Florida International University

Are Spanish and Portuguese mutually intelligible?⁵⁰ Although they have a millennium or so of their own literary, lexical, grammatical, phonological and orthographic traditions behind them, the two languages are often held to be understandable to speakers of the other one.

People who make decisions about public language use seem not to be in agreement. During the World Cup Soccer Championships held in Mexico in 1986, with Spanish language interviews on most television newscasts, TV Globo in Brazil translated all Spanish interviews with subtitles, voice-overs, or consecutive interpretation by the reporter. Another Brazilian network, Manchete, did not translate at all. Television production companies in Colombia customarily broadcast interviews of visiting Brazilians, frequently soccer players, in Portuguese with no translation. The São Paulo International Airport makes virtually all announcements in Spanish, as well as in Portuguese, English, and French. The Rio de Janeiro International airport, on the other hand, does not routinely make announcements in Spanish.

A variety of opinions have been expressed in writing, by both linguists and non-linguists. The British linguist William Entwistle (1953 [31]), in describing what makes two speech forms languages rather than dialects stated that «mutual ease or difficulty of understanding is not the primary consideration. Norwegians and Swedes, Spaniards and Portuguese, can understand each other fairly well in their different languages».

A popular tourist guidebook (Jebsen and Biel, 1986 [329]) states the case for one-way understanding:

Most Portuguese have a fairly good, natural comprehension of spoken Spanish. But be forewarned that the reverse is not the case. Knowing Spanish will put you into a unique position for one-way communication -able to ask directions or make reservations but unable to understand the response.

A similar statement appeared in a recent Associated Press news article (Timberlake, 1989 [2a]): «The languages [Spanish and Portuguese] are closely related but quite different in pronunciation. The Portuguese can generally understand spoken Spanish, but most Spaniards can't understand Portuguese».

The last two statements relate to Continental Spanish and Continental Portuguese, and might not be made the same way for American Spanish and Brazilian Portuguese.

Finally, two specialists in Portuguese, Ellison and Andrews (1969 [259]), take a position for non-mutual-intelligibility:

Almost any informed native speaker of either language has an opinion on mutual intelligibility, generally affirmative but with reservations. Jokes and stories are often told involving misunderstandings in the use of Spanish and Portuguese by visitors from the «other country»⁵¹.

Aside from public language use, anecdotal expressions of opinion and advice to tourists, the question of mutual comprehensibility has a serious side for those involved in teaching Portuguese. One important issue involves teaching technique, particularly Krashen's concept of «input hypothesis» as an element

––––––––   849   ––––––––

of the «Natural Approach» (see Krashen and Terrell, 1983 [32-371). If a beginning Portuguese class consists of exclusively Spanish speakers, 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. If such is the case, a strong argument can be made for setting up special sections or a special intensive course for Spanish-speaking students to take full advantage of the capability of the students.

Perhaps a more problematical question involves the proficiency testing of Spanish-speaking learners of Portuguese. If speakers of Spanish (whether natives or successful second-language learners) understand spoken Portuguese without training or experience with the language, established proficiency guidelines become partially meaningless. For example, the 1987 ACTFL Guidelines include the following description for the Intermediate-Low speaking level:

If the student comes into a beginning Portuguese class with a high degree of understanding ability of the spoken language, then the tasks described above can very probably be accomplished by a reasonably attentive student in a couple of weeks of class, long enough to learn greetings, numbers and other expressions of «the most elementary needs». How then do we measure actual proficiency accomplishments? This is a matter for discussion elsewhere, (Jensen, forthcoming) but it points up the critical nature of the question of mutual comprehensibility.

Although we believe intuitively that there is a lot of mutual intelligibility between the languages, and may often confirm the belief through experience, I conducted a preliminary research project to seek objective answers to the following questions:

1. Are Spanish and Portuguese mutually intelligible?

2. If so, to what extent?

3. Is comprehension more likely in one direction than in the other?

4. If so, which way and to what degree?

5. What effect, if any, do other non-linguistic factors (attitude, age, sex, education and experience with the other language) have on cross-language comprehension?

I prepared a pair of listening-comprehension tests, one with recorded Spanish text and Portuguese questions and the other with Portuguese text and Spanish questions. The audio recordings included the following items, each about three minutes long:

1. A reading on Latin-American urban problems, originally composed in Spanish from a college Spanish textbook (Knorre, et al., 1985 [406-07]) and translated into Portuguese; read by male native speakers for each test.

