stuttering severity. Stuttering severity was determined from video recorded conversa-
tional and reading samples with consideration of duration of disfluent moments and physical
concomitants. Each participants samples were analyzed by the first author using the SSI-3
Table 1. Participant characteristics for adults who do not stutter (AWNS) and adults who stutter (AWS).
Participant Age Handedness Gender Education
Level
Severity PPVT-4 EVT-2 CTOPP-PE CTOPP-BW CTOPP-RDN CTOPP-NWR
1 22 Right Male College ML 123 114 9 13 10 13
2 42 Right Female Graduate
School
SV 105 105 4 10 9 8
3 19 Right Male College ML 117 118 11 12 11 13
4 28 Left Female College SV 106 113 11 6 10 9
5 42 Right Male Masters MOD 107 106 10 9 9 11
6 27 Right Male College VS 86 90 9 4 6 4
7 35 Right Female College ML 107 110 7 11 2 10
8 31 Left Female College ML 114 104 11 10 11 9
9 27 Right Male College ML 96 102 11 10 20 7
10 20 Right Male College VML 113 121 9 10 9 8
11 21 Right Male College VML 127 115 8 10 7 11
12 20 Right Female College ML 91 97 10 13 7 13
13 32 Right Male College VS 109 118 8 10 9 10
14 21 Right Male College N/A 108 102 10 11 12 13
15 25 Left Female College N/A 117 120 12 14 8 12
16 23 Right Male College N/A 128 116 4 11 12 12
17 32 Right Female College N/A 136 120 12 14 9 14
18 27 Right Male College N/A 109 112 11 12 12 14
19 19 Right Female College N/A 98 104 12 12 12 14
20 35 Right Male College N/A 107 104 11 14 9 10
21 27 Right Male College N/A 113 116 11 13 12 13
22 19 Right Male College N/A 126 116 11 10 9 12
23 34 Left Female College N/A 113 104 11 13 16 13
24 39 Right Male Masters N/A 110 115 11 10 12 11
25 25 Right Male College N/A 129 136 12 13 12 11
26 42 Right Female Graduate
School
N/A 103 113 9 13 10 12
Peabody Picture Vocabulary Test-Fourth Edition (PPVT-4): Standard score (M = 100, SD = 15). Expressive Vocabulary Test-Second Edition (EVT-2):
Standard score: (M = 100, SD = 15). Comprehensive Test of Phonological Processing (CTOPP) subtests: Standard score: (M = 10, SD = 2). Severity:
VM = very mild, ML = mild, MOD = moderate, SV = severe, VS = very severe.
doi:10.1371/journal.pone.0151107.t001
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 7/19
[33]. Of the 13 participants who were classified as adults who stutter, two received a rating of
very mild stuttering, six received ratings of mild stuttering, one received a rating of moderate
stuttering, two received a rating of severe stuttering, and two received a rating of very severe
stuttering.
Inter-rater reliability of stuttering severity for speech samples was determined by the first
author and an undergraduate research assistant trained in disfluency count analysis. Eight of
the 26 participants (30%; 4 adults who stutter, 4 adults who do not stutter) were selected at ran-
dom to determine inter-rater and intra-rater reliability. For adults who stutter, inter-rater reli-
ability was within two points on the SSI-3 for the four participants. Thus, the reliability was
found to be Kappa = 0.94. There was 100% agreement for the severity ratings for all four partic-
ipants who do not stutter, with no stuttering-like disfluencies noted during the conversational
samples.
Word Jumble familiarity. Participants self-reported experience with word jumble games
in books, online, or in the newspaper prior to beginning the experimental task. Participants
reported their frequency of completion of word jumble tasks within the last year on a 5-point
Likert scale (i.e., 1 never, 2 = rarely, 3 = occasionally, 4 = a moderate amount, 5 = a great deal).
All 26 participants reported scores of 1 or 2, indicating infrequent exposure to word scramble
activities.
Stimuli Development
Forty English words consisting of 3-, 4-, 5-, and 6-letters (n= 10 per letter length category)
were selected from word lists developed by Snodgrass and Vanderwart [34] and Juhasz, Lai,
and Woodcock [35]. Stimuli were controlled for age of acquisition and visual familiarity as
described by Snodgrass and Vanderwart [34] and Juhasz et al. [35]. The mean age of acquisi-
tion for 3-letter words was 3.02 years. The mean age of acquisition for 4-letter words was 3.13
years. Mean age of acquisition for 5- letter words was 3.25 years and was calculated at 3.82
years for 6-letter words.
Each of the 40 words selected for use in the study was then randomly jumbled using a web-
based application [36]. To generate the randomly jumbled word, the English spelling for each
stimulus item was typed into the Word Scramble Maker. The application randomly transposed
letters within each word to produce a scrambled, or jumbled, stimulus. The same scrambled
stimulus of a word was used across all participants. See Table 2 for the word jumble stimuli
used in the present study.
Procedures
Pre-testing, including standardized measures of expressive language, receptive language, pho-
nological processing, and stuttering severity were complete immediately preceding the admin-
istration of experimental tasks. Experimental tasks were completed in a quiet room over one
session lasting approximately 30 minutes. Participants viewed stimuli on a 35Dell computer
monitor. The distance between the participant, seated on a chair, and the computer screen was
approximately 18 inches. Stimuli were presented using 40 point, Arial, black font on a white
background using Microsoft PowerPoint. Scrambled words were presented one at a time on
the computer display. Participants were asked to silently manipulate the scrambled letters to
form an English word. Participants were given one opportunity to verbally produce each
unscrambled stimulus real word. After the participant successfully unscrambled the letter
string to arrive at a word, or failed to produce a response, he or she manually advanced to the
next scrambled stimulus by left-clicking the mouse. There was no time limit, but participants
were encouraged to respond as quickly as possible to each stimulus.
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 8/19
Participants were read the following directions prior to beginning the task: You will see a
string of jumbled letters appear on the screen. Mentally manipulate or switch around the letters
so that they form a real word. Verbally state your answer as soon as you have unscrambled the
word. Each word jumble only has one correct answer. Please do not mouth, whisper or think
out loud.You only have one opportunity to state your answer. After you have said the word,
advance to the next slide by clicking the mouse. Do you have any questions?One practice
item preceded the experimental tasks. Participants were given the opportunity to ask questions
after the practice item. No participants were observed moving their mouths, whispering, or
thinking aloud during the experimental tasks.
For all participants, the stimuli were presented according to increasing complexity. Thus,
participants were first presented with the 3-letter words, then the 4-letter words, followed by
the 5-letter words, and finally ending with the most challenging, 6-letter words. The order of
words within the letter-length classes was counterbalanced across participants in both groups.
Coding
Participant responses were recorded by hand using a paper score sheet and digitally using a
Shure SM58 microphone placed approximately six inches from the speakers mouth. Responses
were also documented using a Sanyo VPC-HD1010 digital video recorder, an Olympus WS-
321M digital voice recorder, and the record presentation function on Microsoft PowerPoint.
Multiple recording methods were used to ensure that participantsresponses were recorded in
the case of technical difficulty. Each response was coded for accuracy and speech reaction time
(SRT). Participantsattempts to unscramble the word were scored as either being 1) correct, 2)
an error (e.g., stating atomgiven the jumbled stimulus attoom), or 3) no response (e.g., I
cant figure it out.or advancing to the next stimulus without attempting to answer). Stuttered
Table 2. Word jumble stimuli.
3-Letter Words Jumbled Examples 5-Letter Words Jumbled Examples
1. bed edb 1. clock lccko
2. fox ofx 2. glove eogvl
3. sun nus 3. arrow roarw
4. hat ath 4. house hseuo
5. key eyk 5. camel malec
6. y lfy 6. peach epcha
7. and dan 7. eagle eaelg
8. car rca 8. ute ftule
9. gun gnu 9. horse erohs
10. nut tun 10. ruler rurle
4-Letter Words Jumbled Examples 6-Letter Words Jumbled Examples
1. drum urdm 1. tomato attoom
2. moon oonm 2. anyone yennoa
3. sock ksco 3. jacket cjteak
4. bear arbe 4. hotdog gdooht
5. vest etsv 5. turtle eturtl
6. iron roni 6. subway wbsyua
7. nail inal 7. monkey nkmeyo
8. lion oiln 8. batboy baytob
9. kite eikt 9. nger ngeifr
10. sh hs 10. uplift lufpit
doi:10.1371/journal.pone.0151107.t002
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 9/19
productions and responses beginning with non-specific vocalizations (e.g., uhm,er) were
excluded from further analysis. These exclusions occurred three times for adults who do not
stutter and four times for adults who stutter. All responses for each participant were reviewed
offline by the initial scorer and also by a trained research assistant to ensure accuracy in tran-
scription and no discrepancies were found.
