The American opioid epidemic claimed 63,600 lives in 2016 alone. While the public policy challenge is daunting, the problem isn’t that we lack any effective treatment options. The data shows that we could save many lives by expanding medication-assisted treatments and adopting harm reduction policies like needle exchange programs. Yet neither of these policies has been widely embraced.

Why? Because these treatments are seen as indulging an addict’s weakness rather than “curing” it. Methadone and buprenorphine, the most effective medication-assisted treatments, are “crutches,” in the words of felony treatment court judge Frank Gulotta Jr.; they are “just substituting one opioid for another,” according to former Health and Human Services Secretary Tom Price.

And as county Commissioner Rodney Fish voted to block a needle exchange program in Lawrence County, Indiana, he quoted the Bible: “If my people … shall humble themselves … and turn from their wicked ways; then will I hear from heaven, and will forgive their sin.”

Most of us have been trained to use more forgiving language when talking about addiction. We call it a disease. We say that people with addiction should be helped, not blamed. But deep down, many of us still have trouble avoiding the thought that they could stop using if they just tried harder.

Surely would do better in their situation, we think to ourselves. We may not endorse the idea — we may think it is flat-out wrong — but there’s a part of us that can’t help but see addiction as a symptom of weak character and bad judgment.

Latent or explicit, the view of addiction as a moral failure is doing real damage. The stigma against addiction is “the single biggest reason America is failing in its response to the opioid epidemic,” Vox’s German Lopez concluded after a year of reporting on the crisis. To overcome this stigma, we need to first understand it. Why is it so easy to see addiction as a sign of flawed character?

We tend to view addiction as a moral failure because we are in the grip of a simple but misleading answer to one of the oldest questions of philosophy: Do people always do what they think is best? In other words, do our actions always reflect our beliefs and values? When someone with addiction chooses to take drugs, does this show us what she truly cares about — or might something more complicated be going on?

These questions are not merely academic: Lives depend on where we come down. The stigma against addiction owes its stubborn tenacity to a specific, and flawed, philosophical view of the mind, a misconception so seductive that it ensnared Socrates in the fifth century BC.

Do our actions always reflect our preferences?

In a dialogue called the Protagoras, Plato describes a debate between Socrates and a popular teacher named (wait for it) Protagoras. At one point their discussion turns to the topic of what the Greeks called akrasia: acting against one’s best judgment.

Akrasia is a fancy name for an all-too-common experience. I know I should go to the gym, but I watch Netflix instead. You know you’ll enjoy dinner more if you stop eating the bottomless chips, but you keep munching nevertheless.

This disconnect between judgment and action is made all the more vivid by addiction. Here’s the testimony of one person with addiction, reported in Maia Szalavitz’s book Unbroken Brain: “I can remember many, many times driving down to the projects telling myself, ‘You don’t want to do this! You don’t want to do this!’ But I’d do it anyway.”

As pervasive as the experience of akrasia is, Socrates thought it didn’t make sense. I may think I value exercise more than TV, but, assuming no one is pressuring me, my behavior reveals that when it comes down to it, I, in fact, care more about catching up on Black Mirror. As Socrates puts it: “No one who knows or believes there is something else better than what he is doing, something possible, will go on doing what he had been doing when he could be doing what is better.”

Now, you might be thinking: Socrates clearly never went to a restaurant with unlimited chips. But he has a point. To figure out what a person’s true priorities are, we usually look to the choices they make. (“Actions speak louder than words.”) When a person binges on TV, munches chips, or gets high despite the consequences, Socrates would infer that they must care more about indulging now than about avoiding those consequences — whatever they may say to the contrary.

(He isn’t alone: Both the behaviorism movement in 20th-century psychology and the “revealed preference” doctrine in economics are based on the idea that you can best learn what people desire by looking at what they do.)

So for Socrates, there’s no such thing as acting against one’s best judgment: There’s only bad judgment. He draws an analogy with optical illusions. Like a child who thinks her thumb is bigger than the moon, we overestimate the value of nearby pleasures and underestimate the severity of their faraway consequences.

Through this Socratic lens, it’s hard not to see addiction as a failure. Imagine a father, addicted to heroin, who misses picking up his children from school because he’s shooting up at home. In Socrates’s view, the father must be doing what he believes to be best. But how could the father possibly think that?

I see two possibilities. As Socrates’s illusion analogy suggests, the father could be grievously mistaken about the consequences of his actions. Perhaps he has convinced himself that his kids can get home on their own, or that he’ll be able to pick them up while high. But if the father has seen the damaging effects of his behavior time and again — as happens often to long-term addicts — it becomes harder to see how he is not complicit in this illusion. If he really believes his choice will be harmless, he must be willfully, and condemnably, self-deceived.

