Language learners are often faced with a scenario where the data allow multiple generalizations, even though only one is actually correct. One promising solution to this problem is that children are equipped with helpful learning strategies that guide the types of generalizations made from the data. Two successful approaches in recent work for identifying these strategies have involved (i) expanding the set of informative data to include INDIRECT POSITIVE EVIDENCE, and (ii) using observable behavior as a target state for learning. We apply both of these ideas to the case study of English anaphoric one, using computationally modeled learners that assume one’s antecedent is the same syntactic category as one and form their generalizations based on realistic data. We demonstrate that a learner that is biased to include indirect positive evidence coming from other pronouns in English can generate eighteen-month-old looking-preference behavior. Interestingly, we find that the knowledge state responsible for this target behavior is a context- dependent representation for anaphoric one, rather than the adult representation, but this immature representation can suffice in many communicative contexts involving anaphoric one. More generally, these results suggest that children may be leveraging broader sets of data to make the syntactic generalizations leading to their observed behavior, rather than selectively restricting their input. We additionally discuss the components of the learning strategies capable of producing the observed behavior, including their possible origin and whether they may be useful for making other linguistic generalizations.

Manner and locative expressions modifying stative predicates, as in own (something) honestly and (be) quiet in the car, are rare compared to those modifying dynamic predicates, and it has been claimed (for example, in Maienborn 2005 and Katz 2008) that they are systematically excluded on semantic grounds. I argue here that this is not so: in fact, they are perfectly acceptable once the restrictions on them are understood. I propose further that these restrictions take the form of (i) a pragmatic condition that generally bans locative modification of stative predicates, but that may be overridden in certain defined contexts, and (ii) regular semantic incompatibilities between adverbs and stative predicates, which, being semantically ‘impoverished’, have relatively few modifiable semantic features compared to dynamic predicates. These proposals are supported by extensive examples. The conclusions indicate that there is no need to treat states as fundamentally different from other eventualities, whether by invoking Kimian states or by avoiding eventuality variables altogether in their representations.

Old English *motan and Middle English *moten, the ancestors of modern must, are commonly described as ambiguous between a possibility and a necessity reading. I argue instead that in the Alfredian Old English prose, *motan was a nonambiguous ‘variable-force’ modal, with the modal force different from both possibility and necessity. I propose that *motan's variable-force effect was due to the presupposition of a collapse between possibility and necessity. Informally, motan(p) presupposed ‘if p gets a chance to actualize, it will’. I then trace the development of *motan into a modal genuinely ambiguous between necessity and possibility in Early Middle English.

The midpoint pathology (in the sense of Kager 2012) characterizes a type of unattested stress system in which the stressable window contracts to a single word-internal syllable in some words, but not others. Kager (2012) shows that the pathology is a prediction of analyses employing contextual lapse constraints (e.g. *ExtLapseR; no 000 strings at the right edge) and argues that the only way to avoid it is to eliminate these constraints from Con. This article explores an alternative: that systems exhibiting the midpoint pathology are unattested not because the constraints that would generate them are absent from Con, but because they are difficult to learn. This study belongs to a growing body of work exploring the idea that phonological typology is shaped by considerations of learnability.

Many properties of languages, including sign languages, are not uniformly distributed among items in the lexicon. Some of this nonuniformity can be accounted for by appeal to articulatory ease, with easier articulations being overrepresented in the lexicon in comparison to more difficult articulations. The literature on ease of articulation deals only with the active effort internal to the articulation itself. We note the existence of a previously unstudied aspect of articulatory ease, which we call REACTIVE EFFORT: the effort of resisting incidental movement that has been induced by an articulation elsewhere in the body. For example, reactive effort is needed to resist incidental twisting and rocking of the torso induced by path movement of the manual articulators in sign languages. We argue that, as part of a general linguistic drive to reduce articulatory effort, reactive effort should have a significant effect on the relative frequency in the lexicon of certain types of path movements. We support this argument with evidence from Italian Sign Language, Sri Lankan Sign Language, and Al-Sayyid Bedouin Sign Language, evidence that cannot be explained solely by appeal to constraints on bimanual coordination. As the first exploration of the linguistic role of reactive effort, this work contributes not only to the developing field of sign language phonetics, but also to our understanding of phonetics in general, adding to a growing body of functionalist literature showing that some linguistic patterns emerge from more fundamental factors of the physical world.

