VARIABLE MORPHOSYNTACTIC PHENOMENA: EVALUATION AND TEACHING

ABSTRACT

LABOV, 1972

sociolinguistics; morphosyntactic variation; norm; evaluation; teaching

RESUMO

Este artigo visa a analisar como os professores e os falantes da Língua Portuguesa em geral avaliam estruturas morfossintáticas variáveis e qual estatuto cada uma delas recebe (estereótipo, marcador ou indicador). O corpus é constituído por uma redação corrigida por dez professores e outros dez falantes cariocas. A investigação fundamenta-se, sobretudo, nos pressupostos da Sociolinguística Variacionista (WEINREICH; LABOV; HERZOG, 1968; LABOV, 1972) e objetiva (i) definir o estatuto atribuído a certas estruturas morfossintáticas; (ii) verificar se essas estruturas são ou não estigmatizadas, descrevendo o que os indivíduos cultos tomam como norma-padrão e (iii) estabelecer uma relação entre o problema da avaliação e o ensino de Língua Portuguesa. A partir da análise realizada, verificou-se que, no âmbito do estabelecimento da norma-padrão, não há compatibilidade absoluta entre as escolhas dos participantes e o que se propõe usualmente como norma gramatical; em outras palavras, os participantes, professores ou não, usam, para corrigir a redação, critérios particulares em sua idealização de padrão linguístico (norma-padrão), e não apenas o que se divulga em manuais tradicionais (norma gramatical). Com base em tais resultados, o artigo propõe que, em contexto escolar, sejam consideradas as variedades cultas como ponto de partida para o estabelecimento da norma-padrão.

sociolinguística; variação morfossintática; norma; avaliação; ensino

Initial considerations

This article presents the investigation on how Portuguese Language speakers evaluate variable morphosyntactic structures. It also aims at reflecting upon how they estimate them, concerning their prestige/disregard in terms of normative orientation, according to the assumptions of Variationist Sociolinguistics (WEINREICH; LABOV; HERZOG, 1968; LABOV, 1972, 1994). Based on Variation and Change Theory, this study approaches mostly the evaluation of the variants problem, one of the five theoretical questions elaborated by Weinreich, Labov, and Herzog (1968) for scientific treatment of linguistic change.

For the sake of the research, the analysis of a school composition is made, which was elaborated with regard to contemplating a set of morphosyntactic structures seen as discredited, conforming to the grammatical norms, and supposedly considered as mistakes by Education professionals during the correction of compositions written by students. The referred composition that was idealized for the context of the investigation, 8th grade from a public Middle School located in the city of Rio de Janeiro periphery, was corrected by twenty natives, ten of them were Portuguese Language teachers, and the others work in many different areas with a finished or unfinished graduation.

During the data analysis, it is considered, based on Faraco (2008, 2015), the existing differences between the grammatical norm and the supposed norm in which is grounded the correction of the text – therefore, stemming from them the possibility to interpret what would characterize the school standard norm. In other words, despite determined structures were not taken as correct according to the grammatical norm, the one prescribed in traditional manuals, they could not be considered by teachers or by speakers as incorrect, possibly since they productively are already part of the standard varieties meaning a concrete and objective norm used in speaking and even in writing by individuals with a high level of education (educated variety).

Therefore, the investigation has formulated, as a whole, the following goals:

1. observing in a set of morphosyntactic phenomena the most noticeable structures and susceptible of correction by education professionals and other speakers, and which ones are not;

2. from this evaluation, considering the status of specific morphosyntactic structures as possible stereotypes, indicators, or markers, in order to describe what the research participants take as the standard norm;

3. establishing a relationship between the Evaluation Problem and the Portuguese Language teaching, based on the variable morphosyntactic structures analysis, then associate the results to the normative orientations of Vieira (2019a) for the educational context.

Concerning this, the text is developed in four sections, besides this introductory one. In the first section, the theoretical aspects that supported the analysis are summarized. In the second section, a short piece of methodological information, the hypothesis, and the results taken from the analysis of a school composition’s correction is presented. The third section is dedicated to the discussion of results, including reflections concerning Portuguese Language teaching, before the final considerations in the last section of the article.

Theoretical foundations

In order to carry out the current investigation, it was adopted the fundamental tenets of Variationist Sociolinguistics, including the principle of orderly heterogeneity, which posits that systematic language variation is intrinsic to language, and the Evaluation Problem, which proposes that the judgment of the speaker over other variants – which can be investigated through studies on beliefs and attitudes (as the ones proposed by Cyranka, 2011, 2014) – could accelerate or restrain a linguistic change process.

Additionally, since the norm approach is in an educational environment, the study shares, to a certain extent, the concerns of the Educational Sociolinguistics – which, among its assumptions, considers the variable usages in classroom spread over the continua of variation in Brazilian Portuguese (BORTONI-RICARDO, 2005) – or the Pedagogy of Linguistic Variation (FARACO, 2008, 2015). In this regard, the notion of diglossia (MATTOS E SILVA, 2004), just like the proposal of the coexisting and different kinds of knowledge concerning Portuguese Language lessons (BARBOSA, 2007) are also used in this current article and applied in the data analysis.

Concerning the linguistic change, Weinreich, Labov e Herzog (1968) formulate five essential questions to the variationist research: constraints, transition, embedding, evaluation, and actuation. Resuming the main subject of the current investigation, it is assumed that the social evaluation the speakers do, regarding a determined variant, in consonance with the judgment about a variety/norm used by the interlocutor, is extremely relevant in education and practices that have been established in the educational context. Hence, as already presented, if a particular variant is seen as prestigious (“beautiful”, “elegant”, “correct”, and other attributes), such appraisal can contribute to a disseminated and preserved usage. Nevertheless, if the speakers negatively conceive the variant, its usage is appraised as “ugly”, “inelegant”, or “incorrect”. This value judgment can interfere in extending its use and restrain its implementation, consequently delaying the change process.

Considering this, the sociolinguistic studies on beliefs and attitudes are essential, since they seek to understand the variation phenomenon by the awareness of a speaker about language, variety, or even a determined variant. From this viewpoint, in the scope of professional development in Education, Cyranka (2011) explains that:

[...] the study on linguistic beliefs and attitudes, amongst other research subjects, would provide a better understanding and auscultate the difficulties of undergraduate students and professors. On the one hand, this difficulty occurs concerning the problem frequently referred to on acknowledging the legitimacy of linguistic variation or dialect of students, not counterpoising or not simply imposing them the standard norm with no discussion. On the other hand, there is a mistake in adopting writing as the only legitimated language skill. (CYRANKA, 2011, p. 45, our translation)¹.

According to the author, the positive, negative, or neutral attitudes constitute speakers’ reactions towards certain people, situations, and things. In this regard, they consist of what individuals think and feel, revealing how people react to linguistic stimuli presented to them. Also, as stated by Cyranka (2011), beliefs, in turn, consist of personal convictions regarding the speaker’s opinion about the evaluated objects. Through this perspective, the attitudes of individuals towards a communicative event, things, or people come out from their beliefs; in other words, the beliefs and attitudes are related to one another, as the latter is, in some aspects, consequences of the former.

In the scope of linguistic education, it is essential to consider these conceptions in Portuguese Language teaching, as the teacher, depending on the instructions offered concerning linguistic variants, could contribute to certain negative beliefs and attitudes from the student when facing their performance in language use. Taking this fact into account, Cyranka (2014) claims that

[...] whatever the path chosen by the teacher is, in order to take their students to engage in the process of developing reading and writing competencies, teachers must be careful while getting them to, as said before, build up positive beliefs about their potential. The starting point is acknowledging their competence over a complex system that they have acquired naturally and with which they are capable of acting in their social environment. Subsequently, oriented by the school, they must strive for further sophisticated linguistic resources for which they need to be trained and prepared. (CYRANKA, 2014, p. 134, our translation)².

Undoubtedly, it is required for the teacher to opt for a methodology that leads the students to reflect upon the language used by them to understand its variable functioning and be able to use it according to the communicative situation. Based on this proposition, the present research is focused on the attitudes and beliefs of speakers of Rio de Janeiro city concerning certain variable linguistic structures, specifically in the morphosyntactic level.

The treatment to evaluate the variants begins by the typology developed by Labov (1972), according to which the alternating forms have different types based on the appraisal made by speakers that could be denominated as stereotype, indicator, or marker.

