Navigation – Plan du site

Non-native or non-expert? The use of connectors in native and foreign language learners’ texts

Agnieszka Leńko-Szymańska
p. 91-108

Résumés

La recherche décrite dans le présent article porte sur l’usage des connecteurs (mots et expressions de liaison) dans les textes écrits par des locuteurs natifs de l’anglais (auteurs professionnels et étudiants) ainsi que des apprenants avancés de l’anglais qui ont des langues maternelles diverses (français, espagnol, suédois, allemand, russe, polonais et finlandais). L’étude est menée dans le cadre de la rhétorique contrastive et ses hypothèses sont basées sur la théorie de la responsabilité du lecteur par rapport à celle du scripteur (Hinds, 1987). Les résultats montrent que tous les groupes d’apprenants de l’anglais L2 ainsi que les étudiants natifs utilisent les connecteurs bien plus fréquemment que les auteurs professionnels natifs, et que les fréquences d’utilisation des connecteurs par les étudiants natifs se placent au milieu de l’échelle de fréquence d’utilisation par les apprenants de l’anglais L2. Ceci peut suggérer que l’abondance des mots et expressions de liaison est une caractéristique générale des scripteurs non professionnels. Cette observation implique que l’enseignement de l’expression écrite au niveau avancé, qui aujourd’hui préconise un usage intensif des connecteurs dans l’expression écrite académique, devrait être reconsidéré.

Haut de page

Texte intégral

1. Introduction

1Due to the growing importance of English in education, research and global communication, there is an increasing need for people who are not native speakers of English to develop good writing skills in this language. These skills go beyond lexical appropriateness and grammatical accuracy and involve an ability to construct an effective text in English. The study reported in this paper is inspired by this trend. It will investigate how native speakers and advanced learners of English from a variety of linguistic backgrounds use connectors in their writing.

1.1. Connectors

2Connectors form a rather heterogeneous group. To date, linguists have not reached a consensus regarding the very term signifying this class, its definition or its complete list, not to mention a more in-depth analysis of it. For example, Halliday & Hasan (1976) distinguish four types of relations contributing to the grammatical cohesion of discourse: reference, substitution, ellipsis and conjunction. In a later book Halliday (1985:303) defines conjunction as a relation demonstrating different semantic associations among elements of discourse. According to Halliday, the conjunctive relation can be realized by a conjunctive adjunct (an adverbial or a prepositional phrase, e.g. moreover and in addition), or by a conjunction (understood here as a word class, e.g. and). In contrast, Crismore & Farnsworth (1990 cited in Connor, 1996:51) use the term connectors which they define as forming one of the elements of metadiscourse. Connectors can be expressed by one of the following categories : conjunctions, adverbials and prepositional phrases. Finally, in one of the recent descriptive grammars of English, Biber et al. (1999) distinguish linking adverbials which can be expressed by adverbs (e.g. however and finally), prepositional phrases (e.g. for example and on the other hand), or even clauses (e.g. that is to say). They differentiate between linking adverbials and conjunctions which are separate syntactic categories with a similar semantic function.

3In this paper it was decided to follow the definition proposed by Biber et al. (1999) as it makes an explicit distinction between linking adverbials (e.g. however, as a result, to sum up) and conjunctions (e.g. and, but, that, as), and it is the former category that will be the focus of the study reported here. Since the term linking adverbials is rarely used in the literature on second language acquisition (SLA) and English language teaching (ELT), the more neutral terms connectors and linking expressions have been adopted. They have been used to refer only to linking adverbials, with the exclusion of conjunctions.

1.2. Contrastive Interlanguage Analysis

4Granger (1998, 2002) proposed a new model of analyzing learner language. She claims that, in order to gain a full understanding of the factors that influence the different stages of interlanguage, multidimensional comparisons have to be performed. Undoubtedly, the learner language has to be evaluated against the target language in order to see to what extent it stays short of the native norm. However, at the same time the comparisons between learners with different native languages have to be carried out in order to see to what extent the differences between the native norm and the interlanguage are developmental and characteristic of all learners irrespective of their first language. At the same time it is desirable to analyze learners’ language side by side with their first languages (L1) in order to see to what extent the shape of the interlanguage (IL) is influenced by their L1. Figure 1 below illustrates the different paths of comparison which have to be performed in order to gain a fairly complete insight into various factors influencing learner language.

