Gender agreement variation in the Afro-Bolivian determiner phrase: the interplay of social and linguistic factors.
This study investigates gender agreement variation across the Determiner Phrase (DP) in Afro-Bolivian Spanish (ABS), an Afro-Hispanic dialect spoken in Los Yungas, Department of La Paz, Bolivia.
Descriptive articles about ABS have been published during the last years by John Lipski (2006a,b,c,d), who analyzed qualitatively the differences encountered between this dialect and other Spanish varieties; however a quantitative study has not yet been done. The purpose of this work is therefore to fill the gap and shed some light on the linguistic constraints regulating gender agreement; moreover, sociolinguistic inferences will be derived from the data presented.
To accomplish this second goal, it is fundamental to provide a brief historical overview of Afro-Bolivians and their language.
Determining exactly how black communities populated Los Yungas is an historic issue that has not yet been completely solved. The first records attestingthsignificant black slavery in Los Yungas are dated 18th century; they are documents concerning the inventory of goods belonging to some plantations, or sale certificated that attested the passage from one owner to another (Portugal Ortiz 1977: 77-83).
Since that period till 1952, when the Land Reform took place, Afro-Bolivians have been used in Los Yungas as slaves in haciendas. After the Land Reform, the majority of them remained in the region becoming the new owners of small parts of the land that once belonged to the plantation, where they used to work in slavery.
Typically, until 1952 black workers were not allowed to attend school. Several older members of these communities are therefore nearly or totally illiterate. However, after that date, the hacienda system ended and basic public education began to arrive in Afro-Yungueno communities. The study of Spanish at schools resulted in a gradual drop of the traditional dialect by Afro-Bolivians, so that some features of this vernacular have gradually been displaced by Highland Bolivian Spanish (HBS) ones. Due to this relatively recent contact with regional Bolivian Spanish, ABS is undergoing a "change in progress", consisting in the systematic substitution of stigmatized basilectal ABS features with more prestigious HBS ones. As far as ABS gender marking is concerned, this substitution is not random. Rather, what can be observed is the transition from one agreement system to another, according to specific syntactic constrains.
The focus of the present study is on the social and linguistic implications characterizing this transition. The rest of the paper is organized in the following sections: a brief illustration of ABS DP features, a description of data and methodology, results, and conclusion.
2. On Afro-Bolivian DP features
Lipski (2006b) points out five features as quintessentials of Afro-Yungueno DP, which distinguish this dialect from other Spanish varieties: (a) lack of noun-adjective gender agreement; (b) invariant plurals, that is, no plural suffixes on nouns, adjectives, or determiners; (c) use of a single invariant plural definite article; (d) elimination of definite articles in generic constructions; (e) frequently, the retention of plural /s/ only on first element of plural DP.
Cases of (a) and (b) can be exemplified by the following examples: siempre contaba algunos cosa [algunas cosas] 'he always told some things'; esos fiesta [esas fies tas] 'those parties'. Lipski notices that cases of gender and number inflection are more likely to appear on determiner and pre-nominal adjectives, while postnominal adjectives present a lower rate of concord: esa casa chico [chica] "that small house'.
In ABS a robust presence of single invariant plural definite article (lu) is found (c): lu taza [las tazas] "the cups,' lu juamia [las familia] "the families. However, due to this process of "feature substitution" instances in which the article agree in gender and number with the noun can sometimes be heard: las novias 'the girlfriends' Lipski (2006d). The definite article may also be eliminated in generic constructions (d), while it is required in other Spanish dialects: perro ta flojo [los perros estan flojos] "dogs are worthless'.
The last quintessential Afro-Bolivian DP feature reported by Lipski is the frequent retention of plural /s/ only on the first element of plural DPs (e): en idioma antigo di mis abuelo [en el idioma antiguo de mis abuelos] "in the old language of my grandparents'.
Due to the aforementioned features, ABS has been classified by Lipski (2006d: 9) as a case of "DP impoverished agreement". While in standard spec-head agreement processes a certain feature should percolate to all elements c-commanded by the determiner, in cases of "impoverished agreement", concord can be limited to some of them. DP features are normally claimed to percolate up from the noun to the determiner (Grimshaw 1991; 1997); interestingly, no case of post-nominal gender or number concord is found unless pre-nominal elements agree una curva ancha, una curva ancho, un curva ancho, * un curva ancha 'a large curve'; lus guaguas jovenes, lus guagua joven, lus guaguas joven, * lus guagua jovenes 'the young kids' (Lipski 2006d).
My data are generally in line with those reported by Lipski, the only difference that I could notice, mostly in the speech of the eldest informants, is the almost complete lack of gender and number agreement not only on the plural definite article lu but also on quantifiers 'habia harto caballo(s)' [habia hartos caballos] 'there were many hourses', 'todo las cosa' [todas las cosas] 'all the things'.
