- 3 -

Theoretical Linguistics

Lexical Leveling in Four New York City Spanish Dialects: Linguistic and Social Factors

Hunter College and the Graduate Center City University of New York

Five useful collections of articles on the varieties of Spanish spoken in the US. (Elias-Olivares, ed. 1978, Amastae and Elías-Olivares, eds. 1981, Durán, ed. 1981, Fishman and Keller, eds. 1982, Elías-Olivares, Leone, Cisneros, and Gutiérrez, eds. 1985) have appeared since the publication of the landmark annotated bibliography by Teschner, Bills, and Craddock (1975), but there are still many lacunae in our knowledge of the linguistic diversity in US. Latino populations. Only five articles in these volumes report on varieties other than Chicano or Puerto Rican Spanish. The need for linguistic research on other dialects becomes more urgent as the Latino population climbs toward the 25 million mark by the 1990 census, with increasing migrations from diverse Spanish-speaking nations. In New York City there are approximately two million Latinos from all parts of the Spanish speaking world; they constitute 20% of the city's total population. The presence of diverse groups of speakers of one language offers an invaluable opportunity to study the linguistic and socio-cultural effects of extensive contact among diverse dialects.

This study analyzes inter-dialectal contact at the lexical level among the four largest Hispanic groups in New York City -Puerto Ricans, Dominicans, Colombians, and Cubans. Specifically, we sought to investigate whether each group maintained its country's regional lexicon, assimilated that of the largest Spanish speaking group in the city or the most prestigious variety, or produced another, «New Yorker Spanish» lexicon, one shared with the majority of Hispanics⁵⁸.

Seminal research on the linguistic consequence of languages in contact was conducted by Weinreich (1953); Trudgill's more recent work pinpoints two of the principal gaps in our knowledge of dialects in contact:

He argues that a society's incorporation of technological innovations suggests a theoretical framework of great benefit to those concerned with language change, especially in regard to factors which influence the acceptance or rejection of an innovation, e.g., the automobile or the computer. Change, whether technological or linguistic, is the result of the interrelationship between «information», i.e., experience with the new item, and factors which promote or inhibit its acceptance. The most effective manner of receiving information is through personal contact with others. This is especially true regarding language, because, as Trudgill notes, «the media... have almost no effect at all in phonological or grammatical change» (1984:61), although they do contribute to popular acceptance and use of some new vocabulary. Even in lexical diffusion, other more powerful propagators or inhibitors of change include geographic barriers and distances which may isolate one group from another, and socially significant differences which include classes, races, genders, and generations. Of importance linguistically are the similarities and differences of the linguistic

––––––––   1095   ––––––––

codes which are in contact. A critical role is played by language attitudes, which can counteract the impact of the geographic, social, and/or linguistic reality. Attitudes are shaped by the historical, political, and economic experiences of each speech community. Each of these factors is considered in this investigation of dialectical contact at the lexical level in New York City's Spanish(es).

Puerto Ricans, Dominicans, Colombians, and Cubans arrived in the city at different periods and under different circumstances. Puerto Ricans were the first of these Spanish speaking groups to arrive in large numbers, primarily in the post World War II decade, and the only ones to come from a territory that belonged to the United States. Puerto Ricans come and go as U. S. citizens, and their migration fluctuates in accordance with push pull influences of the mainland-vs.-island economy. In 1980, Puerto Ricans constituted 61% of the Hispanic community in New York City (860,000). Since the liberalization of the immigration laws in 1965, other Hispanic groups have been growing at an even faster pace than the Puerto Rican community. There were approximately 400,000 Dominicans, 267,000 Central and South Americans -mainly Colombians-, and 61,000 Cubans legally registered in New York City in 1980 (Hispanic Research Center 1981). These figures can double when we take into account the census undercount of Hispanics, the influx since 1980 of 120,000 Dominicans, 24,000 Colombians and 4,000 Cubans, and the numbers of illegal residents. The official Hispanic population of New York City grew from 16% in 1970 to 20% in 1980; it is expected to reach 30% in the year 2000 and to comprise the largest group (35%) by 2010 (Bouvier and Briggs 1988).

Hispanic residential segregation is increasing throughout the US. (Moore and Pachón 1985), and NYC was found to have the greatest concentration of Spanish speakers (Mann and De Salvo 1985). The principal Latino neighbor hoods house large numbers of two or more different groups. For example, in Manhattan's East Harlem «Barrio», the majority of the residents are Puerto Rican and Black, but Dominicans have also lived there for more than a decade, and Cubans from the Mariel boatlift joined them in 1980, along with increasing numbers of Mexicans and Central Americans in recent years. Evidence of the fact that inter group communication is frequent and intense is the finding that intermarriage among Hispanics is on the rise among the younger generations (Fitzpatrick and Gurak 1979).

In addition to differences in population, periods of and reasons for migration, residential patterns, and rates of intermarriage, other significant factors that affect intergroup contact and communication are race, income, and education. Puerto Ricans, Dominicans, Cubans, and Colombians differ markedly in terms of racial, economic, and educational background; Dominicans have the highest percent of mulattoes (77%), Colombians and other South and Central Americans have the highest labor force participation rate (73.5), and Cubans have the most education (19.8% with four years of college or more) and the highest median family income ($26,858). Puerto Ricans are the most disadvantaged group of all: they have the highest percent of persons who live below the poverty level (33.7%), the lowest labor force participation rate (54.2%), the least education (46% with less than 4 years of high school) and the lowest median family income ($18,132) [U.S. Department of Commerce 1990]. Coupled with other major demographic factors such as median age (Puerto Ricans and Dominicans under 27, Colombians/Central and South Americans 28, Cubans 41), these characteristics affect each community's posture toward issues as diverse as politics, health, education, reproduction, and institutional change (Andrade 1983).

