Neuropsychologia 64 (2014) 282–288

Contents lists available at ScienceDirect

Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia

Cognitive correlates of narrative impairment in moderate traumatic

brain injury
Andrea Marini a,b,n, Marina Zettin c,d, Valentina Galetto c,d
a
Dipartimento di Scienze Umane, University of Udine, Via Margreth, 3-33100 Udine, Italy
b
IRCCS “E. Medea: La Nostra Famiglia”, San Vito al Tagliamento (Pn), Italy
c
Centro Puzzle, Torino, Italy
d
Dipartimento di Psicologia, Università di Torino, Italy

art ic l e i nf o a b s t r a c t

Article history: Traumatic brain injuries (TBIs) are often associated with communicative deficits. The incoherent and
Received 16 December 2013 impoverished language observed in non-aphasic individuals with severe TBI has been linked to a
Received in revised form problem in the global organization of information at the text level. The present study aimed to analyze
10 August 2014
the features of narrative discourse impairment in a group of adults with moderate TBI (modTBI). 10 non-
Accepted 24 September 2014
Available online 2 October 2014
aphasic speakers with modTBI and 20 neurologically intact participants were recruited for the
experiment. Their cognitive, linguistic and narrative skills were thoroughly assessed. The persons with
Keywords: modTBI exhibited normal phonological, lexical and grammatical skills. However, their narratives were
Moderate traumatic brain injury characterized by lower levels of Lexical Informativeness and more errors of both Local and Global
Narrative analysis
Coherence that, at times, made their narratives vague and ambiguous. Significant correlations were
Neurolinguistics
found between these narrative difficulties and the production of both perseverative and non-
Neuropsychology
Language perseverative errors on the WCST. These disturbances confirm previous findings which suggest a deficit
at the interface between cognitive and linguistic processing rather than a specific linguistic disturbance
in these patients.
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Traumatic brain injuries (TBIs) are often associated with commu-
nicative deficits. Factors such as the seriousness of the trauma and
lesion site and extension might alter communicative skills in different
ways (Douglas, 2004; Watt and Douglas, 2006). These individuals
often experience aphasic symptoms that include phonological, lexical
and/or grammatical disturbances. However, even those without
aphasic symptomatology might show altered communicative beha-
viors. Their speech can be incoherent, impoverished (Hartley and
Jensen, 1992; Davis and Coelho, 2004), and pragmatically inadequate
(e.g., Togher et al., 1997; Bond and Godfrey, 1997; Angeleri et al., 2008;
Tu et al., 2011). Such difficulties might be the consequence of a
problem in the global organization of information at the macro-
linguistic level of discourse processing rather than at the lexical and
syntactic (i.e., microlinguistic) level (Brookshire et al., 2000; Coelho,
2002; Body and Perkins, 2004; Le et al., 2011; Marini et al., 2011a;
Carlomagno et al., 2011; Galetto et al., 2013). The high-level linguistic
skills that guide narrative discourse formulation are mediated by a
n findings can be extended to narrative discourse performance in
Corresponding author at: Dipartimento di Scienze Umane, Università di Udine,
Via Margreth, 3-33100 Udine, Italy.
E-mail address: andrea.marini@uniud.it (A. Marini).
wide range of cognitive skills (e.g., working memory, attention and
executive functions; Sim et al., 2013; Mozeiko et al., 2011) that are
usually impaired because of the fronto–temporo-parietal lesions and
diffuse axonal damage induced by the brain injury (Stierwalt and
Murray, 2002; Mendez et al., 2005; Youse and Coelho, 2005; Scheid et
al., 2006; Silver et al., 2009; Erez et al., 2009). Notably, recent evidence
suggests that some of these neural networks are implicated also in
discourse processing (e.g., Coelho et al., 2012; Marini and Urgesi,
2012). There are unresolved issues at both the clinical and theoretical
levels. On the clinical level, the role played by the severity of the
trauma on linguistic skills. On the theoretical level, the linguistic
abilities of persons with TBI have rarely been the object of a
comprehensive micro- and macrolinguistic analysis. On the clinical
level, it is not clear yet how the severity of the trauma affects
language production skills, as the majority of studies have focused
on severe forms of TBI. For example, several studies have shown that
persons with severe TBI have major difficulties in narrative produc-
tion tasks and that in these individuals, story grammar measures
correlate with performance on tests assessing executive functions
(Coelho et al., 1995; Coelho, 2002; Mozeiko et al., 2011; Marini et al.,
2011a; Le et al., 2012). Though interesting, it is unclear whether these
individuals with moderate and mild TBI. To the best of our knowl-
edge, only a few studies have explicitly analyzed this important issue.

http://dx.doi.org/10.1016/j.neuropsychologia.2014.09.042
0028-3932/& 2014 Elsevier Ltd. All rights reserved.
 A. Marini et al. / Neuropsychologia 64 (2014) 282–288 283