ParticipantsSRTs, in milliseconds, were calculated by analyzing participantsverbal
responses using Praat software. The record function on Microsoft PowerPoint simultaneously
created a new audio recording when each word jumble stimulus appeared on the screen. Audio
files were exported from Microsoft PowerPoint and analyzed using Praat software. Thus, SRT
was calculated from the beginning of the recording, which coincided exactly with the appear-
ance of the visual stimulus, to the onset of the participants vocal response for each word jum-
ble stimulus. The first author and a trained research assistant independently, visually inspected
each of the waveforms and spectrograms of the participantsresponses to determine SRT for
each word jumble stimulus. Inter-rater reliability was 90% and intra-rater reliability was 100%.
Results
Statistical analyses were conducted using IBM SPSS Statistics Version 23. Alpha level was set at
0.05. Levenes Test for Equality of Variances was completed for each analysis as well as
Mauchlys Test of Sphericity for each repeated measure analysis. One adult who stutters pre-
sented with SRT that was calculated at greater than two standard deviations above the mean
for adults who stutter and was, thus, excluded from statistical analyses related to SRT in order
meet the ANCOVA assumption of no outliers.
Speech, language, and phonological testing
Independent t-tests conducted on the mean scores demonstrated that the performances of
adults who stutter (M= 107.77; SD = 11.76) and adults who do not stutter (M= 115.15;
SD = 11.35) did not significantly differ for receptive vocabulary; t(24) = 1.63, p= 0.116. The
expressive vocabulary of the adults who stutter (M= 108.69; SD = 9.07) did not significantly
differ from the adults who do not stutter (M= 113.69; SD = 9.21); t(24) = 1.395, p= 0.176.
Likewise, no significant differences were found between the talker groups for CTOPP subtests
Phoneme Elision (PE) or Rapid Digit Naming (RDN) (CTOPP-PE: adults who stutter
M= 9.08, SD = 2.02; adults who do not stutter M= 10.54, SD = 2.15; t(24) = 1.79, p= 0.086;
CTOPP-RDN: adults who stutter M= 9.23, SD = 4.04; adults who do not stutter M= 11.15,
SD = 2.11; t(24) = 1.52, p= 0.142). Independent t-tests conducted on the mean scores demon-
strated that the performances of adults who stutter and adults who do not stutter significantly
differed on CTOPP Blending Words (BW: adults who stutter M= 9.85, SD = 2.51; adults who
do not stutter M= 12.31, SD = 1.44; t(24) = 3.067, p= 0.005). Additionally, a t-test on the
mean scores of CTOPP Nonword Repetition (NWR) revealed significant group differences.
Levenes test for equality of variances was found to be violated for the CTOPP NWR analysis F
(1, 24) = 4.426, p= 0.046. Thus, a t-statistic not assuming homogeneity of variance was com-
puted for CTOPP NWR scores for adults who stutter (M= 9.69, SD = 2.63) and adults who do
not stutter (M= 12.38, SD = 1.26); t(17.25) = 3.33, p= 0.004. To confirm the assumptions of
ANCOVA, no significant interactions between the covariates (i.e., CTOPP BW, CTOPP
NWR) and Word Length or Talker Group variables were found; furthermore, variance infla-
tion factor (VIF) for all variables was below 2.5, indicating no multicollinearity. Thus, perfor-
mances on CTOPP NWR and CTOPP BW subtests were included in statistical models
analyzing word jumble accuracy and SRT as covariates.
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 10 / 19
Word jumble accuracy
As shown in Fig 3, all participants produced attempts to unscramble the 10 stimuli presented
at the 3-letter length. At the 4-letter length, adults who stutter demonstrated fewer attempted
trials to solve the word jumble stimuli compared to adults who do not stutter; however, this dif-
ference was not significant (t(12) = 1.760, p = 0.104), while controlling for CTOPP BW and
CTOPP NWR. Adults who stutter presented with significantly fewer attempted trials compared
to adults who do not stutter, controlling for CTOPP BW and CTOPP NWR, at the 5-letter (t
(15.268) = 3.352, p = 0.04), and 6-letter lengths (t(17.075) = 2.354, p = 0.031). That is, adults
who stutter attempted to solve significantly fewer word jumbles than adults who do not stutter
at the 5- and 6-letter lengths.
A mixed-model repeated measures ANCOVA was conducted with the between-subjects fac-
tor of Talker Group (adults who stutter, adults who do not stutter), a within-subjects factor of
Word Length (3-, 4-, 5-, 6-letter length), and CTOPP NWR and CTOPP BW performances as
covariates. The dependent variable was the percentage of accurately unscrambled real word
responses. Results revealed a significant between-subjects effect for Talker Group F(1,22) =
42.175, p0.000, partial η
2
= 0.657 and a significant interaction between Talker Group and
Word Length F(3,22) = 12.762, p0.000, partial η
2
= 0.367. No significant main effect was
noted for Word Length F(3,22) = 1.003, p= 0.397, partial η
2
= 0.044. Decomposition of the
interaction between Talker Group and Word Length, while controlling for CTOPP BW and
CTOPP NWR, revealed a significant difference between adults who stutter and adults who do
not stutter at the 4-letter (t(22) = 3.828, p0.001, partial η
2
= 0.400), 5-letter (t(22) = 2.70,
p= 0.013, partial η
2
= 0.249), and 6-letter lengths (t(22) = 7.113, p0.0001, partial η
2
= 0.697).
Fig 3. The mean number (+/- two standard errors) of attempted trials for each word jumble task across word-length classes for adults who stutter
(AWS) and adults who do not stutter (AWNS).
doi:10.1371/journal.pone.0151107.g003
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 11 / 19
No significant differences were observed between adults who stutter and adults who do not
stutter, when controlling for CTOPP BW and CTOPP NWR performances, at the 3-letter
length (t(22) = 1.997, p= 0.058, partial η
2
= 0.153). As seen in Fig 4, adults who stutter were sig-
nificantly less accurate solving word jumble tasks at the 4-letter, 5-letter, and 6-letter lengths
compared to adults who do not stutter.
Speech reaction time
A mixed-model repeated measures ANCOVA was conducted with the between-subjects factor
of Talker Group (adults who stutter, adults who do not stutter), a within-subjects factor of
word length (3-,4-, 5-,6-letter length), and CTOPP NWR and CTOPP BW performances
included as covariates. The dependent variable was the participants SRT (in seconds) from the
presentation of the scrambled stimuli to the production of an accurate or inaccurate response.
SRT for no-response trials was not included in this analysis. Results revealed a significant
between-subjects effect for Talker Group F(1,22) = 15.984, p= 0.001, partial η
2
= 0.421 and a
significant interaction between Talker Group and Word Length F(3,22) = 16.410, p0.0001,
partial η
2
= 0.427. There was not a significant main effect for Word Length F(3,22) = 1.237, p=
.303, partial η
2
= 0.053. Decomposition of the interaction between Talker Group and Word
Length, controlling for CTOPP BW and CTOPP NWR, revealed a significant difference
between adults who stutter and adults who do not stutter at the 6-letter word length (t(22) =
4.248, p0.0001, partial η
2
= 0.451). As shown in Fig 5, SRT was significantly slower for adults
who stutter than for adults who do not stutter at the longest word length (6-letter).