2. A reading on Christmas customs, originally written in Portuguese from a college Portuguese textbook (Ellison et al., 1971 [497-98]) and translated into Spanish; read by female speakers on each tape.

3. A reading on Ecuador, originally in Spanish with Portuguese translation; read by female speakers.

4. A television news report consisting of short interviews of two people in each case. The Portuguese test was taken from a TV Globo (São Paulo) report on a suburban tree «murderer». The Spanish test was from a SIN broadcast on political turmoil in Bolivia. Because these were actual air-checks it was decided not to try to reproduce the same text in the other language, but to use two different authentic texts of similar difficulty level and style.

A written test was prepared consisting of five multiple-choice comprehension questions on each reading, given in the informants' native language. The questions were designed to require genuine understanding and could not be answered by just identifying key words in the text and matching them up with written answers. Some questions required a certain amount of mental processing to answer correctly, such as drawing conclusions from facts presented, easy tasks for native speakers, but demanding a high level of comprehension. The written tests were identical in the two

––––––––   850   ––––––––

languages, except for the questions dealing with the TV newscasts.

The written test also included a short questionnaire asking about age, sex, nationality, educational level, contact with the other language, and attitude toward the other language.

I personally administered all tests using the same tape player in each session and with the same oral presentation. The Spanish version was administered in São Paulo to students at the PUC and the USE The Portuguese test was given to students in the English Language Institute of Florida International University, most new arrivals from Latin America⁵². Thirty-nine valid questionnaires were obtained from the Brazilian group and 32 from the Spanish group, with the elimination of responses from informants who had extensive experience with the other language or who were of foreign (e. g., European, Asian or North American) background.

The overall results of the testing are shown in Figure 1, expressed as a simple percentage of correct responses, by group. The global scores are close for the two groups: 58% for the Brazilians and 50% for the Spanish group. Although apparently not large, the difference between the average scores of the two groups is great enough to be statistically significant, at the <.05 level, on both the Pearson's correlation test and the two-tailed t-test for means, meaning that the Brazilians' higher scores are probably not due to chance alone. (See the statistical results in Table 1⁵³.

Figure 1: Average Listening Comprension Scores (per cent)

All Four Readings Combined

Figure 2 shows individual scores for the four texts. On the first two of the four readings, «Urbanization» and the «Christmas Letter», the Brazilians have scores that are considerably higher than the Spanish group, with correlation coefficients quite a bit greater than those reported for the overall score and at a significance level of <.01.

Figure 2: Average Listening Comprehension Scores (per cent)

Each of Four Readings

On the fourth reading, the TV newscasts, correlation is stiff positive for the Brazilians, but to a lesser degree than on the first two and at a <.05 significance level. The difference on this reading is very possibly due to the non-identical nature of the texts in each language, so that the political commentary may well have presented a slightly greater level of inherent difficulty than the tree murderer report.

The results of test number 3, «Ecuador», appear anomalous, with reverse correlation, giving higher scores for the Spanish group than for the Brazilians. The most likely reason for the anomaly is fairly obvious: Spanish-speaking students, Latin Americans all, are much more likely to be familiar with Ecuador than are Brazilians. One of the informants, in fact, was an Ecuadorian, and his perfect «5» was disregarded. (His answers for other tests were included).

On the supposition that the Ecuador reading produced an aberration in the results due to its subject matter, the overall scores were recalculated without the Ecuador results. Figure 3 shows this recalculation of the overall scores (cf. Figure 1). Here with the compounding effect of three tests showing the same direction of correlation, the Pearson's r correlation of .444 is rather high at a significance level of <.01.

Figure 3: Average Listening Comprehension Scores (per cent)

Three Readings (without «Ecuador»)

The attitude question asked informants to give their opinion of the other language, using a five-point scale from «I like it very much» (=5) to «I can't stand to hear it» (=1). Brazilians showed a slightly more positive attitude toward Spanish than did Spanish speakers toward Portuguese, with average scores of 3.13 and 2.97, respectively. However, the difference was significant only at a level of <.20 so that at this point we cannot confidently reject the null hypothesis and recognize an attitude difference between the two groups. There was no significant correlation between attitude and comprehension score.