In one such study, Youse and Coelho (2005) reported on the cognitive
and linguistic profile of 55 participants with moderate and severe TBI
and in both groups, macrolinguistic difficulties on two narrative
production tasks (story retelling and story generation) were observed.
In a previous study by Biddle et al. (1996), a group of moderate to
mild TBI participants produced personal narratives that were less
informative than those produced by a group of healthy individuals.
The group of patients with moderate TBI could not adequately
monitor their narratives and failed to include critical information in
their stories. Overall, the few studies that explicitly explored the
macrolinguistic skills of persons with moderate TBI point to difficul-
ties in the conceptual and linguistic organizations of their narrative
discourse. However, neither of these investigations included accurate
micro- and macrolinguistic analyses. Accumulating evidence suggests
that multi-level procedures for linguistic analysis provide a unique
way to adequately capture the linguistic profile of persons with
communication impairments by identifying deficits that are often
not detectable by standardized testing (Galski et al., 1998; Snow et al.,
1998, 1999; Sherratt, 2007; Jorgensen and Togher, 2009; Marini et al.,
2011b). For example, Marini et al. (2011a) analyzed the cognitive and
linguistic skills in narratives produced by a group of non-aphasic
individuals with severe TBI. Though these individuals showed ade-
quate lexical and grammatical skills, their narrative descriptions were
characterized by several violations of cohesion and coherence due to
frequent interruptions of ongoing utterances, derailments, and intro-
duction of extraneous utterances. As a result, their narratives were
poorly organized, vague and ambiguous. Interestingly, a principal
component analysis revealed that a single factor accounted for the
violations of Global Coherence, reduced propositional density, and
reduced proportion of words that conveyed information. Similarly,
Carlomagno et al. (2011) demonstrated that the production of errors
in cohesion and coherence in narratives corresponded to reduced
informativeness and efficiency. More recently, Galetto and colleagues
(2013) analyzed the narrative abilities of a group of 14 mild TBI non-
aphasic participants showing that even persons with mild TBI
produce more errors of Global Coherence than normal. This finding
lends further support to the hypothesis that macrolinguistic difficul-
ties may arise as a consequence of the broad cerebral lesions that
characterize traumatic injuries, regardless of the severity of the
damage.
Consequently, the current study was designed to analyze the
narrative and linguistic skills of a group of persons with moder-
ate TBI with no sign of aphasic symptoms by adopting a multi-
level procedure for discourse analysis. We hypothesized that this
narrative analysis would reveal the linguistic impairments not
captured by traditional aphasia testing. Furthermore, since the
multilevel procedure for discourse analysis allows clinicians to
explore the complex interactions between the different levels of
linguistic processing (e.g., between verb processing, argument
structure generation, sentence production, and inter-utterance
integration), we expected to find significant correlations between
measures assessing these different levels. Finally, we hypothe-
sized that these patients would experience difficulty with tasks
assessing executive function and attention and that performance
on these tasks correlate with the ability to produce coherent and
informative narrative samples.
2. Methods
2.1. Participants
Thirty Italian-speaking participants formed two age matched groups (Table 1).
The experimental group consisted of 10 patients suffering from moderate traumatic
brain injury (modTBI). Criteria for the selection of TBI participants included
symptom stability, Glasgow Coma Scale score ranging from 9 to 13, and normal
to near-normal performance on the Aachener Aphasie Test (AAT, Italian version,
Luzzatti et al., 1991) (see Tables 1 and 3). Although all participants were living at
home at the time of the study, none were completely independent and all had
caregivers. Three of the participants were employed, four were unemployed, and
three were university students at the time of their participation. After hospital
discharge, participants underwent extensive neuropsychological rehabilitation
aimed at recovering those cognitive functions most affected by brain injury (i.e.,
attention, memory, executive functions, and communicative skills). The treatment,
lasting for 6 months, consisted of individual and group sessions. The aim of each
individual session was treatment of specific cognitive deficits, through a series of
focused exercises. In the group sessions, the focus was mainly on social and
communicative skills. All participants also received language therapy in their post-
acute phase.
The control group consisted of 20 neurologically Healthy Controls (see Table 1).
Inclusion criteria for admission in the control group was normal performance on
Raven's progressive matrices (Raven, 1938) and on a selection of neuropsycholo-
gical tests assessing the functionality of different areas of cognition linked to
language processing (see Section 2.2.1 and Table 2). None of them had a known
history of psychiatric or neurological illness, learning disabilities, or hearing or
visual loss.
Two independent samples t-tests were conducted to compare the age and the
level of formal education in the two groups. There was no significant difference in
age (t (28) ¼  .246; p¼ .530) but the two groups differed in the level of formal
education (t (28) ¼ 3.787; p o.001). For this reason, this variable was included as
covariate in the analyses (see later).
All participants gave their written informed consent to participate in the study
after all procedures had been fully explained. Approval for the study was obtained
from the local ethics committee.
2.2. Procedures
2.2.1. Neuropsychological assessment
The cognitive profile of the participants was obtained by administering tests aimed
at assessing those abilities (attentional and learning skills, long-term memory, and
executive functions) which may affect narrative performance (see Table 2): This
included tests of phonological and semantic verbal fluency, verbal memory (Rey's 15-
word Immediate Recall and Delayed Recall), executive functioning (Wisconsin Card
Sorting Test, WCST perseverative and non-perseverative errors; Heaton et al., 1993),
and the Trail Making Test (Parts A and B; Reitan, 1992). The linguistic skills of the TBI
group were investigated by administering the AAT (see Table 3).
Table 2
Means (and standard deviations) of the neuropsychological performance of the
groups of Traumatic Brain Injured and Healthy Control participants.
Traumatic Brain Healthy p- Eta-
Injured Controls value squared

