Grammatical correctness and communicative ability : a performance analysis of the written and spoken English of Swedish learners

(1)

Inger Bergström

Grammatical Correctness

and Communicative Ability

A Performance Analysis of the Written and Spoken English of Swedish Learners

Umeå 1987

(2)

(3)

(4)

Bergström, Inger: Grammatical Correctness and Communicative Ability. A Per formance Analysis of the Written and Spoken English of Swedish Learners

Department of English, University of Umeå Umeå 1987. x + 223 pp. Monograph

Acta Universitatis Umensis. Umeå Studies in the Humanities 81 ISSN 0345-0155

ISBN 91-7174-285-9

Distributed by Almqvist & Wiksell International, Box 638, S—101 28 Stockholm, Sweden

Abstract

Written and oral material produced by a group of low-achieving learners of English from the 2-year lines of the Swedish upper secondary school was analysed from the perspective of grammatical correctness and communicative ability. The gram matical analysis focussed on the verb phrase and tests included both free produc tion in speech and writing and elicitation tests. Communicative ability was assess ed ‘ ‘objectively* ’ by identifying such parameters as fluency, copiousness and span, and “ holistically” by using non-expert evaluators.

The scores thus obtained were correlated. Grammatical correctness was cor related with communicative ability both in speech and in writing and writing profi ciency was correlated with speech proficiency with respect to both grammatical correctness and communicative ability.

Our findings are that there is a positive correlation between grammatical cor rectness and communicative ability. A remarkable finding is that the percentage of correct verb phrases correlates very weakly with communicative ability in written data. In oral data, the correlation is in fact slightly negative. The learner’s com petence in grammar is reflected in both his written and oral performance. On the other hand, there is no correlation between communicative ability in writing and communicative ability in speech.

The study shows that a working command of a set of syntactic rules is essential for communication. Errors are, however, an integral part of the learning process. The major part of errors are accounted for by the learner’s use of compensatory strategies. Among these low-achievers, communicative ability in conversation is distinct from writing ability.

Key words: grammatical correctness, communicative ability, performance analysis, written and spoken English, classroom instruction, low-achievers, language learning, interlanguage, compensatory strategies, errors, verb phrase, parameters, systematicity, variability, writing proficiency, speech proficiency, elicitation test.

(5)

Acta Universitatis Umensis

Umeå Studies in the Humanities 81

Inger Bergström

Grammatical Correctness

and Communicative Ability

A Performance Analysis of the Written and Spoken English of Swedish Learners

Umeå 1987

(6)

The printing of this volume has been supported by Acta Universitatis Umensis

Editor of the series Per G. Råberg

The Faculty of the Humanities, University of Umeå, S—901 87 Umeå

Printed by Goterna, Box 230, 442 23 Kungälv ISSN 0345-0155

(7)

PREFACE

Having completed this study my sincere thanks are due to those who have, in various ways, helped me to carry it through.

First and foremost I owe my gratitude to my research advisers, Professor Alvar Ellegård and Professor Mats Rydén.

Professor Ellegård has generously and unselfishly guided me from the start to the finish by making detailed, judicious and relentlessly accurate com ments on the many draft versions. His inspiring enthusiasm combined with his unfailing interest in and expert knowledge of language pedagogy have greatly contributed to making my work an exciting and absorbing under taking.

Professor Rydén has provided constant encouragement, assistance and seminal criticism. I am grateful for our many valuable and rewarding discus sions especially on such issues as grammar and categorization.

I also owe my gratitude to Widar Henriksson who advised me on statistical matters, Pat Shrimpton who corrected my English and made several useful suggestions and Marianne Nejati who willingly typed the manuscript so effi ciently despite short notice.

I would like to thank the members of the seminars at the Department of English, Umeå University, for their polite attention to the presentations of preliminary versions of this work.

My work has been made possible by the kind cooperation of colleagues, pupils, and exchange students at Östra gymnasieskolan, Umeå. I am deeply indebted to them all.

A special word of appreciation goes to my friends Inger and Sten Henrysson who have helped me in every possible way, giving both personal and professional support.

Finally, I want to express my heartfelt thanks to my family, Svante, Pia and Anna. Thank you for being helpful, encouraging and loving.

Umeå, April 1987

(8)

(9)

GRAMMATICAL CORRECTNESS

AND COMMUNICATIVE ABILITY

A Performance Analysis of the Written and Spoken English

of Swedish Learners

AKADEMISK AVHANDLING

som för avläggande av filosofie doktorsexamen vid universitetet i Umeå kommer att offentligt försvaras i Humanisthuset, hörsal F,

fredagen den 22 maj 1987 kl. 10.15

av

Inger Bergström

(12)

Bergström, Inger: Grammatical Correctness and Communicative Ability. A Per formance Analysis of the Written and Spoken English of Swedish Learners

Department of English, University of Umeå Umeå 1987. x + 223 pp. Monograph

Acta Universitatis Umensis. Umeå Studies in the Humanities 81 ISSN 0345-0155

ISBN 91-7174-285-9

Distributed by Almqvist & Wiksell International, Box 638, S—101 28 Stockholm, Sweden

Abstract

Written and oral material produced by a group of low-achieving learners of English from the 2-year lines of the Swedish upper secondary school was analysed from the perspective of grammatical correctness and communicative ability. The gram matical analysis focussed on the verb phrase and tests included both free produc tion in speech and writing and elicitation tests. Communicative ability was assess ed “ objectively” by identifying such parameters as fluency, copiousness and span, and “ holistically” by using non-expert evaluators.

The scores thus obtained were correlated. Grammatical correctness was cor related with communicative ability both in speech and in writing and writing profi ciency was correlated with speech proficiency with respect to both grammatical correctness and communicative ability.

Our findings are that there is a positive correlation between grammatical cor rectness and communicative ability. A remarkable finding is that the percentage of correct verb phrases correlates very weakly with communicative ability in written data. In oral data, the correlation is in fact slightly negative. The learner’s com petence in grammar is reflected in both his written and oral performance. On the other hand, there is no correlation between communicative ability in writing and communicative ability in speech.

The study shows that a working command of a set of syntactic rules is essential for communication. Errors are, however, an integral part of the learning process. The major part of errors are accounted for by the learner’s use of compensatory strategies. Among these low-achievers, communicative ability in conversation is distinct from writing ability.

Key words: grammatical correctness, communicative ability, performance analysis, written and spoken English, classroom instruction, low-achievers, language learning, interlanguage, compensatory strategies, errors, verb phrase, parameters, systematicity, variability, writing proficiency, speech proficiency, elicitation test.