Figure 1. Contrastive Interlanguage Analysis: a blueprint for multiple comparisons

Figure 1. Contrastive Interlanguage Analysis: a blueprint for multiple comparisons

In the present study the interlanguages of students with different L1s will be compared with one another and with the native norm. However, the comparisons with students’ L1s cannot be performed directly due to the lack of comparable samples in the learners’ L1s. Instead, the assumptions about the use of connectors in various languages will be extrapolated from the observations and theories put forward within the field of Contrastive Rhetoric in the past 20 years.

1.3. Contrastive Rhetoric

5The status of English as the lingua franca of global communication partly explains the interest that researchers in SLA and ELT have given to the development of good writing skills by learners of English. There is a general agreement among researchers that even very fluent users of English do not manage discourse in the same way as native speakers. Such discrepancies reflect broadly understood cultural differences and are studied within the framework of contrastive rhetoric.

6The claims of contrastive rhetoric can be summarized as follows: “Contrastive rhetoric maintains that language and writing are cultural phenomena. As a direct consequence, each language has rhetorical conventions unique to it” (Connor, 1996:5). When writing in a foreign language, learners show a tendency to transfer not only the linguistic features of their native language but also its rhetorical conventions. These conventions pertain to such factors as the structure or units of texts, explicitness, information structure, politeness and intertextuality (Myers, 2002). As a result, native speakers of a language may find learners’ written discourse ineffective or even incomprehensible.

7One of the first attempts at describing cultural differences in structuring discourse was made by Hall (1976) who distinguished between high-context and low-context cultures. The main difference between the two lies in the need for explanation in discourse: in high-context cultures, characterized by close long-term relationships between group members, much important information can be left implicit, while in low-text cultures, in which group members form a larger number of interpersonal connections of shorter duration, more information must be explicitly stated. A visible manifestation of these cultural differences in writing is the approach the writer adopts for the reader (Hinds, 1987). In some cultures the responsibility for the success of the communicative act, which a text represents, rests with the writer. His/her writing should be as clear and reader-friendly as possible, which means that the ideas have to be laid out explicitly and the text should contain a variety of markers to signal the writer’s stance and to guide the reader through the text. In reader-responsible writing, on the other hand, the responsibility to find his/her way through the text and extract the author’s intentions and ideas is left to the reader. Such texts may not develop in a linear fashion, they may be full of digressions and they may contain fewer overt pointers communicating the author’s stance and guiding the reader. Hinds (1987) and Clyne (1987) made a distinction between cultures which favour reader-responsible or writer-responsible writing. According to these authors, the Anglo-Saxon tradition is reader-oriented whereas other traditions - Oriental and Teutonic - are reader-responsible cultures. The oriental tradition encompasses such languages as Japanese, Korean and Chinese (Hinds, 1987) and the Teutonic tradition includes German (Clyne, 1987), Finnish (Mauranen, 1993), Russian, Polish (Duszak, 1994, 1998) and Czech (Čmejrkova, 1994).

8An important characteristic of a reader-friendly text is the coherence achieved by, among other means, the use of connectors which are explicit markers of relationships between ideas. Thus, based on the observations concerning different writing cultures, it could be assumed that British and American writers would use more connectors than EFL learners coming from the reader-responsible writing traditions.

