The commentaries identify two main reasons for rejecting the term “incomplete acquisition” in the case of heritage language acquisition.
IS “INCOMPLETE ACQUISITION” AN APPROPRIATE CHARACTERIZATION?
The first is that incomplete acquisition fails to accurately characterize the process or outcome of heritage language acquisition. On this distinction, we are happy to accept Montrul and Silva-Corvalán’s clarification that incomplete acquisition is intended to characterize an outcome, rather than a process. It is our firm view that the process of acquisition of a heritage language is not substantively different from any other process of language acquisition. For example, in monolingual L1 acquisition the grammar is incomplete at all stages of acquisition until the process finishes. This is trivial in this particular context because under consistent input conditions, the expected outcome is a complete grammar. In contrast, heritage language acquisition is characterized by exposure to input that is inconsistent or insufficient before an end-state L1 grammar is attained. For this reason, the completeness of the acquisition process in this context is, ultimately, not guaranteed. These uncontroversial similarities and differences between monolingual and heritage child learners that we discuss in our commentary have been ignored by Otheguy when he claims, rather puzzlingly, that we should abandon the label “heritage” altogether.
We can assume that incomplete acquisition, generally, refers to a stage in a general process of acquiring a target grammar, that is, the grammar that generates the data in the individual’s linguistic environment. For heritage speakers, that stage may turn out to be the stable/final stage (which Montrul & Silva-Corvalán refer to as an outcome), or it may not. As we explain, incomplete acquisition is taken to describe just one of the potential outcomes of heritage language acquisition. We have pointed out, as others before us have done, that aspects of heritage language grammars are not necessarily always incompletely acquired. In this respect, we agree with Zyzik that innovations and reanalyses can indeed be found in the grammars of these speakers.
Where we disagree with Montrul and Silva-Corvalán is in their reluctance to refer to such heritage grammars as “incomplete” because, they argue, it is often only an aspect of a grammar that is incomplete. However, we cannot see how this argument holds: if an aspect of a grammar has not had sufficient input to be fully acquired, then the grammar cannot be said to be complete, or fully acquired. In any case, this is an unresolved issue that does not directly impinge on our main arguments for maintaining the term incomplete acquisition.
If incomplete acquisition is to be rejected on the grounds that it does not describe the potential outcome of heritage language acquisition, then there are two possible justifications. The first is that the characterization of the development of the grammar and its relation to input is wrong: this is essentially not about the term per se, but about the theory. This is an open question for the field to resolve; we must reiterate that our critical commentary was never intended to support incomplete acquisition over other possible explanations for heritage language acquisition. As long as the theoretical explanation for certain heritage language grammars remains on the table, then the term remains.
We would also like to emphasize that by accepting incomplete acquisition, no one is supporting the view that heritage speakers' grammars are not natural grammars in their own right. It seems that the interpretation of incomplete acquisition as either a failure to achieve the status of a proper grammar or as only applicable to “rogue grammars” that fail to obey linguistic principles is one that is unique to Bayram, Kupisch, Pascual y Cabo, and Rothman. For everyone else in our field, us included, incomplete acquisition is taken to be compatible with a characterization of functional, coherent, UG-licensed grammars (UG = Universal Grammar). To be clear, we categorically distance ourselves from views of heritage grammars as “rogue,” or not sanctioned by UG (see Schroeder, Reference Schroeder2017; Schwartz & Sprouse, Reference Schwartz, Sprouse and Archibald2000; Thomas, Reference Thomas and Eubank1991, Reference Thomas1995; White, Reference White2003).
The other possible justification for rejecting incomplete acquisition as a description of a potential outcome of heritage language acquisition would be if some alternative term existed that could characterize this outcome more accurately. However, we know of no other term that is more descriptively appropriate for this account of (certain) heritage language grammars and for this reason the term should remain. We have cautioned against adopting terms such as “differential acquisition” and “divergent attainment,” which are less precise and do not offer an appropriate level of scientific accuracy: a grammar that is different/divergent may or may not be incomplete. We agree with Putnam that “unstable” and “unconsolidated” grammars could be better replacements than “divergent”/“different” as they are not as nebulous. He seems to suggest, however, that incomplete acquisition should be replaced by these terms because what we mostly find is heritage grammars being reduced or underspecified. That is, he understands incomplete acquisition to be a marginal aspect of heritage language acquisition that can be easily subsumed under these new terms. It seems to us, however, that this is another argument against the validity of incomplete acquisition to explain heritage language acquisition. If grammars can indeed be incomplete, as Putnam admits, then the need to maintain the term is justified.
IS “INCOMPLETE ACQUISITION” OFFENSIVE?
The second main reason for rejecting the term, and what prompted our critical commentary in the first place, is that “incomplete” as a term is potentially derogatory and can be easily misinterpreted, to the point of causing offense. We agree with Bayram et al. and Zyzik that misunderstanding by those outside our discipline is something we have a duty to consider. However, misrepresentation of theoretical positions from within our own discipline surely deserves attention as well. Otheguy, for instance, goes so far as to claim that target grammars do not exist because a grammar is not a linguistic object in his view, but something that exists in the mind of speakers (in his view, heritage speakers never aim to learn a grammar). For Otheguy, the suggestion of a target (native) grammar can be used to set heritage speakers up to fail. He claims that grammars in the formal sense are objects “against which speakers do or do not measure up” (Otheguy Reference Otheguy, Domínguez, Hicks and Slabakova2019, p. 267). We deliberately chose to frame the discussion of incomplete acquisition clearly within generative linguistic theory and made it clear that issues concerning attitudes and beliefs about language (e.g., whether a bilingual speaker measures up as a native speaker) are tangential to our arguments. Furthermore, Otheguy seems to have ignored that we are concerned only with describing the unconscious grammatical knowledge of speakers and would never aim to evaluate their ability to function as (bilingual) people.
The main principle we defend is that, in striving for scientifically accurate description, researchers should have the freedom to apply terms that are the most descriptively appropriate of the available options. Given this principle, very careful consideration must be given to any proposal to reject specific terms. We chose to focus on incomplete acquisition as this term has in effect already been rejected by some in the field on the basis that it may be misinterpreted, to the point of causing offense. Our aim is to establish whether this is justifiable, and thereby open up a debate that extends beyond the remit of incomplete acquisition to other terms in our field, such as “attrition” for instance, which could also be interpreted negatively. In our view, the basis of the debate is as follows: if incomplete acquisition is a term that better explains the nature of heritage language acquisition compared to competitor terms, then it is difficult to defend that it should be abandoned, in particular if the alternative terms do not offer the same level of accuracy. Furthermore, the proposed alternatives essentially suffer from the same potential problems of misconstrual: Is being perceived as “divergent,” “different,” or “unstable” any better than being perceived as “incomplete”? Under the approach we advocate in the critical commentary, an intermediate position in the terminology debate encourages researchers to consider whether any potential negative impacts of using a term could be avoided by applying an alternative term that is just as descriptively accurate. Those who defend and use incomplete acquisition need to use the term in a precise way, so as not to perpetuate some of the misconceptions that already exist. The message has to be loud and clear: heritage speakers are not deficient or incomplete learners.
Criticisms toward the main tenets of generative language acquisition research are abundant. As Otheguy reminds us, there is much about what we assume that appears to be provocative, at least to those outside our field. In our piece—quite controversially, it seems—we maintain that incomplete acquisition should not be abandoned as it is the term that better describes an important aspect of heritage language acquisition. It cannot yet be replaced by alternative terms that offer the same level of accuracy.
You have
Access
The commentaries identify two main reasons for rejecting the term “incomplete acquisition” in the case of heritage language acquisition.
IS “INCOMPLETE ACQUISITION” AN APPROPRIATE CHARACTERIZATION?
The first is that incomplete acquisition fails to accurately characterize the process or outcome of heritage language acquisition. On this distinction, we are happy to accept Montrul and Silva-Corvalán’s clarification that incomplete acquisition is intended to characterize an outcome, rather than a process. It is our firm view that the process of acquisition of a heritage language is not substantively different from any other process of language acquisition. For example, in monolingual L1 acquisition the grammar is incomplete at all stages of acquisition until the process finishes. This is trivial in this particular context because under consistent input conditions, the expected outcome is a complete grammar. In contrast, heritage language acquisition is characterized by exposure to input that is inconsistent or insufficient before an end-state L1 grammar is attained. For this reason, the completeness of the acquisition process in this context is, ultimately, not guaranteed. These uncontroversial similarities and differences between monolingual and heritage child learners that we discuss in our commentary have been ignored by Otheguy when he claims, rather puzzlingly, that we should abandon the label “heritage” altogether.
We can assume that incomplete acquisition, generally, refers to a stage in a general process of acquiring a target grammar, that is, the grammar that generates the data in the individual’s linguistic environment. For heritage speakers, that stage may turn out to be the stable/final stage (which Montrul & Silva-Corvalán refer to as an outcome), or it may not. As we explain, incomplete acquisition is taken to describe just one of the potential outcomes of heritage language acquisition. We have pointed out, as others before us have done, that aspects of heritage language grammars are not necessarily always incompletely acquired. In this respect, we agree with Zyzik that innovations and reanalyses can indeed be found in the grammars of these speakers.
Where we disagree with Montrul and Silva-Corvalán is in their reluctance to refer to such heritage grammars as “incomplete” because, they argue, it is often only an aspect of a grammar that is incomplete. However, we cannot see how this argument holds: if an aspect of a grammar has not had sufficient input to be fully acquired, then the grammar cannot be said to be complete, or fully acquired. In any case, this is an unresolved issue that does not directly impinge on our main arguments for maintaining the term incomplete acquisition.
If incomplete acquisition is to be rejected on the grounds that it does not describe the potential outcome of heritage language acquisition, then there are two possible justifications. The first is that the characterization of the development of the grammar and its relation to input is wrong: this is essentially not about the term per se, but about the theory. This is an open question for the field to resolve; we must reiterate that our critical commentary was never intended to support incomplete acquisition over other possible explanations for heritage language acquisition. As long as the theoretical explanation for certain heritage language grammars remains on the table, then the term remains.