(13)

2.2.3. The present p e rfe c t...93 2.2.4. Modal auxiliaries...94 2.2.4.1. F u tu rity ...94 2.2.4.2. V olition...96 2.2.4.3. C o n d itio n ...97 2.2.5. Q uestions...98

2.2.5.1. Questions requiring rfo-support... 98

2.2.5.2. Questions not requiring c/o-support... 98

2.2.6. Negated sentences...100

2.2.6.1. Negated sentences requiring d o -su p p o rt... 100

2.2.6.2. Negated sentences not requiring tfo-support... 100

2.3. S u m m ary ...101

3. The communicative perspective... 101

3.1. Introduction...101

3.2. M ethods...103

3.2.1. Number of words per utterance (W /U)... 104

3.2.2. Number of words per topic (W /T)... Ill 3.2.3. Number of points of information per utterance (P i/U )... 117

3.2.4. Percentage of successful exchanges per conversation (% Es)...122

3.2.5. Percentage of English words per conversation (% E n)...131

3.3. S um m ary...133

3.4. Correlation between group mean parameter scores and communicative a b ility ... 138

3.5. Validation of the m easures... 141

3.5.1. D e sig n ... 141

3.5.2. R esu lts...142

4. Correlation between correctness and com m unication... 143

4.1. In tro d u ctio n ... 143

4.2. Total number of verb phrases (Number V P s ) ...144

4.3. Percentage of correct verb phrases (Percent correct V P s ) ... 144

4.4. C onclusions...147

V CORRELATION BETWEEN WRITING PROFICIENCY AND SPEECH PR O FIC IEN C Y ... 148

1. In tro d u ctio n...148

2. Grammatical correctness in written and oral d a t a ... 148

(14)

VI ANALYSIS OF RESU LTS... 156

1. In tro d u ctio n ...156

2. Written d a t a ...157

3. Oral d a t a ... 158

VII SUMMARY AND CONCLUSIONS... 160

APPENDICES ... 163

(15)

Chapter I

INTRODUCTION

1. Aims

In the present study I intend to investigate the written and oral perfor mance in English of a group of Swedish learners exposed to formal classroom instruction in English as a foreign language. I wish to shed some light on such issues as:

— What is the relationship between grammatical competence and com municative ability?

— Is the “ good” performer in writing also a “ good” performer in speech?

— Are classroom learner’s errors in grammar systematic in the same way as are learner’s errors in a natural learning environment?

2. The teaching background

Awareness of the need for adequate language instruction has long characterized educational programmes in Swedish schools. All pupils, in cluding those who are low-achievers are expected to learn English to some extent and to leave school with a reasonable working command of the language.

There are indications (e.g. Balke-Aurell & Lindblad 1980) that several of the ambitions concerned with the acquisition of competence have not been achieved. Nor have the high expectations been fulfilled. This may seem in compatible with the findings of the IEA project (Malmberg 1975), which show that the teaching of English in Swedish schools has been fairly suc cessful, seen in an international perspective. It should be pointed out, however, that all the Swedish students involved in this project were drawn

(16)

from non-vocational “ lines” (p. 7).1 However, about 30% of all Swedish students between the ages of 15—18 are to be found in vocational lines (SCB 1986:1).

There are, unquestionably, problems among certain groups within the school population and a fairly substantial proportion of the students do not seem to attain an efficient working knowledge of the language. It is obvious that we still do not know the ‘ ‘best” way to teach a foreign language, and that the teaching that we do provide is not particularly well-suited to some sections of the school population. The far-reaching organizational reform of the Swedish upper secondary school (Sw. gymnasieskolan2) in 1971 may have had a restraining influence on the development of satisfactory second language (L2) competence. This reform brought together into one educa tional unit the “ vocational” school, the “ continuation” school, and the traditional upper secondary school. It gave rise to problems in the teaching situation as new student populations were integrated into an educational system traditionally based on academic achievement. The problems have attracted attention from scholars who, among other things, have pointed out the pupils’ lack of progress both in their native language and in English (e.g. Axelsson 1977, Grogarn 1979, Balke-Aurell & Lindblad 1980).

3. Language learning

Much of the ineffectiveness of formal L2 teaching/learning may derive from insufficient knowledge of how language learning actually proceeds. Likewise, we know little about the dynamics of the interaction between the learner and his environment. There is also reason to believe that the learner’s own con tribution to the whole learning process has been greatly underestimated.

Some questions have, however, become clearer and recent research findings point in certain directions. First of all, language learning is a developmental process. Even if the stages in this process are as yet not fully understood we can discern a systematicity in the acquisitional pattern of both the LI and L2 learner. A learner’s language has come to be viewed as a system worth study ing in its own right (Corder 1981:5 ff.). Selinker (1972:209 ff.) coined the term ‘ ‘interlanguage’ ’, generally used as a collective term for the learner’s own ver sions of the target language, which reflect his (implicit) hypotheses about the language he is learning (cf. Corder 1981:66). Studies of phases in a language learner’s development, whether they are called “ interlanguage”, “ idiosyn cratic dialects” (Corder 1981:14), “ approximative systems” (Nemser

1 Course programmes at the Swedish comprehensive upper secondary school (Duchenfield 1981:17) (the word “ line” will be used henceforth).

(17)

1971:115 ff.), all claim that a learner’s idiosyncratic sentence is grammatical within the framework of the learner’s developmental grammar.

An increasing interest in interlanguage has brought into focus error analysis, that is, the study of those elements and structures that deviate from the set norm. A learner’s errors are evidence of his interlanguage systems, and a study of these errors will provide insights into the processes of language learning.

In recent years, psycholinguistic research and theory has undergone a ma jor shift, moving away from issues concerning grammatical competence towards issues related to communicative competence and certain aspects of performance carefully avoided by earlier scholars. Hitherto, little has been written on the nature of precisely what it is that must be acquired by the learner to make him capable of using the target language effectively. Judging from most existing textbooks it seems that the focus of L2 instruction is still mainly on formal criteria, and does not involve genuine communication but rather artificial language-like behaviour.

Against this background, a study of the linguistic performance of a somewhat special group of L2 learners in a formal setting may contribute to an improvement in current teaching practices.

3.1. “ Natural” language learning vs.

formal classroom instruction

As stated above, the data in this study were produced by low-achieving learners exposed to formal classroom instruction. A distinction is made be tween classroom instruction and language acquisition in a “ natural” en vironment which includes all kinds of language learning for which no formal provision is made through teaching. “ Natural” language learning thus im plies acquisition without any specific instruction partially outside the con scious awareness of the learner in actual communicative situations (Stern 1983:19). In a natural language environment, the learner constantly has to cope with the target language in various communicative situations offering him opportunities to acquire the language which are not very different from those pertaining to the acquisition of the first language in infancy. In the L2 classroom, as a rule, the focus is on a systematic and analytical study of the target language guided by teaching (Stern 1983:391 ff.).

