Abstract
This article provides qualitative and quantitative analyses of variable gender agreement in Correntino Spanish, the variety of Spanish spoken by both Spanish-Guarani bilinguals and Correntino Spanish monolinguals in the province of Corrientes, Argentina. Drawing on data collected from fieldwork in the province, it will be shown that this variation is conditioned by distance effects and modifier class. Synchronic gender agreement variation in Correntino Spanish is attributed to diachronic source language agentivity effects (Van Coetsem, 1988) given the lack of gender inflection in Guarani. This phenomenon would be unsurprising as a contact effect if found synchronically only in bilinguals; however, its occurrence in monolinguals sets it aside as a rare instance of variable gender agreement in monolingual Spanish. This loosening of gender agreement mirrors the development of gender in Argentine Guarani (Cerno, 2010), and these phenomena taken together shed light on the malleability of gender systems under situations of intense language contact. Together they provide a valuable example of mutual contact-induced changes in gender systems.
1 Introduction
In the Argentine province of Corrientes, stable bilingualism in Spanish and Guarani has characterized the population for centuries. In the modern day, much of the rural population of the province is bilingual to one degree or another. Many Correntinos in these rural regions are L1 speakers of Argentine Guarani, a variety of Guarani unique to the province of Corrientes (also referred to as “Correntinean Guarani”, “Correntino Guarani”, or “guaraní correntino”). Spanish-L1 bilinguals are found throughout the province, and Spanish monolinguals are common in larger cities and towns. The close contact between the two languages in the region has had widespread effects on both grammars in the lexical, phonological, and morpho-syntactic domains. Argentine Guarani is unique among Guarani varieties in the high degree to which it has been influenced by Spanish (Cerno, 2017: 351), and colloquial Correntino Spanish (henceforth called just “Correntino Spanish”) can be considered a Spanish contact variety given widespread Guarani influence.
One salient morpho-syntactic feature of Correntino Spanish is its system of gender agreement in noun phrases. In this article, I analyze variable morphological gender agreement in this variety of Spanish spoken predominantly in rural Corrientes. While many authors have reported Spanish nonagreement phenomena among Spanish contact varieties (Sánchez, 1996; Tesoro, 2002; Lipski, 2006; Michnowicz, 2012), to my knowledge, no analyses provide comprehensive details of the linguistic conditioning of nonagreement or the sociolinguistic profile of those in which nonagreement is found. This study seeks to do both.
I investigate the linguistic factors which condition the use of morphologically masculine gender inflection in feminine noun phrases in Correntino Spanish; I show that this variation is more systematic in some contexts than in others. Distance effects between noun and modifier are important to the patterns of gender morphology in this variety of Spanish, as is the class of the modifier (i.e., whether it is a determiner or adjective); intervening material between noun and modifier decreases the likelihood of agreement, and within adjacent modifiers, determiners are more likely to agree with their noun than are adjectives. I demonstrate that this phenomenon takes place to one degree or another across all speakers of Correntino Spanish; it is found in speakers of all ages, both sexes, and crucially, among both monolinguals and bilinguals. This is in contrast to most of the existing literature on variation in Spanish agreement, which reports similar phenomena among Spanish L2 speakers but not Spanish-L1 bilinguals or monolinguals. I attribute the synchronic situation of gender agreement variation in this Spanish variety to diachronic contact effects, namely source language agentivity effects (Van Coetsem, 1988) resulting from Guarani L1 speakers acquiring Spanish as an L2. The case of Correntino Spanish allows for analysis of synchronic contact phenomena which stem from second language acquisition and of diachronic contact phenomena which serve as engines of language change. It additionally provides specific linguistic conditioning by which gender systems can vary, and ultimately, evolve.
Finally, I link this phenomenon in Correntino Spanish to the opposite phenomenon in Argentine Guarani. While Spanish, a heavily natively gendered language, has lost patterns of gender agreement in Corrientes, Guarani, a natively genderless language, has gained gender in its article system in Corrientes (Cerno, 2010). These facts together shed light on the malleability of gender systems under situations of intense linguistic contact. These two varieties, on opposite poles of a continuum of gender agreement enforcement, have both shifted to a middle ground which is unattested in either of their recent linguistic ancestors.
2 Background
2.1 The Phenomenon
Spanish has a binary system of grammatical gender. Gender in Spanish can be semantic in the case of animate referents, in which natural gender aligns with morphological gender, and purely grammatical with inanimate referents, in which morphological gender is arbitrary (see Corbett, 1991 for a detailed discussion). Standard Spanish has obligatory morphological gender marking in its modifier system, where the gender of the determiner or adjective must align with that of the noun, as seen in (1). The canonical morphological marker of masculine lexical items is -o, while -a is used for feminine lexical items; exceptions to this generalization exist (e.g., the masculine noun día ‘day’ and the feminine noun mano ‘hand’), and some lexical items have ambiguous morphology (e.g., the feminine noun gente ‘people’ and the masculine noun estante ‘shelf’).
As a point of comparison, Argentine Guarani is natively genderless, as are all Guarani languages. While it has acquired gender agreement in its article system (which itself was also borrowed) due to contact with Spanish (as discussed in Section 4.2), it does not have grammatical gender in any other aspect of its modifier system. To reproduce the examples in (1) in Argentine Guarani, as is done in (2), the addition of the words kuimba’e ‘male’ and kuña ‘female’ are necessary given that the gender of the noun cannot be communicated morphologically (Cerno, 2013: 201–202). The modifiers ko ‘this’ and hũ ‘black’ are also morphologically invariable regardless of the sex of the referent.
Correntino Spanish shows variation in whether or not gender agreement is attested in the modifier system. This variation occurs in both Guarani-Spanish bilinguals and Spanish monolinguals of the region. Examples without gender agreement are seen in feminine nouns taking modifiers with canonically masculine morphology; agreement patterns between masculine nouns and their modifiers are identical to those in standard Spanish (i.e., as in (1b) above). This variation is not lexically conditioned; it occurs throughout the lexicon with both frequent and infrequent modifiers. Examples are found in (3), with the relevant morphology in bold.
Corresponding standard Spanish variants of the phrases in (3) are found below in (4), again with the relevant morphology in bold.
Consequently, three out of the four possible combinations of gender alignment are possible in Correntino Spanish noun+modifier phrases. This is illustrated in (5) with the modifier viejo ‘old’. With a feminine noun (as in 5a and 5b) either morphologically masculine or feminine modifiers are possible, whereas with a masculine noun (as in 5c and 5d) only morphologically masculine modifiers are possible; (5d) is ungrammatical.
2.2 Spanish Gender Agreement
There exists an extensive literature on gender in Spanish. As noted by Eddington (2002: 49), most studies fall into one of three categories: (1) the pedagogical approach, (2) the descriptive approach, or (3) the generative approach. In the second category, a second division might be made between studies which aim to account for patterns of gender generally (e.g., Teschner, 1983; Harris, 1991) and those which analyze non-standard patterns of Spanish gender morphology – usually referred to as errors in gender agreement.1 The latter class forms the background relevant to the analysis provided here.
It is assumed that adult L1 Spanish speakers in regions where Spanish is the dominant language have highly consistent systems of gender agreement (Hawkins and Franceschina, 2004). This has been shown experimentally by Alarcón (2011: 340), in which L1 speakers of standard Spanish varieties from Argentina, Chile, Colombia, Spain, Honduras, and Mexico had universally consistent gender morphology in a production task. Similar results were found in a group of 63 monolingual Spanish speakers from Spain, all of whom had gender agreement rates “at ceiling” in a variety of experimental tasks involving gender agreement (Sagarra and Herschensohn, 2010). Antón-Méndez (1999) showed in a production experiment involving 56 L1 Spanish speakers, the majority of whom were monolingual, a gender agreement error rate of 3% (out of 2304 responses). The task in this study was designed to test gender agreement and therefore had stimuli more complex than would be commonly found in naturalistic speech, suggesting that the naturalistic error rate in speakers is likely below 3%. Further evidence is provided by various neurolinguistic studies which have shown that L1 speakers of standard Spanish varieties are highly sensitive to gender agreement errors – even when there is intervening material between the noun and modifier (see Bañón et al., 2012 and the references provided there). Similarly, sensitivity among native speakers to gender agreement is such that in reading tasks, errors induce reanalysis (evidenced by increased reading time, regressive eye movements, and P600 brain responses), which is not the case for L2 speakers (Keating, 2010: 114–115). Acquisition-oriented studies have shown the accuracy of the acquisition of the Spanish gender system in children raised bilingual or trilingual (Eichler et al., 2012; Hager, 2014), in particular when compared to their acquisition of more morphologically complex gender systems, such as that of German. Finally, spontaneous gender agreement errors in Spanish are much less common than number agreement errors (Antón-Méndez et al., 2002: 3; Barber and Carreiras, 2005: 138). This trend has been attributed to the fact that “gender and number information are not processed in the same manner during sentence production” (Igoa et al., 1999: 189).
Many of the previous studies of nonagreement patterns of Spanish gender morphology either address the L1 acquisition process in children (Pérez-Pereira, 1991; Mariscal, 2008), the L2 acquisition process (White et al., 2004; Montrul et al., 2008; Isabelli-Garcia, 2010), or agreement phenomena in code-switching environments (Otheguy and Lapidus, 2003; Liceras et al., 2008; Cantone and Müller, 2008). The underlying theme of this literature is that any deviation from canonical Spanish agreement patterns can be accounted for as some sort of synchronic irregularity (e.g., the complexities of L1 or L2 acquisition).
