Native speakers come in different varieties, and heritage language speakers are one of them. What sets aside monolingually raised native speakers from heritage language speakers is the fact that heritage language speakers grow up in a dual language environment. Because the heritage language is usually a minority language with limited use and access beyond the home, and in many instances no academic support during the school age period, it develops under reduced input and output conditions. These suboptimal input conditions profoundly affect the developing and endstate grammars of heritage speakers in several ways. As Polinsky and Scontras (Polinsky & Scontras, Reference Polinsky and Scontras2019) correctly point out, reduced quantity and different quality of input (combined with economy of online cognitive resources related to low proficiency in the language) trigger grammatical deviations in heritage language grammars – when compared to baseline full-fledged grammars spoken by monolingually raised native speakers in the homeland or in the diaspora. Much linguistic research over the last two decades has been documenting specific grammatical areas that are likely not to develop fully and are deemed vulnerable (such as nominal and verbal morphology or long distance dependencies, for example); alongside areas that do develop fully and are resilient to environmental factors in heritage languages in general (such as core syntax, phonology, some aspects of semantics). Stressing the universality of patterns across heritage languages in general, Polinsky and Scontras (Reference Polinsky and Scontras2019) make the generalization that there are three outcomes of grammatical deviations: avoidance of ambiguity, resistance to irregularity, and shrinking of structure.
We comment on why resistance to irregularity is so commonly seen in heritage language grammars by focusing on overgeneralization of regular morphological patterns. Heritage speakers have smaller vocabularies in the heritage language than the baseline comparison group, and many heritage speakers do not use the heritage language with the same frequency and in a wide variety of contexts as speakers of the baseline group. Vocabulary depends on memory, and the strength and speed of lexical access depends on frequency of forms in the input and of language use (Gollan, Montoya, Cera & Sandoval. Reference Gollan, Montoya, Cera and Sandoval2008). There is evidence that heritage speakers like monolingual child native speakers rely on default morphological patterns in producing verbs and inflected nouns (Gor & Cook, Reference Gor and Cook2010; Gor & Vdovina, Reference Gor and Vdovina2010). Mason (Reference Mason2019) presents similar results in two different experiments studying the processing of verbal morphology in heritage speakers of Spanish. In one experiment, Mason asked heritage speakers of Spanish (n = 24) to complete an untimed written questionnaire to produce, and provide acceptability judgments of, Spanish nonce verbs. Participants nearly always produced regular inflection, and rated regular inflection significantly more acceptable than irregular inflection. In another task, participants were prompted with visually-presented infinitive verbs and asked to respond verbally with an inflected form as soon as possible. Heritage speakers (n = 18) were more accurate with regular forms. Furthermore, accuracy with irregular forms showed a greater influence of the speaker's proficiency and familiarity of the infinitive, indicating that irregular inflection may be more influenced by vocabulary size whereas regular inflection can be applied efficiently even to poorly known words.
What are the theoretical implications of these results? Potential explanations for resistance to irregularity in heritage languages can be captured by two possible theoretical models. One is the Declarative Procedural Model (Ullman, Reference Ullman2001), and the other is the Tolerance Principle (Yang, Reference Yang2016). The Declarative Procedural Model posits that implicit grammatical rules are acquired in early childhood by use of the procedural memory system. This would allow heritage speakers to acquire grammatical rules, including the regular conjugations, just like monolingually-raised native speakers. However, irregularities would be acquired explicitly via the declarative memory system, which depends on language use and experience. While Ullman posits that the procedural memory system remains available into adulthood, heritage speakers tend to have less exposure to the heritage language throughout childhood and adolescence and hence have smaller vocabularies stored in the declarative system, leading to the acquisition, knowledge and use of fewer irregular lexical items.
The Tolerance Principle explains how children overgeneralize and retreat from overgeneralizations as their vocabulary knowledge increases over time, and this model also applies to heritage language grammars. Like child L1 learners, heritage speakers perform mathematical operations on the input to extract rules and exceptions. If the number of exceptions or irregularities is less than a certain number (a proportion of vocabulary size), the learner tolerates the exception and applies the general default regular rule.
In conclusion, resistance to irregularity arises from deficiencies in vocabulary knowledge but also reveals the operation of highly systematic rule-governed behavior in heritage language grammars, even in the absence of optimal input conditions.