Formal classroom instruction must be seen as an artificial way of develop ing language proficiency. It is a process in which supposedly more is needed than simply exposing the learner to the target language. A number of background factors are influential, such as psychological factors and educa tional treatment. The existence of an increasing number of learners in the non-theoretical lines with unsatisfactory educational backgrounds, means

(18)

that it is necessary to find out more about the specific learning capacities which distinguish the low-achievers from the high-achievers. It is only natural that for biological, psychological and other reasons, students should learn language in different ways.

During the late sixties and early seventies certain new developments in language pedagogy occurred which can be regarded as emanating from the view of language as a “ mentalistic” process (Chomsky 1957). The “ creativi ty’ ’ of language was stressed and increasing attention was paid to the learner’s role in the acquisitional process. Several books which appeared in the 70s indicated that the focus was now on the learner and the cognitive processes involved (Oiler & Richards [eds.] 1973).

These studies were mainly carried out with regard to “ natural” second language learning; the emphasis was thus on language acquisition in a natural environment. Collection of empirical data in the classroom tended to be neglected. Consequently language teaching research came to lack direction and there has since been little discussion on research emphasis, nor has a distinct research methodology established itself.

Language teaching in the classroom tends to be a matter of guesswork based on common sense and teacher experience, rather than of knowledge based on research. A better understanding of how language learning comes about is an essential component of any language teaching theory.

3.2 Acquisition vs. learning

Krashen (1977, 1981, 1982) introduced two concepts in L2 learning, namely “ acquisition” and “ learning” . “Acquisition” denotes the subconscious pro cess similar to a child’s way of acquiring his first language, by which L2 learners acquire a language in genuine communicative situations, and ‘ ‘learn ing” denotes a controlled classroom activity which involves explicit instruc tion and consciousness of target language rules on the part of the learner (Krashen 1982:10). Krashen’s acquisition/learning dichotomy has not, however, been taken for granted in this study. Research to date seems to have produced inconclusive evidence of a distinction between acquisition and lear ning, and to me it remains an open question whether the processes underlying various manifestations of linguistic development are in fact qualitatively dif ferent. Since we know very little about the nature of the input, or the way in which it has been transmitted to these learners, we can only say that some, perhaps most, acquisition or learning has been stimulated by teaching, but much is of course independent of any teaching.

Under the concept of “ language learning” I thus subsume both first and second language “ acquisition” and “ learning”.

(19)

4. The present investigation

My analysis is based on written and spoken material produced by a group of low-achievers (see pp. 7—8) selected from the 2-year lines of the Swedish equivalent of the upper secondary school. These students are in general a pro blem as regards academic performance, including English (as taught today). This naturally implies that one must be very careful about drawing conclu sions from this study for language teaching and learning in general. Moreover, although it would have benefited my investigation to have known more about the input factors that influenced the achievements of these learners in English, the original design of my research did not include data about classroom teaching in general or about the exposure of the students to English outside the classroom. These questions are thus outside the scope of this investigation.

The empirical study consists of two parts. The first part is an analysis of written data, that is, essays written by the student group referred to above. In cluded in this part is an analysis of results in an elicitation test (pp. 45—58). The second part contains an analysis of oral data produced by the same group in interaction with peers who are native speakers of the target language. The material has been subjected to a twofold analysis—grammatical and, by means of certain parameters, communicative. The scores obtained by the students in both respects have been correlated. First, the learner’s gram matical competence has been correlated with his communicative ability both as regards writing and speech. Secondly, the learner’s writing proficiency has been correlated with his speech proficiency, both as regards grammatical competence and as regards communicative competence.

(20)

Chapter II

DESIGN OF THE STUDY

1. Introduction

English is the dominant second language in Swedish schools, being part of the curriculum for all Swedish children from the ages of 9 or 10 to 16, which is the official school-leaving age. The major part (c. 90 %) (SCB:U 50 SM 8601) of the school population voluntarily continue to the upper secondary level (Sw.

gymnasieskolan*), which consists of 3-year2 and 2-year lines.

There are five 3-year lines (including the 4-year technical line) and 22 2-year lines. The 3-year lines correspond to the old academic “ gymnasium” . Three of the 2-year lines are “ theoretical” (the social, economic and music lines) with curricula similar to those of the 3-year lines, whereas the remainder are “ vocational”.

English is a compulsory subject in the 3-year and “ theoretical” 2-year lines but optional in the vocational lines. In the 2-year lines there is a choice bet ween a “ general” and a “ special” course (for an analysis of the difference in proficiency between pupils in the 2-year and 3-year lines, see Balke-Aurell & Lindblad 1980).

My analysis is based on written and spoken material produced by a special ly selected group of low-achieving learners mainly from the 2-year vocational lines (Appendix 1).

In Swedish secondary schools the theoretical and vocational lines tradi tionally recruit their pupils from different socio-economic groups (Härnqvist & Svensson 1980:55 ff.). In the school (Östra gymnasieskolan, see below) where the data for this study were collected the building which houses the

1 The comprehensive upper secondary school for 16— 19 age group. In Britain nearest equivalents are open access sixth form* tertiary college and in the United States senior high school (Duchenfield 1981:32).

(21)

vocational lines is an integral part of the school campus. The social boundary between theoretical and vocational lines is, however, comparatively rarely crossed.

My data consist of: (1) essays, (2) elicitation test scores, and (3) conversa tions. Three native speakers of English participated in the conversations (see below).

2. Subjects

The subjects were students drawn from different classes in the 2-year lines (Appendix 1). They were on average 17 years old and at the time of the final tests they were in the second year of their 2-year course. They had studied Eng lish for 8 years and on the last testing occasions they were in their 9th year of study. All of them were selected from a group of low-achievers in English (as taught and assessed in Swedish schools today) singled out by means of a dia gnostic entrance test. This test was (and still is) administered to the cohort of pupils on the 2-year and 3-year lines at the beginning of their first year at the upper secondary level. It is customary to measure pupils’ achievements in vocabulary and grammar in the initial stages of new teaching programmes in English. A test for this purpose designed by the GUME group (Levin 1972) is widely used in Swedish schools at various stages of instruction.

I studied the test scores from a GUME diagnostic test (Appendix 2) admini stered to the whole cohort of beginners at Östra gymnasieskolan in Umeå1 in 1981,1982, and 1983. My subjects were chosen from those starting courses in the 2-year lines in 1982. Mean test scores were compared to see whether the 1982 scores were consistent with those of 1981 and 1983 (see Diagrams 1—3). The test consisted of 80 items distributed equally on vocabulary and gram mar, with a maximum score of 80 points, i.e. one point for each correct answer (Appendix 2). On each of the test occasions (1981 through 1983) there seemed to be a fairly constant number of pupils who scored 30 points or below. If the test had any diagnostic value at all there was reason to believe that the weak ones would be found in this group. At the end of one term’s work I then checked the pupils’ marks as set by their teachers (1 through 5, where 1 indica tes the lowest level of achievement) and it emerged that the pupils graded 1 and 2 both in the 2-year and 3-year lines were invariably found among the low- scorers (scoring 30 points or below). We may conclude that the test predicts success or failure in terms of the standards by which a pupil’s proficiency in English is assessed in our schools today. Likewise, the test items are likely to

1 This school is an upper secondary school with 1553 pupils (in 1982) distributed into 3-year lines (785 pupils), 2-year non-vocational lines (335 pupils), and 2-year vocational lines (433 pupils).

