Editor

JLLT edited by Thomas Tinnefeld
Journal of Linguistics and Language Teaching
Volume 9 (2018) Issue 1
pp. 35-65

The Cognitive Contrastive Approach
as a Functional Method to Enhance
Students’ Language Proficiency? - A Case Study

Thomas Tinnefeld (Saarbrücken, Germany) & Frédérique Grim (Fort Collins (CO), USA)

Abstract (English)
The present article describes a case study targeting at the use of Contrastive Analysis in order to improve students’ foreign language mastery in a functional way. For this purpose, a research study, consisting of a pretest, a teaching phase and a posttest, was carried out at a mid-size American university. The students participating in this study were mostly sophomores in their second year of French. The outcome of the study showed that the mere use of contrastive analysis, without any reference to cognition and communication having been made, was about as efficient as the contrastive approach employed with reference to cognition and awareness-raising.
Keywords: Contrastive Analysis, awareness raising, cognition, pattern drill, communication, French, English
Abstract (Français)
Cet article décrit une étude de cas qui se concentre sur l’utilisation de l’Analyse Contrastive afin d’améliorer la maîtrise d’une langue étrangère par des étudiants de manière fonctionnelle. À cette fin, et à partir d’un pré-test, d’une phase d’enseignement et d’un post-test, une étude a été effectuée dans une université américaine de taille moyenne. Les étudiants participant à cette étude étaient pour la plupart en 2e année universitaire dans des cours de 2e année de français. Les résultats de l’étude montrent que la simple utilisation de l’analyse contrastive, sans aucune référence à la cognition et à la communication, était environ aussi efficace que l’approche contrastive utilisée avec référence à cognition et sensibilisation


Mots-clé: Analyse contrastive, cognition, sensibilisation, communication, français, anglais



1 Introduction
In the present article, a research study on the use of Contrastive Analysis (CA) in a French class at a mid-size American university is described. The target of this case study was to clarify the question of whether CA - in combination with cognitive and communicative elements - represents an efficient way to economically and functionally enhance students’ proficiency in a given foreign language, i.e. French.
In the framework of this study, CA was used in two different ways:
  • as a filter to delimit those structures of the target language that will most probably pose problems to students on the basis of their mother tongue, and
  • as a cognitive method to raise students’ language awareness with respect to selected structures of the target language in a communicative perspective.
In the context of this paper, students’ target language was French, and their mother tongue was English.
At the American university chosen as the academic institution for the study to be carried out, French classes are composed of students whose major is predominantly not French, nor philology in general. Being enrolled in other majors like Business, Communication Studies, Journalism, Liberal Arts, Theatre or Biology, students are, for the most part, focused on the communicative use of language in real-life situations and not so much on the functioning of language. The choice of this university on the one hand and the classes in question on the other also supports the use of CA because the contrastive approach aims to reduce the language knowledge to be taught, to a minimum as it concentrates upon the differences between the two languages and students do not need to be given the full grammatical picture of a structure to be learnt. Instead of presenting all the grammatical implications of a given construction, a relatively short hint to its (possibly parallel) use in students’ mother tongue would do. The actual presentation of grammatical features would then predominantly refer to the similarities or differences between the very structures of French and English.

2 A Brief History of Contrastive Analysis
As the field of Contrastive Linguistics or Contrastive Analysis is a very well-known one and has – with varying intensity – been in the focus of research for decades, it will hardly be necessary to present a literature review at long and at large here. We would therefore like to just point out to some important aspects that are of relevance for the present article.
Contrastive Linguistics (CL), as the theoretical background to be taken into consideration here, and CA as its rather practical version were, on the basis of their potential for foreign language teaching, developed in the 1950s (Lado 1957, Alatis 1968, (Corder 1967, 1974, Krzeszowski 1967, Nickel 1971, Richards 1974, Fisiak 1984 and, a bit later Di Pietro 1971 and Nickel & Nehls 1982), with the presumption that the potential problems learners might have in the mastery of a foreign language could be predicted on the basis of a given learner's mother tongue. The potential prediction of foreign-language errors caused by the interference (e.g. Carroll 1968) of students’ mother tongue, represented the strong claim of CA, and in its wake, grammar pairs were conceived for students to learn foreign languages more rapidly and functionally. In this context, the Contrastive Structure Series by Ferguson (also Ferguson 1968) as the general editor (e.g. Stockwell, Bowen & Martin 1965) were of high importance. In later years, however, two factors determined the further development of CA:
  • The necessity of creating an individual grammar for every language pair (e.g. English-French, English-German, English-Spanish), and this not only in one, but in both directions, i.e. English-French and French-English and so forth. For any language pair concerned, this was considered to be an impossible task to perform;
  • The fact that the strong claim of CA – as expressed by Lado, saying
(…) we can predict and describe the patterns that will cause difficulty in learning, and those that will not cause difficulty, by comparing systematically the language and culture to be learned with the native language and culture of the student. (Lado 1957: vii)
– was too big a promise to be kept, as there existed (and still exist) a considerable number of grammatical errors that could not be predicted on the mere systematic basis of the two languages in question, but rather went beyond that scope. The prediction of errors was therefore modified and transformed into a mere description of errors, which led to the weak claim of CA (Catford 1968). After the somewhat false hope of errors possibly being predictable, the mere explanation of errors was regarded as a considerable disappointment, which then led to a general decline of CA. Klein (1986) rightly summarized the situation when saying:
At present, no one seriously entertains the contrastive hypothesis in its strong sense. This is not to deny, however, that the learner’s knowledge of his first language influences the way in which he approaches and eventually learns a second language. (Klein 1986: 26)
Seen in an objective light, however, the possibility of explaining errors already represents a great chance for the creation of learning materials and the design of language courses. This understanding may have led to a certain renaissance of CA, however at a level that, although much lower in quality and quantity as compared to the hype of the 1950s to the 1970s, is still eminent in terms of the potential outcome of CA for the teaching and learning of foreign languages (e.g. König & Gast 2007), and this is not only true for the field of English, but also¸ for example, for romance linguistics (Roble i Sabater, Reimann & Sánchez Prieto 2016, Domínguez Vázquez & Kutscher 2016).
The fact that CA has always been considered as an important branch of linguistics (Fisiak 1990, Gnutzmann 1990, Krzeszowski 1990, Baumann & Kalverkämper 1992, James 1994) and that it has always been given new impulses (e.g. Dirven 1984, Mair & Markus 1991) is also reflected by its extension to other linguistic fields like lexicology (Hartmann 1991, Baker 1993 Altenberg 2002), syntax (Hill et al. 1991), textology (Hartmann 1997), pragmatics (e.g. Coulmas 1979, Fillmore 1984, Delin, Hartley & Scott 1996), translation (e.g. Friedrich 1969, Coseriu 1990, Ballard 1995, Hatim 1997, Teubert 2002) and various domains of languages for specific purposes (LSP) like terminology (Heltai 1988) and – in recent years – corpus linguistics (e.g. Johansson 1998, Balybina 2013), which only have an indirect influence on language teaching. Extensions to language pairs other than European ones (e.g. Spillner 2016), sometimes also referring to LSP, also reflect this develepment. Extensions of this field to 'contrastive culturality' (Gannuscio 2015) stand for another new development.
The analysis of individual linguistic categories or constructions related to specific language pairs has always been in the focus of contrastive linguistics (e.g., as these studies are legion, Hellinger 1977, Chuquet 1991, Cottier 1992, Colson 1993, De Keyser 1983, Lab 1994, Gallagher 1995, Hoarau 1997, and Godin 1999).
This rudimentary overview shows that CL and CA have, since the beginning of "their" time, been considered as an important factor of influence on foreign language teaching. Yet, they have not been attributed the recognition they actually deserve. One of the targets of our research project therefore is to humbly contribute to the reputation of these linguistic fields.

