By now everyone has seen an advertisement or two for Rosetta Stone, the leading language software out there today. I've gotten the chance to play with three or four of the languages at varying levels of difficulty, enough to form an educated opinion. As a user, I found it intuitive, fun, and visually appealing. The speech recognition software is pretty awesome (although you could basically watermelon your way through the Russian stuff), and most of Rosetta's competitors can't hold a candle to that. But as a linguist, I have a few negative comments.
The operating philosophy of Rosetta Stone is that the best way to learn a language is the way children learn them; immersion using basic sentences and sparse grammar instructions. It seems miraculous, doesn't it, how children learn to speak fluently by the age of six or so without any real help from their environment other than exposure time. Children are the best language learners out there, so they should be our models, right?
Dead wrong. Children learn languages so easily because of the primary language acquisition process, and there is a time limit to it; adults will not soak in complex grammar (grammar is your biggest hurdle) patterns in the same way children do and while learning as an academic subject is boring, awkward, and difficult for many, it's ultimately our best option. We have to tailor our programs not just to what works best theoretically, but what works best practically and for the targeted demographic.
And the idea of teaching examples without grammar works well for some languages. Swahili, for example, isn't grammatically complicated. Someone could learn Swahili as an adult without too many grammar woes. Highly inflected languages, however, are not so easily learned (or acquired) by native speakers of analytic languages like English; we need that grammar background to understand the underlying mechanisms that make sentences work. We have more language 'settings' to change in our minds when languages are more different, and that requires more intensive study (and dedication).
So, as an end result of Rosetta Stone, you will be able to parrot very basic sentences that express minimal abstract thought. A middle-aged language learner wanting to pick up Chinese before they set sail for Hong Kong so they can order in a restaurant or ask for directions may think that's all he needs: basic sentences to say basic things. But unfortunately, languages don't keep simple topics simple grammatically. In fact, since we use them everyday, verbs like "to be", "to know", "to eat", or "to have" are very often the most difficult and irregular.
Verdict: Rosetta Stone is a very cool, very expensive toy. Whether or not NASA uses it is of no interest to me; that doesn't speak to it's effectiveness, only to how much money NASA is willing to spend on bullshit.