Volume 52, Issue s39-s1 - Historica. Special Issue: Linguistic strategies for the construction of ad hoc categories: synchronic and diachronic perspectives, Guest Editors: Caterina Mauri and Andrea Sansò

Ad hoc categorization is the bottom-up abstraction of a category starting from concrete exemplars of the category itself. When we observe linguistic data, we find various phenomena that provide evidence for the ubiquity of such an on-line, goal-driven and context-dependent categorization in everyday communication. Beyond offering concept labels in the form of words, language indeed provides speakers with a great number of strategies to convey reference to a class by naming representative individuals. After providing a semantic and pragmatic account of ad hoc categorization in terms of indexicality, we will survey ad hoc categorization strategies in discourse and across languages: they can be syntactic (lists, general extenders, exemplifying constructions), morphological (heterogeneous plurals, collectives, aggregates, compounds), or in-between (reduplication). We will argue that all these strategies show a similar abstract structure consisting in a categorization trigger, that is, some prosodic, morphological or syntactic element triggering the abstractive inferential process towards the category identification, plus a linguistic expression referring to some overt category member, which is processed as the starting point for abstraction. The diachronic connections between these strategies and the pathways leading to their emergence and conventionalization also speak in favor of their unified treatment.

The aim of this paper is to examine the role of exemplification in categorization processes, that is, how examples can be used in discourse to communicate conceptual categories. Based on data from present-day Japanese and a corpus-driven methodology, it will be shown that exemplifying constructions can be used 1) to refine already explicit categories by contextualizing and actualizing the reference, and 2) to create categories ex novo by triggering associative inferences and abstractive processes. Accordingly, a detailed analysis of the linguistic properties of the examples will be provided in order to identify recurring encoding patterns and correlations with the functions described above. Furthermore, it will be argued that, although any conceptual category can be lexicalized by means of a category label, there exist some interesting correlations between the coding of the examples and the labelling of the category. Finally, we will conclude by showing that the linguistic analysis of exemplification can provide useful insights regarding the modalities in which the human brain categorizes.

The aim of this paper is to study the categorizing function of exemplification in Italian. Based on a corpus study, it will be shown that the strategies used to exemplify are mobilized from a variety of different syntactic levels, ranging from small clauses to lexemes. Though differentiated, exemplifying constructions are all characterized by the same function, that is, to signal that the expressions in their scope are examples and that they have to be processed as triggers of a higher-level category. Exemplifying constructions signal the non-referential status of what is within their scope and presuppose non-exhaustivity, namely, the existence of further alternatives which are contextually associable with the mentioned exemplars. We will first provide a semantic analysis of the constructions under examination (in terms of semantic specificity of examples), and then move on to the distributional features (in terms of the realis/irrealis status of exemplification contexts and the kind of speech act) and textual features (in terms of topic continuity of exemplification). In the conclusions, we will propose the outline of a discourse profile of exemplification as it has emerged from the analysis of corpus data.

Russian, as well as many other languages, displays a wide number of strategies to name non-lexicalized categories, that is, categories which have a conceptual reality, but are not encoded as single lexical items (e.g. fruit, furniture ) or set phrases (e.g. equipment for sport, food for dog ). One way to address this lack of lexical means is to provide one or more concrete exemplars of the intended category, in an attempt to capture the superordinate non-lexicalized concept. This paper aims to provide an overview of the different generalizing means used in Russian for the online construction of exemplar-driven categories, and, at the same time, to explore the interplay between discourse-resources (e.g. exemplification, listing and reformulation) and system-resources (e.g. collective affixation, co-compounding and discourse markers) in the process of categorization. Particular attention will be focused on the description of schematic general extenders, that is, partially lexicalized syntactic patterns, realizing the different extending, generalizing and similative functions of general extenders. The last part of the paper is devoted to a discussion of the processes of grammaticalization and pragmaticalization of the general extender i vse takoe ‘and so on’ (lit. ‘and all such’). The form is interesting because of its discursive multifunctionality.

