See discussions, stats, and author profiles for this publication at: [Link]

net/publication/246427697

Script training and automaticity in two individuals with aphasia

Article in Aphasiology · May 2005

DOI: 10.1080/02687030444000877

CITATIONS READS

95 2,478

4 authors, including:

Gina Youmans Audrey Holland

Long Island University University of Arizona
6 PUBLICATIONS 432 CITATIONS 166 PUBLICATIONS 7,242 CITATIONS

SEE PROFILE SEE PROFILE

Michelle S Bourgeois
University of South Florida
124 PUBLICATIONS 4,343 CITATIONS

SEE PROFILE

All content following this page was uploaded by Michelle S Bourgeois on 18 September 2016.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [University of South Florida]
On: 03 February 2015, At: 12:36
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Aphasiology
Publication details, including instructions for authors and subscription
information:
[Link]

Script training and automaticity in two

individuals with aphasia
a b c d
Gina Youmans , Audrey Holland , Maria Muñoz & Michelle Bourgeois
a
Long Island University, Brooklyn, NY, USA
b
University of Arizona, Tucson, AZ, USA
c
University of Tennessee, Knoxville, TN, USA
d
Florida State University, Tallahassee, FL, USA
Published online: 18 Aug 2010.

To cite this article: Gina Youmans , Audrey Holland , Maria Muñoz & Michelle Bourgeois (2005)
Script training and automaticity in two individuals with aphasia, Aphasiology, 19:3-5, 435-450, DOI:
10.1080/02687030444000877

To link to this article: [Link]

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our
licensors make no representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views expressed in this
publication are the opinions and views of the authors, and are not the views of or endorsed
by Taylor & Francis. The accuracy of the Content should not be relied upon and should be
independently verified with primary sources of information. Taylor and Francis shall not be liable
for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in
relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use
can be found at [Link]
 APHASIOLOGY, 2005, 19 (3/4/5), 435±450

Script training and automaticity in two individuals

with aphasia
Gina Youmans
Long Island University, Brooklyn, NY, USA
Audrey Holland
University of Arizona, Tucson, AZ, USA
Downloaded by [University of South Florida] at 12:37 03 February 2015

Maria L. MunÄoz
University of Tennessee, Knoxville, TN, USA
Michelle Bourgeois
Florida State University, Tallahassee, FL, USA

Background: Automatic production of spoken language is compromised in many individuals

with aphasia whose speech is consistently effortful. Linguistic-based and functional treat-
ments for aphasia both may help to return some automaticity to language production.
However, neither approach specifically targets automaticity of speech production for indi-
vidual with aphasia.
Aims: A treatment protocol was developed to facilitate the automatic spoken production of
trained scripts in specific functional contexts by individuals with aphasia. The purpose of this
study was to measure its treatment outcomes.
Methods & Procedures: Two individuals with non-fluent aphasia participated. A multiple
baseline design across behaviours examined the acquisition of personally relevant short
scripts. Script production was probed at the initiation and termination of treatment, and
during weekly treatment sessions. Scripts were transcribed and analysed according to several
measures of accuracy and fluency.
Outcomes & Results: All scripts were mastered, and scripted speech productions were judged
to have become more automatic based on naturalness and stability of speech, increased
speaking rate, and relatively errorless production. Automatic script production also gen-
eralised to novel conversation partners and novel cues in a limited fashion.
Conclusions: Script training as used here was a successful treatment approach. It may be an
effective, practical, and functional approach to the treatment of aphasic individuals with
moderately spared comprehension and cognitive difficulties and with significant expressive
speech difficulties.

