Scott Schwenter | Ohio State University (original) (raw)

Papers by Scott Schwenter

To appear in a volume on Western Ibero-Romance languages, ed. by Manuel Delicado-Cantero, Fernando Tejedo-Herrero, and Patrícia Matos Amaral, 2025

Alternate ways of expressing sentential negation and variation among these expressions have been ... more Alternate ways of expressing sentential negation and variation among these expressions have been a well-known phenomenon ever since Jespersen's (1917) identification and description of a "negative cycle" in languages such as French and English. However, more in-depth, usage-based study of non-canonical forms of sentence negation has only exploded in the past 20 years or so, as researchers have dug deeper into the discourse-pragmatic motivations for choosing different ways to express sentential negation, negative polarity, negative concord, metalinguistic negation, etc. This chapter surveys some of the main research on this phenomenon in Western Ibero-Romance, with a focus on varieties (including Latin American varieties) of Spanish and Portuguese. The overarching goal is to encourage future research on the range of non-canonical forms and meanings expressed by languages in this part of the world.

We examine the alternation in Brazilian Portuguese 2SG DO pronoun expression between clitic te an... more We examine the alternation in Brazilian Portuguese 2SG DO pronoun expression between clitic te and tonic você using data from an online forced-choice survey completed by 146 native-speaker respondents. Results of mixed-effects logistic regressions show that dialectal subject pronoun preference (tu vs. você) and contrast both play a significant role in conditioning this choice. While te is preferred overall- a preference stronger among users of tu- você is the variant preferred in contrastive contexts. In double contrast contexts, where both the DO and the predicate are contrastive, você is even more strongly preferred. Thus contrast, despite its traditional treatment as binary, shows gradient effects on pronoun choice- the stronger the contrast, the greater the likelihood of você selection. Ultimately, we argue that what counts as the “unmarked” 2SG DO pronoun in BP is context-dependent and can only be determined taking type of contrast and speakers’ subject pronoun preferences into account.

Languages, 2024

We reanalyze the phenomenon of verbal (non)agreement with 2SG tu in a megacorpus of Brazilian Por... more We reanalyze the phenomenon of verbal (non)agreement with 2SG tu in a megacorpus of Brazilian Portuguese compiled from the web. Unlike previous research, which has analyzed sociolinguistic interview data and regional differences, we examine these data with a focus on the internal linguistic factors that constrain the variability. Our analysis of 4860 tokens of tu + verb reveals that non-agreement with the 3SG verb form is by far the most common pattern, 2SG agreement being relatively infrequent. Individual verb lexemes show highly distinct rates of (non)agreement. In addition, the specific tense/aspect/mood forms and main/auxiliary status are likewise significant factors affecting the variation. We conclude that future studies of this phenomenon should not ignore these internal linguistic factors. We situate our study within a group of other recent studies in Romance linguistics which have found that individual verbal and constructional patterns can have diverse effects on morphosyntactic variation.

Both Brazilian Portuguese (BP) and European Portuguese (EP) permit definite third-person null dir... more Both Brazilian Portuguese (BP) and European Portuguese (EP) permit definite third-person null direct objects (DOs) in anaphoric contexts, but differ with regard to the overt DO variant, with the former variety favoring tonic pronouns, the latter clitics. Previous research on language production has shown that the choice between variants in both varieties is constrained by semantic-pragmatic features such as animacy and specificity. We analyze speaker evaluation of these forms in EP and BP, using an experimental acceptability judgement task in which stimuli varied according to DO variant, animacy, and specificity. Data are drawn from the evaluative responses (n= 1752) of 215 Portuguese-speaking participants. Our results demonstrate that the null variant is evaluated most positively overall in both varieties. For EP and BP respectively, the clitic and tonic variants were evaluated most positively with animate specific referents. Our findings show that the patterns of variation previously found in production are reflected in gradient evaluations of anaphoric DOs in EP and BP. This provides support for the hypothesis that the potential shifts in Brazilian Portuguese regarding the transition from clitic to tonic direct object (DO) pronouns, suggest that the overt variant may have adopted characteristics similar to the clitic DOs in EP.

Contemporary Trends in Hispanic and Lusophone Linguistics

From Words to Discourse, 2002

The study of polarity sensitivity has been one of the topics that has sparked greater interest in... more The study of polarity sensitivity has been one of the topics that has sparked greater interest in semantics/pragmatics in recent decades, its popularity due in large part to Klima's (1964) groundbreaking study of negation in English. The current state of the art on this topic most ...

