Education is a problem in Brazil, in terms of its provision, its retention of students, and its overall quality. The Programme for International Student Assessment (PISA) consistently places Brazil near the bottom of the list of the 65 participating countries. In the UN’s Education Development Index, Brazil is ranked in 79th position whilst Portugal is ranked 41st, a position it shares with Chile; Argentina and Uruguay occupy positions 49 and 50, respectively. Although the PISA and other metrics are not without their problems or critics (Zhao, 2021), the quality of public education in Brazil is a persistent issue highlighted by political parties on both sides, official governmental agencies (e.g. Sistema de Avaliação da Educação Básica or Evaluation System of Basic Education) and NGOs (e.g. Todos pela Educação or All for Education). Of particular concern are student completion rates/drop-out rates, which for secondary education (ensino medio) reach a staggering 41 per cent for students who are regularly made to repeat school years (O’Neill & Massini-Cagliari, 2019, p. 46). Overall there are high rates of illiteracy; the official figure stands at 9 per cent of the population but in terms of functional illiteracy, this has been placed at between 77 per cent (Faraco, 2017b, p. 362) and even as high as 90 per cent (Psacharopoulos, 2015).

Issues related to language have often been proffered as not only aggravating factors but decisive factors in the academic failure of students (Soares, 2017a) and school drop-out rates (O’Neill & Massini-Cagliari, 2019) since the language of instruction and assessment is linguistically very far removed from the speech of the great majority of the population, especially the great mass of students from poorer backgrounds. Some have even claimed that the magnitude of these differences warrants Brazilian Portuguese being considered to be the language with the most native speakers that does not have a written system (Perini, 2003).

The difference between the official standard language used as means of instruction and basic form of assessment and the everyday speech of children can be challenging for education systems in every country around the globe but the challenges are considered to be greater in Brazil due to (a) the magnitude of the differences between everyday speech and formal writing and (b) the existence of severe linguistic prejudice against non-standard forms of speech and their speakers, which is correlated with the extreme wealth inequalities in the country.

With reference to (b), numerous PhDs, masters dissertations, academic articles, and books have been published in Brazil addressing the question of linguistic prejudice or wholly dedicated to the topic, in the disciplines of linguistics and education (Bagno, 2002; Leiser Baronas & Pagliarini Cox, 2003; Leite, 2008; Paixão de Sousa, 2010). It is generally accepted that there are two types of Portuguese spoken in Brazil, Educated Brazilian Portuguese and Popular Brazilian Portuguese (Lucchesi, 2002, 2015), which correlate with social class, and that speakers of the latter are at a disadvantage in educational settings since the language of school is closer to the former. The general strategies aimed at tackling problems relating to language and education are therefore almost entirely focused on changing the erroneous beliefs of individuals about the features of Popular Brazilian Portuguese.

I question the effectiveness of this belief-orientated strategy. I conceive language and education problems as forms of symbolic violence (Bourdieu, 1986, 1991) and argue for the need to concentrate not only on changing beliefs but also analysing the material conditions that favour and promote these beliefs and the social structures that sustain them and reinforce social injustices. Nancy Fraser (1995, 2000) identifies two types of remedy to address social injustices, ones based around the concept of Recognition and others around the concept of Redistribution. The former involves the identification and acknowledgement of a particular entity or collective (e.g. lesbians, the working class) and an attempt to change opinions and positively valorise these collectives; the latter involves the restructuring of systems and the redistribution of economic and political resources (e.g. the liberal welfare state, means-tested grants, quotas for underrepresented groups for jobs, etc.). The distinction can also be expressed in terms of the aim of change: cultural opinions vs sociopolitical and economic structures. Fraser (1995) makes the point that in social activism in the United States the overwhelming focus has been on culture and the recognition of identities to the detriment of concerns relating to the political economy, the restructuring of systems, and the redistribution of resources. The overall result is that real change is being hindered. Likewise, drawing on this work, Block (2018) highlights how most language education research is today recognition-oriented and advocates for more consideration of economic and class-based injustices.

In modern Brazil a major challenge for education reforms is the structural inequality of educational provision for the most disadvantaged in society. This ranges from the lack of universal education provision (61% only for secondary education) and the material conditions of schools (precarious water and sewer supplies and inadequate toilet facilities inter alia; (Soares, 2017b)), to the physical effects that poverty and, in some cases, hunger and malnutrition have on learning and school attendance. These factors are crucial structural concerns, but in this article I concentrate on different types of structure: the actual linguistic structure of the standard language used as the means of instruction and basic form of assessment, in opposition to how the language is spoken by the great majority of Brazilians, and the social and educational structures that reinforce and legitimise this standard language. I argue that recognition-orientated strategies aimed at changing attitudes towards non-standard forms of the language need to be accompanied by strategies that advocate for structural changes in (a) the standard language, to make it more readily resemble the actual speech of Brazilians, and (b) how this standard is used as a means of instruction. With respect to the latter, I argue that the school system is rigged to favour those who have access to formal written and spoken language and, essentially, is not designed to teach this variety to non-native speakers but, paradoxically, to assess the extent to which it is acquired. Students then have to try and master this variety alone which, given the gulf between the official standard and speech, represents a gargantuan task for students.

I provide an extended historical overview of the Portuguese language in Brazil from independence in order to establish why there is such a gulf between the written standard taught in schools and actual linguistic usage in Brazil and why non-standard speech is subject to such linguistic prejudice and discrimination. I argue that an appreciation of the historical context reveals how opinions about language were/are based on more deeply ingrained social ideologies, and that these were manifested and reinforced through practices and structures in the context of a newly independent republican Brazil that was grappling with its social realities and desired to affirm itself as a civilised modern state. The historical perspective shows how opinions and structures are often symbiotic and self-reinforcing and that social change is possible but only when attention is paid to the material and the ideological, to structures and opinions.

Conceptual Framework

I frame my discussion on “opinions” around the polar concepts of Civilisation and Barbarism. These antipodes have their roots in the Greek and Roman concepts of “civilised” peoples residing in ordered urban societies governed by the rule of law which provided security and potential progress and prosperity for all its citizens versus those barbarous, primitive, tribal people residing in the untamed hinterlands, who were a law unto themselves and lacked the intellectual and cultural sophistication of the Romans and Greeks. Language has always been a crucial and, at times, deterministic factor in these debates. The term “barbarian”, an antonym of the word politēs [citizen], was used to designate all non-Greek-speaking peoples and its origins lie in the onomatopoeic, discordant, and unintelligible bar-bar noises which foreign languages resembled to Greek ears (O’Neill, 2020, pp. 2–3). These antipodes are being used since multiple and different social distinctions and preoccupations can correspond to the opposing poles, and thus they are not reductive. For example, in Spanish America, where the civilisation-barbarism distinction has been most used, the antipodes have been correlated with urban vs rural, formal vs informal, white vs non-white, and modernity, progress, and industrialisation vs backwardness and underdevelopment. In general, the distinction between what one desires oneself and one’s society to be and the fear of becoming the opposite, the other, generally holds true.

The Language and Education Problem and Proposed Strategies

Brazil is big: all the countries of the EU would fit easily within its 8.5 million square kilometres and, with a population of roughly 820 million, it is the sixth most populous country on the planet. Although a multilingual country, the overwhelming majority of its citizens speak only Portuguese, which displays much variation generally along three different continua: urban vs rural, written vs spoken, careful monitored language vs informal spontaneous language (Bortoni-Ricardo, 2004). Many languages around the world also vary with respect to these continua; however, Brazilian Portuguese stands out since the speech of the educated and affluent who live in the major urban centres does not usually correspond, not even in formal and monitored contexts, to the language used in official administration and, traditionally, that used in teaching materials and assessments. This language corresponds to a “standard norm”, which was artificially created at the end of the nineteenth century (Faraco & Zilles, 2017).

