THE EFFECT OF AFFECT ON FOREIGN LANGUAGE

LEARNING: A REVIEW OF THE ANXIETY RESEARCH

Thomas Scovel
University of Pittsburgh
Although studies of the relationship between affective fac-
tors and language learning proficiency abound in the literature, the
evidence to support such a relationship is difficult to interpret.
Much of the problem resides in the fact that a wide range of
variables are lumped together under the rubric “affect.” An
attempt is made to ameliorate this situation by defining affective
variables in terms of traditional psychological theory and classify-
ing them as a subset of those variables intrinsic to the learner. The
conflicting evidence dealing with one important affective variable,
anxiety, is then examined, and it is shown that ambiguous
experimental results can be resolved if the distinction between
facilitating and debilitating anxiety is drawn. Further classificatory
distinctions are discussed from the abundant experimentation
undertaken by applied psychologists, and an attempt is made to
consider the implications of some of this research for adult
language learning-for some of the new methodologies in EFL as
well as for future research opportunities.

Affective Variables

One does not have to delve deeply into the literature on the
relationship between affective variables and second language learn-
ing to discover that “affect” is a cover term under which is swept
a wide range of disparate constructs and behaviors. Included under
the rubric of affective variables are such various categories as:
“cognitive style” (Brown 1973), “ego boundaries” (Taylor 1974),
“reserved vs. outgoing personality” (Chastain 1975), and “adven-
turesome” (Tucker e t al. 1976). Perhaps the most peculiar candi-
date of all for an affective construct, and, at the same time, the
most indicative of the need for a more precise definition of affect
in second language acquisition, is the variable “conformist protes-
tant ethic” suggest by Scoon (1971). Affective variables have often
been defined as the converse of cognitive variables; that is, they are
everything which impinges on langauge learning which is unrelated
to cognition. Although half a definition is better than none at all,
this interpretation is imprecise; furthermore, it forces us to ac-
complish an even more formidible task-to define cognitive vari-
ables! As Lamendella (in press) has wryly observed, there are two
kinds of researchers who deal with cognition, those who define it

129
130 LANGUAGE LEARNING VOL. 28, NO. 1

erroneously and those who don’t define it at all. In the interests of

avoiding such a nihilistic approach t o the definition of affective
variables, a necessary prelude to the investigation into their rela-
tionship with second language learning is t o attempt to place affect
in a proper perspective-as a subset, along with cognition, whatever
it may be (if it is a t all!), of a larger set of variables, a class that
would include as separate subsets many of the constructs found in
the literature to date, including the examples cited above.
Several investigators have suggested a broad, classificatory
term for this large set of variables of which affect could be
considered an appropriate subset, among them being: “initiating
factors” (Schumann 1976), “perseverance factors” (Swain and
Naiman 1976), and “personality characteristics” (Swain and
Burnaby 1976), but perhaps the most direct and simplest is
“learner variables” (Swain 1977). After reviewing Schwartz’s classi-
fication of motivational factors in learning, Chastain (1976) sug-
gests that there appear to be two different types of factors which
influence the learner; in other words, “learner variables” can be
divided into two subcategories: intrinsic learner variables and
extrinsic ones.
Some students are generally motivated intrinsically; others extrinsi-
cally. Schwartz (1972) lists the intrinsic motivators as anxiety,
need to achieve, self-concepts, and aspirations; and the extrinsic
motivators as sociocultural influences and social reinforcers. ( 7 3)
It can be seen, therefore, that most of the constructs and behaviors
which have been misclassified as affective factors in the literature
can be subsumed under the category “learner variables,” either
intrinsic or extrinsic t o the learner, but affect, if we adhere t o its
traditional definition in psychology, is itself only one variable
within intrinsic learner variables, and, therefore, if we are t o
proceed with an examination of the relationship of anxiety to
foreign language learning, we must first of all realize that we are
talking about only one affective variable among many intrinsic
learner variables.
Another way of viewing this set of factors which comprise the
personality of the language learner is t o adopt a classificatory
scheme from a source quite alien to the traditional literature of
applied linguistics and applied psychology. The advantage of going
so far afield is twofold: it provides us with a fresh perspective of
the variables which we in language learning research have become
so accustomed to, and it serves to reinforce the point of view that
is central to the position of this article, that affect (feeling) is but
one ingredient in the complex and marvellous chemistry that
 SCOVEL 131