TMT-A (seconds) 39.2 (14.2) 54.6 (12.6) .099 .098

Table 1 TMT-B (seconds) 123.3 (89.4) 115.4 (32.3) .105 .094
Means (and standard deviations) of demographic and clinical characteristics of the Phonemic fluency 27.1 (11.2) 35.4 (8.3) .199 .060
groups of Traumatic Brain Injured and Healthy Control participants. Semantic fluency 17.2 (5.7) 24.1 (4.8) .047 .138
WCST (pers err)n 6.4 (6.5) .5 (.2) .001 .402
Traumatic Brain Injured Healthy Controls WCST (non-pers err)n 18.1 (10.3) .2 (.4) .001 .608
Rey's 15-word 43.4 (12.9) 45.6 (6.6) .649 .008
Mean (SD) (Range) Mean (SD) (Range) immediate recall
Rey's 15-word delayed 8.6 (4) 10 (1.8) .211 .057
Age 36.6 (10.8) (22–55) 35.7 (9.5) (21–55) recall
Formal education (years)n 11.0 (2.6) (8–13) 14.7 (2.9) (8–18)
Time after injury (months) 22.4 (7.2) (12–32) — TMT-A: Trail Making Test, Part A; TMT-B: Trail Making Test, Part B; WCST (pers err):
GCS (maximum score: 15) 10.8 (.8) (9–13) — Wisconsin Card Sorting Test, perseverative errors; WCST (non-pers err): Wisconsin
Card Sorting Test, non-perseverative errors.
GCS: Glasgow Coma Scale. n
When the group-related performance was significantly different after Bon-
n
When the group-related difference is significant. ferroni correction (po .006).
284 A. Marini et al. / Neuropsychologia 64 (2014) 282–288
Table 3
Performance of the participants with TBI on the AAT. The analysis of the
spontaneous speech allows to derive a qualitative score ranging from 0 (severe
impairment) to 5 (absence of difficulties). For the other tests of the AAT (i.e., Token
Test, Repetition, Written Language, Naming and Comprehension) from the raw
scores it is possible to derive T-scores. T-scores above 62 suggest minimal or absent
linguistic impairment.
T-Scores Range
Mean (SD)/maximum (SN_scores)
score
Spontaneous speech—communicative 4.1 (.6)/5 3–5
behaviour
Spontaneous speech—articulation 4.5 (.8)/5 4–5
Spontaneous speech—automatic speech 4.5 (.7)/5 5–5
Spontaneous speech—phonology 4.1 (.6)/5 3–5
Spontaneous speech—syntax 4.1 (.6)/5 3–5
Token test 68.2 (3.4)/77 63–73
Repetition 66.7 (3.1)/77 63–71
Written language 67.9 (3.4)/77 62–73
Naming 66.5 (3.3)/77 62–72
Comprehension 68.7 (2.9)/77 63–72
2.2.2. Assessment of narrative abilities
The narrative assessment was performed on the discourses elicited by a
picture-story description task. The stimuli consisted of one single picture (the
“Picnic” scene from the Western Aphasia Battery; Kertesz, 1982) and two cartoon
stories with six pictures each presented on the same page (the stories of the
“Flower Pot” by Huber and Gleber (1982) and of the “Quarrel” by Nicholas and
Brookshire (1993)). The Picnic scene is a complex picture that elicits a sample of
descriptive speech. The two vignettes elicit narrative speech samples. Each
participant described all three stimuli. The order of presentation was counter-
balanced across subjects. Since no normative data are currently available for these
picture descriptions, the narratives produced by the group of participants with TBI
were compared to those producted by the group of Healthy Controls. In order to
avoid poor performance due to short-term memory limitations, the pictures
remained visible until the subject had finished his/her description. Each story-
telling was tape-recorded and subsequently transcribed verbatim by two of the
authors (AM, VG); the transcriptions included phonological fillers, pauses, false
starts, and extraneous utterances. The scoring procedure was performed indepen-
dently by two raters and then compared. Acceptable interrater reliability was
defined as k Z .80, and the two raters, after appropriate training, achieved this level
between them prior to the beginning of the study. The remaining differences were
resolved through discussion.
The stories were subjected to a quantitative, in-depth textual analysis involving
both micro- and macrolinguistic aspects of narrative production. The analysis
focused on four main aspects of linguistic processing: productivity, lexical and
syntactic organization, textual organization, and informative content (Marini et al.,
2011b). Both simple values and ratios were considered to account for the difference
across narratives.
2.2.2.1. Microlinguistic analysis. Productivity was measured in terms of units,
Words, Speech Rate and Mean Length of Utterance (MLU). The unit count inclu-
ded each word, non-word or false start uttered by the speaker. The total number
of phonologically well-formed words (which excluded false starts, phonological
paraphasias and neologisms) was then computed. The number of words was used
to obtain a measure of Speech Rate in terms of words per minute. For each story
description, the total number of utterances was assessed following the criteria
detailed in Marini et al. (2011b); a sequence of words was considered an utterance
when, in the absence of long pauses, it formed a grammatically complete sente-
nce. A ratio measuring the MLU was calculated by dividing the total number of
words by the number of utterances.
Lexical processing was assessed in terms of phonological, semantic and para-
grammatic errors. A percentage of Phonological Errors was obtained by dividing
Phonological Errors (i.e., neologisms, phonological paraphasias and false starts) by
units and multiplying this value by 100. The percentage of Semantic Paraphasias
was calculated by dividing the number of Semantic Paraphasias by the number of
content words and multiplying this value by 100 (Marini et al., 2011b). A semantic
paraphasia occurred when a semantically related word is substituted for a target
word (i.e. substituting the word “mother” for “wife”). The percentage of paragram-
matic errors was calculated by dividing the paragrammatic errors produced by each
participant by the number of words and multiplying this value by 100. Paragram-
matic errors included substitutions of bound morphemes (for example, questo è una
coppia “this [masc] is a couple [fem]”—it should be questa) and function words (for
example, batte da una porta “he is knocking from a door”—da instead of a).
A percentage of Complete Sentences was used to derive a measure of grammatical
organization. This percentage was calculated by dividing the number of gramma-
tical sentences by the number of utterances and multiplying this value by 100. A
sentence was considered grammatically complete if all the arguments required by
the verb were inserted correctly and no omissions or substitutions of free or bound
morphemes could be detected.
2.2.2.2. Macrolinguistic analysis. As for the macrolinguistic organization of dis-
course, the measures included percentages of Local and Global Coherence Errors.
In this study, Local and Global Coherence are not measured in terms of occurrence
of adequate links among the utterances. Rather, they are measured in terms of
countable errors. The percentage of Local Coherence Errors allowed us to quantify
the amount of utterances that contained errors such as topic shifts or unclear
referents. This corresponds to the notion of intersentential cohesion. A topic shift
occurred when an utterance is abruptly interrupted and the following one intro-
duces new information instead of completing the thought. For instance, in the
sequence “he's trying to…/and here he is on the ground” the first utterance remains
unfinished while the second one introduces a new piece of information. Missing
referents were instances in which the referent of a pronoun or the implicit subject
of a verb were not clear or were incorrect. For example, consider the following
sequence of utterances: “Qui stanno litigando furiosamente. Poi dice: (…)” (“Here
they are quarrelling furiously. Then [implicit pronoun] says: (…)”). In the second
utterance there was a missing referent since it was not clear to whom the verb
“dice” referred. The errors of Local Coherence are presented as percentages with
respect to the total number of utterances produced by each speaker for each story
([Local Coherence Errors divided by utterances]n100). It should be noted, however,
that in some cases the production of these errors might not be directly dependent
on the number of utterances, as a person might produce more than one error per
utterance. An inspection of the occurrences of these errors revealed that the maj-
ority of Local Coherence Errors were missing referents (the production of topic
shifts was minimal) and never occurred twice in the same utterance. Errors of
Global Coherence included the production of utterances that were tangential, co-
nceptually incongruent with the story, repetitions or simple fillers. An utterance
was considered tangential when it derailed the flow of discourse with respect to
the information previously provided in the preceding utterances. An example of
this can be seen in the following sequence produced during the description of the
Picnic Picture: /It is a picnic/I like picnics/I have made several picnics in my life/.
Here, the second and third utterances were scored as tangential, as they provided
information that was irrelevant for the task and was simply triggered by a specific
piece of information depicted in the stimulus. An utterance was considered conc-
eptually incongruent when it included pieces of information not present in the
stimulus. For example, in the description of the Quarrel story a patient produced
this sequence: “The man and the woman are quarrelling/He goes in the kitchen/.
Since the man is not going anywhere but is on a sofa, this utterance is conceptually
incongruent with the story. A propositional repetition is a sequence where the s-
peaker repeats concepts that (s)he has already communicated. Therefore, these
utterances do not add any new information to the story. An example occurs in the
following sequence: /The man is walking on the sidewalk/He is on the sidewalk/. In
this case, the second utterance repeats what has already been stated in the prec-
eding one. Finally, a filler utterance is scored whenever the speaker produces an
utterance that does not provide semantic information that furthers the story. An
example occurs in the following sequence: /The man and the woman are eating /
My God, and now? /Ah, yes, I get it/. The last two utterances are considered as filler
utterances as they simply reflect a moment in narration when the speaker is just
thinking about the story and what to say next. These utterances do not provide any
new information. The Percentage of Global Coherence Errors were calculated by
dividing the number of Global Coherence Errors by the number of utterances and
multiplying this value by 100.
2.2.2.3. Analysis of informative content. The informative content of each narrative
was also analyzed in order to obtain a numerical evaluation of lexical-semantic
appropriateness, Thematic Informativeness, and Global Connectedness. Lexical-
semantic appropriateness was determined by counting the total number of Lexical
Information Units (LIUs), which were those content and function words that were
not only phonologically well-formed but also appropriate from a grammatical and
pragmatic point of view (Marini and Urgesi, 2012). Therefore, all those words that
had been classified as Semantic Paraphasias, fillers, paragrammatic errors or were
inserted in tangential utterances were excluded from the count of Lexical Informa-
tion Units. The percentage of Lexical Informativeness was obtained by dividing the
number of LIUs by the number of words and multiplying this value by 100.
Thematic Informativeness was determined by counting the total number of
thematic units. A thematic unit was defined as a single element of content present
in the eliciting stimulus (see Appendix 1 for a list of the thematic units in each of
three stories). The total number of possible thematic units for the three stories had
been previously identified by presenting the pictures to 200 students from the
School of Arts of the Second University of Rome “Tor Vergata” (Marini et al., 2005).
Those thematic units that had been identified by at least 80% of the subjects have
been included in the list. Each element mentioned by the participant received
 A. Marini et al. / Neuropsychologia 64 (2014) 282–288 285

