1

' M ”f'. w

.' . «LN

 

 

 

 

 

 

who“

I}

"19; m2. _

W...”
n -"~...:.»,..>-...

f‘\'.ffrvrrr---rrl -
I ._..’.».,

.—....-.,',... ,

'_""' ... ,.-_.',’._.'. ,,

 

 

 

 

 

 

 

 

 

 

 

l H5819

 

Date

lllllllllllllllllllllllllllllllllllllllllllllllllllll

31293 00896 4540

This is to certify that the

thesis entitled

MONITOR USE BY ADULT LEARNERS OF GERMAN
AS A FOREIGN LANGUAGE

presented by

Laura M Bowman

has been accepted towards fulfillment
of the requirements for

1 .
Master 5 degree in German

”may é %/

Major professor

 

 

F

 

June 15, 1991

 

0-7639

MS U is an Affirmative Action/Equal Opportunity Institution

 

 

  
     

  

LIBRARY
Michtgan State :-
' University 7

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

NWI Q a 20:52

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
c:\cIrc\datedm.pm3—p.1

 

MONITOR USE BY ADULT LEARNERS OF GERMAN AS A FOREIGN LANGUAGE

BY

Laura Marie Bowman

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Linguistics and Germanic, Slavic, Asian and African
Languages

1991

é j¢~ 0775

ABSTRACT

MONITOR USE BY ADULT LEARNERS OF GERMAN AS A FOREIGN LANGUAGE
BY

Laura Marie Bowman

The following Master's Thesis investigates the ability of
beginning learners of German as a foreign language to
successfully monitor their writing in various communicative
situations. The paper sets out to test the hypothesis posited by
Krashen's Monitor Model that Sufficient Time, Focus on Grammar,
and Knowledge of the Grammar Rule are necessary in order for
monitoring to occur.

Thirty-six students of elementary German were asked to
describe slides depicting everyday activities. They were also
asked to complete a fill-in—the—blank grammar excercise and to
fill in a verb-ending chart in order to assess their performance
in more grammar—oriented situations. The language samples were
then evaluated for subject—verb ending agreement.

The study found that there was not a significant difference
in the performance across the various communicative situations.
Possible reasons for these findings, as well as recommendations

for future study, are then suggested.

 

ACKNOWLEDGMENTS

I would like to thank all of the people who, through
their support and guidance helped me in preparing this
document: The members of my advisory committee, friends and
family members. Special thanks to Professor Lovik for his
invaluable suggestions and to my husband Craig for his help
with technical aspects of the paper and for his much-needed
moral support.

TABLE OF CONTENTS

List of Tables

List of Figures
Introduction

Theory and Previous Research
Objective and Hypothesis
Methodology

Scoring

Results

Conclusions

Discussion

Further Study

Conclusion

Appendix A

Appendix B

Appendix C

Works Cited

iv

vi

l4

l6

l9

19

26

27

30

31

33

34

44

45

LIST OF TABLES

Table I: Summary of findings for Tasks 1-4 21

LIST OF FIGURES

Figure I: Scatter Plot showing improvement when
concentrating on grammar versus knowledge of the rule 23

Figure II: Scatter Plot for the percent difference between
Tasks l and 4, versus knowledge of the rule 25

Figure III: Scatter Plot for performance on discrete point
grammar test versus knowledge of the rule 26

vi

 

1. Introduction

This study is based partially on a pilot study in which I
observed behavior contrary to that predicted by Krashen's
(1982) Monitor Model and partially on a study done by
Hulstijn (1982); also discussed in Hulstijn and Hulstijn
(1984) which prompted Krashen to alter one of the hypotheses
in his Monitor Model. More specifically, it examines the
accuracy of the foreign language performance of beginning
learners of German in several communicative situations to
determine whether their behavior is consistent with that

which is predicted by Krashen's theory.

2. Theory and Previous Research

Before examining Krashen's Monitor Model, I will look at
some of the other research into second language acquisition
conducted in the past few decades. The 1960's saw a shift
in the way researchers and psychologists viewed the language
acquisition process (both first and second language
acquisition). Instead of accounting for the language
acquisition process from a behaviorist standpoint, in which
stimulus-response and habit formation play a major role,
psychologists and linguists turned to a mentalist view of
language acquisition. According to mentalist accounts,
every language learner possesses an innate set of linguistic
principles, sometimes referred to as the Universal Grammar,
which determines the form which the sentences of a given

language could take and guides the language learner in

 

2
relating the universal principles of language to the
specific language to which he or she is exposed (Ellis
1986,43).

Several researchers developed similar theories about the
nature of second language acquisition based on the mentalist
views of the process of language acquisition and of language
learners' unique language systems. Selinker's
interlanguage, Corder's idiosyncratic dialects and
transitional competence, and Nemser's approximative systems
are all included here under the heading interlanguage. The
term interlanguage refers to two phenomena. On the one
hand, interlanguage refers to "a separate linguistic system
based on the observable output which results from a
learner's attempted production of a target language norm"
(Selinker 1972, 117). This interlanguage is separate from
the learner's first language (L1) and the second language
(L2) he or she is attempting to learn, thus it represents an
interim stage. On the other hand, interlanguage refers to a
series or a continuum of these linguistic systems which
gradually approach the target language (TL) norm. As Ellis
(1986, 50-1) points out, subsequent investigation into the
interlanguage theory looks at three key features. First,
interlanguage is permeable, meaning that the knowledge which
learners possess at any given time is open to revision and
amendment. Second, interlanguage is continually changing but
the changes evolve gradually. Third, interlanguage is

systematic. Learners select from their store of

3
interlanguage rules systematically rather than haphazardly
and their performance is based on their existing rule
system.

The theories of interlanguage profoundly changed the way
in which errors were viewed. Instead of being regarded as
something to be avoided at all costs, errors became a means
through which to study the development of a language
learner's interlanguage and to gain insight into the
language-learning process. This opened the door to numerous
studies of language—learner interlanguage.