Alternative relations are one of the main relation types between states of affairs and disjunctions are the dedicated morphosyntactic constructions used for encoding them. The standard approach to disjunctions, which assumes two readings, Inclusive and Exclusive, has recently been challenged. A fundamental requirement, shared by the alternatives in all the readings they identified, is that they must be construable as members of a single higher-level category. This requirement also defines specifically one of the readings of disjunctions, or or constructions, HLC (the Higher-level category reading), in which the explicit disjuncts serve only as a means for constructing, often only ad-hoc, a higher-level concept. This concept, rather than the explicit alternatives, is actually the prominent concept profiled by the construction. The study presented here adopts the same bottom-up Usage-Based approach advocated by Ariel and Mauri, and examines the Hebrew or constructions of the Old Testament, given in the Masoretic Tiberian version. It focusses on the question of how those categories are constructed. The data was analyzed adopting a comparative linguistic and textual method, and taking into consideration different philological and diachronic aspects. It reveals that HLC (such as ‘any beast’ for ox or ass , Exodus 23, 4) is the most common reading in this corpus (108/319, 34%). Interestingly, the higher-level categories are extracted from disjuncts of various parts of speech and can be explicit, but in most cases they are inferred. Furthermore, we observe cases which almost equally profile both the single, higher-level category and the alternatives. The methodology adopted revealed some linguistic changes in progress, as well as inclinations of ancient biblical editors and jurists.

Taxonomic nouns such as Spanish tipo appear in constructions that create ad hoc categories, typically in scientific and academic contexts. Some of them develop further uses, notably an adverbial use as an approximator or hedge, as a result of a reduction of the original construction. The aim of this contribution is to trace the origins of constructions that establish ad hoc categories in the case of Spanish tipo , and in particular the role of genitival constructions and apposition in technological, scientific and generally academic texts in the nineteenth century on the basis of corpora. I will further study the subsequent rise of the adverbial approximating functions used for ad hoc categorization and related pragmatic functions such as the use of tipo as a quotative marker in Argentinian youth language. On this basis, two broader issues will be discussed: (i) the relation between ad hoc categories and ad hoc categorization and a possibly different status of ad hoc categories and ad hoc categorization due to proceduralization; and (ii) the striking parallels with cognates from certain Romance languages (Portuguese and Italian) and from Russian and Swedish, as well as the question of whether there has been a parallel, independent development—and if so, why—or whether contact phenomena explain these similarities.

The aim of this paper is to investigate the genre -construction, in which the French noun genre connects a noun phrase (named X) with another constituent (a noun phrase in most cases; named Y). Such a construction has an exemplifying function, and, traditionally, authors consider Y to be an example of a category denoted by X. I show that this traditional analysis is not entirely satisfying as it does not give a comprehensive account of the construction’s semantics. Instead, I show through a wide variety of examples that the genre -construction provides a means of expressing ad hoc categorization. Rather than merely exemplifying a category denoted by X, the construction frequently builds up an ad hoc category based on X and examples provided by Y. This analysis can be further extended to other related constructions so as to give a unified account of such constructions. The data analysed consist of internet users’ messages collected from online forums. The analyses are carried out within the framework of construction grammar.

The expression of ad hoc categories ranges from discourse-level to syntactic and morphological strategies. Considering derivation in particular, it has been observed that also collective suffixes can be used for the identification of context-dependent sets. The aim of this paper is to investigate the relation between collectives and ad hoc categories by focusing on the Italian suffix - aglia and by discussing its diachronic relation with the Latin neuter nouns ending in - ālia . The discussion concerning the notion of collective will take into account a recent proposal that distinguishes between collective nouns, aggregate nouns, and superordinates. As will be shown, aggregate nouns are of particular interest for interpreting some Latin pluralia tantum in - ālia that denote sets of heterogeneous entities with similar properties. The same possibility to derive aggregate nouns characterizes the suffix - aglia that, when attached to proper nouns, can also convey associative meanings, or denote categories including persons and situations that share the named exemplar as a common denominator. As the analysis will show, the development of these functions correlates with the mechanisms of (inter)subjectification.

This paper provides a new description of nominal compounds featuring the noun ādi- ‘beginning’ in Sanskrit. When used as second member of a compound, ādi- is commonly translated as ‘etcetera’ and the resulting compound displays a number of semantic and morphological oddities, as observed in reference grammars. Based on a thorough corpus analysis, we show that ādi- compounds are in fact associated with different functions, and that among other things they can be employed for the task of on-line categorization. Specifically, we argue that they can also encode ad hoc categories. In the second part of the paper, we describe the diachronic process whereby ādi- compounds developed into ad hoc category markers. In so doing, we adopt the perspective of constructionalization and suggest, against the traditional analysis of these forms as compounds, that when meaning ‘etcetera’, ādi- behaves as an affixoid. In this respect, the findings of this paper contribute to the general understanding of the diachronic typology of ad hoc category markers.