1.4. Previous studies

9The use of linking expressions by EFL learners has already been studied by numerous researchers, many of whom work within the framework of corpus linguistics methodology - the studies were of a quantitative nature and they focused either directly on comparing the frequency of connectors in native and EFL texts or on analyzing the frequency of repeated multiword expressions, among which connectors were found to be numerous. The analyzed samples were taken from existing learner corpora containing argumentative essays written by upper-intermediate or advanced learners of English with different first languages. The learners’ production was compared with similar essays written by native-English-speaking students. In some cases the EFL learners’ production was also set against essays written by equivalent students in their first languages. However, the results of these studies lead to ambiguous conclusions. Milton (1998) observed that the majority of multiword units overused by Chinese learners of English were linking expressions such as first of all, on the other hand, all in all, in addition or in a nutshell. A similar observation was made about Polish advanced learners (Leńko-Szymańska, 2006a). Granger & Tyson (1996) analyzed the frequency of linking expressions in essays written in English by French advanced learners and British and American students. They concluded that there were no significant differences between the frequency of connector use by the French and native English writers; however, the French learners tended to choose different expressions from the native speakers. A similar study investigating the use of linking expressions by Swedish advanced learners of English demonstrated that they underuse connectors in comparison with the native norm (Altenberg & Tapper, 1998).

10The studies mentioned above investigate the use of linking expressions by learners from reader-responsible cultures who, according to the assumption made above, should be more reluctant to use linking expressions than English native speakers. Yet, the results do not confirm this assumption. With the exception of Finnish EFL learners, all the other learners used linking expressions as frequently as or even more frequently than English native speakers. A possible explanation for such results could be the choice of native data for comparison.

1.5. Native norm

11All the studies mentioned above compare the production of EFL learners with essays written by British and American students who match the learners in age and educational background. The choice of such a benchmark over standard reference corpora containing published texts has been supported by Granger (1998). She claims that EFL essays are equivalent to native students’ compositions in terms of the authors’ experience and expertise in writing; thus the observed differences will only reflect the disparities in linguistic systems and will not be a result of discrepancies in the level of writing skills.

12However, as pointed out by Leńko-Szymańska (2006b, 2007), a choice of native student data as a base for comparison can be problematic. In the process of second language learning, students of English, particularly at upper-intermediate and advanced levels, usually receive a lot of explicit training in writing which also encompasses the use of metadiscoursal markers. In fact, language learners often receive more writing instruction in English than their British and American counterparts. As a consequence, EFL learners at the very advanced levels can often be more skillful in writing than native students. Moreover, in the process of language instruction at the upper-intermediate and advanced levels students are exposed to numerous authentic or semi-authentic texts, which have not been written by their equivalent native speaking students, but by expert writers. These texts are generally either taken directly or adapted from published sources such as newspapers, magazines or even literary publications. Even though the genres later produced by students do not exactly correspond to the texts they read and study in the classroom, in their writing they are expected to imitate the model of language from these texts. Thus, it seems desirable to include the professional writing in the comparison of EFL learner and native production.

13Of course, the method of comparing non-native essays with expert writing is also not free of problems. Even though there is a strong emphasis on the authenticity of language and tasks in EFL pedagogy, the genres that students produce do not match real-life writing. Argumentative essays are not equivalent to the genres produced by expert writers. Yet, it can be assumed that newspaper and magazine articles, particularly press editorials, as well as literary essays, share many features with student argumentative essays; thus they can form a sufficient base for comparison.

14Therefore, in order to gain a better insight into the factors influencing the development of interlanguage, it seems worthwhile to draw comparisons between EFL learners and both equivalent native students and professional writers.

1.6. Research objectives

15In view of the discussion above, a study was carried out with the following objectives:

  • to compare the frequencies of connectors in argumentative essays written by advanced EFL learners with different L1s;

  • to compare the frequency of connectors used by advanced EFL learners with the frequency of connectors in texts written by native speakers (both novice and expert writers).

2. The study

2.1. Data

16The data used in this study were drawn from three existing corpora containing samples of written English: International Corpus of Learner English (ICLE), Louvain Corpus of Native English Essays (LOCNESS) and Freiburg-London-Oslo-Bergen (FLOB) Corpus. The sections below contain a short description of each of the corpora and of the samples drawn from these corpora for the purpose of the present study.

