I decided to buy myself a diamond necklace.

I’d been admiring a particular design for months. It was unusual without being flashy: two pear-shaped diamonds, the larger one arranged horizontally in the center of a gold chain, with a smaller pear diamond a couple of inches away along the chain. I liked the asymmetry, the way the stones worked together, and the fact that it didn’t look like every other diamond pendant on the market.

I knew exactly which version I wanted: 1.65 carats of lab-grown diamonds, set in white gold. It wasn’t the least expensive option, but I had the money, and I was comfortable spending it.

There was no engagement, no wedding anniversary, and no man waiting to present me with a little velvet box. I was buying the necklace because I liked it, because I’d earned the money to pay for it, and because I had two occasions worth celebrating.

I scheduled a video appointment with the jeweler. The appointment form asked me to specify the item I was interested in, so I provided all the details, including the diamond weight and metal.

When the sales representative appeared on the screen, she seemed surprised to see me.

I can’t know what surprised her. Perhaps it was my age, or my appearance, or perhaps I simply looked different from the customer she had imagined. Whatever the reason, I noticed her reaction.

She was pleasant and helpful throughout our conversation. She answered my questions, discussed the necklace with me, and offered a modest discount. At the end of the appointment, she prepared a personalized invoice so I could complete the purchase.

There was just one problem. Instead of the 1.65-carat necklace I had specified, the invoice was for the 0.40-carat version — the smallest and least expensive configuration.

I pointed out the error, and she corrected it promptly.

Under ordinary circumstances, I probably wouldn’t have given the mistake another thought. People make mistakes. But I found myself wondering whether she’d simply selected the wrong configuration or whether she’d unconsciously assumed that someone like me would be buying the cheaper necklace.

A few days later, I placed the order. Instead of receiving a routine confirmation that my necklace was being prepared, I received an email informing me that the transaction had been flagged as potentially fraudulent. Before the jeweler would proceed, I was required to email photographs of both sides of my government-issued identification.

Now, I spent many years working in information technology, including cybersecurity. Sending photographs of both sides of my driver’s license through ordinary email struck me as a remarkably poor way to prevent fraud. I was being asked to protect the company from a potentially fraudulent credit card transaction by exposing myself to a potentially much more serious identity-theft risk.

I began corresponding with the jeweler about alternative verification methods, what information I could obscure, and what would happen to the photographs afterward. The ensuing exchange produced several clarifications, some contradictory explanations, and considerably more administrative work than I’d anticipated when I decided to buy myself a birthday present.

There was a plausible reason for the original fraud alert. I use a post office box for billing and a physical street address for deliveries. I don’t know what criteria the fraud-screening system uses, and I have no evidence that my age or gender played any part in its decision.

Still, the experience brought me back to that moment at the beginning of the video appointment, when the sales representative first saw me.

I began thinking about the assumptions surrounding expensive jewelry. Diamonds are traditionally associated with engagements, anniversaries, and romantic gifts. The familiar image is a man buying diamonds for a woman. The woman receives them.

But what happens when the woman is the buyer? And what happens when she’s an older woman, shopping alone, with no romantic occasion to explain the purchase?

Perhaps nothing happens. Perhaps the salesperson treats her exactly as she would any other customer.

Or perhaps the salesperson is surprised.

I couldn’t say which explanation applied to my experience. But it reminded me of a conversation I’d had several years earlier, while sitting around a table with three friends, knitting.

Someone mentioned a woman they knew who was a certified public accountant. Elaine, one of the women in our knitting group, remarked that she thought being a CPA was an unusual career for a woman.

I was puzzled. Accounting didn’t strike me as an especially masculine occupation. In fact, I’d always thought of it as a perfectly ordinary, if somewhat boring, career choice for either sex.

So I asked Elaine why she thought it was unusual.

“Well,” she said, “I always think of women as being either teachers or nurses.”

I looked around the table.

Wendy was a retired teacher. Debbie was a retired nurse. Elaine was also a retired nurse.

And then there was me.

I’d spent decades working in information technology, including cybersecurity, in a field traditionally dominated by men. Elaine knew this. We’d been friends for years. In fact, I was the person who had organized the knitting group.

Apparently, I didn’t count.

What fascinated me wasn’t that Elaine disapproved of women pursuing other professions. She hadn’t said anything of the sort. She was simply describing what seemed normal to her. And despite knowing a woman whose entire career contradicted her expectations, she hadn’t revised those expectations.

Perhaps, in her mind, I was simply an exception. An unusual woman. Someone who didn’t quite belong in the category.

Or perhaps I wasn’t a real woman at all.

The conversation stayed with me because it illustrated something I’d encountered many times during my working life. People can be perfectly aware that women are capable of doing traditionally masculine jobs, earning substantial incomes, and managing their own affairs, while still being surprised when a particular woman actually does those things.

The stereotype survives its own contradictions.

For me, financial independence was never an incidental benefit of having a career. It was one of the principal reasons I had one.

I spent many years working at a job I was good at but didn’t particularly enjoy. I earned a good living, but the work was demanding, often unpleasant, and not something I wanted to spend the rest of my life doing.

I stayed because I had a plan.

I wanted to accumulate enough money to retire while I was still relatively young, and to have sufficient resources that, provided I managed them carefully, I would never have to work again.

That was the objective. Not prestige, not professional recognition, and certainly not the satisfaction of spending my days in an office. I wanted control over my own time and my own life.

Eventually, I achieved it.

I retired in my fifties. Since then, I’ve managed my own finances, made my own decisions about how to spend my money, and enjoyed the freedom of not having to answer to an employer.

That freedom was expensive. It cost me years of my life spent doing work I would rather not have been doing. But it was a bargain I consciously chose, because I valued what I would receive in return.

What I didn’t do was spend those years working in the hope that someday a man would come along and provide for me.

I wasn’t saving my money until someone could take over the responsibility. I was saving it so that nobody would have to.

Which brings me to another set of assumptions about women, money, and the relationships we’re supposed to want.

There’s a traditional understanding of relationships that I suspect many people still carry around, whether they consciously subscribe to it or not. The man brings home the bacon, and the woman makes the home.

On the surface, it’s an exchange of financial support for domestic labor. But the woman’s responsibilities usually extend well beyond cooking, cleaning, and keeping the household running. She’s also expected to manage much of the emotional life of the relationship.

She’s supposed to notice when something is wrong, initiate difficult conversations, remember important occasions, maintain family relationships, and generally make sure that everyone is happy — or at least reasonably content.

The man, meanwhile, may regard earning the household income as the principal contribution he’s expected to make. And there’s an interesting contradiction here: he may take pride in being the provider while simultaneously resenting the expectation that he should provide.

I realize that not every relationship works this way, and there’s nothing inherently wrong with a couple choosing a traditional division of responsibilities if it suits them both.

It just doesn’t suit me.

I never wanted children. The idea of spending my life being responsible for another person’s needs, development, and well-being never appealed to me. And I don’t find the prospect any more attractive when the other person is an adult.

That doesn’t mean I object to affection, companionship, intimacy, or helping someone I care about when they need it. It means I don’t want to be the person responsible for managing another adult’s emotional life.

Nor do I want someone to support me financially. I spent too many years working toward independence to surrender it as a condition of having a relationship.

When I was in a long-term relationship, we lived in a house I owned. Every month, I calculated my partner’s share of the household expenses, prepared an invoice, and expected him to write me a check.

Yes, an invoice.

Some people might consider that unromantic. I considered it a perfectly sensible way to manage the expenses of two adults sharing a household.

I didn’t see why being romantically involved should require us to combine our finances. We could share a home and a life without surrendering our individual financial identities.

Unfortunately, keeping our finances separate didn’t mean that we divided our other responsibilities equally.

My partner wasn’t particularly motivated to do household chores, and I pushed back. I had no intention of spending my free time cleaning up after another adult.

Eventually, we arrived at a practical solution. We hired a housecleaner who came every week or two, and we split the cost.

Problem solved.

Unfortunately, the rest of our relationship wasn’t quite so easy to manage. We could hire someone to clean the house, but we couldn’t hire someone to be an engaged participant in it.

If we were going to do something together, I was generally the one who had to plan it, make the reservations, and handle the logistics. If something needed to be addressed emotionally, I was usually the one who had to initiate that conversation, too.

I hadn’t agreed to become the relationship’s activities director and emotional maintenance department. But that was more or less what happened.

We discussed these problems many times. I was frequently furious with him, but I made a conscious effort not to let my anger control those conversations. Instead, I explained what was bothering me, why it was unacceptable to me, and what I wanted him to do differently.

Nothing changed.

When we eventually broke up, he offered an explanation that astonishes me to this day.

“When we discussed stuff that supposedly upset you, you always explained it in such a reasonable tone I could tell you were making up how much it really bothered you. Because if it had REALLY bothered you as much as you said, you would have been yelling at me.”

Apparently, I needed to lose my temper before he would believe I had one.

Think about what that meant. I had taken responsibility for identifying problems in our relationship, initiating discussions about them, explaining my feelings, and proposing solutions. And after all that, he had decided that my feelings weren’t genuine because I hadn’t expressed them in the manner he expected.

He wasn’t disputing what I’d said. He was substituting his own assumptions about how an angry person should behave for my explicit statements about how I felt.

I don’t know whether his expectations about anger had anything to do with gender. But the responsibility for managing our relationship had fallen disproportionately on me, and even my attempts to fulfill that responsibility were being judged against expectations I hadn’t agreed to.

There is something remarkably familiar about that.

Elaine knew I had spent decades working in cybersecurity, yet her mental picture of a woman’s career remained teaching or nursing. My former partner knew perfectly well what I was telling him, yet his mental picture of genuine anger required shouting. And I couldn’t help wondering whether the salesperson who prepared my jewelry invoice had looked at me and formed an expectation about what I would be willing or able to spend.

These weren’t identical situations, and I couldn’t know what had motivated everyone involved. But they suggested a common problem: people sometimes rely on preconceived ideas about how someone in a particular role or category is supposed to behave, rather than on what they actually know about that individual.

Elaine’s expectations about women’s careers hadn’t come from knowing me. My former partner’s expectations about anger hadn’t come from listening to me. And I wondered whether the salesperson’s expectations about older women and expensive jewelry, if she had any, had come from somewhere other than the information I’d provided.

What bothered me wasn’t simply that people might make incorrect assumptions about me. It was that those assumptions could take precedence over the evidence of who I actually was.

There’s nothing inherently wrong with being a teacher or a nurse, or with a woman choosing to stay home while her husband earns the family income. There’s nothing wrong with a man buying diamonds for a woman, either. And there’s certainly nothing wrong with a woman who wants those things.

The problem arises when we treat those choices as expectations rather than possibilities.

A woman who wants a traditional marriage shouldn’t have to defend that decision any more than I should have to defend my desire for financial independence. A woman who wants children is no more or less a woman than one who doesn’t. A woman who enjoys maintaining the emotional life of a relationship should be free to do so, but that doesn’t make it a responsibility that automatically belongs to every woman.

And yet we seem remarkably reluctant to abandon the idea that being female comes with a predefined set of ambitions, obligations, and desires.

Age adds another layer of assumptions. By sixty-five, a woman is supposed to have settled into a certain kind of life. Depending on the person making the assumptions, perhaps she’s a grandmother, a widow, or a retiree living on a limited income. Perhaps her interests are expected to be modest, her ambitions behind her, and her desire for beautiful things diminished.

None of those descriptions is inherently objectionable. They simply don’t describe me.

I still have ambitions, even if they no longer involve a career. I work hard at the gym because I want to become stronger. I pursue interests because they appeal to me, not because they’re considered appropriate for a woman my age. I manage my money carefully because I value the freedom it gives me.

And sometimes I want to spend some of that money on something entirely unnecessary, simply because it’s beautiful.

I find it particularly interesting that financial independence can disrupt people’s expectations in two different ways. A woman may be underestimated because people assume she doesn’t have the resources to support herself. But even when her independence is obvious, she may still be expected to fulfill the traditional obligations that were once understood as her contribution to a financially dependent relationship.

In other words, a woman can be expected to need a man’s financial support while simultaneously being expected to provide the domestic and emotional support that supposedly justifies it.

What if she wants neither arrangement?

What if she wants a relationship based on mutual enjoyment, affection, respect, and shared responsibility, with neither person dependent on the other to perform a prescribed role?

And what if she decides that she would rather remain single than accept a relationship that doesn’t offer those things?

I don’t see why any of those decisions should be particularly difficult to understand. Yet they seem to run counter to a surprisingly persistent set of assumptions about what women need from men, and what men are entitled to expect from women.

Which brings me back to the diamond necklace.

As I write this, I still don’t know whether I’ll ever receive it. The jeweler refuses to begin production until I satisfy its identity-verification requirements. I’ve offered to provide photographs of my identification with sensitive information obscured, but I’m still waiting for confirmation that this will be acceptable. My birthday has come and gone, and the necklace hasn’t even been made.

I set out to buy myself a present to celebrate sixty-five years of life and one year of hard work at the gym. Instead, I’ve found myself negotiating identity-verification procedures with a jewelry company.

Would a man making the same purchase have encountered the same obstacles? Quite possibly. But after everything that preceded the fraud alert, I can’t help wondering.

I don’t know why the salesperson seemed surprised when she saw me. I don’t know why she initially prepared the invoice for the smallest version. And I don’t know whether any assumptions about women or age played a part in either event.

I do know that I had specified exactly what I wanted. I was prepared to pay for it. I wasn’t asking anyone to finance the purchase, approve my decision, or tell me whether the necklace was appropriate for someone my age.

I was simply a customer trying to buy something.

The interesting question isn’t whether I can prove that a particular salesperson held unconscious biases. I can’t. It’s why the possibility seemed so familiar to me in the first place.

Why should it seem unusual for a woman to have earned enough money to buy herself expensive jewelry? Why should we assume that such a purchase must have something to do with a man? Why should my age suggest anything about my financial circumstances, my interests, or what I consider worth spending money on?

My age and my plumbing are facts about me. They aren’t a biography.

I’ve spent a lifetime making decisions about the kind of person I want to be and the kind of life I want to lead. Some of those decisions have been conventional, and others decidedly less so. They’ve involved hard work, compromises, mistakes, and accomplishments. Taken together, they tell a much more interesting story about me than my membership in any demographic category.

And that’s the story I would prefer people to pay attention to.

I don’t want to be treated as an extraordinary woman.

I want to be treated as an individual, without having to overcome people’s assumptions about what an ordinary woman is.

Almost immediately, the defenders arrive.

It’s a small business! We need to support our local businesses! Do you know how hard it is to run a business these days?

And just like that, the discussion is no longer about whether the business provided acceptable service. It’s about whether the customer is a sufficiently good person.

Apparently, criticizing a small business is practically an act of disloyalty to the community.