In (1), I give some of the main linguistic characteristics of present-day unplanned discourse in Brazil and how this contrasts with the standard norm. Note that, although widely attested, these features are not all present in all varieties or consistently used by all speakers of a particular social or geographic variety. Space limitations do not allow for a detailed discussion of how Brazilian Portuguese became so different; suffice to say that it was the result of an almost total disregard for education from colonial to modern times combined with prolonged and extensive (a) contact with other languages, (b) acquisition as a second language by adult speakers (mainly indigenous peoples, African slaves, and foreign immigrants), (c) regular exponential increases in the number of native speakers, and (d) dialect mixing and koinéisation due to population migrations. The linguistics literature is clear in predicting the outcome of such social circumstances: rapid and, at times, transformative linguistic change, particularly towards the simplification of the inflectional morphology (Bentz, Verkerk, Kiela, Hill, & Buttery, 2015; Bentz & Winter, 2013; Lupyan & Dale, 2010; McWhorter, 2007; Nettle, 2012; Nichols & Bentz, 2017; Trudgill, 1986, 2001, 2010, 2011)) and complexification of the phonology as a marker of local identity.

(1) List of the linguistic features characterising the unplanned discourse of a number of varieties of Brazilian Portuguese.


A proclysis in main clauses me fala mais devagar por favor
me deram muitos presentes
o homem me disse que te amava
fala-me mais devagar por favor
deram-me muitos presentes
o homem disse-me que te amava
speak to me more slowly
they gave me many presents
the man said he loved you

B loss of pronominal verbs os guerreiros precipitaram
ela reclinou ao peito dele
os guerreiros precipitaram-se
(ela) reclinou-se ao seu peito
the warriors rushed forward
she rested on his chest

C omission of object pronouns, especially indirect object abra o documento em anexo, leia e envie as suas perguntas por favor abra o documento em anexo, leia-o e envie-me as suas perguntas por favor open the attached document, read it, and send me your questions please

D use of a gente for 1PL subject pronoun instead of etymological nós a gente se vê logo?
a gente não gosta de viajar
(nós) vemo-nos logo?
(nós) não gostamos de viajar
We’ll see each other later?
We don’t like travelling

E substitution of clitic object pronouns for stressed pronouns combined with tendency to express subject pronouns ela viu ele; eu amo ela
eles viram a gente
a gente vai fazer isso
eles falaram para mim
ela deu para eles
vocês falaram/disseram para eles
(ela) viu-o; (eu) amo-a
(eles) viram-nos
(nós) vamos fazê-lo
(eles) falaram-me
(ela) deu-lhes
(vocês) disseram-lhes
she saw him; I love her
they saw us
we are going to do it
they spoke to me
she gave them
you told them

F loss of plural agreement on nouns and adjectives os livro mais ilustrado
os menino inteligente
os livros mais ilustrados
os meninos inteligentes
the most learned books
the intelligent children

G loss of agreement features on verbs os menino pega o peixe
os professor fala muito
os meninos pegam o peixe
os professores falam muito
the children catch the fish
the teachers speak a lot

H mixing of etymological verbal and pronominal systems. Those in bold correspond etymologically to the você paradigm and those in bold italics to the tu paradigm você, fala mais alto, eu não te vejo e não consigo escutar as suas palavras (São Paulo)
tu sabe onde tu nasceu e onde nasceram os teus pais? Me diz por favor e eu te digo onde eu nasci. (Porto Alegre)
você, fale mais alto, não o vejo e não consigo escutar as suas palavras.
(tu) sabes onde (tu) nasceste e onde nasceram os teus pais? Diz-me por favor e te digo onde nasci
you, speak louder, I can’t see you and I can’t to hear your words.
Do you know where you were born and where your parents were born? Tell me and I’ll tell you where I was born

I loss of subjunctive forms você quer que eu saio?
não acho que tem razão
(você) quer que (eu) saia?
não acho que tenha razão
Do you want me to get out?
I don’t think he’s right

J substitution of synthetic future and conditional for the periphrastic forms “going to” and “was going to”, respectively a gente não ia conseguir porque iam pegar a gente e iam cortar as nossas línguas
o que vocês vão fazer? Eu vou dizer para eles que não vou pagar isso
(nós) não conseguiríamos porque nos pegariam e cortariam as nossas línguas
o que farão? Eu direi-lhes que não o pagarei
We wouldn’t make it because they would catch us and cut out our tongues
What will you do? I’ll tell them that I won’t pay it

K existential use of the possessive verb ter [have] instead of the auxiliary haver [have] Tem calças dentro do armário
Tinha muitos livros na biblioteca
Há calças dentro do armário
Havia muitos livros na biblioteca
There are trousers in the wardrobe
There were many books in the library

L reduction of levels of deixis from three to two esse prato foi muito bom!
esse homem quer falar com a gente
não quero nem esse aqui nem esse aí senão aquele lá
este prato foi muito bom!
esse homem quer falar conosco
não quero nem este nem esse senão aquele lá
This dish was lovely!
That man wants to speak to us
I don’t want this one or that one but that one over there

The ability to write and speak the standard norm is seen as a marker of civility in contrast to popular varieties, often considered as barbaric and degenerate forms of the language unsuitable in educational contexts (Britto, 1997; Massini-Cagliari, 2004; O’Neill & Massini-Cagliari, 2019)). The educational system, as pointed out by Soares (2017a) and validated by the research of Patto (2010, 2015), can also perpetuate and reinforce such beliefs, and can even contribute to the academic failure of children from less wealthy families. It can be responsible for the reconceptualisation of linguistic differences as linguistic deficiencies, their non-standard forms considered not only a reflection of cognitive deficiencies originating from insufficient and inadequate linguistic input but also a hinderance to their cognitive development since their non-standard use of language is erroneously assumed to be inadequate for logical and formal reasoning. This logic is in line with the civilisation-barbarism dichotomy: the barbaric speech and ways of the masses are at odds with the civilised environment of the school. The noted academic failure (Soares, 2017a) and drop-out rates of these students (O’Neill & Massini-Calgiari, 2019) merely reinforce the dichotomy and endorse placing them on the side of barbarism. The students fail due to inherent deficiencies in themselves or in the (sub)culture they belong to; their non-standard ways of speaking Portuguese are reflections of this deficiency.

In Brazil there is a wealth of academic publications in both linguistics and education and literacy studies on how to tackle this problem, most underpinned by what Soares (2017a) defines as the theory of differences and the proposal of bidialectalism. Such an approach endeavours to change conceptions of non-standard forms as deficiencies and recategorise them as differences that are neither better nor worse than those of the standard. The onus is on the teacher and the school system to ensure that in the classroom these views are communicated to all students and it is made clear that different forms are appropriate in different contexts. Within such a theory, students who naturally use such forms are therefore taught to become bidialectal and learn the standard forms. This approach is couched within the context of a “transforming school” which is conscious that the distinctive values attributed to different forms of speech are due, in the main, to social and historic factors which disadvantage and discriminate against the poor, but is also conscious that the social and economic progress of the student necessarily depends on the acquisition of the standard. Within a “transforming school”, this acquisition is not conceived as the student needing to accept, adapt, and assimilate to these unjust social structures and mores; rather, the student must be aware that these structures exist but also have sufficient knowledge of the standard in order to be able to exploit and take advantage of the opportunities that this knowledge offers. The ultimate aim of such bidialectalism within a transforming school context is to produce socially aware students who have the linguistic and cultural tools necessary to actively participate as citizens and to combat social inequalities.