1 point. Those thematic units that were not mentioned received a score of 0. The Table 4
percentage of Thematic Informativeness for each story was measured by dividing Results of the microlinguistic analysis for the groups of TBI and Healthy Control
the number of thematic units produced by each participant by the total amount of participants.
thematic units available in each picture and multiplying this value by 100. This
Percentage of Thematic Informativeness can be considered as an index of the Microlinguistic Traumatic Brain Healthy p- Eta-
amount of conceptual/informational content conveyed by the participant for analysis Injured Controls value squared
each story.
A measure of Global Connectedness (see Ash et al. (2006)) was used to measure Words 100.7 (46.5) 98.6 (44.8) .308 .039
the ability of each participant to recognize the main idea of each of the two Speech Rate 102.2 (26.8) 143.6 (25.8) .014 .205
cartoon-stories and to understand that the characters portrayed in the picture MLU 6.4 (1.1) 7.3 (1.3) .676 .007
stimuli were the same across different scenes. For example, in the “Quarrel” story, % Phonological .2 (.4) .3 (.5) .489 .018
the participant should mention that the wife came back home because of a car Errors
accident. In the Flower Pot story, the participant should mention that the man is no % Semantic 2.1 (1.9) .3 (.4) .008 .233
longer angry because the woman who threw the Flowerpot was nice. The narrative Paraphasias
speech samples received 1 point for Global Connectedness if they had understood % Paragrammatic .2 (.4) .1 (.1) .014 .203
the point of the story. They received 0 points if they failed to understand it. We Errors
decided to exclude from this analysis the discourses obtained from the Picnic % Syntactic 62.6 (18.8) 70.1 (11.8) .135 .081
Picture, since this picture elicits samples of descriptive speech and not of narrative Completeness
discourse.
No group-related difference was observable after Bonferroni correction (p o.007).
MLU: Mean Length of Utterance.

3. Results

3.1. Neuropsychological assessment

Table 5
Results of the analysis of the macrolinguistic and informative aspects of narrative
None of the Healthy Control participants fell below the accepted production for the groups of TBI and Healthy Control participants.
cut-off on any of the neuropsychological measures (see Table 2). In
order to investigate significant group-related differences between the Macrolinguistic and Traumatic Brain Healthy p- Eta-
group of Healthy Control participants and that of persons with Informative analysis Injured Controls value squared