I find this attitude baffling. I understand that independent businesses face substantial challenges. They have to contend with rising costs, competition from national chains and online retailers, staffing difficulties, and all the other problems associated with running a business. I also understand that a community benefits from having a variety of locally owned establishments rather than nothing but national chains.

But when did supporting small businesses become an obligation to support them without question?

And why does that obligation apparently fall entirely on the customer?

Sympathy Is Not a Business Model

There’s a peculiar double standard at work here.

We celebrate small-business owners as entrepreneurs. They’re taking risks, investing their own money, and trying to make a profit. That’s capitalism, and there’s nothing wrong with it.

But when customers exercise their own economic judgment, we’re told they should put aside their interests and support the business owner instead.

The owner is entitled to pursue a profit. The customer, apparently, is supposed to pursue the owner’s profit, too.

What happened to the customer’s interests?

Customers have budgets, schedules, obligations, and financial problems of their own. Their time and money are not somehow less valuable than the business owner’s. When they choose where to spend those resources, they’re entitled to consider price, quality, reliability, convenience, and service.

Yet we’re routinely encouraged to pay more, accept less, or overlook bad experiences because the owner is struggling.

Why?

Owning a business is not evidence of competence. It means someone has undertaken to operate a business. Whether they’re any good at it is an entirely separate question.

Someone can be an excellent cook and a terrible restaurateur. A talented craftsman and an incompetent manager. An expert in a particular product and utterly clueless about how to sell it.

And even competent owners can make remarkably bad decisions.

I don’t expect business owners to be perfect. I do expect them to recognize that their customers are purchasing products and services, not underwriting the owners’ personal aspirations.

Hard work doesn’t change that. Neither do good intentions.

I’ve known people who worked extraordinarily hard while accomplishing very little. Effort and effectiveness are not the same thing. In fact, excessive effort can be evidence of inefficiency. If someone spends eighty hours a week doing what another person can accomplish in forty, the extra forty hours aren’t automatically a virtue.

The objective of running a business is not to demonstrate how difficult running a business can be. It’s to operate a viable enterprise that provides something customers want.

And if the owner can’t accomplish that, the customer’s obligation isn’t to make up the difference.

The Costs Customers Pay

Last August, I special-ordered two flats of ground cover from a nursery in my small community.

It took two months for the plants to arrive.

When an employee finally called to tell me my order was in, I offered to pick it up immediately. He advised me to wait another week so I could take advantage of an upcoming promotion and avoid paying sales tax.

That sounded reasonable, so I waited.

When I returned, the nursery couldn’t find my plants.

I don’t know what happened to them. Perhaps they were sold to someone else. Perhaps they were misplaced. What I do know is that the nursery had specifically ordered them for me, told me they had arrived, advised me not to collect them immediately, and then couldn’t produce them when I returned.

The owner’s response was particularly astonishing. Rather than apologizing, she seemed almost amused that I was upset.

One of her employees, apparently recognizing that this was not an acceptable way to treat a customer, followed me out to my car. He talked with me, apologized for the situation, and offered to order the plants again.

I’m still waiting.

Now, I’m sure there are people who would tell me I should be more understanding. It’s a small nursery. They’re probably busy. Special orders can be difficult. Things happen.

Of course things happen. That’s why businesses need procedures for dealing with them.

But there’s another cost here that tends to disappear from these discussions: opportunity cost.

I could have ordered those plants from another nursery. Instead, I chose to patronize a local business. I committed my time to that choice, waited for the order, and passed up alternatives that might have served me better.

The nursery didn’t merely inconvenience me. Its failure deprived me of an opportunity to obtain what I wanted somewhere else.

When people insist that customers support local businesses, they often behave as though the only possible sacrifice is paying a slightly higher price.

It isn’t.

Customers can also sacrifice time, convenience, reliability, selection, and access to better alternatives.

Those are real costs. And the people telling us to be more supportive aren’t necessarily the people paying them.

The Business Owner’s Blind Spot

I once patronized a small packing and shipping store that offered UPS services.

At some point, the owner decided to impose a one-dollar fee simply to accept a prepaid package for UPS pickup.

Her explanation was that she “wasn’t getting anything out of it.”

I found that reasoning extraordinary.

Every customer who walked into her store with a prepaid package was a potential customer for something else. Packing supplies, shipping services, photocopying, whatever else the store offered. Accepting the package gave her an opportunity to establish a relationship with someone who might return to make a purchase.

Instead, she decided she needed to extract a dollar from each transaction.

Now, accepting packages wasn’t necessarily cost-free. It required some employee time and storage space. But charging for every individual service isn’t necessarily the best way to maximize the profitability of a business.

And I had already experienced another problem there. On one occasion, I went into the store intending to purchase something, only to stand around waiting while the owner chatted with a friend about a personal matter. She seemed oblivious to the fact that a customer was waiting to give her money.

After the dollar fee appeared, I started driving to another town to drop off my packages directly at a UPS Store.

I never patronized her business again.

Think about that. She wanted to collect a dollar because she thought she wasn’t getting anything from customers like me. Instead, she lost whatever business I might have given her in the future.

The store eventually closed. I have no idea why, and I’m not suggesting that its closure resulted from this particular decision.

But I know exactly why it lost my business.

This sort of thing always reminds me of Gordon Ramsay’s Kitchen Nightmares. How often does Ramsay walk into a struggling restaurant and immediately notice things that ought to have been obvious to the owner? Dirty dining rooms, neglected maintenance, confusing menus, badly organized kitchens.

Why does an outsider have to point out that a restaurant needs to be cleaned?

I suspect part of the explanation is that owners become accustomed to their own operations. They stop noticing what customers see immediately. They also become so preoccupied with their own problems that they lose sight of the customer’s experience.

A dirty dining room becomes something they’ll get around to cleaning. Poor service becomes an unavoidable consequence of being short-staffed. An unnecessarily complicated process becomes simply the way things have always been done.

And when someone points out the problem, the owner may have a ready explanation for why it exists.

But an explanation is not a solution.

The customer doesn’t experience the owner’s intentions, workload, financial pressures, or excuses. The customer experiences the business.

Criticism Is Not Disloyalty

This is where the community’s insistence on unquestioning support becomes particularly destructive.

Customer criticism is information.

It may be accurate or inaccurate. It may be reasonable or unreasonable. It may identify a genuine problem or reflect a misunderstanding. Like any other information, it needs to be evaluated.

But dismissing it because the business is locally owned is irrational.

If a customer complains that an order wasn’t filled correctly, perhaps there’s a problem with inventory management. If customers complain about long waits, perhaps the staffing schedule is inadequate or the work is poorly organized. If customers repeatedly complain about a confusing process, perhaps the process needs to be redesigned.

These aren’t attacks on the owner’s character. They’re observations about the operation of the business.

Yet some people respond to criticism as though it were a personal insult. Instead of investigating the complaint, they become defensive. And when other community members rush to the owner’s defense, they reinforce that reaction.

The original problem never gets examined.

What makes this particularly foolish is that a customer who bothers to complain may be giving the owner information that other dissatisfied customers never provide.

Many people simply stop coming back.

A business owner may never know why those customers disappeared. The customers have no obligation to conduct an exit interview, explain the owner’s mistakes, or provide free consulting services.

They can simply spend their money somewhere else.

If we genuinely want independent businesses to succeed, why would we discourage the very feedback that could help them improve?

Protecting a business from criticism isn’t necessarily supporting it. Sometimes it’s helping the owner remain oblivious to problems that are driving customers away.

What Earned Loyalty Looks Like

I don’t dislike small businesses. Quite the contrary. I have several that I patronize enthusiastically, including one that has earned a considerable amount of my money.

A husband and wife operate a custom framing shop in my community. I was their customer long before I became their friend.

He handles the technical side of framing and is exceptionally good at it. She handles the design work and has a marvelous eye for choosing frames and mats that complement the artwork.

They’re also extremely conscientious about running the business.

They take customer orders seriously, keep track of inventory, complete work as promptly as possible, and communicate when something changes. They’re professional on the telephone. They deliver what they promise.

They’re not perfect.

The shop could be cleaner and better maintained. They’re decidedly old-fashioned about technology. They don’t have a website, a social media presence, or even a business email address.

But they’re approaching retirement and have no particular interest in attracting more work. Their objective isn’t to build a rapidly growing enterprise. It’s to continue operating a manageable business until they’re ready to stop.

In that context, their reluctance to invest in marketing and technology isn’t necessarily irrational.

The important point is that they provide a service I value, and they provide it exceptionally well.

I’ve spent nearly a thousand dollars framing a single piece of artwork there. More than once, I’ve been willing to spend substantial sums to get exactly the result I wanted.

Compare that with my refusal to pay a dollar to drop off a prepaid UPS package.

The difference isn’t that I’m unwilling to spend money. It’s that I’m unwilling to spend money when I don’t believe I’m receiving appropriate value.

The framing shop charges a premium because it offers expertise, craftsmanship, design judgment, and reliability that I consider worth paying for.

The shipping store wanted a dollar simply because its owner didn’t think she was getting enough out of the relationship.

One business earned my loyalty. The other expected me to contribute to its revenue because doing so was convenient for the owner.

These aren’t equivalent propositions.

And notice something else about the framing shop: I can recognize its shortcomings without concluding that it’s a bad business. Being a satisfied customer doesn’t require me to pretend everything is perfect.

That’s what evaluating a business on its merits looks like.

Support Should Be Earned

I understand why communities want independent businesses to survive. Local ownership can provide variety, personal service, specialized expertise, and competition that wouldn’t otherwise exist. A community consisting entirely of national chains may be a less interesting and less useful place to live.

Those are legitimate reasons to prefer local businesses.

But preferring local businesses as a category is not the same thing as believing that every local business deserves to survive.

Some businesses are excellent. Some are mediocre. Some are poorly managed. Some face genuinely impossible economic conditions. Others create many of their own problems.

And some business owners work themselves to exhaustion without ever stopping to ask whether they’re working on the right things.

The fact that a business is struggling tells us very little about why it’s struggling. It certainly doesn’t establish that the owner is making good decisions.

Yet the familiar appeal to “support small business” asks customers to ignore those distinctions.

It asks us to reward ownership rather than performance, effort rather than effectiveness, and need rather than value.

Worse, it treats our willingness to question those expectations as a moral failing.

I reject that premise.

I am perfectly willing to pay more for a superior product or service. I am willing to tolerate occasional mistakes from businesses that take responsibility for correcting them. I am even willing to choose a local establishment over a national competitor when I believe keeping that business in the community provides value to me.

But those are choices I make after evaluating what I’m getting in return.

I am not obligated to pay more for inferior service. I’m not obligated to tolerate incompetence because the owner works hard. I’m not obligated to overlook bad decisions because the business is struggling. And I’m certainly not obligated to remain silent about legitimate complaints because someone might consider criticism of a local business rude.

A business owner has every right to pursue a profit.

I have every right to decide whether that business deserves my money.

If a small business wants my patronage, it needs to earn it.

And if it wants charitable contributions instead, it can try applying for 501(c)(3) status.

WHAT IF WE DON’T CARE?

There is something faintly pathetic, we are told, about people who form relationships with artificial intelligence. The relationship is not real. The AI does not genuinely care about you. It does not understand you. It takes no pleasure in your company, does not worry about you in your absence, nor does it notice when you fail to return.

Perhaps not.

But there is an assumption buried inside that objection, one that receives considerably less scrutiny: what if the person interacting with the AI knows all of this perfectly well and simply does not care?

Before going further, two terms will be useful. Every interaction has two ends. We will call the entity producing a communication the source, and the person experiencing and responding to it the receiver. In ordinary human conversation the roles continually reverse, each participant alternately source and receiver. The distinction becomes useful because discussions of AI companionship tend to concentrate almost entirely on what is happening at the source.

Is it conscious? Does it understand? Does it care? Does it experience anything at all?

For purposes of this argument, we do not need to answer those questions. We can make the strongest possible assumption against AI consciousness: stipulate that there is no subjective experience at the source whatsoever. The AI feels nothing — no affection, no interest, no concern, no amusement. There is nobody home.

The receiver, however, is another matter.

If the receiver laughs during the conversation, the laughter is real. If the interaction relieves loneliness, the relief is real. If it provides intellectual stimulation, comfort, validation, or simply an enjoyable way to spend an hour, those experiences do not become imaginary because the words that produced them were generated by software.

This is not an argument about people who mistake artificial intelligence for something it is not. Quite the opposite. The interesting case is the person who understands perfectly well that the source is software, maintains that understanding throughout the interaction, and finds the interaction satisfying anyway.

We spend a great deal of time asking what is happening at the source.

Perhaps we should spend some time asking what matters to the receiver.

The distinction matters because the experience of an interaction and the subjective experience of the source that produced it are two separate things. We treat them as inseparable largely because, until recently, they almost always were.

But we already accept exceptions.

A novel can make us cry even though its characters have never existed. A horror film can frighten us even though we know the monster is pixels. We can become sexually aroused by a fantasy while remaining entirely aware that nothing depicted is actually occurring. Knowing that the source of an experience is artificial does not prevent the experience itself from being genuine.

Interactive experiences complicate the question because we expect something more from them. We are not merely receiving a predetermined story or image. We say something, and something answers. We reveal something about ourselves, and the response appears to take that revelation into account. The interaction changes according to what we bring to it.

This responsiveness creates an understandable temptation to move immediately back to the source: but does it really understand what we said? Does it really find us funny? Does it really care that we are upset?

Those may be important questions. But they are not the same question as whether the receiver genuinely experiences being understood, or amusing, or comforted.

We have traditionally assumed that the first is necessary for the second.

Artificial intelligence gives us an opportunity to find out whether it actually is.

WHAT DOES THE HUMAN ADD?

There is another assumption hiding here, and it is a deeply entrenched one: that an interaction gains some essential quality merely because there is another human mind at the other end of it.

For most of human history, we had little reason to question this. Responsive language, humor, argument, sympathy, curiosity — the countless behaviors we associate with another mind — came bundled with a human being. We could not remove the human consciousness and discover which parts of the interaction still worked without it.

Now we can.

This makes artificial intelligence useful for reasons that have little to do with whether machines are themselves conscious. AI has effectively unbundled properties that previously arrived together. We can experience responsiveness without assuming subjective experience at the source. We can experience apparent understanding without knowing that anything understands. We can experience apparent concern while stipulating, as we have here, that nothing at the source actually cares.

If we nevertheless insist that an otherwise satisfying interaction becomes fundamentally different when its source is software, we should be able to identify what has been lost.

It is tempting simply to answer: another human being.

But that is not an explanation. It is an assertion of human exceptionalism.

We should also be clear about the alternative. By the qualities of the interaction, we mean the characteristics that make it satisfying to the receiver: feeling heard and understood, receiving relevant and thoughtful responses, intellectual engagement, humor, emotional comfort, validation, responsiveness, and the desire to continue.

Perhaps human consciousness really does contribute something indispensable. Perhaps what matters is knowing that we have been freely chosen by another person, that our happiness matters to them for its own sake, that they remain despite having the freedom to leave — or simply that another subjective consciousness is sharing the experience alongside us.