(22)

be representative of what is focussed on in current English classroom instruc tion (cf. Balke-Aurell & Lindblad 1980:30).

Test results were obtained from a total of 1863 pupils (1981 through 1983). The 3-year lines and the “ special” courses in the 2-year lines produced, as ex pected, the highest scores. On the whole, the test scores for the 2-year and the 3-year lines differed considerably (Diagrams 1—3).

The pupils in the 2-year lines are usually split up into groups for instruction consisting of a maximum of 30 learners drawn from several different class units. The composition of each teaching group is usually conditioned by the pupils’ choices of vocational courses. The classes in the 3-year lines are undif ferentiated and the average number of pupils is 30. I have used the term “ group” in the diagrams to denote both the composite groups of the 2-year lines and the undifferentiated classes of the 3-year lines.

The subjects for my empirical study were selected from amongst 84 pupils taken from different instructional groups, who were beginners in the 2-year lines in 1982 and who had diagnostic test scores of 30 points or below. Practi cal considerations determined the choice of 35 pupils from this low-achieving group, 8 from amongst those taking the “ special” course and 27 from those taking the “ general” course. Of these 35 pupils 19 were girls and 16 boys (Appendix 1). Oral data from 31 pupils in this group were obtained in inter action with three native speakers of English (see below).

3. Written data

In an empirical study of the type presented here, one usually has to choose between material obtained from discrete-point proficiency tests and commu nication tasks, sometimes called integrative proficiency tests (Spolsky 1978). Integrative approaches to proficiency testing are based on the assumption that language skills are best evidenced and evaluated in contexts where the learners use the target language for communicative purposes. Examples of integrative proficiency tests include tests such as written compositions and oral interviews. The learner is expected to focus on the message, i.e. on what he says rather than how he says it.

Discrete-point proficiency tests are composed of test items, each of which addresses a learner’s skill in controlling certain surface rules of language nor mally related to grammar and/or lexicon. These “ controlled data” are ob tained by means of various elicitation techniques, e.g. tasks of transforma tion, substitution, sentence completion. These tasks have no communicative functions for the learner, who is expected to focus on the linguistic rules requi red to perform the operation.

Integrative communication tasks are assumed to give insights into a learner’s developing grammar. Discrete-point tasks, on the other hand, may

(23)

Diagram 1. Diagnostic test scores 1981 (GUME).

Group mean scores

instruction 3-year lines (mean score for the full cohort 57.0)

— 2-year lines (special course) (mean score for the full cohort 44.4) 2-year lines (general course) (mean score for the full cohort 28.2)

(24)

Diagram 2. Diagnostic test scores 1982 (GUME). Group mean scores

8 0

7 0

5 0

4 0

-3 0 &#-34;

20 -1 0

-Group of instruction — 3-year lines (mean score for the full cohort 59.4)

— 2-year lines (special course) (mean score for the full cohort 49.6) — 2-year lines (general course) (mean score for the full cohort 29.9)

(25)

Diagram 3. Diagnostic test scores 1983 (GUME). Group mean scores

8 0

7 0

6 0

-

504 0

-

30-20

-10

-Group of instruction — 3-year lines (mean score for the full cohort 57.4)

2-year lines (special course) (mean score for the full cohort 48.1) 2-year lines (general course) (mean score for the full cohort 31.9) V = group mean score

(26)

at times give us information about the learner’s meta-linguistic awareness, i.e. his conscious knowledge of rules and forms of the target language.

For the written data in this study, I have chosen essays on selected topics. My assumption is that language produced in communication tasks relates to a learner’s overall level of language development (Adams 1978:277 ff., Bahns & Wode 1980:85 ff.) and is qualitatively and systematically different from that which is produced in a controlled elicitation situation.

A considerable amount of data is generally needed before the range of syn tactic features used by the informant is sufficient to allow assessment of his interlanguage level. Naturally, it is not possible to make any statements about constructions that do not appear in the data. This can be remedied by sup plementing integrative tasks with discrete item elicitation tests as I have done here (see pp. 45—58).

The subjects were asked to write three “ free” compositions of varying types on three different occasions over a period of eight months (Ap pendix 3):1

(1) an essay on a topic of the learner’s own choice (2) a summary in English of a story read out in English (3) a summary in English of a story read out in Swedish

4. Oral data

Our oral data stem from unprepared (open-ended) conversations between native speakers of English and 31 Swedish learners from the group described above.

The native speakers were three exchange students, two from the USA and one from Canada, age peers of the Swedish pupils belonging to the same school albeit not the same lines. The foreign students were pupils in the 3-year theoretical lines (see above).

At the time of the investigation the exchange students had spent a full term in Sweden. They knew a fair amount of Swedish and were fully capable of understanding and speaking Swedish. Naturally, this was unfortunate, all the more so as some of the Swedish students obviously felt uncomfortable in the unfamiliar situation and resorted to Swedish when they realized—as they generally did not at the outset—that the exchange students could understand them. Three of them resorted more or less totally to Swedish. Possibly, there would have been fewer Swedish expressions if the pupils had been forced to use English. Thus the communicative event described here turned out to be less “ natural” than it was meant to be. If possible, I would have preferred a

1 35 pupils wrote two essays each, 24 wrote three each, and three pupils produced four essays each.

(27)

material consisting of conversations between interlocutors belonging to the same social stratum at school and with native speakers unacquainted with the Swedish language. In the actual circumstances, the natural language of com munication would most likely have been Swedish, as that was the language that would have functioned best in the dyads.

However, it seems clear that most pupils tried hard to give their answers in English. Hence, the conversations may after all present a fairly accurate pictu re of their communicative capacities in the type of situation that we aimed at creating.

Data from 31 pupils (out of the 35 subjects involved in this study, pp. 7—8) were obtained. Four members of the original group had to be excluded for various reasons—two were absent due to illness, and two had obtained per- misson to leave school temporarily. The major part of these data were col lected at the end of the spring term of 1983.

The pupils were told that they were supposed to carry on a conversation in English with the exchange students. The foreign students were given a list of unprepared topics to talk about from which they were free to choose. The list proved useful in most of the conversations. It created some “ performance consistency” and made it easier to trace performance characteristics.

5. Procedure

The two types of data described above are analysed from two points of view: (1) the grammatical perspective, by examining all the verb phrases occurr

ing in our written (including elicitation test scores) and oral data in terms of their formation and the range of functions attributed to the various forms

(2) the communicative perspective, by identifying various parameters with a view to assessing the communicative quality of each learner’s written and oral production.

My assessment of the learner’s communicative ability is based on parameters considered to agree with a subjective, “ global” , assessment of the linguistic material. To put my numerical evaluation methods to the test the written and oral productions of some learners were holistically assessed by non-expert evaluators (pp. 79—81 and 141—143) to see if their judgements were reasonably in line with my measures of quantification.

Correctness in grammar has been correlated with communicative ability in our written and oral data. Finally, writing proficiency and speech proficiency have been correlated, as regards both grammatical correctness and com municative ability.

(28)

Chapter III

WRITTEN DATA

1. Introduction

In most classrooms the learner’s attention is probably drawn to the gram matical rather than to the communicative properties of the language presented to him. Any infringement of the grammatical rules has traditional ly been regarded as a serious error in language performance, at least by language teachers. A learner’s language proficiency has generally been evaluated and assessed on the basis of his violations of the rules of grammar.

There are a number of reasons for this priority. There is, for instance, a grammar-translation tradition in our classrooms dating from the days when modern languages were first introduced into our school curricula (Bratt 1977:123 ff.), and use of the behaviouristic audio-lingual method entailed little beyond the performance of drills to practise grammatical structures (Stern 1983:324—327). Furthermore, most language textbooks even today are grammar-based, and often pursue a paradigm for the sake of completeness, regardless of whether or not it agrees with the communicative needs of the learner. In addition, the formal approach to language description has a long tradition manifested in innumerable volumes of grammar whereas the func tional, communicative approach has considerably less documentation. In spite of this emphasis on formal criteria in both language teaching and language description, the stated aim of current L2 teaching is to impart to the learner the ability to use the target language for real-life purposes (Läroplan för gymnasieskolan 1975:169). A crucial question here is what constitutes communicative competence in a given language. How does communicative competence develop in relation to linguistic competence? In what ways do structural and communicative factors interact and to what extent?

We find it comparatively easy to measure the learner’s “ formal” linguistic competence whereas an evaluation of his communicative ability rests on uncertain theoretical grounds.

(29)

written language the writer is primarily concerned with the transference of in formation, that is, the language used is “ message oriented,,. In addition, the writer has no access to immediate feedback being more or less reduced to imagining the reader’s reactions. The reader for his part can only guess at the writer’s intentions on the basis of their shared experience.

In speech, on the other hand, we often focus on the use of language to establish and maintain social relationships. However, the transmission of in formation is a factor common to both written and spoken communication and it is on this particular aspect of language that I have concentrated in my study.

In this section I will analyse our written data formally and from a func tional perspective. The aim is both to investigate the manner in which the learner puts his linguistic knowledge into practice and the correlation be tween each learner’s formal proficiency and his communicative ability as assessed and quantified in my analysis.

2. The Grammatical perspective

2.1. Introduction

For both the written and oral parts of my study, I have chosen to investigate the verb phrase as a representative measure of the learners’ linguistic com petence. I have two reasons for this choice. First, the verb phrase has an obligatory presence in most utterances, explicitly or implicitly. Second, much previous research has been done on how the verbal system is learned. Earlier error analysis research, however, often suffers from a fundamental methodological defect, in that the number of errors made by the learner has not been systematically compared with the number of correct forms of the ex pression he uses. An error analysis cannot be regarded as complete unless all instances of the learner’s non-erroneous realizations of the construction under consideration are also included in the analysis. It is impossible to iden tify the source of an error without knowing the full range of its occurrence in the learner’s production of that particular construction. Hence, if error analysis is to be used as a tool for measuring a learner’s linguistic competence it must be combined with an analysis of the total production (cf. Wagner- Gough 1975:155 ff.).

In the present study, a controlled elicitation test was administered to the learners for the purpose of validating the findings in the essays. These test scores will be dealt with on pp. 45—58.

(30)

2.2. Essays

The following grammatical categories will be accounted for: the simple verb phrase

the copula the perfect modal auxiliaries the progressive

questions and negated sentences

I have tabulated all verbs as part of a finite verb phrase for each pupil (Appendix 4). The verb phrases appear in main clauses as well as in subclauses. All the subclauses occurring in the learners’ essays are accounted for in Appendix 5.

Each learner has been given an identification number from 1 to 35. In ex amples from the essays an identification number is linked with a letter refer ring to the particular essay in which the example is found (Appendix 3). This reference system will be used henceforth.1

The total number of verb phrases registered in our written data is 2703, distributed among the learners as shown in Appendix 4. A comparatively large number of these verb phrases contain violations of various grammar rules, that is, they are functionally non-standard2 in the context in which they appear. A non-standard variant may consist of

(1) a formally correct verb phrase used in an inadequate context (2) a verb phrase, both formally and contextually incorrect.

Examples of variant (1):

The past tense appears in a context in which the reader wbuld naturally expect the present to be the writer’s choice of tense. I have arbitrarily chosen to regard that choice as the present, if the present tense is used in more than 50 °7o of the total number of verb phrases in a particular context.

(15a) When I came home fromschool I started to do my homework. After the dinner I play piano sometimes. I write letters, read a book, and listening to music. Three days a weak I gymnastics to. If I had time on the Tues day I go bath and hope from the trampolens (= diving boards). On the Sundays I ran four kilometers. I hate to be sick (I am sick now) becouse then I can’t do all this.

The phrases in italics are in the past tense. The remainder of the verbs in the essay are in the present tense; in a total of 19 simple verb phrases the present tense occurs 15 times, that is, more than 50%.

1 This system o f identification is also used in our oral data.

(31)

A modal auxiliary occurs in the wrong context:

(lib) . . . and he died when he should save Kristins’ life. The base is used for the 3rd pers. sing, pres.:

(27a) Her mother feel worry but her father think that the boy in the bushes only is a fantasy person. Christine call him Harry and talks with him everyday.

Examples of variant (2): The ing-form is used for the past:

(19c) We driving towards the town but the night was comeing very quick. So we slept over in my friends beatch buggy.

Some verb phrases are the learners’ own creations: (3b) The Godfather affittetet a people . . .

Mis-spellings of grammatical morphemes in the verb phrase based on pro nunciation have been overlooked. Hence, askt for asked\ happend for hap

pened, stopt for stopped, etc. have been regarded as standard variants (see fur

ther, pp. 20—21). On the other hand, overgeneralized forms of irregular verbs have been regarded as non-standard (cf. Taylor 1975:73 ff.). Diged for dug,

telldTor told, writed for wrote have thus been classified as non-standard. Mis

spellings of lexical words have been disregarded as long as the words are identifiable in the context in which they appear, for instance thougt for

thought, meat for meet, by for buy.