How does extensive contact among such diverse groups affect each group's active knowledge of lexical variety in surrounding Spanish dialects, and the ability of individual speakers to produce forms that are part of other groups' lexicons? We asked 194 Hispanics (73 Puerto Ricans, 51 Colombians, 50 Dominicans, and 20 Cubans, see Table I in Appendix) to identify 25 objects, presented in drawings, as part of a 1-2 hour sociolinguistic interview adapted from Labov (1984)⁵⁹ . The items included common objects which are lexically different for one or more of the nationalities studied. Interviewees were asked how they referred to each item in everyday conversation, and whether or not they knew of any other group's term for it. Quite aware of the limitations of direct questions which elicit vocabulary items out of context in a formal situation, despite every attempt to overcome

––––––––   1096   ––––––––

the «observer's paradox» (Labov 1984), we take the results as evidence of the speakers' most active vocabulary, and not of their passive or complete knowledge. After submitting all the responses to chi-square analysis by the SPSS-X computer program, we found that differences in age, gender, education, years in the US., and Spanish or English proficiency, largely proved insignificant in determining the subjects' responses. National origin determines their choice of vocabulary, with some revealing additions and exceptions.

How do competing terms accommodate each other, or how and why does one win out? We may assume that a lexical item's chances of survival in New York's melting pot are assured if it is shared by several national groups, but that is not always the case, and it is even more difficult to determine the fate of those words/expressions which are part of only one or two national lexicons. The lexical synonyms offered by our respondents revealed five different scenarios with implications for lexical loss, maintenance, and leveling. The words that appear in each category are presented as follows:

Numbers alone, or before a slash mark (/), represent the percent of the total group which gave this word as its everyday preference, followed by the percent of the occurrences of that word which were produced by that group. Thus, the last line of category I, #3 cartera (below): Col 58% (bolso 34/51, + All-Cu) is to be read as: 58% of all Colombians interviewed said cartera was their common word for pocketbook, but 34% preferred bolso, and 51% of all occurrences of bolso in the data came from Colombians, A (+) indicates that the group(s) that follow(s) produced there mainder of occurrences of this lexical item, and a (-) precedes the names(s) of the group(s) that never used the item. In this example, all of the other groups had some members who also preferred bolso except for the Cubans; none of them produced it.

I. Majority Unity

1. collar «necklace»: Cu 95%, DR 87%, Col 86%, PR 80%

2. cadena «chain necklace»: Cu 95%, PR 91%, DR 83%, Col 75%

3. cartera «pocketbook»: Cu 87%, DR 80%, PR 64%, Col 58% (bolso 34%/51%, All-Cu)

Three words, collar, cadena, cartera, were cited as the common term by the majority of all four groups for necklace, chain, and pocket book. Although members of each group offered synonyms for each of these, e.g., one third of the Colombians preferred bolso for pocketbook instead of camera, the majority of each group did employ the same word. The absence of significant differentiation confirms that these terms are part of the standard Spanish lexicon in New York, the bulk of which is shared by most of the nearly 400 million speakers of Spanish worldwide. We can predict that they are in no immediate danger of replacement, indeed, the Colombian bolso may be hard pressed to keep its place in view of the preference for cartera in every group. The speakers' in-group casual conversations revealed that informal situations are most likely to produce the regional variants, but each group also has access to the more general lexical item for formal and/or out-group settings.

In this category, three of the four national groups gave the same terms more often than any other, but one group preferred another term. The influence of larger dialect region membership is apparent in #s 13-15, e.g., Puerto Ricans, Dominicans, and Cubans walk on an acera «sidewalk», ride the guagua «bus»; and sit on muebles «furniture», demonstrating

––––––––   1097   ––––––––

their Antillean unity in opposition to the Colombians' preference for andén, bus, and sofá⁶⁰. This Antillean agreement is broken in #s 10-12, because Colombians join Puerto Ricans and Dominicans in favoring carro «car», escritorio «desk», and florero «vase», in contrast to the Cubans' preference for máquina, buró, and búcaro. For #s 4-9, Puerto Ricans are alone in their significant preference for china «orange;» matre(s) «mattress», pantallas «earrings;» pimiento «hot pepper», chavos «money», and mahones «bluejeans» Dominicans did not produce any word that was favored by a majority of their group only, but speakers from that country produced the most examples of words which the others did not mention at all. Among these Dominican Spanish items, 100% of which were mentioned by Dominicans only, are tarro «garbage can» (#17); pasillo, contéin, and calzado as three synonyms for sidewalk (#13); cuarto «money» (#8); aritos «earrings» (#6); and fuerteazul «bluejeans» (#9). The majority of Dominicans did offer other more widely known words as their everyday terms for these items. Similarly, although the majority of Colombians said they commonly used cartera and dinero, they produced the majority of bolso (51%) for pocketbook (#3) and 95% of the occurrences of plata in reference to money (#8). Also, Cubans have other ways of saying máquina (#10), but they are the only group for whom this translation of car is customary. Puerto Ricans were the only participants who sometimes borrowed the English words for #12 «vase» (phonologically adapted as [beis] or, rarely, [base]) and for #s 19 and 20 in Category III, [ayglase(s)] «eyeglasses» and [hafslip] «halfslip», but they were the only group that did not prefer [gins], using mahones for bluejeans (#9). Factors that influence the preference for loans, such as years of residence in the US., are discussed under Anglicisms, below.

