ABSTRACT: while less study has been done on the relation of gender to syntax than to some other levels of language, such as phonology and pragmatics, some evidence indicates that gender might be relevant in adverbial clause usage in English. This study investigates the relationship of speaker and interlocutor gender to the usage of four types of adverbial clauses in Spanish in an ex- perimental setting. The results show that women and men in sin- gle-gender dyads produce very similar proportions of each kind of adverbial clauses. In mixed-gender dyads, a comparison of wo- men and men again shows a very similar proportion of clauses. However, when comparing the single-gender to the mixed-gender speech we see that both men and women use more conditional clauses in the mixed-gender context. These preliminary results suggest that the function of adverbial clauses with regard to gen- der is much different in Spanish than in English.

KEYWORDS: adverbial clause, gender, syntax, mixed-gender, sin- gle-gender

RESUMEN: si bien se han realizado menos estudios sobre la re- lación del género con la sintaxis que con algunos otros niveles del lenguaje, como la fonología y la pragmática, algunas pruebas indican que el género podría ser relevante en el uso de cláusu- las adverbiales en inglés. Este estudio investiga la relación del hablante y el género interlocutor con el uso de cuatro tipos de cláusulas adverbiales en español en un entorno experimental. Los resultados muestran que las mujeres y los hombres en díadas de un solo género producen proporciones muy similares de cada tipo de cláusulas adverbiales. En las díadas mixtas, una comparación de mujeres y hombres nuevamente muestra una proporción muy similar de cláusulas. Sin embargo, al comparar el discurso de un solo género con el género mixto, vemos que tanto hombres como mujeres usan más cláusulas condicionales en el contexto de géne- ro mixto. Estos resultados preliminares sugieren que la función de las cláusulas adverbiales con respecto al género es muy diferente en español que en inglés.

PLABRAS CLAVE: cláusula adverbial, género, sintaxis, género mixto, género único

INTRODUCTION

A considerable body of research has been developing on the performance of gender through language in various speech communities. (For an overview of the field, see Eckert and McConell-Ginet, 2013 and Holmes & Meyerhoff.

2008, inter alia). However, significant gaps remain in our knowledge of this topic. For one thing, language levels have been investigated unevenly. For example, while a good deal of research is building up in gender-linked phonological variation, very little has been studied on gender-linked syntactic variation. (For some of the rare examples, see Aijon Oliva 2013 and Cheshire 2005.) Another imbalance concerns work in various languages. The performance of gender in English has been studied much more than in other languages. Work such as Elhindi and McGarry (2013) and the Gender Across Languages series shows that considerable progress is being made on some other languages, such as Japanese; however a great deal more cross-linguistic work is needed. This study attempts to address these gaps by investigating adverbial clauses in relation to gender in Spanish.

Adverbial clauses and gender in the literature. Our choice of this particular aspect of syntax is motivated by the work of Britta Mondorf. Mondorf (2004) reports an investigation of the London Lund corpus, a collection of naturally-occu- rring British English, containing 100 spoken texts of 5000 words each. She focused on four types of adverbial clauses: causal, conditional, purpose, and concessive. This taxonomy is explained in Section 2.3, below; see Table 1 for exam- ples. Comparing the frequencies of these clauses in the speech of women and men, she found that the women used more causal, conditional, and purpose clauses than the men, while the men used more concessive clauses. Mondorf ’s explanation for this difference is that women use the three clause types that mitigate the speaker’s commitment to the truth of the proposition, while men tend to use more concessives, which strengthen the commitment. Thus, the women use adverbial clauses to perform femininity by referencing tentativeness, while men use them to perform masculinity by referencing certainty.

A further question she investigated concerned the gender composition of the group. Previous research in the social psychological framework (e.g. Takano 1998, Hogg 1985, Mulac et al. 1988) had found that many features linked to the

FECHA DE RECEPCIÓN:

JUNIO 2017

FECHA DE ACEPTACIÓN:

NOVIEMBRE 2017

performance of gender tended to occur more frequent- ly in same-sex interaction than in mixed-sex interaction. That is, speakers tend to match their speech to that of their conversational partners by using the gender-linked features more frequently with interlocutors of the same sex and less frequently with interlocutors of the opposite sex. However, other research such as Brownlow, Rose- mond, & Parker (2003) and McGarry (2004), shows that this phenomenon does not apply to all language features in all situations. The opposite pattern has also been ob- served, where speakers in mixed-sex conversation heigh- ten certain differences, i.e. the women use features linked to the performance of femininity more in the mixed-gen- der context than in the single-gender context, and/or the men in the mixed context increase their use more of the features linked to masculinity, compared to all-male speech. For more discussion on this issue, see Weatherall and Gallois (2003). In the corpus, Mondorf found this pa- ttern: differences in adverbial clause usage that appeared when she compared the all-female and all-male speech generally appeared more intensely when she compared the women’s and men’s speech in the mixed-gender con- text. This was true with regard to casual, purpose, and concessive clauses. Her analysis indicated that both se- xes participated in this divergence; i.e. the men increased their use of concessive clauses and reduced their use of causal and purpose clauses, and the women decreased their use of concessive clauses and increased their use of causal and purpose clauses.

