Introduction
The study of learner characteristics or
individual differences (IDs) has a long-standing interest in the field of interlanguage
pragmatics (ILP) as factors affecting
pragmatic competence (e.g., Kasper &
Schmidt, 1996; Kasper & Rose, 2002;
Kuriscak, 2010). Among the ID factors
examined, L2 proficiency has accumulated
the most research. Many studies took a
cross-sectional design and compared
pragmatic competence across proficiency
levels (e.g., Kuriscak, 2010; for a review,
see Kasper & Rose, 2002; Kasper & Röver,
2005). However, the literature is rather
limited when other ID factors are concerned.
In Kasper and Rose (2002), only a small
amount of ID research is cited. These studies
examined such factors as age (Kim, 2000),
gender (Rintel, 1984; Kerekes, 1992),
motivation (LoCastro, 2001), and social
identity (Iino, 1996; Siegal, 1996). A decade
after Kasper and Rose's book, a few studies
have updated the list, including: Shimura's
(2003) studies on personality and pragmatic
competence, Takahashi’s (2005) study on
motivation and pragmatic learning,
Taguchi's (2008a, 2008b) studies on the
effects of lexical access skill and working
memory in pragmatic comprehension,
Yates's (2005) study on the effect of gender,
and Davis's (2007) study on the role of
subjectivity in pragmatic choice. However,
very few longitudinal studies have examined
that examined the effects of ID factors on
pragmatic development (e.g., Matsumura,
2003).
This study expanded the existing literature
by investigating the impact of multiple ID
factors on L2 learners' longitudinal change
in pragmatic competence. Japanese learners
of English completed a speaking test
measuring their ability to produce speech
acts (requests and opinions) three times
during one academic year. The study
measured three ID variables (proficiency,
motivation, and lexical access) and
examined their effects on pragmatic change.
Background
Because large individual variations are
typically found among L2 learners in their
success in L2 acquisition, the study of
individual differences (IDs) has been a
paramount area of SLA research that
explains observed individual variations in
L2 learning (DÖrnyei, 2005, 2009; Ellis,
2005). The field of ILP has followed this
tradition and accumulated a body of research
that examined the effect of individual
difference on pragmatics ability. However,
the factors examined to date are largely
concentrated on L2 proficiency. Numerous
cross-sectional studies compared L2
pragmatic performance across different
proficiency levels determined by
standardized exams, grade level, or length of
formal study (e.g., Al-Gahtani & Roever,
2012; Dalmau & Gotor, 2007; Félix-Brasdefer, 2004, 2007; Garcia, 2004a;
Geyer, 2007; Pinto, 2005; RÖver, 2005;
Taguchi, 2007a, 2011; Yamanaka, 2003;
Xu, Case, & Wang, 2009). These studies
have revealed that high proficiency
generally leads to better pragmatic
performance but it does not guarantee a
native-like performance. For example, a
recent study by Al-Gahtani and Roever
(2012) used a role play task to examine
sequential organization in the production of
requests by Arabic learners of L2 English.
Participants formed two proficiency groups
based on their course levels and
performance on proficiency measures. They
completed three role play tasks with a native
speaker interlocutor. Results revealed a
positive effect of proficiency on learners’
sequential organization of requests with
regard to pre-expansions (e.g., greetings and
summons prior to request) and insertion of
expansions (e.g., negation about timing and
details of the request).
Aside from proficiency, other ID factors
have been addressed only sparsely in L2
pragmatics research. Kasper and Rose's
(2002) review cited only a handful of studies
that treated IDs as their central investigative
concern, emphasizing that the review should
be "best read as a strong invitation to
research on individual differences in
learning L2 pragmatics" (p.278). These
studies examined ID factors such as gender
(Kerekes, 1992; Rintel, 1984), age (Kim,
2002), and social identity (Iino, 1996;
Siegal, 1996). A decade after Kasper and
Rose's book, the field of ILP has expanded
the body of studies that situated ID as the
central investigation. Below I will review
recent studies on individual differences in
ILP appeared after Kasper and Rose’s book.
Due to the space constraint, I will limit my
review to quantitative studies, which are
directly relevant to my study.