This theory of differences within the context of a “transforming school” was first put forth by Magda Soares in 1986 in her seminal and extremely influential book Língua e escola. Thirty-three years later, however, it must be concluded that linguistic prejudice still exists in Brazil and therefore this theory of differences has not been particularly effectual. The reasons for this are manifold and could be independent of the validity of the theory.1 I suggest, however, that the strategy of the theory of differences is flawed since it is premised on the desire and necessity to change people’s beliefs in the hope that future generations may attempt to change the structures and practices which generate, reinforce, and perpetuate these beliefs. Such an approach, in which non-standard forms of language are not inferior to standard forms but merely conceived of as being different, is typical of most Brazilian sociolinguistic studies. Note also that scholars in education studies have drawn upon and been inspired by such studies. These studies have been criticised, however, for being sociologically naïve (O’Neill & Massini-Cagliari, 2019, pp. 47–52) and, whilst valid and useful, of limited effectiveness since they aspire to achieve social change via a “principle of error correction” (Lewis, 2018) whereby the focus is on academics sharing their objective and scientific knowledge and changing the beliefs of individuals but not analysing the political, historical, and social factors that sustain and reinforce such beliefs and the material structures that favour and promote them.

In Brazil, there has been much excellent sociolinguistic research on non-standard, stigmatised forms of Brazilian Portuguese, the majority inspired by the work of the American sociolinguist William Labov. The studies typically take a “scientifically objective view” of language and, via sophisticated methods, systematically map out the usage of non-standard forms, explain how they are correlated with certain broad social categories (usually social class) and how they are linguistically justified within the linguistic system of the non-standard speaker. In general, such studies foreground the variable of social class and thus reduce variation within Brazilian Portuguese to a crude dichotomy between popular and educated varieties (Lucchesi, 2017; Mattos & Siva, 2004). The general take-home message from most studies is that “the different varieties of a language are not ugly or beautiful, right or wrong, good or bad, elegant or inelegant; they are simply different” (Fiorin, 2002, p. 114). Such statements fail, however, to understand the important role which value plays in language and how language can be related to other social ideologies and preoccupations. That is, some forms of a language are socially better and are socially more elegant, and are socially much more useful in being perceived as successful, educated, and even beautiful (see also Cameron, 2012 and her concept of verbal hygiene). Thus, sociolinguists and scholars in education who draw on the conclusions of sociolinguistic research are largely ignored, since such “scientifically objective” views are out of sync with those of the general public, who perceive the societal value of language. More worrying, however, is that the views of academics “are interpreted as opinions from the far left which blindly advocate the acceptance of all types of diversity” (O’Neill & Massini-Cagliari, 2019, p. 50). This situation can lead to entrenched, polarised beliefs. A case in point is the famous public outcry and rejection of an officially endorsed pedagogical textbook aimed at introducing an inclusive, non-discriminatory approach to language variation within the classroom (O’Neill & Massini-Cagliari, 2019, pp. 41–44); the strategy of the author was underpinned by the theory of differences and the proposal of bidialectalism and within the public debate the book was deemed to be “academic trash dressed up as cultural avant guard” by a “circle of false intellectuals” and “enemies of good Portuguese” who were “barbarizing the country with a primitive language” and wanted to return it “to a tribal system where each person spoke how they wanted” (for an analysis of the public comments, see Leiser Baronas and Pagliarini Cox (2003)).

In short, many modern sociolinguistic studies can be viewed as lacking with respect to their “socio” dimension (see similar criticisms by Severo & Görski, 2017, pp. 122–123). This is not just true of Brazil but of the discipline more widely and has been highlighted by Fishman (1970; 1971; 1972; 1985; 1991, p. 127), who makes a distinction between a sociology of language and sociolinguistics, and notes how sociolinguistics increasingly became more “linguistically” orientated towards language variation and change and increasingly more removed from sociological and political concerns. Education researchers have drawn on sociolinguistic studies, internalised the distinction between Educated and Popular Brazilian Portuguese, and focused their efforts on changing opinions towards popular varieties of the language and fostering (very laudable) non-discriminatory teaching practices towards students who speak such varieties.

All this is good and necessary in a society, such as Brazil, in which the poor are generally neglected and even denigrated. However, as revealed by the work of Bourdieu (1896, 1991), societies are often structured in economic and social ways which reinforce the hierarchical relationships and value judgements between standard and non-standard forms, and the education system can contribute to this. Moreover, even if opinions towards non-standard forms were to improve drastically within educational contexts, apart from a (very valuable) heightened sense of worth and dignity for speakers of these non-standard forms, would such a change mean that the education system in Brazil would improve dramatically? Would speakers of these non-standard forms be more successful in assessments based on the standard norm? Would they see their employment opportunities improve? And would Brazil cease to be characterised by huge social and wealth inequalities? I think not (for similar arguments for English, see Block (2018)).

In fact, in the PISA data mentioned at the start of this article, in all countries analysed around the globe both illiteracy and drop-out rates were correlated with socio-economic status, and children from socio-economically privileged backgrounds consistently outperform disadvantaged students. Surprisingly, however, for a country with such wealth inequalities as Brazil, there was no significant difference with respect to the attainment of the two groups of children when compared with other OECD countries (OECD Brazil, 2018, p. 5). That is, across all countries disadvantaged children perform worse, but this is no worse with respect to socio-economically privileged students in Brazil than in other countries, mutatis mutandis. Thus, the level of education is generally low across the board in Brazil with only 50 per cent of students attaining at least Level 2 proficiency in reading compared to the OECD average of 77 per cent. These data would seem to suggest that there are fundamental structural problems with education in Brazil, which affect most students, and low attainment is not merely a result of negative opinions towards the speech of the disadvantaged classes.2 Some of these structural matters may be related to methods of teaching and assessment, but with specific reference to language, I suggest that a fundamental structural problem is the significant distance between the standard norm which still prevails and the actual linguistic practices of almost all Brazilian students. That and the doubts over whether the education system actually provides sufficient and adequate teaching for this written variety to be learned or merely assesses the extent to which it has been acquired. These are what I would identify as the linguistic “structural” problems facing language and education in Brazil.

In what follows, I provide a historical overview of how these structural problems arose and how they are intimately related to other social concerns related to the civilisation-barbarism dichotomy.

Historical Sociolinguistic Overview of the Portuguese Language in an Independent Brazil

To understand the dynamics of language and society in an independent Brazil it is necessary to go back to the eighteenth century, when there was a marked increase in Brazilian students at the University of Coimbra, a hotbed of enlightened thought and philosophy (Maxwell, 1997, p. 104). This institution played a fundamental role in the construction of the Brazilian political elite, who were characterised as being “uniform with respect to their ideas and training” (Carvalho, 2006, p. 21), and who undoubtedly shared a common way of speaking Portuguese.

Regarding thinking about language, in early modern Europe the civilisation-barbarism dichotomy was expressed as one between savants [learned] and rustiques [rustics]. Linguistically, the savants had knowledge of many languages whilst the rustiques either knew only one remote and obscure language or spoke only one language, badly (Bouza Álvarez, 2018, p. 19). As early as 1574 the Portuguese grammarian Nunes de Leão (1530–1608) denounced certain semantic usages as pertaining to the plebeus, also referred to as idiots (Burke, 2004, p. 29), and Bouza Álvarez (2018, p. 45) characterises the views about the speech of the common people in early modern Europe as speaking “blindly and foul-mouthed” and threatening the order of things.

The Brazilian students at Coimbra would have been exposed to such linguistic ideologies and would have striven to emulate the linguistic model of the metropolis. Upon return to Brazil, these elites provided a model for spoken and written Portuguese within their social networks but within the colony as a whole they constituted an “island of literate men in a sea of illiterates” (Carvalho, 2006, p. 65). The numbers of like-minded people, with what must have been similar linguistic practices and literacy abilities, expanded exponentially with the arrival of the royal court in Brazil in 1808 as a result of the Napoleonic invasion of Portugal. These Portuguese-born nobles and their administrative and social retinue changed the social and political scene of Rio de Janeiro. Ideologically, the observations in (2) suggest that their arrival added to the general and well-attested sense of Brazil and its inhabitants being degenerate in contrast to themselves and the metropolis (Russell-Wood, 2002). Note, however, that many of these new arrivals would, in time, merge with the Brazilian elites and therefore reinforce the idea of the elites (and their particular way of speaking Portuguese) being civilised as opposed to the barbarous masses.