17International Corpus of Learner English is a commercially available learner corpus compiled at the Université Catholique de Louvain in Belgium. The corpus contains argumentative essays written by advanced learners of English with different mother tongues. The essays are on average 500 words long and they were written by third and fourth year university students in English departments around the world. The corpus consists of 11 sections, each corresponding to a different first language and containing about 200,000 running words, equivalent to about 400 essays. Random samples of 100 essays were drawn from the following sections of the corpus: German, Swedish, French, Spanish, Russian, Polish and Finnish. The languages were selected to represent the major European language families: Germanic, Romance, Slavonic and Finno-Ugric languages.

18The LOCNESS Corpus was compiled in parallel with the ICLE corpus at the Université Catholique de Louvain to serve as a native benchmark. It contains over 300,000 running words and is made up of essays written by British secondary school students in A-level exams and by British and American university students. For the purpose of the study a random sample of 100 essays written by British university students was drawn from the corpus.

19The FLOB (Freiburg-London-Oslo-Bergen) Corpus is one of the commercially available reference corpora of British English. It contains one million running words of published texts from the beginning of the 1990s. The corpus consists of 15 sections (marked with letters from A to R), which correspond to different genres of written language. Section B of the corpus containing press editorials was drawn from the corpus for the study since it was assumed that this genre is the closest equivalent to students’ argumentative essays. The entire section, made up of 187 essays, was used in the study.

20The choice of British English as a benchmark for learners’ production was motivated by the fact that this variety of English is most often used as a linguistic model in language instruction in Europe.

21All the samples of writing used in the study are summarized in the table below:

Table 1. Summary of the data used in the study

Writers’ Native Language

Type of data

Source

No. of texts in the sample

Total size of the sample (in running words)

Average text length (in running words)

Finnish

student essays

ICLE

100

75 037

750

Swedish

student essays

ICLE

100

53 548

535

German

student essays

ICLE

100

36 682

367

French

student essays

ICLE

100

61 793

618

Spanish

student essays

ICLE

100

61 243

612

Polish

student essays

ICLE

100

62 841

628

Russian

student essays

ICLE

100

56 845

568

British

student essays

LOCNESS

100

53 412

534

British

press editorials

FLOBB

187

54 893

294

2.2. Analysis

22The analysis performed in this study was quantitative and involved a comparison of frequencies of connector use in each sample. This type of analysis allows one to trace the underuse and overuse of linking expressions by learners but disregards the appropriateness of use. The choice of method was motivated by the fact that the data used in the study came either from the native speakers or from very advanced learners of English. It has been pointed out by several researchers (e.g. Ädel, 2008) that interlanguage at the advanced stage of development contains few explicit errors and is characterized, rather, by underuse or overuse of certain linguistic elements. Although it cannot be ruled out that certain uses of linking expressions in the learner and even British novice samples are erroneous, it has been assumed that such errors are rather infrequent and that in any case they are not relevant to the investigation of the readiness to employ connectors in writing, which is the focus of the study.

23The data were analyzed with the concordance package Wordsmith 4 (Scott, 2004). First, concordance lines for each linking expression in each sample were generated and then examined manually to delete those lines which contained the search word used with a non-connective meaning and function. For example, the word yet is a linking expression only when used at the beginning of a clause; when used in the final position it functions as an adverbial of time. Next, the frequency of all connectors in each sample was calculated. Finally, the ten most frequent linking expressions in chosen samples were analyzed in more detail.

2.3. Analyzed connectors

24In order to calculate the frequency of connectors in the samples, a complete list of linking expressions was necessary. Such a list was found in Biber et al. (1999), who provide the most comprehensive inventory of connectors in the reviewed literature on the topic. Still, the list was appended with three variants of linking expressions: to sum up, that is to say and nonetheless. On the other hand, the decision was made to exclude the connector then from the analysis. The omission was motivated by the fact that then is a highly polysemous word which can take on a variety of functions and identifying these functions sometimes required arbitrary decisions which could influence the results of the study. Table 2 below lists 80 connectors used in the study.