Several studies, collectively called "morpheme studies,“
attempted to discover whether there was an invariant order
in which language learners acquire certain grammatical
features, which would in turn suggest that there might be
universal processing strategies at work in second language
learners. Dulay and Burt used the Bilingual Syntax Measure
(Burt, Dulay and Hernandez 1973) to study the oral
production of English of 55 Chinese and 60 Spanish-speaking
children in order to determine whether children with
dissimilar native languages would produce a similar sequence
of acquisition of the 11 morphemes they tested. They found
that "the sequences of acquisition of 11 functors obtained
for Spanish and Chinese children are virtually the same"
(Dulay and Burt 1974, 49). From this they concluded that
there is "strong evidence that children exposed to natural
L2 speech acquire certain structures in a universal order"

and that it is the L2 system, not the L1 system, which is

4
responsible for guiding the L2 acquisition process (Dulay
and Burt 1974, 50-52). This last conclusion implies that L1
interference plays a smaller role in the language learner's
errors than was previously thought.

In a second study, Bailey, Madden and Krashen performed a
similar experiment on adult second language learners. Their
subjects were seventy—three adults (ages 17 to 55) from
twelve different native language backgrounds. They also
used the Bilingual Syntax Measure developed by Dulay and
Burt to elicit data and set out to test the following two
hypotheses:

(1) adults learning English as a second language will show
agreement with each other in the relative difficulty of
functors in English.

(2) the adult rankings will be similar to that of the
child learning English as a second language, rather than to
that of children learning English as a first language
(Bailey, Madden and Krashen 1974, 237).

Both of these hypotheses were supported by the study and
the results were very similar to the findings of the
Dulay/Burt study. The researchers also found that "while
casual observation affirms that errors due to mother tongue
interference do occur in second language learning in adults,
our data imply that a major source of errors is intra-
rather than inter-lingual, and are due to the use of
universal language processing strategies" (Bailey, Madden

and Krashen 1974, 242). On the basis of their findings,

5
especially the similar difficulty orders, they suggested the
use of a "natural syllabus” in formal (classroom) language
instruction. This idea was later rejected by Krashen (1985).

A 1978 study done by Larsen~Freeman, however, brought the
discovery of a natural sequence of acquisition into
question. She used five different tasks—-reading, writing,
listening, imitating and speaking (using the Bilingual
Syntax Measure)—-to elicit data from twenty-four adult
learners of English as a second language from four native-
language backgrounds. She hoped to discover, among other
things, if the same morpheme order would be found to exist
if different data collection procedures were utilized.

The results of this experiment showed that, although
native language background did not have a significant effect
on the accuracy order of English morphemes, the morpheme
sequence was not the same for all of the tasks. However,
the oral production tasks (speaking and imitating) did
produce morpheme sequences similar to that found by Dulay
and Burt. This finding brings up the question of
variability in interlanguage.

Anyone who has ever taught a second or foreign language
has inevitably noticed that a given student's performance
varies with different tasks. On a written examination, for
example, a student might do very well on the multiple-choice
grammar section, but lose many points on the essay portion.
Or one might be a very poor speaker, but do very well on

grammar exercises. Some of this variability in performance

6
is undoubtedly due to the difficulty of each of the tasks
involved, but much of it is due to the language learning
processes at work within the learner.

As Ellis (1986, 75) notes, at any given stage of
development, a learner's interlanguage system contains a
number of competing rules. One rule guides performance on
one occasion, while another rule may guide performance on a
different occasion. Variability which can be explained with
reference to either the linguistic or the situational
context of use is called contextual variability. Another
type of variability, individual variability, has to do with
individual learner differences such as age and motivation.
Finally, free variability consists of the learner's
haphazard use of two or more alternate forms which exist in
his interlanguage. For the purposes of this study we will
be primarily concerned with contextual variability.

Many theories of second language acquisition have
attempted to account for and explain variability in
interlanguage. The Monitor Model of second language
acquisition, developed by Krashen (1982), is probably the
most influential theory to be developed in the past decade.
Krashen's theory, which encompasses many areas of the
language acquisition process, has received both praise and
criticism from other researchers in the field. Rather than
examine the entire theory, we will deal only with the
Monitor hypothesis, which offers an explanation for

variability in interlanguage.

7

The Monitor Model posits two separate systems which
learners use to gain and utilize knowledge of a second
language. According to this theory, the process of
acquisition is responsible for initiating utterances while
the process of learning is responsible for modifying these
utterances to make them more correct grammatically. In
other words, learned features of a language are available to
the learner only as a Monitor, which may or may not be
employed by the learner depending on the communicative
situation.

Krashen states that the Monitor Model attempts to
account for discrepancies in oral and written second
language performance and that:

"the model predicts that the nature of second
language performance errors will depend on whether
monitoring is in operation. Errors that result
from performance based on the acquired system
alone will be consistent across
learners/acquirers, regardless of first language,
as acquisition is guided by universal principles.
Errors that result from situations in which
monitoring is possible will be more idiosyncratic,
as they will reflect each learner's conscious
mental representation of linguistic regularities
in the target language." (Krashen 1977,).

According to Krashen (1982, 16) three requirements are
deemed necessary for Monitor use: (1) sufficient time to
enact the Monitor, (2) explicit knowledge of the
grammatical rule, and (3) focus on the form of the message
rather than on the meaning. (Krashen (1985) retracts time

as a necessary condition for Monitor use. See discussion

below).

Krashen's theory makes intuitive sense when one applies it
to a typical language learner. When learners are
concentrating on expressing an idea, they are not as
concerned with correct grammar and thus their performance
contains more errors than it would if they were
concentrating on grammatical form. Also, if learners are in
a hurry or are not allowed much time to express an idea,
they will not have time to concentrate on grammar and will
therefore make more errors. If given sufficient time, one
would expect their performance to improve. Finally, one
would not expect language learners to be able to correct
their errors if they have not yet learned the grammar rule
in question. Studies have been done, however, which
challenge these seemingly logical assumptions.

Grezel (1986) reports finding that in some instances
subjects perform less accurately in certain linguistic
areas, e.g. morphology, in situations which should be
favorable for monitoring, e.g. a written proficiency
exercise, than they do in situations which should not be
conducive to monitoring, e.g. a dialog. He attempts to
explain this phenomenon, which he calls "negative
monitoring," in terms of cognitive development as well as in
terms of Krashen's hypothesis of individual variation in
Monitor use.

Hulstijn (1982) investigates the effect of sufficient time
and focus on form on the correct use of two Dutch word-order

rules. He also examines the relationship between explicit

9
rule knowledge and correct rule application. Since the
present study is, in part, modeled after the Hulstijn and
Hulstijn study, we will examine this study in more detail.
The experiment by Hulstijn and Hulstijn involved 32 adult
learners of Dutch as a second language from nine native
language backgrounds. The two rules which they focused upon
were inversion of subject and finite verb in main clauses
(INV) and the placement of the finite verb in final position
in subordinate clauses (VF).

