You are on page 1of 22

Student and Instructor Beliefs and

Attitudes about Target Language Use,


First Language Use, and Anxiety:
Report of a Questionnaire Study
GLENN S. LEVINE
Department of German
400 Murray Krieger Hall
University of California, Irvine
Irvine, CA 92697
Email: glevine@uci.edu

This article presents the results of an anonymous, Internet-based questionnaire study on target
language (TL) and first language (L1) use in university-level foreign language (FL) classes.
The participants were 600 FL students and 163 FL instructors. The goals of the study were to
develop preliminary components of a descriptive model of TL and L1 use and explore the
relationships between TL use and student anxiety about TL use. It was hypothesized that
(a) amounts of TL use would vary according to constellation of interlocutors and communi-
cative contexts, and (b) the amount of TL use overall would correlate positively with student
anxiety about it. The results support the first hypothesis. The second hypothesis was not
supported. Analysis of the data revealed a negative relationship between reported amounts of
TL use and reported TL-use anxiety. Tenets are offered for maximizing TL use while at the
same time granting to L1 pedagogically sound functions.

THERE ARE LIKELY FEW FOREIGN LAN- clusive TL use, there appear to be two opposing
GUAGE (FL) instructors who have not developed positions. Though they expressed it only in pass-
an individualized approach to classroom target ing, Krashen and Terrell (1983) stated that the
language (TL) and first language (L1) use.1 This natural approach is “based on the use of the lan-
approach can be influenced by pedagogical train- guage in communicative situations without re-
ing, knowledge of the second language acquisi- course to the use of the native language” (p. 9).
tion (SLA) literature, official policy, and class- The choice of the word “recourse” is indicative of
room experience, yet often it appears to be based the stigma associated with the use of the L1 in the
primarily on classroom experience and intuitions communicative FL classroom (Cook, 2001). This
about what feels right. Questions of how much TL view of classroom language use also appears evi-
the instructor uses and in what contexts, as well as dent even in much of the SLA literature, espe-
how much TL the instructor expects students to cially studies dealing with input and interaction
use, become inextricably linked to all other class- (e.g., Brooks, 1990; Ellis, 1999; Gass, 1997;
room practices. Yet despite the ubiquitous nature Johnson, 1995), in which codeswitching or the
of the TL-use issue, relatively little attention has use of the L1 versus the TL often does not enter
been paid to TL and L1 use in the literature on in- into the discussion. Empirical works dealing with
structed adult second language (L2) acquisition. TL use as its own construct that largely support
When researchers specifically address the issue natural-approach assumptions about TL use in-
of classroom TL use, especially the notion of ex- clude the studies by Duff and Polio (1990) and
The Modern Language Journal, 87, iii, (2003) Polio and Duff (1994).
0026-7902/02/343–364 $1.50/0 On the other side of the issue, represented by
©2003 The Modern Language Journal Antón and DiCamilla (1999), Atkinson (1987,
344 The Modern Language Journal 87 (2003)
1993), Auerbach (1993), Belz (2003), Blyth decisions about TL use, as well as in designing
(1995), Castellotti (2001), Cook (1999, 2001), observational and experimental research.
Franklin (1990), Nation (1997), and Nizegorod-
cew (1996), it is argued that the “recourse-to-L1” EMPIRICAL STUDIES OF TARGET
position has created a classroom dynamic in LANGUAGE USE
which the use of the L1 is at best discouraged,
and at worst stigmatized. In reaction to what they For the sake of brevity, this review of the litera-
view as misguided and unrealistic pedagogical ture is limited to a small selection of the works
principles, these scholars have staked a claim for that deal specifically with TL use as a construct.
a sanctioned role for the learner’s L1 in the lan- Furthermore, with the exception of Cook’s
guage classroom. (2001) article, this review is also limited to em-
Unfortunately, with the exception of the few pirical studies. For a more complete overview of
empirical studies that I describe in the literature the literature on TL and L1 use, the reader is
review that follows, both sides of this debate often referred to Turnbull and Arnett (2002).
base their assumptions and arguments largely on One of the earliest studies to break ground on
intuitions about best practices, anecdotal evi- classroom TL use was conducted by Guthrie
dence, and personal classroom experience. (1984). Exploring the question of optimal class-
Whereas these nonempirical works are well-in- room conditions for L2 acquisition, the re-
formed sources of information, they do not suf- searcher investigated the TL use of 6 university
fice in the face of the pedagogical and curricular French instructors, all of whom reported teaching
ramifications of the issue. As Macaro (2001) and with a communicative approach. Through an
Turnbull and Arnett (2002) pointed out, the fact analysis of 10 hours of instruction, Guthrie discov-
of the matter is that, to date, we have relatively ered that even in a multisection course at a single
little empirical evidence as to the amount or na- institution there is a great degree of variability in
ture of TL versus L1 use upon which to make the amounts and purposes of TL and L1 use by
sound pedagogical and policy decisions. instructors. Yet overall, most instructors used the
The present research report begins to address TL a great deal of the time. Of the 6 instructors, 5
this problem. The study described here repre- apparently used the TL 83% to 98% of the time
sents the first part of a multipart study of TL and (average of the two lessons recorded for each in-
L1 use in the FL class. Presented here are the structor). The students’ use of the TL was more
results of an anonymous questionnaire com- variable, ranging from 68% to 93% (average of the
pleted by 600 undergraduate FL students and 163 two lessons). Yet it is interesting that the students
FL instructors. The purpose at this stage was to in the classes recorded apparently did not talk very
identify what students and instructors believe much overall (in either language); they spoke
goes on with the TL, with a central focus on the ranging from 9% to 24% of the time.
perceptions and experiences of the students, as Noting the dearth of empirical evidence on TL
well as on variables associated with more or less use, Duff and Polio (1990) undertook a qualita-
TL use. Of particular interest was the relationship tive investigation of instructors’ use of the TL in
between reported amounts of TL use and re- 13 different university-level language classes. In
ported anxiety about TL use. In addition, this the ratio of TL to L1 use, the authors found an
article attempts to offer a first (but by no means even broader range of TL to L1 use than Guthrie
last) response to Macaro’s (2001) challenge that did, from 10% to 100%. Duff and Polio suggested
“future research needs to establish some princi- several factors that they believed affected the
ples for codeswitching in FL classrooms by under- amount of instructor TL use, namely, language
standing its functions and consequences” (p. type, departmental policy on TL use, lesson con-
545). Therefore, in the Discussion and Conclu- tent and objectives, pedagogical materials, and
sions section, I offer some preliminary tenets for formal training. They noted that the instructors
classroom codeswitching that appear to be indi- opted to use the L1 in many administrative or
cated by the questionnaire data. other situations in which the TL could have been
The reader is asked to keep in mind through- used. In their opinion, using the L1 in these con-
out that the data described are not “reality,” that texts deprives students of many good opportuni-
is, they are not actual recordings of TL use such ties to hear and process the TL for a range of
as those collected by Duff and Polio (1990) or by communicative functions. Finally, most of the stu-
Guthrie (1984); rather they are students’ and in- dents claimed to be satisfied with the status quo
structors’ estimations of and beliefs about TL use. of their particular class, regardless of the re-
Yet I suggest that these perceptions and beliefs corded amount of instructor TL use.
can be a powerful tool in making pedagogical In their work 4 years later, Polio and Duff
Glenn S. Levine 345
(1994) returned to the data recorded for the When instructors initiated a switch to the L1, it
earlier study and considered the primary ques- appeared that they did so for the sake of effi-
tion, When do teachers tend to use English, ciency and expediency, or to impose discipline or
rather than the TL, and for what functions? The keep control of the group (students were 11- to
researchers focused on what they saw as one of 14-year-olds). Considering the low proportion of
the reasons for the students’ general lack of suc- L1 use overall, the author concluded that there is
cess in the L2 acquisition endeavor: “Students are little reason to advocate the exclusion of the L1
simply not engaged in meaningful interaction in from the classroom. In accord with the assertions
the FL during class time” (p. 313). Through a made by Cook (2001), the question is raised
discourse analysis of judgment samples of the whether the prevailing taboo against the L1
data, they considered the type of TL available to might not stifle reflective practice among instruc-
students vis-à-vis “the contexts in which English is tors (p. 545), that is, that the dogma of exclusive
used in lessons instead of the target language” TL use limits what we could accomplish in the
(pp. 313–314). Their analysis revealed that many classroom. Still, in line with Polio and Duff, Ma-
instructors appeared to engage in intrasentential caro suggested that rather than going the path of
codeswitching, especially of specific lexical items. least resistance by switching to the L1, certain
Also, instructors often dealt with communication contexts present themselves as opportunities to
breakdowns by switching to English rather than expand students’ vocabulary and hence the
negotiating in the TL. Last, the instructors ap- range of functions of the TL. Macaro stated that
peared reluctant to teach grammar or engage in we need “a framework that identifies when refer-
classroom management in the TL. It is interesting ence to the L1 can be a valuable tool and when it
that some instructors appeared unaware of “how, is simply used as an easy option” (p. 545).
when, and the extent to which they actually use Rather than quantifying the amount of TL and
English” (p. 320). Some instructors apparently L1 use, Antón and DiCamilla (1999) favored in-
believed they used the TL exclusively and “urged vestigating the discourse functions of L1 during
the students to use it too, but would not necessar- L2 classroom interaction and took a Vygotskian,
ily do so themselves” (p. 320). interactionist approach to the issue. The re-
As did Polio and Duff, Nzwanga (2000) also searchers studied the verbal interaction of 5 pairs
found that “despite their effort to avoid it, both the of adult learners of Spanish as they worked col-
teachers and students appealed to CS [codeswitch- laboratively on three Spanish writing tasks. It was
ing] for a number of reasons” (p. 103). Still, in the concluded from the transcripts that “L1 use pro-
14 hours of recorded university-level French class- vides, through collaborative dialogue, an oppor-
room conversation, instances of L1 use by the in- tunity for L2 acquisition to take place” (p. 237).
structor and students were relatively scarce (p. The L1 serves not just as the code that learners
73). Indeed, in the classes observed, the instruc- resort to when language ability proves inade-
tors explicitly forbade the use of the L1. From his quate, rather it serves three main functions: “con-
detailed discourse analysis, Nzwanga concluded struction of scaffolded help, establishment of in-
that whereas communicative approaches to in- tersubjectivity, and use of private speech” (p.
structed L2 acquisition may dictate maximal or ex- 245). The L1 is not simply a metalinguistic tool,
clusive TL use, it appeared that the L1 did and rather it is a “means to create a social and cogni-
should have a role to play. Specifically, code- tive space in which learners are able to provide
switches, which the author acknowledges were re- each other and themselves with help throughout
ally discourse switches (p. 104), were most likely to the task” (p. 245). In this detailed, qualitative
occur during pair or group work (presumably analysis, the authors showed that the relationship
among the students), before a quiz, or during the between the L1 and the L2 in the classroom can-
presentation stage (presumably by the instructor). not be viewed in simple terms of how much TL
Macaro (2001) also was interested in how versus how much L1 is used. Underscoring the
much L1 instructors use, why they claim to use it, inextricable nature of language and thought and
and what factors appear to influence their deci- rejecting the notion of a simple prohibition of
sion to use it. Specifically, the author investigated the L1, Antón and DiCamilla, it appears, concur
how the decision to use the L1 is influenced by with Macaro that a principled framework is
beliefs about TL and L1 use, pedagogical train- needed in which the L1 serves as one of many
ing, or governmental or institutional policy. As communicative tools.
did Guthrie and Nzwanga, Macaro discovered Though not based on empirical evidence,
that overall, very little L1 was used in the classes Cook’s (2001) arguments for allowing the L1 in
recorded, either by the instructor or the students. the language class deserve attention here, for I
346 The Modern Language Journal 87 (2003)
believe that the position he has taken accords to ety (e.g., Bailey, Onwuegbuzie, & Daley, 2000;
some extent with the findings of Antón and Di- Horwitz et al., 1986; MacIntyre, 1995; MacIntyre
Camilla, Macaro, and others. In addition, several & Charos, 1996; Young, 1990) addresses many
of his observations were employed directly in social and personality variables, but learner anxi-
constructing the questionnaire for the present ety about TL use (whether the TL is used by the
study. After sketching the predominant “mono- instructor or the student) is seldom taken into
lingual principle” of 20th century L2 instruc- consideration.
tion—the belief that only the L2 has a place in For this study, it was hypothesized that the
the classroom (p. 404)—Cook attempted to show amount of instructor and student TL use corre-
that the prevailing motivations that have been lates positively with the students’ sense of anxiety
offered in support of the classroom as a virtual L2 about TL use (i.e., more TL use co-occurs with
environment (Macaro, 2001) do not in them- greater TL-use anxiety). Three points support the
selves preclude a role for the L1. Furthermore, he assumption of a positive relationship between
asserted that this position “has prevented lan- these two variables. First, a classroom dynamic in
guage teaching from looking rationally at ways in which “resorting” to L1 use is considered a short-
which the L1 can be involved in the classroom” coming or lapse of appropriate behavior—argu-
(p. 410). He then described some alternative ap- ably the prevalent dynamic in many university FL
proaches that overtly include L1 in the classroom, classes—could likely lead to anxiety about TL use
such as the “Key Two-Way Model” and the “New for a considerable number of learners. Second,
Concurrent Method.” The latter example, appar- the literature on FL anxiety points toward a posi-
ently advocated by Cook, “acknowledges code- tive relationship between TL use and anxiety.
switching as a normal activity and encourages the Young (1990) found that students generally be-
students to see themselves as true L2 users, at come anxious when they have to use the TL in
home in both languages” (p. 412). At the same front of others. Horwitz et al. (1986) suggested a
time, this method encourages teachers to use the relationship between FL anxiety and unwilling-
students’ L1 (Spanish) “when concepts are im- ness to interact verbally. And whereas this ten-
portant, when the students are getting distracted, dency is, of course, influenced by factors such as
or when a student should be praised or repri- personality traits or overall willingness to commu-
manded” (p. 412). Note that it is exactly in these
nicate (see MacIntyre & Charos, 1996; MacIntyre,
sorts of situations that others (e.g., Polio & Duff,
Clément, Dörnyei, & Noels, 1998), anecdotal evi-
1994) have argued that the L2 is needed. Cook
dence also appears to support a positive relation-
(2001) then detailed specific situations in which
ship: A common reason for using the L1 and
he believed the L1 could be used: to check mean-
allowing students to use it is that it helps lower
ings of words, explain grammar, organize tasks
learners’ affective filter (Krashen, 1982). One in-
and give directions, maintain discipline, adminis-
structor who completed the questionnaire in the
ter tests, and carry out classroom group activities.
present study offered the following comment:

TARGET–LANGUAGE–USE ANXIETY There are some students in the classroom which [sic]
feel comfortable with FL and using it (although not
Along with investigating amounts and contexts all the time). We need to understand that as an in-
of TL use by instructors and students, this study structor we [sic] have to be careful not to frustrate
also considered the construct “TL-use anxiety,” our students, especially those just starting with FL.
regarded here as a debilitating state anxiety Depending on the composition of the class, I evaluate
(Horwitz, Horwitz, & Cope, 1986; MacIntyre, how much English I need to use/not use. I truly try
1995; Scovel, 1991). There are three reasons for to use only Spanish in my classes. My purpose is for
focusing on this variable. First, other variables the students to feel comfortable in all areas of FL and
that influence how much TL and L1 instructors to continue in the program (if possible).
choose to employ, such as pedagogical training or
official policy, have been dealt with recently (Ma- Polio and Duff (1994) also presented a statement
caro, 2001; Turnbull & Arnett, 2002). Second, made by a German instructor who believed that
TL-use anxiety is a variable based on the learner’s “if you want to create some sort of relaxed atmo-
perceptions and experiences, whereas most stud- sphere . . . I think this is hard to do in German
ies on TL use to date have focused primarily on only” (p. 318). I suggest that these comments
the verbal behavior and attitudes of the instructor accord with the opinion of many FL instructors
and only secondarily on that of the students. and support the notion that in many FL class-
Third, the literature on language learning anxi- room situations, greater TL use may be accompa-
Glenn S. Levine 347
nied by greater anxiety about TL use for many two a priori hypotheses followed from these re-
learners. search questions:
H1 Reported amounts of TL use differ ac-
RESEARCH QUESTIONS AND HYPOTHESES cording to constellation of interlocutors and
communicative contexts.
The empirical studies described above shed
H2 Reported amounts of TL use by instructor
light on many aspects of the relationship between
and students correlate positively with students’
TL and L1 use in the FL class. Yet each of them is
sense of anxiety about TL use.
based on a fairly small sample, or on language
classes at a single institution. As a consequence, Based on these hypotheses, two a priori con-
there is some question about the generalizability structs were assumed for the purposes of design-
of the findings. Also, by and large these studies ing and analyzing the questionnaires. I called
focused primarily on the verbal behavior of the these Target Language Use (TLU) and Target-
instructor. In order to develop an adequate de- Language-Use Anxiety (TLA).
scriptive model of TL/L1 use, a study was needed After describing the questionnaire study and
that would focus on the verbal behavior of the the procedure, I present the characteristics of the
students as well as that of the instructors, include participant sample and then address each of the
greater numbers of participants from more lan- research questions in turn. Thereafter, I discuss
guage learning situations (see Rifkin, 2000), and the findings and their pedagogical implications.
take into account not only overall amounts of
each language but also (a) constellations of inter-
METHODS
locutors and (b) contexts and functions of com-
munication. Therefore, an important first step is The Questionnaire
to determine, by using a sizable sample, what
instructors and students believe goes on in the A questionnaire was devised that would mea-
classroom in these two areas. Specifically, this sure first- and second-year FL students’ and in-
study provides a cross-sectional description of stu- structors’ (a) estimations of the quantity of TL
dent and instructor verbal behavior as it relates to use in different classroom contexts in university-
TL and L1 use. A second goal is to identify covari- level FL classes, (b) beliefs about the importance
ates of TL use, and particularly, variables that of TL use, and (c) beliefs about student anxiety
relate to amounts of TL use or TL-use anxiety, or experienced through TL use, also with regard to
both. Although the range of variables that could specific classroom contexts. Two versions of the
be analyzed is large, the present study focuses on questionnaire were created, one intended for stu-
a few demographic (age, gender, language back- dents, and one for instructors (see Appendixes A
ground) and classroom variables (expected and B).
grade, level of motivation, TL-use strategies in- The first version of the questionnaire was con-
struction, instructor characteristics). To meet the verted to Web format and piloted with a small
two goals, the following research questions were group of students and instructors in order to
formulated: refine it.2 Based on student and instructor com-
ments, changes were made in the wording of
1. What do students and instructors believe several items, and several items were deleted alto-
goes on in the FL classroom with the TL and the gether. The final student questionnaire con-
L1? Specifically, tained seven sections for a total of 63 items. The
1a. How does reported TL use differ for dif- final instructor questionnaire contained six sec-
ferent interlocutors? tions for a total of 58 items. Cronbach’s alpha for
1b. How does reported TL use differ in dif- each a priori construct (TLU and TLA) in both
ferent communicative contexts? the instructor and student questionnaires was .78
1c. How do instructors and students perceive or above, suggesting that the reliability of the
TL-use anxiety, particularly in different com- instrument was satisfactory.3
municative contexts?
2. How do reported amounts of TL use and
TL-use anxiety relate to each other and to per- Procedure
sonal or classroom variables?
Beginning at midterm of Fall 2000 (October),
Although I considered the study exploratory and the final versions of the questionnaires were
identified most of the initial patterns through posted on the Internet, linkable through a gate-
data mining procedures (e.g., factor analyses), way site. An appeal for participation was made
348 The Modern Language Journal 87 (2003)
through email and listserv announcements to FL the respondents, 88.3% reported expecting an A
instructors and language program directors or a B. An additional 10% expected a C, and only
throughout the United States and Canada 1.7% expected a D or an F. This distribution of
(roughly 240 direct email appeals plus several the self-selected sample is addressed in the Re-
listserv announcements). The appeal asked FL sults and Analysis and the Limitations sections of
instructors to complete the anonymous, Web- this article.
based questionnaire and to ask their students to Next, the students were asked to rate how mo-
do the same.4 No systematic attempt was made to tivated they felt they were to acquire the FL. Over
send letters to all U.S. universities and colleges, half of them, 56.6%, reported “extremely high”
but an attempt was made to send several letters to or “high” motivation, 34.4% reported “moderate”
all 50 states and to each English-speaking prov- motivation, and just 8.1% reported “weak” or “ex-
ince of Canada. Except through the listservs tremely weak” motivation.
(which obviously reached a larger audience), ap- The students also were asked about their in-
peals were made to 4-year colleges and universi- structors. Women comprised 73% of the stu-
ties only. dents’ instructors; 27% were men. Slightly less
At the beginning of each questionnaire, re- than half of the instructors, 48.2%, were native
spondents were informed that the questionnaire speakers of the FL; 37.5% were graduate student
was completely anonymous unless they chose to instructors; 34% were professors; 16.1% were lec-
identify themselves (by providing an email ad- turers; and for 12.3% of the instructors, students
dress). The respondents were asked to complete were unable to provide a rank.
the questionnaire alone and in one sitting. Finally, the students were asked about their
instructors’ approach to TL use. In this section of
the questionnaire, 59% of the students reported
RESULTS AND ANALYSIS
that their instructor regularly made expectations
Characteristics of the Sample: Students clear regarding TL use; 33% stated that their
instructors did so at the beginning of the term;
Between October and early December of 2000, only 8% reported that their instructors never
responses were received from an even 600 stu- made expectations clear. Similarly, 72.7% of the
dents.5 Of the respondents 59% were male, 41% students reported that their instructors spent
were female, 80.8% were in the 18- to 21-year-old class time working through or discussing commu-
age range, and 14.1% were in the 22- to 30-year- nicative strategies to facilitate TL use; 12.7% re-
old age range. Most of the students (89%) called ported that this sort of assistance was provided at
themselves native speakers of English; 15% the beginning of the term; 14.7% stated that their
claimed to be bilingual; 64.9% were in the first instructors never focused on these strategies.
year of FL instruction; and 35.1% were in second-
year courses. The student respondents came Characteristics of the Sample: Instructors
from 21 different states and three Canadian prov-
inces. Responses were received from 163 instructors
Most of the respondents were students of of first- or second-year FL classes. Of these in-
French (60.5%; n ⫽ 361), German (24.3%; n ⫽ structors, 68.9% were female; 31.1% were male.
145), and Spanish (11.9%; n ⫽ 71). For the re- The different age ranges were well represented
maining languages in the list (see Appendix A, among the instructors. The geographical distri-
item 11), very few students responded, with 4 bution of the instructors was greater than that
students checking “other.” among the students. They came from 31 different
In addition to providing demographic infor- U.S. states and 4 Canadian provinces. Among the
mation, the students were also asked to share instructors, as for the students, the three “most
information about their FL classes and their in- commonly taught” languages were most strongly
structors. I focus on the results that were used as represented, though the proportions were differ-
variables in the analysis of the data, as follows: ent: 44.2% of the instructors taught German;
expected grade; language learning motivation; 27% taught Spanish; and 17.2% taught French.
instructor’s gender, academic rank, and native Just over one third of the instructors (34.4%)
speaker status; instructor’s stated expectations re- claimed to be native speakers of the language
garding TL use, and frequency of strategies in- they were teaching.
struction regarding TL use. With regard to the academic rank of instruc-
The students were asked what grade they ex- tors, almost half of them (49.1%) were graduate
pected to receive in their current FL course. Of teaching assistants, 12.9% were “long term” lec-
Glenn S. Levine 349
turers, 11% were tenured faculty, 10.4% were Research Question 1
tenure-track faculty, and 9.8% were adjunct lec-
turers. When asked about their professional What do students and instructors believe goes on in
training, 88.2% of the instructors reported hav- the FL classroom with the TL and the L1?
ing received training in FL pedagogy. The years The questionnaire data for both students and
of teaching experience among them varied instructors clearly support the first hypothesis
greatly: 19.6% had been teaching less than 1 that the reported amounts of TL use differ ac-
year; 16% had taught for 1 to 2 years; 25% had cording to the constellation of interlocutors and
taught between 3 and 10 years; and 29.4% had communicative contexts. In assessing the amount
taught more than 10 years. Finally, a large ma- of the TL and the L1 used in the classroom in
jority (74.8%) of the instructors reported that order to reply to this research question and to its
they employ a communicative approach/meth- subquestions, the respondents were asked to esti-
odology. mate TL use by considering three interlocutor
constellations—(a) instructor with student(s),
(b) student(s) with instructor, (c) student(s) with
student(s)—and three communicative con-
Target Language Use and Target-Language-Use texts—(a) topic/theme-based activities (i.e., text-
Anxiety book activities), (b) discussion of grammar and
In the following sections I will report the fre- usage, and (c) discussion of tests, quizzes, other
quencies of estimated TL use overall among stu- assignments.
dents and instructors, TL use in specific con-
texts, student and instructor opinions about TL Research Question 1a
use, and TL-use anxiety among students. These
results are organized around the research ques- How does reported TL use differ for different inter-
tions posed earlier. While reading the responses locutors?
of the instructors and students, the reader Dealing with each of these rubrics in turn, Fig-
should keep in mind that in many instances, the ure 1 represents graphically student and instruc-
frequencies measure somewhat different things. tor estimates of amount of TL use by different
Whereas the students were often asked about sets of interlocutors (the frequencies repre-
their own personal experiences (e.g., How much sented by this chart are presented in Appendix
TL do you use?), the corresponding instructor C). For the first category, instructors with stu-
questionnaire item often asked instructors to es- dents, about 60% of students reported that their
timate how much TL their students used (as a instructors use the TL 80% to 100% of the time.
group). It is interesting to note that only about 44% of

