Introduction
Coda /s/ is an extensive topic within the field of Hispanic Linguistics, in general, and in Puerto Rican Spanish, in particular. Nonetheless, there is a scarcity of research that specializes in the attitudes that Puerto Rican speakers attribute to the coda /s/ variants. Previous research has attributed perceived sexuality to coda /s/ variants (Mack, 2011, 2016; Walker et al., 2014). However, what we know about other attitudes associated to coda /s/ variants in Puerto Rican Spanish come indirectly from studies that examines attitudes regarding different Spanish dialects (Mojica De León, 2014). Mojica De León (2014) notes that /s/ aspiration does not have negative attitudes in Puerto Rico. Furthermore, even though some studies on the attitudes toward /s/ variants in other dialects have shown that weakened /s/ variants are highly stigmatized (Fuster, 2012; Rojas, 2012) and used to identify regional varieties considered less prestigious (Fuster, 2012), other studies show positive evaluations towards weakened variants (Cedergren, 1973, as cited in Blas-Arroyo, 1999, p. 60). Therefore, this chapter aims to fill this research gap by examining both the production patterns and the attitudes towards the different variants of coda /s/ in Puerto Rican Spanish. Following Mojica De León (2014), we hypothesize that /s/ variants may not be as stigmatized as what has been reported for other dialects, given the high frequency of weakened variants (e.g., [h] and ∅), throughout different social classes and stylistic contexts in Puerto Rico (Galarza & Sedó, 2016; Holmquist, 2011; López-Morales, 1980; Luna, 2010; Poplack, 1980; Terrell, 1978; Valentín-Márquez, 2006; Willis et al., 2015).
This chapter works under the assumption that phonetic variants can index social information, such as age, gender, nationality, etc. (Foulkes & Docherty, 2006). Additionally, we argue, following Eckert (2008), that indexed social information is dynamic, constructed locally, in daily interactions, and may vary across populations. This means that phonetic variants may index social information, but this information is constantly reinterpreted by speakers. Consequently, the same phonetic variant may have diverse social meanings in different dialects.
Previous investigations on coda /s/
Coda /s/ has been extensively studied in Spanish because of its widespread dialectal, social, and stylistic variation. This variability stems from coda /s/ lenition (Bybee, 2000; Hualde, 2014; Lipski, 1999; Widdison, 1993), but also has been attested in onset position (Lipski, 2011). Previous studies show that coda /s/ weakening is more likely to occur in preconsonantal contexts, followed by prepausal contexts, and prevocalic contexts (Bybee, 2000; Lipski, 1999; Widdison, 1993). Furthermore, dialects may differ regarding in which /s/-lenition stage they are. For instance, there are /s/-maintaining dialects, such as Latin American Highlands and Northern-Central Spain, transitional dialects, such as coastal Latin American zones, and advanced dialects, such as Caribbean dialects and Andalusia (Bybee, 2000; Hualde, 2014; Lipski, 1994, 2011). Thus, many Spanish dialects evidence coda /s/ lenition with varying degrees.
Even though coda /s/ is in a process of change in Spanish ([s] → [h] → ∅, Ferguson, 1990), it is mediated by social factors (Alba, 2010; Emmanuelli, 2000; Lafford, 1986; Lynch, 2009; Samper Padilla, 2011) as well as by linguistic factors (Brown, 2009; Bybee, 2000; Caravedo, 1987; File-Muriel, 2009; File-Muriel & Brown, 2011; Fontanella de Weinberg, 1973; Lipski, 1999; Lynch, 2009; Poplack, 1980; Terrell, 1978; Valentín-Márquez, 2006). Additionally, previous investigations have documented a wide range of coda /s/ variants, which includes [s], [z], [h], [ɦ], [ʔ], and ∅ (Brogan & Bolyanatz, 2018; Filimonova & Berkson, 2016; Galarza & Sedó, 2016; Luna, 2010; Sedó et al., 2020; Schmidt & Willis, 2011; Torreira, 2006; Valentín-Márquez, 2006; Widdison, 1993; Willis et al., 2015; Willis et al, 2016). Furthermore, /s/ lenition is also subject to compensatory effects, both in the preceding vowel, as well as in a following consonant (Amastae, 1989; Bishop, 2007; Figueroa, 2000; Galarza et al., 2014; Hochberg, 1986; Resnick & Hammond, 1975; Scrivner, 2014).
It is worth mentioning that, even though coda /s/ variation differs cross-dialectally, several investigations have found that men and lower socioeconomic classes tend to reduce or elide coda /s/ more frequently than women and higher social classes. This pattern has been documented in Las Palmas, Spain (Samper-Padilla, 2011), Barranquilla, Colombia, (File-Muriel, 2019), Concepción, Chile (Rogers, 2020), and Cubans living in Miami (Lynch, 2009), to mention a few studies. Furt...