Domínguez, Hicks, and Slabakova (Reference Domínguez, Hicks and Slabakova2019) argue that the term incomplete is a scientifically valid one to describe the grammars of adult or adolescent heritage speakers. They correctly distinguish between the issue of the scientific validity of the term and the deleterious consequences that incompleteness may have when misunderstood by the public, or by scholars or practitioners in language-related fields. The following remarks, then, address only the theoretical validity of the concept, and leave for another day the matter of communication with others. My critique encompasses incomplete acquisition whether the term used is incomplete or the more recent divergent, which does not repair the conceptual flaws that undermine both terms.
I have offered a critique of incomplete acquisition (Otheguy, Reference Otheguy, Tortora, den Dikken, Montoya and O’Neill2016) in foundational terms that are not addressed by Domínguez et al. But ignoring foundational problems puts at risk the theoretical coherence of any enterprise, including the one at issue. Domínguez et al. articulate their position in terms of the native speaker, the heritage speaker, the target grammar, and the baseline control group. Domínguez et al. ask us to accept these concepts without motivating them, and yet it is precisely the dubious validity of these entirely a priori notions that invalidates the thinking Domínguez et al. want to defend in the service of incompleteness.
THE SO-CALLED HERITAGE SPEAKER
The seemingly natural terms native speaker and nonnative speaker are known to mark a sociopolitical distinction much more than a scientific one (Bonfiglio, Reference Bonfiglio2010). Those who accept these terms have never spelled out exactly what they mean, nor how the distinction between them is made. But even if we were to grant Domínguez et al. the terms native and nonnative, Domínguez et al. would still be left with the problem that no one has justified peeling off heritage speakers from the native category and setting them up as a distinct group. These speakers spent their first four or five years of life, just like everyone else, exposed to a language for which they, like everybody else, should be presumed to have formed a grammar that at some point became whatever all the other grammars became (we should be happy to say they became complete if anyone offered to make sense of that term). The unjustified assertion that when these adults were children they were heritage, not native, speakers, and that this status carries over to today appears to be an arbitrary terminological move that lacks force in supporting claims that their grammars are incomplete. Linguists don’t think of the grammars of any other group of native speakers as incomplete, and removing the label native from them does not constitute an argument for making an exception with these ones. To be convincing, researchers in this area, Domínguez et al. included, would need to abandon the initial labeling of heritage, and scientifically build an empirically supported and theoretically sound argument for why we should think of these native speakers differently from the way we think of others. As they try this, it will be good to keep in mind that nothing says that monolinguals born in the areas where a language is the official or majority one are its only native speakers.
Domínguez et al. discuss native speakers of the so-called heritage type together with speakers engaged in the process of second language acquisition (SLA), as if we could all agree that these two groups form a natural class, which they do not (did L2 learners of Russian spend their first five years of life listening to, and speaking with, Russian parents in Brighton Beach, Brooklyn?). Because it seems natural to say that at many stages L2 learners lack a complete grammar, bringing together L2 speakers with so-called heritage speakers smuggles into the discussion an illegitimate argument for the concept of incomplete acquisition, which may be defensible in SLA but unsustainable for the native speakers called heritage speakers.
THE TARGET GRAMMAR
Domínguez et al. appear comfortable with the notion of a target grammar that native speakers dubbed heritage speakers have failed to reach. But there is no reason to believe that child language acquisition involves a conceptual trip that reaches (or not) a target grammar. As far as anyone can tell, adult grammars are not objects transferred from parents to children, and neither are they targets that children reach or fail to reach. Rather, children build grammars anew each generation. There is justification for the notion of input, but not for that of a target.
The unjustified notion of a target, then, is another illegitimate argument purporting to support incomplete acquisition. But let us suppose that we granted Domínguez et al. their target. The question would remain: What target? We are confidently told by Domínguez et al., that “Spanish-English bilinguals in the US have been found to exhibit a reduced tense-aspect-mood verbal system” (p. 242). Reduced in comparison to what? In English, did the speakers that, in recent memory, abandoned the distinctions between who and whom and less and fewer have a reduced system? And did they, in consequence of this reduction, need to be described as having an incomplete grammar? The answer is no; the English grammars with and without the distinctions are different, but the ones without them are not divergent from anything or incomplete with respect to anything. The speakers that developed them simply did not include these items in their newly built, conceptually independent linguistic edifice. There was no target for their grammars to diverge from, nothing with respect to which their grammars are incomplete.
One cannot help but think that Domínguez et al. believe that there are out there linguistic objects with names like “English,” “Russian,” or “Spanish,” the characteristics of whose grammars are well known and against which speakers do or do not measure up. But there are no such linguistic objects. Grammars reside in the minds of individuals, and even if we were to grant the faulty thinking in terms of targets, we have been given no evidence that, say, speakers of Spanish born in Los Angeles were ever aiming (in the subconscious way that humans build their grammars as children) for anything that merits being called the grammar of Spanish.