FIGURE 1
Comparison of Student and Instructor Estimates of Target Language Use among Different Interlocutors
350 The Modern Language Journal 87 (2003)
the instructors claimed to use the TL with stu- 80% to 100% of the time. Similarly, 29.4% of the
dents 80% to 100% of the time. The most strik- instructors (n ⫽ 160) reported that their students
ing feature of this result is that whereas the fre- switched to English 80% to 100% of the time. An
quencies differed for each group, the pattern of additional 21.2% of the students reported that
responses was nearly identical for the students they switched to English 60% to 80% of the time.
and the instructors. In other words, there is an Correspondingly, 29.4% of the instructors re-
indication that in 40% to 60% of FL classes the ported that students switched to English 60% to
instructor used the TL 80% to 100% of the time. 80% of the time. In short, for communication
By and large, this result accords with the find- among the students, the unmarked code for “off-
ings of Guthrie (1984), Nzwanga (2000), and record” communication (Hancock, 1997) ap-
Macaro (2001). pears to be the L1 for a strong majority. It also
For the next category, when asked to estimate appears that the students and instructors concur
the amount of the TL used by students when in their estimations of this aspect of classroom
speaking with their instructors, the two groups verbal behavior.
reported similarly. Only about 17% of the stu-
dents reported that they used the TL 80% to Research Question 1b
100% of the time with their instructors; corre-
spondingly, only 15% of the instructors reported How does reported TL use differ in different commu-
that their students used the TL 80% to 100% of nicative contexts?
the time when speaking to them. For the final Research question 1b is related to question-
category, student to student communication, the naire items that differentiate among communica-
two groups differed more in their responses, yet tive contexts. Figure 2 represents graphically the
each group reported that less of the TL was used student and instructor estimations of TL use in
than in the other two categories. In sum, the three typical classroom communicative contexts
frequencies presented in Figure 1 indicate that (frequencies are listed in Appendix D).6 It is
the TL was used most by instructors when speak- noteworthy that 53.5% of the instructors esti-
ing to students, less by students when speaking mated that the TL was used 80% to 100% of the
with their instructors, and still less overall when time for topic-based activities (i.e., language text-
students spoke with other students. book activities), whereas only 28.4% of the stu-
With regard to learner–learner TL use, one dents made the same estimate. By and large,
additional result should be reported. When asked there was a gap in all three contexts between
to estimate the percentage of the time they student and instructor estimations, with instruc-
switched to English as soon as a partner or group tors estimating greater amounts of TL use overall.
activity was completed, 36.7% of the students (n Also noteworthy is the continuum from
⫽ 594) reported that they switched to English topic/theme-based activities, to discussion of

