LÉO RAMOS CHAVESLabeling himself a country bumpkin because he was born in Araçatuba and grew up in São José do Rio Preto, the linguist Ataliba Castilho enjoys it when language changes. One of the most recent examples of such change is the transformation of the invariant Portuguese pronoun que [which] into a variant, as in Ques pessoas? [which-es people], a novelty identified on social media by one of Castilho’s doctoral students at the University of Campinas (Unicamp). In the professor’s opinion, “when grammarians tried to get people to obey rules, they didn’t see that they were telling Brazilians to shut up.” Castilho’s criticism of grammarians does not mean that he is averse to the norms of standard Portuguese, but only that he appreciates the language’s geographic and historical boundaries.
Professor at São Paulo State University (Unesp) from 1961 to 1975, Unicamp from 1975 to 1991, and the University of São Paulo (USP) from 1993 to 2006, Castilho has led major research projects that helped define the identity of Portuguese as spoken in Brazil. The most recent was the project “Towards a History of Brazilian Portuguese” (PHPB, in its Portuguese acronym), which brought 200 researchers together from around the country; their consolidated results are now being published. To better understand Brazilian Portuguese, Castilho developed a multisystem approach to language, which is a method of analysis where any linguistic utterance calls into simultaneous play four systems, which should be viewed integrally: lexicon, grammar, semantics, and discourse.
Castilho is married to Célia Maria Moraes de Castilho, likewise a linguist, and has three children and four grandchildren. His frequent travels took him to a conference in Lisbon in July 2017 and he will be attending another, in Coimbra, in October 2017. He received the Pesquisa FAPESP team at his home in Campinas, near Unicamp, from which he retired in 1991. He continues to work there as a voluntary collaborating professor.
| Age |
| 80 |
| Specialty |
| Linguistics |
| Education |
| Undergraduate degree in the classics (1959) and PHD in linguistics (1966) from the University of São Paulo |
| Institution |
| Unicamp |
| Scientific production |
| 62 articles; master’s and doctoral advisor; author of Nova Gramática Do Português Brasileiro (Contexto, 2010), among other books |
How are things going with “Towards a History of Brazilian Portuguese,” one of your most recent projects?
That project began at USP in 1987. We had initiated studies on spoken language in the 1970s and when the results started emerging, I asked my colleagues where all that had come from. So we began reconstructing the history of the establishment and development of Brazilian Portuguese, using the same method applied in earlier projects: forming work groups, setting timetables, holding national seminars, and publishing results. Ten volumes of studies were released, and now the consolidated results are coming out in 12 volumes, seven of which should be published in 2017. The first will be volume four, on the history of nouns, adjectives, and prepositions; Célia Lopes, of the Federal University of Rio de Janeiro, is the editor. Did everything come from European Portuguese, or did we invent something? A PhD student whom I’m advising at Unicamp, Flávia Orci Fernandes, searched social media and found plural markers on the pronoun que [which], which no longer is an invariant. Nowadays people say and write ques pessoas [which-es people]. How do we explain this dynamic aspect of language? We have to document and explain it. Changes in language create new rules. The educated haven’t accepted this yet, but it’s a matter of time; they’ll eventually accept it. The language is entering a new stage.
What other changes are you seeing?
Our studies have indicated that the plural will no longer be expressed by adding the morpheme ‘s’ to the end of a noun but by adding it solely to the article—as in os menino [the-s boy]—which is the result of a phonological mechanism involving open syllables. It seems that consonants are changing more than vowels. Isn’t fazer [to do] turning into fazê? Other final consonants are dropping off, like duas vez [two time] instead of duas vezes [two times]. Both words are marked for the plural, which is redundant; it might be more economical to mark only the first element and leave the other alone. French still maintains final consonants in writing but not in speech anymore, as in l’enfant, les enfants [the child, the children]—the ‘s’ isn’t heard in spoken French. We believe we’ll end up with another language, unlike Portuguese from Portugal. The two languages are still the same, but they’re already displaying many differences. Using statistical methods to project the pace of these changes, we find that 200 years from now, the Brazilians and the Portuguese won’t be able to understand each other anymore (see Pesquisa FAPESP Issue No. 230).
What do you find most fascinating about Brazilian Portuguese?
The incredible variations in accents and ways of using the language. This is a consequence of our history. Since the Portuguese came over in a number of waves, each region developed its own modality of the same language. My wife, Célia, has been studying who these Portuguese were and where they came from. They began arriving in São Vicente in 1532, and it was in what is now the state of São Paulo that the Portuguese began colonizing Brazil. The Portuguese reached Rio de Janeiro, Salvador, and Recife in the 16th century. In the 18th century, Santa Catarina and Rio Grande do Sul received more waves of Portuguese, from the Madeira Islands and the Azores—not the continent, like the previous century. To learn what the Portuguese language was like in the 16th and 17th centuries, Célia looked at the authors of wills left by the Portuguese. She found that 20% of the authors were new Christians—Sephardic Jews—who had gone to Portugal after being expelled from Spain. Few people knew how to write back then. She separated out the writings by these Jewish immigrants, who lived in isolation. The women never left the house and their Portuguese was very conservative. The men had a more conservative side, used inside the family, and another, less conservative, used in doing business. Célia compared the writing of the new Christians and the Portuguese and separated the era’s modern language—used by the Portuguese—from the conservative version, a linguistic variation. She also studied specific features of grammar to uncover the historical basis of caipira [rural or, literally, hillbilly] Portuguese, which preserves 15th-century Portuguese. Our pronunciation known as the caipira ‘r’ is a remnant of the conservative Portuguese spoken then.
Is the caipira ‘r’ 500 years old?
No one knows for sure. Some say that Indians from the Paraíba Valley used this ‘r’, but this explanation would only hold if it were proven that the tribe extended over São Paulo and followed frontier adventurers into the interior. Other phonologists think the caipira ‘r’ is a feature that didn’t develop within Portuguese phonology but might have appeared naturally, as a consequence of the phonological system, and not through contact with the Indians. A phonological system is a system of sounds. There are always tendencies to combine sounds that account for the changes. One tendency in Brazilian Portuguese has to do with how sibilants are dealt with. It’s what distinguishes São Paulo natives from people living in Rio de Janeiro, the South, and the Northeast. Where does the palatalization of the final ‘s’ come from, as in ash crianssash [as crianças; the children]? From the tendency to palatalize the sibilant. This has gone a step farther now, because palatal vowels are becoming diphthongized, as in aish pessoaish (as pessoas; the people).
Why were Portuguese and grammar classes always so boring and full of rules?
Because linguistics didn’t exist yet, just school grammar. Grammarians take a very narrow view of language. They see themselves as having to defend language against change. Linguists also talk about rules because they feel it’s their duty to explain them, but they don’t stop there, and they see the rules as fascinating. Why is it like that? Was it always? Is it like that all over? Grammarians stick to the application of rules and see new things as wrong and old as right. The problem is that when grammarians tried to get people to obey linguistic norms, they didn’t see that they were telling Brazilians to shut up. As if they were saying: “You have to speak and write according to the rules. Don’t speak wrong!” And because people are afraid to fail, they don’t speak or write very much. This is a consequence of preaching traditional grammar. It hasn’t been a good thing. People have to feel at ease about expressing themselves, taking part in discussions, developing a democratic spirit. Language belongs to the speaker, not to the grammarian. We learn language with the speaker, as he speaks it, and we try to figure out why he’s speaking like this or that. Did the expression come from Latin? Was it created here? And we can use the structures, not just classify and name them. What’s the point of only knowing the sub-classes of subordinate clauses? It’s not a scientific attitude, an attitude of discovery, to say that someone is speaking incorrectly. Linguistics has replaced the order to shut up with the pleasure of scientific discovery. It was only through linguistics that our vision expanded and we began considering any topic worthy of study.
How did this change come about?
