It is common knowledge today that applied contrastive analysis, after having been very popular in the 50s and 60s, is regarded as inadequate and is no longer in use or present in teaching methods and materials. Contrastive analysis was used especially with the audiolingual method, in conjunction with structural linguistic theory, aural-oral procedures, and behaviorist psychology. From the behaviorists came the idea of allowing the students to develop the habit of transfer from the native language to the target language -via conscious attention to contrastive analysis-, rather than fighting this interference. But as foreign language teaching methods moved more towards the communicative and natural approaches, which are the leading methods today, contrastive analysis fell more and more into disrepute.

This paper intends to show that contrastive analysis, while in fact presenting serious problems when applied to languages that are not closely related, can be of pedagogical benefit when used in teaching a language that is close to the learner's native -or previously acquired- language.

Following Fisiak (2-3), there is a distinction to be made between theoretical and applied contrastive studies: whereas the former look for the realization of a universal category in languages A and B, applied studies «investigate how a given category present in language A is presented in language B». This implies that applied contrastive studies deal more with the comparison of surface representations of languages than do theoretical ones. Since we are dealing in this paper with two very closely related languages, Spanish and Portuguese, and the learning of the latter starting from Spanish, we will be concerned with the expression at the surface of basically the same set of grammatical concepts which are common to the two languages.

On the theoretical side, contrastive analysis was mostly treated in conjunction with various linguistic models, such as the generative-transformational grammar model. (cf. Di Pietro, 1971) But whereas in America contrastive analysis was criticized severely both from the theoretical and the pedagogical point of view (cf. Report of the 1968 Georgetown Roundtable), in Europe it did not lose its following: in fact, a number of organized projects were undertaken where English was described contrastively with respect to German, Serbo-Croatian, Polish, Romanian, Hungarian, Finnish, French, Swedish, Danish, etc., and there are articles being written on it to this day.

What are the main points of criticism leveled against contrastive analysis? In his article entitled «What a contrastive grammar is not, if it is», Eric P. Hamp (in: Alatis 137-47) states that contrastive analysis is necessarily bound to a polysystemic approach because all it can ever achieve is a comparison of subsystems once it gets into details, since each language is in its complexity a «system of systems» (141). He gives the example, taken from Halliday, McIntosh and Stevens, of comparing the possessive adjectives in English, French and Russian, and concludes: «The contrastive relation at issue remains far from formalized and largely anecdotal» (142). It seems that he can find as the only suggested criterion that of translation (143). Likewise, he dismisses Di Pietro's comparison (in his talk at the same Roundtable) of the phonetic r in Italian and English and states that it «will not necessarily bring out any good result at all in a pedagogical contrastive treatment» (145), as interesting as the comparison may be at the theoretical level. Hamp also rejects the idea of practicing contrastive analysis according to a specific linguistic model when it comes to the pedagogical application: «The rules of transformational grammar and their content do not by any means necessarily mirror the paths of movement through the speaker's head» (146). Or, as Marton formulates it, «grammatical operations as described by transformational derivations cannot be adopted as psychological models of the speaker's/hearer's performance... It seems that in an actual communication act the speaker/ hearer relies more heavily on a number of situational cues and expectancies than on any purely   —789→   grammatical analyses» (Nickel 185). These statements express the stance taken by the taxonomic/behavioristic linguistic school. Hamp advocates practical error analysis on the basis of actual observation in the classroom: «If we want to teach languages better, I would say, why not start with a careful and searching consideration of the learning mistakes that are made, and proceed to a body of increasingly predictive statements. Such bodies of statements might, of course, look nothing like a formally derived contrastive statement» (146).

All the above objections to contrastive analysis are, in my opinion, more or less valid in regard to this branch of linguistics in general, especially when languages with very different syntactical and semantic configurations are compared for pedagogical purposes. But for the case discussed in the paper, Portuguese for Spanish speakers, the situation is quite different, and contrastive analysis can be very useful in such a course.

These two languages are very closely related; in fact, their differences are mainly phonological and morphological. Teachers of either Spanish or Portuguese to English speakers know how problematic it is to make the students acquire an understanding, for example, of the use of the different aspects of the past, of the subjunctive mood, of ser and estar, etc. Since these basic semantic concepts are alike in Spanish and Portuguese, they can be left untouched, given that the English speaking student has already acquired them and the Spanish speaker knows them, and the phonological and morphological differences can be taught with the help of, among other things, simple contrastive comparisons.