It is possible that these nationally specific but minority lexical choices may fall by the wayside in favor of the word that is preferred by the majority of their co-nationals, particularly if they are also endorsed by other groups, or they may be kept as markers of in-group, intimate, speech styles. We must caution that the process of incorporation, loss, or maintenance is not predictable by a simple mathematical formula which calculates the number of people who know a term versus those who ignore it, because of a number of social and economic realities that impinge upon communication and linguistic change, leveling, and/or diffusion. We return to these factors below.

III. Two For/Two Against

A. Two groups prefer one term, two others each prefer one or more different terms.

(1) Puerto Ricans and Dominicans prefer one term

16. birthday cake = bizcocho:

(PR 71/54, DR 67/36)

(Cu [kei(k)] 86/56, + All)

(Col pastel 31/60, + All)

(Col ponqué 24/100)

17. garbage can = zafacón:

(PR 60/60, DR 63/36, +All)

(Cu latón 57/71, + All-PR)

(Col caneca 50/100)

(Col tarro 19/100)

18. banana = guineo:

(PR 67/59, DR 86/40, + Col)

(Cu plátano 75/48, + All)

(Col banano 63/80)

(2) Puerto Ricans and Cubans prefer one term

19. eyeglasses = espejuelos:

(PR 50/54, Cu 91/16, + All)

(Col gafas 49/57, + All)

(DR lentes 45/47, + All)

(3) Puerto Ricans and Colombians prefer one term

20. halfslip = enagua(s):

(PR [-s] 69/46, +DR)

(Col [+s] 71/35)

(Cu saya/sayuela 84/80)

(DR mediofondo 47/90)

In #s 16-20 the majority of two groups agree on the same term and each of the other two prefers a distinct term. There is no clear regional pattern to the groupings that are for or against a particular lexical item. The majority of Puerto Ricans and Dominicans use the same words for birthday cake, garbage pail, and banana, but the Colombians and Cubans differ in their responses. For Cubans a birth day cake is the anglicism [kei] or [keik]/[keike]/[keiki] but for Colombians it is a pastel or ponqué. A garbage can (#17) is a latón for Cubans and a caneca for Colombians, while a banana (#18) is a Cuban plátano and a Colombian banano. In numbers 19 and 20, the pair of nationalities that is united is different in each case. For eyeglasses (#19), the majority of Puerto Ricans and Cubans share espejuelos in contrast to the Dominicans' lentes and the Colombians' gafas. For halfslip (#20), Puerto Ricans and Colombians share a version of enagua(s), but the Dominicans prefer mediofondo and the Cubans alternate between saya and sayuela⁶¹.

B. Two preferences, opposite priorities

21. pig = puerco plus cerdo for 60% of all groups

puerco = DR55/39, Col 13/9

Cu 50/12, PR 36/40

cerdo = DR 24/18, Col 40/24

Cu 29/8, PR 37/46

22. grocery store = bodega plus tienda for 75% of all, -DR

bodega = Cu 76/16, Col 2/1

PR 37/45

DR 44/38

tienda = Cu9/2, Col 73/44

PR 42/51

DR 3/3

(DR pulpería = 38/100)

Another variation of Two For/Two Against is found in IIB. Many synonyms co-existed for pig and grocery store, but in each case two account for most of the responses of all four groups: puerco and cerdo for pig, and bodega and tienda for grocery. Two countries ranked the frequency of these terms in reverse order, i.e., Dominicans produced puerco first (55%) and cerdo last (24%), and Colombians produced cerdo first (40%) and puerco last (13%). Similarly, Cubans mentioned bodega more frequently than tienda (76% vs. 9%) but Colombians preferred tienda over bodega (73% vs. 2%). The continued existence of both leading competitors for #s 21 and 22 is favored by the fact that the largest group, Puerto Ricans, used both of them at similar rates: puerco (36%) and cerdo (37%), bodega (37%) and tienda (42%).

Two words, #s 23 clothespin and 24 bobby pin, triggered extensive inter- and intra- group variation. These semantically linked words are both referred to as pinche by Puerto Ricans and pincho by Dominicans, but both occasioned the largest number of synonyms (n = 22) in the study, many of which were shared by all the countries. In addition to the number of competing variants, other factors which may accelerate the loss of these items are analyzed under Frequency and Semantic Weight, below.

This category, which represents the maximum amount of dialectal distinction, is limited to one word in this list, and this phenomenon, i.e., each nationality prefers a different term, does not occur frequently in the full lexicon of the four groups. In this example, the Colombian cometa, the Cuban papalote, the Dominican chichigua, and the Puerto Rican chiringa are each maintained as distinct ways of labeling «kite», but the English word, some times pronounced as [kay], has made incursions in the Puerto Rican and Dominican communities, for reasons which we will analyze in our discussion of Anglicisms.