3   The Study
3.1 Presumptions
On the basis of the aforementioned reflections, we presume that CA has a higher impact when used in combination with cognitive language learning (i.e. awareness raising with regards to the learner's mother tongue that may interfere with the foreign language to be learnt) and the explicit explanation of the communicative functions of given language structures. This implies that when employed on its own, CA is not as efficient as it is in combination with the said elements.
This presumption expresses our expectation that CA needs to be backed up by other elements that come into play and that benefit from drawing students’ attention to specific factors of language learning and to the fact that communication is the very end of any human being's effort of learning a foreign language. CA, then, as important as it may be for the learning of foreign languages, should not stand alone: it should be supported by cognitive and functional elements that go beyond its own scope.

3.2 Pre- and Posttest
The study described here consisted of a pre- and a posttest and of a teaching phase (Section 3.5) in between these two.
Pre- and posttest, which were created on the basis of the grammatical and lexical systematicity of the two languages involved as well as on the outcome of classroom observation, comprised the following elements:
  • information about students (name, age, nationality, etc.),
  • a C-test,
  • a translation task (English-French),
  • faulty sentences to be corrected, and
  • a free composition task.
The information students were asked to indicate about themselves were:
  • their name (or a nickname to be kept constant in pre- and posttest so as to identify students discreetly),
  • a C-test of 30 gaps so as to approximately define students’ language proficiency level (Appendix I),
The C-test is a test type developed in the 1980s at the then Duisburg University1 (Germany) (Klein-Braley & Raatz 1982). It is a student- / examinee-friendly test which examines all the four basic language skills, i.e. reading, writing, listening, and - indirectly – speaking. This test allowed us to evaluate students’ performance in the subsequent exercises in a relatively reliable way;
  • The translation task (Appendix II) comprised five sentences or very short dialogues (i.e. two utterances each), which referred to some French lexemes that are easily confounded with English words by native speakers of English. These lexemes were:
    • okay vs. d’accord
In classroom observations in the classes examined, it was found that students, at times, translated the English utterance Are you okay? into French as *Tu es d’accord?
  • to attend a class vs. assister à une classe / un cours
The French sentence *J’attends la classe du professeur Dubois being linguistically correct, is contextually wrong, meaning I am waiting for his class to begin instead of the intended meaning (I attend / am attending Professor Dubois' class).
  • To rest vs. rester
American students show a tendency to translate the English verb to rest by the French verb rester.
  • to be finished vs. avoir terminé
American students often times translate the sentence I am finished by Je suis fini, this French sentence being both grammatically wrong and communicatively unacceptable.
  • to pass an exam vs. réussir à un examen
The French collocation passer un examen exists, but has a totally different meaning, i.e. ’to succeed in an exam’, and not just ‘to sit for an exam’ as is the case in English.
These grammatically wrong sentences refer to the structures researched upon in this study, namely object pronouns, the negation, the passé composé and the imparfait as well as the passif pronominal or voix moyenne.

3.3 Grammatical Background
The structures mentioned above can be expected to pose syntactic problems to native speakers of English as they are partly or even totally different in the two languages.
The French object pronouns are different from English in terms of form and syntactic position. In French, a sentence like You can give it (e.g. the book) to him would be rendered as Vous pouvez le lui donner, the pronoun lui replacing the indirect object (the person) and the pronoun le replacing the direct object. This twofold difficulty, i.e. the syntactic position and the lack of a preposition in the case of the indirect object, represents an obstacle for English native speakers to master.
The French imparfait (terminologically ‘Past Tense’) and the French passé composé (terminologically ‘Present Perfect’) do not reflect their English “equivalents” as far as this is implied by terminology. The French imparfait is rather used in analogy with the English Past Continuous, and the French passé composé is used in accordance with the Simple Past. Students need to develop a certain awareness of this tense system.
The French voix moyenne, as in the sentence Ce livre se vend bien, is another structure analysed in the present study. This structure exists in English as well, with the only difference that the reflexive pronoun does not figure in it, as in the sentence This book sells well. English-speaking students are therefore expected to form French sentences that are analogous to the English sentence, i.e. *Ce livre vend bien, which would be understandable to French native speakers, but which would not be well-received, as the mistake shows up immediately. Mastering this construction is therefore necessary so as to be well-reputed among French natives. Apart from that, the voix moyenne also represents an easily accessible way for students to express passive meaning without mastering the passive itself. This means that by knowing how to use the voix moyenne, students can easily enter the realm of the passive voice without facing all its morphological problems: the conjugation of the auxiliary être through the different tenses, and the forms and agreement of the respective past participles, which the classical passive entails. There is, however, a set of hesitations to the voix moyenne because it can only be used with unanimated nouns and (in most cases) with an adverb or an adverbial phrase.

3.4 Experimental Group and Control Group
The present study was traditionally designed and comprised an experimental group and a control group. Both groups represented two parallel second-year second-semester courses at the American university in question. Students’ general level of French was a low intermediate one.
The experimental group comprised 16 students, and the control group 14 students. Of these students, however, only nine students (experimental group) and 10 students (control group), respectively, were Americans.
In a strictly contrastive approach - which presupposes one and same mother tongue for the whole population of students, this one being English spoken by American natives -, only the American students could be counted in, the other students who did not represent the core of the groups having to be excluded. Thus, in the given situation, only nine students per group could be included in our study as this one was the smaller number of American students in the two groups and therefore represented the reference value. Among these students, one of the American students in the experimental group and two of the American students in the control group did not take part in the posttest. This means that, after all, a population of eight students in every group could be considered as the basis for this study.
This relatively small population of students did not permit any generalisations of findings. Accordingly, with regards to such a small population, is was not meaningful to carry out a statistical analysis, as none of the given statistical methods, e.g. the paired t-test, would be capable of rendering results of quantitative importance. Yet, it will be interesting to take a closer look at the figures retrieved – in awareness of the fact that these will have to be analysed very cautiously. That said, it is of general interest to look at the microstructure of these data in such a way that the students' individual developments these figures imply will be taken into consideration. In this vein, the present, predominantly qualitative study, is likely to provide some interesting insights.
With regards to the structure of experimental and control group, some personal information was first requested.
As far as students’ age was concerned, seven students in the experimental group were between 18 and 20 years old, and one student was aged 30. In the control group, all the eight students were between 18 and 20 years old. This means that in terms of age, the two groups were comparable.
With regards to students’ majors, the following subjects were indicated:
Experi-
mental Group
Major
Number of Students
Control Group
Major
Number of Students
Biochemistry
1
Biology
2
Business
1
Business (HR)
1
Communi-cation Studies
1
International Studies
1
Environmental Studies
1
Journalism
1
English
1
Theatre
1
Journalism
1
Undeclared
1
Liberal Arts
1


Undeclared
1


Table 1: Students’ Majors
In spite of some diversity of students’ majors, there are undeniable points that the two groups had in common:
  • None of the students was a French major. This means that both groups were in toto constituted of non-experts of French;
  • Only one student (in the experimental group) was a language major (English);
  • In both groups, two students each were enrolled in natural sciences (biology, biochemistry, environmental studies);
  • Two students (experimental group) and four students (control group) were enrolled in communication-oriented subjects (Communication Studies and Journalism on the one hand, and International Studies, Journalism and Theatre on the other).
These points assured the qualitative comparability of experimental and control group.
In this context, it is interesting to note that only one student (in the experimental group) was a language student, enrolled in English.
In terms of the number of years during which students had already learnt French, the following situation was retrieved:
Experimental Group
Control Group
Duration
(in years)
Number of Students
Number of Students
2-3
3
2
4-5
4
5
6-7
1
1
Table 2: Duration of Learning French
The above table shows that experimental and control group were highly comparable as far as the duration during which students had learnt French is concerned. If a duration of up to three years of learning French is defined as a short one, there were three and two students, respectively, who fell into this category. Four and five students, respectively, had learnt French during a medium duration of four to five years. One student each had learnt French for a long period of time (six or seven years, respectively).
Students’ beginning ages of learning French - a category which is of secondary importance, but which may still be interesting to consider - varied from 12 and 19 in the experimental group and from 13 to 18 in the control group. From this perspective, students in both groups were highly comparable as well.