moderate TBI, the scores obtained by each participant on the % Local Coherence Errorsn 14.5 (8.0) 3.2 (3.5) .003 .285
neuropsychological assessment were entered into a one-way % Global Coherence Errorsn 19.5 (11.7) 6.9 (6.8) .012 .211
between-groups analysis of covariance (ANCOVA) with Group (1. % Lexical Informativenessn 69.0 (12.1) 85.5 (7.8) .004 .265
Healthy Controls; 2. participants with Moderate TBI) as the indepen- % Thematic Informativeness 56.2 (18.8) 66.3 (11.6) .745 .004
Global Connectedness .9 (.3) 1.00 (.00) .248 .049
dent variable and the neuropsychological scores as dependent vari-
ables (Trail Making-A, Trail Making-B, Semantic Fluency, Phonological n
Asterisks indicate when the group-related performance was significantly
Fluency, perseverative and non-perseverative errors at the WCST, Rey different after Bonferroni correction (p o .025 for the macrolinguistic analysis;
Immediate and Delayed recall of a word list). The participants' years p o.017 for the analysis of informative content).
of formal education were used as the covariate in this analysis. The
level of statistical significance was defined as po.006 after Bonferroni
correction for multiple comparisons (alpha o.05/8 – alpha o.006).
adjusting for the years of formal education, the two groups produced
After adjusting for the years of formal education, the two groups
descriptions that were comparable with respect to number of Words,
differed only on the production of perseverative [F (1; 27)¼18.151;
Speech Rates and MLU. Furthermore, no significant group-related
po.001; η2p ¼ .402] and non-perseverative errors [F (1; 27)¼ 41.936;
differences were noted in the production of phonological or mor-
po.001; η2p ¼.608] at the WCST.
phological errors. For the TBI group, the production of semantic
On all the subtests of the AAT assessing general linguistic
errors approached significance (% Semantic Paraphasias: [F (1; 27)¼
functions, the group of participants with moderate TBI had a
8.193; po.008; η2p ¼ .233]).
performance well above the cut-off for normality (see Table 3).
This confirmed that they were not aphasic.
3.2.2. Macrolinguistic analysis
3.2. Assessment of narrative abilities The results from the macrolinguistic analysis are reported in
Table 5. The scores obtained from the macrolinguistic analysis
The results from the narrative analysis will be presented in two were entered into a one-way between-groups ANCOVA as depen-
separate sections: microlinguistic analysis (including measures of dent variables (% Local Coherence Errors and % Global Coherence
productivity, and both lexical and grammatical processing) and Errors) with Group (1. Healthy Control participants; 2. ModTBI
macrolinguistic analysis (including the analysis of structural and participants) as fixed factor. The participants' years of formal
functional aspects of discourse production). The narrative scores of education were used as a covariate in this analysis. The level of
each participant with TBI are presented in Appendix 2. statistical significance was defined as p o.025 after Bonferroni
correction for multiple comparisons (alpha o.05/2). After adjust-
ing for the years of formal education, the group of persons with
3.2.1. Microlinguistic analysis
moderate TBI produced more Errors of Local [F (1; 27) ¼ 10.771;
The results from the microlinguistic analysis are reported in
po .003; η2p ¼.285] and Global Coherence [F (1; 27) ¼7.224;
Table 4. The scores obtained from the microlinguistic analysis entered
po .012; η2p ¼.211].
into a one-way between-groups ANCOVA as dependent variables
(Words, Speech Rate, Mean Length of Utterance, % Phonological
Errors, % Semantic Paraphasias, % Paragrammatic Errors and % 3.2.3. Analysis of informative content
Syntactic Completeness) with Group (1. Healthy Control participants; The results from the analysis of the informative content of the
2. ModTBI participants) as the independent variable. The participants' narratives produced by the participants are also reported in
years of formal education were used as a covariate in this analysis. Table 5. The scores obtained with this analysis were entered into
The level of statistical significance was defined as po.007 after a one-way between-groups ANCOVA as dependent variables (%
Bonferroni correction for multiple comparisons (alpha o.05/7). After Lexical Informativeness, % Thematic Selection, and Global
286 A. Marini et al. / Neuropsychologia 64 (2014) 282–288
Connectedness) with Group (1. Healthy Controls; 2. Participants
with Moderate TBI) as fixed factor. The participants' years of
formal education were used as a covariate in this analysis. The
level of statistical significance was defined as p o.017 after
Bonferroni correction for multiple comparisons (alpha o.05/3).
After adjusting for the years of formal education, the group of
persons with moderate TBI produced fewer appropriate words (%
Lexical Informativeness: [F (1; 27) ¼9.722; p o.004; η2p ¼ .265]) but
adequate levels of % Thematic Selection and Global Connectedness.
3.3. Correlations between neuropsychological and narrative scores
The relationship among the macrolinguistic scores that were
found to be impaired in the group of participants with moderate
TBI was explored using the Pearson product–moment correlation
coefficient (see Table 6). There were strong negative correlations
between the percentage of Lexical Informativeness and the per-
centages of both Local and Global Coherence Errors. Furthermore,
in order to explore the possibility that the reduced levels of
informativeness and discourse organization found in the group
of participants with moderate TBI were correlated with the low
performance on the WCST, new correlations were performed
between the number of non-perseverative and perseverative
errors on the WCST and the narrative measures found impaired
in the group of participants with moderate TBI (i.e., % Local
Coherence Errors, % Global Coherence Errors, and % Lexical
Informativeness). The analyses showed that the % of Local Coher-
ence Errors was positively correlated with the production of both
perseverative and non-perseverative errors. The % of Global
Coherence Errors was positively correlated only with the produc-
tion of non-perseverative errors. Finally, the % of Lexical Informa-
tiveness was negatively correlated with the production of both
perseverative and non-perseverative errors.
4. Discussion
This study focused on the analysis of cognitive, linguistic and
narrative skills in a group of persons with moderate TBI without
aphasic symptoms. Their performance was compared to that of a
group of healthy participants. The linguistic and narrative assess-
ments included procedures for the analysis of micro-, macrolin-
guistic and informative aspects of discourse production. Overall,
the participants with moderate TBI had difficulty with Local
Coherence, Global Coherence, and Lexical Informativeness but
were not impaired in reporting the content of the stories and
picture. Their persisting impairments of executive functions cor-
related with the percentages of local and Global Coherence errors
and the reduced production of Lexical Informativeness.
The participants with moderate TBI were all in the chronic
phase with a minimum of 12 months post-onset and had under-
gone extensive rehabilitation. Nonetheless, they had still difficul-
ties in selecting appropriate strategies to cope with the Wisconsin
Table 6
