Productivity in Italian word formation: a variable-corpus approach

Abstract

The quantitative approach to morphological productivity first proposed by Baayen crucially refers to the relation between the number of hapax legomena formed with a given a‰x occurring in a su‰ciently large corpus and the total number of tokens of that a‰x sampled in the corpus. Most criticism against this measure focuses on its neglecting the role played by frequency in the evaluation of productivity. As an improvement of Baayen's procedure, a variable-corpus approach is proposed. Accordingly, the productivity values should be calculated at equal token numbers for di¤erent a‰xes instead of taking the di¤erent token numbers which result from sampling the whole corpus for all a‰xes, as in Baayen's works. This implies that variably-sized subcorpora must be sampled to compare a‰xes displaying di¤erent frequencies. On the basis of a 75-million-token newspaper corpus, the productivity values for several Italian a‰xes in the deverbal and deadjectival domain are calculated. The resulting rank proves linguistically plausible, avoiding the overestimation of productivity for low-frequency affixes typically occurring in fixed-corpus calculations. As a further advantage, the procedure proposed here makes it possible to deal satisfactorily with two problematic aspects usually neglected in previous investigations, namely, the quantitative impact of (i) allomorphies and lexicalizations and (ii) inner-cycle derivations on productivity measures.

On the double nature of productivity in inflectional morphology

“…As also observed by Baayen and Renouf (1996), the daily issues of a newspaper are an optimal test bed to investigate productivity exactly because of its increasing character. The corpus worked out in Gaeta and Ricca (2002, 2006) is structured in 36 text chunks corresponding to the monthly issues of the newspaper, so that they can be computed separately. This allows us to extract data for affixes displaying different frequencies computing different text chunks.…”

“…Gaeta and Ricca 2002, 2006, the P-index was calculated for a number of Italian suffixes which basically belong to the deverbal and to the deadjectival domain:…”

“…In particular, the so-called variable-corpus approach developed in Gaeta and Ricca (2006) on the basis of the original procedure suggested by Baayen (1989) will be shown to provide an optimal frame for investigating the productivity of inflectional morphology. Relying on data taken from the verbal morphology of a richly inflecting language like Italian, it will be shown that the variable-corpus approach to productivity can shed light on this intricate question by concretely revealing the double nature of inflectional morphology.…”

Faut-il brûler les dictionnaires ? Ou comment les ressources numériques ont révolutionné les recherches en morphologie

“…La plupart du temps, les ressources utilisées pour cette tâche de comptage sont des archives de journaux (cf. Baayen & Lieber, 1991 ;Baayen & Renouf, 1995 ;Gaeta & Ricca, 2006 ;Grabar et alii, 2006), mais il pourrait aussi s'agir de corpus réalisés à partir de la Toile (cf. § 3.3.2.).…”

How to Live Naturally and not be Bothered by Economy

“…A good example is again provided by the Italian nominalizing suffi xes -mento and -(z)ione (for a more detailed analysis, see Gaeta 2002a: 66-77;2005a). The fi rst one displays a much lower type (and token) frequency than the second, but it is more productive (at least relying on a quantitative measure of productivity such as the one proposed by Baayen 2001 and revised by Gaeta & Ricca 2006). It does not strike us as a surprise that -mento derivatives also are fairly more transparent than -(z)ione formations.…”