In the first part of the study, the subjects were asked to
perform an oral story-retelling exercise under four
different conditions. In the IF condition
(Information/Fast) they were asked to pay attention to the
information which they had to reproduce and at the same time
they were instructed to speak as fast as they could. In the
IS condition (Information/Slow) they could take as much time
as they wished, still concentrating on information. In the
GF and GS conditions (Grammar/Fast and Grammar/Slow)
subjects were instructed to pay as much attention as
possible to grammatical correctness with and without the
time constraint.

The subjects listened to stimulus texts (in Dutch) through
headsets. They were then required to retell the content of
the texts in Dutch. They were also shown a phrase which
would force the use of the required type of sentence
structure (either INV or VF) and were told to begin their

response with this phrase. For the two fast conditions, the

lO
experimenter told the subjects to respond as quickly as
possible and used a stopwatch to measure the amount of time
taken. In the slow conditions, no stopwatch was used and
the subjects were advised to take as much time as they
needed. In the two information conditions, the subjects
were told to concentrate on the contents of the stimulus
texts and the experimenter informed them that their
responses would be scored on the basis of information only.
In the grammar conditions, the subjects were told to pay
attention to their grammar and were informed that their
responses would be scored for grammatical errors but not for
information errors.

From this experiment the researchers found that
concentration on grammatical correctness improved both INV
and VF performance, but the presence of time pressure had no
effect at all. They concluded that "Focus of Attention on
grammar resulted in better performance, which is indicative
of increased monitoring, irrespective of the Time Pressure"
(Hulstijn and Hulstijn 1984, 34).

The second part of the study investigated the relationship
between explicit or nonexplicit knowledge of a grammar rule
and the successful application of this rule under different
circumstances. Immediately following the first part of the
experiment, the subjects were interviewed in order to assess
their ability to explicitly verbalize the two word-order
rules as well as their ability to judge whether a sentence

with the corresponding grammatical structure was correct or

11
incorrect and why. The latter was done in order to test the
subjects' nonexplicit knowledge of the rule. The results of
this interview were then compared with the results from the
first part of the experiment, specifically with the
differences in correct scores between the I/F condition
(presumably the condition under which the subjects would
have the greatest difficulty producing correct forms) and
the G/S condition (the condition under which the subjects
would probably have the least difficulty producing correct
forms).

The researchers found that only a few of the subjects had
explicit knowledge of the INV rule and even fewer had
explicit knowledge of the VF rule. In general, the subjects
with explicit rule knowledge outperformed the subjects
without such knowledge. However, when rule knowledge was
compared with the percent difference between the I/F and 6/3
conditions, they found that improvement in performance for
learners with explicit rule knowledge was generally not
greater than for learners without this knowledge. In terms
of grammatical accuracy the researchers concluded that "the
learners lacking explicit knowledge did not profit less than
the learners with explicit knowledge from the absence of
time pressure and from a focus on grammar" (Hulstijn and
Hulstijn 1984, 39). This finding is contrary to Krashen's
claim that knowledge of the grammar rule is necessary for
successful Monitoring to occur.

In a pilot study which I completed in March, 1989, I also

12

noticed performance which runs contrary to that predicted by
Krashen's Monitor Model, specifically the predictions
concerning the effect of concentration on grammar, the
presence of time pressure and knowledge of the grammatical
rule on learners' grammatical accuracy. The study compared
the grammatical accuracy of the performance of eighteen
first-year students of German at Michigan State University
in three communicative situations. My hypothesis was that
they would perform best in the situation which was most
conducive to monitoring, and worst in the situation which
was least conducive to monitoring.

The experiment, which focused upon seven verb—ending and
verb-placement rules in German, consisted of four parts.
The subjects participated in an oral interview in which they
were first asked several questions dealing with everyday
topics (e.g. school, holidays, leisure activities, etc.) and
were then asked to describe two pictures from a magazine.
This activity was termed "casual speech" (CS) since the
focus was on communicating information and no mention was
made of grammar, with the exception that the subjects were
asked to use complete sentences when possible. Immediately
following the interview, the subjects were asked to write
responses to similar questions as well as to write a
description of two more pictures. This activity was termed
"casual writing" (CW). Again, no mention was made of using
correct forms, but the subjects were asked to use complete

sentences. The final two parts of the experiment were

13

conducted a few days later. For this task, the subjects
were given a series of sentences to translate. The
sentences were grouped according to the grammar feature
being tested, and following each section the subjects were
asked to state the rule for that grammar feature in their
own words. These two activities were called "formal
writing" (FW) and "knowledge of rule" (KR) respectively.

The interviews were then transcribed and all of the data
were analyzed and scored for the following grammatical
features: conversational past, modals, placement of the
verb following subordinating conjunctions, placement of the
verb following coordinating conjunctions, future tense,
general conjugation of verbs in the present tense, and
placement of the verb in second position. The subjects‘
renditions of the rules involved were then assessed as to
their correctness and completeness. The expanded hypothesis
was that: l) the subjects would perform most accurately in
the translation exercise, 2) they would perform least
accurately in the interview, and 3) the subjects would
perform "somewhere in between" in the casual writing
exercise. It was also hypothesized that subjects without
explicit knowledge of the rule involved would not perform
significantly better in the tasks which were more conducive
to monitoring than they did in the CS task, while the
subjects who knew the rule would perform better in these
situations.

In many cases the results of the experiment confirmed the

14
hypotheses. There were some interesting exceptions,
however. In the cases of verb placement after coordinating
conjunctions and present tense verb conjugation, the
subjects performed most accurately in the CS condition. In
cases of future and verb-second placement, they performed
most accurately in the CW condition. In addition, there
were several cases of subjects who could not articulate a
certain rule, but who produced this grammar feature more
accurately in conditions conducive to monitoring, as well as
subjects who did articulate a rule, but produced the
corresponding grammar feature most accurately in the CS

situation.