FIGURE 2
Student and Instructor Estimations of Target Language Use in Different Contexts
Glenn S. Levine 351
grammar and usage, to discussion of tests or discernable is that in the face of generally high
other assignments. The TL is used most for levels of reported TL-use anxiety, a strong major-
topic/theme-based communication, less overall ity of both students and instructors (approxi-
for communication about grammar, and less still mately 63% in each group) agreed or strongly
for communication about tests, quizzes, and as- agreed with the statement that it is a “rewarding
signments. The indication is that the L1-use con- and worthwhile challenge” to “have to use the FL
ditions advocated by Cook (2001) may reflect the to communicate (rather than fall back on En-
way many language classes are actually con- glish).”7 This result suggests that whereas stu-
ducted. dents may feel anxious about TL use, many ap-
pear to appreciate its importance in the L2
acquisition process.
Research Question 1c
Turning to the responses to items about TL-use
How do instructors and students perceive TL-use anxiety in particular contexts, Figure 4 presents
anxiety, particularly in different communicative con- some interesting patterns, the most intriguing of
texts? which is that in these questionnaire items, the
For research question 1c, there were three students and instructors revealed quite different
items on the questionnaire designed to elicit in- perceptions (for frequencies, see Appendix F).
structor and student perceptions and estimations By and large, a small proportion of students and
of TL-use anxiety in general and in different com- instructors (23.5% and 17.4%, respectively)
municative contexts. Figure 3 represents the re- agreed or strongly agreed that students feel anx-
sponse frequencies for these items (frequencies ious during topic/theme-based activities. Yet, for
listed in Appendix E). Looking only at the per- the items on anxiety felt during TL communica-
centage of each group that agreed or strongly tion about grammar, tests, and administrative in-
agreed with the statement of a given item, there formation, the instructors perceived a higher
are three very interesting patterns. First, a sizable level of anxiety in students than the students
minority of the students (around 40% overall) themselves reported feeling. Even with topic/
agreed with the statement that using the TL theme-based activities, a greater percentage of
made them feel anxious. Second, it appears that instructors than students appeared to perceive
instructors may perceive higher anxiety about TL anxiety in the students. Still, these results should
use in students than students themselves report be regarded with caution, given that each student
feeling. This pattern can be seen in the higher was being asked how he or she personally experi-
proportion of instructors who agreed or strongly enced TL-use anxiety, whereas the instructors
agreed with the first two statements (59.9% and were asked to estimate TL-use anxiety among
45.7% respectively). The third interesting pattern their students as a group.

FIGURE 3
Student and Instructor Agreement with Statements about Target-Language-Use Anxiety in General
352 The Modern Language Journal 87 (2003)

FIGURE 4
Student and Instructor Agreement with Statements about Target-Language-Use Anxiety in Particular
Communicative Contexts

Research Question 2 mately 15% of the variance. Correspondingly, the


same inverse relationship obtained for instruc-
How do reported TL use and TL-use anxiety relate to tors, which accounted for approximately 17.5%
each other and to personal or classroom variables? of the variance.
The most interesting result of the analysis is At the outset, there were many additional vari-
that the second hypothesis—the reported ables that were expected to correlate with TLU.
amount of TL use would correlate positively with Some of these variables were analyzed as interval
TL-use anxiety—was not supported. Rather, a sig- variables and some as nominal variables. “Motiva-
nificant negative relationship was found between tion” and “strategies instruction” were treated as
these two variables. In order to test H2, a factor interval variables. For example, one would expect
analysis8 on all numeric questionnaire items in low levels of motivation to be associated with low
both student and instructor data sets was carried TL use and vice versa. Low, or no, strategies in-
out.9 Two of the factors extracted represented the struction in TL use can be assumed to be associ-
a priori constructs TLU and TLA and were used ated with higher anxiety among students.
in the present analysis. Corresponding factors “Year of instruction,” “bilingual background,”
were extracted from both the student and the and “expected grade” were treated as nominal
instructor data sets.10 These factors included all variables. The first two are of course dichoto-
the items related to TL use, including use with mous, but “expected grade” was treated as cate-
different interlocutors and in different contexts. gorical because, although it turned out not to be
Because the exact items in each factor differed the case, it was expected that both the best and
slightly for each group, no inferential tests could the worst students would give similar responses
be performed that would directly compare the about TL-use anxiety (because both might be un-
two groups.11 The mean scores of questionnaire concerned about TL use, but for opposite rea-
items from each factor were employed as indexes sons). Thus, any significant effect would have
in analyzing relationships between the two fac- been missed by a linear analysis. Table 2 presents
tors. the means of each of these nominal variables.
Simple least squares regression analyses of the These results confirm intuitive expectations that
relationship between TLU and TLA were carried TL use tends to be higher, and anxiety lower, for
out for each group. The results of these analyses (a) students in the second year of instruction,
are presented in Table 1. For the students, a sig- (b) students with a bilingual background, and
nificant inverse relationship was found between (c) students who expect a higher grade.
the constructs, which accounted for approxi-
Glenn S. Levine 353
TABLE 1
Least Squares Regression Analysis of Target Language Use and Target-Language-Use-Anxiety: Students and
Instructors

Target-Language-Use Anxiety
Parameter
Estimate SE t p R2
Target Language Use Students (n ⫽ 590) –.36 .035 –10.16 ⬍ .0001 .149
Intercept 3.84 36.66 ⬍ .0001
Instructors (n ⫽ 161) –.30 .053 –5.86 ⬍ .0001 .177
Intercept 4.23 24.31 ⬍ .0001

Among the additional variables considered, or any other personal or classroom variables and
significant positive relationships (␣ ⬍ .05) with the two constructs. These other variables in-
TLU were found for year of instruction (first or cluded instructor’s gender, age group, native-
second year), expected grade, reported level of speaker status, formal pedagogical training, years
motivation, and frequency of strategies instruc- of experience, academic rank, stated expecta-
tion about TL use (see Table 3). Likewise, several tions about TL use, and frequency of strategies
variables were expected to correlate with TLA, instruction about TL use.
and again, we found this to be the case: five vari- To summarize, several personal and demo-
ables demonstrated a significant relationship graphic variables were found to correlate signifi-
with TLA (the four variables mentioned above cantly with both TLU and TLA. The next issue to
and the variable student bilingual vs. monolin- be addressed was whether any of the variables
gual background). It is important to note in these that were found to stand in a significant relation-
results, however, that only expected grade and ship to either of the two factors did a better job at
reported motivation accounted for more than a predicting TLA than the factor TLU itself. When
nominal amount of the variance. TL-use anxiety was modeled using the student-
For all other student personal and classroom data variables listed in Table 3, the main effects of
variables tested, no significant relationships were each of the variables dissipated when TLU was
found. In the student data, these variables in- introduced as a predictor of TLA. The same oc-
cluded student gender, age group, university size, curred with the instructor data set. In short, the
instructor’s gender, instructor’s native speaker strongest predictor of students’ TL-use anxiety
status, instructor’s rank, and instructor’s stated was the amount of TL use itself.
expectations about TL use.
The same examination of the instructor data DISCUSSION
revealed a significant relationship only between
year of instruction and TLU (p ⫽ .007); no other The main goals of this study were first to take
significant relationship was found between these initial steps toward a generalizable, descriptive

TABLE 2
Means of the Nominal Variables Year of Instruction, Bilingual Background, and Expected Grade

Mean Target Mean Target-


Language Use Language-Use Anxiety
Year of Instruction
First Year 2.67 2.91
Second Year 2.99 2.71
Bilingual Background
No 2.77 2.88
Yes 2.87 2.59
Expected Grade
A 2.97 2.52
B 2.69 3.05
C or Lower 2.19 3.67
354 The Modern Language Journal 87 (2003)
TABLE 3
Variables that Correlated Significantly with Either Target Language Use or Target-Language-Use Anxiety:
Students Only

