THE DRIFT/CONTACT DEBATE IN THE HISTORY OF BRAZILIAN PORTUGUESE

ABSTRACT:

This paper resumes the debate over the prevalence of the drift factor, considered as strictly internal, or the contact factor, considered as strictly external in the constitution of Brazilian Portuguese. On the empirical level, it focuses on the contact with Bantu languages. On the theoretical level, the resumption favors the epistemological bases of cognitive linguistics, where the adequacy of two new branches is under discussion: a cognitive contact linguistics and a cognitive sociolinguistics. This paper recapitulates the association between contact and creolization in order to dissociate them; it rejects the thesis that contact is restrict to lexical influence, with a view to associating the drift/contact dichotomy to the lexicon/grammar dichotomy; and it proposes a contact hypothesis based on domain-general processes. Reformulating the debate in these terms leads to the conclusion that contact motivates compatibilization among conceptualizations, for the matter being of cognition contact, not language contact.

KEYWORDS:
Cognitive Linguistics; Cognitive Contact Linguistics; Cognitive Sociolinguistics; Contact; Drift; Relativism; History of Brazilian Portuguese

RESUMO:

Este artigo retoma o debate acerca da prevalência do fator deriva, considerado como estritamente interno, ou do fator contato, considerado como estritamente externo, na constituição do português brasileiro. No plano empírico, a retomada privilegia o contato com línguas do ramo banto. No plano teórico, circunscreve-se em bases epistemológicas da linguística cognitiva, em que se discute a adequação de dois desdobramentos: uma linguística cognitiva de contato e uma sociolinguística cognitiva. O artigo recapitula a associação entre contato e crioulização, tendo em vista dissociar os dois fenômenos; rejeita a tese de que o contato se restringe à influência lexical, tendo em vista associar a dicotomia deriva/contato à dicotomia léxico/gramática; e propõe uma hipótese do contato baseada em processos de domínio geral. Reformular o debate nestes termos permite concluir que o contato motiva compatibilizações entre conceptualizações, por se tratar de cognições em contato, e não de línguas em contato.

PALAVRAS-CHAVE:
Linguística Cognitiva; Linguística Cognitiva de Contato; Sociolinguística Cognitiva; Contato; Deriva; Relativismo; História do Português Brasileiro

Introduction

The contact among Portuguese and languages from the Bantu branch of the Niger-Congo family (NURSE; PHILIPPSON, 2003; LUCCHESI, 2004) as constituent of the Brazilian Portuguese has attracted the interest of linguists and historians (FIORIN; PETTER, 2008; ALMEIDA, 2014, 2019; AVELAR; GALVES, 2014; NEGRÃO; VIOTTI, 2014; MAGALHÃES, 2018; AVELAR, 2019). Among the languages that belong to that branch, only a few were spoken by the enslaved peoples brought to Brazil (ALMEIDA, 2014, 2019; LUCCHESI, 2004; PESSOA DE CASTRO, 2012), which imposes the careful identification of which Bantu languages got in contact with that version of Portuguese along three centuries in different areas of what would come to be the Brazilian territory: Kikongo, Kimbundu, and Umbundu.¹

Such interest takes part of a long agenda of discussion about Brazilian Portuguese being as it is due to natural evolution from European Portuguese or contact with indigenous and African languages.

It is estimated that approximately 1.175 indigenous languages (RODRIGUES, 1993) were spoken in Brazil when the Portuguese arrived. Although they came to Brazil in 1500, the process of language transplantation really began as from the decade of 1530, when the king D. João III – called the colonizer for that very reason – draws a policy to occupy and administrate the new lands, with the division of the country into Hereditary Captaincies. Also in the decade of 1530, the traffic of slaves begins and brings speakers of around 200 to 300 languages (PETTER, 2006).

Today, after more than 500 years of many and diverse histories of language contact, there is no African language spoken as a native language by any Brazilian community, about 150 to 180 indigenous languages are spoken by approximately 260,000 Indians, and the vast majority of Brazilian population speaks a language that is the historical continuation of Portuguese, which Brazilian contemporary linguistics has denominated Brazilian Portuguese. (OLIVEIRA; LOBO, 2009, p. 7, our translation)²

Therefore, the contact with Bantu languages can be addressed ab ovo (considering the numerous and long historical studies and the equally numerous and long contact studies) or in medias res (conciliating contact as a historical phenomenon with a theoretical framework). The second option is chosen in this paper, where I recapitulate the drift/contact debate in order to approach it according to the foundations of cognitive linguistics, aligned with different studies that have developed a cognitive contact linguistics (ZENNER, 2013; NOËL, 2015; ZENNER; BACKUS; WINTER-FROEMEL, 2018).

The debate has become classic in the studies of Brazilian Portuguese history: the configuration of the Portuguese language in its Brazilian variety is due to “[…] a natural process of evolution by means of which internal mechanisms of the language motivate some changes; or to a rupture of the natural process caused by the contact of the language with other languages”³ (NEGRÃO; VIOTTI, 2012, p. 315, our translation)? It is usually associated with issues such as what, after all, Brazilian Portuguese is; the consideration of the language as Brazilian, not Portuguese; the homogenization of varieties of Brazilian Portuguese by the adjective “Brazilian” likewise the homogenization of varieties of European Portuguese by the adjective “European”, which could put in check labels like BP or EP; the relationship between variation and change; among others that lead to a vast bibliography.

Therefore, this paper is not dedicated to reviewing the history of Brazilian Portuguese in its distinctive properties from those of the European Portuguese or any other variety. It is dedicated to the re-examination of some theoretical assumptions about the history of Brazilian Portuguese, with a view to formulating a new pro-contact position that, based on the foundations of cognitive linguistics, dissolves the dichotomy between drift and contact and focuses on domain-general principles.

Foundations and procedures across the debate

The first foundation of cognitive linguistics I invoke is relativism. By doing so, I end up interweaving two debates: drift vs contact, and universalism vs relativism.

The former, at the interface between linguistics and history, is part of the agenda of scholars devoted to the history of Portuguese, a language that established contact with indigenous and African languages. Among many issues, scholars of indigenous languages tend to emphasize that Brazil has never been a monolingual country – an issue where the distinction between official language and natural language is crucial – and to identify the indigenous languages in risk of extinction, with a view to a linguistic policy that preserve those languages. This is one agenda of investigation. In turn, linguists who study African languages, in general, and Bantu languages, in particular, take on the challenge of identifying which languages came into contact with the Portuguese version of that time, which was European in Brazil, not yet Brazilian for being prior to contact, nor European as in Portugal, in an effort of periodization and categorization usually controversial. This is another agenda of investigation.

The latter, in the interface between linguistics and philosophy [of language] (MARTINS, 2011), concerns the relation between language and thought. Is thinking independent of language or constituted by it? Roughly speaking, the universalist thesis claims that thinking stands on its own, it can be overlaid with language for the sake of expression or not, whereas the relativist thesis claims that language is constituent of thinking, i.e., thinking does not stand on its own, there is only thinking through a natural language.

Langacker (1994) says that the issue behind the theses can be mistakenly formulated if the existence of the terms “language”, “thought”, “culture” and “cognition” is taken for the existence of separate, non-overlapping entities, besides not being empirical. These are the representations of the understanding the author rejects (Figure 1.a) and the understanding the author defends (Figure 1.b):

Figure 1
Relativism in cognitive grammar

Based on that version of relativism, I attribute part of the controversy concerning the drift/contact debate to the understanding that restrains the phenomenon to languages. According to that version, some aspects of language are cultural, some aspects are not, but the aspects that do not have a cultural origin may have a cultural manifestation. The encyclopaedic character of meaning, a foundation of cognitive linguistics, results precisely from the overlapping of language and culture as facets of cognition in one of its models, named cognitive grammar (LANGACKER, 1987, 1994, 2008, 2009).

