Andrew Jackson is criticizing Secretary of State Condoleeza Rice for referring to the border between Turkey and Kurdistan. Given one geographical and cultural entity that the name 'Kurdistan' refers to, he is right. Kurdistan is a region that is larger than the region within Iraq occupied by Kurdish people. Parts of it are in other countries, including most importantly a part that is in Turkey, and thus there could be no border between Turkey and that Kurdistan.
However, the term 'Kurdistan' is ambiguous, which Wikipedia's disambiguation page for the term demonstrates. One legitimate use of the term 'Kurdistan' is to refer to the province of Kurdistan in Iraq. One might argue that it would be politically inexpedient to refer to the border that way when talking to someone in Turkey, but she was speaking in the U.S. Congress about problems that in context very clearly had to do with Iraqi Kurdistan, not the larger Kurdish region. In context, then, what she said was not inaccurate and not an error. There is in fact a border between Turkey and the Kurdistan she was referring to, and anyone who knows the name of that semi-autonomous region in Iraq would have been able to register that she meant that Kurdistan rather than the larger, culturally-identified, non-political region that Andrew has in mind.
Update 3-3-07 10:54 pm: The State Department has issued a statement. Andrew describes it as follows: "Condi Rice backsteps and clarifies that northern Iraq is not Kurdistan, no matter what they want to call themselves." But what he links to sounds like the State Department is explaining what she meant and not taking it back. All it says is that she was referring to the region of Iraq that goes by that name. It's thus more akin to my defense of her statement than to an apology or retraction.
What I don't understand is the Turkish government's pretense that they don't know about this region of Iraq right next door to them that is going by this name. Their statement does what Andrew predicted they would do. It takes her statement to be referring to something that, in context, they should know better than to take it to refer to. It seems to me to be something like Canada saying that they won't recognize anything referring to their border with the Great Lakes region of the U.S., since the Great Lakes extend into Canada. Thus any statement about what might be called the Great Lakes region must include the Canadian Great Lakes region, even if the context makes it clear that it's about the Great Lakes region of the U.S. Such a policy of interpretation is linguistically indefensible. Terms that refer to regions can be context-sensitive, such that the region they refer to (and whether it is a political entity, a cultural group, a geographical region, etc.) can depend on the context of utterance.
Update (March 6): Andrew has a further comment on this discussion:
Jeremy emphasizes logic, but anyone who travels extensively in the Middle East knows that the Middle East is not logical. It is all local, pragmatic politics and PKK terrorism in Turkey’s specific case. Again, if Rice did not know the stand alone name “Kurdistan” has to the Turkish ears, then I can see why we are in the Iraqi mess we are in.
Well, I don't think we should cater to ignorance or to assumptions that have their foundation in immoral views about the Kurds. Irrationality may be at the root of the Turkish government's response to this, but that doesn't mean we should give in to such irrationality and insist on misusing language or refusing to recognize a region of a country they have no business interfering with. There are times to use language to bridge gaps, and there are times not to give in to evil by doing so. I think this is a case where the latter is the best response.