Selecciona una palabra y presiona la tecla d para obtener su definición.

  —160-161→  

ArribaAbajo- 3 -


ArribaAbajo Linguistics

  —162→  
Contemporary Trends in Hispanic and Luso-Brazilian Linguistics: Part 2 (continued from December 1989)


Prepared by Dr. Jorge Guitart95



ArribaAbajo Current Issues in Studies of Language Contact

Carmen Silva-Corvalán


University of Southern California



Introduction96

In the past thirty years, sociolinguistic research has been concerned, among other questions, with the examination of language change (v. Romaine 1982). Labov (1972, 1981 a & b), in particular, has successfully challenged the traditional structuralist position, which argued that change in progress could not be observed, by developing the necessary methodological techniques to identify and study possible changes in progress in apparent or real time. During this period, one of the most debated questions within the field of historical and sociohistorical linguistics has been the interaction between internal linguistic factors and external social forces in what Weinreich, Labov & Herzog (1968) call the actualization or motivation of linguistic change.

Neogrammarians as well as post-Saussurean structuralists, including generativists among these, view change as motivated and governed by internal factors. Structuralism conceives language as a system whose elements are defined by the place they occupy in opposition to other elements or as a system controlled by parametric rules and universal principles. This concept of language underlies the explanation of change in relation to the existence of structural spaces and of incomplete or unbalanced correlations within the system, on the one hand, and, on the other hand, on the basis of processes of reanalysis essentially motivated by rule opacity. Also internal are considered a number of cognitive factors which constrain possible changes; for instance, it is suggested that changes which may cause comprehension difficulties will tend to be resisted (cf. Martinet 1962). By contrast, sociolinguistics focuses on the social forces which shape language structure and use as well as on internally motivated variation. Sociolinguists have shown that language is inherently and systematically heterogeneous and variable, and that the seeds of change he precisely in the existence of this variation. In regards to change, therefore, one of the general principles postulated by Weinreich, Labov & Herzog (1968: 188) states that «linguistic and social factors are closely interrelated in the development of language change. Explanations which are confined to one or the other aspect, no matter how well constructed, will fail to account for the rich body of regularities that can be observed in empirical studies of language behavior».

Despite changes of perspective and methodological advances, however, Labov (1982) notes that we are still far from reaching a solution to the problem of causality. The question of what forces, be they internal or external, have motivated a specific change at a given time and place continues to haunt sociolinguists as well as structuralist historians. In this respect, Malkiel (1967: 1228) observes   —163→   that the answers proposed depend to a large extent on the linguist's theoretical inclination. Accordingly, the linguist will «favor either external or internal factors, ... [gradual improvement or deterioration of structure or mere regrouping of the structure's constituents] ... either an explanation allowing for the intervention of speakers, at varying levels of consciousness, in the events affecting their speech or the rival explanation operating with unguided clashes of blind forces»:

In the context of this continuous controversy, increasing attention has been focused on pidgin and creole languages, on the acquisition of first and second languages, and on the linguistic phenomena which develop in situations of societal bilingualism and multilingualism. Developing and receding languages as well as maintenance in language contact lend themselves to the examination of hypotheses about linguistic change because they are characterized by constant and rapid changes which may be observed as they arise and spread in the linguistic and social systems. Among the important theoretical issues dealt with by those concerned with these studies have been the universality of the linguistic processes characteristic of these situations, and the role that a model, primary or superordinate language may play in the shaping of the developing, secondary or subordinate language -as against the possibility of autonomous developments constrained by the linguistic system of the languages in question and/or by human cognitive processes.

In regards to the issue of universality, it has been observed that some of the phenomena characteristic of bilingualism and multilingualism, namely simplification, overgeneralization, transfer analysis, and convergence, are indeed attested across different situations of linguistic stress. This observation has motivated the proposal of a number of corresponding theories (i. e., theory of transfer, theory of simplification, etc.) valid for all these language situations (e. g. creolization, language acquisition, language loss). In what follows, I proceed to examine and define these universal phenomena.




Definitions and Theories

Researchers have not yet come to an agreement about the meaning of simplification. Indeed, Ferguson (1982: 58) has pointed out that «the notions "simplicity" and "simplification" are among the most elusive concepts used in the characterization of language». However, he adds that despite this difficulty, there is general consensus regarding what in fact constitutes linguistic simplification. Included as such are, for instance, reduction of the inventory of linguistic forms, semantic range, or language functions, and the elimination of alternative structures at certain levels (59).

I have proposed (Silva-Corvalán, to appear a) a definition of simplification as a complex process which implies also rule generalization in the sense that a given form is being expanded to a larger number of contexts. Simplification, then, involves the higher frequency of use of a form X in context Y (i. e., generalization) to the expense of a form Z, usually in competition with and semantically closely related to X, where both X and Z existed in the language prior to the initiation of simplification. Thus, X is an expanding form, while Z is a shrinking/contracting form. The final outcome of simplifications is loss of forms, i. e., a simplified system with fewer forms, and possibly, though not necessarily, loss of meanings.

This definition of simplification appears to correspond to the notion of overgeneralization (cf. Preston 1982, Silva-Corvalán, to appear b) as the more extensive use of a form than expected in ordinary practice. The only difference is that simplification explicitly refers to contraction, that is, the less frequent use of a competing form. Overgeneralization, on the other hand, may affect contexts where no corresponding competing form exists, i. e., where we may have XY extending to X or viceversa (e. g.,  + Vi -> se + Vi). A related term is regularization, which I use to refer to those cases where the forms extended or overgeneralized correspond to those with a wider structural distribution in the language in question.

Transfer is undoubtedly a controversial notion as well. It may be simply defined as the incorporation of language features from one language into another, with consequent restructuring of the subsystems involved (cf. Weinreich 1974). The problem is, however, that non-lexical transfer is difficult to prove. Still, we consider that transfer may have occurred whenever one or more of the following phenomena is present in the data: a] The replacement of a form in language S with a form   —164→   from language F, or the incorporation from language F into language S of a form (with or without its associated meaning) previously absent in S (this is usually referred to in the literature as «borrowing»); b] The incorporation of the meaning of a form R from language F, which may be part of the meaning of a form P in S, into another form, structurally similar to R, in system S (cf. Weinreich's 1974: 30 «extension or reduction of function»); c] The higher frequency of use of a form in language S, determined on the basis of a comparison with more conservative internal community norms (cf. Klein-Andreu 1986; Mougeon et al. 1985, Silva-Corvalán 1986), in contexts where a partially corresponding form in language F is used either categorically or preferentially; and d) The loss of a form in language S which does not exist in the system of F [cf. Weinreich's «neglect» or «elimination» of obligatory categories]. Transfer leads to, but is not the single cause of, convergence, defined as the achievement of structural similarity in a given aspect of the grammar of two or more languages, assumed to be different at the onset of contact [cf. Gumperz & Wilson 1977]. Indeed, convergence may result as well from internally motivated changes in one of the languages, most likely accelerated by contact, rather than as a consequence of direct interlingual influence (Silva-Corvalán 1986).