What are the implications and repercussions of this lexical variety in New York City Spanish? We must proceed with caution before we consider the predictions this data will allow us to make about the future of Spanish in that city. Even with a study limited to 25 lexical items in 4 Spanish dialects, the variation is broad and the groupings are complex; more words from more countries would produce additional configurations, further complicating any prognosis. While we would be wise to avoid any conclusions about the fate of even this limited sample of vocabulary in NYC, it is possible and profitable to analyze the list in the light of the principal factors which play a crucial role in interdialectal contact.

Although we have cautioned against the mechanical application of a purely quantitative approach, Trudgill (1984) and Labov (1966) are correct in maintaining that exposure via personal contact is essential to linguistic accommodations; thus it follows that the total number of people who know and use a term in everyday speech is of critical significance in its extinction or extension. This amount of exposure approach to lexical leveling would

––––––––   1099   ––––––––

lead us to predict that when common lexical items are in conflict, the one employed by the largest Spanish speaking group in the area, e.g., Puerto Ricans in the Northeast, will be adopted. While numbers are important, they are not sufficient to guarantee the accuracy of a prediction about a particular word's chances for incorporation in the larger Spanish speaking community's lexicon because of the need to take into account other variables of a linguistic and social nature that promote or inhibit acceptance. We found three linguistic factors and three social factors which constitute barriers or incentives to co-existence, replacement, and/or specialization of meaning or style.

The extent to which a lexical item is part of the most common vocabulary stock affects its stability, i.e., frequent vocabulary is more stable, infrequent vocabulary is «less stable, more subject to oblivion and replacement» (Weinreich: 57). From our data we can see that the words for clothesline and bobby pin each produced nearly two dozen competing terms. These words are linked to everyday activities, but they are used infrequently today in New York City because the items themselves are going out of style, given the advent of blowdried hairdos and automatic clothes driers. The impact of this increasingly lower semantic weight is also related to a social variable, gender, because males were hard pressed to come up with words for these items, since they are linked to the female do mains of hairstyling and clothes washing. This factor has the effect of halving the pool of frequent users of the words. Moreover, the manufacture of these products in technologically advanced nations means that few indigenous terms existed for them in the home countries, if any. One possible outcome is that the memory of these vanishing cultural artifacts will be captured by the English words that their exporters used for them in the homeland and in NYC, as has occurred with newer and still popular items, e.g., blender «blender» and [bak)i)un 'kliner] «vacuum cleaner.»

Weinreich disallowed the importance of structural resistance to assimilation, including the impact of homonyms, maintaining «that there exists structural resistance other than the recognizability of transferred forms has not been proven» (op. cit.: 62), and that «potential homonymy as a source of resistance to transfer... must be regarded with scepticism.» (op. cit.: 61). Our study provides evidence to the contrary, corroborating Trudgill's position that linguistic structure does exert an influence, particularly in regard to homonyms. The tendency to avoid homonyms is in keeping with the fact that a linguistic system avoids the loss of contrasts, so as to minimize the possibility of miscommunication. When speakers of different dialects come in contact, it is often the homonyms they share that make them most aware of the fact that they do not speak identical codes, and anecdotes about gaffs and misunderstandings, particularly regarding taboo terms, are plentiful in our recordings. That words as innocuo us for most Spanish speakers as bicho «insect», vente «come» (used pseudo reflexively), and papaya «papaya fruit» refer to the male sex organ, orgasm, and the female sex organ in Puerto Rico, the Dominican Republic, and Cuba respectively, as well as in other populations, is always a source of amazement and/or amusement to speakers when they are first encountered. Everyone who is aware of them makes an effort to avoid such homonyms when members of other nationalities are present. The majority of our interviews believed it appropriate to avoid or translate taboo terms in cross dialectal conversations, but they also recognized that there was no way to predict when another term might cause similar out rage. Almost everyone had a favorite story to tell in this regard, and even non-taboo homonyms that caused confusion were recalled as part of their experiences with other Latinos in NYC. To illustrate, the following incident about a misunderstanding with a Puerto Rican co-worker was recounted by a 55 year old Colombian female who has lived in the U.S. for ten years:

The Colombian knew there was a homonym conflict because the meaning of [wawa] (as she pronounced guagua «bus») that she knew did not make any sense in the context of her Puerto Rican co-worker's conversation, but she did not challenge the term or ask for an explanation. Although it refers to a wild animal, possibly a prairie dog, in her Spanish, she interpreted it as another form of transportation, unknown to her, and it was clarified only when she used it in contradiction to bus. The attempt to make sense of an item by taking the context into account is a prevalent discourse strategy in the repair of cross dialectal miscommunication.