When comparing plural and gender marking, Lipski claims that the later is weaker than the first one, even though both may disappear in Afro-Bolivian speech. This would be due to the fact that numeric features are easier to acquire than gender ones as they mark semantically a prominent distinction (one versus many), whereas grammatical gender concord is almost always semantically empty (Lipski 2006a). In line with the same principle, I noticed that plural --s realization is not required neither on the definite article lu nor on quantifiers. In Standard Spanish the masculine plural determiner undergoes a stem vowel change (el [flecha diestra] los), which already indicate plurality, -s marking therefore does not affect the meaning of the determiner. The same is true for ABS, with the only difference that in the most basilectal variety, lu(s) is the only plural article, used for both masculine and feminine nouns. Due to contact with regional varieties of Spanish, some instances of plural feminine (las) were encountered in the speech of my informants, but however lu(s) was always conveying plurality. Differently from Standard Spanish, where plural --s on quantifiers generally distinguishes countable from uncountable nouns mucha agua 'much wather' vs. muchas botellas 'many bottles', in traditional ABS such a distinction is not present and quantifiers, non-inflected neither for gender nor for number, can be used to convey both meanings mucho agua 'a lot of water', mucho botella(s) 'a lot of bottles'.
Cases concerning quantifiers and lu(s) seem to violate the pre-nominal to post-nominal percolation order (Grimshaw 1991; 1997), however they do not. Such elements in traditional ABS already convey plurality (1b3b,5) without having to be followed by --s marking. Besides, the quantifier phrase (QP), which is external to the DP (5), does not contradict Grimshaw's model. In fact, while tokens like (4) are not found, only in cases like (5) the pre-nominal to post-nominal gender percolation order does not affect Q. On the other hand, when Q immediately precedes N, if percolation is present (complete or partial), the element in D position will be affected (la-3a,1b-3b).
For these reasons it can be said that in ABS agreement is weak and locally constrained. It gets checked first on the element in the determiner position, it can percolate to the post-nominal elements, and in the grammar of certain speakers --the eldest ones-, it is not specified in quantifiers and in the plural definite article lu.
For the purpose of this article, I will not focus on points (b) and (e) (Lipski 2006b), leaving for another study the case of number marking. I will rather analyze gender agreement across the DP (a, c, d) for both canonically and non-canonically feminine marked words, for pre-nominal and post nominal elements, and for grammatical categories.
3. Data and Methodology
2604 tokens were extracted from a corpus of 13 recorded interviews I taped during summer 2008, for a total of almost 13 hours of conversation with Afro-Bolivian speakers residing in the communities of Tocana, Mururata and Chijchipa, North Yungas. The interviews were conducted by letting the speaker talk about any topic of their liking and asking them follow-up questions, in line with the principle of Tangential Shift (Labov 1984:37). The goal was therefore to attempt to reduce the Observer's Paradox (Labov 1972) as much as possible.
The speakers were selected in order to have individuals belonging to different groups according to the external factor groups selected, namely: Generation, Gender, Level of Education, and Mobility.
Within the group Generation, I contrasted four different categories of speakers according to their age: 2140; 41-60; 61-80; 81+. What I expected to find was a progressive increase in the DP gender disagreement in older generations as a result of the lack of education, the racial segregation, and the haciendasystem, which kept Afro-Bolivians as slaves till the 1952, year of the Land Reform.
Within the factor group Gender the categories contracted were obviously Male vs. Female. Gender has often proved to be a significant factor in cases of variation due to change in progress (Labov 1990). Moreover, women in Afro-Bolivian communities are those who sell agriculture products at the market and have therefore greater contact with the outside word and with HBS.
As far as the Level of Education is concerned, three different groups were identified:
A) Illiterate speakers
B) Speakers that attended the primary school
C) Those that have higher education beyond primary school
Education is a crucial factor; the exposure of members of the black community to HBS taught at school had an effect on traditional ABS. This change should therefore be reflected in the presence of a higher DP gender agreement in the speech of the most educated speakers.
Mobility is another important element. Since the approximation towards the Bolivian Spanish norm originated from external linguistic pressure, it must be expected that Afro-Bolivians who spent more time outside the communities (at least one year) would present a higher level of gender concord than those who did not.
My analysis took into consideration also internal/linguistic factors, namely: the Coordinated Element Position in the DP (pre-nominal/post-nominal), the Transparency of the noun, and the Grammatical Category of the element.
The Coordinated Element Position is crucial for agreement probability according to the pre-nominal to post-nominal percolation order proposed by Grimshaw (1991; 1997).
Besides, transparent nouns are contrasted with nontransparent ones. As in the process of acquiring a language speakers tend to rely on its regularities, I am assuming that feminine transparent nouns ending with the prototypical feminine marker --a (e.g. casa 'house') would present a higher level of agreement than those being not transparent e.g. nacion 'nation', etc.