Which leads us to the second, even more damning possibility: Perhaps the father knows the consequences shooting up will have on his children, but he doesn’t care. If his choice cannot be ascribed to ignorance, it must reveal his preferences: The father must care more about getting high than he cares about his children’s well-being.

If Socrates’s model of the mind is right, these are the only available explanations for addictive behavior: The person must have bad judgment, bad priorities, or some combination of the two.

Our philosophy of addiction shapes our treatment of it — whether we realize it or not

It’s not exactly a sympathetic picture. But I suspect it underlies much of our thinking about addiction. Consider the popular idea that someone with addiction has to hit “rock bottom” before she can begin true recovery. In the Socratic view, this makes perfect sense. If addiction is due to a failure to appreciate the bad consequences of getting high, then the best route to recovery might be for the person to experience firsthand how bad those consequences really are. A straight dose of the harshest reality might be the only cure for the addict’s self-deceived beliefs and shortsighted preferences.

We could give a similar Socratic rationale for punishing drug possession with decades in jail: If we make the consequences of using bad enough, people with addiction will finally realize that it’s better to be sober, the thought goes. Once again, we are correcting their flawed judgment and priorities, albeit with a heavy hand.

Socrates’s view also makes sense of our reluctance to adopt medication-assisted treatment and needle exchange programs. These methods might temporarily mitigate the damage caused by addiction, but on the Socratic view, they leave the underlying problem untouched.

By giving out clean needles or substituting methadone for heroin, we may prevent some deaths in the short term, but we won’t change the skewed priorities that caused the addictive behavior in the first place. Worse, we may “enable” someone’s bad judgment by shielding her from the worst effects of her actions. In the long run, the only way to save addicts from themselves is to make it harder, not easier, to pursue the lifestyle they so clearly prefer.

Is Socrates right? Or can we find a better, more sympathetic way of thinking about addiction?

To see things differently, we need to question the fundamental picture of the mind on which Socrates’s view rests. It is natural to think of the mind as a unified whole and identify ourselves with that whole. But this monolithic view of the mind leads to the Socratic view of addiction. Whatever I choose must be what my mind wants most, and so what want most. The key to escaping the Socratic view, then, is to realize that the mind has different parts — and that some parts of my mind are more me than others.

The “self” is not a single, unitary thing

This “divided mind” view has become popular in both philosophy and psychology over the past 50 years. In psychology, we see it in the rise of “dual process” theories of the mind, the most famous of which comes from Nobel laureate Daniel Kahneman, who divides the mind into a part that makes judgments quickly, intuitively, and unconsciously (“System I”) and a part that thinks more slowly, rationally, and consciously (“System II”).

More pertinent for our purposes is research on what University of Michigan neuroscientist Kent Berridge calls the “wanting system,” which regulates our cravings for things like food, sex, and drugs using signals based in the neurotransmitter dopamine. The wanting system has powerful control over behavior, and its cravings are insensitive to long-term consequences.

Berridge’s research indicates that addictive drugs can “hijack” the wanting system, manipulating dopamine directly to generate cravings that are far stronger than those the rest of us experience. The result is that the conscious part of a person’s mind might want one thing (say, to pick his kids up from school) but be overruled by the wanting system’s desire for something else (to get high).

You might be hoping for me to draw you a picture of the brain with “The Self” outlined in thick black ink: a country with its own sovereign territory. Things aren’t quite that simple. Though some parts of the brain (prefrontal cortex) appear to be Selfier than others (cerebellum), conscious and unconscious processes are too deeply intertwined for us to expect to find a clean neurobiological break between them.

The question of how to find the self in the mind is more a philosophical question than a neurobiological one. Even if we had a high-definition map of every neural firing in your brain, we would still have to take a stand on what in this flurry of electrical activity constitutes you.

Over the past half-century, philosophers have turned to this question with new vigor, trying to make sense of the idea that some of a person’s desires (to get sober and care for her children) represent what she cares about — her true self — in a way that other desires (to get high) do not.

The desires that represent my true self are, on different theories, the desires that I want myself to have (Harry Frankfurt), the desires that align with my judgments of what is valuable (Gary Watson), the desires that cohere with my stable life plans (Michael Bratman), or the desires that are supported by rational deliberation (Susan Wolf).