A separate literature on irregular agreement analyzes synchronic nonagreement phenomena in Spanish heritage speakers. Montrul (2013: 174) notes that “[u]ndoubtedly, the linguistic area most noticeably affected in heritage language grammars is inflectional morphology”. This is evidenced by studies which have shown systematic patterns of gender nonagreement in heritage speakers, with Montrul et al., (2008) for instance, showing nonagreement rates of 11–15% across three different experimental tasks. Nonagreement patterns have been demonstrated among heritage speakers of other languages with grammatical gender as well, e.g., Russian (Polinsky, 2008) and Norwegian (Lohndal and Westergaard, 2016). Anderson (1999: 392), in a discussion of linguistic outcomes of L1 attrition, notes that because Spanish grammatical gender cannot be reliably recovered morphologically, due to irregularities in the paradigm, “grammatical gender is a feature of the noun […]. It is this characteristic of grammatical gender that makes it susceptible to loss.”
Other studies address variable gender agreement in situations which are not minority language environments or environments of language shift – such as that of Corrientes. The unifying theme of this literature is that gender agreement variation seems to exclusively happen in situations of language contact.2 One of the more well-studied examples is that of Afro-Bolivian Spanish, spoken by descendants of slaves in Los Yungas, Department of La Paz, Bolivia. This variety of Spanish is also reported to have variable gender agreement, as well as other features which distinguish it from standard Bolivian Spanish (Lipski, 2006, 2008; Sessarego, 2010, 2011; Sessarego and Gutiérrez-Rexach, 2011). Afro-Bolivian Spanish has been described as “creole-like” (Lipski, 2008), and its socio-historical development has led to various grammatical features which are uncommon in Spanish generally.
Relaxed gender agreement is also documented in the Andes in regions where Quechua is spoken. Sánchez (1996) reports nonagreement among speakers of Quechua who acquired Spanish as adults. Babel (2014: 60) similarly notes that in the Andes “[v]ariable number and gender agreement, including variation in the interpretation of collective nouns, is often associated with speakers who have contact influence.” Escobar (2001: 87) notes that in the writings of Quechua-Spanish bilinguals in Peru in the colonial period, “gender agreement tends to be masculine between the Art+N and the N+Adj.” The extent to which gender nonagreement in the Andes is found in Spanish monolinguals – complicated by the gradient nature of the categories “monolingual” and “bilingual” there – is unclear, as are the details of the linguistic conditioning of the reported patterns which have been described.
Michnowicz (2012) discusses the case of a regional variety of Spanish in close contact with Yucatec Maya. The reported agreement effects in this variety have broad scope, as determiners and object pronouns are affected, implying an across-the-board relaxing of gender agreement unlike what is seen in Corrientes. Crucially, gender nonagreement is reported among bilinguals of this variety but not among monolinguals.
Tesoro (2002) presents the results of an analysis of a corpus of Guatemalan Spanish composed of written texts and spoken speech; data is provided from both Spanish monolinguals and multilinguals. She reports relaxed gender agreement among noun-adjective relationships in which morphologically feminine modifiers are used in masculine contexts and vice versa. It is not immediately clear whether the monolinguals in her corpus produced patterns of nonagreement, given that no analysis of monolinguals vs. multilinguals is provided, and the precise details of the patterns of nonagreement are not given.3
Balam (2014) reports nonagreement in Northern Belizean Spanish, but the details and scope of the patterns are not given. This phenomenon is reportedly most prevalent among Kriol-dominant trilinguals who speak Kriol, English, and Spanish; adolescent and post-adolescent speakers of Belizean Spanish exhibit standard Spanish agreement patterns.
Spanish-lexicon creoles, namely Chabacano, Papiamentu, and Palenquero, do not have systems of gender agreement (see Grant, 2007: 178; Kouwenberg and Murray 1994: 19; and Lipski, 2015: 1146, respectively) and thus are not subject to the variation reported in non-creole Spanish varieties.4
Looking at Spanish globally, variation in gender agreement in stable Spanish-speaking communities, insofar as I am aware, has only been reported in situations of contact in which multilingualism is common. The case of Correntino Spanish is not unique in this regard, but I am not aware of other Spanish varieties, setting aside “creole-like” varieties, in which monolinguals exhibit variation in gender agreement in noun-modifier relationships. As Escobar (2012: 69) points out in a discussion about grammatical contact features in Spanish contact varieties, “it is important to distinguish features found only in second-language speakers of Spanish who have the Amerindian language as their first language.” This allows us to differentiate synchronic contact phenomena related to second language acquisition on the one hand from diachronic contact phenomena which serve as engines of language change on the other; the former phenomena can only become the latter if the sociolinguistic conditions are met for such features to become conventionalized within a speech community and consequently form part of the linguistic input for L1 acquisition in subsequent generations.
2.3 Spanish-Guarani Contact
There exists a small literature on the linguistic outcomes of the contact between Spanish and Guarani in Corrientes. Abadía de Quant, (1996) provides a general overview of the Guarani influence on Correntino Spanish and briefly addresses gender, mentioning “incorrecta concordancia” (‘incorrect agreement’) in the Correntino Spanish noun phrase (Abadía de Quant, 1996: 208). Cerno (2010) analyzes the adoption of Spanish articles into Argentine Guarani, contrasting the synchronic grammars of it and Paraguayan Guarani regarding their article systems. Guillán (2010) and Sandoval (2006) both discuss a few primarily morphosyntactic effects of Argentine Guarani on Correntino Spanish, in particular the borrowing of specific morphemes (e.g., interrogative markers, evidentials, etc.). Fernández (2008) discusses institutional factors with respect to this situation of language contact, while Fernández (2015) broadly analyzes the effects of this contact on the construction of identity. These studies comprise the totality of the literature specifically aimed at the contact situation in Corrientes as far as I am aware, and with the exception of Cerno (2010) I am unaware of any study which has systematically analyzed any of the contact effects in either language.5
In morphology, all the features of [Colloquial Paraguayan Spanish] – except for voseo and leísmo – are induced by contact with Guaraní. They can be classified in two types: those which calque Guaraní morphology and those which use Guaraní morphology. The first type includes: the lack of number and gender agreement in the noun phrase, induced by the non-marking of these categories in the indigenous language.
Secondly, Granda (1996: 78) makes mention in a footnote of “the very notorious weakening in Paraguayan Spanish of the gender agreement system of the standard Spanish noun phrase, a fact which is a result of, as an exclusive driving factor, the non-existence in Guarani of gender marking” (my translation).6 Lastly, Herring (2012) provides a more focused analysis of gender in Paraguayan Spanish. She ultimately provides evidence for the system of agreement in Paraguay being similar to that of non-contact varieties of Spanish via a corpus analysis of 1,135 tokens in which only 26 (2.3%) showed nonagreement; she concludes that there were no apparent social or linguistic explanations for the observed nonagreement. Thus, with regard to the existing literature, the nature and patterns of gender (non)agreement in Paraguayan Spanish are unclear.
3 Analysis
3.1 Data
The data informing this study were collected during fieldwork in the province of Corrientes during the (southern hemisphere) winters of 2017 and 2018. These data come from a corpus of formal, recorded sociolinguistic interviews with 24 participants. There were 8 female and 16 male participants, and the participant age range spanned 35–86. On the basis of self-reporting, there were 13 L1 Guarani speakers and 11 L1 Spanish speakers. Four of the L1 Spanish speakers considered themselves monolingual (speakers 10, 11, 14, and 18 in Table 1); all other participants considered themselves bilingual. The speech analyzed in this study comprised approximately 14 hours of recordings.
All recordings were made within either the department of San Miguel or General Paz, and the majority took place in the primary field site of Loreto (in San Miguel). Figure 17 shows a map of the province, and Loreto can be seen in the central northern region of the province, near the Paraná River (not shown on the map) which forms part of the natural boundary separating Argentina and Paraguay.
It is important to mention that I observed this phenomenon in many speakers in Corrientes who were not included in this study, including completely monolingual children. For the quantitative analysis, however, I only included speakers for whom I had recorded sociolinguistic interviews. I took notes in a field notebook whenever I encountered the phenomenon, but these speakers for whom I only had tokens of nonagreement (i.e., speakers who were not formally interviewed) were excluded from the analysis to avoid skewing the data. Qualitative statements here about the phenomenon include these excluded data, but the quantitative analysis does not.
Table 1 provides the sex, age, birthplace, and L1 of each speaker included in the quantitative analysis. “Birthplace” indicates not just birthplace but also that the individual was raised in that location. All birthplace locations are within the departments of General Paz and San Miguel as shown in Figure 1 except that of speaker 4; Ituzaingó is a town within a department of the same name (located immediately to the east of the department of San Miguel).
The province of Corrientes; the superimposed map shows it within Argentina. The adjacent departments of General Paz and San Miguel, respectively from west to east, are highlighted; the town of Loreto is indicated by the small circle.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
The province of Corrientes; the superimposed map shows it within Argentina. The adjacent departments of General Paz and San Miguel, respectively from west to east, are highlighted; the town of Loreto is indicated by the small circle.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
The province of Corrientes; the superimposed map shows it within Argentina. The adjacent departments of General Paz and San Miguel, respectively from west to east, are highlighted; the town of Loreto is indicated by the small circle.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
The envelope of variation, i.e., the context in which variation occurs (Cameron and Schwenter, 2013: 466), for this study includes all grammatically feminine Spanish nouns with an accompanying modifier with overt gender morphology, given that the gender agreement variation is only found in feminine np s. “Modifier” here is used as an umbrella term to mean any adjective or determiner – crucially excluding articles. I did not find variation in other contexts. Article agreement patterns, for both definites and indefinites, are identical to those in standard Spanish, as are agreement patterns between morphologically masculine forms and their modifiers.8 On these criteria, noun phrases such as those in (7) would not be included on the respective bases of the presence of a definite article, indefinite article, and non-overt gender morphology; noun phrases such as those in (8) would be included.