Correlation between accuracy and SRT
An evaluation was made of the linear relationship between SRT and accuracy for each talker
group using Pearsons correlation, controlling for CTOPP BW and CTOPP NWR scores. As
depicted in Fig 6, for adults who stutter, at the 6-letter length, the two variables, accuracy and
SRT, were not correlated, r(10) = 0.06. However, for adults who do not stutter, at the 6-letter
Fig 4. The mean percentage (+/- two standard errors) of accurate responses for each word jumble task across word-length classes for adults who
stutter (AWS) and adults who do not stutter (AWNS).
doi:10.1371/journal.pone.0151107.g004
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 12 / 19
word length, accuracy and SRT demonstrated a strong negative relationship r(11) = -0.75. In
other words, for adults who do not stutter, as accuracy increased, SRT decreased, but for adults
who stutter, no relationship between accuracy and SRT was observed. No other meaningful
correlations were noted for 3-letter, 4-letter, or 5-letter word lengths for either talker group.
Discussion
The present study extended investigations of phonological working memory and stuttered
speech by using orthographic rather than auditory stimuli. A word jumble task was chosen
because the metalinguistic operations underlying successful performance requires an ability to
translate sequences of graphemes into their phonemic equivalents, in reiterative fashion, in
search of, and eventual access to, a lexical word. Results revealed that adults who stutter were
significantly, and progressively, less accurate than adults who do not stutter at solving word
jumbles at the 4-letter, 5-letter, and 6-letter lengths, and moreover, demonstrated significantly
longer SRT at the longest word length. Thus, rather than seeing a uniform deficiency in phono-
logical processing, processing deficits emerged when the level of cognitive load was increased.
This section will focus on potential areas of processing deficiencies in phonological working
memory for adults who stutter during their performance on this demanding word jumble task.
Grapheme to phoneme conversions and phonological encoding: Effect
of cognitive load on working memory
Previous research has speculated that phonological encoding differences in persons who stutter
may contribute to their difficulty establishing and maintaining fluent speech production,
Fig 5. The average speech reaction time in seconds (+/- two standard errors) for each word jumble task across word-length classesfor adults who
stutter (AWS) and adults who do not stutter (AWNS).
doi:10.1371/journal.pone.0151107.g005
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 13 / 19
particularly when cognitive load is taxed (e.g., [37], [13]). An example of a cognitive load effect
can be seen in a rhyming study conducted by Weber-Fox and colleagues [38]. They compared
rhyming performance between adults who do and do not stutter when only visual orthographic
input was provided. Participants manually selected a yesor nobutton to indicate whether
or not the two visually presented words rhymed. Out of the four conditions, the condition con-
sidered to be the most phonologically challenging was the only one to reveal talker group dif-
ferences. In this condition, participants were presented with two words that were
orthographically similar but did not rhyme (e.g., costand most). Adults who stutter were
significantly slower to respond than their fluent peers in this condition. Weber-Fox et al. [38]
contend that their findings do not demonstrate that adults who stutter have fundamental pho-
nological processing deficits but that their ability to phonologically encode may be atypically
vulnerable when the cognitive demand increases. In the word jumble task, performance differ-
ences between adults who do and do not stutter were noted at the longest word lengths (i.e., 5-
letter and 6- letter words). The series of phonologically-dependent operations involved in solv-
ing a complex word jumble may recruit a high cognitive load and, thus, may have negatively
impacted both the accuracy and SRT of adults who stutter in the current study.
Sub-vocal stimuli manipulations
Nonword repetition and phoneme elision paradigms, which require sub-vocal rehearsal, have
demonstrated significant performance differences in adults who stutter relative to adults who do
not stutter (e.g., [8], [12], [13]). Such tasks heavily depend on sub-vocal rehearsal (i.e., keeping
the stimuli in short term memory storage In the present task, sub-vocal rehearsal is not required
as defined by Baddeley [24] because the participant has continual visual access to the word as he/
she is attempting to unscramble it. However, the participant is required to nonvocally manipulate
Fig 6. The correlation between accuracy and speech reaction time (in seconds) at the 6-letter word length for adults who stutter (AWS) and adults
who do not stutter (AWNS).
doi:10.1371/journal.pone.0151107.g006
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 14 / 19
the stimulus in a trial and error fashion to arrive at a lexically appropriate solution. The numer-
ous trial and error rearrangements, including assessing if the reordered letter strings match a lexi-
cal word, provides a robust processing load on phonological encoding and rehearsal networks. A
similar cognitive load effect was seen in Byrd et al. [13]. Adults who stutter, relative to fluent con-
trols, demonstrated significant difficulty manipulating a novel phonological string, even when
vocal output was not required. Unlike the visual stimuli used in the present study, Byrd and col-
leagues [13] provided auditory input only. Thus, deficits in covert manipulation of phonological
strings, irrespective of modality (i.e. verbal versus visual), appear to be relevant in explain and
understanding the performance of adults who stutter on such demanding tasks.
Lexical access
The nature of the word jumble task requires participants to hold the changing phonological
strings in the output buffer while simultaneously accessing the lexicon to determine if their solu-
tion to the paradigm is congruent with a real English word. Adults who stutter exhibited both
increased SRT and decreased accuracy at the longest (6-letter) word length, whereas, adults who
do not stutter demonstrated increased SRTs for 6-letter strings, but also increased accuracy. In
other words, adults who do not stutter appeared to present with a speed-accuracy trade-off, such
that increased SRT was required in order to produce the accurate solution to the word jumble
task at the 6-letter length. Conversely, adults who stutter did not appear to benefit from increased
SRT. That is, the slowness of their responses did not correspond with the accuracy of their
responses. The absence of the speed-accuracy tradeoff for adults who stutter suggests that lexical
access and/or the organization of the mental lexicon may be compromised in adults who stutter.
Nevertheless, as has been previously argued by Hakim and Ratner [9]it is difficult to know
whether weaknesses in responding to the tasks reflect difficulty in encoding the input, storing it
in memory, or accessing it efficiently (p.194).More specific to the present study, it is challenging
to parse out these individual processes (i.e. grapheme to phoneme conversion, phonological
encoding, sub-vocal manipulation of serial ordering of segmental elements, lexical access) that
take place nearly simultaneously as part of completing the word jumble paradigm.
Speech Motor and Additional Considerations
It is also possible that adults who stutter demonstrated a temporal instability in motor pro-
gramming that resulted in slower and less accurate word jumble responses. Although stuttered
responses were excluded from analysis, it is possible that adults who stutter demonstrated
increased response times due to avoidance of anticipated stuttering. Previous research using
nonword repetition and phoneme elision tasks with adults who stutter lends support to the
assumption that there is an interplay between phonological encoding and motor programming
(e.g., [13], [39], [40]). Future research could include follow-up questions to determine if partic-
ipants detected difficulty responding due to anticipated disfluencies and/or if they used any flu-
ency-enhancing strategies that may have impacted the onset of their responses.
Additional research is warranted to better understand the various subsystems involved in
motor speech, lexical access, and phonological processing for adults who do and do not stutter.
Perhaps, neuroimaging data can be used to further explain the role of phonological processing
in stuttered speech.
Word Jumble Processing in Adults who Stutter: Possible Neural
Correlates
In the past decade there has been an upsurge in brain-based studies reporting structural (white
matter integrity, gray matter volumes), functional (overactive and underactive brain regions),
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 15 / 19
and neuro-pharmacological (imbalances of D1 and D2 dopamine receptors in the basal ganglia)
anomalies in persons who stutter (see Craig-McQuaide et al. [41] for meta-analyses). Collec-
tively, the compromised neural areas, networks, and chemical regulators implicated in stutter-
ing involve cortical/ subcortical areas performing speech motor planning and execution, and
their subsequent integration with a phonological interfacee.g., the speech sound map.Neu-
rons in the speech sound map area are hypothesized to encode sound, syllable, and word
sequences to be articulated [42]. This section of the discussion will provide an overview of dif-
fusion tractography studies in persons who stutter, focusing on region-to-region white matter
connectivity within the dorsal stream [43]. The dorsal stream contains the superior longitudi-
nal fasciculus and the arcuate fasciculus. Of particular importance to the present study, these
pathways are thought to sustain the dorsal phonological route underlying grapheme-phoneme
decoding([44], p. 935).