Takahashi (2005) examined the effect of
motivation on L2 English learners' ability to
notice target request-making expressions in
written dialogues. Japanese students of
English completed a motivation
questionnaire and then completed a
noticing-the-gap task on request forms. The
degree of the learners’ awareness of the
target pragmalinguistic forms was assessed
through a retrospective questionnaire.
Results showed that more motivated learners
noticed more target forms and were more
metapragmatically aware than less
motivated learners.
Although Takahashi’s study is probably the
only existing study that measured
motivation in relation to pragmatics
learning, motivation has been examined
extensively with a broader construct of L2
ability (see DÖrnyei, 2005, for a review).
For example, drawing on Gardner's (1985)
concept of integrative motivation (i.e., desire
to interact with members in L2 community),
Yashima (2002) developed a survey that
measures motivational orientation (long-range goals for learning a language) among
Japanese learners of L2 English. Yashima
(2000) and Yashima et al (2004) found that
integrative and instrumental orientation to
English study predicted motivational
intensity, which in turn led to higher L2
proficiency. Hence, the orientation toward
L2 study seems to be closely related to
motivation, and seizes both integrative and
instrumental orientations toward L2
learning.
On the other hand, very few studies have
examined the effect of personality in
pragmatic competence. Shimura (2003)
examined the relationship between
personality and pragmatic competence in the
speech act of advise-giving. Japanese
students of English completed the task of
writing an advice letter in a formal situation.
Linguistic strategies were analyzed
according to three response categories:
direct, hedged, and indirect advice. There
was a significant effect of personality on the
choice of strategies: introversion types used
more direct expressions than extroversion
types. Although Shimura’s findings are
promising, at the moment there is no
theoretical model that describes how
personality dimensions are related to various
aspects of SLA (DÖrnyei, 2005). Hence, the
use of personality as a variable calls for a
caution.
Other ID factors that recently entered the
ILP research are cognitive factors. Taguchi
(2008a, 2008b) assessed the extent to which
accurate and speedy comprehension of
conversational implicature is associated with
lexical access skill and amount of language
contact. Forty-four college students of
English completed three measures over a
four-month period: (1) the pragmatic
listening test, (2) the lexical access test, and
(3) the language contact survey measuring
the amount of time spent in four language
skills. Results showed that lexical access
skill and the amount of time spent on
speaking and reading significantly correlated
with gains in comprehension speed, but not
with gains in accuracy of pragmatic
comprehension.
These findings revealed that the lexical
access skill could the factor that directly
affects pragmatic comprehension. Pragmatic
comprehension involves the lower-level
processing of attending to and assigning
meaning to linguistic stimuli, as well as
higher-level processing of supplementing
linguistic information with non-linguistic
information to derive meaning. Lexical
access speed is considered to form one of
the lower-level processes that contribute to
pragmatic comprehension. However, the
role of lexical access in production of
pragmatic functions has not been attested.
Yet, the significant relationship between
lexical access and oral fluency found in
previous research (Segalowitz & Freed,
2004) implies that the ability to access word
meaning quickly could influence fluent
production of pragmatic functions.
As described above, previous research
showed that certain ID factors explain
individual variations in L2 pragmatic
competence and development. However,
there are several gaps in the existing
literature. First, previous studies were
almost exclusively confined to a cross-sectional, single-moment design by
examining the relationship between ID
factors and pragmatic competence at given
point of time, and very few studies have
addressed the role of IDs from a
developmental perspective. Although a large
volume of research has examined the
relation between proficiency and
pragmatics, very few studies have employed
longitudinal research design to examine
pragmatic development in relation to
changing L2 proficiency. In order to gain a
more complete understanding of the impact
of learner characteristics in pragmatic
development, more studies should employ a
longitudinal design.
Second, most previous studies
operationalized pragmatic competence as
accurate, appropriate comprehension and
production of pragmatic functions, and
fluency aspect of pragmatic performance
(speedy processing of pragmatic functions)
has been neglected in the analysis of ID
effect. This is a serious neglect, considering
that a growing number of recent studies
have measured both knowledge (i.e.
accuracy and appropriateness) and
processing (i.e. fluency) in pragmatic
performance. These studies revealed that
knowledge and processing dimensions are
distinct from one another: they do not
correlate with each other; they exhibit
different developmental rates; and they are
affected differently by varying learning
contexts, the amount of language contact,
and certain cognitive variables (Taguchi,
2007b, 2008a, 2008b).