Speech production is partly automatic. Once an intended meaning is in place, gramma-

tical and phonological encoding of the message and its articulation flow automatically,
rapidly, and without conscious effort or attention (Levelt, 1989; Seuren, 1978). However,
for individuals with expressive language difficulties due to aphasia, speech is no longer

Address correspondence to: Gina Youmans, Long Island University, Department of Communication Dis-
orders, 1 University Plaza, Metcafe Building Suite 257, Brooklyn, NY 11201, USA. Email: ggb5466@[Link]

# 2005 Psychology Press Ltd

[Link] DOI:10.1080/02687030444000877
 436 YOUMANS ET AL.

automatic. For them, speaking becomes an effortful struggle to find and produce words to
convey ideas.
Functional treatment approaches emphasise communication within contexts that are
natural and personally relevant to the speaker (Aten, 1986; Holland, 1991; Hopper &
Holland, 1998). Functional treatments can effectively increase communicative compe-
tence, particularly within the contexts that were practised in therapy (Aten, Caligiuri, &
Holland, 1982; Hinckley, Patterson, & Carr, 2001). Because functional communication
therapy focuses on communication of whole messages in a versatile, realistic fashion, this
approach does not easily lend itself to the massed, specific repetition necessary to
establish automaticity of specific language skills or elements of conversation. However,
emphasis on the whole message may shift an aphasic speaker's conscious attention away
from specific language component skills, such as word finding and grammar, and con-
Downloaded by [University of South Florida] at 12:37 03 February 2015

sequently permit preserved ability to produce language automatically to surface effort-

lessly. This deflection of conscious attention away from specific components of language
production may be a generally unrecognised strength of functional approaches to aphasia
treatment.
Traditional component-based approaches to aphasia treatment also do not specifi-
cally target automatic speech production, although they may indirectly facilitate auto-
maticity. Traditional approaches generally involve drilling of specific compromised
component language skills such as naming or grammatical rule use (for example,
Hickin, Best, Herbert, Howard, & Osborne, 2001; Mitchum, Haendiges, & Berndt,
1995; Nickels, 2000). These approaches view language as a group of component
skills, and language disorders such as aphasia as impairments in a subset of these
component skills (LaBerge & Samuels, 1974; Logan, 1978). Traditional approaches
parallel a modal theory of automaticity, which claims that normally automatic, compo-
nent processes can be singled out, practised, and re-automatised. Component-based
approaches to aphasia therapy have been reasonably successful in effecting positive
change in the language abilities of aphasic individuals (Hinckley et al., 2001;
Jacobs & Thompson, 2000; Nickels, 2000). However, they often invoke active self-
monitoring, and conscious application of specific linguistic rules. To the extent that
an approach focuses attention on specific language components such as grammar or
conscious rule use, it may promote a self-conscious stance in aphasic speakers, for
whom every word must be considered and weighed. Such an exacerbation of the voli-
tional, conscious production of language runs counter to a return to normal, auto-
matic language production.
Although both functional and traditional aphasia therapy approaches may facilitate a
degree of automatic speech production, the reinstatement of automatic speech is not
typically an overtly stated goal of aphasia treatment. Holland and colleagues have
recently developed a script-training approach to aphasia treatment that focuses directly on
re-injecting islands of automatic natural language production into the speech of indivi-
duals with aphasia. (For a succinct description of the approach, see Holland & Ramage,
2004). Script training first identifies and then practises personally relevant and functional
scripts. Scripts are performed in natural, conversational contexts and practised in a whole
task fashion. However, script training also incorporates elements of more traditional
approaches in that clients practise their speaking script in a repetitive, cue-based, massed
drilling fashion. Script training methods arise from an instance theory of automatisation
(Logan, 1988). Instance theory suggests that automaticity is achieved by retrieving
memories of complete, context-bound, skilled performances. In this view, tasks are not
broken down and practised as component skills. Instead, they are practised in an holistic,
 SCRIPT TRAINING AND AUTOMATICITY 437

integrated fashion. Because they are context bound, they are not expected to generalise
easily to untrained contexts.
In the present study, two participants with aphasia intensively practised speaking
scripts as monologues and subsequently in conversational contexts. This study also
probed generalisation by assessing script production with untrained communication
partners. The purpose of the study was to determine if the participants attained automatic
speech production as measured by relatively errorless speaking, increased speaking rate,
and relative consistency in their ability to use the scripts.

