Abstract This study investigates variable number agreement within the determiner phrase (DP) in two non-standardized Latin American Spanish varieties: Afro-Mexican Spanish (AMS) and Afro-Bolivian Spanish (ABS). Results show that both phonological and morphosyntactic factors shape agreement, although the varieties differ structurally. While AMS and ABS occupy distinct positions along a continuum of contact-induced grammatical restructuring, both display reduced agreement systems consistent with vernacular universals – recurring linguistic tendencies across non-standard varieties that reflect shared cognitive and social constraints. This perspective aligns with research on agreement impoverishment in the Spanish DP. Integrating a variationist approach with formal and functional theories of morphophonological reduction, the study advances understanding of agreement in underexplored Spanish varieties. Findings further indicate that grammatical simplification in AMS and ABS extends beyond the DP, mirroring broader patterns of structural reduction characteristic of vernacular speech worldwide.
Abstract This study deals with mood variability in Spanish, addressing whether semantic or non-semantic effects best predict mood selection, and if the Spanish subjunctive is disappearing. Focusing on epistemic adverbials of doubt and probability, over 7,000 occurrences of variable mood selection in Argentinian and Mexican Spanish, 1850s–2001, were analyzed using Bayesian mixed-effects logistic regression. Both semantic ( epistemic adverbial, aspect, presence of modal auxiliaries , and polarity ) and non-semantic ( distance between the matrix and the verb, conditional probability of the verb infinitive and of the verb form , and morphological similarity ) predictors were analyzed in interaction with real time and variety . The results show that while both semantics and non-semantics constrain mood selection, semantic effects are more limited in scope. Moreover, subjunctive usage has increased over time. I argue that in both varieties change is attributable to concomitant developments in the epistemic adverbial system itself: as indicative-favoring acaso falls into disuse, the adverbials that remain frequent are overwhelmingly variable in mood selection. In Argentinian Spanish, where these remaining adverbials are stably variable, probablemente – previously restricted to the indicative – shifts towards variable mood selection by analogy with the dominant pattern. In Mexican Spanish, the adverbial system itself is more unstable, which delays this consolidation. The two varieties thus diverge in the actualization of change as a consequence of differing input frequencies of contexts permitting the subjunctive. A comprehensive account of mood must therefore integrate both semantic and cognitive determinants, while also bearing in mind that the constructions governing mood selection may themselves be in flux.
Abstract This study investigates the perception of Brazilian Portuguese (BP) mid vowel contrasts (/e/-/ɛ/; /o/-/ɔ/) by English speakers who learned Portuguese as an L2 in a naturalistic setting and had recently returned to the United States. L2 learners took part in three experiments: an AXB discrimination task, a binary word categorization task, and a sentence acceptability judgment task. Results from the experiments indicate that experienced L2 Portuguese listeners successfully perceived these contrasts on the AXB task, reaching ceiling effect. On the subsequent two tasks their performance was above chance, although below that of the L1 control participants. The results suggest that even though one to two years of naturalistic immersion foster significant progress in L2 categorization of Brazilian Portuguese mid-vowel contrasts, adult learners may require further exposure and/or explicit phonetic training to fully establish the phonetic categories needed for native-like accuracy.
Abstract This study provides an empirical and quantitative cross-dialectal comparison of the stressed vocalic inventories of seven Portuguese varieties and contributes to the description of under-documented ones from Africa. Aiming to verify the hypothesis that the phonetic realization of stressed vowels varies between varieties for the same vowel category, we compared the phonetic realization of stressed oral vowels in European (continental), Brazilian, and African (Angola, Guinea-Bissau, Cape Verde, São Tomé and Príncipe, Mozambique) Portuguese varieties, using the Spoken Portuguese – Geographical and Social Varieties corpus as a source for naturalistic data, with a sample of five speakers per country. For each recording, all seven stressed vowels were labeled (initial n = 5,296), followed by formant measurements and outlier removal (final n = 5,111). The Euclidean distance between vowels was also calculated and used to evaluate the distance between mid-high and mid-low vowels. In order to cope with the unbalance in our data and to account for individual variation across informants, a linear mixed-model was fit to the data, with fixed effects for country , vowel , and sex and random effects for speaker and word . We found that there are similarities between all the varieties as regards a tendency for mid vowel merger, but also differences in the quality of stressed vowels across varieties, which could be due to language contact. More and suitable data are needed in order to verify the role played by language contact.
Abstract In contemporary Spanish, the verb decir ‘say’ presents idiosyncratic behavior regarding subject expression and position. In spoken Peninsular Spanish, decir strongly favors postverbal subjects, unlike other transitive or ditransitive verbs. The present paper focuses on the diachrony of the verb form dice / dize , tracing back the syntactic patterns in diachronic data extracted from the Corpus del Español from the 13th to the 21st century. Our aim is to elucidate whether postverbal subjects of dice/dize have become more frequent over time, or whether they rather can be considered a vestige of word order patterns that were more productive in old Spanish, as the residual V2 hypothesis predicts. The data suggest that while null subjects have become more common and postverbal subjects less common in general, this is due to the increasing use of specific constructional patterns, partly related to the frequency of syntactic operations like quotative inversion. The idiosyncratic morphosyntactic tendencies of the verb decir reflect the subject types and discourse functions of the verb that remain unchanged throughout the centuries.
