This study explores the implementation of critical thinking via metacognition in linguistics courses. It employs surveys to examine strategies used by students in two courses, Morphosyntax and Field Methods, devoted to the development of analytical skills in linguistics. We hypothesized that the application of metacognition surveys would enhance students' awareness of techniques that promote critical thinking and active learning. Two surveys built in as core components in each course were deployed at different points during the semester. Students' responses indicate that metacognition surveys can help students and instructors gain greater awareness of learning concerns and capabilities and identify areas for intervention.

I describe a typological gap in case and agreement alignment in ditransitive constructions. In languages in which verbal agreement is controlled by the subject and at most one object, object case and agreement in ditransitive constructions do not exhibit all logically possible combinations of alignment. I show that this typological gap follows from assumptions about the structure of ditransitive constructions (recipients c-command themes) and the interaction of morphological case and agreement (case marking restricts agreement). These assumptions derive exactly and only the attested patterns of alignment. I also argue that the typological gap in ditransitive constructions has a parallel in transitive constructions, providing further support for the proposals made here.

Negation in Tuparí (Tupían; Brazil) is an exclusively nominal category: verbs must enter into a nominalized form to accept the negator -'om and must undergo a subsequent process of reverbalization so as to combine with tense and evidential morphology. These category-changing processes leave -'om in a low position in the clause, and scopal evidence confirms that negation is also interpreted low. In keeping with the low structural position of -'om, the same negative strategy known from finite matrix clauses appears in nonfinite embedded contexts as well.

Tuparí shows that negative phrases exhibit more crosslinguistic variation than standardly assumed: they may appear in either the nominal or verbal extended projection. This finding is not compatible with cartographic efforts to strictly circumscribe the distribution of NegP within the clause. Like nominal tense in Tupi-Guaraní and other languages, in Tuparí a grammatical category normally associated with the verbal domain instead surfaces within the nominal one. For the purpose of typological comparison, the Tuparí facts highlight the need for classifications of negation that take into account both constructional asymmetries between affirmative and negative clauses and individual negator morphemes' selectional and categorical properties.

The German gender system is known for its complexity, and there is a persistent misconception that it is largely arbitrary, and hence a challenge for the typology of gender systems. In response, we construct a database of more than 30,000 German nouns and show that a boosting tree model achieves a predictive success of 96%. Even more surprising, the model performs at 87% when trained on just the 100 most frequent nouns. We thus demonstrate that the complex German system fits into a typologically well-known scheme, being a combination of semantic and formal assignment principles. In addition to our success with the specific problem, we show the value of statistical modeling for typologists and reflect on what exactly we can learn from these techniques.

In many different domains of language acquisition, there exists an apparent learnability problem to which innate knowledge of some aspect of Universal Grammar (UG) has been proposed as a solution. The present article reviews these proposals in the core domains of (i) identifying syntactic categories, (ii) acquiring basic morphosyntax, (iii) structure dependence, (iv) subjacency, and (v) the binding principles. We conclude that, in each of these domains, the innate UG-specified knowledge posited does not, in fact, simplify the task facing the learner.

In this response to commentators on our target article ‘Child language acquisition: Why universal grammar doesn't help’, we argue that the fatal flaw in most UG-based approaches to acquisition is their focus on describing the adult end-state in terms of a particular linguistic formalism. As a consequence, such accounts typically neglect to link acquisition to the language that the learner actually hears, instead assuming that she is able, by means usually unspecified, to perceive her input in terms of high-level theoretical abstractions.

This article explores the variation surrounding the semi-modals be going to and gonna. While gonna is frequently mentioned alongside be going to, it remains under-described in traditional grammars and academic literature. However, recent studies within Construction Grammar suggest that gonna may represent an independent construction, prompting a reconsideration of other variants within the be going to / gonna paradigm such as gon and imma, which appear to derive directly from gonna and no longer from be going to. In light of recent work, what have traditionally been regarded as mere ‘phonetic realizations’ or ‘orthographic variants’ may in fact play a more significant role in the formation and definition of constructions, raising questions about the structure of constructional networks. This article analyzes the immediate syntactic environment of the variants to account for both the variation of forms and the status of such forms. The study is conducted using two corpora that are particularly prone to showing linguistic innovations and language change: a spontaneous spoken corpus and a web corpus. Findings indicate that shorter variants often involve elision of be and that gonna is more grammaticalized than going to, based on the types of verbs they precede.

