Based on diachronic data extracted chiefly from the available lexicographic sources and historical corpora of Polish, this paper aims at determining whether the initial stage of the adverbialization of indefinite quantifiers of nominal origin typically involves extent modification, degree modification being a posterior development. The results of an investigation into the evolution of the functional status of the commonly used quantifiers trochę 'a bit' , odrobinę 'a bit' , as well as masę 'a lot' indicate that prior to establishing themselves as degree modifiers, the items function as extent modifiers, i.e. duratives or frequentatives. In their earliest adverbial attestations recorded in the analysed material, the quantifiers at issue modify the duration or frequency of the action denoted by the associated verbal element, or, if the pertinent verb encodes a punctual event, the duration of the resultant state, and only later do they start to combine with scalar predicates, i.e. degree verbs as well as gradable adjectives and adverbs, including adverbials in the form of prepositional phrases. Exceptional in this respect is masę 'a lot' , as it (still) appears incapable of serving as a degree intensifier.

While the attachment of diminutive morphology to concrete nouns, gradable adjectives and adverbs, as well as interjections has already received a well-merited share of attention in Polish, diminutivization of vague quantifiers remains empirically understudied. The present paper takes a first step towards filling in this gap by reporting on a corpus-based investigation of the numeralized partitive garść ‘handful’ and its diminutive variant Garstka ‘handful.dim’. The results of a collocational analysis of both forms corroborate the hypothesis that diminutivization further enhances scalar implications inherent in the base ‘small size’ item, as reflected in the diminutive form’s significantly higher frequency of quantifier attestations. Apart from exhibiting a substantially greater proportion of quantifier uses, the latter element displays an overwhelming predilection for animate N2-collocates, which suggests that diminutivization may not only intensify a paucal quantifier’s expressivity but also lead to conspicuous changes in its distributional profile.

The aim of this paper is to contrast the near-synonymous Polish classifiers kupa ‘heap’, sterta ‘pile’, and stos ‘stack’, all of which encode upward-oriented arrangements of objects or substances and thus prototypically combine with concrete inanimate nouns, by means of a collocational analysis conducted on naturally-occurring data derived from the Na­tional Corpus of Polish. The results of the empirical investigation point to a tendency for kupa ‘heap’ to combine predominantly with mass nouns denoting amorphous, frequently natural, stuff, whereas sterta ‘pile’ and stos ‘stack’ exhibit a pronounced predilection for count N2-collocates referring to artefacts. In a similar vein, while both sterta ‘pile’ and stos ‘stack’ typically stand for aggregates formed by a volitional human agent, it is not infrequent for kupa ‘heap’ to classify portions of substances whose shape is a result of the forces of nature or merely constitutes a by-effect of activities intended to achieve goals other than arranging stuff into units. What differentiates between sterta ‘pile’ and stos ‘stack’, however, is that constructional solidity appears a more salient feature of the latter item, hence its capability of applying to vertical collections of entities marked by an orderly internal structure.

On the basis of corpus-derived data, the present paper examines the collocational patterns of the singular and the plural forms of a pair of etymologically and semantically related quantifying nouns (QNs), namely English heap and its Polish equivalent kupa 'heap'. The primary aim is to determine their respective levels of numeralization, operationalized as the frequency of co-occurrence with animate and abstract N2-collocates in purely quantifi cational uses, in an attempt to establish whether, and to what extent, the addition of the plurality morpheme bears on the grammaticalization of a nominal of this kind into an indefi nite quantifi er. Following the observations arrived at by Brems (2003Brems ( , 2011, the hypothesis is that pluralization should yield a facilitating effect on the numeralization of nouns referring to large quantities by amplifying their inherent scalar implications. The results demonstrate that whereas heaps indeed exhibits a higher percentage of such numeralized uses than heap, kupy 'heaps' has turned out to be grammaticalized in the quantifying function to a markedly lesser degree than kupa 'heap'. It is argued that this apparently aberrant behaviour of kupy 'heaps' can nonetheless be elucidated in terms of the specifi city of numeralization in Polish, since at its advanced, morphosyntactic stage, the process in question affects solely the singular (accusative) forms of QNs.

While the grammaticalization of English size nouns into vague quantifiers has already received a considerable amount of scholarly attention, their subsequent syntactic expansion beyond the nominal domain remains an under-researched area. In particular, little has hitherto been written about the possible factors contributing to the emergence of additional adverbial uses of such items. Based on synchronic corpus data, this paper therefore aims to partially fill in this gap by providing an analysis of the adverbialization patterns of nine nominal forms of this kind, namely bit, scrap, shred, heap, heaps, load, loads, lot, and lots, whose empirical tokens have been classified into six categories: (i) verbal inherent modification, (ii) verbal extent modification, (iii) adverbial ambiguous, (iv) object-pronominal, (v) adjectival modification of positives, and (vi) adjectival modification of comparatives. The results demonstrate that in the verbal domain, most of the analyzed forms reveal a preference for pronominal uses, in which they function as an argument of the verb rather than a genuine degree adverb, while in the adjectival domain, a majority of the items, especially 'large size' nouns, exhibit a conspicuous propensity to combine with the comparative forms of adjectives/adverbs. Moreover, it is shown that there exists a strong positive correlation between the items' respective degrees of grammaticalization in the quantifier function and their extents of adverbialization, operationalized as the proportion of pertinent attestations in corpus samples. Thus, the study underscores the role of frequency in grammaticalization on the one hand, and points to the importance of paradigmatic analogy on the other.