Regarding the concept of stereotype, it is characterized by linguistic forms socially marked, labeled ostensibly by society, creating structures that are strongly stigmatized. It has to be said that such variants – that tend to be the object of linguistic correction – are stigmatized under the perspective of an individual speech community, being the attribution of this status highly acceptable. Relying on the judgment (point of view) of the interlocutor, a variant can be stereotyped negatively, characterizing a person as, for example, uncultured, incapable, less intelligent, or even overly educated, archaic, among other possibilities. The indicators, in consonance with Labov (1972), differently from the stereotypes, are not perceived in the consciousness level, composing structures that are not exclusively framed in a social or stylistic matrix. Therefore, they are not liable to be corrected. Compliant with Gorski & Coelho (2009, p.82, our translation), indicators “present a scarce evaluative strength. They are traces that do not have social diversity and, generally, present a regular distribution over different social strata.”³. The markers consist of variable phenomena that are not usually stigmatized. However, they present a social distribution (by age group, for example) and stylistic variation (level of proximity between speakers, for example). In other words, they are variants that generally adapt to the socio-interactive context, experiencing frequent adaptations in the speech style level. The nature of variants as stereotypes, markers, and indicators⁶ guides the correction analysis of the variable morphosyntactic structures presented in this article.

Considering the subjective evaluation of each structure, it is assumed that the motivations to it derive from a complexity of factors related to the multivariate reality of language. Bortoni-Ricardo (2005) has been sought to represent such complexity starting by the formulation of three continua of variation in Brazilian Portuguese (BP): the rural-urban, the orality-literacy, and the stylistic monitoring.

The rural-urban continuum – of diatopic nature – encompasses geographically isolated rural varieties to urban variety, such as expressions of the “rurbano”, which is an intermediate area, with both urban and rural features at the same time. In the orality-literacy continuum – that has features related to diastratic/social variation and also diamesic (from the most educated/cultivated/literate to the least educated/more popular/more oral) –, there is, on one hand, the elements whose production is mediated by the written language, following a fixed and artificial pattern. On the other hand, the elements are related to the social practices of orality.

With respect to stylistic monitoring, the continuum includes the level of attention demanded by speaking, which can be usually related to the level of formality of the communicative situation: from the most controlled/monitored, at one end, to the least controlled/monitored, at another end. In this continuum, it must be considered not only the level of attention but also the planning done by the speaker to their interaction once they tend to pay more attention to their speech in more formal and monitored contexts. According to the author, at the end of each of these continua there may be exclusive variants, variants which would be discontinued, representing the prototype of a diatopic variety of a specific modality, or a register.

In the analysis presented, the three continua are considered, once the school, hypothetically, tends to privilege what is taken as urban, monitored, and related to the written culture. However, one characteristic related to the continuum of stylistic monitoring is privileged, owing to the analysis that stems from a school composition in which the students have to use proper structures to a monitored and formal writing, the one supposedly based on models found in traditional manuals. What happens is that, in the school context, it is conceived as modeling an idealization of language that assumes, against all scientific evidence, the invariance as a pattern. This model results in the adoption of a concept of an inflexible and static standard norm.

Considering the polysemy and the conceptual plurality referred to the term norm, it must be clarified the definitions adopted regarding this issue. The term is related to two relevant general fields: normality (referred to usual linguistic facts, ordinary, regular in a speech community, to how language is actually used), and the normativeness (corresponded to the idea of prescription, of modeling or exemplary, of how a language supposedly has to be used).

In the normality level, it is needed to recognize that the language is composed of many forms, once heterogeneously constituted by a group of varieties used generally by the people of a particular speech community. Each one of these norms can vary naturally, due to linguistic factors (structural) or extralinguistic, such as the geographic origin of the speaker, socioeconomic status, level of education, the context of interaction in which they are (formal/informal), age, sex, and others.

Concerning the existent norms/varieties, on the one hand, there is the popular norm, which is supposedly used by speakers of semi-educated or uneducated families, people who do not have proper access to formal education and are not familiarized with the written culture. On the other hand, educated variety/norm, the one currently used by speakers of urban social classes with a high level of education, especially in monitored situations, is usually associated with people with more familiarity with the writing modality. Obviously, this norm is variable, especially if it is considered the continuum speaking-writing. In this regard, Vieira & Freire (2014) and Vieira & Lima (2019) observed the research results on various morphosyntactic phenomena in various textual genres of writing modality, journalistic and academic domains, demonstrating a substantial variation in the scope of the so-called educated variety.

In the normative level, on the other hand, there lies the need for language standardization, in consonance with the usual conception of a unitary and homogeneous entity. According to Faraco (2008), the standard norm concept is an abstract codification, an idealization that, although initially proposed to classic literature writing, comprehends a social ideology in which speakers elect structures like those that must be used. In other words, it is the result of a social attitude towards language, creating a selection in the way of speaking, among the various existing ones in the community and the establishment of certain structures as being the prestigious form, or the “correct” form of speaking. It is worth to notice that the standard norm (ideal level), at first, should be related to the educated variety/norm (of use), once that, according to what Vieira (2013) claims, the users of the latter would be the supposed creators of the former. It is known that, however, “the inevitable strengths that operate historically in favor of variation and change create a natural distancing between these norms.” (VIEIRA, 2013, p. 61, our translation)⁵.

In general terms, it can be stated that one of the standard norm models, widespread socially, is codified in the grammatical norm, the one that, according to Faraco (2008), is present in the good old normative instruments, such as grammar books and dictionaries developed by scholars and philologists of Portuguese Language. The author highlights that the grammatical codification registered in textbooks pay special attention to the standard usage facts, especially the literary one. Frequently, he also presents the traditional usages and specific innovations of language, advising the speaker that gives preference to more conservative structures, mainly in determined contexts, considered classics.

The inflexible segment and proposition of certain structures, recommended or not by the grammatical norm, however, results in what Faraco, in fine irony, named strict norm, characterizing certain normative models strictly constituted by

[...] dogmatic principles that do not find support in facts, even in the good old normative instruments. Still, they reinforce the culture of error, and it has been preventing a proper study of our standard variation/common/standard. This norm (real usage, therefore) and the comments of our good grammarians usually contradict such dogmatic principles. Even though they are repeated as if they were absolute truth and taken as a reason to humiliate, embarrass, and harm people. (FARACO, 2008, p. 92, our translation)⁶

As stated by the author, this normative ideology predominates, unfortunately, in school discourse, in common sense, and in the media, having prestige in Brazilian society and leading to the perpetuation of linguistic prejudice. Besides, it supports the culture of error and affects the speakers’ plural domain of language.

Hence, the education reality in many Brazilian schools, as Mattos e Silva (2004) attests, presents some kind of diglossia, meaning a polarization between the speech of students (popular norm/vernacular) and the standard norm that they must supposedly learn through literacy, an ideal language based on a grammatical norm, created by following the Lusitanian standards and not according to the Brazilian linguistic reality. It can be understood that the shorter the model of this patronization disseminated in schools is, the longer will be the distance between the norm(s) of usage and the norm idealized by whomever they are.

Given the distance between the standard norm idealized in the school context, inspired or not in the grammatical norm, and the norms/varieties of use, from the most cultivated/educated to the most popular, it is essential the importance that the teacher has a vast knowledge not only concerning about what it is presented in the school materials/instructions and grammar textbooks but also about the scientific descriptions of BP varieties. In this direction, Barbosa (2007) develops the three pieces of knowledge that should live together in Portuguese Language lessons in order to have the teacher working the variable phenomena in class productively, situated socially, and without ignoring the scientific progress: the spoken vernacular norm of usage knowledge, the descriptive-prescriptive knowledge and the descriptive-scientific knowledge.

The knowledge over the vernacular norm of usage is “apart from innate competence, it is shared by their community or region. It is the social knowledge conceived primarily in family environment” and, after, shared by their community (BARBOSA, 2007, p. 37, our translation)⁷. According to the author, such knowledge is going to interact with the writing tradition found in schools. The descriptive-prescriptive knowledge, the traditional grammar one, is the knowledge provided by the Greek-Roman model of linguistic approach, according to which “any linguistic variant that does not belong to a certain group of variants selected as the standard will be the incorrect variant for formal use” (BARBOSA, 2007, p. 38, our translation)⁸. About the descriptive-scientific knowledge – the academic knowledge, acquired through research, reading of books and scientific articles – is the one constituted by various linguistic theories, which enables teachers to understand language, explain problems and contradictions found in traditional grammar. It also leads teachers to understand that description/prescription based on this model is dedicated to literary writing, not to the entire reality of the language.