But before we look at some of these differences in some detail, I would like to point out a few particulars about teaching Portuguese to Spanish speakers. How similar these two languages are was demonstrated recently in a paper by John B. Jensen, «On the Mutual Intelligibility of Spanish and Portuguese» (848-52), who found the capability of understanding when listening to an audiotape in the other language by natives of either language to be about 50%-60%. His work also supports the common belief that Portuguese speakers understand Spanish better than viceversa, although the difference was «not overwhelming». Jensen concludes that a course of Portuguese for Spanish speakers is the ideal setting for Krashen and Terrell's «Natural Approach» because «the instructor can immerse the students in natural Portuguese at a fairly high level from the start, confident that not only will they understand much of what is said, but that understanding will set in motion the natural acquisition process» (849).

In my view, the main criterion here is what we determine the goals of such a course to be, and I am concerned specifically about college level. If we want the Spanish speaker to be able to understand Portuguese or even to make himself understood in Portugal or Brazil, according to the above findings (which can be improved by simplifying the language and speaking slowly) then it seems hardly worth the trouble of such a course; the Spanish speaker will understand most of what is said in Portuguese and will be understood by simply speaking Spanish. If, on the other hand, we aspire to a certain degree of proficiency in reading and writing, and if we want the student to actually speak Portuguese as correctly as possible, then it seems useful, especially at the adult level, to point out the basic phonological and morphological differences so that the student may then make use of them creatively by process of analogy and generalization. Certainly there is the danger of faulty analogy and overgeneralization. That is why the main emphasis should be on those differences that have few or no exceptions, such as Spanish -ción corresponding to Portuguese -ção, or a limited number of alternate possibilities, such as Spanish ll which can convert in to ch, l, pl, pr: llave-chave, ella-ela, llano-plano, llanto-pranto (as listed by Takeuchi 182).

As W. Marton (Fisiak 147-70) points out and I think this is true especially for closely related languages-: «As the process of comparison is going to take place anyway, it is better to make it conscious and channel it to profitable uses, at the same time preventing distortion resulting from uncontrolled assimilation» (150). Marton also quotes Jakobovits's hypothesis «based on certain principles of transfer operation, and particularly on Osgood's three laws of interlist similarity. The hypothesis predicts that with unrelated languages a coordinate setting will yield less negative transfer than a compound setting, but that with related languages a compound setting will yield more positive transfer than a coordinate setting» (160). The compound setting tries «to make the native language a matrix of reference for the acquisition of elements and relations in the target language» (159).

As to the similarity of Spanish with Portuguese, N. Takeuchi, in his article about Spanish for Brazilian students, writes that these two are the Romance languages «que presentan mayor   —790→   transparencia lingüística». And he quotes Leo Gibson Ribeiro: «Son mellizas, aunque no siamesas» (Takeuchi 181). He calls the lexical similarity «el gran tropiezo», the confusion in the students' mind being caused by the fact that there are many identical or very similar words together with relatively few lexical differences. The student begins looking for a rule-oriented system behind the differences: «Ya los adultos lanzan mano del razonamiento y, consciente o inconscientemente, buscan explicaciones a los fenómenos lingüísticos» (182).

This is precisely the point where the language learner can make use of a contrastive grammar which Marton describes as «one of the reference books for language learning, whenever he really needs it» (Fisiak 168). In the case of Portuguese for Spanish speakers this «reference book» (such as one I experimentally elaborated for such a course), turns out to be not only relatively short, but it can also abide by the desirable condition that it concentrate on Portuguese and avoid, for the most part, the Spanish equivalent, at least where the morphology is concerned, as will be explained later. In our particular situation we can treat the student as an advanced learner, as described by Marton (Nickel 190-91) in that a Spanish speaker «is psychologically ready for theoretical explanations, since he already has the necessary conceptual framework in his mind into which new information can be fitted».

