Following Cavalli-Sforza, we call these units “semes” rather than “memes” (Dawkins 1976, Durham 1991, Boyd and Richerson 1985) because “seme” comes from “sign” and emphasizes the symbolic nature of culture.
Semes were apparently first mentioned in a 1970 book by Roland Barthes called "S/Z" - a three hundred page dissection of a short story written in French.
However, they were subsequently adopted by Cavalli-Sforza, Marc Feldman and others. Feldman explains the term in a 2007 presentation here. He presents "semes" as the cultural equivalent of "genes".
Marcus Feldman said in 2014 (16 minutes in):
I call them semes because I don't like memes, memes gives come impression of biology which I do not want to do and semes comes from semiotics.
Of course, culture is part of biology, so biologically-inspired terminology is completely appropriate.
I think "semes" are more-or-less dead terminology now. Not all culture is symbolic, so the death of "semes" seems to be no great loss. However, IMO, it is fascinating historical tit-bit that Cavalli-Sforza and Marc Feldman join all the other students of cultural evolution that publicly toyed around with "-eme" words.