Esta é uma resenha da mesa-redonda Gramática de Construções: estabilidade, variação e mudança, composta pelos professores Marcia dos Santos Machado Vieira, Edson Rosa Francisco de Souza e Taísa Peres de Oliveira, coordenada por Marcos Luiz Wiedemer. Na atividade, que ocorreu no dia 18 de julho de 2020, promovida pelo evento Abralin Ao Vivo, a fim de dar conta dos três tópicos centrais da mesa – estabilidade, variação e mudança – foram expostos resultados de pesquisas que têm sido desenvolvidas no Brasil, fundamentadas na abordagem construcionista, que compreende a construção como um pareamento convencionalizado de forma e significado, como unidade simbólica. Ficou evidente o alcance que essa abordagem já tem em nosso país, sobretudo no tocante à variação e à mudança.


In this roundtable, entitled Construction Grammar: stability, variation and change[1], composed by professors-researchers Marcos Luiz Wiedemer (UERJ), Marcia dos Santos Machado Vieira (UFRJ), Edson Rosa Francisco de Souza (UNESP) and Taísa Peres de Oliveira (UFMS), analysis discussions based on Construction Grammar approach were presented. This theoretical perspective understands language, according to Traugott and Trousdale (2013, p. 1[2]), as a network of relations between constructions, and these are defined as “conventional symbolic units”, since, as they are shared among speakers, they start to generate a pairing between form (phonetic, morphological, syntactic properties) and meaning (semantic, pragmatic, discursive properties) in the speaker's mind.

To start the roundtable discussion, Marcos Luiz Wiedemer presented an overview of three moments, non-excluding, of the Construction Grammar. The first one, in the 1980s, dedicated to the phenomena of stability/systematicity and to the synchronic description; the second one, already in the 21st century, which starts to deal with linguistic change and incorporates the diachronic description, as well as methods of analysis from Corpus Linguistics; and the third one, more current, which made room to variation, experimental methodology and the discussion of theoretical and methodological concepts.

As for the variation, one of the main topics of the discussion, its positioning, central or peripheral, was denounced, depending on the researcher’s view, who may or may not consider the heterogeneous character of the language. However, even if the adopted view is idealized, one that understands the language as homogeneous, its heterogeneous nature will stand out before the researcher. This statement is in line with what Leino and Östman (2005, p. 192[3]) declare, for whom it is not possible to ignore variation, within the scope of the Construction Grammar, if this approach’s objective is to deal with all the constructions of a language, whether central or peripheral.

In Brazil, several researches are already being carried out aiming at the data treatment and description capable of illustrating the variation from a constructional perspective, following the examples of the works developed and guided by two of the speakers, Marcos Wiedemer and Marcia Machado (cf. MACHADO VIEIRA; WIEDEMER, 2018, 2019[4,5], WIEDEMER; MACHADO VIEIRA, 2018[6]). Both of them have raised discussions that involve variation both in terms of competition and in regards to coexistence between variants, i.e., alternatives licensed by the network of which they are part. For this, they have been working on variation in two different ways: (i) variation by analogy/alignment of independent constructions; and (ii) variation by matching (col)lexemes to construction slots. For this purpose, Machado Vieira e Wiedemer (2020, forthcoming[7]) present the following methodologies for the variation treatment within the scope of the Construction Grammar: (i) by symbolic similarity; (ii) by discursive paradigm; and (iii) by alloconstruction and metaconstruction, which received prominence in their speeches.

The variation by alloconstruction and metaconstruction corresponds to possibilities of representing, “by comparability, configurational similarity or symbolic similarity among constructs/uses licensed by two or more constructional patterns of the system, based on an alignment of attributes (of form and/or function) of these” (WIEDEMER; MACHADO VIEIRA, 2018, p. 48[6]). In other words, according to Cappelle (2006[8]), it is the way of conceiving two patterns, not as a derivative of the other, but as variant realizations of a partially under-specified construction, i.e., as alloconstructions. In turn, metaconstruction is the theoretical construct that neutralizes differences and conceptualizes the similarities between constructions.