Of course, larger noun phrases in which articles and other modifiers are present, e.g., la casa blanca ‘the white house’, are counted as a modifier token with the article being ignored; the article is ignored given that, as described above, article+noun combinations show universal agreement for all speakers.9 No proper nouns or fixed phrases involving a feminine noun and modifier were found in the corpus (e.g., La Casa Rosada ‘the pink house’, i.e., the executive mansion and office of the President of Argentina).
736 total tokens within the envelope of variation were identified. The overall rate of nonagreement in the data is 17% (124/736), leaving an agreement rate of 83%. This rate is highly influenced by inter-speaker variability (see Figure 2 above) and by the fact that the interviews with some speakers were longer than those with others. The two longest interviews, for instance, were with two individuals with high rates of agreement. When controlling for token count by speaker, the mean nonagreement rate between speakers is 22%, leaving an agreement rate of 78%, i.e., agreement is consistently more common than nonagreement. It is noteworthy that some speakers, for instance 9 and 2, have agreement rates approximating 50%, while other speakers have agreement rates higher than 90%. Every speaker interviewed had some instances of nonagreement; except for speaker 12, for whom there is a very low token count. Speaker individual agreement rates are shown left to right from highest rate to lowest rate in Figure 2.
Morphological gender agreement rates by speaker (mean rate: 83%, 612/736). The y-axis represents the number of observations; the x-axis represents individual speakers left to right from highest rate to lowest rate of agreement (speaker numbers below each column correlate to those in Table 1). Agreement rates for each speaker are given at the top of the column in percentages.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Morphological gender agreement rates by speaker (mean rate: 83%, 612/736). The y-axis represents the number of observations; the x-axis represents individual speakers left to right from highest rate to lowest rate of agreement (speaker numbers below each column correlate to those in Table 1). Agreement rates for each speaker are given at the top of the column in percentages.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Morphological gender agreement rates by speaker (mean rate: 83%, 612/736). The y-axis represents the number of observations; the x-axis represents individual speakers left to right from highest rate to lowest rate of agreement (speaker numbers below each column correlate to those in Table 1). Agreement rates for each speaker are given at the top of the column in percentages.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
3.2 Variables
The dependent variable in this study is gender agreement, a categorical variable (agreement vs. nonagreement) which was contrast-coded. I coded for both social and linguistic independent variables. The social independent variables coded for were:
– speaker age (continuous)
– speaker sex (categorical)
– speaker L1 (categorical)
The linguistic independent variables coded for were all categorical:
– animacy (animate vs. inanimate)
– priming (whether or not the same pattern – either agreement or nonagreement – took place twice within a five-clause span)
– definiteness (definite vs. indefinite)10
– adjacency (immediately adjacent vs. non-adjacent)
– modifier class (determiner vs. adjective)11
– noun-modifier order (N+M vs. M+N)
– noun morphology (canonical feminine marking, i.e., ending in -a vs. non-canonical marking, i.e., everything else)12
It should be noted that no single statistical model can contain all the variables coded for given the relative scarcity of data. Accordingly, various variables were not included in the final model. The decision of what variables to keep was made on the bases of: (1) which variables were theoretically most expected to impact gender agreement (on the basis of prior literature), and (2) which variables impressionistically seemed most important via the use of contingency tables, random forests, and conditional inference trees. Additionally, some pairs of variables could not be included in the same model due to issues of collinearity. For instance, I coded for both modifier class and noun-modifier order in spite of their likely collinearity, given that determiners most often appear before the noun (e.g., este ‘this’ in este ave rojo ‘this red bird’) while adjectives most often appear after the noun (e.g., rojo ‘red’ in este ave rojo ‘this red bird’). In the end, three independent variables were included in the final statistical model: animacy, adjacency, and modifier class.
3.3 Results
Given the categorical dependent variable, I built a logistic mixed-effects regression model to predict the realization of gender agreement from animacy, adjacency, modifier class, and all their interactions. I began by running a maximal effects structure model; given that the variables are all within-subjects variables, this is necessary to avoid violation of the assumption of independence (Barr et al., 2013). The maximal effects structure model did not converge. Given this, I systematically followed the advice of Barr et al., (2013) to simplify the model until a model which converged was produced. I first removed all correlations among the random effects and reran the model. This model did not converge, so I then removed random intercepts for speaker. Again the model did not converge, so I removed random effects with the lowest variance until arriving at a model which converged. That model included a random slope for adjacency by speaker but no intercept. I then ‘stepped down’ using log-likelihood comparisons in order to find the simplest model which accounted for the patterns in the data. This model included adjacency, modifier class, and their interaction, but not animacy, as including animacy did not significantly improve the model’s fit. The final model produced the results seen in Table 2.
We can see the significant variables visually represented in Figures 3 and 4. Figure 3 shows agreement ratios looking at adjacency of the modifier alone. Figure 4 shows the agreement ratios for the significant interaction of adjacency and modifier class across the four possible groups: adjacent adjectives, non-adjacent adjectives, adjacent determiners, and non-adjacent determiners.
Rates of morphological gender agreement by adjacency only. The y-axis represents the number of observations. Percentages within each class are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Rates of morphological gender agreement by adjacency only. The y-axis represents the number of observations. Percentages within each class are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Rates of morphological gender agreement by adjacency only. The y-axis represents the number of observations. Percentages within each class are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Rates of morphological gender agreement across all combinations of adjacency and modifier class. The y-axis represents the number of total observations. Percentages within each of the four divisions are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Rates of morphological gender agreement across all combinations of adjacency and modifier class. The y-axis represents the number of total observations. Percentages within each of the four divisions are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Rates of morphological gender agreement across all combinations of adjacency and modifier class. The y-axis represents the number of total observations. Percentages within each of the four divisions are found at the top of each column.
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
As corroborated by the logistic regression, Figure 3 demonstrates that adjacency is important overall. Modifiers are less likely to agree with their nouns if there is any intervening linguistic material between them; this is illustrated in (9), produced by speaker 11 (see Table 1), where immediately adjacent correntina agrees with historia but non-adjacent lindo does not.
Figure 4 demonstrates that the effect of adjacency differs by modifier class. Within adjacent modifiers, determiners are more likely to agree with their noun than adjectives; this is illustrated by comparing (10a) and (10b), the former produced by speaker 7 and the latter by speaker 19 (see Table 1), where the adjacent determiner esa agrees with música but the adjacent adjective auténtico does not. Adjacent determiners show the highest rates of agreement at 95.6%; they have both the highest token count of the four divisions seen in Figure 4 as well as the largest discrepancy in terms of the agreement ratio.
Given that determiners in Spanish generally precede the noun while adjectives usually follow the noun, it might be the case that the effect of modifier class here is not actually of modifier class but rather of sensitivity to word order. To test this, I went through the same stepwise process described above to build a model that included noun-modifier order in place of modifier class. This model yielded significance of adjacency only (i.e., noun-modifier order and the interaction between adjacency and noun-modifier order were not significant). This suggests that modifier class is more important to determining gender agreement than noun-modifier order, and that adjacency is consistently important overall given its significance in both models.
3.3.1 Speaker L1
While the social variables were left out of the formal statistical model, the role of speaker L1 is worth looking at descriptively. Of the 736 tokens, 228 were produced by self-reported Guarani L1 speakers, while the remaining 508 came from Spanish L1 speakers. Of the Guarani L1 group, the overall rate of agreement was 73.7% (168/228) while for Spanish L1 speakers the agreement rate was 87.4% (444/508). The four Spanish-L1 speakers who considered themselves monolingual, speakers 10, 11, 14, and 18, had agreement rates of 94%, 94%, 67% and 78% respectively. Mean agreement rates by L1, with Spanish L1 monolinguals separate from Spanish L1 bilinguals, are seen in Figure 5.
Mean rates of gender agreement in Correntino Spanish by speaker L1
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Mean rates of gender agreement in Correntino Spanish by speaker L1
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Mean rates of gender agreement in Correntino Spanish by speaker L1
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
These rates were calculated by taking the mean agreement rate of each speaker within the three groups, and then taking the mean of those means. This helps control for the fact that token counts varied by speaker. This comparison is valuable but has some limitations; the number of speakers within each group varies (4 Spanish L1 monolinguals, 7 Spanish L1 bilinguals, and 13 Guarani L1 bilinguals) as do the token counts within each group (125, 383, and 228 respectively). Therefore, we cannot confirm statistically that the behavior of these three groups is different; nonetheless, the data do demonstrate that Spanish L1 speakers – including monolinguals – have moderately high nonagreement rates. Given the rates of agreement of 97–100% in Spanish monolinguals found in the previous studies discussed above, Correntino Spanish speakers – in particular the monolinguals – have higher-than-expected rates of nonagreement when compared to monolinguals of standard Spanish varieties.
4 Discussion
In light of the widely held assumptions that long-distance dependencies tax computational resources (e.g., Gibson 1998) and that memory resources vary among individuals (e.g., Daneman and Carpenter 1980), the distance that mediates a controller noun and its target adjective and the availability of memory resources needed to check agreement between them may be key determinants in whether native-like processing of gender is achievable in a L2.