METHOD
Participants
Downloaded by [University of South Florida] at 12:37 03 February 2015

Two adults, MN and FG, with chronic non-fluent aphasia secondary to stroke were
selected for script training. MN was a 43-year-old female, premorbidly right-handed, who
was 12 years post-onset of an embolic stroke with middle cerebral artery occlusion. The
stroke resulted in severe, non-fluent aphasia and right hemiparesis. Administration of the
Western Aphasia Battery (WAB; Kertesz, 1988) prior to script training classified MN
with Broca's aphasia (AQ: 48.7). Comprehension was impaired for complex and abstract
sentences in conversation; her speech output was halting, telegraphic, and agrammatic.
She had no dysarthria and although she had difficulty initiating speech, and demonstrated
occasional articulatory groping, she did not have other distinguishing apraxic behaviours.
MN had participated in script training 1 year prior to this study, using a slightly different
paradigm and with different scripts.
FG was a 60-year-old, premorbidly right-handed male, 31 years post-onset of a CVA
which resulted in chronic aphasia and right hemiparesis. A WAB administered prior to
treatment indicated conduction aphasia (AQ: 68.8). In conversation, FG comprehended
moderately complex conversational speech; his speech was non-fluent due to word-
finding difficulties, repetitions, and frequent self-corrections of errors. He had mild
dysarthria, characterised by imprecise articulation. He had no characteristics of apraxia of
speech. No specific information about lesion localisation is available for either MN
or FG.

Procedures
All sessions were held in a handicap-accessible room at the University of Arizona. Each
participant generated functional, personally relevant topics. When three topics were
selected, the script was written, and presented to the participant for their input concerning
its wording. Script length was limited to three or four short, relatively simple sentences.
The scripts are presented in the Appendix. Participants attended 30±45-minute script-
training sessions three times weekly. Details of the method are described below

Script training procedures. Scripts were trained one phrase at a time. A cueing
hierarchy was used to train new material: phrase repetition, choral reading of phrases
with the clinician, and then independent production. Written phrases were available for
the participant to consult. When the participant could produce a newly trained phrase
independently at least 20 consecutive times, the next script phrase was added to pre-
viously mastered portions of that script. In addition to treatment sessions, participants
were expected to practise scripts at home, for 15 minutes per day. Participants were
given audio cassettes containing client- and clinician-produced versions of the script
 438 YOUMANS ET AL.

portions being mastered. Both patients kept records of their home practice that indicated
daily compliance. Throughout script training, verbatim production and repetition was
practised, and verbatim recitation was the standard for mastery.
Once a script was mastered, generalisation training was initiated. The clinician pur-
posefully varied her responses and comments to help the participants make their scripts
more resilient and more flexible. At this point, monologue scripts were practised in
conversational form. In addition, seven different novel conversation partners, (students
and faculty members) participated in conversations. All partners had been trained in
supported communication, and all had considerable experience communicating with
persons with aphasia. Novel conversation partners were informed of the general topic
before script generalisation training, but had no precise knowledge of scripts' content. At
the end of generalisation training for a given script, novel partners were instructed to try
Downloaded by [University of South Florida] at 12:37 03 February 2015

to ``throw'' the script participant by saying something appropriate but surprising during
the conversation.

Data coding and analysis. At the beginning of each baseline and treatment session,
participants conversed on each script topic, or attempted to produce each learned script
from memory, without clinician feedback, and before the initiation of treatment. Two
minutes was allowed for the participant to converse on each script topic. These probes
were transcribed and coded for the following dependent variables: percent script correct,
number of errors produced, and speaking rate. For each session, a single data point on
each variable for each script topic was obtained from the probe transcript and graphically
plotted for visual inspection.
Percent script correct was calculated by dividing the number of script words produced
by the total number of words in the target script. Percent script correct was the primary
dependent variable; phase change decisions were based on this variable. Appropriate
substitutions and circumlocutions were not accepted as correct in accordance with
automaticity training, in which verbatim repetition is stressed.
Errors were defined as non-communicative word or phrase repetitions, fillers, pauses
of 3 seconds or more, and unrecognisable utterances. A frequency count of errors per
probe was calculated. Although circumlocutions and real-word substitutions were not
counted as correct, neither were they counted as errors.
Speaking rate. The duration of each script or untrained script topic was timed, and a
word per minute rate was calculated. Words were defined as any real word or recogni-
sable word approximation that was not an error.