Results from the univariate correlation analysis between linguistic (% Local Coherence Errors, % Global Coherence Errors and % Lexical Informativeness) and
neuropsychological scores in the TBI group.
% Global Coherence Errors
% Local Coherence Errors .560 (p o.001)
% Global Coherence Errors –  .924 (p o.001)
% Lexical Informativeness – –
WCST (pers err) – –
Scores are presented as Pearson' r-indexes and (p-values). Significant correlations are highlighted in bold.
Card Sorting Test where they produced a number of perseverative
and non-perseverative errors. This confirms the presence of
persisting impairments of executive functioning. This is not a
new finding. Indeed, executive functions are usually impaired in
persons with TBI of varying levels of seriousness (Body et al., 1999;
Stierwalt and Murray, 2002; Body and Perkins, 2004; Galetto et al.,
2013) and might affect cognitive processes such as those required
in the speech elicitation task employed in this experiment (e.g.,
Snow et al., 1998; Snow and Douglas, 2000; Youse and Coelho,
2005; Coelho et al., 2013). We will return to this issue later in the
discussion.
In terms of narrative performance, the persons with mild TBI
produced discourses with a good amount of words, adequate Speech
Rate and normal MLU. Their lexical selection skills were mildly
weakened, which is demonstrated by their production of slightly
more Semantic Paraphasias than the control participants (2% of the
total amount of uttered words). However, they could still appro-
priately retrieve the phonological and morphological information
stored in their mental lexicon (i.e. they did not produce more
phonological or paragrammatic errors than the control participants).
Overall, the microlinguistic analysis confirmed the results from the
AAT. They did not have aphasic symptoms that might have affected
their narrative performance. Even the slightly reduced word retrieval
efficiency might have been a side effect of other difficulties; the TBI
might have altered resource allocation and content organization
abilities (see also Gillis (1996), Murdoch and Theodoros (2001),
Adamovich (2005) and Coelho et al. (2013)) as well as the ability to
plan future propositions and inhibit irrelevant ones (Levin et al.,
1981; Ylvisaker et al., 2001; Barrow et al., 2003, 2006; King et al.,
2006a, 2006b; Hough, 2008). Cartoon-story description tasks such as
those used in the current study might have increased the demands of
message level processing thus complicating the process of lexical
selection in these patients (Linebarger et al., 2004).
The analysis of macrolingusitic and informative aspects of dis-
course production allowed exploration of the potential link between
executive functioning and narrative production skills. The partici-
pants with moderate TBI experienced significant difficulties in
keeping their discourse coherent even when they had the ability to
both mention a sufficient number of story events and to recognize
the main idea of the story. They produced several errors of both Local
(mostly words with missing referents) and Global Coherence. Since
the errors of Local and Global Coherence were strongly correlated
with the % of Lexical Informativeness, it is possible that the reduced
informative ability of these patients was linked to difficulty in taking
into account the listener's perspective and refraining from introdu-
cing irrelevant pieces of information in their narratives. In other
words, this reduced communicative effectiveness may stem from a
failure to monitor the process of narrative production. This high-level
linguistic ability is mediated by executive functions (Sim et al., 2013;
see also Carlomagno et al. (2011) for similar findings on a group of
persons with severe TBI). According to Miyake and colleagues (2000)
and Mozeiko et al. (2011), at least three specific executive compo-
nents might be involved in the process of message planning,
% Lexical Informativeness WCST (pers err) WCST (non-pers err)
 .577 (po .001) .390 (p o .033) .737 (po .001)
.305 (p ¼.102) .584 (po .001)
 .366 (p o .047)  .594 (po .001)
– .821 (po .001)
 A. Marini et al. / Neuropsychologia 64 (2014) 282–288
monitoring and execution: shifting (or attention switching), involved
in the generation of complete episodes within a narrative discourse
and in the selection of informative words; updating, required to recall
former episodes or episodic contents for accurate organization of the
story; and inhibition, important for monitoring the production of
extraneous comments and derailments while telling a story. In the
current study, the production of non-perseverative errors correlated
positively with the production of errors of both local and global
coherence and negatively with the production of lexical information
units. Since performance on the WCST can be considered as a
measure of the ability to shift from one concept to another to
generate complete episodes, the correlations found in the current
study suggest that at least some of the errors of Local and Global
Coherence produced by the group of participants with TBI were
linked to difficulty in shifting between the concepts that they had
correctly identified. However, the production of perseverative errors
also correlated negatively with the production of lexical information
units and positively with the production of errors of Local Coherence.
This suggests that other executive functions might be compromised
in these participants. These compromised executive functions may, in
turn, affect the participants' ability to inhibit the insertion of
comments and words whose referent had not been previously
introduced. Overall, these data support the possibility that the
reduced narrative skills observed in persons with moderate TBI are
an epiphenomenon of a disturbance affecting their executive skills.
This relation between macrolinguistic difficulties and executive
functions parallels what has been found in studies on neurodegen-
erative diseases. For example, Ash et al. (2006) found that a subgroup
of non-aphasic patients with frontotemporal dementia with disor-
ders of executive functioning exhibited difficulties of narrative
organization that were related to their executive deficits. However,
unlike the participants with TBI, these patients could not even
generate the main idea of the story. The participants with TBI
showed a preserved ability to extract all necessary information about
the events of the story from the picture stimuli but were unable to
appropriately organize this information at the macrolinguistic level.
This organizational difficulty likely triggered the production of errors
of Local and Global Coherence that, in turn, reduced the levels of
Lexical Informativeness. This interpretation is consistent with pre-
vious studies where a connection between executive impairments
and narrative disturbances in TBI speakers has been hypothesized
(e.g., Hincliffe et al., 1998; Tucker and Hanlon 1998; Ylvisaker et al.,
2001; Coelho, 2002; Galetto et al., 2013). For example, in their “High
Level Language Hypothesis”, Hincliffe and colleagues (1998) sug-
gested that the communicative deficits observed in persons with TBI
might be considered as the outcome of more general cognitive
difficulties rather than a direct effect of a primary linguistic deficit.
Namely, impairments of executive functions might account for
macrolinguistic deficits such as the inability to organize, plan and
produce a discourse and to appropriately link the concepts that make