There are, then, two problems. One, that it makes no sense to speak of target grammars. Two, that even if it did, there is no one thing called English, Russian, Spanish, and so forth whose grammars the so-called heritage speakers aimed at and missed. These individuals listened, for years, to their individual parents (the details of whose individual grammars are not always presented in the literature) and constructed a grammar that, yes, was different from that of their parents. But in doing this, and if the concept of the native speaker is to be licensed, they behaved like the typical native speakers that they are.
THE CONTROL GROUP
Domínguez et al. justify the misconceived comparisons found in the study of incomplete acquisition with the argument that, as in any human science, the analysis of one group requires comparing it to a control group. But the scientist needs to think through whether in the particular type of research involved it is coherent to think of control groups and, if the answer is yes, needs to make sure that the chosen controls make sense. In the case of language acquisition, the coherence of working in terms of controls has not been established. And the particular ones that are usually chosen, as far as I’ve been able to tell, are speakers who cannot be reasonably compared with so-called heritage speakers. When the grammars of working class, bilingual native speakers from the Bronx are compared to those of upper-class, monolingual native speakers from Guadalajara, Mexico who are studying in the United States, we’re facing a strong candidate for the apples and oranges award.
INCOMPLETENESS ALL AROUND
Domínguez et al. buttress their claim for incomplete acquisition by saying that the term is used, without incurring any criticism, when applied to small children or to L2 speakers. If we grant that their acquisition is often incomplete, why argue against the concept for the speakers at issue here? The answer is that here we are talking about adolescents and adults, not children. It may make sense to think of a 3-year-old as having an incomplete grammar in the sense that she’s clearly still constructing it. But it does not make sense to think in this way about the same person when she’s 13 or 30. By then, we assume the process is well along, and that her grammar is, in its fundamentals, the same as that of people who grew up in essentially similar circumstances (i.e., with one or two parents with no speech or hearing disabilities who were speakers of the language). When the linguist looks at the outcome, this bilingual 30-year-old’s grammar of Arabic, or Russian, or Spanish, or what have you, may very well look different from that of her parents. But a different grammar, as we have seen, is what we always expect.
The notion of incomplete or divergent acquisition requires a category of speaker, the heritage speaker, that has not been motivated; a target grammar that does not exist; a process of aiming at this grammar that no child engages in; a notion of control group whose general coherence has not been worked out, and that does not seem sensible in the particulars; and an equation with L2 speakers that is false. On such a cracked foundation the concept of incomplete acquisition seems unconvincing and its defense by Domínguez et al. unjustified.
You have
Access
Domínguez, Hicks, and Slabakova (Reference Domínguez, Hicks and Slabakova2019) argue that the term incomplete is a scientifically valid one to describe the grammars of adult or adolescent heritage speakers. They correctly distinguish between the issue of the scientific validity of the term and the deleterious consequences that incompleteness may have when misunderstood by the public, or by scholars or practitioners in language-related fields. The following remarks, then, address only the theoretical validity of the concept, and leave for another day the matter of communication with others. My critique encompasses incomplete acquisition whether the term used is incomplete or the more recent divergent, which does not repair the conceptual flaws that undermine both terms.
I have offered a critique of incomplete acquisition (Otheguy, Reference Otheguy, Tortora, den Dikken, Montoya and O’Neill2016) in foundational terms that are not addressed by Domínguez et al. But ignoring foundational problems puts at risk the theoretical coherence of any enterprise, including the one at issue. Domínguez et al. articulate their position in terms of the native speaker, the heritage speaker, the target grammar, and the baseline control group. Domínguez et al. ask us to accept these concepts without motivating them, and yet it is precisely the dubious validity of these entirely a priori notions that invalidates the thinking Domínguez et al. want to defend in the service of incompleteness.
THE SO-CALLED HERITAGE SPEAKER
The seemingly natural terms native speaker and nonnative speaker are known to mark a sociopolitical distinction much more than a scientific one (Bonfiglio, Reference Bonfiglio2010). Those who accept these terms have never spelled out exactly what they mean, nor how the distinction between them is made. But even if we were to grant Domínguez et al. the terms native and nonnative, Domínguez et al. would still be left with the problem that no one has justified peeling off heritage speakers from the native category and setting them up as a distinct group. These speakers spent their first four or five years of life, just like everyone else, exposed to a language for which they, like everybody else, should be presumed to have formed a grammar that at some point became whatever all the other grammars became (we should be happy to say they became complete if anyone offered to make sense of that term). The unjustified assertion that when these adults were children they were heritage, not native, speakers, and that this status carries over to today appears to be an arbitrary terminological move that lacks force in supporting claims that their grammars are incomplete. Linguists don’t think of the grammars of any other group of native speakers as incomplete, and removing the label native from them does not constitute an argument for making an exception with these ones. To be convincing, researchers in this area, Domínguez et al. included, would need to abandon the initial labeling of heritage, and scientifically build an empirically supported and theoretically sound argument for why we should think of these native speakers differently from the way we think of others. As they try this, it will be good to keep in mind that nothing says that monolinguals born in the areas where a language is the official or majority one are its only native speakers.
