Doctors, psychologists, ethicists and others, along with our society at large, debate whether “the ends justify the means.” But nobody debates whether “the means justify the ends.” There is no point even looking for an answer to a question that is patently silly. For now, just hold that thought, please.

Medical ethics can be very challenging. There are the difficulties of interpreting “do everything” in desperate situations where heroic effort is on one side of a line, and futility on the other. There are the challenges of doing “no harm,” while taking great risks. There are challenges of optimal resource allocations for the greater good. There are challenges related to tradeoffs between beneficial and adverse effects, particularly with high-risk surgical and critical care procedures. In this context, the question of ends justifying means comes up routinely.

Perhaps the most vivid and obvious illustration is any variation on the theme of euthanasia. Those who believe it is the work of medical practice to protect life view all such variants as wrong, if not anathema. If, however, the work of medicine is to preserve dignity, and autonomy — the case for assisted dying can be made, at least under narrowly-defined circumstances. It can be a case where the ends — relief from suffering, death with dignity — might justify the means.

The question has far-ranging implications for the whole field of ethics. One school of thought, for example, is that whatever achieves the greatest good for the greatest number is “right.” This is referred to as utilitarianism, and while few real-world ethicists espouse it in pure form, they do invoke its principles.

The extreme contrary view, deontology, stipulates that some things are wrong just because they are wrong — no matter what effects they exert. Again, the pure practice of this probably doesn’t exist, but it informs the “do ends justify the means” debate.

Psychology experiments famously reframe the “ends versus means” debate by presenting a scenario where a great deal of good can be done, such as saving a whole group of people, but only by doing intentional harm — such as killing an individual.

There are good reasons why the debate endures, and is to some extent insoluble. There may be no single right answer.

But again, no one wrestles with the reciprocal question — “do the means justify the ends?” — and with good reason. If you are getting bad outcomes, what point could there possibly be in “justifying” the means that lead to the ends you don’t want?

In a world where means are used to justify ends, there might be means to treat the nausea of pregnancy. For those affected by it, those would be welcome means, indeed. And for those with more severe forms of pregnancy-related nausea and vomiting, they might even be truly important means.

But, as has in fact proven true in the past, those means might produce serious unintended consequences, in the form of birth defects. In a world that sensibly asks “do ends justify means?” while just as sensibly avoiding “do means justify ends?” the response to this is rather obvious. Doing what seems like a good idea stops being a good idea when it produces bad outcomes. A treatment for pregnancy-related nausea that produces common, serious birth defects would not be justifiable. The abandonment of thalidomide for this purpose demonstrates that this is not just hypothetical. In the real world, bad ends unjustify well-intended means.

And now we come to the reason for this ramble. My hope, if not quite my belief, is that we might constructively look at the vexing issue of gun control through this same lens. We do so, of course, in the immediate aftermath of background checks failing to make it through the Senate.

The roiling debate about the Second Amendment seems to hinge on where one places one’s semantic emphasis. Those opposed to any regulation of gun sales emphasize “shall not be infringed.” Proponents of gun control emphasize the subordination of that clause to “a well regulated Militia.”

I have opinions about this, and you may as well, but since we are unlikely to resolve any differences of opinion about the language of the amendment here and now, let’s not try.

Rather, let’s consider this: The language of the amendment, however it is interpreted, is about means. Some manner of access to guns for some portion of the citizenry is the means, and something like defense against tyranny and protection of liberty the presumably intended ends.

Clearly, the ends could justify the means. If more guns of all kinds freely accessible to all meant more liberty, more security, less risk of tyranny — then the means might well be justified, and the fuss would end.

But the means cannot justify bad ends. If the consequences of interpreting the Founders’ means one way are ill and unintended, such as the massacre of schoolchildren without better protection of liberty of defense against tyrrany, then the means — whatever their original intentions — are subject to reconsideration, no less than thalidomide. It in no way tramples the rights of pregnant women to have their nausea treated when we abandon a drug that causes birth defects. Bad ends, however unintended, unjustify means, however well-intended.

We might better confront the gun control debate with data, gathered in a non-partisan manner, about the ends we are getting. We could make a systematic effort to look for all potential good, and all potential bad, ensuing from the status quo just propagated on the floors of the U.S. Senate. If we don’t even look for such data, it implies someone doesn’t want to know the ends we are getting, and that is an always ominous sign of ulterior motives and cowardice. We must know the effects of our actions to be qualified judges of our conduct.

Whether ends justify means will remain, in particular contexts, a legitimate and challenging debate for the foreseeable future. But in a world where means justify ends, and unintended consequences don’t matter — the very concept of ethics has met a very mean end already.

David L. Katz is the founding director, Yale-Griffin Prevention Research Center.