In today's NY Times, there is a piece on Tiananmen Square [see also Slate]--amazing that it has already been 20 years. The pictures of the man in front of the tank remind me of the central question when the military ends up playing a role in the domestic politics of the society--will the military use force and against whom? China differed from the transitions going on at the same time in Eastern Europe as its military used force against its people. The Romanian "revolution" had some violence, but the military ultimately sided with the people against the Securitate and the Ceauşescus regime.
If the Chinese military had stood aside, things may have turned out differently. The problem for dissenters and for governments is that it is never really clear what the military, or, to be more precise, enough of the relevant elements, will do at the moment of decision. There is plenty of academic work on coups, including the factors associated with coup attempts, coup-proofing strategies, and the like. Indeed, coups were the original focus for much of the literature on civil-military relations.
I am still a relative new-comer to civil-military relations, but, thus far, I have not been convinced by any one argument or by the body of research that we know much about this key tipping point.
Of course, this raises another question--when do we want a military to intervene? Would the loss of civilian control of the Chinese military in 1989 have been a good thing? I am always reminded of the commander of the Soviet strategic rocket forces calling the mobile missiles back to base in the 1991 coup to prevent them getting into the wrong hands (I think I am remembering this right) even though he had no authority to do so. So, the good news is that this guy made the right call for securing the weapons. But the bad news is that he acted without any authorization--loss of control. We got lucky, I suppose.
And then there is Pakistan.