creates the personality. The model I am thinking of is adopted

from the rich tradition of Buddhist philosophy. Like all major
religious literature, Buddhists texts have dealt with the essential
nature of man and the make-up of human personality. Thus, it is
not surprising to discover that the personality has been thoroughly
investigated, and, according to Buddhist belief, it is classified into
five skandhas or “mind-body complexes” (Conze 1967: 107).
As the stars in a constellation do not really belong together, but it
is we who have arranged them into an arbitrary unit, so also our
“personality” is a mere conventional grouping of disparate ele-
ments, all of which belong to one of the five groups, known as the
skandhas.
Table 1 categorizes intrinsic learner variables into the five skandhas
of Buddhist philosophy, citing an example from language learning
research for each of the personality characteristics (learner vari-
ables) which relate t o language acquisition. Under this scheme,
affective variables, if they are to be viewed correctly, are those
factors which deal with uedane (feelings), the emotions of pleasure
and displeasure that surround the enterprise of a task such as
second language learning.
It should be clear from this introductory discussion that
affective factors are those that deal with the emotional reactions
and motivations of the learner; they signal the arousal of the
limbic system and its direct intervention in the task of learning.
Having made this point, we may now proceed t o an examination
of one of the most important affective variables identified in
learning tasks-anxie ty .

TABLE 1
Some Intrinsic Learner Variables: Classified as skandhas, the five
components of the mind-body according t o
Buddhist philosophy

Skandhas English Term Example of Research on this Variable

rupa body lateralization and the critical period
for language learning
savskarci impulses integrative vs. instrumental
motivation
savjiid perceptions and the conscious use of the “monitor”
cognition by adult language learners
vijiiana consciousness
or soul empathy
vedana feelings or
affective variables anxiety
132 LANGUAGE LEARNING VOL. 28, NO. 1

Anxiety and Foreign Language Learning

The research into the relationship of anxiety to foreign

language learning has provided mixed and confusing results, im-
mediately suggesting that anxiety itself is neither a simple nor
well-understood psychological construct and that it is perhaps
premature t o attempt to relate it to the global and comprehensive
task of language acquisition. Some studies have revealed incomplete
correlations between anxiety and measures of language proficiency;
Swain and Burnaby, for example, in their study of English-speaking
French immersion children found a negative correlation between
anxiety and one measure of the children’s proficiency in French,
but, at the same time, found no other significant correlations,
either negative or positive, with any other proficiency measures.
Tucker et alia also found that anxiety (in this case, one of a
duster of variables in a factor analysis) correlated again with one
measure of French proficiency, but not with any of the three
other criterion measures of language proficiency which they ex-
amined. It is perplexing to find anxiety implicated with one skill at
one level, but never with other skills‘ at different levels. Other
studies have revealed complete correlations, that is, there is a
consistent relationship between the academic performance of a
language student in the classroom and an anxiety measure, but
these correlations directly contradict the results obtained with
other students or other languages. Backman (1976) found that the
two worst English-learning Spanish speakers in her study scored the
highest and the lowest on the anxiety measure she utilized. In a
larger study, Chastain (1975) found a negative correlation between
French audio-lingual method student scores on tests and anxiety,
but, in contradiction, he discovered a positive correlation between
anxiety and the scores of German and Spanish students using the
traditional method. Chastain accurately identifies the crux of the
problem in these anxiety studies when he states that: “perhaps
some concern about a test is a plus while too much anxiety can
produce negative results” (160).
If Chastain’s paper indicates the direction toward which
anxiety/language studies should move, it is an article by Kleinmann
(1977) which actually takes a step in the right direction, first by
defining two different types of anxiety, and then seeking to
discover a correlation between these two different constructs and
measures of second langauge learning behavior. Kleinmann was
interested in examining the relationship between the syntactic
structures in English that are avoided by foreign students and
 SCOVEL 133

syntactic structures of the students’ native languages-the hypothe-

sis being that the English structures which contrasted most mark-
edly with the native language of the student would be avoided
most frequently. After examining the English output of Arabic and
Spanish students using a variety of tests, Kleinmann was able to
confirm his hypothesis. Of secondary interest to him was the
relationship between the avoidance behavior of the subjects and
anxiety, the assumption being that “facilitating anxiety” would
encourage learners t o employ the very English structures that their
native language group would tend to avoid. Table 2 summarizes
some of the results obtained by Kleinmann and indicates that, in
general, facilitating anxiety does function in the predicted manner.
Apparently, the Spanish students who scored high on facilitating
anxiety measures (e.g. “Nervousness while using English helps me
do better”) frequently used infinitive complements and direct
object pronouns in English, structures that were avoided because of
their difficulty by the other Spanish-speaking students. In the same