Will this speaker and other Colombians replace their bus [bus] with guagua as time goes by? The preference for guagua in the Puerto Rican, Dominican, and Cuban communities suggests that its adoption by Colombians is inevitable, if we consider numbers only. On the other hand, the presence of a Colombian homonym in a different semantic field and the similarity of Colombian [bus] to English [b^s] may prove effective counterforces. The process of lexical diffusion proceeds on an individual, not wholesale basis, because it depends on the interplay of different linguistic and social factors in the case of each word, and each individual. To take a different example, pantalla means a movie or TV screen as well as lampshade in all varieties of Spanish, but for Puerto Ricans it is also the word for earrings. Colombians, Dominicans, Cubans and other Spanish speakers have grown accustomed to distinguishing their homonyms in context, but when they come into contact with Puerto Ricans, they must incorporate a third meaning. Although Puerto Ricans understand and can use the more general word for earrings, aretes, 73% of them responded to the picture with pantallas; that figure undoubtedly is higher in informal conversations. Puerto Ricans who are aware of other groups' conflicts with their term may switch to aretes in their presence, but since pantallas has no taboo connotations, most continue to use it. As a result, it was one of the variants most frequently cited as an example of Puerto Rican vocabulary by speakers of the other dialects. There also is evidence that members of other groups are adopting pantallas as their own: 10% of its occurrences were produced by non Puerto Ricans. In most cases, these were Dominicans who had lived in Puerto Rico or had lived among Puerto Ricans for years, but it was also adopted by some Colombians and Cubans.

On some occasions, an individual's attempt to accommodate another group's term may lead to misassignment of the word to another object or semantic field. For example, anyone who has be en around Puerto Ricans, especially at Christmastime, hears frequent mention of a favorite national dish, pasteles. Since pastel is the word for cake in several regions of the Spanish speaking world, e.g., Colombia and Mexico, it is not uncommon for others to designate it as the Puerto Rican word for cake, thus keeping it in the field of food, but misassigning it to the wrong food.

The varied patterns of alternative regionalisms, taboo terms, multiple homonyms and overlapping semantic fields that result from the lexical variation among New York's Spanish speakers may seem to present formidable barriers to communication, but strategies for resolving conflicts in meaning become part of their verbal repertoire. One alternative for neutralizing lexical conflict that is available to dialect speakers in a bilingual setting is to have recourse to loans from the dominant language. Thus, Latinos in New York turn to English in order to understand each other's Spanish.

Powerful socioeconomic and cultural forces stimulate borrowing whenever two cultures are in contact; the borrowing by the subordinate group's language from that of the dominant group is always significantly greater than the other way around. Studies from all over the world and within many US. language minority groups have documented this uneven process, and the examples in New York Spanish, often labelled «Spanglish», have received wide attention (Acosta-Belén 1975, Milán 1982, Poplack 1979, Zentella 1985, Otheguy 1988). The factors that contribute to word borrowing are varied and overlapping. Most items reflect a cultural reality that is new or different, e.g., [kei/keike/keiki] «cake», [bobipín] «bobby pin», [ŷins/bluŷines] «(blue)jeans.»

––––––––   1101   ––––––––

Others may be facilitated or even triggered by a similarity in the phonological and/or morphological structure of a word in the dominant language that makes it sound like a possible subordinate word, e.g., the form of matre(s) «mattress» is similar to that of madre(s) «mother(s).» Still others may be linked to a reduction in the number of syllables, e.g., [beis] replaces florero for «vase», and, certainly, the role that prestige plays in the adoption of the dominant vocabulary cannot be over looked. The results of this study suggest another reason why, despite widespread condemnation of loans by every group, the practice continues.

Anglicisms can play the role of neutralizer between competing dialectal variants because the prestigious outside language acts as the lingua franca that resolves the conflict without favoring one group at the expense of the other. This is not to suggest that Cubans, Colombians, Dominicans and Puerto Ricans come to adopt anglicisms only after their arrival in NYC; natives in Cuba eat [kei], in Colombia they ride the [bus], and in the Dominican Republic and Colombia they wear [ŷins]. Once in the U.S., their encounters with other products and a different social reality, coupled with the intense pressure and desire to learn English, encourage more borrowing. Our contention that an additional pressure that favors borrowing is the role that loans play as neutralizers of dialectal conflict is supported by the results for the term for kite (#25) in this study. Each of the four nationalities is known to have a different word for kite, and the majority of each group gave its regional variant when asked for their usual way of referring to a kite, but it seems that the longer a group has been in the U.S., the less they use the homeland's term: 60% of the Puerto Ricans mentioned chiringa, 76% of the Dominicans say chichigua, 86% of the Cubans say papalote, and 92% of the Colombians say cometa. In lieu of their native word, 13% of the Puerto Ricans and 3% of the Dominicans claimed to use the English word kite, in contrast to the Cubans and Colombians who preferred other Spanish synonyms only.

Logically, more intense exposure to English over longer periods of time tends to result in the incorporation of more loan words. García, Otheguy and Fernández (1987) found that second generation Cubans incorporated more loans in their speech than members of the first generation. Since Puerto Rican and Dominican communities in New York have the highest proportion of youngsters, we might assume age to be the most significant factor in our Puerto Rican and Dominican subjects' choice of kite, further strengthened by the fact that kites usually are linked to the young. Yet in our sample population, there were less Puerto Ricans (31%) and Dominicans (32%) under the age of 26 than Colombians (39%), and the majority of the Puerto Ricans (59%) and Dominicans (58%) had lived in the U.S. for less than 15 years. Language proficiency appears to be a more relevant factor than age, but in contrasting ways for the groups that were studied. No member of a group whose great majority claimed either «very good» or «excellent» Spanish proficiency, i.e., Colombians and Cubans (84% each), mentioned kite, regardless of individual self-ratings. Puerto Ricans, who claimed the highest English proficiency (20% excellent and 37% very good), produced 87% of the occurrences of kite. These patterns are in keeping with the expectation that greater Spanish proficiency is related to adherence to Spanish synonyms, whereas greater familiarity with English paves the way for word borrowing. In the case of the Dominicans, however, low self-ratings in both English (46% fair, 30% poor) and Spanish (56% fair, 4% poor) point to the role played by linguistic insecurity in the adoption of anglicisms as possible status enhancers, and as a way to avoid criticism of their Spanish.