A change in a language L is considered to have accelerated in the speech of a group X when both the number of context types (as opposed to tokens of the same context), and the frequency of use of the innovation in a token count in the various contexts are higher as compared to an older-age group Y in the same speech community as X, and this increase is in turn higher than any possible increase identified in the speech of a group P, as compared to an older-age group Q in the same speech community as P. This definition follows from the assumption that changerelated variation observed in apparent time (Labov 1972), i. e., across generations, and crucially, in the case of language contact, across the bilingual continuum, may reflect stages of diffusion in real time.

Analysis is the process which underlies either the preferential use or the creation of analytical or periphrastic constructions as opposed to synthetic ones. This accounts, for instance, for the much more frequent use of the periphrastic future and conditional forms (ir a + Inf.) rather than the corresponding synthetic ones (-, -ría) in all varieties of Spanish.

At least three of these processes, simplification, overgeneralization, and analysis, which characterize change in full-fledged languages as well, may be accounted for on the basis of cognitive and intralinguistic factors (v. Silva-Corvalán, to appear b), thus avoiding the question of language permeability to foreign influence. There is general consensus, however, that intensive language contact is a powerful external promoter of language change. Nevertheless, perhaps due to the fact that the acknowledgment of contact-induced changes poses a threat to some of the methods of historical linguistics, it is only rather recently that linguists have challenged the view that a grammatical system is impermeable to direct transfer of elements which do not correspond to its internal structure or tendencies of development (cf. Jakobson 1938). In this regard, Thomason & Kaufman (1988) argue that there is evidence from a large number of contact situations in support of their hypothesis that «it is the sociolinguistic history of the speakers, and not the structure of their language, that is the primary determinant of the linguistic outcome of language contact» (35).

While basically in agreement with this hypothesis, my research indicates (v. Silva-Corvalán 1987, to appear a and b) that even under conditions of intense contact and strong cultural pressure speakers of the receding language simplify or overgeneralize grammatical rules but do not introduce elements which cause radical changes in the structure of the language. Rather, these changes occur step by step across the bilingual continuum and in real time. Since ultimately they may lead to the development of a language which is fundamentally different from its non-contact ancestor, I have argued in favor of a slightly different hypothesis: the sociolinguistic history of the speakers is the primary determinant of the ultimate linguistic outcome of language contact; the structure of the languages involved, to a large extent constrained by cognitive and interactional processes, governs the diffusion of the innovative elements in the linguistic systems. This hypothesis accounts for the changes attested in numerous situations of language maintenance and/or shift with normal language transmission across generations.



  —165→  
The Case of Spanish-English Contact in the U. S.

U. S. Spanish illustrates both maintenance (at the societal level) and shift to English (at the individual or family level). This, plus the fact that the sociohistorical information regarding this contact is recoverable, makes it a singular situation. The study of Spanish-English contact has made and will most certainly continue to make valuable contributions for the advancement of historical linguistics and sociolinguistics. Indeed, cross-generational examinations of the linguistic behavior of bilinguals (in communities such as that of Los Angeles, for instance) have thrown light upon the nature of the processes underlying both possible stable changes in language maintenance as well as changes characteristic of language shift or loss97.

It has been shown (v. Dorian 1981, Elías-Olivares 1979, Silva-Corvalán, to appear a) that in these situations of societal bilingualism an oral proficiency continuum may develop between two languages in contact. This continuum resembles in some respects a creole continuum inasmuch as one can identify a series of lects which range from standard or full-fledged Spanish to an emblematic use of Spanish and, vice versa, from full-fledged to emblematic English. Speakers can be located at various points along this continuum depending on their level of dominance in one or other of the languages. Accordingly, studies which claim to be describing a given aspect of bilinguals' competence will be invalid if they do not take into account the place that these bilinguals occupy in the continuum. Indeed, the absence of this information constitutes a serious methodological problem which weakens the validity of numerous published studies of Spanish-English bilingualism.

The case of language attrition has been referred to as «creolization in reverse» by Trudgill (1984) in his study of the Arvanitika speech community of Greece, a term which captures the observation that while certain pidgins and creoles move toward a higher number of grammaticalized distinctions, in language loss the reverse is true, e. g., speakers at the lower stages rely on contextual and/or lexical strategies to communicate certain grammatical meanings (e. g., temporal or aspectual distinctions).

The type and degree of loss found in pidgins, however, is different from that found in language attrition in situations of language shift. As observed by Dorian (1978, 1981) in East Sutherland Gaelic, in these situations a language dies «with its morphological boots on», i. e., with a certain degree of complexity which pidgins do not seem to develop. I have observed the same phenomenon in the Los Angeles community. Indeed, speakers with the lowest levels of Spanish proficiency are most fluent in English and do not need to use Spanish for any practical purposes. Therefore, the little Spanish they use does retain some verbal inflections, gender, number and case markers, and prepositions. The extralinguistic factors which seem to account for the amount of attrition attested at these lowest levels include steadily decreasing use of Spanish, or its restriction to the domain of the family and close friends; subjective attitudes towards Spanish and Hispanics in general; and instrumental or affective motivation for maintaining Spanish (e. g., as a symbol of ethnic identity).

It is evident that the long and sustained contact between English and Spanish in the U. S. has given rise to numerous changes in the system of Spanish (and no doubt also in the English spoken in Spanish communities), some of which have been investigated and reported in the literature. At first, research focused mainly on phonological and lexical aspects. By contrast, since the mid-seventies the field has moved forward and has explored morphological and syntactic aspects as well as the very interesting phenomenon of code-switching (v. various articles in Amastae and Elías-Olivares 1982; Bergen 1989; Bills 1974; Bowen and Ornstein 1976; Elías-Olivares 1983; Fishman and Keller 1982; Hernández-Chávez, et al. 1975; Peñalosa 1980; Sánchez 1983; Wherritt and García 1989, among others). Furthermore, it has been concerned with the processes which underlie the phenomena examined at different stages of bilingualism, and with the theoretical implications of the results of the studies conducted. To illustrate these concerns, I will discuss some of the work that I have been conducting in the Spanish-English bilingual community of eastern Los Angeles. Constrained by length limitations, I will refer only briefly to the work of other scholars in this context.




The Data

The studies to be presented here are based   —166→   on samples of data obtained through recordings of conversations between the author and 50 Mexican-American speakers from three groups, all residing in the eastern section of Los Angeles. Group 1 includes speakers born in Mexico, who immigrated to the U. S. after the age of eleven. Group 2 encompasses speakers either born in the U. S. or who have immigrated from Mexico before the age of six. Group 3 also comprises speakers born in the U. S.; in addition, at least one parent must respond to the definition of those in group 2.

Only those speakers in group 1 have received more than three years of formal instruction in Spanish, in Mexico, but no one in this group has completed secondary education. The level of education in English varies in every group: two speakers in group 1 have received no formal education in this language, one speaker in group 3 has achieved a doctoral degree, and the remaining eleven have completed or will soon complete secondary education.

