On the website Big Think, John Horgan argues that war is just a cultural practice that humankind could eventually abandon, unless we keep infecting ourselves with the "war virus." If one state is infected by proneness to war, his argument runs, its neighbors may have no choice but to adopt similar measures to avoid being conquered.

In Horgan's words, as reported by Mark Cheney, "Imagine your neighbor is a violent psychopath who is out for blood and land. ... You would have few options but to embrace the ways of war for defense. So essentially your neighbor has infected you with war."

It's an arresting use of language, perhaps, but social Darwinism should have taught us to be wary of misplaced biological analogies in the study of world politics. Viral infections spread by specific, well-known mechanisms - e.g., they take over the DNA of cells and replicate themselves - and that's not remotely like the mechanism Horgan is identifying.

He's actually describing a situation in which an external threat forces the leaders of neighboring states to rationally adopt policies and strategies to ensure their survival. That's not how viruses spread: You don't catch a cold because you've decided the only way to protect yourself against your sneezing neighbor is to start sneezing with him.

The logic Horgan is pointing to here is the "security dilemma" that realists have been talking about ever since John Herz, an international relations scholar. In a world with no institution to protect states from each other, each is responsible for its own security. Because states cannot know each other's intentions with certainty, they have to prepare for the possibility that their neighbors may do something nasty. So they invest in armed forces or look for powerful allies, especially if they think the possibility of trouble is high. And once they do that, others have to worry about them in turn.

This is the "tragedy" of great-power politics identified by the political scientist John Mearsheimer, and it's a much better explanation for security competition - and war - than an analogy to microbes.

To be fair, Horgan's larger point is simply that war is not a biological necessity; it is a specific political or cultural response to certain conditions and thus, in theory, could be gradually abandoned. This theme has been developed at length by authors John Mueller and Steven Pinker.

I agree with Pinker that overall human violence has declined significantly over several centuries, but I remain agnostic about the larger claims of a long-term reduction in international violence.

That trend is driven almost entirely by the absence of great-power warfare since 1945, which may have multiple, overlapping causes (bipolarity, nuclear weapons, the spread of democracy, etc.) whose persistence is hard to forecast.

The absence of great-power warfare is good, because it can do the greatest harm. What we're seeing instead are protracted conflicts among warlords, insurgents, and weak states (think Sudan or Colombia), and wars of choice waged by the United States and other powerful states in strategic backwaters, mostly against adversaries that we don't think can do much in response. At least, we hope not.

Stephen M. Walt is a professor of international relations at Harvard University. He wrote this for Foreign Policy.