3.5 Teaching Phases
Any research on CA faces a fundamental logical problem: the linguistic items presented to subjects, i.e. students, need to be oriented to (similarities and differences of) their mother tongue. Without this conditio sine qua non, contrastive research cannot be executed. This, however, means that implicitly, i.e. without verbally mentioning these similarities or differences, the contrastive perspective is always present. And this presence may have been felt by those students who formed the control group. Those in the experimental group would be confronted with contrastive phenomena anyway, that means they would explicitly be informed about the mechanisms, such as transfer and interference, that come to fruition in this context. With both groups being – implicitly (control group) and explicitly (experimental group) - confronted with phenomena that they may, consciously or not, have come across in their previous learning processes anyway, it was, for this logical reason alone, not unlikely that the results obtained in both groups might not be too different.
With these theoretical reflexions in mind, we designed the teaching phases, which lasted 50 minutes in every group and were delivered by the male researcher. The only information students in the control group were given was the presentation of the fact that we human beings normally do not learn any foreign language from scratch, i.e. without any previous knowledge about our mother tongue, but that we do have to develop a “feel” for the foreign language to be learnt because this language has typical features that are different from those of our native language. No mention was made about contrastive linguistics nor contrastive analysis so that students did not have the slightest idea as to the background of this study .
Students in the experimental group were given exactly the same information. After that, they were asked the following questions which were discussed in detail:
Which ones are the biggest difficulties for you in French grammar?
Consider these French sentences: What is wrong in them?
Cette porte ouvre sans bruit. (reflexive verbs)
Hier, quand je rentrais à la maison, je trouvais une lettre de mon amie. (tenses)
Je pense Martine est ma meilleure amie. (que conjonctif (= ‘that’))
L’année dernière, j’allais en Espagne. J’ai l’impression que c’est un pays très intéressant. (tenses)
Mon impression avère être réaliste. (reflexive verbs: ‘proves to be’)
Now think of other problems of that kind. What other sources of errors in French can you think of?
Which French structures will not be problematic for native speakers of English? What do you think?
This was the contrastive part of the teaching phase.
In addition, students’ awareness was raised with regards to the constructions dealt with in this study, using easily accessible English:
  • For the field of false friends, the instructor presented the general problem of two words, one from students’ mother tongue and one from the foreign language, that looked very similar, this similarity often being transferred to the potential meanings of these words - a process which then goes wrong, resulting in negative transfer (i.e. interference).
  • For the negation, its basic functions were presented:
    • expressing different types of meanings to express a given perspective like ne...pas (‘not’), ne... plus (‘no more’ or ‘no longer), ne...pas du tout (‘not at all’), and ne...jamais (‘never’), ne...personne (‘nobody’) or ne... pas encore (‘not yet’). These equations were identified as being highly similar and therefore safely usable in French and English;
    • expressing oneself in a differentiated way, using various no ways;
    • expressing politeness, e.g. by saying Ce n’est pas tout à fait correct (‘This is not quite right’) instead of bluntly saying C’est faux (That’s wrong’).
  • Object pronouns were presented as having the function of:
    • shortening sentences;
    • providing information that is already known, as the persons or objects they refer to must have been mentioned previously;
    • creating exclusion to outsiders, as the persons or objects they refer to must be known to all interlocutors for them to understand what is meant;
and as being systematically different in French as far as their forms and positions in the sentence are concerned, this systematicity making any direct transfer from English to French impossible.
  • The voix moyenne, was explained as:
    • representing a short and easily accessible way to express the passive;
    • enabling students to express passive meaning without mastering the genuine passive;
    • a construction that is easy to handle but is not very frequent and comes with some restrictions (limitation to unanimated nouns, action verbs and the presence of an adverbial phrase; tendency to being used in spoken French).
  • The tenses imparfait and passé composé were presented in terms of their “magic” methodological questions, i.e.
    • What was the situation like? and What was going on (before something else happened)? for the imparfait and
    • What happened at that moment? for the passé composé,
this usage being transferable from English to French.
The procedure we followed in the experimental group represented what we call the cognitive contrastive approach.
For the treatment of these grammatical constructions, we used pattern drill in both groups because it would ensure a sufficiently mechanical way that would exclude any cognitive elements other than the ones explained in the experimental groups as outlined above. In the control group, no cognitive elements were referred to.
In the following section, the results of our study will be presented and analysed.

4 Results and Analysis
4.1 C-Test
Among the general information students were asked to indicate, there was their personal estimation of their own ability of French. They were also presented a C-test so that their actual language skills could be objectively defined. Table 4 shows the results of this C-test:
Student
Experimental Group
(30 = 100%)2
Student
Control Group
(30 = 100% )
Ada3
19 - 63.3% (E)4
Alana
14 - 46.7% (U-I)
Beth
10 - 33.3% (L-I)
Caldwell
21- 70,0% (L-I)
Brenda
17 - 56,7% (U-I)
Joan
25 - 83,3% (L-I)
Caroline
14 - 46.7% (L-I)
Maddie
14 -46.7% (U-I)
Charlene
3 - 10.0% (L-I)
Nathasha
8 - 26.7% (E)
Chuck
21 - 70.0% (L-I)
Sandra
19 - 63.3% (U-I)
Elaine
26 - 86.7% (L-I)
Sonia
21 -70.0% (U-I)
Anna
17 - 56.7% (E)
Stella
17 - 56.7% (L-I)
Table 3: C-Test Results
The results imply that students' self-estimations (in brackets) differ in such a way that students of the control group thought more highly of their own language competence than those of the experimental group did.
What is interesting - as a side result apart from any contrastive considerations - is the fact that students’ self-estimations and their actual language performance did by far not always go together, which becomes clear when the different language levels and the percentages attained in the C-test are correlated. By grouping the various language proficiency levels as listed below, we were able to interpret students’ self-estimations:
Elementary: 0% - 20%5
Upper-Intermediate: 51% - 80%
Lower-Intermediate: 21% - 50%
Advanced: 81% - 100%
Only three and two students, respectively, (Beth, Brenda, Caroline in the experimental group, and Sandra and Sonia in the control group) estimated their own French ability realistically. These rates hint at the fact that asking students for their self-estimations in terms of their own foreign language skills is only of limited explanatory power6.
The level of the two courses students attended was lower-intermediate or B1 / B2 in terms of the European Framework of Reference for Languages. B2 representing a lower intermediate level, with a range from 21% to 50% postulated here, we see that seven students of the experimental group and all the eight students of the control group either had this level or were even more proficient than expected. This result shows that the two language courses taken as a basis for this study can be considered as qualitatively homogeneous.