To provide a framework for interpreting neuroimaging studies completed with persons who
stutter, it is critical to first understand the general assumptions underlying the relationship
between the dorsal stream and phonological processing. Saygin et al. [23] reported composite
scores on measures of phonological awareness (elision, blending, nonword repetition, letter
knowledge) positively correlated with white matter volume of the left arcuate fasciculus in typi-
cally developing children. With that in mind, it is interesting to note that Chang et al. [18]
reported lower tract densities in adults who stutter in connections between Brocas area and
both left primary motor cortex and premotor cortex. Chang and Zhu [19] also reported lower
fractional anisotrophy (FA) values in children who stutter in basal ganglia-to-thalamus-to-cor-
tex loops as well as connections between posterior superior temporal regions and the left infe-
rior frontal gyrus. While FA is a fairly non-specific biomarker of fiber neuropathology, lower
FA values are associated with slower speed and reduced quality of signal transmissions, or less-
than-ideal synchronization of signals between relevant brain regions [21].
A more recent investigation by Connally et al. [21] reported lower FA values in the arcuate
fasciculus bilaterally, as well as an inverse correlation between FA value and stuttering severity
in the left angular gyrus. Cai et al. [17] established a region-to-region connectivity matrix in
adults who stutter showing a wide array of inferior connections in several regions of left peri-
Rolandic sensorimotor cortex and left premotor areas, most notably between left ventral pre-
motor cortex and mid-regions of primary motor cortex. Correlational analyses between sever-
ity of stuttering and white matter connectivity in the speech network revealed that all
significant correlations were negativelow FA values correlated with high stuttering severity.
Additionally, Cieslak et al. [20] reported three white matter tracts that significantly differen-
tiated adults who stutter from fluent controls: (1) the left arcuate fasciculus connecting the left
inferior temporal gyrus to the left insular cortex was missing 39% of the tractvolume, (2) the
right arcuate fasciculus was also missing streamlines (38% of the tract) connecting inferior
temporal gyrus to the supramarginal gyrus and surprisingly, (3) a novelleft temporo-striatal
tract connecting left inferior temporal gyrus to the left accumbens area of the basal ganglia was
found only in adults who stutter (7 of the 8 participants who stutter). This white matter path-
way overlaps with the uncinate fasciculus, which has been functionally linked to auditory
working memory/sound recognition([45], p. 3538). Thus, findings from diffuse tensor tracto-
graphy studies provide a neural basis for behavioral differences in phonological working mem-
ory for persons who stutter relative to fluent controls.
Additional supportive evidence of network dysfunction in frontal speech motor areas in
adults who stutter has been reported in a whole-head magnetoencephalography (MEG) map-
ping study [22] using a delayed reading (isolated nouns) paradigm. The temporal patterning of
cortical activations during single word reading in adults who stutter were found to be reversed
relative to fluent controls. Within the initial 400 milliseconds after seeing a word, timing order
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 16 / 19
of cortical activation in fluent controls proceeded from left inferior frontal cortex (articulatory
programming) to left lateral central sulcus and dorsal premotor cortex (motor preparation); in
contrast, adults who stutter showed an early left motor cortex activation followed by a long
delayed left inferior frontal signal. This implies that the functional synchronization between
preparation and execution of an articulatory motor sequence in persons who stutter is out of
phase. High temporal resolution methodologies, such as MEG recordings, are essential to fur-
ther identify subtle serial ordering anomalies in activating the speech production system.
Taken together, behavioral data from the current study and previously reported neuroimaging
results support the idea that differences in phonological processing and/or lexical access may
contribute to the persistence and maintenance of stuttered speech production.
Conclusion
Through the use of behavioral tasks such as nonword repetition and phoneme elision,
researchers have demonstrated that phonological working memory may contribute to stuttered
speech production (e.g., [5]; cf., [7]). The current investigation extended past research by taxing
phonological working memory using phonologically scrambled visual stimuli (i.e. word jumble
task). As the cognitive load increased, adults who stutter attempted to solve significantly fewer
word jumble tasks compared to fluent controls. Of the trials attempted, adults who stutter were
significantly less accurate in their completion of word jumble tasks at the 4-letter, 5-letter, and
6-letter lengths compared to adults who do not stutter. Furthermore, adults who stutter dem-
onstrated increased SRT at the 6-letter word length as compared to fluent controls. Addition-
ally, adults who do not stutter demonstrated a significant negative correlation between
accuracy and SRT at the longest word length; however, adults who stutter did not appear to
present with this speed-accuracy trade-off. These results may be interpreted to further support
the notion that differences in various aspects of phonological working memory, including
visual-to-sound conversions, lexical access, and sub-vocal manipulations, may uniquely com-
promise the speech fluency of adults who stutter. Various neural underpinnings for both pho-
nological processing deficits and speech motor control deficits may overlap and be responsible
for the on-offfluency problems characterizing stuttering.
Supporting Information
S1 Data. Accuracy and speech reaction time data for adults who do and do not stutter.
(XLSX)
Acknowledgments
We would like to sincerely thank the Michael and Tami Lang Stuttering Institute and Dr. Jen-
nifer and Emanuel Bodner for their endowed support of our research efforts. We would also
like to thank Elizabeth Hampton, M.S., CCC-SLP and Zoi Gkalitsiou, M.A., CCC-SLP who
assisted us with the recruitment and testing of participants as well as the graduate students
who assisted with the testing, and data collection/analysis process. Most of all, we would like to
thank the adults who do and do not stutter who were willing to give their time to participate in
this study and help us to further our knowledge of the underlying nature of stuttering.
Author Contributions
Conceived and designed the experiments: HS CTB MM. Performed the experiments: MM.
Analyzed the data: MM HS CTB. Wrote the paper: MM HS CTB.
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 17 / 19
References
1. Bloodstein O, Bernstein Ratner N. A handbook on stuttering. 6th ed. Clifton Park, NY: Delmar Cen-
gage Learning; 2008.
2. Conture E. Stuttering: Its nature, diagnosis, and treatment ( 3rd Edition). Boston: Allyn and Bacon;
2001.
3. Guitar B. Stuttering: An integrated approach to its nature and treatment. Baltimore: Williams & Wilkins;
2013.
4. Yairi E, Seery C. Stuttering: Foundations and Clinical Applications. Upper Saddle River, NJ: Pearson
Education; 2011.
5. Bajaj A. Working memory involvement in stuttering: Exploring the evidence and research implications.
Journal of Fluency Disorders. 2007; 32: 218238. PMID: 17825670
6. Sasisekaran J. Exploring the link between stuttering and phonology: A review and implications for treat-
ment. Seminars in Speech and Language. 2014; 35: 95113. doi: 10.1055/s-0034-1371754 PMID:
24782273
7. Nippold MA. Stuttering and Phonology: Is there an interaction? American Journal of Speech-Language
Pathology. 2002; 11: 99110.
8. Byrd CT, Vallely M, Anderson J, Sussman H. Nonword repetition and phoneme elision in adults who do
and do not stutter. Journal of Fluency Disorders. 2012; 37: 188201. doi: 10.1016/j.jfludis.2012.03.003
PMID: 22682320
9. Hakim H, Ratner N. Nonword repetition abilities of children who stutter: An exploratory study. Journal of
Fluency Disorders. 2004; 29: 179199. PMID: 15458830
10. Ludlow CL, Siren K, Zikria M. Speech production learning in adults with chronic developmental stutter-
ing. In Hulstijn W., Peters H. F. M., & van Lieshout P. H. H. M. (Eds.), Speech production: Motor control,
brain research and fluency disorders (pp. 221230). Amsterdam: Elsevier; 1997.
11. Sasisekaran J. Nonword repetition and nonword reading skills in adults who do and do not stutter. Jour-
nal of Fluency Disorders. 2013; 38: 275289. doi: 10.1016/j.jfludis.2013.06.001 PMID: 24238389
12. Sasisekaran J, Weisberg S. Practice and retention of nonwords in adults who stutter. Journal of Flu-
ency Disorders. 2014; 41: 5571. doi: 10.1016/j.jfludis.2014.02.004 PMID: 25173457
13. Byrd CT, McGill M, Usler E. Nonword repetition and phoneme elision in adults who stutter: Vocal and
nonvocal performance differences. Journal of Fluency Disorders. 2015; 44: 1731. doi: 10.1016/j.
jfludis.2015.01.004 PMID: 25680736
14. Brocklehurst PH, Corley M. Investigating the inner speech of people who stutter. Evidence for (and
against) the Covert Repair hypothesis. Journal of Communication Disorders. 2011; 44: 246260. doi:
10.1016/j.jcomdis.2010.11.004 PMID: 21208627
15. Postma A, Kolk H, Povel DJ. Speech planning and execution in stutterers. Journalof Fluency Disor-
ders. 1990; 15: 4959.