3. Objective and Hypothesis

The Monitor Model has been the focus of much debate in
recent years, but little testing has been done to determine
whether the hypotheses of the Monitor Model can be supported
by empirical evidence. Although the findings of the 1989
pilot study are interesting, it would be premature to draw
any conclusions from them. Because the subjects produced
very few examples (either correct and incorrect) of some of
the grammatical features, it is difficult to predict how
they would perform if they were to produce more examples of
these features. Also, pronunciation was not scored in the
CS condition. Even if it had been scored it would be
difficult to determine whether the variations in

pronunciation reflected errors in the grammatical system of

15
the subjects' interlanguages. For these and other reasons,
it was advisable to conduct a follow-up study done in order
to make the results of the pilot study more valuable and
significant.

There is also a need for a follow—up to the study done by
Hulstijn (1982). The findings of Hulstijn's study were
significant enough to prompt Krashen to alter his Monitor
Model. Krashen (1985) no longer includes time as a
necessary condition for Monitor use. A willingness to
modify one's theory on the basis of new information is
commendable, but Krashen's reaction to the Hulstijn study is
a bit surprising simply because it is based solely on the
work of these two researchers. It seems that more
investigation into the relationship between time pressure
and Monitor use would be necessary for the theory to be
altered. The purpose of the present study is, therefore, to
provide some additional research in the area of Monitor use
in the presence of time pressure and in the relationship
between explicit rule knowledge and Monitor use in various
communicative situations. More specifically, this study
examines the written performance of beginning German
students in situations which should be conducive to
monitoring with situations in which monitoring should be
difficult. The goal is to determine the degree to which
they are indeed able to monitor their speech. It also
compares the results of this experiment with Hulstijn's

(1982) findings.

16

The hypotheses which will be tested in this thesis are:
(1) Subjects' grammatical accuracy will be similar in the
presence of time pressure and when time pressure is removed,
(2) Subjects' grammatical accuracy will be greater when they
are told to concentrate on grammar than when they are told
to concentrate on information, (3) The grammatical accuracy
of subjects who know the grammatical rule will improve more
when the focus is on grammar than will the grammatical

accuracy of subjects who do not know the rule.

4. Methodology

For this study I decided to use written data in order to
compare the results obtained from different modes of speech
(Hulstijn and Hulstijn used oral data) as well as to avoid
ambiguities as to the accurateness of responses due to
pronunciation errors. The grammatical rule focused upon in
this experiment is subject—verb ending agreement in German.
This grammar feature was chosen because, in a pilot study
performed in 1989, I found that accuracy for verb
conjugation (including both subject-verb ending agreement
and correct verb-stem formation) was actually higher in
situations which should not be conducive to monitoring than
in situations which should favor monitoring.

The subjects for the present study were thirty-six first—
year students of German at Michigan State University. The
study aims to answer the following four questions: 1) Do

the subjects perform differently in the presence of time

17
pressure than in the absence of time pressure? 2) Do the
subjects perform differently when concentrating on
information than when concentrating on grammar? 3) What is
the relationship between explicit knowledge of the rule and
successful monitoring? 4) What is the relationship between
knowledge of the rule and performance on a discrete-point
grammar exercise?

The study was conducted as follows: The subjects were
collectively shown slides depicting scenes from everyday
life in Germany such as shopping, eating, studying, etc. A
total of twelve slides were shown, three slides for each of
Tasks 1— 4. A brief description of each of the slides is
given in Appendix A, page 30. For Task 1, before each slide
was shown, the subjects were told to describe the slide,
trying to write as much about the picture as they could.
They were also told that they had only one minute to
complete this task. For Task 2, the subjects were asked to
describe each slide, again trying to write as much about the
picture as they could. This time they were told that they
could take as much time as they needed, and that the
experimenter would call time when it looked as though
everyone had written a sufficient amount. The purpose of
Tasks 1 and 2 was to get the subjects to concentrate on
information in the presence and in the absence,
respectively, of time pressure.

For Task 3, the subjects were asked to describe each

slide in writing, and to make their descriptions as

18
grammatically correct as possible. They were also told that
they only had one minute to complete this task. For Task 4,
the subjects were asked to describe each slide in writing
and to make their descriptions as grammatically correct as
possible. 'This time, they were told that they could take as
much time as they needed, and that the experimentor would
call time when it appeared that everyone had written a
sufficient amount. The purpose of Tasks 3 and 4 was to get
the subjects to concentrate on grammar in the presence and
in the absence, respectively, of time pressure.

Following this part of the experiment the subjects were
given a fill—in—the-blank exercise in which they were asked
to correctly complete twenty German sentences by filling in
the correct form of a German verb using English cues (Task
5). This was done to satisfy Krashen's (1982) notion that it
might take an activity as focused on grammar as a discrete-
point grammar exercise to activate the Monitor. Finally,
the subjects were asked to fill out several verb—endings
charts in order to assess their explicit knowledge of the
verb-ending rule (Task 6). A duplicate of the forms and
instructions used in the experiment described above is given

in Appendix B, pages 31—40.

Summary of Tasks:

Task 1—-time pressure present, focus on information
Task 2——time pressure absent, focus on information
Task 3——time pressure present, focus on grammar
Task 4——time pressure absent, focus on grammar

Task 5——discrete-point grammar exercise

Task 6——verb—ending chart

19

5. Scoring

The responses were then scored in the following manner:
Only subject-verb ending agreement was scored and only verb
endings were evaluated for accuracy. If the stem of a verb
was misspelled, or if the wrong verb was used, it was not
counted as wrong as long as the ending was correct. A verb
was not scored if the subject could not be determined. The
number of correct verbs for each task was divided by the
total number of verbs (both correct and incorrect) for that
task in order to get a "percent correct" for each task.
This procedure was also used for the fill—in-the-blank
exercise and the verb—ending charts. If a word was not
filled in for the fill—in-the-blank exercise, it was not
scored, because it was possible that the subject simply did
not know the correct German equivalent of the English cue.
If a blank was left empty for the verb-ending charts,
however, it was counted as incorrect, since the German verb
was provided. After the scores for each individual subject
were calculated, mean scores for the group were calculated.
A table of the raw scores for each subject is given in

Appendix C, page 41.