Multidisciplinary Approaches to Bilingualism in the Hispanic and Lusophone World, 2017

Variation and Evolution, 2020

The Routledge Handbook of Variationist Approaches to Spanish, 2021

This article examines a case of syntactic variation that has not been previously noted and studie... more This article examines a case of syntactic variation that has not been previously noted and studied: the Spanish neuter clitic lo and the null direct object (DO) in two dialects of the language, Mexican and Peninsular. The phenomenon of null DOs (a.k.a. “object drop” or “null direct object pronominalization”) in Spanish has been traditionally considered to be highly restricted: null DOs are only permissible when the noun in question is non-referential and non-countable, i.e. mass nouns as in (1a) and bare plurals (1b) (Campos 1986; Clements 1994, 2006):

Journal of Pragmatics, 2019

We describe the system of confirming affirmative responses in Brazilian Portuguese (BP) on the ba... more We describe the system of confirming affirmative responses in Brazilian Portuguese (BP) on the basis of a corpus of natural spoken dialogues between interlocutors that share a high degree of familiarity. While the BP response system has been characterized as an echo system (Sadock and Zwicky 1985), the unmarked option being a verb that echoes the verb in the antecedent utterance, our analysis reveals that this characterization only applies to polar question antecedents. Using inferential statistical modeling, we demonstrate that the echoicity of a verbal response crucially depends on the speech act of the antecedent. The use of echoic responses is more likely for antecedents in which the speaker displays a low degree of commitment to the truth of the utterance than for antecedents with a high degree of commitment. Our analysis also reveals that it is necessary to distinguish two specific verbal response types e e 'is' and t a 'is' e from other verbal responses. Whereas e has been conventionalized as a multipurpose affirmation particle, t a is typically used to respond to orders or proposals, which is why e and t a are significantly less probable to be used as echoes than other verbal responses.

Insubordination, 2016

When I wrote a dissertation that focused (mainly) on the Spanish conditional marker si 'if' and i... more When I wrote a dissertation that focused (mainly) on the Spanish conditional marker si 'if' and its myriad contexts of use (Schwenter 1999), I didn't realize that I'd stumbled upon what (now years later) people would end up calling "insubordination" (e.g. Evans 2007; Verstraete et al. 2012). While there were already a number of papers on somewhat similar phenomena in other languages in the mid-to late-1990s the term "insubordination" was not being used in the widespread way it is now and there was certainly no vision for a symposium dedicated to the broader phenomenon. In this presentation I will trace the work that preceded mine on insubordination in Spanish conditionals, from the very old, yet extremely insightful comments, of the Venezuelan grammarian Andrés Bello in 1847 to the more modern precursor of my work (Almela Pérez 1985). I will also provide a general sketch of what my findings were on the use of subordinate-marked si clauses as independent declaratives, with an emphasis on their very particular discourse function in dialogue contexts. I will go on to examine in more depth the prosodic properties of "independent si-clauses" (Schwenter 1996), which were not a focus of my prior work. Finally, I will also explicate the consequences of the (supposed) insubordination process for the creation of a new adversative connective si identical in form but very different in function from the conditional marker (Schwenter 2000). The development of this adversative function has led to other, related uses of si that ultimately can be related back to the insubordination process. The more general question for discussion to emerge from my talk will be this: What happens to subordination markers once they begin to appear in contexts of insubordination?

The Handbook of Portuguese Linguistics, 2016

Annual Meeting of the Berkeley Linguistics Society, 2008

A Romance Perspective on Language Knowledge and Use, 2003

A scalar propositional negative polarity item in Spanish Javier Gutiérrez-Rexach and Scott A. Sch... more A scalar propositional negative polarity item in Spanish Javier Gutiérrez-Rexach and Scott A. Schwenter The Ohio State University . Introduction: Propositional NPIs In this paper, we study the semantics and pragmatics of a negative polarity item (NPI) in Spanish that has ...

Studies in Language Variation, 2010

To appear in a volume on Western Ibero-Romance languages, ed. by Manuel Delicado-Cantero, Fernando Tejedo-Herrero, and Patrícia Matos Amaral, 2025

Languages, 2024

Contemporary Trends in Hispanic and Lusophone Linguistics

From Words to Discourse, 2002

Multidisciplinary Approaches to Bilingualism in the Hispanic and Lusophone World, 2017

Variation and Evolution, 2020

The Routledge Handbook of Variationist Approaches to Spanish, 2021

Journal of Pragmatics, 2019

Insubordination, 2016

The Handbook of Portuguese Linguistics, 2016

Annual Meeting of the Berkeley Linguistics Society, 2008

A Romance Perspective on Language Knowledge and Use, 2003

Studies in Language Variation, 2010

OSUCHiLL, 2024

(1) A cara dela é muito linda 'Her face is very pretty' 'Guy/Dude' (Masculine N) (2) Esse cara é ... more (1) A cara dela é muito linda 'Her face is very pretty' 'Guy/Dude' (Masculine N) (2) Esse cara é bastante inteligente 'That guy is quite intelligent' Vocative (used with males or females) (3) Ô cara, vem cá e me ajuda! 'Hey dude, come here and help me!'