In another model, cognitive construction grammar, Lakoff (1987, p. 334-335) adopts a more widespread version of relativism:

Am I a relativist? Well, I hold views that characterize one of the hundreds of forms of relativism.

[…]

As a linguist, I am interested in the grammar of languages, especially in what concepts are grammaticized in the languages of the world and what concepts are not. I am also interested in what it means for a concept to be grammaticized, and it is here that the use issue arises. Whorf was right in observing that concepts that have been made part of the grammar of a language are used in thought, not just as objects of thought, and that they are used spontaneously, automatically, unconsciously, and effortlessly. As a cognitive scientist, I am interested not only in what our concepts are but also in how they are used.

Such conception of relativism reminds the Wittgensteinian inspiration of cognitive linguistics, another controversial topic that is part of another agenda of investigation. I limit myself to tracking the origin of the position adopted by cognitive linguistics, without discussing the degree of fidelity of the theory to Wittgenstein (LAKOFF, 1987; MARTINS, 1999). The inspiration brought hereafter refers to the conception of a natural language as evidence of a form of life:

As regards linguistic practices, Wittgenstein embraces not a naturalist determinism, but a cultural relativism (e.g. MS 109 58), which follows from the conceptual relativism of the autonomy of language. The latter denies merely that our forms of representation are subject to metaphysical standards, a putative essence of reality, not that they may be subject to pragmatic standards. However, it is based on the idea that each form of representation lays down its own standards of rationality, which implies that even pragmatic justifications are internal to particular language-games. (GLOCK, 1996, p. 126)

Therefore, that is a version of relativism that “[…] stresses the intertwining of culture, world-view and language” (GLOCK, 1996, p. 124).

The version of relativism adopted by cognitive grammar and the version adopted by cognitive construction grammar share the conception of cognition as situated, which, in cognitive linguistics, associates the rationality inscribed in the grammar of a natural language to the perspective from which the speakers of that language conceive of the world:

Perspective, especially deixis, is perhaps the most obvious and most commented upon of the construal operations. Particularly for spatial descriptions, perspective is essential, and its dependence on the relative position and viewpoint of the speaker is well known. But perspective is also found in nonspatial domains: we have a perspective based on our knowledge, belief and attitudes as well as our spatiotemporal location. The closest cognitive property to perspective taken broadly is probably the philosophical notion of our situatedness in the world in a particular location – where location must be construed broadly to include temporal, epistemic and cultural context as well as spatial location. (CROFT; CRUSE, 2004, p. 58, emphasis in original).

Accordingly, to assume that “[…] meaning is perspectivist (does not objectively reflect the world, but models it, construes it some way or from a perspective and thus from different perspectives) […]”⁴ (SOARES DA SILVA, 2006, p. 7, our translation) presupposes the situated character of cognition, which, in turn, is associated with the distributed character of cognition, for “[…] it is important to highlight the agency of individuals in the process as a whole”⁵ (NEGRÃO; VIOTTI, 2012, p. 316, our translation). The situatedness and the distribution of the cognition of speakers of Brazilian Portuguese embed, as a consequence, beliefs and attitudes of speakers of those Bantu languages in this intersection between drift/contact debate and universalism/relativism debate, given the tension between what is universal and what is relative: “If language both shapes and reflects human experience, then language is as historical as that experience: while part of the human experience is universal and biologically species-specific, another part is historical and cultural” (GEERAERTS, 2011, p. 333).

This being a paper intended to discuss the constitution of Brazilian Portuguese as a result of contact with some Bantu languages, the position as for the former debate is expected to be pro-contact. This being a paper also committed with the foundations of cognitive linguistics, the position as for the latter debate is expected to be pro-relativism.

From those starting points, it would be natural to work based on them, describing some aspect of Brazilian Portuguese through contact in a relativist fashion. However, this paper aims to discuss the terms by means of which those starting points should be formulated. Regarding the former debate, my purpose is to dissolve the drift/contact dichotomy. With regard to the latter, my proposal is to no longer speak of language contact, but of cognition contact.

Drift vs contact, lexicon vs grammar, social vs structural

Avelar and Galves (2014, p. 242-243, our translation) point out the dichotomic mood of the traditional formulation of the drift/contact debate:

The debates over that issue have developed in a polarized fashion, with defenders of what we can call drift hypothesis and contact hypothesis in each pole – the former claim that all of the grammatical properties of Brazilian Portuguese were brought from Portugal, no new property is due to interlinguistic contacts; the latter claim that Brazilian Portuguese have grammatical properties that emerged as a consequence of contact, especially in the constitution of its popular varieties.⁶

Maintaining the polarization, this paper would follow contact hypothesis, as opposed to drift hypothesis, according to which “[…] in Brazilian Portuguese there is no specific grammatical influence of any African language”⁷ (NARO; SCHERRE, 2007, p. 182).⁸

In other words, for Naro and Scherre, the characteristics of contemporary Brazilian Portuguese are natural developments of archaic and classic Portuguese. For that reason, they cannot and must not be explained as a consequence of the contact of Portuguese with indigenous languages from Brazil or the languages brought by enslaved Africans, or as the result of a process of imperfect transmission in the acquisition of the language. Although the authors acknowledge the importance of the contact with Africans and Indians in the constitution of Brazilian culture, their position seems to be that three centuries of multilingualism have not significantly impacted the formation of the language currently spoken in Brazil.⁹ (NEGRÃO; VIOTTI, 2012, p. 320, our translation).

However, more than adopting one or another hypothesis, the theoretical framework that guides this paper should dissolve the polarization, which emerges along with others in a scenario it behoves me to trace back. The intended outcome is a contact hypothesis in which contact determines the path of drift. For that, let me recapitulate the history of the concepts involved.

Câmara Jr. (1986) attributes the term drift to Sapir, in avoidance of the teleological character of the term evolution:

EVOLUTION – Set of changes that a language undergoes in its internal history. The term was adopted by the 19^th century, following natural sciences, in which “evolution” means the gradual and continuous growth of an organism to the fullest. Many linguists reject or at least avoid the term because strictly there is no growth in languages, only changes, much less the march to fullness.

[…]

In this particular linguistic sense, evolution opposes to borrowing, a change caused by the adoption of elements originated in a different language.

[…]

Instead of the term, Sapir created another one – drift, which registers only the interconnection of changes towards a clear direction (Sapir, 1954, 165). (CÂMARA JR., 1986, p. 113-114, our translation)¹⁰.