6. Results
The scores were then organized in order to compare
performance under time pressure with performance without

time pressure, and performance when concentrating on

20

information with performance when concentrating on grammar.
The scores for Tasks 1 and 3 were averaged and compared with
the average of the scores for Tasks 2 and 4 to examine the
effect of time pressure on grammatical accuracy. The scores
for Tasks 1 and 2 were averaged and compared with the
average of the scores for Tasks 3 and 4 to examine the
effect of focus on information and focus on grammar on
grammatical accuracy. While Hulstijn (1982) used only group
mean scores for his comparisons, I will use both group mean
scores and individual scores in order to get a more detailed
look at how the subjects performed on the various tasks.

First the subjects' performance with time pressure and
without time pressure will be examined (Table l). Hulstijn

(1982) found that the presence of time pressure had no

effect on grammatical accuracy. In my study, this was
partially confirmed, as well. The mean of the scores for
Tasks l and 3 combined was 86.7. The mean of the scores for

Tasks 2 and 4 was 87.4, meaning that, the presence of time
pressure had no effect on the subjects' levels of
grammatical accuracy when viewed collectively. However, if
individual scores are examined, one finds that most of the
subjects performed differently in the presence of time
pressure than they did when time pressure was removed. The
difference is not uniform, however, in that some subjects
performed better (had greater grammatical accuracy) when
time pressure was removed, some performed worse when time

pressure was removed, and some performed essentially the

21
same under both conditions. Thirteen subjects scored five
or more percentage points higher in the absence of time
pressure than they did in the presence of time pressure.
Twelve subjects scored five or more percentage points lower
in the absence of time pressure than they did in the
presence of time pressure. Eleven subjects had scores which
varied by fewer than five percentage points in the two
categories. Thus, for roughly one—third of the subjects the
presence of time pressure had a positive effect on
grammatical accuracy, for one-third of the subjects it had a
negative effect on grammatical accuracy, and for roughly
one-third of the subjects it had no significant effect on

grammatical accuracy.

Table I: Summary of findings for Tasks 1-4:

 

Group Means

Focus On Focus on

Info. Grammar

Task 1 Task 3 Mean
With Pressure 87.4 86.0 86.7

Task 2 Task 4 Mean
Without pressure 86.0 88.7 87.4
Mean 86.7 87.4 0.7=diff.

 

 

 

The results for the difference in the subjects'

22
performance when concentrating on information and when
concentrating on grammar were similar to the results for the
difference in performance when under time pressure and when
time pressure was removed. Again, the mean of the scores
for Tasks 1 and 2 was 86.7 and the mean of the scores for
Tasks 3 and 4 was 87.4, suggesting that concentration on
grammar had no effect on grammatical accuracy. Out of the
thirty-six subjects, twelve scored at least five percentage
points higher when concentrating on grammar as opposed to
concentrating on information. Twelve subjects scored at
least five percentage points lower when concentrating on
grammar and the scores of twelve subjects varied by fewer
than five percentage points in the two categories. Thus,
for one-third of the subjects, concentration on grammar had
a positive effect on grammatical accuracy, for one-third it
had a negative effect, and for one— third it had no
significant effect on grammatical accuracy. These results
are different from the findings of Hulstijn, who concluded
that focus on grammar had a positive effect on grammatical
accuracy. However, Hulstijn notes that subjects whose
scores tended to be higher (and who also tended to have more
explicit knowledge of the grammar rule) improved less than
those subjects whose scores tended to be lower, and states
that this may be due to the fact that there is less room for
improvement for these subjects. This phenomenon could play
a role in the present study, too, as the group means are

already quite high. If one examines individual scores one

23
notes that as the subjects' explicit knowledge of the
grammar rule increased, their scores when focusing on
information and grammar tended to be more consistent (Figure
I). Figure I shows the difference between the mean of the
scores on Tasks 3 and 4, in which subjects were
concentrating on grammar, and the mean of the scores on
Tasks 1 and 2, in which subjects were concentrating on
information, plotted against their knowledge of the grammar

rule.

 

 

O

QT~+ ——+— A—I— —: -1,
i
I

9
ll) 2 z 9
lm l O
pmvcment w ICl'I 5
concentrating on 6 x O :
mmmar v ' "“ 1 §
.- l
‘J I e 9 z
:04 9 .
I
3 Q
751 9
. i t
-30 i Q
80 52 84 {)5 EB 90 ‘32 94 9G 93 103

Knowledge of the rule

 

 

 

Figure I: Scatter Plot showing improvement when
concentrating on grammar versus knowledge of the rule.

The next item considered was the relationship between the
subjects' explicit knowledge of the grammar rule and their

ability to successfully monitor their writing. To

 

24
accomplish this the scores for Task 1, in which time
pressure was present and the focus was on information, were
compared with the scores for Task 4, in which time pressure
was absent and the focus was on grammar. The percentage
difference between these two scores was compared with the
subjects' explicit knowledge of the rule using the Pearson
Product Moment correlation. (Ellis (1986, 125) summarizes
the Pearson Product Moment Correlation as being a
statistical procedure used to establish the the degree of
fit, or correlation, between two sets of measurements
relating to two separate variables. It helps the researcher
to establish whether changes in the measurements of one of
the variables is related to changes in the measurements of
the other variable. It is important to note that the
Pearson Product Moment Correlation is used to describe the
relationship (or lack thereof) between two variables, but
does not in any way suggest that a change in one of the
variables causes a change in the other variable.) The
correlation coefficient for Tasks 1 and 4 was -.113, which
suggests almost no relationship between knowledge of the
grammar rule and successful monitoring under monitor-
conducive situations (Figure II). Hulstijn's study produced

similar results in this comparison.

25

 

r I
1.2 ‘1
i O
1 4
0.8 7:
0.6 4
Percent Dilicrcnce E .
Between Tests III 0.4 i, 6
and 4 i
f 0.2 i Q. 9 Q
l o o O
O 4— ——r—+—- MI- ~——I————I——~1~——°———¢-—‘— ‘ :— z
m l 6 9 6 9
l
fifie
I) 75 60 ”Q :0 rt 80 85 90 95 1?
Knowledge of the Rule

 

 

 

Figure II: Scatter Plot for the percent difference between
Tasks l and 4, versus knowledge of the rule.

The final relationship which was examined was the
relationship between knowledge of the grammar rule and
performance on the discrete—point grammar exercise (Task 5).
This was done to test Krashen's notion that it might take an
exercise as focused on grammar as a discrete-point grammar
exercise to activate the Monitor. Using the Pearson Product
Moment formula the correlation coefficient for this
comparison is .545, which suggests a modest, positive

relationship between these two variables (Figure III).