Table 2. Connectors analyzed in the study (following Biber et al., 1999)

Category

Subcategory

Linking adverbials

Enumeration/addition

Enumeration

first, second, third, fourth, firstly, secondly, thirdly, fourthly, in the first/second/third/fourth place, first of all, for one thing, for another thing, to begin with, to start with, next, lastly

Addition

in addition, further, similarly, also, by the same token, furthermore, likewise, moreover, at the same time, what is more, as well, too

Summation

in sum, to conclude, all in all, in conclusion, overall, to summarize, in a nutshell

Apposition

Restatement

which is to say, in other words, that is, i.e., namely, specifically

Example

for instance, for example, e.g.

Result/interference

therefore, thus, consequently, as a result, hence, in consequence, so

Contrast/concession

Contrast

on the one hand, on the other hand, in contrast, alternatively, conversely, instead, on the contrary, in contrast, by comparison

Concession

though, anyway, however, yet, anyhow besides, nevertheless, still, in any case, at any rate, in spite of that, after all

Transition

by the way, incidentally, by the way

2.4. Results

25The frequencies of connectors in the analyzed texts are presented in Table 3 below. Due to the fact that the samples varied greatly in length, it is important to point out here that a comparison of raw frequency values would not be meaningful. For this reason, the last column of the table presents the normalized values, i.e. the average frequencies per 10,000 running words.

Table 3. Frequencies of connectors in the analyzed samples

Sample

Total sizeof the sample(in running words)

Average text length (in running words)

Total number of connectors in the sample

Average number of connectors per 10,000 running words

Finnish

75 037

750

868

116

Swedish

53 548

535

475

89

German

36 682

367

216

59

French

61 793

618

692

112

Spanish

61 243

612

596

97

Polish

62 841

628

821

131

Russian

56 845

568

457

80

British students

53 412

534

606

113

British press editorials

54 893

294

226

41

In order to facilitate the interpretation of the results, the normalized frequencies are presented in Figure 2 below.

Figure 2. Graphic representation of the average frequencies

Figure 2. Graphic representation of the average frequencies

26The results demonstrate that the average frequencies of connector use vary widely among the samples. The highest rate of linking expressions was found in the Polish essays with the normalized frequency three times higher than in the British expert writing. The average frequency of use of connectors by native-speaking students falls in the middle range of results with some groups of learners surpassing the British students and some others lagging behind them. Professional British writers were the most reluctant to employ linking expressions in their texts.

27The results also indicate that there might be some relationship between text length and the frequency of connector use. The texts by British expert writers and German students are the shortest among the compared samples and, at the same time, the average frequencies of linking expressions are the lowest in these samples. On the other hand, Polish and Finnish learners, whose essays are the longest, employed the largest number of connectors (as measured in normalized values). However, due to the insufficient quantity of data in this study, this observation cannot be verified statistically.

28Table 4 lists the 10 most frequently used linking expressions in the two native samples as well as in two learner samples drawn from different frequency ranges. It also presents the percentage of each of the 10 connectors in the total number of linking expressions in each sample. The last row of the table presents the cumulative percentage of the 10 most frequent connectors in the total number of linking expressions in each sample.

Table 4. Ten most frequent connectors in the chosen corpora

Polish

Swedish

British student

British expert

%

%

%

%

1.also

18,39

1. also

26,53

1. also

29,04

1.also

30,53

2. however

16,20

2. however

11,16

2. however

20,96

2. so

11,06

3. therefore

7,55

3. so

6,95

3. so

10,56

3. however

11,06

4. for example

6,09

4. therefore

6,32

4. therefore

9,57

4. yet

7,96

5. so

4,63

5. for example

5,47

5. for example

6,44

5. instead

3,98

6. thus

4,14

6. thus

3,58

6. yet

3,30

6. therefore

3,54

7. moreover

3,41

7. on the other hand

3,16

7. thus

3,14

7. thus

3,54

8. on the other hand

3,05

8. instead

3,16

8. firstly

1,98

8. after all

2,65

9. consequently

2,56

9. furthermore

2,95

9. instead

1,82

9. first

2,21

10. nevertheless

2,31

10. for instance

2,53

10. in conclusion

1,32

10. in the first place

2,21

Cumulative percentage

68,33

Cumulative percentage

71,79

Cumulative percentage

88,12

Cumulative percentage

78,76

The table shows that there is an overlap in connector use between the samples. However, the concession connector yet is not listed among the 10 most frequently used linking expressions in EFL texts; the learners preferred to use the contrast connector on the other hand. The learners frequently used additive connectors moreover and furthermore, and the British writers (both experts and students) preferred enumeration expressions first, firstly, and in the first place. It is also interesting to note that all the student writers (both native and non-native) frequently used the connectors for example and for instance, which are missing from the expert writing.

