The ongoing sit-ins in front of parliament by enthusiasts of the Pakistan Tehreek-e-Insaaf (PTI) and Pakistan Awami Tehreek (PAT) and the way their leaders and the government ultimately rushed to GHQ to settle ‘issues’ have generated debate in the media, if not academia, about the state of relations between the civilians, i.e. parliamentarians, and the military. There are those who argue that the federal and Punjab governments led by the Sharifs have conceded political and strategic ground to the military by adopting a complacent and reconciliatory approach in terms of agreeing to the military’s bidding on domestic and foreign policy. Similar concerns are raised both by critics and supporters of the PTI and PAT about Imran Khan and Tahirul Qadri’s meeting with the Chief of Army Staff (COAS). The critics view this as part of the ‘script’ while supporters take it as an unnecessary development that provided the anti-PTI-PAT elements with enough material to dub them the king’s parties in the making. More importantly, there is also a tiny minority of intellectuals, including Dr Ayesha Siddiqa, that has termed the above-mentioned a “soft coup” whereby the military, familiar with abrogating the constitution, adopted a middle way in terms of pressuring the government through protesting ‘proxies’ to do political and other damage to the existing parliamentary dispensation. This author agrees with the notion of a soft coup but there is a need to highlight its ontological basis in existing major studies on civil-military relations. To begin with, sociological perspectives as presented by Huntington and Janowitz sought to look at the civil-military nexus from the societal rather than institutional dimension. Regarding the question of military intervention, these works heavily emphasised the importance of normative factors, i.e. “subjective civilian control”. However, Huntington introduced and applied the concept of “objective civilian control” and could not exclude the ideological factor to maintain a balanced relationship between the praetorian and the civil. More importantly, despite their pioneering position within the domain of the civil-military relations theory, such sociological perspectives were criticised even on empirical grounds. Comparative literature, on its part, tended to take a cross-cultural view of civil-military relations in particular and military intervention in general. The exponents of this literature (Finer, Nordlinger, Perlmutter) relied on cultural rather than purely institutional and political variables. Moreover, this literature, while aiming at a comparative analysis of different cross-national cases, seriously compromised the significance of context. For example, it is very difficult to understand how a contextually different case from Latin America can be compared with that of the Middle East. Similarly, from a structural (threat) perspective, Michael Desch has attempted to explain the intriguing question of civilian control over the military while using cross-cultural cases for this purpose. Nevertheless, despite the significance of this study, its assumptions could not be applied to the Pakistani case on empirical grounds. The same is the case with the above-mentioned sociological and comparative studies that have a limited scope in terms of explanatory potential and generalisability. More importantly, the studies produced on Pakistan’s politics in general and civil-military relations in particular, mostly fall into the sociological category. A couple of them, produced either by non-Pakistani authors or Pakistanis based in the west, give the impression of being comparative but methodological and conceptual considerations are overlooked. The most recent one, The Army and Democracy by Aqil Shah, suffers from the foregoing despite carrying large empirical details. Nevertheless, this work has arrived to make one understand the military’s mindset with regards to democracy and democratisation. In the view of this author, the fundamental question while explaining civil-military relations should be to address why the military intervenes in politics. Of course the ‘how’ question has to necessarily be taken into account since it guides the choice of scientific methods. The first part of the puzzle (the ‘why’ question) is indubitably about adopting a philosophical position that the existing accounts are not ignorant of. However, to my surprise, none of the existing research takes a non-structural and/or non-sociological view of politics, the state and civil-military relations in Pakistan. Hence, considering other ontological positions as important as the structure is in social sciences, this author prefers agency (to make things happen) to structure, due to the following. It accords top priority to the agency of actors; they are capable of making things happen. Sociological perspectives believe in the precedence of structure, thus underemphasising the significance of agency. After all, it is the actors’ agency that makes or breaks structure in the first place. The agency theory of civil-military relations assumes that the actors, both principal and agent, are rational. They have a clear concept of costs and benefits. They tend to maximise their benefits and minimise their costs during strategic interaction. The non-agency theories pay little attention to this fact and, instead, give more weight to factors such as culture, norms, values, etc. Moreover, the agency theory attempts to explain the puzzle central to civil-militantly relations, namely, the civilian control of the military. In this respect, it assumes actors’ interests as central to the puzzle. Their interests mould their set of preferences that, in turn, produce a set of actions. If the civilians are vigilant and are able to devise and implement an oversight mechanism, they enhance control in their favour in terms of winning over the preference of the military. If not, the military’s preference prevails and gets out of control. Thus, the agency theory is, in my view, capable of answering the puzzle. On the other hand, sociological literature accords more importance to exogenous variables whose impact, under complex situations, is difficult if not impossible to identify and measure. In other words, it provides a researcher with both theoretical and empirical tools to measure the patterns of interaction between the principal and agent and to explain and possibly predict the nature and direction of their relationship. Thus, it enables one to explain why crises and coups do and do not occur. If one applies the rationality oriented agency theory to the current political impasse in Pakistan, one may draw the following conclusions. First, the military, since 1977, stays a principal actor because its preference has prevailed in the realm of politics, economy, society, foreign policy and even media. Second, the agents, both the government and its opponents, ultimately chose to look up to GHQ as arbiter. This further helped the latter consolidate its principal-ship that was slightly contested during 2007-2013. Third, the military thought it irrational to opt for a hard coup since its policy choices seem to have been agreed to by the government. In sum, the military agency is not constrained. This shall remain so until civilians realise and work on their agency. The writer is a DAAD fellow. He holds a PhD in political science from Heidelberg University. He has authored Military Agency, Politics and the State in Pakistan (2013). Currently, he works as assistant professor at Iqra University, Islamabad. He tweets @ ejazbhatty