With this 'reference only' use in mind, we can refute L. Newmark's accusation that «in structurally oriented textbooks and courses, contrastive analysis leads to structural drills designed to teach a set of specific 'habits' for the well-formation of utterances, abstracted from normal social context» (51). Since, as Jensen pointed out, the instructor can (and should) «immerse the students in natural Portuguese at a fairly high level from the start», the contrastive grammar reference book will serve as a means to find explanations for what is heard: «The most important role of contrastive analysis -or rather, of the data obtained by contrastive analysis- is explanatory rather than predictive» (Catford in: Alatis 160). At the same time, it can help prevent what Lee calls «interference both from [the learner's native language] and, at every stage, from what has already been taught and absorbed» (Alatis 187). This takes us back to Marton's suggestion of making profitable use of the inevitable comparison of the two languages, «preventing distortion from uncontrolled assimilation» (Fisiak 150). The reference use of a contrastive grammar is a precautionary measure of remedying what Nickel calls «the greater difficulty to 'encode' than to 'decode' in a foreign language -especially one that is closely related to one's own» (Nickel 75). (Incidentally, an additional interference can occur when the student knows or is learning another Romance language, such as Italian or French.) All of the above allows us to conclude that, used together with the -in our case possible- unilingual immersion and contextualized teaching method, contrastive analysis proves to be a useful and valuable tool.

As to the practical application of contrastive analysis in a situation where the students' native language was closely related to the target language, we have the example of W. Boeddinghaus's experience of teaching German to both English and Africaans speakers (Nickel 21-31). He favors initial passive learning by having the students read sentences that contain words that they recognize and making them recover the meaning of the other words in the sentence by the context. He stresses the similarities between the two languages and advocates the use of a contrastive grammar which contains partial information about differences. By stretching the passive learning phase out, he intends to strengthen the students' self-confidence in the target language, thus perhaps trying to counteract what Takeuchi calls their tendency to overmonitor themselves (Takeuchi 183). Boeddinghaus admits to behaviorist ideas (success breeds success) and does not indicate how the transition from passive learning to active communication is supposed to be achieved. But he does exemplify a possible way of putting to good use the lexical similarities between two closely related languages for the purpose of learning.

***

The experimental workbook I devised for a course of Portuguese for Spanish speakers at the college level was written entirely in Portuguese. It contains chapters on phonological differences and also some on such basic sentence components as possessives, demonstratives, articles, etc. An initial indication that only differences between the two languages were mentioned in this book made the contrastive intentions clear to the student from the outset. Also, to cover the differences in verb morphology, a verb book was recommended as reference for verb conjugation in Portuguese.

The phonological aspects included: pronunciation rules for certain sounds (e. g. defining values for vowels in Portuguese which have multiple phonetic shapes per vowel, as opposed to basically just one per vowel in Spanish); and an itemization of correspondences between the two languages such as Spanish -ción- Portuguese -ção (atención-atenção), intial ll- ch- [llegar-chegar] (as one of the alternatives mentioned above); and -ue-o- (muerte-morte), -ie-e- (tierra-terra) where the students will not find the reverse problem experienced by Portuguese speakers who learn Spanish, as described by Takeuchi (182), who overgeneralize and produce sentences like: 'quiero un pueco de cueca',' instead of: 'quiero un poco de coca'.

The morphological differences involved mainly verb forms and other grammatical categories such as pronouns, prepositions, conjunctions, etc. The only syntactical aspects discussed were the uses of the personal (conjugated) infinitive and the future subjunctive in Portuguese. This workbook was not used in class, instead the material was incorporated in conversational exercises. The students were exposed to nothing but Portuguese from the very beginning of the course (one semester with three 50-minute class periods a week). Most of the time in class was spent with conversation, readings of short stories, watching videos, and games.

Toward the end of the course I audio-taped four class sessions in order to get a general idea on how much interference from Spanish there still was. This was not meant to be a statistical account, but rather a means of assessing which parts of the contrastive analysis that was consciously brought to the students' attention had been absorbed in a way that it became part of the active communication process. The results showed how the students had concentrated on certain differences, trying to use the Portuguese forms almost all the time, and had problems with other discrepancies. Here are some examples.

The different values of vowel were still not pronounced correctly, but there was only one mistake in all four sessions regarding the Spanish diphthongs -ue-, -ie-: one student said 'cuenta' instead of 'conta'. The intervocalic -s- was often not voiced (someone pronounced the Portuguese word 'ocioso' with the -c- and the -s- sounding alike). The verb forms were sometimes incorrect, but not through interference from Spanish; for example one student said 'fora' instead of 'era'. Hardly any mistakes were made with the morphology of contractions of prepositions with articles, demonstratives or personal pronouns (such as de + o = do), or even with the most difficult one to recover, por + o = pelo.