Target-Language-Use Anxiety Target Language Use


Nominal Independent Variables n F eta p n F eta p
Year of Instruction (df ⫽ 1) 584 6.6 .011 .011 591 15.5 .026 ⬍.0001
Expected Grade (df ⫽ 2) 581 61.9 .18 ⬍.0001 588 2.85 .07 ⬍.0001
Bilingual Background (df ⫽ 1) 582 8.09 .014 .005 589 0.84 .001 .36
Target-Language-Use Anxiety Target Language Use
Interval-Level Independent Variables n F R2 p n F R2 p
Motivation 584 123.63 .17 ⬍.0001 592 71.1 .11 ⬍.0001
1 ⫽ high 5 ⫽ low
Instructor Strategies Instruction 579 21.11 .035 ⬍.0001 586 8.34 .014 .004
0 ⫽ never
1 ⫽ at beginning
2 ⫽ regularly

model of TL and L1 use, one that included all Although this continuum of TL to L1 use
classroom members and began to represent the across communicative contexts is not surprising,
communicative complexity of the FL class, and it offers an interesting picture of the FL class that
second, to identify covariates of TL use and TL- deserved closer scrutiny in light of the answer
use anxiety. The first research question required uncovered to the second research question,
a description of what students and instructors which required exploration of covariates of TL
believed goes on in the FL class with the TL and use and TL-use anxiety. Several personal and
the L1. In accord with Guthrie (1984), Macaro classroom variables in the student data were
(2001), and Nzwanga (2000), it appears that most found to stand in a significant relationship with
instructors used the TL a great deal of the time the factors TLU and TLA.12 Year of instruction,
overall. This study found that the students used expected grade, reported level of motivation, and
the TL less than the instructors did overall but the frequency of strategies instruction about TL
that they used it more when speaking to their use all correlated with how much TL was used in
instructors than when speaking with other stu- the classroom overall. In other words, it is not
dents. Where students and instructors differed in surprising that the tendency to report higher lev-
their estimations, though, was in their percep- els of TL use overall was found among students
tions of student TL-use anxiety. Although the pat- who (a) were in the second year of instruction,
terns of response for both groups were similar, (b) expected a higher grade, (c) expressed
the data point to the possibility that instructors higher motivation to learn the FL, and (d) had
may perceive higher levels of TL-use anxiety instructors who engaged in more frequent strate-
among students (in general) than students them- gies instruction about TL use. Variables that pre-
selves report. This finding holds in particular for dicted TL-use anxiety included these same four
TL-use anxiety in different communicative con- variables, plus student bilingual background.
texts. This finding alone may lead us to call into This finding suggests that students who come
question the prevailing norm, as described by from bi- or multilingual backgrounds may tend to
Cook (2001), whereby the L1 is used for discuss- feel less anxious about TL use than students from
ing grammar, class assignments, course policies, monolingual backgrounds. This characteristic ac-
and the like. cords with intuitive expectations. Still, the fact
In analyzing reported amounts of TL use in that four of these five variables predict both TLU
different communicative contexts, we learned and TLA points toward the important relation-
from both groups that the TL appears to be the ship between these two factors themselves.
unmarked code for most students and instructors And indeed, the strongest predictor of TL-use
for textbook-type activities, whereas the tendency anxiety appears to be the amount of TL use itself
toward L1 use is greater for communicating in both data sets. The second hypothesis, which
about grammar or usage and tests or assignments. asserted a positive relationship between TL use
Glenn S. Levine 355
and TL-use anxiety, was rejected because a signifi- interaction. Therefore, any curricular decisions
cant negative relationship was identified between based on this study should be made with caution.
these two constructs (factors). This finding indi- A second area of concern that may limit the
cates that students who reported higher TL use in strength of these findings relates to the repre-
their FL classes tended to report lower levels of sentativeness of the sample. Clearly, the self-se-
anxiety about TL use. Correspondingly, instruc- lected respondents were those who had both ac-
tors who reported higher levels of TL use in their cess to the Internet and the willingness to
classes tended to perceive lower levels of TL-use complete the form. This self-selection excluded
anxiety in their students. The important implica- by default large numbers of FL students and in-
tion of this finding is that greater TL use may not structors who either did not make regular use of
translate into greater anxiety for many learners the Internet, or simply did not wish to participate
and that many students feel comfortable with in the study. In addition, the self-selected student
more TL use when that is what they are used to. sample obviously was made up of more highly-
From these findings, three pedagogical ten- motivated, higher performing students than a
ets—I refrain from committing to calling them random sample would have involved. Although
principles at this point—for TL and L1 use in the the significant relationships found between ex-
classroom come into focus. The first I call the pected grade and motivation and the two factors
Optimal TL Use Tenet (adopting Macaro’s [2001] TLU and TLA indicate that similar relationships
term), which asks instructors to accept the notion would have obtained if the sample had been
that the L1 apparently serves numerous functions more representative, some caution is indicated by
in the FL class, and that denying a role to the L1 the skewed nature of the sample.
would appear to be a futile endeavor. In short, as A third potential limitation is that there may be
advanced by Blyth (1995), instructors should ac- confounding variables unmeasured in the ques-
cept the idea that the FL class is a multilingual tionnaires or unaccounted for in the analyses,
environment. variables that may predict student TL-use anxiety
Resolving not to deny the L1 a role, then, leads better than TL use itself. For example, whereas
to a second tenet, which I provisionally term the students’ expected grades correlate with how
Marked L1 Tenet. Using the L1 simply for the sake much TL they or their classmates report using, it
of reducing anxiety or increasing efficiency, as may be that if students in a given class are doing
Cook suggested, is likely untenable, according to well overall, then the instructor and students may
this study’s findings. Rather, with regard to cur- feel more comfortable using more TL. The ques-
rent practice in many FL classes, the avid use of tionnaires contained no means of teasing out this
the L1 as the unmarked code is what may engen- level of detail about the students’ FL classes. Re-
der TL-use anxiety among many students. In- lated to this point is also the issue of minimal
structors might rather strive to create situations information about respondents in this anony-
in the classroom in which the L1 could serve mous, Internet survey format. For instance, the
meaningful pedagogical functions (Macaro, questionnaires did not gather information about
2001) yet remain, relative to the TL, a marked personality traits or learning styles. And although
code. This sort of approach would obviate the we have self-reports about general motivation and
outright stigmatization of L1 use, which may be a expected grade, the instrument also did not allow
source of TL-use anxiety. for the collection of actual FL performance mea-
A third tenet, which I call the Collaborate Lan- sures. Last, with regard to the questionnaire, I also
guage Use Tenet, delegates to students an active had no way to ensure the honesty of the respon-
role in managing TL and L1 use in the FL class- dents. These limitations are intrinsic to the anony-
room. Instructors can seek to formalize the rela- mous Internet questionnaire, yet I believe that the
tionship between the two codes, in order to “cre- advantages of the methodology used still out-
ate,” in essence, bilingual norms that tend to weigh these potentially confounding factors, be-
develop organically in multilingual environments cause in the end, it was both fruitful and interest-
outside the classroom. ing to collect a large amount of data from students
and instructors at literally dozens of universities
LIMITATIONS throughout the United States and Canada.

Several limitations of the present study should CONCLUSION


be kept in mind. In the broadest sense, the find-
ings are based on respondents’ perceptions and This study has sought to contribute to a de-
beliefs and not on samples of actual classroom scriptive model of TL and L1 use in the FL class-
356 The Modern Language Journal 87 (2003)
room. The results indicate that despite the pre- project. This article is dedicated to the memory of Ag-
vailing “monolingual principle” in U.S. FL nes Engels and Ben Nelson.
classes, both the TL and the L1 appear to serve
important functions. As pointed out by Macaro
NOTES
(2001), our task as instructors is to identify effec-
tive pedagogical principles that both acknow-
1 The term TL as used in this article, as in most litera-
ledge and support the classroom as the multilin-
gual environment that it is. To that end, a ture on the topic, refers to oral production in the for-
number of important issues are raised by the find- eign language classroom, by both the instructor and the
students. When relevant, though, I will indicate whether
ings that need to be explored and understood
it is the instructor or students being referred to in dis-
more fully. First and foremost, in light of the
cussing TL use. With regard to the use of the conven-
significant negative relationship between TL use tional L1, the term is used in this text in opposition to
and anxiety, a study would be called for that L2, target language (TL) or foreign language (FL). It is
would compare directly students’ and instructors’ understood that in U.S. foreign language classes L1
perceptions and attitudes with actual observation means English. At the same time, no assumption is
of classroom verbal behavior. Also, the reasons made that English is the L1 of all students in FL classes.
why instructors choose to use the L1 and allow The term is used for the sake of simplicity of expression,
students to do so deserves further investigation, in order to avoid more cumbersome phrasing. A more
along the lines of Macaro’s (2001) study but for a appropriate term might be (societal) “dominant lan-
guage” or DL, but it is not my intention to initiate new
greater range of language learning situations.13
terminological conventions in the present article. In
Focusing as well on the student as this study has
addition, the terms TL, FL, and L2 will be used fairly
done, a study also would be welcome that would interchangeably here, though TL and FL generally are
probe more thoroughly students’ attitudes, anxi- used in contexts in which classroom communication is
eties, and beliefs about TL and L1 use. Along described or discussed, whereas L2 is most often used
these lines, the points of similarity and difference when referring to the phenomenon of second language
between student and instructor responses to acquisition.
2 The student form was administered to 6 students,
questionnaire items about student TL-use anxiety
deserve more detailed investigation. Finally, the the instructor form to 2 graduate teaching assistants in
findings of this study underscore the need for my own German program. Each person completed the
questionnaire on the Internet and then engaged in a 10-
concrete, theoretically motivated guidelines,
to 15-minute conversation with the researcher. Each was
based perhaps on the three pedagogical tenets
asked questions about the form and its items, such as
offered here, for FL classroom codeswitching, the following: What items were not clear or confusing?
guidelines that indicate which sorts of code- What items were disturbing or annoying? Was the
switching behaviors facilitate L2 acquisition and length of the questionnaire acceptable? Were the direc-
which behaviors undermine it. tions clear? and so forth.
3 Details of the factor analysis used to identify the

items associated with each construct are presented in


the Results and Analysis section. In the student ques-
ACKNOWLEDGMENTS
tionnaire data, TLU had a Cronbach’s alpha coefficient
of .86, while TLA had an alpha of .85. In the instructor
I am extremely grateful to the Humanities Center at questionnaire data, TLU had a Cronbach’s alpha coeffi-
the University of California, Irvine for its generous sup- cient of .92, and TLA had an alpha of .78.
port of each stage of the project. I would like to thank 4 Given that the appeals included personal acquain-