In the definition above, it is possible to identify:

1. the opposition between the internal and the external triggering factor for language change, which presupposes that the factors cannot co-occur and, if the external prevails, it is restricted to lexicon, conceived of as separate from grammar; and

2. the presence of an author who influenced the history of linguistics as a defender of the relativist thesis, Edward Sapir, which, on its own, makes the pro-drift thesis compatible with relativism, a foundation assumed by theories recognized for continuities and also by theories recognized for discontinuities (structural linguistics, functional linguistics and cognitive linguistics).¹¹

Câmara Jr. (1986, p. 126, our translation) associates the concept of evolution to that of genius of the language in relativist terms:

GENIUS OF THE LANGUAGE – Traditional way (cf. Leoni, 1858) of designating the general characters of the grammar of a language in its formal and functional oppositions. From a diachronic outlook, it refers to a general sense of evolution, which Sapir called drift (Sapir, 1954). (CÂMARA Jr., 1986, p. 126).¹²

That concept is also present in the definition of word: “That is – as Sapir says – ‘a form that takes in as much or as little of the conceptual material of the whole thought as the genius of the language cares to allow’ (Sapir, 1921, 33)” (CÂMARA Jr., 1989, p. 93, our translation).¹³

Having recapitulated the relativist character of the concept of drift, the position defended by Naro and Scherre (2007) proves adequate to that foundation of functional linguistics. However, two other frameworks interfere in the drift/contact debate: sociolinguistics and generative linguistics.

Pagotto (2007) situates the drift/contact debate in a period of the studies of the Brazilian Portuguese history when the structural was contrasted with the social, the structural being of interest of generative linguistics, then guided by government and biding model (GALVES, 1995), and the social being of the interest of sociolinguistics, then guided by the theory of language change (WEINRICH; LABOV; HERZOG, 1968). The mindset behind the contrast gave rise to an object that blurred what was already at stake:

1. in the foreground, the contrast between two models of categorization – the Aristotelian, adopted by generative linguistics, and the radial (to which Labov contributed enormously), adopted by functional linguistics and cognitive linguistics:

   Much of the work on the part of the psychologist Eleanor Rosch on natural categories (2) suggests a prototype semantics. For the point I am making, the ‘naturalness’ of the categories is not so much the issue; but that helps. The prototype semantic notion I have in mind is this: I can know a square more or less directly; a trapezoid I know, however, in the first instance anyway, as a square that got distorted in a particular way.

   Related to these questions is what some researchers see as the problem of determining linguistic category boundaries. This work is exemplified in some recent studies of William Labov's (3). Knowing the category cup (as opposed to glass or bowl) is recognizing such properties as the ratio between the circumference of the opening and the height of the container, having one handle, being made of opaque vitreous material, being used for consumption of liquid food, being accompanied with a saucer, tapering, and being circular in cross-section.¹⁴ (FILLMORE, 1977, p. 56).

2. In the background, the contrast between two models of cognition – modular, adopted by generative linguistics, and connectionist, adopted by functional linguistics and cognitive linguistics:

   As opposed to the symbolic paradigm that (…) conceives cognitive processing as chains of symbolic representations linearly manipulated, connectionist paradigm claims that cognitive processing is based on clusters of units or nodes connected in dynamic networks in which units elicit or inhibit other units.¹⁵ (FRANÇA; FERRARI; MAIA, 2016, p. 179, our translation).

In other words, the label sociolinguistics stood out over the proposal of a model of categorization. This accident in the history of linguistics made it seem that variation is a social phenomenon despite its cognitive origin with a social manifestation. It also set aside that the putative structural/social opposition is the simplified and distorted version of the opposition between disembodied cognition and Aristotelian categorization, on one hand, and embodied cognition and radial categorization, on the other.

The “[…] foundation [of cognitive sociolinguistics] within the theoretical framework of Cognitive Linguistics […]”¹⁶ (SOARES DA SILVA, 2009, p. 192, our translation) seems to be able help to overcome that accident by “[…] showing how the cognitive and the social dimension of language are intrinsically related […]”¹⁷ (SOARES DA SILVA, 2009, p. 192, our translation), if only for the very project of cognitive linguistics as socio-cognitive linguistics (ALMEIDA, 2005; SALOMÃO, 2009). The point I insist to make is that, regardless of the adjective that goes with “sociolinguistics”, the commitment with radial model of categorization is constitutive, which proves wrong the reception that led to the separation between structural and social.

Due to that accident, the cognitive character of the drift/contact debate in its origin has been being neglected. The prevalence of drift would entail that “[…] Brazilian Portuguese would be just the natural and gradual continuation of a change process that is common to all Indo-European languages and echoes the morphological losses of Latin”¹⁸ (PAGOTTO, 2007, p. 462-463, our translation). Consequently, discussing the creole origin of Brazilian Portuguese would make sense within the limits of pidginization, creolization and dialectalization.

Creolization would point to the distance between European and Brazilian Portuguese, whereas decreolization to the proximity. To be or not a creole would be associated with having or not been influenced by the languages with which Brazilian Portuguese had had contact. That is why defenders of drift assume that contact restrains to lexical borrowings, it never reaches structural level, which, therefore, discards creolization and implies that the history of Brazilian Portuguese is different from that of European Portuguese, but not as a result of contact: “[…] the contact of Portuguese with several languages would have favoured the emergence of a dialectal variety, not a creole”¹⁹ (PAGOTTO, 2007, p. 465, our translation).

[…] every time the hypothesis of a creole origin for Brazil's Portuguese is discarded, the role that interlinguistic contact would have played in its constitution is indirectly discarded. It is for that reason that I think it is important to keep the issue alive, as a latent nuisance, a sting deep in the skin, lest studies of purely linguistic nature take the centre stage. Or, in other words, without interlinguistic contact in colonial Brazil and imperial Brazil, Brazilian Portuguese would not have the characteristics it has today. The latter formulation brings to the foreground the role that Amerindians and Africans played in the constitution process of Brazil's Portuguese, without falling in the trap of assuming it all as the result of a single pidgin and/or a single creole language.²⁰ (PAGOTTO, 2007, p. 467-468, our translation).

I consider the proposal against “studies of purely linguistic nature” taking the centre stage as an invitation to contemplate studies of historical nature (SLENES, 1992, 1995, 2018; ALMEIDA, 2014, 2019). Acknowledging that “Brazilian Portuguese would not have the characteristics it has today” makes it inevitable, for the purposes of this paper, to recapitulate the drift/contact debate, not for the sake of replacing drift hypothesis by contact hypothesis, but for the sake of:

1. re-establishing the original affinity between the concept of drift and relativism, in order to clarify why the drift hypothesis was defended by functional linguistics;

2. identifying the participation of the creolization factor in the debate as alien to the foundations of the theories involved but related to the procedure of separating internal and external history;

3. harmonizing drift and contact so conceived with the conception of language and culture as facets of cognition; and

4. understanding and overcoming the mistake of opposing social and structural, either through a cognitive contact linguistics, or through the recovery of the original connection between sociolinguistics and radial model of categorization.

Each of these incidental steps would be the subject of a paper. The step most eligible to contribute to the purposes of this paper allows me to discard creolization without discarding the role of contact, in the terms by Pagotto (2007) above. The series of studies that rejects the dichotomy between internal and external history (NEGRÃO; VIOTTI, 2012, 2014; VIOTTI, 2013) is aligned with the series of studies that denies to creolization the status of a linguistic phenomenon (MUFWENE, 2000, 2007, 2010, 2015).

Mufwene (2015) includes in his rejection of creolization as a linguistic phenomenon the rejection of the radial categorization of properties supposed to characterize a creole, which are precisely those taken into account by the scholars who defended contact hypothesis associated with creolization. Obviously, rejecting the phenomenon presupposes rejecting any approach to the phenomenon.

By different paths, this paper converges with the author's position: for him, what literature has conventionally called creolization is not a linguistic phenomenon; in this paper, even assuming that creolization were a linguistic phenomenon, being a domain-specific process would make it less capable of explaining a supervening process such as language change by contact. The conclusion is: contact yes, creole no.

