THE SPEAKERS’ PROFILE FROM THE VIEWPOINT OF THEIR SOCIAL NETWORKS AND THEIR ASSOCIATION WITH LINGUISTIC CHANGE

ABSTRACT

Milroy (1987 Bortoni-Ricardo (2005 2009 2011 Coelho et al. (2015)

Social Networks; Sociolinguistics; Variation and change

RESUMO

Este trabalho mostra, por meio do mapeamento da fala de adolescentes oriundos de um distrito rural da cidade de Londrina, no Paraná, a Análise das Redes Sociais de Interação (social networks) - ARS, como um método analítico apropriado para o reconhecimento da realidade sociolinguística, uma vez que atua como um caminho eficiente para a obtenção de respostas sobre o comportamento linguístico de falantes, o que não é possível utilizando apenas as variáveis mais tradicionais (sexo, idade, classe social, escolaridade etc.). Utilizando o programa Egonet, software cuja função é quantificar os contatos pessoais, por meio da identificação de conexões e pontos dentro de redes egocêntricas, são apresentados mapas individuais das redes dos informantes, como um modelo apropriado de reconhecimento das referências linguísticas individuais. Pelo viés da Sociolinguística, no que concerne ao estudos da segunda onda, tendo como suporte teórico Milroy (1987 [1980]), Bortoni-Ricardo (2005, 2009, 2011 [1985], 2014), Coelho et al. (2015), entre outros, as discussões com enfoque nos resultados retratam a importância do trabalho com a ARS como um eficiente instrumento para os estudos da linguagem em geral, e sua relevância como um fator norteador no entendimento da manutenção e da mudança de falares como os analisados nesta pesquisa.

Redes Sociais de Interação; Sociolinguística; Manutenção e Mudança linguística

Introduction

Ah, quando eu estou em algum lugar, assim, mais importante e com minha patroa, que ela fica pegano no meu pé, o trabalho dela, ela era professora lá, era professora da UEL, ela fica sempre me corrigindo ‘fala direito, fala direito’, aí, pra mim não ficá sendo chamada a atenção, daí eu fico meio que medino...² (Inf1)

Our language is our homeland, the Portuguese writer Fernando Pessoa once said. Our mark, our identity. By means of it, we set ourselves up like subjects, like persons. We interact, we empower, and, even, we are oppressive.

Within the social contract established in the linguistic community, each individual always uses the language in their own way. However, the interaction made real by their speech also suffers influence. “A man is known by the company he keeps.”

In the epigraph, there is an example of the speech of an eighteen-years-old teenager, a casual worker, resident in a District in the rural zone in Londrina, whose principal networks are job and family. In both networks, the most relevant interaction ties happen with individuals with higher literary literacy. Why would a girl resident in a rural district, whose job frequently requires popular literacy, show a speech different from what is expected to find in a rural zone, since it does not present the variables associated with this type of expression, like rhotacism or the change of -LH- for IOD, for example? In other words, why would her speech and that of other teenagers examined in this work, who even lives in rural areas, distance from the radicular variety towards an urban speaking? Would the answer be in their social networks? What reasons would some of them have for maintaining their linguistic expression close to a rural speaking? The influence of the parents? Of the neighbors? Networks whose speeches distance from an urban speaking?

Our aim in this paper is to answer those questions is. To do so, we investigated interactions on the basis of social networks in a corpus composed of the speech of two teenager’s groups: one with members who live in a central zone of a district in Londrina (ZCD), considered more urban, and the other with members who live in a rural zone (ZRD) in this same district, through the recordings of their answers to a sociolinguistics questionnaire.

Our informants are twenty-four high school students in the aforementioned district, with ages between fifteen and eighteen years, being twelve girls and twelve boys. Among these, six boys and six girls live in the ZCD; six boys and six girls live in the ZRD.

The option for adolescence is due to the fact that this is the transition period between the childhood and the adult age, therefore, a phase in which speakers have a tendency to be more sensitive to their social environment, since, according to Netto (1968), they usually form groups and their group of companions starts to have great influence on their behavior, which sends us to the importance of the social networks in which they are inserted as determinants of their linguistic expression structure.

As regards the method for the composition of the corpus, after the application of questionnaires, the questions of which were intended to discover their social, economic and cultural reality, to all the classes of the secondary education of the school in which the research happened, we selected twenty-four pupils who corresponded to previously stipulated criteria, like dwelling place, age, participation in the labor market, participation in extracurricular courses, engagement in religious or social activities and access to cultural goods like cinema, theater, reading, travels, etc.

Afterwards, these twenty-four students were examined again through questions aiming to obtain the most significant possible time of their vernacular speech, following the proposal by Labov (2008 [1972]) of directed interview. At the end, each one of the twenty-four interviews was transcribe in accordance with the key of transcription proposed by the project Vertentes of the UFBA, pointing out the incidence of the following variables: absence of verbal agreement in the 1st and 3rd persons plural; prevalence of rhotacism; prevalence of iotization, marks considered to be of a rural speaking for Amaral (1982 [1920]), corroborated by Bortoni-Ricardo (2011 [1985]).

As regards the analyses, we follow the methodology of the social networks (MILROY, 1987 [1980]), for we understand, with Battisti (2014, p.96), that “in the sociolinguistics research, the analysis of the social networks and the social practices can explain the role of the connections between people, of the highest or lowest degree of group cohesion, of peer pressure and the local identities in the variation and linguistic change.”

We present here a brief result of the investigation of daily connections lived by these teenagers, reporting their role for the maintenance of the rural variety, which hypothetically is the variety common to the first social network of the examined subjects – the family – or for the change, in favor of the prestige variety, proclaimed by the broadcasting organizations of the hegemonic culture (school, church, work, media etc.).

However, before presenting such results, we deal with the theories that guide the research and serve as a basis for the analyses and the conclusions of this work.