Experimental design

A multiple baseline design across behaviours was used to assess the effectiveness of
script training on script production, as measured by the primary dependent variable of
percent script correct. Experimental control was established by gains in percent script
correct occurring only upon initiation of training for each specific script, and not before.
In addition, the variables of speaking rate and error production were examined for all
sessions, to more fully explore the process of script learning, and to assess automatisation
of scripted speech production. Throughout all conditions, dependent variable measures
were collected from the script spoken at the beginning of each session, without any
cueing or feedback, as previously described.
 SCRIPT TRAINING AND AUTOMATICITY 439

Baseline. Participants attempted to speak on each untrained script topic, without

clinician feedback. Stable baselines were defined as no visible upward trend upon
inspection of the data, across a minimum of three sessions.

Script training. Baseline cessation and script training initiation were contingent upon
the baseline stability of the script topic in question, and upon mastery of any previously
trained script material. The criterion for script mastery was independent script produc-
tion, without cueing or feedback, at the beginning of a session, at 90% script correct, for
at least two consecutive sessions.

Maintenance and generalisation. Following script training, performance data on

dependent variable measures continued to be collected and plotted for visual inspection.
Downloaded by [University of South Florida] at 12:37 03 February 2015

During Maintenance, participants continued to practise script material at home, and to

review previously learned scripts in treatment sessions. In addition, generalisation
training was begun. Generalisation probes were given at the beginning of each session.

Reliability
Probe data were transcribed and triple checked for accuracy by the first author. A total of
30% of the transcripts were re-coded for reliability of errors, speaking rate, and percent
script correct by an independent coder. Transcripts for reliability coding were randomly
selected across baseline, treatment, and generalisation sessions. For MN, speaking rate
was coded with 97% agreement, errors with 94% agreement, and percent script correct
with 97% agreement. For FG, coder agreement was 87% for speaking rate, 84% for
errors, and 97% for percent script correct.

Social validity
To validate the effectiveness of script training, nine independent unfamiliar coders rated a
representative baseline and maintenance sample from each script for each participant.
Two coders were SLP students familiar with aphasia; the other seven were unfamiliar
with speech disorders. Samples were selected randomly from a subset of three to four
probes judged by the first author to be representative of either baseline or maintenance
performance on the script in question; markedly poor or superb performances were not
included. These samples were presented to coders in a fixed random order. Coders rated
each script sample for speaking rate, naturalness, and informativeness. Raters marked a 7-
inch unscaled line that had descriptive adjectives at either end. Pre and post ratings were
not included on the same page. Ratings were assessed by measuring the distance (in mm)
from the leftmost point of each line to the mark made by the rater. Paired t-tests (a =
0.05) were used to compare these baseline and maintenance ratings, and a Bonferroni
correction was applied.

RESULTS

Script-training measures
Percent script correct. MN and FG reached the criterion for mastery for all three of
their scripts, as shown in Figures 1 and 2 respectively. Percent script correct scores
remained low and relatively stable during the baseline condition, and improved markedly
for each topic with script training. The criterion for script mastery was 90% accuracy in
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 1. Percent of script words produced correctly across sessions by MN, on each topic. White triangles
represent generalisation probes.

440
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 2. Percent of script words produced correctly across treatment sessions by FG, on each topic. White
triangles represent generalisation probes. Arrow indicates new topic 3 introduction, which occurred due to
preference of FG.

441
 442 YOUMANS ET AL.

percent script correct across at least two consecutive sessions. Time required for mastery
of all three scripts was similar for both participants: MN mastered three scripts in 25
sessions, FG in 26 sessions. MN mastered scripts 1, 2, and 3 in 8, 11, and 6 sessions,
respectively. FG mastered scripts 1, 2, and 3 in 10, 11, and 5 treatment sessions,
respectively. At the end of training, script performances stabilised; participants
consistently maintained their ability to produce the scripts with 98 to 100 % accuracy.