Rio is described [by those who accompanied the court] as an “‘inferno’; a Babylon corrupted by the pernicious effects of slavery; a land of perdition; a Godless land whose peoples were libertines, listless, physically and morally weak, and degenerate”.

(Russell-Wood, 2002, p. 110)

Regarding structures, the capital of the empire was now based in Brazil and the king passed numerous laws opening up trade and commerce with other nations. Brazil also now had a fully approved royal printing press, which dynamised intellectual life due to the circulation of ideas and the creation of public opinion (Pessoa, 2003, p. 176). Shortly after declaring independence from Portugal (1822) the (now) Brazilian emperor approved the establishment of two law faculties (in Olinda and São Paulo) which took over from Coimbra as centres for learning and the production of the elite. The elite were a socially dominant group of people who, like their forerunners from Coimbra, had similar outlooks and values (Carvalho, 2006) and their way of speaking Portuguese constituted a model for the different urban regions. They were also influenced by the philosophy of Positivism and influential in the establishing the Brazilian Republic. It is in this period that strong opinions regarding language become aligned with power and are manifested in structures.

Positivism is a philosophical outlook that holds that just as natural laws govern the universe, so too do they govern human societies and therefore scientific knowledge and reason can and should be used to control society and social life. In positivist philosophy, social change was a very top-down process whereby strong leadership by a select group of elite scientists and professionals could lead to improved social change and modernisation. As Reid (2014) explains, “All this struck a chord with sections of Brazil’s incipient middle class, especially among the army officers, teachers, engineers and doctors to whom positivism offered a role as self-appointed apostles of national development” (p. 81). The extent of the effects of positivist thought in Brazil is visible in the motto of “Order and Progress” which figured on the new national flag. In Republican Brazil the civilisation-barbarism debate was expressed in these positivist terms whereby civilisation was “Order and Progress” and its antipode, barbarism, was not only the lack of progress but associated with ideas of degeneration, prevalent in late nineteenth-century Brazil (Borges, 1993, p. 235). However, whilst the ideal was order and progress for the whole nation, the social and geographical proximity of the elites to the masses contributed to a sense of a ubiquitous threat of degeneration towards barbarism and therefore the need not so much to civilise the masses but to safeguard the nation against their harmful, corrupting effects.

There was also a racial and racist dimension to this idea of degeneracy and barbarism. Bear in mind that Brazil received more African slaves than any other country and over a longer time period (some 4.9 million between 1500 and 1866: Reid, 2014, p. 52). When the Portuguese court arrived in Rio in 1808, the population of the country was calculated at 2,323,386 people (Schwarcz & Starling, 2015, p. 209) of which 28 per cent were classed racially as whites, 66 per cent as brown or blacks (slaves 38%, free blacks and mulattoes 28%), and 6 per cent as Amerindians (Reid, 2014, p. 58). Moreover, from the arrival of the royal court to the cessation of the Atlantic slave trade, the influx of African slaves each year was steady, continuous, and increasing (24,410 a year from 1800 to 1810, 32,770 from 1810 to 1820, and 43,140 a year from 1820 to 1830: (Lucchesi, 2017, p. 370)). Thus, when slavery was finally abolished in Brazil in 1888, blacks and people of mixed race heritage were the overwhelming majority of the population. The republican elites, influenced by the combination of positivism and other pseudo-scientific social and racial theories (e.g. social Darwinism), subscribed to the tenets of scientific racism and thus race analysis “became dominant and dogmatic” (Borges, 1993, p. 44). The African element was considered as a degeneration of Brazilian blood and a determinative factor in the poverty, illiteracy, and ill health of the Brazilian masses (Borges, 1993; see also Schwarcz (1993)). Some prominent elites even went as far as to suggest that Afro-Brazilians should not be considered citizens or that they were not mentally fit to stand trial in court (see Borges, 1993).

Inspired by positivist ideals, the educated elites adopted the role of physicians to a sick society forever on the brink of degenerating into barbarism and away from order and progress. The mass of Brazilian people was considered somehow deficient and degenerate and in danger of degenerating even further and, more worryingly for the elites, of affecting the rest of society and frustrating attempts at national progress. As noted by Pino (1997) of Latin America in general, “National progress thus depended on forcibly assimilating these ‘barbarians’ into the dominant culture or physically removing them from society”, since otherwise the infection may spread.

Importantly, the Brazilian Republic not only had the ideas but also the power to convert these ideas into policies, practices, and social structures. Thus, Sevcenko (1995) notes how reforms of the first republic led to the transformation of public spaces in its capital, Rio de Janeiro; the popular masses were forcibly evicted and banished to the outskirts or the hills (where the future favelas would form). The promotion and funding of white European migration coupled with a ban against Asian and African immigration was also inspired by this civilising ideology and desire to stave off the spread of degeneration, as were programmes of small pox inoculation. Borges (1993) notes how this rhetoric of degenerations “turned into the central, near-official ideology of the conservative, oligarchical Republic (1889–1930)” whereby the fear and threat of the degeneration of the whole nation “furnished an ideological common denominator to diverse reform proposals of republican governments” (p. 240). It is in this social and ideological context that a codified standard language came to be adopted that was not entirely representative of the speech of the elites but based on the models of the old metropolis. The linguistic differences of the Portuguese spoken in Brazil was another mark of the degeneration of the Brazilian character and, in order to safeguard against this, a goal of what language should be was fixed.

The Language Dimension to Order and Progress/Civilisation and Barbarism

Unfortunately, there is very limited source material about the spoken linguistic practices of Brazilians from colonial, imperial, and republican times. The French botanist Auguste de Saint-Hilaire, who travelled to Brazil in the first half of the nineteenth century, noted that Brazilian Portuguese was different from that spoken in Europe and that “its inflections were not very varied”, pointing, perhaps, to a simplification of the morphological inflectional system (Saint-Hilaire, 1974, p. 155). Although there are few testimonies of the actual linguist features of the speech of the masses, we know that in the nineteenth century it was considered to be full of “gross mistakes” and characterised as “a language of savage people, uncivilised language, a defiled language” (Santos Silva, 2012, pp. 108–109). It is at this point that racial and linguistic prejudices merge since the speech of people of African descent and especially African slaves was vilified; Araripe Júnior (1848–1911), a founding member of the Brazilian Academy of Letters, referred to “the warped speech of the Africans”. Moreover, it was also widely accepted that this community was responsible for the speech defects in non-black Brazilians, as attested by an article by the Brazilian editor of a local Recife newspaper, O Carapuceiro, in 1842 (3), the speech of a Brazilian senator (4), and the Portuguese grammarian and honorary member of the Brazilian Academy of Letters, Cândido de Figueiredo (5).


