How is lexical similarity really computed?
Looking at the following 3 excerpts from Ethnologue:
http://www.ethnologue.com/show_language.asp?code=fslLexical similarity: 43% with American Sign Language [ase] in an 872-word list.
http://www.ethnologue.com/show_language.asp?code=aseLexical similarity: 57% between modern ASL and French Sign Language (LSF) [fsl] on a comparison of 872 signs. Although the 2 are related, ASL has undergone substantial creolization (Woodward 1975, 1976).
http://www.ethnologue.com/ethno_docs/introduction.asp#dialectLexical similarity. The percentage of lexical similarity between two linguistic varieties is determined by comparing a set of standardized wordlists and counting those forms that show similarity in both form and meaning. Percentages higher than 85% usually indicate a speech variant that is likely a dialect of the language with which it is being compared. Unlike intelligibility, lexical similarity is bidirectional or reciprocal.
So, why the 43% and 57% figures? Is one an error, or (since 57% is 100%-43%) actually giving lexical difference? Or is there another reason why the figures are different, such as using different wordlists, or comparing different dialects?
http://www.ethnologue.com/show_language.asp?code=fslLexical similarity: 43% with American Sign Language [ase] in an 872-word list.
http://www.ethnologue.com/show_language.asp?code=aseLexical similarity: 57% between modern ASL and French Sign Language (LSF) [fsl] on a comparison of 872 signs. Although the 2 are related, ASL has undergone substantial creolization (Woodward 1975, 1976).
http://www.ethnologue.com/ethno_docs/introduction.asp#dialectLexical similarity. The percentage of lexical similarity between two linguistic varieties is determined by comparing a set of standardized wordlists and counting those forms that show similarity in both form and meaning. Percentages higher than 85% usually indicate a speech variant that is likely a dialect of the language with which it is being compared. Unlike intelligibility, lexical similarity is bidirectional or reciprocal.
So, why the 43% and 57% figures? Is one an error, or (since 57% is 100%-43%) actually giving lexical difference? Or is there another reason why the figures are different, such as using different wordlists, or comparing different dialects?