Marcia Machado, in her presentation entitled Constructional variation: perspective and reality, treated variation as a subject that has generated provocations and further research. To talk about the provocations, she brought, among other points of discussion, the conceptions of (a) variants envelope, idea of ​​the same truth value between different units, understanding the variation as inherent in the language; (b) functional comparability, as a condition for the study; (c) principle of non-synonymy, when she asks, in the Construction Grammar space, whether it is possible to talk about variation and consider this principle at the same time; (d) iconicity, questioning this one-to-one relationship, in a way that only has a semantic or pragmatic meaning; and (e) arbitrariness.

In the research exhibition, in order to deal with compatibility among variants, similarities and differences and functional and formal factors, she exhibited, above all, studies under development in the Predicar project - Formation and expression of complex predicates and predictions, since Machado Vieira (2016[11]). Many of these researches have been dedicated to variation as one of the central points, as exemplified by two studies1, one on constructions [support verb + particle “-se” + the definite article “uma” + the preposition “de” or “por” X], about the support verbs give (“dar”), do (“fazer”), pass (“passar”), pose (“posar”) and take (“tirar”) followed by a prepositioned nominal element, and another on intensifying constructions X of Y, in which slot X is filled with a lexeme of color, as purple/green of anger or red of embarrassment (“roxo/verde de raiva” ou “vermelho de vergonha”). In these and other researches of the project, various types of analysis have been used, such as frequency analysis, multivariate analysis (KLAVAN, 2012[9]), collostructional analysis (STEFANOWITSCH, 2013[10]) and, still, experimental researches.

Edson Rosa dedicated his speech to the presentation of the work entitled A diachronic study of intensifying micro-construction [[x] “pra caramba” (as heck)] in Portuguese. First, he answered the following question: "What moves the speaker?" His answer was directed to the intensification treatment, which led him to affirm that we, speakers, make assessments, whether intensified or not, about what surrounds us, as people, feelings and places, a fundamental characteristic present in the most diverse communicative situations. He also stated that looking at language from a usage-based perspective is to understand that the linguistic structure reflects the structure of the speaker's experience/world perspective. It is worth noting, based on his statement, that more than reflecting the experiences of the speakers, the language also generates meanings and contributes to this experience.

His objective was to analyze, based on the assumptions of the constructional approach, the intensifying micro-construction [[x] “pra caramba” (as heck)] in Portuguese, instantiated by the subschema [[x] Prep + N], which motivated him to (i) verify the formal and functional motivations that led/leads the speaker to choose this peripheral micro-construction instead of a simple one to intensify and (ii) analyze the process of construction change, in view of the schematicity, the compositionality and the productivity. Of this work, it is worth highlighting, among other points, (a) the observations about the simple and peripheral constructions and their relationship with the intensity gradation and the treatment of hyperbolic intensity, since there is intensification in both, although in different degrees; (b) the fact that the construction [[x] “pra caramba” (as heck)], based on the Corpus do Português data, is a recent intensification strategy, considered innovative in the language; and (c) this construction is little stigmatized when compared to [[x] “pra caralho” (as fuck)], very marked, and [[x] “pra cacete” (as hell)], in a medium way.

Therefore, it was concluded that, as for the formal properties, such construction is fixed, since it is usually positioned to the right of the element modified by it; it does not allow inflection of number, gender and person, and its use in the diminutive and in the augmentative is not very productive; and the preposition [“para” (for)] appears, almost every time, in its reduced form “pra”. Regarding functional properties, the author stated that the construction [“pra caramba” (as heck)] can act as an intensifier by modifying verbs, adverbs, adjectives, names and sentences, transmitting a positive value. In addition, its frequency is higher in textual genres of news and newspaper articles that present contexts of oral interaction.