These and other vices have many causes, the principal one being without a doubt the family dealings with African blacks, who are, as a rule, our first teachers. After this, there is the neglect of the state our spoken language, which before we scorned; we are content to speak it as we learned it from our nannies, almost always slaves. (Pessoa, 2003, p. 106)


The best thing to come from the abolition of slavery is for us to rid ourselves of that barbarous race which ruins our customs, the education of our children, the progress of our industry and all that could be useful, and because of them we have even lost our pure language. (Faraco, 2018, p. 44)


The thousands or millions of blacks, whom the colonisation of Brazil called from Africa to America, naturally and gradually adopted Portuguese words, constructing sentences in their own way; and the paid workers and colonists, certainly fewer in number than those who worked in the fields and on the rocks, were more concerned with their rubber and coffee than grammatical invasions, and so they idly allowed themselves to become imbued with fads of speaking which were constantly harming their ears. These fads went from the fields to the cities and, when the Brazilian grammarians realised their folly, many seemed to be of the opinion that it was too late to correct, and others, like Mr Paulino de Brito, tried to evict from the city the grammar of the blacks. (Figueiredo, 1909, p. 120, qtd. in Santos Silva, 2012, p. 175)

As with the degeneration of the Brazilian character, the degeneration of the Portuguese language spoken in Brazil, although never far from issues relating to race (Borges, 1993), was not directly associated with any particular colour but with the non-urban illiterate masses who were more racially mixed and described as “men without quality, the dregs of society, parasites of the social tree” but at the same time acknowledged as being “the great mass of the population” present also in the cities as the “urban rabble” (Faraco, 2017b, p. 138; Mattos Ilmar, 1987). These people had also contributed to the deformation of the Portuguese language, producing the degenerate form of communication which some referred to as “Brazilian”. The ideas of some of the literate urban class about this form of speech are revealed in the words, reproduced here in (6), of the character Lobo, nicknamed “the Grammar”, a worker at a Rio newspaper in the 1909 novel by Lima Barreto, Recordações do escrivão Isaías Caminha [Memories of the Scribe Isaías Caminha].


“Brazilian, doctor!” the Grammar said quietly. What is spoken here is not language, it isn’t anything: it’s a rubbish heap of filth. If Brother Luís de Sousa were to rise from the dead, he would not recognise his beautiful language in that confused mess, in that diabolic mixture of French, English, African, and Indian tongues, the screeching degenerate confusion of screams, shouts, and silences […] it’s hell! (qtd. in de Assis (2020)

Note that here the identification of the speech as a mixture of French and English reveals that it was not just the language of the lower classes which was looked down upon but also the speech of the literate class. We have already seen in (3) that in 1842 the editor of a local Recife newspaper, O Carapuceiro, makes a number of comments about the contemporary use of Portuguese. Pessoa (2003) notes that even though he reserves most scorn for what he calls “crude language” and/or “Luso-African gibberish”, he also decries the speech of educated young men who do “not know how to speak” due to their mixing of French with Portuguese, and even the speech of the most highly regarded in society, who, he complains, speak “incorrectly and crudely”.

This gradation in the degeneration of the Portuguese language spoken in Brazil, much like gradations of skin colour, posed a threat of contagion and of becoming the barbaric other, and so steps needed to be taken to place educated speech, and writing in particular, firmly on the pole of civilisation. It is in this context that the elites came to use particular forms of speech which were purposely distanced from local practices and aligned with their status as elites (for the educated speech of São Paulo, see Santos Silva, 2012). The variations in the speech of the elite, which were mostly related to lexical items and the positioning of clitics, came to be distinctive markers of particular political and ideological affinities in opposition to others (Oliveira, 2019). These small differences in speech sparked a debate about the standard language of Brazil. Note that considerations of incorporating the speech of the masses are totally absent from these debates.

A key figure was the politician and avid abolitionist Rui Barbosa (1849–1923), a graduate of the law faculties of both Olinda and São Paulo. Regarding language, as illustrated by his quotations in (7), he was a purist, associated “good” language with good thinking, manners, and taste, and was a fierce critic of different ways of writing Portuguese:


Language is the way in which we channel our ideas, when it is not drunk from the cleanest, most crystal-clear, purist channels, the thoughts of those who use it will not flow in a pure, crystal-clear, and clean way.

Those in politics, administration, and the press have regularly been, in all areas (all the more amongst us!), the ones who have most powerfully corrupted language and good taste.

(Barbosa (1902, p. 302) qtd. In Pinto (1978, p. 383))

Barbosa famously fiercely criticised (in a 560-page document) how the new Republican Civil Code had been drafted, highlighting its “use of the vernacular” and categorising it as a “crude, indigestible, and crippled piece of work”. And, when it was redacted (on the basis of his criticisms) by his old teacher the eminent scholar and grammarian Ernesto Carneiro Ribeiro, he produced another similar diatribe against it, stating the following:


If the law is not specific, it cannot be fair: “Legis tantum interest ut certa sit, ut absque hoc nec justa esse possit”. To be specific, however, it needs to be precise, well-defined, clear. And how can it be clear, if it has drained through the impure dregs of a language that is nought but a deluge. Why not let it be revealed in that purified and transparent language that tradition has refined throughout the ages? To aspire to clarity, simplicity, and precision without a suitable vocabulary and an exact grammar would be to desire the outcome without the means. Clarity in how laws are styled “depends, at the same time, on logic and on grammar”, says Bentham, “sciences that one needs to totally understand in order to adequately write laws

(Barbosa (1902, p. 302))

Language deemed appropriate for writing and public speaking is conceived as something which has been elaborated historically and has fixed forms and definitions, and which, as with science, people must strive to learn and fully understand. These ideas were prevalent in Europe from early modern times, when what Burke (2004, p. 90) has characterised as “the anxiety of instability” meant that variation in language was perceived negatively. Civilisation was related to order and rules, and implied following a code of behaviour. Language was included within this behaviour and thus changing or inhibiting one’s natural linguistic tendencies and speaking in another way could be conceived as part of a “civilising process” (Burke, 2004, p. 90). For Barbosa and others in the late nineteenth century, speaking and writing was about learning and striving to better oneself through the acquisition of something which was not originally known or natural but could be acquired through hard work and education. This gold standard to which to aspire was also a way for the elites to protect themselves from general degenerate linguistic tendencies. In Europe, linguistic purism was often related to the desire to have a language and society which were morally, socially, and ethnically pure (Burke, 2004, pp. 141–159). Given the social demographics of the Brazilian Republic, this was not possible from the standpoint of the white(r), Europe-centric elites. However, the aim of a non-degenerate and civilised society was there and it could be achieved through a combination of European immigration, education, and laws, and by having an official, civilised, pure form of Portuguese as the standard language, which not only the masses but also the elites had to strive to perfect, and which would translate into clearer and better thinking.

The purists won the debate, their victory sealed with the founding of the Brazilian Academy of Letters in 1897, of which Rui Barbosa was a founding member. The members of the Academy became the gatekeepers of what was considered institutionalised Brazilian literature and they encouraged and legitimised a type of Portuguese which was really quite distant linguistically from the speech of the great majority, if not virtually all, of the population, but remained the ideal goal for personal and national order and progress. They also liaised with printing presses and educational authorities to plan cultural activities to spread this linguistic norm (Lehmkuhl Coelho, Oliveira, Monguilhott, & Gorski Severo, 2014, p. 29). The creation of the Academy is just one example of how, in this period, linguistic ideologies were being physically materialised in practices and institutions. Another example of this is credentialisation, which “became an unwritten prerequisite for public positions” and was essential for certain other middle-class professions. The result was that social advancement was dependent on accredited knowledge to write and speak the standard variety and, as noted by Borges in (9), this had effects on social mobility, especially for non-white citizens:


The credentialization closed off some of the ladders of upwards mobility for autdidact plebians. For example, the profession of schooolteacher, which had been filled by black and brown men, was increasingly occupied by white women graduated from teachers’ academies. And through some informal exlusionary practice that has not been identified (though it was probably simple preference for the white sons of European immigrants) the school system, the system of artisan careers such as typesetting, and the patronage system recruited fewer colored intellectuals. As a result, by the generation of 1930 there was a perceptible decline in the presence of mulato and negro intellectuals. (1996, p. 43)

This relationship between credentials, social mobility, and language is represented in the aforementioned novel by Lima Barreto. The eponymous narrator and protagonist, Isaías Caminha, a mulatto born in the interior of Rio de Janeiro state, moves to the capital in search of a better life, and imagines, as illustrated in (10), that if he were to gain the qualification of “doctor”, this would not only automatically improve his social standing but his mental and linguistic abilities, too:


Ah! I’d be a Doctor! I would redeem myself for the original sin of my humble birth, I would soften the deeply felt, harrowing, and all-encompassing torture of my colour […] Bound in the folds of my diploma would be the respect and esteem of all. Sure in the majesty of my own person, with it I, as a man, would walk more upright in life. I would not stutter or stumble with my words, I would be able to speak freely, to pronounce loudly my thoughts which would be writhing around in my brain […] Ah! Doctor! Doctor! To be called so in the streets, in the squares, on the roads, in the parlours, receiving the well-wishing of people: Doctor, how was it? How are you, Doctor? That would be superhuman. (qtd. in de Assis (2020))

This novel also demonstrates how language use could be directly aligned and influenced by financial concerns since, in the novel, the editor of a Rio newspaper is forced to change his policy of publishing articles in a language more similar to that spoken in the city, having been accused of discriminating against Portuguese-born journalists and the fact that the Portuguese community in the city was particularly influential and bought most of the advertising space in the newspaper. Although the novel is a work of fiction, the linguistic issues discussed are corroborated by the results of historical research which have revealed the constant preoccupation at this time with language use and, more importantly, how this translated to material structures and practices of policing and censorship (Silva, 1998). A case in point is the life of the novel’s author, Lima Barreto, who Borges (1996) refers to as “an emblem of the exclusion of non-white students from the doutor route into the intelligentsia and privileged public employment” (p. 49). Barreto, like many non-white students, left his studies and abandoned the teaching academy “because one professor failed him repeatedly” (Borges, 1996, p. 49). As is common in modern Brazil, social discrimination was correlated with skin colour but not exclusively defined by it since other students of colour did graduate from the academy. As exemplified in his novels and in the quotation below in (11), in response to being compared to the novelist Machado de Assis (also a mulatto), Barreto was outspoken and very critical of the artificiality and importance of language in the republic and was unwilling to acquiesce to the invented norms of a largely white elite. It is not surprising that his literary genius was not acknowledged by his contemporaries and that he was never granted entry to the Academy of Letters. This fact is another example of how language use came to be materialised in exclusionary practices in the republic.


Machado wrote in fear of the [the grammarian Feliciano de] Castilho and hiding what he felt, so as not to lower himself; I’m not afraid of being spanked by Feliciano and I write with great fear of not saying everything that I want and feel, without calculating whether I lower myself or whether I exalt myself. (Borges, 1996, p. 49)

Sadly, the preoccupation with the correctness of the Portuguese language and the importance of credentials did not translate into a more effective public education programme to teach citizens this form of speaking or writing. To the contrary, it actually caused a loss of interest and disregard for public primary education since resources and efforts were directed more towards providing further education for an already literate class of individuals (Pessoa, 2003, p. 162). There were very few schools in Brazil, resulting in extremely high illiteracy rates. Also, the quality of teaching and the conditions of those schools that did exist were poor (Scarato, 2016, p. 220). And, although, new teaching methods from England and France had been adopted in Brazil since imperial times, these were to benefit the elites, which is ironic since in Europe they were conceived as ways of expanding education to the poor. As Scarato (2016) notes,

The idea of modernising Brazil sought inspiration in European ideas, but adapted them to the colonial [now imperial] context. Poor people in Brazil were confined to the “waiting room of history”, while the elites guided a new nation; it was clearly more important to refine the elites (equal to Europeans in shaping Europe as a bastion of modernity) than to educate the poor. (p. 244)

Note that while this excerpt refers to imperial times, the situation did not substantially improve in the republic, despite the calls of Rui Barbosa for educational reform.

Thus, in the Brazilian Republic, opinions, power, and structures aligned with the result that ideas about society and language came to be materialised in social practices and systems of practices to constitute what Foucault termed “technologies of power” (Foucault & Sheridan, 2020): the material ways in which the conduct of people is shaped and controlled so as to produce certain desired effects (civilisation, order, progress) and averting certain undesired ones (barbarism, disorder, degeneration) (see also Rose, 1999, p. 52; Lehmkuhl Coelho et al. (2014)). However, in the Brazilian Republic these technologies and structures (educational infrastructure, materials) were not robust enough or backed up with the resources necessary in order to be successfully implemented. Moreover, the underlying ideas were ambivalent; civilisation was at the same time a quality to be extended to the masses but also a virtue which needed to be safeguarded against the degenerate influences of those masses. In terms of language and education, this translated into establishing a written standard for Portuguese that was very far removed from the speech of most people and not providing a robust and universal system of education. In the next section I show how in the modern period, despite the “democratisation” of education, language was still a means of social exclusion since this artificial standard was integrated and accepted in the education system but no pedagogical measures were ever consistently put in place to help students achieve active fluency in it. The distance between the educational standard norm and everyday speech has always meant that acquisition of this form of language is extremely difficult. As explained throughout, I identify these factors as the central structural issues of the language and education problem in Brazil.

Modern Brazil

From the 1930s onwards, the Portuguese language was central to the construction of a Brazilian national identity (Oliveira, 2000).3 Education also became highly centralised and governmentally controlled. The emphasis was on teaching students “correct” Portuguese on the basis of “good” writers and there was a heightened formal concern with the regulation and control of the syllabus and teacher training programmes (Lehmkuhl Coelho et al., 2014, p. 29). In this way the idealised language of the late empire and early republic, now referred to as the norma-padrão [standard norm], became entirely legitimised, institutionalised, and policed within the education system: learning Portuguese in school meant learning this “standard norm”, despite it becoming increasingly distanced from the newly developing urban varieties.

However, education was still the reserve of the rich: as late as 1950, the average Brazilian worker only had 1.8 years of schooling (Reid, 2014, p. 113), and it was not until the 1960s with what has been called the “democratisation of education” (Soares, 2002, p. 166) that students from less privileged backgrounds had more access to education, albeit underfunded and lacking materials and fully qualified teachers (Pietri, 2010, p. 70). At this point the education system underwent quantitative changes with respect to the number of students it had to cater for and also qualitative changes in the type of student and their linguistic repertoire. No longer were the students exclusively from the most privileged classes with access to a written culture and perhaps a way of speaking more closely resembling the “standard norm”.

The education system adapted by providing specific content for Portuguese language classes based not only on knowledge about the language (i.e. the grammar) but also on the study of texts (Soares, 2002, pp. 166–167). However, these texts, which were “merely passages from consecrated authors” (Castro dos Santos, 2014, p. 11), were never the main focus of study: this was reserved for the metalinguistic, explicit passive knowledge of the “standard norm”. The focus was not on ensuring that students acquired active competence in this “standard norm” through rigorous training but merely that they had passive metalinguistic knowledge of it and were familiar with the technical terminology to describe it. Unsurprisingly, most students did not learn how to effectively reproduce this way of writing and speaking, which led, in the 1970s, to what has been termed a national “wake-up call to the language crisis” (Soares, 1998). Britto (1997, p. 100) notes how, at this time, there was “a generalised outcry, even in conservative sectors, due to the increasing difficulty experienced by students who had finished secondary education in reading and writing in a correct, clear, and well-articulated way.”