Given these general findings in this large English cor- pus, we now ask whether the effects are replicable with regard to other languages. To study the relation to both speaker gender and group composition, we elicited data from native speakers of Spanish in single-gender and mixed-gender dyads and compared the frequency of the four types of adverbial clauses they produced. Starting with the assumption that language features are inherent- ly indeterminate and can be used for many different pur- poses in many different contexts, we attempt to ascertain if the frequency patterns relating adverbial clause usage to social gender in English also appear in Spanish and, if so, whether the clauses are being used to perform similar functions.

The results suggest that Spanish is quite different from English in this regard. In the single-gender dyads, there were no differences in the frequencies of the clauses be- tween the male and female dyads. In the mixed-gender dyads, the speech of the women and the men also exhi-

the mixed-gender (MG) dyads were the same ones that participated in the single-gender (SG) dyads. Each dyad was given a conversation prompt on a topic chosen to be accessible to all participants and not associated with a specific gender and asked to converse for at least five minutes. To avoid any effect of participants discussing the same topic twice, we gave different prompts to the different types of groups. The single-gender dyads were given this prompt:

Dentro de los valores que tenemos como seres humanos,

¿cuál es su concepto de la amistad?

The mixed-gender dyads were given this prompt:

Hable por lo menos durante 5 minutos sobre el carnaval de

Guaranda y lo que significa para su pueblo.

2.2. Analysis

The conversations were audio-recorded and transcribed. We then identified every instance of the four types of adverbial clauses, following Mondorf (2004). The three criteria she sets out for adverbial clauses generally, with reference to English, are these:

the clause must be introduced by a subordinating con-

junction, e.g. because, unless, or in order that.

The clause must have a subject and a finite verb.

The clause must not be a subject or an object in a bigger clause.

She then establishes a taxonomy of four types of adver- bial clauses, based on meaning, and provides a list of su- bordinators that can begin each type of clause. For exam- ple, causal clauses can begin with because, since, for, as, seeing that, as much as, or in case. To obtain a parallel list of Spanish equivalents, we relied on meaning and syntax. Examples of the four kinds of clauses from our data are shown in Table 1.

TABLE 1. Examples of adverbial clause types from data

causal … por qué no decirles del mundo entero a que visiten mi provincia ya que es una de las mejores del Ecuador en lo que es la tradición del carnaval? ‘why not tell the whole world to visit my province because it’s one of

the best of Ecuador in the tradition of carnival?’

sex interaction but it does not necessarily index either femininity or masculinity.

METHODOLOGY

Participants and data collection

The participants in the study were undergraduate stu- dents in their early 20s at a public university in Ec- uador. All were native speakers of Spanish; some also spoke one or more indigenous languages. We arranged them in three groups: 13 female dyads, 12 male dyads, and 16 mixed-gender dyads; most of the speakers in

purpose … nosotros tenemos que seguir con esas cosas para que no se pierda la tradición.

‘we have to continue those things so that we don’t lose tradition.’

concessive …amigo verdadero es el que aunque a uno le duela le dice las verdades …

‘true friend is one who even though it hurts tells you the truth’

Figure 4. Clauses in Men’s Speech in SG and MG Dyads

Once the clauses were identified, we compared the obser- ved counts to the expected counts. The expected counts, based on the assumption that there would be no diffe- rence between male and female usage, were obtained by multiplying the observed counts, relative to speaker or setting, by the proportion of words attributable to the speaker gender or setting. For example, in the SG dyads, women produced 7734 words and men produced 8221, meaning that women produced 48.47% of the words produced in the SG setting. There were 126 causal clau- ses produced in the SG setting. Therefore, the expected number for the women’s causal clauses in the SG setting is .4847 x 126 = 61.07. If the women actually produced more than this number, their proportion of these clauses is higher than the men’s. Table 2 shows the total number of words obtained.

TABLE 2. Words produced by speaker gender and context

SG dyads MG dyads Total

Women 7734 5706 13440

Men	8221	5666	13887
Total	15955	11372	27327

Hypotheses

Based on previous research, we make the following hy-

potheses:

In the MG interaction, the women will produce more cau- sal, conditional, and purpose clauses than the men, and the men will produce more concessive clauses than the women.

In the SG settings, the same differences will appear, but they will not be as great.

