Bananas are an intriguing fruit. Quick growing, and tall, tropical herbs, rather than real trees, known pretty much everywhere today from the most temperate climes, as a typical and inexpensive table fruit, while in other places they serve as starchy staple alongside or even instead of tubers or cereals (and as the base for beer-brewing). Because most cultivated bananas are seedless hybrids tracking them archaeologically is difficult. Various lines of archaeological, linguistic and ethnobotanical evidence were pulled together a couple of years ago in a journal issue, while this year saw some updated syntheses, or at least attempts at synthesis. The origins of cultivated bananas seems to have focused on New Guinea, and the First Act can be regarded as the dispersal throughout most of tropical Asia. A PNAS article this summer, with 18 co-authors, across botany, linguistics and archaeology provided a model of integrating genetics and historical linguistics (with rather more limited archaeology) for tracking the early evolution, diversification of bananas, mainly in SE Asia [pdf]. I still have issues with how the hybridization between A and B genomes took place. The authors postulate an anthropogenic dispersal of M. balbisiana (B), and do not really deal with any potential role of South Asian balbisiana in hybrid bananas. Both parts of India (Orissa through Assam) and Sri Lanka have wild Musa balbisiana; Sri Langa has reports of wild M. acuminata too. It seems clear that Pleistocene (to early Holocene) humans in Sri Lanka were using and probably consuming wild bananas. In addition to the seeds from Beli-lena cave reported years ago by Kajale (in Harris and Hillman's volume Foraging and Farming). To this can be added new phytolith finds, in well-dated stratigraphic context from Batadomba-lena, published in Journal of Human Evolution this past summer. Ongoing phytolith analyses from other Sri Lanka caves, including some work here in London, will have more ancient Musa phytoliths to report soon. It may well be that wild Musa use in Sri Lanka was a dead-end with regards to early cultivars, but it seems premature to rule it out entirely.
Second act: the introduction of bananas and plantains to Africa. Probably the best general account of this is still to be found in the 1999 article by De Langhe and De Maret, and it was also addressed in several of the papers in the 2009 Ethnobotany Research and Applications issue on bananas. But for a summary that places this in the wider context of the translocation of crops, weeds and commensal animals across the Indian ocean, from Asia to Africa see the paper I jointly authored with some other members of the Sealinks project, published in Antiquity this past summer, " Across the Indian Ocean: the prehistoric movement of plants and animals." The evidential lynch-pin for the a pehistoric/Iron Age translocation of bananas (or plantains) remains a single site in Cameroun with reported phytolths, Nkang. This limited evidence, of course, until or unless more is found may be open to critique-- which has been coming from some quarters of Africa archaeobotany. For the latest installment see the recent Neumann et al article in Quanternary International. A short blog here.
Third act: Gobailization via refigerator vessels and 20th century AD supermarket culture. For some account of the modern technology involved in the mass shipping and then ripening of supermarket bananas, specifically in New York city, see this recent blog at Edible Geography.The book Banana: the fate of the fruit that changed the World by Dan Koeppel, deals with this and much, much more; and I have discovered he has his own banana blog.