TABLE 2
Correlations between two measures of anxiety and tendency to use
English structures normally avoided b y native language
group (adapted from Kleinmann, 1977)

Use of English
Use of English structures structures normally
normally avoided by avoided by Arabic
Spanish speaking EFL speaking EFL
students students

infinitive direct present

complement object pron. passive progressive

high
facilitating sig. cor. sig. cor. - -
Spanish anxiety <
( p .01) (p < .01)
speaking
students high
debilitating - - - -
anxiety
high
facilitating - - sig. cor. #
Arabic anxiety <
(P .05)
speaking
students high
debilitating - - - -
anxiety

- no significant correlation obtained, as predicted

# no significant correlation obtained, despite prediction
134 LANGUAGE LEARNING VOL. 28, NO. 1

manner, Arabic students who scored high on facilitating anxiety

measures tended to use the passive (the present progressive proved
t o be an exception) more frequently than their Arabic-speaking
peers, who usually avoided this structure because of its divergence
from Arabic syntax. Kleinmann summarizes his conclusions about
the relationship between this affective variable and language acqui-
sition behavior with the following important statement.
The findings suggest avoidance operating as a group phenomenon,
but within the particular avoiding group, use of the generally
avoided structure is a function of facilitating anxiety levels of the
group’s members. This finding is not inconsistent with a study
conducted by Chastain (1975), who found a significant negative
correlation between test anxiety and final course grade in a
university audio-lingual French-as-a-foreign-languageclass. Obvi-
ously, the anxiety there had a debilitating influence. But Chastain
also implied a facilitating influence of anxiety based on his
findings that anxiety was a significant predictor of success in
learning Spanish as a foreign language. The evidence, therefore,
seems to support the notion that certain affective measures
influence learner behavior in a foreign language. (105)

The studies by Kleinmann and Chastain indicate the manner

in which an affective variable, in this case, anxiety, can be
appropriately investigated, but it behooves us to examine other
ways in which anxiety can be viewed, not as a simple, unitary
construct, but as a cluster of affective states, influenced by factors
which are intrinsic and extrinsic t o the foreign language learner.
For this multidimensional perspective of anxiety, we must turn
from language learning research and direct our attention on some
of the work that has been done on anxiety in applied psychology.

Anxiety Research in Applied Psychology

Anxiety is commonly described by psychologists as a state of

apprehension, a vague fear that is only indirectly associated with
an object (Hilgard, Atkinson, and Atkinson 1971). Because anxiety
is clearly an emotional state, it is generated through the arousal of
the limbic system, the primitive, subcortical “chasis” of the
cerebrum, which plays an important, though indirect, role in many
kinds of human enterprises, including communication (Lamendella
1977). Anxiety is usually measured in one of three ways: by
behavioral tests, where the actions of a subject are observed (floor
pacing by a father in a maternity waiting room); by the subject’s
self-report of internal feelings and reactions (“I feel uneasy when
awaiting the delivery of my first child.”); or by physiological tests,
where measures of heart rate, blood pressure, or palmar sweating
 SCOVEL 135

