Culturally Specific Messaging and the Explanation of Contact in Impacted Bilinguals

Abstract

A sketch is offered of a framework that would abandon the familiar notion of a language and the accompanying question of whether it has changed under contact. The framework would focus instead on speakers and on the linguistic consequences of people contact. Speakers in contact settings are not failing or deviating from a language’s norm while attempting to say the same things that are said in non-contact settings; rather, they are succeeding at saying different things. New arrivals face vast differences in the conceptualization of referents between their home precursor setting and the new encounter setting. These differences in conceptualization give rise to large numbers of changes in what speakers say. In most cases, these new things they say are just that, new speech or new messaging with no change in the grammar. But in a minority of cases, the new messaging does have linguistic, that is, grammatical consequences. Changes in the grammars of people in contact thus result not only, and perhaps not primarily, from formal copying or modeling but are responses to new conceptualizations prevailing in the new environment. The distinction between expressions reflecting only new conceptualizations, and those reflecting new conceptualizations and new grammar carries theoretical implications for the way linguists think about the grammars of bilinguals. And it carries applied implications for the way educators think about the linguistic performance of bilingual students, especially in social settings where they are minoritized. Data are drawn from the speech of Latin Americans and their descendants in New York City and other U.S. locales.

1. Introduction

The present paper deals directly with this volume’s topic of the interaction between language and culture by focusing on encounters between peoples of different cultural and linguistic backgrounds, especially those brought together by population movements. It offers a brief sketch of a framework that would abandon the familiar notion of a language and the accompanying question of whether it has changed under contact. The framework would focus instead on speakers in an encounter and on the question of whether their individual grammars have changed. My immediate interest is in the language of Latino immigrants in New York City (NYC) and their descendants. Some of the data come from the formal interviews of the Otheguy–Zentella Corpus or OZC (Otheguy and Zentella 2012), others from my first-hand participant observations.

The paper rests on the simple idea that speakers in contact settings are not failing or deviating from a norm while attempting to say the same things that are said in non-contact settings; rather, they are succeeding at saying different things. The central thesis of the paper is that newly arrived Latino immigrants in NYC face vast differences between the conceptualizations of referents in Latin America and the, for them, new conceptualizations of essentially the same referents in the US. These differences in conceptualization between the two societies give rise to large numbers of changes in what NYC Latinos say to one another when using the linguistic resources they bring from Latin America. In most cases, these new things they say are just that, new speech or new messaging with no change in the grammar. But in a minority of cases, the new messaging does have linguistic, that is, grammatical consequences. Changes in the grammars of immigrants in a contact setting thus result not only, and perhaps not primarily, from formal copying or modeling but are responses to new conceptualizations prevailing in the new environment. The distinction between expressions reflecting only new conceptualizations, and those reflecting new conceptualizations and new grammar carries theoretical implications for the way linguists think about the grammars of bilinguals. And it carries applied implications for the way educators think about the linguistic performance of bilingual students, especially in social settings where they are minoritized.

The linguistic forms that are noticeable in NYC Latinos lend themselves to a preliminary division into the familiar categories of: (i) lexical units such as apoinmen, bildin, lánlor, which are usually not found in Latin America and are adapted forms of appointment, building, landlord; and (ii) morphosyntactic units used for purposes not generally found in Latin America, as in la próxima vez cuando me pongo bravo “next time when I get angry” [OZC Informant 417P], where indicative inflection pongo is used in NYC even though subjunctive inflection ponga would have been more likely in the speaker’s place of origin. I take it as given that cases of (i) are obvious instances of a grammatical difference (in this case a lexical difference) between speakers in NYC and elsewhere. So I focus here on cases like (ii), and I ask which of them represent simple differences in messaging following from different conceptualizations between the US and Latin America and which represent actual differences in grammar.

2. Theoretical Terms and Assumptions

A number of theoretical positions inform the proposal that is sketched out in this paper.

Linguistic consequences of people contact. The linguistic behaviors in (i) and (ii) above are usually studied under the rubric of languages in contact, a type of research long practiced in many parts of the world but that solidified its name and received renewed impetus from the publication in the US of Uriel Weinreich’s ([1953] 1967) book by that title (for recent work and references on language contact see Mufwene and Escobar 2022). But rather than languages in contact, the topic can be better conceived of as the linguistic consequences of people contact, for three reasons. First, because it is obviously people who are in contact, not the social abstractions constituted by named languages. Second, because conceiving of the field in terms of languages takes the focus off the behavior of concrete speakers to the detriment, as we will see, of the theoretical coherence of the enterprise. And third, because, for all its scientific bona fides and the good intentions of its proponents, the conception in terms of languages in contact often redounds, as I discuss below, to the disadvantage of bilingual speakers, especially minoritized bilinguals.

Impacted bilinguals. I call the emergent or established bilinguals in whom either speech or grammar shows the consequences of exposure to new conceptualizations impacted bilinguals. It happens in encounter settings that, for reasons of origin, education, personality, profession or identity, there are always individuals in whom neither speech nor grammar is affected. They are not included in this discussion. Impacted bilinguals, then, are the vast majority of encounter speakers in whom we detect the effects of new conceptualizations at least in speech, that is, in the content of what they say, and often as well in language, that is, in the content of their grammars.

