DEGREES OF SUBJUNCTIVE VITALITY AMONG MONOLINGUAL SPEAKERS OF PENINSULAR AND ARGENTINIAN SPANISH

This investigation seeks to expand the current understanding of mood use and vitality among monolingual native speakers of two varieties of Spanish. It focuses on establishing a cross-dialectal analysis of mood use in semi-spontaneous oral production. A total of 112 adult monolingual native speakers of Spanish participated in the study (N=56 from Rosario, Argentina and N=56 from Toledo, Spain). A controlled interview was administered to elicit the frequency and vitality of subjunctive use. Results indicate that the use of present subjunctive in complement clauses was found stable and vital among older speakers (51-73 years old) in both regions. However, while the youngest generation (18-30 years old) shows prevalence of non-subjunctive forms in Rosario, this age group favors the use of present subjunctive in Toledo. A similar pattern is observed with middle-aged participants. Results indicate an overall reduction of present subjunctive use among younger generations in Rosario, and vitality in every age group in Toledo.


Introduction
A recent proliferation of studies investigating mood use as determined by linguistic and sociolinguistic variables such as gender, education, socio-economic status, and geographical region have indicated both the debilitation and the vitality of subjunctive forms.With regard to varieties of Spanish spoken in the U.S., nonsubjunctive forms have been found predominant, a phenomenon that has been interpreted -in most cases-as a sign of simplification (Lynch 2000;Martínez Mira 2010;Mikulski 2010;Montrul 2009;Silva-Corvalán 1991, 1994, 2001).Similarly, some varieties outside of the U.S. also have shown debilitation in the use of past forms and prevalence of the indicative in contexts in which the subjunctive would have been expected (López Morales 1989;Molina 2008;Serrano 1995).Contrarily, maintenance and vitality of subjunctive use has been observed in several regional varieties (Fernández-Ulloa & Portillo Mayorga 2000;García 2011;Gudmestad 2006Gudmestad , 2010Gudmestad , 2012;;Geeslin & Gudmestad 2010;Kowal 2007;Lastra & Butrageño 2012).
The present paper seeks to expand current understanding of mood use among monolingual native speakers (NSs).It analyzes monolingual speakers' behavior concerning mood alternation in two varieties that are currently underexplored, namely central Argentinian Spanish and central Peninsular Spanish.The study focuses on determining frequency of use of subjunctive forms in subjunctivetriggering contexts in semi-spontaneous oral production.The variables under evaluation are tense, gender, social status, and generation.
Concerning the broad category of Spanish in the U.S., specifically those varieties spoken by heritage speakers, previous research has identified a gradual trend of mood simplification, which is explained by the predominance of indicative forms in contexts in which subjunctive forms would be expected (Lynch 2000;Martínez Mira 2010;Mikulski 2010;Montrul 2009;Silva-Corvalán 1991, 1994, 2001).The process of simplification has been described by Silva Corvalán (1994) as the occurrence of a given form in contexts in which another form is expected.Silva Corvalán (1994aCorvalán ( , 1994bCorvalán ( , 2001) ) claims that if the simplification process is completed, certain forms are lost, resulting in a simplified system, which may or may not render meaning loss.
Debilitation of subjunctive use has also been reported among speakers of Latin American and Peninsular varieties.Molina (2008) observed the prevalence of nonsubjunctive forms among speakers of the northern Castilian region, in Spain.Results indicated a duplication of the conditional both in protasis and apodasis.While the conditional is normative in the apodasis, the protasis is a context in which the subjunctive would be expected.The study also indicated subjunctive use in contexts in which only the imperfect of the indicative mood would be typically accepted.Serrano (1995) identified a similar pattern among older female speakers of La Laguna (Spain), in which the imperfect indicative appears in lieu of the present or imperfect subjunctive as in: (1) Si tenía doce hijos, los atendía a todos and (2) Si fuera más jóven, yo me lo compraba.This author pointed out that the overgeneralization of subjunctive use is also evident in Latin American varieties (Bello 1860;Serrano 1991).
In contrast, García (2011) observed high frequency of subjunctive forms in adverbial constructions with quizá and quizás.She analyzed six different regional varieties as registered in CREA (Corpus de Referencia del Español Actual) [Reference Corpus of Current Spanish].Results revealed that the adverbs quizá and quizás do not behave as synonyms and that temporal reference has a major impact on both adverb selection and mood selection.In the case of future temporal reference, when using quizás, the subjunctive is preferred, while it is disfavored when using quizá.Nevertheless, while both adverbs behave similarly with past and present temporal references, the subjunctive is favored in present constructions and disfavored in past constructions.With regard to the variable regional variety in determining mood selection, Argentina was placed on the higher end of the spectrum favoring subjunctive use; while Venezuela was the region that least favored it.
Similarly, the imperfect subjunctive appeared to display high levels of vitality in a study conducted by Fernández-Ulloa & Portillo Mayorga (2000), which analyzed university entry exams.Results indicated that subjunctive use was not determined by any of the variables under evaluation (gender, place of origin, and monolingual vs. bilingual speakers).Morphologically, it was established that -ra was common in the spoken domain, whereas -se was found more frequently in the written domain.
Subjunctive vitality was also found in studies conducted with NSs currently residing in the U.S. (Geeslin & Gudmestad 2010;Gudmestad 2010Gudmestad , 2012)).However, in terms of frequency, Geeslin and Gudmestad (2010) indicated that their native speaker (NS) participants favored non-subjunctive forms significantly more than subjunctive forms.A total of 8,716 tokens were identified for the data collected during an oral elicitation interview, of which 586 (6.7%) were subjunctive forms.Gudmestad (2010) analyzed whether NS mood selection was influenced by variables such as semantic categorization, irregular morphology, time reference, and hypotheticality, by means of three tasks.Results indicated that NSs were more prone to use the subjunctive with irregular forms (60.5%).In addition to word-level results, she found that participants used the subjunctive more frequently in constructions of volition or complement clauses (94.8%).Lastra and Butragueño (2012) have also found that the subjunctive was widely used, particularly by young, educated women.Their analysis of Corpus sociolingüístico de la Ciudad de México [Sociolinguistic Corpus of Mexico City] indicated that two-thirds of subjunctive occurrences were in the present tense, and there was a significant predominance of subjunctive use in complement clauses.Gudmestad (2006Gudmestad ( , 2010Gudmestad ( , 2012) ) and Silva Corvalán (1994) also noted this predominance, indicating that volitional constructions (such as Mi hermano quiere que le compres un regalo para Navidad) are the most stable, frequent, and robust across varieties.