It is believed that the union of these three pieces of knowledge be essential for Portuguese teachers to understand the language functioning, based on a diversified theoretical-descriptive approach (structuralism, socio-cognitive, textual-discursive, and others), and as a result, developing the knowledge of students. In the scope of usage and evaluation of variable morphosyntactic structures, it is understood the possibility of, from the sociolinguistic knowledge perspective, raising awareness of students about the natural and systematically heterogeneous character of the language, enabling them to master it successfully, consonant with the situation (or context) in which the communicative act takes place.

A study on the evaluation of variable structures: methodological aspects and results

In order to study the subjective variable morphosyntactic phenomena, the investigation has covered various steps regarding methodology.

Firstly, students of the 8th grade of a public middle school, located in the city of Rio de Janeiro, wrote a composition whose theme is related to the protests that occurred in 2013 all over the country due to the price increase of public transportation fares, corruption, low salaries, and others.

Secondly, the authors observed these compositions, collecting the most frequent variable morphosyntactic phenomena in each of them. Based on this collection, a text was elaborated which presented such phenomena, so that, in a later moment, ten teachers of the public system of Rio de Janeiro and ten other carioca speakers were asked to correct them, according to what was considered as correct by them. It is relevant to observe that the group of non-teacher carioca speakers is composed of people who have a finished or unfinished graduation, and they came from various work fields, not necessarily related to the educational one. Referring to the procedure, used to aim at becoming possible to consider the evaluation issue proposed by Weinreich, Labov, and Herzog (1968), it is necessary to clarify that each person received one composition copy and corrected it separately, without having any access to the evaluation of the other participants. It is worth mentioning that they corrected the text intuitively, meaning that they made it based on what they judged to be correct, not consulting any manual or book.

In the last step, the correction of each one of the participants was analyzed, observing which structures were considered incorrect frequently, which ones were not corrected most of the times, and which ones were not noticed by any of them, including the teachers, hence, not being susceptible to correction. It must be pointed out that, despite a simple correction or not of a phenomenon does not make a variant an indicator, marker, or stereotype. It is possible to verify, under the recurrence of behavior from the evaluators related to the corrections made concerning each construction, a particular pattern of data analysis. Although it is not proposed to be a rigorous quantitative analysis, the observation of this standard, especially in qualitative terms, allowed the interpretation of the data to be substantiated.

Based on the procedures already presented, it is expected to be possible to evaluate, by the analysis of the composition corrections, which is conceived as the standard norm, according to the idealization of educated speakers. The hypothesis guiding the analysis is that the structures that are not identified as problematic and do not constitute a frequent object of correction – independently of being conceived as incorrect or not according to the grammatical norm – could not be stigmatized. In this case, it means that morphosyntactic structures are commonly used in speaking and even in the writing of educated people or, in other words, in educated varieties/ norms (of use).

Interpreting the results, the structures in question were, respectively, related to the categories of stereotypes, markers and indicators. Therefore, it was possible to evaluate the closeness or the distance between the standardization revealed in the correction (the standard norm in question) and that proposed in traditional manuals (grammatical norm). Moreover, the usage of vernacular varieties that are already accepted and the ones disregarded was observed, upon a possible case of diglossia, in some aspects, between the norm considered in the subjective level – standard norm / grammatical norm – and the varieties/norms of use. Finally, the analysis enabled establishing a relationship between the data collected, and Portuguese Language teaching, evaluating a possible alternative to the norm(s) approach in the educational context.

The composition submitted to the evaluation of teachers during the investigation is shown in the chart below:

Chart 1
– Composition submitted to evaluation by carioca readers

O Brasil acordou Há algum tempo vem ocorrendo manifestações por todo o Brasil, pois o povo brasileiro estão revoltado com a atual situação do país. No início, a intenção da população era conseguir a diminuição do preço da passagem dos transportes públicos, que tinham aumentado R$ 0,20. É importante dizer que as manifestações foi organizada nas redes sociais pelo povo, que o Brasil precisa para tornar-se um país melhor. Vale notar que houveram manifestações em pontos importantes de cada estado, com pessoas que estavam com as caras pintadas com as cores do Brasil, e com cartazes e faixas pedindo por melhorias, principalmente, na educação e saúde. Na realidade, a população se revoltou porque quer mais do que a diminuição do preço da passagem, se deseja menos corrupção e mais profissionais na área de saúde e educação recebendo um salário justo, ao invés de gastar milhões na construção de estádios. Porém, além dos manifestantes que reivindicam por um Brasil melhor, também haviam entre eles, muitos vândalos que infiltraram-se nos protestos com intuito de quebrar patrimônios públicos e particulares, colocar fogo em carros e até mesmo furtar produtos de lojas que foram arrombadas por eles mesmos. Se pode dizer que as passeatas que apresentaram algum tipo de vandalismo foram repreendida pela polícia, que reagiu usando spray de pimenta, bombas de efeito moral e balas de borracha. Felizmente, mesmo com os conflitos entre os vândalos e a polícia, se alcançou alguns dos propósitos que o povo almejou-os, como por exemplo, a diminuição das tarifas do transporte público, aprovação do texto final que tornará a corrupção um crime hediondo, 75% dos royalties do petróleo destinado para a educação, entre outras conquistas. Logo, se pode perceber que através de manifestações, os brasileiros, que lutando por seus direitos, o povo reconhece ele como batalhadores, conseguem atingir os seus objetivos e mostrar para os governantes que o papel deles é representar a população e as suas necessidades, e não utilizar o poder deles para benefício próprio utilizando o dinheiro público para fins particulares.

Source: Author’s elaboration.

In this text, the four phenomena disposed to be observed according to the subjective evaluation of the reader were the following: agreement, pronominal placement, use of the verb “haver” as impersonal, and use of adjectives and/relative clauses. Each of these variable morphosyntactic phenomena is considered by how they are presented according to the grammatical norm. This description served as a base for interpreting the correction of such structures by the research participants.

Concerning agreement, the following passages of the composition can be considered deviant data from traditional prescription – the one stating that the verb form must follow, concerning number, the same form of the NP:

“(...) o povo brasileiro est-ão revoltado (...)”

the people Brazilian are-PRS.3PL outraged-M.SG

‘Brazilian people are outraged’

“(...) as manifestações f-oi organizad-a (...)”

the protests was-PASS.PST.3SG organized-PTCP.F.SG

‘the protests were organized’

“(...) as passeatas que apresentaram algum tipo de vandalismo foram repreendid-a (...)”

the marches that presented any type of vandalism were reprehended-PTCP.F.SG

‘the marches that presented any type of vandalism were reprehended’

“(...) se alcanç-ou alguns dos propósitos (...)”

‘se’-PASS achieved-PST.3SG some of the intentions

‘some of the intentions were achieved’

“(...) preço da passagem dos transportes públicos, que tinh-am aumentado (...)”

price of fare of transportation public, that have-PST.3PL raised

‘price of public transportation fare, that has raised’

“(...) 75% dos royalties do petróleo destinad-o (...)”

75% of the royalties of the petroleum destined-PTCP.M.SG

‘75% of the royalties of the petroleum destined’

Therefore, according to the grammatical norm, in each of these clauses, the verbs (finite or participle) should agree with the head of the NP in question (in this case, “povo”; “preço”; “manifestações”; “passeatas”; “propósitos” and “75% dos royalties”, respectively) in number and/or person.

Through the analysis of the corrections and, consequently, the evaluations done by teachers and other speakers, it is verified that the already mentioned morphosyntactic phenomenon, in general terms, is highly susceptible to correction, in a higher or lower number of structures, suggesting a typical pattern of the stereotype category. However, it is important to mention that the structures in question were analyzed differently in terms of degrees of acceptance, signaling that determined constructions are possibly more stigmatized than others. Thus, the most disregarded structures and that effectively should be taken as stereotypes are the following: “(...) o povo brasileiro estão revoltado (...)”; and “(...) as manifestações foi organizada (...)”. Considering this, nineteen and eighteen participants, respectively, indicated inaccuracy.

These clauses are followed by “(...) preço da passagem dos transportes públicos, que tinham aumentado (...)”, and “(...) as passeatas que apresentaram algum tipo de vandalismo foram repreendida (...)”. Both of them were corrected by twelve participants. It could be noticed that the distance between the subject and the verb / nominal form of the verb, especially when interposed a PP with a plural or singular number expression, incompatible with the head of the NP, could affect the acceptability of the non-standard structures. Among the sentences analyzed concerning the agreement, the sentence “(...) 75% dos royalties do petróleo destinado (...)” received the lowest number of corrections, with only nine participants evaluating the structure as incorrect, that is, less than a half of the participants noticed that there was no standard agreement in this passage.

