Linguistic variation and second language Spanish

Abstract

Research on progressive and habitual aspectual marking in Spanish provides a productive test case for second language variation research. In Spanish, two forms can be used to reference ongoing action at speech time (simple present and present progressive) while English predominantly makes use of the present progressive (Torres Cacoullos 2000). Nevertheless, the patterns of use and their acquisition are complex. One key difference for advanced non-natives (as compared to natives) is related to the range of lexical bases (e.g., estar ‘to be,’ andar ‘to walk,’ venir ‘to come,’ etc.) that form the progressive (Fafulas 2015). Additionally, the alternation between the simple present and the progressive is conditioned by a host of linguistic factors, such as lexical aspect, the presence of an adverb, clause type, and animacy, among others (Geeslin & Fafulas 2012). Finally, differences exist between these same two forms cross-linguistically for habitual aspectual marking such that the simple present is more likely to occur in English in habitual contexts than in Spanish and this constitutes another context for these variable forms (Fafulas 2012). Thus, the predictions for English-speaking learners are highly context-dependent. The present study uses a written contextualized task designed to examine patterns of selection across the categories of lexical aspect (e.g., activities, statives, etc.) in combination with the influence that the presence or absence of adverbial phrases may have. We employ a cross-sectional design, collecting data from learners from multiple levels of enrollment, and compare these to native speakers of Spanish and to native speakers of English (tested in English). Our analysis provides an account of the patterns attested in English, in Spanish, and across levels of second language Spanish development.

Variability and individual differences in L2 sociolinguistic evaluations: The GROUP, the INDIVIDUAL and the HOMOGENEOUS ENSEMBLE

“…This can more generally aid in facilitating a clearer understanding concerning the extent to which learners integrate different varieties into their multivarietal repertoires based on their explicit experiences with, exposure to, and proficiency in the respective varieties. Second, as Geeslin et al (2013) and Kanwit (2022) note, variationist SLA research must more aggressively address how socioaffective factors such as motivation affect L2 learners' sociolinguistic repertoires. To our knowledge, the results in George (2014) provide the lone exploration concerning the extent to which quantitatively captured motivational factors explain differential outcomes in sociolinguistic competence, though her results focused on differences in production in L2 Spanish.…”

“…In this section and addressing RQ2, subject-level variability is thus subjected to exploratory Bayesian analyses to determine whether the variability in the present subjects' L2 evaluative judgements is explainable as a function of standard/dialect proficiency, exposure, and motivation. Moreover, these analyses address the calls for research in Geeslin et al (2013) to (a) probe the extent to which variability is linked to linguistic and extralinguistic individual factors and (b) explore individual differences variables beyond proficiency and exposure.…”

“…Such findings from a variationist viewpoint are, of course, encouraging, as they would posit that group estimates provide an accurate picture of individual variation behavior in use. As Geeslin et al (2013) lamented, however, the aforementioned investigations strictly considered either one or two proficiency levels, thus obscuring how individual variability shifts (i.e., grows, declines) dynamically across proficiency levels. Geeslin et al (2013) addressed this issue in more detail and underscored that individual norms of variation are, indeed, prone to change across, for example, proficiency levels such that variability decreases with increasing proficiency, though it does not disappear entirely.…”

“…This is problematic inasmuch as one predominant goal of applying inferential statistics is to be able to generalize a set of observations to a wider population that the same sampling procedure would produce (i.e., external validity). However, drawing such population-level conclusions is oftentimes difficult to achieve, particularly in research focusing on the acquisition of variation and in SLA as a whole, given the tremendous individual variation in L2 learning outcomes and, by extention, the inherent heterogeneity of acquisition and development (Dörnyei, 2006;Ellis, 2004;Geeslin et al, 2013). The field of applied linguistics has begun addressing such issues, particularly within the meta-theoretical framework of complex dynamic systems theory (CDST; for a recent review of the CDST research in SLA, see Hiver et al, 2022), and there are increasing calls for innovative methods to more accurately transfer findings to wider populations (see, e.g., Lowie & Verspoor, 2019;.…”

“…Even if the overall variability cannot be explained by any particular individual differences, the results will address whether ensembles-that is, subgroups of similarly behaving individuals in terms of their evaluative judgements-are differently influenced by (certain levels of) individual differences (i.e., are the findings at the subgroup level and those of the individuals in said subgroups mutually inferable?). In view of the increasing calls for more person-centered approaches to influencing factors in sociolinguistic development (e.g., Ender, 2019;Geeslin et al, 2013;Howard, 2012;Kinginger, 2008;van Compernolle, 2019), these topics warrant much-needed investigation and can inform the variationist SLA landscape as to the dangers of taking group-level estimates at face value without also assessing the learner individuality present in the data. To this end, this study is guided by the following exploratory research questions:…”