Elsewhere, I have documented the extent to which Colombians, Cubans, and Puerto Ricans contribute to Dominican linguistic insecurity by their widespread rejection of Dominican Spanish (Zentella 1987). In this study, there is evidence of their rejection of Dominican lexicon, and a greater propensity to assimilate the regional terms of other groups. Some members of every group (except Cubans) have adopted the Colombian word for kite, cometa, some of every group (except Puerto Ricans) have adopted the Cuban papalote, and some from every group (except Colombians) have adopted the Puerto Rican chiringa. Only the Dominicans have adopted from every other group without exception, and theirs is the only word for kite (chichigua) that was not adopted by any members of the other groups. These results alert us to the impact of social variables which may be more powerful in the process of dialect leveling than

––––––––   1102   ––––––––

the structural linguistic ones we have cited.

Class, education, and race shape the attitudes of one group towards another and towards features of their linguistic codes. Negative attitudes inhibit the spread of certain features just as positive attitudes promote their adoption. In this regard, it is relevant to recall Labov's famous analysis of New York City's English, in which he attributed the confinement of linguistic changes in the city to the fact that it was «a great sink of negative prestige» (1966: 499). Among the groups that we studied, and in New York in general, Puerto Ricans and Dominicans are the poorest, the least educated, and the darkest Latinos in the city; they are discriminated against as individuals and as a group, and so is the variety of Spanish that they speak. It is difficult to tease apart the relative weight of each of the social barriers, and their negative impact is further compounded by the fact that Caribbean Spanish is known for consonant deletion or aspiration, particularly of syllable final -s, e.g., lo(s/h) libro(s/h), assimilation, e.g., cerca [sekka], epenthesis, e.g., fuiste>fuistes, and metathesis, e.g., denle>delen. This Spanish, descended from that brought to the Caribbean from southern Spain by Andalusian colonizers and characteristic of the Spanish of the lowlands of Latin America (Canfield 1981), has been labelled «radical» in contrast to the «conservative» dialects that are found in the highlands (Guitart 1982). But as Guitart has noted, even the most conservative dialects display some radical phonology, especially in informal settings, e.g., the -s is lost before alveolar trills, e.g., los rojos>lorrojos, and intervocalic -d- is variably dropped in the past participle, e.g., cansado>cansao. Nor is it accurate to portray radical dialects as categorical deleters of segments or features, e.g., when -s occurs before vowels it is variably retained in Puerto Rican and Cuban Spanish (Terrell 1982). Even when -s deletion is widespread, the meaning of plurality and/or tense is conveyed by other syntactic elements (Poplack 1981). Nevertheless, the preservation of features that recall those of northern Spain's Castilian model in conservative dialects, e.g., Colombian highland Spanish, strongly favors prestige for those varieties of Spanish⁶². The fact that Cuban Spanish does not suffer the same wide spread condemnation that is voiced against Puerto Rican and Dominican Spanish, despite its linguistic similarities, reveals the overriding power that social factors have in the face of linguistic ones. Even when speakers of the higher status groups have radical phonology, they can evaluate their dialect positively and express negative attitudes towards the others. This is obvious in the remarks of a 21 year old Cuban female (15 years in the U.S.) who did not hesitate to admit her prejudice against Dominican Spanish: «Yo siempre pensaba que hablaban feo» «I always used to think they spoke ugly», and against Puerto Rican Spanish: «Yo odio como hablan los puertorriqueños.» «I hate how Puerto Ricans talk.» Although her own Spanish is replete with non standard and/or radical features, e.g., r-l alternation, loss of syllable final -s, and code switches, she is very proud of it, and of Cuban Spanish in general:

This affirmation of dialect pride is in contrast to the negative Dominican evaluations of their dialect, e.g., 80% believe that their Spanish should not be taught in schools, 35% of these because it is «incorrecto» or «malo» (Zentella 1987). Since race, education, and class are inextricably linked in New York's Hispanic communities, e.g., the more middle class Cuban and Colombian migrants also tend to be the better educated and lighter skinned, their varieties of Spanish are not as stigmatized as that of their darker, poorer Caribbean sisters and brothers.

For language teachers the results presented here have a number of obvious pedagogical implications related to the fundamental linguistic notion of dialectal equivalence which can make for exciting and relevant class sessions.

––––––––   1103   ––––––––

Students can collect similar lists from local speakers and/or literature, and compare them to regional varieties of English in order to come to grips with the nature of norms, variation, acceptability, propriety, etc.