are taken and these are assumed t o be correlated to the subject’s

emotional state.
Because the limbic system can trigger a variety of physiologi-
cal responses through the autonomic nervous system, physiological
measures have long been used as an easily quantifiable indicator of
a subject’s emotional state. A study by Schaffer (1947) undertaken
on combat aviators over three decades ago, indicated that physio-
logical reactions associated with anxiety varied from rapid pulse
rate (most common) to soiling pants (least common)! Despite the
illusion of empirical objectivity conveyed by the physiological
experimentation, there is enormous individual variation in the
physiological reactions of subjects under different states of emo-
tional arousal. Schnore (1959), for example, discovered that pat-
terns of physiological responses varied considerably among indi-
viduals (e.g. one subject might exhibit a lower heart rate than
another during a “high arousal” task, but, at the same time, a
higher degree of forearm tension than the other subject). The most
promising avenue of research into the correlation of physiological
measures of emotional states t o human behavior seems to lead to
the athletic arena, where quality of performance is measured
almost entirely by active physical behavior. It is common, there-
fore, t o encounter studies such as that undertaken by Nideffer and
Yock (1976) which measured the relationship between amount of
sweating during pre-race warmups and success in actual races by
competitive swimmers. It was found that the higher the arousal
level of the swimmers during the pre-race periods, as measured by
increased palmar sweating, the poorer their results in actual races.
If palmar sweating is an accurate measure of affective arousal, and
if this emotional state is akin to anxiety, then it appears that
increased emotional activity has an adverse effect on the ability to
perform demanding physical activities at optimal levels of success.
Even though language learning is largely a cerebral, rather than a
Physical endeavor, it is related in many ways to the acquisition of
athletic skills, especially the neuromuscular task of speaking (Sco-
vel 19?3), and it might be worthwhile to investigate the possible
relationship between physiological measures of emotional arousal
and success in foreign language performance, especially in articula-
tory tasks.
The other two measures of affective arousal, paper and pencil
tests of behavior and self-reports, are not as easily quantifiable as
the physiological tests, but they do have an advantage in that they
are much more precise in focussing in on a specific affective
construct, say anxiety, than the physical measures which can only
136 LANGUAGE LEARNING VOL. 28, NO. 1

assume to be related to affective involvement; in addition, these

behavioral measures are easy to administer t o large groups of
subjects. For these reasons, they have been used more abundantly
in applied psychology than the physiological tests. These behavioral
tests are numerous and include, among others, the following: the
Achievement Anxiety Test (AAT), the Taylor Minifest Anxiety
Scale (MAS), the Yale Test Anxiety Scales (TAS), the State/Trait
Anxiety Inventory (STAI), the Test Anxiety Scale for Children
(TASC), the Children’s Manifest Anxiety Scale (CMAS), and the
Digit Span Test. Unfortunately, as Ekeman, Martin, and Meyers
(1972) point out, the reliability between these behavioral measures
of anxiety and the physiological measures remains unsubstantiated.
The relationship between paper and pencil measures of anxiety
and physiological measures have been studied for years and
..
contradictory evidence is the rule. Generally, there is a low
correlation between clinically rated anxiety, self-rated anxiety, and
psychometric anxiety. (427)
There has been a large amount of research undertaken in
educational psychology which attempts to assess the impact of
anxiety, as measured by these many behavioral tests, on academic
performance. As might be expected, the more that researchers have
investigated the topic, the more complex the relationship between
anxiety and classroom performance has grown, and the models first
constructed to account for affect in learning have been encum-
bered with increasing numbers of intervening variables. Among the
most important of these are: the subject studied or tested at
school, the children’s level of intelligence, the difficulty of the
learning skill under investigation, and the degree of familiarity the
children have with the learning task. It has been discovered, for
example, that higher states of anxiety facilitate learning at upper
levels of intelligence whereas they are associated with poorer
preformance at lower I& levels, ceteris paribus (Verma and
Nijhawan 1976); furthermore, increased anxiety is likely to im-
prove performance at later stages in a learning activity, but
conversely hinders academic performance at earlier stages of the
same activity (Beeman, Martin, and Meyers 1972). Based on an
earlier model of emotional drive by Taylor (1956) and Spence
(1958), which claimed that high emotional drive facilitates per-
formance when correct response reinforcement is stronger than
competing response reinforcement (usually occurring at initial
stages of learning), Spielberger (1966) has presented an elaborate
model which effectively integrates intelligence, stage of learning,
and difficulty of task into an examination of the impact of anxiety
 SCOVEL 137

on learning performance. In brief, the model claims that high

anxiety (HA) facilitates learning when the task is relatively easy
but leads to decrements in performance when the task becomes
more difficult. This means that for low I& students with HA
motivation, simple tasks will be learned well with little or no initial
exposure, intermediate tasks will be learned well after increased
exposure, and that difficult tasks will probably be learned poorly,
if at all, even after periods of long exposure. For high I& students
with HA drive, simple and intermediate tasks will be learned well
irrespective of exposure, and for difficult tasks, although there
might be a decrement in performance at the initial stages of the
task, HA will eventually facilitate high levels of performance after
a period of increased exposure. Consequently, Spielberger’s model
suggests that whereas HA will eventually provide positive motiva-
tion for high I& students learning a difficult skill, it will provide
negative reinforcement to the low I& student, even under con-
tinued exposure to the activity to be acquired.
In an investigation of 7th grade children studying a variety of
subjects in twelve different Australian schools, Gaudry and Fitz-
gerald (1971) confirmed the hypotheses proposed by Spielberger
and implicated a cognitive factor (IQ) as an intervening variable in
any consideration of the affective construct, anxiety. This led them
to conclude:
... while high anxiety was associated with slightly higher perform-
ance for the most able children, it was associated with lower
performance at all other levels except the centre group where the
mean scores were identical. (161)