Grammatical Category was the last factor group considered. Coordinated Element Position alone can give us an idea of the percolation pattern, however as noted before, certain elements (quantifiers and the plural definite article lu(s) even though tend to occur pre-nominally, in ABS are often not subjected to gender and number agreement processes. Coding for Grammatical Category is therefore needed in order to get a better picture of the gender agreement patterns found in the ABS DP.
My data were analyzed using GoldVarbX program (Sankoff , Tagliamonte, & Smith 2005), which calculates probabilities for the application of a given rule. Results were extracted by operazionalizing the 2604 tokens in one GoldVarbX analysis containing both internal/linguistic factors and external/social ones.
All external factor groups selected resulted to be significant but one, Gender (Table1).
Education resulted to be the most significant factor group (Range 57). Afro-Bolivians who attended secondary education or higher (C) present reduced disagreement phenomena (Factor Weight.10), on the other hand, those who attended only primary school (B) or had no education at all (A) show higher levels of lack of concord, reflected in their relatively higher Factor Weights, .58 and.67, respectively.
Generation is another crucial factor (Range 54), with 80+ strongly favoring disagreement (Factor Weight.66) and 21-40 disfavoring it (Factor Weight.23). These data reflect the presence of a change in progress pushing ABS in the direction of HBS. Young generations did not experienced the segregation imposed by the haciendasystem and had more chances to have contact with the Spanish spoken outside the community. These elements, in addition to the stigmatization attached to the Afro vernacular, are pushing the younger members of the community to quickly replace the basilectal features with more prestigious HBS ones.
Mobility is the third significant factor group (Range 18). Speakers who spent more than one year outside the community present higher levels of DP gender agreement (Factor Weight.38) if compared with those who never moved or only for a shorter period of time (Factor Weight.56). It is a result that had to be expected, as the underlying reason for this change in progress is linguistic pressure originated outside the community.
Gender did not seem to play a significant role in regulating the variation. I expected women to lead the process of change (Labov 1990), especially because of their work at the market. However, it seems that the differences in contact with HBS induced by male/female subdivision of work are not enough as to affect the phenomenon significantly.
As far as internal factors are concerned, all factor groups selected resulted to be significant (Table 2).
Grammatical Category is the most significant factor group (Range 60). As factor weights were very similar, I decided to collapse quantifiers and indefinite articles under the group Indefinite, and demonstratives, possessives and definite articles under the group Definite. Definite elements disfavor disagreement (Factor Weight .27), while indefinite ones favor it (Factor Weight.82). This could be due to their relative position within the DP. In fact, definite elements are claimed to be in lower projections, closer to the noun, than indefinite ones (Bosque & Gutierrez-Rexach 2009). Adjectives are found in between (Factor Weight.72) and favor disagreement (Factor Weight.72).
Results for Position indicate that it is statistically less likely to find post-nominal concord (Factor Weight .86) than pre-nominal one (Factor Weight.44). This findings, combined with the fact that post-nominal agreement is not found unless also pre-nominal one is present, further back the claim that features percolate up from the noun to the determiner and eventually to the other projections. Moreover, adjectives (N 492), which taken together show a factor group of.72, when sub-categorized and run according to their location in the DP, bring additional support to such model: post-nominal adjectives (N 220, Factor Weight.65), post-nominal adjectives (N 272, Factor Weight.90).
Transparency proved to be significant too (Range 10). Disagreement on elements receiving gender assignation from prototypical feminine nouns (Factor Weight.48) is lower than on those whose gender assignation depend on non-prototypical ones (Factor Weight.58).
This study offers a quantitative approach to variable gender agreement across the DP in ABS, an Afro vernacular dialect spoken in Los Yungas, Department of La Paz, Bolivia. Results indicate a clear case of change in progress, consisting in the systematic substitution of basilectal Afro-Bolivian features with contemporary Bolivian Spanish ones.
The underlying reasons pushing Afro-Bolivian speech in the direction of a more prestigious Spanish variety are essentially the stigmatization of the Afro variety and the increasing contact with a more prestigious Spanish dialect. Contact with Bolivian Spanish increased sensibly after 1952, year of the Bolivian Land Reform, which freed Afro-Bolivians from slavery and introduced education in the black communities. These changes, which affected the socio-economical scenario of black Bolivia during the last six decades, are reflected in the speech of the members of its community. For this reason Gender, Mobility and Education proved to be significant factor groups affecting gender agreement.