16. Sasisekaran J, De Nil LF. Phoneme monitoring in silent naming and perception in adults who stutter.
Journal of Fluency Disorders. 2006; 31: 284302. PMID: 16982085
17. Cai S, Tourville JA, Deryk S, Perkell J, Guenther FH, Ghosh S. Diffusion imaging of cerebral white mat-
ter in persons who stutter: Evidence for network-level anomalies. Frontiers in Human Neuroscience.
2014; 8: 118.
18. Chang SE, Horwitz B, Ostuni J, Reynolds R, Ludlow C. Evidence of left inferior frontal-premotor struc-
tural and functional connectivity deficits in adults who stutter. Cerebral Cortex. 2011; 21: 25072518.
doi: 10.1093/cercor/bhr028 PMID: 21471556
19. Chang SE, Zhu DC. Neural network connectivity differences in children who stutter. Brain. 2013; 136:
37093726. doi: 10.1093/brain/awt275 PMID: 24131593
20. Cieslak M, Ingham RJ, Ingham JC, Grafton S. Anomalous white matter morphology in adults who stut-
ter. Journal of Speech, Language, and Hearing Research. 2015; 58: 268277. doi: 10.1044/2015_
JSLHR-S-14-0193 PMID: 25635376
21. Connally EL, Ward D, Howell P, Watkins K. Disrupted white matter in language and motor tracts in
developmental stuttering. Brain & Language. 2014; 131: 2533.
22. Salmelin R, Schnitzler A, Schmitz F, Freud H. Single word reading in developmental stutterers and flu-
ent speakers. Brain. 2000; 123: 11841202. PMID: 10825357
23. Saygin ZM, Norton ES, Osher DE, Beach S, Cyr A, Ozernov-Palchik O, et al. Tracking the roots of read-
ing ability: White matter volume and integrity correlate with phonological awareness in prereading and
early-reading kindergarten children. Journal of Neuroscience. 2013; 33: 1325113258. doi: 10.1523/
JNEUROSCI.4383-12.2013 PMID: 23946384
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 18 / 19
24. Baddeley A. Working memory and language: An overview. Journal of Communication Disorders. 2003;
36: 189208. PMID: 12742667
25. Ludlow C, Siren K, Zikria M. Speech production learning in adults with chronic developmental stutter-
ing. In Hulstijn W., Peters H. F. M., & van Lieshout P. (Eds.), Speech production: Motor control, brain
research and fluency disorders (pp. 221230). NY: Elsevier; 1997.
26. Sasisekaran J, De Nil LF, Smyth R, Johnson C. Phonological encoding in the silent speech of persons
who stutter. Journal of Fluency Disorders. 2006; 31: 121. PMID: 16405985
27. Guidelines for audiologic screening. American Speech-Language-Hearing Association. 1997. Avail-
able: http://www.asha.org/policy.
28. Measuring Visual Acuity. University of Texas Health Science Center at San Antonio 2001. Available:
http://www.uthscsa.edu
29. Dunn LM, Dunn DM. Peabody Picture Vocabulary Test, Fourth Edition (PPVT-4). Circle Pines, MN:
American Guidance Service; 2007.
30. Williams K. Expressive Vocabulary Test, Second Edition (EVT-2). Circle Pines, MS: American Guid-
ance Service; 207.
31. Wagner R, Torgesen J, Rashotte CA. Comprehensive Test of Phonological Processing (CTOPP). Aus-
tin, TX: PRO-ED SAGE Publications; 1999.
32. Yairi E, Ambrose N. Early childhood stuttering. Austin: Pro-Ed, Inc.; 2005.
33. Riley G. Stuttering Severity Instrument for Children and Adults ( 3rd edition). Tigard, OR: C.C. Publica-
tions; 1994.
34. Snodgrass JG, Vanderwart M. A standardized set of 260 pictures: Norms for name agreement, image
agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning
and Memory. 1980; 6: 174215.
35. Juhasz BJ, Lai Y, Woodcock ML. A database of 629 English compound words:ratings of familiarity, lex-
eme meaning dominance, semantic transparency, age of acquisition, imageability, and sensory experi-
ence. Behavior Research Methods. 2014; 1: 116.
36. Word Scramble Maker [software]. Available: http://www.theteacherscorner.net
37. Bosshardt HD. Differences between stutterersand nonstutterersshort-term recall and recognition per-
formance. Journal of Speech, Language, and Hearing Research. 1993; 36: 286293.
38. Weber-Fox C, Spencer RM, Spruill JE, Smith A. Phonological processing in adults who stutter:
Electrophysiological and behavioral evidence. Journal of Speech, Language, and Hearing Research.
2004; 47: 12441258. PMID: 15842008
39. Namasivayam AK, van Lieshout P. Investigating speech motor practice and learning in people who
stutter. Journal of Fluency Disorders. 2008; 33: 3251. doi: 10.1016/j.jfludis.2007.11.005 PMID:
18280868
40. Smith A, Sadagopan N, Walsh B, Weber-Fox C. Increasing phonological complexity reveals height-
ened instability in articulatory coordination in adults who stutter. Journal of Fluency Disorders. 2010;
35: 118. doi: 10.1016/j.jfludis.2009.12.001 PMID: 20412979
41. Craig-McQuaide A, Akram H, Zrinzo L, Tripoliti E. A review of brain circuitries involved in stuttering.
Frontiers in Human Neuroscience. 2014; 8: 120.
42. Guenther FH, Ghosh S, Tourville JA. Neural modeling and imaging of the cortical interactions underly-
ing syllable production. Brain & Language. 2006; 96: 280301.
43. Hickok G, Poeppel D. Dorsal and ventral streams: a framework for understanding aspects of the func-
tional anatomy of language. Cognition. 2004; 92: 6799. PMID: 15037127
44. Vandermosten M, Boets B, Poelmans H, Sunaert S, Wouters J, Ghesquiere P. A tractography study in
dyslexia: Neuroanatomic correlates of orthographic, phonological and speech processing. Brain. 2012;
135: 935948. doi: 10.1093/brain/awr363 PMID: 22327793
45. Dick AS, Tremblay P. Beyond the arcuate fasciculous: consensus and controversy in the connectional
anatomy of language. Brain. 2012; 135: 35293550. doi: 10.1093/brain/aws222 PMID: 23107648
From Grapheme to Phonological Output
PLOS ONE | DOI:10.1371/journal.pone.0151107 March 10, 2016 19 / 19
... Different profiles of stuttering may emerge over the course of the life span and include overt difficulties with fluent speech production, negative feelings and thoughts, and work and social situation avoidance (e.g., Bloodstein et al., 2021;Yaruss, 2007;Yaruss & Quesal, 2004). Specific to linguistic factors thought to influence stuttering onset or persistence, many investigations have suggested that adults who stutter (AWS) have weaker language skills than adults who do not stutter (AWNS) in areas including lexical access and retrieval (e.g., McGill et al., 2016;Newman & Bernstein Ratner, 2007;Pellowski, 2011;Wingate, 1988), syntax (e.g., Cuadrado & Weber-Fox, 2003;Kleinow & Smith, 2000;Spencer et al., 2009;Tsiamtsiouris & Cairns, 2013), and phonological processing (e.g., Byrd et al., 2012Byrd et al., , 2015Castro et al., 2017;McGill et al., 2016). ...
... Different profiles of stuttering may emerge over the course of the life span and include overt difficulties with fluent speech production, negative feelings and thoughts, and work and social situation avoidance (e.g., Bloodstein et al., 2021;Yaruss, 2007;Yaruss & Quesal, 2004). Specific to linguistic factors thought to influence stuttering onset or persistence, many investigations have suggested that adults who stutter (AWS) have weaker language skills than adults who do not stutter (AWNS) in areas including lexical access and retrieval (e.g., McGill et al., 2016;Newman & Bernstein Ratner, 2007;Pellowski, 2011;Wingate, 1988), syntax (e.g., Cuadrado & Weber-Fox, 2003;Kleinow & Smith, 2000;Spencer et al., 2009;Tsiamtsiouris & Cairns, 2013), and phonological processing (e.g., Byrd et al., 2012Byrd et al., , 2015Castro et al., 2017;McGill et al., 2016). ...
