Skip to main content
School of Languages, Linguistics and Film

Pedro De Lima has been awarded a PhD for Research in Linguistics.

Thesis title: Language Variation and Social Positioning in a Brazilian Terreiro.


We are pleased to announce that Pedro De Lima has been awarded a PhD for Research in Linguistics. The date of the award is 31-Jul-2023.

Thesis title:
Language Variation and Social Positioning in a Brazilian Terreiro.

Supervised by:

  • Erez Levon
  • Devyani Sharma


Thesis summary:

Language Variation and Social Positioning at a Brazilian terreiro

Pedro Eduardo de Lima


In my thesis, I examine how language variation correlates with the social structuring of a community of Candomblé, an Afro-Brazilian spiritualistic religion. This specific religious setting was chosen because of its wide openness to people of different ages, sexualities, and schooling, some of the social factors I aimed to include in the investigation. My main goal was to determine if and how linguistic and especially social factors correlate to the plural configuration within the noun phrase (NP) – the dependent variable in the study – in the speech of followers of a terreiro, which is the place for religious practices in Candomblé.

In standard Brazilian Portuguese (STA), all pluralisable items in plural NPs are marked, e.g. “A-S casa-S grande-S” (“The big houses”). In more informal contexts, this standard redundant nominal number agreement is optional, as in “A-S casa grande”. Users of the nonstandard variant (-Ø) are often a target of linguistic prejudice. Aiming to investigate the correlation between this plural configuration in the speech of participants and social factors, I conducted an ethnography at a terreiro in Salvador/Brazil. It included intensive participant-observation for 5 months, field notes, and semi-structured sociolinguistic interviews with 18 followers. Participants were grouped by age (younger/older), schooling (lower/higher), sexuality (gay/straight), and their role within the terreiro (rodantes if they can experience a trance-like state and be inhabited by orixás – the gods in the religion, and ogans if they cannot).

Quantitative analyses revealed that some linguistic factors influenced the plural configuration within the NP as independent fixed effects. It was also demonstrated that, despite the collinearity role-sexuality among participants, sexuality did not show an effect over NP plural marking in the dataset, whereas role did. The analyses further revealed statistically significant effects of interactions between social factors over the dependent variable, namely age and schooling, role and schooling, and role and age. Among other findings, the analyses of these interactions showed that: (i) older participants with higher education use -Ø more than the younger; (ii) ogans with higher education favour -Ø more than rodantes with the same schooling; (iii) older rodantes privilege -Ø more than the younger ones. Findings (i) and (iii), further considered with field notes examination, suggest that standard language was less valued in the past at the terreiro than it is now.

Qualitative analyses of the interviews and of field notes explored the reasons why the interactions found take place as they do. Such analyses suggest that informants take up positionings through metapragmatic evaluations, i.e. evaluations of their own and others’ plural marking of NPs. This results in an agentive linguistic shaping of different identifications (the social factors considered here) by participants, and such identifications shaping is intrinsically embedded in the terreiro’s social structuring. For example, considering the interaction between age and role, the favouring of STA by younger rodantes more than by younger ogans is explained by the fact that the first have more contact with outside society and engage in further education more than the latter. Also, rodantes’ favouring of STA over -Ø is seen by themselves as an important and available tool which they use, aiming to grow within the religion. Additionally, STA use integrates rodantes’ aspirational practices seeking to eventually become a Babá – the leader of a terreiro. Such phenomena suggest that, in participants’ perspective, mastering standard language, more specifically the standard pluralisation of NPs, is indexically linked to what is considered a good Babá. On the other hand, having lower schooling and limiting one’s professional life to their duties at the house is often considered as part of being an ogan. Further, ogans cannot envisage one day becoming leaders of terreiros, as this is only available for rodantes. These features pertaining to the role of ogan help explain why they favour the -Ø variant more than do younger rodantes. STA use is not part of what constructs the role of ogan at the place. As a result, ogans do not seem to find sufficient reasons to seek mastering STA use, like pursuing further education, for example. Therefore, as a social meaning, low schooling helps producing the role of ogan and also results out of it. Consequently -Ø becomes indexical of younger ogans.

In other words, the plural configuration within the NP helps construct and maintain the terreiro’s social and hierarchical organisation and functioning. Ultimately, the study conducted helps explain how the terreiro is sociolinguistically structured. This work brings contributions to sociolinguistics, especially in intersection with studies of identifications and those with  underrepresented speech communities. Furthermore, as a direct social contribution, the study gives more visibility to Candomblé, a religion which is a target of stigma and prejudice in Brazil despite its positive influence on Brazilian arts, culture, and society more broadly.

Keywords: sociolinguistics, noun phrase, Candomblé, age, schooling, role



You can read Pedro's full thesis here:




Back to top