But a major concern should be to help students understand the repercussions of dialectal contact for the course of the development of Spanish in the U.S., and the role of linguistic and social factors in the process of dialectal accommodation, change, and diffusion. Will Puerto Rican and Dominican variants be avoided at all costs because of their speakers' lower status? Will Cuban and/or Colombian variants take precedence because of their greater economic power? Do linguistic factors play a determining role in the direction of accommodation? Guitart indicates that in his experience speakers of radical dialects tend to imitate conservative speakers, not viceversa. However, I have found conservative speakers who became radical speakers because of their identification with a group of radical speakers, either because of personal relationships, e.g., some Mexicans who married into Puerto Rican families, or political ideology, e.g., South American community workers in Dominican neighborhoods. Although the role of attitude is crucial, it is necessary to recognize that often covert attitudes belie overt expressions of negativity, as has been documented in analyses of the continued maintenance of non standard dialects despite their prevalent rejection. Members of speech communities prize the non standard forms of their link to important affective variables such as sincerity and trustworthiness, and/or their correlation with coolness or toughness, and incorporate them in their speech for these reasons (Labov 1972). Corroboration of this view in this study came from the comments of several Dominicans who linked the pronunciation of syllable final -s in male speech to effeminacy and/or putting on airs; they refer to it as «hablar fisno.» Thus, although the pronunciation of -s is lauded overtly as correct in the Dominican community, its deletion may be linked covertly to masculinity, and being down to earth.

The degree of speakers' integration into social networks that speak another dialect is related to the extent of linguistic assimilation that is possible as well as to the types of variables that are assimilated. The success with which speakers are able to reproduce accurately the syntactic as well as phonological and lexical variables of another group's dialect is presently being questioned. In Philadelphia, a very black-sounding Italian American teenager had adopted much more of Black English Vernacular (BEV) phonology than its syntax (Labov 1980). In NYC, the greatest amount of linguistic assimilation of Puerto Rican youths to BEV occurred in the speech of those who had more Black friends, yet most retained a distinctively Puerto Rican variety of English (Wolfram 1978). We can expect similar outcomes for Latinos in New York City, i.e., those who are members of dense and multiplex networks (Milroy 1980) which include an other Spanish speaking group sometimes will adopt features of that dialect, particularly the lexicon, using the appropriate forms that allow them to express different «acts of identity» (Le Page and Tabouret Keller 1985) in different settings/situations. There will be others, like the aged whites in Gullah-speaking territory (Rickford 1985), who will never accommodate to the surrounding dialect, despite extensive exposure and even dialect swamping, because they cannot overcome the racial and socioeconomic barriers that divide the groups. On the other hand not all linguistic accommodation should be interpreted as evidence of cultural assimilation, or even acceptance; in some cases it may merely be a practical manipulation of the linguistic resources available in order to meet special needs. Such is the case of conservative dialect speakers who adopt the Puerto Rican dialect in an attempt to avoid persecution by immigration authorities (since all Puerto Ricans are U. S. citizens). The Office of the Commonwealth of Puerto Rico often is called upon to determine if someone is trying to pass as a Puerto Rican; they ask questions about the purported hometown if the speaker cannot be distinguished by dialect alone. Nor is the maintenance of one's own dialect necessarily a repudiation of other groups and their ways of speaking; indeed, pride in one's own dialect may coincide with interest in and respect for others' ways of speaking, in keeping with the vital cultural norm of «respeto» (Lauria 1964). Finally, exact predictions about the course of linguistic leveling that are based on the mechanistic application of demographic or linguistic principles cannot hold up under the scrutiny of sociolinguistic research with an ethnographic base. Given the inherent variability of language

––––––––   1104   ––––––––

and its flexible adaptation to different settings, speakers, styles, and power relationships, only the close observation and analysis of interactions among different groups of Spanish speakers will enable us to illuminate the particular configuration of variables that are relevant in each community and each instance of dialect contact. Future research must shift its focus away from the present preoccupation with the extent to which the influence of English has led to deviations from norms established in/by/for officially Spanish-speaking homelands, in order to capture the process of creation of new norms within U.S. Hispanic speech communities. The contribution made to this process by the new experience of intensive inter-Latino communication is of particular importance. In this paper we have at tempted to identify some of the social barriers which, along with demographic and linguistic factors, must be taken into consideration in an effort to describe and explain the very different realities that are captured by the beguilingly simple phrase, «aquí se habla español.»

Acosta-Belén, E. 1975. «Spanglish: A Case of Language in Contact.» In M. Burt and H. Dulay (eds.), New Directions in Second Language Learning, Teaching, and Bilingual Education. Washington, DC: TESOL. 151-58.

Amastae, J. and L. Elías-Olivares (eds.). 1983. Sociolinguistic Aspects of Spanish in the United States. London: Cambridge University Press.

Andrade, S. J. (ed.). 1983. Latino Families in the United States: A Resource Book for Family Life Education. New York: Planned Parenthood Federation of America.

Ash, S. and J. Myhill. 1986. «Linguistic Correlates of Inter-Ethnic Contact.» In D. Sankoff (ed.), Diversity and Diachrony. Amsterdam: John Benjamins. 33-45.

Bouvier, L. and V. Briggs (eds.). 1988. The Population and Labor Force of New York: 1990-2050. Washington, DC: Population Reference Bureau.

Canfield, D. L. 1981. Spanish Pronunciation in the Americas. Chicago: University of Chicago Press.

Durán, R. (ed.).1981. Latino Language and Communicative Behavior. Norwood, NJ: Ablex Press.

Elías-Olivares, L. (ed.). 1983. Spanish in the U.S. Setting: Beyond the Southwest. Riverside, CA: National Clearinghouse for Bilingual Education.