A speaker-centered framework and the idiolect. The study of the linguistic consequences of people contact adopts a speaker-centered framework, making use of the underutilized notion of the idiolect. The idiolect is the speaker’s own grammar, his or her inventory of phonological, lexical, and morphosyntactic features, created de novo starting from earliest childhood, on the basis of exposure to the speech of family and community. Speaker-centered theory assumes the cognitive reality of these grammatical units. But it recognizes no empirical or theoretical advantage in conceiving of them as grouped into a named language. Idiolects, then, are not personal versions of the named languages with which society deems speakers to be affiliated. Instead, idiolects are, from the strictly linguistic point of view, unaffiliated collections of phonological, lexical, and morphosyntactic features that help individuals communicate with other individuals who, to greater or lesser extents, share similar features.

Meaningful grammar. Basing the framework on the speaker and not the language stems from a conception of grammar as a tool for communication. From all appearances, it is communication that is the goal of the speaker even if it obviously cannot be the goal of the named language. The instrumental conception of grammar leads to a semiotic conception of its units, which must be meaningful if they are to facilitate communication. I thus assume not only the widely accepted idea that words have specifiable meanings but also the less consensual position that grammatical forms and constructions likewise do so. The position that grammar is directly meaningful is adopted by many theories, especially by those situated within what Butler and Gonzálvez-García (2014) call the functional-cognitive space. In the position adopted here, idiolectal grammar consists of words, affixes, inflections, and syntactic orders and positions, all of which may be meaning-bearing units.

The tool and its uses. The idea that grammar is a tool for communication and that its units are meaningful leads to a distinction between the structure of the tool and the uses to which it is put. The metaphor of the tool highlights that when these tools are deployed for new purposes the tool itself does not necessarily change. The heel used to hammer a nail into the wall maintains the unchanged shape of a part of the shoe. The stapler used as a paperweight keeps the structure of a device to clip papers together. When immigrants bring linguistic forms to a new society and use them for new expressions, they do it sometimes relying on the unchanged meanings of these forms, and only sometimes on newly created meanings for them.

Encoded meaning versus on-going messaging. To distinguish between these two possibilities as I aim to do here, it is useful to draw from theories that distinguish sharply between encoded meaning (the tool) and on-going messaging (the uses of the tool), as does for example the Columbia School (Diver [1975] 2012, [1995] 2012; Huffman 2001). Encoded meanings are the semantic contents of the individual lexical and morphosyntactic forms of the language. Messaging on the other hand has to do with the communicative intentions of speakers, which end up corresponding loosely to the contextually conditioned abductive inferences of listeners. Encoded meanings are the grammar, whereas on-going messaging is outside of grammar and pertains to speech or language use.

In the familiar example, the shivering person who says freezing cold now has used the meanings of cold, freezing, now and of the adjective-noun order. The different resulting inferences of “I’m feeling cold” or “you better close that window again” are aspects of the messaging that in different situational, personal, and physical contexts are derived from those meanings. Similarly, according to Stern (2022) the grammatical form -self has the encoded meaning Insistence on an entity. In an utterance like he hurt himself, the traditional notion of reflexive is not a meaning encoded in the grammar but, at best, one way to look at an aspect of on-going messaging.

In a speaker-centered approach where all linguistic forms are potentially meaningful, the use of language involves acts of selection between the meanings available in the idiolect. But there is no implication that the choice is always a fully conscious one, as linguistic self-awareness and monitoring are influenced by the moment, the context, the frequency of the selected form, its greater or lesser frequency of co-occurrence with other selected forms, etc. In addition, there is no implication that these not always fully conscious choices of meaningful lexical and grammatical forms are guided only by the exigencies of description and reference in an objectivist semantics. For, as well, the choices reflect the expression of point of view, comment on the scene, attitude toward an event, personal identification, social belonging, judgments of context, speaker and hearer characteristics, etc.

In the study of the linguistic consequences of people contact within a framework that distinguishes encoded meaning from on-going messaging, it is only sometimes the case that a fully tested meaning hypothesis is available. In the four cases discussed below, the encoded meanings will be those that I have judged to be essentially correct, drawn from literature published under different theories, requiring only that the analyses be sufficiently mindful of the meaning—message distinction so as to provide answers to the question about contact being asked here.

Message and reference. I also adopt the distinction, adapted from Frege (1948), between message and reference. In my usage, reference has to do with entities, properties, conditions, or events out in the experiential realm that speakers point to, or make salient, in particular acts of speaking. Frege’s examples were the expressions evening star and morning star, which communicate somewhat different messages, say different things about, what is in fact a single referent, the planet Venus. Lexical and grammatical units, then, neither have nor make reference. It is speakers who make references (different ones at different times) aided by the meanings of these units.

Encounter settings and conditions. My interest here is in speakers in encounter settings or encounter conditions. Encounters are geographic or social dimensions where people of different cultural and linguistic traditions come into contact. Depending on historical circumstances, these speakers are immigrants, exiles, refugees, expatriates, invaders, occupiers, enslavers, invaded, occupied, enslaved. I focus on a particular people in a particular kind of encounter (Latinos in NYC) but the framework advanced here is intended to be of general applicability. And I concentrate on the encounter’s earliest participants, but what I have to say applies in many cases as well to their descendants. I use the term precursor for areas of origin of encounter speakers where people contact is not a prominent feature. Many aspects of NYC life can be usefully thought of as settings of encounter for impacted bilinguals. In contrast, their Latin American homelands are their corresponding precursor societies.

It is sometimes straightforward to think of encounter and precursor as the names of geographic locations. But it is often better to think of them more abstractly as sociolinguistic dimensions or fields. This allows for taking into account conditions not always tied exclusively to geography, such as the speakers’ socioeconomic standing, level of education, or disposition toward assimilation. For example, highly educated Latin Americans recently arrived in NYC or otherwise closely identified with their societies of origin may in some cases exhibit messaging preferences and choices of encoded meaning that are more usefully thought of as reflecting precursor dimensions rather than encounter ones.

The analytical question that arises in the encounter. The question raised in this paper is whether or not the speech behavior of impacted bilinguals (or their descendants) in the encounter reflects differences between the forms and encoded meanings of their idiolects and those of speakers in the corresponding precursor fields. My answer to this question will avoid established terms like interference, simplification, and cognitive load, as well as formal notions like structural convergence, calquing, and feature activation. It will avoid, too, the familiar position that impacted bilinguals are making mistakes or that they are in some way exceptional halting users of an incomplete grammar. The answer will rest instead on differences in the conceptualization of the same or closely similar referents between the precursor and encounter settings, and their consequences in either communication or grammar, or in both.

3. Messaging Differences Related to Meaning Differences

General comparisons between societies often show differences in conceptualization that lead to differences in messaging by their speakers. These cross-cultural, conceptually driven differences in messaging are of two kinds: (i) differences in conceptualization that have led to, and are detectable in, differences in inventories of encoded meanings, and (ii) differences in conceptualization that are only detectable in messaging differences, not in differences in meaning inventories. Even though my focus is on differences in conceptualization of the second type, a brief reminder of differences in encoded meaning makes for instructive prologue.

The observation is old and widely accepted in language-centered linguistics that the meanings of words in different languages are in many cases not congruent, and that the same is true of the meanings of grammatical forms (Saussure [1916] 1972/1983; Wierzbicka 1997). In studying Karok in Northwestern US, where cardinal directions are expressed in relation to the Klamath River, Bright and Bright (1965, 251ff) report a meaning Away from the river for the word maruk. This meaning is not found in the words of Indo-European languages, where cardinal orientations are usually pegged to the movements of the sun. Closer to home, the usages of the English bring and Spanish llevar suggest that their encoded meanings are not congruous, as llevar is Carry from ego’s current situation whereas traer is Carry toward ego’s current situation. This makes them both different from the English bring, whose meaning allows for both directions. Similarly, English toe almost certainly has a different encoded meaning from Spanish dedo, which can apply to digital extremities of both feet and hands.

In order to make these accounts applicable to the speaker-centered framework, one should note that it is not the Karok language that conceptualizes cardinal orientations in terms of the river, but the Klamath people who do, and who have, in their grammars, encoded meanings that reflect this conceptualization. And it is not that Spanish makes a distinction between traer and llevar missing in English, but rather that most Latin Americans, Spaniards, etc., have lexicons where a semantic difference operates that is inoperative in the lexicons of most monolingual Australians, Britons, US North Americans, etc.

The meaning inventories of speakers of different cultural groups have been shaped, then, by these conventionalized conceptualizations of the referential world. The point to note is that the different conceptualizations are primary, and that differences in encoded meaning are due to them. In saying that the Klamath people speak of cardinal points in terms of upriver, downriver, toward the river and away from the river, Bright and Bright (1965, passim) make the point well, claiming that between the Klamath and other people there is a difference in what they call conceptual structure. Conceptual structure, they maintain, is “best regarded not as part of language itself, but as part of nonlinguistic culture, albeit a part having especially close relations with language” (1965:257n). The key idea is that the same objects, conditions, and situations appear to be envisioned differently by speakers in different cultural settings, that is, they appear to have been conceptualized differently. Importantly, this means that these are not simply matters of lexical difference. It is not simply that in the idiolects of the Anglophone, Francophone, Garophone, Hispanophone, and Karokphone peoples, one finds that there is this, or that, meaning that has, or has not, been lexicalized elsewhere. Rather, it is more primarily the case that over the course of their histories, these societies have developed certain conventionalized conceptualizations that over time have been formalized into the encoded meanings of words (conceptualizations about cardinal orientations, the bringing and taking of objects, etc.).

Several points in the exemplification above need stressing. First, note the distinction between the encoded meanings in the speakers’ idiolects and the conceptualization that has prompted the use of these meanings. On a particular occasion, it is not the same to say one is traveling east as it is to say that one is traveling away from the river, a messaging difference directly connected to the difference of encoded meaning between east and maruk. Second, note that the discussion has relied on the notion of reference as distinguished from both encoded meaning and on-going messaging. When two speakers who are standing east of the Klamath River aim to push farther east, the one who speaks in terms of going farther maruk is making the same reference as the one who speaks in terms of going farther east, even though their messaging is different. Likewise, the speaker who in the US talks about applying nail polish on their toes is messaging differently from the speaker in Venezuela who talks about applying it to “the fingers of the foot”. But they are both making the same reference, polishing the same body part. Reference, or better yet, referring is something speakers do with the encoded meanings of their words, not a meaning component of the words themselves.

4. Messaging Differences Not Related to Meaning Differences

In the comparisons made above, the society’s particular conceptualizations have penetrated all the way into the speakers’ lexicons, becoming encoded in meanings that are found in one community of speakers but not in others. It would be useful to refer to these differences in conceptualization as cross-cultural conceptual gaps. As in the examples just discussed above, the ones to be offered presently also reveal a conceptual gap. But in what follows the gap will not have to do with meanings that are unavailable in a particular society (the meaning of maruk is not available to the non-Klamath). Rather, the conceptual gap will now separate two societies that have available meanings that could have been chosen to express the same conceptualization of a referent, but that were not chosen because the referent is conceptualized differently.