From the study of sign languages we know that the visual modality robustly supports the encoding of conventionalized linguistic elements, yet while the same possibility exists for the visual bodily behavior of speakers of spoken languages, such practices are often referred to as ‘gestural’ and are not usually described in linguistic terms. This article describes a practice of speakers of the Brazilian indigenous language Nheengatú of pointing to positions along the east-west axis of the sun’s arc for time-of-day reference, and illustrates how it satisfies any of the common criteria for linguistic elements, as a system of standardized and productive form-meaning pairings whose contributions to propositional meaning remain stable across contexts. First, examples from a video corpus of natural speech demonstrate these conventionalized properties of Nheengatú time reference across multiple speakers. Second, a series of video-based elicitation stimuli test several dimensions of its conventionalization for nine participants. The results illustrate why modality is not an a priori reason that linguistic properties cannot develop in the visual practices that accompany spoken language. The conclusion discusses different possible morphosyntactic and pragmatic analyses for such conventionalized visual elements and asks whether they might be more crosslinguistically common than we presently know.

This article discusses in detail two cases of even-marked negative polarity items (NPIs) in Greek and Korean that are not scalar or exhaustive. This prima facie paradoxical finding suggests that EVEN-marking is not always an indicator of scalarity—and, at least in the case of the Korean and Greek NPIs discussed, even is grammaticalized as a nonscalar NPI marker. We propose that the nonscalar NPIs are antispecific indefinites with referential vagueness, which is a form of ignorance best captured as nonexhaustive variation in the potential values of the NPIs (Giannakidou & Quer 2013). We also show that the difference in Greek and Korean between scalar and nonscalar NPIs is reflected in prosody: scalar NPIs are ‘emphatic’, and nonscalar NPIs are ‘nonemphatic’; we therefore conclude that prosodic prominence, not even, signals scalar structure. The fact that not all NPIs are scalar or exhaustive falsifies theories claiming that exhaustivity is the source of all NPIs (Chierchia 2006, 2013).

A comparison of speakers’ treatment of two categorically unattested phonotactic structures in Cochabamba Quechua reveals a stronger grammatical prohibition on roots with pairs of ejectives, *[k'ap'u], than on roots with a plain stop followed by an ejective, *[kap'u]. While the distribution of ejectives can be stated as a single restriction on ejectives preceded by stops (ejective or plain), *[-cont, -son][cg], speakers show evidence of having learned an additional constraint that penalizes cooccurring ejectives more harshly, *[cg][cg]. An inductive learning bias in favor of constraints with the algebraic structure of*[cg][cg] is hypothesized (Marcus 2001, Berent et al. 2002, Berent et al. 2012), allowing such constraints to be preferred by learners over constraints like *[-cont, -son][+cg], which penalize sequences of unrelated feature matrices.

The verbal suffixes of Seri (a language isolate of Sonora, Mexico) divide the lexicon into classes of unparalleled complexity. The paradigm has only four forms, which mark subject number and aspect (or event number), yet there are over 250 distinct types in a corpus of just under 1,000 verbs. This relation of forms to types means that by information-theoretic measures this is among the most complex inflection class systems yet studied. In part this complexity is due to the sheer wealth of allomorphs and the freedom with which they combine within the paradigm; however, these properties can be found in all inflection class systems of any complexity. The unique property of Seri is that although the suffix morphology and the morphosyntactic paradigm have the same featural content, the two systems are not directly coordinated. Both suffix morphology and verbal morphosyntax are based on the concatenation of markers of plurality, and an increase in the morphological marking of plurality reflects a morphosyntactic accumulation of subject and predicate plurality (i.e. aspect). In this sense, morphology is a direct exponent of featural content. But there is no consistent mapping between the two systems, and the precise calibration between morphological form and morphosyntactic function must be lexically specified; it is this specification that increases dramatically the number of inflectional types. Seri therefore represents a middle ground between the conceptual extremes of morphosyntactically motivated and morphologically autonomous morphology that serve as a basis for much of our theory building.

The quantifiers most and more than half pose a challenge to formal semantic analysis. On the one hand, their meanings seem essentially the same, prompting accounts that treat them as logically equivalent. On the other hand, their behavior diverges in a number of interesting ways. This article draws attention to some previously unnoticed contrasts between the two and develops a novel semantic analysis of them, based on principles of measurement theory. Most and more than half have logical forms that are superficially equivalent (per Hackl 2009), but that place different requirements on the structure of the underlying measurement scale: more than half requires a ratio scale, while most can be interpreted relative to an ordinal scale or one with a semiordered structure. The latter scale type is motivated by findings from psychophysics and by psychological models of humans’ approximate numerical abilities. A corpus analysis is presented that confirms the predictions of the present account.