These speakers were contacted beforehand and were aware of my interest in bilinguals and «the life of Spanish in the community». I established a friendly relationship with most of the speakers (for instance by maintaining at least some telephone contact between the recording sessions), especially so with a number of families, which included two and sometimes three different generations of immigrants. The speakers were recorded twice, with an interval of about six months between the two recordings, for periods ranging from 75-100 minutes each time.

The goals of my studies required the use of careful sociolinguistic fieldwork methodology in order to succeed in obtaining comparable data across speakers, while at the same time maintaining an atmosphere of «social conversation» during the recordings. Crucial to a study of the impact of language contact on both form and function of such aspects as verb morphology, for instance, is spoken data ranging along a wide variety of topics and discourse genres which ensure the creation of contexts for the use of every form of the verb system hypothesized to be available in the language of first generation immigrants. Despite a few drawbacks, the data obtained from every speaker represent the variety of topics/discourse genres expected. These include, but are not limited to: exposition of past events, narratives of personal experience, route directions, descriptions of people and places, argumentative discourse, and discourse about hypothetical situations.

In the following section, I examine the verb system of Los Angeles bilinguals.




Tense in Spoken Los Angeles Spanish: Simplification and Loss

Following Comrie (1985), I consider tense to be «the grammaticalization of location in time» (see, in particular, Ch. 1) and include as such those forms discussed under the labels of preterite, pluperfect, future, non-finite forms, etc., regardless of the fact that these forms are usually used to convey mood and aspect distinctions rather than strictly tense. I also follow Comrie in the classification of tense forms into absolute, absolute-relative, and relative98. In Spanish, absolute tenses include present, preterite, imperfect, future, and present perfect indicative (semantically «past with present relevance», Comrie, 77-82). Absolute-relative tenses encompass the conditional (as future in the past), the present and imperfect subjunctive, and indicative and subjunctive compound forms. Relative tenses comprise the non-finite forms.

Hypothesizing the possibility of transfer from English implies making predictions about the permeability of the Spanish verb system to change and to influence from another language. At the same time, however, there exist other factors which must be taken into account and which may or may not predict similar results. Indeed, although these distinct factors are related to phenomena of a different nature: on the one hand, cognitive (e. g., higher cognitive complexity of certain types of discourse, (v. Levelt 1979, Silva-Corvalán 1987) and social (e. g., absence of normative pressures, specific communicative needs, the nature of verbal interaction), and on the other, intra- and interlinguistic, it is usually the case that more than one, or even all, may be motivating and constraining a specific process of change.

Thus, consideration of socially motivated circumstances makes it reasonable to expect that simplification and loss would affect first those forms used in contexts of higher hypotheticality or weaker assertiveness, i. e., conditional and subjunctive forms. However, hypothetical texts are cognitively more complex to produce99, a fact which makes a cognitive explanation for the occurrence of the phenomenon   —167→   quite plausible. Even further, at least simplification of conditional morphology could also have intralinguistic motivation since it is attested both in first generation immigrants' speech and in non-immigrant varieties in Mexico. Finally, given that English lacks subjunctive tenses and the conditional is not marked by bound morphology, it seems justifiable to invoke the impact of interlinguistic factors as well.

Conversely, cognitive and interlinguistic factors appear to predict contradictory results in the case of the preterite-imperfect opposition. Indeed, the cognitively motivated principle of semantic transparency, defined by Slobin (1977: 186) as a tendency «to maintain a one-to-one mapping between underlying semantic structures and surface forms, with the goal of making messages easily retrievable for listeners», held to be responsible for a number of linguistic phenomena characteristic of language acquisition, attrition, loss, development, change, predicts the retention of the preterite-imperfect opposition. By contrast, direct transfer from English may lead us to expect the neutralization of an opposition not marked by bound morphology in this language.

To identify patterns of simplification and loss in the verb system, illustrated in exx. 1-2, I have compared the linguistic behavior of second and third generation speakers, groups 2 and 3, with that of those in group 1, i. e., with the speech of members of the same bilingual community.

(1) A: ¿Cómo ves tu vida en México si en vez de vivir aquí te hubiera tocado vivir allá?

B: Bueno, pues, iba a tener que acostumbrarme a las costumbres de allá.


(Alb, m60, II, ELA73A20)100                


A: «How do you see your life in Mexico if instead of living here you had to have lived there?»

B: «Well, I would've had to have gotten used (Lit.: was going to have to get used) to the way things are there».


(2) Se comunicó con el «police department» a ver si teman uno que estaba interesado en ser «teacher», so me llamaron a mí.


(Rro, m46, III, ELA 36A465)                


«He called the police department to see if they had anyone who might be (Lit.: was) interested in being a teacher, so they called me».




The methodology followed in the analysis of the data allowed us to confirm that the bilingual continuum referred to above is clearly represented in even a small sample of speakers such as the one included here. With only minor exceptions, the progression of simplification (S) and loss (L) of verb morphology falls into nine stages, implicationally ordered as shown in Table 1, such that if an individual uses the forms in (I), it may be assumed that the forms listed in (II)-(IX) will form part of his verb system and will not be affected by processes of simplification. Conversely, loss of present subjunctive and imperative (stage IX) implies as well the absence of the forms listed in (I) through (VIII), and the simplification of the preterite-imperfect opposition, as indicated in (III) and (V).

    Table 1. Stages of simplification and loss

  1. L: future perfect
    morphological conditional (tense function)
    S: conditional perfect
  2. L: conditional perfect
    S: morphological future
  3. L: morphological future
    S: pluperfect indicative
    imperfect subjunctive
    pluperfect subjunctive
    preterite (with closed list of stative verbs)
  4. L: pluperfect indicative
  5. L: pluperfect subjunctive
    present perfect subjunctive
    perfect infinitive
    morphol, conditional (hypothetical function)
  6. S: imperfect indicative
    present and
    imperative
  7. L: imperfect subjunctive
  8. S: present perfect indicative
  9. L: present perfect indicative
  10. L: present subjunctive
    imperative
  • L = loss
  • S = simplification


I have used a sort of «apparent time technique» in establishing these nine stages, assuming that the individual system most divergent from the norms of first generation immigrants reflected the most advanced stage of simplification and loss, while intermediate degrees of divergence would reflect various intermediate stages of simplification and loss. As expected, given the existence of a bilingual continuum, speakers in groups 1 to 3 do not fall neatly into each stage. For instance, speakers in group 1 may be at stage (I) or (II), and speakers in group 3 may be at any stage below (V).

It is arresting to note that language attrition in societal bilingualism is in fact to a large extent the mirror image of development in creolization, and in first and second language acquisition. That is to say, in acquiring the   —168→   verb system of Spanish, and indeed of various other languages (v. Brown 1974; Klein 1985; Slobin 1986), learners go through stages of development which are the reverse of loss: the earlier tense forms to be acquired are present and preterite (cf. Jacobsen 1986), while future, conditional and compound tenses are acquired much in the same order in which they are lost across the bilingual continuum. This correspondence may in fact reflect the freezing, at different levels of development of grammatical proficiency, of the bilingual's secondary language. However, the possibility of loss at the individual level, though not documented in my studies, could not be dismissed a priori.