Aside from the above-mentioned factors which seem to regu-

late the effect of affect on academic performance, be they
non-affective intrinsic factors such as intelligence, or extrinsic
factors such as level of difficulty of the material learned, another
important fact to be taken into account is the point considered
earlier in the discussion of Kleinmann’s study, where it was
emphasized that anxiety itself is not a simple, unitary construct
that can be comfortably quantified into either “high” or “low”
amounts. On the one hand, some researchers feel that momentary
anxiety should be distinguished from a more permanent predisposi-
tion to be anxious, and that this dichotomy would help to account
for some of the conflicting results of previous anxiety studies. This
sentiment led Spielberger, Gorsuch, and Lushene (1970) to develop
a measure which distinguishes between state anxiety (SA) and trait
anxiety (TA), a distinction which appears to resolve some of the
discrepancies of earlier studies which failed to consider this para-
138 LANGUAGE LEARNING VOL. 28, NO. 1

meter of subject behavior (Hodges and Spielberger 1969). On the

other hand, some researchers have felt it important to define
anxiety in terms of the potential effect it has on performance;
hence, Alpert and Haber (1960) developed the Achievement Anx-
iety Test to identify the amount of facilitating and/or debilitating
anxiety a subject possess. The authors of this instrument are quick
to point out that these constructs are not extremes on a con-
tinuum, but are independent of each other:
In fact, these two constructs of debilitating and facilitating anxiety
may be uncorrelated. Thus, an individual may possess a large
amount of both anxieties, or of one, but not the other, or of none
of either. The nature of this correlation can be determined
empirically following the construction of two such independent
.
measures of anxiety . . (213)

Anxiety Motivation in Foreign Language Learning

The facilitating/debilitting measure constructed by Alpert

and Haber presents an attractive path down which future research
on the effects of anxiety on foreign language acquisition might
proceed, providing that a sincere attempt is made by language
researchers to control or account for the many intervening vari-
ables that have already been mentioned. Indeed, there are several
current trends in EFL which suggest that the field in which applied
linguists labor is fertile for research on affect to take root and
blossom-especially research on anxiety, as long as it is well-
designed, empirical, and, if the pun can be forgiven, well-grounded.
The attractiveness of this binary approach to anxiety lies in
its common sense viewpoint that learning, whatever the activity
might be, is enhanced by both positive and negative motivation. A
good performance, in music, in sports, or in language learning,
especially the overt social act of speaking, depends on enough
anxiety to arouse the neuromuscular system to optimal levels of
performance, but, at the same time, not so much that the complex
neuromuscular systems underlying these skills are disrupted. In this
sense, these two aspects of anxiety fulfill a similar function to that
of the two complementary components of the autonomic nervous
system, the sympathetic and the parasympathetic; one arouses, the
other depresses-ach working together and in balance to keep the
organism in tune with its ever-changing environment. So it is with
facilitating and debilitating anxiety in the normal learnereach
working in tandem, serving simultaneously to motivate and to
warn, as the individual gropes t o learn an ever-changing sequence
 SCOVEL 139

