Peter Ackema

The middle construction and the syntax-semantics interface *

Abstract English and Dutch middle verbs have a grammatical subject that is the corresponding transitive verbs object. In this paper we will account for this fact by invoking the properties of a pre-syntactic level of semantic representation and its interplay with syntax proper. We will argue that the grammatical subject of a middle is its actual external argument, and we will propose a model of projection of arguments that allows for this. We will show that other special properties of middle constructions follow from the way the verbs logical subject is represented at the pre-syntactic level of representation. In particular, it will be shown that the Affectedness Condition on middle formation is not a condition on the type of argument that can appear as the middle verbs grammatical subject, but that its effects and some exceptions to it naturally follow from this representation. Our model will also allow an account of Dutch impersonal and ‘adjunct’ middles.

CONTEXT-SENSITIVE SPELL-OUT

This paper deals with a class of morphological alternations that seem to involve syntactic adjacency. More specifically, it deals with alternative realizations of syntactic terminals that occur when a particular phrase immediately follows a particular head. We argue that this type of allomorphy is not conditioned by a syntactic adjacency condition. Instead, it is found when the head and phrase in question are contained in the same prosodic phrase at the interface that connects syntax and phonology (PF). We illustrate our approach with six case studies, concerning agreement weakening in Dutch and Arabic, pronoun weakening in Middle Dutch and Celtic, and pro-drop in Old French and Arabic.

Effects of Short-Term Storage in Processing Rightward Movement

There is a striking asymmetry between leftward and rightward movement in syntax: whereas leftward movement can in principle be unbounded, rightward movement is subject to very strict locality conditions. There are two possible approaches to explaining this asymmetry. One can either assume that some syntactic principle disfavours rightward movement, or that some mechanism having to do with sentence processing is responsible. In this chapter we will argue that a processing approach to limitations on rightward movement is more fruitful. In particular, we will argue that the human parser cannot process certain instances of rightward movement because the introduction of an antecedent-trace relation leads to a conflict with information about the parse which is already stored in short-term memory before this relation can be established. Similar problems do not occur in cases of leftward movement.

Conflict resolution in passive formation

In this paper we will argue that a typology of passive formation can be derived from the optimality-theoretic interaction between three well-known constraints. These are the Extended Projection Principle (VP must have a subject), Stay (do not move) and Parse (elements from the input should occur in the output). The analysis will not only account for different forms of passives (personal passives versus impersonal passives) but also for languages which lack passives (of certain verbs or altogether). To account for this we employ the socalled null parse, a candidate without structure.

Agreement Weakening at PF: A Reply to Benmamoun and Lorimor

Benmamoun and Lorimor (2006) dispute the claim made in Ackema and Neeleman 2003 that certain agreement alternations in Standard Arabic, and various related phenomena, can successfully be analyzed in terms of postsyntactic spell-out rules that are sensitive to prosodic structure. In this reply, we argue that the data discussed by Benmamoun and Lorimor do not warrant their conclusion, and in fact provide further evidence in favor of our original analysis.

Word-Formation in Optimality Theory

A grammar can be defined as the set of principles that distinguish the possible morpheme combinations, word combinations and sound combinations in a language from the impossible ones. In traditional grammars a possible word, sentence or syllable is one that satisfies all the principles pertaining to it. Data may be accounted for by a conspiracy of principles, but the principles themselves do not compete with one another. No principle is violated in order to avoid violating another principle. In fact, no principles are violated at all in a grammatical sentence; violation of even a single grammatical principle inexorably means ungrammaticality. In recent years, theories of grammar have come up in which this no longer holds true, in particular Optimality Theory (henceforth OT) (see Prince and Smolensky 2004 [1993]). This theory emphasizes the role of competition in determining which forms are grammatical and which are not. The crucial question is which of a number of forms that compete for the realization of a particular concept satisfies the principles of grammar better than the others (where ‘better’ is defined in a precise way, to be discussed shortly). This will be the grammatical structure. This implies that grammatical structures can violate principles of grammar – as long as there is no competitor that does better. This also implies that different principles of grammar can impose demands on structure that are in direct conflict (meaning that in any structure at least one of them will be violated). Let us sketch the outlines of an OT-style grammar in a bit more detail. Such a grammar consists of two components. The first is a device, called GEN(erator), that determines how elements can be combined into a structure. The demands that GEN imposes on structures cannot be violated. (Thus, there remains room for inviolable principles in OT). Below, we will assume a minimal GEN component for morphology, one in which an operation of merger is applied to morphemes, so that hierarchically ordered structures for words are built. This parallels the building of structure in syntax, but the morphological GENerator is distinct from the syntactic one and builds structures specifically for the sub-word level. In other words, we assume word structures are not the result of operations in phrasal syntax such as head movement. For a defense of such a specific ‘word syntax’ component to build

Morphological selection and representational modularity

In much work on word formation, the morpho-phonological properties of morphemes are strictly separated from their semantic and morpho-syntactic properties. A morpheme is not a unit taken from the lexicon and combined with other morphemes. Rather, its semantic and morpho-syntactic features are inserted in the semantic and morpho-syntactic components respectively, while only the morpho-phonological component contains its overt form. The three representations thus formed must of course be related, something which is achieved by a set of mapping principles. A range of proposals along these lines can be found in Sproat (1985), Anderson (1992), Halle & Marantz (1993), Beard (1995) and Jackendoff (1997), amongst others. We will refer to models of this type as representationally modular (borrowing a term from Jackendoff).1

Semantic versus syntactic agreement in anaphora: The role of identity avoidance

General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and investigate your claim.