OSUCHiLL, 2022

Do native speaker judgments of the naturalness of the presence/absence of the IO clitic follow th... more Do native speaker judgments of the naturalness of the presence/absence of the IO clitic follow the accessibility predictions of Pericchi et al. (2020)? Given that prior studies focus heavily on dar, are there differences between verbs and the naturalness of the presence/absence of the IO clitic? While it is well-known that doubling is nearly always considered natural with IO NPs in Spanish, how natural is the lack of doubling and what conditions favor it? Methods Experimental Design Online experimental survey created in Qualtrics Rating task of "naturalness" of sentences on a 5-point Likert scale Disseminated through social networks and personal contacts to native Spanish speakers

Linguistic Society of America Annual Meeting, 2022

While the Labovian variable is a key construct in variationist research, the locus of social eval... more While the Labovian variable is a key construct in variationist research, the locus of social evaluation of linguistic forms has been argued to be at the level of the individual variant with regard to morphophonological (Campbell-Kibler 2011) and lexico-grammatical (Dinkin 2016; Maddeaux & Dinkin 2017) variables. Yet what happens when examining a set of competing grammatical forms that can be considered to make up one abstract variable in the Labovian sense? In this research, we analyze social evaluation of two competing types of direct object pronouns in spoken Brazilian Portuguese (BP) and demonstrate the importance of examining the roles of both the broader variable in question (e.g., clitic vs. tonic pronouns) as well as the individual variants (e.g., lhe, clitic ‘you’) that instantiate the variable, to better understand the role of social perceptions in linguistic variation.

Anaphoric DOs in BP can be expressed by normative clitic pronouns or non-standard tonic pronouns. Some clitic DOs are not naturally learned, but instead are only acquired in school. Previous work has described a prescriptive preference for clitic pronouns and negative attitudes toward their tonic counterparts (Perini 2010), with the former possessing an implied “correctness” by virtue of their institutional associations (Bagno 2009). We therefore hypothesized that BP speakers will tend to evaluate clitics as more closely related to education and formality, as well as negative social characteristics associated with “correctness.” Nevertheless, we also hypothesized that despite seeing general trends in social perceptions of tonic versus clitic pronouns (i.e., the broader variable), individual variants within each type would also show diverse speaker evaluations. Data were collected through an online survey on Qualtrics where participants evaluated written discourses by fictional speakers via six different adjectives (educated, intelligent, formal, friendly, irritating, pedantic), using a five-point Likert scale. These adjectives were used in dos Santos & Mendes’ (2019) study of the BP subjunctive and adapted for our project. Participants evaluated 32 survey items (16 target, 16 distractors). Target items were presented with 16 scenarios and varied according to pronoun type (clitic or tonic).

The responses of 160 native Portuguese speakers who completed the survey were analyzed using mixed-effects linear regression in R (R Core Team 2017), with survey item and verb as random effects. The dependent variable, the evaluation of fictional speakers via the previously mentioned adjectives, was z-score transformed by survey respondent for each adjective to obviate differences due to scale compression and scale skew (cf. Schütze & Sprouse 2014). The best-fit models for all adjectives included pronoun as a significant predictor of respondent evaluations. Specifically, fictional speakers were statistically significantly more likely to be evaluated negatively (more irritating, pedantic, formal) when using clitic pronouns, as well as of higher education status (more educated, intelligent). By contrast, speakers employing tonic pronouns were consistently rated more positively (more friendly, less irritating and pedantic) but less educated and intelligent than those using clitics. Despite these general trends in perceptions of tonic versus clitic DO pronouns, our analysis also revealed important interactions between social evaluation and individual variants. In particular, our best-fit models for formal, irritating, and pedantic all showed significant interactions between pronoun type and person and number. Most strikingly, we found that the 2SG clitic pronoun lhe was significantly more likely to be evaluated as formal, irritating, and pedantic than all other pronoun variants.

These findings illustrate the critical roles of both traditional Labovian sociolinguistic variables, in this case clitic versus tonic DO pronouns, and also individual variants as sites of variation in social evaluation. We find that there is a broad distinction in evaluation in terms of the variable, i.e., pronoun type, but at the same time there are more fine-grained differences associated with their individual variants. We conclude that both the variable and individual variants can play a role in speaker perception and evaluation.

References

Bagno, Marcos. 2009. Não é errado falar assim! São Paulo: Parábola Editorial.

Campbell-Kibler, Kathryn. 2011. The sociolinguistic variant as a carrier of social meaning. Linguistic Variation and Change 22: 423-441.

Dinkin, Aaron J. 2016. Variant-centered variation and the like conspiracy. Linguistic Variation 16(2): 221-246.

Maddeaux, Ruth, and Aaron Dinkin. 2017. Is like like like? Evaluating the same variant across multiple variables. Linguistics Vanguard 3(1).