This article follows a change in pronunciation of word-medial intervocalic /t/ in New Zealand English, as it unfolds over 120 years. Data are analyzed in the context of questions about the role of experience-based lexical representations and their potential impact on the time course of sound change in progress. Three major results are reported. First, frequent words lead the change. Second, the distributions of individual words affect their participation in the change: words favored by younger speakers are produced with newer variants. Finally, the topic of conversation affects which variant is favored: older topics elicit older variants. Together, these findings provide evidence that phonetic distributions of word-level representations are implicated in the course of sound change.

Since Du Bois's (1987b) seminal paper, ergative alignment in morphosyntax has been claimed to correlate with a characteristic constellation of argument realization in discourse: both intransitive subjects (S) and transitive objects (P) serve to introduce new referents via full noun phrases (NPs), while transitive subjects (A) are dispreferred for this function and are thus mostly realized as pronouns or zero (e.g. Dixon 1995, Du Bois et al. 2003, Goldberg 2004). This ergative patterning in discourse is generally accounted for in terms of information-management strategies employed by speakers in dealing with the cognitive demands of introducing and monitoring referents in discourse. These claims have recently been questioned by Everett (2009), whose data (English and Portuguese) show no support for the claimed ergative bias in discourse and raise doubts about explanations in terms of information management. The present article subjects the claims of an ergative bias in discourse to more rigorous testing, drawing on the largest database compiled to date (nineteen spoken-language corpora from fifteen typologically diverse languages), and assesses the explanatory frameworks. We find that, with the exception of Du Bois's original Sakapultek data, there is very little evidence for the postulated ergative pattern in natural spoken-language discourse crosslinguistically. Although our findings do confirm low levels of full NPs in the A role (Du Bois's ‘Non-lexical A’ constraint), we concur with Everett (2009) that the semantic feature [±human] provides an empirically more sound and conceptually more economical account than earlier explanations framed in terms of information management. Finally, we address the plausibility of emergentist claims for a diachronic link between ergative alignment in morphosyntax and information flow in discourse. The raw data used in this article and extensive exemplification of the methodology employed are available as online supplementary materials.

This article is concerned with the analysis of ‘short’ or ‘fragment’ answers to questions, and the relationship between these and the hypothesis of DIRECT COMPOSITIONALITY (DC) (e.g. Montague 1970). DC claims that the syntax and semantics work ‘in tandem’ to prove expressions well formed, while at the same time assigning them a meaning (a model-theoretic object). DC makes it difficult to state any kind of identity condition for ‘ellipsis’ and would hence lead one to suspect that short answers do not contain hidden linguistic material. This article argues that they indeed do not. Rather, as proposed in Groenendijk & Stokhof 1984, the question and short answer together form a linguistic unit, which I call a Qu-Ans, whose semantics gives the proposition that is understood as following from the pair. Three new arguments are adduced for the Qu-Ans analysis over one making use of silent linguistic material, and a core class of traditional arguments for silent linguistic material are answered. Moreover, it is shown that many of the traditional arguments for silent linguistic material themselves presuppose a non-DC architecture. If (as is claimed) these arguments do not hold, the Qu-Ans analysis of short answers actually supports the DC view, under which no use is made of logical form, and no use is made of representational constraints on structure.

We compare the complexity of idiosyncratic sound patterns involving American English /ɹ/ with the relative simplicity of clear/dark /l/-allophony patterns found in English and other languages. For /ɹ/, we report an ultrasound-based articulatory study of twenty-seven speakers of American English. Two speakers use only retroflex /ɹ/, sixteen use only bunched /ɹ/, and nine use both /ɹ/ types, with idiosyncratic allophonic distributions. These allophony patterns are covert, because the difference between bunched and retroflex /ɹ/ is not readily perceived by listeners. We compare this typology of /ɹ/-allophony patterns to clear/dark /l/-allophony patterns in seventeen languages. On the basis of the observed patterns, we show that individual-level /ɹ/ allophony and language-level /ɹ/ allophony exhibit similar phonetic grounding, but that /ɹ/-allophony patterns are considerably more complex. The low complexity of language-level /l/-allophony patterns, which are more readily perceived by listeners, is argued to be the result of individual-level contact in the development of sound patterns. More generally, we argue that familiar phonological patterns (which are relatively simple and homogeneous within communities) may arise from individual-level articulatory patterns, which may be complex and speaker-specific, by a process of koineization. We conclude that two classic properties of phonological rules, phonetic naturalness and simplicity, arise from different sources.

