Heritage speakers are a fascinating group of bilinguals with a unique profile. Living abroad as immigrants of the second generation, they speak the language of their own speech community (the heritage language) at home, and the societally dominant language in most other domains. What exactly they know about their heritage language continues to fascinate the research community as well as teachers and other practitioners working with this group. The different contributions cover a large variety of studies into heritage languages spoken in Europe and North America (including Chinese, Norwegian, Russian, Spanish and Turkish). The volume makes a key contribution to the description and explanation of variability in the outcomes of heritage language acquisition, taking into account a wide range... of factors which impact on language acquisition. As comparisons are frequently made with monolinguals and foreign language learners, the volume is also highly relevant for researchers working in monolingual language acquisition and foreign language learning and teaching.;;By comparing two generations of speakers in China and the Netherlands, we investigated whether Wenzhounese Chinese as spoken by heritage speakers in the Netherlands might be subject to change due to its contact with the Dutch language. To this end, we considered how nouns referring to already mentioned referents were encoded, hypothesizing an increase in overtly marked already mentioned referents in the speech of second generation speakers in the Netherlands. Also, shifts in the use of classifier and demonstrative constructions were investigated as carriers of definite meanings. The data showed clear generation effects, especially in the use of classifier and demonstrative constructions, but no location effects. Although the interpretation of the data is compromised by the fact that Wenzhounese as spoken China is under threat, the study suggests that innovations found in heritage speakers may not necessarily be related to the heritage scenario, but can also reflect more general contact effects or internally motivated changes.;;This study investigates the spoken performance of UK based adult heritage speakers (HSs) and first generation immigrants (LBs) of Turkish descent regarding the distribution of evidentiality encoded in the past tense system of Turkish. Additionally, we trace the effect of input quality and quantity on the development of the heritage language (HL). The analyses suggest that the ability to differentiate between different evidential structures is unstable in HL grammars. There are no signs of qualitatively modified input available to the HSs through the LBs. This performance is instead linked to the amount of input, which has been found to be insufficient to compensate for the detrimental effects of early bilingualism on the HL. Findings are discussed within the premises of available approaches to heritage language bilingualism.;;Previous studies (e.g., van Osch & Sleeman 2016; Perez-Cortes 2016) have found that heritage speakers (HSs) of Spanish produce less lexically-selected subjunctive mood morphology than Spanish-dominant speakers. It remains unclear, however, whether the HSs’ tendency to produce less subjunctive mood than Spanish-dominant speakers is attributable to representational differences (e.g., Montrul 2002, 2008), input quality differences (e.g., Pires & Rothman 2009; Pascual y Cabo & Rothman 2012), or as yet unidentified factors. The present study addresses this question by testing the effect of lexical frequency on advanced proficiency HSs’ productive (Experiment 1) and receptive (Experiment 2) knowledge of lexically-selected subjunctive mood in Spanish. Results of Experiments 1 and 2 indicate that advanced proficiency HSs are both (a) highly accurate with subjunctive mood and yet also (b) significantly less accurate with lower frequency verbs. Given these findings, as well as the categorical subjunctive production of the Spanish-dominant bilingual control group, it is argued that HSs’ differences from dominant speakers may be (partially) attributable to gaps in lexical, rather than morphosyntactic knowledge.;;This study investigates two word order phenomena in Norwegian heritage language spoken in the US, subject shift (SS) and object shift (OS). SS and OS occur in syntactic environments where (pronominal) subjects and objects may either precede or follow negation. This paper explores to what extent these two phenomena in Heritage Norwegian are affected by the factors frequency and structural similarity/difference. As subjects are frequently shifted, while objects are not, SS is expected to be robust and OS vulnerable. There is generally no structural overlap between English and Norwegian in these cases; thus, cross-linguistic similarity or difference should not play a role, except in one context: questions with auxiliaries or be , in which the two languages allow both orders ( is he not /is n’t he ), but have opposite preferences. The results show that OS is somewhat vulnerable, but SS is also affected, in that both proficient and less proficient speakers seem to overuse the word order preferred in English. We thus speculate that all heritage speakers may be affected by cross-linguistic influence in situations with complete structural overlap.;;This paper examines the effect of language contact on the knowledge of Spanish gender assignment and agreement in adult second language learners and simultaneous bilinguals (heritage speakers