2.2.1« The simple verb phrase

The simple verb phrase, that is, the present or the past forms of the main verb (and the copula), is the most frequently required verb construction in our data: the present is required 469 times and the past 882 times in the context (in all 1351 exx.) (Table 1).

Table 1. The present and past tenses (copula excluded)

Function Standard Non-standard Total Present 258 (55%) 211 (45%) 469 Past 606 (69%) 276 (31%) 882 Total 864 (64%) 487 (36%) 1351 Table 2 shows how the standard functions of the present and past tenses are expressed by the learners. It is evident from the table that several functionally non-standard variants have been used to realize one tense in a context. Fur

(32)

thermore, some forms appear in more functions than others, such as the base and the wg-forms for instance.

Table 2. Learners ’ use o f variants to express the p resen t and p a st tenses

Standard function

Pupils’ forms

Simple

Base 5-marker past Be + ing Ing Other Total

% standard Present 3 sg 115 40 23 3 5 3 189 21 other 2 1 8 a) 7 9 12 18 16 280 78 Total present 333 47 32 15 23 19 469 59 Past 208 10 606 11 14 33 882 69 Total 541 57 638 26 37 52 1351 64 % standard 40 70 95 0 0 0

a) Boldface figures are realizations according to the standard.

The following variants are used to realize the present and the past tenses: (1) The base of the verb (541 exx.):

The base is the most frequent form. It occurs 218 times to express what should be, according to the standard, the present (non-third pers. sing.). In addition, it appears 115 times as a non-standard variant of the 3rd pers. sing., and 208 times for the past.

The unmarked base far outnumbers the s-marked verb for the 3rd pers. sing. (115 exx. compared to 40 correct forms) and is the predominant non standard variant for the past tense. These findings are consistent with the findings in the oral material (p. 88):

(5a) It was a mother and a dady and theirs dother she was adopte. The dother had a friend his names was Harry, and she talk to him evry day, but nobody could see him exept the dother. The mother think that Harry was straing so she must no how he was. She walk to the adoption agency and ask where hers dother come from. The lady to the adoption agency said at (= that) she come from London from a (slum kvarter (= slum area)). The mother walk there and there was a old lady like (= who) told her about Harry.

(2) The 5-marker (17 exx.):

The 5-marker is overgeneralized in functions other than the 3rd pers. sing, where, in accordance with our principles, we should expect the present tense (7 exx.) and the past tense (10 exx.):

(28b) On the winter we are some girls how plays ice-hockey and in the summer we plays football. I’m allso play tennis and runs.

(33)

(26b) I decided to go to a restaurant and eat dinner. Then I goes to the city and by the c o a t. . . Then I goes to a cinema and saw a movie . . .

(3) The past-tense marked verb occurs correctly 606 times (247 regular and 359 irregular):

In 32 cases the past-tense marker is used for non-standard realizations of the present tense, that is, the past tense is interpreted here as functionally non standard, according to our principles, since it appears in a context where the present predominates (cf. discussion above, p. 16):

(28a) Kristin is playing in the garden. She is talking to a rose bush and cold it Harry. Her mother is worried and cold her in. Her mother tells her father that Kristin has a fantasy friend but he is only laughing at it. (32a) Mrs James went to see where Christine has lived. When she is there she

see a white rose bush in the garden. A woman who is living on the first floor is looking out and ask what Mrs James is doing there.

The examples below illustrate the learners’ use of the past and fluctuations between various realizations of the tense. It should be pointed out that the learners almost invariably use the correct tense marking for the copula (p. 21 ff.) and the auxiliaries, whereas the lexical verbs (in focus here) are often unmarked for tense:

(16a) This story is about an adoptid girl. She is 5 year old. One day when she was in the garden she begin to talk whit a boy she could Harry. She saw Harry in the rose bush.

(19a) Sudden the mother think about the dother and she hurried to the shool but the girl was away, she ask the teather wher the girl was, he said his brother come and get the girl, and after that the family never sean the girl again.

(4) Be+ing (26 exx.) is used incorrectly for the present 15 times and for the past 11 times:

(8a) Every day Christin was talking to the rose bush . . . When he was jum ping he landed in the rose bush.

(20a) When she was beginning the school she didn’t want whitout her brother, the big rose bush out on the yard, she was talking with very often.

(5) The ing-form (without be) occurs 37 times, of which 23 realize what should be the present and 14 what should be the past:

(15b) When I not training I play, read books, go to movie and go out dancing. (13a) Latary (= later) miss Johns walking to the hous Cristine and hirs fami

(34)

(6) Other variants (classified under “ Other” ) occur 52 times:

(6b) My name is P.R. my age is 16 years old. Vm live in Umeå and Pm born here.

(13b) /racoram wteN ordm aling-Um eaevrydayitis5m ileswan(= one) way. Before I started in this line I going svets (= welding) in Vännäs.

V m + the base occurs 12 times and it’s -I- 5-marked main verb 3 times. Both

forms are here assumed to be reflections of “ chunk” learning from early stages of L2 learning (Hakuta 1974).1

Some verb phrases (5 exx.) contain redundant uses of was. Here, was may be seen as a past tense “ particle” or part of intended but disfigured be+ing constructions:

(20a) She’s brother died but Kristin was escaped.

(8a) One day her mother saw that Chris was talked to the rose bush in their garden.

Overregularized forms of irregular verbs appear 13 times:

(12a) But then we drived over a main (= mine) which had been buried in the hot sand.

(34a) They diged tanuls (= tunnels) under the ground . . . they kild 30 man and catched 22 man . . .

In 15 instances, a disfigured variant of an irregular verb realizes the past tense: (14c) A family in London, had a adoptiv girl, who names Christine. The girl,

spoken about a boy who names H a rry . . . The Mother gone to an adop

tion agency. There a women said “ Christine had a brother, who was dead, and he called H arry” .

Spelling of simple present and past forms

Various types of mis-spelling in the verb are registered below, in the present tense (13 exx. or 2 % out of 598 exx.) and in the past (103 exx. or 16 % of 638 exx.) (Table 3).

1 A learner’s use o f memorized sequences o f the language he is learning has also been discussed in terms o f “ formulas” (Jespersen 1924, Fillmore 1976), “ prefabricated routines and patterns” (Krashen & Scarcella 1978), “ linguistic routines” (Hymes 1971), “ fossilizations” (Selinker 1972), etc., as opposed to language generated either by interlanguage rules or by rules o f the target language.

(35)

Table 3. Spelling o f past tense forms

Correct Incorrect Total

Regular verbs 204 46 250

Irregular verbs 331 57 388

Total 535 103 (16%) 638

Four groups of mis-spellings may be distinguished:

1. Confusion of homophones; by for buy; no for know, sad for said. 2. Overgeneralization of spelling rules; plaid, staid by analogy with paid,

cryed for cried, planed for planned, stoped for stopped.