... Lexical access must operate in an efficient and timely manner to retrieve words while speaking (Levelt, 2001). In behavioral studies, AWS have been found to be less accurate on lexical decision (Castro et al., 2017), word jumble (McGill et al., 2016), and word naming tasks (Newman & Bernstein Ratner, 2007;cf. Hennessey et al., 2008). ...
Article
Purpose Language abilities have long been thought to be weaker in adults who stutter (AWS) compared to adults who do not stutter (AWNS). However, it is unknown whether modality affects language performance by AWS in time pressure situations. This study aimed to examine lexical access and retrieval abilities of AWS in oral and typed modes. Method Fifteen AWS and 15 well-matched AWNS completed computer-administered letter fluency tasks. Adults were asked to orally produce words that began with one of two letter targets and type words that began with one of two alternate letters. Conditions were counterbalanced across participants. Results Generalized linear mixed-effects models were evaluated to determine the effects of group (AWS/AWNS), mode (oral/typed), and expressive vocabulary on letter fluency performance. Group predicted letter fluency such that AWS generated fewer items on both the oral and typed letter fluency tasks. Mode did not impact letter fluency results. Expressive Vocabulary Test scores predicted letter fluency similarly in both AWS and AWNS. Conclusions AWS were not penalized by oral task demands. AWS generated fewer items on the letter fluency tasks regardless of response mode, suggesting that they have weaker lexical access abilities. Furthermore, better expressive vocabulary skills were associated with better letter fluency performance in both groups.
... In this study, we anticipated that both groups would perform better in the neutral compared to the phonological condition (e.g., Byrd et al., 2015b;Mueller et al., 2003;Sweet et al., 2008;Vallar & Baddeley, 1984). Furthermore, based on evidence suggesting that AWS demonstrate deficits in lexical access, phonological loop, and central executive (e.g., Byrd et al., 2015aByrd et al., , 2015bMaxfield et al., 2012;McGill et al., 2016), we predicted that AWS would be significantly less accurate and slower than AWNS in the phonological condition at the 2-and 3-back levels, the most cognitively demanding of the required tasks. ...
... The lack of a phonological facilitation effect in AWS may suggest that AWS did not demonstrate a phonological priming effect or their priming effect was very weak to reach significance. This account can be supported by the reported difficulties of AWS with lexical access, phonological encoding (e.g., Byrd et al., 2015a;McGill et al., 2016;Sasisekaran & Weisberg, 2014;Weber-Fox et al., 2004) and weaker spreading activation abilities (e.g., Byrd et al., 2015b;Maxfield et al., 2012), For example, several studies have supported differences in phonological priming between AWS and AWNS (e.g., Maxfield et al., 2012;Maxfield et al., 2015;Wijnen & Boers, 1994;cf. Hennessey et al., 2008), which suggest that, compared to AWNS, AWS may require more overlapping phonological information between the prime and the target in order to demonstrate a phonological priming effect (e.g., Wijnen and Boers, 1994). ...
... Individual variability in participant mRT data and the varying degree of nonverbal and verbal processing each participant engaged into could have influenced the results. However, it needs to be noted that our sample size of 30 participants (i.e., 15 per group) is a common sample size used in other behavioral experimental paradigms investigating phonological encoding and phonological working memory in AWS (e.g., McGill et al., 2016, n = 13 participants per group; Weber-Fox et al., 2004, n = 11 participants per group), as well as in similar studies that have employed an N-back task with other clinical and typical populations (e.g., Braver et al., 2001, n = 28 healthy young adults; Wright et al., 2007, n = 9 adults with aphasia). Nevertheless, a larger sample size would better allow us to capture individual variability within each group. ...
Article
Purpose The purpose of this study was to investigate working memory in adults who do (AWS) and do not (AWNS) stutter using a visual N-back task. Processes involved in an N-back task include encoding, storing, rehearsing, inhibition, temporal ordering, and matching. Methods Fifteen AWS (11 males, 4 females;M = 23.27 years, SD = 5.68 years) and 15 AWNS (M = 23.47 years, SD = 6.21 years) were asked to monitor series of images and respond by pressing a “yes” button if the image they viewed was the same as the image one, two, or three trials back. Stimuli included images with phonologically similar (i.e., phonological condition) or phonologically dissimilar (i.e., neutral condition) names. Accuracy and manual reaction time (mRT) were analyzed. Results No difference was found between AWS and AWNS in accuracy. Furthermore, both groups were more accurate and significantly faster in 1- followed by 2- followed by 3-back trials. Finally, AWNS demonstrated faster mRT in the phonological compared to neutral condition, whereas AWS did not. Conclusion Results from this study suggest different processing mechanisms between AWS and AWNS for visually presented phonologically similar stimuli. Specifically, a phonological priming effect occurred in AWNS but not in AWS, potentially due to reduced spreading activation and organization in the mental lexicon of AWS. However, the lack of differences between AWS and AWNS across all N-back levels does not support deficits in AWS in aspects of working memory targeted through a visual N-back task; but, these results are preliminary and additional research is warranted.
... Stutterers' degraded speech perception appears to be primarily attributed to their aberrant lexical access to the mental lexicon (Lescht et al., 2022;McGill et al., 2016;Newman & Bernstein Ratner, 2007). The gating experiment delicately controls the length of each gate and evaluates the amount of acoustic information needed to access the mental lexicon (Grosjean, 1996;Wiener & Ito, 2016). ...
... Meanwhile, there are multiple functions, such as conceptualization, selection, and determination of a lexicon (Levelt, 2001;Levelt et al., 1999), involved in lexical access. Although not measured directly, some cognitive functions related to lexical access could remain abnormal in IWS (Maxfield et al., 2015;McGill et al., 2016;Pellowski, 2011), especially considering their poorer performance than IWNS when the full word was played in this study. For example, it has been articulated that IWS have difficulties in selecting a target word among a cohort of candidate words (Lescht et al., 2022;Maxfield, 2020). ...
Article
Purpose: Previous studies have shown that individuals who stutter exhibit abnormal speech perception in addition to disfluent production as compared with their nonstuttering peers. This study investigated whether adult Chinese-speaking stutterers are still able to use knowledge of statistical regularities embedded in their native language to recognize spoken words and, if so, how much acoustic information is needed to trigger this information. Method: Seventeen stutterers and 20 typical, nonstuttering controls participated in a gating experiment. All participants listened to monosyllabic words that consisted of syllables and lexical tones and were segmented into eight successive gates. These words differed in syllable token frequency and syllable-tone co-occurrence probability in line with a Chinese spoken word corpus. The correct syllable-only, correct tone-only, correct syllable-tone word, and correct syllable-incorrect tone responses were analyzed between the two groups using mixed-effects models. Results: Stutterers were less accurate overall than controls, with fewer correct syllables, tones, and their combination as words. However, stutterers showed consistent and reliable perceptual patterns triggered by statistical information of speech, as reflected by more accurate responses to high-frequency syllables, high-probability tones, and tone errors all in manners similar to those of nonstuttering controls. Conclusions: Stutterers' atypical speech perception is not due to a lack of statistical learning. Stutterers were able to perceive spoken words with phonological tones based on statistical regularities embedded in their native speech. This finding echoes previous production studies of stuttering and lends some support for a link between perception and production. Implications of pathological, diagnostic, and therapeutic conditions of stuttering are discussed.
... In terms of speech planning difficulties, a considerable amount of data implicate weaknesses in phonological encoding early in life as a key area of compromise (e.g., [7][8][9], for an opposing viewpoint, see [10,11]), particularly for children whose stuttering persists into adulthood (e.g., [12][13][14][15]). These data are further supported by studies which indicate that many adults who stutter (AWS) perform more poorly than adults who do DOI: 10.1159/000485657 not stutter (AWNS) when completing experimental tasks that rely on efficient phonological processing (e.g., nonword repetition [16][17][18]; silent error monitoring [19]; phoneme elision [16,17]; word jumble tasks [20]; silent rhyme judgment [21,22]; silent phoneme monitoring [23][24][25]). Although differences in the phonological abilities of AWS are not unequivocal (cf. ...