Speaking rate. As shown in Figures 3 and 4, over the course of treatment MN and
FG generally demonstrated an immediate slowing of speaking rate followed by an
increasing, but variable, speaking rate. MN's baseline average speaking rate of 13 wpm
on topic 1 increased to 90 wpm at the end of maintenance. For script 2, her average
baseline speaking rate of 24 wpm increased to 74 wpm, and for topic 3, MN's average
Downloaded by [University of South Florida] at 12:37 03 February 2015

baseline speaking rate of 32.5 wpm increased to 111 wpm. FG's average baseline
speaking rate for topic 1 of 38.6 wpm increased to 93 wpm at the end of maintenance. For
topic 2, his baseline speaking rate of 31.8 wpm increased to 91 wpm, and for topic 3, his
baseline of 35.8 wpm increased to 80 wpm. These results are presented in Figures 3
and 4.

Errors. The error results are presented in Figures 5 and 6. The total number of errors
produced during each script was variable from session to session. Although a decrease in
errors did not consistently coincide with implementation of script training, there was a
decrease in error production by the end of script maintenance for both participants. MN
repeatedly produced errorless scripts on topics 1 and 3 across maintenance sessions, and
flawlessly produced script 2 once during maintenance. FG achieved several errorless
productions of his first two scripts. Error production for FG on script topic 3 increased on
the last treatment session, at the end of script training.

Generalisation
Generalisation probe data are plotted in Figures 1±6. Participants were fairly successful at
generalising their scripted speech, but their performances varied widely. All scripts were
produced at 80% to 100% accuracy in generalisation conversations. For scripts 2 and 3
for MN and scripts 1 and 2 for FG, the most rapid speaking rates occurred during
generalisation.

Social validity
Social validity ratings were significant for all three measures. MN's speech was rated as
significantly more natural, t(26) = 78.32; p < .01, more rapid, t(26) = 75.33; p < .001,
and more informative, t(26) = 78.53; p < .001, at maintenance than at baseline. For FG,
ratings of speech naturalness, t(26) = ±2.92; p < .007, speaking rate, t(26) = 72.60; p =
.015, and informativeness, t(26) = 74.60; p < .001, were also significant.

DISCUSSION
The results of this study suggest that script training was an effective treatment for these
participants with non-fluent aphasia. Both mastered all scripts, and produced scripts from
memory consistently, with 97±100% accuracy. As a result of training, speaking of these
scripts improved from a halting, ineffective, obviously impaired style at baseline to fluid,
competent speech, produced flexibly in conversation with novel partners. Independent
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 3. Speaking rate in words per minute across sessions for MN, on each topic. White triangles represent
generalisation probes.

443
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 4. Speaking rate in words per minute across sessions for FG, on each topic. White triangles represent
generalisation probes. Arrow indicates new topic 3 introduction, which occurred due to preference of FG.

444
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 5. Number of errors produced by MN across sessions, for each topic. White triangles represent
generalisation probes.

445
Downloaded by [University of South Florida] at 12:37 03 February 2015

Figure 6. Number of errors produced by FG across sessions for each topic. White triangles represent gen-
eralisation probes. Arrow indicates new topic 3 introduction, which occurred due to preference of FG.

446
 SCRIPT TRAINING AND AUTOMATICITY 447

raters coded trained script performances as significantly more informative, more natural
sounding, and spoken at a significantly more normal rate than were the baseline script
performances.
This study suggests that massed practice of specific short phrases or sentences can
result in the automatic, effortless production of phrases, even in individuals with mod-
erate to severe expressive speech difficulties. Theorists generally concur that as a skill
becomes automatic, performance becomes rapid, relatively errorless, and stable (Sega-
lowitz & Segalowitz, 1993). Experiments that specifically establish automaticity of
skilled performance generally examine the speed-up and stabilisation of brief response
latencies associated with new learning (Logan, 1988; Segalowitz & Segalowitz, 1993).
Response latency does not appear to be an appropriate measure of script learning.
However, measures of error production, percent script correct, and speaking rate were
Downloaded by [University of South Florida] at 12:37 03 February 2015