Such results call into question the status of stereotype with no plural mark when it is before not only partitive constructions (in this case, expressing percentage), but also before an intervening PP in the singular form – “do petróleo”. It seems that constructions of this nature – with the components in agreement distant or affected by the interpretation of the elements inserted in the noun phrase – are characterized, in general, as sociolinguistic markers, considering that they are susceptible to high monitoring so that they are used or conceived as incorrect forms from the perspective of the grammatical norm in some cases.

The corrections of the clause “(...) se alcançou alguns dos propósitos (...)” indicate another standard of behavior in the evaluation: only two participants, both teachers, noticed that there was no standard agreement between the supposed subject (“alguns dos propósitos”) and the verb (alcançou) of this construction. This structure was possibly characterized, for most of the evaluators, as an indicator. It is a construction of the called synthetic passive voice, that is, according to the grammatical norm, formed by the direct transitive verb in the third person (plural or singular) plus the particle se used as passive. Hence, in consonance with the grammatical norm, the correct form is “(...) alcançaram-se alguns dos propósitos (...)”. In this kind of structure, including the specificity of the called passive, the position of the subject, postponed to the verb, also affects the use and the interpretation of the data. Concerning this, contemporary scientific descriptions (MATEUS et al., 2003; RAPOSO et al., 2013, among others) have proposed labeling it as an unaccusative construction, whose subject is actually an intern argument; consequently, the interpretation of this component as the direct object of the clause justifies the non-standard agreement.

In the following table, it is possible to observe quantitative results in detail, as well as it is noticeable which structures were more corrected by teachers and which ones were more corrected by other speakers.

Based on Table 1, it is possible to notice, in general, that, concerning the sentences in question, a significant difference in evaluation between teachers and other participants did not occur. In any case, it is verified that the structures in which the no marking of the standard agreement is more evident (the first two) are identified by practically all individuals (95%/90%), teachers or not. The structures in which unfavorable factors of marking – distance and the interference of intervenient elements (LEMLE; NARO, 1977; VIEIRA; BAZENGA, 2013, 2015) – are present, as in the case of the sentences 3, 4, and 5, are identified in approximately half of the cases or a bit more, oscillating between the most identified of the non-agreement by teachers (sentence 3) or by other participants (sentences 4 and 5). Finally, it is, in fact, the last structure which is not identified as susceptible of correction by most of the participants, possibly for not being stigmatized in the educated variety/norm; in any case, only two teachers (10%) – individuals with technical knowledge on such structure – were able to identify and correct it.

Table 1
– Number of corrections of structures with the non-standard agreement in the school composition

Structures	Teachers	Other speakers	Total
1. “(...) o povo brasileiro estão revoltado (...)”	9	10	19/20 (95%)
2. “(...) as manifestações foi organizada (...)”	8	10	18/20 (90%)
3. “(...) as passeatas que apresentaram algum tipo de vandalismo foram repreendida (...)”	8	4	12/20 (60%)
4. “(...) preço da passagem dos transportes públicos, que tinham aumentado (...)”	5	7	12/20 (60%)
5. “(...) 75% dos royalties do petróleo destinado (...)”	3	6	9/20 (45%)
6. “(...) se alcançou alguns dos propósitos (...)”	2	0	2/20 (10%)

Source: Author’s elaboration.

With regards to the position of the unstressed oblique pronoun concerning only one verb form – in proclisis (before the verb), mesoclisis (in the middle of the verb), or enclisis (after the verb) –, the grammatical norm recommends, as a general rule, the use of the clitic in post-verbal position. The use of proclisis is justified only in contexts with negative words or expressions; with subordinate conjunctions (such as que); adverbs; relative, demonstratives and indefinite pronouns; in interrogative sentences; in exclamatory or optative sentences; with verbs in gerund preceded by the preposition em (ROCHA LIMA, 2011). Mesoclisis must only be used in structures without the presence of a clitic attractor and when the verb is in the future of the present or in the future of the past.

Although variationist’ studies (VIEIRA, 2002; NUNES, 2009; CORRÊA, 2012; VIEIRA; FREIRE, 2014; VIEIRA, 2016) have widely demonstrated the preference for the pre-verbal placement in Brazilian speaking, independently of the presence of a traditional attractor, whether in popular varieties or educated varieties, the traditional rules of placement are usually present in didactic materials, in Portuguese Language classes, and in several orientations for proofreaders, which favors the existence of another standard of placement in Brazilian writing modality. Vieira & Freire (2014, p.101, our translation) summarize the behavior of Brazilian variety as the following:

In short, in constructions with only one form with clitics effectively acquired in BP (<me, te, se> reflexive), the data revealed that the proclisis is generalized. With the clitics that related more specifically to the educated domain, two of them guarantee the enclisis survival, though not much productive: <o,a(s)> and <se> in constructions of the indefinite agent. In writing, there is another system of pronoun placement with strong morphosyntactic conditioning, sensitive to the verb context in the initial position, to such actions of the traditional attractor particles, and the clitic type. In any case, the difference in literate Brazilian writing is relevant in relation to what is stated in the grammatical norm, regarding two characteristics that can be highlighted: (i) the noticeable presence of the proclisis in contexts without the traditional attractors, in particular before the subject; and (ii) the presence, less noticeable but not inconsiderable, of the enclisis in contexts with traditional attractors⁹.

In the composition analyzed in the present research, there are cases of proclisis and enclisis¹⁰ in some of the morphosyntactic contexts already referred. The following constructions were registered in contexts which grammatical norm recommends the enclisis: (i) verbal form (phrase) in the very beginning of the period (“Se pode dizer [...]”); (ii) verbal form in the beginning of the clause is delimited by comma in relation to the previous one (“[...], se deseja menos corrupção e mais profissionais na área de saúde e educação recebendo um salário justo, ao invés de gastar milhões na construção de estádios.”); (iii) verbal form (phrase) after an adverbial form separated by comma (“Logo, se pode perceber [...]”); (iv) verbal form after no attractor adjuncts (“Felizmente, mesmo com os conflitos entre os vândalos e a polícia, se alcançou alguns dos propósitos que o povo almejou-os [...]”); (v) verbal form right after the subject (“Na realidade, a população se revoltou [...]”).

In the first-mentioned clauses, the use of the enclisis is justified since the use of proclisis at the beginning of the sentence is not recommended; consequently, the supposedly correct form would be “Pode-se dizer que as passeatas (...)”, and “(...) deseja-se menos corrupção (...)”. In the third and fourth examples, the enclisis would be expected, since there are only adverbial adjuncts before the verb. In the third sentence, according to Rocha Lima (2011, p. 545), the proclisis is not used after short adverbs separated by comma. The supposedly correct form would be, thus, “Logo, pode-se perceber (...)”. In the fourth sentence, the use of the enclisis is recommended, since no attractor is registered before the verb but only an adverb of manner (“felizmente”) and a concessive adjunct (“mesmo com os conflitos entre os vândalos e a polícia”). Then, the expected should be “(...) alcançaram-se alguns dos propósitos (...)”. Finally, in the last sentence, the traditional norm determines that the pronoun must be placed after the verb (enclisis), since no attractor is justifying the use of proclisis. Concerning the grammatical norm, the correct form should be, then, “(...) a população revoltou-se (...)”.

The composition presents, also, the occurrence of one case of required proclisis, according to the grammatical norm: verbal form preceded by one clitic attractor, the relative pronoun que, as it could be verified in “(...) vândalos que infiltraram-se nos protestos (...)”. In this case, the use of the proclisis is recommended, since the relative pronoun que, acting as one of the traditional attractors, determines that the unstressed pronoun should precede the verb. Therefore, the correct form should be “muitos vândalos que se infiltraram”.

The analysis of the corrections shows that there is a distinctive identification by the evaluators concerning the application of the rules of pronominal placement in monitored contexts. In the following table, the analyzed structures in consonance with the evaluation of the participants can be observed.

Table 2
– Number of corrections of structures concerning pronominal placement in a school composition

Structures	Teachers	Other speakers	Total
1. “Se pode dizer (...)”	5	9	14/20 (70%)
2. “Logo, se pode perceber (...)”	4	7	11/20 (55%)
3. “(...) se deseja menos corrupção (...)”	3	5	8/20 (40%)
4. “Felizmente, mesmo com (...), se alcançou alguns dos propósitos (...)”	1	4	5/20 (25%)
5. “Na realidade, a população se revoltou (...)”	0	0	0/20 (0%)
6. “(...) vândalos que infiltraram-se (...)”	3	2	5/20 (25%)

Source: Author’s elaboration.

Considering the pronominal placement, the results displayed in Table 3 reveal, firstly, that it is not confirmed that teachers act in greater consonance with the traditional rules of pronominal placement. In reality, some structures – “(...) e alcançou alguns dos propósitos (...)”; “(...) se deseja menos corrupção (...)”; “Se pode dizer (...)”, and “(...) se pode perceber (...)” – were even more corrected by other speakers. Apparently, the language users evaluate the morphosyntactic constructions related to the order of pronominal clitics by following intuitively their beliefs of which ones would be prestigious, independently of such understanding conforms with the grammatical norm, whose detailing is, as an assumption, frequently ignored by Brazilians.

Table 3
– Number of corrected structures concerning the use of “haver” as an impersonal verb

Structures	Teachers	Other speakers	Total
“(...) houveram manifestações (…)”	4	2	6/20 (30%)
“(...) haviam entre eles, muitos vândalos (...)”	5	4	9/20 (45%)

Source: Author’s elaboration.

In any case, the traditional proposal is noticed in certain specific contexts, especially in absolute first position in a sentence for being the structure with the highest rate of correction. The result suggests that, in terms of data interpretation, regarding writing modality, the proclisis in this context seems stigmatized, representing a sort of stereotype, for which reason fourteen participants (70%) noticed that this structure should be wrong and it was corrected. Although there has been no evidence for further generalizations in the scope of the present study, it is presumed that, based on the results of other investigations (VIEIRA; FREIRE, 2014), the fact that it is a modal verbal periphrase followed by the clitic se could have favored the enclisis for this context.

In relation to the status of the preferential order of the clitics in the contexts 2, 3, and 4, all with verbs after the comma, preceded by adjuncts or another clause, the evaluation of the phenomenon received from the participants seems compatible with a sociolinguistic marker. The constructions “(...) se pode perceber (...)”, “(...) se deseja menos corrupção (...)”, and “(...) se alcançou alguns dos propósitos (...)” were corrected by eleven (55%), eight (40%), and five (25%) participants, respectively. Concerning that a pre-verbal variant in these contexts is, in spoken modality, the preferential option, in the categorical and semi-categorical level, it is also absolutely acceptable that they are not conceived as inappropriate in writing modality, once they are part of Brazilian educated variety. Therefore, at most, the use of enclisis in monitored situations indicates variation according to the register (formal/informal) and the modality (writing/oral), which is once more correlated to a marker behavior effectively.

Regarding the improper usage of the proclisis in consonance with the grammatical norm, it is important to notice that there was a case in which none of the informants took the pronominal order as incorrect. This fact occurred in the clause “(...) a população se revoltou (...)”, whose pronoun must be, according to the grammatical tradition, after the verb (enclisis), as it is preceded by only a subject NP, and not by a clitic attractor element. In this case, everything indicates that the absence of corrections would imply that the pre-verbal variant works as an indicator to the participants, and consequently, it would not involve even one case of stylistic variation in the use of this structure by educated individuals.

Lastly, concerning the use of enclisis being considered traditionally incorrect, in the clause “(...) muitos vândalos que infiltraram-se (...)”, instead of “(...) muitos vândalos que se infiltraram (...)”, it was observed that, based on the participants’ evaluations, the corrections were made by only five people (25%) among the twenty participants. It seems that there is a case of stylistic variation for many users of the language. Actually, certain speakers evaluate this use as stigmatized and, for this reason, follow what is prescribed in the grammatical norm. While others, due to the phenomenon of hypercorrection¹¹ or the absence of knowledge over what is expected in traditional grammar, prefer to use the enclisis even in contexts in which its use is not recommended.

In order to conclude the analysis dedicated to the pronominal placement in monitored contexts associated with the sociolinguistic results concerning the phenomenon in BP, it can be proposed that, in general terms, this is diamesic and register variations. Although in spoken modality the use of the pre-verbal variant works as an indicator, given its generality in any variety of BP, the clitics order in controlled educated writing modality can even have the status of a marker. Though the fact that BP has essentially a proclitic character, independently of a supposed attractor existing or not in the clause, certain individuals, with a high degree of literacy and in highly monitored situations, can adopt, when writing in specific textual genres such as the academic and journalistic ones, another standard for the use of pronominal placement and, then, even be aware of some morphosyntactic contexts in consonance with the approach of the grammatical norm. Nevertheless, it is known that employing the traditional rules is always limited and partial, according to the status of a marker. Studies on usage have revealed that, besides the enclisis being the most evident at the beginning of the sentence, the observance of the proclisis attractor elements, even being limited to a more controlled writing modality, is rare and partially applied.

A third phenomenon observed in the composition was the use of the verb haver with the meaning of “exist”, in the context of an impersonal verb, that could be evaluated in the following excerpts:

“(...) houveram manifestações (...)”

existed-PST.3PL protests

‘there were protests’

“(...) haviam entre eles, muitos vândalos (...)”

existed-PST.3PL among them, many riots

‘there were many riots among them’

According to the grammatical norm, the structure does not present a referential subject, since the NP that follows the verb is the direct object, in such a way that the verb must not be in the plural form, but just in the singular. The correct forms would be the structures:

“(...) houve manifestações (...)”

existed-PST.3SG protests

‘there were protests’

“(...) havia entre eles, muitos vândalos (...)”

existed-PST.3SG among them, many riots

‘there were many riots among them’

The following table contains the data obtained from the correction done concerning the use of “haver” as an impersonal verb by the twenty participants, teachers and other speakers.

Concerning the status this phenomenon can have, the number of corrections (six and nine, respectively) suggests moderate to a low perception of the deviant standard to the grammatical norm. In any case, it is suggested that it can be understood by a particular group of participants, teachers or not, as a marker, meaning that the use of haver with no plural mark would be recognized as appropriated to formal writing (VIEIRA; FREIRE, 2014).

In addition, it should be emphasized that the phenomenon in discussion is closely related to what it is conceived usually as linguistic hypercorrection; in order to meet the insistent requisites of the verbal agreement rule, they mistakenly apply it to the impersonal verb context.

Eventually, the use of the called relative clauses was analyzed. Based on the traditional approach, they are adjective subordinate clauses introduced by a relative pronoun and function as an adnominal adjunct of a noun or a preceded pronoun, when a restrictive relative, or as an apposition, when an appositive clause. In the generativist framework adopted by Mateus et al. (2003), three strategies of relativization are proposed: the standard, in which the relative clause is composed by a relative pronoun and a preposition (when required by the verb in the embedded clause), as in:

O livro de que ele gosta

The book of-PREP that-REL he likes

‘The book that he likes’

And two non-standard processes that give rise to the resumptive strategy and the chopping strategy. The former is characterized by the absence of the preposition (when necessary) that should follow the relative pronoun, as in:

O livro que ele gosta

The book that-REL he likes

‘The book that he likes’

While the latter is characterized by repeating (copying) the referent of the previous clause in the embedded structure, as in:

O livro que ele gosta dele

The book that-REL he likes of-PREP it-OBL

‘The book that he likes’

The standard relativization strategy, regarded by the traditional approach, is disseminated as an object of learning in Basic Education institutions. As a result, its use is proposed especially for interactions that demand a high degree of monitoring, mostly registered in studies on Brazilian variety (VIEIRA, 2017, 2019b) in formal writing modality. In turn, the chopping strategy acts as a non-marked structure, being the construction generally used by an individual, educated or not, in several spoken situations. In turn, the strategy with resumptive pronouns is more susceptible to be stigmatized due to the duplication of the referent, and it is apparently used in specific interactive situations. As a consequence, it is not largely registered.

In the analyzed composition, two examples of resumptive pronoun strategy with different forms were found; these pronouns do the resumption of the referent already indicated by the relative pronoun:

“(...) que o povo almejou -os (...)”

that-REL the people desired them-OBL

‘that people desired’

“(...) que, lutando por seus direitos, o povo reconhece eles como batalhadores (...)”

who-REL, fighting for their rights, the people recognize they-PERS.PRO as strugglers

‘who, fighting for their rights, people recognize as strugglers’

As it was observed, they are constructions in which the precedent of the relative pronoun was referred back (copied) by the clitic “os” and by the personal pronoun “eles”, respectively. In the first case, the relative pronoun, referred back by the unstressed pronoun “os”, functions as the direct object of the verb “almejar”, in the subordinate clause. The standard correspondent relative clause should be, then:

“(...) propósitos que o povo almejou (...)”

intentions that-REL the people desired

‘intentions that people desired’

In the second excerpt, the relative pronoun “que”, referred back by the personal pronoun “eles”, functions as the direct object of the verb “reconhecer” in the subordinate clause. The standard correspondent relative clause should be, thus:

“(...) os brasileiros, que, lutando por seus direitos, o povo reconhece como batalhadores (...)”

the Brazilians, who-REL, fighting for their rights, the people recognize as strugglers

‘Brazilians who, fighting for their rights, people recognize as strugglers’

Another example of relative clause is present in the excerpt:

“(...) X que o Brasil precisa para tornar-se um país melhor (...)”