Perini, Mário A. 2002. Modern Portuguese: A reference grammar. New Haven: Yale University Press.

Santos, Wendel Silva dos, and Ronald Beline Mendes. 2019. Sounding competent: Effects of mood alternation. Paper presented at NWAV 48, Eugene, Oregon.

Schütze, Carson. T., and Jon Sprouse. 2014. Judgment data. In Research methods in linguistics, ed. Robert J. Podesva and Devyani Sharma, 27-50. Cambridge: Cambridge University Press.

Hispanic Linguistics Symposium, 2021

According to normative descriptions, the alternation in Spanish between the conjunctions y/e and ... more According to normative descriptions, the alternation in Spanish between the conjunctions y/e and o/u is purely phonetically motivated such that the sequences y + [i] and o + [o] require the use of e and u respectively. Mejías-Bikandi (2019) argues that this alternation is also subject to pragmatic restrictions, such that in coordinated sentences y and o are preferred to convey sequence and consequence (in the case of y) or of a preferred alternative (o), while e and u are less acceptable in such contexts, even when the phonetically-based prescriptive norm requires their use. While the validity of these claims is difficult to determine given that they rely on a small set of primarily constructed examples, here we offer an approach based on a larger body of real usage data and taking into account the frequency of individual constructions (Goldberg 2019) to demonstrate that several other factors are also relevant to the selection of y/e or o/u.

We analyze data from the web corpus esTenTen18 (Kilgariff & Renau 2013; 17.5 billion tokens) and show that even in formal written language speakers often use y and o in contexts where prescriptive norms call for the dissimilating allomorphs, as in (1-3). However, speakers also extend use of e and u beyond prescribed domains, resulting in hypercorrections (Labov 1972) as in (4-8). We propose that the association of e and u with formality and adherence to grammatical norms promotes this extension; however, this alone does not account for the disparate rates of hypercorrect e/u across constructions. Instead, in keeping with predictions of usage-based grammar (Bybee 2010), the strong association between u and otro/a/os/as, which stems from the high frequency of this sequence, leads to entrenchment of u with otro even when intervening material renders u non-standard (6-8). The hundreds of tokens of e otro/a/os/as in the corpus demonstrate that constructional association of u with otro also extends to its homolog.

The use of y and o together (e.g. y/o, y/u) shows even more clearly the role of constructional association. The sequence y/o + [o] is more frequent than the normative y/u + [o] (Table 1), thereby suggesting that some speakers treat this conjunction pairing not as two elements, the second of which is subject to phonetic alternation, but rather as an independent construction with a cognitive representation distinct from its component parts. Sporadic instances of e/o (9) with no phonetic motivation further highlight the role of hypercorrection in conjunction choice; again, however, constructional associations are necessary to fully explain the variation observed. The appearance of e/u before [i] (10) is motivated by the fact that when the more restricted allomorph e is used, its counterpart u is also activated, revealing their shared cognitive link.

Our results show that failure to respect the phonetic motivation of e and u is not, as some have claimed, a phenomenon restricted to spoken language (Butt & Benjamin 2013: 454). Instead, it is a usage-based regularity that results in a number of constructional schemata that preserve y and o in phonetically non-normative contexts, as well as license e and u in contexts where they are not phonetically or pragmatically motivated (cf. Mejías-Bikandi 2019). These patterns therefore provide evidence that Spanish speakers create and store constructions that group individual conjunction allomorphs with the words or phrases they most frequently conjoin.

Examples
(1) En medio de crisis, recesiones, desempleo, incertidumbre y incluso angustia, Chile sigue creciendo (lanuevaopcion.cl; 4,107 tokens of y incluso)
(2) Además ofrece servicios de vuelos programados, ‘chárter,’ carga y correo con destinos nacionales y internacionales (aeropuertosmexico.com; 1,255 tokens)
(3) Los centros escolares o otros colectivos que así lo deseen pueden participar en estas iniciativas guiadas (elcorreogallego.es; >21,000 tokens)
(4) El parásito era un pequeño organismo constituido de una sola célula, que generalmente se alimenta de algas, bacterias e que incluso se puede comer a otros como él (revistahypatia.org)
(5) En el '3' hay diferencias en sus trazos de inicio superior e terminación inferior y en las curvaturas, principalmente en la del centro (filateliaargentina.org)
(6) La calificación estará integrada por: participación, casos, exámenes u cualquier otra actividad asignada por los facilitadores (uan.mx)
(7) Cada región ha impuesto en su cultura términos propios para las charlas informales, que en otras no tienen ningún significado u poseen otro (deconceptos.com)
(8) Se ruega también que procuréis dar razones de por qué preferís a uno u a otro (sedice.com)
(9) ...cómo establece el equilibrio entre las necesidades actuales y la proyección de futuro de la empresa e/o institución (vlex.es)
(10) No hay ninguna política que genere condiciones favorables para la procreación de hijos e/u hijas... (derechos.org)