In Correntino Spanish, feminine nouns require morphological agreement from accompanying modifiers variably along a continuum, as represented in Figure 6.
Continuum of gender agreement enforcement in Correntino Spanish
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Continuum of gender agreement enforcement in Correntino Spanish
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Continuum of gender agreement enforcement in Correntino Spanish
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Articles are a class apart in that gender agreement is consistent among all speakers. Adjacent (non-article) determiners form a separate class for which agreement is favored but in which variation is still attested. A third class, for which variation in the agreement system is freer, is composed of adjectives and non-adjacent (non-article) determiners.
There are three possible interpretations of the data presented here and what they tell us about the Correntino Spanish grammar:
1. The grammatical assignment of gender at the lexical level is variable, i.e., speakers do not have rigid assignment of gender in their nominal lexicon.
2. Gender is rigidly assigned at the lexical level, but the grammatical constraints requiring gender agreement in general have been lessened to the point where agreement is variable in some contexts.
3. Gender is rigidly assigned at the lexical level and agreement is still enforced, but morphologically masculine forms can serve as a gender-neutral alternative.
These data favor the third interpretation as the appropriate analysis. Interpretation 1 cannot account for the fact that the article+noun agreement system in Correntino Spanish (including both definites and indefinites) is identical to that of standard Spanish. All the irregularities in the gender patterns of standard Spanish nouns (i.e., nouns with ambiguous gender morphology, feminine nouns with -o endings, masculine nouns with -a endings, etc.) are present in the Correntino Spanish system.13 I found no instances of gender nonagreement between articles and nouns in the data; that agreement of this type is strictly enforced suggests all nouns still have clear gender assignment for Correntino Spanish speakers.
This same line of reasoning also eliminates interpretation 2 as a potential analysis. Although logically possible, the notion that patterns of agreement are being loosened in general cannot account for the obligatory article+noun agreement in Correntino Spanish. Additional evidence against this account comes from the fact that, as described in Section 3.1, morphologically masculine modifiers can optionally appear with feminine nouns but not the other way around (e.g., cabeza rojo ‘red head’ is grammatical but not *pájaro roja ‘red bird’). If the rules governing agreement in the Correntino Spanish grammar were loosened generally, we would not expect such rule-governed variation. That is, we would expect instances of morphologically masculine nouns being modified by morphologically feminine modifiers as is reported by Tesoro (2002) for a variety of Guatemalan Spanish (discussed in Section 2.2).
These pieces of evidence point to the third interpretation, that of morphologically masculine forms serving as a gender-neutral option. Speakers appear to be treating morphologically masculine forms as forms which can agree with nouns of either gender, given (1) that gender agreement is still rigidly enforced in certain domains of the Correntino Spanish noun phrase, and (2) that we see variation involving canonically masculine morphology but not feminine morphology. This can be seen as a constrained (i.e., non-global) loosening of the rules of agreement, and the distance effects indicate that as the cognitive load of tracking gender agreement gets higher (with intervening linguistic material between the noun and its modifier) speakers can mitigate this with optional use of a gender-neutral modifier. Across-the-board non-enforcement of agreement rules should lead to less systematic patterns. Such systematicity suggests the evolution of these modifiers from having masculine morphology and masculine grammatical gender to having masculine morphology but optionally ambiguous grammatical gender.14
One subtle but important distinction between the behavior of these optionally ambiguous modifiers and previous definitions of “default” gender in Spanish is worth consideration. Harris (1991: 44) for instance, notes “I interpret unmarked or default gender literally as the absence of any information about gender in lexical entries.” Such a definition is not applicable to the phenomenon in Correntino Spanish given that strictly enforced article+noun agreement indicates the direct presence of gender information on the noun. The Correntino Spanish patterns are therefore not theoretically identical to the masculine-as-default phenomenon discussed in standard Spanish, in which masculine modifiers are used when the noun is inherently genderless.15
Similarly, ubiquitous article+noun agreement also theoretically distinguishes the patterns here from accounts of the underspecification of gender in nouns (see, for instance, Opitz and Pechmann, 2016 for such an analysis in German). Nouns in Correntino Spanish, as in standard Spanish, are fully specified for gender. If we assume Harris’ (1991: 44) notion that “[c]oncord targets like determiners and adjectives must be genderless in the lexicon (presumably universally); they acquire gender only through concord”, then analysis of these modifiers in Correntino Spanish as being underspecified for gender is redundant. They have merely expanded the range of nouns which they can grammatically appear with.
Finally, it is worth considering the unique behavior of articles (as shown in Figure 6). The special treatment of articles by the Correntino Spanish grammar may plausibly be due to their privileged status within the noun phrase. Article+noun combinations are highly susceptible to chunking phenomena because of their frequent adjacent appearance, and this effect is exaggerated in Spanish given that adjectives prototypically follow nouns and therefore intervening material between a noun and its article is even less frequent than it is in languages such as English. Evidence for this comes from L1 acquisition of the Spanish noun phrase. López Ornat (1997) shows how children younger than two have already acquired article+noun phrases, in which the “articles” are linguistically relevant vowels (or “proto-articles” for Lleó, [1998]) which are later bootstrapped into their fully morphosyntactically meaningful forms. Other studies have demonstrated sensitivity to gender-marked articles in L2 speakers of Spanish (Lew-Williams and Fernald, 2010; Grüter et al., 2012), providing further evidence of learner – be they L1 or L2 – sensitivity to article+noun combinations. The privileged status of articles in the Correntino Spanish grammar is likely a manifestation of their privileged status in acquisition generally.16
4.1 Contact Effects
I here take the stance which Gómez Rendón (2008: 149) takes for Paraguayan Spanish, i.e., that these patterns of variable gender agreement are a product of Spanish contact with Guarani, a natively genderless language. This account is grounded in previous studies of two different types: (1) those showing that Spanish gender systems vary in other parts of the Hispanosphere in situations of intense contact with genderless languages, and (2) those showing that monolingual varieties of Spanish outside situations of intense contact have been shown to have extremely consistent patterns of gender agreement (see Section 2.2).
With regard to contact effects, Van Coetsem (1988) distinguishes two different phenomena, those governed by recipient language agentivity (i.e., borrowing) and those governed by source language agentivity (i.e., imposition).17 Phonological adaptation of borrowed forms is an example of recipient language agentivity, given that an input form comes from the source language and is then adapted by the recipient language grammar. Source language agentivity, however, is exemplified by cases where an L2 speaker of a given language uses features from their L1 grammar in the L2 system, e.g., as Van Coetsem (1988: 11) points out, if a native speaker of French pronounces English pit as [pit] and not [phɪt]. In this latter case, linguistic material is not outright borrowed, but rather grammatical features of the acquired system undergo modification as a result of the native system.
Given that the phenomenon here is found in monolingual speakers of Spanish in Corrientes, it cannot be synchronically analyzed as a case of source language agentivity (i.e., imposition). The most parsimonious account of this phenomenon is that it has diachronic roots in imposition, i.e., it is an instance of contact-induced change (Thomason and Kaufman, 1988) which arose sometime previous to the current situation of widespread bilingualism in Corrientes. It is easy to envision the historical scenario in which large numbers of the Guarani-speaking population were learning Spanish as a second language. The influence of their native Guarani grammars could have mediated their acquisition of the Spanish gender agreement system, as often happens when L1 speakers of a genderless language acquire a language with gender (Franceschina, 2005: 116–120), e.g., when L1 English speakers acquire Spanish. With a sufficiently skewed L2-to-L1 speaker ratio, the conditions might not have formed to allow for the complete acquisition of the standard Spanish system at the community level. Consequently, the variability in agreement patterns, with its source in L2 acquisition of Spanish, became the linguistic input for L1 learners of this variety, and Spanish monolinguals began to exhibit this variation, at which point it ceased being a contact feature (except in the diachronic sense). The Spanish gender system is particularly vulnerable in L2 acquisition given that it is not strictly speaking necessary for comprehension and therefore would not have been one of the more pressing features for L2 learners to acquire in order to communicate. Under this interpretation, the contact effects on the transmission of Spanish gender resemble the development of English; the diachronic reduction in English morphology and other grammatical features has been argued to stem from the acquisition of English as an L2 by Old Norse speakers and, later, Norman French speakers, whose imposition effects became an aspect of monolingual, standard English (McWhorter, 2002).
4.2 Contact-Induced Gender Agreement in Argentine Guarani
In light of the fact that Guarani languages are natively genderless, the article system of Argentine Guarani is noteworthy for its productive use of gender (Cerno, 2010). Argentine Guarani uses both definite singular articles from Spanish, i.e., masculine el and feminine la. On the basis of the generalizations provided by Cerno (2010), which are additionally confirmed by Argentine Guarani data I have collected, el is the default article, while la is limited in scope.18 As seen in Table 3, la appears in cases in which the noun either (1) has feminine semantic gender (e.g., ryguasu ‘hen’, guáina ‘girl’) or (2) is a borrowing from Spanish with feminine grammatical gender in the source form (e.g., ekuéla ‘school’, pare ‘wall’).19
This phenomenon extends to the Argentine Guarani indefinite article system as well, in which both Spanish masculine un and feminine una are used (with the same patterning as el and la). It is, however, confined to the singular; the plural definite article lo is used as a gender-neutral article in Argentine Guarani (Cerno, 2010: 7).