Linguistics started in Brazil in the 1970s. It was quite an upset, because prior to that, only grammarians studied the language. In terms of linguists, there were Joaquim Mattoso Câmara Júnior [1904-1970], in Rio de Janeiro; Theodoro Henrique Maurer Júnior [1906-1979], in São Paulo; and Rosário Farâni Mansur Guérios [1907-1987], in Curitiba. It was Mansur who steered Aryon Dall’Igna Rodrigues [1925-2014] towards indigenist studies when he said: “In Portuguese, you’re only going to rediscover what we already know, but not in indigenous languages, because they’re not Indo-European and they have completely different solutions and categories. That would be something truly new.” He was right; in terms of structure, indigenous languages have categories that nobody can imagine. And Aryon inaugurated indigenous linguistics in Brazil. A linguist from the United States, Daniel Everett, came to Unicamp in the 1970s and then studied the language of the Pirahã community, in the state of Amazonas. He butted heads with U.S. linguist Noam Chomsky, who said that what we call recursion is universal. Recursion refers to the fact that we can apply the same grammar rule multiple times when building sentences. In Portuguese, whenever I want to put something in the plural, I add an ‘s’. But Everett didn’t find any recursion in the Pirahã language—which also has no words for numbers or colors. Chomsky had to recognize this exception to a rule he thought was general. Brazilian linguists can contribute a lot to the general theory of languages. Today there are 160 indigenous languages in Brazil. There used to be twice as many, but the Indian speakers have died. Of these languages, only 60 have been described.
In the 1970s, you began the project “Standard Urban Portuguese” (NURC). Tell us about it.
The idea was to describe standard spoken Portuguese. It was surprising, because standard Portuguese contained a lot that had been condemned by grammarians, showing that our catalogue of “mistakes” didn’t take into account how the Portuguese language is really used in Brazil. The project began with a Spanish professor from El Colegio de México, in Mexico City, Juan Miguel Lope-Blanch [1927-2002]. He worked in dialectology—in other words, he described the language of rural regions—until he realized that people were migrating to cities. So he decided to do urban dialectology. In the 1960s, Lope-Blanch launched the idea of studying the standard urban language spoken in federal capitals, not only Spanish but other languages across the Americas and in Portugal. This proposal reached Brazil through another dialectologist, Nelson Rossi [1927-2014], of the Federal University of Bahia, who compiled the first linguistic atlas of Brazil. At a meeting in São Paulo in 1969, Rossi said that Lope-Blanch wanted to study the language of federal capitals, but in Brazil, our capital, Brasilia, wouldn’t serve as an example of relevant linguistic fact, because it was very new. So we chose five state capitals: Recife, Salvador, Rio de Janeiro, São Paulo, and Porto Alegre. We followed precisely the same methodology as the original project. NURC did a very good job recording interviews and then doing transcriptions, but it fell short in its descriptions. A huge corpus was compiled from 1,500 hours of recordings. But when it came to describing structures—phonology, morphology, syntax—it didn’t work, because the questionnaire used in the interviews lacked theoretical consistency; it was cumbersome. I applied the questionnaire to study aspect and verb tenses, and I saw that it wouldn’t produce any results because each question corresponded to one theory, different from the theory underpinning the next question. I wrote a paper about how the questionnaire wasn’t workable for the most important step: description, knowledge. In 1981, I read the paper at a conference at Cornell University, in the United States. I thought Lope-Branch would kill me, because he was always very forceful when he had something to say, but I was bowled over. Do you know what he said? “You’re right.” When we couldn’t reach a decision, we’d put the matter to a vote, as if science were a democracy. It’s not. You’ve got to be coherent with your concepts and not decide by vote. After what he said, I came to the conclusion that we needed to leave the project. But at the same time, we couldn’t just ignore the wonderful quantity of data produced by NURC.
Personal archives
Castilho with classmates João de Almeida (left) and Wladimir Olivier (right) at their graduation from the USP School of Letters in 1959Personal archivesSo what did you do?
