This thematic issue almost did not happen. One of us (JH) was almost killed two days after the deadline for article submissions. As a pedestrian on a sidewalk minding his own business, he was struck by a car that ran a red light and lost control after a collision. So when we write that we are delighted to be writing this introduction, over one year later, we both really mean it.

Following the Articulatory Model of Handshape (Keane 2014), which mathematically defines handshapes on the basis of joint angles, we propose two methods for calculating phonetic similarity: a contour difference method, which assesses the amount of change between handshapes within a fingerspelled word, and a positional similarity method, which compares similarity between pairs of letters in the same position across two fingerspelled words. Both methods are validated with psycholinguistic evidence based on similarity ratings by deaf signers. The results indicate that the positional similarity method more reliably predicts native signer intuition judgements about handshape similarity. This new similarity metric fills a gap in the literature (the lack of a theory-driven similarity metric) that has been empty since effectively the beginning of sign-language linguistics.

This paper examines a key difference between constraint conjunction and constraint weight additivity, arguing that the two do not have the same empirical coverage. In particular, constraint conjunction in weighted probabilistic grammar allows for superadditive constraint interaction, where the effect of violating two constraints goes beyond the additive combination of the two constraints’ weights alone. A case study from parasitic tone harmony in Dioula d'Odienné demonstrates superadditive local and long-distance segmental feature similarities that increase the likelihood of tone harmony. Superadditivity in Dioula d'Odienné is formally captured in Maximum Entropy Harmonic Grammar by weighted constraint conjunction. Counter to previous approaches that supplant constraint conjunction with weight additivity in Harmonic Grammar, information-theoretic model comparison reveals that weighted constraint conjunction improves the grammar's explanatory power when modelling quantitative natural language patterns.

Behavioural findings indicate that English, Mandarin and Korean speakers exhibit gradient sonority sequencing preferences among unattested initial clusters. While some have argued these results support an innate principle, recent modelling studies have questioned this conclusion, showing that computational models capable of inducing generalisations using abstract phonological features can detect these preferences from lexical statistics in the three languages. This paper presents a computational analysis of the development of initial clusters in Polish, which arguably presents a stronger test of these models. We show that (i) the statistics of Polish contradict the Sonority Sequencing Principle (SSP), favouring sonority plateaus, (ii) models that succeed in the other languages do not predict SSP preferences for Polish and (iii) children nonetheless exhibit sensitivity to the SSP, favouring onset clusters with larger sonority rises.

Opaque phonological patterns are sometimes claimed to be difficult to learn; specific hypotheses have been advanced about the relative difficulty of particular kinds of opaque processes (Kiparsky 1971, 1973), and the kind of data that is helpful in learning an opaque pattern (Kiparsky 2000). In this paper, we present a computationally implemented learning theory for one grammatical theory of opacity, a Maximum Entropy version of Stratal OT (Bermúdez-Otero 1999, Kiparsky 2000), and test it on simplified versions of opaque French tense–lax vowel alternations and the opaque interaction of diphthong raising and flapping in Canadian English. We find that the difficulty of opacity can be influenced by evidence for stratal affiliation: the Canadian English case is easier if the learner encounters application of raising outside the flapping context, or non-application of raising between words (e.g. life with [ʌɪ]; lie for with [aɪ]).

This paper presents a Maximum Entropy learner of grammars and lexicons (MaxLex), and demonstrates that MaxLex has an emergent preference for minimally abstract underlying representations. In order to keep the weight of faithfulness constraints low, the learner attempts to fill gaps in the lexical distribution of segments, making the underlying segment inventory more feature-economic. Even when the learner only has access to individual forms, properties of the entire system are implicitly available through the relative weighting of constraints. These properties lead to a preference for some abstract underlying representations over others, mitigating the computational difficulty of searching a large set of abstract forms. MaxLex is shown to be successful in learning certain abstract underlying forms through simulations based on the [i]~[Ø] alternation in Klamath verbs. The Klamath pattern cannot be represented or learned using concrete underlying representations, but MaxLex successfully learns both the phonotactic patterns and minimally abstract underlying representations.

I consider the use of computational simulations in phonology, and the benefits and dangers of making abstractions, or failing to make abstractions. I argue that the potential of simulation studies is not yet realised as it could be.

A computational notion of locality, based on forbidden substructures of a fixed size, is applied to autosegmental representations, and tone-association patterns are argued to be local. This is significant for phonological theory, for two reasons. First, this notion of locality provides for an explicit theory of tonal well-formedness that is superior to previous explanations in that it makes clear, restrictive typological predictions. Second, it provides a clear path for understanding how these patterns can be learned. A brief survey of major tone-association patterns shows that association generalisations which are edge-based (Mende and Hausa), quality-specific (Kukuya) or positional (Northern Karanga Shona) are all local in this way. This is contrasted with previous explanations of the typology, which require global reference to the directionality of association, and can thus overgenerate.

Domains play an integral role in linguistic theories. This paper combines locality domains with current models of the computational complexity of phonology. The first result is that if a specific formalism – strictly piecewise grammars – is supplemented with a mechanism to enforce first-order definable domain restrictions, its power increases so much that it subsumes almost the full hierarchy of subregular languages. However, if domain restrictions are based on linguistically natural intervals, we instead obtain an empirically more adequate model. On the one hand, this model subsumes only those subregular classes that have been argued to be relevant for phonotactic generalisations. On the other hand, it excludes unnatural generalisations that involve counting or elaborate conditionals. It is also shown that strictly piecewise grammars with interval-based domains are theoretically learnable, unlike those with arbitrary, first-order domains.

This article presents a selection of methods to analyse, compare, verify and formally prove properties about phonological generalisations. Drawing from both well-known and recent results in the domains of model checking and automata theory, a useful methodology for automating the task of comparing analyses and inventing counterexamples is explored. The methods are illustrated by practical case studies that are intended to both resolve concrete issues and be representative of typical techniques and results.