On the Development of Grammatical Number

Abstract

Almost all cultures around the world exploit language resources to construe experiences of phenomena in quantity terms (e.g., cardinals, grammatical number) yet the evolutionary processes underlying the development of grammatical (singular, dual, and trial) and lexical number in language is understudied. This study tests the hypothesis that grammatical number developed from cardinal number through a case study of dual marking in the Para Nyungan language family. Factors that motivate the development of dual grammatical number are examined. Number, as a category of grammar, describes the regular encodings of our experiences of phenomena in exact quantity terms (as singular, dual, trial, paucal and plural values) through morphological, syntactic, and/or phonetic (e.g., tones, clitics) configurations. Beyond grammatical number, the majority of the languages of the world exploit lexical number — such as cardinals (e.g., one, two, three, four) and ordinals (e.g., first, tenth) — for exact quantification. Some studies hypothesise that the dual, trial, and paucal grammatical number values evolve from the cardinal numbers two, three, and four (Corbett, 2000; Aikhenvald, 2018) with evidence from language families (Austronesian and Sino Tibetan) where the number features are often marked. Yet, the crucial role of shared descent in influencing such a development (Macklin & Round, 2022) is not central in these reports since examples are extracted from only a few of the languages that mark these grammatical number features. The hypothesis also assumes that lexical number (specifically, cardinals) evolves prior to grammatical number. The evidence that plurality (Grammatical number) is marked in the 907 sampled languages with lexical numbers (high limit of cardinals) but not the reverse seems to support the conclusion that lexical number precedes the development of grammatical number (Overman, 2021). The evolution of grammatical number and cardinals in languages can however be studied independently. And the results from such independent examination of evolution could shed light on the relation between these two ways of quantifying as well as socio-cultural factors underline their relation. The present study tackles the biases in sampling by focusing on language families where the grammatical number features of interest dominate. The data for the study has been taken from Grambank and Numeralbank, global databases respectively on the grammatical and numeral systems of languages worldwide. The Atlantic Congo dominates (with 222 (52%) of the 425 languages) families marking singular grammatical number on nouns (Grambank). Austronesian dominates (204 languages - 33%) the 616 languages that mark dual grammatical number on pronouns. Pama-Nyungan has the majority of languages marking dual on nouns (25% — 26 of 184 languages). Lastly, the trial grammatical number the least frequently marked grammatical number (8 out of 2,214 languages mark trial number) is dominant in Austronesian (with 4 languages). The study focuses examines the evolutionary time for cardinal numbers and dual marking in the Pama Nyungan language family. The numerals of these languages were extracted from numeralbank. Example 1: Development of dual marking in the Yanyuwa clade LANGUAGE DUAL MARKER CARDINAL TWO SOURCE OF INFLUENCE Yanyuwa _wiya, rri_ Kanymarda, Binary system: 3 = 2 +1, 4 = 2+2 Anyirarra Coordination: marda [Also, too, and] Dual relates to animacy Warluwarra _wiya kutya, guca Borrowed from Yanyuwa Barlarnu _wiya gutyiya, guciya Borrowed from Yanyuwa Yindjilandi _wiya ɡulharra ɡutharra Borrowed from Yanyuwa The results from the preliminary phylogentic reconstruction of cardinals and dual markers in Pama-Nyungan languages shows that (1) cardinals developed first and (2) dual grammatical number developed much latter. The cardinal number two usually grammaticalises as dual in the languages which mark dual number. Socio-cultural practices, including counting practices and objects counted, suggests that the grammaticalisation process is influenced by cultural factors. In other words, the findings shows that the development of dual grammatical number from the cardinal two is motivated by cultural practices. The example (1) shows that the independent development of dual in the Yanyuwa clade was motivated by the binary numeral system, the conception of addition ("Marda" [and] a clip from Kanymarda [two] used as coordinator) and the animate objects that were counted (animacy effect). The remaining languages in the clade borrowed the dual marker from Yanyuwa. The results raises questions about the possibilities for languages to mark grammatical number beyond five given similar sociocultural practices. While the presence of birth order naming seems a positive evidence of such grammatical number development, the feature is less popular in the family for any strong conclusions to be drawn. Moreover, the few instances of quantifying birth order on nouns of the languages, the birth order labels did not exceed the numeral limit of the language family. This preliminary results shows that grammatical number marking in Pama Nyungan develops from cardinal numbers and has implication for the evolution of grammatical number, the cognitive scientific theory of subitizing and the evolution of cardinals.