Who Owns Language?
This may be partly a case where a procedural violation -- the Mapuche were not consulted directly by Microsoft in the process of producing the Mapudungan versions of the software -- is being fought on the territory of the substantive outcome. But I think there's also something to the substantive case. (Indeed, here it's difficult to disentangle the two, since the Mapuche's objection seems to be not so much to the very idea of a Mapudungan version of Word as it is to Microsoft making a Mapudungan version of Word.)
Defenders of Microsoft make both deontological and consequentialist claims. Deontologically, they point out that the idea of group ownership of language is absurd within our Western system. The usual rebuttal is to argue that rights (or at least some rights, of which property rights would be the clearest case) are culturally relative. I'm more interested here in the consequentialist case* -- how does Microsoft making a Mapudungan version of Word hurt the Mapuche? Or more generally, how does an outsider's use of an element of a culture harm insiders?
My answer depends on three main concepts: structuration, diversity of values, and power. Structuration refers to the fact that social institutions evolve through use. A language is thus not a fixed object that can be picked up, used, and put back the way it was. The popular descriptivist position in linguistics -- words and grammatical structures mean whatever people use them to mean -- is a correct structurationist position**. By diversity of values, I mean that different people have different ideas about what society should be like. Ceteris paribus, it's better for a given person's values to be realized than not. One's pursuit of those values will be constrained by the available institutions, but they will also shape how one uses those institutions, and hence what those institutions look like when one is done with them.
Finally, power refers to the fact that different people and groups have different abilities to reshape institutions in accordance with their use of them. Problems arise when inequalities of power align with (real or potential) differences in values. The minority (in power, and often numerically as well) then finds their ability to achieve their values limited, because they have limited influence over the social instutions available to them. This is a problem even when the difference is merely numerical -- while it may be fair in each instance taken in isolation for the larger group to get its way, when taken as a whole a persistent minority will end up getting outvoted every time. The solution here is autonomy -- to separate the institutions used by the majority and the minority, so that the majority's use of their version does not affect the version used by the minority. This goes some way toward explaining the emergence of subcultures and the fierce defense of existing cultural diversity.
Thus, when Microsoft makes a Spanish version of Word, it's little threat to most of the Spanish-speaking community for two reasons. On the one hand, Microsoft's values with respect to the Spanish language are not likely to be that divergent from those of most Spanish-speakers. Second, Microsoft's power vis-a-vis the 400 million Spanish speakers is comparatively limited -- indeed, Microsoft is largely at the mercy of the general public's usage and the prounouncements of Spanish grammarians. But both of those factors tilt against the Mapuche. It's far more likely that Microsoft will have different values from the Mapuche, and it's reasonable for an oppressed group to be especially suspicious of one of the world's biggest corporations on this count. And Microsoft's power to define Mapudungan is greatly exaggerated vis-a-vis a small and disempowered group like the Mapuche. Thus group rights to language sovereignty (and by similar arguments, rights to sovereignty over other cultural products) are absurd in the case of "big" languages like English and Spanish, but may be a legitimate defense mechanism in the case of "small" languages like Mapudungan.
Bringing the question of power into the discussion, however, raises yet another difficult problem -- establishing the legitimacy of the Mapuche's desire to limit outsiders' use of their language. We want to let the Mapuche decide when and how their language may be used, rather than presume to decide on their behalf what would be good for them. But this presents us with a Scylla and Charybdis situation. On the one hand, in recognition of our own limited understanding of the situation and our disproportionate power, we want to avoid an imperialistic use of our own ideals of legitimacy to judge claims made by Mapuche individuals or groups. But on the other hand, we also want to avoid a naive assumption that Mapuche views on this issue are internally uncontested or that we can treat the traditional leadership of the tribe as legitimately speaking for everyone (assuming we can even rely on our own understanding of what that traditional distribution of authority is).
*Since I think any assertion of a right must have an underlying consequentialist justification, though of course the relativistic argument could be used to deny the relevance of the sort of utilitarianism I'll apply.
**Though this should not be taken to the extreme of denying the validity of debate over the proper use of language. While Platonic sorts of arguments about the transcendental correctness of certain meanings are invalid, pragmatic arguments -- "we should use these words in this way because it allows us to make certain useful distinctions" -- are still fine. Indeed, such pragmatic arguments merely articulate what structuration tells us will be happening inevitably.