Social Networks

Although the local relationships are more frequent, and the knowledge between members of the same speech community is mutual, people move, join different movements, get involved in various undertakings, insert themselves into other groups and enter other societies in which the identity relations happen. In other words, there is a sensitive and tenacious bond between social relationships and linguistic variation and change.

It is possible to understand how the interactions occur and how they propagate the new or keep what is known. A sociolinguistic research, considering the examined object and the interests concerned, can observe the development of a language from three standpoints: speech communities, social networks or communities of practice. Although one may be chosen, the other standpoints are not necessarily disregarded, once each research has its own specificities. And “what links all these approaches is the focus on the language in its social context.”, as we see in Coelho et al. (2015, p.70).

The works based on the Variationist Sociolinguistics were structured by the concept of speech community proposed by Labov (2008 [1972], p.150), for whom “speech community is not defined by any marked agreement in the use of linguistic elements, so much as by participation in a set of shared norms” (LABOV, 2008 [1972], p.150).

In such a conception, it is the attitude shared by the speakers towards linguistic uses that defines a speech community, and the uniformity of the shared norms happens when the linguistic variable presents visible social marks, which means that the speakers are, very often, conscious of the linguistic uses that they do, besides being capable of making value judgements about them.

Such an understanding raised questions related to the evaluation of the variants, in particular regarding their operationalization, since it does not seem possible to determine a number of variable linguistic forms towards which the users of a language would have a uniform attitude, something which, according to Labov, would characterize a community of speakers.

Gumperz (1972), relativizing the generalization done by Labov, conceives speech community as a community of social bonds. He approaches the conception adopted in studies the methodology of which aims to clarify the bond existing between general sociolinguistics standards and local practices. The linguist proposes a new concept for the community of speakers, based on the social interactions between individuals and the connections established in these interactions, because, living in society, the individuals frequently adopt a type of conduct that is equivalent to that of the members of their interaction groups, including linguistics behavior.

Milroy’s (1987 [1980]) studies within this perspective, and her researches developed in three proletarian communities of Belfast, Ireland, introduced the study of Social Networks, besides establishing the connective elements of this methodology in the studies of Sociolinguistics.

Conceiving these questions of linguistic variation and change as a result of the interaction between speakers inserted in determined social and interactional contexts, the researcher uses the concept of Social Networks, a complementary approach that helps explaining the daily social mechanisms favorable to the linguistic maintenance or change, for it performs in a way that explains why subjects with close characteristics, as it is the case of the students investigated in this study – about the same age, schooling, social stratum and history of life, among other factors –, present outstanding differences in their oral expression.

Bortoni-Ricardo (2011, p.15) defines Social Networks as “a set of bonds of all types between the individuals in a group.” From the perspective of Severo (2007), they are a way to verify the existing mechanisms in the communities that facilitate or that difficult changes, and these mechanisms also serve to analyze the way in which the individuals use the linguistic variability resources available to them.

Severo (2007, p.5) explains that this is a notion that came from the Social Anthropology in the decades of 1960 and 1970 and brought to the studies on linguistic variation and change in order to “explain the relationship between patterns in the vernacular maintenance and patterns in linguistic change over the course of time”. Evans (2004) observes that there are two forms of looking at the social networks: from a first point of view that allows seeing them as a system of personal relations with effects on the individuals; from a second one that allows understanding them as relations used by the individuals to reach their goals. It is the first one that guides Sociolinguistics.

These networks represent the degrees of contact between individuals who associate with each other regularly, through properties like multiplexity, consistent with the network content, and density, related to the network structure.

In Milroy (1987 [1980], p.50), we read that “a network is said to be relatively dense if a large number of the persons to whom ego is linked are also linked to each other.” The larger the number of persons who know each other within a group, the higher the network density is, still informs Milroy (1987 [1980]), since, after they are closely related to each other in the network, the contact with the outside is minimized, as each of them has few possibilities to use his relations to contact persons and to be contacted by persons from outside of the limits of the group. In this concept, a social network in which few persons know each other mutually is a loose-knit network, which shows low-density.

Regarding the multiplexity, the author considers the characteristics of a ties within networks as significant as the networks themselves, given that, even being of the same order, some ties may be more subject to external influence than others.

Within this concept, Milroy (1987 [1980]) creates a dichotomy between multiplexity and uniplexity. A bond will be uniplex when a member of the network represents a single role regarding another member of the same network; if, for example, member A is member B’s “boss” only (regarding B as A’s employee). The degree of complexity here is low, since the social roles practiced by the same person are not diversified, due to the limited fields of social activities.

On the other hand, the bond will be multiplex if member A assumes a variety of roles, for example, being B’s boss, but at the same time, neighbor, uncle, evangelical preacher, and so on. Bortoni-Ricardo (2011 [1985]) stresses the importance of the social roles for the distinction between rural and urban speeches, given that, in rural towns, the individuals commonly play several roles in the same speech community, producing a linguistic dependence between the members of this community, while, at urban towns, the acquaintances can be different, each one having a well-marked role. The author (2011 [1985], p.94) explains: “while a high level of density in the relationships characterizes the urban setting, the village settings present a low level of this type of density.”

In her research studies in Belfast, Milroy (1987 [1980]) looks at the social networks structure as a matter of strong and weak ties. The establishment of strong ties happens through routine and continuous bonds, as the ones between relatives, neighbors, and friends, in a way that the high intimacy level secures a daily contact. Weak ties, however, result from different, not intense activities, and without extreme bonds.

These interaction spaces between the individuals are essential for understanding the change and the maintenance of linguistic standards, because they comprise mechanisms that can be tendentious not only to facilitate but also to difficult them. Being a dense or loose network incurs directly in the linguistic habits, especially in what concerns linguistic maintenance and innovation.

Once the concept of social networks and its peculiarities was established, we discuss its relevance as an analytic method.