Table 1: Frequencies of use of y/o and y/u according to the following orthographic element
y/u + o*: 9,512 y/u + ho*: 694 Total: 10,206 (49.4%)
y/o + o*: 8,889 y/o + ho*: 1,564 Total: 10,453 (50.6%)

References
Butt, J. & C. Benjamin. 2013. A new reference grammar of modern Spanish. 5th ed. London: Routledge.
Bybee, J. 2010. Language, usage and cognition. New York: Cambridge University Press.
Goldberg, A. 2019. Explain me this: creativity, competition, and the partial productivity of constructions. Princeton: Princeton University Press.
Kilgarriff, A. & I. Renau. 2013. esTenTen, a vast web corpus of Peninsular and American
Spanish. Procedia-Social and Behavioral Sciences 95. 12-19.
Labov, W. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.
Mejías-Bikandi, E. 2019. Gradual conventionalization of pragmatic inferences. The y/e ando/u alternation in Spanish. International Review of Pragmatics 11. 222-234.

NWAV, 2021

Research on Spanish indirect objects (IOs) has focused nearly exclusively on clitic doubling: the... more Research on Spanish indirect objects (IOs) has focused nearly exclusively on clitic doubling: the co-occurrence of a dative clitic (e.g. le) with a coreferring indirect object NP (noun phrase) (1a). By contrast, the lack of doubling in the presence of an IO NP has been called “very rare” (RAE 2009) and set aside as a register phenomenon, used “when a formal tone is required” (Butt & Benjamin 2013). Recent studies looking at written texts, however, reveal that IO doubling (1a) is far from categorical, and in fact non-doubling (1b) remains the overall norm (Aranovich 2016; Pericchi et al. 2020). Our research asks the question: Given that doubling is so frequent in spoken language, why is it so infrequent in written registers?
Data are drawn from the Spanish Web 2018 corpus (esTenTen18; Kilgarriff & Renau 2013). We searched for tokens of eight frequent ditransitive verbs that included overt NP direct and indirect objects in postverbal position. We randomly sampled 300 eligible tokens for each verb (N=2400) to permit greater generalization of our findings (cf. Buchstaller & Khattab 2013). We coded each token for the co-occurrence of a dative clitic and predictor variables (IO number, definiteness, animacy, specificity). Overall, we find clitic doubling only 24.4% (n=562) of the time, but a high degree of variability between verbs. Comunicar showed the lowest rate of clitic doubling (7.4%; n=22), while mandar (36.7%; n=110) showed the highest rate in our sample.
We built a series of stepwise mixed-effect logistic regression models with verb as a random intercept to obviate idiosyncrasies between verbs. Our best-fit model included IO number, definiteness, animacy, and specificity as significant effects. Singular IOs were significantly more likely to co-occur with a dative clitic (26.2% n=421) than plurals (17.5%; n=136). Animate IOs were more likely to occur with a dative clitic (28%; n=393) than inanimate IOs (23.4%, n=77) or human collectives (12.5%; n=84). Specific IOs were more likely to occur with a clitic (25.9%; n=429) than nonspecific IOs (17.2%; n=128). And finally, definite IOs were more likely to occur with a clitic (24.1%; n=505) than indefinite IOs (16%; n=49). Similar factors have been found to influence IO doubling in spoken Spanish (Rinke et al. 2019).
Our data also reveal several constructions that deviate from the probabilistic preferences noted above (cf. Pijpops et al. 2018), where specific verb + direct object (DO) combinations promote enhanced or decreased predictor effects. For example, 9.3%(n=28) of tokens for the verb dar consisted of the dar + lugar collocation. Importantly, these tokens take more inanimate IOs (82.1%; n=23) than does dar with other DOs (57.1%; n=156). Furthermore, 0% of the IOs in the dar lugar construction co-occur with a dative clitic, compared to 27.8% (n=76) of dar tokens with other DOs.
Out findings differ crucially from previous work on several fronts. First, we examine written language as opposed to speech, which bears on the characteristics of the indirect objects themselves. Studies of spoken data have noted rates of specific IOs and animate IOs above 80% (e.g., Rinke et al. 2019); our data show just 69% (n=1654) and 58.4% (n=1401) of specific and animate IOs, respectively. Furthermore, our analysis expands on research that considered one verb only (Aranovich 2016) by considering the idiosyncratic nature of individual verbs as well as constructional differences affecting the kind of IOs with which they co-occur. These considerations allow greater generalization of our findings in order to understand the multiple factors governing clitic (non-)doubling in Spanish.

