This paper presents the results of a corpus-based analysis of a special type of modification of the English (as) Adj as NP similes. The modification involves filling the property slot with a cognate noun-adjective compound, i.e., a compound adjective consisting of the original adjective and the noun representing the original source of comparison, and inserting a new source of comparison into the construction (red as blood vs blood-red as a raw steak). Our data come from three sources: the Corpus of Contemporary American English (COCA), the iWeb Corpus, and material published on the Google website. Using quantitative methods we first explored whether there is a relationship between various distributional and formal features of the “original” as-similes and their likelihood of exhibiting this type of modification behavior. We then performed a quantitative and qualitative analysis of the semantic and, to a lesser extent, discourse-related features of authentic examples of this type of modification. Our results indicate that while these modifications are not abundant, the as-similes that have been found to modify in this way are significantly different from the as-similes that have been found not to modify in this way, on a number of formal and distributional features. The analysis of the semantic and discourse-related features of the modifications themselves revealed (i) the typical semantic domains of the three nominal entities featured in the modified simile: the original source, the new source, and the target, including the semantic fit among the domains of those entities; (ii) the typical semantic domains of the properties for which the three nominal entities are compared, including the semantic fit among the domains of those properties; and (iii) the typical text varieties accommodating these modifications. The latter results confirm, and to some extent elaborate on, some earlier findings about the semantic and discourse-related profiles of similes at large.

No abstract

In his typological study of the genesis of diminutive and augmentative suffixes Grandi (2011) attributed the development of augmentative suffixes in Romance and Slavic languages to four sources: agentive/ pejorative derivational suffixes, collective suffixes, relational suffixes, and – in Slavic languages – suffixes deriving place nouns. This paper seeks to establish how the Croatian suffix –ara fi ts into Grandi’s (2011) system. According to Babić (2002: 130, 132), suffixations in –ara primarily denote places but there is a small number of suffixations with augmentative or pejorative meaning. Skok (1971: 49–52), in turn, describes this Pan–Slavic suffix under the entry for the agentive suffix –ar, which goes back to Latin –arius. This paper sought to identify any conceptual indications in our synchronic data that the augmentative/ evaluative meanings of the suffix may have developed from any of the four meanings, viz. agentive/ pejorative, collective, relational and locative. The paper leans on usage–based models of language and Rainer’s (2005) model of semantic change in word formation involving two mechanisms: reinterpretation and approximation. Results indicate that the augmentative and other evaluative meanings of the suffix may have developed from its agentive/pejorative meaning through a complex interplay of reinterpretation and approximation, with some influence of metaphor and metonymy and structural factors. Our results are submitted as hypotheses to be tested in future by dedicated diachronic studies.

This paper explores the polysemy of the Croatian verbal prefix od- ‘(away) from’ from a Cognitive Grammar perspective (Langacker, 1987, 1991, etc.; Taylor, 2002). Despite the seemingly inordinate semantic heterogeneity of od-prefixed verbs, we argue that the different uses of od- are semantically motivated and that this has at least as much to do with the existence of a single abstract sanctioning schema as it does with the motivated extensions from the prototype, viz. spatial ablativity, or other category members. In this way we depart not only from the deep-rooted Croatian tradition, which tends to ignore any semantic motivation among different uses of a single prefix (e.g. Babić, 1986), but also from those cognitive linguistic approaches which rely on meaning chains, i.e. categorization by extension, to account for the polysemy of prefixes, prepositions and particles (e.g. Janda, 1985, 1986, 1988; Šarić, 2006a, 2006b, 2008, etc.).

IgA-vAskulItIs sA zAhvAćAnjem bubregA u odrAsloj dobI:IskustvA Iz klInIčke bolnIce dubrAvA, zAgreb ana gudelj gračanin 1 , tea mikula 2 , ivica horvatić 3 , luka torić 3 , majda golob 1 , Jasminka dobša 4 , matea liskij 2 , gabrijela buljan 2 , Karla draženović 2 , matej nedić 5 , danica galešić ljubanović 6 , Krešimir galešić 31 division of clinical Immunology, Allergology and rheumatology, department of Internal medicine, school of medicine university of zagreb, clinical hospital dubrava, zagreb, croatia / zavod za kliničku imunologiju, alergologiju i reumatologiju, klinika za unutarnje bolesti medicinskog fakulteta sveučilišta u zagrebu, klinička bolnica dubrava, zagreb, hrvatska; 2 school of medicine of the university of zagreb, zagreb, croatia / medicinski fakultet sveučilišta u zagrebu, zagreb, hrvatska;3 division of nephrology and dialysis, department of Internal medicine, school of medicine university of zagreb, clinical hospital dubrava, zagreb, croatia / zavod za nefrologiju i dijalizu, klinika za unutarnje bolesti medicinskog fakulteta sveučilišta u zagrebu, klinička bolnica dubrava, zagreb, hrvatska;