Seriously, some quick wiki'ing can introduce you to the problem.
- Department of State: "The term "terrorism" means premeditated, politically motivated violence perpetrated against noncombatant [including off-duty military] targets by subnational groups or clandestine agents, usually intended to influence an audience."
- Department of Defense: "Calculated use of unlawful violence to inculcate fear; intended to coerce or intimidate governments or societies in pursuit of goals that are generally political, religious, or ideological."
So of course the DoD's definition includes attacks on military, because then they're relevant to the war on terror. So then people who fight ISAF in pitched battles are terrorists whether or not they kill civilians. Except for the countries in ISAF where DoD's definition doesn't fly. It gets really fascinating when you look outside of NATO, though. I can't find it for the life of me online, but I believe Iran defines terrorism as "an act of Zionism" or something along those lines. There is no international, interstate, acceptance of An Issue of Terrorism. Not even close. And if you wanted to bet that the UN's definitions have been lawyered beyond all comprehension, you'd be right.
The use of an amorphous definition to fight an amorphous problem allows repressive regimes plenty of room to operate. I'm not as well-versed in Central Asian sub-state actors yet, but the old Turkish Hezbollah was, by many smart folks, believed to be a government creation to fight the PKK. And, I mean, there's the whole history of the ISI. And though that stuff gets conspiracy-theorist in a hurry, the very existence of dark corners is a function of this vague and confused definition of terrorism.
I mentioned the terrorism degree before, and it stands to say that the folks I got it from also put out the Global Terrorism Database, which is a terrific amount of fun. If you're interested, you can download an excel file of all their data, which allows folks like us to crowdsource terrorism research. All of this is good, good, stuff.
At the same time, there are some gaps to fill. Gaps that exist because of the aforementioned amorphousness of the issue the GTD is there to study. And I hate to harp on Andijan, but I view it as a litmus test of someone's knowledgability of Central Asia: If you know about the events and the questions surrounding it, there's a good chance you're pretty knowledgable about the rest of the issues surrounding Central Asia (or at least are in a good position to learn more). And the GTD? Well, it don't know Andijan:
Thousands of armed protestors stormed a jail in an effort to free 23 members of the Akramia religious group in Andijan, Uzbekistan. At least nine people were killed and numerous others were injured in the incident. The government of Uzbekistan believed that the Akamia religious group was in contact with the outlawed, radical Islamic party Hizb-ut-Tahrir. No group claimed responsibility for the attack.So here we go again. It's not GTD's fault, per se. The point of the database is to create a set of datapoints, and this datapoint has the right date, the right location, and all of that. There's just a stark lack of sincerity about the underlying issue. This weird dichotomy exists between the "With us or against us" belief and the "Let countries figure out how to deal with their own problems" I'm honestly not sure how Karimov landed on a side, Kadyrov landed on a side, and all of that.
The tools to study exist, but there is no consensus on what to study. The whole inchoate beginning of Terrorism Studies as a field relies on an ability to define terrorism like one can define history, anthropology, or geology. Withouth that, it's just political posturing and think tanks. I'm not sure what the better solution is.