ExPortLi, 2021

Few grammars acknowledge the full variation in direct object pronouns in Brazilian Portuguese (BP... more Few grammars acknowledge the full variation in direct object pronouns in Brazilian Portuguese (BP), which allows for both normative clitic and non-standard tonic variants. Bagno (2009) notes that clitics are not learned naturally, but only once children go to school, implying the “correctness” but also artificial nature of these pronouns. Perini (2010) describes a prescriptive preference for clitic pronouns and negative attitudes towards their tonic counterparts. Nevertheless, no prior studies have investigated empirically speaker evaluations of these competing pronoun variants. Therefore, in this project we created a perception experiment of direct object pronouns in spoken Brazilian Portuguese, in order to investigate the role of attitudes and social evaluations in language variation (cf. Evans & Iverson 2007; Campbell- Kibler 2009). We hypothesized that it would reveal clear patterns of evaluation towards the variants, and specifically that normative clitic forms would be evaluated more negatively than tonic forms.
Data were collected through an online survey on Qualtrics where participants evaluated fictional speakers in terms of six different adjectives (educated, intelligent, formal, friendly, irritating, pedantic), using a five-point Likert scale. These adjectives were used in dos Santos & Mendes’ (2019) study of the BP subjunctive and adapted for our project. Participants evaluated 32 survey items (16 target items, 16 distractors). Target items were presented with 16 scenarios varied according to pronoun type (clitic, tonic), and utterance type (monologue, dialogue), in order to assess the role of the language internal factors in sociolinguistic evaluation, as well as respondent age, gender, socioeconomic class, level of education, and state of origin.
The responses of 160 native Portuguese speakers who completed the survey were analyzed using mixed-effects linear regression in R, with survey item and verb as random effects. The dependent variable, survey respondents’ evaluation of fictional speakers via the previously-mentioned adjectives, was z-score normalized to account for disparate scale usage. The best-fit model of our data selected pronoun type and utterance choice as significant predictors of respondents’ evaluation of speakers. Specifically, fictional speakers are significantly more likely to be evaluated negatively (irritating, pedantic, formal) when using clitic pronouns, as well as of higher education status (educated, intelligent). By contrast, speakers employing tonic pronouns were consistently rated friendlier and less irritating/pedantic than those using clitics. With regard to specific pronouns, clitic nos was evaluated as more irritating than tonic a gente, while the accusative use of lhe received the most negative evaluations. Regarding utterance type, speakers using clitics were rated more negatively in dialogues than in monologues. None of the external factors were significant.
Given the wide gulf between the grammars of normative BP and actual spoken BP, our study provides a new direction for future research on competing grammatical variants in this variety and their subjective evaluation by native speakers. More generally, our results suggest that the role that social evaluation plays in usage preferences should be re-assessed based on studies linking subjective attitudes with grammatical choices (dos Santos & Silva 2019; Squires 2014).

Verb and Context (Alicante), 2021

Typically when languages have more than one dedicated imperative, the distinct forms correspond t... more Typically when languages have more than one dedicated imperative, the distinct forms correspond to different person/number combinations. But what happens when there is variation between two forms for the same person/number combination, e.g. 2nd person singular? This talk looks at two languages, Brazilian Portuguese (BP) and Argentine Spanish (AS), which illustrate this situation. While AS has this option only in negative imperatives, where vos and tú form imperatives alternate (cf. Johnson 2013, 2016), BP has this option in both affirmative and negative contexts, where the historical imperative tu form for 2SG alternates with the 3SG present subjunctive form (cf. Lamberti & Schwenter 2019). Through a combination of methodologies, including the analysis of naturally-occurring examples and the results of experimental surveys, I show that the variation between the two forms can be accounted for in terms of the semantic-pragmatic notion of presumed settledness (Hoff & Schwenter in press), i.e. speaker confidence about how the future will unfold, and can be operationalized in terms of contextual characteristics such as linguistic markers of certainty, temporal immediacy, and temporal specificity. More broadly, I suggest that these characteristics are relevant to accounting for contrasts between imperative forms across languages such as those with "delayed/deferred imperatives" (Aikhenvald 2014) and to the cross-linguistic observation that prohibitives tend to be more irrealis than imperatives (van der Auwera & Devos 2012).

OSUCHiLL, 2021

Data were collected through an online survey on Qualtrics where participants evaluated fictional speakers in terms of six different adjectives (educated, intelligent, formal, friendly, irritating, pedantic), using a five-point Likert scale. These adjectives were used in dos Santos & Mendes’ (2019) study of the BP subjunctive and adapted for our project. Participants evaluated 32 survey items (16 target items, 16 distractors). Target items were presented with 16 scenarios varied according to pronoun type (clitic, tonic), and utterance type (monologue, dialogue), in order to assess the role of the language internal factors in sociolinguistic evaluation, as well as respondent age, gender, socioeconomic class, level of education, and state of origin.