The women will produce more causal, conditional, and purpose clauses and fewer concessive clauses in the SG setting than they do in the MG setting. The men will pro- duce fewer causal, conditional, and purpose clauses and more concessive clauses in the SG setting than they do in the MG setting. Thus, in the SG settings, the differences will be heightened, and in the MG settings the greater si- milarity will be attributable to change on the part of both genders.2.3.

RESULTS

Mixed-gender context

In the aggregate of the MG dyads, 11,372 words were produced. The women’s speech accounted for 50.18% (5706 words), while the men’s speech accounted for

49.82% (5666 words). The observed actual values (AV) and expected values (EV) for each kind of clause, based on these figures, are shown in Figure 1.

Figure 1. Clauses in mixed-gender dyads

Differences between men’s and women’s speech in this context are small and statistically insignificant. Thus, Hypothesis 1, that in the MG interaction the women will produce more causal, conditional, and purpose clauses than the men, and the men will produce more concessive clauses than the women, is not supported.

The number of purpose and concessive clauses is very small, which may reduce observable effects, but the fin- ding regarding those two types is concordant with the finding regarding causal and conditional clauses.

Single-gender context

In the aggregate 15,955 words of the SG dyad speech, the ratio of women’s speech is 48.47% (7734 words), as mentioned above, while the men’s speech constitutes

51.53% (8221 words). Figure 2 shows the actual and ex-

pected counts for each kind of clause in this setting.

Figure 2. Clauses in single-gender dyads

Again, we see very similar proportions for women and men regarding each kind of adverbial clause. Thus, Hypo- thesis 2 is also disconfirmed. In neither female nor male SG speech does any kind of clause appear to be more pre- valent.

Comparing the single-gender and mixed-gender setting

Recall that the third hypothesis is somewhat contingent on the first two.

The prediction that the women will produce fewer of the proposition-mitigating clauses and more of the propo- sition-strengthening clauses in the MG context, and the men will do the converse is based on the premise that the women and men when speaking together will decrease the differences between their SG speech.

Given that no differences have been found in the clause usage of women and men in either context, this hypothe- sis is problematic.

However, to find any patterns related to differences of ei- ther gender between the two setting types, we compare the speech of women in the SG context to women in the MG context and men in the SG context to men in the MG context.

Of the 13,440 total words produced by the women,

57.54% (7734) were produced in the SG context, and

42.46% (5706) were produced in the MG context.

Figure 2 shows the actual and expected numbers of clau-

ses produced by the women in the two settings.

Figure 3. Clauses in Women’s Speech in SG and MG Dyads

The apparent tendency for the women to produce more causal clauses in the MG context is not statistically sig- nificant. However, the difference concerning conditional clauses is significant at p<.01. The women used conditio- nal clauses more in the mixed dyads than in the female dyads.

Of the total 13,887 words produced by the men, 59.20% (8221) were produced in the SG context, and 40.80% (5666) were produced in the MG context. Figure 4 shows the actual and expected numbers of clauses produced by the women in the two settings.

Figure 4. Clauses in Men’s Speech in SG and MG Dyads

A pattern very similar to that in the women’s speech appears. Again, there is a non-significant trend towards producing more causal clauses in the MG context and a significant (p<.005) increase in the number of conditio- nal clauses in the MG context. Thus, Hypothesis 3, which predicted converse trends between the women and the men, is not supported. Rather, they exhibit parallel trends, resulting in their exhibiting similar usage frequencies in both types of context.

DISCUSSIONS

This study has attempted to explore whether Spanish speakers use adverbial clauses to perform gender in a way that resembles the strategies that English speakers use. Striking differences have emerged. While previous research suggests that among English speakers women use clauses to index tentativeness with regard to the pro- positions they express and men use them to express con- fidence, no indication of these phenomena appear in our Ecuadorean Spanish data. The rates at which men and women use each type of adverbial clause when talking to each other is extremely similar, as are the rates evidenced when comparing female and male single-gender speech.

The theoretically possible interpretation that Ecuado- rean Spanish speakers do not perform gender through language is not tenable in light of research such as that found in the edited volume of Rincón (2004). Nor does it seem likely to us that gender was markedly less salient in the experimental dyads than in normal conversation. Rather, we believe that the results suggest that despite the syntactic similarity of English and Spanish gram- matical clauses, the Spanish speakers perform gender more by means of other linguistic features. Whether we can go further and say that the tentativeness-confidence dimension is less important in gender roles in Ecuador than in Britain is not clear to us, given our current state of knowledge. It seems equally likely that these charac- teristics are comparable in importance but are indexed through different language features. This question might be studied through further research examining other fea- tures previously found to index tentativeness in English and other languages, such as tag questions. Moreover, Mondorf examined other aspects of adverbial clauses, such as their sentential position in regard to the main clause and their epistemic categorization. More detailed analysis of our data and other Spanish data in regard to these aspects might yield a different understanding,