4.2 Translation Task
The first task that was relevant in contrastive terms was a translation task. It comprised five sentences to be translated from English into French. These sentences were the following ones:
  • Julie had a bad cold last week. Is she okay now?
The item in question whose translation we were interested in was the word okay (Section 3.1 for this one and the four subsequent sentences). Against a contrastive background, it might be expected that students would translate this word by the French d’accord. As a matter of fact, students were not informed that this was the relevant word.
  • A: Are you going to attend Professor Miller’s class?
B: I don’t know yet.
In this short dialogue, the verb attend was the word whose translation we suspected to pose contrastive problems to students. Contrastively speaking, students‘ translation might be the French verb attendre (*Vas-tu attendre la classe du professeur Miller?)
  • I am very tired. I have to relax a little bit.
The word of relevance in this sentence was the verb relax, which students might translate as rester. Here, we did not use the verb to rest on purpose, so as not to make things too blunt for students. The idea here was to find out whether students would use the verb rester although the contrastively relevant false friend was not evoked.
  • Unfortunately, I cannot meet with you now because I still have a lot of work. But I’ll be finished soon.
The salient structure here is the verb finish, which, from a contrastive perspective, might be translated by *Je serai fini.
  • A: I’m extremely worried because I’m so bad at mathematics.
   B: Don’t worry, you are much better than you think. You will definitely pass the exam     next week.
The structure in the focus is to pass the exam, which, in a contrastive view, might be translated as *passer l’examen.
The structures chosen for the translation task were closely related to those that had been dealt with in the teaching phase. The results of correct translations delivered by students in the pre- and the posttest are listed below. In the experimental group, the situation was as follows:
Task 1: Translation
Students
(Experimental Group)
Pretest
Posttest
Ada
4 / 5
4 / 5
Beth
3 / 5
1 / 5
Brenda
2 / 5
1 / 5
Caroline
0 / 5
2 / 5
Charlene
2 / 5
1 / 5
Chuck
5 / 5
2 / 5
Elaine
4 / 5
4 / 5
Anna
2 / 5
1 / 5
Table 4: Translation Task: Pre- and Posttest Results (Experimental Group)
In the control group, the situation was as follows:
Task 1: Translation
Student
(Control Group)
Pretest
Posttest
Alana
1 / 5
0 / 5
Caldwell
2 / 5
0 / 5
Joan
3 / 5
1 / 5
Maddie
4 / 5
4 / 5
Natasha
5 / 5
1 / 5
Sandra
1 / 5
2 / 5
Sonia
3 / 5
3 / 5
Stella
4 / 5
2 / 5
Table 5: Translation Task: Pre- and Posttest Results (Control Group)
Both in experimental and control group, one students improved her pretest result in the posttest, two students’ performances remained unchanged, and five students obtained results that were not as good as the ones obtained in the pretest.
Thus, when looking into the results at an individual level, we will find that several students made some promising progress from pre- to posttest7. We will therefore limit our further analysis to these very students.
In the first item of the translation task, Caroline, who was the only student of the experimental group to improve her performance from pre- to posttest, did not come up with any translation of the sentence Is she okay now?. In the posttest, however, she translated this sentence as Est-ce qu'elle est bien?. As for the fourth item of the translation task - I’ll be finished soon -, she delivered the contextually inadequate translation Je finis bientôt in the pretest, and the adequate translation Je vais finir bientôt in the posttest. This performance displayed a clear improvement as compared to the one shown in the pretest.
In the control group, there was also one student, Sandra, who, at an equally low level, obtained a better result in the pretest than in the posttest. Not only did she correct the error in Item 5 right, forming the correct sentence Tu recevras une bonne note, but also reproduced this correction in the posttest and additionally corrected Item 1 in form of Est-ce qu'elle est bien?. This performance also represents a slight improvement although not a single allusion to cognitive or communicative aspects had been made to the students of this group.
Error correction was the next task students were asked to work on.

4.3 Error-Correction Task
The second task students were presented with contained faulty sentences that needed to be corrected. This task is briefly described below:
Item 1 referred to the voix moyenne, the wrong version being Ce livre est très joli. Il vend vraiment bien, which was to be corrected as Ce livre est très joli. Il se vend vraiment bien (the correct form being underlined).
Three items dealt with objective pronouns:
  • In Item 2, the second sentence Tu lis le livre que je voudrais avoir. Me le donne, s’il te plaît needed to be corrected as Tu lis le livre que je voudrais avoir. Donne-le-moi, s’il te plaît.
  • Item 5 was the following utterance: Quand je t’ai prêté mon livre préféré, je ne savais pas que tu le garderais tellement longtemps. Me le rends tout de suite, s’il te plaît. The second (faulty) sentence was to be transformed into Rends-le-moi tout de suite, s’il te plaît.
  • Item 9 featured the same grammatical phenomenon in the following short dialogue:
A: Si tu veux, je peux vendre mon téléphone portable à toi à un très bon prix.
B: Ce serait formidable. Je serais très heureuse si tu lui vendais à moi.
and needed to be corrected into
A: Si tu veux, je peux te vendre mon téléphone portable à un très bon prix.
B: Ce serait formidable. Je serais très heureuse si tu me le vendais.
The focus of three items was the use of imparfait and passé composé, which was exemplified by the incidence scheme:
  • Quand j’ai rangé mes choses, ma mère entrait dans ma chambre (Item 6)
This sentence needed to be corrected as follows: Quand je rangeais mes choses, ma mère est entrée dans ma chambre.
  • Item 8 also featured these two tenses and also evoked the incidence scheme, however not in one, but in two different sentences: Le soleil a brillé. Il y a eu une ambiance fantastique. Mais tout d’un coup, un bruit me réveillait de ma rêverie. The correct version these sentences needed to be transformed into was: Le soleil brillait. Il y avait une ambiance fantastique. Mais tout d’un coup, un bruit m’a réveillé(e) de ma rêverie.
  • Item 10 also focused on two tenses presented in a sequence, but without any reference to the incidence scheme: Avant-hier, Frank me promettait de m’aider. Hier, il ne le savait plus. The correct version of the first sentence was: Avant-hier, Frank me m’a promis / m’avait promis de m’aider. Hier, il ne le savait plus.
Item 3 and Item 7 focussed on the subjonctif.
  • Item 3 featured the sentences Je crois pas que tu me dis la vérité8 to be modified into Je ne crois pas que tu me dises la vérité.
  • In item 7, the subjonctif was dependent on the verb croire: Je crois pas que je pourrais participer à la fête ce week-end9, where the correct form should have been: Je ne crois pas que je puisse participer à la fête ce week-end.
  • Item 4 was a short dialogue:
             A: Est-ce tu as perdu quelque chose?
             B: Non, je n’ai perdu rien,
     
          in which B’s reply had to be corrected as follows: Non, je n’ai rien perdu.

Students’ performance in the experimental group can be gathered from Table 6 below:

Task 2: Correction of Errors
Students
(Control Group)
Pretest
Posttest
Ada
3 / 10
2 / 10
Beth
1 / 10
4 / 10
Brenda
1 / 10
5 / 10
Caroline
0 / 10
2 / 10
Charlene
3 / 10
6 / 10
Chuck
2 / 10
3 / 10
Elaine
3 / 10
3 / 10
Anna
3 / 10
2 / 10
Table 6: Error-Correction Task: Pretest and Posttest Results (Control Group)
Microstruturally speaking, i.e. with respect to the individual performance of students, in the experimental group, five out of eight students improved their performance from pre- to posttest, three of them showing considerable improvements, i.e. by three or four points, respectively. One student's performance remained constant, and two students did not perform as well as they had done in the pretest, losing one point each as compared to their pretest results.
In the control group, the results were even slightly better as is documented in Table 7:
Task 2: Correction of Errors
Students
(Control Group)
Pretest
Posttest
Alana
2 / 10
2 / 10
Caldwell
2 / 10
4 / 10
Joan
3 / 10
3 / 10
Maddie
3 / 10
5 / 10
Natasha
5 / 10
1 / 10
Sandra
1 / 10
4 / 10
Sonia
3 / 10
4 / 10
Stella
2 / 10
5 / 10
Table 7: Error-Correction Task: Pretest and Posttest Results (Control Group)
Five students performed better than in the pretest, one student showed a constant performance, and one student performed (much) worse than in the pretest.
With respect to the qualitative performances shown in the experimental group, we will take a closer look at Brenda, Beth and Charlene, as these three students showed the highest quantitative improvements.
Brenda's quantitative improvement was the highest from one to five out of ten adequate error corrections. Whereas in the pretest, she had only managed to correct one single mistake, trying to solve five out of the ten errors offered for correction and giving up on the other five, in the posttest, she gave five correct answers out of ten. Apart from the adequate correction she had offered tor Item 1 (Je ne crois pas) in the pretest and which she repeated in the posttest, she got Items 4 (Je n'ai rien perdu), Item 6 (Ma mère est entrée) and Item 7 (Je ne crois pas) right, which stands for a clear qualitative improvement.
Beth did not provide any answer to eight of the ten items offered in the pretest. She only corrected Item 3 (Je ne crois pas) right. In the posttest, she repeated this correction, adding Je rangeais and *ma mère a entrée10 in Item 6, Je ne crois pas in Item 7 as well as Il y avait and un bruit m'a réveillé in Item 8. All these corrections represent a positive development as compared to the student's pretest, which may be considered as being due to the awareness raising processes in terms of cognition and communication that students in the experimental group were provided with.
The same is true for Charlene, who had got three items right in the pretest - Item 3 (Je ne crois pas), Item 7 (Je ne crois pas) and Item 8 (Il y avait) - and who gave four correct answers in the posttest: Il se vend bien in Item 1, again Je ne crois pas in Item 3 and Item 7, and je rangeais and – like Beth - *ma mére a entrée in Item 6. These corrections also stand for a qualitative improvement and, last but not least, reveal the successful acquisition of the voix moyenne, which had been new to students.
Students in the control group were also rather successful in the completion of this task. Again, the three students whose performance showed the biggest differences between pre- and posttest will be focused upon here.
Sandra and Stella made the biggest leap, improving their results by three points each. Sandra only corrected the error in Item 3 (Je ne crois pas) in the pretest. In the posttest, she successfully added the corrections of Items 4 (Je n'ai rien perdu), 6 (je rangeais – *ma mère a entré), and 7 (Je ne crois pas).
Just like Sandra, Stella only got Item 3 (Je ne crois pas) in the pretest right. In the posttest, she successfully added the corrections of Items 6 (je rangeais - ma mère est entrée), 7 (Je ne crois pas), 8 (il y avait...), and 10 (Frank m'a promis) to that one. This performance undeniably shows that Stella had made progress from pre- to posttest.
Maddie improved her pretest result by two points. Whereas in the pretest, she got Items 3 (Je ne crois pas), 4 (Je n'ai rien perdu) and 6 (je rangeais – *ma mère a entré) right, in the posttest, she corrected the items 2 (Tu me le donne ), 4 (Je n'ai rien perdu), 5 (Tu me le rends), 6 (je rangeais – ma mère est entrée), and 9 (tu le vendais à moi11) successfully. Even if she could not maintain the adequate correction of Item 3 in the posttest, this result also hints at some learning progress made.
The qualitative improvements documented here stand for undeniable learning effects in both groups.
The next task was closely related to error correction as it was about explaining the nature of the faulty elements in the corresponding sentences. These are the results retrieved in this part of our experiment: 
Task 3: Explanation of Errors
Students
(Experimental Group)
Pretest
Posttest
Ada
2 / 10
4 / 10
Beth
1 / 10
4 / 10
Brenda
1 / 10
5 / 10
Caroline
0 / 10
1 / 10
Charlene
3 / 10
6 / 10
Chuck
2 / 10
2 / 10
Elaine
2 / 10
1 / 10
Anna
3 / 10
2 / 10
Table 8: Error-Explanation Task: Pretest and Posttest Results (Experimental Group)
The explanation of language errors, i.e. the conscious awareness of language problems and, what is more, the ability to verbalise them, can be considered a highly complex task. This high task complexity explains the low overall results obtained by students, with only Charlene, who obtained six points in the posttest, attaining more than half of the available points.
As a whole, our figures show that five of the eight students in question improved their performance from pre- to posttest. One student's performance remained stable, and two students performed worse in the posttest than in the pretest. In addition, the positive leaps students made (e.g. Charlene from three to six points, Beth from one to four and Brenda from one to five points, respectively) were bigger than the deteriorations of performance shown by Elaine and Anna, who only lost one point each from pre- to posttest.
Thus, the teaching phase with its three elements - the contrastive approach, combined with cognitive and communicative elements – may be considered as qualitatively successful.
In the control group, the situation was as follows:
 
Task 3: Explanation of Errors
Students
(Control Group)
Pretest
Posttest
Alana
2 / 10
2 / 10
Caldwell
0 / 10
1 / 10
Joan
3 / 10
3 / 10
Maddie
3 / 10
4 / 10
Natasha
5 / 10
1 / 10
Sandra
1 / 10
3 / 10
Sonia
3 / 10
4 / 10
Stella
3 / 10
6 / 10
Table 9: Error-Explanation Task: Pretest and Posttest Results (Control Group)
Just like in the experimental group, five students improved their respective results obtained in the pretest. With one exception - Stella, who attained three points more in the posttest than she had scored in the pretest -, there was just a limited gain of (one or two) points from pre- to posttest. Thus, although the number of students scoring better in the post- than in the pretest was the same in both groups, the quality of improvement in the control group was not as remarkable as that of the experimental group.
At an individual level, this result shows that the contrastive approach, which represented the starting point in both groups but which was not further elaborated to the control group, was successful, but it was not quite as successful as the combination of contrastive analysis, contrastive awareness raising and the explanation of the communicative functions of the constructions in question, which was applied in the experimental group. At that level, students did make progress.
The last task students were asked to perform was to write a composition on the topic Mes vacances les plus inoubliables (‘The most memorable vacations’), which could be regarded as broad and general enough for everyone to be able to write a short text about. The question that we were particularly interested in was the number of contrastive errors that students made. Furthermore, we wanted to examine whether the number of contrastive errors made by students in the posttest would improve as compared to the number of contrastive errors made in the pretest.
Before analysing students’ quantitative performance, it will be necessary to address the difference between a 'contrastive' error and a 'general' error. A general error is here defined as any mistake that people (natives or non-natives) make when using a given language, i.e. their mother tongue or a foreign language, and which is not occasioned by another language. A contrastive error is a mistake that (mostly non-native) speakers make when using a foreign language, which is (most probably) occasioned by their mother tongue.
One example each may clarify this differentiation: When an English native speaker utters the French sentence Ce livre, je le te donne tout de suite (instead of the correct sentence Ce livre, je te le donne tout de suite), this error represents a general one because the structure in question, i.e. the sequence of object pronouns in the French sentence, is an internal problem of French grammar and not used in English in the same way.
When, however, a speaker writes Hier, j'allais au théâtre, this error can be classified as a contrastive one because the construction is the same one as in English (Yesterday I went to the theatre) and can therefore be seen as a trigger for the mistake in the French sentence, in which the passé composé should be used instead of the imparfait (Hier, je suis allé au théâtre).
The error rate was calculated on the basis of an error coefficient, i.e. the number of errors made in 100 words. In order to maintain the highest possible level of objectivity, the individual errors were not qualitatively differentiated, all being counted as a full error (1.0) each and not as 0.5 or even 0.25 errors in minor cases.
The results of the experimental group in pre- and posttests are displayed in the table below: 
Task 4: Composition (Error Coefficient)
Students
(Experimental Group)
Pretest
Posttest
Ada
6.2
0
Beth
0
5.5
Brenda
1.7
3.7
Caroline
0
0
Charlene
2.8
2.2
Chuck
4.3
1.2
Elaine
9.1
3.6
Anna
3.0
0
Table 10: Composition Task: Pretest and Posttest Results (Experimental Group)
Of the eight students forming the experimental group, five improved their composition result in terms of contrastive errors from pre- to posttest. One student's performance remained the same, but at a very high level, and two students performed worse in the posttest, one of them (Beth) with a considerable deterioration (from 0.0 to 5.5 errors in 100 words). The improvements of three students (Ada Chuck, and Anna) were considerable (from 6.2 to 0.0, and 4.3 to 1.2, 3.0 to 0.0, respectively). Clearly, the two biggest improvements were qualitatively more extensive than the biggest deterioration (Beth: from 0.0 to 5.5).
In the control group, results were as follows:
Task 4: Composition (Error Coefficient)
Students
(Control Group)
Pretest
Posttest
Alana
2.3
0
Caldwell
1.5
6.5
Joan
2
2,8
Maddie
3
2
Natasha
1.8
4.3
Sandra
3.2
2.3
Sonia
1.2
1.112
Stella
0
0
Table 11: Composition Task: Pretest and Posttest Results (Control Group)
Four of the eight students of the control group improved their pretest results in the posttest. The performance of three students was worse than the one they had attained in the pretest, and one student performed exactly the same way in both tests. This means that, as compared to the experimental group, one student less could improve his or her result and showed a rising tendency in terms of error coefficients. Relatively speaking, Alana's performance was the best one in the two groups combined, the other improvements being rather small-scale ones (Maddie: from 3.0 errors to 2.0 and Sandra: from 3.2 errors to 2.3). As far as the respective deteriorations are concerned, there was a considerable one (Caldwell from 1.5 errors to 6.5), which represents an even larger deterioration than those documented for the experimental group (Natasha from 1.8 errors to 4.3 and Joan from 2.0 errors to 2.8). Overall, the improvements made by students in the control group were - with one single exception - less far-reaching than those of the experimental group, which means that the latter was more successful in this composition task.
Our results show a clear tendency for students at an individual level: in both groups, students could reduce the number of contrastive errors considerably - without even knowing what a 'contrastive' error was, thus, without having any means of consciously influencing this result. In the experimental group, the contrastive error coefficients went down more drastically from pre- to posttest than in the control group. Qualitatively speaking, this means that the French texts students generated read less English than before and that they became more "authentic".
On the basis of the previous results, the following deductions can be made:
  • As students in both groups made some progress, the contrastive approach - which was used as a starting point in the control group and as the essential part of a deeper approach, combined with cognitive and communicative elements, in the experimental group - can generally be attributed a positive effect.
  • Even when used implicitly, the contrastive approach proved to be functional for foreign language learning.
  • When used explicitly and in combination with cognitive and communicative elements, its functionality increased slightly more.
This means that CA has a high potential for language learning when used alone and implicitly and an even slightly higher potential when used explicitly and in combination with cognitive and communicative considerations.