On the other informant questions, no significant correlation appeared between comprehension score and education or age. For sex and experience with the language, however, Pearson's r showed significant correlations. Because of possible interactions with the factor GROUP, a stepwise multiple

––––––––   851   ––––––––

regression test was run of GROUP, SEX, and EXPERIENCE with SCORE dependent. Table 2 shows the results of that test. GROUP and EXPERIENCE entered the model as significant predictors for SCORE, with 27% of the variation explained by them, while SEX showed no significance in predicting comprehension score. (The two groups were not balanced for sex, with most of the males in the Spanish group). The correlation of .304 at the <.01 level of significance, therefore, indicates that experience with the other language (use in the family, study, travel, etc.) is a positive factor in comprehension, surely no surprise.

The results suggest a conditional affirmative response to the first questions posed at the outset: Yes, Spanish and Portuguese are mutually intelligible, but at a level of only about 50% to 60%, at least as measured on this type of passive listening to electronically reproduced voices. It is precisely this type of listening, as mentioned above, that is involved in airport announcements and TV interviews, where policy decisions have to be made.

In a face-to-face conversational setting results may be rather different, perhaps aided by the visual contact and the direct transmission of the voice, but hindered by the need to respond actively upon but a single hearing and by a more casual speech style than that used in recorded readings.

It is obvious that level of comprehension is greatly affected by subject matter, and probably many other factors, as evidenced in the reversal of the direction of correlation shown by reading number three and by the variance among the other readings as well.

As to the third question, concerning the direction of higher comprehension, this work supports the common belief that Portuguese speakers understand Spanish better than viceversa. However, the difference is not overwhelming, and can be greatly affected by individual factors. The tendency is confirmed

––––––––   852   ––––––––

statistically, but it is perhaps an exaggeration (at least for Brazil) to assume that Portuguese speakers «have a natural comprehension ability of Spanish» but that their Portuguese responses will not be understood by Spanish speakers, as our guidebook indicates.

The informant background questions suggest that the only factor that correlates with comprehension is past contact with the language. A larger sample would be necessary to confirm any subtle difference that might exist for age, sex, or education.

The attitude question may present a surprise in not showing distaste for Spanish among Portuguese speakers, since many Brazilians, at least in the United States, express an aversion to Spanish. It is quite possible that this negative feeling is not common in Brazil, rather appearing among Brazilians after they come to this country and tired of being taken as Spanish speakers and being expected to speak that language. A survey of language attitudes among Brazilians outside of Brazil would be in order.

This research supports the cause of educators who fear the inappropriateness of standard ACTFL Guidelines in the proficiency testing of Spanish-speaking Portuguese students. The 40 to 50 percent of comprehension score attained by untutored listeners would have an enormous invalidating effect on any testing program that ignored the factor. On the other hand, the same level of comprehension can be a very positive force in the Natural Approach classroom and would support the designation of special classes designed to take full advantage of it.

ACTFL [American Council on the Teaching of Foreign Languages]. « ACTFL Proficiency Guidelines 1986». In Defining and Developing Proficiency: Guidelines, Implementations, and Concepts. Ed by Heidi Byrnes and Michael Canale. Lincolnwood, IL: National Textbook Company, 1987. 15-24.

Ellison, Fred P. and Norwood Andrews. «Portuguese in the High Schools». In A Handbook for Teachers of Spanish and Portuguese. Ed by D. D. Walsh. Lexington, MA, 1969.

Ellison, Fred. P. et al. Modern Portuguese. New York: Knopf, 1971.

Entwistle, William. Aspects of Language. 1953.

Jebsen, H. and Steven H. Biel, eds. Let's Go: The Budget Guide to Spain, Portugal and Morocco. New York: St. Martin, 1986.

Jensen, John B. «Evaluating Portuguese Performance of Spanish-Speaking Students». In Negotiating for Meaning: Papers in Foreign Language Teaching and Testing. Ed. by Dale Koike and Antônio Simões. Austin: TX. Papers in Foreign Language Education. 1989, forthcoming.

Knorre, et al. Pumas de Partida. 2nd ed., Instructor's ed. New York: Random House, 1985.

Krashen, Steven and Tracy D. Terrell. The Natural Approach: Language Acquisition in the Classroom. Oxford and New York: Pergamon, 1983.

Timberlake, Cotten. «Portugal Still Wary of Closest Neighbor: Report from Portugal». The Miami Herald, May 31, 1989, p. 2A.

Hispania [Publicaciones periódicas]. Volume 72, Number 4, December 1989