of new facts about the environment. It is not at all remarkable

that these two aspects of anxiety are so directly linked to the
limbic system, the source of all affective arousal, because they
appear to be emotional correlates of two of the four basic drives
which are generated by this primitive portion of the human brain.
Facilitating anxiety motivates the learner to “fight” the new
learning task; it gears the learner emotionally for approach be-
havior. Debilitating anxiety, in contrast, motivates the learner to
“flee” the new learning task; it stimulates the individual emotion-
ally to adopt avoidance behavior. Kleinmann’s work is a good
example of the interplay between these two basic drives of the
limbic system as manifested by the cooperative functioning of two
different types of anxiety in a language learning task. Although the
difference between the students’ native language and the target
language encouraged them to avoid (to “flee” from) the English
structures which were different, those students who scored high on
facilitative anxiety were emotionally equipped to approach (to
“fight” in primitive terms) the very structures that their peers
tended t o avoid.
Another consideration that should be incorporated into the
affective research is the Monitor Model for adult language acquisi-
tion. Krashen (1976) has sought to distinguish between two
different types of activity in which adults engage when they
attempt to master a’ new language. Adult learners can pick up a
language informally and unconsciously, in a manner similar to
children. This process, called Zunguuge acquisition, contrasts with
language learning, the formal, conscious study of a language more
commonly undertaken by adults than by children. It is Krashen’s
contention that in this latter sphere of endeavor, adults have an
opportunity to monitor their language output, using the rules and
patterns they have been formally taught, to edit their target
language production. Krashen’s model presents yet another variable
to be accounted for in our investigation of the effect of anxiety on
language learning performance. Regardless of what specific mea-
sures of affect are used, I would imagine that anxiety is more
directly implicated in the formal activity of language learning than
in the informal enterprise of language acquisition. Be that as it
may, the monitor theory should be incorporated into any model
concerning the effect of affect on foreign language learning, for it
deals with the intrinsic learner variables that are part and parcel of
the learner’s personality, and, as such, have a bearing on the
individual’s affective motivation.
Currently, there is great interest in the psychological motiva-
140 LANGUAGE LEARNING VOL. 28, NO. 1

tion of adult foreign language learners, especially in terms of the

students’ relationship to their teachers and their learning environ-
ment. This interest has encouraged the development of several new
methods: among them, the Silent Way (Gattegno 1972), Counsel-
ing-Learning (Curran 1976), and Suggestopedia (Lozanov 1973). As
Stevick (1976) has pointed out, these methods are to be com-
mended in that they recognize the importance of affective motiva-
tion in the task of foreign language learning. It is fair to claim that
all of these methods are interested in controlling and regulating
affective motivation, in particular, anxiety, and in a vague way,
proponents of these methods are making claims that are supported
by Kleinmann’s anxiety results. Blatchford (1976), for example, in
a discussion of the Silent Way, refers to the negative effects of
“tenseness” (debilitating anxiety?) but to the salubrious benefits of
“tension” (facilitating lanxiety?) in the EFL classroom. I would like
to emphasize, however, that the results of the anxiety research in
applied psychology do not provide a ringing endorsement of these
new methodologies; rather, the results indicate what language
teachers have known all along, that students learn better in a
supportive, non-threatening environment. As Brown (1977) has
pointed out in his fair but forthright review of Counseling Learn-
ing, the enormous magnitude and variety of affective and cognitive
variables that intervene at both the individual and the cultural level
in language learning, make it difficult, if not impossible, to
prescribe one single method as the most congruent with the
emotional and intellectual needs of the majority of students.
The conclusion might read like a good news-bad news joke,
and if that is so, I trust that the news is not so much amusing as it
is informative. The good news is that we are able to isolate
affective variables in our research into the psychology of language
acquisition; this is illustrated by Kleinmann’s study which mea-
sured the effects of anxiety on language learning performance in a
well-circumscribed experiment. The bad news is that the deeper we
delve into the phenomenon of language learning, the more complex
the identification of particular variables becomes. As this paper has
suggested, before we begin to measure anxiety, we must become
more cognizant of the intricate hierarchy of learner variables that
intervene: the intrinsicfextrinsic factors, the affective/cognitive
variables, and then the various measures of anxiety and their
relationship t o these other factors. But the overwhelming intricacy
of these intertwining systems should not deter us from the task of
trying to discover natural patterns and continuities, for, at the very
least, we will realize even more profoundly and with even deeper
 SCOVEL 141

respect than before, the marvellous act that our students so subtly
perform in front of us day by day, the act of inheriting someone
else’s language and culture.

