As far as I know and can research in 5 minutes, there isn't even a concise term for it in the computing world where it's still common in certain programming languages; it's just a "lowercase underscored identifier". You might hear of it as "Hungarian notation", but that term relates almost exclusively to the identification of the type and scope of a variable identifier in the name of the identifier itself; "li_local_integer" is Hungarian notation but so is "lintLocalInteger".
The use, obviously, is to avoid whitespace, which is in many languages a de facto division between code elements (keywords, identifiers, sometimes operators). While "camelCasing" and "PascalCasing" as conventions for similar situations have the eponymous terms, there isn't a single catchy term for identifiers that use underscores.
Perhaps "lower_score", "Cap_Score" and "UPPER_SCORE" could be introduced, but as the use of underscores to separate identifiers in actual code is a deprecated style in most of the popular languages, it's unlikely to catch on.
using_underscoresas opposed tocamelCase), Wikipedia calls it underscore-based (Wikipedia, "CamelCase," under "History->Computer Programming->The 'Lazy Programmer' theory") – zpletan Apr 2 '12 at 21:22