4.4 Free Composition Task
To round off the presentation of our data, it appears to be fruitful to take a closer look at students‘ writing performance as far as contrastive considerations are concerned. For this purpose, the performance of one student from the experimental and two students from the control group will be focused upon.
For the experimental group, Chuck’s text will be analysed. In his pretest free-composition exercise, he wrote a text on a trip to California that he took with his family several years ago.
Chuck (pretest):
Il y a 16 ans, j'ai travaillé au Californie pour visiter ma famille.13
The two errors featured in this example are the use of the verb travailler (‘to work’) for the English verb to travel and that of the verb visiter (‘to visit’) in a context where the construction rendre visite would be used in French. These two errors are clearly induced by the influence of the student’s mother tongue. Such errors represent serious mistakes because in the case of travailler, French natives who do not master English would not understand what the speaker wants to express: they would think of work having been done rather than a trip having been undertaken. The second error (visiter) does not really represent a source of potential misunderstandings because the intended meaning is much easier to grasp for natives, as the same verb is used in connection with places, but not in connection with people as is the case here.
In comparison with his pretest performance, the only contrastive error Chuck made in his posttest was the following one:
Chuck (posttest):
Nous (...) avons trouvé une place qui prépare des nourritures. (i.e. a place that prepares...)
In this error, which is clearly caused by the student’s mother tongue, the English syntax shines through (a place which prepares food)14. This error does not reflect good style nor is it good French. However, the intended meaning is easy to understand.
The description of these errors made by Chuck in his pre- and his posttest shows a clear upward tendency of his performance at a qualitative level.
A comparable tendency can be noted for the following two students, who belonged to the control group. Maddie wrote a short text about a trip to London with her fellow students and friends:
Maddie (pretest):
Quand nous sommes arrivés au Heathrow, j'étais très fatigué. Ma amie et moi, nous décidions de dormir pour quelqu'un minutes avant nous devions de partir avec le groupe.
In this example, the use of the French imparfait is incorrect. Instead, the passé compose needs to be used as a new action sets in. This French usage unduly generalizes the use of the English Past Tense, which is normally employed for any kind of actions that happened in the past. This incorrect usage of French is thus directly derived from the English, the error thus becoming a contrastive one. Tense errors of this type generally confuse French natives because they have to think twice so as to reconcile the actions described and state-oriented tense that is used to verbalize them. Thus, such errors can be regarded as serious ones.
Remarkably, in her posttest, Maddie only made one error with regards to tenses, and this does not represent a content-related error like the previous one, but “only” a morphological error (*nous nous sommes promenés):
Maddie (posttest):
J'aimais les structures grandeurs pour la famille royale. Ensuite, nous avons promené dans la défilé au le jour nouvelle.
What is even more, Maddie predominantly selected the correct tenses for the sequences of actions she described in her second text:
j’ai voyagé - nous sommes partis – je suis arrivée – j’étais très fatigué – nous avons visité – il y a avait – mon lieu favori était – j’aimais – nous avons promenés – il y avait – mon voyage … était … inoubliable
This consistently correct usage of tenses clearly represents a learning effect from pre- to posttest in a grammatical domain that represented an important part of this study.
Another evidence of progress made by students even in a field that was not part of the study is the one shown by Alana, who wrote a short text about a trip to Paris:
Alana: (pretest):
Le musée je préfère est le Louvre.
This usage of the French relative clause is incorrect because unlike the English, the French relative pronoun – in this case que – is mandatory. The error is therefore clearly caused by students’ underlying mental English grammar. Whereas this error, which is rather a serious one as French natives will get stuck when having to process the information, is the only contrastive one that Alana made in her pretest, there is no contrastive error at all in her posttest, which also stands for an improvement in her performance.
We see, then, that altogether, some learning improvement could be documented with regards to students’ practical approach towards contrastively relevant features of French. Even if there is only limited evidence of this kind because the two groups of students examined were rather small, these results appear to be promising: contrastive considerations seem to work out well regardless whether they are made explicit or left implicit. Should such evidence be generalizable one day, this would show how important contrastive considerations can be for the process of learning a foreign language.

5 Discussion and Conclusion
In the present study, the contrastive approach has proved to be nearly as implicitly successful as it is explicitly. Its mere use has the potential to foster learning processes in students even when this approach is not explained to them. Additional awareness raising by explaining its potential effects to students and by elaborating upon the cognitive and communicative sides of learning in general and the linguistic items in question in particular may have even more positive effects.
Even in our present time, with all its new developments in the field foreign language methodology, the use of the contrastive approach seems to be a highly promising and efficient way of teaching foreign languages. This may be the case because of the natural phenomenon that human beings compare structures of new languages to those of their mother tongue so as to have a sound basis to start from.
According to our results, the positive effect of the contrastive approach especially shows up in the free use of language, i.e. in free composition exercises, rather than in guided exercises featuring individual grammatical constructions. In translation exercises, the contrastive approach is not very successful because the negative transfer from the mother tongue to the foreign language is highly dominant there.
The very limitation of the present study is represented by the fact that the number of students forming the experimental and the control group, respectively, was extremely small. However, we refrained from including more students into our experiment, not considering even those who were of a language origin other than English but had lived in the U.S. for a certain number of years and had thus been under sufficient English influence to be counted in. Yet, these students would not have matched our research purpose satisfactorily, and their inclusion into this study might have watered down its potential results. Consequently, we decided to exclude them.
The present study has the following fortes:
  • The experimental and the control group were very homogeneous:
  • All students were of American origin, were born in the U.S. and were therefore undoubtedly native speakers of (American) English;
  • All students were not French majors (nor minors) and had a comparable "layman" status with regards to this language.
Against this background, the present study provides a solid methodological basis. In addition, the statements uttered with regards to students' individual results have been made with caution.
The findings of this study will have to be confirmed by further research projects including bigger student groups and covering other language pairs. The finding, however, that the contrastive approach seems to implicitly work in surroundings that are not optimal, i.e. that do not consider communicative and / or cognitive factors, is a remarkable one and, if confirmed by other studies, offers an immensely functional teaching approach.