There is no need to take here the further step of inquiring why it is so often the case that in different societies there are differences in conceptualization that lead to the deployment of different encoded meanings to make essentially the same reference. Interesting as such an inquiry may be, it is beside the point here. The reason for the cross-cultural non-congruousness of meanings that we just saw, and for the non-congruousness of conventionalized conceptualizations that we will now see, is not relevant to the question whether encoded meanings have changed in the encounter. It is enough to know that, when comparing societies, rampant non-congruousness is seen to be the case for both encoded meanings and conceptualizations. We do not need to now press further and ask why this is the case.

Conventionalized differences in conceptualization and the resulting messaging are of central importance for the study of contact. For the motivation, and therefore the deeper explanation, for many of the linguistic consequences of people contact must be sought in messaging habits that responded to culturally dictated conceptualizations in precursor settings and their frequent abandonment and replacement in the encounter by different messaging motivated by different conceptualizations. In settings of encounter, impacted bilinguals often tend to set aside conventionalized precursor conceptualizations and to replace them with those of their out-group neighbors. In the examples that follow, NYC Latinos have started to move away from messaging in relation to particular objects, entities, conditions and situations in the way one does in Latin America, and have started to do it in the way one does it in the US. Four cases (supplemented at the end by a larger list) of this type of US messaging by Latinos in NYC are analyzed in this section. In the next section, we will then see that the new messaging is achieved in some cases through the use of the same lexical and morphosyntactic meanings that underlie speech in Latin America, whereas in others it involves transformed meanings assigned to existing Latin American forms.

Each of the four examples that follow contains three utterances, labeled (a), (b), and (c). The utterances in (a) and (b) are made up by me and, for readers who need it, serve as translations of each other. The (c) utterances are real NYC data, reflecting widespread usage among Latinos in NYC. In all the (a), (b), (c) utterances the referent is the same. But (a) reflects the prevailing US conceptualization of the referent, while (b) reflects the favored conceptualization of the same referent in Latin America. Then, (c) reflects the adoption by Latin Americans in NYC of US conceptualizations. For the cases in (1), (2), (3), the encoded lexical or grammatical meanings used in the encounter (c) utterance are familiar from Latin American grammars. That is, the closing of the conceptual gap in the encounter has been achieved in (1), (2), (3) using meanings familiar from the precursor field, so that there is no linguistic difference between Latin American and NYC idiolects. But for (4), I will argue that in the documented (c) utterance there is a form whose encoded meaning is not found in the idiolects of the corresponding Latin American precursor field.

Consider example (1):

(1a) My last name is Smith;

(1b) Mi apellido es Crespo.

Utterances (a) and (b) are used to message about the same referent, the patrilineal surname. Both of the societies where Smith and Crespo live place the surname at the end of the onomastic syntagm, John Smith, Juan Crespo, not at the beginning, as is done for example in China (the cases in Latin America where the mother’s surname is tacked on are not at issue now). The words name and last have congruous meanings in nombre and último. The conceptual gap between (a) and (b) reflects the fact that in Smith’s society the patrilineal surname is conceptualized as a last name, but not so in Crespo’s, even though there too it usually does go last. The difference in messaging between (a) and (b) thus reflects the conceptual gap between the two societies with respect to the patrilineal surname referent. Now consider expression (c), which constitutes familiar usage among NYC’s impacted bilinguals:

(1c) Mi último nombre es Crespo [505P].

The conceptual gap separating (1b) from (1a) disappears in (1c). The messaging in (1c) is now on the basis of the same conceptualization as in (1a). The use of the words último “last” and nombre “name” by Latin American immigrants in NYC to refer to the patrilineal surname is a case of when in Rome do as the Romans do. It reflects a tendency to say in New York what other New Yorkers say. In my terms, (1c) reflects the tendency to message, not on the basis of the conceptualization guiding messaging in the precursor setting, but on the one that guides messaging in the encounter.

In example (1), the difference between messaging and conceptualization is easy to perceive. Messaging involves the specific, momentary, and contextually dependent communication between language users in context. In contrast, a conceptualization is the society’s standing conception of entities, conditions, and situations. The conceptualization of family names as last names leads to different messaging, not only in My last name is Smith, but also in:

I’ve noticed that he says he’s Henry Joseph, so I guess he’s a Mr. Joseph, not a Mr. Henry

Don’t be confused: The Xi in Xi Jinping is not his first name

In these cases, a conceptualization has been a stable guide post that leads to many particular kinds of messaging and the consequent choice of encoded meanings in relation to a referent, as Bright and Bright (1965) explained above. But in other cases, the conceptualization appears to be less general, guiding the choice of meaning in a more localized manner, so that the distinction between conceptualization and messaging is blurred. In what follows, we will use the term conceptualization in this specific way, dispensing for now with the term message.

Consider now example (2):

(2a) They’re not going to have face-to-face classes;

(2b) No van a tener clases presenciales.