One of the methods to address this problem was to ensure that composition in Portuguese was compulsory for the national exams for access to higher education. Thus, the solution to the problem was focused on assessing whether the skill had been acquired instead of ensuring that the teaching was appropriate to acquire that skill. Unsurprisingly, this did not correct the problem; but it did highlight that “knowledge of grammar does not guarantee the student being able to produce a good and acceptable composition” (Britto, 1997, p. 101). There was, therefore, a shift away from a conception of language as an aesthetic expression, with a focus on rhetoric and teaching grammar and its concomitant terminologies (Soares, 1998, p. 169), towards the view of language as a form of communication, an expression of Brazilian culture, and an instrument for development. The name of the subject even changed in primary education, from “Portuguese Language” to “Communication and Expression” in the early years and then “Communication in Portuguese Language” in later years. However, the Portuguese language was still equated with the “standard norm” and behind the communicative approach was a positivist view of language as a behaviour which had to be perfected so that students could both express and receive messages efficiently (Soares, 1991, in Saviani, 2013, p. 379). This authoritarian form of education was conceived to be heavily based on repetition and copying existing models of writing and so one could say that there was a change towards drilling and teaching the “standard norm”. There was also a focus on looking at a broader range of texts which were not only from literature but also journalistic texts and texts used in advertising. Additionally, and problematically, there was an increased openness to the value of spoken language and the use of non-linguistic forms of communication.

These last points, together with the dropping of the term “Portuguese”, sparked strong opposition to the reforms, especially from the media and some teachers. These people had successfully mastered this standard norm and it formed part of their symbolic power (Kramsch, 2021), a sign of their civilisation and marker of their status as elites or of their rise up the social ladder. Meserani (1995, p. 19) notes how the media referred to these changes as being “the end of civilisation and the start of a new barbarism” (p. 19). The result was that the subject reverted to being called Portuguese language. As to exactly how Portuguese was taught, although in the textbooks there was an increased emphasis on studying both oral and written texts of different types, styles, and registers, there was a lack of specific methodologies or centrally endorsed practices to follow. Moreover, there were no official guidelines inviting reflection upon how to teach Portuguese and not even basic notions of what it meant to teach students their native language (Veit Holme, 2019, pp. 35–36). The result was that “it was down to each teacher to choose the contents and methods which were best suited to their own reality” (Veit Holme, 2019, p. 36), which usually resulted in reproducing how they themselves had been taught (i.e. the metalinguistic grammatical tradition).

This period also saw the introduction of ideas from modern linguistics into teaching training programmes (change of structures) and the result was that over the next two decades there was an appreciation of the diverse nature of language and how the speech of students represented legitimate linguistic systems of Brazilian Portuguese which were simply different with respect to the standard norm. Teachers of Portuguese with such training and/or some background in linguistics were therefore “fearful to say that they taught grammar” and so claimed that instead they were teaching “the educated norm” or “standard Portuguese”. Soares Gomes (2019, p. 22) notes, however, that it was not clear what was actually meant by “teaching the educated norm” and that there was a contradiction between discourse and teaching practice, the latter continuing with the standard norm.

There is an ongoing conflict between these different norms. The standard norm is dogmatic whereas the educated norm is nebulous and ill-defined: it ranges from being as arbitrarily prescriptive as the standard norm, in which case it has been dubbed the “short norm”, to encompassing any and all types of diatopic “educated” usage (Faraco & Zilles, 2017, pp. 184–186), which in a country as vast and populous as Brazil means that it encapsulates numerous different ways of speaking and writing. Indeed this “educated norm” is characterised as not being homogenous or uniform and by speech being markedly different from writing (Costa Freire, 2020, pp. 661–662). The central problem, however, is that although there have been numerous publications by linguists about different specific linguistic phenomena which are becoming generalised and even systematic in the educated speech of different Brazilian cities, there is a lack of comprehensive, structured, codified accounts of the different educated norms of specific cities, not to mention one for the entire country (see similar criticisms from Faraco (2017a, p. 23); Moura Neves, 2002, p. 239). So, the aforementioned contradiction between teachers claiming that they were teaching the “educated norm” whilst continuing to teach the “standard norm” is not surprising—what norm would they teach? To what would they refer students in cases of doubt? What teaching materials would they use? An actual codified guide to serve as a (non-prescriptive) reference point and for the elaboration of materials seems to be an essential constitutive element for another type of educational and linguistic reality.

Brazilian linguists realised this and the second half of the twentieth century was marked by much research on Brazilian Portuguese resulting in the publication of a number of grammars/compendia (Bagno, 2011; Bechara, 2015; Castilho et al., 1991; Cunha & Cintra, 1985; Faraco & Moura Neves, 2005; Moura Neves, 1999; Rocha & Silva Neto, 1984). However, these were not all specifically attempting to codify the educated norm but were more flexible and less rigid versions of the traditional standard norm, which differed according to their target audience (linguists, students, teachers), their level of prescriptiveness, and the grammatical generalisations considered valid for Brazilian Portuguese. Take for example the treatment of the passive construction with se of the type se vendem casas [houses are sold]: Bechara (2015, p. 575), Cunha and Cintra (1985), Rocha and Silva Neto (1984, p. 390), and Faraco and Moura Neves (2005, p. 354) all describe this construction and emphasise the need for the verb to agree in number with its internal argument (vendem-se casas not vende-se casas; see also Costa Freire, 2020, p. 664). However, Bagno (2011, p. 812) claims that this structure does not exist in Brazilian Portuguese and classifies its inclusion in grammars as “entirely inappropriate”, arguing that “it should be abandoned once and for all, along with the bizarre agreement it implies” (Bagno, 2011, p. 807).

In sum, in Brazil there is now a proliferation of norms: the aforementioned new trend of grammars correspond to what have been termed “grammatical norms” (Faraco, 2008), which are more flexible than the dogmatic traditional standard norm, which exists alongside the vague but much visible educated norm which, when it becomes more prescriptive, is termed the short norm. However, in educational policy it is not clear which norm is being endorsed and promoted. In the preliminary version of the official 1995 National Curriculum Parameters it is stated that the aim of teaching of the Portuguese language is to “break with the markedly traditional ideology that had permeated mother tongue education” and that “a new conception of language and speech is being instituted” (qtd. in Soares Gomes, 2019, p. 23). As to what this “traditional ideology” might be, in another section it is defined as “the excessive value given to normative grammar and the insistence on rules and exceptions with the concomitant prejudice against oral forms and non-standard varieties” (Soares Gomes, 2019, p. 24). It would seem, therefore, that there was a move from the standard norm towards one of the more flexible grammatical norms but nowhere is it specified what this change actually means apart from more textual analysis. Likewise, the most recent policy document, the National Basic Common Curriculum (2017), makes it clear that students ought to have knowledge of the standard norm but nowhere is it defined or reference made to materials or compendia which describe and prescribe it and how it is different from the traditional standard norm and the educated norm, which, as Costa Freire (2020, p. 660) has pointed out, can be the cause of much confusion in educational settings.

For an outsider the situation is truly baffling and one can only imagine the difficulties experienced by students and teachers. Even within the linguistics literature, and specifically Faraco (2008), who makes constant reference to the standard norm and distinguishes between it and the grammatical norms, it is not clear what specific materials/authorities the author is using as a point of reference. His book contains numerous examples of phrases which highlight the differences between these two norms but the only direct references are to ill-defined notions such as the “grammatical tradition” and “classical syntax” (see also Costa Freire, 2020, pp. 662–663).

It must be concluded, therefore, that the standard norm is an idealisation which some people know but most only have general and vague ideas about due to their exposure to educational materials and official documents written using this norm. In this way the standard norm is not that dissimilar from the educated norm. This norm for which, to my knowledge, there is no authoritative grammar or language manual, supposedly corresponds to the spoken linguistic habits of the educated and to the written language in (some) newspapers and (some) other formal materials. The fact that the educated norm has not been systematically codified, however, means that it has no existence independent of its use and so it cannot be formally and systematically learned or taught but only acquired. In the first instance this would only be possible if one were brought up in a family which spoke this variety or, failing this, through immersion by reading materials supposedly written in this norm or associating with people who have the spoken variety. The onus is then on the individual child/student to work out what the rules and generalisations of this variety are. This situation means that some are automatically privileged from birth with the educated way of speaking, whilst those who are not have no direct assistance to learn it; the best they can hope is to learn the standard norm in school and figure out how it overlaps with and differs from the more widely used educated norm.