Appendices
Appendix I: C-Test
On dit que le travail c’est la santé, mais quand on fait réellement le calcul du temps que l’on consacre à son travail, il y a de quoi se poser de vraies questions existentielles ! Selon un_______ étude bas_______ sur un______ durée moye_______ de vi_______ de 78 an_______, nous pass_______ environ u_______ an d_______ notre vi_______ pour no_______ rendre a_______ travail e_______ pas moi_______ de 90 000 heu_______ sur not_______ lieu d_______ travail.
Ou_______, nous pass_______ notre vi_______ à trava_______. Nous travai_______ parce qu’_______ faut s______ loger, s_______ nourrir, ma______ aussi po_______ payer no_______s sorties, décou_____ le mon_______. Bref améliorer, ou en tout cas maintenir, un certain confort de vie.

Appendix II: English-French Translation
1. Julie had a bad cold last week. Is she okay now?
2. A: Are you going to attend Professor Miller’s class?
   B: I don’t know yet.
3. I am very tired. I have to relax a little bit.
4. Unfortunately, I cannot meet with you now because I still have a lot of work. But I’ll be finished soon.
5. A: I’m extremely worried because I’m so bad at mathematics.    
B: Don’t worry, you are much better than you think. You will definitely pass the exam
next week.

Appendix III: Error-Correction Task
Find the mistakes in the following sentences. Say why this language usage is wrong.
1. Ce livre est très joli. Il vend vraiment bien.
2. Tu lis le livre que je voudrais avoir. Me le donne, s’il te plaît.
3. Je crois pas que tu me dis la vérité.
4. A: Est-ce tu as perdu quelque chose?
B: Non, je n’ai perdu rien.
5. Quand je t’ai prêté mon livre préféré, je ne savais pas que tu le garderais tellement longtemps. Me le rends tout de suite, s’il te plaît. (object pronouns).
6. Quand j’ai rangé mes choses, ma mère entrait dans ma chambre.
7. Je crois pas que je pourrais participer à la fête ce week-end.
8. Le soleil a brillé. Il y a eu une ambiance fantastique. Mais tout d’un coup, un bruit me réveillait de ma rêverie.
9. A: Si tu veux, je peux vendre mon téléphone portable à toi à un très bon prix.
B: Ce serait formidable. Je serais très heureuse si tu lui vendais à moi.
10. Avant-hier, Frank me promettait de m’aider. Hier, il ne le savait plus.

    References
Alatis, J. E. (ed.) (1968). Contrastive Linguistics and Its Pedagogical Implications. Report of the Nineteenth Annual Roundtable Meeting on Linguistics and Language Studies (Monograph Series on Languages and Linguistics; 21). Washington D.C.: Georgetown University Press.
Altenberg B. & S. Granger (2002). Recent trends in cross-linguistic lexical studies. In: Altenberg & Granger (eds.) Lexis in Contrast. Amsterdam: Benjamins, 3-48.
Baker, M. (1993). Corpus linguistics and translation studies: Implications and applications. In: Baker, Mona / Gill Francis & Elena Tognini-Bonelli (eds.). Text and technology. In honour of John Sinclair. Philadelphia & Amsterdam: John Benjamins, 233-250.
Baumann, K.-D. & H. Kalverkämper (eds.) (1992). Kontrastive Fachsprachenforschung. Tübingen: Gunter Narr Verlag.
Ballard, M. (ed.) (1995). Relations discursives et traduction. Lille: Presses Universitaires de Lille.
Balybina, Marina (2013). Phraseologismen in der spanischen und italienischen Sprache – eine korpusbasierte kontrastive Analyse. Hamburg: Diplomica.
Carroll, J. B. (1968). Contrastive Linguistics and Interference Theory. In: Alatis, J. E. (ed.) (1968). Contrastive Linguistics and Its Pedagogical Implications. Report of the Nineteenth Annual Roundtable Meeting on Linguistics and Language Studies (Monograph Series on Languages and Linguistics; 21). Washington D.C.: Georgetown University Press, 113-122.
Catford, J. C. (1968). Contrastive analysis and language teaching. In: Alatis, J. E. (ed.) Contrastive Linguistics and Its Pedagogical Implications. Report of the Nineteenth Annual Roundtable Meeting on Linguistics and Language Studies (Monograph Series on Languages and Linguistics; 21). Washington D.C.: Georgetown University Press (Monograph Series on Languages and Linguistics, 21), 159-173.
Chuquet H. (1991). Problèmes soulevés par la traduction de certains adverbes évaluatifs de l'anglais vers le français. In: Travaux linguistiques du CerLiCO. Presses Universitaires de Rennes 2, 161-188.
Colson, J.-P. (1993). L'usage moderne du passé composé en néerlandais et en français: une étude contrastive. In:Taal en leren. Een bundel artikelen aangeboden aan prof. dr. E. Nieuwborg. (Bibliothèque des cahiers de l'Institut de Linguistique de Louvain.) Louvain-la-Neuve: Peeters, 129-138.
Corder, S. P. (1967). The Significance of Learners' Errors. In IRAL, 4, 161-169.
Corder, S. P. (1974). Error Analysis. In: Allen, J. P. B. & S. Pit Corder (eds.). Techniques in Applied Linguistics (The Edinburgh Course in Applied Linguistics:3). London: Oxford University Press (Language and Language Learning), 122-154.
Coseriu, E. (1990). Science de la traduction et grammaire contrastive. In: Linguistica Antverpiensia XXIV, 29-40.
Cottier, E. (1992). L'opérateur causatif "make" et ses traductions françaises: dissymétrie avec le francais "faire". In: Guillemin-Flescher, J. (1992), 79-125.
Coulmas, F. (1979). Kontrastive Pragmatik. Kongressbericht der 9. Jahrestagung der Gesellschaft für Angewandte Linguistik. Stuttgart.
De Keyser, X. (1983). Contrastieve Kern Grammatica Engels-Nederlands. Alfabetisch geordend. Antwerpen.
Delin, J. / Hartley, A. & D. R. Scott (1996). Towards a contrastive pragmatics: Syntactic choice in English and French instructions. In: Language Sciences 18, 3-4, 897-931.
Di Pietro, R.J. (1971). Language Structures in Contrast. Newbury House.
Dirven, R. (1976). A redefinition of contrastive linguistics. In IRAL 14,1, 1-14.
Domínguez Vázquez, María José & Silvia Kutscher (eds.) (2016). Interacción entre gramática, didáctica y lexicografía. Estudios contrastivos y multicontrastivos. München: De Gruyter.
Ferguson, Charles A. (1968). Contrastive Analysis and language Development. In: Alatis, J. E. (ed.). Contrastive Linguistics and Its Pedagogical Implications. Report of the Nineteenth Annual Roundtable Meeting on Linguistics and Language Studies (Monograph Series on Languages and Linguistics; 21). Washington D.C.: Georgetown University Press , 101-112.
Fillmore, Charles J. (1984). Remarks on contrastive pragmatics. In: Fisiak (ed.) (1984), 119- 141.
Fisiak, Jacek (1984). Contrastive linguistics: prospects and problems. Berlin / Amsterdam & New York: Mouton De Gruyter.
Fisiak, Jacek. (1980). Theoretical issues in contrastive linguistics. Amsterdam: Benjamins.
Friederich, W. (1969). Technik des Übersetzens: Englisch-Deutsch. München: Max Hueber.
Gallagher, J. (1995). L'effacement des connecteurs adversatifs et concessifs en français contemporain. In: Ballard, M. (ed.), 201-220.
Gannuscio Vincenzo (ed.) (2015). Kontrastive Perspektiven im deutschen Sprach- und Kulturerwerb. Frankfurt a. M. & Bern. Wien: Lang.