 

 

 

100 ‘r . O
9’ ,i o o 9
e2: ; 3 o o
p: I o A
1.) -; 2 v
W.) : g G
Discrete PuiuL 7,. i g g 9
Grammar Test ‘ i
w :9
7" 1 9
i .
60 7‘1 Q
i
J‘) T 9
so .1. —.-.§~_- _4- _ .
30 03 6O L5 7L ”1 83 n, 93 93 100
KIMWk-dgv of the Rule

 

 

Figure III: Scatter Plot for performance on discrete point
grammar test verus knowledge of the rule.

7. Conclusions

From these data the following conclusions can be drawn: 1)
As Table l and Appendix C show, both collectively and on an
individual basis, there is no correlation between the
presence or absence of time pressure and grammatical
accuracy. 2) As Table l and Appendix C show, both
collectively and on an individual basis, there is no
relationship between grammatical accuracy and concentration
on grammar versus concentration on information. 3) As
Figure II shows, there is no correlation between explicit
knowledge of the grammar rule and successful monitoring when
other conditions favor monitoring. In other words, the

performance of subjects with explicit knowledge of the rule

 

 

27
does not improve more under monitor—conducive conditions
than that of subjects without explicit knowledge of the
grammar rule. 4) As Figure III shows, there is a moderate,
positive correlation between explicit knowledge of the rule
and performance on the discrete—point grammar exercise. The
first conclusion agrees with the findings of Hulstijn, who
found that the presence of time pressure had no effect on
grammatical accuracy. This study also found no correlation
between the presence of time pressure and grammatical
accuracy. The second conclusion, however, does not agree
with the findings of Hulstijn. Whereas he concluded that
concentration on grammar results in higher grammatical
accuracy than does concentration on information, our data
show that there is no correlation between concentration on
grammar versus information and grammatical accuracy. The
third conclusion also agrees with Hulstijn's findings, since
in both studies explicit knowledge of the grammar rule did
not result in increased improvement when conditions favored
monitoring. Finally, the fourth conclusion implies that,
since there seems to be a moderate, positive correlation
between knowledge of the grammar rule and performance on the
discrete—point grammar exercise, it might indeed take a task
as focused on form as a discrete—point grammar exercise to

activate the Monitor.

8. Discussion

Before examining the implications of the results of this

28
study, some of the weaknesses of the study shall be
examined. These weaknesses could have skewed the results of
the study, causing them to be incongruent with the results
of the Hulstijn and Hulstijn study and with the hypotheses
of the Monitor Model.

One of the weaknesses of this study was the fact that the
mean percentage correct for all of the tasks was quite high
(86% or higher). This left very little room for improvement
as the tasks became more and more conducive to monitoring.
The high percentage of correct verb endings suggests that
the subjects had nearly mastered this grammar feature within
the framework of these types of writing exercises. Oral
data or data utilizing listening or reading comprehension
might produce entirely different results. For follow-up
studies one might use a different, more advanced grammatical
feature which would allow more room for improvement when
monitoring. Another possibility would be to screen the
subjects for the study, as Hulstijn and Hulstijn (1984) did,
in order to test only those who had not mastered the
specific grammar feature being studied.

Another weakness of this study (and others like it) is
that one has no way of knowing whether the Monitor was
indeed activated in the conditions which were monitor—
conducive. Given the high percentages of correct verb
endings, it is also possible that the subjects in this study
monitored their speech in all of the conditions, even those

in which monitoring would be more difficult. We must be

29
careful that our conclusions reflect this uncertainty as to
whether the Monitor is activated in the various
communicative situations.

A third, and perhaps the most important weakness of the
study is the fact that the study was conducted with a very
small sample of students who were at a beginning level of
language proficiency. In addition, since the first-year
German courses at MSU are all taught using a communicative—
based teaching style in a natual approach, the subjects may
have had a disproportionate amount of practice in speaking
skills. In order for any far—reaching conclusions to be
drawn, more studies on subjects with varying levels of
proficiency and from various teaching—methodology
backgrounds must be performed.

Finally, as with any study, if more data had been obtained
from each of the subjects, a greater difference in
performance on each of the tasks might have been produced by
the subjects. Perhaps a larger-scale study or a more in-
depth study on fewer subjects would have produced results
different from ours.

Despite these weaknesses, there are some important
implications of this study. The fact that there was such a
low correlation between the subjects' knowledge of the
grammatical rule and their improvement when conditions
favored monitoring could give language teachers some insight
into the performance and progress of their students. Quite

often teachers assume that once a student has learned a

30
rule, his or her performance concerning that particular rule
should be almost error—free, especially on tasks in which
they are allowed as much time as they need and are told to
be careful about grammar errors. If the students make
errors even though they know the rule, many teachers
conclude that they simply were not being careful. The
results of this study suggest that this kind of thinking may
be flawed, at least when it comes to beginning learners of a
foreign language. Students at this level of foreign
language proficiency may not yet be capable of applying the
grammatical knowledge that they have to communicative

situations.

9. Further Study

As stated earlier, although much debate has surrounded
Krashen's (1982) Monitor Model, very little empirical
testing has been done to determine its validity. The
Hulstijn and Hulstijn (1984) study, the Grezel (1986) study
and the present study represent just a fraction of the
research possibilities concerning the subject of monitoring
and the Monitor Model. In addition to the suggestions posed
when discussing the weaknesses of this study, further
research directly related to the present study would include
a similar study of more advanced subjects to see whether
their monitoring skills were more advanced than those of
beginning students. Or one could test the subjects' ability

to monitor their own written performance in a written self-

31
correction exercise. This type of study would shed some
light on students' ability to find and correct their own
errors on a longer writing passage based on their knowledge
of the grammatical rule being tested. Or one could look at
subjectS' monitoring ability when discussing or writing
about different topics or to different audiences to
determine the role that inhibition plays in the ability to

monitor speech.

10. Conclusion

The present study was conducted in order to discover
whether the hypotheses of Krashen's (1982) Monitor Model
would be supported by research. Although the results of our
study did not directly support the notion that sufficient
time, knowledge of the grammatical rule and focus on grammar
are necessary in order for language students to monitor
their speech, they also by no means proved this notion to be
false. Krashen's Monitor Model is a difficult theory to test
because it deals with processes occurring deep within the
human brain and fails to consider such factors as individual
learning style and learners' previous exposure to the target
language within the context of various teaching methods.
Nonetheless it is still a valuable theory which has
dramatically changed the field of foreign and second
language acquisition as well as the majority of foreign
(second) language classrooms in this country. To fail to

test this theory further before abandoning or building upon

32
it would be a disservice to the field of foreign and second

language acquisition.