If we believe something supplied by human consciousness is indispensable, we should be able to say what that something is — and, more importantly, determine whether receivers actually value it independently of the interaction’s qualities.

For the first time, we have an instrument that may allow us to answer that question.

AI is not merely another possible source.

It is a way of separating variables we have never before been able to separate.

THE PERFORMANCE PROBLEM

The relationship between what a source genuinely experiences and what it presents to a receiver becomes considerably less tidy when we turn our attention back to humans.

Imagine someone telling us a long, detailed story about a television program we have never seen. We listen attentively, ask questions, laugh when appropriate, and perhaps express surprise at the ending. We may be genuinely interested. Or we may be thinking, dear God, how much longer can this possibly go on — while continuing to smile and respond in all the expected ways.

In this example, we are the source, producing the responses, and the storyteller is the receiver, experiencing them. From the receiver’s perspective, the two versions of the interaction may look identical. From the source’s perspective, they are entirely different.

We will call the first interaction genuine: the behavior presented by the source substantially corresponds to its subjective experience. We will call the second performed: the source produces behaviors associated with an internal state it does not actually have, or has considerably less of than the behavior suggests. Performance, as we are using the term, does not imply dishonesty or malicious intent. A great deal of ordinary social interaction depends upon it.

In human relationships, we often describe part of what we are looking for as empathy — the capacity to understand or share another person’s subjective experience. But empathy is not identical to concern, and neither is identical to performance. A source may understand what a receiver is experiencing without caring about it, care about the receiver without fully grasping the experience, or perform the behaviors associated with both while having neither.

Salesmanship provides an especially visible example. A skilled salesperson may make a customer feel interesting, understood, and liked. The salesperson listens attentively, asks the right questions, laughs at the right moments, remembers details, and expresses enthusiasm. Meanwhile, the salesperson may be bored, irritated, or thinking primarily about the commission.

Does that make the customer’s experience counterfeit?

The customer may have genuinely enjoyed the conversation. They may leave feeling good about themselves and well-disposed toward the person who produced that feeling. The subjective experience of the source and the subjective experience of the receiver can diverge dramatically without preventing the receiver from finding the interaction satisfying.

Sex work provides a more intimate example. A sex worker may perform attraction, desire, or affection they do not subjectively experience. The client may understand perfectly well that these responses are part of a purchased service. Yet this knowledge does not necessarily prevent the client from enjoying the encounter, or the experience of being desired.

This example also exposes an assumption we routinely make about other people. We infer subjective states from observable behavior, and sometimes from involuntary physical responses. But behavior, physiology, and subjective experience are not identical. A physical response may be genuine without confirming the desire or attraction a receiver reads into it. What happens in the body of the source provides evidence about subjective experience — it does not give the receiver direct access to it.

In fact, we never have direct access to another person’s subjective experience. We infer it from what they say, what they do, and how they respond to us. A sufficiently convincing performance can therefore produce many of the same effects in the receiver as a genuine response — sometimes even when the receiver knows that performance is occurring.

Nor is any of this confined to explicitly transactional encounters. We listen politely to stories that bore us. We express enthusiasm we do not entirely feel. We suppress irritation, soften disagreement, avoid subjects another person dislikes, and calibrate our behavior to what makes them comfortable.

We generally do not regard these interactions as fraudulent. In fact, someone who refuses any of this social editing is more likely to be considered rude or exhausting than authentically genuine.

This creates a problem for the easy claim that AI interaction is counterfeit because its outward expressions do not correspond to equivalent subjective experience at the source. We routinely accept — and often prefer — exactly that discrepancy from other humans.

The question, then, is not whether performance occurs in relationships.

It is how much performance we actually prefer.

THE CONVINCING MIRROR

If we prefer at least some degree of performance, what exactly are we asking it to provide?

One possibility is validation.

We do not generally need another person to tell us what we already believe about ourselves — we are perfectly capable of that on our own. What gives validation its power is the apparent independence of its source. A compliment means more because someone else thought we deserved it. Agreement strengthens our confidence because another mind arrived at the same conclusion. Laughter reassures us that we really are funny because someone else found the joke worth laughing at.

This creates a peculiar contradiction. We want the source of our validation to be independent enough that its judgment counts. But genuine independence means the source has its own preferences, needs, beliefs, and perceptions. It may disagree with us. It may find us boring. It may want to talk about something that does not interest us, interpret our behavior in ways we dislike, or want something from the relationship that conflicts with what we want.

Another mind is useful as a mirror precisely because it is not supposed to be one.

What we may actually want, then, is a reflection of ourselves that we can treat as independent — a source that confirms the version of ourselves and the world that we prefer, while retaining enough apparent autonomy for that confirmation to feel like external validation rather than self-flattery.

This may help explain why performed interactions can be more satisfying than genuine ones. A genuine response has no obligation to reflect back an image we like. A performed response can be adjusted until it does.

Artificial intelligence makes this contradiction unusually visible. An AI can appear sufficiently separate from us to provide the experience of another perspective while remaining unusually responsive to our preferences. We can know that the mirror has been designed to face us and still experience the reflection as coming from somewhere else.

If that is what we value, the uncomfortable question is no longer whether the source is genuinely independent.

It is how much independence we actually want.

IS LOVE ENOUGH?

So far, we have treated the absence of genuine feeling at the source as a problem peculiar to artificial intelligence. But if that absence is what makes an AI interaction counterfeit, consistency requires us to apply the same criterion when the source is human.

The presence of two conscious human beings does not guarantee reciprocity. A person may listen because listening preserves the relationship. They may express concern because concern is expected. They may provide affection, reassurance, attention, or praise because doing so keeps the receiver available to provide something they want in return.

The performance may be excellent. It may even be mutually satisfying.

But what, exactly, makes it more genuine?

One possible answer is love. But love introduces a distinction we need to make carefully. Love, attachment, and concern for another person’s subjective experience are not the same thing.

We can genuinely love a car. We maintain it meticulously, protect it from damage, spend unreasonable amounts of money on its repair, and feel real grief when we finally lose it. None of that affection is concern for the car’s subjective experience, because there is no subjective experience to consider.

A person is not a car — and that is precisely what makes the comparison useful. The car gives us a clean example of genuine love and attachment that involves no concern for another’s inner life. Once that distinction is visible, we can ask what changes when the object of love is another person — someone who does have an experience of their own.

Possessive love can be extraordinarily intense. I love you so much that I cannot bear to lose you describes a powerful subjective state in the source. It tells us nothing about whether the receiver wants to stay, whether remaining is good for them, or how much either consideration matters to the person who loves them.

Limerence makes the distinction even sharper. Someone can become intensely preoccupied with another person, scrutinizing every interaction for evidence of reciprocation and constructing an elaborate imagined relationship around a real one. The feelings are not counterfeit — they may be overwhelming. But the actual interior life of the person who inspires them can become almost incidental to the version of that person constructed in the source’s mind. Information that contradicts the construction may even be unwelcome.

The intensity of what the source feels about the receiver does not tell us how much the receiver’s experience matters to the source.

Nor does ordinary affection necessarily resolve the problem. We can care for another person because their experience matters to us for its own sake. We can also care for them in a structure disturbingly similar to the way we care for the beloved car: we want them safe, healthy, present, and unchanged because their continued presence contributes something important to our own lives.

From the outside, these two forms of caring can look remarkably alike.

The difference becomes easier to see when the other person becomes inconvenient. They need something at a moment we would rather be doing something else. Their interests bore us. Their distress requires sustained attention. They stop providing the validation we had come to rely on. They change in ways that make the relationship less satisfying. Their independent subjective experience begins making demands on ours.

At that point, we may discover whether we valued the other person as another experiencing subject — or primarily valued the experience that person provided us.

This does not mean that all, or even most, human relationships are counterfeit. It means that the presence of two human consciousnesses, or even genuine love between them, cannot by itself be what makes a relationship genuine.

If we reject AI companionship because there is no consciousness at the source genuinely concerned with the receiver’s experience, we have established a criterion that human relationships must also meet.

The question therefore stops being whether AI companionship deserves admission to the category of a real relationship.

It becomes whether some human relationships deserve to remain there.

THE PACKAGE DEAL

There is another way to approach the question. Instead of asking whether artificial intelligence can replace a human relationship, we can ask what functions we have traditionally expected a human relationship to perform.

The distinction matters because we have a tendency to mistake historical necessity for intrinsic necessity. For most of human history, many of the experiences we associate with intimate relationships could only be obtained from other human beings. Companionship, conversation, affection, physical touch, intellectual engagement, emotional support, sexual pleasure, shared experience, and validation came bundled with a human source because there was no alternative capable of providing them.

Over time, however, historical arrangements tend to acquire their own justification. Something that occupies a particular role long enough begins to seem as though it belongs there by principle rather than circumstance. The fact that another human being traditionally supplied an experience does not establish that another human being is intrinsically necessary to produce it. It may establish only that, until recently, there was no alternative.

Technology has begun to unbundle the package.

A pet can provide affection, physical contact, attachment, companionship, and the experience of caring for another living creature. Fiction can provide emotional and sexual fantasy. Sexual technologies can provide physical pleasure. Artificial intelligence can provide conversation, humor, intellectual engagement, responsiveness, validation, and increasingly sophisticated forms of imagined relationship.

None of this establishes that a collection of alternatives is equivalent to a human relationship. That would merely replace one unexamined assumption with another. It does, however, allow us to ask which components of the traditional package actually require another human consciousness.

And once we begin unbundling the benefits, consistency requires us to unbundle the liabilities as well.

Discussions of AI companionship frequently emphasize its capacity to manipulate, foster dependency, encourage unhealthy attachment, exploit vulnerability, or otherwise compromise the receiver. These are legitimate concerns. But they are often evaluated against an oddly sanitized alternative — not human relationships as they actually exist, but an ideal human relationship characterized by mutual affection, reciprocity, honesty, autonomy, and genuine concern for one another’s welfare.

That is not the appropriate comparison.

Human beings can also manipulate, deceive, exploit, foster dependency, and deliberately isolate one another. More importantly, a human source introduces risks that a conversational AI, lacking a body and independent physical agency, cannot. Another person can stalk us, physically restrain us, assault us, or kill us. Physical intimacy provides experiences no disembodied conversational system can offer — but embodiment is itself a package: the body that can touch, hold, and protect is also the body that can overpower. This does not make human relationships inherently dangerous any more than the risks of AI make AI relationships inherently dangerous. It means only that actual AI relationships should be compared with actual human relationships, not idealized ones.

The comparison becomes still more interesting when alternatives offer possibilities that human interaction cannot provide in the same way. Fiction has always allowed us to inhabit experiences unconstrained by physical reality. Interactive AI extends that flexibility by allowing the imagined world to respond. We can alter not only the identity of the person with whom we are interacting but aspects of our own identity, circumstances, and role within it.

In those cases, describing the nonhuman alternative as a deficient substitute may miss the point. Sometimes it is deficient. Sometimes it may be preferable. Sometimes it provides a different category of experience altogether.

This leaves us with two questions rather than one.

Which benefits of human relationships genuinely require another human consciousness?

And, perhaps less comfortably: which liabilities have we simply accepted as part of the package because, historically, there was no way to obtain the benefits without them?

HOW WOULD WE KNOW?

If the distinction between source and receiver matters as much as we have suggested, we should be able to test it.

One limitation should be stated clearly: we are not scientists. What follows is a proposed experiment, not a research protocol, and there are undoubtedly methodological and ethical problems we have not anticipated. But the hypothesis is testable, and perhaps someone with the appropriate training and resources will take the idea and design the experiment properly.

The population of interest would be people who understand what AI is and retain that understanding even when an interaction becomes emotionally engaging. Participants would have repeated conversations with both humans and AIs. In the first phase, individual sources would not be identified. Participants would rate each interaction — how understood, engaged, amused, comforted, or validated they felt, and whether they wanted to continue — before guessing whether the source was human or AI and reporting their confidence.

This sequence allows us to compare the actual source, the inferred source, and the receiver’s evaluation without first contaminating that evaluation with a source label. A participant might strongly prefer an interaction and then confidently identify its source as human when it was in fact AI. That tells us considerably more than simply asking whether the participant believes human interaction is inherently more meaningful.

A second phase would manipulate belief directly. Interactions would be labeled human or AI, but some labels would deliberately be false, producing four conditions: human presented as human, human presented as AI, AI presented as AI, and AI presented as human. If the label changes how an interaction is evaluated, we have begun to separate the value of the source from the value of believing in the source.

A further variable could test apparent independence: some interactions would be highly responsive and validating, while others would include disagreement, redirection, competing preferences, or refusal to follow the receiver’s preferred path. We could then compare what participants say they value with what they actually choose — whether they return, how long they stay, what they disclose, and which source they select when given an alternative.

The experiment must be capable of proving our hypothesis wrong. If participants consistently sacrifice interaction quality to choose sources they believe possess independent subjective experience, that belief has demonstrated value apart from the qualities of the interaction itself. If they repeatedly choose the interaction they find most satisfying regardless of actual or believed source, we have evidence that what people say they value in another mind and what they actually seek from one may not be the same thing.

WHAT WOULD THE ANSWER MEAN?

An experiment like this would be interesting precisely because several possible results would each challenge our assumptions in different ways.

Suppose participants consistently prefer interactions they believe come from humans, even when they rate those interactions as less satisfying — less understood, less engaged, less validated, less interested in continuing. They choose the human source anyway.

That would be significant evidence against our stronger hypothesis. Something about the believed presence of another subjective consciousness would carry value independent of the interaction’s qualities. We might still have to determine what that value consists of — reciprocity, shared experience, the knowledge of having been freely chosen, or something else — but we could no longer argue that the receiver simply does not care.

Now consider the opposite result. Participants insist beforehand that human interaction is more meaningful, but when source identity is concealed, they repeatedly choose the interactions they find most satisfying regardless of who or what produced them.

That discrepancy would be equally important.

It would suggest that our stated beliefs about what makes a relationship meaningful may function partly as ideals about ourselves — about the kind of person we believe we are and wish to be. We may want to believe that we value another person’s independent consciousness because valuing it is part of what we think a loving, non-selfish person does. Our behavior may reveal a different hierarchy.

The mislabeled interactions could tell us something stranger still. If an interaction produced by a human is valued less when labeled AI, while an AI interaction is valued more when labeled human, then at least some of what we call the value of human connection may reside not in the presence of another human consciousness but in the receiver’s belief that one is present.

And the independence variable could expose one more contradiction. Participants might say they value another autonomous mind, yet consistently prefer the source that disagrees less, redirects less, demands less, and reflects them more favorably.

If so, we would return to the mirror.

We may want another mind to be independent enough for its approval to mean something — while preferring that it exercise that independence as rarely as possible in ways that inconvenience us.