29The cumulative percentages of the 10 most frequent connectors in the samples indicate that British writers, particularly students, tend to rely on the most frequent linking expressions to a larger extent than do EFL writers. This implies that the use of linking expressions by non-native speakers is more varied than that of British writers.

3. Discussion

30The current study yielded the same results as the earlier research discussed in the introduction. Similar to Altenberg & Tapper (1998), it demonstrated that Swedish advanced learners use fewer connectors than native-speaking students. It confirmed the observation by Granger & Tyson (1996) that French learners do not differ in the frequency of connector use from their British counterparts. The results also stayed in tune with the conclusion by Leńko-Szymańska (2006a) that Polish learners overuse linking expressions. In particular the observation concerning Polish students is significant because, whereas Altenberg & Tapper and Granger & Tyson employed the same corpora as the ones used in this study (although not the same samples drawn from these corpora), Leńko-Szymańska analyzed a different corpus of learner English and compared it with a different collection of native student writing. This shows that the observed differences between the essays produced by students with different L1 backgrounds are not due to chance and represent a fairly robust characteristic of their writing.

31The results demonstrate the inadequacy of the assumptions about the use of connectors made at the outset of the study and extrapolated from the theory concerning reader/writer responsibility. British writers (both students and experts) did not use more linking expressions than did all the groups of advanced learners, and in addition they tended to rely on a more narrow selection of connectors than did foreign language students. Another significant finding is a discrepancy in the frequency of connector use by the students coming from the same writing tradition. Polish learners used twice as many linking expressions in their essays as did German writers even though both Polish and German are classified by researchers as belonging to the same reader-responsible tradition of writing (Clyne, 1987; Duszak, 1994, 1998). This observation provides further evidence that the approach to the reader is not directly reflected in the writer’s use of connectors. Interestingly, the frequencies of connector use by learners with related first languages (Swedish vs. German, Polish vs. Russian) also differed greatly. If a particular type of language could express coherence by other linguistic means than linking expressions, the results would be similar for all the languages belonging to this language group. Thus, the reason for the significant differences between the frequencies of connector use by learners coming from different L1 backgrounds cannot be established in this study.

32The most important observation made in the study is the large discrepancy in connector use between two native corpora containing British student and expert texts. In fact, the professional writers used linking expressions much less frequently than did any of the students, both native and non-native, and their preferences in the connector choice were less varied than the choices made by EFL learners. This may indicate that professional writers achieve coherence in their texts without an abundant use of linking expressions. It can be hypothesized that they choose and structure their arguments more effectively than do inexperienced writers; thus the reader does not need many overt markers to follow the reasoning of the writer. A qualitative study would have to be carried out in order to verify this hypothesis. Nevertheless, an important implication of this study is that the use of linking expressions by advanced learners of English is not influenced by the metadiscoursal conventions of their L1, and that a general characteristic of novice writing is an abundant use of connectors. This conclusion is also confirmed by findings reported by Altenberg & Tapper (1998) and Leńko-Szymańska (2007) concerning the frequency of linking expressions in texts by Swedish and Polish novice writers when writing in their first language. The frequency of connector use was comparable to the native English rate for Swedish writers or much higher than the native English rate in the case of Polish students.

33If the overuse of connectors is a general characteristic of student writing, it is possible that it is teaching-induced. For example, Kaszubski (1998) remarks that pedagogical materials used for writing instruction at advanced levels in Poland devote a lot of attention to the use of linking expressions. It can be hypothesized that the consequence of such extensive treatment in the classroom is students’ belief that if their essays contain a large variety and number of connectors, they will receive higher marks, which can lead to the forced and unnatural use of linking expressions, as in the example below drawn from the Polish sample.