... To reiterate, the purpose of the present study was to replicate and extend the findings of Ozdemir et al. [33] in AWS rather than clarify this conceptual debate, and behavioral data may be insufficient to attribute causality exclusively to perception or exclusively to production. Nonetheless, in light of the findings of Lu et al. [70], it is possible that the differences observed in the present study support the larger body of data that implicate speech production issues in persons who stutter [12][13][14][15][16][17][18][19][20][21][22][23][24][25], rather than the perceptual monitor itself. Further neurophysiological examination may reveal that differences in monitoring in AWS, or lack thereof, are more closely associated with difficulties in perception rather than production in AWS. ...
Article
Full-text available
Background/aims: Previous studies employing a variety of tasks have demonstrated that adults who stutter (AWS) pre-sent with phonological encoding differences compared to adults who do not stutter (AWNS). The present study examined whether atypical preverbal monitoring also influenced AWS performance during one such paradigm - the silent phoneme monitoring task. Specifically, we investigated whether monitoring latencies for AWS were accelerated after the word's uniqueness point - the phoneme that isolates the word from all lexical competitors - as observed for AWNS when monitoring internal and external speech. Methods: Twenty adults (10 AWS, 10 AWNS) completed a silent phoneme monitoring task using stimuli which contained either (a) early uniqueness points (EUP), (b) late uniqueness points, or (c) no uniqueness point (NUP). Response latency when identifying word-final phonemes was measured. Results: AWNS exhibited the expected uniqueness point effect when monitoring internal speech; word-final phonemes were accessed more rapidly for words with EUP than NUP. In contrast, AWS did not differ in the phoneme monitoring speed. That is, AWS did not exhibit the expected uniqueness point effects. Conclusion: Findings suggest that inefficient or atypical preverbal monitoring may be present in AWS and support theories that implicate the internal speech monitor as an area of deficit.
... Previous work demonstrates that stutterers exhibit atypical lexical selection, phonological encoding, and phonetic encoding, as evidenced by stutterers demonstrating slower and less accurate responses than non-stutterers on language tasks (e.g., Castro et al., 2017;Lescht et al., 2022;McGill et al., 2016;Pelczarski et al., 2019;Sasisekaran et al., 2006;Vincent, 2017). Our study showed that words that place greater demands on these linguistic processes are more likely to be stuttered, which is possibly due to atypical language planning in stutterers. ...
Article
Purpose: Previous work shows that linguistic features (e.g., word length, word frequency) impact the predictability of stuttering events. Most of this work has been conducted using reading tasks. Our study examined how linguistic features impact the predictability of stuttering events during spontaneous speech. Methods: The data were sourced from the FluencyBank database and consisted of interviews with 35 adult stutterers (27,009 words). Three logistic regression mixed models were fit as the primary analyses: one model with four features (i.e., initial phoneme, grammatical function, word length, and word position within a sentence), a second model with six features (i.e., the features from the previous model plus word frequency and neighborhood density), and a third model with nine features (i.e., the features from the previous model plus bigram frequency, word concreteness, and typical age of word acquisition). We compared our models using the Area Under the Curve statistic. Results: The four-feature model revealed that initial phoneme, grammatical function, and word length were predictive of stuttering events. The six-feature model revealed that initial phoneme, word length, word frequency, and neighborhood density were predictive of stuttering events. The nine-feature model was not more predictive than the six-feature model. Conclusion: Linguistic features that were previously found to be predictive of stuttering during reading were predictive of stuttering during spontaneous speech. The results indicate the influence of linguistic processes on the predictability of stuttering events such that words associated with increased planning demands (e.g., longer words, low frequency words) were more likely to be stuttered.
... Heitmann et al (18), stated that individuals with stuttering have significant impairment in their skills of focused attention. McGill et al (21), noted that individuals with stuttering have deficits in phonological processing including vision to sound conversions and lexical access. Therefore, in this study these deficits in individuals with stuttering would have in turn lead to deficits in visual working memory capacity in individuals with stuttering thus leading to a poor recall of the visually presented sentences. ...
Article
Full-text available
It is common in literature to relate stuttering with some other deficit that interferes with communicative functions. Working memory comprises the system of human memory dedicated to both temporary storages of phonological detail and allocation of cognitive resources necessary for forming lasting memories. In this study we have analyzed the performance of individuals with stuttering on various working memory tasks. The aim of study is to compare the working memory abilities in individuals with stuttering and individuals with normal fluency on various working memory tasks. A total of 30 individuals with stuttering and 30 individuals with normal fluency in the age range of 18 - 40 years participated in the study. The Working Memory domain will be assessed using The Manipal Manual for Cognitive Linguistic Abilities (MMCLA) which consists of auditory word retrieval, auditory letter and number recall, auditory word list recall, auditory delayed sentence recall, visual practice recall, visual letter and number recall, visual word list recall and visual delayed sentence recall. Results revealed that the individuals with normal fluency had superior performance compared to the individuals with stuttering. Hence, it's helpful to understand the involvement of working memory in stuttering and incorporate working memory training along with the conventional fluency therapy.
... Various aspects of phonological processing (e.g., vision-to-sound conversions, sub-vocal rehearsal and lexical access) were found to differ between persons who stutter and fluent controls (Byrd, McGill, & Usler, 2015). For example, adults who stutter were slower and less accurate than fluent adults in a word jumble task, requiring them to silently manipulate scrambled letters to form a real word (McGill, Sussman, & Byrd, 2016). Although speech production is not involved in this task, rather the ability to sub-vocally rehearse the possible real words, persons who stutter show difficulties in its performance. ...
Article
People show better memory for words read aloud relative to words read silently, the Production Effect (PE). Vocalisation at study makes the produced (aloud) words more distinct than the non-produced (silent) words, hence more memorable. Such encoding distinctiveness is related to the additional processing of aloud words that is later used during retrieval. This study investigated the PE in dysarthric adults, characterised by speech production difficulties. Their memory performance (recognition) was compared to a group of healthy adults. Results showed a PE for both groups. The production benefit was significantly larger for the dysarthric adults, despite their overall memory performance being reduced relative to controls. The results demonstrate long-term verbal memory deficits in dysarthria, and suggest that vocalisation (although impaired) may assist in remembering. Hence, vocalisation may be used in intervention contexts with this population, to compensate for memory decrease.
... Stuttering is a multifactorial communication disorder that interrupts the forward flow of Speech production and in phonological working memory may be one of the factors that contribute to the difficulties in persons who stutter have establishing and maintaining fluent speech, particularly when presented with cognitively demanding tasks [1]. The growth in the stuttering appears during the initial school i.e. between 2-4 years and stuttering can be quite variable, especially in childhood. ...
Article
Objective: Stuttering is a complex communication and developmental speech disorder wherein forward flow of speech is interrupted by sound repetitions, words, prolongation of sounds and psychological and social effects. Most of the therapeutic approaches may focus more on cognitive, behavioral or psychological therapy. The aim of this study was to innovate the novel stuttering therapy procedure and to evaluate its efficiency in adults who stutters in accomplishing spontaneous fluent speech. Methods: The current proposal was designed as the pilot study and the participants were selected based on 3 treatment groups i.e., (i) speech-hand synchronization (SHS) (ii) Camperdown programme (CP) and (iii) control group CG). The equal number (n=10) of participants were selected in all the 3 groups and treatment sessions was carried out for 50 minutes per day for 10 weeks (5 days/week). Results: The stuttering severity instrument-4 was used to measure the scores of pre and post treatment. The overall, assessment of the speakers’ experience of stuttering, locus of control of behavior and speech satisfaction rating scale. Conclusion: The results of the current study conclude non-significant alterations and huge similar outcomes within the SHS and CP groups. This could be due to difference in the superior programme between SHS and CP, in terms of fluency, participants quality of life and satisfaction and internal locus of control.
... Based on imaging data (for a review see Neef et al., 2015), and behavioral studies (e.g. Byrd et al., 2015;McGill et al., 2016;Smith et al., 2010) demonstrating subtle vulnerabilities in the speech production systems of PWS, it is likely that there is more inherent conflicting activation, or noise, in the prearticulatory neural activity of PWS. Increased conflicting activation would increase the chances that the monitoring system would respond with cognitive control processes (e.g. ...