While among the external factors considered gender did not show particular significance, the internal factors selected resulted to have an important effect on the concord phenomenon. On one hand, Transparency of nouns resulted to be indicative of a higher rate of cross DP gender agreement. This is motivated by the fact that in the process of acquiring a language, speakers rely on its regularities to build their grammar and in the case of Spanish gender, the ending morphemes --a for feminine and --o for masculine represent two almost always reliable indicators. On the other hand, the absence of tokens indicating post-nominal agreement preceded by prenominal non-agreeing elements strongly suggests that DP features percolate up from the noun to the determiner (Grimshaw 1991: 37-39; 1997) and can eventually reach the extended projection. However, as noticed, in the most traditional dialect quantifiers and the definite plural article lu(s) seem to be exempt from agreement processes. Besides, variation also appears to depend on the definiteness of the element considered, which indirectly indicates projection proximity to the noun (Bosque & Gutierrez-Rexach 2009).
From a theoretical perspective, my results document a case of change in progress, in which an Afro-Hispanic vernacular approximates to a more prestigious Spanish dialect. The process is driven by social factors through a path, which is highly constrained by syntactic ones.
BOSQUE, Ignacio and Javier GUTIERREZ-REXACH. Fundamentos de Sintaxis Formal. Ediciones Akal: Cambridge University Press, 2009.
GRIMSHAW, Jane. Extended Projection. MS. Brandeis University. 1991.
-- "Projection, Heads, and Optimality." Linguistic Inquiry 28 (1997): 373-422.
LABOV, William. "Language in the Inner City." Philadelphia: Univ. of Pennsylvania Press, 1972.
-- "Field Methods of the Project on Linguistic Change and Variation." Baugh and Sherzer (Eds.), Language in Use, Englewood Cliffs: Prentice Hall, 1984. 84-112.
-- "The Intersection of Sex and Social Class in the Course of Linguistic Change. Language Variation and Change 2 (1990): 205-254.
LIPSKI, John. "Afro-Bolivian Language Today: The Oldest Surviving Afro-Hispanic Speech Community." Afro-Hispanic Review 25.1 (2006a): 179-200.
--. "Morphosyntactic Implications in Afro-Hispanic Language: New Data on Creole Pathways." Presented at NWAV-35, Columbus, Ohio, November 10, 2006b.
--. El dialecto afroyungueno de Bolivia: en busca de las raices el habla afrohispanica. Revista Internacional de Linguistica Iberoamericana 3.2. (2006c): 137-166.
--. Afro-Bolivian Spanish and Helvetia Portuguese: Semi creole Parallels. Papia 16 (2006d): 96-116.
PORTUGAL ORTIZ, Max. "La esclavitud negra en las epocas colonial y nacional de Bolivia." La Paz: Instituto Nacional de Cultura 1977
SANKOFF, David, Sali TAGLIAMONTE and Eric SMITH. Goldvarb X: A Variable Rule Application for Macintosh and Windows. Department of Linguistics, University of Toronto. 2005.
Ohio State University
Table 1. Variable rule analysis of the contribution of exter- nal factors to the probability of lack of gender agreement in Afro-Bolivian DP (Total = 2604; Log likelihood = -637.3592; Total Chi-square = 377.9828; Chi-square/cell = 2.0591; Signif- icance = 0.039; Input = 0.031) Factor % Lack EDUCATION Weight Agreement N % data A (illiterate) .67 20 1124 43 E (primary) .58 11 993 38 G (secondary +) .10 0 487 19 Range 57 GENERATION 80+ .66 21 651 25 61-80 .61 18 604 23 41-60 .54 11 671 26 21-40 .23 1 678 26 Range 54 MOBILITY Inside .56 18 1774 68 Outside .38 2 830 32 GENDER Female [.51] 13 1203 46 Male [.48] 12 1401 54 Table 2. Variable rule analysis of the contribution of internal factors to the probability of lack of gender agreement in Afro-Bolivian DP (Total = 2604; Log likelihood = -637.3592; Total Chi-square = 377.9828; Chi-square/cell = 2.0591; Significance = 0.039; Input = 0.03) Factor % Lack Weight Agreement N % data GRAMMATICAL CATEGORY Indefinite (Quantif., Indef Art.) .82 40 636 25 Adjectives .72 34 492 19 Definite (Poss., Dem, Def. Art.) .27 2 1456 56 Range 60 POSITION Post-Nominal .86 49 300 12 Pre-Nominal .44 8 2304 88 Range 42 TRANSPARENCY Non-Transparent .33 20 373 22 Transparent .48 11 2031 78 Range 10
|Printer friendly Cite/link Email Feedback|
|Publication:||Revista Iberoamericana de Linguistica|
|Date:||Jan 1, 2009|
|Previous Article:||El fenomeno linguistico de la ironia desde la perspectiva pragmatica en relacion a la fraseologia en el aula de e/le.|
|Next Article:||La comunidad coreana en Tailandia aspectos sociales y linguisticos.|