X that-REL the Brazil needs to become a country better

‘that Brazil needs to become a better country’

A chopping relative clause is presented in this excerpt. In this passage, according to traditional grammar, the antecedent of the subordinate clause is functioning as the verb complement of the clause and, as a result, the relative “que” should be preceded of the preposition “de”, since the verb “precisar” demands a prepositional complement. It could be stated that this clause is described as a chopping clause because the preposition (in this example, “de”) that should follow the relative pronoun is not employed. The standard relative equivalent should be the one found in:

“(...) o povo de que o Brasil precisa para tornar-se um país melhor (...)”

the people of that-REL the Brazil needs to become a country better

‘the people that Brazil needs to become a better country’

The following table shows the number of corrections done by teachers and by other speakers regarding the already mentioned strategies of relativization.

Table 4
– Number of corrections of structures regarding the use of resumptive strategy and chopping strategy

Structures	Teachers	Other speakers	Total
“(...) que o povo almejou-os (...)”	9	6	15/20 (75%)
“(...) que, lutando por seus direitos, o povo reconhece eles como batalhadores (...)”	3	3	6/20 (30%)
“(...) X que o Brasil precisa para tornar-se um país melhor (...)”	0	0	0/20 (0%)

Source: Author’s elaboration.

The corrections done by the participants (fifteen, in total) present a high level of identification of the supposed incorrectness that occurs in the clause “(...) que o povo almejou-os (...), suggesting the phenomenon should be conceived as a stereotype. All participants that corrected it wrote it in the form considered the correct one, eliminating the unstressed pronoun that seemed to sound excessive and inadequate.

The clause “(...) que, lutando por seus direitos, o povo reconhece eles como batalhadores (...)” was corrected by only six informers (30%). Although this number of corrections was inferior to the expected, the low identification of the structure can be related to several factors, as the distance between the resumptive pronoun and the relative one, or even the type of structure in focus that has, besides the pronoun “eles”, another complement (“como batalhadores”), making the use of accusative “ele” sound more natural (DUARTE, 1986; SANTANA, 2016). Despite the number of corrections was relatively low, the declarations of the participants – that although they did not know how to correct such structure and, for that reason, they opted for keeping it in its original form, they thought the construction sounded strange, and they would not write it that way – they indicate that this construction can also be characterized as a marker or even a stereotype.

In relation to “(...) X que o Brasil precisa para tornar-se um país melhor (...)”, that presents a chopping relative clause, it was verified that none of the participants corrected the clause by using the preposition “de”. This result leads to the conclusion that such structure, produced in a partially categorical form in vernacular Brazilian speaking (VIEIRA, 2017), should be identified as an indicator, reaffirming that its use is already part of the educated variety/norm; its use would already sound so natural that it is not seen as incorrect, including formal texts.

Systematization and discussion of results

Throughout the analysis, it was possible to interpret how some of the registered variable morphosyntactic phenomena in a school composition was evaluated by various educated speakers, among them teachers. Besides, they suggested which status each phenomenon would have, using the classification proposed by Labov (1972) – stereotype, marker, indicator.

The following graphic presents a general overview of how the corpus, considering all variable morphosyntactic structures, was corrected by teachers and other speakers.

Image 1
– Display of the corrected structures/not corrected by the participants

Based on the chart, it is possible to notice that more than half of the linguistic structures (56,47%) following the norm were not corrected. Furthermore, it was observed, comparing the behavior of teachers and other speakers, that the identification of the constructions out of the standards of the grammatical norm occurred similarly: teachers: 21.18%; other speakers: 22.35%.

Through this perspective, regarding the evaluation problem, developed by Weinreich, Labov and Herzog (1968), it is possible to notice that participants evaluate the linguistic variants through the interpretation of the data in the present analysis, electing as correct or adequate probably the ones that would be in accordance with the norm already disseminated in current educated spoken norm (educated variety/norm). Then, the standardization idealized by them (standard norm) does not necessarily conform to the model proposed in the codified norm in grammars and Portuguese dictionaries (grammatical norm). In other words, the standard norm – regarded in the evaluation of the participants by following their beliefs concerning which structures should be more prestigious or not – would not be necessarily in accordance with the grammatical norm.

It was then possible to verify that specific structures are more stigmatized than others –, such as the absence of standard agreement between the subject and the verb (especially when next to each other and in the direct form), whose evaluation indicated the most identified structures as incorrect, receiving, consequently, the status of a stereotype. On the other hand, it was observed that although some constructions are not expected as correct according to the grammatical norm, they are not effectively stigmatized, considering that they constitute highly productive variants in the speech of all varieties, including the most educated ones. As a result, they would be considered as neutral forms. Thus, they are neither corrected by teachers nor restrained in school and academic contexts. Those are called, for example, passive voice agreement or chopping relative clauses. Then, they could have the status of an indicator. Between these two possibilities, several constructions (related to the agreement, pronoun placement, the use of impersonal “haver” and the resumptive strategy of relativization) were partially identified by the participants. This probably happens owing to the cases that already belong to the educated spoken language that would be noticed and/or avoided only in highly monitored writing situations, enabling to attribute them the status of a sociolinguistic marker.

It is worth to highlight that, due to the conflicts between the corrections of the teachers and the grammatical norm prescription for each studied usage, it was possible to observe that what teachers often use as their source of knowledge to correct linguistic structures is, in fact, their evaluation, their idealization of linguistic standard (the idealization of standard norm), their vernacular knowledge – which is similar to the various educated speakers – and not only the descriptive-prescriptive knowledge, which includes the grammatical norm knowledge, according to Barbosa (2007)´s proposal.

Consequently, it is possible to state that, although the educated norm in use is composed by the linguistic usage of a social group situated at one end of the literacy continuum, there is no total correspondence between it and the grammatical norm, not even between it and the standard norm constituted abstractly. Indeed, there are some similarities between these instances, specifically in situations where the structures utilized correctly by participants coincide with the proposals of the grammatical norm – consequently, becoming a parameter to the correction of the structures posed to an evaluation in the school composition.

Seeking to establish a relation between the theoretical background that based the research and the analysis of the compositions, it is possible to present some final considerations concerning Portuguese Language teaching. Regarding this, first, it is crucial to understand how the linguistic variation is conceived in some educational institutions and even by some professionals in the field, in order to be able to envision solutions for the so-called linguistic variation pedagogy, according to Faraco (2008, 2015).

Firstly, it is important to mention that schools and Portuguese teachers, although they state theoretically the principle of systematic heterogeneity in terms of language, admit a practice that promotes the myth that the Portuguese language is homogeneous and invariable. As a result of such practice, they design activities where students must learn only the grammatical norm and use the structures prescribed as appropriate in all social-communicative instances – even though it does not conform to the texts read and produced in written and spoken modalities in the classroom.

In this perspective, a pedagogical practice is perpetuated, which means that everything not following what is prescribed by the grammatical norm – that is, in turn, characterized by presenting an idealized model of “correct” language – must be disregarded and taken as incorrect. Due to this equivocate point of view, teachers usually turn out to be spokespeople of what is called here strict norm, when what is present in the grammatical norm – which is constituted for specific purposes – has been radicalized, as if it could be applied to the entire language system. Indirectly, it contributes to the perpetuation of linguistic prejudice, supporting the culture of error, and the access of students to the vast domain of the language is made difficult once it leads Brazilians to adopt negative beliefs and attitudes about their own linguistic variety.

Considering the official orientations for the Portuguese Language teaching and the advances in Variationist Sociolinguistics applied to education, proposals must be developed, and they must effectively include BP as a heterogeneous reality, which presents different varieties/norms of use that apply to various modalities and registers. Therefore, taking into account that the general purpose of Portuguese Language teaching is developing reading competence and text production, it is unavoidable to deal with the complex linguistic diversity of the country and allow students to realize that language varies not only in the spatial level, from rural to urban, but also in the orality-literacy level, as well as in the stylistic monitoring level. By the language domain, according to these continua, the student can identify which linguistic variants to use in each textual genre and each social-communicative situation, in order to have the autonomy to choose the acceptable form within their repertoire, considering the situation in which they are involved.