Not only is there a large degree of correspondence regarding emergence and disappearance of tense-mood-aspect markers overall, but also with respect to the development and loss of verbal inflections with different types of verbs. Indeed, a comparison of processes which affect some preterite and imperfect forms in the Los Angeles data with those observed by Andersen (n. d.) in the development of these forms in the Spanish as a second language of two English speaking children (at ages 8-10, and 12-14) shows that in both situations simplification of preterite and imperfect morphology affects stative and non-stative verbs differently (v. Silva-Corvalán, to appear b), such that prototypical stative verbs (e. g., ser, estar, tener), are marked with imperfect morphology in the past, regardless of the perfectivity of the context.

The changes affecting the preterite-imperfect opposition starting in group 2 are in fact of enormous interest because they appear to indicate that speakers view this opposition as more basically one of stative versus non-stative aspect rather than imperfective versus perfective aspect. Thus, imperfect morphology is attached to stative verbs, while preterite morphology is attached to activity/non-stative verbs regardless of the perfectivity of the context. This is exemplified in 3 and 4.

(3) a. Después mi hermano era (imp.) el que iba (imp.) a misa

b. y entregó (pret.) el sobre.


(Son, f19, III, ELA66B540)                


a. «Afterwards my brother was the one who would go to church

b. and handed in the envelope».


(4) a. Porque este mejicano no sabía (imp.) el inglés,

b. no más habló (pret.) español,

c. pero era (imp.) muy bravo y muy macho.


(Dan, m45, III, ELA43B455)                


a. «Because this Mexican didn't know (imp.) English,

b. he only spoke (pret.) Spanish,

c. but he was (imp.) very tough and very macho».




The very interesting parallels across different situations of language contact and acquisition give evidence of the generality of the processes and phenomena which underlie changes and restructuring of verbal systems. In regards to parallels with other varieties of Spanish in the U. S., I am aware of only one study, Torres 1989, which has employed a methodology similar to mine so as to allow adequate comparison. Indeed, Floyd (1978) notes the methodological differences in her exhaustive review of the literature relevant to verb usage in Southwest Spanish and states that «the studies surveyed differ greatly in purpose, methodology, analysis, and presentation of the data; such differences tend to inhibit any precision in comparisons across the investigations» (87). It seems necessary, therefore, to develop uniform methods of research across Hispanic communities so that we may be able to identify what is common to all varieties of contact Spanish in this country.

In regards to the impact that English may have had on the verb system of Spanish, examination of the progression of simplification and loss, summarized in Table 1, leads us to conclude that the impact from English is only indirect. This is, evidence does not seem to be sufficient to conclude that contact with a typologically different language would have led to different results. Note that direct influence from English does not justify (a) the early simplification and loss of pluperfect indicative101; 5 (b) the loss of perfect infinitive [at stage V]; (c) the loss of present perfect [at stage VIII] as compared to the later loss of present subjunctive [stage IX]; nor, finally, (d) the retention of imperfect indicative down to very low points in the continuum. Thus, as previous studies have shown (Dorian 1978, 1981; Silva-Corvalán 1986, among others), some changes occur rather as a result of reduction of both exposure to and use of a full-fledged variety of a subordinate language in contact with a superordinate one.

Torres's (1989) examination of the use of subjunctive forms by Puerto Rican speakers in El Barrio (East Harlem) in New York leads her to conclude as well that impact from English appears to be only indirect. By contrast,   —169→   Torres argues that the differences she finds between first and second generation speakers are not significant, despite her own observation that there is «a difference in the distribution of subjunctives between first and second generation Puerto Ricans» (70). Quite interestingly, she interprets the non-occurrence of imperfect subjunctive in the protasis of conditional clauses as due to «pragmatic changes» rather than to loss of form. Note, however, that although the use of the imperfect indicative in a context where first generation Puerto Ricans would use imperfect subjunctive indicates that the imperfect indicative is undergoing some sort of semantic-pragmatic extension or change, this pragmatic change is not necessarily the cause of the non-occurrence of subjunctive; the new meaning may simply be an a posteriori interpretation of the meaning of this form, once the form has taken over the specific context of conditional protases. It is certainly the case that the imperfect subjunctive form has not disappeared from the system, but it is also the case that its use or distribution is being reduced, such that the imperfect indicative is now taking over contexts previously reserved for the imperfect subjunctive. This is a classical example of what we consider form simplification, which in many cases goes hand in hand with the expansion of a competing form.

The indicative-subjunctive opposition has been studied in the speech of 9 speakers from Los Angeles, 3 in each generational group, by Ocampo (to appear). His methodology differentiates categorical contexts, i. e., those which do not allow the indicative, e. g., purpose clauses (El dinero es para que compres [subj.] sellos «The money is for you to buy stamps») from contexts which allow both moods, obviously with corresponding semantic-pragmatic meaning differences (Si tengo/tuviera tiempo te ayudo/ayudaría «If I have/had time I help/would help you»). His results show the following progression of reduction of use of the subjunctive:

Group 1 Group 2Group 3
Categorical contexts100% 91% 62%
Variable contexts79%60%22%



These results are interesting because they offer support for the hypothesis that simplification affects first non-obligatory contexts, i. e., those where the choice between two or more forms is possible in order to convey subtle meaning differences associated with these forms. Neutralization of such oppositions, therefore, would involve some degree of semantic loss in a secondary language which may not be called upon to express these meaning differences as frequently as if it were the primary means of communication in a given group.

Turning to the question of convergence with English, it seems clear that despite the rather drastic changes that Spanish has gone through, there is only one which approximates it to English, the loss of the subjunctive mood. However, this happens at the very low stages of the continuum. Since the variety of Spanish spoken at these stages is very unlikely to be transmitted to later generations, it will most likely disappear with its speakers. One may expect, therefore, that at the societal level the grammar of U. S. Spanish will remain quite close to that of its ancestor varieties.




Verbal Periphrases: Analysis

In regards to analysis, a process which supports Slobin's principle of semantic transparency, I would like to discuss the development of periphrastic and auxiliary constructions affecting ser «be», estar «be», and ir «go».

Most bilinguals in groups 2 and 3 in Los Angeles use the preterite fue «went/was», a homophonous form, exclusively with the active meaning of went, (as in ex. 5). In addition, they use the imperfects of ser/estar, i. e., era/ estaba «was», in both perfective and imperfective contexts (as in exx. 6-7), while the imperfect form of ir, iba «would go/was going/went», starts to lose its lexical meaning and is used almost exclusively as an auxiliary in the ir a + Infinitive «be going to + Inf.» construction (as in ex. 1 above). In turn, the lexical meaning of iba is expressed by means of the imperfect past progressive (estaba + -ando «was + -ing», as shown by ex. 8.

(5) Y estábamos esperando a mi 'amá -porque ella fue a llevar mi hermano a la dentista.