The domain-general processes proposed by cognitive grammar are: association, automatization, schematization, and categorization (LANGACKER, 2008). The only domain-general process proposed by cognitive construction grammar is categorization (LAKOFF, 1987). Detailing each one would drive me away from the purposes of this paper, where I call upon the foundations of cognitive linguistics in order to foment a cognitive contact linguistics that is able to formulate a contact hypothesis free of the polarization pointed out by Avelar and Galves (2014) or the association with creolization pointed out by Pagotto (2007).

Contact hypothesis: new affinity between universalists and relativists

A significant number of works has been pursuing to demonstrate that some syntactic properties of Brazilian Portuguese (FIORIN; PETTER, 2008; OLIVEIRA; LOBO, 2009; OLIVEIRA; SOLEDADE; SANTOS, 2009; AVELAR; GALVES, 2014; AVELAR, 2019), including the most surprising in what comes to argument structure, like left-dislocation, are present in the Bantu languages spoken by the slaves brought to Brazil.

Incidentally, left-dislocation in Kimbundu has long been identified by Givón (2001) as the origin of the passive construction in that language, with a view to characterizing his proposal of a functional approach to language and a typological approach to grammar:

L-dislocation plus impersonal ‘they’

Finally, in some languages, such as Kimbundu, a passive construction arose diachronically from, and still resembles structurally, a left-dislocation construction plus the impersonal pronoun ‘they’, as in (Givón 1995: Ch. 3):

• (24)

  1. Passive

     Nzua a-mu-mono kwa meme

     John they-him-saw by me

     ‘John was seen by me’

     (lit.: ‘John, they saw him by me’)

  2. L-dislocation

     Nzua, aana a-mu-mono

     John children they-him-saw

     ‘John, the children saw him’

  3. Impersonal-subject with L-dislocation

     Nzua, a-mu-mono

     John they-him-saw

     ‘John, they saw him’ (anaphoric/active)

     ‘John, he was seen’ (impersonal/passive)

(GIVÓN, 2001, p. 22).

This fact in the history of Kimbundu leads straightforwardly to the drift/contact debate in the terms herein stated for:

1. contrasting the convergence regarding the procedures of adopting the radial model of categorization and the diachronic method of analysis with the divergence of separating grammar and lexicon between functional and cognitive linguistics;

2. presenting a case of drift in a Bantu language (the debate, in its traditional formulation, does not take into account what drifts and contacts may have resulted in the varieties of the languages that came into contact with the Portuguese that would come to be Brazilian, being those languages idealized as exempt from the same factors considered for the description of Brazilian Portuguese);

3. involving the Bantu language that, by means of contact, may have introduced left-dislocation in the Portuguese that would come to be Brazilian; and

4. eliciting the discussion of how a universalist theory can conciliate the explanatory power provided by the notion of contact with the explanatory power provided by notion of linguistic universals as from the identification of a share property, vis-à-vis a relativist theory, for which “(…) each language has an internal private form. It distinguishes itself from the others through the set of categories it comprises and the peculiar features that grade each one of the categories apparently common to another language or others”²¹ (CÂMARA Jr., 1989, p. 126, our translation).

In other words, it is not necessary to go as far as left-dislocation. Based on that version of relativism, predominant in the history of linguistics, grammatical categories are apparently common for serving the “[…] worldview (al. Weltansicht) embodied therein […]”²² (CÂMARA Jr., 1989, p. 116, our translation). According to the cognitive understanding defended in this paper, apparently common for serving different cognitions, different forms of life.

Maintaining the focus on constructions such as left-dislocation in Kimbundu and Brazilian Portuguese, the commitment with any relativist theory inheres in investigating that construction in each language as private and tracking it diachronically. Consequently, the evolution of left-dislocation into the passive construction in Kimbundu makes the passive construction of that language incommensurable with the passive construction of any other language and illustrates a case of drift, confirming the original affinity between drift and relativism.

The tendency to describe structures of Brazilian Portuguese based on contact with Bantu languages overcomes the original mistake of regarding contact as restricted to borrowings. Nevertheless, the commitment with classical model of categorization, modular model of cognition and universalism leads to an investigation about the role of contact that allows the identification of shared products. The question that arises is whether Brazilian Portuguese and those Bantu languages could not have some grammatical properties in common despite contact, exactly as more than one language has a passive construction. Universalism ends up weakening the explanatory power of the description.

For Avelar and Galves (2014, p. 243-244, our translation):

[…] certain singularising grammatical features of Brazilian Portuguese in the group of romance languages are due to the effect of interlinguistic contact between speakers of Portuguese and speakers of African languages (especially, Bantu languages). What we are calling effect outlines, in general, two different situations: (a) transfer of sentence patterns and morphosyntactic properties from the African languages to the Portuguese acquired as a second language by the Africans, and (b) properties triggered by Africans’ difficulty in reproducing the features of the target-language (Portuguese, in this case) during acquisition.²³

We agree in attributing certain singularizing features of Brazilian Portuguese to contact between speakers of Portuguese and speakers of Bantu languages. I just emphasize that speakers are cognoscent subjects, conceptualizers. The tendency to not fill in the internal argument (null object) can be considered one of these features (BEARTH, 2003).

Incidentally, the very influence of contact in the lexicon is one of the causes of those features, since overcoming the dichotomy between lexicon and grammar does not imply shifting the interest from lexicon to grammar (MAGALHÃES, 2018). What it implies is a radical revision of lexical influence: from marginal because lexicon would be external to grammar to central because lexicon and grammar are integrated by the submission to the same domain-general processes.

The implications of contact in “[…] the acquisition of Portuguese as a second language in a context of domination […]”²⁴ (PAGOTTO, 2007, p. 476, our translation) are at the core of our agreement, but, instead of calling my attention to shared products, they point to the cognitive conditions for second language acquisition in different types of colonization (MUFWENE, 2010). Consequently, my relativist approach privileges an agenda of investigation not satisfied by the arduous but insufficient identification of shared grammatical properties, an agenda that takes variation for a cognitive phenomenon and devotes to identifying conceptualizations (FABISZAK; HEBDA, 2011), perspectivizations (BATORÉO; FERRARI, 2015; MAGALHÃES, 2018), categorizations (LAKOFF, 1987; LANGACKER, 1987, 2008; BYBEE, 2010), and the very conditions of domination with the interdisciplinary support of history (SLENES, 1992, 1995, 2018; ALMEIDA, 2014, 2019).

However, we disagree on the reach of a universalist approach to contact, either because the theoretical framework, generative linguistics, claims that different languages can fix the same value for the same parameter, or because historical investigation (ALMEIDA, 2014) demonstrates that some contacts have never happened, as the supposed one between Zulu and Portuguese. The fact that the contact has never happened is, in turn, a cognitive evidence: for universalists, it is an evidence that principles and parameters are universal, that each language is a cluster of different values, that two languages that have never came into contact can exhibit the same value for a given parameter; for relativists, that each language comprises a worldview, that two languages that have never came into contact can exhibit apparently common constructions, that the so-called “transfer of sentence patterns and morphosyntactic properties” results from the match of conceptualizations, once, as emphasized by Mufwene (2010, p. 359), “transfer is not ineluctable”.

Therefore, the relativist framework both preserves that different languages share identical properties with or without contact and predicts that the contact between speakers of Bantu languages that arrived to Brazil as slaves and speakers of that variety of Portuguese in Brazil promoted the approximation of forms of life, capable to motivate linguistic influence, not to preview. That is why I understand that what comes into contact are not languages, but cognitions.