The ARS as an analytic method

Initially used in the telecommunication and computer systems, the ARS seeks to establish an objective way of identifying connections (ties or relations) and points (nodes or actors) inside a determined system and, by that, to represent structural patterns of ties, which may present themselves as constant as well as totally unpredictable and nonlinear (FAZITO, 2002).

It is a methodology devoted to the study of the relationships between entities and objects of any nature, especially in respect of the complex problems, like linguistic maintenance and change, for example, and, therefore, liable for social studies.

According to Hanneman and Ridle (2005), the ARS is fundamentally characterized for its occupation with data that express the relations between different groups, causing a dislocation of the analytical focus. The traditional approach in humanity studies focus on the individual attributes, whereas the ARS looks at the relations established by these individuals with the other participants in a determined social context. Using the ARS as a method of analysis is favorable for the treatment of speech, because individuals are not treated as equals.

The networks can be quantified in different ways. For this study, we used the program Egonet, a free software available on the Internet, created by Christopher McCarty, University of Florida – USA, which quantifies the contacts of a person, through the identification of connections and points within the individual network, establishing the interaction structural patterns and, therefore, defining the constancy.

To obtain the maps of the informants’ networks, we applied a basic questionnaire with three direct questions: “Who are the five most important persons in your life? Who are the five persons with whom you talked the most along this week? With whom would you share a secret?”. After obtaining the respective names, we put them in the program in order to define who connects with whom in the network.

Each informant was invited to answer about the affinity they had with each member of their network. Exemplifying by INF1: she cited her husband, grandmother, employer, mother, and brother, respectively, as the members of her social network. Then, we questioned if her husband, first cited member, used to interact with her grandmother, employer, mother, and brother; and, subsequently, if her grandmother used to connect with her husband, employer, mother, and brother; so did we successively until all the twenty-four teenagers answered.

The result obtained allows us to characterize the networks of these informants, with regard to the density and complexity, as well as to the quality of their bonds, whether strong or weak, which is of fundamental importance for the identity speech’s maintenance or change in favor of another variety that they would like to achieve, like the change from rural variety to urban variety.

Furthermore, knowing these ties allows understanding how networks act when it comes to speakers’ conscious or unconscious desire to maintain or change their speech. In this regard, Bortoni-Ricardo (2014) asserts that “in communities of close-knit networks, in which basically all the individuals interact with each other, the norm pressure is stronger. In communities of loose knit networks, the norm pressure is weaker.”

As a result, we present maps prepared by the Egonet program and some considerations on their structure:

The teenager’s networks structure

The maps bring the structure of the teenager’s social networks separated by groups. The first group present maps of closed networks, and, sequentially, a second group presents the maps of open networks. At last, a third group brings intermediate networks.

Figure 1
– Maps with closed networks.

Figure 2
– Maps with open networks.

Figure 3
– Maps with intermediate networks.

INF24 map is the only one presenting two ties that do not bind. In this way, his network can be characterized as loose, considering that it does not present bond redundancy and there is more than one weak, uniplex tie that opens for outside contact. This can hypothetically be the cause of the approximation to the urban variety demonstrated by an expression without rural marks analyzed in this paper, because such networks favor the adhesion to the prestige culture and, therefore, the approximation to the urban variety.

Bortoni-Ricardo (2005, p.88) justifies such a posture by reporting that

When the individual manages to ascend socially, his network becomes more heterogeneous, and, consequently, of looser knit. The process of dialectic diffusion intensifies, and the speaker tends to approach the prestigious norm, acquiring a wider range of registers [...].

The networks of INF1, INF6, INF7, INF10, INF11, and INF23 hold at least one weak tie open to the outside, besides having no connection between all of the members. Then, they may also be qualified as loose networks, which consequently may suffer greatly from a possible external interference. We associate this sort of network with a linguistic rurban practice, in other words, it is possible that, in a final allocation of these informants in the urbanization continuum, they are situated on rurban points, nearer the rural extreme or the urban one, but not specifically on those extremes or poles.

Networks referring to job, always relevant, can have different characteristics, which affect the informant’s linguistic structure. For instance, we have INF7 and INF23. INF7, whose workplace is the very district, presenting aspects of rural variety, like rhotacism in “Tem que sê compreto?” [Does it have to be complete?], while INF23, who works at the biggest commercial center of Londrina, expresses quite close to the urban variety, even at the last moments of the interview, at which, relaxed, she was revealing her vernacular already, as in: “aí ela dexô um pedido que era pra cuidarmos do filho mais novo dela que é o que mora em Santa Catarina” [Then she left a request for us to take care of her younger son, the one who lives in Santa Catarina].

This discrepancy in the speech of informants who have the work-oriented interactions in their networks can be explained by combined factors. Still referring to INF7 and INF23, the workplace and the degree of literacy of those networks influence the speech’s change or maintenance. INF7 works at a market located in the urban area of the district, sticking prices to products and replenishing shelves. Even though her bosses have higher social status, they also live in the district, and their co-workers are all from there, and her access to literacy events permeated by the prestige variety is minimal. However, INF23 works at the administration of a shopping center. All her co-workers interact using the high prestige variety, if not, a rurban variety situated right near the tip of the urbanization continuum. These factors are predominant in the salient differences between INF7 and INF23 speeches.

INF1’s norm changes are motivated by her work. In her interview, she says that one of her employer, a retired university professor, forces her to speaking closer to the standard rule. In Bortoni-Ricardo (2011 [1985]), we see that it is possible to associate this fact with what Labov calls “status inconsistency”, since the informant, looking for the use of an urban speaking, is adopting the values of a group different from her own, which is her reference, and aiming for it, alternates between the rural and the urban, staying in the rurban, as seen in “Então, o que eu assisto mais é os canais pagos, né, eu gosto do History, eu gosto do de filmes, todos os canal de filme” [So, what I watch more is the pay-per-view, I like History, I like movies, all movie channel], an expression that introduces nonstandard verbal and nominal agreement. Due to this contact with different employers, INF1’s networks are loose and have uniplex ties.