The study
This investigation seeks to establish whether there are dialectal differences in the frequency of use of the present subjunctive in complement clauses in oral semispontaneous production.It specifically focuses on determining whether subjunctive forms are favored in contrast with non-subjunctive forms in contexts in which the subjunctive would be expected.This study evaluates both linguistic variables, such as frequency of use of complement clauses with the present subjunctive -and sociolinguistic variables, such as gender, generation, and region.The following research questions guided this investigation: -Is subjunctive use vital in semi-spontaneous oral production as indicated by NSs in Argentina and Spain?-Are degrees of frequency of subjunctive use contingent upon extralinguistic factors such as age, gender, region?

Participants
A total of 112 participants recruited in Toledo, Spain and Rosario, Argentina took part in the study.Participants were men and women ranging from 18 to 73 years of age.They were monolingual speakers of their respective varieties of Spanish who resided in regions where Spanish is not in contact with any other languages.

Data Collection and Analysis
Private face-to-face interviews were conducted by the researchers.After participants received information about the nature of the investigation, they were asked to complete a demographic questionnaire.
As part of the subjunctive eliciting interview, participants were provided with a written prompt containing a scenario (see appendix), a set of five questions designed to elicit the use of subjunctive in complement clauses, and four distracter questions.The prompt was adapted to reflect dialectal differences in reference to voseo or tuteo in the respective regions.
For the data analysis, participants were distributed in three age groups (18-30, 31-50, and 51-73) to capture generational differences.Additionally, since eight participants in the Toledo group considered themselves bilingual, they were excluded from the data poll.
Data were coded [+SUBJ] or [-SUBJ] for each answer provided by the participants.Each question in the prompt (a total of five) was considered a subjunctive triggering token.All [-SUBJ] items were further coded as infinitive or hypothetical constructions, as these forms appeared to have been frequently used in lieu of subjunctive forms.

