Identifying cultural-specific pattern - Githubissues

clics / clicsbp

CLDF dataset on Body Part Colexifications

Creative Commons Attribution 4.0 International

1 stars 0 forks source link

Identifying cultural-specific pattern #34

Closed AnnikaTjuka closed 1 year ago

AnnikaTjuka commented 2 years ago

The goal is to explain the outliers that do not confirm to the adjacency hypothesis, i.e., colexifications with non-adjacent body parts and non-colexifications of adjacent body parts.

Data:

Restrict the data set to the most well-attested concepts across languages and a restricted number of diverse language families (i.e., 15 language families)
Only data with orthographic profile

Output:

Matrix with the information on whether a colexification is attested in a language or not (also used for #21)
Labels for shape and function

Colexification	Attested (0/1)	Shape	Function	Language
ANKLE-WRIST	1	0	1	German
FINGER-TOE	1	1	0	German

Analysis:

Power analysis (#21): Which colexifications and languages can we integrate to make the networks as comparable as possible?
Language family as a predictor
Clusters should arise based on language families
Pairwise distance measure to compare the similarity of the networks (dendrogram + heat map)
Preferences of language families for colexifications based on shape versus function

Figure:

Dendrogram + heat map
Body part outline (either with #24 or graphics department)

LinguList commented 1 year ago

Here, I do not see where our extended "cognate set" approach comes in: if we label in a language N that the community recurs there, we do not offer the information on whether the colexification belongs to a cluster X or a cluster Y inside the language family.