taken, not only as indicators of script acquisition but also to assess informally the ability
to speak mastered script material automatically. Both participants repeatedly spoke all
mastered script material faster and with few, if any, errors. In addition, script perfor-
mance stabilised for all topics during maintenance. This speed-up in performance,
relatively errorless production, and stability of performance all suggest that scripted
speech did become automatic, particularly when compared to the relatively poor per-
formances on these measures during baseline. Automatic speech should be effortless. The
absence of struggle and self-correction in mastered script productions, and the relative
ease with which these individuals produced their scripts, also suggest that participants
achieved this goal.
During script maintenance and generalisation, participants continued to practise their
mastered scripts at home. These scripts were reviewed periodically. In addition, mastered
scripts were practised in less constrained and more spontaneous situations. Speaking rate
generally continued to improve with this additional, post-mastery practice. Generalisation
training was an important component of this script-training study, because the final goal
was the functional use of automatic, scripted speech in conversation. Specific practice in
conversation was particularly important to those scripts initially trained in monologue
form, because although monologue practice facilitated initial learning, rote monologue
production is seldom useful in daily communication. Both participants used their scripts
successfully in conversations with novel partners. It is likely that the automatisation of
scripts combined with specific generalisation practice promoted an ability to use scripts
flexibly and functionally in conversation.
Measures of script production, such as percent script correct, speaking rate, and
number of errors, became quite variable during generalisation, although overall script
performance during generalisation was good. This variability was interpreted as the
perturbation of a relatively stable performance by new performance demands inherent in
conversing with novel partners. In general, participants appeared capable of meeting such
challenges least in part, because automatisation of scripted speech allowed them to focus
on their comprehension of their partner's speech, and then to respond appropriately to
their partner's unexpected and novel conversational cues.
This study suggests that script training resulted in acquisition of islands of automatic
speech that could be produced on a specific topic with some degree of flexibility. Script
training may be appropriate for individuals who desire to speak relatively normally (if in
a limited fashion) on a certain topic or in a social situation that is very important to them
personally.
Although functional, compensatory communication strategies are always important for
daily survival, the strengths of a communication approach such as script training, which
 448 YOUMANS ET AL.

emphasises a return to relatively normal verbal communication, even within a very

limited context, should not be underestimated. The script topics chosen by the current
participants testify to the importance, at least to these individuals, of being able to
produce even limited relatively normal speech. MN wished to be able to simply ask her
significant other about his feelings, and about his day, and to tell him in turn how she was
feeling and about her own daily experiences. She felt that her limited ability to produce
one- to two-word utterances with accompanying gestures was very unsatisfactory for
these daily personal exchanges, which are so important to personal relationships and
which normal communicators take for granted. And although FG was able to commu-
nicate through gesture and limited single-word writing in a functional manner during
grocery store outings, he was unsatisfied with this mode of communication, and speci-
fically wished to be able to ask for what he wanted, to relax and not worry about being
Downloaded by [University of South Florida] at 12:37 03 February 2015

understood or getting a message across. Script training was both practical and functional
for these participants. Time for script mastery ranged from five to eleven sessions.
Targeting a specific short script for massed practice is fairly practical, therefore, and
results appear to be achievable in 2±4 weeks of therapy occurring three times weekly, for
individuals with abilities similar to those in the present study.

Future directions
Research is needed to further evaluate the effects of this treatment protocol for indivi-
duals with aphasia. Additionally, the utility of this approach for individuals with more
significant cognitive or comprehension impairments might be probed. Subsequent
research might focus on the type of generalisation practice that might enhance or max-
imise the maintenance of scripts. Research that explores the limitations and functional
applications of this treatment approach will help establish script training as a treatment
option for individuals with aphasia.