(Vir, f18, II, ELA17A1709)                


And we were waiting for my mom -because she had gone to take (Lit.: went to take) my brother to the dentist.


(6) [R: ¿Y la otra vez, qué pasó?]

H: Era (imp.) en el seventh grade, con un muchacho del sixth grade.


(H, m21, II, ELA11A240)                


«[R: And the other time, what happened?] It was in seventh grade, with a guy from sixth grade».

(7) [Produced at the end of a narrative about the speaker's prom.]

  —170→  

Y estaba (imp.) muy bonito el prom.


(Rra, m20, II, ELA50B395)                


And the prom was real nice.

(8) Pasó un día cuando estaba caminando pa' la casa. (Grupo 2)

«It happened one day when I was walking home».




Table 2 summarizes these changes.

Iba a + Inf., then, replaces the bound conditional morpheme -ría and achieves paradigmatic regularity: va a +Inf. «is going to + Inf». = future, iba a + Inf. «was going to + Inf». = past of future.

It is not possible for me to ascertain on the basis of the data studied whether all these changes occur independently of one another or whether one change has triggered some kind of chain reaction in the system. Whichever the case may be, the trend is, as in full-fledged language changes, toward a one-to-one relationship between certain forms and their semantic content. The semantic content which remains is the one which is most frequently associated with the form in communication. Further, when full lexical items (e. g., iba) reduce semantically into grammatical forms, the lexical and paradigmatic gap left is filled in by a periphrastic construction.

It is clear that speakers treat stative and non-stative verbs differently. What is not so clear is whether this is due to the innateness of this distinction (as proposed by Bickerton 1982, Brown 1974, etc.), or to the fact that statives occur much more frequently in imperfective form (cf. Andersen n. d.) in everyday communication. Our study cannot throw any light upon the question of innateness, but it does suggest that the inherent imperfective meaning of certain verbs and their high frequency of occurrence in imperfective contexts favor the loss of the corresponding preterites, and thus a consequent change in the meaning of the «new» imperfect forms. At this stage, the loss of preterite morphology is lexicalized to occur with «prototypical» statives, but it is in theory possible for this loss to continue spreading through the verb lexicon in a manner which would reverse stages of development of preterite and imperfect morphology.




Placement of Verbal Clitics: Transfer

It has been proposed (Klein-Andreu 1986; Mougeon et al. 1985) that one of the manifestations of transfer is the higher frequency of use of a form in language S, determined on the basis of a comparison with more conservative internal community norms, in contexts where a partially corresponding form in language F is used either categorically or preferentially. In this respect, Klein-Andreu has observed that in contexts in which both the simple present and the present progressive are used in Spanish while only the progressive is allowed in English, Puerto Rican bilinguals in New York employ the present progressive more frequently than the Spanish speaking monolinguals.

I have examined this type of transfer through a study of the position of verbal Clitics (Cls) in constructions with verbal periphrases. Even though it is not entirely clear to me how comparable Spanish Clitics may be to oblique pronouns in English, I assume cross-linguistic equivalence when in a given English sentence, the oblique pronoun translates the Spanish verbal clitic102. Accordingly, lo is considered to be equivalent to him, unstressed, in exx. 9-10:

(9) Lo vi ayer.

him saw-1s yesterday

«I saw him yesterday».

(10) Quería verlo.

wanted-1sg see him

«I wanted to see him».



Simplification, and transfer allow us to advance certain hypotheses with respect to the behavior of Cls along the continuum. Simplification justifies expecting a trend towards the categorical occurrence of one of the two possible alternative positions for Cls in constructions with verbal periphrases, either preverbal (ex. 11a) or postverbal (ex. 11b).

(11) a. Lo puedo hacer mañana.

it can-1st do tomorrow

b. Puedo hacer lo mañana.

can-1sg do it tomorrow

«I can do it tomorrow».



Transfer, on the other hand, should result in the preference for postverbal placement of the Cl (ex. 11b). This type of transfer, which at first evidences itself not in ungrammaticality but in an increased frequency of use of parallel structures, is proposed by Klein-Andreu   —171→   (1986: 7) to be the most likely to occur, as well as the most likely to become part of the community language norms. Furthermore, though previous research has shown that, as compared to free morphemes, bound morphology is more resistant to change (cf. Poplack, 1978; Weinreich, 1974)103, the presence or absence of inflectional markings in one language is also cited as one of the possible features to be affected by the presence or absence of corresponding inflections in the contact language (cf. Meisel, 1983; Weinreich, 1974). Thus, as inflections, Cls may be a plausible site for transfer from English to affect at least both the position and the actual occurrence of Cls.




Clitic Position

When Cls refer to an argument of an infinitive or a present participle in a verbal periphrasis with a finite «semi-auxiliary» verb, they may variably occur before the finite verb. This is the only variable context for clitic placement in Spanish (as exx. 11 a & b illustrate). Otherwise, with imperative, infinitive, and present participle forms the Cl is postverbal, while it must be preverbal with finite forms.

It is interesting to note that none of twenty speakers studied violate the categorical constraints on pre- or postverbal placement of Cls. In regards to those utterances which allow one of two positions (ex. 11), on the other hand, our study indicates that, contrary to what a naive view of transfer might predict, postverbal placement is less frequent in the speech of bilinguals. Furthermore, the variables which simply favor preverbal placement in Spanish dominant bilinguals (Gr. 1) and in monolinguals appear as almost categorical contexts for this order in Group 2, i. e., the Spanish of bilinguals moves in the direction of strengthening Spanish internal trends rather than English patterns (cf. With the extension of estar, discussed below).

Previous studies (v. Landa 1989, Myhill 1988) have shown that the pre- or postverbal position of the clitic is to a large extent constrained by the semantics of the matrix or «semi-auxiliary» verb. Thus, preverbal placement is strongly favored when this verb retains little or none of its basic meaning, as illustrated in exx. 12 and 13.

(12) Matrix verb: venir (basic meaning)

Pepe viene a entrevistarme hoy.

«Pete's coming to interview me today»

(13) Matrix verb: venir (grammaticalized meaning)

Pepe me viene molestando por años ya.

Pete's been/kept bothering me for years



In exx. 12 and 13 the Cl could have been placed in pre- or postverbal position. It is interesting to note, however, that the preverbal position is favored when the matrix verb conveys epistemic meanings, progressive aspect, and future tense, three meanings which Bybee (1985) has shown to be expressed most frequently by means of bound morphology in natural languages. The quantification of this preference in data from four Spanish speaking countries is summarized in Table 3104.

The fact that clitics are a Romance creation, and that their position with respect to the verb did not become fixed in Spanish before the 15th century justifies considering their variable position in verbal periphrases a syntactic-semantic change in progress, which involves the gradual grammaticalization of a number of verbs105. The trend to place the clitic in front of the matrix verb would have, therefore, an internal motivation and would be controlled by semantic factors. Despite the existence of a parallel construction with a postverbal clitic pronoun in English, the internally motivated opposite trend in the direction of preverbal placement is accelerated in Los Angeles, as Table 4 indicates.