Both (2a) and (2b) are used to make reference to the typical classroom setup that was common in many countries but abandoned during the 2020 COVID-19 pandemic. But the conceptualization of this same referent is different in different societies and consequently so are the encoded meanings selected. In (2a), the conceptualization is of instructors and students so situated as to be able to look at each other directly; in (2b), it is that of students occupying, and being present in, the same space as the instructor. As a result of this difference in conceptualization, the meaning of face is chosen for (2a) but the parallel meaning of cara is not chosen for (2b). That is, the available linguistic resources of speakers in Latin America are not used because the conceptualization of the classroom in Latin America is different. Now consider the familiar NYC usage of (2c):

(2c) No van a tener clases de cara-a-cara [504U]

Here, again, (2c) closes the conceptual gap between (2a) and (2b). The NYC speaker is now using their familiar word cara under the guidance of the US conceptualization, saying in NYC what other New Yorkers say while making use of linguistic resources brought from Latin America (cara, de, a).

The objection may be made that, in the cases just discussed in (1) and (2), a progression has been presented from general conceptualization to intended message to meaning selection. The objection can be that this has it backwards, or that it is circular, since I in fact have used the encoded meanings I have assumed for cara, face, last, name, último to learn about the intended message in each case. But I am describing the situation from the point of view of the speaker, not the linguist. Speakers clearly make meaning choices in the service of priorly intended messages that, in some cases, reflect prevailing conceptualizations. This is why it makes sense to think of meaning selections by speakers as being guided by the way referents have been conceptualized. To be sure, in theoretical works engaged in basic research, establishing the exact nature of the aspect of the conceptualization that prompted the choice of a meaning by the speaker is itself a demanding analytical effort. But for the point being made here, there is no circularity in my recognizing a particular conceptualization on the basis of what I can tell about the encoded meanings of the forms selected. There can be little doubt that, while making the same reference, speakers in Latin America and NYC are saying something different, expressing a different conceptualization, when they respectively say apellido or último nombre, and when they say presencial or cara-a-cara. And there can also be little doubt that último nombre and cara-a-cara are reproducing the conceptualizations that guided meaning selection in last name and face-to-face.

Informed by the theoretical position of meaningful grammar, my analysis for the next two cases, which involve grammatical forms, is the same as for the two previous ones, which involved lexical forms.

Consider (3):

(3a) He’ll buy it for you when he gets back from dropping her off at daycare;

(3b) Te lo va a comprar cuando vuelva de dejarla en daycare.

The when/cuando clauses in (3a) and (3b) both make reference to a future event. But in (3b), the cuando event has an element of uncertainty that the when event does not have in (3a). The different conceptualizations are most clearly revealed by the differences in the encoded grammatical meanings chosen for (3a) and (3b), in parallel to the different meanings chosen for (2a) and (2b) to express the conceptualizations related to looking at faces versus being present.

In (3b), the speaker has chosen a subjunctive inflection vuelva. For the encoded meaning of this inflection, a widely recognized source reports that “not a small number of linguists have replaced the traditional contexts of modal alternation with a belief in the non-assertive nature of the subjunctive” (RAE (Real Academia Española) and AALE (Asociación de Academias de la Lengua Española) 2009, p. 1869, my translation). Along similar lines, the Latin subjunctive has been analyzed as calling the event into question or indicating the possibility that it may not have occurred (Diver [1992] 2012, pp. 182–86). This suggests that the use of subjunctive vuelva in (3b) follows from a conception of the future event of returning as one involving an element of uncertainty.

In contrast, in the meaning of the inflection he gets in (3a) there is no element of certainty or uncertainty (cf. the zero inflection in, for example, I insist that he get back right away, where there would be such an element). As an additional consideration, it must be kept in mind that for speakers of (3a) most of what one could call subjunctive forms with meanings of questionable occurrence are homophonous with indicative forms with meanings of past time (cf. if he calls vs. if he called). This may be motivating speakers in (3a) to avoid when he got back from daycare, which would have given rise to an overall inference of a past occurrence, leading to an incoherent utterance. Be that as it may, the fact remains that the future event in (3b) has been presented with an irrealis element that is not found in (3a), creating a conceptual gap.

Now consider the familiar usage by NYC impacted bilinguals:

(3c) Te lo va a comprar cuando vuelve de dejarla en déiquear [Informant 503U].

Here, (3c) closes the conceptual gap between (3a) and (3b). Informant 503U (we will call her Reggie) has adopted the US conceptualization of the when event. The event is now seen as indifferent to considerations of certainty. As in (3a), where the choice of indicative he gets ignores the available he get0 because the messaging does not involve considerations of certainty, now, in (3c) the choice by Reggie of indicative vuelve ignores the available subjunctive vuelva because the messaging, following from the US conceptualization, is also unrelated to the question of certainty.

The point about (3c) closing the conceptual gap between (3a) and (3b) requires taking notice of the research by Schwenter and Hoff (2020), which shows that the use of subjunctives (like vuelva) varies considerably between countries and regions in Latin America and Spain. I read their finding to be telling us that whether to conceptualize occurrences in terms of relative certainty is a matter of local convention; and that it is this localized range of the conceptualization that leads to differences in the decisions by speakers in different Latin American and Spanish societies to use a subjunctive or an indicative. Still, my (3b) is the correct comparison for the analysis of (3c) because Reggie is the NYC-born daughter of parents from Cuba who would show a strong tendency to use subjunctive vuelva here, providing Reggie the data on the basis of which she developed her grammar of verb modality. For this informant, and for many like her in NYC, it is clear that (3c) is a case of closing the conceptual gap between (3a) and (3b), her usage reflecting the adoption of a US conceptualization that was not part of the precursor conditions that provide the relevant comparison.

