Volume 6, Issue 1

The present study investigates the conceptualization of our bodily orientation in a quantitative corpus-based approach. Specifically, we identify correlated target domains with respect to each spatial dimension and clarify to what extent the physical symmetry of the spatial dimension plays a role in the metaphorical conceptualization. Based on the symbolic nature of constructions, we examine the correlation patterns of the NPs and the locatives in Mandarin Locative Construction “ zai4 + NP + ( zhi1 ) + LOCATIVE” through three major statistical analyses, i.e., covarying collexeme analysis, hierarchical clustering and principal component analysis. The results show that only FRONT/BACK dimension displays symmetrical metaphorical extension to similar metaphorical domains. The distributional patterns of the locatives have far-reaching implications for the embodiment of conceptual metaphors. It is concluded that the (a)symmetry of metaphorical patterns along each spatial dimension may be attributed to our recurring (a)symmetrical daily interaction and bodily experiences with the surrounding physical environment.

This study addresses the statistical analysis of a phenomenon in Russian verbal paradigms, a suffix shift that is spreading through the paradigm and making it more regular. A problem that arises in the analysis of data collected from the Russian National Corpus is that counts documenting this phenomenon are based on repeated observations of the same verbs and, moreover, on counts for different parts of the paradigms of these same verbs. Unsurprisingly, individual verbs display consistent (although variable) behavior with respect to the suffix shift. The non-independence of the elementary observations in our data has to be taken into account in the statistical evaluation of the patterns in the data. We show how mixed-effects modeling can be used to do this in a principled way, and that it is also necessary to do so in order to avoid anti-conservative evaluation of significance.

Studies of modifier-noun compounds have indicated that they tend to follow regular semantic patterns (e.g., Downing, Language 53: 810–842, 1977; Warren, Acta Universitatis Gothoburgensis. Gothenburg Studies in English Goteborg 41: 1–266, 1978). The results of several psycholinguistic studies have supported the hypothesis that people rely on statistical knowledge about how nouns tend to be used in combination in order to facilitate the interpretation of novel compounds (e.g., Gagné & Shoben, Journal of Experimental Psychology: Learning, Memory and Cognition 23: 71–87, 1997; Maguire, Maguire & Cater, Journal of Experimental Psychology: Learning, Memory, and Cognition 36: 288–297, 2010; Storms & Wisniewski, Memory and Cognition 33: 852–861, 2005). We conducted a series of corpus analyses in order to establish the salience and reliability of semantic patterns in English compounds. These analyses demonstrated that similar concepts tend to appear in combination with similar sets of nouns. In addition, categorizing combinations according to the semantic category of the modifier and head revealed consistent regularities in productivity reflecting the likelihood of plausible relationships. These findings support the idea that statistical knowledge about semantic patterns in compounding can be used to facilitate the interpretation of novel compounds. The implications for existing theories and models of conceptual combination are discussed.

Where the variables selected for cluster analysis of linguistic data are measured on different numerical scales, those whose scales permit relatively larger values can have a greater influence on clustering than those whose scales restrict them to relatively smaller ones, and this can compromise the reliability of the analysis. The first part of this discussion describes the nature of that compromise. The second part argues that a widely used method for removing disparity of variable scale, Z-standardization, is unsatisfactory for cluster analysis because it eliminates differences in variability among variables, thereby distorting the intrinsic cluster structure of the unstandardized data, and instead proposes a standardization method based on variable means which preserves these differences. The proposed mean-based method is compared to several other alternatives to Z-standardization, and is found to be superior to them in cluster analysis applications.