REFERENCES
Alpert, R. and Haber, R. 1960. Anxiety in academic achievement situations.
Journal of Abnormal and Social Psychology 61.207-215.
Backman, N. 1976. Two measures of affective factors as they relate to
progress in adult second-language learning. Working Papers in Bilingualism
10.100-122.
Beeman, P., Martin, R., and Meyers, J. 1972. Interventions in relation to
anxiety in school. Anxiety:Current Trends in Theory and Research.
edited by C. Spielberger, New York: Academic Press.
Blatchford, C. 1976. My silent way experience-one model for training
teachers and students. Paper presented to the Xth Annual TESOL
Convention. New York.
Brown, H. D. 1973. Affective variables in second language acquisition.
Language Learning 23.231-244.
Brown, H. D. 1977. Some limitations of Counseling-Learning/Community
Language Learning models of second language teaching. TESOL Quarterly
11 .ooo-000.
Chastain, K. 1975. Affective and ability factors in second language acquisition.
Language Learning 25.153-161.
Chastain, K. 1976. Developing Second Language Ski1ls:Theory to Practice.
Chicago: Rand McNally.
Conze, E. 1967. Buddhist Thought in 1ndia:Three Phases of Buddhist Phi-
losophy. Ann Arbor: University of Michigan Press.
Curran, C. 1976. Counseling-Learning in Second Languages. Apple River:
Apple River Press.
Gattegno, C. 1972. Teaching Foreign Languages in Schoo1:the Silent Way.
New York: Educational Solutions.
Gaudry, E. and Fitzgerald, D. 1971. Test anxiety, intelligence, and academic
achievement. Anxiety and Educational Achievement. edited by Gaudry
and Spielberger. Sydney: John Wiley and Sons.
Hilgard, E., Atkinson, and Atkinson. 1971. Introduction to Psychology. New
York: Harcourt, Brace, and World.
Hodges, W. and Spielberger, C. 1969. Digit span:an indirect indicant of trait
or state anxiety? Journal of Consulting and Clinical Psychology
33: 430-434.
Kleinmann, H. 1977. Avoidance behavior in adult second language acquisition.
Language Learning 27.93-107.
Krashen, S. 1976. Formal and informal linguistic environments in language
acquisition and language learning. TESOL Quarterly 10.157-168.
Lamendella, J. 1977. The limbic system in human communication. Studies in
Neurolinguistics. Volume 3. edited by Whitaker and Whitaker, New York:
Academic Press.
Lamendella, J. in press. The Early Growth of Cognition and Language: a
Neuropsychological Approach.
Lozanov, G. 1973. Suggestology. Nauka I Izkistvo. Sofia (in Bulgarian; English
translation pending with Gordon and Breach Science Publishers, New
York).
142 LANGUAGE LEARNING VOL. 28, NO. 1

Nideffer, R. and Yock, T. 1976. The relationship between a measure of

palmar sweat and swimming performance. Journal of Applied Psychology
61.376-378.
Schaffer, L. 1947. Fear and courage in aerial combat. Journal of Consulting
Psychology 11.137-143.
Schnore, M. 1959. Individual patterns of physiological activity as a function
of task differences and degree of arousal. Journal of Experimental
Psychology 58.117-128.
Schumann, J. 1976. Second language acquisition research:getting a more
global look at the learner. Papers in Second Language Acquisition. edited
by Brown. Ann Arbor.
Schwartz, L. 1972. Educational Psycho1ogy:Focus on the Learner. Boston:
Holbrook Press.
Scoon, A. 1971. Affective influences on English language learning among
Indian students. TESOL Quarterly 5.285-291.
Scovel, T. 1973. Language learning as a sport. Education 93.84-87.
Spence, K. 1958. A theory of emotionally based drive (D) and its relation to
performance in simple learning situations. American Psychologist
13.1 31-141.
Spielberger, C. (editor). 1966. Anxiety and Behavior. New York: Academic
Press.
Spielberger, C., Gorsuch, and Lushene. 1970. State-trait anxiety inventory.
Palo Alto: Consulting Psychologist Press.
Stevick, E. 1976. Memory, Meaning and Method:Some Psychological Perspec-
tives on Language Learning. Rowley : Newbury House Publishers.
Swain, M. 1977. Future directions in second language research. Proceedings of
the L. A. Second Language Research Forum. UCLA:Department of
English-ESL section.
Swain, M. and Burnaby, B. 1976. Personality characteristics and second
language learning in young children: a pilot study. Working Papers in
Bilingualism 11.115-128.
Swain, M. and Naiman, N. 1976. Discussion of Schumann’s paper. Papers in
Second Language Acquisition. edited by Brown. Ann Arbor.
Taylor, B. 1974. Toward a theory of language acquisition. Language Learning
24.23-35.
Taylor, J. 1956. Drive theory and manifest anxiety. Psychological Bulletin
53.303-320.
Tucker, R., Hamayan, and Genesee. 1976. Affective, cognitive, and social
factors in second language acquisition. Canadian Modern Language Re-
view 32.214-226.
Verma, P. and Nijhawan, H. 1976. The effect of anxiety, reinforcement, and
intelligence on the learning of a difficult task. Journal of Experimental
Child Psychology 22.302-308.