Attn:
Discovery News
BBC News
et al.
We all know how obsessed you are with the oldest of anything. But please, stop. You are doing a grave disservice to languages by saying things like ‘A tribal language thought to have existed for 65,000 years has disappeared forever’ or ‘Languages in the Andamans are thought to originate from Africa. Some may be 70,000 years old.’ This is utter nonsense, even if you find a scholar to tell you otherwise. All languages are always changing, and although some may change more rapidly than others, the idea that a language could persist essentially unchanged for multiple millennia is pure bunk. The idea that we can assign exact ages to languages (other than recent inventions) is even more ridiculous. The only ancient languages are the ones actually spoken in the past, and all currently spoken languages have equally long histories. It is a great tragedy that the Bo language has gone extinct, as it is when, every other week or so, another language goes extinct on this planet. It is a tragedy regardless of how long the language has been spoken, because it represents the end of a particular part of the modern world’s cultural diversity. Your attempt to sensationalize this story by exoticizing indigenous peoples as primitives lost in time is unwelcome and counterproductive. Let me help:
The Bo (Aka-Bo) language was a member of the Northern branch of the Greater Andamanese language subfamily. With its extinction, only one Greater Andamanese language, A-Pucikwar, has any remaining known speakers, and it is highly endangered. The ten Great Andamanese and three South Andamanese languages are all related to one another, although the exact relationships among them remain unclear, but there is no known relationship between the Andamanese languages and any other languages of the world. Their importance for linguistics is that they may represent descendants of the languages of the original migrants to the Andaman Islands many millennia ago, and if we were able to reconstruct the Proto-Andamanese language, potentially to better understand the population and migration history of the Indian Ocean. Their importance for their remaining speakers is inestimably greater.
Glossographia 
You say:
‘The ten Great Andamanese and three South Andamanese languages are all related to one another’
Can you provide a source for that? I thought there had been virtually no serious historical/comparative work on these languages, largely due to lack of data and the huge problem of separating out inheritance from borrowing. There is Blevins (2007), which claims to provide evidence of proto-Ongan, a postulated ancestor to Jarawa, Onge and sibling to proto-Austronesian, but most researchers feel it’s highly dubious.
I agree with your feelings about the misleading statements about Bo being so many 1000’s of years old, but I think they’re trying to say that the language lineage has been isolated for that long. There is no word that I know of for ‘a language and all of its ancestors back to when it was a common ancestor to something else’-maybe we could talk about the Andaman clade being so many 1000’s of years old (if they do form a clade).
The only source I know is the Blevins 2007, which you mention. However, my understanding is that it was generally agreed that the two branches were related – otherwise, don’t we have two 70,000 year old language families hanging around on the Andamans? In any case, I absolutely agree that the issue of limited data coupled with the issue of borrowing makes establishing specifics very difficult.
With regard to the ‘ancient’ label, I do think that any linguist would understand what is meant, but any layperson or journalist is not likely to reach that conclusion. For that matter, the 70,000 figure is basically made up – there is no linguistic evidence whatsoever to support it. Given the frequently racist history of the discipline and the assumption that ‘primitives’ are unchanging relics of deep prehistory, we (as specialists) need to address the problem.
As you say, we don’t know how long they’ve been there, it could be much less than 70,000 years. And yes, while positing two language families seems less likely than one, we don’t know, so we can’t say. I think it’s fair enough to say it’s possible that they’re related, but the work hasn’t been done to demonstrate it (and probably won’t be). Re this issue Blevins (2007) says:
‘…there remains no persuasive evidence that of a family relationship between Jarawa-Onge and the Great Andaman languages.’
Re the ‘ancient’ business, as a linguist I find that kind of thing very frustrating. All the more so because I find so many fellow specialists (eg in anthroplogy, archaeology, sociology…) have no more knowledge about language than non-specialists, and that includes ideas such as this ‘ancient language’ business.