Câmara Jr. (1989, p. 102, our translation), in the classical discussion about prefixation being part of derivation or compounding, presents a case of inflectional prefix:

In other languages, the prefix is the usual type of inflection affix. Thus, in the African languages of Bantu group, it is through prefixes that the noun class (cf. § 65), and concomitantly singular or plural, is marked. See the very word bantu “men”, where –ntu is the semanteme (which never occurs isolated) and ba- is the pl. pref. of animated class, in opposition to the sing. pref. mu- in muntu “a man” (cf. § 65).²⁵

The case serves as an excellent counterexample to the supposed contact with Zulu. In spite of the contact between Portuguese and Bantu languages, what would come to be what is currently taken for Brazilian Portuguese has no inflectional prefix. That structural data, however, does not lead to denying the influence of contact on grammar.

Indeed, transfer is not ineluctable. In the terms proposed in this paper, it is licenced by the compatibilization among conceptualizations distributed in the cognitions in contact.

Bybee (2010), in turn, addresses the fact that languages that do not belong to the same family exhibit the same pattern of grammaticalization in which a future construction (time) arises from a movement construction (space). The author recapitulates her 1994 work, where she analysed a sample of 76 languages. The grammaticalization steps revised in light of domain-general processes permit to understand similarities among unrelated languages with no reference to contact: according to functional linguistics, by the attention to discourse-text universals (KATO, 1989) or “social factors that shape grammar” (BYBEE, 2010); according to cognitive linguistics, by the conciliation between the universal character of cognitive processes and the relativism characterized as the overlapping of language, culture, cognition, and thought above (LANGACKER, 1994; CASASANTO, 2013).

The centrality of conceptualization in cognitive linguistics is due to the conception of meaning, not only as encyclopaedic, but also as perspectivist, as highlighted above.

Though it is a mental phenomenon, conceptualization is grounded in physical reality: it consists in activity of the brain, which functions as an integral part of the body, which functions as an integral part of the world. Linguistic meanings are also grounded in social interaction, being negotiated by interlocutors based on mutual assessment of their knowledge, thoughts, and intentions. As a target of analysis, conceptualization is elusive and challenging, but it is not mysterious or beyond the scope of scientific inquiry. (LANGACKER, 2008, p. 4).

If the negotiation among interlocutors leads to the distributed dimension of cognition as speakers of the same language, it becomes even more important among interlocutors who conceive of the world differently for they speak different languages. It is in this sense that dissolving drift/contact dichotomy opens the way for scientific inquiry on conceptualization that can characterize a cognitive contact linguistics, devoted exactly to the exploration of contact among cognitions, not in socially symmetric interaction synchronically observable, but in asymmetric social interaction diachronically reconstructable from scarce data.

As for categorization, since the description of Dyirbal in Lakoff (1987), it reveals that women, fire and dangerous things can be together despite no common property, as opposed to what classical model states. When the author contrasts younger and older speakers of that language, he recapitulates chaining and experiential domains, two of the organizing principles of radial categorization in cognitive construction grammar, as the ones that establish new cognitive relations and the formation of categories. In these terms, language change is an epiphenomenon in relation to the change of principles that organize categorization.

In this paper, such principles gain great importance for allowing a cognitively motivated definition of drift, since Lakoff (1987) draws that case of change and ulterior demise of Dyirbal as the result of loss of chaining links from one generation to another with no participation of contact. Revised from the angle proposed in this paper, that description could be situated in the realm of drift as governed by a domain-general process.

Again, as in the evolution from L-dislocation into passive in Kimbundu described by Givón (2001), drift without contact. Or without the identification of contact? Very different was the history of Brazilian Portuguese.

Perspectivization, eventually, allows to track the beliefs and attitudes of the cognitions that became compatible in the configuration of Brazilian Portuguese. Cognitive linguistics regards perspectivization as a prerequisite for conceptualization, which leads back to the first item in the agenda of investigation proposed here.

Final remarks

This paper aimed to demonstrate that drift and contact necessarily interweave when contact happens, once abandoned dichotomies shared even by theories that refute one another, like that between lexicon and grammar, internal history and external history. For that, it identified in the debate about the role of drift or contact in the history of Brazilian Portuguese assumptions that deserve reflection.

One assumption was that acknowledging the role of contact would entail acknowledging the occurrence of creolization. Following the literature that rejects the dichotomy between internal and external history and that which defines the foundations of cognitive linguistics, I defend the dissociation between contact and creolization, be it for creolization not qualifying as a linguistic process, be it for, at best, qualifying as a domain-specific process.

Another assumption was that contact would exert only lexical influence. That assumption, in turn, embeds another one: the separation of lexicon and grammar. Once rejected lexicon/grammar dichotomy, the influence of contact extends to the language as a whole, it is not restricted to one of its components, and lexical influence gains a different epistemological status.

Because the influence of contact has been being acknowledged by studies committed to the classical model of categorization, this paper had to address universalism, on which that model lies, in contrast with relativism, on which radial model is based. That is because the convergence as for the influence of contact on grammar does not suffice to explain why languages that came into contact do not transfer certain morphosyntactic patterns or why the same morphosyntactic patterns exist in languages that have not came into contact.

The next step was to propose that contact involves cognitions, not languages, which requires moving from the realm of products (the languages) to the realm of processes, especially the domain-general ones. Consequently, it also requires identifying the domain-general processes in each model of cognitive linguistics and test their explanatory power in an agenda of investigation.

The latent patterns inherited by Portuguese in its previous configuration must have been influenced by the successive contacts promoted by the traffic of enslaved Africans for centuries. Therefore, neither contact nor slavery can be treated as homogeneous events. The acknowledged need of historical support will require refinements and delimitations that shed more light on a phenomenon that developed from 16^th to 19^th century with specific characteristics in each moment, including the population amounts shifted, the arrival points in Brazilian territory and the languages spoken by each amount.

This paper takes part of an agenda that addresses these issues from the viewpoint of linguistics, in general, and cognitive linguistics, in particular. The theoretical framework has, as its bases, the commitment with dissolving dichotomies created under different assumptions, in an arrangement of cards where pushing one is enough to drop all the others. Here, the cards named universalism, Aristotelian categorization, and modularity are replaced by the cards named relativism (predominant in linguistics before cognitive linguistics), radial categorization (obscured by the label sociolinguístics and shared by different theories), and connectionism, in order to drop another dichotomy incompatible with the theoretical framework: the one that separates, instead of harmonizing, drift and contact.

Some clarifications were necessary. First, this is about dissolving drift/contact dichotomy, not endorsing the relevance of contact on the polarized assumption. Second, the very definition of drift has a relativist character. Third, the model of cognitive linguistics named cognitive construction grammar focuses on principles that organize radial categorization and applies it to language change, addressed as what can be defined as the result of drift, while the model called cognitive grammar considers categorization as one of the domain-general processes. By doing so, the theory articulates the principles that organize categorization with domain-general processes, one of which is categorization.

This process, conceived as radial by all cognitive linguistics’ models, and the other domain-general processes seem to be the key to a relativist approach of the relations among language, culture and cognition. It is based on this key that a cognitive contact linguistics and a cognitive sociolinguistics seem to be viable.

Acknowledgements

I especially thank Maria Lucia Leitão de Almeida (Universidade Federal do Rio de Janeiro) for provoking me to think about contact and the wonderful conversations about the subject; Marcos Leitão de Almeida (Northwestern University) for carefully reading the first and the almost latest version of this paper, making suggestions, and sharing an original paper; and Anderson Salvaterra Magalhães (Universidade Federal de São Paulo) for the equally careful readings and suggestions.