The present study aimed to fill these gaps in
the literature and examined the effects of ID
factors in changing pragmatic competence.
Pragmatic change was traced in two aspects:
appropriateness and fluency of speech act
production. Changes in these aspects were
analyzed in relation to three ID factors:
proficiency, orientation toward learning
(integrative and instrumental), and lexical
access skill. These variables were selected to
represent a range of individual (both
cognitive and affective) and contextual
factors that were found to affect pragmatic
competence in the previous literature. The
study was guided by the research question:
Do individual difference factors affect
changes in appropriate and fluent speech act
production?
Methods
Participants
Participants were 48 Japanese students of
English as a foreign language (EFL) in an
English-medium university in Japan.
1
In the
school all courses are taught in English, 50-60% of the instructors are foreign nationals,
and 10-15% of the student population are
international students. All first-year students
live in a dormitory with international
students. The participants (hereafter EFL
learners) were first semester Japanese
students enrolled in the intensive English
program. There were 16 males and 32
females, ranging in age from 18 to 21 with
an average age of 18.33 (SD=.66).
2
Three students had experienced living in U.S.A.
for one month. From class observations,
interviews with instructors, and textbook
analyses, it was concluded that the
participants did not receive focused
instruction on pragmatics.
Instrument
This study examined the development of
pragmatic production – the ability to convey
intentions appropriately and fluently in
speech acts. A computerized oral discourse
completion test (oral DCT) was developed
to examine this ability. Participants read
situational descriptions and produced two
speech acts: requests and opinions. These
two were selected after consulting Garcia's
(2004b) analysis of naturalistic
conversations in university settings. Garcia
analyzed conversations across three
registers: conversations between a professor
and student, conversations among study
group members, and service encounter
conversations. She found that speech acts of
directives (request) and expressives
(opinions) are common in the corpora. From
the examples in the corpus data, request and
four opinion situations were adapted for this
study.
Oral rather than written DCT was selected
because this study measured fluency as part
of the construct of pragmatic competence.
However, I acknowledge the weakness of
the DCT instrument. While the DCT format
was necessary to collect a large amount of
data at once, DCT has been criticized
because it lacks authenticity and participants
have more time to plan their responses than
in face-to-face interaction (Geluykens,
2007). These limitations should be kept in
mind when interpreting the present findings.
Target speech acts of requests and opinions
were divided into two different situational
categories based on three contextual factors:
interlocutors' power difference (P), social
distance (D), and the size of imposition (R)
(Brown & Levinson, 1987). In one situation
type, the power relationship was equal, the
distance between the interlocutors was
small, and the degree of imposition was
small (PDR-low). In the other situation type,
the listener had greater power, the
interlocutor distance was larger, and the
degree of imposition was also large (PDR-high). See Table 1 for sample speech acts.
Appendix A contains the copy of the
instrument.
The length of situational descriptions was
controlled across test items. The number of
words used in each description ranged from
55 to 57 with a mean of 55.55 (SD=.60).
The vocabulary used to write descriptions
came from the top 3,000 words in the
JACET (Japan Association of College
English Teachers) basic word list (JACET,
2003). Two versions of the test were
prepared and used alternatively to minimize
the practice effect. The versions differed in
proper nouns used, object names, dates and
times, and conversation topics in the
scenarios. The order of the items was
randomized each time.
The final version of the instrument had a
total of 14 items: four PDR-low speech acts,
four PDR-high speech acts, four filler items,
and two practice items. The oral DCT was
computerized using the Revolution software
(Runtime Revolution Ltd., 1997). The
situations were presented in written form on
the screen. The instrument was piloted with
25 native English speakers and 12 ESL
students prior to the main study.
Evaluation of speech acts: Appropriateness
and fluency
Participants' speech acts were evaluated on
appropriateness and fluency.
Appropriateness was defined as the ability to
perform speech acts at the proper level of
politeness, directness, and formality. It was
assessed using a five-point rating scale
ranging from 1 (very poor) to 5 (excellent)
(see Appendix B). Four native speakers of
English evaluated the samples.
3
Interrater
reliability was r=.92. About 2.2% of the
samples that had two points off in rating
were discussed in the follow-up meetings.
