Researching the origins of counting systems, I came across the question I cannot seem to find an answer for: what is the typological distribution of languages that consider thumb a finger (5 fingers on hand) vs. those that do not (4 fingers, thumb is not a finger). English is in the latter category, although since recently, due to its status as a de facto international language, it seems to have become more fluid on this.
Lexically, there seems to be not a clear-cut distinction among even language families. Danish (Germanic) has tommelfinger for the thumb, and Russian (Slavic) has большой палец bol'shoy palec, “great finger”, while Slovenian (also Slavic) uses the same root palec for “thumb,” and has a non-cognate prst for “finger.” Romance languages seem to prefer a separate root for the thumb, after Latin, but again, Romanian has deget mare, “great finger” for the thumb, making it a “thumb-unaware” language. There are signs, such as remnants of the dual declension of Russian numerals for two to four and Latin octo, and the cognacy of nine and new, suggesting that PIE placed more weight on counting in fours and not fives when its counting developed (a relation to the number of fingers seems likely to be a factor), but the modern daughter languages fluctuate between 4 and 5 rather non-systematically.
Of course, having a separate root lexeme for the thumb may be tangential to finger counting. It is possible that of “thumb-aware” languages, some count 5 fingers, despite having a separate root for the thumb, and some do 4. Cf. English has pinky but it is still unquestionably a finger. This is the main, the most interesting question of my research: how many fingers languages count on a hand.
Unfortunately, WALS does not have this category. There are distributions for languages that lexically distinguish hand and finger (topics 130A and B), but this monumental atlas seems shy on semantic topics in general, so this question may be even out of their scope.
Can anyone help with references to the research on the typological classification of languages into “4-finger” and “5-finger” categories? A wide, methodologically consistent review would be the best (WALS is a great example). As I am dealing with the development of counting, references to reconstructed and ancient languages would also be very helpful.
Splitting the latter (5-finger) category into those having and not having a root lexeme for thumb is secondary and less important to me currently, but I believe, for the general nature of SE questions, would also lie in the scope. This should be related in any case.