• 1
  The frequent mention of these three languages is due to statistical relevance, but the number of languages spoken by the enslaved peoples is impossible to track (OLIVEIRA; LOBO, 2009).
• 2
  Original: “Calcula-se que, quando da chegada dos portugueses ao Brasil, aproximadamente 1.175 línguas (RODRIGUES, 1993) seriam faladas pela população indígena. Embora tenham aportado no Brasil em 1500, o início do processo de transplantação da sua língua ocorrerá, sensivelmente, a partir da década de 1530, quando o rei D. João III – por isso mesmo chamado de o colonizador – traça, com a divisão do país em Capitanias Hereditárias, uma política para povoar e administrar as novas terras. Também na década de 1530, dá-se início ao tráfico de escravos que para aqui trará falantes de, aproximadamente, 200 a 300 línguas (PETTER, 2006). Passados, hoje, mais de 500 anos de muitas e diversas histórias de contato lingüístico, não há língua africana sendo falada como nativa por nenhuma comunidade lingüística brasileira, sobrevivem cerca de 150 a 180 línguas indígenas, faladas por uma população de aproximadamente 260.000 índios, e é a esmagadora maioria da população brasileira falante nativa de uma língua que é continuadora histórica da língua portuguesa e a que a lingüística brasileira contemporânea tem designado de português brasileiro” (OLIVEIRA; LOBO, 2009, p. 7).
• 3
  Original: “[…] um processo natural de evolução pelo qual mecanismos internos à língua motivam algumas mudanças; ou uma ruptura do processo natural, ocasionada pelo contato de uma língua com outras línguas” (NEGRÃO; VIOTTI, 2012, p. 315).
• 4
  Original: “[…] o significado é perspectivista (não reflecte objectivamente o mundo, mas modela-o, constrói-o de determinada maneira ou perspectiva e, assim, de muitas perspectivas diferentes) […]”. (SOARES DA SILVA, 2006, p. 7).
• 5
  Original: “[…] é importante realçar a agência dos indivíduos no processo como um todo” (NEGRÃO; VIOTTI, 2012, p. 316).
• 6
  Original: “Os debates em torno dessa questão têm se dado de forma polarizada, com defensores do que podemos chamar de hipótese da deriva e hipótese do contato ocupando cada um dos lados da polarização – os primeiros argumentam em favor de que todas as marcas gramaticais do português brasileiro foram trazidas de Portugal, não havendo nenhuma propriedade inovadora devida à ação de contatos interlinguísticos; os segundos defendem que o português brasileiro apresenta características gramaticais que emergiram como consequência do contato, em particular no que diz respeito à constituição das suas variedades populares” (AVELAR; GALVES, 2014, p. 242-243).
• 7
  Original: “[…] no português do Brasil inexiste influência gramatical específica de qualquer língua africana” (NARO; SCHERRE, 2007, p. 182).
• 8
  This position has always been present in Brazilian lexicological tradition: “[…] borrowing, despite being considered in terms of a high dimension, only occurs, in fact, in the realm of lexicon or vocabulary. Thus, there seems to be no languages that receive from others phonological or morphological features” (CARDOSO; CUNHA, 1978, p. 138).
• 9
  Original: “Em outras palavras, para Naro e Scherre, as características do português brasileiro contemporâneo são desenvolvimentos naturais do português arcaico e clássico. Por isso, elas não podem nem devem ser explicadas como fruto do contato do português com as línguas indígenas do Brasil ou com as línguas trazidas pelos escravos africanos, nem como o resultado de um processo de transmissão imperfeita no processo de aquisição de língua. Embora os autores admitam a importância do contato com os povos africanos e indígenas na constituição da cultura brasileira, seu entendimento parece ser o de que três séculos de multilinguismo não impactaram sobremaneira a formação da gramática da língua falada hoje no Brasil” (NEGRÃO; VIOTTI, 2012, p. 320).
• 10
  Original: “EVOLUÇÃO – Conjunto de mudanças que sofre uma língua em sua história interna. O nome foi adotado nos meados do séc. XIX, a exemplo das ciências naturais, onde “evolução” significa o crescimento gradual e paulatino de um organismo até atingir a plenitude. Muitos lingüistas rejeitam ou pelo menos evitam o termo, porque na língua não há a rigor um crescimento, mas apenas mudanças e, muito menos, a marcha para a plenitude. […] Neste sentido lingüístico, particular, a evolução se opõe ao empréstimo, que é uma mudança proveniente da adoção de elementos provenientes de outra língua distinta. […] Em lugar do termo, Sapir lançou outro – deriva (ing. drift), que assinala apenas o encadeamento das mudanças numa direção nítida (Sapir, 1954, 165)” (CÂMARA JR., 1986, p. 113-114).
• 11
  In Lemos de Souza (2016), I emphasize that epistemological continuities and discontinuities must be considered in foundations and procedures. The acknowledged affinity between functional linguistics and cognitive linguistics, for example, is confirmed by the assumption of relativism as a foundation but does not produce consensus on the role of contact. On the other hand, the acknowledged opposition between generative linguistics and cognitive linguistics in all of their foundations does not prevent the consensus on the role of contact, as detailed below.
• 12
  Original: “GÊNIO DA LÍNGUA – Maneira tradicional (cf. Leoni, 1858) de designar os caracteres gerais da gramática (v.) de uma língua nas suas oposições formais e funcionais (v. função). Do ponto de vista diacrônico, refere-se a um sentido geral da evolução (v.), a que Sapir denominou deriva (Sapir, 1954)” (CÂMARA Jr., 1986, p. 126).
• 13
  Original: “É – como diz Sapir – ‘uma forma que chama a si grande ou pequena parte do pensamento integral na medida em que o gênio da língua se compraz em o permitir’ (Sapir, 1921, 33).” (CÂMARA Jr., 1989, p. 93). I must say that the exact formulation by Sapir (1921, p. 33) is slightly different: “The word is merely a form, a definitely molded entity that takes in as much or as little of the conceptual material of the whole thought as the genius of the language cares to allow.”
• 14
  Footnotes (2) and (3) are, respectively, “ROSCH, E., ‘Natural categories’, Cognitive Psychology 4, (1973), pp. 328-350” and “LABOV, W., ‘The boundaries of words and their meanings’, in C-J; N. BAILEY and R. SHUY, (eds.), New Ways of Analyzing Variation in English, Georgetown University Press, 1973”.
• 15
  Original: “Ao contrário do paradigma simbólico, que […] concebe o processamento cognitivo como cadeias de representações simbólicas que são manipuladas linearmente, o paradigma conexionista postula que o processamento cognitivo se baseia em complexos de unidades ou nós conectados em redes dinâmicas em que umas unidades excitam ou inibem outras” (FRANÇA; FERRARI; MAIA, 2016, p. 179).
• 16
  Original: “[…] fundamentação [da sociolinguística cognitiva] no enquadramento teórico da Linguística Cognitiva […]” (SOARES DA SILVA, 2009, p. 192).
• 17
  Original: “[…] mostrar como intrinsecamente se relacionam as dimensões cognitiva e social da linguagem […]” (SOARES DA SILVA, 2009, p. 192).
• 18
  Original: “[…] o português brasileiro seria apenas a continuação natural e gradual do processo de mudança que é comum a todas as línguas indo-européias e que no caso do português encontra eco nas perdas morfológicas do latim” (PAGOTTO, 2007, p. 462-463).
• 19
  Original: “[…] o contato do português com diversas línguas teria propiciado o surgimento de uma variedade dialetal, não um crioulo” (PAGOTTO, 2007, p. 465).
• 20
  Original: “[…] todas as vezes que se descarta a hipótese de origem crioula para o português do Brasil, descarta-se, por tabela, o papel que o contato interlingüístico teria tido na sua constituição. É por esta razão que penso ser mais interessante deixar a questão viva, como um incômodo latente, uma farpa lá no fundo da carne, para que os estudos de natureza puramente lingüística não dominem exclusivamente a cena. Ou, dizendo de outra maneira, sem o contato interlingüístico no Brasil Colônia e no Brasil Império, o português brasileiro não teria as características que teria hoje. Esta última formulação traz para o plano central o papel que ameríndios e africanos tiveram no processo de formação do português do Brasil, sem que caiamos na armadilha de supô-lo todo como fruto de um único pidgin e/ou uma única língua crioula” (PAGOTTO, 2007, p. 467-468).
• 21
  Original: “[…] cada língua tem uma forma interna privativamente sua. Distingue-se das demais pelo grupo de categorias que encerra, e pelos traços peculiares que matizam cada uma das categorias aparentemente comuns a mais outra ou outras línguas” (CÂMARA Jr., 1989, p. 126).
• 22
  Original: “[…] visão do universo (al. Weltansicht) ali consubstanciada […]” (CÂMARA Jr., 1989, p. 116).
• 23
  Original: “[…] certas marcas gramaticais singularizadoras do português brasileiro no conjunto das línguas românicas se devem à ação dos contatos interlinguísticos estabelecidos entre falantes de português e de línguas africanas (em particular, línguas bantas). O que estamos chamando de ação indica, em linhas gerais, duas situações diferentes: (a) transferência de padrões frásicos e propriedades morfossintáticas das línguas africanas para o português adquirido como segunda língua pelos africanos e (b) propriedades desencadeadas por dificuldades em reproduzir, por parte dos africanos, as marcas da língua tomada como alvo (no caso, o português) quando da sua aquisição” (AVELAR; GALVES, 2014, p. 243-244).
• 24
  Original: “[…] a aquisição do português como segunda língua, em condições de dominação […]” (PAGOTTO, 2007, p. 476).
• 25
  Original: “Já noutras línguas o prefixo é o tipo usual de afixo flexional. Assim, nas línguas africanas do grupo bântu, é por prefixos que se assinala a classe do nome (cf. § 65) e concomitantemente o singular ou plural. Haja vista a própria palavra bantu ‘homens’, onde –ntu é o semantema (que nunca aparece isolado) e ba- o pref. pl. da classe pessoal, em oposição ao pref. sing. mu- de muntu ‘um homem’ (cf. § 65)” (CÂMARA JR., 1989, p. 102).

