Volume 6, Issue 2

Words which frequently co-occur in language (‘collocations’) are often thought to be independently stored in speakers' minds. This idea is tested here through experiments investigating the extent to which corpus-identified collocations exhibit mental ‘priming’ in a group of native speakers. Collocational priming is found to exist. However, in an experiment which aimed to exclude higher-order mental processes, and focus instead on the ‘automatic’ processes which are thought to best reflect the organisation of the mental lexicon, priming is restricted to collocations which are also psychological associates. While the former finding suggests that collocations found in a large corpus are likely to have psychological reality, the latter suggests that we may need to elaborate our models of how they are represented.

Whole-word frequency effects are shown to exist in what appears to be a completely regular system, the spelling of prefix-final /z/ in Russian. Russian prefixes that underlyingly end in /z/ ( roz -, bez -, iz -) end in [s] on the surface when followed by a voiceless consonant. According to the rules of Russian orthography, the surface form, rather than the underlying form, must be reflected in the spelling. However, spelling errors reflecting the underlying form often occur, especially for the prefix bez -. The present paper reports that the error rate, either in natural typing on the web or in a classroom dictation task, for a given word is negatively correlated with the frequency of the word, suggesting that Russian writers rely, to a significant extent, on memory of complete orthographic forms as opposed to the orthographic rule. The frequency effect holds even within the set of regular inflectional variants of a single lexeme, with more frequent wordforms showing lower error rates. The evidence demonstrates a high degree of reliance on whole-form lexical retrieval even in what appears to be a regular system that is explicitly taught to the writers throughout their schooling in a morphologically rich language and thus provides support for the use of lexical retrieval even when it is not necessary (Baayen et al., Dutch inflection: The rules that prove the exception, Kluwer, 2002, Butterworth, Lexical representation, Academic Press, 1983, Bybee, Morphology: A study of the relation between meaning and form, John Benjamins, 1985, The phonology of the lexicon: Evidence from lexical diffusion, CSLI, 2000, vs. DiSciullo and Williams, On the definition of word, MIT Press, 1987, Pinker, Science 253: 530–535, 1991). However, reliance on retrieval is argued to be especially strong when there is a relatively long period of temporary uncertainty regarding which rule is applicable during processing (see also Albright, Lexical and morphological conditioning of paradigm gaps, Equinox, 2009, Barca et al., Reading and Writing 20: 495–509, 2007, Burani et al., Psychonomic Bulletin and Review 13: 346–352, 2006). The importance of temporary uncertainty and resulting rule competition suggests that the regular/irregular distinction needs to be reconsidered as even fully “regular” systems may feature rule competition due to temporary uncertainties about rule applicability. Reliance on retrieval may go largely undetected in Russian during schooling because teaching and test materials focus on the spelling of frequent words, which can be either produced by rule or retrieved as wholes. The largely complementary methodological challenges in studying lexical frequency effects in corpus and experimental data are discussed.

In this study we examine ten hypotheses related to stylistic features of two Semitic languages, Hebrew and Arabic. Our assumption was that these hypotheses would enable us to discriminate texts written originally in Hebrew from those translated from Arabic. The ten hypotheses take into account a contrastive analytical view as well as recent research in Translation Studies. Being cognate languages, Hebrew and Arabic share morphological, phonetic and semantic features; consequently, certain Arabic forms trigger similar forms in Hebrew. We see this as accounting for the predominance of interference in the case of this (cognate) language pair and as evidence of an overarching hypothesis, whereby translation between cognate languages will entail interference, often superseding the norm of adherence to target-language standards. Eight of the ten hypotheses were borne out, to varying extents.

This paper tackles the issue of the validity and reliability of coding discourse phenomena in corpus-based analyses. On the basis of a sample analysis of coherence relation annotation that resulted in a poor kappa score, we describe the problem and put it into the context of recent literature from the field of computational linguistics on required intercoder agreement. We describe our view on the consequences of the current state of the art and suggest three routes to follow in the coding of coherence relations: double coding (including discussion of disagreements and explicitation of the coding decisions), single coding (including the risk of coder bias, and a lack of generalizability), and enriched kappa statistics (including observed and specific agreement, and a discussion of the (possible reasons for) disagreement). We end with a plea for complimentary techniques for testing the robustness of our data with the help of automatic (text mining) techniques.

Direct objects in the Dutch middlefield can either precede adverbs or follow them. This word order variation is traditionally labeled scrambling. Based on a corpus study of scrambling in spoken Dutch, we show that pronouns scramble almost categorically, whereas indefinite and definite objects scramble hardly at all. The observed effect of definiteness cannot be reduced to the influence of grammatical weight, in this way establishing both factors as independent determinants of word order variation. A closer investigation of proper noun objects shows that their position relative to the adverb is influenced by their animacy, length and stress. We argue that the ordering of elements in the Dutch middlefield is to be understood in terms of planning considerations on behalf of the speaker such that use of the unscrambled order buys him as much time for articulation of the direct object as possible.