0
A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems
Richard Sproat, 2014. Language.
Sproat re-tests the entropy-based measures that Rao and Lee used to argue that the Indus signs and Pictish symbols are writing, on a wider range of linguistic and non-linguistic corpora. He finds the published methods do not work as stated; the two measures that do separate the classes put both the Indus and Pictish systems on the non-linguistic side.
What it's used for here
Three methods surveys cite it as the main caution against treating entropy or Zipf statistics as proof that a symbol system encodes language.
Cited by on this site 3 records
3 methods surveys
SRCSignal processing, linguistics, inductive methods
Cite it
Sproat, R. (2014). A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems. Language. https://doi.org/10.1353/lan.2014.0031
@article{sproat2014statistical,
author = {Sproat, Richard},
title = {A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems},
journal = {Language},
year = {2014},
doi = {10.1353/lan.2014.0031},
url = {https://doi.org/10.1353/lan.2014.0031}
}TY - JOUR TI - A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems AU - Sproat, Richard PY - 2014 JO - Language DO - 10.1353/lan.2014.0031 UR - https://doi.org/10.1353/lan.2014.0031 ER -
Sproat, Richard. “A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems.” Language (2014). https://doi.org/10.1353/lan.2014.0031.