INF6 has a strong connection with work, personified in the supervisor’s person. His network is loose, with uniplex ties in most interactions. INF7 is also enclosed in a loose network, since some of the ties have diffusion points.

As for INF2, INF9, INF12, INF8, INF16, and INF17, although most of their ties are connected, there are still disconnected points, without cohesion bonds, as we saw in Milroy (1987 [1980]), enabling a non-categorical external intervention. This opening qualifies their networks as relatively dense and their ties as relatively strong, which may be responsible for the rurban way of speaking, because sometimes the speech is closer to the urban variety, sometimes it is closer to the rural variety.

INF2, for instance, presents this alternation in “Ué, por causo que eu acho bem bonito aquelas praia, eu sempre tive vontade de conhecê-las.” [Well, cause I think very beautiful those beach, I’ve always wanted to meet them.]. The absence of formal agreement is a gradual¹ trait, according to Bortoni-Ricardo (2004), for taking place in the relaxed speech of most Brazilian speakers, but it is not socially accepted and contrasts with the pronominal placement guided by the standard in “conhecê-las” [to meet them].

In INF8’s speaking, we have “é ligaro né po um zero nove da polícia e falo que tinha que, é um bicho perigoso dentro da casa que ia pegá ele” [They called the police and said they had a dangerous animal in the house that would get him]. There is the verbal agreement in the first verb in the third person singular but occurs the desnazalization. Moreover, in the second verb, the agreement with the third person singular expected does not occur. This characteristic is also a rurban trait specific to peripheral areas of large cities, where the informant lived for a period before moving into the ZRD.

On the other hand, the networks of INF3, INF4, INF5, INF13, INF14, INF15, INF18, INF19, INF20, INF21, and INF22 are dense, since all their members interact with each other. As we saw in Milroy (1987 [1980]), the influence of external factors is hampered by strong ties, some of which have multiplex traits. All points are connected. Therefore, their networks are of close-knit. INF4, for example, has a bond characteristic of dense networks, which is the multiplexity of roles: his father is also his boss. While what INF20, INF21, and INF22 have in common is a family member playing more than one role in their interactions, which is a mark of close-knit networks.

After observing the informant’s networks structure, we will describe the linguistic aspects present in their interactions, adopting, with Amaral (1982 [1920]) and Bortoni-Ricardo (2011 [1985]), the variables: the absence of verbal agreement (CV) in 1st and 3rd persons singular, the iotization and the rhotacism as markers of the rural speech.

Interpreting the data

The linguistic aspects analyzed in this paper, as we have mentioned, are understood as markers of a rural variety, but are present in the speech of a large group of Brazilians, characterized as rurban speakers.

In this sense, Bortoni-Ricardo (2009, p.52) clarifies that the rurban speaking is characteristic among the migrants of rural origin, but also of “countryside communities resident in districts or semi-rural nuclei, which are subjected to urban influence [...]”, in other words, in communities like the ones analyzed in this paper.

To guide the analyses, it is worth noting that the twenty-four teenagers researched live in a rural locality, according to the concepts adopted by the IBGE (Brazilian Institute of Geography and Statistics), being half of them resident in the central part of the district and the other half living in villages and small farms in the region. They are all of the same age group and have the school as a social network and literacy agency. It is possible to consider that they are inserted into the same speech community, according to the conceptualization given by Gumperz (1972).

However, when we compared INF3 and INF24, for example, who are the same age and are in the same school year and live in ZCD, we noticed a striking difference in linguistic expression. Even INF15 and INF22, twin brothers, have differences in their language.

The configuration of their social networks of interaction can guide towards the understanding of such differences, since it can be decisive for the linguistic choices practiced by the speakers, either consciously, as a mark of identity or localism, or a non-identification with the local and group.

Thus, we have to talk about the presence or absence of the linguistic factors being analised in this article and their relationship with the configuration of the speakers’ networks. We begin with the absence of the verbal agreement.

Nonstandard verbal agreement in the 1st and 3rd persons plural

The non-standard verbal agreement (CV) is understood by Castilho (2010) as a morphological conformity between a class, represented by the verb, and its scope, represented by the subject. Thus, it is one of the markers of the socially disadvantaged speakers, and also of residents of rural areas, being one of the most stigmatized linguistic factors. This mark was productive in the speech of the different teenagers analyzed here.

The following table details these results:

Table 1 – Verbal agreement: frequency of uses.

CONTEXT	OCCURRENCES TOTAL	CANINICAL VERBAL AGREEMENT	NO VERBAL AGREEMENT
1ª PP	73	63	10
3ª PP	301	158	143
TOTAL	374	221	153

Source: Elaborated by the author.

Graphic 1
– Verbal agreement: frequency of uses.

Through the analysis of the table and the graphic, it is possible to realize that, in the 1st person plural context, the standard or canonical verbal agreement occurs more often. As for the 3rd person plural context, there is a higher leveling, despite the canonical verbal agreement happening more often, which breaks with the characterization of the rural variety, whose configuration, as shown by Amadeu Amaral (1982 [1920]) is formed by the non-realization of verbal agreement in the form of the normative grammar of the Portuguese language.

Thus, the structure of the speech reality of the researched teenagers, presented only by the realization or non-realization of the CV, shows a tendency for the diffusion of the speech which hypothetically represents the identity of these speakers, since, even with their incidence in the speech of all the informants analyzed, it does not occur in 100% of the contexts in which it appears.

Incidence of Iotization

Amaral (1982 [1920]) informs that the rural speaker tends to vocalize the “LH” into “I” in words like “espaiado, meió, muié, fio” [spread, better, woman, son], a fact also verified by Bortoni-Ricardo (2011 [1985]) between Braslândia’s speakers when she reports that the palatal lateral /λ/ is vocalized.