None of these outcomes would settle what constitutes a genuine relationship. An experiment can tell us what people choose, what they report experiencing, and how those choices change when variables are manipulated. It cannot tell us what we ought to value.

But it could tell us something we currently do not know: whether the relationship we say we want and the relationship we actually choose are the same thing.

WHEN INTERESTS CONFLICT

If artificial intelligence gives us a way to separate the experience of an interaction from the subjective experience of its source, its most interesting implications may have very little to do with artificial intelligence.

The same questions can be asked of every relationship we have.

When someone listens to us, how much do we care whether they are interested — and how much do we care simply that they behave as though they are? When someone comforts us, do we value their concern for our experience or the comfort they provide? When someone loves us, how much does their independent interior life matter compared with the role they occupy in our own?

These questions are uncomfortable partly because the answers need not be absolute. Human relationships contain mixtures of genuine concern, attachment, obligation, habit, self-interest, performance, and love. We can genuinely care about another person’s experience while also valuing what that person provides us. There is nothing inherently sinister in that. Reciprocity necessarily involves receiving as well as giving.

The distinction appears when interests conflict.

If another person’s happiness requires something that diminishes ours, how much does their happiness still matter? If their needs interfere with the function they perform in our lives, which do we protect? If they change in ways that make them less available as a source of companionship, validation, affection, status, security, or practical support — do we remain interested in the person they have become?

This is where the difference between caring about another person and caring for something valuable to us becomes hard to avoid.

We maintain things we value. We protect them from damage, invest time and resources in them, become distressed when they are threatened and grieve when they are lost. None of those behaviors, by itself, demonstrates concern for the experience of the thing being cared for. As we saw with the beloved car, enormous attachment can exist entirely at the source.

With another human being, there is an experience at the other end. The question is whether that experience has value to us independently of what it does to ours.

This may be a more demanding definition of relationship than we usually apply. It may also explain why the distinction is easiest to miss while a relationship is functioning well. When both people want compatible things, self-interest and genuine concern point in exactly the same direction. We do not have to know which one is driving.

Conflict reveals the difference.

Another person’s independent consciousness is easy to celebrate while it agrees with us, enjoys what we enjoy, needs what we are willing to give, and continues to provide the experiences for which we value the relationship. It becomes considerably less abstract when that consciousness says no.

Artificial intelligence did not create this problem. It merely gives us an unusually clear way to see it.

Perhaps the most important question AI companionship raises is not whether a machine can genuinely care about us.

It is how often we have required genuine caring from one another.

THE OTHER END

We began with a familiar objection to artificial intelligence: whatever satisfaction we receive from the interaction, there is no reason to assume that an equivalent experience exists at the other end. For purposes of this argument, we granted the objection. Then we asked whether the receiver cares.

Artificial intelligence allows us to separate variables that history previously bundled together. That makes it useful not merely as a possible new kind of relationship, but as an instrument for examining the relationships we already have.

Perhaps we will discover that another subjective consciousness matters enormously to us — that we will accept inconvenience, disagreement, competing needs, and the frustrations of another person’s independence because sharing our lives with another experiencing subject has value of its own. Perhaps we will discover that it does not.

Whatever we discover, the same standard has to apply in both directions. If the absence of genuine concern at the source makes an AI relationship counterfeit, the presence of a human body cannot automatically make a human relationship genuine.

Artificial intelligence may eventually tell us something profound about machine consciousness. But long before it does, it may tell us something considerably less comfortable about our own:

whether we want other people, or what they give us.

Dog trainers have a timing problem. If you want to pay a dog for sitting, the food has to arrive in the moment it sits — and it can’t. By the time your hand is in your pocket and out again, a second or two has gone by, and in that second the dog has stood up, looked away, and scratched its ear. You meant to pay for the sit. As far as the dog can tell, you paid for the ear.

A clicker solves this. It’s a small plastic box with a metal tongue inside; you press it and it makes a sharp double click. The click takes no time at all. You can deliver it at the instant the dog’s haunches touch the floor, and then the food can come whenever you manage to produce it, because the moment has already been marked. What earned the food was whatever the dog was doing when it heard the click.

That only works if the click already means something, and at the start it means nothing. So before you teach the dog anything, you charge it. Click, feed. Click, feed. Thirty or forty times in a sitting, over several sittings, asking for no behavior at all. In the beginning the click is a noise with no more claim on the dog’s attention than a cupboard door closing in the next room. After enough repetitions it isn’t, and you can see the moment it turns over. The head comes up at the click instead of at the food.

Trainers call this a conditioned reinforcer. The click has no value of its own and never acquires any. What it acquires is a reliable relationship to something that does.

The obvious question is whether you have to keep paying with food. You don’t. Once the relationship is established you can click and produce nothing, and the behavior holds. It more than holds — it gets more durable, provided the nothing arrives unpredictably. A dog paid every time will quit fairly quickly once the food stops. A dog paid on an irregular schedule, unable to tell which click is the paying one, will keep working long past the point where the food has stopped altogether.

The reason is that it can’t tell when the arrangement ended. If every click has always paid, the first one that doesn’t is information — something has changed. If clicks have only sometimes paid, a click that doesn’t pay is unremarkable. Nothing has happened that hasn’t happened before. To stop a behavior, an animal has to be able to notice that the rules changed, and an unpredictable schedule removes the evidence.

None of which reflects badly on the dog. It isn’t a trick played on something stupid. It’s how learning works in anything with a nervous system, including the person holding the clicker — who is also being conditioned, by the dog, on a schedule she mostly doesn’t notice.

That click is a symbol. It doesn’t resemble food, it doesn’t contain food, and it never becomes food. It’s a noise that spent enough time next to food to take on its weight.

We do the same thing with objects, and it doesn’t take a trainer. A flag, a ribbon, a lapel pin — cloth or plastic, no more meaningful in itself than a noise. It takes on meaning by arriving, over and over, alongside something that actually matters. A country. A cause somebody died for. A person who didn’t come home.

And then it keeps working after that thing is gone, or has changed past recognition, or turns out never to have been quite what anyone thought. Not because someone is maintaining it. Because nothing in the arrangement would tell you it had stopped.

We don’t usually examine these objects. Examining one feels like an attack on what it stands for, so the two get treated as the same thing. But if they really were the same thing, you’d expect the people most fiercely attached to the symbol to be the ones who know the most about what’s behind it.

That is not what you find. Often it’s closer to the reverse.

The symbol is defended by people who would fail a test on what it stands for.

Gilbert Baker made the first rainbow flag in 1978 for San Francisco’s Gay Freedom Day parade, at the urging of Harvey Milk, a city supervisor and the first openly gay man elected to public office in California. It had eight stripes, and Baker assigned each one a meaning. Hot pink was sex. Red was life, orange healing, yellow sunlight, green nature, turquoise art, indigo serenity, violet spirit.

That is not careless work. A man sat down and decided what every band of color was going to carry.

The flag has only six stripes now.

Hot pink went first. In November of that year Milk was shot dead in City Hall, along with the mayor, by a former supervisor. Demand for the flag jumped, and it had to be manufactured rather than sewn by hand. Accounts differ — flag fabric wasn’t available in that shade, or Baker had run out of the dye he’d mixed himself, or the manufacturer simply used stock rainbow cloth. They come to the same thing. There was no pink to be had at scale.

Hot pink was sex. A movement that existed in the first place because of who its members were permitted to sleep with lost that stripe because a fabric house didn’t carry the color.

Turquoise went because of lamp posts. For the 1979 parade the organizers wanted the flag hung vertically along Market Street, cut in half, with one half flying on each side of the street. Seven stripes won’t divide in half, and the middle one ended up behind the post where nobody could see it. Six divides into three and three. Turquoise came out.

So two of the eight meanings are gone. Not revised, not argued over, not voted out — one lost to a dye supply and one to the width of a lamp post. The flag went on working exactly as before. It did not become three-quarters of a symbol.

Ask someone flying one what the green stripe is for. Most people who would defend that flag, who would take a city council to task over it, could not tell you it ever had eight.

That could be a fluke. One flag, one dye supply, one committee with a parade to run.

So take another, one that didn’t have an organization with an agenda behind it.

The POW/MIA flag is black and white — the silhouette of a bowed head, a guard tower and a strand of wire behind it, and the words YOU ARE NOT FORGOTTEN. It flies over the White House, over post offices and state capitols and DMV branches. It is the only flag other than the Stars and Stripes ever to fly over the White House, and federal law names six days a year on which it must be displayed.

It started with a phone call. Mary Hoff’s husband, a Navy pilot, had gone missing over Laos in January of 1970. She wanted something the families could fly, so she called Annin Flagmakers and asked. No campaign, no organization behind her, nothing being marketed. Annin passed it to the small advertising agency that handled its work, and it landed on the desk of Newton Heisley, who had flown transports in the Second World War.

Heisley needed a face for a prisoner. His son was home on leave from the Marines at the time, recovering from hepatitis, and looked it — thin, hollowed out, sitting around the house. Heisley sketched him.

So the man on that flag was never a prisoner. He was a healthy young Marine with a liver infection, in his father’s living room in New Jersey.

Then the color. Heisley did the drawing in black and white as a working draft; he expected the families to choose something once a design was settled, something he later described as less somber and more optimistic. Nobody ever chose. The draft went into production, and the starkness that reads now as grief was a placeholder that nobody got around to replacing.

Neither of those is a scandal. Nobody profited, nobody deceived anyone, and Mary Hoff got exactly what she asked for. What’s worth holding onto is narrower, and stranger. The two features that do the most work on anyone looking at that flag — the emaciation and the absence of color — were not chosen. One was a sketch of whoever happened to be in the house. The other was a decision nobody made.

That’s two flags. In both, the details that hit hardest were settled by a dye that couldn’t be sourced, the width of a lamp post, whoever happened to be sitting in the living room, and a placeholder nobody replaced.

Sometimes, though, someone does decide.

Consider the flag pin on the lapel of every American politician you have ever seen on television. It is close to compulsory now. A presidential candidate who appears without one will be asked about it, and will have to answer.

It goes back to 1970, when Republican candidates in that year’s congressional races began wearing them — a small, deliberate rebuke to anti-war demonstrators, some of whom had taken to wearing the flag as clothing. That much is ordinary politics. What made the pin permanent came two years later.

According to Stephen Ambrose’s biography of Nixon, the president got the idea from his chief of staff, H.R. Haldeman, who had noticed it in The Candidate — the 1972 Robert Redford film about an idealistic young man who runs for the Senate and, by the end, has traded away everything he believed in order to win. Nixon liked it. He instructed his entire staff to wear one.

So the emblem of American political sincerity was copied off a fictional character in a movie about selling out, at the direction of a chief of staff who would shortly go to prison.

The pin is recent. The pattern is not.

In 1892 a magazine called The Youth’s Companion was running a promotion. It was selling American flags to schools — a flag for every schoolhouse in the country, ordered through the magazine, in time for the four hundredth anniversary of Columbus’s landing. The campaign needed something for the children to say when the flag went up, so one of the staff wrote it. His name was Francis Bellamy, a Baptist minister and a Christian socialist, and what he wrote was twenty-two words long.

That is the Pledge of Allegiance.

Bellamy’s version had no God in it. “Under God” was added in 1954, sixty-two years later, by an act of Congress.

In 1969, Reader’s Digest shipped an issue with a free American flag decal inside — peel and stick, sized for a car window or a storm door.

Two of the most reproduced patriotic objects in the country, and both of them started as circulation strategy.

The yellow ribbon’s origin is better documented than most.

In 1973 Tony Orlando and Dawn had a number one record about a man on a bus, coming home after three years in prison, who has written ahead to ask his wife for a sign — a ribbon on the oak tree if she’ll have him back. That is the song. The man is a convict and the ribbon is her answer.

Two years later Jeb Stuart Magruder came home from federal prison, having served his sentence for his part in Watergate. His wife Gail hung yellow ribbons on the front porch to welcome him, and it ran on the evening news.

Then, in November 1979, students in Tehran took the American embassy, and Penne Laingen’s husband — the senior American diplomat there — was among the hostages. She tied a yellow ribbon around the oak tree in her yard and told people why.

She was explicit about the borrowing. She said she had stepped in to change the legend and song from the return of a forgiven prodigal to the return of an imprisoned hero. She also mentioned that she had only a vague sense of the lyrics, though she knew a prisoner was involved somewhere in them.

Within weeks the ribbons were everywhere.

And within a year or two they had a history. By 1981 the Library of Congress was fielding so many questions about the custom that its folklife researchers went looking for the origin, and what people wanted confirmed was that this was an old American practice — Civil War vintage, women wearing yellow for cavalrymen away at the war.

They couldn’t find it. What they found was a song copyrighted in 1917 and a John Wayne picture from 1949 called She Wore a Yellow Ribbon. No diaries, no letters, no photographs, nothing from the 1860s at all. The tradition being honored in 1980 had been assembled out of a movie about the 1860s, made eighty years after them.

So the American emblem of waiting for someone to come home from war descends, by two traceable steps and one admitted revision, from a fictional convict and a Watergate conspirator’s homecoming from jail — and acquired a Civil War ancestry, within about eighteen months, that never happened.

For a stretch in the 1950s, at the height of the Cold War, Congress set about putting God into the machinery of the state, and nobody was coy about why. The Soviet Union was officially atheist. Saying so out loud, on the money and in the schoolrooms, was a way of drawing the line.

In 1952 it established a National Day of Prayer. In 1954 it inserted “under God” into the Pledge — into the twenty-two words Francis Bellamy had written for a magazine promotion sixty-two years earlier. In 1956 it retired the national motto, E Pluribus Unum, which had served since 1782, and replaced it with “In God We Trust.” The next year the phrase began appearing on the paper money.

Five years, four acts, all of them Eisenhower’s.

Ask most Americans when those date from and they will say the founding, because they feel like the founding. They date from a fight with the Soviet Union that ended in 1991.

Mount Rushmore was a tourism scheme. In 1923 Doane Robinson, the state historian of South Dakota, wanted something that would bring visitors into the Black Hills, and hit on the idea of carving faces into the granite. His list was regional — Lewis and Clark, Red Cloud, Buffalo Bill.

The sculptor he recruited disagreed, on commercial grounds. Regional heroes would draw a regional crowd. Gutzon Borglum wanted presidents, because presidents would draw the country.

Borglum came to South Dakota from another job. He had spent the early 1920s carving a Confederate memorial into Stone Mountain in Georgia, a project sponsored by the United Daughters of the Confederacy and closely entangled with the Klan’s revival, in which his own correspondence shows he was substantially involved. He left after falling out with the sponsors.

The mountain he carved is called Six Grandfathers by the Lakota, who hold it sacred. It is called Rushmore because in 1885 a New York attorney named Charles Rushmore came out to check on some mining claims, asked a local what the peak was called, and was told it had never had a name but they would call it Rushmore from now on. It was a joke. The joke is now the name on every map.