A very interesting clue as to motivation for the Spanish speakers in our study to use the adverbial clauses appears when we note that the similarities in usage do not hold within genders across interlocutor type. Rather, the wo- men and men achieve the similarities between their sa- me-gender and cross-gender conversation by effecting parallel increases in conditional and causal clauses. Whi- le the causal clause difference is statistically insignificant for both genders, as well as for the genders combined, the similarity of the increase between genders is still striking. The results are different from Mondorf ’s in that in her data the women used fewer rather than more of these clauses in the MG setting. However, in another sen- se our results are similar because they suggest that the speakers match their speech to that of their partner. This of course still leaves open the question of what function the adverbial clauses are performing, but it suggests that in the mixed setting the speakers feel the need to perform

that function more frequently.

We must be careful, though, about attributing the parallel increase in causals and conditionals to the gender compo- sition of the dyad. Although we tried to assign similar to- pics of conversation to the single-gender and mixed-gen- der groups, it is still possible that the mixed-gender topic induced higher use of causal clauses than the single-gen- der topic. Other aspects of the conversations could also be responsible for the shift. More studies are needed to clarify this point.

Concerning purpose and concessive clauses, no such differences appeared. However, the small total number of these kinds of clauses produced could have reduced observable effects. A study with more participants might yield different results in this regard. Among the other li- mitations of the study, one that must be mentioned is the experimental setting. A study of a large corpus of natura- lly occurring Spanish speech would be a large step in she- dding more light on the questions explored in this study.

CONCLUSIONS

The relationship of gender to language is complex and locally constructed, and it intersects and interacts with features of the speaker and of the context. We are very far from a complete picture. Mondorf identified one small area, the use of adverbial clauses in relation to the gender of the speaker and the interlocutors, in which an oppor- tunity presents itself to fill in the canvas. This study has taken up the task and accomplished one more increment. Our results demonstrate that Spanish speakers finely at- tune their use of clauses to match the speech of their in- terlocutors, and speakers in mixed-gender dyads address each other with a higher proportion of conditional clau- ses than speakers in single-gender dyads. More research is needed to interpret this finding in terms of qualities that index gender. Studies that account for speaker age, socioeconomic status, and familiarity with the inter- locutor are called for, as well as studies addressing in- tra-speaker variation. Since we have found results much different than those found by Mondorf in English, it will be crucial that the additional research include the analy- sis of other languages as well as other contexts.

BIBLIOGRAPHIC REFERENCES

Aijon Oliva, M. (2013). On the meanings and functions of grammatical choice: The Spanish first-person plural in written-press discourse. Pragmatics, 23, 573-603.

Brownlow, S., Rosemond, J., & Parker, J. (2003). Gen- der-linked linguistic behavior in television inter- views. Sex Roles, 49, 121-132.

Cheshire, J. (2005). Syntactic variation and beyond: Gender and social class variation in the use of dis- course-new markers. Journal of Sociolinguistics, 9,

479-508.

Eckert, P., & McConnell-Ginet, S. (2013). Language and Gender. (Cambridge University Press, Ed.). Cam- bridge.

Elhindi, Y., & McGarry, T. (2013). Gender-linked variation across languages. (Common Ground, Ed.), Cham- paign,.

Hogg, M. A. (1985). Masculine and feminine speech in dy- ads and groups: A study of speech style and gender salience. Journal of Language and Social Psychology,

4, 99-112.

Holmes, J., & M, M. (2008). The Handbook of Language and Gender. (MA: Blackwell, Ed.). Malden.

McGarry, T. (2004). Speaking of topic introductions in the ladies’ auxiliary: A single-gender and mixed-gender comparative analysis. University of South Carolina, Columbia, SC.

Mondorf, B. (2004). Gender differences in English syn- tax. (M. Niemeyer, Ed.), Linguistische Arbeiten 491 (Tübingen).

Mulac, A., Wiemann, J. , Widenmann, S. , & Gibson, T . (1998). Male/female language differences and ef- fects in same-sex and mixed-sex dyads: The gen- der-linked language effect and mutual influence. Communication Monographs, 35, 315-355.

Rincón, A. (2004). Perspectiva de género en la comuni-

cación e imagen corporativa.

Takano, S. (1998). A quantitative study of gender differ- ences in the ellipsis of the Japanese post-position- al particles –wa and –ga: Gender composition as a constraint on variability. Language Variation and Change, 10, 289-323.

Weatherall, A., & Gallois, C. (2003). Gender and Identity: Representation and Social Action. In J. Holmes & M. Meyerhoff. (MA: Blackwell, Ed.), The Handbook of Language and Gender (Malden).