Quotative be like is a rapid global innovation, yet no evidence pinpoints when it arose, under what circumstances, or the consequences of its emergence. Using a data set spanning four cities and two hemispheres, we document systemic regularity across time and space. The results force us to confront three issues: the uniformitarian principle, the criterion of face-to-face contact in the diffusion of language change, and the nature of language as a complex adaptive system. Be like is an outlier, it has had a major impact on the linguistic system, and it can only be rationalized by hindsight, demonstrating the possibility of significant random events outside the predictable structures and processes in language. We conclude by suggesting that be like is a (linguistic) black swan event (Taleb 2010).

While published linguistic judgments sometimes differ from the judgments found in large-scale formal experiments with naive participants, there is not a consensus as to how often these errors occur nor as to how often formal experiments should be used in syntax and semantics research. In this article, we first present the results of a large-scale replication of the Sprouse et al. 2013 study on 100 English contrasts randomly sampled from Linguistic Inquiry 2001-2010 and tested in both a forced-choice experiment and an acceptability rating experiment. Like Sprouse, Schütze, and Almeida, we find that the effect sizes of published linguistic acceptability judgments are not uniformly large or consistent but rather form a continuum from very large effects to small or nonexistent effects. We then use this data as a prior in a Bayesian framework to propose a small n acceptability paradigm for linguistic acceptability judgments (SNAP Judgments). This proposal makes it easier and cheaper to obtain meaningful quantitative data in syntax and semantics research. Specifically, for a contrast of linguistic interest for which a researcher is confident that sentence A is better than sentence B, we recommend that the researcher should obtain judgments from at least five unique participants, using at least five unique sentences of each type. If all participants in the sample agree that sentence A is better than sentence B, then the researcher can be confident that the result of a full forced-choice experiment would likely be 75% or more agreement in favor of sentence A (with a mean of 93%). We test this proposal by sampling from the existing data and find that it gives reliable performance.

Northern Paiute uses clause chaining to express certain temporal relations between clauses, which are conveyed by temporal subordinators such as after and while in English. Rather than a subordination structure, however, I show that clause chaining in Northern Paiute has an underlying coordination structure. I propose that temporal relations between the clauses in a chain arise, in part, from verbal morphology conveying relative tense. In Northern Paiute, relative tense can be bound in a coordination structure, as in an embedded clause in other languages (Ogihara 1994, 1995, 1996, Abusch 1997). In addition, I argue that this semantics is enriched pragmatically to produce a ‘forward-moving’ temporal interpretation characteristic of narrative discourse (Kamp & Rohrer 1983, among others). This in-depth investigation of one language raises questions about the syntax and semantics of clause chaining in other languages.

The conventional wisdom regarding the diachronic process whereby phonetic phenomena become phonologized appears to be the ‘error accumulation’ model, so called by Baker, Archangeli, and Mielke (2011). Under this model, biases in the phonetic context result in production or perception errors, which are misapprehended by listeners as target productions, and over time accumulate into new target productions. In this article, I explore the predictions of the hypocorrection model for one phonetic change (prevoiceless /ay/-raising) in detail. I argue that properties of the phonetic context underpredict and mischaracterize the contextual conditioning on this phonetic change. Rather, it appears that categorical, phonological conditioning is present from the very onset of this change.

This article addresses a phenomenon of long-standing interest: the existence of child-specific phonological patterns that are not attested in adult language. We propose a new theoretical approach, termed the A(RTICULATORY)-MAP model, to account for the origin and elimination of child-specific phonological patterns. Due to the performance limitations imposed by structural and motor immaturity, children’s outputs differ from adult target forms in both systematic and sporadic ways. The computations of the child’s grammar are influenced by the distributional properties of motor-acoustic traces of previous productions, stored in episodic memory and indexed in the eponymous A-map. We propose that child phonological patterns are shaped by competition between two essential forces: the pressure to match adult productions of a given word (even if the attempt is likely to fail due to performance limitations), and the pressure to attempt a pronunciation that can be realized reliably (even if phonetically inaccurate). These forces are expressed in the grammar by two constraints that draw on the motor-acoustic detail stored in the A-map. These constraints are not child-specific, but remain present in the adult grammar, although their influence is greatly attenuated as a wide range of motor plans come to be realized with a similar degree of reliability. The A-map model thus not only offers an account of a problematic phenomenon in development, but also provides a mechanism to model motor-grammar interactions in adult speech, including in cases of acquired speech impairment.