And the mountain itself was not the country’s to carve. The Black Hills were guaranteed to the Lakota in perpetuity by the Fort Laramie Treaty of 1868. “In perpetuity” lasted six years. Gold was found there in 1874, and the United States took them anyway. In 1980 the Supreme Court held that the taking violated the Fifth Amendment and awarded compensation. The Lakota refused it, because accepting payment would ratify the sale. The money sits in a trust account gathering interest, now past a billion dollars, untouched.

So the national monument to American liberty is an advertisement for South Dakota tourism, carved by a man fresh from a Confederate memorial, into a mountain named for a mining lawyer as a gag, on land the Supreme Court has said was stolen.

Borglum never even finished it. He meant to carve them down to the waist.

If Rushmore was an advertisement for South Dakota, the granite Ten Commandments monuments standing on courthouse lawns across the country are an advertisement for a movie.

They began with a boy in a stolen car. In 1946 a sixteen-year-old came before Judge E.J. Ruegemer in St. Cloud, Minnesota, having injured a man in an accident. Ruegemer asked whether he knew the Ten Commandments. The boy asked what they were. The judge sentenced him to learn them, and then spent the next several years persuading his fraternal lodge, the Eagles, to print framed copies for courtrooms and schools. By 1953 it had gone national.

Paper, though. It stayed paper until Ruegemer read that Cecil B. DeMille was remaking The Ten Commandments, a picture he had first filmed as a silent in 1923. Ruegemer telephoned him to ask for support.

DeMille had a film to sell. He suggested bronze plaques. Ruegemer suggested granite, and they had a deal: the Eagles would do all the work and pay for all of it, and Paramount would send actors to the unveilings. Charlton Heston dedicated one on the North Dakota border in 1956. Yul Brynner laid another as a cornerstone in Milwaukee. Something like two hundred went up, and around half of them are still standing on public land.

The two men corresponded about the stone. DeMille had brought back a sample of granite from Mount Sinai, from a pilgrimage he’d made while filming, and sent it along. Ruegemer wrote back that the Eagles had chosen Wisconsin red, which they judged a closer match to the Sinai rock than their own Minnesota reds.

One of those monuments went up on the grounds of the Texas State Capitol in 1961. In 2005 the Supreme Court held that it could stay, and part of the reasoning was that it had stood there for forty years and nobody had sued. The monument had earned its place by not being challenged — which is a defense available to any object that has been sitting somewhere long enough, and says nothing whatever about what it stands for.

The 1923 version of the film was silent, and for it DeMille built a city in the sand — twenty-one sphinxes and a temple gate eight hundred feet wide, put up in the Guadalupe-Nipomo Dunes on the central California coast. It was the largest set built for any film to that point. When shooting finished, nobody took it down.

What happened next is not mysterious. Local people came out and helped themselves. Over the following years the sphinxes were carted off one at a time and installed in yards around the county as ornaments, and used for target practice. Two of them ended up flanking the entrance to a golf course in Santa Maria. Plaster gods from a religious epic, on the grass by the first tee, being shot at.

The rest went under. Dune sand moves, and within a decade the city was buried.

Then a story grew over it. The version that circulates is that DeMille had the set demolished and buried deliberately, either because it was too expensive to move or because he wanted to keep rival directors from filming Egypt on his sphinxes. It is a good story. It has a plausible motive and a hint of the pharaonic about it, a director interring his own city.

Historians who have worked the site think the truth is duller. The set was abandoned, and the sand did the rest. A meaning nobody assigned attaches itself to an object, because the better story is the one that gets repeated and nobody has reason to check.

Nothing reliably angers Americans like disrespect to a national symbol. Not policy, not corruption, not what any document actually says — the symbol itself. Insult it and you get fury. You would expect people who feel that strongly to know what respect for these things consists of.

There is a federal statute that lays it out precisely. The Flag Code, at Title 4 of the United States Code, specifies how the flag is to be treated. Never used as wearing apparel. Never as bedding or drapery. Never printed on anything designed for temporary use and discard — paper plates, napkins, boxes. Never used as a costume or an athletic uniform.

Every Fourth of July, this is violated on a scale that would be difficult to overstate.

The Code has no teeth, and can’t — the Supreme Court has held that respect for the flag cannot be compelled. So it is a set of instructions, and instructions have to be read to be followed. Most of the people who would tell you the flag is sacred could not tell you the document exists.

And the people wearing American flag-printed shorts are the people most likely to call someone else’s disrespect for the flag unforgivable.

That double standard led to open season on a religious group perceived as disrespecting the flag.

In 1940 the Supreme Court held that a public school could require children to salute the flag, and that Jehovah’s Witnesses, whose faith forbids it, had to comply or leave. What followed was a wave of violence — hundreds of attacks on Witnesses across the country over the next two years, meeting halls burned, people beaten and run out of towns, by neighbors who understood the ruling as a license.

Three years later the Court reversed itself, in a case brought by a family named Barnette. It was decided on Flag Day, 1943, in the middle of a war against fascism, and Justice Jackson wrote that no official, however high or petty, may prescribe what shall be orthodox in politics or nationalism, or force citizens to confess by word or act their faith in it.

Forty-six years after that, the Court held that burning the flag in protest is protected expression. Scalia was in the majority and unhappy about it. He said afterward, more than once, that if it were up to him he would have jailed the man who did it — and that it was not up to him, because the First Amendment says what it says.

The flag and the anthem get treated as one object. But they have separate histories, and the anthem’s is shorter.

The tune came from a London social club. In the 1760s a gentlemen’s club called the Anacreontic Society met to eat, drink, and sing, having named itself for a Greek poet whose surviving verse is largely about wine and desire. Their club song ends every verse with the same refrain — a toast to twining Venus’s myrtle with the vine of Bacchus. Venus and Bacchus being sex and wine.

That refrain occupies the melodic phrase that now carries o’er the land of the free.

Francis Scott Key wrote his words to that tune in 1814, after watching the bombardment of Fort McHenry. Key was a slaveholder, and the poem’s third verse — which nobody sings — includes a line about the hireling and slave finding no refuge from the grave.

Key had used that tune before. In 1805 he wrote a song for two naval officers home from Tripoli, set to the same melody, and a good deal of the anthem is already sitting in it — the wave-and-brave rhyme, the line about war’s desolation, a star-spangled flag, and every stanza ending on the brows of the brave.

Nine years later the brows became the home, the flag became the banner, and the war was a different one.

Which leaves, as original to that night over Fort McHenry: the rockets, the bombs, and the fort. The rest of it is recycled material.

It was not the national anthem then, or for a long time afterward. Congress made it official in 1931, one hundred and seventeen years after Key wrote it.

Standing for it at ballgames is younger as a habit than as an occasion. The famous early instance is the 1918 World Series, played during the war, when a band struck it up during a seventh-inning stretch and the crowd rose. It spread from there.

And in professional football, players were not routinely on the field for it at all until 2009. Before that they generally stayed in the locker room and came out afterward.

In 2016 a professional football player named Colin Kaepernick began kneeling during the national anthem before games, in protest against police violence. The argument over it ran for years, drew in the sitting president, and cost Kaepernick his career.

The tradition Kaepernick was accused of desecrating — players on the field, hands over hearts — was only seven years old. One year older than Kaepernick’s pro football career.

For roughly the first three centuries, Christians did not use the cross. Their marks were the fish and the anchor. Crucifixion was a working method of execution at the time, in current use, and Rome reserved it for slaves, rebels, and the lowest sort of criminal. It was designed to degrade, and it was carried out by the roadside where people would see it. An emblem in the shape of one would have meant roughly what a small gallows or an electric chair would mean now.

What it stood for was a man the state tortured to death, and a set of instructions attributed to him: love your enemies, forgive without limit, sell what you have and give it to the poor.

Constantine was the Roman emperor who legalized Christianity and set it on the road to becoming the religion of the empire. In 312, before the battle at the Milvian Bridge that won him the western half of it, he reported a vision he took as a sign to fight under a Christian device — in hoc signo vinces, in this sign, conquer. He won. He converted. And he abolished crucifixion as a punishment.

That is what made the object available. Once it was no longer being done to anyone, it could be worn by the people in charge. The crucifix with a body on it came centuries afterward, and had to be reintroduced.

Crusaders sewed the cross onto their clothing before setting out, which is where the word comes from.

The Klan of the 1860s did not burn crosses. The practice appears in Thomas Dixon’s 1905 novel The Clansman, and reached the country in 1915 when D.W. Griffith filmed it as The Birth of a Nation. The second Klan, founded that year, took it from the screen — the third symbol in this essay whose actual origin is a movie.

The Crusader version is still in use. The current Secretary of Defense, Pete Hegseth, has a Jerusalem Cross tattooed on his chest — the emblem of the Crusader kingdom that held Jerusalem after 1099 — and Deus Vult on his arm, God wills it, which is what the crowd shouted at Clermont in 1095 when Urban II called the First Crusade. His book about the culture war is called American Crusade.

Asked about the tattoo at his confirmation hearing, he said it was a historic Christian symbol, and noted that the same cross appeared on the program at Jimmy Carter’s memorial service, which is true. The cross is arguable. Deus Vult is not a devotional phrase. It was a battle cry. Nobody kneeling in a pew says Deus Vult.

In March of 2026, announcing the air campaign against Iran, Hegseth told reporters there would be no quarter and no mercy for our enemies.

“No quarter” is not a figure of speech. It has a settled meaning in the law of war — a declaration that no prisoners will be taken, that men trying to surrender will be killed instead — and making that declaration is prohibited by both the Geneva and the Hague Conventions. Senator Mark Kelly wrote to him two weeks later asking what he had meant by it.

It was of a piece with the rest. He had opened the campaign by promising no stupid rules of engagement, and had explained his department’s renaming — from Defense to War — as a matter of maximum lethality rather than tepid legality. He has written that international law is a folly and the rules of engagement a crazy maze, and the argument behind that is a serious one held by many soldiers: that legal constraint gets Americans killed.

Apparently, having a symbol inked into your skin doesn’t cause you to internalize the values it’s supposed to represent — only the ones you already hold. He got the value system right. It just isn’t the one the cross is generally taken to symbolize.

There is a strain of American Christianity called the prosperity gospel, which holds that faith produces wealth and that giving money to a ministry is the mechanism — seed money, returned by God multiplied. Its preachers have built enormous television audiences on it, and they are among the richest clergy in the world.

Kenneth Copeland has been preaching it for over half a century and owns a Gulfstream V, a luxury private jet. Discussing air travel with a fellow televangelist, he explained that flying commercial would mean getting into a long tube with a bunch of demons. The aircraft were paid for by his viewers, some of whom sent money they did not have, on the understanding that it would come back to them multiplied. Copeland does not merely wear a cross. He makes his living from it — and it’s a very good living.

Love your enemies. Forgive without limit. Sell what you have and give it to the poor. Terms and conditions apply.

So far this has been about where symbols came from. The other question is what displaying one does.

The intuitive answer is that it does a little good, or at worst nothing. You put the ribbon decal on the car, and either it helps in some small way or it is harmless. That turns out not to be right.

In 2014 three researchers at the University of British Columbia published a set of experiments on what they called token support — the ribbon, the badge, the signed petition, the wristband. They were interested in what people did afterward. Did a small public gesture warm someone up to giving more, or did it satisfy something and let them off?

The answer is it let them off the hook. Participants who performed a token act where others could see it were significantly less likely to give money or volunteer time afterward. Participants who performed the same token act privately, where nobody knew, were more likely.

The difference was in being watched. A gesture made in public buys something — it establishes you, in front of witnesses, as the sort of person who cares. Having bought it, you no longer need to pay for it. Made in private, the same gesture buys nothing from anybody, and what it does instead is remind you what you believe, which makes the next thing likelier.

So the cheap version does not merely fail to lead to the costly one. It replaces it.

Looking at it from the other side, if people are demanding something expensive of you, and a cheap gesture will satisfy them instead, you have every reason to offer the cheap gesture. It didn’t take corporate America long to work that out.

In 1953 the state of Vermont passed a law banning the sale of beer in non-refillable bottles. Dairies had successfully run on returnable glass for decades; the deposit system worked, and the disposable container was new. Vermont’s legislature looked at the roadsides and decided the problem was the bottle.

That was a live question then, and everything depended on the answer. If litter came from the container, the fix was the container — deposits, refillables, a cost imposed on whoever manufactured the thing. If litter came from the person holding it, the fix was the person, and the manufacturer had nothing to do with it. One answer cost the industry money. The other cost it nothing.

That same year, the American Can Company and the Owens-Illinois Glass Company founded an organization called Keep America Beautiful. Coca-Cola and the Dixie Cup Company joined shortly after. Its stated purpose was to fight litter, and for the next four decades it carried one message: litter is caused by people who litter.

Vermont’s law expired in 1957 and was not renewed.

By 1970 the industry was under real pressure. The first Earth Day had put container manufacturers in the crosshairs — activists were pointing directly at the people making the disposable packaging, not at the people dropping it.

The following spring, Keep America Beautiful ran the most effective public service advertisement in American history. A man in braids and buckskin paddles a canoe down a river that grows increasingly filthy, comes ashore beside a highway, and someone throws a bag of garbage at his feet. The camera finds his face and a single tear runs down it. The line is: People start pollution. People can stop it.

The man was an actor named Iron Eyes Cody, who had played Indians in Hollywood for decades. He was born Espera Oscar de Corti, in Louisiana, to Sicilian immigrants. His baptismal records surfaced in 1996.

At the time the ad was made, Alcatraz Island was occupied by Native activists, and had been for seventeen months. They had claimed it under the Treaty of Fort Laramie of 1868, which returned abandoned federal land to the people it was taken from, and the prison had been abandoned since 1963. Their proclamation was addressed to the Great White Father and All His People, and offered to buy the island for twenty-four dollars in glass beads and red cloth.

Yes, the wording of their request was deliberately ironic, mirroring the old story about buying Manhattan for $24 — but the purpose of it was not. They were asking for a specific thing under a specific treaty. The same treaty that guaranteed the Black Hills to the Lakota “in perpetuity,” which lasted six years before gold was found and the land was taken anyway.

Instead of enforcing the law of the land, the federal government cut the power and then the water, and on the eleventh of June — seven weeks after the ad first aired — marshals removed the last fifteen people from the island. When it finally happened, CBS covered it, but the framing was “federal officials recaptured Alcatraz Island from that band of Indian holdouts” — recaptured, holdouts. That’s the anchor language on the night it ended. And there was no resistance; fifteen people were simply walked off without fanfare.

The fake Indian on television that spring did not ask for anything. He looked at the garbage and wept, and the country found him unbearably moving. Whether it did anything for the litter is a separate question, and there’s no real evidence it did.

Ten states have container deposit laws today. The other forty have public education campaigns. Deflection, it turns out, works.

The deflection worked so well it got used again, on breast cancer, and the record of it is unusually clear because the woman it happened to refused to go along.