Results
A total of 934 utterances were analyzed.Each complement clause-triggering question was considered an opportunity for subjunctive production.Of the possible [+SUBJ] tokens, the total number of utterances coded as [+SUBJ] was 399.These utterances were constructions of the type: (1) Subjunctive in complement clauses: Interviewer: _Entonces,¿Qué le recomendás/recomiendas? Participant: _Le recomiendo que hable con su jefe.Eso primero, después tendrá que ver qué pasa.
The majority of those utterances (N=252) were produced by participants in Toledo, and fewer (N=147) were produced by participants in Rosario.Simple frequency calculations determine that 63% of the complement clauses with present subjunctive were produced by speakers from Toledo and 37% were produced by speakers from Rosario.This preliminary analysis indicates that speakers in Toledo favored the use of the subjunctive in complement clauses when a complement clausetriggering context was presented to them, while speakers in Rosario used the subjunctive in complement clauses significantly less, and favored other constructions as well as non-subjunctive forms.
A repeated-measure ANOVA was performed afterwards, with the three types of utterances previously identified as the within factor, and age (3 levels), gender (2 levels), and region (2 levels) as the between factor.
The use of the present subjunctive in complement clauses was found stable and vital in older speakers (51-73 years old) both in Toledo and Rosario, as most complement-clause triggering questions were answered with constructions such as those seen in (1) above.Conversely, results indicate debilitation of subjunctive use, particularly in younger speakers.This trend is significantly more apparent in Rosario than in Toledo.
In the youngest group (18 to 30 years old), participants from Toledo favored the use of the present subjunctive in complement clauses, followed by the infinitive and hypothetical constructions.Contrarily, participants from Rosario favored the use of the infinitive, followed by hypothetical constructions, and the present subjunctive in complement clauses.This is consistent with the overall regional pattern: a more conservative tendency toward the norm in Toledo.
The higher percentage of answers containing constructions with infinitives, such as Le recomiendo hablar [I recommend her/him to talk], in the youngest group of speakers from Rosario does not indicate a semantic change.However, it could indicate not only a debilitation of subjunctive use, but also a syntactic simplification.The frequency of [+SUBJ] and [-SUBJ] utterances among the youngest participants (Age Group 1) is graphically illustrated in Figure 1, in which the label utterance 1 indicates constructions with the subjunctive as in (1), utterance 2 indicates constructions with the infinitive as in (2), and utterance 3 indicates hypothetical constructions as in (3).

Figure 1. Token distribution in younger participants
With regard to the middle-aged participants (31 to 50 years old), those in Toledo also significantly favored the present subjunctive in complement clauses, followed by the infinitive and hypothetical constructions.In Rosario, middle-aged participants produced a similar number of complement clauses with the present subjunctive (for which the mean (M) was 2.13) and constructions with the infinitive (M=2.04), and considerably less hypothetical constructions (M=0.75).The frequency of [+SUBJ] and [-SUBJ] utterances among middle-aged participants (Age Group 2) is graphically illustrated in Figure 2, in which constructions with subjunctive, as in (1) are labeled under utterance 1, utterance 2 indicates constructions with the infinitive as in (2), and utterance 3 indicates hypothetical constructions as in (3).

Figure 2. Token distribution in middle-aged participants
Contrasting behavior is observed in the older participants.While there is a significant effect of region in the younger groups, the older participants produced a high number of complement clauses, regardless of their place of origin.Most of the older participants in both Toledo and Rosario produced constructions with the present subjunctive instead of other semantically equivalent or semi-equivalent alternatives.
The only discrepancy identified between older participants in Toledo and older participants in Rosario was that constructions with the infinitive were the prevalent second choice in Rosario, whereas participants in Toledo favored hypothetical constructions when not using the subjunctive in complement clauses.The behavior of the older participants in Rosario is similar to that of the middle-aged group in this same geographical area, with regard to second and third choices.The frequency of [+SUBJ] and [-SUBJ] utterances among the oldest participants (Age Group 3) is graphically illustrated in Figure 3, in which the label utterance 1 indicates constructions with the subjunctive as in (1), utterance 2 indicates constructions with infinitive as in (2), and utterance 3 indicates hypothetical constructions as in (3).