More important than the differences between these views is one critical similarity: These philosophers are united in rejecting the Socratic view. None of them thinks that what I really want is just a matter of what desire wins out over my behavior. To see what my true self wants, we should look not to my actions but to my reflective judgments about the kind of person I want to be and the life I want to lead.

Putting these two strains of thought together, we can see the heroin-addicted father in a different light. As the father decides whether to shoot up or go pick up his kids, two parts of his mind are battling for control: the part that wants heroin more than anything else, and the part that cares far more about his kids. But the father is not a mere bystander in this conflict: He is a participant in it. The father is fighting on the side of the part that cares about his children.

Drugs that reduce cravings don’t “enable” addiction. They give people with addiction an ally.

I would go further and say that the father is the part of his mind that cares more about his children. For if we asked him to tell us what, on reflection, he really cares about, he would say that he wants to get sober and take care of his kids. And in this case, words speak louder than actions.

When the desire for heroin unfortunately wins out, that doesn’t mean that the father cares more about getting high than he cares about his children. It means that he lost the struggle: His behavior is being controlled by a part of his mind that is not his true self.

This is the possibility Socrates failed to recognize: A person might judge one thing to be best and yet do another. The plight of addiction is that of having a powerful part of your mind push you relentlessly and automatically toward behaviors you do not actually want to do. An addicted person behaves the way she does not because she has bad judgment or skewed priorities, but because she is blocked from acting on her true values by her supercharged “wanting system.”

I don’t mean to suggest that no one ever endorses the choice to do drugs. Indeed, as the philosopher Hanna Pickard has argued, addictive behavior is often initiated and maintained by the purposes it serves in someone’s life, often as self-medication for physical or psychological trauma. Nor am I saying that addictive behavior is compulsive, irresistible, or completely out of the person’s control. After all, many people manage to recover from addiction without the help of medication or even clinical intervention.

The messy truth about addiction is that it lies somewhere in between choice and compulsion. Addictive cravings work in much the same way as the cravings that everyone experiences — for Netflix or chips, say. They do not simply take over one’s muscles like an internal puppeteer. Instead, they pull one’s choices toward the craved object, like a psychological kind of gravity.

But as Berridge’s research suggests, the neurochemical effects of addictive drugs make the cravings addicts experience far, far stronger than those the rest of us have to contend with in our daily lives. It may not be impossible to resist these cravings, but it is extraordinarily difficult. And given how hard it is to resist cravings of normal strength — just think of those bottomless chips — we should not blame someone with addiction for failing to overcome her neurobiologically enhanced cravings.

This is why addiction is not a moral failure. The addicted person need not be shortsighted or selfish; she may have the very same priorities as anyone else. Nor need she be any worse at self-control than the rest of us are. She is just faced with cravings that are far harder to resist.

Seeing addiction this way also helps us think more clearly about treatment. Emphasizing the bad consequences of using, whether by pushing someone to rock bottom or by threatening her with prison, is ineffective because the part of the mind that drives addiction can overpower thoughts about consequences.

The problem is not that a person with addiction does not understand the consequences of her actions, but that she is unable to use this understanding to control her behavior. Thus, we should not be worried about “enabling” her addiction by protecting her from its worst effects — for example, by providing her with clean needles.

The paradigm shift is most dramatic for medication-assisted treatment. While the Socratic view paints these treatments as crutches that leave the basic problem unaddressed, the divided mind view shows this to be wrongheaded. If the source of addiction is overly strong automatic cravings, then the most direct way to treat addiction would be to weaken or satiate these cravings in a non-damaging way.

And that is exactly what methadone and buprenorphine do. By satiating the wanting system’s cravings, these medications put the addicted person back in the driver’s seat, allowing her to control her life again.

Plato himself eventually came to understand that the mind was more divided than his teacher thought. While he always used Socrates as his star character, Plato began to strike out on his own in later work. And so it is revealing that in one of his later dialogues, the Phaedrus, Plato takes a different view. The soul, Plato writes, is like a chariot.

The charioteer, Reason, tries his best to guide the chariot along the road of virtue. But his horse, Appetite, is stubborn, “deaf as a post” and may gallop off the road at any moment. “Chariot-driving in our case,” Plato concludes, “is inevitably a painfully difficult business.” If we take that to heart, maybe we will start giving the addicted what they need to get their lives back under control.

Brendan de Kenessey is a fellow in residence at the Edmond J. Safra Center for Ethics at Harvard University. This fall, he will join the faculty of the philosophy department at the University of Toronto. Find him on Twitter @BrendanKenessey.