This use of both Spanish el and la is synchronically unattested in Argentine Guarani’s closest relative, Paraguayan Guarani,21 in which the Spanish definite article la is used as a gender-neutral form and not as a marker of feminine grammatical gender (Gynan, 2017: 87).22 Paraguayan Guarani np s such as la karai ‘the gentleman’ are ungrammatical in Argentine Guarani. Likewise, we see the same pattern with indefinites, in which Paraguayan Guarani uses the native form peteĩ as a gender-neutral indefinite article (Cerno, 2010: 8). The greater presence of Spanish-origin features in the Argentine Guarani grammatical system is one of several distinguishing features between it and Paraguayan Guarani, and nominal gender in the former but not the latter, visible in the borrowed article system, is an example of this.
4.3 Theoretical Implications
The fact that Argentine Guarani developed and has maintained a system of grammatical gender in effect mirrors the findings here that the canonical Spanish system of obligatory gender agreement is loosened in Correntino Spanish. That is, the pre-contact patterns of agreement in both languages – obligatory gender agreement in Spanish and no agreement in Guarani – have both shifted to become like the opposite language. On a continuum of gender agreement enforcement, both languages have taken steps away from their native polar positions of the continuum toward the center in which agreement is neither entirely enforced nor entirely absent. The latter has developed patterns of gender agreement where there originally were none, and the former has relaxed obligatory gender agreement such that canonical patterns of agreement are not enforced in all environments. These parallel but ultimately distinct contact effects provide interesting insight into how intense language contact can affect gender systems, which feeds into larger conversations about the evolution of gender systems generally.
[O]n the historical side of things, it must be admitted that a lot of what we believe we know about gender systems and how they change is inferential. There is very little direct observation of change, except for some reshuffling of distinctions, including especially the loss of distinctions, and virtually no observation of how they arise in the first place and get to be so pervasive across large swaths of the grammar.
While “direct observation of change” might not be an appropriate characterization of the data and analysis presented here, we can demonstrate that both gender-related changes, that of Correntino Spanish and that of Argentine Guarani, have taken place since contact. This contact situation provides an example of not only a (partial) loss of distinctions, but also a direct instance of the genesis of a gender system. While Argentine Guarani has not developed gender morphology, cross-linguistically it is not uncommon for gender systems to exist in which underlying gender is morphologically irrecoverable and only visible through modifier agreement (Herbert, 1985: 172). Language contact is one avenue for the genesis of such a system. In saying that we have “virtually no observation of how [gender systems] arise in the first place”, Janse et al. were likely referring more directly to language-internal gender development; however, the role of contact as a vehicle for evolution in gender systems is a necessary component of a more thorough understanding of gender generally (see also Trudgill, 1999 and Clamons, 1995).
The data here also inform our understanding of gender agreement hierarchies cross-linguistically. The fact that the Correntino Spanish system is sensitive to the determiner/adjective distinction is in line with a variety of previous studies. As Franceschina (2005: 114) notes, “[S]tudies of child sla (e.g., Boyd, 1975; Van Naerssen, 1986) and adult sla (e.g., Bartning, 2000; Bruhn de Garavito and White, 2000; Chini, 1995; Dewaele and Véronique, 2001; Fernández García, 1999; Finnemann, 1992) show that det+N agreement is overall more accurate than N+adj or adj+N agreement, as in L1 development.” It has been proposed that determiners, and articles in particular (as discussed above), are encoded during the acquisition process in the lexical representation of nouns and therefore act as privileged elements with regard to their role in the acquisition of gender systems (Carroll, 1989: 572). The Correntino Spanish data here are consistent with this analysis.
There are also ways in which Correntino Spanish does not pattern according to previously discussed gender agreement hierarchies. Franceschina (2005: 114) also notes for instance that “gender agreement accuracy between the noun and the definite article typically emerges and is mastered before agreement between the noun and the indefinite article (e.g., Bartning, 2000; Chini, 1995; Hawkins, 1998; Müller, 1994, 2000; Sabourin, 2001), as in L1 development”. The Correntino Spanish grammar however makes no distinction with regard to agreement patterns between definite and indefinite articles. Additionally, Sessarego and Gutiérrez-Rexach (2011) provide the gender agreement ranking in Figure 7 for Afro-Bolivian Spanish.23
Gender agreement ranking in Afro-Bolivian Spanish. Adapted from Sessarego and Gutiérrez-Rexach (2011:482).
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Gender agreement ranking in Afro-Bolivian Spanish. Adapted from Sessarego and Gutiérrez-Rexach (2011:482).
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
Gender agreement ranking in Afro-Bolivian Spanish. Adapted from Sessarego and Gutiérrez-Rexach (2011:482).
Citation: Journal of Language Contact 14, 3 (2021) ; 10.1163/19552629-14030005
While there are similarities between this ranking and that in Figure 6, there are differences as well, such as the fact that Correntino Spanish demonstratives and definite articles do not pattern together, the Correntino Spanish grammar is not sensitive to the weak/strong quantifier distinction, and adjective position with respect to the noun does not impact agreement likelihood. Thus, while cross-linguistic generalizations can (and should) be made,24 the minor details of gender agreement hierarchies exhibit inter-variety (or inter-language) variation, something unsurprising given the common occurrence of intra-variety variation. This latter kind of variation also leads to questions about the stability of such systems. Sessarego and Gutiérrez-Rexach (2011) show evolution in the Afro-Bolivian system over three generations, such that speakers who are 80+, 51–80, and 21–50 have in effect three different (although similar) grammars. While the Correntino Spanish system does not exhibit such demarcated generational distinctions, the long-term stability of occupying middle ground on a continuum of gender agreement enforcement is still an open question, and further movement along this continuum for either Correntino Spanish or Argentine Guarani will make for fruitful future research.
5 Conclusion
In this study, I have provided an account of the variation in gender agreement in the Correntino Spanish np. Feminine nouns in this variety of Spanish can optionally take modifiers with canonically masculine morphology, and the data here suggest that modifiers with masculine morphology can serve as gender-neutral modifiers in this variety. Variation is not random, and immediately adjacent modifiers are significantly more likely to agree with their nouns that non-adjacent modifiers. Additionally, within adjacent modifiers specifically, determiners are more likely to agree with their noun than adjectives.
This effect is presented here as a diachronic imposition phenomenon, i.e., a contact effect from Argentine Guarani, which is a natively genderless language. Synchronically however, this phenomenon cannot be exclusively attributed to contact effects resulting from the acquisition process given that it is present in Spanish monolinguals. The case of Correntino Spanish highlights the relationship between synchronic contact phenomena related to second language acquisition and diachronic linguistic change generally.
This study has brought to light a previously uninvestigated instance of morphosyntactic variation in an under-documented variety of Spanish. The case of Correntino Spanish allows insight into how contact effects can become part of a monolingual grammar. Other Spanish contact situations which have had similar effects on Spanish gender agreement, such as the cases of Yucatec Spanish and North Belizean Spanish, have not resulted in effects on the Spanish of monolinguals. Setting aside “creole-like” Spanish varieties, the case of Correntino Spanish represents a rare case of contact effects causing a change in patterns of gender agreement among Spanish monolinguals. This case study also illustrates specific linguistic mechanisms by which gender systems can vary – namely via distance effects resulting from intervening linguistic material between a noun and its modifier and via sensitivity to lexical classes such as adjective and determiner.
Via comparison with the contact-induced development of gender agreement in Argentine Guarani, this study paints a larger picture of the malleability of gender systems in environments of intense language contact. When Spanish and Guarani first came into contact in Corrientes, the system of the former was characterized by mandatory gender agreement in the noun phrase, while that of the latter had no gender agreement of any kind. Centuries later, Argentine Guarani is unique among Guarani varieties as having a gendered article system, reflecting the development of underlying gender in nouns, and Correntino Spanish is rare among Spanish varieties in having developed gender-neutral modifier morphology. The situation of widespread bilingualism in Corrientes has led to a kind of meeting in the middle between the two languages regarding gender agreement requirements, and it provides a valuable example of mutual contact-induced changes in gender systems.
References
Abadía de Quant, Inés. 1996. Guaraní y español: Dos lenguas en contacto en el Nordeste argentino. Signo y Seña 6: 199–233.
Abbott, Barbara. 2004. Definiteness and indefiniteness. In Laurence R., Horn and Gregory, Ward (eds.), Handbook of Pragmatics: 122–149. Oxford: Blackwell Publishers.
Alarcón, Irma V. 2011. Spanish gender agreement under complete and incomplete acquisition: Early and late bilinguals’ linguistic behavior within the noun phrase. Bilingualism: Language and Cognition 14(3): 332–350.
Anderson, Raquel T. 1999. Loss of Gender Agreement in L1 Attrition: Preliminary Results. Bilingual Research Journal 23(4): 389–408.
Antón-Méndez, María Inés. 1999. Gender and number agreement processing in Spanish: PhD dissertation. Tucson. University of Arizona.
Antón-Méndez, Inés, Janet L., Nicol and Merrill F., Garrett. 2002. The Relation between Gender and Number Agreement Processing. Syntax 5(1): 1–25.
Babel, Anna. 2014. The role of context in interpreting linguistic variables. Boletín de Filología 49(2): 49–85.
Balam, Osmer. 2014. Notes on the History and Morphosyntactic Characteristics of Spanish in Northern Belize. Kansas Working Papers in Linguistics 35: 79–94.
Bañón, José Alemán, Robert, Fiorentino and Alison, Gabriele. 2012. The processing of number and gender agreement in Spanish: An event-related potential investigation of the effects of structural distance. Brain Research 1456: 49–63.
Barber, Horacio and Manuel, Carreiras. 2005. Grammatical Gender and Number Agreement in Spanish: An erp Comparison. Journal of Cognitive Neuroscience 17(1): 137–153.
Barr, Dale J., Roger, Levy, Christoph, Scheepers and Harry J., Tily. 2013. Random effects structure for confirmatory hypothesis testing: Keep it maximal. Journal of Memory and Language 68: 255–278.
Bartning, Inge. 2000. Gender agreement in L2 French: Pre-advanced vs. advanced learners. Studia Linguistica 54(2): 225–237.
Boyd, Patricia A. 1975. The Development of Grammar Categories in Spanish by Anglo Children Learning a Second Language. TESOL Quarterly 9(2): 125–135.
Bruhn de Garavito, Joyce and Lydia, White. 2000. L2 acquisition of Spanish dp s: The status of grammatical features. In Catherine, Howell, Sarah A., Fish and Thea, Keith-Lucas (eds.), Proceedings of the 24th Annual Boston University Conference on Language Development: 164–175. Somerville: Cascadilla Press.
Cameron, Richard and Scott, Schwenter. 2013. Pragmatics and Variationist Sociolinguistics. In Robert, Bayley, Richard, Cameron and Ceil, Lucas (eds.), The Oxford Handbook of Sociolinguistics, chap. 23, 464–483. Oxford: Oxford University Press.
Cantone, Katja Francesca and Natascha, Müller. 2008. Un nase or una nase? What gender marking within switched dp s reveals about the architecture of the bilingual language faculty. Lingua 118(6): 810–826.
Carroll, Susanne. 1989. Second-Language Acquisition and the Computational Paradigm. Language Learning 39(4): 535–594.
Cassano, Paul. 1971. Substratum hypothesis concerning the Spanish of Paraguay. Neophilologus 55(1): 41–44.
Cassano, Paul. 1973. Retention of Certain Hiatuses in Paraguayan Spanish. Linguistics 11(109): 12–16.
Cerno, Leonardo. 2010. Spanish Articles in Correntinean Guaraní: A Comparison with Paraguayan Guaraní. STUF - Language Typology and Universals (Sprachtypologie und Universalienforschung) 63(1): 20–38.
Cerno, Leonardo. 2013. El Guaraní Correntino: Fonología, Gramática, Textos. Frankfurt am Main: Peter Lang.
Cerno, Leonardo. 2017. Aspects of Dialectal Diversification of Guarani in Paraguay and Corrientes: Contact between Two Given Languages in Different Settings. In Bruno, Estigarribia and Justin, Pinta (eds.), Guarani Linguistics in the 21st Century, chap. 11: 348–378. Leiden: Brill Publishers.
Chini, Marina. 1995. Un aspect du syntagme nominal en italien L2: Le genre. Acquisition et Interaction en Langue Étrangère 5: 115–142.
Choi, Jinny K. 1998. Languages in contact: A morphosyntactic analysis of Paraguayan Spanish from a historical and sociolinguistic perspective: Georgetown University dissertation.
Choi, Jinny K. 2000. [-Person] Direct Object Drop: The Genetic Cause of a Syntactic Feature in Paraguayan Spanish. Hispania 83(3): 531–543.
Choi, Jinny K. 2001. The genesis of voy en el mercado: The preposition en with directional verbs in Paraguayan Spanish. Word 52(2): 181–196.
Clamons, Robbin. 1995. How Recent Contact Erased Ancient Traces in the Gender Systems of the Oromo Dialects. In Proceedings of the Twenty-First Annual Meeting of the Berkeley Linguistics Society: General Session and Parasession on Historical Issues in Sociolinguistics/Social Issues in Historical Linguistics: 389–400. Berkeley: Berkeley Linguistics Society.
Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press.
Corbett, Greville G. 2006. Agreement. Cambridge: Cambridge University Press.
Cuervo, María Cristina and Natalia, Mazzaro. 2013. Duplicación de la negación en el español de Corrientes. In Laura, Colantoni and Celeste Rodríguez, Louro (eds.), Perspectivas teóricas y experimentales sobre el español de la Argentina, 159–176. Madrid: Iberoamericana.
Daneman, Meredyth and Patricia A., Carpenter. 1980. Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior 19(4): 450–466.
Dewaele, Jean-Marc and Daniel, Véronique. 2001. Gender assignment and gender agreement in advanced French interlanguage: A cross-sectional study. Bilingualism: Language and Cognition 4(3): 275–297.
Eddington, David. 2002. Spanish Gender Assignment in an Analogical Framework. Journal of Quantitative Linguistics 9(1): 49–75.
Eichler, Nadine, Veronika, Jansen and Natascha, Müller. 2012. Gender acquisition in bilingual children: French-German, Italian-German, Spanish-German and Italian-French. International Journal of Bilingualism 17(5): 550–572.
Escobar, Anna María. 2001. Contact features in Colonial Peruvian Spanish. International Journal of the Sociology of Language 149: 79–93.
Escobar, Anna María. 2012. Spanish in Contact with Amerindian Languages. In José Ignacio, Hualde, Antxon, Olarrea and Erin, O’Rourke (eds.), The Handbook of Hispanic Linguistics, chap. 4, 65–88. Hoboken: Blackwell Publishing.
Estigarribia, Bruno. 2017. Guarani Morphology in Paraguayan Spanish: Insights from Code-Mixing Typology. Hispania 100(1):47–64.
Fernández, Karina Laura. 2008. Lenguas en contacto: el caso guaraní/español en la República Argentina. Revista de Lenguas para Fines Específicos 13–14: 39–50.
Fernández, Karina Laura. 2015. Contacto guaraní-español en Corrientes y Buenos Aires: La lengua como un espacio de construcción de identidades. PhD diss., Buenos Aires Universidad Católica Argentina.
Fernández-García, Marisol. 1999. Patterns of gender agreement in the speech of second language learners. In Javier, Gutiérrez-Rexach and Fernando, Martínez-Gil (eds.), Advances in Hispanic Linguistics: Papers from the 2nd Spanish Linguistics Symposium: 3–15. Somerville: Cascadilla Press.
Finnemann, Michael D. 1992. Learning Agreement in the Noun Phrase: The Strategies of Three First-Year Spanish Students. International Review of Applied Linguistics in Language Teaching (IRAL) 30(2): 121–136.
Franceschina, Florencia. 2005. Fossilized Second Language Grammars: The acquisition of grammatical gender. Amsterdam: John Benjamins Publishing Company.
Gibson, Edward. 1998. Linguistic complexity: locality of syntactic dependencies. Cognition 68: 1–76.
Gómez Rendón, Jorge. 2008. Typological and Social Constraints on Language Contact: Amerindian Languages in Contact with Spanish: PhD diss., Amsterdan. University of Amsterdam.
Granda, Germán de. 1979. Calcos sintácticos del guaraní en el español del Paraguay. Nueva Revista de Filología Hispánica, 28(2): 267–286.
Granda, Germán de. 1980a. Préstamos morfológicos del guaraní en el español del Paraguay (I). Revue de Linguistique Romane 44(173–174): 57–68.
Granda, Germán de. 1980b. Préstamos morfológicos del guaraní en el español del Paraguay (ii). Revue de Linguistique Romane 44(175–176): 247–265.
Granda, Germán de. 1996. Interferencia y convergencia sintácticas e isogramatismo amplio en el español paraguayo. International Journal of the Sociology of Language 117(1): 63–80.
Grant, Anthony P. 2007. Some aspects of np s in Mindanao Chabacano: Structural and historical considerations. In Marlyse, Baptista and Jacqueline, Guéron (eds.), Noun Phrases in Creole Languages: A multi-faceted approach, chap. 6, 173–204. Amsterdam: John Benjamins Publishing Company.
Grüter, Theres, Casey, Lew-Williams and Anne, Fernald. 2012. Grammatical gender in L2: A production or a real-time processing problem? Second Language Research 28(2): 191–215.
Guillán, María Isabel. 2010. La variedad de español del nea: Huellas del contacto español-guaraní. In Actas del IX Congreso Argentino de Hispanistas: El Hispanismo ante el Bicentenario, La Plata: Universidad Nacional de La Plata.
Gynan, Shaw N. 2011. Spanish in Contact with Guaraní. In Manuel, Díaz-Campos (ed.), The Handbook of Hispanic Sociolinguistics, chap. 17, 353–373. Oxford: Blackwell Publishers.
Gynan, Shaw. 2017. Morphological Glossing Conventions for the Representation of Paraguayan Guaraní. In Bruno, Estigarribia and Justin, Pinta (eds.), Guarani Linguistics in the 21st Century, chap. 3, 86–130. Leiden: Brill Publishers.
Gynan, Shaw Nicholas & Ernesto Luís López, Almada. 2020. The glottal stop in Guaraní and Paraguayan Spanish. In Rajiv, Rao (ed.), Spanish Phonetics and Phonology in Contact: Studies from Africa, the Americas, and Spain, 227–261. Amsterdam: John Benjamins Publishing Company.
Hager, Malin. 2014. Der Genuserwerb bei mehrsprachig aufwachsenden Kindern: Einelongitudinale Untersuchung bilingualer und trilingualer Kinder der Sprachenkombinationen deutsch-französisch/italienisch/spanisch, französisch-italienisch/spanisch und deutsch-spanisch-katalanisch. PhD diss., Wuppertal. University of Wuppertal.
Harris, James W. 1991. The Exponence of Gender in Spanish. Linguistic Inquiry 22(1). 27–62.
Hawkins, Roger. 1998. The inaccessibility of formal features of functional categories in second language acquisition. Paper presented at the Pacific Second Language Research Forum, Tokyo, Japan, March 1998.