I came back disillusioned by the project’s direction, but then I got the idea of reaching out to the best linguists in Brazil, who weren’t taking part in NURC yet. I reached out to Mary Kato and Rodolfo Ilari, of Unicamp; Leda Bisol, of PUC [Pontifical Catholic University], in Rio Grande do Sul; Luiz Antônio Marcuschi [1946-2016], of the Federal University of Pernambuco; and a number of others. I explained that NURC had hit a wall because its methodology lacked consistency. I asked if they would agree to do a grammar using material from NURC. So then they asked me what kind of grammar I wanted. I said, “Not what kind I want; what kind we want. It’ll be a group effort.” So I launched the idea that we each go off to our own corner and write a paper entitled “My grammar as I conceive of it”—precisely in that stilted language. Groups with theoretical affinities gathered spontaneously—generativists, functionalists, structuralists, each in its own corner. Each group wrote something that ended up forming a work group for the project, and we divided up according to these sets of ideas and planned to meet once a year to discuss our results, in all frankness. After our discussions, we each took what was left of our papers—because the discussions were really hard-hitting—and then we redid the essays. They were published next, forming a nine-volume collection. Then came the job of consolidating the results of the grammar itself. The first of the eight-volume Gramática do Português Falado [Grammar of spoken Portuguese] was released in 2006 by Unicamp’s university press. It was entitled A Construção do Texto Falado [Construction of the spoken text] and was edited by Clélia Jubran, of Unesp, and Ingedore Koch, of Unicamp. They devised a theory to explain the features of oral texts. Seven volumes came out in a second edition, published by Contexto between 2012 and 2016. As a result, Brazilian Portuguese became the only romance language in the Americas to have its standard version broadly documented and described.
And what are its features?
Spoken language is hesitant, interrupted, redundant, unplanned, fragmented, incomplete, not thought out much, with little informational density and short simple sentences. We talk and create at the same time. Another feature are discourse markers, like tá? [okay] and né? [right], always at the end of sentences. So there’s a placement rule. Spoken language displays different regularities than written Portuguese. Another thing is that we learn the spoken modality first and then the written language. It seems like a silly detail but it makes all the difference. Written language comes later, imposed on us, because we learn spoken language in our family and written at school. All of this makes for a tremendous difference between the two modalities. Traditional grammar is only concerned with written language. If the description of a language is concentrated on its written form, then I’d be grabbing the point of arrival not of departure, and I’d be taking as true a bunch of misconceptions about how the language works. I was bitten by this bug in the 1980s, and I asked myself what theory I could draw from it. I devised the multisystem approach to language, which I used as the foundation for Nova Gramática do Português Brasileiro [New grammar of Brazilian Portuguese]. I reworked it for a presentation at the 11th Lusistanistentag—the day of scholars of the language, literature, and culture of Portuguese-speaking nations—held two years ago in Aachen, Germany.
How do you define the multisystem approach?
It’s very simple. All of linguistics has always revolved around three axes: phonetics and phonology [the study of speech sounds], morphology [the study of word forms], and syntax [the study of relations between words]. In addition to this system, which constitutes grammar, we have semantics, which deals with meaning; discourse, which is how people compose a text; and lexicon, that is, words. So we’re talking about language as a set of four systems: lexicon, grammar, semantics, and discourse. To describe a phenomenon thoroughly, I have to rely on these four systems: lexicon, semantics, grammar, and discourse. Can this be done by one person alone? No. The study of a language has to be a group effort. This is the corollary of NURC, the Gramática do Português Falado, and the project “Towards a History of Brazilian Portuguese,” which also fosters exchange between people with different views, including sociolinguists, generativists, functionalists, and cognitivists.
How do you bring people with different perspectives together?
I respect different ways of thinking. And when you show respect, you bring people together. No one wants to get beaten up by others. Science isn’t meant for that purpose, but for uniting people in the discovery of knowledge. Some people find it easier to do this and others, harder. When things get stuck and opinions seem irreconcilable, sometimes I’ve got to remember: “And our obligation, our duty?” I bring out my Presbyterian discourse. I was raised in that church in São José do Rio Preto.
Who was your first Portuguese teacher and what was she or he like?
It was Amaury de Assis Ferreira [1920-1995], father of Brazilian television host Amauri Jr. He was a really good teacher. He read and studied a lot, and he’d get all excited showing us the books he’d bought. I’d go to his house once in a while; my dad was an electrician and he’d go to change the element in his electric stove. He’d call me over and show me his library and the books he’d bought. He really enjoyed what he did. I thought, “I want to be a guy like this.” I had other wonderful teachers in São Paulo, like Theodoro Maurer, my doctoral advisor. Quiet type, skinny, the son of Swiss immigrants. Working alone, he wrote one of the world’s most thoroughgoing works on the grammar and syntax of Vulgar Latin. He also engaged in another type of leadership, which I discovered by accident when I was walking in the neighborhood of Consolação, in São Paulo. As I passed a house, I heard some Presbyterian hymns. I peeked in the door and there was Maurer and a professor of philology, Isaac Nicolau Salum [1913-1993]. They were studying the Gospel of St. Matthew, everybody reading Greek! I was studying the classics, Latin and Greek, and I was amazed to see the professors discussing in the original Greek. I asked him, “Why don’t you invite your students?” “I can’t,” he said. “It’s a lay university.” In addition to being a professor of romance languages and an evangelical pastor who knew Latin, Greek, and Hebrew, he was president of the São Paulo chapter of a political party. And he didn’t say anything about this at the university.
