Last year a friend, on a major national standards committee, wrote to me regularly complaining that it was difficult getting the things (information, documentation) he needed out of Microsoft on Office Open XML, and trying to figure out effective strategies for this. My advise to him was to first rid himself of any idea that Microsoft thought people exist: there are corporations or organizations in interesting classes, or nothing. So personalized attention would only come by first getting on their radar somehow as one head of a larger Hydra that warranted their attention. (I see Wikipedia calls Hydra anancient nameless serpent-like chthonic water beast: great!)
I think this is something that is pretty much common to most large US corporations, and indeed, to the organizational behaviour of people in most large institutions. As a theory it does explain a lot: why for example Microsoft would be so resistant to fixing bugs that did not effect new sales.
My problem, if it is a problem, is that I don't basically don't believe in the existence of institutions and corporations, only of people. So when I hear people talking about "elephants in the room" (meaning large abstractions) I tend to see the elephants as Javanese shadow puppets being carted around by individuals.
While I don't remotely agree with Maggie Thatcher's conclusion that There's no such thing as entitlement, unless someone has first met an obligation, at a fairly basic level I see what she means that There is no such thing as society. However, corporations and other ghosts do have a kind of existence (e.g. existence in the ideal), and it is of course useful shorthand to give in to the fiction and not lift the corporate veil to see the individuals behind.
But it is difficult, for me at least. I don't think I am being original at all in pointing out that giving abstractions a reality is a prerequisite of bigotry, demonization and idealization (in the day-to-day sense.) But what can we do?
Perhaps the problem is that English orthography, like XML, does not have enough tokens to represent the semantic status of nouns (especially proper nouns) for fictional/conceptual things. Perhaps we need delimiters for abstractions and fictional beings: so "Shock jock" Rex Flex said '"Liberals" want to sell "the country" to "bin Ladenites"'.
But would it always make things clearer? What about a sentence like '"IBM" disavowed that the opinions of its bloggers were anything to do with its marketing policy' for example? Where the implication was that it was really "A spokesperson for corporate officers of IBM disavowed.." therefore only containing only people.
The flood of blogs this week, with titles like Is Google becoming evil? perplex me: if "Google" does not exist, then this really is, simplistically, Are the decision makers at Google becoming evil? But more reasonably the writer is probably meaning something more like Are the decision makers at Google switching over to making decisions which will have effects that I regard as evil? Once you start to couch things in terms of the people involved, it seems that many times simplistic sentences are revealed being based on lots of tacit assumptions. The danger of these kinds of simple seeming sentences is that they can get used as if they were simple statements of fact, used in syllogisms etc.
On the other hand, bringing things down to people can also mean that prejudices that were safely deflected onto fictional beings can get targeted to people. So perhaps it might be too extreme to ban fictional beings, even legal fictions, and abstractions from discourse entirely.