3. Idiosyncratic interpretations of L2 spelling; p o t for put, could, cold for

called, route for wrote, thold for told, tray for try; skrimed for scream ed, kome for come.

4. The ed-morpheme spelt phonetically; happend, kild (= killed), askt (= asked), lookt (= looked), playd{= played), cald (= called), pro-

mist { = promised), anserd(= answered).

Observations here seem, on some crucial points, to be in agreement with observations of developmental phases in the early stages of LI learning (Brown 1973, de Villiers & de Villiers 1973) as well as L2 acquisition in a natural environment (Bailey, Madden & Krashen 1974, Fathman 1975, Larsen-Freeman 1975, Rosansky 1976, etc.). The effects of the learner’s use of simplification strategies show up, for instance, in an extensive use of the un marked base to cover, in addition to its standard function, the 3rd pers. sing, pres, and the past tense (cf. Corder 1981:79—86, Schumann 1982:337 ff., Butterworth & Hatch 1978:245, Meisel 1980:13—41, etc.). The ing-form (with or without be) seems to be regarded as an unmarked variant of the base used for present and past tenses. Furthermore, there are tendencies to overgeneralization (e.g. irregular verbs inflected in the past tense on the pat tern of regular verbs). Some constructions may be interpreted as reflections of “ chunk” learning from early stages of L2 training (e.g. the use of Vm + the base of the main verb).

2.2.2. The copula

The description of the copula in non-complex verb phrases (be as part of a complex verb phrase is not dealt with here) follows the same principles as the analysis of the present and past tenses above. A fundamental difference be tween the use of the copula and the lexical verb is, however, that in their essays the learners almost invariably use the correct tense with the copula. This is

(36)

illustrated in the examples below, all of which are taken from the same learner (no. 32):

(32a) When she come to the school Christine wasin’t there. The teatcher told that Christine’s brother was cashing her.

(32b) Some days we were out for looking at the town and all the things around it. I have the camera with me and there were many things to photograph The last day was going to pack done all the things we have with oss ( = us) for the holiday. The most of oss were wery sad to have to leave the mountain, because it has been a funny holliday and all of oss have enjoy it.

The copula invariably takes the “ correct” tense, that is, the tense that is tex- tually predominant in that particular part of the essay.

Some of the cases of non-standard uses of the copula may be derived from the frequent Swedish use of the past tense form of the copula to express for instance a personal involvement of some kind (Svartvik & Sager 1977:88): (7b) I was in London now. The weather is very fine, the sun shining and the

the birds flig h t. . . It was very much with tourists here. If we have time over we shall go to piccidily cirkus . . .

(1) There are 267 occurrences of the present copula, of which 29 are non standard (Table 4) (Appendix 4).

Table 4. The use o f the presen t copula

Pupils* forms Standard

functions A m A re Is Other Total standard

2 sg 3 sg plur 100 Present Total 44 29 187 7 267 89 % standard 100 59 95 0

a) Boldface figures are realizations according to the standard.

Is is the most frequent form of the copula in the data (cf. Oral data p. 92)

occurring 187 times. Of these 10 are incorrect. A m appears 44 times (used cor rectly in all instances). Finally, are occurs 29 times, of which 12 are non standard. Examples of non-standard use of the present copula:

(37)

(Ila) Now after nine years she are dead and we have only good memory’s of her.

(13a) hiess (= his) adoption mum be worrid about this.

In example (13a) LI might have exerted some influence; be and the Swedish verb bli have a phonetic as well as a semantic resemblance.

(2) The total number of past copulas is 328, of which 53 are non-standard (Table 5) (Appendix 4).

The predominant past form is was, used 280 times in varying functions; 13 times for the 1st pers. sing., 238 times for the 3rd pers. sing., according to the standard, and, erroneously, in the plural 29 times. The plural of the past copula caused the learners problems. Were is required 53 times; on 29 occa sions there are violations of the S-V concord rule (Table 5).

Table 5. The me o f the past copula Pupils’

forms S ta n d a r d ^ ^

functions Was Were Other Total

% standard 1 sg 13a) 2 1 16 81 Past 2 sg 3 sg plur 238 29 19 24 2 259 53 92 45 Total 280 45 3 328 84 % standard 90 53 0

a) Boldface figures are realizations according to the standard. Examples of non-standard use of the past copula:

(16b) Her parents was in town on a p a rty . . . When the steps was in the house she take a knife and went to the door.

(24b) I didn’t found another name, so it just been “ Snutten” .

The copula is different from lexical verbs. It has little independent meaning and is therefore useful as a substitute for a lexical verb irrespective of its mean ing. It is also the first and probably the most frequently used verb in the initial stages of beginners’ courses, and its various forms are most probably learnt by heart as lexical items. Correct use of the copula, with its varying forms, re quires, however, an awareness of concord rules, rules which are non-existent in the learners’ native grammar.

Some of the learners (nos. (1), (7), (11), (16), and (20)) use the copula extens ively. For pupil (1) the copula appears more frequently than lexical verbs, no. (7) uses the phrase in 40 % of the total number of verbs, no. (11) 45 %, no. (16) 36 %, and learner no. (20) 49 %, respectively (Appendix 4).

(38)

2,2,3. The perfect

The present and past perfect appear 192 times in contexts where the construc tions are required.

(1) The most common non-standard variants are have/has/had/ + the base of the main verb (58 exx.) (Table 6):

(18a) oh dear I have forget to tell my name is Peter.

(27a) She run to school but the teatcher says that Christines brother already

have catch her.

(28b) When I come home from school the clock is five. And after I have eat I goes out meating frieands and sometime play ice-hockey.

(9a) When the mother come at school to fetch Kristin, she wasn’t there. The teatcher said that an boy had fetch her.

(26c) When her father come home late in the evning her mother said that Mary had talk with a person.

(2) H ave/has/had + the ing-form occurs 12 times (Table 6): (7b) To day we have eating in a reastaurang with very good food. (28a) Her parens had taking thers live.

(3) The perfect is realized by other, non-standard, constructions (16 exx.) (listed under “ Other” in Table 6), for instance by be + past + ing, by an in correct past participle of an irregular verb, or by a past-tense form of the verb: (17a) I looked at the watch and its was three so I get to the school but Kristin

wasint there and the teacher sad that her brother was feeching her. (32a) When she come to the school Christine wasin’t there. The teatcher told

that Christine’s brother was cashing her.

(27b) We lost against Furunäs, but who cares we had took the second place in the cup . . .

There are 5 instances of incorrectly formed past participles of irregular verbs (included under “ Other” ) based on the pattern of a regular verb, as in

braked (34d) for broken, taked (19b) for taken.

(4) H ave/has/had + the past participle (or the base) is used in functions other than the present or past perfect in 9 instances (Table 6):

(5b) I just forgett that I have disapperd mine money. And now have I travel I to Spaine, I ’m so happy. Becuase I have money.