REFERENCES

Aten, J. L. (1986). Functional communication treatment. In R. Chapey (Ed.), Language intervention strategies in
adult aphasia (2nd ed.). Baltimore, MD: Williams & Wilkins.
Aten, J. L., Caligiuri, M., & Holland, A. L. (1982). The efficacy of functional communication therapy for
chronic aphasic patients. Journal of Speech and Hearing Disorders, 47, 93±96.
Hickin, J., Best, W., Herbert, R., Howard, D., & Osborne, F. (2001). Phonological therapy for word-finding
difficulties: A re-evaluation. Aphasiology, 16, 981±999.
Hinckley, J., Patterson, J., & Carr, T. (2001). Differential effects of context-based and skill-based treatment
approaches: Preliminary findings. Aphasiology, 15, 463±476.
Holland, A. L. (1991). Pragmatic aspects of intervention in aphasia. Journal of Neurolinguistics, 6(2), 197±211.
Holland, A. L., & Ramage, A.E. (2004). Learning from Roger Ross: A clinical journey. In J. F. Duchan & S.
Byng (Eds.), Challenging aphasia therapies. Hove, UK: Psychology Press.
Hopper, T., & Holland, A. (1998). Situation-specific training for adults with aphasia: An example. Aphasiology,
12(10), 933±944.
Jacobs, B. J., & Thompson, C. K. (2000). Cross-modal generalisation effects of training noncanonical sentence
comprehension and production in agrammatic aphasia. Journal of Speech, Language, & Hearing Research,
43(1), 5±20.
Kertesz, A. (1988). Western Aphasia Battery. New York: The Psychological Corporation.
LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading.
Cognitive Psychology, 6, 293±323.
Levelt, W. (1989). Speaking: From intention to articulation. London: The MIT Press.
Logan, G. D. (1978). Attention in character classification: Evidence for the automaticity of component stages.
Journal of Experimental Psychology: General, 107, 32±63.
 SCRIPT TRAINING AND AUTOMATICITY 449

Logan, G. D. (1988). Toward an instance theory of automatisation. Psychological Review, 95(4), 492±527.
Mitchum, C. D., Haendiges, A. N., & Berndt. R. S. (1995). Model-guided treatment to improve written sentence
production: A case study. Aphasiology, 7, 71±109.
Nickels, L. (2000). Therapy for naming disorders: Revisiting, revising and reviewing. Aphasiology, 16(10/11),
935±979.
Segalwitz, N. S., & Segalwitz, S. J. (1993). Skilled performance, practise, and the differentiation of speed-up
from automatisation effects: Evidence from second language word recognition. Applied Psycholinguistics,
14, 369±385.
Seuren, P. A. M. (1978). Grammar as an underground process. In A. Sinclair, R. J. Jarvella, & W. J. M. Levelt
(Eds.), The child's conception of language. Heidelberg: Springer.

APPENDIX
Script topics and contents
Downloaded by [University of South Florida] at 12:37 03 February 2015

MN Script 1: Talking about Feelings (Conversational training form)

How are you feeling? I feel happy. Are you angry? You look confused.

(Words in italics are examples of words practised interchangeably. Other words practised within this script:
aggravated, mad, sad, wonderful, great.)
MN Script 2: Talking about our day (Conversational training form)

How was your day? I went to Desert Fitness. I went to Albertson's. I worked on my computer. I worked on
my painting.

(Other words practised: group, therapy, Foothills mall, the party, Thunder Brewery.)
MN Script 3: Party Hostess (trained initially in monologue form)

Hi, thank you for coming! What can I get you to drink? We have beer, and wine and soda. Please have
something to eat.

FG Script 1: Childhood in Hawaii (trained initially in monologue form)

I was born and raised in Hawaii. I grew up in Lanaki. I had to climb fifty four stairs to get to my house. It had
a beautiful view of the island.

FG Script 2: Conversational Questions (Conversational training form)

Hi, how are you? Where are you living? Where are you from? Where are you working?

FG Script 3: Asking for Grocery Store Items (Conversational training form)

Where is the juice? Where is the meat? How much is the yoghurt? How much is the milk?

(Other words practised: cantaloupe, melon, bread, fruit, cheese, cottage cheese, cereal.)
 Downloaded by [University of South Florida] at 12:37 03 February 2015

View publication stats