The case of (3) is analogous to those of (1) and (2). In the (b) utterances in those examples, meanings were used that were suitable to express the conventionalized Latin American conceptualizations (apellido, presencia, subjunctive inflection) and that ignored the availability of certain other meanings (those of cara, nombre, último, indicative inflection). But in the encounter, the newly adopted conceptualizations of family names as last names, of classroom arrangements as cross-seating, and of future when events as unrelated to matters of certainty made useful those hitherto ignored meanings. In all three instances, the (c) utterance is a case of speakers reaching into the existing inventory of Latin American linguistic resources in order to express US conceptualizations.

Now consider example (4):

(4a) On the second trip, I remember meeting my grandmother;

(4b) En el segundo viaje, yo recuerdo haber conocido a mi abuela

Here, the difference in conceptualization has to do with the greater directness, immediacy or salience of the event of meeting grandmother in (4a) than in (4b). Consider choices of lexical meanings first. In classifications of action verbs versus stative verbs, meet would belong to the former and conocer to the latter. In (4), action meet and stative conocer are translation equivalents, but meet is still more of an action than conocer, which in other contexts easily stands translation as stative know. I take it that actions are more directly noticeable than states, the messaging in meeting grandmother being thus more salient and immediate than in haber conocido a mi abuela. With regard to grammatical meanings, the inflections chosen in the two utterances, remembering versus haber conocido, appear to have different meanings. Neither form encodes the semantic substances of number, person, tense, and mood that are associated with forms interpreted as verbs or occurrences. But the form -ing is part of the System of Vividness (Huffman 1989). The form -ing, with a meaning of More Vivid, is opposed to -en, with a meaning of Less Vivid (Huffman points to utterances like the biting dog vs. the bitten dog). No such semantic substance appears to be part of the meaning of the form -er, or of -er + ido, further suggesting that the meaning choices made in the two utterances follow from the more immediate or salient conceptualization in (4a) than in (4b).

Consider now (4c), which is common among NYC impacted bilinguals:

(4c) En el segundo viaje, yo me recuerdo conociendo a mi abuela [201U]

The US conceptualization, now adopted in NYC, has a similar effect as in the previous cases, closing in (4c) the conceptual gap between (4a) and (4b). In NYC, informant 201U (let us call him Roberto) moves toward the more salient US messaging by replacing the form -er (or -er + ido) with the form -ndo, to be discussed below.

Once again, it is worth stressing that (3) and (4) are like (1) and (2). Choices of encoded meanings reflect cross-cultural differences in conventional conceptualization whether the meanings are lexical or grammatical. This is so whether the conceptualizations have to do with family names or classroom arrangements, as in (1) and (2), or with the certainty or directness of occurrences, as in (3) and (4). In both sets of cases, the different conceptualizations that led to different meaning choices in the (a) and (b) utterances give rise to a conceptual gap between the precursor societies that meet in the NYC encounter. And in both sets of cases, the gap is bridged in the (c) utterances.

The (c) utterances above are but a very small sample of the repeated practice among Latinos to conceptualize a particular referential situation in NYC, not in the way that their ancestors, or they themselves, did in the precursor setting, but in the way of their encounter neighbors (neighbors who peculiarly think of family names as being last, of regular classes as containing faces looking at one another, of future when events as indifferent to matters of certainty, etc.). Added to the many unique lexicalizations (cf. traer vs. llevar), US immigrants and their descendants move in a speech landscape filled with ways of thinking about referents that did not exist in Latin America. Time and again, what they say about a referent in the US encounter is often not what is said about it under precursor conditions. As dinner becomes ready in the precursor field, diners are summoned to eat with a la mesa, but in the encounter they are often told that la comida está servida (cf. dinner is served); at the end of the meal in the precursor setting, the diners will use their cucharitas de postre but in the encounter they will often use cucharitas de té (cf. tea spoons). And there may be talk over dinner about a criminal that was condemned to cadena perpetua, discussed in the encounter as getting vida en prisión (cf. life in prison). When children play a board game in the precursor setting, they often say me toca a mí but in the encounter one more often hears es mi turno (cf. it’s my turn). And a promise of returning an object in the precursor setting is te lo doy de vuelta or te lo devuelvo, but in the encounter it is te lo doy para atrás (cf. I’ll give it back to you). And on and on. In a vast number of cases, the linguistic features brought from Latin America are adapted for messaging that parallel those of the US, rooted in what have come to be very similar or identical conceptualizations of referents.

The discussion of (1) through (4) above has illustrated the reason that there can often be linguistic consequences to people contact (the peoples of the encounter select meanings based on conceptualizations that differ from those of the precursor society). But I have not yet answered the question whether or not these expressions actually show linguistic consequences. I address this question in the next section.

5. Linguistic Consequences of People Contact

Answering the question about the linguistic consequences of people contact requires that we distinguish between coded meanings of the precursor idiolects that have been, so to speak, good enough to express US conceptualizations and coded meanings that have not. When they have been good enough, people contact has had no linguistic consequences. When they have not, people contact has had consequences, as the idiolects of the immigrants, and more often those of their children, have changed to meet the new communicative demands and these changes are inherited by subsequent generations. What we need to know, then, is whether the encoded meanings that the impacted bilinguals rely on to produce, for example, the último nombre of (1c) or the cuando vuelve of (3c), both of which reflect the same process of conceptual-gap closing, are the same meanings that came into the encounter from Latin America.