Dr. Chris Hanks of the UC Irvine Center for Statistical tances of the researcher, a dummy variable regression
Consulting for his expertise, sound judgment, patience, was conducted in order to determine whether response
and dedication to the study, Dwayne Pack for his assis- patterns were different for acquaintances of the re-
tance in creating the Internet-based questionnaire, and searcher and their students. No effects were found.
Karen Lowe for her efficient administrative assistance. I 5 In reporting frequencies in terms of percentages in

offer special gratitude to Lourdes Ortega for her careful the following, note that n is slightly variable from item
reading and critical comments on an earlier draft of this to item, because most questions did not receive a 100%
paper. I would like to thank the four anonymous review- response.
ers for their extensive comments. I also extend my 6 Questionnaire items relating to TL use outside of

thanks to the FL instructors and students who partici- class are not included in the present analysis.
pated in the study. Finally, I would like to thank Caroline 7 Student form item 59; instructor form item 54.

Grace, Ursula Levine, Grit Liebscher, Erwin Tschirner, 8 For a concise description of factor analysis as a tool

and the Arabic teaching staff at the Defense Language for analyzing questionnaire data, see Bacon and Finne-
Institute for their encouragement and helpful com- mann (1990), p. 461.
ments and feedback during the development of the 9 PROC FACTOR in SAS version 8.2 was used. The
Glenn S. Levine 357
method was Principle with Varimax rotation. The eigen- Lee (Eds.), Second language acquisition/foreign lan-
values were as follows: student TLU 8.2; instructor TLU guage learning (pp. 153–169). Clevedon, UK: Mul-
6.9; student TLA 2.99; instructor TLA 2.26. tilingual Matters.
10 Only respondents who answered at least two thirds Castellotti, V. (2001). La langue maternelle en classe de
of the questions in both factors were employed in analy- langue étrangère [The native language in a foreign
ses using these factors. language classroom]. Paris: CLE International.
11 Questionnaire items that loaded on each factor in Cook, V. (1999). Going beyond the native speaker in
each group are as follows: TLU students 28–33; TLU language teaching. TESOL Quarterly, 33, 185–209.
instructors 21–29; TLA students 58–63; TLA instructors Cook, V. (2001). Using the first language in the class-
52–58. The text of these items can be found in Appen- room. Canadian Modern Language Review, 57,
dixes A and B. 402–423.
12 As yet, no explanation can be offered for why most
Duff, P. A., & Polio, C. G. (1990). How much foreign
of the instructor personal and classroom variables did language is there in the foreign language class-
not correlate significantly with the two factors. It may be room? Modern Language Journal, 74, 154–166.
that instructors’ perceptions of TL use are simply not Ellis, R. (1999). Learning a second language through inter-
determined by years of experience, pedagogical train- action. Amsterdam: John Benjamins.
ing, or their native-speaker status, or it may be that a Franklin, C. E. M. (1990). Teaching in the target lan-
different instrument would be needed to gauge these guage: Problems and prospects. Language Learning
relationships. Journal, 2, 20–24.
13 At the very least, a replication of the present study
Gass, S. M. (1997). Input, interaction, and the second lan-
involving languages other than French, German, and guage learner. Mahwah, NJ: Erlbaum.
Spanish surely would contribute to a more complex and Guthrie, E. M. L. (1984). Six cases in classroom commu-
accurate picture of student and instructor attitudes and nication: A study of teacher discourse in the for-
perceptions. eign language classroom. In J. P. Lantolf & A.
Labarca (Eds.), Research in second language learning:
Focus on the classroom (pp. 173–194). Norwood, NJ:
REFERENCES Ablex.
Hancock, M. (1997). Behind classroom code switching:
Antón, M., & DiCamilla, F. J. (1999). Socio-cognitive Layering and language choice in L2 learner inter-
functions of L1 collaborative interaction in the L2 action. TESOL Quarterly, 31, 217–235.
classroom. Modern Language Journal, 83, 233–247. Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). For-
Atkinson, D. (1987). The mother tongue in the class- eign language classroom anxiety. Modern Language
room: A neglected resource. English Language Journal, 70, 125–132.
Teaching Journal, 41, 241–247. Johnson, K. E. (1995). Understanding communication in
Atkinson, D. (1993). Teaching in the target language: A second language classrooms. Cambridge: Cambridge
problem in the current orthodoxy. Language University Press.
Learning Journal, 8, 2–5. Krashen, S. (1982). Principles and practice in second lan-
Auerbach, E. R. (1993). Reexamining English only in guage acquisition. New York: Pergamon Press.
the ESL classroom. TESOL Quarterly, 27, 9–32. Krashen, S., & Terrell, T. D. (1983). The natural approach:
Bacon, S. M., & Finnemann, M. D. (1990). A study of the Language acquisition in the classroom. Oxford, UK:
attitudes, motives, and strategies of university for- Pergamon.
eign language students and their disposition to Macaro, E. (2001). Analyzing student teachers’ code-
authentic oral and written input. Modern Language switching in foreign language classrooms: Theo-
Journal, 74, 459–473. ries and decision making. Modern Language Jour-
Bailey, P., Onwuegbuzie, A. J., & Daley, C. E. (2000). nal, 85, 531–548.
Correlates of anxiety at three stages of the foreign MacIntyre, P. D. (1995). How does anxiety affect second
language learning process. Journal of Language and language learning? A reply to Sparks and Gan-
Social Psychology, 19, 474–490. schow. Modern Language Journal, 79, 90–99.
Belz, J. A. (2003). Identity, deficiency, and first lan- MacIntyre, P. D., & Charos, C. (1996). Personality, atti-
guage use in foreign language education. In C. tudes, and affect as predictors of second language
Blyth (Ed.), The sociolinguistics of foreign-language communication. Journal of Language and Social Psy-
classrooms: Contributions of the native, the near-native, chology, 15, 3–26.
and the non-native speaker (pp. 209–248). Boston: MacIntyre, P. D., Clément, R., Dörnyei, Z., & Noels,
Heinle. K. A. (1998). Conceptualizing willingness to
Blyth, C. (1995). Redefining the boundaries of language communicate in a L2: A situational model of L2
use: The foreign language classroom as a multilin- confidence and affiliation. Modern Language Jour-
gual speech community. In C. Kramsch (Ed.), Re- nal, 82, 545–562.
defining the boundaries of language study (pp. Nation, P. (1997). L1 and L2 use in the classroom: A
145–183). Boston: Heinle. systematic approach. TESL Reporter, 30, 19–27.
Brooks, F. B. (1990). Foreign language learning: A so- Nizegorodcew, A. (1996). The function of the native
cial interaction perspective. In B. VanPatten & J. F. language in foreign language teacher-learner in-
358 The Modern Language Journal 87 (2003)
teraction. Jyvaskyla Cross-Language Studies, 12, Scovel, T. (1991). The effect of affect on foreign lan-
209–214. guage learning: A review of the anxiety research.
Nzwanga, M. A. (2000). A study of French-English code- In E. K. Horwitz & D. J. Young (Eds.), Language
switching in a foreign language college teaching environ- anxiety: From theory and research to classroom implica-
ment. Unpublished doctoral dissertation, The tions (pp. 15–23). Englewood Cliffs, NJ: Prentice
Ohio State University, Columbus. Hall.
Polio, C. G., & Duff, P. A. (1994). Teachers’ language Turnbull, M., & Arnett, A. (2002). Teachers’ uses of the
use in university foreign language classrooms: A target and first languages in second and foreign
qualitative analysis of English and target language language classrooms. Annual Review of Applied Lin-
alternation. Modern Language Journal, 78, 313–326. guistics, 22, 204–218.
Rifkin, B. (2000). Revisiting beliefs about foreign lan- Young, D. J. (1990). An investigation of students’ per-
guage learning. Foreign Language Annals, 33, spectives on anxiety and speaking. Foreign Lan-
394–420. guage Annals, 23, 539–553.

APPENDIX A
Student Questionnaire Items

Note. A formatted version of this questionnaire can be found at http://www.humanities.uci.edu/german/


GLEVINE/Student_Form.html

Part 1. Biographical Information


1. Would you be willing to be contacted by the researchers in order to provide more detailed information? If so,
please provide your email address: (Optional!)
2. State/Province of Residence (2-letter abbreviation, please)
3. Type of College or University
4. Gender
5. Age group
6. Primary area of university study
7. What is your native language or language group? (Note: If you have more than one native language, please mark
the language that you speak most proficiently)
8. Did you grow up in a bi- or trilingual household?
9. Did you grow up primarily in an English-speaking environment or country?
10. Are you proficient now in a FL other than the one you are currently studying?
11. What FL are you studying (If you are currently studying more than one FL, base all your answers in the
remainder of this questionnaire on one of the FLs)?
12. What year of the language are you currently taking?
13. What grade did you earn in the FL course you most recently completed?
14. What grade do you anticipate earning in the FL course you are currently completing?
15. Is your current FL instructor a native speaker of the FL?
16. What gender is your current FL instructor?
17. What rank/status is your current FL instructor?