Domínguez et al. discuss native speakers of the so-called heritage type together with speakers engaged in the process of second language acquisition (SLA), as if we could all agree that these two groups form a natural class, which they do not (did L2 learners of Russian spend their first five years of life listening to, and speaking with, Russian parents in Brighton Beach, Brooklyn?). Because it seems natural to say that at many stages L2 learners lack a complete grammar, bringing together L2 speakers with so-called heritage speakers smuggles into the discussion an illegitimate argument for the concept of incomplete acquisition, which may be defensible in SLA but unsustainable for the native speakers called heritage speakers.
THE TARGET GRAMMAR
Domínguez et al. appear comfortable with the notion of a target grammar that native speakers dubbed heritage speakers have failed to reach. But there is no reason to believe that child language acquisition involves a conceptual trip that reaches (or not) a target grammar. As far as anyone can tell, adult grammars are not objects transferred from parents to children, and neither are they targets that children reach or fail to reach. Rather, children build grammars anew each generation. There is justification for the notion of input, but not for that of a target.
The unjustified notion of a target, then, is another illegitimate argument purporting to support incomplete acquisition. But let us suppose that we granted Domínguez et al. their target. The question would remain: What target? We are confidently told by Domínguez et al., that “Spanish-English bilinguals in the US have been found to exhibit a reduced tense-aspect-mood verbal system” (p. 242). Reduced in comparison to what? In English, did the speakers that, in recent memory, abandoned the distinctions between who and whom and less and fewer have a reduced system? And did they, in consequence of this reduction, need to be described as having an incomplete grammar? The answer is no; the English grammars with and without the distinctions are different, but the ones without them are not divergent from anything or incomplete with respect to anything. The speakers that developed them simply did not include these items in their newly built, conceptually independent linguistic edifice. There was no target for their grammars to diverge from, nothing with respect to which their grammars are incomplete.
One cannot help but think that Domínguez et al. believe that there are out there linguistic objects with names like “English,” “Russian,” or “Spanish,” the characteristics of whose grammars are well known and against which speakers do or do not measure up. But there are no such linguistic objects. Grammars reside in the minds of individuals, and even if we were to grant the faulty thinking in terms of targets, we have been given no evidence that, say, speakers of Spanish born in Los Angeles were ever aiming (in the subconscious way that humans build their grammars as children) for anything that merits being called the grammar of Spanish.
There are, then, two problems. One, that it makes no sense to speak of target grammars. Two, that even if it did, there is no one thing called English, Russian, Spanish, and so forth whose grammars the so-called heritage speakers aimed at and missed. These individuals listened, for years, to their individual parents (the details of whose individual grammars are not always presented in the literature) and constructed a grammar that, yes, was different from that of their parents. But in doing this, and if the concept of the native speaker is to be licensed, they behaved like the typical native speakers that they are.
THE CONTROL GROUP
Domínguez et al. justify the misconceived comparisons found in the study of incomplete acquisition with the argument that, as in any human science, the analysis of one group requires comparing it to a control group. But the scientist needs to think through whether in the particular type of research involved it is coherent to think of control groups and, if the answer is yes, needs to make sure that the chosen controls make sense. In the case of language acquisition, the coherence of working in terms of controls has not been established. And the particular ones that are usually chosen, as far as I’ve been able to tell, are speakers who cannot be reasonably compared with so-called heritage speakers. When the grammars of working class, bilingual native speakers from the Bronx are compared to those of upper-class, monolingual native speakers from Guadalajara, Mexico who are studying in the United States, we’re facing a strong candidate for the apples and oranges award.
INCOMPLETENESS ALL AROUND
Domínguez et al. buttress their claim for incomplete acquisition by saying that the term is used, without incurring any criticism, when applied to small children or to L2 speakers. If we grant that their acquisition is often incomplete, why argue against the concept for the speakers at issue here? The answer is that here we are talking about adolescents and adults, not children. It may make sense to think of a 3-year-old as having an incomplete grammar in the sense that she’s clearly still constructing it. But it does not make sense to think in this way about the same person when she’s 13 or 30. By then, we assume the process is well along, and that her grammar is, in its fundamentals, the same as that of people who grew up in essentially similar circumstances (i.e., with one or two parents with no speech or hearing disabilities who were speakers of the language). When the linguist looks at the outcome, this bilingual 30-year-old’s grammar of Arabic, or Russian, or Spanish, or what have you, may very well look different from that of her parents. But a different grammar, as we have seen, is what we always expect.
The notion of incomplete or divergent acquisition requires a category of speaker, the heritage speaker, that has not been motivated; a target grammar that does not exist; a process of aiming at this grammar that no child engages in; a notion of control group whose general coherence has not been worked out, and that does not seem sensible in the particulars; and an equation with L2 speakers that is false. On such a cracked foundation the concept of incomplete acquisition seems unconvincing and its defense by Domínguez et al. unjustified.