WASHINGTON — Just what exactly is the military?

On one level, this question has an obvious answer. “The military” is “the armed forces,” which in this country essentially means the active-duty Army, Navy, Air Force and Marines, together with their reserves and the National Guard. (Yes, yes, under certain circumstances the Coast Guard could be considered part of the military, and then there’s the Merchant Marine, and the Public Health Service, and even a bunch of uniformed officers with commissions from the National Oceanic and Atmospheric Administration — did you know that? — but let’s keep it simple for now.)

Sticking with the obvious, if we know who’s in the military, then presumably we know what the military is: the military is what it does. In other words, military functions are those functions performed by members of the military.

Military analysts refer to the ratio of combat versus non-combat troops as the “tooth to tail” ratio. In 2007, the Army’s Combat Studies Institute published a fascinating study by John McGrath, who found that the U.S. military’s tooth-to-tail ratio has declined substantially over the last century.

During World War I, for instance, the United States initially fielded about twice as many combat troops as support troops, for a 2-to-1 tooth-to-tail ratio. By 1945, as World War II wound down, that had changed; only about 40 percent of troops in the European theater were combat troops, while the rest were headquarters, administrative, logistics and support troops of varying kinds (giving a tooth-to-tail ratio of roughly 2-to-3). By 1953 — in Korea — the tooth-to-tail ration was 1-to-3. By the 1991 Gulf War, it was even lower: McGrath estimates it as 1-to-3.3. During the Iraq War, the ratio of combat to non-combat troops deployed ticked up slightly, but primarily as a function of the increased use of civilian contractors.

McGrath — himself a retired Army Reserve officer — concludes that “combat elements have progressively declined as a proportion of the total force since 1945.” And “[A]s the percentage of combat troops deployed declines, it raises the question of whether such a deployment is, in fact, a military deployment at all, or some other type of operation.”

That’s a vital question.

Go back to my initial query: Just what is the military?

Let’s complicate matters some more. McGrath’s important study defined combat troops not by whether troops actually engaged in combat, but by rather by job description: thus, for instance, he counts as combat troops all “company size and above units of infantry, armor, cavalry, field artillery, air defense, artillery, attack and assault aviation and combat engineers … special operations forces” and so on.

But in Iraq and Afghanistan, those combat troops spent a great deal of their time engaged in activities far removed from combat. They engaged the enemy when needed (the high casualty rates for troops in combat-related military occupational specialties make this painfully clear), but also found themselves doing everything from building schools to encouraging women’s participation in economic activity.

The stated rationale for such seemingly not-very-militaryish activities was clear: to “win” in Iraq and Afghanistan, the United States needed to win the hearts and minds of the population. As Lt. Gen. William Caldwell put it in a 2008 Military Review article, “The future is not one of major battles and engagements fought by armies on battlefields devoid of population . . . victory will be measured in far different terms than the wars of our past. The allegiance, trust and confidence of populations will be the final arbiters of success.”

In a world in which critical threats to U.S. national security may come from airline passengers armed only with box cutters, from cyberspace or from a virus deliberately transmitted, it’s inevitable — and necessary — that our troops will spend more and more time on activities that don’t much resemble traditional forms of combat.

Complicating matters even more, the decline in the military’s tooth-to-tail ratio has been paralleled by a rise in civilian organizations (public and private) engaging in what look suspiciously like traditional military activities. The CIA has gone kinetic, for instance, with paramilitary forces that engage in direct action, often working hand in hand with military special operations forces. And for-profit private military companies increasingly place civilian contractors in jobs that resemble combat positions in all but name.

All this leads me to echo McGrath’s question: When is a military deployment not a military deployment? Or: when does a military stop being a military?

There aren’t just academic questions. Whether (and how much) the civilian-military gap matters depends greatly on how we categorize what the military is doing. In fact, much of what we think we know about how to run our military — how to sustain it and constrain it, how to divvy up roles and missions, funding and authorities between the military and other entities — depends on our ability to know what it is that we mean when use the term “the military.”

If “the military” increasingly performs civilian functions, for instance, then maybe it doesn’t matter that much if the State Department has fewer resources — maybe our focus should just be on ensuring that the military performs those formerly civilian functions well. Conversely, if civilian entities such as the CIA perform “military” functions, then maybe we need to rethink how we hold the CIA accountable for its activities, which are far less transparent than those of the military. More generally, how do we make sense of civil-military relations — and civilian control of the military — when the boundaries between the “civilian” and “military” categories grow ever more blurry?

In a recent guest post on Tom Ricks’ Foreign Policy blog, Mackubin Thomas Owens wrote, “The line between military and civilian is not impermeable. Success in national security requires that civilians have an ongoing say in military affairs,” while “the military has to be at the policy and strategy table” as well.

That’s wise advice. But if we can’t define “military affairs” with any clarity, or reliably distinguish it from “policy” or “strategy,” can we act on it?

Rosa Brooks is a law professor at Georgetown University.