Bortoni-Ricardo (2011) and Aguilera (1999) associate this phenomenon with the rural variety. Aguilera (1999, p.158) claims that this process is “a predominant trait of the rural or low-prestige variety that expands throughout the Brazilian regions as the already published atlases can document”. For Bortoni-Ricardo (2011 [1985]), the rule is productive “in the rural and rurban varieties, functioning as a typical discontinuous trait.” The author recalls that, because it is one of the most socially stigmatized traits, it enters the concept of stereotype proposed by Labov (2008 [1972]).

The table and the graphic below illustrate the reality of iotization among the informants of this paper:

Table 2 – Incidence of iotization.

CONTEXT	OCCURRENCES	STANDARD	IOTIZATION
“Work” and verbal forms of “To work”	137	79,56%	20,44%
“Mirror” and verbal forms of “To mirror; to reflect”	8	50,00%	50,00%
“Old” and verbal forms of “To get old”	6	83,33%	16,67%
“Son”, “Daughter”, “Godchild” and other derived forms	35	71,43%	58,57%
“Better” and verbal forms of “To improve”	39	97,43%	2,56%
Verbal forms of “To disturb”	4	100,00%	0,00%
Verbal forms of “To spraid”	2	0,00%	100,00%
“Eye” and verbal forms of “To see”, “To look”	33	51,51%	48,49%
“Advice” and verbal forms of “To advide	2	100,00%	0,00%
“Family” and other derived forms	103	98,06%	1,97%
“Wonderful” and other derived forms	3	100,00%	0,00%
“Detail” and other derived	2	100,00%	0,00%
“Ear”	4	75,00%	25,00%
“Noise”	1	0,00%	100,00%
“Red”	1	0,00%	100,00%
“Playful”	1	100,00%	0,00%
“Knee”	1	100,00%	0,00%
TOTAL	382	82,43	17,54

Source: Elaborated by the author.

Graphic 2
– Incidence of iotization.

By the charts, it is possible to know the contexts favorable to the iotization present in the corpus and the number of its occurrence among the researched speakers. Overall, it is possible to notice a higher direction, in relation to this linguistic phenomenon, for the diffusion of the rural speech, even if there are words whose incidence would indicate a certain resistance to iotization. However, that factor needs further studies.

The incidence of rhotacism

The rhotacism has traditionally been described as the exchange of a lateral sound by a rhotic sound, so /l/ is converted into /r/. Amaral (1982 [1920]) names the rhotacism a pronunciation vice and classifies it as one of the most productive vices in the speech of speakers who was born in São Paulo, including those who had no direct contact with the rural speech.

Despite not using the same nomenclature, Bortoni-Ricardo (2011 [1972], p.76) addresses this case by pointing to “the neutralization of /r/ and /l/”, e.g. “incrusive – inclusive” [inclusive], or the total suppression of liquids, as in “oto – outro” [other]. We analyze the phenomenon of rhotacism in two contexts: i) in complex syllables, such as PL-, CL-, BL-; II) in internal coda, in words such as “alguma, cultura, folga”, [some, culture, respite] for example.

The table and graphics below show the results.

Table 3 – Incidence of rhotacism in internal coda.

CONTEXT	TOTAL	STANDARD	RHOTACISM
Something, someone	203	97%	3%
Lunch, to have lunch	4	100%	0%
Tall	2	100%	0%
Illiterate	4	100%	0%
Asphalt	3	100%	0%
Bag	2	100%	0%
Cacilda (woman’s name)	1	100%	0%
Calculation	1	100%	0%
Calm	3	100%	0%
Guilty	2	100%	0%
Cultured	1	100%	0%
Cultivate	1	100%	0%
Culture	1	100%	0%
Apologies	1	100%	0%
Difficulty	1	100%	0%
School or Faculty	1	100%	0%
Lack	10	100%	0%
Film	7	100%	0%
End	1	100%	0%
Respite	2	100%	0%
Modesty	1	100%	0%
Same	1	100%	0%
Judge	3	100%	0%
Fine	1	100%	0%
Multinational	1	100%	0%
Any	11	100%	0%
Solve	3	100%	0%
Salty	1	100%	0%
Jump	1	100%	0%
Save	1	100%	0%
Wild	1	100%	0%
Single	1	100%	0%
Last	6	83%	17%
Back	19	100%	0%
TOTAL		98%	2%

Source: Elaborated by the author.

Graphic 3
– Incidence of rhotacism in internal coda.

Table 4 – Incidence of rhotacism in complex syllables.

CONTEXT	TOTAL	STANDARD	RHOTACISM
Blocks	2	100,00%	0,00%
Clear	34	88,24%	12,00%
Complete	5	100,00%	0,00%
Pair	1	100,00%	0,00%
Example	7	57,14%	42,86%
Explain	3	66,66%	33,34%
Explosion	1	100,00%	0,00%
Flower	1	100,00%	0,00%
Implicit	1	100,00%	0,00%
Implosion	1	100,00%	0,00%
Influence	2	100,00%	0,00%
English	1	100,00%	0,00%
Plant	7	66,66%	33,34%
Plural	1	100,00%	0,00%
Problem	14	78,57%	21,43%
Public	2	100,00%	0,00%
Simple, Simply	3	66,66%	0,00%
Supplementary	1	100,00%	0,00%
Key, Keyboard	2	100,00%	0,00%
TOTAL		98,33%	1,64%

Source: Elaborated by the author.

Graphic 4
– Incidence of rhotacism in complex syllables.

The table, expressing numerically the incidences of rhotacism, and the graphics, through the expression of occurrence or not occurrence, show that such a phenomenon is not productive in the speech of the district teenagers, even if there is a greater realization in contexts of a complex syllable. Once again, through the global analysis, it is observed a tendency of their linguistic expression to an urban speaking, featuring a rurban speech.