For the cases with one point off, the average
score was assigned as the final score. In
order to maintain consistency in rating,
speech samples used in the norming session
included samples taken from different data
collection sessions. Fluency was
operationalized as speech rate and was
measured as the number of words spoken
per minute. False starts and repetitions were
excluded from word count.
Measures for individual differences (ID)
Three ID factors were measured: general
proficiency, orientation toward English
study, and lexical access skill. All factors
were measured multiple times during the
study period, conforming to the current
process-oriented approach that individual
differences factors are not fixed and change
over time (DÖrnyei, 2009).
Proficiency
Participants' proficiency was measured with
the institutional TOEFL (ITP TOEFL),
consisting of three sections: listening,
grammar, and reading. The ITP TOEFL was
given three times at about same timing with
the oral DCT. Different versions of the test
were used to avoid practice effect.
Orientation to English study
This study used a portion of Yashima's
(2002) survey to measure orientation toward
English study because the measure was
developed specifically for Japanese learners
of English who were also the target
population in this study. Yashima used 12
items to measure Japanese EFL learners’
specific orientations toward studying
English. Adapting Gardner and Lambert’s
(1972) motivation framework, half of the
items measured the degree of integrative
orientation toward learning English (i.e.,
learning English to develop friendship with
English speakers), and the other half
measured the degree of instrumental
orientation (i.e., learning English for utility-
based purposes).
5
Students rated the degree
to which each statement matched their
reason to study English on a 7-point scale.
See examples (The items are in Yashima,
2002, p.66):
As a reason to study English:
1. It will allow me to get to know various
cultures and people.
2. It will be useful for a future career.
Lexical access test
Lexical access skill was operationalized as
the ability to make speedy lexical
judgments. It was measured by a
computerized word recognition test called
lexical access test (LAT) adapted from
Segalowitz and Freed (2004). The LAT had
40 frequent English words (e.g., "tiger"),
and participants made quick judgments on
whether each word that appeared on a
computer screen referred to a living or
nonliving object. The words were
considered familiar to the Japanese
participants because they appear in the basic
vocabulary list in school textbooks
authorized by the Japanese Ministry of
Education. In addition to English lexical
access, Japanese lexical access was
measured in order to control individual
differences coming from L1 processing. The
test was given three times at same timing
with the oral DCT with a different version
each time.
Data collection procedures
The oral DCT was given individually three
times during one academic year: Time 1
(April), Time 2 (July), and Time 3
(December). Two versions of the oral DCT
were used across three time points (see
instrumentation section). Students put on
headphones with a microphone attached and
read directions in English with Japanese
translations. They were told to read each
situational scenario and respond as if they
were in a real situation and performing the
role. They had two practice items. Each item
started with a situational scenario on the
computer screen. They were allowed to take
as much time to read the scenario and
prepare for the speech act. When they were
ready, they clicked on the "continue" button.
Planning time was measured between the
moment when the situational scenario
appeared on the computer screen until the
moment when the participants clicked on the
"continue" button. Once they clicked the
button, the scenario disappeared and the
message "start speaking" appeared on the
screen. After they finished the item, they
moved on to the next item. The computer
recorded their speech.
The LAT was given to the participants using
the same computers. The participants read
instructions on the screen in Japanese. After
practicing four items, they completed the
test items. When a word appeared on the
screen, they made a quick judgment on
whether the word referred to a living or
nonliving object by pressing the key '1' for
'living' and '2' for 'nonliving,' which were
adjacent to each other on the keyboard.
Response time was measured between the
moment when the target word appeared on
the screen and the moment when the
participants made a judgment and pressed
the number key. The computer recorded all
responses and their latencies. After the
participants finished the oral DCT and LAT,
they completed a paper-and-pencil version
of the survey measuring learners’ orientation
to English study. The survey was given
twice, at Time 1 and 3. Participants took
about 10 minutes to complete the survey.
The ITP TOEFL was given on separate
days, approximately one week after they
completed the oral DCT.
Data analysis procedures
This study examined the effects of learner
characteristics on the ability to produce two
speech acts (requests and opinions)
appropriately and fluently. Appropriateness
was evaluated on an interval scale between 0
and 20 across two situation types: PDR-low
(requests and opinions combined; k=4, scale
of 0-20) and PDR-high speech acts (requests
and opinions combined; k=4, scale of 0-20).
Fluency was assessed as speech rate (words
per minute).