The responses of 160 native Portuguese speakers who completed the survey were analyzed using mixed-effects linear regression in R, with survey item and verb as random effects. The dependent variable, survey respondents’ evaluation of fictional speakers via the previously- mentioned adjectives, was z-score normalized to account for disparate scale usage. The best-fit model of our data selected pronoun type and utterance choice as significant predictors of respondents’ evaluation of speakers. Specifically, fictional speakers are significantly more likely to be evaluated negatively (irritating, pedantic, formal) when using clitic pronouns, as well as of higher education status (educated, intelligent). By contrast, speakers employing tonic pronouns were consistently rated friendlier and less irritating/pedantic than those using clitics. With regard to specific pronouns, clitic nos was evaluated as more irritating than tonic a gente, while the accusative use of lhe received the most negative evaluations. Regarding utterance type, speakers using clitics were rated more negatively in dialogues than in monologues. None of the external factors were significant.

Given the wide gulf between the grammars of normative BP and actual spoken BP, our study provides a new direction for future research on competing grammatical variants in this variety and their subjective evaluation by native speakers. More generally, our results suggest that the role that social evaluation plays in usage preferences should be re-assessed based on studies linking subjective attitudes with grammatical choices (dos Santos & Silva 2019; Squires 2014).

Null Objects from a Crosslinguistic and Developmental Perspective, 2021

While considerable previous research (e.g. Cyrino & Matos 2002; Kato & Raposo 2005; Duarte & Cost... more While considerable previous research (e.g. Cyrino & Matos 2002; Kato & Raposo 2005; Duarte & Costa 2013) has centered on the syntactic differences between null direct objects in European (EP) and Brazilian Portuguese (BP), Schwenter (2014) and Sainzmaza- Lecanda & Schwenter (2017) have offered analyses of naturally-occurring spoken data from both dialects showing that the underlying constraints on the variable use of null versus overt forms are similar in the two varieties. Indeed, both the constraint types and their ordering in the two dialects are nearly identical, with the obvious difference in the pronominal forms for third person referents across dialects: clitic (EP) versus tonic (BP). Moreover, this research showed clearly that the null object was the most frequent and unmarked option for direct objects in both dialects, in that it corresponded to direct objects with prototypical semantic/pragmatic features (-animate, -specific), while overt forms (clitic or tonic) were employed to mark direct objects with non-prototypical characteristics.

An open question in null object research, however, is whether speakers evaluate the acceptability of null forms vis-à-vis their overt pronominal counterparts in a positive fashion, and what semantic/pragmatic factors affect their judgments. For Portuguese, an additional question is whether, given the similarity in usage constraints across EP and BP (Schwenter 2014), speakers of the two dialects are also parallel in their evaluations of the different forms. Thus, are null objects stigmatized in either dialect? Are clitics ever preferred over null objects by EP speakers? Do BP speakers evaluate null objects as more acceptable than tonic pronouns for human referents?

We tested the hypothesis that null objects for third-person direct object referents in EP and BP are evaluated positively, and overall more positively than their overt pronoun counterparts, using an online experimental survey in Qualtrics consisting of 10 target items (plus 10 fillers) whose acceptability was rated on a seven-point Likert scale. We manipulated the animacy and specificity of the survey referents, following the well-known effects that these features have on anaphoric direct object realization in both dialects (Cyrino 1997; Schwenter & Silva 2003; Schwenter 2014).
Approximately 100 Brazilians and 100 native speakers from Portugal responded to the survey. Mixed-effects linear regression of participants' z-score normalized responses confirmed that, in addition to the research demonstrating the usage parallels between EP and BP (Schwenter 2014), speakers of the two dialects also evaluate null objects in nearly identical and, overall, highly positive fashion. Indeed, even though BP shows a much greater rate of null objects overall (ibid.), EP speakers rate them at comparable levels of acceptability as BP speakers do. Furthermore, while the features of animacy and specificity had clear effects on the acceptability of overt pronouns (tonic in BP and clitic in EP), these features were statistically insignificant when evaluating null objects, which were always rated higher than their overt counterparts. These results thus provide corroborating evidence for a shared grammar of object marking across Portuguese dialects, and more specifically for a system that takes null realization as the preferred variant for all third- person direct objects.

HLS-El Paso, 2019

Espera with this function often appears in the singular, even with plural interlocutors, though t... more Espera with this function often appears in the singular, even with plural interlocutors, though this is not always the case:

HLS-El Paso, 2019

The second-person plural pronouns vosotros and ustedes have received scant attention when compar... more The second-person plural pronouns vosotros and ustedes have received scant attention when compared to their singular counterparts tú/vos/usted. Until Morgan and Schwenter (2016), the purported symmetry between 2SG and 2PL pronouns in Castilian Spanish had never been questioned. They showed that in Spain there is clear asymmetry between the forms: choice of usted in the singular is often mirrored not by ustedes but by vosotros in the plural. Even in Andalusian dialects where both pronouns are found, verbal agreement is overwhelmingly with the vosotros form (e.g. ustedes queréis) and vosotros is replacing ustedes more generally (Jaime Jiménez 2018). Meanwhile, vosotros is not used natively by any speakers in Latin America, where ustedes is the only 2PL pronoun available in naturally-occurring speech.
Here, we examine the roles of the "minority" 2PL pronoun in each region: vosotros in Latin America and ustedes in Spain. Using examples from oral and print sources, as well as qualitative commentary from native speakers from both regions, we show that the functional range of these pronouns is similar, albeit not (yet) identical, in their "foreign" varieties.
Ustedes in Spain is mainly used in formulaic sequences or formal/scripted situations (Morgan and Schwenter 2016). In our online survey of over 200 native speakers of Peninsular Spanish administered via social media, respondent commentary regarding usage of ustedes was of two types. First, respondents failed to comprehend the question about the use of the 2PL form, and instead interpreted it as referring to the 2SG form, leading to comments about what kinds of individuals and/or situations would merit tú or usted. We interpret this result as a kind of "plural blindness" on the part of speakers, who appear to have little access to intuitions about use of ustedes. Of course, it may be that these respondents use ustedes so rarely that they don't have clear intuitions to share, and these were the second type of comments. Speakers stated that they "never" use ustedes, that it is restricted to formal or classroom situations, or even that it was typical of Latin American speakers, and by implication not identifiable as a "native" Peninsular feature.
Vosotros in Latin America has even less currency than ustedes in Spain, of course, but the parallels are noteworthy. While verb conjugations corresponding to the former are simply not productive, speakers are still exposed to them. Not surprisingly, speakers evaluate these pronominal and verb forms as elegant, deferential, and archaic (Morgan et al. 2017)—quite the opposite of their value in Spain (but ironically aligning with Peninsular reactions to modern voseo, in spite of the etymological and even phonological overlap of the two forms). Possessive vuestro does enjoy some very limited productivity in Latin American Spanish, but accompanies ustedes and its 3PL morphology rather than the erstwhile vosotros (or os). This one morphological relic of an otherwise obsolete 2PL system thus takes its place beside European Portuguese possessive vosso, which similarly outlived its historical 2PL paradigm and was recast to accompany vocês.
Our findings buttress the argument that 2PL pronominal forms are not mirror images of their 2SG counterparts. Indeed, all modern Ibero-Romance varieties show a strong preference for just one 2PL pronoun in spite of the shared history of an apparently symmetrical paradigm (Morgan and Schwenter 2018). Nevertheless, they reserve a space in their grammar for other forms that, while limited in productivity, never quite disappear.

INAR 05-Sheffield, 2019

Morgan and Schwenter (2016, 2018) showed that Romance languages with both singular and plural T/V... more Morgan and Schwenter (2016, 2018) showed that Romance languages with both singular and plural T/V pronouns are asymmetrical, in that the use of the plural V forms is consistently less frequent and more constrained than the corresponding singular forms. For example, Castilian Spanish speakers, who have the vosotros (T) and ustedes (V) distinction in the plural, often select the T form when the group of addressees is mixed between T/V in the singular, and may even do so when all the individuals in a group would be addressed as V in the singular.
Here, we examine two seemingly intractable cases for this research: European Portuguese (EP) and Romanian. In EP, speakers often use"descriptive pronouns" (Amaral 2018), 3PL NPs employed as 2PL V forms, as in Os pais querem café? ('Do you [the parents] want coffee?'). Speakers also use an "avoidance" strategy in the plural: since both the T (vocês) and the V pronoun (os senhores, os pais, etc.) are conjugated as 3PL, speakers can and often do "opt out" of choosing a pronoun. Romanian, like French, allows a 2PL pronoun (and 2PL conjugation) to do double duty as a 2SG V form, but the paradigm is further complicated by two additional 2PL pronouns, one of which variably takes 2PL or 3PL verbal morphology. These facts, together with the null subject preference found in certain contexts in Ibero-Romance, rendered previous results unconvincing. For both EP and Romanian, therefore, we designed a new survey instrument, recasting the address forms in the scenarios as objects of prepositions, where pronoun avoidance would be ungrammatical.
Our results permit more accurate comparisons with the other Romance languages analyzed previously. We illustrate with added clarity that, despite the other options available in EP and Romanian, both languages show asymmetries similar to those uncovered in their Romance brethren.

OSUCHiLL, 2019

Research on related (imperative) DMs 7 English Spanish Other languages Brinton (2001): look (also... more Research on related (imperative) DMs 7 English Spanish Other languages Brinton (2001): look (also see, listen) Pons Bordería (1998): oye, mira Waltereit (2002): Italian guarda Tagliamonte (2019): wait Fagard (2010): mira Fagard (2010): Portuguese olha, Italian guarda, French regarde, Romanian uite Oliveira (2015): Portuguese espera aí / peraí