Part 2. FL Use in Previous FL Classes You Have Taken (if applicable; H.S. or College levels)
18. If you have been involved in a class that has been conducted entirely (or almost entirely) in the FL, what term
(semester/quarter) were you first exposed to this situation?

Part 3. Use of the FL in Your Current FL Class


All responses you provide in the remainder of this questionnaire pertain only to the FL course you are currently
taking.
19. I would rate my level of overall motivation to learn the FL as [level].
—-extremely high—-high—-moderate—-weak—-extremely weak
20. Please mark the best/primary reason you are learning this FL:
university requirement only
university requirement and personal interest in the language/culture
personal interest in the language only
it is my family heritage language
don’t know
21. Do you intend to pursue advanced study of the FL and/or study or work abroad?
Glenn S. Levine 359
22. I regularly seek out opportunities to use or hear the FL outside of my FL class (e.g., FL coffee hour, friendships
with native speakers, FL TV or radio).
23. My instructor has made expectations regarding the use of the FL in the classroom explicit by discussing them
with us.
24. My instructor has spent class time working through or discussing communicative strategies that will help
students communicate in the FL.

For Questions 25–38 choose from: 0%–20%; 20%–40%; 40%–60%; 60%–80%; 80%–100%
25. My FL instructor uses the FL to communicate with students about __% of the time in the classroom (select one;
this includes time spent on activities and time spent in between activities).
26. My fellow students use the FL to communicate with the instructor about __% of the time in the classroom (select
one; this includes time spent on activities and time spent in between activities).
27. My fellow students use the FL to communicate with other students about __% of the time in the classroom (select
one; this includes time spent on activities and time spent in between activities).
28. I use the FL to communicate with my instructor about __% of the time in the classroom (select one; this includes
time spent on activities and time spent in between activities).
29. I use the FL to communicate with my fellow students about __% of the time in the classroom (select one; this
includes time spent on activities and time spent in between activities).
30. I use the FL to communicate within topic-based/thematic activities about __% of the time we spend on those
activities.
31. I use the FL to communicate about grammar and usage about __% of the time we spend on discussing or
working on these.
32. I use the FL to communicate about tests, quizzes, and other assignments about __% of the time we spend
discussing these.
33. I use the FL to communicate with my instructor outside of class time (e.g., office hours, in the hall, before or
after class) about __% of the time.
34. I use the FL to communicate with fellow FL students outside of class time (e.g., in the hall, before or after class,
over coffee) about __% of the time.
35. While working with a partner or group in my FL class, I switch to English as soon as we are through with a
particular activity about __% of the time.
36. I understand what my instructor is saying in the FL about __% of the time.
37. When I do not understand what my instructor is saying in the FL, I request in the FL that she or he repeats or
clarifies about __% of the time.
38. When I do not understand what a fellow student is saying in the FL, I request in the FL that she or he repeats
or clarifies about __% of the time.

Part 4. Your Opinions about FL Use in the Classroom


Mark the degree to which you agree or disagree with the following statements (1 ⫽ strongly disagree; 5 ⫽ strongly agree):
39. I believe that the more FL students use the FL in the classroom, the better they will be at communicating in the
FL.
40. I believe that I must use the FL a great deal in the classroom in order to master the language.
41. I believe that there are no situations in which the first language (i.e., English) should be used in the classroom
(i.e., total immersion in FL classes is best).
42. I believe that the instructor and students should use only the FL to learn about grammar and usage of the FL.
43. I believe that the instructor should use only the FL when giving directions for activities.
44. I believe that the instructor and students should use only the FL to discuss course policies, attendance, and other
administrative information.
45. I believe that, regardless of how much FL students choose to use, the instructor should use the FL at all times
in the classroom.
46. I believe that the FL students should use only the FL the entire time they are in the classroom with both the
instructor and fellow students, even when not working on a specific activity.

Part 5. Your Course Grade and Classroom Communication (in the FL or in English)
Rate the importance of each type of communication in terms of its role in students’ success in the course (i.e., a good
grade).
Mark the degree of importance of each type of communication on a scale of 1 ⫽ least important to 5 ⫽ most important.
Note: each item may be given its own ranking; there can be more than one 5, 4, 3, etc.
47. FL topics (family, weather, FL culture, literature, study abroad, sports, hobbies, daily routines, etc.)
48. FL grammar and usage (verb conjugations, word order, agreement, idioms, vocabulary, etc.)
49. Details about tests, quizzes, and other assignments (how much will be covered, format of test sections, etc.)
360 The Modern Language Journal 87 (2003)
50. Instructions for classroom activities
51. Administrative information (office hours, upcoming events, important dates, etc.)

Part 6. FL Acquisition and Classroom Communication (in the FL or in English)


Rate the importance of each type of communication in terms of its role in students’ successful acquisition of the FL
(i.e., regardless of course grade earned).
Mark the degree of importance of each type of communication on a scale of 1 (least important) to 5 (most important).
Note: each item may be given its own ranking; there can be more than one 5, 4, 3, etc.
52. FL topics (family, weather, FL culture, literature, study abroad, sports, hobbies, daily routines, etc.)
53. FL grammar and usage (verb conjugations, word order, agreement, idioms, vocabulary, etc.)
54. Details about tests, quizzes, and other assignments (how much will be covered, format of test sections, etc.)
55. Instructions for classroom activities
56. Administrative information (office hours, upcoming events, important dates, etc.)

Part 7. Your Feelings about FL Use in the Classroom


Mark the degree to which you agree or disagree with the following statements (1 ⫽ strongly disagree; 5 ⫽ strongly agree):
57. I generally feel anxious using the FL.
58. I generally find trying to communicate in FL frustrating.
59. I view it as a rewarding or worthwhile challenge when I have to use the FL to communicate (rather than fall back
on English).
60. I feel uncomfortable or anxious speaking the FL during activities about FL topics (family, weather, FL culture,
literature, study abroad, sports, hobbies, daily routines, etc.).
61. I feel uncomfortable or anxious using the FL when working on, discussing, or asking questions about grammar
and usage (e.g., verb conjugations, word order, agreement, idioms, vocabulary).
62. I feel uncomfortable or anxious using the FL when working on, discussing, or asking questions about tests,
quizzes, and other assignments (how much will be covered, format of test sections, etc.)
63. I feel uncomfortable or anxious using the FL when discussing or asking questions about administrative informa-
tion (office hours, upcoming events, important dates, etc.).

APPENDIX B
Instructor Questionnaire

Note. A formatted version of this questionnaire can be found at http://www.humanities.uci.edu/german/


GLEVINE/Instructor_Form.html

Part 1. Biographical Information


1. Would you be willing to be contacted by the researchers in order to provide more detailed information? If so,
please provide your e-mail address: (Optional!)
2. State/Province of Residence (2-letter abbreviation, please)
3. Type of College or University
4. Gender
5. Age group
6. Primary area of university study (of your most advanced degree)
7. What is your native language or language group? (Note: If you are native in more than one language, please
mark the language that you speak most proficiently)
8. Did you grow up in a bi- or trilingual household?
9. Did you grow up primarily in an English-speaking environment or country?
10. What is your status as a FL instructor?
11. What FL are you currently teaching (If you are currently teaching more than one FL, base all your answers in
the remainder of this questionnaire on one of the FLs)?
12. What year of the language are you currently teaching?
13. Are you a native speaker of the language you are currently teaching?
14. How many years have you been teaching the FL?
15. What lower-division FL levels have you taught?
16. Have you received formal training in FL teaching/pedagogy?
If yes to 16, then please answer 16a–c
16a. How much formal training have you had in FL teaching (workshops, courses, internships, etc.)?
Glenn S. Levine 361
16b. How would you characterize your most important formal training as a FL instructor? (if you received no formal
training, please leave blank)
16c. How would you characterize the primary FL approach/methodology in which you were trained? (if you received
no formal training, please leave blank)
17. Have you taught your FL using only the FL at all times (as opposed to some mixture of the FL and English)?
If yes to 17, then please answer 17a–b.
17a. When you engaged in the exclusive use of the FL in your class, was the practice dictated/prescribed to you by a
supervisor?
17b. When you engaged in the exclusive use of the FL in your class, did you find the experience productive and
enjoyable?
18. When you engaged in the exclusive use of the FL in your class, did your students generally appear to find the
experience productive and enjoyable?

Part 2. Use of the FL in Your Current FL Class


All responses you provide in the remainder of this questionnaire pertain only to the FL course you are currently
teaching.
For the following items (19–20), please choose: at the beginning of the term; regularly throughout the term; never
19. I have made my expectations regarding the use of the FL in the classroom explicit by discussing them with
students.
20. I have spent class time working through or discussing communicative strategies that will help students commu-
nicate in the FL.