up a coherent story. Coelho (2002) reported significant correlations
between scores on the WCST and measures assessing the ability to
produce a complete discourse; better performance on the WCST was
associated with the production of more episodes in story retelling
tasks and therefore to the ability to organize a story. Similarly, Galetto
et al. (2013) showed a statistically significant negative correlation
between the production of non-perseverative errors on the WCST
and the percentage of lexical information units produced in the
narrative sample in a group of persons with mild TBI. Therefore, the
accurate multilevel analysis used in the present study allowed us to
have a comprehensive picture of the narrative skills of persons with
moderate TBI and to extend these considerations well beyond the
simple analysis of the ability to generate a well-formed set of
episodes.
In conclusion, even if based on a small sample of participants with
moderate TBI, the results from the current study have important
287
clinical and theoretical implications for the understanding of com-
municative disorders observed in individuals with mild TBI. From a
clinical point of view, this study confirms the usefulness of a multi-
level procedure for discourse analysis. This methodology allows both
clinicians and scholars to have a comprehensive picture of the
linguistic skills of persons with brain injuries. In story description
tasks, speakers tend to be more fluid communicators and to make use
of several linguistic skills. Consequently, procedures of discourse
analysis allow scholars and clinicians to explore the interaction
among several processing levels (e.g., between processes of lexical
selection and access, grammatical unification and inter-utterance
integration). Furthermore, this analysis incorporates measures not
included in traditional linguistic assessment tasks, such as indexes of
informativeness and of the individual's ability to connect the different
utterances via coherent connections. It is noteworthy that the
qualitative assessment of the spontaneous speech produced by the
TBI participants on the AAT did not register any particular difficulty in
these individuals (see Table 3), whereas a comprehensive multi-level
methodology of discourse analysis such as the one employed in this
study showed the presence of significant difficulties in this group.
This study also raises some interesting theoretical issues about the
cognitive processes underlying the phases of message planning and
monitoring. Namely, it suggests the possibility of significant involve-
ment of executive skills in these preliminary processes of message
generation. More specifically, our data suggest that difficulty in
shifting between concepts (Miyake et al., 2000) and in inhibiting
irrelevant responses (either perseverative or non-perseverative
errors) played a major role in the macrolinguistic impairment
observed in the participants with TBI. Finally, these results provide
a clearer understanding of the nature of the communicative impair-
ment in persons with TBI in general and moderate TBI in particular.
Confirming previous reports, the group of non-aphasic participants
with TBI had normal lexical and grammatical skills and produced a
normal amount of information content in their narratives (Stout et al.,
2000; Carlomagno et al., 2011). However, this information was not
correctly organized at the macrolinguistic level as evidenced by the
fact that they produced fewer lexical information units and more local
and Global Coherence errors than the group of healthy participants.
Unfortunately, this study included only a limited number of persons
with moderate TBI. This suggests that these data need to be replicated
on larger cohorts of individuals with moderate TBI in order to draw
firm conclusions on the cognitive processes underlying the narrative
difficulties observed in these patients. Hence, further research is
required to explore the impact of these factors on the communicative
impairment in persons with moderate traumatic brain injury.
Role of funding source
Funding for this study was provided by IRCCS “E. Medea”. The
IRCCS “E. Medea” had no further role in study design; in the
collection, analysis and interpretation of data; in the writing of the
report; and in the decision to submit the paper for publication.
Aknowledgements
We would like to thank Amy Henderson and Stephen Kintz for
the linguistic revision of the article.
Appendix A and Appendix B. Supporting information
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.neuropsychologia.
2014.09.042.
288 A. Marini et al. / Neuropsychologia 64 (2014) 282–288
References
Adamovich, B.L.B., 2005. Traumatic brain injury. In: LaPointe, L.L. (Ed.), Aphasia and
Related Neurogenic Communication Disorders, 3rd ed. Thieme, New York.
Angeleri, R., Bosco, F.M., Zettin, M., Sacco, K., Colle, L., Bara, B.G., 2008. Commu-
nicative impairment in traumatic brain injury: a complete pragmatic assess-
ment. Brain Lang. 107, 229–245.
Ash, S., Moore, P., Antani, B.S., McCawley, B.A., Work, B.S., Grossman, M., 2006.
Trying to tell a tale. Discourse impairments in progressive aphasia and
frontotemporal dementia. Neurology 66, 1405–1413.
Barrow, I.M., Hough, M., Rastatter, M.P., Walker, M., Holbert, D., Rotondo, M.F., 2003.
Can within-category naming identify subtle cognitive deficits in the mild
traumatic brain.injured patient? J. Trauma Inj. Infect. Crit. Care 54, 888–897.
Barrow, I.M., Hough, M., Rastatter, M.P., Walker, M., Holbert, D., Rotondo, M., 2006.
The effects of mild traumatic brain injury on confrontation naming in adults.
Brain Inj. 20, 845–855.
Biddle, K.R., McCabe, A., Bliss, L.S., 1996. Narrative skills following traumatic brain
injury in children and adults. J. Commun. Disord. 29, 447–469.
Body, R., Perkins, M.R., 2004. Validation of linguistic analyses in narrative discourse
after traumatic brain injury. Brain Inj. 18, 707–724.
Body, R., Perkins, M.R., McDonald, S., 1999. Pragmatics, cognition and communica-
tion in traumatic brain injury. In: MCDonald, S., Togher, L., Code, C. (Eds.),
Communication Disorders Following Traumatic Brain Injury. Psychology Press,
Hove.
Bond, F., Godfrey, H.P.D., 1997. Conversation with traumatically brain injured
individuals: a controlled study of behavioural changes and their impact. Brain
Inj. 11, 319–330.
Brookshire, B.L., Chapman, S.B., Song, J., Levin, H.S., 2000. Cognitive and linguistic
correlates of children's discourse after closed head injury: a three-year follow
up. J. Int. Neuropsychol. Soc. 6, 741–751.
Carlomagno, S., Giannotti, S., Vorano, L., Marini, A., 2011. Discourse information
content in non-aphasic adults with brain injury: a pilot study. Brain Inj. 25,
1010–1018.
Coelho, C., Liles, B.Z., Duffy, R.J., 1995. Impairments of discourse abilities and
executive functions in traumatically brain-injured adults. Brain Inj. 9, 471–477.
Coelho, C.A., 2002. Story narratives of adults with closed head injury and non-brain
injured adults. J. Speech, Lang. Hear. Res. 45, 1232–1248.
Coelho, C., Le, K., Mozeiko, J., Krueger, F., Grafman, J., 2012. Discourse production
following injury to the dorsolateral prefrontal cortex. Neuropsychologia 50,
3564–3572.
Coelho, C.A., Le, K., Mozeiko, J., Hamilton, M., Tyler, E., Krueger, F., Grafman, J., 2013.
Characterizing discourse deficits following penetrating head injury: a preli-
minary model. Am. J. Speech-Lang. Pathol. 22, S438–S448.