As Table 4 shows, group 1 speakers favor preverbal position as compared with speakers   —172→   in the four countries examined with almost all verbs. This preference for preverbal position moves further among second generation bilinguals. Speakers in group 3, on the other hand, though still behaving according to this general tendency, have slightly lower percentages of preverbal Cls than those in group 2. This general result needs to be examined in more detail. However, and based on a comparison of other features of group 3 Spanish with group 2, namely verb morphology and the extension of estar «to be», we note that, as a group, Mexican-Americans in group 3 do not seem to continue along the lines of a «natural» historical development of their ancestors' language. There is a break at this point in the continuum, such that the language of speakers below this point moves much further away from the norms of group 1 speakers, and shows signs of stronger convergence with the dominant contact language. This is especially evident in the case of speakers who have «revived» their Spanish through two or three semesters of Spanish language instruction in high school.

Clitic placement, then, permits us to offer evidence against transfer, in the case of naturally developed bilingualism, when the parallel structure in the dominant language is not characterized by analyticity. In regards to Klein-Andreu's hypothesis, our study suggests that, if it is to be saved, it should be modified to incorporate a condition on the type of parallel structure likely to be transferred/ preferred, as follows: in situations of intensive bilingualism, the higher frequency of use in a subordinate language S of morphological and/or syntactic parallel structures (PS) in a superordinate language F will occur only if the PS in F corresponds to a preferred structure in S prior to the initiation of contact with F.

It should be clear that this condition on PS transfer corresponds to what has been discussed both in terms of acceleration of change and overgeneralization. Therefore, I propose that a theory of simplification, cognitively and intralinguistically motivated (cf. Dorian 1980; Ferguson 1982: 59; Silva-Corvalán, to appear a and b), which predicts the loss of morphosyntactic and lexical variables in subordinate contact languages, accounts more appropriately than transfer for the tightly constrained postverbal placement of Cls, the extension of estar (discussed below), and the preference for progressive over simple forms observed by Klein-Andreu.




Ser-Estar along the Continuum: Acceleration of Changes

In addition to the principles of semantic transparency and generalization, the theory of simplification is based on the principle of distance. According to this principle, if a language system has several forms in the same syntactic-semantic sphere, the form which is farthest away from the speaker, in the sense that it refers to objects or events which are the farthest from him in his objective (e. g., actual distance) or subjective (e. g., possibility of actualization) world, will tend to be lost first and acquired later. Interactional, and perhaps also memory factors underlie this principle: since speakers tend to speak about themselves and their immediate objective world rather than about distant and hypothetical situations, the infrequently used forms, if any, will disappear first, while the more frequently used ones will be acquired earlier or retained longer.

Both this principle of distance and the hypothesis which proposes that language contact accelerates processes of change already present in the language of first generation immigrants are illustrated in what I have called the extension of estar (v. Silva-Corvalán 1986).

The particular phenomenon which I refer to is the extension of estar to contexts previously limited to ser, both copulative verbs translated as «to be». This trend, which affects the use of copulas in predicate adjective constructions, has characterized the development of Spanish since at least the 12th century, and it appears to have accelerated in Los Angeles. Exx. 14-17, produced during conversational descriptions of people and objects, are illustrative.

(14) (Mi abuelita) es blanca. Ni es gorda ni es delgada. Esta bien.


(M, m17, I)                


«(My grandmother) is fair-skinned. She's neither fat nor slim. She's okay».


(15) Una de esas recá-, recámaras es el master bedroom, el más grande. Y el otro está pequeñito.


(H, m21, II)                


«One of those bed-, bedrooms is the master bedroom, the larger one. And the other one is (estar) small».


(16) Teníamos otro cabanete (sic) allá arriba -pero estaba muy largo, y no cabía la hielera, y no estaba, no estaba ancho.


(M, f33, III)                


«We had another cupboard up there -but it was (estar) too long, and the refrigerator wouldn't fit, and it wasn't (estar), it wasn't (estar) wide».


(17) Mi papa era un hombre -muy alto. «Todos los Campas son altos, -como me dijo mi tío- menos   —173→   usted, Daniel». /I: ¿Te dijo?/ ¡El cabrón! Y yo le dije p'atrás: «Pero yo estoy inteligente y muy guapo y no te puedo tener todo».


(D, m45, III)                


My father was a very tall man. «All the Campas are (ser) tall,-as my uncle once told me- except you, Daniel». /I: He really told you so?/ The son of a bitch! And I told him: «But I'm (estar) intelligent and very handsome and I can't have everything»




Observe that in ex. 14, a group 1 speaker links the subject referent to its attributes using ser. In equivalent discourse contexts, in exx. 15-17 speakers in groups 2 and 3 use estar to introduce the attributes. In ex. 17 a speaker in group 3 uses estar with the adjective inteligente, which only in sarcastic exclamations would allow estar in the speech of those in group 1.

Examples 15-17 illustrate the diffusion of estar to contexts in which other varieties of modern Spanish allow only ser. This extension to new contexts represents a more advanced stage in a continuous process of syntactic-semantic extension of the copula estar throughout the history of Spanish. The change in general Spanish may be clearly observed by comparing exx. 18 and 19 from the Poem of Mio Cid in its original version and in a modern Spanish version:

(18) a. Old Span.: «Dios, que alegre era (ser) -tod cristianismo»

b. Mod. Span.: «Dios, que alegre estaba (estar) -todo el cristianismo»



This process of extension of estar characterizes all varieties of Spanish. It is in Mexico, however, where the change appears to be more advanced, as illustrated in exx. 19-21, which in most other Spanish speaking countries would be constructed with ser.

Michoacán, Morelia (exx. from Gutiérrez 1989):

(19) ...y ahora vivimos allí en Prados Verdes en las casas de Infonavit, están chiquitas, pero están bonitas.

...and now we live in Prados Verdes in the Invonavit houses, they are small (estar), but they are pretty (estar).

(20) ...dicen que allá [en U. S. A.] crecen mucho y aquí estamos muy enanos.

«... they say that over there [in the U. S.] they grow a lot and here we are (estar) extremely short».

(21) ...la que me gustó mucho fue ésta, la de Rambo /¿Rambo?)/ Sí, está muy buena esa película.

...the one I liked a lot was this one, Rambo /Rambo?/ Yes, that movie is (ser) very good.



In view of these historical and dialectological facts, it would not be correct to explain the trend towards using only estar with predicate adjectives in Los Angeles as being a result of influence from English. In other circumstances this would have been a possible explanation, given that English has only one copula which frequently corresponds to both ser and estar in Spanish. If there is any influence, it would be only indirect, simply a favoring factor in the neutralization of ser and estar

Without entering into any detailed discussion of the explanations proposed for the ser/ estar opposition, I would like to point out that, clearly, the choice between ser and estar does not depend on syntactic or lexical decisions only. Indeed, the acceptability of either copula may not be determined by the researcher on the basis of the relevant sentence alone; the extended discourse and shared knowledge between the interlocutors must be taken into account.