REFERÊNCIAS

• ALMEIDA, M. A. L de. African voices from the Congo coast: speech communities, meanings and the process of identification in the slave ship Jovem Maria (1850). Journal of African History, Cambridge, p. 167-189, 2019.
• ALMEIDA, M. A. L de. Vozes centro-africanas no Atlântico-Sul (1831-c.1850). In: LIMA, I. S.; CARMO, L. do (ed.). História social da língua nacional 2: diáspora africana. Rio de Janeiro: Nau Editora, 2014. p. 73-104.
• ALMEIDA, M. L. L de. A lingüística sociocognitiva e os anguladores: uma nova perspectiva para fenômenos tidos como marginais. In: HENRIQUES, C. C.; SIMÕES, D.; VALENTE, A. (org.). Língua portuguesa: reflexões sobre descrição, pesquisa e ensino. Rio de Janeiro: Europa, 2005. p. 56-67.
• AVELAR, J. Sobre o papel do contato linguístico nas origens do português brasileiro. In: GALVES, C.; KATO, M.; ROBERTS, I. (org.). Português brasileiro: uma segunda viagem diacrônica. Campinas: Ed. da UNICAMP, 2019. p. 57-92.
• AVELAR, J.; GALVES, C. O papel das línguas africanas na emergência da gramática do português brasileiro. Linguística: Revista da ALFAL, [s.l.], v. 30, n. 2, p. 241-288, 2014.
• BATORÉO, H.; FERRARI, L. Quão fundo é o fundo? Perspectivação no caso das expressões com “ao fundo” em português europeu e português do Brasil. Estudos Linguísticos, Lisboa, v. 10, 2015.
• BEARTH, T. Syntax. In: NURSE, D.; PHILIPSSON, G. (ed.). The Bantu languages. London: Routledge, 2003. p.121-142.
• BYBEE, J. Language, usage, and cognition. Cambridge: Cambridge University Press, 2010.
• CÂMARA Jr., J. M. Princípios de linguística geral. 7. ed. Rio de Janeiro: Padrão, 1989. Original de 1941.
• CÂMARA Jr., J. M. Dicionário de lingüística e gramática. 13. ed. Petrópolis: Vozes, 1986. Original de 1977.
• CARDOSO, W.; CUNHA, C. Estilística e gramática histórica: português através de textos. Rio de Janeiro: Tempo Brasileiro, 1978.
• CASASANTO, D. The role of language in metaphor. In: BORKENT, M.; DANCYGIER, B.; HINNELL, J. (ed.). Language and the creative mind: new directions in cognitive linguistics. Cambridge: Cambridge University Press, 2013. p. 3-18.
• CROFT, W.; CRUSE, A. Cognitive linguistics. Cambridge: Cambridge University Press, 2004.
• FABISZAK, M.; HEBDA, A. Cognitive historical approaches to emotions: pride. In: WINTERS, M.; TISSARI, H.; ALLAN, K. (ed.). Cognitive linguistics research. Berlin: Boston: Mouton De Gruter, 2011. p. 261-297.
• FILLMORE, C. Scenes-and-frames semantics. In: ZAMPOLLI, A. (ed.). Linguistic structures processing. Amsterdam: New York: Oxford: North-Holland Publishing Company, 1977. p. 55-81.
• FIORIN, J. L; PETTER, M. (org.). África no Brasil: a formação da língua portuguesa. São Paulo: Contexto, 2008.
• FRANÇA, A. I; FERRARI, L.; MAIA, M. A linguística no século XXI: convergências e divergências no estudo da linguagem. São Paulo: Contexto, 2016.
• GALVES, C. Princípios, parâmetros e aquisição da linguagem. Cadernos de Estudos Linguísticos, Campinas, n. 29, 1995.
• GEERAERTS, D. Prospects for the past: perspectives for cognitive diachronic semantics. In: WINTERS, M.; TISSARI, H.; ALLAN, K. (ed.). Cognitive linguistics research. Berlin: Boston: Mouton De Gruter, 2011. p. 333-356.
• GIVÓN, T. Syntax: an introduction. Amsterdam: Philadelphia: John Benjamins, 2001. v.1. Original de 1984.
• GLOCK, H-J. Dicionário Wittgenstein. Tradução: Helena Martins. Revisão técnica: Luiz Carlos Pereira. Rio de Janeiro: Jorge Zahar, 1998.
• KATO, M. Formas de funcionalismo na sintaxe. DELTA, São Paulo, v. 14, n. esp., 1998.
• LAKOFF, G. Women, fire, and dangerous things: what categories reveal about the mind. Chicago: The University of Chicago Press, 1987.
• LANGACKER, R. Investigations in cognitive grammar. Berlin: New York: Walter de Gruyter, 2009. (Cognitive linguistics research, 42).
• LANGACKER, R. Cognitive grammar: a basic introduction. Oxford: Nova York: Oxford University Press, 2008.
• LANGACKER, R. Culture, cognition, and grammar. In: PÜTZ, M. (ed.). Language contact and language conflict. Amsterdam: Philadelphia: John Benjamins, 1994. p. 25-54.
• LANGACKER, R. Foundations of cognitive grammar: theoretical prerequisites. Stanford: Stanford University Press, 1987. v.1.
• LEMOS DE SOUZA, J. Fundamentos epistemológicos da linguística cognitiva. In: ÁLVARO, P.; FERRARI, L. (org.). Linguística cognitiva: da linguagem aos bastidores da mente. Rio de Janeiro: Brasil Multicultural, 2016. p. 2-23.
• LUCCHESI, D. A língua mina-jeje no Brasil: um falar africano em Ouro Preto do século XVIII. DELTA, São Paulo, v. 20, n. 1, 2004.
• MAGALHÃES, A. Pejoração e constituição do léxico brasileiro: um estudo semântico acerca de bantuísmos na interface da análise dialógica do discurso e da linguística cognitiva. 142f. 2018. Relatório de pesquisa (Pós-Doutorado em Língua Portuguesa) -Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2018.
• MARTINS, H. Três caminhos na filosofia da linguagem. In: MUSSALIM, F.; BENTES; A. (org.). Introdução à linguística: fundamentos epistemológicos. 5. ed. São Paulo: Cortez, 2011. v.3. p. 439-473.
