Should marijuana be legal, for either medical or recreational use? I think the best initial answer to this is: It’s a crummy question! We are good at those.

It’s a crummy question, because it calls for answers based on unsubstantiated opinion. Answering it does not invoke or even encourage any relevant evidence, or precedent. So what would a better question be? How about: On what basis should any particular substance be legal for either medical or recreational use?

There are many advantages to this new question. For one thing, since it has no emotionally-charged word like “marijuana” in it, it avoids provocation. It invites thought, rather than potentially thoughtless passion, the proverbial knee-jerk response. For another, it almost requires consideration of whether, and how, the question has been answered already. For instance, alcohol and tobacco are currently legal for recreational use: Why? What criteria pertain in these cases?

As for medical use: Benzodiazepines, such as Valium, are legal. Not only are these drugs potentially addictive, but they are among the few addictive substances (along with alcohol) from which withdrawal can be lethal. Benzos are incomparably more dangerous than marijuana. Why are these drugs legal and in current use? What are the relevant criteria?

Similarly, Dilaudid — a synthetic version of morphine, related to it and heroin, and many times more potent than either, is legal and in current use. Cocaine is legal and in current use in every hospital emergency room. Why? What are the relevant criteria? Asking whether or not marijuana might have legitimate medical application is a lot less provocative once we’ve conceded that cocaine already does. (It is used, by the way, in a dilute solution to treat severe nosebleeds.)

Should assisted suicide be legal? This is another poor question, inviting nothing but emotional responses and perhaps some religious moralizing. A better question is: Are there any circumstances in which allowing death to occur represents the best means of alleviating pain and preserving dignity? We should wrestle with that one — thinking of ourselves or the person we love best in the world in the hot seat — first. Then, we might constructively move on to: What, exactly, do we mean by “allowing”?

Even such words as “marijuana” and “suicide” are tepid in comparison to “abortion,” a topic I broach only rarely, with caution and some degree of trepidation. But even where passions are most inflamed, we might turn down the heat by asking better questions. What evidence-based approaches most reliably reduce the frequency of abortions, legal or otherwise, in any given society? Since reducing the demand for abortions is good for all concerned, we might constructively start a discussion there — and might manage to avoid calling one another names or throwing things.

A recent, and already notorious, meta-analysis by Katherine Flegal and colleagues at the CDC suggests that death rates do not necessarily rise or fall with body weight. Asking, in reaction to this — is obesity important after all? — is misguided and silly. Good questions are: Who were the thin and heavy people with higher and lower death rates? Is extra body weight sometimes helpful — and if so, when? Is extra body weight sometimes harmful — and if so, when? We in fact have answers to all of these questions — but they are lost in a haze of hyperbolic nonsense if we fail to ask them.

And then, there is the vexing issue of the lingering, post-Newtown moment: What of that Constitutionally-protected right to bear arms? Asking this question, or any question remotely like it, is a surefire way to get everybody reaching for their respective triggers. It, and all variations on the theme of asking what the Founders really meant, may be diverting for Constitutional scholars — but it’s mostly a boondoggle for the rest of us.

The exact words of the Second Amendment are: “A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.”

What are good questions that derive from this? Two occur to me, and they are as obvious as they are simple: Which people? And, what arms?

I trust we can agree, from the most devout pacifist to the NRA leadership, that the Second Amendment cannot possibly mean all people. It can’t mean prison inmates, incarcerated for violent crime. It probably can’t mean felons on parole after violent crime, either. It can’t mean people committed in a psychiatric ward for paranoid schizophrenia. It can’t mean 2-year-olds. I don’t think any of this is even remotely controversial.

But if “the people” does not, and cannot, mean all people, and if the Founders did not further specify which people — then that is a question we are obligated to ask and answer. Which people? Once we agree it requires an answer, a potentially constructive dialogue ensues.

And, similarly, what arms? Those who feel the Founders anticipated our era of Bushmasters must concede that if so, they envisioned nuclear weapons as well. The Founders either were omniscient and prescient, or they weren’t. If they were, and they did envision nuclear weapons, why didn’t they say: arms except nuclear weapons? Did they mean individuals should have a right to nuclear arms? Is this a case even the NRA wants to make?

If not, the Founders left it to us to determine what arms made sense. So that becomes a good question: What arms do make sense? We don’t have to answer it. We just have to recognize the legitimacy of the question.

Again, I suspect that across the spectrum of opinion here we can agree that individual citizens should not possess a biological weapon, such as smallpox, capable of wiping out the entire population indiscriminately. Individual citizens should not have nuclear missile launch codes. No need to go on, the point is clear. The right to bear arms can’t possibly mean all arms. So we are invited to ask: What arms? Again, the question could lead to constructive dialogue unencumbered of hostility.

When I write about the factors contributing to epidemic obesity and ill health, as I often do, I routinely get pushback. Those who think everything is a matter of personal responsibility — that we would all be thin and healthy if we weren’t lazy gluttons — consider me a member of the nanny-state food police. Those who believe in the extreme of environmental determinism aren’t shy about calling me a food industry apologist, when I suggest that our food supply may have something to do with demand.

But the barrage of resistance and reproach from those who don’t like my answers leads me perennially back to the same conclusion. I think we are wonderfully adept as a culture — at asking all the wrong questions.

Epidemiology should trump ideology — mine, as well as yours. For our efforts to do good for real people in the real world, they have to be based on the actual evidence of what our actions do to and for actual people in the real world — not hypothetical abstractions born of hazy hope or morbid fantasy.

Gertrude Stein famously told us: “A difference, to be a difference, must make a difference.” Data-driven public policy would address what differences our differences of opinion actually make, and thereby give us a far better platform for action, based on better answers.

But, of course, our only hope of moving in that promising direction begins — with asking the right questions.

David L. Katz is the founding director of Yale University’s Prevention Research Center.