Article
Researchers, clinicians and people who stutter (PWS) find it difficult to adequately explain the variability across contexts in the frequency and severity of stuttering. The purpose of this paper is to present the Speech and Monitoring Interaction (SAMI) framework of stuttering that incorporates potential deficits within the speech production system and a contextually modulated monitoring system to provide a biologically plausible explanation for the contextual variability of stuttering. SAMI incorporates trait and state factors within the speech production system and the monitoring system that can account for stuttering variability. The specificity of the neural substrates of the monitoring system and how the monitor interacts with the speech production system can be used to drive hypotheses based research and computational modeling of the contextual variable stuttering. The framework can also be used clinically to inform clients of the social and emotional factors contributing to their stuttering; while highlighting the importance of addressing these factors in therapy.
Article
Full-text available
Presents a standardized set of 260 pictures for use in experiments investigating differences and similarities in the processing of pictures and words. The pictures are black-and-white line drawings executed according to a set of rules that provide consistency of pictorial representation. They have been standardized on 4 variables of central relevance to memory and cognitive processing: name agreement, image agreement, familiarity, and visual complexity. The intercorrelations among the 4 measures were low, suggesting that they are indices of different attributes of the pictures. The concepts were selected to provide exemplars from several widely studied semantic categories. Sources of naming variance, and mean familiarity and complexity of the exemplars, differed significantly across the set of categories investigated. The potential significance of each of the normative variables to a number of semantic and episodic memory tasks is discussed. (34 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
Article
Full-text available
Developmental stuttering is now generally considered to arise from genetic determinants interacting with neurologic function. Changes within speech-motor white matter (WM) connections may also be implicated. These connections can now be studied in great detail by high angular resolution diffusion MRI. Therefore, diffusion spectrum imaging (DSI) was used to reconstruct streamlines to examine white matter connections in persons who stutter (PWS) and normally fluent controls. WM morphology of the entire brain was assayed in eight right-handed male PWS and eight similarly aged right-handed fluent males. WM was exhaustively searched using a deterministic algorithm that identifies missing or largely misshapen tracts. To be abnormal, at least one third of a tract (defined as all streamlines connecting a pair of gray matter regions) was required to be missing in seven out of eight subjects in one group and not in the other group. Large portions of bilateral arcuate fasciculi, a heavily researched speech pathway, were abnormal in PWS. Conversely, all PWS had a prominent connection in the left temporo-striatal tract connecting frontal and temporal cortex that was not observed in nonstuttering control subjects. These previously unseen structural differences of WM morphology in classical speech-language circuits may underlie developmental stuttering.
Article
Full-text available
Stuttering has been the subject of much research, nevertheless its etiology remains incompletely understood. This article presents a critical review of the literature on stuttering, with particular reference to the role of the basal ganglia (BG). Neuroimaging and lesion studies of developmental and acquired stuttering, as well as pharmacological and genetic studies are discussed. Evidence of structural and functional changes in the BG in those who stutter indicates that this motor speech disorder is due, at least in part, to abnormal BG cues for the initiation and termination of articulatory movements. Studies discussed provide evidence of a dysfunctional hyperdopaminergic state of the thalamocortical pathways underlying speech motor control in stuttering. Evidence that stuttering can improve, worsen or recur following deep brain stimulation for other indications is presented in order to emphasize the role of BG in stuttering. Further research is needed to fully elucidate the pathophysiology of this speech disorder, which is associated with significant social isolation.
Article
Full-text available
Three lines of research have been pursued in the literature to study the link between phonology and stuttering: (1) effects of phonological complexity on the location (loci) of stutter events; (2) outcomes of standardized test measures in children who do and do not stutter; and (3) studies of phonological encoding in children and adults who stutter. This review synthesizes findings from these three lines of research to address the purported link between phonology and stuttering and its potential implications for stuttering treatment. Results from the loci studies offer some support for the role of phonological complexity in the occurrence of stuttering. Studies of performance in standardized tests of phonology have not identified differences between children who do and do not stutter. Studies of phonological encoding have been equivocal in reporting differences between children and adults who stutter and those who do not stutter. Several cautions are raised in interpreting the findings from the discussed studies, and despite the mixed findings, some implications for treatments are considered.
Article
Full-text available
Deficits in brain white matter have been a main focus of recent neuroimaging studies on stuttering. However, no prior study has examined brain connectivity on the global level of the cerebral cortex in persons who stutter (PWS). In the current study, we analyzed the results from probabilistic tractography between regions comprising the cortical speech network. An anatomical parcellation scheme was used to define 28 speech production-related ROIs in each hemisphere. We used network-based statistic (NBS) and graph theory to analyze the connectivity patterns obtained from tractography. At the network-level, the probabilistic corticocortical connectivity from the PWS group were significantly weaker than that from persons with fluent speech (PFS). NBS analysis revealed significant components in the bilateral speech networks with negative correlations with stuttering severity. To facilitate comparison with previous studies, we also performed tract-based spatial statistics (TBSS) and regional fractional anisotropy (FA) averaging. Results from tractography, TBSS and regional FA averaging jointly highlight the importance of several regions in the left peri-Rolandic sensorimotor and premotor areas, most notably the left ventral premotor cortex (vPMC) and middle primary motor cortex, in the neuroanatomical basis of stuttering.
Article
The purpose of the present study was to enhance our understanding of phonological working memory in adults who stutter through the comparison of nonvocal versus vocal nonword repetition and phoneme elision task performance differences. For the vocal nonword repetition condition, participants repeated sets of 4- and 7-syllable nonwords (n=12 per set). For the nonvocal nonword repetition condition, participants silently identified each target nonword from a subsequent set of three nonwords. For the vocal phoneme elision condition, participants repeated nonwords with a target phoneme eliminated. For the nonvocal phoneme elision condition, participants silently identified the nonword with the designated target phoneme eliminated from a subsequent set of three nonwords. Adults who stutter produced significantly fewer accurate initial productions of 7-syllable nonwords compared to adults who do not stutter. There were no talker group differences for the silent identification of nonwords, but both talker groups required significantly more mean number of attempts to accurately silently identify 7-syllable as compared to 4-syllable nonwords. For the vocal phoneme elision condition, adults who stutter were significantly less accurate than adults who do not stutter in their initial production and required a significantly higher mean number of attempts to accurately produce 7-syllable nonwords with a phoneme eliminated. This talker group difference was also significant for the nonvocal phoneme elision condition for both 4- and 7-syllable nonwords. Present findings suggest phonological working memory may contribute to the difficulties persons who stutter have establishing and/or maintaining fluent speech. Educational Objectives: (a) Readers can describe the role of phonological working memory in planning for and execution of speech; (b) readers can describe two experimental tasks for exploring the phonological working memory: nonword repetition and phoneme elision; (c) readers can describe how the nonword repetition and phoneme elision skills of adults who stutter differ from their typically fluent peers. Copyright © 2015 Elsevier Inc. All rights reserved.
Article
Since the work of Taft and Forster (1976), a growing literature has examined how English compound words are recognized and organized in the mental lexicon. Much of this research has focused on whether compound words are decomposed during recognition by manipulating the word frequencies of their lexemes. However, many variables may impact morphological processing, including relational semantic variables such as semantic transparency, as well as additional form-related and semantic variables. In the present study, ratings were collected on 629 English compound words for six variables [familiarity, age of acquisition (AoA), semantic transparency, lexeme meaning dominance (LMD), imageability, and sensory experience ratings (SER)]. All of the compound words selected for this study are contained within the English Lexicon Project (Balota et al., 2007), which made it possible to use a regression approach to examine the predictive power of these variables for lexical decision and word naming performance. Analyses indicated that familiarity, AoA, imageability, and SER were all significant predictors of both lexical decision and word naming performance when they were added separately to a model containing the length and frequency of the compounds, as well as the lexeme frequencies. In addition, rated semantic transparency also predicted lexical decision performance. The database of English compound words should be beneficial to word recognition researchers who are interested in selecting items for experiments on compound words, and it will also allow researchers to conduct further analyses using the available data combined with word recognition times included in the English Lexicon Project.