The verb phrase in italics in (5b) seems to be used to realize the notion of Futurity.

(27a) Before Harry she has wanted to start school real bad but now she wants to be with Harry.

(39)

In (27a) the present perfect is apparently used in a context where the past tense is required, possibly under the influence of Swedish.

Spelling of the past participle morpheme

The spelling of the regular past participle morpheme varies (cf. the spelling of the simple past, pp. 20—21); in 5 cases the spelling reflects the pronunciation of the morpheme as in ju m p t (28a), (35b) for jum ped and kild (34d) for killed.

On the whole, the learners use the present and past tenses of the auxiliaries in contexts where they are required. The notion of present perfect is realized 45 times out of a total of 47 occurrences (96 %) and the past perfect 101 times out of a total of 145 occurrences (70%). The contrast between present and past may thus be said to have been observed. The second element of the con struction, the past participle, however, causes various problems as we have seen above. There are 58 instances out of a total of 201 (29%) of the base replacing the past participle in the present and past perfect. There are 17 in stances of irregular verbs used in this way: take (3 exx.), fin d (3 exx.), do (2),

eat (2), catch (2), forget (1), get (1), make (1), meet (1), sleep (1). There are 29

required past participles expressed by the base which consist of verbs that all correspond to Swedish verbs of the first conjugation (Hultman 1975:73). In these verbs the past participle morpheme is not usually pronounced in the Swedish dialect used by the learners. The learners’ first language may thus have been influential here. Furthermore, the past participle is replaced by the

ing-form 12 times, which probably indicates that the form has been regarded

as an unmarked variant here (cf. p. 21).

2.2.4. Modal auxiliaries

Modal auxiliaries are used to realize various concepts of Modality in 282 in stances, of which 91 are functionally non-standard (Appendix 6):

Futurity1 (108 exx., of which 45 are non-standard) Ability/Possibility (59 exx., of which 3 are non-standard) Volition2 (10 exx., of which 3 are non-standard)

Obligation/Necessity (25 exx., of which 9 are non-standard) Condition (63 exx., of which 25 are non-standard)

Other concepts of Modality (17 exx., of which 6 are non-standard) (Appendix 6).

1 Included here for practical reasons since ‘future time’ is usually rendered by modal auxiliaries and since the realization o f Futurity often contains a modal component.

(40)

T ab le 6. R ea li za ti o n s of th e pr es en t an d pa st p e rfe c t V. vo «O CO ^ cN r- vo I I I I IO t ^ I ^ I 'sO CO VO CO <N OS l/-.' CO VO rj-OiJ VI CO o v_{ej <U}'t: Ph Qh 5 <N rf V ,"H O (N VO o r- o 1 o vo o O o CO VO o <N o 00 vo vvD ov T—t »o Tf VO <N CO tT

1

TO

3 s

_o CÄ H 1 T3 § •5 o ’S_o c o •■a §> « I 73 «

(41)

§

o S è 'S

t

Sì

a

fScScii g X_co H 00 in rf rt O to IO 0\ ON v£> NO NO io co co on io so Tt (N CO NO IO NO 00 VO IO Tf »—i CO »-H <N I I t> Tf O I 00 I I IO 1—i CO I I I I o I I I I I co I I co I no I I co io O I I so <N Tt —< I l> I co <N <N I I I a i i i i i_CN o o <D £ o * e i •3 & 3 3 3 -2 tu < ö c o -e -2 o ^ .2 « ^ M O Ä o 5 > O u o ON 2 29 7 2 53 3 2 6 2 r-5 9 o VO ON 09 1f-h (N 09 3 2 00 7 4 5 2 7 6 (N (N T3 ctt TO To tal <Z) a) ve rb p h ra se , us ed ac co rd in g to th e st an da rd (st.) b) ve rb p h ra se , vi ol at in g so m e st an da rd ru le s (no n-s t.) c) in cl ud in g w an t to re al iz e th e co ncept

(42)

In the following I will comment on some of the most frequent realizations of Modality in our written material (Table 7).

2.2.4.I. Futurity

The concept of Futurity is expressed by the following markers:

/be/going to (22 exx., including 4 exx. of omitted ôe-auxiliary and

3 exx. of the variant gonna/gona)

will/would (25 exx.)

shall/should (1st pers.) (12 exx.) shall/should (non-first pers.) (18 exx.)

other choices (31 exx.)

In all, there are 45 examples of non-standard realizations of Futurity (Table 7). Examples of non-standard realizations:

(1) Would:

(9c) At the summerholiday I should work at a shop called “ Hem och Fritid”. But I should only work for four weeks. I work at this shop at the winter-weekends too. After I have work I would go to our sum merhouse.

In the example above the modal is expressive of something that is going to happen in the future. The auxiliary can be said to be a realization of Futurity. On the other hand, it seems as if the writer wanted to express something more than Futurity—there may be implications of hope and desire as regards the fulfilment of the event.

(2) Shall/should (non-first pers.):

(27a) Soon shall she start school, before Harry she has wanted to start real bad but now she wants to be with Harry. When she shall go to bed she smiles when she look out in the garden.

(5a) She should pick up hers dother three aclock. So she rush away to the school but she wasn’t there.

The frequent use of should in particular is most probably due to transfer of the Swedish modal skulle, which has a wider range of application than its English counterpart.

(3) The simple present (included under “ Other” ) (Table 7): (8b) Than the women said: — Okey I tell you her bakeground.

(11a) Maybe, sometimes in the future we get a new dog but it can never be in the same manner.

Grammatical correctness and communicative ability : a performance analysis of the written and spoken English of Swedish learners

Inger Bergström

Grammatical Correctness

and Communicative Ability

Umeå 1987

Acta Universitatis Umensis

Umeå Studies in the Humanities 81

Inger Bergström

Grammatical Correctness

and Communicative Ability

Umeå 1987

PREFACE

CONTENTS

GRAMMATICAL CORRECTNESS

AND COMMUNICATIVE ABILITY

A Performance Analysis of the Written and Spoken English

of Swedish Learners

Inger Bergström

Chapter I

INTRODUCTION

1. Aims

2. The teaching background

3. Language learning

3.1. “ Natural” language learning vs.

formal classroom instruction

3.2 Acquisition vs. learning

4. The present investigation

Chapter II

DESIGN OF THE STUDY

1. Introduction

2. Subjects

3. Written data

8 0

7 0

5 0

4 0

-3 0 &#-34;

20

-1 0

8 0

7 0

6 0

-

504 0

-

30-20

-10

4. Oral data

5. Procedure

Chapter III

WRITTEN DATA

1. Introduction

2. The Grammatical perspective

2.1. Introduction

2.2. Essays

2.2.1« The simple verb phrase

2.2.2. The copula

2,2,3. The perfect

2.2.4. Modal auxiliaries

1

3 s

§

t

a