Davis, G.A., Coelho, C.A., 2004. Referential cohesion and logical coherence of
narration after closed head injury. Brain Lang. 89, 508–523.
Douglas, J., 2004. The evidence base for the treatment of cognitive–communicative
disorders after traumatic brain injury in adults. In: Reilly, S., Douglas, J., Oates, J.
(Eds.), Evidence-Based Practice in Speech Pathology. Whurr, London, pp. 59–82.
Erez, A.B., Rothschild, E., Katz, N., Tuchner, M., Hartman-Maeir, A., 2009. Executive
functioning, awareness, and participation in daily life after mild traumatic brain
injury: a preliminary study. Am. J. Occup. Ther. 63, 634–640.
Galetto, V., Andreetta, S., Zettin, M., Marini, A., 2013. Patterns of impairment of
narrative language in mild traumatic brain injury. J. Neurolinguist. 26, 649–661.
Galski, T., Tompkins, C., Johnston, M.V., 1998. Competence in discourse as a measure
of social integration and quality of life in persons with traumatic brain injury.
Brain Inj. 12, 769–782.
Gillis, R.J., 1996. Traumatic Brain Injury. Butterworth-Heinemann, Singapore.
Hartley, L.L., Jensen, P.J., 1992. Three discourse profiles of closedhead-injury
speakers: theoretical and clinical implications. Brain Inj. 6, 271–282.
Heaton, R.K., Chelune, G.J., Talley, J.L., Kay, G.G., Curtiss, G., 1993. Wisconsin Card
Sorting Test Manual, Revised and Expanded. Psychological Assessment Resour-
cesm, Odessa.
Hincliffe, F.J., Murdoch, B.E., Chenery, H.J., 1998. Toward a conceptualization of
language and cognitive impairment in closed-head injury: use of clinical
measures. Brain Inj. 12, 109–132.
Hough, M.S., 2008. Word retrieval failure episodes after traumatic brain injury.
Aphasiology 22, 644–654.
Huber, W., Gleber, J., 1982. Linguistic and nonlinguistic processing of narratives in
aphasia. Brain Lang. 16, 1–18.
Jorgensen, M., Togher, L., 2009. Narrative after traumatic brain injury: a comparison
of monologic and jointly-produced discourse. Brain Inj. 23, 727–740.
Kertesz, A., 1982. Western Aphasia Battery. Grune and Stratton, New York.
King, K.A., Hough, M.S., Vos, P., Walker, M.M., Givens, G., 2006a. Word retrieval
following mild TBI: implications for categorical deficits. Aphasiology 20,
233–245.
King, K.A., Hough, M.S., Walker, M.M., Rastatter, M., Holbert, D., 2006b. Mild
traumatic brain injury: effects on naming in word retrieval and discourse. Brain
Inj. 20, 725–732.
Le, K., Coelho, C.A., Mozeiko, J., Krueger, F., Grafman, J., 2011. Measuring goodness of
story narratives: implications for traumatic brain injury. Aphasiology 25,
748–760.
Le, K., Coelho, C., Mozeiko, J., Krueger, F., Grafman, J., 2012. Predicting story
goodness performance from cognitive measures following traumatic brain
injury. Am. J. Speech-Lang. Pathol. 21, S115–S125.
Levin, H.S., Grossman, R.G., Sarwar, M., Meyers, C.A., 1981. Linguistic recovery after
closed head injury. Brain Lang. 12, 360–374.
Linebarger, M.C., McCall, D., Berndt, R.S., 2004. The role of processing support in
remediation of aphasic language production disorders. Cogn. Neuropsychol. 21,
267–282.
Luzzatti, C., Willems, K., DeBleser, R., 1991. Aachener Aphasie Test (Versione
Italiana). Organizzazioni Speciali, Firenze.
Marini, A., Urgesi, C., 2012. Please, get to the point! A cortical correlate of linguistic
informativeness. J. Cogn. Neurosci. 24, 2211–2222.
Marini, A., Carlomagno, S., Caltagirone, C., Nocentini, U., 2005. The role played by
the right hemisphere in the organization of complex textual structures. Brain
Lang. 93, 46–54.
Marini, A., Andreetta, S., Del Tin, S., Carlomagno, S., 2011a. A multi-level
approach to the analysis of narrative language in. Aphasiology 25, 1372–1392.
Marini, A., Galetto, V., Zampieri, E., Vorano, L., Zettin, M., Carlomagno, S., 2011b.
Narrative language in traumatic brain injury. Neuropsychologia 49, 2904–2910.
Mendez, C.V., Hurley, R.A., Lassonde, M., Zang, L., Taber, K.H., 2005. Mild traumatic
brain injury: neuroimaging of sport-related concussions. J. Neuropsychiatry
Clin. Neurosci. 17, 297–303.
Mozeiko, J., Le, K., Coelho, C., Krueger, F., Grafman, J., 2011. The relationship of story
grammar and executive function following TBI. Aphasiology 25, 826–835.
Murdoch, E., Theodoros, D.G., 2001. Physiological assessment of tongue function in
dysarthria following traumatic brain injury. Logop. Phoniatr. Vocology 26,
51–65.
Miyake, A., Friedman, N.P., Emerson, M.J., Witzki, A.H., Howerter, A., 2000. The
unity and diversity of executive functions and their contributions to complex
“Frontal Lobe” tasks: a latent variable analysis. Cogn. Psychol. 41, 49–100.
Nicholas, L.E., Brookshire, R.H., 1993. A system for quantifying the informativeness
and efficiency of the connected speech of adults with aphasia. J. Speech Hear.
Res. 36, 338–350.
Raven, J.C., 1938. Progressive Matrices: A Perceptual Test of Intelligence. H.K. Lewis,
London.
Reitan, R.M., 1992. Trail Making Test. Reitan Neuropsychology Laboratory, Arizona.
Scheid, R., Walther, K., Guthke, T., Preul, C., Von Cramon, D.Y., 2006. Cognitive
sequelae of diffuse axonal injury. Arch. Neurol. 63, 418–424.
Sherratt, S., 2007. Multi-level discourse analysis: a feasible approach. Aphasiology
21, 375–393.
Silver, J.M., McAllister, T.W., Arciniegas, D.B., 2009. Depression and cognitive
complaints following mild traumatic brain injury. Am. J. Psychiatry 166,
653–661.
Sim, P., Power, E., Togher, L., 2013. Describing conversations between individuals
with traumatic brain injury (TBI) and communication partners following
communication partner training: using exchange structure analysis. Brain Inj.
27, 717–742.
Snow, P.C., Douglas, J., Ponsford, J.M., 1998. Conversational discourse abilities
following severe traumatic brain injury: a follow up study. Brain Inj. 12,
911–935.
Snow, P., Douglas, J., Ponsford, J., 1999. Narrative discourse following severe
traumatic brain injury: a follow-up study. Brain Inj. 13, 529–551.
Snow, P.C., Douglas, J.M., 2000. Subject review: conceptual and methological
challenges in discourse assessment with TBI speakers: towards an under-
standing. Brain Inj. 14, 397–415.
Stierwalt, J.A.G., Murray, L.L., 2002. Attention impairment following traumatic brain
injury. Semin. Speech Lang. 23, 129–138.
Stout, C.E., Yorkston, K.M., Pimental, J.I., 2000. Discourse production following mild,
moderate and severe traumatic brain injury: a comparison of two tasks. J. Med.
Speech-Lang. Pathol. 8, 15–25.
Togher, L., Hand, L., Code, C., 1997. Analysing discourse in the traumatic brain injury
population: telephone interactions with different communication partners.
Brain Inj. 11, 169–190.
Tu, L.V., Togher, L., Power, E., 2011. The impact of communication partner and
discourse task on a person with traumatic brain injury: the use of multiple
perspectives. Brain Inj. 25, 560–580.
Tucker, F.M., Hanlon, R.E., 1998. Effects of mild traumatic brain injury on narrative
discourse production. Brain Inj. 12, 783–792.
Watt, A.J., Douglas, J.M., 2006. Interpreting facial expression and communication
competence following severe traumatic brain injury. Aphasiology 20, 707–722.
Ylvisaker, M., Szekeres, S., Feeney, T., 2001. Communication disorders associated
with traumatic brain injury. In: Chapey, R. (Ed.), Language Intervention
Strategies in Aphasia and Related Neurogenic Communicaion Disorders, 4th
ed. Lippincott Williams, & Wilkins, Philadelphia, PA.
Youse, K.M., Coelho, C.A., 2005. Working memory and discourse production abilities
following closed-head injury. Brain Inj. 19, 1001–1009.