The effects of ID factors on pragmatic
change were examined by using hierarchical
linear modeling (HLM), with "time" as an
independent variable, pragmatic abilities
(i.e., appropriateness and fluency) as
dependent variables, and the ID and context
measures as covariates. This was
accomplished by using a mixed model
approach that revealed whether the
covariates have a significant main effect on
pragmatic change or a significant interaction
effect with "time" on pragmatic change.
HLM was used for two reasons. First, it is
appropriate for data collected from intact
classes without random sampling
(Raudenbush & Bryk, 2002). Second, it can
be used for covariates that form repeated
measures data (data taken from the same
individuals over multiple times). Because
the covariates were measured at multiple
times in this study, HLM allowed us to
examine the effect of covariates that may
change over time. Normality of distributions
of residuals was checked by inspecting Q-Q
plots. Because planning time data and
lexical access response times data were not
normally distributed, following Tabachnick
and Fidell (2001), a logarithmic
transformation was performed before
submitting the data to statistical analyses.
The alpha-level was set at .05.6
Results
Descriptive statistics
Tables 2 and 3 display descriptive statistics
of the two aspects of speech act production
analyzed in this study: appropriateness
scores and speech rate. There was a large
discrepancy between low- and high-imposition speech acts in both measures at
all time points, confirming the distinct
differences between the two situation types.
Low-imposition speech acts were easier and
faster to produce than high-imposition
speech acts. Paired-sample t-test results
confirmed the situational differences for all
variables, at all time points.
Notes. Low- and high-imposition situations
include speech acts of requests and opinions
combined. Appropriateness was assessed on a
five-point scale ranging from 1 to 5.
Notes. Speech rate refers to the average number
of words spoken per minute.
Tables 4 to 6 display descriptive statistics of
individual differences (ID) factors. As
shown in Table 4, the learners demonstrated
an increasing proficiency over time. Paired-sample t-test results revealed significant
differences for all time contrasts, t=-18.34
(p=.000) at Time 1-2, and t=-4.91 (p=.000)
at Time 2-3.
Table 5 displays descriptive statistics for
orientation toward English study. This
variable showed no significant change
between Time 1 and 3: t=-.42, p=.68 for the
integrative orientation, and t=.69, p=.50 for
the instrumental orientation. At both times,
the learners' integrative orientation was
higher than their instrumental orientation.
Notes. Six items of the seven-point Likert scale
measured integrative and instrumental
orientation. The mean refers to the average. A
larger number means a greater degree of
orientation.
The learners showed improvement in their
lexical access speed (Table 6). Paired-sample t-test showed significant difference
in the response times at Time 1-2 (t=3.33,
p=.002) and at Time 2-3 (t=4.06, p=.000).
Effects of individual differences and
language contact on pragmatic change
Statistical analyses were performed to
examine the effects of ID factors on
learners’ changing ability to produce speech
acts. The first ID factor assessed was general
proficiency. The mixed-model revealed no
significant main effect of proficiency on
changes of appropriateness or fluency,
neither PDR-high- or low speech acts.
Findings indicate that pragmatic abilities
measured did not change over time
corresponding to the changes in the ITP
TOEFL score.
Regarding the orientation toward English
study, the mixed-model revealed the
significant main effect of integrative
orientation on the appropriateness score of
PDR-low speech acts, F=14.72, p=.000
(Cohen’s d=.50 for Time 1-2; Cohen’s
d=.97 for Time 2-3). Results suggest that the
students who reported studying English
because of their interest in social interaction
with English speakers showed a greater gain
in their ability to produce PDR-low speech
acts appropriately. Instrumental orientation,
however, had no significant effect on any of
the pragmatic abilities measured in this
study.
Finally, the mixed-model revealed
significant main effect of lexical access skill
on speech rate, F=9.62 (p=.003) for PDR-low speech acts (Cohen’s d=1.32 for Time
1-2 and .79 for Time 2-3) and F=9.32
(p=.003) for PDR-high speech acts (Cohen’s
d=1.40 for Time 1-2 and .47 for Time 2-3).
Results indicate that the faster the lexical
access became, the faster the speech rate
became, regardless of the situation type.
7
Discussion
This study found that the change in different
aspects of pragmatic competence
(appropriateness and fluency) over different
task situations (PDR-high and low) was
affected differently by different ID factors.
