LEADhoC - Linguistic Expression of Ad hoc Categories

The LEAdhoC project aims to analyze the linguistic expression of a basic cognitive process, namely the construction of ad hoc categories (Barsalou 1983 and onwards). Ad hoc categories are constructed to achieve communicative goals and depend on context for their interpretation (e.g. "activities one can do at home"), are usually less established in memory and less easily mastered by speakers than common categories. While the latter can typically be expressed by fairly short conventional linguistic means (words, e.g. clothing, or small phrases, e.g. vegetarian food), ad hoc categories often do not come with ready-made linguistic labels. Their identification in discourse nonetheless crucially depends on verbalization, i.e. the linguistic strategies that speakers systematically employ to refer to the process of ad hoc category building. These strategies typically involve the explicit naming of one or more exemplars, that the addressee processes as pointers for conjuring up the ad hoc category: e.g., in interpreting 'let's play tennis, take a walk, or something like that', the addressee has to construct the category 'activities we could do outdoor', in order to imagine further alternatives. However, the role of exemplars and the degree of context-dependence may vary in the process of ad hoc category building. For instance, in order to construct the category 'people and situations revolving around Berlusconi' from the derived noun 'Berluscon-ame' (Italian), Berlusconi has to be taken as an exemplar, but also as a property shared by all members, and no access to the specific speech situation is required, but rather a general knowledge of Italian politics. From a structural point of view, a shallow cross-linguistic survey shows great variation in the constructions encoding ad hoc categories, ranging from transparent discourse-level constructions such as “general extenders” (e.g. English ‘or stuff like that'), to synthetic, less transparent means, such as non-exhaustive connectives (e.g. Japanese -ya), dedicated plurals (e.g. similative plurals, Daniel 2000), derivational affixes and special types of reduplication.Although ad hoc categories are ubiquitous in our everyday cognition, no systematic analysis of their linguistic realizations has ever been made. The aim of the LEAdhoC project is to answer the following core theoretical questions, through three complementary research directions: 1) What types of constructions are attested to express ad hoc categories? Are there universal patterns and correlations between the type of ad hoc category, the degree of context-dependence and particular morphosyntactic features? TYPOLOGICAL SURVEY -The sampling procedure will be compliant with current standards in typological research. Due to the highly discursive nature of the phenomenon, descriptive grammars will be integrated by a metalinguistic questionnaire for language experts, a context/translation questionnaire for educated native speakers, and wherever possible by the analysis of naturally occurring texts. 2) What grammatical resources are recurrently mobilized to convey this function? DIACHRONIC ANALYSIS - The diachronic analysis is meant to provide (i) a comprehensive diachronic typology of the sources of the constructions encoding ad hoc categories and (ii) in-depth analyses of the pathways leading to them, with a focus on Romance languages. 3) Are there different aims and different ways in which ad hoc categories may be built in discourse? CORPUS-BASED STUDY - The analysis will minimally be conducted for English and Italian, based on a comparable conversational corpus that will be specifically designed and implemented for the project. A comprehensive picture of how languages encode (and speakers use) ad hoc categories may have theoretical implications for philosophical and psychological research on human rationality and conceptual processing, with potential applications for artificial intelligence and machine learning.