However, there is no evidence that the educational system provides sufficient teaching to ensure active competence in the standard norm, instead of mere passive competence and understanding. Take for example nominal and adjectival plural agreement: in the standard norm where lack of agreement is fiercely condemned, these are categorical musts, but they are variable rules in Brazilian Portuguese overall (Brandão, 2013), which correlate strongly with levels of education and social class (Araújo, 2014, 2016; Araújo & Sousa, 2019). The ability to naturally and effortlessly produce these agreements is a skill that cannot be acquired via mere access to texts which display them; it is a skill which needs to be trained and drilled via exercises and practice. Some proposals have been put forward by linguists based on structuring learning where the focus is on doing exercises specifically based on the contexts where standard agreement is most likely to be omitted (Araújo & Sousa, 2019, p. 93; Lemle & Naro, 1977). Such proposals, however, have not been adopted into any official teaching policy.


The official standard norm for Brazil arose in a context in which the elites of the new republic were grippling with the social realities of their newly independent country and were looking for ways to ensure that it and its citizens could advance towards order and progress and not slip into degeneration and barbarism. The linguistic dimension of the solution was to establish a standard norm which was spoken naturally by hardly any native Brazilians but which represented a gold standard to aim for and one which was considered to produce positive cognitive benefits. This standard norm was then accepted and legitimised within the education system and in society more generally. However, universal education provision in Brazil was, until very recently, deficient and so the great majority of the population did not have access to this standard norm: as late as the 1980s the average Brazilian worker only had 3.9 years of schooling and in the mid-1990s it was approximately 6.6 years (3.3 years in the north-east), which was low by international standards (Reid, 2014, p. 113). Moreover, there is no evidence that the education system ever effectively taught students to achieve active, advanced proficiency in the standard norm through rigorous processes of drilling and teaching. When the authorities became aware of students’ lack of proficiency the solution was to ensure that there was an assessment to test proficiency instead of changing teaching practices to help students achieve it. The linguistic distance between the speech of the great majority of Brazilians and the educational standard norm combined with a lack of (a) an appreciation of the difficulties in acquiring a different form of speech and (b) teaching policies and methodologies to ensure active competence in the standard, is identified as a significant structural problem relating to language and education in Brazil.

The great majority of publications related to the language and education debate in Brazil are recognition-oriented and focused on matters of linguistic prejudice and discrimination. These approaches encourage a (re)valorisation of non-standard forms of Portuguese within the same educational and societal structures. These strategies have even succeeded in having official government directives explicitly denounce linguistic prejudice and promote respect for other varieties of the language. This is a positive development and one which I fully support and endorse. However, with regard to the structural problems identified in this article, the recognition-oriented strategies seem to have led to further ambiguities and lack of clarity with regard to the importance of the standard norm in educational policy and what exactly it is that students should learn. In the documents pertaining to the National Basic Common Curriculum, the standard norm is referred to on a number of occasions and it is repeatedly emphasised that students should have knowledge of the standard norm, be able to recognise it and identify the ways in which it is different from the speech of students. However, it is not clear whether students are expected to be proficient in this way of writing and speaking; the extract below highlights the ambiguities.


Knowledge of the language, the other semioses, and the standard norm should not be understood as a list of content disassociated from linguistic practices, but as ways to invite reflection as to how language works in the context of these practices. The selection of skills in the N[ational] B[asic] C[ommon] C[urriculum] is related to this basic knowledge so that the student can take ownership of the linguistic system that organises Brazilian Portuguese.(Governo do Brasil, 2017, p. 135)

The point, however, is that in a country so large and populous as Brazil there are various linguistic systems which organise Brazilian Portuguese. From the excerpt above it is not clear that the function of education is to ensure that the standard norm is the privileged linguistic system being taught since it could be understood that a child simply needs to be able to identify the standard norm and the contexts in which it is used. In fact, the curriculum documents only explicitly state on one occasion that students should know how to “write texts correctly, in accordance with the standard norm” (idem, p. 183). On the other occasions in which the standard norm is mentioned as being a necessary active competence this is with reference to students learning how to produce texts and the need for these to be adapted to the standard norm. However, this stipulation is often added at the end of a list of other skills and seems to be an afterthought. For example, for written reports it appears last, after the ability to handle and manipulate images and audio files (idem, p. 139 and p. 175).

In sum, it seems that in the most recent policy documents there is a reticence against explicitly emphasising the importance of being able to effectively write in accordance with the standard norm. This could perhaps be due to the lobbying and/or influence of linguists. However, it remains an open question whether such unclear legislative policies with respect to the standard norm will benefit students and have an effect on linguistic usage and tolerance of non-standard forms outside educational contexts. Bourdieu (1991; see also Faraco, 2017b, p. 85) noted how the imposition of French in France was successful not due to changes in official policy but to structural changes in the socio-economic makeup of society and how the French language correlated with material resources and social advancement. These structural and material factors are of crucial importance and can have significant effects on language use and language ideologies since social structures often inform and shape beliefs. The following question therefore needs to be posed: in Brazil if educational success, as determined by passing assessments, and social advancement, loosely defined as the ability to secure a job that confers prestige and guarantees financial security, are dependent on the active use of either the standard norm or adapting it towards alignment with the educated norm, then should not the school system ensure that students possess this skill instead of merely having knowledge of it and being able to recognise texts written in the standard? More research is necessary to establish whether this skill is being effectively taught. Its acquisition represents a gargantuan task for the child/student due to the linguistic distance between the outdated and unclear norms and the actual speech practices of Brazilians. In this article I have argued that these factors represent important structural problems for language and education in Brazil.

I would like to conclude by affirming that I am in no way against the struggle to change the prejudicial opinions towards non-standard forms of Portuguese and their speakers. I have myself contributed to this endeavour, suggesting that linguistic prejudice was contributing to school drop-out rates (O’Neill & Massini-Cagliari, 2019) and bemoaning the lost opportunity during the modernist Anthropophagic movement for a reconfiguration of opinions on Brazilian Portuguese in general (O’Neill, 2020). Cultural-oriented efforts aimed at changing opinions and recognising non-standard varieties as legitimate forms of the language are essential but in order to be effective they need to (a) appreciate how these opinions are related to deeper societal concerns and (b) be combined with efforts to change structures and offer different models since these undergird much of the social and economic dynamics which determine and perpetuate value judgements about and hierarchies of different varieties of a language.

The task of creating alternative codified regional norms with their concomitant official pedagogical materials is both a linguistic and political task and one which is daunting and challenging. Linguistically, this is due to the extent of linguistic variation attested and the work involved in discovering the systematicities. Politically, such an enterprise threatens the unity of the Portuguese language in Brazil and challenges the hegemony and tradition of the standard norm, which has fervent devotees and adherents in the media and the education system (Faraco, 2001; 2008; 2017b, p. 363). This standard norm, which one has to strive to learn, is related to concepts of civility, order, good conduct, and good management. The inability to master the standard or a disregard for the grammatical laws is related to concepts of barbarism, degeneracy, pandemonium, and bad conduct, placing the individual outside civilised society (O’Neill & Massini-Cagliari, 2019, pp. 42–43). Moreover, the grammatical standard forms the foundation of entire industries and professions and, more importantly, it is the institutionalised and legitimised standard of the country. It is understandable, therefore, why the general strategy in linguistics and education research has been to adopt cultural-oriented approaches. However, it is doubtful whether these alone will help resolve problems related to language and education in Brazil. The historical overview I have provided here shows how opinions and structures are often symbiotic and self-reinforcing and that social change is possible but usually when attention is paid to the ideological and the material, to opinions and robust and effective structures.