No effect of general proficiency on
pragmatic change was somewhat surprising,
considering that a large body of existing
literature clearly suggests an influence of
proficiency on pragmatic performance.
Previous cross-sectional studies generally
found that higher proficiency learners
perform pragmatic functions more
appropriately and fluently, and employ more
target-like pragmalinguistic forms than
lower proficiency learners (e.g., Dalmau &
Gotor, 2007; Félix-Brasdefer, 2004, 2007;
Garcia, 2004a; Geyer, 2007; Pinto, 2005;
RÖver, 2005; Taguchi, 2007a, 2011;
Yamanaka, 2003; Xu et al., 2009). There are
several interpretations for the contradicting
findings from this study. First, proficiency
could differentiate levels of pragmatic
performance but may not be the factor that
influences developmental changes. Previous
research largely explored the proficiency-pragmatics relations in a cross-sectional,
single-moment design by comparing learner
groups of different proficiency levels.
Hence, in a longitudinal design, proficiency
may exhibit different degrees of influence
on pragmatic competence. The findings
suggest that a threshold-level proficiency is
necessary to perform pragmatic functions,
but proficiency alone is not sufficient for
learners to make further progress toward a
full mastery of pragmatic abilities. In
addition, the length of this study (eight
months) might have affected the results. In
earlier cross-sectional studies that compared
pragmatic competence across different
proficiency levels, the difference in learners’
proficiency between lower and higher
proficiency groups was much greater than
the differences between the learners’
proficiency across time spans in this study.
If not proficiency, what factors affect
pragmatic development? This study revealed
different effects of IDs across different
dimensions of pragmatic abilities.
Appropriateness of PDR-low speech acts
was affected by integrative orientation.
According to Gardner (1985), an integrative
orientation involves a wish to develop an
understanding and possibly become part of
the target language culture. In Yashima
(2002), integrativeness (termed as
"intercultural friendship orientation")
predicted motivation, which in turn led to L2
proficiency. In the present study, this same
orientation had a positive impact in the
development of PDR-low speech acts,
indicating that a positive affective
disposition toward the L2 community and a
desire to interact and identify with its
members was a key mediating factor for the
attainment of PDR-low speech acts. The
PDR-low situations used in this study were
informal situations involving talking to
friends about routine matters. Making a
small request to a friend or giving a personal
assessment on a subtle matter were probably
frequent in the bilingual campus
environment and were part of daily
communication among the learners studied
here. Integrative orientation explained this
type of speech act’s gains probably because
one's needs and desire to become part of the
target language culture led to greater access
to the L2 community, which involved plenty
of opportunities to practice the target PDR-low speech acts.
Appropriateness of PDR-high speech acts,
on the other hand, was not affected by the
integrative orientation probably because
formal communication opportunities (e.g.,
talking to a professor on a serious matter)
were not as frequent. As a result, learners'
interest in social interaction with members
in the L2 community did not matter as much
for the development of PDR-high speech
acts because the type of social interaction in
the community did not provide as many
opportunities to observe or practice formal
language. This was somewhat supported by
learners’ responses in the language contact
survey I administered as background survey.
Average amount of time spent talking to
teachers (presumably involving formal
speech) was less than one hour per week,
while that of talking to friends (presumably
involving informal speech) was 2.35 hours
per week. However, regrettably, this study
was not able to examine the actual level of
formality involved in interaction with
teachers versus with friends. Because no
variables in this study revealed significant
impact on PDR-high speech act, it remains
for future research to reveal precise factors
that affect the development of this type of
speech act.
Gains on the fluency of speech acts (i.e.,
speech rate) were significantly affected by
lexical access skill. The learners with faster
lexical access showed a faster speech rate
regardless of the situation type. These
findings are consistent with previous
findings that revealed a significant
relationship between lexical access skill and
processing speed in pragmatics. Taguchi
(2007b) found that lexical access skill was
associated with response times of pragmatic
comprehension, but not with accuracy of
comprehension. Similarly, in Taguchi
(2008b), lexical access speed and language
contact significantly correlated with gains in
pragmatic comprehension speed, but not
with gains in accuracy. These previous
findings suggest that the lexical access skill
– the ability to assign meaning quickly –
serves as a component process that affects
the comprehension speed of pragmatic
meaning. The present study adds to the
previous findings in that speedy lexical
access could promote fluent production of
speech acts. The present results are also
consistent with Segalowitz and Freed (2004)
who found significant correlation between
oral fluency and lexical access speed,
indicating that the ability to access word
meaning quickly could enhance oral fluency
in general.