Nevertheless, this proposal is not incompatible with providing access to students, in the intersecting of these continua, to the expressions considered cultivated, that turn out to determine a specific privileged model of behavior admitted as the standard norm, socially constructed, usually, from the urban environment, in formal situations, in writing modality. At this point, the standard norm conception cannot be automatically associated with the grammatical norm, which was revealed in this present investigation.

Therefore, it is proposed that the success of pedagogical orientations regarding the standard norm depends on, to a large extent, a comprehensive teacher training regarding the knowledge (spoken vernacular norm knowledge, the descriptive-prescriptive knowledge, and the descriptive-scientific knowledge) proposed by Barbosa (2007). Undoubtedly, from the joint actions of these pieces of knowledge, educators can have as a source of knowledge, besides their own experience as language speakers, the contents proposed by traditional grammars, dictionaries, and teaching materials, besides the vast production displayed in scientific materials.

Based on this extension of scientific production, Vieira (2019a, 2019b) developed a proposal for the achievement of the good compliance of school normative orientations that, most of all, follows what was suggested by Faraco (2015): that the standard norm to be taught in schools reflects the educated varieties. Then, the author formulated basic orientations to the type of the variants in evaluative terms, among which stand out, considering the analyzed data in this article, the following three:

1. do not offer any normative orientation, in case of a variant that works as an indicator not only in speaking but also in writing in various genres (for instance, the proclisis after the subject);

2. offer express orientation over the social value of the variants in terms of prestige or disregard, in case the variant represents a stereotype, situation strongly associated to linguistic/social prejudice (such as certain structures without standard agreement);

3. offer specific orientations for production and reading, in case of sociolinguistic markers (such as the standard relativization strategy; the impersonal haver), bearing in mind that the use and/or the identification of these structures will guarantee the access of students to various textual genres, with varied degrees of stylistic monitoring.

Besides these situations, Vieira (2019a) informs about determined structures that are not naturally used in contemporary speaking and writing. However, they are defined as archaic variants that usually circulate in determined discursive traditions, such as the literary or religious ones. In this case, according to the author, “the orientations should be restricted to the level of data reception (identification by reading/listening – from the religious or literary domain, for example), and not to production, under penalty of promotion of archaisms.” (VIEIRA, 2019a, p. 257, our translation)¹².

As it can be observed, in a pedagogy of linguistic variation scientifically grounded, it is crucial that the work with the standard norm – that should always be dynamic and renewable, as language is – be articulated to sociolinguistic results. Although the total systematization of the results in relation to all grammatical facts is not available, studies on variable morphosyntactic phenomena – not only related to the employment and the productivity of the variants, but also its subjective evaluation, as it was proposed – are promising in the sense of providing the needed information for teaching practice.

Final considerations

Throughout this article, which stemmed from the axiom that language is a heterogeneous and variable system, it was possible to observe that speakers can evaluate several variants as prestigious or stigmatized. In other words, it can be noticed that, in parallel with the linguistic variation framework, it also figures the evaluation one, that it is constructed not only by typical linguistic characteristics but also in consonance with the beliefs of the speakers.

In the face of that evaluation process of the several norms/varieties that compose BP, it was possible to perceive that, by means of the analysis of structures supposedly considered incorrect, even in the scope of normative idealization, there is no uniformity, since there is a lack of absolute correspondence between the standard norm (what the educated speakers idealize, in the abstract level, as prestigious/correct) and the grammatical norm (what is proposed in traditional normative instruments – grammars and dictionaries – as correct).

It results from the natural coexistence of these scopes of standardization that teachers should take into account the flexibility of the standard norm, that, consisted of a set of cultivated expressions, should conform with the several socio-communicative situations, materialized in a plurality of spoken and written textual genres. By the advances in science, especially in the area of Theory of Language variation and change, it does not sound even reasonable that the pedagogical guidelines adopt as a principle that only one normative direction must be applied unequivocally in every moment the speaker interacts, as if such codification of a norm could represent language in its totality.

Admitting the exuberance of linguistic variation – in the regional, social and stylistic, in speaking and writing – and make it a pedagogical object, does not mean undermining the relevance of the social idealization in the standard norm. Nonetheless, in conforming to what the exercise of evaluation has demonstrated in the present article, such standardization cannot be naively equipped at the pedagogical level, with an objective and limited list of constructions, supposedly correct or incorrect. The construction of a standard norm – that, not being a variety, has the purpose of homogenization and by principle is established as opposed to variation – is as complex as each norm/variety in use. Due to such complexity, it is unavoidable to conclude that it is by the usage of the variants of the educated varieties that should emerge naturally the idealization of standard norm, which accommodates the same space for selection and flexibility in each context seeking to its application.

REFERÊNCIAS

• BARBOSA, A. G. Saberes gramaticais na escola. In: VIEIRA, S. R.; BRANDÃO, S. F. (org.). Ensino de gramática: descrição e uso. São Paulo: Contexto, 2007. p. 31-54.
• BORTONI-RICARDO, S. M. A língua portuguesa no Brasil; Um modelo para a análise sociolingüística do português brasileiro. In: BORTONI-RICARDO, S. M. Nós cheguemu na escola, e agora?: Sociolingüística e Educação. São Paulo: Parábola Editorial, 2005. p. 31-52.
• BORTONI-RICARDO, S. M. Educação em Língua Materna: a Sociolingüística em sala de aula. São Paulo: Parábola, 2004.
• CORRÊA, C. M. M. de L. Cliticização pronominal na região metropolitana do Rio de Janeiro: a interface Sintaxe-Fonologia. 2012. 183 f. Dissertação (Mestrado em Letras Vernáculas) – Faculdade de Letras, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2012.
• CYRANKA, L. F. de M. Avaliação das variantes: atitudes e crenças em sala de aula. In: MARTINS, M. A.; VIEIRA, S. R.; TAVARES, M. A. (org.). Ensino de português e Sociolinguística. São Paulo: Contexto, 2014. p. 133-156.
• CYRANKA, L. F. de M. Atitudes linguísticas de alunos de escolas públicas. In: CYRANKA, L. F. de M. Dos dialetos populares à variedade culta: a Sociolinguística na escola. Curitiba: Appris, 2011. p. 43-51.
• DUARTE, M. E. L. Variação e sintaxe: clítico acusativo, pronome lexical e categoria vazia no Português do Brasil. 1986. Dissertação (Mestrado em Linguística Aplicada e Estudos da Linguagem) – Pontifícia Universidade Católica, São Paulo, 1986.
• FARACO, C. A. Norma culta brasileira: construção e ensino. In: ZILLES; A. M. S.; FARACO, C. A. (org.). Pedagogia da variação linguística: língua, diversidade e ensino. São Paulo: Parábola, 2015. p. 19-30.
• FARACO, C. A. Norma culta brasileira: desatando alguns nós. São Paulo. Parábola Editorial, 2008.
• GORSKI, E. M.; COELHO, I. L. Variação linguística e ensino de gramática. Working Papers em Linguística, Florianópolis, v. 10, n. 1, p. 73-91, 2009.
• LABOV, W. Principles of linguistic change. Oxford: Cambridge: Blackwell, 1994.
• LABOV, W. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press, 1972.
• LEMLE, M.; NARO, A. J. Competências básicas do português. 1977. Relatório final de pesquisa apresentado às instituições patrocinadoras Fundação Movimento Brasileiro de Alfabetização (Mobral) e Fundação Ford.
• MATEUS, M. H. M. et al. (org.) Gramática da língua portuguesa. 5. ed. rev. e aum. Lisboa: Editorial Caminho, 2003.
• MATTOS E SILVA, R. V. O português são dois. Variação, mudança, norma e a questão do ensino do português do Brasil. In: MATTOS E SILVA, R. V. O português são dois: novas fronteiras, velhos problemas. São Paulo: Parábola Editorial, 2004. p. 128-151.
• NUNES, C. da S. Um estudo sobre a ordem dos clíticos em complexos verbais no PB e no PE. 2009. 256 f. Dissertação (Mestrado em Letras Vernáculas) – Faculdade de Letras, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2009.
• RAPOSO, E. et al. (org.). Gramática do Português. Lisboa: Fundação Calouste Gulbenkian, 2013. v. I e II.
• ROCHA LIMA, C. H. da. Gramática normativa da língua portuguesa. 49. ed. Rio de Janeiro: José Olympio, 2011.
• SANTANA, J. M. C. P. de. Diagnose e ensino de pronomes: um estudo sobre a retomada anafórica do objeto direto de terceira pessoa no Português brasileiro. 2016. 205 f. Dissertação (Mestrado em Letras Vernáculas) – Faculdade de Letras, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2016.
• VIEIRA, M. de F. A ordem dos clíticos pronominais nas variedades urbanas europeia, brasileira e são-tomense: uma análise sociolinguística do Português no início do século XXI. 2016. 238 f. Tese (Doutorado em Letras Vernáculas) – Faculdade de Letras, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2016.
• VIEIRA, S. R. Para uma norma-padrão flexível no contexto escolar. In: MACHADO-VIEIRA, M. dos S.; WIEDEMER, M. L. (org.). Dimensões e pesquisas sociolinguísticas. São Paulo: Blucher, 2019a. p. 243-264.
• VIEIRA, S. R. Ensinando gramática em três eixos: conectivos e conexão de orações. In: SEMINÁRIO DO GRUPO DE PESQUISA CONECTIVOS E CONEXÃO DE ORAÇÕES, 2., 2019, Niterói. Anais [...], Niterói: Letras da UFF, 2019b. p. 57-79. v. 1, n. 2.
• VIEIRA, S. R. Ensino de Português e o contínuo fala-escrita: o caso das estratégias de relativização. Percursos, Florianópolis, v. 18, p. 8-35, 2017.
• VIEIRA, S. R. Sociolinguística e ensino de Português: para uma pedagogia da variação linguística. In: MARTINS, M. A.; TAVARES, M. A. (org.). Contribuições da Sociolinguística e da Linguística Histórica para o ensino de língua portuguesa. Natal: EDUFRN, 2013. p. 53-90. (Coleção Ciências da Linguagem Aplicadas ao Ensino, V).
• VIEIRA, S. R. Colocação pronominal nas variedades europeia, brasileira e moçambicana: para a definição da natureza do clítico em Português. 2002. 441f. Tese (Doutorado em Letras Vernáculas) – Faculdade de Letras, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2002.
• VIEIRA, S. R.; BAZENGA, A. A concordância da terceira pessoa plural: padrões em variedades do Português. In: VIEIRA, S. R. (org.). A concordância em variedades do Português: a interface Fonética-Morfossintaxe. Rio de Janeiro: Vermelho Marinho, 2015. p. 29-75.
• VIEIRA, S. R.; BAZENGA, A. Patterns of third person verbal agreement. Journal of Portuguese Linguistics, Lisboa, v. 12, n. 2, p. 7-50, 2013. Disponível em: https://doi.org/10.5334/jpl.67 Acesso em: 10 fev. 2022.
  » https://doi.org/10.5334/jpl.67
• VIEIRA, S. R.; FREIRE, G. C. Variação morfossintática e ensino de Português. In: MARTINS, M. A.; VIEIRA, S. R.; TAVARES, M. A. (org.). Ensino de Português e Sociolinguística. São Paulo: Contexto, 2014. p. 81-114.
• VIEIRA, S. R.; LIMA, M. D. A. de O. (org.). Variação, gêneros textuais e ensino de Português: da norma culta à norma-padrão. Rio de Janeiro: Letras UFRJ, 2019.
• WEINREICH, W.; LABOV, W.; HERZOG, M. Empirical Foundations for Theory of Language Change. In: LEHMANN, P.; MALKIEL, Y. (ed.). Directions for Historical Linguistics. Austin: University of Texas Press, 1968. p. 95-188.