Phonotactic patterns are commonly constrained by morphology. In English, for example, non-homorganic nasal–stop sequences are disallowed within morphemes but may occur across morpheme boundaries. This article demonstrates that similar effects of morphology on phonotactics can be found with non-concatenative morphology, even though they involve morphological domains that are more difficult to identify on the surface. Specifically, vowel alternation in a class of Egyptian Arabic verbs is affected by gradient phonotactic restrictions on consonant–vowel co-occurrence. However, such restrictions are only active in the imperfective form (e.g., [-rgaʕ] ‘return.ipfv’), not the perfective (e.g., [rigiʕ] ‘return.pfv’). Using a lexicon study and a wug test, I show that this pattern is in fact bounded by morphological domains and is reliably generalised by speakers when deriving novel forms. I compare accounts of this effect that differ on whether they require abstract morphosyntactic representations and non-concatenative morphemes and discuss their implications.

An avenue for the progress of areal linguistics in South America is the investigation of the geographical distribution of specific features, such as the expression of sociative causation. Sociative causation is a particular type of causation where the causer not only makes the causee do an action but also participates in it (Shibatani & Pardeshi, 2002). Guillaume & Rose (2010) hypothesize that dedicated sociative causative markers are an areal feature of South America, in particular western South America. The aim of the present paper is to reassess the spatial distribution of these markers based on a large worldwide sample of 325 languages. The results show that dedicated sociative causative markers are significantly more frequent in South America compared to the rest of the world.

In the introduction for their recent state-of-the-art volume on English at the grassroots, Meierkord and Schneider (2021) point out the recurrent problem of Creolistic study not being fully incorporated into the World Englishes paradigm, arguing, like Mufwene (1997; 2001) and others, that English-based Creoles are best viewed as varieties of English ‘and, as such, require their integration into existing models and theories, too’ (11). Further work which seeks to overtly integrate Creole varieties within studies of English at the grassroots – the ‘new player in the World Englishes paradigm’ (Buschfeld 2001, 25) – has not been quickly forthcoming, though, with most of the work in the field focusing on ‘typical’ multilingual settings. In an attempt to remedy this, the current paper discusses the language situation in Trinidad, the last island in the Caribbean’s Lesser Antilles. In Trinidad, Trinidadian English Creole (TEC) and Trinidadian English (TE) interact in a complex where English might be best viewed as a second dialect (ESD), rather than in one of the prototypical ENL, ESL, or EFL situations of acquisition or use (cf. Deuber 2014). After an exploration of the limited research that has been done on language use and social class in Trinidad, this paper compares those previous findings on morphosyntactic features with new data from short semi-structured interviews conducted with speakers who can be described as grassroots.

Languages in contact commonly leave an imprint on one other. The most straightforward of these imprints to identify is MAT-borrowing, which results in clearly identifiable lexical items of one language (the donor language) being used in utterances of another language (the recipient language). This stands in contrast with PAT-borrowing, which does not involve any such incorporation of “other language” material but rather results in the reshaping of existing structures of the recipient language on the model of the donor language. This type of language change is therefore arguably more “invisible” to speakers since no easily identifiable “other language” material is present.

This study presents a detailed examination of PAT-borrowing in Guernésiais, the Norman variety spoken in Guernsey (British Channel Islands), which is now at an advanced state of language shift. It also highlights a major difference between MAT- and PAT-borrowing, namely that, whereas MAT-borrowing can only be explained with reference to the dominant language, PAT-borrowing can on occasion admit an internal explanation.

This study examined the variability of language profiles in Spanish–English bilingual children with and without developmental language disorder (DLD). The data included 529 children between the ages of 5 and 10 years. Eighty-eight of these children were identified as having DLD. A latent profile analysis was conducted based on children’s morphosyntax and semantics performance in Spanish and English. The optimal model identified five different profiles, illustrating the heterogeneity in bilingual development. Children with DLD were observed across all profiles, but most were classified in the only two profiles where lower morphosyntax than semantic performance was observed across languages. These results show the variability in both bilingual children with and without DLD. Additionally, the hallmark deficit of DLD in morphosyntax was confirmed, with the morphological weakness being observed in each of the bilingual children’s languages. Children’s background factors (age, maternal education and language exposure) were associated with profile characteristics.

To determine the source of transfer in third language acquisition (L3A), we tested the effects of grammatical feature typology and L2 Spanish proficiency on the comprehension and production of grammatical [gender] and [number] in the early stages of L3A of Portuguese. We distinguish between the two features based on their participation in the lexical-conceptual structure of the lemma and its interaction with the morpho-syntactic derivation. L1 English speakers were tested on their knowledge of the features in both their L2 and their L3 through a grammaticality judgment task and an elicited production task. Results show that L3 learners transfer only some features, specifically [gender] rather than [number], suggesting a fine-grained divide in feature compositionality between the structural ([gender]) and semantic ([number]) features. We also found facilitative transfer only after a threshold acquisition in L2, in support of the Threshold Hypothesis. For beneficial transfer of a feature, mere knowledge of the L2 structure was found to be sufficient. However, a higher generalized L2 proficiency threshold was found to predict high L3 accuracy.