In conclusion, responding to Kasper and
Rose's (2002) claim that individual
differences and L2 pragmatic development
is the most under-researched area, this study
investigated the effect of three ID factors on
gains in pragmatic production. Results
showed that there was no single ID factor
that had a significant effect on all aspects of
pragmatic competence. The findings suggest
that the aspect of pragmatic competence and
individual characteristics interact with one
another. The interaction gleaned in this
study is two-fold: gains in fluency of
pragmatic competence (speech rate) was
affected by the cognitive variable, namely
lexical access, while gains in the
appropriateness of pragmatic competence
was affected by the affective variable,
namely learners’ orientation toward target
language community (integrative
orientation). These findings lend support to
the recent claim that individual attributes are
multi-componential (DÖrnyei, 2009). There
is a combined operation of mixed individual
factors on change in language abilities, and
it was found to be true in pragmatic
competence in this study.
Limitations of the study and directions
for future research
This study has several limitations that need
to be addressed in future research. First,
because the ID factors examined in this
study were no way exhaustive, future
research should explore a greater number of
affective and cognitive factors to expand the
scope of ID research. With a larger sample
size, future research could use a different
statistical method, such as structural
equation modeling and path analysis, to
visualize the hierarchy and direction of
interaction among the ID factors in their
effect on pragmatic achievement. Similarly,
this research is limited in that it examined
the development of one type of pragmatic
function, namely the production of speech
acts, using a small-scaled instrument. With a
larger test battery, future research should be
able to expand the scope of the target
pragmatic features and track down the
change of different pragmatic sub-competencies in relation to learner-specific
and contextual factors.
Finally, this study pursued a quantitative,
group-based analysis of IDs focusing on the
central tendency of certain characteristics. In
future research, a qualitative, individual-level analysis focusing on idiosyncratic
deviations from the group average would be
useful in investigating pragmatic
development from a perspective of a socially
situated individual process. When combined,
the group and individual-level findings will
mutually inform each other with the synergy
between them illuminating the complex
intersect between individuals, context, and
changes in pragmatic abilities.
Notes
1. In addition to the L2 learners, 24
native speakers of English completed
the oral DCT and provided base-line
data. Due to the space limit, the data
is not reported here.
2. The sample was skewed toward
women because the male-female
ratio in the institution is three-to-seven. I acknowledge the possibility
that findings from this study were
gender-biased.
3. Four raters of mixed cultural
background: an Australian white
male and female, an African-American male, and a female
Japanese-American, evaluated the
samples. They had little background
in Applied Linguistics or related
field, and had limited experience in
teaching English. They were not
instructors of the participants.
4. Gardner (1985) distinguishes
between motivation and orientation.
Orientation refers to the long-range
goals for learning a language, while
motivation refers to the effort
learners are prepared to make to
learn L2. This study measured
orientation following Yashima
(2002).
5. The reliability for the orientation
survey was .78 in Yashima (2002).
In my study it was .73.
6. Although I consulted with a statistics
expert to ensure the appropriateness
of the use of HLM with the current
sample size of 48, results should be
interpreted with caution due to the
small sample size. Separate mixed-model analyses were performed for
individual covariates the covariates
did not correlate significantly. When
there was no interaction effect but
main effect, the model was adjusted
by re-running HLM without
interaction effects. The main effect
of covariate was confirmed in all
cases. Effect size (Cohen’s d) was
calculated by dividing parameter
estimate by the standard deviation of
dependent variable. Model fit was
checked by inspecting the residuals-covariate scatter plot.
7. As a post hoc analysis, the
participants’ lexical access in their
L1 (Japanese) was measured and
assessed in relation to their L2
lexical access (English). There was
also a significant interaction effect
between English lexical access and
Japanese lexical access on speech
rate: F=9.52 (p=.003) for PDR-low
and F=9.29 (p=.003) for PDR-high
speech acts. Hence, the effect of
English lexical access on speech rate
depended on the Japanese lexical
access speed.