In addition, the speech of any single individual is characterized by a certain amount of variation between a more restricted selection of estar, as in ex. 14, and a less restricted one, as in exx. 15-17. That is to say, the stage of complete neutralization of ser and estar with predicate adjectives has not been reached at any point of the bilingual continuum examined, nor in Mexico either.

The hypothesis regarding acceleration of change is supported by a comparison of the results of my study in Los Angeles, with those obtained by Gutiérrez (1989) in Morelia. The general percentages are presented in Table 5.

In Los Angeles, the overall percentage of innovative estar is 34%, while in Morelia it is only 16%. Furthermore, in each of the 3 groups studied in Los Angeles, the younger speakers show a higher percentage of innovative uses of estar. Thus, in a family where there are younger second- or third-group members, these will show marked differences   —174→   in regard to the use of estar when compared with older first- or second-group members of the family. Table 5 shows this generational difference for one family in each site: innovative uses of estar increase from 29% to 59% in the Los Angeles family. By contrast, in the Morelia family they decrease from 22% to 10%, possibly as a result of the higher level of formal education reached by the son in this family.




Conclusions

The uniform speech behavior of the younger speakers in groups 2 and 3, but particularly in group 2 with respect to estar as well as in regards to the verb system suggests that a number of changes affecting Los Angeles Spanish have become stable features of this variety, i. e., we are witnessing the birth of a new dialect or variety of Spanish.

As for the role of English in the shaping of the development of this variety, my conclusion is that in the different areas of the grammar examined (estar, the verb system, and clitic placement) direct influence from English, though possible, is difficult to prove. This does not altogether preclude the possibility of direct transfer from English, which is evident at the level of the lexicon, nor of the type of transfer which involves the higher frequency of use of parallel constructions.

There is nevertheless no evidence that the grammar of Spanish is headed toward imminent collapse, despite the intensive pressure from English. The continuous «renovation» of Spanish through the arrival of new waves of immigrants secures the necessary input to keep Los Angeles Spanish from drifting away entirely from other Spanish dialects. The possibility of continuous contact with full-fledged varieties of Spanish distinguishes the Mexican-American community from other communities which have undergone or are undergoing language shift and eventual language death (such as those investigated by Dorian 1981, Gal 1979, Hill & Hill n. d., among others). A similar observation has been made by Elías-Olivares, González-Widel, and Vargas (n. d.) with respect to the Mexican-American community of Chicago. These authors state that the stabilized use of Spanish in this community «is reinforced by the circulatory migration patterns of Mexican Americans between Chicago and Mexico, and by the increasing need to interact with more recent immigrants and a large population of Puerto Ricans, Central and South Americans living in other areas of Chicago» (9), and cite further supporting evidence from Attinasi (1985) for Mexican-Americans in Northwest Indiana, and from Pedraza (1985) for Puerto Ricans in New York City.

I would like to stress the importance of these observations in understanding the nature of intergenerational studies of Spanish in Los Angeles. The data I have collected in this community show that language shift may occur across generations at the family level. At the societal level, however, the situation appears to be similar to those of Chicago, Indiana, and New York. Speakers who belong in group 3, who do not seem to continue along the lines of a natural historical development of their ancestors' language, are unlikely to pass on Spanish to their descendants. Group 2 speakers, on the other hand, may keep Spanish at home and thus promote the continuous development of a variety which will eventually be recognized as specifically Los Angeles Spanish.



  —175→  
WORKS CITED

Amastae, J. and L. Elías-Olivares, eds. 1982. Spanish in the United States: Sociolinguistic Aspects. New York: Cambridge University Press.

Andersen, Roger W (n. d.) «Interpreting Data: Second Language Acquisition of Verbal Aspect». Ms., UCLA.

Attinasi, John. 1985. «Hispanic Attitudes in Northwest Indiana and New York». Spanish Language Use and Public Life in the U. S.: Sociolinguistic Aspects, ed. by L. Elías-Olivares et al., Berlin: Mouton, 27-58.

Bergen, John, ed. 1989. Spanish in the United States: Sociolinguistic Issues. Washington, DC: Georgetown University Press.

Bickerton, Derek. 1982. Roots of Language. Ann Arbor, MI: Karoma.

Bills, Garland, ed. 1974. Southwest Areal Linguistics. San Diego: Institute for Cultural Pluralism.

Bowen, J. Donald and J. Ornstein, eds. 1976. Studies in Southwest Spanish. Rowley, MA: Newbury House.

Brown, Roger. 1974. A First Language, 2nd printing. Cambridge, MA: Harvard University Press.

Bybee, Joan. 1985. Morphology: A Study of the Relation Between Meaning and Form. Amsterdam: John Benjamins.

Comrie, Bernard. 1985. Tense. Cambridge: Cambridge University Press.

___. 1978. «The Fate of Morphological Complexity in Language Death». LQ 54: 590-609.

Dorian, Nancy. 1980. «Maintenance and Loss of Same-Meaning Structures in Language Death». Word 31: 39-45.

___. 1981. Language Death. Philadelphia: University of Pennsylvania Press.

Elías-Olivares, Lucía. 1979. «Language Use in a Chicano Community: A Sociolinguistic Approach». Sociolinguistic Aspects of Language Learning and Teaching, ed. by J. B. Pride. Oxford: Oxford University Press, 120-34.

___. ed. 1983. Spanish in the United States: Beyond the Southwest. Washington, DC: NCBE.

Elías-Olivares, L., M. V González-Widel and L. Vargas (n. d.) «Variational Use of the Subjunctive in Chicago "La Villita"». Ms., University of Illinois at Chicago.

Ferguson, Charles A. 1982. «Simplified Registers and Linguistic Theory». Exceptional Language and Linguistics, ed. by Loraine Obler and Lise Menn, New York: Academic Press, 49-66.

Fishman, J. A. and G. D. Keller, eds. 1982. Bilingual Education for Hispanic Students in the United States. New York: Teachers College Press.

Floyd, Mary B. 1978. «Verb Usage in Southwest Spanish: A Review». The Bilingual Review / La Revista Bilingue 5: 76-90.

Gal, Susan. 1979. Language Shift: Social Determinants of Linguistic Change in Bilingual Austria. New York: Academic Press.

García, E., F. van Putte and Y. Tobin. 1987. «Cross-linguistic Equivalence, Translatability, and Contrastive Analysis». Folia Linguistics 21: 373-405.

Givon, Talmy. 1976. «Topic, Pronoun and Grammatical Agreement». Subject and Topic, ed. by C. N. Li, New York: Academic Press, 149-88.

Gumperz, John and R Wilson. 1977. «Convergence and Creolization». Pidginization and Creolization of Languages, ed. por Dell Hymes, Cambridge: Cambridge University Press, 151-67.

Gutiérrez, Manuel. 1989. «Español de Michoacán / Español de Los Angeles». Tesis doctoral inédita, University of Southern California.

Hernández-Chávez, E., et al., eds. 1975. El lenguaje de los Chicanos. Arlington: VA: Center for Applied Linguistics.