You have
Access
Native speakers come in different varieties, and heritage language speakers are one of them. What sets aside monolingually raised native speakers from heritage language speakers is the fact that heritage language speakers grow up in a dual language environment. Because the heritage language is usually a minority language with limited use and access beyond the home, and in many instances no academic support during the school age period, it develops under reduced input and output conditions. These suboptimal input conditions profoundly affect the developing and endstate grammars of heritage speakers in several ways. As Polinsky and Scontras (Polinsky & Scontras, Reference Polinsky and Scontras2019) correctly point out, reduced quantity and different quality of input (combined with economy of online cognitive resources related to low proficiency in the language) trigger grammatical deviations in heritage language grammars – when compared to baseline full-fledged grammars spoken by monolingually raised native speakers in the homeland or in the diaspora. Much linguistic research over the last two decades has been documenting specific grammatical areas that are likely not to develop fully and are deemed vulnerable (such as nominal and verbal morphology or long distance dependencies, for example); alongside areas that do develop fully and are resilient to environmental factors in heritage languages in general (such as core syntax, phonology, some aspects of semantics). Stressing the universality of patterns across heritage languages in general, Polinsky and Scontras (Reference Polinsky and Scontras2019) make the generalization that there are three outcomes of grammatical deviations: avoidance of ambiguity, resistance to irregularity, and shrinking of structure.
We comment on why resistance to irregularity is so commonly seen in heritage language grammars by focusing on overgeneralization of regular morphological patterns. Heritage speakers have smaller vocabularies in the heritage language than the baseline comparison group, and many heritage speakers do not use the heritage language with the same frequency and in a wide variety of contexts as speakers of the baseline group. Vocabulary depends on memory, and the strength and speed of lexical access depends on frequency of forms in the input and of language use (Gollan, Montoya, Cera & Sandoval. Reference Gollan, Montoya, Cera and Sandoval2008). There is evidence that heritage speakers like monolingual child native speakers rely on default morphological patterns in producing verbs and inflected nouns (Gor & Cook, Reference Gor and Cook2010; Gor & Vdovina, Reference Gor and Vdovina2010). Mason (Reference Mason2019) presents similar results in two different experiments studying the processing of verbal morphology in heritage speakers of Spanish. In one experiment, Mason asked heritage speakers of Spanish (n = 24) to complete an untimed written questionnaire to produce, and provide acceptability judgments of, Spanish nonce verbs. Participants nearly always produced regular inflection, and rated regular inflection significantly more acceptable than irregular inflection. In another task, participants were prompted with visually-presented infinitive verbs and asked to respond verbally with an inflected form as soon as possible. Heritage speakers (n = 18) were more accurate with regular forms. Furthermore, accuracy with irregular forms showed a greater influence of the speaker's proficiency and familiarity of the infinitive, indicating that irregular inflection may be more influenced by vocabulary size whereas regular inflection can be applied efficiently even to poorly known words.
What are the theoretical implications of these results? Potential explanations for resistance to irregularity in heritage languages can be captured by two possible theoretical models. One is the Declarative Procedural Model (Ullman, Reference Ullman2001), and the other is the Tolerance Principle (Yang, Reference Yang2016). The Declarative Procedural Model posits that implicit grammatical rules are acquired in early childhood by use of the procedural memory system. This would allow heritage speakers to acquire grammatical rules, including the regular conjugations, just like monolingually-raised native speakers. However, irregularities would be acquired explicitly via the declarative memory system, which depends on language use and experience. While Ullman posits that the procedural memory system remains available into adulthood, heritage speakers tend to have less exposure to the heritage language throughout childhood and adolescence and hence have smaller vocabularies stored in the declarative system, leading to the acquisition, knowledge and use of fewer irregular lexical items.
The Tolerance Principle explains how children overgeneralize and retreat from overgeneralizations as their vocabulary knowledge increases over time, and this model also applies to heritage language grammars. Like child L1 learners, heritage speakers perform mathematical operations on the input to extract rules and exceptions. If the number of exceptions or irregularities is less than a certain number (a proportion of vocabulary size), the learner tolerates the exception and applies the general default regular rule.
In conclusion, resistance to irregularity arises from deficiencies in vocabulary knowledge but also reveals the operation of highly systematic rule-governed behavior in heritage language grammars, even in the absence of optimal input conditions.
Author ORCIDs
Silvina Montrul, 0000-0001-6011-5959