• 1
  [...] o estudo das atitudes e das crenças linguísticas, entre outros focos de pesquisa, possibilitaria melhor compreender e auscultar a dificuldade de graduandos e de professores. Essa dificuldade se dá, por um lado, em relação ao problema, frequentemente aludido, do reconhecimento da legitimidade da variedade linguística, ou dialeto, dos alunos, não lhes contrapondo, ou não lhes impondo simplesmente, sem reflexão, a norma culta; por outro lado, há o equívoco de adotarem apenas a modalidade escrita da língua como a única legítima. (CYRANKA, 2011, p. 45)..
• 2
  [...] qualquer que seja o caminho trilhado pelo professor, no sentido de levar seus alunos a se engajarem no processo de desenvolvimento de competências de leitura e de escrita, requer o cuidado de também levá-los, como dito anteriormente, a construir crenças positivas sobre essa sua capacidade. O ponto de partida é o reconhecimento de que são usuários competentes de um sistema complexo que eles próprios adquiriram naturalmente e com o qual são capazes de atuar em seu meio social. A partir daí, orientados pela escola, devem se empenhar em conhecer recursos linguísticos mais sofisticados, para os quais precisam estar treinados e preparados. (CYRANKA, 2014, p. 134).
• 3
  “apresentam escassa força avaliativa. São traços que não identificam uma diversidade social e, geralmente, apresentam uma distribuição regular nos diferentes estratos sociais.” (GORSKI; COELHO, 2009, p. 82).
• 4
  Gorski & Coelho (2009), aside from presenting the Labovian definition of these two phenomena, they give examples to illustrate the difference among them, and it is going to be highlighted the following ones: as the stereotype, socially stigmatized forms such as nós fumo (in place of nós fomos); as the marker, the alternation between tu and você, that has different stylistic and social features, depending on various factors; as the indicator, the variation between peixe/pexe ou dinheiro/dinhero, rarely noticed among different varieties.
• 5
  “as forças inevitáveis que operam historicamente em prol da variação e da mudança geram um natural distanciamento entre essas normas.” (VIEIRA, 2013, p, p. 61).
• 6
  [...] preceitos dogmáticos que não encontram respaldo nem nos fatos, nem nos bons instrumentos normativos, mas que sustentam a cultura do erro e têm impedido um estudo adequado da nossa norma culta/comum/standard. Esta norma (o uso real, portanto) e os comentários dos nossos bons gramáticos costumam contradizer tais preceitos dogmáticos. Apesar disso, eles são repetidos como se fossem verdades absolutas e são tomados como justificativa para humilhar, constranger e prejudicar as pessoas. (FARACO, 2008, p, p. 92).
• 7
  “para além da competência inata, é compartilhado por sua comunidade ou região. É o saber social da língua, recebido primeiramente no âmbito familiar” (BARBOSA, 2007, p, p. 37).
• 8
  “qualquer variante linguística que não pertença a um dado conjunto de variantes eleitas como padrão seria uma variante errada para o uso formal” (BARBOSA, 2007, p, p. 38).
• 9
  Em síntese, nas construções com uma só forma verbal com clíticos que são efetivamente adquiridos no PB (<me, te, se> reflexivo), os dados revelam que se dá a próclise generalizada. Com os clíticos que se relacionam mais especificamente ao domínio culto, dois deles garantem a sobrevivência da ênclise, ainda que pouco produtiva: <o, a(s)> e <se> em construções de indeterminação do agente. Na escrita, tem-se outro sistema de colocação pronominal com forte condicionamento morfossintático, sensível ao contexto de verbo em posição inicial, certa atuação das partículas atratoras tradicionais, além do tipo de clítico. De todo modo, é relevante a diferença da escrita do letrado brasileiro em relação ao que se propõe na norma gramatical, tendo em vista duas características que podem ser destacadas: (i) a presença expressiva de próclise em contextos sem tradicionais atratores, em particular diante de sujeito; e (ii) a presença, menos expressiva mas não desprezível, de ênclise em contextos com tradicionais atratores. (VIEIRA; FREIRE 2014, p, p. 101).
• 10
  The mesoclisis was not investigated because we judged it as not a priority since its use is restricted to past periods of the language or specific traditions of determined textual genres (VIEIRA, 2002).
• 11
  According to Bortoni-Ricardo (2004), hypercorrection or ultra-correction is a phenomenon that occurs when a language user tries to adjust to the standard norm and overcorrects oneself in certain contexts in which such correction is not appropriated. In other words, the hypercorrection is the correction of what is already correct. Then, such correction is actually incorrect.
• 12
  “as orientações devem estar restritas ao campo da recepção dos dados (reconhecimento pela leitura/escuta – do domínio religioso ou literário, por exemplo), e não da produção, sob pena de promoção de arcaísmos.” (VIEIRA, 2019a, p. 257).

Publication Dates

• Publication in this collection
  30 Mar 2022
• Date of issue
  2022

History

• Received
  10 June 2020
• Accepted
  23 Aug 2020