Hill, Jane H. and K. C. Hill. (n. d.) Speaking Mexicano. Ms., Vols. 1 & 2.

Jacobsen, Teresa. 1986. «¿Aspecto antes que tiempo? Una mirada a la adquisición temprana del español». Adquisición de lenguaje. Adquisição da linguagem, ed. by J. Meisel, Frankfurt: Vervuert, 97-114.

Jakobson, Roman. 1938. «Sur la théorie des affinités phonologiques entre des langues». Actes du Quatrième Congrès Internationale de Linguistes. Copenhagen: Einar Munksgaard, 48-59.

Klein, Wolfgang. 1985. Second Language Acquisition. Cambridge: Cambridge University Press.

Klein-Andreu, Flora. 1986. «La cuestión del anglicismo: apriorismos y métodos». Thesaurus, Boletín del Instituto Caro y Cuervo, Tomo XI.: 1-16.

Labov, William. 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.

___. 1981a. «What Can Be Learned about Change in Progress from Synchronic Description?» Variation Omnibus, ed. by D. Sankoff and H. Cedergren. Edmonton, Canada: Linguistic Research, 177-99.

  —176→  

___. 1981b. «Resolving the Neogrammarian Controversy». LQ 57: 267-308.

___. 1982. «Building on Empirical Foundations». Current Issues in Linguistic Theory. Vol. 24, ed. by W. P. Lehmann and Y. Malkiel. Amsterdam/ Philadelphia: Benjamins, 17-92.

Landa, Alazne. 1989. «Posición de los clíticos en construcciones con perífrasis verbales en el habla culta de Caracas». Ms., University of Southern California.

Levelt, Willem. 1979. «Linearization in Discourse». Ms. Nijmegen: Max-Planck-Institüt für Psycholinguistik.

Malkiel, Yakov. 1967. «Multiple Versus Simple Causation in Linguistic Change». To Honor Roman Jakobson: Essays on the Occasion of His Seventieth Birthday, Vol. 2. The Hague: Mouton, 1228-246.

Martinet, André. 1962. A Functional View of Language. Oxford: Oxford University Press.

Meisel, Jürgen. 1983. «Transfer as a Second Language Strategy». Language and Communication 3: 11-46.

Mougeon, Raymond, E. Beniak and D. Valois. 1985. «A Sociolinguistic Study of Language Contact, Shift, and Change». Linguistics 23: 455-87.

Myhill, John. 1988. «The Grammaticalization of Auxiliaries: Spanish Clitic Climbing». B. L. S.

Ocampo, Francisco. (to appear). «El subjuntivo en tres generaciones de hablantes bilingües». Spanish in the United States: Sociolinguistic Issues, ed. by John Bergen. Washington, DC: Georgetown University Press.

Otheguy, R., O. García and M. Fernández. 1989. «Transferring, Switching, and Modeling in West New York Spanish: An Intergenerational Study» I. J. S. L., special issue No. 79: U. S. Spanish: The Language of Latinos, ed. by Irene Wherritt and Ofelia García, 41-52.

Pedraza, Pedro. 1985. «Language Maintenance in the New York City Puerto Rican Community: Some Ethnographic Evidence». Spanish Language Use and Public Life in the U. S.: Sociolinguistic Aspects, ed. by L. Elías-Olivares et al., Berlin: Mouton, 59-72.

Peñalosa, Fernando. 1980. Chicano Sociolinguistics. Rowley, MA: Newbury House.

Poplaek, Shana. 1978. Quantitative Analysis of Constraints on Code-Switching. New York: Center for Puerto Rican Studies Working Paper No. 2.

Preston, Dennis R. 1982. «How to Lose a Language». Interlanguage Studies Bulletin 6: 64-87.

Romaine, Suzanne. 1982. Socio-historical Linguistics: Its Status and Methodology. Cambridge: Cambridge University Press.

Sánchez, Rosaura. 1983. Chicano Discourse. Rowley, MA: Newbury House.

Silva-Corvalán, Carmen. 1981. «The Diffusion of Object-Verb Agreement in Spanish». Papers in Romance 3: 163-76.

___. 1985. «Modality and Semantic Change». Historical Semantics-Historical Word Formation, ed. by Jacek Fisiak, Berlin: Mouton, 547-72.

___. 1986. «Bilingualism and Language Change: The Extension of estar in Los Angeles Spanish». LQ 62: 587-608.

___. 1987. «Tense-Mood-Aspect Across the Spanish-English Bilingual Continuum». Variation in Language: NWAV-XV at Stanford, ed. by K. Denning, S. Inkelas, F. Mc-Nair-Knox, and J. R. Rickford. Stanford: Department of Linguistics, 395-410.

___. (to appear a) «Spanish Language Attrition in a Contact Situation with English». First Language Attrition: Structural and Theoretical Perspectives, ed. by H. W. Seliger and R. Vago. Cambridge: Cambridge University Press.

___. (to appear b) «Cross-generational bilingualism: Theoretical Implications of Language Attrition». Cross-Currents in Second Language Acquisition and Linguistic Theories, ed. by T. Huebner and C. A. Ferguson, Amsterdam: John Benjamins.

Slobin, Dan. 1977. «Language Change in Childhood and in History» Language Learning and Thought, ed. by John Macnamara, New York: Academic Press, 185-214.

___. ed. 1986. The Cross-linguistic Study of Language Acquisition, Hillsdale, NJ: Lawrence Erlbaum.

Thomason, Sarah and T Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press.

Torres, Lourdes. 1989. «Mood Selection Among New York Puerto Ricans». I. J. S. L., special issue No. 79: U. S. Spanish: The Language of Latinos, ed. by Irene Wherritt and Ofelia García, 67-77.

Trudgill, Peter. 1983. On Dialect. Oxford: Blackwell.

Weinreich, Uriel. 1974. Languages in Contact. Eighth printing. The Hague: Mouton.

Weinreich, U., W. Labov and M. Herzog. 1968. Empirical Foundations for a Theory of Language Change. Directions for Historical Linguistics, ed. por W. P Lehmann and Y. Malkiel, Austin, University of Texas Press, 95-195.




DATA SOURCES EXAMINED

Esgueva, M. y M. Cantarero, eds. 1981. El habla de la Ciudad de Madrid. Materiales para su estudio. Consejo Superior de Investigaciones Científicas, Instituto «Miguel de Cervantes», Madrid.

Instituto de Filología «Andrés Bello». 1979. El habla culta de Caracas. Materiales para su estudio. Ediciones de la Facultad de Humanidades y Educación, Caracas, Universidad Central de Venezuela.

Rabanales, A. y Lidia Contreras, eds. 1979. El habla culta de Santiago de Chile. Materiales para su estudio. Anejo #2 del Boletín de Filología, Dpto. de Lingüística y Filología, Universidad de Chile.

Universidad Nacional Autónoma de México. 1976. El habla popular de la ciudad de México. Materiales para su estudio. Publicaciones del Centro de Lingüística Hispánica 4, México, UNAM.