According to a ‘radical orphanage’ approach, non-restrictive relative clauses are not part of the syntactic representation of the sentence that contains them. It is an appealing view, and seems to capture some important properties of non-restrictive relative clauses. Focusing mainly on empirical shortcomings, this paper aims to show that the appeal of such approaches is illusory. It also outlines an empirically superior ‘syntactically integrated’ account.

The classification of that is, or (in other words), and or rather as reformulation markers would seem to suggest that the utterances they introduce achieve relevance in the same way. However, the examination of a range of discrepancies between reformulations introduced by or, on the one hand, and reformulations introduced by that is, on the other, suggests that any account of the pragmatics of reformulation must be a non-unitary one. In this paper, I build on Burton-Roberts' (1993) suggestion that the reformulations introduced by or are meta-linguistic in character, and show how these can be distinguished from the reformulations introduced by that is, which must be analysed at the level of conceptual representation. I also show how this distinction corresponds to a distinction between the different ways in which a parenthetical construction may be pragmatically integrated with its host. As Potts (2005) would predict, parenthetical that is-reformulations are not themselves part of the truth-conditional content of their hosts. However, in contrast with or-reformulations, they communicate information about the propositional content of their hosts, and in this way can contribute to the identification of their truth-conditional content at the level of pragmatic interpretation.

Welsh vowel mutation is a purely positional vowel alternation, the effects of which serve to obscure phonemic contrasts between three vowels in the system, namely barred-i, schwa and [u]. The theoretical interest in this alternation stems from the surface orientation of current phonological theory: is such a non-surface-true state of affairs amenable to plausible modelling in an optimality-theoretic framework, or are the relevant relationships best accounted for through lexical listing? In this paper I argue that a straightforward optimality-theoretic account is available and that this account is simpler than any of its predecessors. The analysis differs from previous derivational analyses (e.g. Thomas 1979, 1984; Williams 1983; Bosch 1996) in various ways, including the underlying values of some of the vowels involved, the avoidance of ad hoc extrinsic rule ordering, and the lack of reliance on intermediate representations. Furthermore, reference to phonological position alone is sufficient, with no need to refer either to stress or to morphological complexity. The correct results emerge primarily through the interaction of a high-ranking structural constraint prohibiting schwa in a final syllable, an input–output faithfulness constraint on vowel features, and a constraint prohibiting a high central rounded vowel, [u with bar through].

A common view in theoretical syntax and computational linguistics holds that there are no grammatical restrictions on multiple center-embedding of clauses. Syntax would thus be characterized by unbounded recursion. An analysis of 119 genuine multiple clausal center-embeddings from seven ‘Standard Average European’ languages (English, Finnish, French, German, Latin, Swedish, Danish) uncovers usage-based regularities, constraints, that run counter to these and several other widely held views, such as that any type of multiple self-embedding (of the same clause type) would be possible, or that self-embedding would be more complex than multiple center-embedding of different clause types. The maximal degree of center-embedding in written language is three. In spoken language, multiple center-embedding is practically absent. Typical center-embeddings of any degree involve relative clauses specifying the referent of the subject NP of the superordinate clause. Only postmodifying clauses, especially relative clauses and that-clauses acting as noun complements, allow central self-embedding. Double relativization of objects (The rat the cat the dog chased killed ate the malt) does not occur. These corpus-based ‘soft constraints’ suggest that full-blown recursion creating multiple clausal center-embedding is not a central design feature of language in use. Multiple center-embedding emerged with the advent of written language, with Aristotle, Cicero, and Livy in the Greek and Latin stylistic tradition of ‘periodic’ sentence composition.

In this paper, we explore the possibility that machine learning approaches to natural-language processing being developed in engineering-oriented computational linguistics may be able to provide specific scientific insights into the nature of human language. We argue that, in principle, machine learning results could inform basic debates about language, in one area at least, and that in practice, existing results may offer initial tentative support for this prospect. Further, results from computational learning theory can inform arguments carried on within linguistic theory as well.