(1)

As far as water pollution is concerned, every big city has a purification plant clearing the contaminated water. There is such a plant near Poznan as well. However, the technologies used there are obsolete. Nevertheless, the city council doesn’t have any funds for the modernization of this object. Still, a purification plant itself will not solve the ongoing problem of water contamination on a big scale. It only fights against the effects not the causes. Thus, we should rather find other ways of disposing waste liquids than changing our rivers into sewage ditches. But so far nobody has created a better solution. <ICLE-PO-POZ-0005.1>

Thus, this study should be followed up with a qualitative analysis of connector use by advanced learners of English in order to investigate whether linking expressions are employed to indicate real semantic relationships between ideas or whether they are used inappropriately and compensate for a lack of adequate structure in the development of an argument.

34One more observation deserves further investigation. As noted in the results section, the outcomes of the study seem to indicate that there might be some relationship between text length and the frequency of connector use; this relationship, however, could not be explored further due to the insufficient quantity of data. The related literature does not address this issue. Thus, it remains an open question whether this relationship is due to chance or represents a robust characteristic of written English. If the latter were the case, the results of this study would have to be reinterpreted.

4. Conclusions

35This study aimed to investigate the use of linking expressions by advanced learners of English. It was hypothesized that language students’ production falls short of native speaker norms due to the transfer of writing conventions from students’ L1 writing cultures. The results demonstrate in fact that students tend to overuse linking expressions in their texts and that this is a general characteristic of novice writing in English (both as a first and second language), which might be induced, at least partly, by instruction. An open question remains as to whether or not the same phenomenon can also be found in novice writing in other languages, like French for example.

36The outcome of the study bears important implications for the teaching of English as a foreign language. The pedagogical materials for writing instruction at the intermediate level and above encourage students to use linking expressions in their texts. However, this may lead to the exaggerated frequency of connectors. While it is counterintuitive to suggest that students should not be taught how to use linking expressions, it seems reasonable to recommend that their role in the construction of coherent discourse should be de-emphasized. Instead, other means of building coherence should be highlighted in the teaching process – constructing a logical argument in which semantic relationships between ideas are self-evident and where the reader requires few overt markers to follow it. Such a skill might be much more difficult to teach, but as a result, essays written by English learners may be not only more native-like, but also more expert-like.

Haut de page

Bibliographie

ÄDEL, A. 2008. Metadiscourse across three varieties of English: American, British and advanced learner English. In U. Connor, E. Nagelhout & W. Rozycki (eds.) Contrastive Rhetoric: Reaching to intercultural rhetoric, 44-62. John Benjamins, Amsterdam.

ALTENBERG, B. & M. TAPPER 1998. The use of adverbial connectors in Swedish learners’ written English. In S. Granger (ed.) Learner English on Computer, 80-93. Longman, London and New York.

BIBER, D., S. JOHANSSON, G. LEECH, S. CONRAD, & E. FINEGAN1999. Longman Grammar of Spoken and Written English. Pearson Education Limited, Harlow.

CLYNE, M. 1987. Cultural differences in the organization of academic texts: English and German.Journal of Pragmatics n° 11, 211-247.

CONNOR, U. 1996. Contrastive Rhetoric: Cross-cultural Aspects of Second Language Writing. Cambridge University Press, Cambridge.

CRISMORE, A. & R. FARNSWORTH 1990. Metadiscourse in popular and professional science discourse. In W. Nash (ed.) The Writing Scholar, 118-136. Sage, Newbury Park, CA.

ČMEJRKOVA, S. 1994. Non-native (academic) writing. In S. Čmejrkova, F. Danes, and E. Havlova (eds.) Writing vs. Speaking: Language, Text, Discourse, 303-310. Gunter Narr Verlang, Tübingen.

DUSZAK, A. 1994. Academic discourse and intellectual styles. Journal of Pragmatics n° 21, 291-313.

