The way you tell if a country like Iran is actively working on a nuclear bomb is that it diverts uranium to weapons purposes. Iran has not done that, as the IAEA repeatedly affirms. Almost certainly, if Iran were seriously working on a bomb, it would kick international inspectors out altogether. ... It is likely that Iran wants “nuclear latency,” or the “Japan option.” That would involve knowing how to construct a bomb in short order if the country was ever directly menaced with an invasion and regime change a la Iraq.If Latour's philosophy is as strictly actualist as Graham Harman claims in Prince of Networks then it seems that any ANT constructed on this basis cannot really understand the above. (Indeed, given that Harman embraces a rather strict actualism I'm not sure if he could adequately articulate the above in his terms.)
If Cole is correct and Iran are not seeking a functioning atomic bomb but simply a latent capacity to build a bomb then the bare actuality of their project misses that the goal of the process is virtual. Iran is apparently constructing not a coherent material object but rather a virtual capacity to bring into existence a coherent material object (that is dependent upon a pre-existing arrangement of many objects, knowledges, etc.).
If there is any principle that is fundamental to Latour's work (at least as Harman articulates it) it is that if something makes something else happen then it is an actor. It doesn't matter if that actor is fictitious, material - whatever. Moreover, to act is to exist; no action, no existence. (c.f. Nietzsche.)
Should it really make a difference whether the actor is actual? This is where things get complicated as actual derives from the Latin actus meaning simply 'act.' What then is the relation of the virtual to the actual if to exist is to act and virtual things apparently can act?
(Perhaps it's the opposition of actual and virtual that is the problem here.)
I'm not sure I'm in a position even to formulate this question properly.