Pothipath, Vipas

2018-05-14T10:15:30Z

2018-05-14T10:15:30Z

2008

The principles of economy and distinctness in language seem to be key selection pressures for language evolution. Accordingly, to express the exact number of things, humans might be expected to use cardinal numeral-noun constructions (CNNCs) consisting of just two constituents, namely a noun (N) representing quantified things and a cardinal numeral (NUM) representing the number of the quantified things (for example, English three sheep). However, the structural patterns of CNNCs used in a number of languages spoken today are not that simple, and have seemingly redundant constituents, typically non-singular markers (NSG) and numeral classifiers (CLF). CNNCs observed in the world's languages also appear to show a diversity of structural patterns despite the fact that simple structures like English three sheep seem very practical. This observation brings up two related major goals of this thesis. The first is to reveal structural types of cardinal numeral-noun constructions of singularity (CNNCsg) and cardinal numeral-noun constructions of non-singularity (CNNCnsg)- The other major goal is to hypothesize a possible evolutionary scenario for CNNCs since their emergence till the modern era.

This thesis approaches these two issues by exploring CNNCs in 241 languages representing 101 language groups (i.e. language families, language isolates and pidgins and Creoles) across the globe through reference grammars to ensure the greatest range of possible attested structural patterns of CNNCs. This cross-linguistic survey demonstrates that, with regard to CNNCsg, the world's languages are divided into two major types, namely {N,NUM} and {N,NUM,CLF} with relatively few other possibilities. In relation to CNNCnsg, the world's languages are divided into four major types, namely {N,NUM}, {N,NUM,NSG}, {N,NUM,CLF} and a mixture of {N,NUM} and {N,NUM,NSG} with some other less common possibilities. The historical origins of these structural types are then investigated, using evidence from old written records together with theoretical approaches, especially grammaticalization. Finally, it is found that the various structural patterns of CNNCs discovered can be considered in the light of a hypothetical evolutionary ladder. Hence, with cross-linguistic comparison integrated with diachronic approaches, hypothesized evolutionary trajectories of CNNCs are postulated. It is conjectured that the construction consisting of a noun plus a word with a numerical interpretation such as the words meaning 'alone' or 'company' may represent a possible initial stage of CNNCs. From that stage onwards, CNNCs have split into many types over time. The development is reversible in terms of structural complexity, and idiosyncratic in some cases. Besides, the contributory factors in the development of CNNCs involve a quantifying function, a non-quantifying function, and a mixture of both.

Based on the study of evolution of CNNCs, this thesis also discusses the nature of language by comparing language change with biological evolution in some major aspects. The comparison suggests that language is strikingly similar to biological organisms in general, perhaps rather than to other cultural artefacts. Overall, this thesis contributes to current studies of the complexity and diversity of human language(s)

http://hdl.handle.net/1842/29947

The University of Edinburgh

Annexe Thesis Digitisation Project 2018 Block 18

Already catalogued

Typology and Evolution of Cardinal Numeral-Noun Constructions

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy