How Do You Classify a Disease That Won't Sit Still?

Here’s something strange: two people can both be told they have “breast cancer,” but their cancers might be more genetically different from each other than either is from a completely different kind of cancer—say, ovarian cancer. And a tumor the size of a pea in one person’s prostate might grow slowly for decades without ever causing trouble, while the same-looking tumor in someone else could become deadly within a year. So when doctors call both of these things “cancer,” are they talking about the same disease? Or are they using one word for many different things that just happen to share some features?

This isn’t just a question about labels. It matters for real decisions. Should someone with a slow-growing prostate growth get surgery that could leave them incontinent or impotent? Should researchers spend billions of dollars looking for a “cure for cancer” when there might be hundreds of different cancers that need hundreds of different treatments? How do you even decide what counts as cancer in the first place?

What Is Cancer, Anyway?

You might think this is an easy question. Cancer is when cells grow out of control, right? But that definition turns out to be surprisingly slippery.

For one thing, lots of cells in your body grow in controlled ways—that’s how you heal a cut or grow during puberty. And some cells that grow “out of control” aren’t cancer. A wart is a growth, but it’s not cancer. A benign tumor is a mass of cells that aren’t behaving normally, but it won’t spread or kill you.

For another thing, you can have cancer cells in your body without ever knowing it. Autopsies have shown that many people who died of completely different causes had small prostate or thyroid cancers that never caused any symptoms. About half of men over 60 have some prostate lesions that look like cancer under a microscope. If having those cells counts as having the disease, then millions of people have a disease they’ll never notice. If it doesn’t count, then what does?

And here’s where it gets really tricky: cancer doesn’t just have one cause. Some cancers are triggered by smoking, some by viruses (HPV can cause cervical cancer, hepatitis C can cause liver cancer), some by inherited genes, some by hormones, some by things we don’t understand at all. The “same” cancer in the same organ can have completely different genetic signatures, respond to different treatments, and progress at different speeds.

So when philosophers and scientists try to define cancer, they run into a problem: either the definition is so broad that it includes things that aren’t really disease, or it’s so narrow that it leaves out things that clearly are.

Why Classification Matters: The Case of Prostate Cancer

Imagine you’re a 65-year-old man. You go for a routine checkup and a blood test shows some numbers that look a little off. Your doctor does a biopsy and finds a small cluster of abnormal cells in your prostate. The diagnosis: prostate cancer.

But here’s what the doctor might not tell you right away: that small cluster of cells might never cause you any problems. It might sit there for twenty years without growing, without spreading, without affecting your health in any way. You’ll probably die of something else first—a heart attack, a car accident, old age.

If you get treated—surgery, radiation—you might end up with serious side effects: inability to control your bladder, inability to have erections, chronic pain. These side effects could make the rest of your life much harder than the cancer itself ever would have.

So what should happen? Some philosophers, drawing on a theory of disease called the “biostatistical theory,” argue that this small growth shouldn’t count as a disease at all. On this view, disease means impairment of normal function. If the growth isn’t actually interfering with how your body works, it’s not a disease—it’s just a risk factor. Like having high cholesterol. You’d monitor it, but you wouldn’t treat it the same way you’d treat a heart attack.

Other philosophers push back. They say that calling something a disease is always a judgment that involves values. When we decide that a growth is “cancer,” we’re also deciding that it matters enough to do something about. That decision isn’t purely scientific—it depends on how cautious we want to be, how much risk we’re willing to accept, and what we think a good life looks like.

This debate isn’t academic. Hundreds of thousands of men get diagnosed with prostate cancer every year. Many get treatments that permanently damage their quality of life. If we changed how we classify those early-stage growths—if we called them something other than “cancer”—it could change medical practice overnight. But it could also mean that some people who would have been caught early enough to save their lives get missed.

Are Diseases “Natural Kinds”?

Philosophers have a term for categories that really exist in nature, independent of human interests: natural kinds. Gold is a natural kind. All gold atoms have the same number of protons. If you find something that has 79 protons, you can be sure it’s gold. Water is a natural kind. Every water molecule is H₂O. If you’ve found a genuine sample of water, you know exactly what you’ve got.

Cancer is not like that.

Some philosophers argue that cancer is not a natural kind at all, because there’s no single set of features that all cancers share and only cancers share. Different cancers are caused by different things, progress in different ways, respond to different treatments. The category “cancer” might just be a convenient label for a bunch of loosely related problems—useful for organizing hospitals and research funding, but not a real thing that exists in nature.

Other philosophers disagree. They argue that cancers do share something important: they all involve cells that have acquired certain “hallmarks”—abilities that normal cells don’t have. These hallmarks include things like: growing without needing signals from other cells, ignoring signals that tell cells to stop growing, avoiding programmed cell death (apoptosis), and the ability to spread to other parts of the body (metastasis). This list of hallmarks was developed by two cancer researchers, Hanahan and Weinberg, who argued that these are the real defining features of cancer.

On this view, cancer is what philosophers call a “homeostatic property cluster kind.” That’s a fancy way of saying: there’s a cluster of properties that tend to go together, and there are underlying mechanisms (like mutations to certain genes) that cause this cluster to hang together. Not every cancer has every property, but most cancers have most of them, and the properties are connected in ways that aren’t just random.

But even this view has problems. Recent research has shown that some of these hallmarks can be present in normal cells that never become cancerous. And the order in which mutations happen matters. Depending on which of two genes gets mutated first, patients with the same type of blood cancer end up with very different outcomes. The same “cause” can lead to cancer in one person and nothing in another.

Three Different Theories of What Causes Cancer

If you can’t agree on what cancer is, you probably can’t agree on what causes it. And indeed, there are at least three major competing theories.

The Somatic Mutation Theory says that cancer starts when cells acquire mutations—changes in their DNA—that disrupt the normal controls on cell division. These mutations can be inherited (like the BRCA gene mutations that raise breast cancer risk) or acquired over a lifetime (like from smoking or UV radiation). Over time, cells accumulate enough mutations to break free from normal regulation and become cancerous. This has been the dominant view for the last few decades, and it’s the reason why so much cancer research focuses on genetics.

The Tissue Organization Field Theory says something quite different. Its advocates, Ana Soto and Carlos Sonnenschein, argue that cancer isn’t primarily a disease of genes at all. Instead, they say, cancer starts when the normal organization of tissues breaks down. Cells communicate with each other through their physical environment—the extracellular matrix, the way tissues are structured. When that communication gets disrupted, cells start behaving badly. Mutations come after the breakdown, not before. On this view, you can’t understand cancer just by looking at genes; you have to look at the whole tissue environment.

The Cancer Stem Cell Theory says that tumors are organized like organs, with a small population of “stem cells” that keep dividing and producing all the other cancer cells. Most cancer cells can only divide a limited number of times, but cancer stem cells can keep going indefinitely. If you kill the main tumor but leave the stem cells alive, the cancer will come back. This means that treatments need to target stem cells specifically—and that measuring how many stem cells a tumor contains might be crucial for understanding how dangerous it is.

Which theory is right? The honest answer is that nobody really knows, and it might turn out that different theories are right for different kinds of cancer. Some philosophers argue that these theories aren’t really competing. They might be complementary—each capturing a different part of a very complex picture. The somatic mutation theory might explain how cancer starts, the tissue organization theory might explain how it progresses, and the stem cell theory might explain why it comes back after treatment.

Cancer as an Evolutionary Process

Here’s another way to think about cancer, and it’s one that connects to something you already know about: evolution. You’ve probably learned about how species evolve through natural selection—individuals with traits that help them survive and reproduce pass those traits on. Something similar happens inside your body when you have cancer.

Cancer cells divide and mutate. Most mutations are harmless or harmful to the cell, but occasionally one gives a cell an advantage. Maybe it lets the cell grow faster, or ignore signals that would normally tell it to die, or resist chemotherapy. That cell’s descendants will outcompete the others. Over time, the tumor becomes more aggressive, more dangerous, harder to treat.

This is evolution, but on a much shorter timescale and inside a single body. It’s also a kind of tragedy from a “multilevel selection” perspective. Your body is a collection of trillions of cells that mostly cooperate. They do their jobs, stay in their places, and die when they’re supposed to. Cancer cells are cheaters—they break the cooperative rules. From the perspective of natural selection at the level of individual cells, cancer cells are highly successful. From the perspective of you, the whole organism, they’re catastrophic.

Some scientists have wondered whether cancer is actually a byproduct of how multicellular life evolved. Billions of years ago, single-celled organisms started cooperating to form collectives. Those collectives eventually became multicellular organisms like us. But the mechanisms that hold these cellular societies together are never perfect. Cancer might be the price we pay for being multicellular at all.

This perspective leads to some interesting questions. If cancer is inevitable—a built-in risk of being a complex organism—then maybe we should approach it differently. Instead of trying to “cure” it in the way we cure an infection, maybe we should focus on managing it, slowing it down, living with it the way we live with aging.

Is This Just a Scientific Question?

By now you might be thinking: okay, this is complicated, but eventually scientists will figure it out, right? They’ll sequence enough genomes, run enough experiments, and settle the debate.

But many philosophers argue that the debate about cancer classification isn’t just scientific. It’s also, in important ways, normative. That means it involves judgments about values, about what matters, about what we should do.

When a doctor decides whether to call a slow-growing prostate growth “cancer” or “a risk factor,” that decision involves value judgments. How important is it to avoid overdiagnosis and unnecessary treatment? How important is it to catch every possible case, even if that means some people get treated for things that would never have hurt them? These aren’t questions that science can answer on its own.

When researchers decide how to classify different cancers, they’re also deciding how to allocate research funding, which treatments to develop, which patients are “similar enough” to be included in the same clinical trial. These decisions affect real people’s lives.

And when we talk about “fighting a war on cancer,” as President Nixon declared in 1971, we’re making choices about how to spend enormous amounts of money. Billions of dollars have gone into genetic research, into sequencing cancer genomes, into developing targeted drugs. Has this been the right investment? Some critics say that despite all this spending, cancer mortality rates haven’t dropped as much as promised. Others argue that the focus on high-tech treatments has distracted from simpler, more effective interventions like reducing smoking, improving diet, and making screening available to everyone.

There’s also a question about whether our classification of cancer might be influenced by economic interests. Cancer drugs are incredibly expensive—the average new cancer drug costs about $150,000 per year. Pharmaceutical companies have strong incentives to define cancer broadly, to push for more screening, to encourage treatment even for early-stage growths that might never cause problems. That doesn’t mean they’re wrong, but it does mean we should be thoughtful about who benefits from different ways of classifying disease.

So What’s the Answer?

If you were hoping for a neat conclusion, you’re probably disappointed. Philosophers haven’t agreed on exactly what cancer is, or how to classify it, or which theory of its causes is correct, or who should decide these things. They probably never will agree completely.

But here’s what the debate shows us. Cancer is strange in a way that most things aren’t. It’s not like gold, where you can run a simple test and know exactly what you’ve got. It’s not like an infection, where a specific germ causes a specific disease. Cancer is messy, heterogeneous, dynamic, and deeply entangled with the fact that we are evolved, multicellular, aging organisms.

Maybe the best way to understand cancer is not to look for a single definition or theory, but to accept that different definitions and theories are useful for different purposes. If you’re a surgeon deciding whether to operate, you might care about the tumor’s size and location. If you’re a geneticist trying to develop a targeted drug, you might care about specific mutations. If you’re a public health official trying to reduce cancer deaths, you might care about patterns of smoking and screening across whole populations. If you’re a patient trying to decide what to do, you might care about all of these things and more.

The philosopher John Dupré called this view “promiscuous realism”—the idea that there are many legitimate ways to carve up the world, none of them the single “correct” one. Cancer might be a case where this view makes the most sense. There’s no one right way to classify it. But that doesn’t mean any classification is as good as any other. Different classifications serve different purposes, and some serve their purposes better than others.

The hard work is figuring out which classification we need in which situation, and being honest about the values and priorities that guide our choices.

Appendices

Key Terms

Term	What it does in the debate
Natural kind	A category that really exists in nature, not just something humans made up for convenience
Hallmarks of cancer	A set of abilities that cancer cells typically acquire (like unlimited growth and ability to spread)
Homeostatic property cluster kind	A category where a group of features tend to hang together because of shared underlying mechanisms
Somatic Mutation Theory	The dominant view that cancer starts when cells accumulate DNA mutations that disrupt normal controls
Tissue Organization Field Theory	The alternative view that cancer starts when communication between cells and their physical environment breaks down
Cancer Stem Cell Theory	The view that tumors contain a small population of “stem cells” that keep producing all the other cancer cells
Overdiagnosis	Diagnosing a condition that would never have caused symptoms or harm in a patient’s lifetime
Multilevel selection	The idea that natural selection can operate at different levels simultaneously (individual cells, whole organisms, groups)

Key People

Hanahan and Weinberg – Two cancer researchers who developed the influential “hallmarks of cancer” framework, listing the capabilities that cancer cells typically acquire.
Ana Soto and Carlos Sonnenschein – Biologists who challenged the dominant mutation-based view of cancer, arguing that cancer starts with breakdowns in tissue organization, not genes.
John Dupré – Philosopher who argued for “promiscuous realism”—the idea that there are many legitimate ways to classify things in nature, not just one correct system.

Things to Think About

If you were a doctor, would you tell a patient they have “cancer” if the growth is so slow that it will probably never cause problems? What are the costs and benefits of telling them vs. not telling them?
The “tissue organization theory” says that if you put cancer cells into a healthy tissue environment, they sometimes become normal again. What does this suggest about what “causes” cancer?
Cancer is sometimes described as “evolution within the body.” If that’s true, why can’t we just let it evolve and see what happens? What makes this kind of evolution different from the evolution of species?
The cost of a new cancer drug averages $150,000 per year. Who should decide whether that’s worth it—scientists, doctors, patients, insurance companies, or government regulators?

Where This Shows Up

Medical debates about prostate cancer screening – The same questions about what counts as disease and when to treat are happening right now in doctors’ offices and hospital ethics committees.
Discussions about precision medicine – When you hear about “personalized” cancer treatments based on genetic sequencing, you’re seeing the consequences of how scientists classify different cancers.
Evolutionary biology – Cancer is increasingly studied as an evolutionary process, connecting it to broader questions about how cooperation and cheating work in biological systems.
Public debates about health care costs – The question of whether expensive cancer treatments are worth their price depends partly on how we define “effective treatment” and what we count as success.