Volume 62, Issue 4

This volume is concerned with questions raised by two apparently conflicting observations on the inventory of expressions (and in particular, lexical items) in natural languages. On the one hand, it is widely assumed that the rules of grammar include so-called blocking constraints, which prohibit the co-existence of forms that are identical in meaning, but differ in form. On the other hand, it can be shown that many natural languages exhibit examples of non-blocking and at least some degree of linguistic variation involving exactly this type of (morphological) doublets which can be said to compete for realizing one and the same meaning. In the following we are going to introduce the different types of blocking that are discussed in the literature as well as non-blocking effects to set the contributions in this volume in their context.

We propose a syntactic approach to apparent blocking effects in the realization of definiteness marking in the Scandinavian languages. The claim is that the differences in definiteness marking can be attributed to a requirement that a definiteness feature ([DEF], a property of N) must be located at the left edge of the DP phase in order to be PIC-accessible for probes outside of the DP. As a result, [DEF] can be spelled out on N if N is the only element within DP and [DEF] is therefore part of DP’s edge domain (giving rise to suffixal marking). In contrast, the presence of an (overt) adjectival modifier (at the left edge of DP) requires feature movement of [DEF] to D, which is then realized as a prenominal article (with additional spell-out of the lower copy of [DEF] in Swedish). The paper also addresses the (slightly different) behavior of definiteness marking in the context of relative clauses and certain issues pertaining to the interpretation of the different strategies.

This paper deals with (non-) blocking phenomena arising in the context of morphological voice marking patterns, such as active versus passive (as in e.g. Latin), or active versus non-active (as in e.g. Albanian and Modern Greek). Starting from the observation that across languages with voice (or voice-related) marking non-blocking effects arise only with unaccusatives (specifically, with so-called “alternating unaccusatives”), but not for instance with passives or reflexives, the paper presents a variety of facts that converge in showing that this situation is due to a difference in the respective featural make-up of passives and unaccusatives. The central claim that is defended is that the passive morphology is being reanalysed as a morpheme that relates not just to an [–external argument] feature, but to a [+activity] feature (in the sense of Kallulli 2007) in a predicate structure (here: v 0 ) also containing a [–external argument] feature.

Dutch simplex nouns ending in schwa have two regular plural forms, e.g. bode-s/bode-n ‘messengers’. This means that neither of the two allomorphs -s and -n is blocked. Such regular non-blocking is non-existent in the closely related languages German and Afrikaans. To identify the reasons for this divergence, the history of plural marking in the three languages is compared. Based on the historical observations it is argued that the doublets in Dutch are retained due to the output principle of plurals ending in a trochaic foot, according to which both -s and -n may apply without preference on stems with non-final stress.

This paper examines the factors that play a role in blocking and non-blocking in verb formation in Modern Hebrew. Hebrew verbs are formed in prosodic templates (so-called binyanim) that consist of a vocalic template and affixes in some of them. I address the criteria for the selection of such templates with regard to verb innovation, relation between existing forms and variation of forms. I will show that the selection of a particular binyan is the result of an interaction between markedness and faithfulness constraints on the one hand and thematic-syntactic considerations on the other hand. While certain constraints block the formation of verbs in some binyanim, there are cases of non-crucial ranking of constraints that gives rise to free variation of existing forms.

The semantics of natural gender in animate nouns is modeled in the framework of bidirectional Optimality Theory (OT). This allows for the interaction of lexical, conceptual and contextual constraints and for a straightforward treatment of the effect of blocking in this domain. Two versions of bidirectional OT are discussed and related to each other in terms of Blutner’s (2007) notion of fossilization.