Among the linguistic forms that we have been studying in (1c) and (2c) above, the words cara, nombre, último, and the adjective-noun positioning in último nombre (and probably also the form a in cara a cara) are cases of people contact with no linguistic consequences. In these cases the NYC speakers are just saying something different about the referent than is said in Latin America. But the meanings they are using are the same. That is, these lexical and grammatical units almost certainly have the same meaning in NYC idiolects as they still do in the precursor settings, as their current uses in Latin America suggest. Note in particular that the phrase último nombre itself is not new in NYC, and that it is perfectly normal to use it in Latin America to refer, for example, to a name at the bottom of a list (e.g., el último nombre de la lista es Manolo “the last name on the list is Manolo”). Whether the referential purpose for which último nombre is used is the entry at the bottom of a list or a surname is not a matter of grammar but of conceptualization and reference. The same is true for cara a cara. The phrase is normal in Latin America to express the conceptualization of a type of private conversation. And it is not a matter of grammar that one now also uses it in NYC, under the US conceptualization, to refer to a type of classroom meeting. In (1c) and (2c), the tool (the grammar) has remained the same as speakers use it differently.

The same analysis applies to (3c). Similar to cara, último, and adjective-noun order, it is almost certain that the meaning of the indicative form in cuando vuelve de dejarlo is the same as the meaning of the indicative in Latin American idiolects. The only thing of note in (3c) is the adopted US conception of the future event in terms that say nothing about assertiveness or questionability. Once the future event of returning is thus conceptualized, then the appearance of the indicative in cuando vuelve in (3c) is unremarkable. The speaker in NYC is just saying something different from the speaker in Latin America, and because the message is different, a different meaning is used. The parallel is clear. In (1c) and (2c), with último nombre and cara a cara, there have been no lexical consequences to people contact because the precursor lexical meanings have done the job of expressing the US conceptualizations; in (3c), there have been no morphosyntactic consequences to people contact because the precursor grammatical meaning has likewise been good enough for the job. In both the lexical and morphosyntactic cases, the old tool remains unchanged as it is put to new uses.

But in (4c), the situation appears to be different. In me recuerdo conociendo a mi abuela, the evidence suggests that we have a case where the meaning of the precursor idiolects fails at the new task of expressing the adopted encounter messaging. The goal of expressing the US conceptualization about meeting grandmother as a salient or direct event appears to have been achieved in (4c), but with an altered meaning for -ndo. That is, the meaning of -ndo appears to be different in the grammar of NYC-born Roberto than in the grammar of his parents and other Cuban-born speakers. Now, there is not available a fully worked out meaning analysis of -ndo in the precursor idiolects parallel to the study of -ing discussed above. But we do have traditional efforts that describe -ndo as always and in every instance of use being imperfective (RAE (Real Academia Española) and AALE (Asociación de Academias de la Lengua Española) 2009, p. 2075). In contrast, there is no reason to believe that there is any notion of imperfectivity in -ing, whose only meaning, as we saw above, is more Vivid. This allows -ing to be usable for both punctual and progressive messaging, whereas -ndo, in its precursor meaning, is usable only for progressive messaging.

Yet, and this is the key point, the event of meeting grandmother in (4c) is not a progressive occurrence but a punctual one. It describes the moment of meeting her. (In Roberto’s interview in the OZC, there were four other punctual -ndo out of a total of 18, so this was for him an established usage). In other words, in (4c) conociendo does not have a continuing interpretation as it must have if -ndo were the same in NYC as in Cuba. It seems, then, that the reason that Roberto can use -ndo for punctual events is that his -ndo has a new encounter meaning that has shed the precursor element of imperfectivity, and that, like -ing, has very likely a meaning of Vividness. The conclusion, then, is that the meaning of -ndo in Roberto’s idiolect is almost certainly not the same as in those of his parents, but has probably been equalized with that of the -ing of his neighbors. In (4c), we very likely have an instance of a consequence of people contact, not simply in conceptualization, but in the encoded meanings of Roberto’s idiolect.

6. Named Languages in Encounter Linguistics: Theoretical and Applied

In the preceding explanations, the reader will have recognized the near total absence of the terms English and Spanish, corresponding to the general claim that the concept of a named language does not serve the linguist well when studying the linguistic consequences of people contact. In settings of encounter, the wide coverage of named language terms makes it difficult to distinguish differences in messaging (stemming from differences in conceptualization) from systemic linguistic differences in coded meaning. Relying on the named language, the cuando vuelve of (3c) would be regarded as no different from the conociendo a mi abuela of (4c). Under the language-centered approach, the differences in usage between Latin America and NYC that are noticeable in (3c) and (4c) would, in both cases, lead to the conclusion that the Spanish of NYC is different from that of Latin America. Yet, we know that grammatically this is only true of (4c). Thus, the named language becomes an obstacle to our understanding of contact speakers and the way they meet their communicative needs.

Moreover, by making it difficult to distinguish innovative messaging (based on a new conceptualization) from innovative grammar, the named language forecloses the possibility of explaining the latter by the former. When every bit of knowledge about elements of a communication is seen as part of knowledge of a language (when knowing for example that Smith and Crespo are called last names in the US is regarded as part of English), there is nothing left outside of language to serve as explanans. This is important because it is precisely the conceptual structure that Bright and Bright (1965) insightfully located outside of language that, as we have seen, serves to differentiate between new speech using old grammar and new speech using new grammar.