Sometimes the students simply used whole Spanish words in place of Portuguese ones, like 'perro' instead of 'cão'. What Di Pietro quotes Grauberg as saying, namely that «early problems appear to be more of a structural nature while later ones are primarily lexical» (Nickel 74), seems to have proven correct for this particular course, considering how few morphological mistakes were made as compared to the, albeit not too frequent, use of Spanish words.

Another noticeable type of mistake was made with genders of words, for example the use of o instead of 'a 'a árvore', clearly an interference from Spanish. Sometimes a student would use a Spanish word, such as in one case 'niño', however later in the same conversational turn she used the correct word 'menino' without having been corrected by the teacher. Occasionally there was overgeneralization, for example of the possibility in Portuguese of putting the pronoun after the conjugated verb, which is not always correct, for example in the case of 'não usa-se', i. e. when the verb is negated. Another instance of interference from Spanish was the overuse of reflexives, for example 'morrer-se'. But this was one case not treated in detail from the contrastive point of view because this would only have been possible through a complete list of these verbs, and I did not want to practice contrastive analysis by having the students memorize lists of lexical items.

The written performance by these students was outstanding. Here the combination of the contrastive grammar book and the constant exposure in class to spoken Portuguese and the readings showed remarkable results. The students did in fact make very active use of their contrastive workbook and easily applied all the descriptive rules of contrast it contained. The compositions written in class without any help (grammar, verb book or dictionary) were made up of grammatically well-formed sentences with very few mistakes.

All written homework was almost without mistakes. Here may come into effect what Takeuchi observed with Portuguese-speaking Brazilian students learning Spanish: that the similarity between the two languages leads to a more conscious effort not to make mistakes: «Se nos parece que la relativa identidad lingüística de las dos lenguas hace que los estudiantes se monitoren demasiado, hasta eviten exponerse a la posibilidad   —792→   del error» (183). To overcome this tendency in oral communication, the students must be exposed to hearing and reading in the target language as much as possible and must feel the incentive to communicate actively during most of the time in class, so that they gradually lose fear of making mistakes.

I have tried to point out through some discussion of principles and examples from a specific teaching experience that contrastive analysis should not be globally rejected. The literature on this subject shows a surprisingly small amount of works written on the comparison of closely related languages (cf. «A Bibliography of Applied Contrastive Studies» in: Fisiak 243-79). And, in my view, in the case of Spanish and Portuguese, there are valid reasons for incorporating contrastive analysis when speakers of one of these languages are trying to learn the other, simply because the similarities far outnumber the differences and the comparisons can be restricted to surface representations of basically the same grammatical concepts and diverging phonetic realizations from common roots. In the particular case discussed above, a course of Portuguese for Spanish speakers at the university level, some students even sought help in their knowledge of old Spanish, which they recognized as being even more similar to Portuguese than modern Spanish. Thus perhaps contrastive analysis should -for synchronic purposes- even be allowed to incorporate certain aspects of diachronic analysis, which is, in fact, nothing else but contrastive analysis of two very closely related languages.

Alatis, James E., ed. Report of the Nineteenth Annual Round Table Meeting on Linguistics and Language Studies. Washington: Georgetown University Press, 1968.

Di Pietro, R. J. Language Structures in Contrast. Rowley, Mass.: Newbury House, 1971.

Fisiak, Jacek, ed. Linguistics and the Language Teacher. Oxford: Pergamon Press, 1981.

Jensen, John B. «On the Mutual Intelligibility of Spanish and Portuguese». Hispania 72 (1989): 848-52.

Newmark, Leonard. «How not to Interfere with Language Learning». Methods that Work. Ed. John W. Oller, Jr. and Patricia A. Richard-Amato. Cambridge: Newbury House, 1983. 49-54.

Nickel, G., ed. Proceedings of the 3rd AILA Congress, Copenhagen 1972, Vol. 1. Heidelberg: Julius Groos Verlag, 1974.

Takeuchi, Nair Nodoca. «La semejanza con la lengua materna: tropiezos para el aprendizaje del español». Revista Letras (Paraná, Brasil) 33 (1984): 181-85.

Hispania [Publicaciones periódicas]. Volume 74, Number 3, September 1991