We would also like to emphasize that by accepting incomplete acquisition, no one is supporting the view that heritage speakers' grammars are not natural grammars in their own right. It seems that the interpretation of incomplete acquisition as either a failure to achieve the status of a proper grammar or as only applicable to “rogue grammars” that fail to obey linguistic principles is one that is unique to Bayram, Kupisch, Pascual y Cabo, and Rothman. For everyone else in our field, us included, incomplete acquisition is taken to be compatible with a characterization of functional, coherent, UG-licensed grammars (UG = Universal Grammar). To be clear, we categorically distance ourselves from views of heritage grammars as “rogue,” or not sanctioned by UG (see Schroeder, Reference Schroeder2017; Schwartz & Sprouse, Reference Schwartz, Sprouse and Archibald2000; Thomas, Reference Thomas and Eubank1991, Reference Thomas1995; White, Reference White2003).
The other possible justification for rejecting incomplete acquisition as a description of a potential outcome of heritage language acquisition would be if some alternative term existed that could characterize this outcome more accurately. However, we know of no other term that is more descriptively appropriate for this account of (certain) heritage language grammars and for this reason the term should remain. We have cautioned against adopting terms such as “differential acquisition” and “divergent attainment,” which are less precise and do not offer an appropriate level of scientific accuracy: a grammar that is different/divergent may or may not be incomplete. We agree with Putnam that “unstable” and “unconsolidated” grammars could be better replacements than “divergent”/“different” as they are not as nebulous. He seems to suggest, however, that incomplete acquisition should be replaced by these terms because what we mostly find is heritage grammars being reduced or underspecified. That is, he understands incomplete acquisition to be a marginal aspect of heritage language acquisition that can be easily subsumed under these new terms. It seems to us, however, that this is another argument against the validity of incomplete acquisition to explain heritage language acquisition. If grammars can indeed be incomplete, as Putnam admits, then the need to maintain the term is justified.
IS “INCOMPLETE ACQUISITION” OFFENSIVE?
The second main reason for rejecting the term, and what prompted our critical commentary in the first place, is that “incomplete” as a term is potentially derogatory and can be easily misinterpreted, to the point of causing offense. We agree with Bayram et al. and Zyzik that misunderstanding by those outside our discipline is something we have a duty to consider. However, misrepresentation of theoretical positions from within our own discipline surely deserves attention as well. Otheguy, for instance, goes so far as to claim that target grammars do not exist because a grammar is not a linguistic object in his view, but something that exists in the mind of speakers (in his view, heritage speakers never aim to learn a grammar). For Otheguy, the suggestion of a target (native) grammar can be used to set heritage speakers up to fail. He claims that grammars in the formal sense are objects “against which speakers do or do not measure up” (Otheguy Reference Otheguy, Domínguez, Hicks and Slabakova2019, p. 267). We deliberately chose to frame the discussion of incomplete acquisition clearly within generative linguistic theory and made it clear that issues concerning attitudes and beliefs about language (e.g., whether a bilingual speaker measures up as a native speaker) are tangential to our arguments. Furthermore, Otheguy seems to have ignored that we are concerned only with describing the unconscious grammatical knowledge of speakers and would never aim to evaluate their ability to function as (bilingual) people.
The main principle we defend is that, in striving for scientifically accurate description, researchers should have the freedom to apply terms that are the most descriptively appropriate of the available options. Given this principle, very careful consideration must be given to any proposal to reject specific terms. We chose to focus on incomplete acquisition as this term has in effect already been rejected by some in the field on the basis that it may be misinterpreted, to the point of causing offense. Our aim is to establish whether this is justifiable, and thereby open up a debate that extends beyond the remit of incomplete acquisition to other terms in our field, such as “attrition” for instance, which could also be interpreted negatively. In our view, the basis of the debate is as follows: if incomplete acquisition is a term that better explains the nature of heritage language acquisition compared to competitor terms, then it is difficult to defend that it should be abandoned, in particular if the alternative terms do not offer the same level of accuracy. Furthermore, the proposed alternatives essentially suffer from the same potential problems of misconstrual: Is being perceived as “divergent,” “different,” or “unstable” any better than being perceived as “incomplete”? Under the approach we advocate in the critical commentary, an intermediate position in the terminology debate encourages researchers to consider whether any potential negative impacts of using a term could be avoided by applying an alternative term that is just as descriptively accurate. Those who defend and use incomplete acquisition need to use the term in a precise way, so as not to perpetuate some of the misconceptions that already exist. The message has to be loud and clear: heritage speakers are not deficient or incomplete learners.
Criticisms toward the main tenets of generative language acquisition research are abundant. As Otheguy reminds us, there is much about what we assume that appears to be provocative, at least to those outside our field. In our piece—quite controversially, it seems—we maintain that incomplete acquisition should not be abandoned as it is the term that better describes an important aspect of heritage language acquisition. It cannot yet be replaced by alternative terms that offer the same level of accuracy.