33

Appendix A

Following is a description of each of
slides used in Tasks 1 through 4.

Task 1

Slide 1: Two women standing in front of an
concession stand. A sign reading
frites" hangs near the window.

outdoor
“Pommes

Slide 2: A classroom with several college-aged

students sitting at their desks.
instructor is not visible.

The

Slide 3: Two men sitting at a table at an outdoor
cafe. A waitress is standing there,

waiting to take their order.

Task 2

Slide 1: Two women standing outside a fruit
market. They are looking at a display of

fresh fruit outside the market.

Slide 2: A man and a woman walking down the street
conversing. A child 5 or 6 years old

walks with them.
Slide 3: A family at the dinner table.

Task 3
Slide 1: A meat market. Several customers
waiting to be served.

are

Slide 2: A young person (gender unclear) in the

kitchen loading a dishwasher.

Slide 3: A man and woman sitting at the table
about to eat. The man is asian and they

appear to be eating asian food.

Task 4
Slide 1: Two women standing outside a fish

market.

A sign saying "Fische" hangs outside the

door.

Slide 2: A man and two children standing outdoors
next to a car. The man is about to pick

up one of the children.

Slide 3:Four people sitting at an outdoor
restaurant, talking and laughing.

34
Appendix B

PICTURE DESCRIPTION

Task One: Write as much as you can about the
picture. One minute time limit.

1.

35
Appendix B

Task Two: Write as much as you can about the
picture. No time limit.

1.

36
Appendix B

Task Three: Concentrate on correct grammar. One
minute time limit.

1.

Appendix B

Task Four:

time limit.

1.

37

Concentrate on correct grammar.

No

38
Appendix B

Task Five

Please fill in the blanks with the appropriate
form of the German equivalent of the English verb
provided.

Ein Geschaft

1. Zwei Manner (come) in das
Geschaft.

2. Sie (want) neue Kleider.

3. Ein Mann (finds) ein blaues Hemd.
4. Das Hemd (is) nicht sehr teuer
(expensive).

5. Er (buys) das Hemd.

Ein Restaurant

1. Eine Frau und ihr Mann (go) ins
Restaurant.

2. Sie (have) einen groBen Hunger.
3. Die Frau ______________(sees) den Kellner.

4. Die Frau (speaks) mit dem
Kellner.

5. Der Mann und die Frau (wait) auf

das Essen.

Auf der StraBe

1. Viele Autos (are) auf der
StraBe.

2. Ein Junge (stands) an der Ecke
(corner).

3. Ich (think), er

(waits) auf den Bus.

4. In funf Minuten (is) der Bus da.

39
Appendix B

Mein Geburtstag

1. Heute (is) main Geburtstag.
2. Meine Mutter (makes) einen
Schokoladekuchen.

3. Mein Bruder und meine Schwester
(have) Geschenke fur mich.

4. Mein Vater (brings) ein groBes
Paket nach Hause.

5. Ich (eat) ein wunderbares
Abendessen.

In der Kneipe

Karl: "Ich (would like) ein Bier."

Lotte: "Du (can) kein Bier haben, du
(are) noch nicht 21.

Karl: "Ja, aber ich (have) das
Ausweis (I.D.) von meinem Bruder. Er

(is) schon 23."

Lotte: "Ach, gut. Wieviel (costs)
ein Bier? Ich (want) auch eins."
Karl: "Ich (pay for) das. Du

(have) nicht so viel Geld."

 

Appendix B

Task Six

Please complete the
sagen

ich sag___
du sag___
er sag___
wir sag___
ihr sag___
Sie sag___

sein

Vielen Dank!

40

following verb charts:

finden wollen
ich find ich
du find du
er find___ er
wir find wir
ihr find ihr
Sie find Sie
haben
ich
du
er
wir
ihr
Sie

41
Appendix B

Bowman Experiment—-Instruction Sheet

Please read the following instructions aloud
before administering the experiment.

For this experiment you will be asked to
describe a number of slides. Before you are shown
each slide, you will be told what you are to
concentrate on when writing your descriptions.

For some of the slides you will be given a
specific time limit in which to write your
descriptions. For the other slides, there will be
no time limit. The instructor will stop you when
it appears that each student has written a
sufficient amount. For each slide, please try to
keep writing until the instructor tells you to
stop. You may be as creative or factual in your
descriptions as you like. All of your descriptions
should be in German.

An example of a creative description might
sound like this: Here is a bakery. Mrs. Jones is
in the bakery. She wants to buy an apple strudel
and a loaf of bread. The clerk says they don't
have any more strudel today. Mrs. Jones becomes
very angry and says she will not come to this
bakery any more.

An example of a factual description might
sound like this: Here is a bakery. There are
three people in the bakery. One woman is about 35
years old. She is wearing a yellow shirt and blue
jeans. She has her son with her. He is wearing
green pants and a red shirt. The other person is
an old man. He is fat.

In some instances you will be asked to
describe the scene from th point of view of one of
the people in the slide. For this task you may
want to write a conversation. Your description
may sound like this: I am in the bakery. I see a
man with a white shirt and blue pants. I think he
needs help with his package. I ask him: "do you
need help?" He says: "Yes, please."

Don't worry if you have trouble with
vocabulary or if you think your German isn't good
enough. This exercise was meant to assess the
performance of students at your level of
competence. Your descriptions don't have to be
very elaborate. Just do your best. Don't forget
to write in German! Thank you!

42
Appendix B

Bowman Experiment
Instructor's Sheet

Please read the following instructions before
showing the respective slide. You may repeat the
instructions if necessary.

Slide #1

Try to say as many things about this picture as
you can. You will have only one minute to write
this description.

Slide #2

Try to say as many things about this picture as
you can. You will have only one minute to write
this description.

Slide #3

Try to say as many things about this picture as
you can. This time, describe the picture from the
point of View of the man in the dark blue jacket,
in other words, you are the man in the dark blue
jacket. You will have only one minute to write
this description.

Slide #4

Try to say as many things about this picture as
you can. This time, you will have no specific
time limit. The instructor will call time when it
appears that most of the students have written a
sufficient amount.

