new methods from the evaluation study

[LinguList]

There are three more functions to be added officially to lingrex now, for a paper accepted with minor modifications.

• a function to compute cross-semantic cognate statistics
• a function to compute "normal" cognates from salient cognates (annotated with a column for morpheme glosses)
• a function to compute b-cubed f-scores to assess the degree of variation resulting from strict vs. loose cognate coding

[LinguList]

This has been addressed, but we could update the documentation.