 Gnutzmann, C. (1990). Kontrastive Linguistik. Frankfurt a. M.: Peter Lang.
Godin, P. (1999). Rhématisation du sujet grammatical en néerlandais et les correspondants en français et en anglais: étude de cas. In: Le Langage et l'Homme XXXIV, 1, 25-44.
Hartmann, R. R. K. (1991). Contrastive linguistics and bilingual lexicography. In: Hausmann, F.J. et al. (eds.) Woerterbuecher / Dictionaries / Dictionnaires. International Encyclopedia of Lexicography (Vol. III). Berlin / Amsterdam & New York: De Gruyter, 2854–2859
Hartmann, R. R. K. (1997). From contrastive textology to parallel text corpora: Theory and applications. In: Hickey R. & S. Puppel (eds.). Language History and Linguistic Modelling. A Festschrift for Jacek Fisiak. Berlin / Amsterdam & New York: De Gruyter, 1973-1987.
Hatim, Basil (1997). Communication across cultures: translation theory and contrastive text linguistics. Exeter: University of Exeter Press.
Hellinger, M. (1977). Kontrastive Grammatik Deutsch / Englisch. Anglistische Arbeitshefte 14. Tübingen: Niemeyer.
Heltai, P. (1988). Contrastive analysis of terminological systems and bilingual technical dictionaries. In: International Journal of Lexicography Vol. 1 (1), 32-40.
Hill, S. et al. (ed.). (1991). Bilingual Grammar of English-Spanish Syntax. New York: University Press of America.
Hoarau, L. (1997). Étude contrastive de la coordination en français et en anglais. Gap: Ophrys.
James, Caldwell (1994). Contrastive analysis. Harlow: Longman.
Johansson, S. (1998). On computer corpora in contrastive linguistics. In: Cooper, W. R. (ed.). Compare or contrast? Current issues in cross-linguistic research. (Tampere English Studies 6). Tampere: University of Tampere 269-89.
Klein, Wolfgang (1986). Second Language Acquisition. Cambridge: Cambridge University Press.
Klein-Braley, Christine & Ulrich Raatz (1982). Der C-Test: Ein neuer Ansatz zur Messung von allgemeiner Sprachbeherrschung. Bochum: AKS.
König, E. & V. Gast (2007). Understanding English-German Contrasts. Berlin: Erich Schmidt Verlag.
Krzeszowski, T. P. (1967). Fundamental Principles of Structural Contrastive Studies. In: Glotto-didactica, 2, 33-39.
Krzeszowski, T. P. (1990). Contrasting languages: The scope of contrastive linguistics. Berlin: Mouton de Gruyter.
Lab, F. (1994). Traduire le prétérit: imparfait ou passé simple? In: J. Guillemin-Flescher (ed.). Linguistique contrastive et traduction, Tome 3. Gap: Ophrys, 57-85.
Lado, R. (1957). Linguistics across cultures: Applied linguistics for language teachers. University of Michigan Press: Ann Arbor.
Mair, Christian & Manfred Markus (eds.) (1991). New Departures in Contrastive Linguistics. Proceedings of the Conference Held at the Leopold-Franzens-Universität of Innsbruck, Austria, 10-12. Innsbruck: Universität Innsbruck.
Nickel, G. (ed.) (1971). Papers in contrastive linguistics. Cambridge: Cambridge University Press.
Nickel, G. & D. Nehls (eds.). (1982). Error analysis, contrastive linguistics and second language learning. Heidelberg: Groos.
Richards, J.C. (ed.) (1974). Error Analysis. Perspectives on second language acquisition. London: Longman.
Robles i Sabater, Ferran / Daniel Reimann & Raúl, Sánchez Prieto (2016). Sprachdidaktik Spanisch – Deutsch. Forschungen an der Schnittstelle von Linguistik und Fremdsprachendidaktik. Tübingen: Narr Francke Attempto.
Spillner, Bernd (2016) (ed.). Kontrastive Fachsprachenforschung Deutsch-Chinesisch. Tübingen: Edition Julius Gross im Stauffenburg Verlag.
Stockwell, Robert P. / J. Donald Bowen & John W. Martin (1965). The Grammatical Structures of English and Spanish. Chicago: University of Chicago Press.
Teubert, W. (2002). The role of parallel corpora in translation and multilingual lexicography. In: Altenberg & Granger (eds.) (2002), 189-214.


Authors:
Professor Thomas Tinnefeld
Applied Languages
Saarland University of Applied Sciences
Business School
14, Waldhausweg
66123 Saarbrücken
Germany
E-Mail: thomas.tinnefeld@htwsaar.de


Frédérique Grim, PhD.
Associate Professor of French and Second Language Acquisition
Colorado State University
Department of Foreign Languages and Literatures
Fort Collins, CO 80523-1774, USA
E-mail: Frederique.Grim@colostate.edu


___________________

1  The university changed its name several times, now being the Duisburg-Essen University.

2  In order not to put too much time into the collection on supplementary data like students’ language proficiency level, the C-test, which should actually have 100 items, was reduced to 30 items. Although this scope is not an optimal one, it still elicits data that can be considered a sufficiently sound basis as long as they represent nothing but an information supplement to this study, which is the case here.

3  To keep students’ identities discreet, the names used here are not students’ real ones, neither the nicknames they may have indicated, but randomly used names.

4 In brackets: Students’ self-estimations: E=Elementary, L-I= Lower Intermediate; U-I=Upper Intermediate, the hyphen having been added for better readability. In cases when students were not sure about their language ability and made indications like “Upper / Lower Intermediate”, the respective lower level is documented here.

5  As not a single student indicated the level Intermediate, the levels, which should actually comprise 20 percentage points each, have been adjusted, with the Lower-Intermediate level being extended upwards by 10 points, and the upper-intermediate level being extended downwards by ten points.

6  Of course, the basis on which this statement is made is far too limited to be generalised. However, it can be expected that the results obtained in a far bigger and, then, more representative group would not be much different.

7  This does not mean that these students came up with the best results whatsoever. On the contrary, those who performed at a quantitatively high level, keeping this level from pre- to posttest, but not improving it, may have done even better. However, our aim here is to show how students who reduced their number of mistakes from pre- to posttest, did so qualitatively.
     In the same vein, we do not aim at exhaustiveness and will only analyse those answers given by students which are worth analysing here and not all the correct answers students may have given in the context of a certain task.

8  Although such an utterance can be heard in spoken French, it is not grammatically correct and should therefore not be taught.

9   See the previous footnote.

10 This correction is, of course, not right in the very sense of the word, but it has to be stressed that the student identified the tense, i.e. the passé composé, adequately, which, compared to a zero answer beforehand, already represents a considerable progress.

11 This sentence is not quite what could have been expected, but at least, it is not totally inacceptable in French, although it would normally imply a contrast (... non pas à Pierre, mais à moi).

12 This improvement by 0.1 points cannot be counted in because the difference between pre- and posttest is too small.

13 In the following examples, the contrastive errors are underlined. Any other errors that figure in these examples will not be identified nor discussed.

14 This is true in spite of all the collocational problems of the personification of the noun place that this underlying construction would represent, i.e. the logical impossibility for a place to 'prepare' something.

Posted by JLLT