of Spanish), all residing in the Geneva area of New York State. The data comes from 27 English-speaking learners of Spanish and 27 bilingual speakers, who completed a grammatical judgment task (GJT) and an oral elicitation production task (OPT). In particular, the paper investigates whether the successful acquisition of gender is dependent on the extent of exposure to the target language. The results show successful acquisition of gender assignment and agreement in all groups. In addition, the findings indicate that the extent to which Spanish is used seems to affect the gender accuracy of Spanish L2 learners as opposed to heritage speakers, who perform at ceiling in the GJT and oral task regardless of the frequency of heritage language ( hl ) use.;;How does vocabulary in the heritage language develop? Does the social environment of the community have an influence? This chapter presents empirical results regarding the development of expressive and receptive vocabulary in the heritage language and analyzes the effects of amount of exposure, use, socio-economic status, dominance, and community on the acquisition of vocabulary in the heritage language. A Russian-German and a Turkish-German speaking sample are compared: 211 children at the age of 6–10 years were tested with a standardized picture naming task in a cross-sectional design. The results show a good receptive mastery and a limited expressive command of vocabulary with large individual differences, and only a slight development in the timespan of four years. Between the communities we find systematic variation, which we attribute to social and pragmatic differences with moderator and mediator effects. Possible limitations of the results are discussed with respect to cross-linguistic test effects.;;The present study explores the effects of literacy support on the linguistic and cognitive skills among (non-)heritage speakers. We investigate 70 children speaking Albanian (L1) and Greek (L2) who are divided into three groups according to whether they receive literacy support in their L1, L2 or in both languages. For testing of the proficiency we applied an expressive vocabulary task, two working memory tasks, a non-verbal intelligence task and a sentence repetition task (SRT). The findings suggest that good levels of biliteracy established through bilingual education positively influence the child’s linguistic and cognitive performance. Furthermore, the lack of working memory effects on the children’s SRT performance emphasizes the importance of biliteracy development and its educational support, which together with vocabulary knowledge contribute to bilingual development.;;This study investigates semantics-morphosyntax and pragmatics-morphosyntax interfaces in separate experiments, yet examining the same structure, namely the optional verb number marking in Turkish. We tested a group of bilingual heritage speakers of Turkish, whose dominant language is German. Optional use of the overt verb number marking in Turkish interacts with semantic and pragmatic properties of the plural subject of the sentence. The interaction of optional verb number marking with these properties is tested separately in two different experiments, using the Magnitude Estimation technique. The results showed that the bilingual speakers treat both interface types differently from the monolingual speakers. More precisely, the bilingual speakers make finer distinctions regarding the semantic and pragmatic notions that were put into test. This sensitivity results in a semantically and pragmatically constrained pattern, which is in line with both language-specific descriptions and cross-linguistic tendencies. This outcome is taken to suggest that the nonconvergence in the bilingual data stems from a high sensitivity to the semantic and pragmatic properties that constrain the use of the morphosyntactic structure under investigation. The research results are further evaluated from a processing based linguistic framework, namely Modular Online Growth and Use of Language (MOGUL).;;The paper investigates pitch level and span in a group of German L1-English L2 late bilinguals in comparison to two monolingual control groups. The late bilinguals had moved to Vancouver, Canada in adulthood, and had been living in Vancouver for an average of 40 years. The results indicate that the bilingual males increased their pitch in both English and German, and widened their pitch span, therefore indexing non-aggressive, friendly behaviour, but deviating from both monolingual pitch norms. Thus, the results offer evidence that pitch changes are at least in part dependent on the social and political environment in which they are embedded, as a low pitch level is associated with dominance and aggression which would boost the negative image of the Vancouver German community due to their ethnic origin after WWII.;;This study concerns whether extended exposure to a second language would lead to first language (L1) attrition, i.e. changes of L1 linguistic behavior/knowledge. An acceptability judgement task, which examined the perceptive knowledge of perfective and durative aspect marking in Mandarin Chinese, was employed, and the performance of 14 Mandarin-English bilinguals in the UK was examined. The results did not suggest that the bilinguals showed L1 attrition in perceiving perfective/durative aspect marking. The paper also discusses how research on heritage language acquisition could benefit from L1 attrition research.