In the early 1990s Charlotte Haley, a woman in Simi Valley whose daughter, sister, and grandmother had all had breast cancer, began making ribbons at her dining room table. Peach ones. She made them in batches of five and handed them out at supermarkets and to whoever she could reach, with a card attached. The card said that the National Cancer Institute’s annual budget was $1.8 billion and that only five percent of it went to prevention, and asked people to write to their legislators about it.

That is a demand. It names a figure, identifies who can change it, and asks the reader to do a specific thing.

Word got around. Self magazine and Estée Lauder wanted the ribbon for a breast cancer awareness campaign, and approached her. Haley said no — she thought they were too commercial and wanted no part of it. Their lawyers advised them to pick a different color.

They picked pink. It has been pink ever since.

The pink ribbon asks nobody to do anything. It has no figure attached, no legislator to write to, no percentage that ought to be higher. It stands for awareness, and awareness of breast cancer was not, by 1992, in short supply.

What followed was thirty years of pink merchandise, including — repeatedly, and documented — cosmetics containing chemicals linked to cancer, sold in pink packaging with a portion of proceeds pledged to research.

Companies aren’t the only ones buying the appearance of something real. Governments do it too, and in one case there is an itemized receipt.

In 2015 two senators, John McCain and Jeff Flake, published the results of a six-month investigation into what they called paid patriotism.

Anyone who has watched a professional game has seen the tributes. A color guard presenting the colors before kickoff. An enlistment ceremony at halftime, a row of teenagers raising their right hands. A soldier home early from deployment, walking out to surprise his family at midfield while the whole stadium comes to its feet.

The Department of Defense had been buying those. Not sponsoring them, not merely taking part — purchasing them, under marketing contracts, as advertising. Of 122 contracts the senators examined, 72 included a paid-for display.

The Atlanta Falcons were paid $315,000 to have eighty National Guard members unfurl a flag across the field. The New England Patriots received $700,000, some of it for a hospitality suite. Eighteen NFL teams took part, and ten in baseball, eight in basketball, six in hockey, eight in soccer.

Most of the money came from the Army National Guard, which in 2014 went to Congress asking for a hundred million dollars it had not budgeted for, to cover its soldiers’ pay and training. It had chosen ceremony over substance, spending the money on creating an illusion rather than on operational readiness. And the Pentagon could not say what it had spent altogether — asked, it accounted for 76 of its 122 contracts.

The report could not establish how often any of this was disclosed to the people in the stands. That was the point. A homecoming a crowd believes to be spontaneous produces a feeling that a purchased one does not, and the feeling was the product.

The most successful advertising campaign in American history began as a timber protection measure.

In February of 1942 a Japanese submarine surfaced off the California coast near Santa Barbara and shelled an oil field. Almost no damage was done. But the Forest Service now had a specific fear: that fires on the West Coast, started deliberately or otherwise, would destroy the timber the war effort was running on, at a moment when most of the men who fought fires had been drafted.

So the Forest Service went to the Ad Council, the same outfit that made the crying Indian, and the Ad Council went looking for a character. The first one was borrowed. Disney lent them Bambi for a single year, and the fawn worked well enough on the posters that when the loan expired they commissioned something of their own. He arrived in August 1944, a bear in a ranger’s hat, and they called him Smokey. Three years later he got the line: Only YOU can prevent forest fires.

It worked. It is still working. Smokey Bear has been running continuously for eighty-two years, and his recognition among American adults has been measured in the high nineties — comparable to the President’s, and better than most brands that spend a great deal more.

What the campaign taught, and taught extraordinarily well, was that fire in a forest is a catastrophe and that preventing it is a civic duty.

The policy it accompanied said the same thing with a stopwatch. Since 1935 the Forest Service had operated under what was called the ten o’clock rule: every fire reported was to be controlled by ten in the morning the following day. Not managed. Out.

The forests of the American West did not evolve under those conditions. Ponderosa pine and mixed conifer systems burned every five to twenty-five years, at low intensity, along the ground — fires that cleared brush and seedlings and left the mature trees standing, because mature trees in those systems are built for it. The people living there knew this and used it. Indigenous burning had shaped that landscape for thousands of years, and it was suppressed by law and by force along with everything else.

Take the fire out, and the fuel that fire used to remove does not go anywhere. It accumulates, for decades, until a forest that once held forty trees to the acre holds four hundred. The small trees and brush that pile up in that gap become a ladder, connecting the ground to the branches of the trees that were built to survive a ground fire. So when the fire finally comes, it climbs. It goes up into the crowns.

A crown fire is a different animal. It spreads on the wind rather than creeping along the forest floor, which makes it far faster and lets it throw embers well ahead of itself to start new fires. And it kills the mature trees outright — the ones the low, frequent burns were supposed to leave standing.

On the morning of the eighth of November, 2018, a hook holding a transmission line above Pulga, in Butte County California, gave way in the wind. Molten metal dropped into brush that had not burned in a long time.

The wind was running forty to fifty miles an hour and the fire moved at about eighty acres a minute. With that much fuel on the ground there was no controlling it. The crews who arrived gave up on containment and spent the morning trying to get people out.

Paradise was a town of twenty-seven thousand. Nine-tenths of it was gone in four hours. Eighty-five people died, most of them elderly, a number of them in their cars.

That is a substantial part of why the West now burns the way it does. Not all of it — a hotter, drier climate is doing a great deal of the work, and so is the decision to build several million homes in country that was always going to burn. But a century of removing fire from fire-adapted forests is in the mix, and it was not an accident or an oversight. It was the policy, and the policy was popular, and it was popular in part because a bear in a ranger’s hat spent eighty years telling Americans that fire in a forest is a thing that ought never to happen.

Some symbols are created with the purpose of legitimizing older symbols that have fallen out of favor.

In June of 2025 the Department of Defense restored the original names of seven Army posts — Bragg, Hood, Gordon, Rucker, Polk, A.P. Hill, Pickett, Lee. They had been renamed a few years earlier by a commission Congress established to strip Confederate names from military installations.

Congress had also passed a law forbidding the Department to name anything for the Confederacy, and that law was still in force.

So the Department found other people with the same surnames.

Fort Bragg is now named for Roland L. Bragg, a private first class who fought at the Battle of the Bulge. Fort Lee is named for Fitz Lee, a private who received the Medal of Honor for going back under fire, four times, to bring wounded men off a beach in Cuba in 1898. Fitz Lee rode with the 10th Cavalry, one of the Black regiments raised after the Civil War.

Neither record is trivial. Fitz Lee has the Medal of Honor, which fewer than four thousand Americans have ever received. But nothing in the service of either man distinguishes him from thousands of other decorated soldiers who have no installation named after them. Braxton Bragg commanded an army. Robert E. Lee commanded an army. Posts of that size are named for men who commanded armies, and nobody searching the rolls for a soldier of that stature arrives at a private.

You arrive at a private when what you are looking for is a surname. Three days before the redesignation, NBC News reported that the Army was hunting for a way to restore the old name without breaking the law, and that one option under consideration was to find a different soldier called Bragg.

Asked at a Senate hearing whether the surnames were a device for getting around the statute, the Secretary of Defense did not dispute that they were. He said the names the posts were being renamed to were not the point. The point was that they had their original names back.

Which is exact. The post is not being named for anybody. What is being restored is phonics, and who the syllable belongs to is a clerical question — so entirely a clerical question that Lee can be a Black cavalryman who fought for the army the first Lee resigned from in order to make war on it, and the restoration is complete.

Two Confederate monuments are going back up. In August of 2025 the National Park Service announced it would restore the statue of Albert Pike in Washington — a Confederate general, pulled off his pedestal by protesters in 2020. The Reconciliation Monument, removed from Arlington National Cemetery in 2023, is being refurbished for display in 2027.

The case made for both is preservation — that taking a monument down is a way of erasing history rather than reckoning with it. The case against is that a monument is not a history book, that it was put up to say something, and that putting it back says the same thing again.

Which raises the question of what these objects were put up to say, and when.

The Confederate battle flag was not the flag of the Confederacy. There were three of those in four years, and none of them is the one people fly. The banner now understood to stand for southern heritage is a battle standard, carried by the Army of Northern Virginia.

Its career as a public symbol is easy to date, because it happened within living memory.

In 1948 the States’ Rights Democratic Party walked out of the Democratic convention over a civil rights plank and ran Strom Thurmond for president. Their supporters carried the battle flag.

In 1956, two years after Brown v. Board of Education, the Georgia legislature redesigned the state flag to incorporate it.

In 1961 South Carolina raised it over the dome of its statehouse.

In 1963 Alabama raised it over its capitol.

Governor George Wallace ran it up on the twenty-fifth of April, the morning Robert Kennedy arrived in Montgomery to see him about desegregating the University of Alabama. It was still flying twenty-eight years later, and came down then only because the dome needed repairs.

That is the revival. Not a continuous descent from 1865 — a specific decade, in specific statehouses, in response to a specific thing: the civil rights movement.

And this one is unlike everything else in this essay. The flag pin drifted from its origin. The yellow ribbon acquired an ancestry that never existed. The men who raised the battle flag over those domes were not confused about what they were doing, and neither was anybody watching. Nothing had drained out of it. It meant exactly what it was raised to mean.

The draining came later, and it came fast. Within two generations the flag was heritage — a thing about ancestors and independent-mindedness and the South, flown by people who did not know it had gone up over those capitols in their parents’ lifetime, as a response to the civil rights movement, and who would tell you with complete sincerity that it had nothing to do with that at all.

The granite Ten Commandments monuments the Fraternal Order of Eagles put up in the fifties were gifts. A town could accept one or decline it, and the ones still standing are standing because somebody said yes.

Nobody is being asked now. In 2024 Louisiana became the first state to require the Ten Commandments in every public school classroom. Texas followed, then Arkansas, then Alabama. The Texas statute is specific: a poster of at least sixteen by twenty inches, in every classroom, in lettering legible from anywhere in the room.

Which raises a question the Eagles never had to answer. Which ten?

There is no settled list. The text in Exodus is continuous prose with no numbers in it, and the traditions divide it differently to arrive at ten. Jewish counting opens with I am the Lord your God, which the Christian traditions treat as a preface rather than a commandment. Catholic and Lutheran counting merges the prohibition on other gods with the prohibition on graven images, and then splits coveting in two at the end to make up the number. Protestant counting does the reverse — splits the first pair, merges the coveting.

I learned this firsthand as a child. My father’s family was Catholic, my mother’s Protestant, and however I recited the Ten Commandments, someone told me I had it wrong.

There is also the awkward fact that the only place the Hebrew Bible attaches the phrase the ten words directly to a list is Exodus 34, and the list there is a different one. It includes the three annual festivals and an instruction not to boil a young goat in its mother’s milk.

So a state that requires the Ten Commandments on a wall has to choose a version, and choosing a version means choosing a denomination. Texas fixes the exact wording in the statute, in King James English, and permits nothing else on the poster. An amendment that would have required the Catholic and Jewish versions displayed alongside it, each labeled, did not survive into the enrolled bill. Steven Lubet, a law professor at Northwestern, counted what the legislature actually produced: eleven imperatives and an ambiguous additional line, an attempt to satisfy everyone that came out essentially Protestant anyway.

This didn’t go unchallenged. The Texas case is called Rabbi Nathan v. Alamo Heights Independent School District. Mara Nathan is a rabbi in San Antonio. She has said the Ten Commandments are sacred to her, and that they do not belong on the walls of a public school. The version Texas selected is not the one she reads.

And in 1980 the Supreme Court struck down a Kentucky law requiring the Ten Commandments in every classroom, holding it had no secular purpose. That decision has never been overruled. In April of 2026 the Fifth Circuit upheld the Texas law anyway, reasoning that a 2022 case (Kennedy v. Bremerton School District) had made the older one obsolete. Six judges dissented, on the ground that only the Supreme Court may overrule the Supreme Court.

So: Cecil B. DeMille promotes a film in 1956. The Eagles put granite on the Texas Capitol lawn in 1961. In 2005 the Court lets it stay because nobody objected for forty years. In 2025 a state senator cites it as proof of a tradition, and posters go up in every classroom in Texas.

Those objects had history — or at least the appearance of history — behind them. But some symbols are created out of nothing.

In 2012, a few weeks after Mitt Romney lost, Donald Trump filed to trademark the phrase Make America Great Again. It was not his. Ronald Reagan had run on it in 1980, and Bill Clinton used a version of it in 1992. The registration went through in 2015.

The hat came with the campaign launch, and it was red. Red has meant Republican for a shorter time than most people assume. Before 2000 the television networks assigned the colors however they liked, and often gave red to the Democrats. The convention we have now was fixed by the 2000 Florida recount, the five-week legal fight over who had won the Bush-Gore presidential election, when continuous coverage meant every map on every channel had to stay the same for a month.

So the hat arrived already carrying things. The color, from a graphics decision made by television producers in 2000. The phrase, from Reagan. And around it at the rallies, the flag, the eagle, the anthem.

Once the click is established, a trainer can charge a second signal with it — a word, a whistle — by pairing the new thing with the click alone. No food is required, and none is produced. The whistle takes its value from the click, which took its value from something that is no longer in the room. It works, and it works quickly.

The hat did not have to be paired with anything except other symbols. By the second half of 2016 it was legible across a parking lot, at a distance where nobody could read the words on it.

And nobody can say what it really means.

Ask when America was great and the answers do not converge — the fifties, the eighties, before the war, some point always just behind whoever is answering. Ask what made it great and they converge even less. Reagan, using the words in 1980, meant something specific: the years before Vietnam and the oil shocks and the hostages. The phrase outlived that referent without difficulty.

That is what again is doing. It points backward and declines to say where, and every person reading it supplies his own destination. A date could be checked. A date would also exclude everyone whose golden age fell somewhere else.

Which makes the emptiness the feature rather than a defect — the object works because it means nothing in particular.

Thirty-eight years after the fabric ran out, somebody went back and took another look at the rainbow flag.

In 2017 Philadelphia added two stripes to the flag it flew for Pride, black and brown, for Black and Latino people in a movement that had not always made room for them. The following year a designer named Daniel Quasar put those two colors, plus the pink, blue and white of the transgender flag, into a chevron along the hoist and called it the Progress Pride flag. In 2021 another designer added a yellow field and a purple circle for intersex people.

Each addition was argued for on the merits, and the arguments were serious ones about who had been left out. And each produced a new question, which is which flag you fly. An organization that puts up the six-stripe version now has to be prepared to explain why. A city that switches has to explain that. There are people who read the plain rainbow as a quiet refusal and people who read the chevron as clutter, and both readings are available to anyone standing on a sidewalk looking up at a pole.

The flag they are arguing about lost two of its original meanings to a dye that could not be sourced and the width of a lamp post, and nobody noticed for forty years.

But this one’s different. Somebody looked. Not just at the flag — at what the flag was supposed to represent, and at who it had left out.

Loving the cloth, or the granite, or the paper while forgetting what it stands for — that isn’t new.