DUSZAK, A. 1998. Academic writing in English and Polish: comparing and subverting genres. International Journal of Applied Linguistics n° 8, 191-213.

GRANGER, S. 1998. The computerized learner corpus: a versatile new source of data for SLA research. In S. Granger (ed.) Learner Language on Computer, 3-18. Longman, London and New York.

GRANGER, S. 2002. A bird’s-eye view of computer learner corpus research. In S. Granger, J. Hung & S. Petch-Tyson (eds.) Computer Learner Corpora, Second Language Acquisition and Foreign Language Teaching, 3-33. John Benjamins, Amsterdam.

GRANGER, S. & S. TYSON 1996. Connector usage in the English essay writing of native and non-native EFL speakers of English. World Englishes n°15, 19-29.

HALL, E. 1976. Beyond Culture. Doubleday, New York.

HALLIDAY, M.A.K. 1985. An Introduction to Functional Grammar. Edward Arnold, London.

HALLIDAY, M.A.K. & R. HASAN 1976. Cohesion in English. Longman, London.

HINDS, J. 1987. Reader versus writer responsibility: A new typology. In U. Connor & R.B. Kaplan (eds.) Writing across Languages: Analysis of L2 Text, 141-152. Addison-Wesley, Reading, MA.

KASZUBSKI, P. 1998. Enhancing a writing textbook: a national perspective. In S. Granger (ed.) Learner English on Computer, 172-185. Longman, London and New York.

LEŃKO-SZYMAŃSKA, A. 2007. Past progressive or simple past?The acquisition of progressive aspect by Polish advanced learners of English. In E. Hidalgo, L. Quereda & J. Sanatana (eds.) Corpora in the Foreign Language Classroom, 253-266. Rodopi, Amsterdam and New York.

LEŃKO-SZYMAŃSKA, A. 2006a. The curse and the blessing of mobile phones – a corpus-based study into American and Polish rhetorical conventions. In A. Wilson, D. Archer & P. Rayson (eds.). Corpus Linguistics Around the World, 141-154. Rodopi, Amsterdam and Atlanta.

LEŃKO-SZYMAŃSKA, A. 2006b. Problemy metodologiczne Kontrastywnej Analizy Interjęzyka. In A. Duszak, E. Gajek & U. Okulska (eds.)Korpusy w angielsko-polskim językoznawstwie kontrastywnym: teoria i praktyka, 115-142.Universitas, Kraków.

MAURANEN A. 1993. Cultural Differences in Academic Rhetoric. Peter Lang, Frankfurt am Main.

MILTON, J. 1998. Exploiting L1 and interlanguage corpora in the design of an electronic language learning and production environment. In S. Granger (ed.) Learner English on Computer, 186-198. Longman, London and New York.

MYERS, G. 2002. Contrastive rhetoric and academic discourses: an institutional view. Paper presented at the 2nd International Contrastive Linguistics Conference, Santiago de Compostela, 26 October.

SCOTT, M. 2004. Oxford WordSmith Tools 4.0. Oxford University Press, Oxford.

Haut de page

Table des illustrations

Titre Figure 1. Contrastive Interlanguage Analysis: a blueprint for multiple comparisons
URL http://aile.revues.org/docannexe/image/4213/img-1.png
Fichier image/png, 5,9k
Titre Figure 2. Graphic representation of the average frequencies
URL http://aile.revues.org/docannexe/image/4213/img-2.png
Fichier image/png, 6,4k
Haut de page

Pour citer cet article

Référence électronique

Agnieszka Leńko-Szymańska, « Non-native or non-expert? The use of connectors in native and foreign language learners’ texts », Acquisition et interaction en langue étrangère [En ligne], 27 | 2008, mis en ligne le 30 septembre 2011, consulté le 19 novembre 2017. URL : http://aile.revues.org/4213

Haut de page

Auteur

Agnieszka Leńko-Szymańska

a.lenko@uw.edu.pl

Haut de page

Droits d’auteur

© Tous droits réservés

Haut de page