Finally, Taísa Peres defined as her goal, in the speech entitled Network models in the treatment of conditionals, to present a construction network that would consider that networks can emerge, expand, reorganize. For this, she exposed characteristics of the vertical model of inheritance and of the horizontal network model, in order to show that the association between both is more effective than the treatment of just one of them. The first model, developed by Goldberg (1995, 2006[13,14]), understands that the constructions are based on generalizations of real statements, and the network is seen as an organization of schemes, subschemas, micro-constructions and constructs. The second model, proposed by Velde (2014[12]), understands that schemas and subschemas do not represent discrete expressions; in fact, a construction can be partially motivated by its neighbors.

The participant illustrated the association between these models, using the category of conditionals in Portuguese, signaling that this list is not easily closed. She also highlighted the advantages of working with both models, such as the possibility of (i) explaining functionality as a result of using items that instantiate existing schemes; (ii) showing that subschemas can also inherit properties from neighboring subschemas, and not just from the highest hierarchical levels; (iii) and understanding that it is possible to deal with a multidimensional network model, in order to explain the relationship between the different semantic and pragmatic domains that support a given construction.

Through such presentations, and through the researches that served as a base and are located at the most current moment in the Construction Grammar, it was evident that the variation has been highlighted, the theoretical and methodological concepts have been questioned and expanded and experimental researches are already being planned and elaborated. Thus, the roundtable illustrated how rich and pertinent are the contributions that have been made from the constructional perspective to the variation and change treatment, both of these characteristics inherent to the language.


  1. GRAMÁTICA de construções: estabilidade, variação e mudança. Mesa-redonda apresentada por Marcia dos Santos Machado Vieira, Edson Rosa Francisco de Souza e Taísa Peres de Oliveira, sob coordenação de Marcos Luiz Wiedemer, 2020. 1 vídeo (2hs24min 44s). Publicado pelo canal da Associação Brasileira de Linguística. 2020.
  2. Cappelle B. Particle placement and the case for “allostructions”. Constructions. 2006; 1
  3. Goldberg A. E. Constructions: a construction grammar approach to argument structure. The University of Chicago Press: Chicago; 1995.
  4. Goldberg A. E. Constructions at work: the nature of generalization in language. Oxford University Press: Oxford; 2006.
  5. Klavan J. Evidence in linguistics: corpus-linguistic and experimental methods for studying grammatical synonymy. Institute of Estonian and General Linguistics - University of Tartu, 2012.
  6. Leino J, Östman J.-O.. Construction and variability. Grammatical Construction. Back to the roots. John Benjamins Publishing Company: John Benjamins Publishing Company; 2005.
  7. Machado Vieira M. S. Variação e mudança na descrição construcional: complexos verbo-nominais. Revista Linguística. 2016; EspecialDOI
  8. Machado Vieira M. S, Wiedemer M. L. Sociolinguística variacionista e Gramática de Construções: os desafios e as perspectivas de compatibilização. In: Machado Vieira M. S, Wiedemer M. L. Dimensões e experiências em Sociolinguística. Blucher: Blucher; 2019.
  9. Wiedemer Marcos Luiz, Machado Vieira Marcia Dos Santos. LEXEMAS E CONSTRUÇÃO: ATRAÇÃO, COERÇÃO E VARIAÇÃO. Caderno Seminal. 2018; 30(30)DOI
  10. Machado Vieira M. S, Wiedemer M. L. A variação no modelo construcionista da Linguística Funcional-Cognitiva. Anais do XXXII ENANPOLL. a sair.
  11. STEFANOWITSCH A. Collostructional analysis. The Oxford Handbook of Construction Grammar. 2013.
  12. Traugott E. C, Trousdale G. Constructionalization and Constructional Change. Oxford University Press: Oxford; 2013.
  13. Van de Velde F. Degeneracy: The maintenance of constructional networks. Extending the scope of Construction Grammar. De Gruyter Mounton: De Gruyter Mounton; 2014.
  14. Wiedemer M. L, Machado Vieira M. S. Sociolinguística e gramática de construções: o envelope da variação. Sociolinguística: estudos de variação, mudança e atitudes linguísticas. Ed. da Unicentro: Ed. da Unicentro; 2018.