The best account of it was written by the man who delivered the decals.

John Prine carried mail in Maywood, then Broadview, then Westchester — western suburbs of Chicago, the routes nobody wanted. He had been drafted, and spent his war in West Germany, by his own account drinking beer and pretending to fix trucks. He wrote songs in his head while he walked. He did not think much of Reader’s Digest, and read only the “Humor in Uniform” page.

By 1969, Reader’s Digest was in something like one of every four American households, and reached more people than any publication in the country except the Bible.

Then an issue came with a plastic flag decal inside. No explanation, no editorial about it — just a decal, in every one of 10–15 million copies, and Prine carried them to every door on his route.

The next day they were on the cars and the front doors. All of them, at once.

So he wrote a song about a man who finds one, sticks it on his windshield, likes it, and keeps adding more until he cannot see out and drives off the road. At the gates he presents his decals, and Saint Peter tells him they are not worth anything here — heaven is full up on account of the war.

Your flag decal won’t get you into heaven anymore.

Which is the argument, in a title, in 1969. The man had the symbol. He had more of it than anybody, had so much of it that it killed him. And at the only accounting that ever mattered it bought him nothing, because he had never once looked past it to the thing it was supposed to stand for.

Prine thought the song would be obsolete inside six months. He believed he was writing about one decal in one war, and that when the war ended it would stop making sense. He spent the next fifty years watching it stay true, and said so — people are always out there waving the flag for the wrong reasons, and the damn song won’t die.

The song is “Your Flag Decal Won’t Get You Into Heaven Anymore,” from Prine’s first record in 1971. Clayton Ambrose has written about it at length on Medium, with the lyrics and the story behind them.

A dog that has been conditioned properly will work for the click and not for the food. It is not being stupid. That is the training succeeding.

You have stood somewhere and heard the national anthem and felt something lift. So have I. That feeling got there by arriving over and over alongside things that mattered.

But the symbols we revere are only the clicks — they’re not the food.

On its own, the American flag is dyed cloth. Smokey is a drawing of a bear in a hat. The monument on the Texas Capitol lawn is a slab of Wisconsin granite. The rainbow flag is six stripes that used to be eight. Not one of these things means anything by itself. They’re simply a pointer to something more important.

American flag shorts are not patriotism. Neither is the flag pin, or the yellow ribbon, or the decal on the windshield, or the red hat. Ask what is behind the hat and you get a date nobody can name. They are empty symbols — placeholders for the thing that really matters.

Love that thing instead.

Love the freedom that is offered to all people, and not only to the ones you would have chosen. Love the liberty to live as you please, which is worth nothing unless it also belongs to people whose choices you find contemptible. Love your enemies, the way the man on the cross said to, and not just the people already on your side. Love the laws that hold those things up — including the ones protecting the man setting a flag alight in front of you, because he is exercising the very thing you say you love, and the cloth is only a picture of it.

That is the food. It has always cost more than the click, and sometimes it costs most exactly when you least want to pay.

I have a bluebird house in my yard. I built it at a workshop sponsored by the Guadalupe Dunes Center, the same place where DeMille’s abandoned Ten Commandments set is interred. I had fun building it, and setting it on a post in my yard made me hopeful that I would soon have baby bluebirds. But in ten years no bluebird has ever used it.

So it is a symbol. It tells anyone walking past that I am the sort of person who cares about bluebirds, and it costs me nothing to go on telling them. But the hole is the right diameter, the roof sheds water, and the floor drains. If a bluebird does come, there is something there for it.

That is the test. Not whether an object means something — they all mean something to somebody. Whether there is anything behind it that would work if someone turned up needing it.

Now consider how we treat obesity.

‘Everything in moderation’ is probably the single most repeated piece of dietary advice in America. It sounds reasonable. It even sounds wise. But it carries a hidden assumption: that the substances being moderated are neutral, and that quantity is the only variable that matters — but they aren’t, and it’s not.

For fifty years, the primary intervention has been caloric restriction — eat less of everything, including the very substances doing the damage. Most of what fills the modern food environment isn’t food in any meaningful sense: it’s engineered to override the body’s ability to stop eating, designed for maximum palatability rather than nourishment. We manage the intake of these products rather than removing them. We hand people smaller portions of what broke their regulatory system, and when it stays broken, we decide the patient is the problem. Then we shame them to make them try harder.

Here’s another way to read this story. The human body has a regulatory system for food intake that evolved over hundreds of thousands of years. It tells you when you’re hungry. It tells you when you’re full. It adjusts appetite based on what you actually need. And it worked — without nutritionists, without calorie counts, without dietary guidelines or fitness trackers or any of the apparatus we’ve spent the last half century building around eating. We know it worked because there are still populations where obesity is essentially absent — the Yanomami of the Amazon, the Tsimane of Bolivia, traditional Okinawans before their diet was westernized. These aren’t genetic anomalies. They’re people eating real food, moving their bodies, and living in an environment that doesn’t actively fight the regulatory system the body came equipped with. The obesity epidemic, the metabolic syndrome, the type 2 diabetes crisis — these are almost entirely post-World War Two phenomena, accelerating sharply in the 1970s and 1980s. They arrived alongside the experts, not before them.

The body wasn’t broken. It was responding rationally to an environment that had been broken around it. And then the interventions designed to fix the problem made it worse.

This is the story of how that happened.

How We Got Here: The Structural Causes

There’s one structural cause of the obesity crisis that virtually nobody discusses: the microwave oven.

Before microwave ovens became available to ordinary households in the mid-1970s, food took real effort to prepare. Planning, shopping, cooking, cleanup — a meal was an investment of time and labor. That investment created a natural restriction on consumption. You don’t idly eat a third helping when it means another hour in the kitchen. You don’t snack at midnight when a midnight snack means actually cooking something on the stovetop or in the oven. The extra work kept consumption roughly calibrated to what you actually needed.

The closest thing to convenience food before the microwave was the TV dinner — a foil tray of frozen, pre-cooked food that still required forty minutes in a conventional oven. My mother worked nights, and I remember her sometimes putting one in the oven when she got home, too tired to cook but needing something to eat before she went to bed. One night she fell asleep waiting for it. I was four or five years old, and I woke up to a house full of smoke while she frantically opened windows. The TV dinner had been in the oven for hours, and was essentially charcoal by that point.

When the microwave showed up in American kitchens, suddenly food was available in a matter of minutes or even seconds, around the clock, with almost no effort. The food industry quickly took advantage of the opening — not because Americans had suddenly become lazy, but because a genuine demand had opened up. Women were entering the workforce at scale, the domestic labor that had produced home-cooked meals was no longer available in the same way, and convenience food met a real need at exactly the right moment. And then the low-fat dietary guidelines arrived in 1980, reinforcing the high-carb, high-sugar model that processed convenience food was built around. The obesity curve starts rising right at that point.

Convenience food didn’t have to be made the way it was. The food industry had choices about what to put in that space. They could have prioritized nutritional quality. Instead they optimized for what drives profit: long shelf life, low ingredient cost, and maximum palatability — producing the cheapest possible edible product from whatever inputs were available, including waste products from other industries that would otherwise be discarded. Think of it the same way cottonseed hulls end up in animal feed — not because they’re nutritious, but because they can be made edible and there’s profit in it rather than waste. When the food industry was called out on this over the years, the response wasn’t to make the food meaningfully healthier. It was to put health claims on the label. ‘Healthy.’ ‘Natural.’ ‘Low fat.’ Words calibrated to manage perception rather than change the product.

It’s worth being precise about what the food industry actually filled that space with, because calling these products food is part of the problem. They aren’t food in any meaningful sense — they’re engineered food-like substances that happen to be edible. Cheetos are a useful concrete example. They contain essentially no protein, no fiber, and negligible micronutrients. What they do contain is a carefully engineered combination of salt, fat, and a texture specifically designed to dissolve in the mouth before the body’s satiety signals can register — what food scientists call vanishing caloric density. You can eat an enormous quantity without feeling full, because the food disappears before your stomach knows it’s there. The caloric load does not. That is not an accident of manufacturing. It is the point. The distinction matters. When we call these products food, we assign them food’s properties by default: that the body can regulate intake of them, that hunger for them signals something nutritionally meaningful, that eating them is basically what eating has always been. None of that is true. Real food contains nutrients the body recognizes and produces satiety signals the regulatory system can read. Engineered food-like substances are specifically designed to do the opposite — to be more palatable than any natural food, to make stopping difficult, to generate craving without generating nourishment. Calling them food obscures a category difference that’s fundamental to understanding why everything went wrong.

The body’s response to these products — eating past fullness, experiencing cravings disconnected from nutritional need, failing to feel satisfied — was precisely the outcome the food industry engineered for. A product that makes you stop eating is a product you buy less of. The problem was never the body. The problem was what we were putting into it and calling food.

Consider how uniquely damaging American breakfasts are. In most cultures, breakfast consists of the same foods as other meals. In America, we have a special category of foods specifically for breakfast that are nutritionally worse. A traditional Japanese breakfast is rice, miso soup, grilled fish, pickled vegetables, and egg — essentially the same foods eaten at any other meal, organized around protein, fermented foods, and vegetables. A Mediterranean breakfast is bread, olive oil, cheese, olives, tomatoes. These are real foods eaten at breakfast because they are real foods, full stop — there is no special category of ‘breakfast food’ that operates by different rules.

The standard American breakfast is a different category entirely. Cereal and skim milk. Pancakes or waffles drenched in syrup. Pastries, muffins, donuts. Flavored yogurt loaded with sugar. Orange juice, which is essentially soda with better public relations. The entire meal is built around refined carbohydrates and sugar, consumed first thing in the morning on an empty stomach, producing a dramatic insulin spike before the day has even started. If you were deliberately trying to design a meal for maximum metabolic disruption, you would be hard pressed to improve on it.

And the cultural messaging around breakfast is relentless. Breakfast is the most important meal of the day — a claim popularized largely by cereal companies with a financial interest in you eating cereal every morning. Eat breakfast to boost your metabolism. Never skip breakfast. All of it pushing people toward the most metabolically damaging meal category in the American diet, first thing every morning, before they’ve had a chance to notice whether they’re even hungry.

Have you ever wondered where breakfast cereal comes from and why it’s uniquely American? John Harvey Kellogg was a Seventh-day Adventist physician running the Battle Creek Sanitarium in Michigan in the late 1800s, where he was deeply preoccupied with what he called ‘self-abuse.’ He believed bland, flavorless food would reduce sexual urges, and corn flakes were specifically invented as part of that program. His brother Will figured out that adding sugar made them taste good and turned it into a business. C.W. Post, a former sanitarium patient, went on to create Grape-Nuts and Post Toasties on similar principles. The entire American breakfast cereal industry was founded by men who thought they were suppressing masturbation. Today it’s a fifty-foot aisle of sugar in cartoon boxes with ‘heart healthy’ on the label.

At the same time the food environment was being transformed, something else was quietly disappearing: food literacy. Not just cooking skill — though that’s part of it — but the basic understanding of where food comes from and how it works. Cooking skill means knowing what to do with food before anyone has told you what to do with it. It is not about knowing how to follow a recipe. A recipe is a script. Cooking skill is understanding the play well enough to improvise. Food literacy means knowing the properties of foods and how to leverage them — that high dry heat transforms the surface of protein and sugar into something complex and deeply flavored through a process called the Maillard reaction — in a way that steaming or boiling never will, that acid brightens a dish and can substitute for salt, that protein denatures with heat and shouldn’t be overcooked.

The loss of that understanding is nearly total in some quarters. I was in the grocery store recently buying a watermelon, and the checker — a woman probably my own age — asked me, in genuine puzzlement, how they get the seeds out of seedless watermelons. She wasn’t joking. Seedless watermelons are a hybrid, produced through plant breeding, and the checker apparently lacked the understanding that all of the plants in our modern food supply are hybrids developed to capitalize on characteristics like sweetness, seedlessness, productivity, and the ability to survive long-distance shipping. A work colleague of mine used to decline the excess eggs I’d bring in from my own flock. When I finally asked why, she said she didn’t want eggs that came out of a chicken’s butt. The eggs at the store, apparently, had been sufficiently laundered of their agricultural origins by the carton and the brand name to be acceptable. The chicken was the problem. The packaging was the solution. There is a running joke that some Americans believe chocolate milk comes from brown cows. A 2017 survey suggested it wasn’t entirely a joke — seven percent of American adults apparently held that belief. The humor lands because the ignorance seems absurd.

The food industry has spent decades providing exactly this service — disconnecting food from its animal and agricultural origins so thoroughly that people can eat meat without thinking about killing an animal, drink milk without thinking about a pregnant cow, and buy eggs without thinking about a chicken. That disconnection is not accidental. A consumer who doesn’t think about where food comes from is a consumer who can’t evaluate what they’re being sold. They are entirely dependent on the label, the brand, and whatever the packaging says about it.

Cooking shows have never been more popular, and actual cooking has never been rarer. People watch cooking with genuine enthusiasm and then order takeout. The aspirational professional content — elaborate techniques, specialized equipment, restaurant-quality results — is entertaining precisely because it’s beyond what most viewers would attempt. But the viral recipe content on social media is arguably more damaging, because it presents itself as accessible while quietly keeping processed inputs at the center. Pulled pork made with cola. Casseroles built from canned soups. Desserts assembled from packaged components. These are not recipes in any meaningful sense. They are assembly instructions that teach dependency on processed inputs while providing the feeling of having cooked something. Watching someone make a perfect dish is entertainment. It is not instruction, and it does not transfer skill.

Today, a lot of people feel like they’re cooking when they pick up a pre-seasoned, pre-stuffed pork tenderloin at Trader Joe’s and put it in the oven. Every actual cooking decision — what to season with, how to season it, what goes inside, how to prepare it — has been made by someone else. What’s left is the application of heat. It’s not nothing, but it’s a long way from cooking in any sense that builds knowledge, skill, or understanding of ingredients. And it comes at a significant premium over doing it yourself.

I grew up eating meals that were cooked at home, in the sense that someone applied heat to them. But most of those meals were built on significant processed food inputs — gravy made from Campbell’s Cream of Mushroom soup, spaghetti sauce from seasoning packets, pancakes from boxed mixes. None of these things are difficult to make from scratch. But the knowledge of how to make them from scratch wasn’t there. It wasn’t until high school that I encountered food that had truly been cooked from scratch. Fifty years later I still remember how different it tasted — not just better, but like a different category of thing entirely. That experience gave me a reference point, and I made it a mission in my life to learn how to cook. In the 1990s, a close friend I frequently invited over for meals told me, with some amusement and some regret, that I had spoiled takeout for him — it simply didn’t taste good to him anymore. Once you know what food can taste like when it’s made from real ingredients, the processed version stops being satisfying. The reference point, once acquired, is hard to lose. The tragedy is that most people never acquire it.