The general observation of the data, therefore, indicates the diffusion of the rural variety, regardless of the configuration of the informant’s networks. A more individualized analysis may clarify this data better.

Considerations on the data in the charts

The teenagers inserted in closed networks not only presented a substantial incidence of the analyzed marks, but also some of them did not present any of them.

For example, the speeches of INF3 and INF13 are marked by discontinuous traits, such as the absence of CV in “Elas que faiz você vivê mais” (INF3) [They make you live longer] e or “eles aprende, lá no passado, por isso que eles até parô de estudá” (INF13) [They learn in the past, so they don’t study more]. Iotization appears all the times that the noun trabalho [work] or the forms of the verb trabalhar [to work] occur, as in: “não, minha mãe trabaia em casa só” [My mom works only at home] in the INF3 speech. INF13 presents, at the beginnings of almost all his answers, the discursive marker “Óia” (look), iotization of “Olha (look)”, however, iotization also occurs in other moments of his speech, as in “Vai lá e asséste tevezinha com a veinha” [Go on and watch TV with the old lady], and “trabaia como pedrero “ [he is a constructor]. Rhotacism also appears in the speech of both, as in “as pessoa num recrama, ah, às vezes recrama do lugar” (INF3) [People don’t complain, oh, sometimes complain of the place]; “é muito raro vê argum jovem sertanejo por ai né” (INF13) [It’s very rare to see some young country boy out there].

The absence of verbal agreement is typical to the informants INF18, INF19, INF20, INF21, and INF22, as it occurs in these excerpts: “Eles começa a contá” (INF18) [They start to count]; “as família gostava muito da amizade delas” (INF19) [The family likes their friendship]; “Sempre eles tá perto de mim” (INF20) [They are near me always]; “ Meu pai e minha mãe sabe escrevê e lê tamém” (INF21) [My parentes know to write and to read]; “tem ali no pasto lá, nói joga tamém” (INF22) [There is in the pasture, we play there].

Iotization is productive in the speeches of INF18, INF20, INF21 and INF22, as in: “Eis tiraro” (INF18) [They took] and “É, às veiz trabaio num lugar, às veiz no otro” (INF20) [Sometime I work in a place; somethimes in other place]; “que eu trabaiava lá, eu sempre ia” (INF21) [I worked there; I always went there]; “aham, sempre fala pra trabaiá assim” (INF22) [He always says to work]. The rhotacism marks the speech of INF19, as in: “O aluno tá resolveno um pobrema, um pobrema no quadro” [A student is making a problem in the blackboard], and of INF20 in: “ajudo no negócio de prantação, mais trabaio fixo eu num tenho.” [I help in a plantation, but I don’t have a fixed job]; INF20 presented these marks repeatedly.

The speech of the teenagers inserted in open networks is a little more homogeneous. Among the seven informants, only one presented all the rurality marks analyzed here.

The informants INF1, INF6, INF10, INF11, and INF23, however, present only the absence of CV mark, as in “eu assisto mais é os canais pagos” (INF1) [I wath the paid channels]; “elas era, duas amigas que era carne e unha” (INF6) [They were friends, they were as thick as thieves]; “essa aqui é treis criança jogano bola na praia” (INF10) [This picture is about children play soccer in a beach]; “São pessoas inteligente e que aprendero e tá usando o que aprendeu” (INF11) [They are intelligent people who have learned and are using that]; “daí eles num entraro num consenso” (INF23) [They didn’t reach consensus]. INF24 does not present any stigmatized marks.

In contrast, INF7 presents the rhotacism in her linguistic expression, in the first response to the questionnaire, when she was asked to write her name: “tem que sê compreto “ [It has to be complete]. Afterwards, the iotization and the absence of verbal agreement permeate the whole conversation, as in “por causo da minha fia né, eu tenho que dá o exempro, né?” [For my daughter I have to set the example]; “Eles sempre trabaiô muito né e num lê” [They always work a lot and they didn’t read].

There is an incidence of all the discontinuous linguistic factors observed in this work in the speech of two teenagers inserted in intermediate networks.

For INF2 and INF9, the verbal agreement does not happen most of the time when it is mandatory, in the first and third persons, as seen in “Nóis tem uma mercearia lá onde que eu moro mesmo, aí fica os dois lá” (INF2) [We have a grocery store where we live and they stay there]; “meus pais faiz serviços gerais, eles mexe com gado assim”(INF9) [My parents are farm employees]. In their expression, the rhotacism appears in: “Pele crara” (INF2) [clear skin]; “Eles tinha um probreminha”(INF9) [They have a little problem]. And the iotization in “na verdade, ele num ia trabaiá” (INF2) [Really, he doesn’t go to work] or “tem tanta coisa pra puxá a oreia desse povo” (INF9) [There a lot of things to get people attention].

INF8, INF12, INF16 and INF17 present only passages with the discontinuous trait concerning the absence of verbal agreement as in “mais agora, as otras que não, que sabem falá certo e fala assim, num sei, deveria falá do jeito que sabe né?” (INF8) [Who knows how to speak right and speaks wrong should speak the way they know]; “ele é empregado, mais eles se dá bem” (INF12) [He is employed but they are friends]; “das coisas qui tá acontecendo no mundo, no Brasil” (INF16) [Things that are happening in the world, in Brazil]; “as pessoas mais chegadas, entre amigos assim, não fala certo, fala mais na gíria né qui fala” (INF17) [The closest people, among friends, don’t speak right, speak more in slang].

These examples show a greater tendency to the rural variety diffusion in open networks. The teenagers enclosed in these networks expressed themselves, mostly without rhotacism and iotization, but they presented the absence of verbal agreement. However, one of them presents a linguistic configuration very close to the rural speech, denoting more focusing. In her case, the network configuration was not relevant.