Elías-Olivares, L., E. A. Leone, R. Cisneros and J. Gutiérrez (eds.). 1983. Spanish Language Use and Public Life in the United States. Berlin: Mouton.

Fishman, J. A. and G. Keller (eds.). 1982. Bilingual Education for Hispanic Students in the U.S. New York: Teachers College, Columbia University.

Fitzpatrick J. and D. Gurak. 1979.Hispanic Intermarriage in New York City: 1975. Bronx, NY: Fordham University, Hispanic Research Center.

García, O., R. Otheguy and M. Fernández. 1987, October. «The Spanish of First and Second Generation Cuban-Americans in West New York.» Paper delivered at El Español en los E.E.U.U. -VIII, University of Iowa.

Guitart, J. 1982. «Conservative vs. Radical Dialects in Spanish: Implications for Language Instruction.» In J. Fishman and G. Keller (eds.) 1982. Bilingual Education for Hispanic Students in the U.S. New York: Teachers College, Columbia University. 167-90.

Hispanic Research Center. 1981, April-July. Hispanic American Diversity [special issue], Research Bulletin 4.2-3. Bronx, NY: Fordham University.

Labov, W. 1966. The Social Stratification of English in New York City. Washington, DC: Center for Applied Linguistics.

——. 1969. The Logic of Non-Standard English. J. Alatis (ed.), Monograph Series on Language and Linguistics 22. Washington, DC: Georgetown University Press.

——. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press.

——. 1980. «Is There a Creole Speech Community?» In A. Valdman and A. Highfield (eds.), Theoretical Orientations in Creole Studies. New York: Academic Press. 369-88.

——. 1984. «Field Methods of the Project on Linguistic Change and Variation.» In J. Baugh and J. Sherzer (eds.). Language in Use: Readings in Sociolinguistics. Englewood Cliffs, NJ: Prentice Hall. 28-53.

Labov, W. and W. Harris. 1986. «De Facto Segregation of Black and White Vernaculars.» In D. Sankoff (ed.), Diversity and Diachrony. Amsterdam: John Benjamins. 1-25.

Lauria, A. 1964. «Respeto, Relajo, and Interpersonal Relations in Puerto Rico.» Anthropological Quarterly 37: 53-67.

Le Page, R. B. and A. Tabouret-Keller. 1985. Acts of

––––––––   1105   ––––––––

Identity. Cambridge: Cambridge University Press.

Mann, E. S. and J. J. Salvo. 1985. «Characteristics of New Hispanic Immigration to New York City: A Comparison of Puerto Rican with Non-Puerto Rican Hispanics.» Research Bulletin: Hispanic Research Center, Fordham University 8.1-2: 1-10.

Milán, W. G. 1982. «Spanish in the Inner City: Puerto Rican Speech in New York.» In J. Fishman and G. Keller (eds.) 1982. Bilingual Education for Hispanic Students in the U.S. New York: Teachers College, Colombia University. 191-206.

Milroy, L. 1980. Language and Social Networks. Baltimore: University Park Press.

Moore, J. and H. Pachón. 1985. Hispanics in the United States. Englewood Cliffs, NJ: Prentice-Hall, Inc.

Poplack, S. 1979. «Sometimes I'll Start a Sentence in Spanish y termino en español: Toward a Typology of Code-Switching.» In J. Amastae and L. Elías-Olivares (eds.), Spanish in the United States: Sociolinguistic Aspects. Cambridge: Cambridge University Press, 1982. 230-63.

——. 1981. «The Notion of the Plural in Puerto Rican Spanish: Competing Constraints on /s/ Deletion.» In W. Labov (ed.), Locating Language in Time and Space. New York: Academic Press. 55-68.

Rickford, R. 1985. «Ethnicity as a Sociolinguistic Boundary.» American Speech 60.3: 90-125.

Terrell, T. D. 1982. «Current Trends in the Investigation of Cuban and Puerto Rican Phonology.» In J. Amastae and L. Elías-Olivares (eds.). 1982. Spanish in the United States: Sociolinguistic Aspects. London: Cambridge University Press. 47-70.

Teschner, R. V, G. D. Bills and J. Craddock (eds.). 1975. Spanish and English of United States Hispanos: A Critical Annotated Linguistic Biliography. Arlington, VA: Center for Applied Linguistics.

Trudgill, P. 1984. On Dialect. New York: New York University Press.

——. 1986. Dialects in Contact. Oxford: Basil Blackwell.

United States Department of Commerce, Bureau of the Census. 1990, May. Hispanic Population in the U.S. (P-20, Nº. 444). Washington, DC: U.S. Government Printing Office.

Weinreich, U. 1979. Languages in Contact. New York: Linguistic Circle of New York (9th printing of 1953 originally published by The Hague: Mouton).

Zentella, Ana Celia. 1982, April-August. «Spanish and English in Contact in the U.S.: The Puerto Rican Experience.» In E. Chang Rodríguez (ed.), The Spanish Language in the Western Hemisphere. Special issue of Word 33.1-2: 41-57.

——. 1987, October. Linguistic Security/Insecurity Among New York's Latinos. Paper presented at Symposium on Spanish and Portuguese Bilingualism, University of Colorado, Boulder.

Hispania [Publicaciones periódicas]. Volume 73, Number 4, December 1990