Discussion and future research
Previous research on subjunctive use has indicated different trends observed in monolingual and bilingual speakers.Overall, the non-categorical use of the subjunctive among NSs has been established.Our findings are in line with previous research, since they also reveal a non-categorical use of the subjunctive in monolingual speakers.Results show a distinct pattern in each one of the two regions analyzed.While the use of the present subjunctive in complement clauses is vital in Toledo, it is significantly less frequent among participants in Rosario.All age groups in Toledo responded to complement clause-triggering questions using mainly the present subjunctive in complement clauses, as in (1), whereas only the older participants in Rosario significantly preferred that construction.Debilitation in the use of the present subjunctive is observed in Rosario among middle-aged speakers, and reduction among younger speakers, for whom the present subjunctive in complement clauses comprised only 1 in 5 possible tokens (M=1).These findings are in line with previous research pointing to a gradual loss of the subjunctive (Martínez Mira 2010;Mikulski 2010;Montrul 2009;Silva-Corvalán 1991, 1994, 2001).Further research could explore whether there is vitality or debilitation of subjunctive use in other constructions, and incorporate additional extralinguistic variables, such as socioeconomic status or education, in the analysis.
The pervasiveness of constructions with the infinitive among younger speakers is noteworthy.For participants from Toledo, complement clause-triggering questions were mainly answered with complement clauses; however constructions with the infinitive were the second most frequent choice.In Rosario, on the other hand, young and middle-aged participants primarily answered complement clause-triggering questions with constructions using the infinitive.We can argue that semantically both constructions are almost identical: Le recomiendo hablar con su jefe vs. Le recomiendo que hable con su jefe, and the use of the infinitive does not posit a meaning loss.Syntactically, the use of the infinitive is often a resource to avoid subject co-reference.For instance, when the subject in the matrix coincides with the subject in the subordinate clause (*Recomiendo que yo haga ejercicio a diario), the construction is generally not accepted (Kempchinsky 1986(Kempchinsky , 2009)), and one way to circumnavigate the co-reference is to use the infinitive instead (Recomiendo hacer ejercicio a diario).This alternative, however, is not without ambiguity, given that it could carry meaning dissymmetry.While in *Recomiendo que yo haga ejercicio a diario, the subject+agent in the matrix is recommending something for him or herself, in Recomiendo hacer ejercicio a diario, the subject+agent in the matrix could be recommending something in general, not just to him/herself.
One last consideration could be made regarding the use of hypothetical constructions such as those previously seen in (3).According to Sánchez-Naranjo (2014), this type of conditional structure that may have the imperfect subjunctive in the protasis and the conditional in the apodosis indicates that the events described are counterfactual, or false (2014).In the case of constructions such as: Interviewer: _¿Qué le sugerís/sugieres?Participant: _Y…yo le diría que primero escribiera una carta a recursos humanos.
the event is not counterfactual, but rather hypothetical.Moreover, the hypotheticality implied could be interpreted as the inability of the speaker to imagine the situation described as real.This could be a limitation in our interview instrument.The prompt was not formulated to characterize the scenario as a non-real situation (what would you do if…/imagine that…), but rather to help the participant think about the scenario as something that is currently happening to him/her.We can speculate that responses such as those seen in ( 3) above could be attributed to the inability of the participant to think about the scenario as a real, current situation.An alternative interpretation is that, for certain participants, there was certain degree of equivalence in the meaning conveyed by more factual answers comprised of complement clauses and less factual answers comprised of hypothetical constructions.
The overextension of the subjunctive in protasis and apodosis (Molina 2008;Serrano 1991Serrano , 1995) ) was not observed in our participants.Further research is needed to determine the pervasiveness of this phenomenon that has been reported in previous research.

Conclusion
This study examined subjunctive use in complement clauses by monolingual NSs of Spanish of two regional varieties.It represents the first study that reports results concerning the use and vitality of the present subjunctive in complement clauses among speakers of Argentina and Spain in oral semi-spontaneous production.The study includes a comparison of the frequency of subjunctive use, as determined by region, generation and gender.Results indicate that the use of the present subjunctive in complement clauses is prevalent over alternatives such as the infinitive or hypothetical constructions in Toledo.Thus, it is a vital, robust phenomenon in this area.However, young and middle-aged participants in Rosario use non-subjunctive forms more frequently.Therefore, there seems to be an overall debilitation of present subjunctive use in complement clauses among younger generations in Rosario, while a strong presence of present subjunctive use in complement clauses in every age group in Toledo.

Figure 3 .
Figure 3. Tokens distribution in older participants

Table 1 .
Demographic information