Lexicostatistics away from the armchair: handling people, props and problems

Abstract

In the last decade or so, historical linguistics has witnessed a renewed interest in the application of quantitative techniques to comparing and classifying languages. Among these quantitative techniques, approaches of a lexicostatistical nature enjoy particular popularity, possibly due to the relative ease of manipulating lexical data, and the convenience of a ‘fixed’ list with predefined meaning slots. However, the issue of how exactly these slots should be filled in the first instance has been curiously ignored. An attempt is made here to remedy this shortcoming by proposing an explicit data-collection strategy, which advocates fieldwork with several native speakers. This novel methodology is described in some depth, and the results from a pilot study are presented. These reveal considerable variation when lexicostatistical data are obtained from multiple speakers.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 96 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.