Hawkins, Roger and Florencia, Franceschina. 2004. Explaining the acquisition and non-acquisition of determiner-noun gender concord in French and Spanish. In Philippe, Prévost and Johanne, Paradis (eds.), The Acquisition of French in Different Contexts: Focus on Functional Categories: 175–205. Amsterdam: John Benjamins Publishing Company.
Herbert, Robert K. 1985. Gender systems and semanticity: Two case histories from Bantu. In Jacek, Fisiak (ed.), Historical Semantics – Historical Word-Formation: 171–198. Berlin: De Gruyter.
Herring, Elizabeth. 2012. Grammatical Gender in Paraguayan Spanish. Paper presented at the Ninth Annual Graduate Student Conference on Luso-Brazilian and Hispanic Linguistics, Literature and Culture, Indiana University, 24 March 2013.
Herring, Elizabeth. 2017. La in Paraguayan Spanish and Guaraní. Poster presented at the 2017 Hispanic Linguistics Symposium, Texas Tech University, 26–28 October 2017.
Igoa, José Manuel, José Eugenia, García-Albea and Rosa, Sánchez-Casas. 1999. Gender-number dissociations in sentence production in Spanish. Rivista di Linguistica 11(1): 163–196.
Isabelli-Garcia, Christina. 2010. Acquisition of Spanish Gender Agreement in Two Learning Contexts: Study Abroad and At Home. Foreign Language Annals 43(2): 289–303.
Janse, Mark, Brian D., Joseph and Gunther, De Vogelaer. 2011. Changing gender systems: A multidisciplinary approach. Folia Linguistica 45(2): 237–244.
Keating, Gregory D. 2010. The effects of linear distance and working memory on the processing of gender agreement in Spanish. In Bill, Van Patten and Jill, Jegerski (eds.), Research in Second Language Processing and Parsing, 113–134. Amsterdam: John Benjamins Publishing Company.
Kouwenberg, Silvia and Eric, Murray. 1994. Papiamentu. Munich: Lincom Europa.
Lew-Williams, Casey and Anne, Fernald. 2010. Real-time processing of gender-marked articles by native and non-native Spanish speakers. Journal of Memory and Language 63(4): 447–464.
Liceras, Juana M., Raquel Fernández, Fuertes, Susana, Perales, Rocío, Pérez-Tattam and Kenton Todd, Spradlin. 2008. Gender and gender agreement in bilingual native and non-native grammars: A view from child and adult functional-lexical mixings. Lingua 118(6): 827–851.
Lipski, John M. 2006. El dialecto afroyungueño de Bolivia: en busca de las raíces del habla afrohispánica. Revista Internacional de Lingüística Iberoamericana 4(2): 137–166.
Lipski, John M. 2008. Afro-Bolivian Spanish. Madrid: Iberoamericana Vervuert.
Lipski, John M. 2015. From ‘more’ to ‘less’: Spanish, Palenquero (Afro-Colombian creole) and gender agreement. Language, Cognition and Neuroscience 30(9): 1144–1155.
Lleó, Conxita. 1998. Proto-articles in the acquisition of Spanish: Interface between phonology and morphology. In Ray, Fabri, Albert, Ortmann and Teresa, Parodi (eds.), Models of Inflection, 175–195. Tübingen: Max Niemeyer Verlag.
Lohndal, Terje and Marit, Westergaard. 2016. Grammatical Gender in American Norwegian Heritage Language: Stability or Attrition? Frontiers in Psychology 7, 344.
López-Ornat, Susana. 1997. What Lies in between a Pre-Grammatical and a Grammatical Representation? Evidence on Nominal and Verbal Form-Function Mappings in Spanish from 1;7 to 2;1. In Ana Teresa, Pérez-Leroux and William R., Glass (eds.), Contemporary Perspectives on the Acquisition of Spanish, 3–20. Somerville: Cascadilla Press.
Mariscal, Sonia. 2008. Early acquisition of gender agreement in the Spanish noun phrase: starting small. Journal of Child Language 35: 1–29.
McWhorter, John H. 2002. What happened to English? Diachronica 19(2): 217–272.
Michnowicz, Jim. 2012. The Standardization of Yucatan Spanish: Family Case Studies in Izamal and Mérida. In Kimberly, Geeslin and Manuel, Díaz-Campos (eds.), Selected Proceedings of the 14th Hispanic Linguistics Symposium, 102–115. Somerville: Cascadilla Proceedings Project.
Milsark, Gary Lee. 1974. Existential sentences in English. PhD diss., Massachusetts Massachusetts Institute of Technology.
Montrul, Silvina. 2013. Bilingualism and the Heritage Language Speaker. In Tej K., Bhatia and William C., Ritchie (eds.), The Handbook of Bilingualism and Multilingualism: 168–189. Malden: Wiley-Blackwell.
Montrul, Silvina, Rebecca, Foote and Silvia, Perpiñán. 2008. Gender Agreement in Adult Second Language Learners and Spanish Heritage Speakers: The Effects of Age and Context of Acquisition. Language Learning 58(3): 503–553.
Morínigo, Marcos Augusto. 1990. Raíz y destino del guaraní. Asunción: Biblioteca Paraguaya de Antropología, Universidad Católica de Asunción.
Müller, Natascha. 1994. Gender and Number Agreement within dp. In Jürgen M., Meisel (ed.), Bilingual First Language Acquisition: French and German grammatical development: 53–88. Amsterdam: John Benjamins Publishing Company.
Müller, Natascha. 2000. Gender and Number in acquisition. In Barbara, Unterbeck, Matti, Rissanen, Terttu, Nevalainen and Mirja, Saari (eds.), Gender in Grammar and Cognition: 351–400. Berlin: De Gruyter.
Opitz, Andreas and Thomas, Pechmann. 2016. Gender Features in German: Evidence for Underspecification. The Mental Lexicon 11(2): 216–241.
Orozco, Antonio Millán. 1970. Anomalías de concordancia del nombre en el español de la ciudad de México. Anuario de Letras: Lingüística y Filología 8: 125–146.
Otheguy, Ricardo and Naomi, Lapidus. 2003. An adaptive approach to noun gender in New York contact Spanish. In Rafael, Núñez-Cedeño, Luis, López and Richard, Cameron (eds.), A Romance Perspective on Language Knowledge and Use, 209–232. Amsterdam: John Benjamins Publishing Company.
Palacios Alcaine, Azucena. 2000. El sistema pronominal del español paraguayo: Un caso de contacto de lenguas. In Julio Calvo, Pérez (ed.), Teoría y práctica del contacto: el español de América en el candelero, 123–143. Madrid: Iberoamericana Vervuert.
Pérez-Pereira, Miguel. 1991. The acquisition of gender: What Spanish children tell us. Journal of Child Language 18(3): 571–590.
Pinta, Justin and Jennifer L., Smith. 2017. Spanish Loans and Evidence for Stratification in the Guarani Lexicon. In Bruno, Estigarribia and Justin, Pinta (eds.), Guarani Linguistics in the 21st Century, chap. 9, 285–314. Leiden: Brill Publishers.
Polinsky, Maria. 2008. Gender under Incomplete Acquisition: Heritage Speakers’ Knowledge of Noun Categorization. Heritage Language Journal 6(1): 40–71.
Pruñonosa, Manuel. 2000. Algunos rasgos fónicos de interferencia del guaraní en el español del Paraguay. In Julio Calvo, Pérez (ed.), Teoría y práctica del contacto: el español de América en el candelero, 113–122. Madrid: Iberoamericana Vervuert.
Sabourin, Laura L. 2001. L1 effects on the processing of grammatical gender in L2. In Susan H. Foster-Cohen and Anna, Nizegorodcew (eds.), EUROSLA Yearbook, vol. 1, 159–169. Amsterdam: John Benjamins Publishing Company.
Sagarra, Nuria and Julia, Herschensohn. 2010. The role of proficiency and working memory in gender and number agreement processing in L1 and L2 Spanish. Lingua 120(8): 2022–2039.
Sánchez, Liliana. 1996. Word order, predication and agreement in dp s in Spanish, Southern Quechua and Southern Andean bilingual Spanish. In Karen, Zagona (ed.), Grammatical Theory and Romance Languages: Selected Papers from the 25th Linguistic Symposium on Romance Languages, 209–218. Amsterdam: John Benjamins Publishing Company.
Sandoval, Gabriela González. 2006. Las tranferencias del guaraní en la conversación colloquial de Corrientes. In Wolf, Dietrich and Haralambos, Symeonidis (eds.), Guaraní y “Mawetí-Tupí-Guaraní”: Estudios históricos y descriptivos sobre una familia lingüística de América del Sur, 415–432. Berlin: LIT.
Sessarego, Sandro. 2010. On gender and number agreement in the determiner phrase: the Afro-Bolivian Spanish case. Sintagma 22: 131–147.
Sessarego, Sandro. 2011. On the Status of Afro-Bolivian Spanish Features: Decreolization or Vernacular Universals?In Jim, Michnowicz and Robin, Dodsworth (eds.), Selected Proceedings of the 5th Workshop on Spanish Sociolinguistics, 125–141. Somerville: Cascadilla Proceedings Project.
Sessarego, Sandro and Javier, Gutiérrez-Rexach. 2011. A Minimalist approach to gender agreement in the Afro-Bolivian dp: Variation and the specification of uninterpretable features. Folia Linguistica 45(2): 465–488.
Teschner, Richard V. 1983. Spanish Gender Revisited: -z Words as Illustrating the Need for Expanded Phonological and Morphological Analysis. Hispania 66(2): 252–256.
Tesoro, Ana Isabel García. 2002. El español en contacto con lenguas mayas: Guatemala. In Azucena, Palacios and Ana Isabel, García (eds.), El indigenismo Americano III, 31–60. Valencia: Universitat de València.
Tesoro, Ana Isabel García. 2010. Español en contacto con el tzutujil en Guatemala: Cambios en el Sistema pronominal átono de tercera persona. Revista Internacional de Lingüística Iberoamericana 8(1): 133–156.
Tessen, Howard W. 1974. Some Aspects of the Spanish of Asunción, Paraguay. Hispania 57(4): 935–937.
Thomason, Sarah G. and Terrence, Kaufman. 1988. Language Contact, Creolization, and GeneticLinguistics. Berkeley: University of California Press.
Thun, Harald. 2006. “A dos mil la uva, a mil la limón”: Historia, función y extensión de los artículos definidos del castellano en el guaraní jesuítico y paraguayo. In Wolf, Dietrich and Haralambos, Symeonidis (eds.), Guaraní y “Mawetí-Tupí-Guaraní”: Estudios históricos y descriptivos sobre una familia lingüística de América del Sur, 357–414. Berlin: LIT.
Trawick, Sonya and Jim, Michnowicz. 2019. Glottal Insertion before Vowel-Initial Words in the Spanish of Asunción, Paraguay. In Gregory L. Thompson and Scott M. Alvord (eds.), Contact, Community, and Connections: Current Approaches to Spanish in Multilingual Populations, chap. 6, 147–174. Wilmington: Vernon Press.
Trudgill, Peter. 1999. Language contact and the function of linguistic gender. Poznań Studies in Contemporary Linguistics 35: 133–152.
Van Coetsem, Frans. 1988. Loan Phonology and the Two Transfer Types in Language Contact. Dordrecht: Foris.
Van Naerssen, Margaret. 1986. Hipótesis sobre la adquisición de una segunda lengua, consideraciones inter-lenguaje: comprobación en el español. In Jürgen M., Meisel (ed.), Adquisición del lenguaje – Aquisição da linguagem, 139–155. Frankfurt: Vervuert.
Velázquez-Castillo, Maura. 1994. Possessive constructions in Paraguayan Spanish: the influence of Guaraní. RLA: Romance Languages Annual (6): 607–613.
White, Lydia, Elena, Valenzuela, Martyna, Kozlowska-Macgregor and Yan-kit Ingrid, Leung. 2004. Gender and number agreement in nonnative Spanish. Applied Psycholinguistics 25: 105–133.
Winford, Donald. 2007. Some issues in the study of language contact. Journal of Language Contact Thema 1: 22–39.
Note that this is distinct from cases where agreement is variable due to a conflict between semantic and syntactic gender assignment. See Corbett 2006: 214–218 for a discussion of how patterns of variable gender agreement of this type fit into larger crosslinguistic patterns of agreement (i.e., Corbett’s Agreement Hierarchy).
Counterexamples exist outside of the context of purely morphosyntactic operations such as those discussed here. Orozco (1970) for instance discusses the use of nosotros ‘we [masculine]’ and uno ‘one [masculine]’ when the referent is a woman.
Tesoro (2010) does provide instances, although scarce, of gender nonagreement in the direct object system of monolingual speakers of Guatemalan Spanish. Comparisons between this and Correntino Spanish are not productive however given widespread leísmo (i.e., the use of gender-neutral le as an invariable direct object pronoun where standard Spanish uses masculine lo or feminine la) among Correntino speakers.
Non-systematic use of both masculine and feminine Spanish definite articles is however reported in Mindanao Chabacano (Grant, 2007: 178).
Cuervo and Mazzaro (2013) provide a noteworthy analysis of patterns of negation in Correntino Spanish which are linked to Guarani contact. Bilingualism and language dominance are excluded from their study however, which is only concerned with Guarani contact as part of the diachronic context to their synchronic analysis of other social and linguistic factors.
Original text: “el muy notorio resquebrajamiento en el español paraguayo del sistema de concordancias genéricas existente en el sintagma nominal del español común, hecho en el que ha actuado causalmente, como factor propulsor exclusivo, la inexistencia en guaraní de marcas de género.”
Both images are taken from Wikimedia Commons and are used under the terms of the gnu Free Documentation License; the national map (https://commons.wikimedia.org/wiki/File:Provincia_de_Corrientes,_Argentina.png) is unaltered while the provincial map (https://commons.wikimedia.org/wiki/File:Map_of_Corrientes_Province.svg) is altered for the purposes of this article.
There were actually two instances of this reverse pattern in the corpus, in which a masculine noun took morphologically feminine modifiers:
That is, both la casa blancaand la casa blanco are possible but not *el casa blanco or *el casa blanca.
Whether or not a given np was classified as definite or indefinite was based on the typology in Abbott 2004.
“Determiner” here is used to mean demonstratives (e.g., este ‘this [masculine, singular]’), possessive determiners (e.g., nuestro, ‘our [masculine, singular]’), and quantifiers (e.g., muchos ‘many [masculine, plural]’). Other canonical determiners were excluded due to their not having variable gender morphology in Spanish, e.g., cada ‘each’ which is gender-neutral in spite of its ending in -a, i.e., cada mujer ‘each woman’, cada hombre ‘each man’, *cado hombre. Additionally, as mentioned before, articles are excluded due to their not showing variation.
The possibility of other morphophonological factors playing a role was also considered, such as whether or not a Spanish noun with Guarani phonological shape would be less likely to show agreement. Given that in many ways the native phonologies of the two languages overlap, that Spanish phonological structures have been imported into Guarani (Pinta and Smith, 2017), and that in my experience bilingual speakers are sometimes unaware of the source of a given lexical item if it appears in both languages, such factors were left out of this analysis.
The singular exception is the noun idioma ‘language’ which is irregular and masculine in standard Spanish (i.e., el idioma, un idioma, etc.) but feminine for many speakers of Correntino Spanish (i.e., la idioma, una idioma).
A reviewer raises the point of the various inflectional classes in Spanish. The canonical -o/-a class (e.g., bueno ‘good [masculine]’, buena ‘good [feminine]’) contrasts with the -e class (e.g., interesante ‘interesting’); lexical items in the latter class are invariable regardless of the gender of the noun phrase. Additionally, there are borrowed morphemes in Spanish which are also invariable, e.g., the Greek-origin agentive -ista. The variation in Correntino Spanish in some ways resembles these invariable cases, yet is different in the sense that it occurs in instances in which Spanish has the morphological means by which to communicate gender; the -e class and morphemes like -ista are invariable because gender inflection in these contexts is not morphologically possible (*-isto for instance does not exist). The variation in Correntino Spanish concerns morphologically masculine modifiers, i.e., instances in which -o adjectives, for instance, can function in both masculine and feminine noun phrases. There is no morphological class in standard Spanish in which -o forms can act invariably, making an explanation of the Correntino Spanish phenomenon on the basis of analogy to a standard Spanish inflectional class unlikely.
Consider Harris’ (1991: 43) example: Tienes demasiados “paras” en ese párrafo [You have too many “paras” in that paragraph]. Here the metalinguistically mentioned preposition para ‘for’, which cannot have inherent gender, is modified by the morphologically masculine (and plural) modifier demasiados.
Special thanks to an anonymous reviewer for insightful discussion along these lines.
As pointed out by Winford (2007: 26), this latter category is roughly used in the same way as the terms “interference via shift”, “transfer”, “indirect diffusion”, and “substratum influence” elsewhere in the literature.
This is further supported by Morínigo (1990: 111) who provides el as the Argentine Guarani equivalent to Paraguayan Guarani’s la, although erroneously seeming to imply that el is used for both masculine and feminine nouns: “…la adopción del artículo español la (el en el dialecto correntino) como señalador del nombre singular masculino o femenino…” […the adoption of the Spanish article la (el in the Correntino dialect) as a marker of both masculine and feminine singular nouns…] (my translation).
One unexpected form is el mesã ‘the table’ (Cerno, 2010: 7), which comes from Spanish la mesa and therefore might have reasonably been expected to appear in Argentine Guarani as la mesã, following the pattern of ekuéla and pare in Table 3.
The status of these variable cases is unresolved. Whether they have truly ambiguous grammatical gender, and whether or not there are other forms like them, are questions for future research.
Use of both el and la is however attested in Paraguayan Guarani from the mid 19th century (Thun, 2006: 389–391), around the time that it and Argentine Guarani were undergoing a period of linguistic divergence (Cerno, 2013: 34). This suggests that the use of both el and la in Argentine Guarani could be a retention of a pre-Argentine Guarani development and not an innovation. That is, the disappearance of el from Paraguayan Guarani took place after it and Argentine Guarani had diverged, leaving Argentine Guarani unaffected by this change.
It is worth mentioning that in both Argentine and Paraguayan Guarani, the use of Spanish-source definite articles differs from definite article use in Spanish. In addition to use as an article they have a variety of other functions, e.g., as a relative pronoun, in genitive np s, etc. (see Cerno, 2010 for Argentine Guarani; see Thun, 2006 and Herring, 2017 for Paraguayan Guarani).
“Weak” quantifiers are those which can occur in existential sentences (e.g., some, many) while “strong” quantifiers are those which cannot (e.g., all, most); see Milsark 1974.
See for instance Corbett (2006: 214–218) for a crosslinguistic discussion of gender as it relates to Corbett’s Agreement Hierarchy.