Slide #5
Try to say as many things about this picture as
you can. You will have no specific time limit.

Slide #6
Try to say as many things as you can about this
picture. This time, describe the picture from the

point of View of the woman in the black-and—white
striped shirt. You will have no specific time
limit.

Appendix B

Slide #7

As you describe the following slide, please pay
attention to your grammar. Try to make your
sentences as grammatically correct as you can.
You will have only one minute to write your
descrptions.

Slide #8

Again, please pay attention to your grammar. You
will have only one minute to write your
descrptions.

Slide #9

Again, please pay attention to your grammar. This
time, describe the picture from the point of view
of either the man or the woman in the picture.

You will have only one minute to write your
descriptions.

Slide #10
Again, please pay attention to your grammar. This
time you will have no specific time limit. The

instructor will call time when it appears that
most of the students have written a sufficient
amount.

Slide #11
Again, please pay attention to your grammar. You
will have no specific time limit.

Slide #12

Please pay attention to your grammar. This time,
describe the picture from the point of View of the
man with the beard. You will have no specific time

limit.

\OOOQONLDIDWNH

Appendix C

Info/

Fast
Task 1

75
83
67
80
100
67
100
88
91
100
100
60
100
75
82
100
100
100
43
82
100
86
100
100
75
100
100
67
83
100
82
88
89
92
93
100

44

Info/

Slow
Task 2

50
91
89
33
100
87
89
83
86
93
100
90
75
77
89
94
83
82
91
89
100
85
94
100
90
91
100
82
55
76
95
64
100
100
94
100

Gram/
Fast
Task3

6O
71
88
86
83
67
100
67
100
100
100
78
100
54
63
88
75
100
67
86
88
56
90
100
100
100
83
100
100
100
89
57
100
100
100
100

Gram/ Dis.
Slow Test
Task 4 'Task 5
64 62
92 97
83 79
73 72
100 76
83 76
93 86
86 85
95 93
94 97
100 93
92 79
100 97
67 9O
80 93
79 90
8O 9O
92 9O
91 76
92 83
100 77
75 92
100 97
92 100
85 88
100 100
92 100
92 69
100 54
100 69
93 93
77 50
89 8O
79 82
93 100
93 93

Ending
Chart
Task 6

8O
96
92
52
96
80
92
88
96
92
84
88
92
92
84
92
92
100
88
88
92
84
100
100
84
100
96
84
56
84
96
88
100
88
100
96

45

Works Cited

Bailey, N., C. Madden, and S. Krashen. 1974.
‘Is there a "natural sequence" in adult
second language learning?‘ Language
Learning 24: 235—44.

Burt, M., H. Dulay, and E. Hernandez. 1973.
Bilingual Syntax Measure. New York:
Harcourt Brace Jovanovich.

Corder, S. 1971. ‘Idiosyncratic dialects and
error analysis.‘ International Review of
Applied Linguistics IX: 149-59.

Dulay, H. and M. Burt. 1974. ‘Natural
sequences in child second language
acquisition.‘ Language Learning 24:
37-53.

Ellis, R. 1986. Understanding Second Language
Acquisition. Oxford: Oxford University

Press.

Grezel, J.E. 1986. ‘Stil und Stadium in der
Zweitsprachent— wicklung: Zwei Erklarungen
fOr negatives Monitoring.‘ Osnaercker

Beitrage zur Sprachtheorie 34: 36-45.

Hulstijn, J.H. 1982. Monitor use by adult
second language learners. Ph.D.
dissertation, University of Amsterdam.

Hulstijn, J.H. and W. Hulstijn. 1984.
‘Grammatical errors as a function of
processing constraints and explicit

knowledge.‘ Language Learning 34: 23-43.
Krashen, S. 1977. ‘The Monitor Model for adult
second language performance.‘ In Burt,

Dulay and Finocchiaro (Eds.), Viewpoints
on English as a Second Language. New York:
Regents.

Krashen, S. 1982. Principles and Practice in
Second Language Acquisition. Oxford:
Pergamon.

Krashen, S. 1985. The Input Hypothesis:
Issues and Implications. New York:
Longman.

 

 

46

Larsen—Freeman, D. 1978. ‘An explanation for
the morpheme accuracy order of learners of
English as a second language.’ Language

Learning 26: 125-135.

Nemser, W. 1971. ‘Approximative systems of
foreign language learners.‘ International
Review of Applied Linguistics IX: 115—23.

Selinker, L. 1972. ‘Interlanguage.'
International Review of Applied Linguistics
X: 209—30.

 

 

 

I’IICHIGQN STQTE IV LIBRQRIE‘

II222I III IIIII "III I I I

31293008964540

- . ..‘w.a.-<.‘1.-..
a."

..m-v~.-..-..‘...n
"mu-(~-

.. .4." “'- . ton-«~21-
.._...».w u
w

'5..7."'.....
.......u3..'..
.'\--ug .1. "“33
...r ....,.......‘..'...
u :4..~...--..‘
. ”may“...
.. .4 .p .‘.
._. .. . .. . .rg...
‘l. I‘IIOIIIIUIIYA' 9 '2'. " : i Il-erlull a“: l'l'l.‘i‘i::.|
37"...”.4..‘.....L.a...... .0. .3"... ~ ..-_..,\'.'.u..2.
~-n-:.:,.w '.‘ ll'l"::i" ---\l'.\'.
. .. ._:....:w

~ “may
‘ .

«(man-4: .
a,_u,.-.~~.
~u .: q-n-
...uX. luv-t

.-- n-.~I‘c‘~- va-

. .
., .w u -- up... I." ,. .',,.,..."..2, ..,"'...-..2'.'..
I-OV t that, >--- u. ‘wl'l'l‘ll‘ |.~\W-~z.y....-
.. .-.. . .- .- ... . - . .,.,~ .
wiv'x. ..v. - u .-. . .
. v 1..
.
. . ‘
”law-n ...- v."-.1.yu."2‘, . - . .
"nut-v.1...- .-.--..:....'..... »:-":“-.-.....
4." .Wuuu \ H“.
n

 

|
u- .a .-
:.‘.,..;;‘1..:; nu‘uau‘: 0
‘.'III1!\IIVl-~|[lnll
0v-
1... u
:‘.n:lrlw.l\.l
..""