Blend, “the mix everyone likes!”: a blendscription of the process in the lexicon of brazilian portuguese"

Abstract

The linguistic creativity in word formation has been the subject of study by various linguists. Among the creative processes of word formation is the blend, originating from two pre-existing words to create a new word, with the suppression of segmental or syllabic material from at least one of them, and in certain cases, overlapping segments (PEREIRA, 2016; GONÇALVES, 2019), as seen in examples like "chafé" (chá + café) and "Debochara" (deboche + Bechara). Being formed from two bases, the phenomenon exhibits similarities with the process of compounding, as in examples like "cavalo-marinho" and "pontapé." However, compounds are formed through concatenation (VILLALVA, 2020), while blends are formed through non-concatenation (BEVILACQUA; SILVA, 2021). Thus, in this work, we propose an analysis of blends to provide further contributions on this complex phenomenon that is quite prevalent in the daily lives of speakers. For this purpose, we have structured the research around three analytical dimensions: i) an investigation into the morphological status of the phenomenon within the language, through an examination of its lexicalization process; ii) an analysis of the perception of native speakers' knowledge and use of blends; iii) an analysis of the accentual pattern, with the aim of understanding whether speakers access accentual information in the creation of blends. Three main questions guide this study: 1) What is the morphological status of the blend in the lexicon? 2) What is the level of knowledge and use of blends by native speakers? 3) Is the accentual pattern of blends related to their types of formation? As a preliminary response to these questions, we propose the following hypotheses: that the blend is a word-formation process independent of the compounding process that combines two words to generate a third word; that the knowledge of the meaning of blends is associated with the capacity for semantic decomposition by native speakers and usage is linked to the social context in which the blend was created; and that blends exhibit distinct accentual patterns based on their formation types: lexical interposition (e.g., "promochila" << promoção + mochila), truncated combination (e.g., "chocotone" << chocolate + panetone), and sublexical substitution (e.g., "roubartilhar" << roubar + compartilhar). Our goals are, therefore, to suggest that the blend has a morphological status in the lexicon independent of compounding, to investigate the level of knowledge that speakers have regarding blends, and to examine the accentual patterns of blends, considering their formation types. To this end, we compiled a database of 750 blends to discuss the morphological status of the phenomenon, drawing from studies conducted up to the present moment of this thesis's preparation. We considered aspects involved in their formation, such as phonological, morphological, syntactic, and semantic factors. For the perception test, we designed a Google Forms survey to assess native speakers' knowledge of blends. To investigate the rhythmical dimension, we employed Acoustic Phonetics to understand the accentual algorithm of the phenomenon through relative duration. Our results indicated that blends differ from compounds because their formation involves prosodic considerations and can exhibit phonological processes like haplology. Additionally, blends are mostly endocentric, while compounds maintain their phonological integrity and are mostly exocentric. Perception-wise, the results showed that blends are more contextual, as they tend to be better known and used in the environment in which they were created. The results also indicated that there is no difference in the accentual patterns of different blend types. Furthermore, the findings demonstrated that speakers tend to access accentual information in the creation of a blend, leading us to infer that the phenomenon has two primary accents based on prosodic composition (ULRICH; SCHWINDT, 2020).