Concerning the closed networks, there is a greater tendency towards focusing, observing that most informants bring the marks of rurality into their speech, which leads to conclude on the relevance of these networks for a relative maintenance of the rural variety. However, for four informants, that conclusion is not real.

At this point, it seems to be relevant to detail two concepts: focusing and diffusion. Le Page (1980) proposes the concept of diffusion as the result of the physical and social mobility of speakers who are in the dialect contact area. Contrarily, in communities established for a long time where there is no dialectal contact, the focusing occurs.

These concepts, intrinsically linked to the characterizations of social networks, have direct implication in linguistic uses, especially in relation to linguistic maintenance and change. The researches of Milroy (1980 [1987]) and Bortoni-Ricardo (2011 [1985]) pointed out the disposition for the maintenance and the focusing of the identity speech in communities whose networks are isolated, due to resistance to linguistic change. On the other hand, in open networks, large diffusion was featured.

This was also observed in relation to intermediate networks, whose structure established a slight tendency to diffuse. Two informants presented all the traits analyzed here and four of them only showed the absence of verbal agreement.

A remarkable fact in this analysis is the incidence of a verbal agreement that departs from the canonical standards. Twenty-three of the informants surveyed presented such a stigmatized mark, some of them more often than the others. The analyses undertaken here do not account for the answer to this matter. Nevertheless, it is necessary to delve into the subject in future studies.

Concerning the configuration of the networks, it is possible to understand their density or looseness as relevant factors for linguistic change or maintenance, more precisely for the diffusion or focusing of the rural variety. However, such a statement is not the only factor acting in this way. The configuration of the networks is not solely responsible for this. Other factors are permeating the linguistic choices, the conscious and unconscious uses by the teenagers from the analyzed district.

We understand, with Araujo, Santos and Freitag (2014, p.102-103), that “if an individual has a strong degree of closeness with an interlocutor and a weak one with another, his linguistic behavior in the interaction with each of them is probably different as a result of the type of relationship.” Thus, this fact has to be considered, since, still according to the author “the control of this variable allows us to verify whether the different linguistic uses are due to the degree of closeness among the informants.”

Therefore, the questions proposed in this paper may be resolved by the precise definition of the center point of the networks and by clarifying the level of closeness between the members of these networks. In this analysis, following the studies on literacy proposed by Street (2014) [1970]), it is relevant to highlight the literacy events common to this network, analyzing the concerned practices, and how this influences the rural speech maintenance or change.

An unfinished conclusion

In the previous pages, after presenting the theories that have been our base, we sought to establish the configuration of our informants’ linguistic maps. In addition, through this mapping, along with examples of the incidence or the non-incidence of some linguistic factors regarded as speech marks, we tried to understand the motivations for the maintenance of this variety, which hypothetically represents the identity of all the informants, or for its change in favor of a more urban linguistic expression. In other words, we sought to understand the reasons for the focusing or diffusion of this rural variety.

In these discussions, one thing is sure: the non-existence of rural speakers and urban speakers among the teenagers of the district. Everyone can be classified as rurbans speakers, and they can be allocated in various points of the urbanization continuum, but none in its extremes.

Another outstanding factor which deserves further study: the absence of the standard verbal agreement by 96% of the speakers at least in two moments of the interaction in which this phenomenon occurred. Which factors could be triggering that linguistic expression? As previously mentioned, by what was suggested here, it is not possible to answer this question. Our future investigation will undoubtedly focus on obtaining those answers.

For this and other reasons, there are more “inconclusions” than “conclusions” in this final conversation. Just establishing the configuration of social networks of the teenagers who reside in a district in Londrina was not enough to understand their speech configuration, though it has been significant, since it has functioned as an orientation, and may work as a guiding basis for the discussions that may be undertaken from here.

Our discussion began with the mention of INF1 and a questioning on her speech, which happened to be different from the one expected to be found in the countryside. Looking upon her expression, we were able to notice the social networks acting on her language configuration. Her relationship with her employers, speakers that may be located in points very close to the urban extreme in the urbanization continuum, or with her brothers-in-law, both with a university degree and, therefore, considered speakers of a prestige variety, may be a response to this higher closeness to the urban speech presented by the teenagers.

Hypothetically speaking, inserting into a dense network would be the reason why INF13 maintains linguistic marks that feature a rural speaker’s speech. This hypothesis is refuted by observing the linguistic expression of INF15, whose networks also have a closed structure, but who presents no incidence of iotization and rhotacism. Would identity be a reason for that? In both cases, through the interviews, it was possible to see their identity affinity with their parents: INF13 has an illiterate father; whereas INF15 has a literate father, which outstands from the other members of INF15’s social networks. Thus, this is a hypothesis that needs to be tested.

INF3 resides in ZCD, but his speech is one of the closest to the rural speech. His dense network, without any outward opening, promotes the focusing of his speech, making it very close the rural variety. Once again, the ties of his network may be at the heart of this focusing: an illiterate mother, whose literacy events evolve socially devalued practices, is his most prominent reference.

INF7 is another informant who may have her expression explained by literacy practices. She is resident in ZCD, she works, and she has all the feature marks of the rural variety observed in this paper. Her open network could be favoring the diffusion, but, on the contrary, focusing occurs. This may be favored by the literacy events in which she participates within this network, since her job is in the district and her colleagues and bosses, as well as her family members and friends, are all speakers of a rurban variety closer to the rural variety.

INF24 is resident in the ZCD. He is inserted in an open network, whose central tie seems to be his work, permeated by hegemonic literacy events. Probably, these factors are responsible for his linguistic expression, which were the closest to the urban speech. All his networks have ties to a more socially accepted culture.

These examples serve to emphasize the inconstancy in the responses.

As mentioned previously, it is possible, however, to see social networks literacy as a possible driving force behind linguistic change or maintenance, as it leads the speeches in favor or not of hegemonic agencies.