In contrast to the traditional approach based on the named language, the speaker-centered framework facilitates maintaining the distinction between language structure and language use. In such a framework, the use of language is not a supplement, often called pragmatics, to the more central level of structure. Rather, usage is an integral input to the process of grammatical analysis, because it is in the usage of speakers that the linguist finds the data, and the object of explanation, for the hypothesized coded meanings. And because it is through the direct participation of the language user in the inferential process that links meaning to messaging that an understanding of the observed distributions of linguistic forms is achieved. The speaker-centered framework is thus particularly suited for the study of the idiolects of the encounter. It provides a direct role to the user of the language who can adopt new conceptualizations using existing meanings, thus, limiting cases of grammatical difference between encounter and precursor settings to instances where precursor meanings cannot account for encounter usages, as in Roberto’s case (4c).

Sponsored by the named language, the unwarranted expansion of language contact and change that results from herding into the grammar facts belonging to culture and communication is exacerbated by the assumption that while lexis is meaningful, grammar is rule-based. The concept of a language as a reified object—in this case Spanish—existing out there in the world encourages the notion that its orderliness, as that of any other object, derives from rules, such as the rule that dictates that subordinate clauses with future reference headed by cuando take the subjunctive (as do those headed by para que “in order to”, antes de que “before”, etc.). This rule is then violated in the syntactic convergence or structural calque found when an indicative is used in cuando vuelve in (3c). But we now know that the subjunctive was being used in Latin America all along in response to the conceptualization of the future occurrence as containing an element of uncertainty, not in response to a rule. When exposed to a different conceptualization of the future event in when he gets back in (3a), the encounter speaker readily switched to the indicative in cuando vuelve in (3c). This suggests that it is not that the reified object Spanish has undergone a rule change but that speakers have engaged in a reconceptualization expressed through unchanged grammar. The named language and its “rules” thus contribute to the difficulty of distinguishing real linguistic consequences of people contact, such as reflected in recuerdo conociendo in (4c), from the much more frequent case of expressions reflecting cultural adaptation without language change. In short, another problem with terms like English and Spanish in discussions of encounter settings is that they encourage a rule-based theoretical discourse that is not sufficiently precise to distinguish the units of the grammar of impacted bilinguals from the conceptualizations that are driving and explaining its use.

Before ending, a possible criticism is worth considering from the point of view of usage-based grammar (Bybee 2010). As frequent combinations of individual meaningful units, the items in all the (c) utterances above could be analyzed as being cognitively enregistered as chunked wholes in NYC idiolects. As such, they would enjoy easier access and retrieval than in Latin America. Thus, something of a structural nature would in a way be different among Latinos in NYC after all. Among them, for example, último nombre, would be a newly entrenched chunk based on a similarly deeply enregistered last name chunk. However, usage-based theory does not deny the compositionality of frequent combinations, does not deny, in my terms, that they are still made up of the individual meaningful units that constitute the grammar. The deployment by bilinguals of existing combinations like último nombre to engage in new messaging in the service of new conceptualizations suggests that last name, for all its high frequency, has maintained a strong compositionality. For, it is precisely the robust compositionality of last + name that allowed the recognition by bilinguals that led them to start referring to the surname as an último + nombre. Moreover, an analysis that would view último nombre as a single últimonombre unit would find it difficult to account for its rise among Latinos in NYC because it could not appeal to the new conceptualization with which it has been explained here. The chunk analysis would have to think of últimonombre as simply a loan. It could describe the new usage, but could not explain it.

A final point about named languages. Through the confounding of grammatical, cultural, and communicative elements, and the resulting expansion of what constitutes contact-induced linguistic change, the named language tends to disadvantage impacted bilinguals, especially minoritized ones, particularly in educational settings. The named language, as an object endowed with its own ontology and governed by rules, has facilitated the return in several branches of linguistics of what is, in its fundamentals, a prescriptive approach. As part of it, in the US, the notion of a named language has made possible the removal of the title of native speaker from bilingual students, and its replacement by the term heritage speaker. Heritage-language learning now names a field that addresses the need to improve and correct student grammars in ways that are not essentially different from what is done with second language learners.

More generally, the speaker-centered approach adopted here recognizes that the named language has encouraged many scholars, in both education and linguistics, to perceive the speech of impacted bilinguals in marked terms; to perceive it, that is, as an alternative, defective version of something that is somehow less marked and more authentic. This something is usually the named language of the corresponding precursor dimensions. It is thus the language name shared by precursor and encounter that promotes this essentially prescriptive view, taking the focus off the speaker and placing it on the named language abstraction, which in this way becomes susceptible to having different versions that can be adjudged of greater or lesser legitimacy.

In the case of NYC and the US more generally, what many psychologists and educators armed with the notion of a named language have seen as incorrect or incomplete Spanish, the speaker-centered approach sees as precisely the opposite, a tribute to adaptability and creative use of idiolectal linguistic resources. In the speaker-centered approach that dispenses with the notion of a language called Spanish, we saw that nothing of grammatical interest had taken place in our examples (1c) through (3c). And that what we saw in (4c) was simply a case of differences between speakers living in different societies that the language-centered approach should better take care to describe as a dialectal difference rather than a mistake. The problem of the named language, then, is that the creative use of unchanged idiolectal features to express the new adaptive messaging of the encounter is incorrectly seen as a systemic linguistic change, deserving of at best reluctant indulgence and at worst, especially in school settings, negative evaluation and reproach. Dispensing with the language, and with language contact, allows for a better understanding of the linguistic consequences of people contact. And it allows as well for a more science-based and also more humane treatment of impacted bilinguals.