Did you attend USP on a scholarship from the City of Rio Preto? I did. My family was quite humble and I had to pay room and board in São Paulo. A classmate of mine said that the City of Rio Preto was awarding scholarships to students who got into either USP or what was then called the National University of Rio de Janeiro. Shortly after I was accepted, in 1956, I took my enrollment form in and they gave me a scholarship on the spot. It was 200 cruzeiros, paid out in a lump sum, to last the whole year. Inflation rose in the third year and the scholarship only covered half the year. Then I started teaching, to supplement these funds. In 1959 and 1960, I taught Portuguese at Francisco Roswell Freire, a state high school in São Miguel Paulista, and in 1960 I taught Latin at the Suzano state high school and teacher’s college, in Greater São Paulo. I loved the experience. São Miguel was an industrial neighborhood, and for its students, the school was their way of getting into a profession and out of that world. They treated the teachers really well. The first thing I noticed when I got to São Miguel was that the school had no library. How could I teach Portuguese without a library? But there was a kind of school association that raised money from the parents who could afford to give. I asked the principal if I could use the money to buy books and he said yes. I lived on Guaianazes Street, across from the publishing house Editora Nacional, and I’d buy the books at a discount and leave them at school, where a student looked after them. I’d buy historical novels by Paulo Setúbal [1893-1937] and Monteiro Lobato [1882-1948] and some titles from the Book Club collection. The students liked that a lot. I met my wife at São Miguel. She was studying there, although she wasn’t my student.
What did you teach at São Miguel?
I copied my teacher from Rio Preto. I tried to give lively classes. I assigned students homework, showed appreciation for what they did, put the pressure on when something wasn’t right. I didn’t teach much grammar. I did what the program required, but emphasized reading. I was later invited to work at what would become the Marília campus of Unesp. A small town, a small group, with people searching for their paths. The Latin professor, Enzo Del Carratore, had been my classmate at USP. We were all young and tried to establish a work program. We looked at what they were studying at USP. Historical linguistics? So let’s do descriptive linguistics. They were focusing on written language? So let’s study spoken language. That defined our lives, because we soon decided what to do. We wanted to research different topics. At our invitation, great linguists—Maurer, Mattoso Câmara Júnior, Nelson Rossi—came to read their papers.
What was your participation at the Portuguese Language Museum?
In 2004, Jarbas Mantovanini, who worked at the Roberto Marinho Foundation, came to USP, presented the museum project, and said he wanted to ask two things of me. First, he wanted me to offer some ideas for the museum; then to draw up a timeline on the history of Portuguese. Aryon was going to cover the part on indigenous languages, while Yeda Pessoa de Castro, of the Federal University of Bahia, would address African languages. Jarbas told me to invite whoever I wanted to. I invited Mário Viaro and Marilza de Oliveira, both from USP, to cover the other parts. Jarbas asked me how I wanted to represent the timeline, using films or panels. I preferred panels, because films were going to be shown on the other side of the room. I delivered the project, and he liked it. “It’s all very lovely, but I’m going to replace the last panel with a mirror. Everybody is going to travel that incredible 2,000-year history and when they get to the end, they’ll see themselves.” You know, he hit the nail on the head. A lot of the folks who saw themselves after taking this historical journey broke into tears. A colleague from Minas Gerais, Maria Antonieta Cohen, at first went to see the museum and then to see the people when they got to the mirror. She asked me: “Why do you suppose they cry?” I thought about it a lot. People certainly cried because they saw their identity there. What is the Portuguese language? It’s me, who now represents this whole journey. Language is my identity.