For the following items (21–33), please give an estimate, choose from: 0%–20%; 20%–40%; 40%–60%; 60%–80%;
80%–100%
21. I use the FL to communicate with my students about __% of the time in the classroom (select one; this includes
time spent on activities and time spent in between activities).
22. My students use the FL to communicate with me about __% of the time in the classroom (select one; this
includes time spent on activities and time spent in between activities).
23. My students use the FL to communicate with each other about __% of the time in the classroom (select one;
this includes time spent on activities and time spent in between activities).
24. I use the FL to communicate with students within topic-based/thematic activities about __% of the time we
spend on those activities.
25. I use the FL to give directions for activities exclusively (i.e., no translation) in the FL about __% of the time.
26. I use the FL to communicate with students about grammar and usage about __% of the time we spend on
discussing or working on these.
27. I use the FL to communicate with students about tests, quizzes, and other assignments about __% of the time
we spend discussing these.
28. I use the FL to communicate with students about administrative information (course policies, announcements,
deadlines, etc.) about __% of the time we spend discussing these.
29. I use the FL to communicate with students outside of class time (e.g., office hours, in the hall, before or after
class, etc.) about __% of the time.
30. While students are working with partners or groups in my FL class, they switch to English as soon as they are
through with a particular activity about __% of the time.
31. My students understand what I am saying in the FL about __% of the time.
32. When my students do not understand what I am saying in the FL, they request that I repeat or clarify about __%
of the time.
33. When one of my students does not understand what a fellow student is saying in the FL, she or he requests that
repetition or clarification about __% of the time.

Part 3. Your Opinions About FL Use in the Classroom


Mark the degree to which you agree or disagree with the following statements (1 ⫽ strongly disagree; 5 ⫽ strongly
agree)
34. I believe that the more FL students use in the classroom, the better they will be at communicating in the FL.
35. I believe that in order to really master/acquire the FL, students must use the FL a great deal in the classroom.
36. I believe that there are no situations in which the first language (i.e., English) should be used in the classroom
(i.e., I believe that total immersion in FL classes is best).
37. I believe that only the FL should be used to learn about grammar and usage of the FL.
38. I believe that only the FL should be used to discuss tests, quizzes, and other assignments.
39. I believe that only the FL should be used to discuss course policies, attendance, and other administrative
information.
362 The Modern Language Journal 87 (2003)
40. I believe that, regardless of how much FL students choose to use, the instructor should use the FL at all times
in the classroom.
41. I believe that the FL students should use only the FL the entire time they are in the classroom with both the
instructor and fellow students, even when not working on a specific activity.

Part 4. Students’ Course Grades and Classroom Communication (in the FL or in English)
Rate the importance of each type of communication in terms of its role in students’ success in the course (i.e., a good
grade).
Mark the degree of importance of each type of communication on a scale of 1 (least important) to 5 (most important).
Note: each item may be given its own ranking; there can be more than one 5, 4, 3, etc.
42. FL topics (family, weather, FL culture, literature, study abroad, sports, hobbies, daily routines, etc.)
43. FL grammar and usage (e.g., verb conjugations, word order, agreement, idioms, vocabulary, etc.)
44. Details about tests, quizzes, and other assignments (how much will be covered, format of test sections, etc.)
45. Instructions for classroom activities
46. Administrative information (office hours, upcoming events, important dates, etc.)

Part 5. FL Acquisition and Classroom Communication (in the FL or in English)


Rate the importance of each type of communication in terms of its role in students’ successful acquisition of the FL
(i.e., regardless of course grade earned).
Mark the degree of importance of each type of communication on a scale of 1 (least important) to 5 (most important).
Note: each item may be given its own ranking; there can be more than one 5, 4, 3, etc.
47. FL topics (family, weather, FL culture, literature, study abroad, sports, hobbies, daily routines, etc.)
48. FL grammar and usage (e.g., verb conjugations, word order, agreement, idioms, vocabulary, etc.)
49. Details about tests, quizzes, and other assignments (how much will be covered, format of test sections, etc.)
50. Instructions for classroom activities
51. Administrative information (office hours, upcoming events, important dates, etc.)

Part 6. What You Believe Students Feel about FL Use in the Classroom
Mark the degree to which you agree or disagree with the following statements (1 ⫽ strongly disagree; 5 ⫽ strongly agree):
52. I believe that my students generally feel anxious using the FL, especially in the first year of instruction.
53. I believe that my students generally find it frustrating communicating in the FL.
54. I believe that my students generally view it as a rewarding and worthwhile challenge when they have to use the
FL to communicate (rather than fall back on English).
55. I believe students generally feel uncomfortable or anxious speaking the FL during activities about FL topics
(family, weather, FL culture, literature, study abroad, sports, hobbies, daily routines, etc.).
56. I believe students generally feel uncomfortable or anxious using the FL when working on, discussing, or asking
questions about grammar and usage (e.g., verb conjugations, word order, agreement, idioms, vocabulary, etc.).
57. I believe that students generally feel uncomfortable or anxious using the FL when working on, discussing, or
asking questions about tests, quizzes, and other assignments (how much will be covered, format of test sections,
etc.).
58. I believe that students generally feel uncomfortable or anxious using the FL discussing or asking questions about
administrative information (office hours, upcoming events, important dates, etc.).
Glenn S. Levine 363

APPENDIX C
Frequency (and Percentage) of Student and Instructor Estimates of Target Language Use among Different
Interlocutors

Percentage of Time
Interlocutors 0–20% 20%–40% 40%–60% 60%–80% 80%–100%
Instructor with Students Students 3 22 66 145 363
(n ⫽ 599) (0.5) (3.7) (11) (24.2) (60.6)
Instructors 3 6 33 49 72
(n ⫽ 163) (1.8) (3.7) (20.3) (30) (44.2)
Students with Instructor Students 68 104 168 153 105
(n ⫽ 598) (11.4) (17.4) (28.1) (25.6) (17.5)
Instructors 24 34 45 36 24
(n ⫽ 163) (14.7) (20.9) (27.6) (22.1) (14.7)
Students with Students Students 146 155 163 98 34
(n ⫽ 596) (24.5) (26) (27.4) (16.4) (5.7)
Instructors 33 45 40 27 17
(n ⫽ 162) (20.3) (27.8) (24.7) (16.7) (10.5)

APPENDIX D
Frequency (and Percentage) of Student and Instructor Estimations of Target Language Use in Different
Contexts

Percentage of Time
Context of TL Use 0–20% 20%–40% 40%–60% 60%–80% 80%–100%
Topic/Theme Students 48 78 132 166 168
(n ⫽ 592) (8.1) (13.2) (22.3) (28) (28.4)
Instructors 6 9 16 43 85
(n ⫽ 159) (3.8) (5.7) (10) (27) (53.5)
Grammar Students 88 140 160 142 68
(n ⫽ 598) (14.7) (23.4) (26.8) (23.7) (11.4)
Instructors 35 30 21 33 43
(n ⫽ 162) (21.6) (18.5) (13) (20.4) (26.5)
Tests, etc. Students 149 150 127 96 71
(n ⫽ 593) (25.1) (25.3) (21.4) (16.2) (12)
Instructors 39 35 24 26 38
(n ⫽ 162) (24.1) (21.6) (14.8) (16) (23.5)
364 The Modern Language Journal 87 (2003)

APPENDIX E
Frequency (and Percentage) of Student and Instructor Agreement with Statements about
Target-Language-Use Anxiety in General

Level of Agreement
Strongly Strongly
Disagree Agree
1 2 3 4 5
Anxious Using the FL Students 49 125 178 167 70
(n ⫽ 589) (8.3) (21.2) (30.2) (28.4) (11.9)
Instructors 6 19 40 49 48
(n ⫽ 162) (3.7) (11.7) (24.7) (30.3) (29.6)
Frustrating Students 52 151 169 124 93
(n ⫽ 589) (8.8) (25.6) (28.7) (21.1) (15.8)
Instructors 12 33 43 43 31
(n ⫽ 131) (7.4) (20.4) (26.5) (26.5) (19.2)
Worthwhile Challenge Students 19 64 132 209 166
(n ⫽ 590) (3.2) (10.9) (22.4) (35.4) (28.1)
Instructors 4 18 37 55 47
(n ⫽ 161) (2.5) (11.2) (23) (34.1) (29.2)

APPENDIX F
Frequency (and Percentage) of Student and Instructor Agreement with Statements about
Target-Language-Use Anxiety in Particular Communicative Contexts

Level of Agreement
Strongly Strongly
Disagree Agree
Anxious in FL 1 2 3 4 5
During Topic/Theme Students 90 204 155 87 51
(n ⫽ 587) (15.3) (34.8) (26.4) (14.8) (8.7)
Instructors 37 66 30 24 4
(n ⫽ 161) (23) (41) (18.6) (14.9) (2.5)
Discussing Grammar FL Students 57 169 153 137 74
(n ⫽ 590) (9.7) (28.6) (25.9) (23.2) (12.6)
Instructors 7 22 36 52 45
(n ⫽ 162) (4.3) (13.6) (22.2) (32.1) (27.8)
Discussing Tests, Quizzes Students 61 156 142 138 92
(n ⫽ 589) (10.4) (26.5) (24.1) (23.4) (15.6)
Instructors 7 16 34 47 57
(n ⫽ 161) (4.4) (9.9) (21.1) (29.2) (35.4)

You might also like