• MARTINS, H. Metáfora e polissemia no estudo das línguas do mundo: uma apresentação não-representacionista. Orientação: Margarida Basilio. 190f. 1999. Tese (Doutorado em Linguística) - Universidade Federal do Rio de Janeiro, Rio de Janeiro, 1999.
• MUFWENE, S. The emergence of creoles and language change. In: BONVILLAIN, N. (ed.). The Handbook of Linguistic Anthropology. London: Routledge, 2015. p. 345-368.
• MUFWENE, S. SLA and the emergence of creoles. Studies in Second Language Acquisition, Cambridge, v.32, p. 359-400, 2010.
• MUFWENE, S. Population movements and contacts in language evolution. Journal of Language Contact, Leiden, v.1, p. 63-91, 2007.
• MUFWENE, S. Creolization is a social, not a structural, process. In: NEUMANN-HOLZSCHUH, I.; SCHNEIDER, E. (ed.). Degrees of restructuring in creole languages. Amsterdam: Philadelphia: John Benjamins, 2000. p. 65-83.
• NARO, A.; SCHERRE, M. Origens do português brasileiro. São Paulo: Parábola, 2007.
• NEGRÃO, E.; VIOTTI, E. Contato entre quimbundo e português clássico: impactos na gramática de impessoalização do português brasileiro e angolano. Lingüística, Montevideo, v. 30, 2014.
• NEGRÃO, E.; VIOTTI, E. Em busca de uma história linguística. Revista de Estudos Linguísticos, Belo Horizonte, v. 20, n. 2, 2012.
• NOËL, D. Cognitive contact linguistics as an essential ingredient of diachronic construction grammar. In: INTERNATIONAL COGNITIVE LINGUISTICS CONFERENCE (ICLC-13), 13., Newcastle. Proceedings […], Newcastle: Northumbria University, 20-25 jul. 2015. Disponível em: https://core.ac.uk/download/pdf/38074502.pdf Acesso em: 23 jul. 2020.
  » https://core.ac.uk/download/pdf/38074502.pdf
• NURSE, D.; PHILIPPSON, G. (ed.). The Bantu languages. London: Routledge, 2003.
• OLIVEIRA, K.; LOBO, T. Introdução: ou sobre como a África, no Brasil, avista a escrita. In: LOBO, T.; OLIVEIRA, K. (org.). África à vista: dez estudos sobre o português escrito por africanos no Brasil do século XIX. Salvador: EdUFBA, 2009. p. 6-49.
• OLIVEIRA, K.; SOLEDADE, J.; SANTOS, V. de S. Concordância nominal. In: LOBO, T.; OLIVEIRA, K. (org.). África à vista: dez estudos sobre o português escrito por africanos no Brasil do século XIX. Salvador: EdUFBA, 2009. p. 255-316.
• PAGOTTO, E. Crioulo sim, crioulo não: uma agenda de problemas. In: CASTILHO, A. de; MORAIS, M. A.; LOPES, R.; CYRINO, S. (org.). Descrição, história e aquisição do português brasileiro. São Paulo: FAPESP; Campinas: Pontes, 2007. p. 461-482.
• PESSOA DE CASTRO, Y. Renato Mendonça e “A influência africana no português do Brasil”, um estudo pioneiro de africanias no português brasileiro. In: MENDONÇA, R. A influência africana no português do Brasil. Brasília: FUNAG, 2012. p. 15-28.
• SALOMÃO, M. Teorias da linguagem: a perspectiva sociocognitiva. In: MIRANDA, N.; SALOMÃO, M. (org.). Construções do português do Brasil: da gramática ao discurso. Belo Horizonte: Ed. da UFMG, 2009. p. 20-32.
• SAPIR, E. Language: an introduction to the study of speech. New York: Harcourt; Brace, 1921.
• SLENES, R. Metaphors to live by in the diaspora: conceptual tropes and ontological wordplay among Central Africans in the middle passage and beyond. In: ALBAUGH, E.; DE LUNA, K. (org.). Tracing language movement in Africa. New York: Oxford University Press, 2018. p. 343-363.
• SLENES, R. Malungu, Ngoma Vem!: África encoberta e descoberta no Brasil. Cadernos do Museu da Escravatura, Luanda, Angola, v. 1, p. 1-24, 1995.
• SLENES, R. Malungu, Ngoma Vem!: África coberta e descoberta no Brasil. Revista USP, São Paulo, v. 12, p. 48-67, 1992.
• SOARES DA SILVA, A. A sociolinguística cognitiva: razões e escopo de uma nova área de investigação linguística. Revista Portuguesa de Humanidades: Estudos Literários, Braga, v.13, p.191-212, 2009.
• SOARES DA SILVA, A. O mundo dos sentidos em português: polissemia, semântica e cognição. Coimbra: Almedina, 2006.
• VIOTTI, E. Mudança linguística. In: FIORIN, J. L. (org.). Linguística?: Que é isso? São Paulo: Contexto, 2013. p. 137-180.
• WEINRICH, U.; LABOV, W.; HERZOG, M. Empirical foundations for a theory of language change. In: LEHMANN, W.; MALKIEL, Y. (ed.). Directions for historical linguistics: a symposium. 3. ed. Austin: University of Texas Press, 1968. p. 95-195.
• ZENNER, E. Cognitive contact linguistics: the macro, meso and micro influence of English on Dutch. Orientação: Dirk Geeraerts. Co-orientação: Dirk Speelman. 420f. 2013. Tese (Doutorado em Linguística) - Katholieke Universiteit Leuven, Leuven, 2013.
• ZENNER, E.; BACKUS, A.; WINTER-FROEMEL, E. (ed.). Cognitive contact linguistics: placing usage, meaning and mind at the core of contact-induced variation and change. Berlin: Boston: De Gruyter Mouton, 2018.

Publication Dates

• Publication in this collection
  07 Sept 2020
• Date of issue
  2020

History

• Received
  21 July 2018
• Accepted
  11 May 2019