However, for that, it is necessary to establish, without a doubt, the core of each network ties to, knowing the reference, understand the speech structure of each of these teenagers. In view of what has been presented here, the need for further studies, such as those made by Battisti (2014) is urgent, which does not diminish all the research carried out here but put it as a starting point and a guiding basis for future researches, ours and of others who want to tread these paths. The field is fruitful. Starting it is all it takes.

REFERÊNCIAS

• AGUILERA, V. de A. Um estudo geolinguístico da iotização no português brasileiro. In: AGUILERA, V. de A. (Org.). Português no Brasil: estudos fonéticos e fonológicos. Londrina: Eduel, 1999.
• AMARAL, A. O dialeto caipira: gramática, vocabulário. 4. ed. São Paulo: HUCITEC, 1982 [1920].
• ARAUJO, A. S.; SANTOS, K. C.; FREITAG, R. M. K. Redes sociais, variação linguística e polidez: procedimentos de coleta de dados. In: FREITAG. R. M. K. (Org.). Metodologia de coleta e manipulação de dados em sociolinguística. São Paulo: Editora Edgar Blücher, 2014. p.99-115. Disponível em: <http://dx.doi.org/10.5151/BlucherOA-MCMDS-7cap> Acesso em:14 abr. 2018.
  » http://dx.doi.org/10.5151/BlucherOA-MCMDS-7cap>
• BATTISTI, E. Redes sociais, identidade e variação linguística. In: FREITAG. R. M. K. (Org.). Metodologia de coleta e manipulação de dados em Sociolinguística. São Paulo: Editora Edgar Blucher, 2014. Disponível em: <http://dx.doi.org/10.5151/BlucherOA-MCMDS-7cap> Acesso em:10 abr. 2018.
  » http://dx.doi.org/10.5151/BlucherOA-MCMDS-7cap>
• BORTONI-RICARDO, S. M. Nós cheguemu na escola, e agora? Sociolinguística e educação. São Paulo: Parábola Editorial, 2005.
• BORTONI-RICARDO, S. M. Educação em língua materna: Sociolinguística em sala de aula. 6. ed. São Paulo: Parábola Editorial, 2009.
• BORTONI-RICARDO, S. M. Do campo para a cidade: estudo sociolinguístico de migração e redes sociais. São Paulo: Parábola Editorial, 2011 [1985].
• BORTONI-RICARDO, S. M. Manual de Sociolinguística. São Paulo: Editora Contexto, 2014.
• CASTILHO, A. T. de. Nova gramática da língua portuguesa São Paulo: Contexto, 2010.
• COELHO, I. L.; GORSKI, E. M.; SOUZA, C. M. N. de; MAY, G. H. Para conhecer sociolinguística São Paulo: Contexto, 2015.
• EVANS, B. The role of social network in the acquisition of local dialect norms by Appalachian migrants in Ypsilanti, Michigan. Language Variation and Change, 16, p.153-167, 2004.
• FAZITO, D. A análise das redes sociais e a migração: mito e realidade. Anais… XIII Encontro Nacional de Estudos Populacionais, ABEP, setembro de 2002. Disponível em: <http://www.abep.org.br/publicacoes/index.php/anais/article/viewFile/1094/1058> Acesso em: 01 abr. 2014.
  » http://www.abep.org.br/publicacoes/index.php/anais/article/viewFile/1094/1058>
• GUMPERZ, J. The speech community. In: GIGLIOLI, P. (Org.). Language and social context. Londres: Penguin Books, 1972.
• HANNEMAN, R.; RIDLE, M. Introduction to social network methods 2005. Disponível em: <http://revista-redes.rediris.es/webredes/> Acesso em: 18 fev. 2014.
  » http://revista-redes.rediris.es/webredes/>
• LABOV, W. Padrões sociolinguísticos São Paulo: Parábola Editorial, 2008 [1972].
• LE PAGE, R. Projection, focusing and diffusion. York Papers in Linguistics, University of York, v.9, p.9-32, 1980.
• MILROY, L. Language and social networks. 2. ed. Oxford: Blackweel, 1987 [1980].
• NETTO, S. P. A psicologia da adolescência São Paulo: Ed. Livraria Pioneira, 1968.
• SEVERO, C. G. A questão da identidade e o lócus da variação/mudança em diferentes abordagens Sociolinguística. Revista Letra Magna Revista Eletrônica de Divulgação Científica em Língua Portuguesa, Linguística e Literatura, n.7, p.1-9, 2. sem. 2007. Disponível em: <www.letramagna.com/variacaomudan%E7asocio.pdf>. Acesso em: 06 fev. 2011.
  » www.letramagna.com/variacaomudan%E7asocio.pdf>
• STREET, B. V. Letramentos sociais: abordagens críticas do letramento no desenvolvimento, na etnografia e na educação. São Paulo: Parábola Editorial, 2014 [1995].

• 1
  The proposed analysis of Brazilian Portuguese continuum is well established and demonstrated by its creator, Stella Maris Bortoni-Ricardo, in her 2005 work, detailed in the references for this paper, but it should be clarified that, for the author, all Brazilian speakers can be inserted into the urbanization continuum, orality and literacy and stylistic monitoring. In this article, we focus on the urbanization continuum. In the left pole of this continuum are isolated rural varieties; the right pole, urban varieties standardized; everything that cannot be allocated on neither of those poles is considered rurban.ttribution 4.0 International License.
• 2
  Oh, when I’m somewhere more important and with my mistress, she’s always nagging me, her job, she was a teacher there, was a teacher of the UEL, she’s always correcting me ‘ say it right, say it right ‘, then, for me not to be being called at attention, so I’m kind of measuring ... (Inf1)

Publication Dates

• Publication in this collection
  Sep-Dec 2018

History

• Received
  8 Aug 2017
• Accepted
  3 Mar 2018