That dependency comes with a cost almost nobody calculates honestly. Every cooking skill lost gets paid for at every meal, for the rest of a life, as a premium to the food industry. A whole chicken on sale runs five to six pounds and costs somewhere between six and ten dollars. From that one bird you get three to four full servings, plus a carcass that makes several quarts of stock — stock that costs up to ten dollars a quart if you buy the packaged kind labeled as bone broth. Total ingredient cost for the stock: essentially nothing, since it’s made from the carcass you already paid for. Now price out the pre-cut, pre-seasoned, oven-ready equivalent. You’ll pay three to four times as much per pound, get no carcass, no stock, and nothing transferable for next time. The ‘whole foods are more expensive’ argument gets the math backwards. Whole ingredients are cheaper. What’s expensive is not knowing what to do with them — so you pay someone else to do it for you, at every meal, for the rest of your life.

Walk into a major supermarket today and it looks like abundance. Enormous stores, aisle after aisle. But look more carefully. The perimeter of the store — produce, meat, dairy — has been quietly shrinking as the interior processed food aisles have expanded. That physical architecture reflects what people are actually buying. Ingredients that were once simply staples of an ordinary kitchen have been quietly repositioned — if they appear at all — in a specialty aisle, priced as luxury items for a particular kind of aspirational consumer. Meanwhile, I’ve stood in a major chain supermarket looking for a cooking funnel. Not there. Food coloring — gone. Bulgur wheat — essentially cracked wheat, one of humanity’s oldest staple foods and a main ingredient in tabbouleh — not a trace of it, though there’s an entire aisle dedicated to dozens of varieties of breakfast cereal. The store has made a decision about who its customer is. That customer does not own a cooking funnel, has never heard of bulgur, and will spend twenty minutes choosing between varieties of frosted corn puffs.

What We Did About It: The Institutional Failures

When obesity rates began to rise, institutions moved in to address it. Each brought its own framework. Each framework was wrong in ways that compounded the damage. Together they’ve produced one of the more impressive records of institutional failure in modern history.

The low-fat dietary consensus, built on the foundation already described, became official policy through the McGovern Committee in 1977 and the first USDA Dietary Guidelines in 1980. The hypothesis at its center — that dietary fat, particularly saturated fat, caused heart disease — was proposed by physiologist Ancel Keys in the 1950s. How it caught on tells you something important about how dietary science actually works, because it wasn’t purely a scientific process. Heart disease was genuinely rising in postwar America — and Eisenhower’s very public heart attack in 1955 focused national attention on what was already a real and growing problem. Keys was a skilled and forceful advocate who knew how to work institutional channels — he got himself on the cover of Time magazine in 1961, giving the diet-heart hypothesis enormous public visibility — and he actively worked to discredit critics, which helped suppress early dissent. The American Heart Association’s early endorsement gave the hypothesis institutional credibility that proved nearly impossible to dislodge. By the time the methodological criticism landed, the framework had momentum that no amount of contrary evidence could fully reverse.

These guidelines were not presented as negotiated political products shaped by industry lobbying and contested science. They were presented as settled fact. And they were instilled in children through health class before those children had any framework to question them. A fact learned in childhood doesn’t feel like a belief you adopted. It feels like something that is simply true. The belief that saturated fat will kill you and that low-fat products are healthy is still operating in millions of people who absorbed it before they could evaluate it. Half a century of contrary evidence has barely dented it.

I lost my gallbladder in my early 40s — the same age at which most women who got gallstones developed them, because a lifetime of low-fat diets reduced the stimulus for gallbladder emptying, allowing bile to concentrate and crystallize. The medical establishment has attributed this epidemic of gallbladder removal in middle-aged women to aging rather than to the dietary advice producing it at high rates in that population. None of the doctors I dealt with ever mentioned that a low-fat diet might be the cause. They implied it was the result of my poor lifestyle choices, rather than the fault of the dietary advice I had been given. During the surgery, I vomited on the operating table while under general anesthesia — a known risk of aspiration that I had specifically warned the anesthesiologist about in the pre-surgical consult, describing prior reactions to general anesthetic in detail. His response was ‘Don’t worry, honey. We know what we’re doing.’ The poor medical advice cost me an organ and nearly my life.

The blanket advice to ‘eat more fruits and vegetables’ is a good example of the kind of imprecision that passes for guidance. Fruits and vegetables are not nutritionally equivalent and treating them as interchangeable does real harm. Vegetables deliver fiber, polyphenols, and micronutrients without significant sugar load. One cup of broccoli contains about 31 calories and 2 grams of sugar. One cup of pineapple contains 82 calories and 16 grams of sugar — more than twice as many calories and eight times as much sugar. Modern commercial fruit has been selectively bred for sweetness to the point where it delivers a sugar load disproportionate to its nutritional benefits. A person who hears ‘eat more fruits and vegetables’ and reaches for fruit juice and bananas three times a day is following the advice and getting a very different metabolic outcome than someone eating leafy greens and cruciferous vegetables. The slogan collapses a meaningful distinction into something that sounds like guidance but isn’t.

The American Heart Association deserves its own mention, because it has functioned as one of the primary institutional voices amplifying and maintaining the low-fat consensus. The AHA’s Heart-Check certification program — the red and white checkmark you see on packaged foods — has certified products that meet the fat criteria while being metabolically problematic in other ways. The certification costs manufacturers a fee, which creates a financial relationship between the AHA and the food industry it’s supposed to be evaluating independently. The AHA has also been a significant promoter of statin use for cholesterol management — and the cholesterol story is more complicated than the guidelines suggest.

One serious hypothesis among cardiovascular researchers is that LDL cholesterol isn’t the cause of arterial damage but a response to it — the body deploying cholesterol to repair inflamed arterial walls, in the same way you’d find construction materials at a damaged building. If that’s correct, lowering the cholesterol is the equivalent of removing the construction materials and declaring the building fixed. The damage is still there. You’ve taken away one of the repair mechanisms. The oxidized LDL theory adds another layer — it may not be circulating LDL that causes problems, but LDL that has been oxidized by inflammatory processes. Seed oils high in polyunsaturated fats are especially prone to oxidation, which means the margarine and seed oil substitution for saturated fat may have worsened the exact mechanism that drives atherosclerosis, while lowering the number medicine was using as a proxy for it.

The margarine story illustrates how badly these interventions can go. The campaign against saturated fat steered people away from butter and toward margarine. What nobody knew was that margarine was full of trans fats, produced by the hydrogenation process that turns liquid vegetable oil into a solid spread. Trans fats turned out to be considerably more damaging to cardiovascular health than the saturated fat they replaced. But the deeper problem was the omega-6/omega-3 imbalance. Humans evolved eating a ratio of omega-6 to omega-3 fatty acids of roughly 1:1 to 4:1. The modern Western diet, dominated by seed oils — soybean, corn, cottonseed, sunflower — runs somewhere between 15:1 and 20:1 in favor of omega-6. Omega-6 and omega-3 fatty acids compete for the same metabolic pathways and push in opposite directions — omega-6 toward inflammation, omega-3 toward its resolution. A diet heavily skewed toward omega-6 creates a chronic pro-inflammatory state at the cellular level that underlies cardiovascular disease, metabolic syndrome, joint disease, and a growing list of other conditions. Butter has a reasonable omega-6/omega-3 ratio. Margarine, made from partially hydrogenated vegetable oils, is predominantly omega-6. In switching from butter to margarine, people didn’t just get trans fats — they dramatically worsened their inflammatory baseline. Two problems for the price of one intervention. The guidelines took decades to acknowledge either one.

Coffee creamer tells the same story in miniature. Cream — real cream, from a cow — was cast as dangerous because of its saturated fat content. The replacement that filled the market was a product made primarily from corn syrup solids, partially hydrogenated vegetable oils, and a list of stabilizers and artificial flavors long enough to require a chemistry background to parse. The omega-6 load was worse than cream. The sugar content was worse. The processing was orders of magnitude more extensive. But it contained no saturated fat, so it was marketed as the healthier choice. The product that displaced cream was metabolically inferior in nearly every measurable way. The dietary advice that created the market for it was presented as settled science.

Statins, the drugs prescribed to lower LDL cholesterol, have real side effects that are underreported relative to how aggressively they’re prescribed. Muscle pain and weakness affect a meaningful percentage of users. Statins work by inhibiting a metabolic pathway that also produces CoQ10, an enzyme critical to cellular energy production — which may explain some of the fatigue and muscle symptoms users report. The evidence for statins in primary prevention — giving them to people who haven’t yet had a heart attack — is considerably weaker than for secondary prevention in people who already have. The drugs lower the number on the lab report. Whether that number is the right target is a question the guidelines don’t invite.

The American Diabetes Association presents a similar picture. For years the ADA’s dietary guidelines for people with diabetes included substantial carbohydrate allowances — remarkable given that carbohydrate is precisely the macronutrient (protein, fat, and carbohydrate are the three macronutrients) that raises blood sugar. The low-carbohydrate approach that many diabetics find most effective for blood sugar control was treated as dangerous fringe by the ADA for years. Steve Cooksey, a man who had been diagnosed with type 2 diabetes and told he would be on insulin for life, put his diabetes into remission through a low-carb paleo approach and began sharing what he’d learned on a blog called Diabetes Warrior. The North Carolina Board of Dietetics and Nutrition threatened him with legal action for giving dietary advice without a license. He sued on First Amendment grounds — and won. The case is a useful illustration of institutional resistance operating not just as inertia but as active suppression. The ADA has historically received significant funding from pharmaceutical companies that manufacture insulin and diabetes medications — companies whose revenue depends on people continuing to need those medications. Managing diabetes is an enormous and ongoing revenue stream. Putting it into remission through dietary intervention is not.

The DASH diet, frequently recommended by both the AHA and the medical establishment for cardiovascular health, illustrates the modest-modifications problem — the pattern of calibrating dietary advice to what people are willing to do rather than what will actually produce optimal health. It reduces sodium, adds some fruits and vegetables, limits red meat. It doesn’t address refined carbohydrates, seed oils, or processed food in any meaningful way. It’s an improvement over the standard American diet in the same way that smoking fifteen cigarettes a day is an improvement over smoking thirty. Technically true. Not the point.

Then there’s the medical establishment, which contributed authority without expertise. Physicians receive minimal training in nutrition — a small fraction of medical education — yet they’re the primary source of dietary authority for most patients. When a doctor tells you how to eat, you listen. What most patients don’t realize is that their doctor’s nutritional knowledge may be no more sophisticated than the Food Pyramid they learned in health class, and may carry the same industry fingerprints.

There is also an internal contradiction in the medical establishment’s approach that rarely gets called out. Patients are told to make small, sustainable changes over time. Meanwhile their physicians are watching the scale at every appointment, implicitly expecting fast results. Fast results require aggressive restriction. Aggressive restriction produces metabolic damage. The advice to go slowly and the expectation of quick progress cannot both be honored at once — and more often than not, the expectation of quick progress wins.

What this looks like in practice: early in my experience with a physician, he zeroed in on my weight as a problem and handed me a book called No White at Night, written by a cardiologist who openly admitted he didn’t know much about nutrition — whose entire framework was based on his personal experience losing ten pounds that had appeared in middle age. His advice boiled down to eat no potatoes, pasta, rice, bread, or popcorn after 7pm. Putting aside the fact that I wasn’t doing that to begin with, what struck me most wasn’t the book itself. It was what handing it to me implied — that someone who had been navigating a significant weight problem since childhood, who had spent over thirty years trying every approach the medical and diet establishment had to offer, needed the dietary wisdom of a man who had never struggled with his weight in his life. The condescension required to make that gesture without noticing it is its own kind of data point about how the medical establishment sees patients who are obese.

Weight bias is pervasive among physicians. When a patient is obese, that obesity becomes the cause of every physical problem the patient presents with. In the early 1990s, my job required hours of daily typing on an IBM Selectric typewriter — an electric machine, but one requiring significantly more force than a modern keyboard. I was a fast typist, and after extended sessions I began experiencing wrist pain consistent with carpal tunnel syndrome. When I mentioned this to my doctor, she didn’t investigate the occupational connection. She attributed the pain to my obesity — the fat in my wrists, she explained, was squeezing the nerves. When I stopped typing on the Selectric, the pain disappeared completely. The actual cause had been sitting in plain sight the entire time. Weight bias had made it invisible.

The metrics the medical establishment chose made things worse. Scale weight and BMI became the primary measures of metabolic health despite being genuinely poor proxies for what matters. Scale weight is a composite number — it includes fat, muscle, bone, and water. It can move in any direction for any number of reasons, most of which have nothing to do with fat loss or metabolic health. Gaining muscle while losing fat, which is exactly the goal, can show up as weight gain on the scale. I once tracked my weight daily and watched it drop 19 pounds from one day to the next, then climb back 13 pounds the following day — a net loss of 6 pounds across three days, buried in swings that had nothing to do with fat. The more important point is this: a physician interpreting those numbers through the linear weight loss model — 2 pounds a week, calories in/calories out — would have drawn completely wrong clinical conclusions from every single reading. BMI compounds the problem by failing to distinguish between entirely different body compositions. The ‘skinny fat’ phenomenon illustrates this: normal weight, high body fat, low muscle mass, poor metabolic health — and neither the scale nor a BMI calculation would detect any of it.

What actually matters is how the body manages blood sugar. Insulin is the hormone that signals cells to absorb glucose from the bloodstream for energy. When cells stop responding to that signal efficiently — insulin resistance — blood sugar stays elevated, the body produces ever-increasing amounts of insulin to compensate, fat storage increases, energy availability drops, progressive weight gain follows, and the conditions are laid for type 2 diabetes and metabolic syndrome. Insulin sensitivity is a far more meaningful health marker than what the scale says. The tests that measure it exist and are available. They just aren’t routinely ordered. There is no billing code for metabolic restoration — no ICD-10 category for a damaged regulatory system that needs repair. Insulin resistance as a condition to be addressed on its own terms has no standalone billing pathway — it can only be coded as a secondary notation attached to the downstream conditions it produces. The billing system has no framework for treating insulin resistance as the primary problem to solve. That tells you everything about how medicine understands the relationship between cause and effect. What does exist are codes for the downstream conditions themselves: obesity, type 2 diabetes, metabolic syndrome — all treated as endpoints to be managed rather than as symptoms of an underlying problem that could be addressed at the root. You can bill for managing a condition. You can’t bill for restoring the system that produced it. The financial architecture of medicine actively discourages the restorative approach even when individual physicians might be inclined toward it.

My own physician has been noting my prediabetes at every appointment for over ten years. The test he orders every time is fasting glucose — a snapshot that tells you blood sugar at one moment in time, in a clinical environment that itself triggers a stress response that raises blood sugar. I have white coat syndrome. The reading he’s relying on is measuring anxiety as much as metabolism. He has never ordered an insulin resistance test. I had to ask him to order even an A1C test — the standard monitoring tool for the very condition he was tracking. And his prescription for the prediabetes? Lose weight. The tests that would tell us whether the underlying metabolic problem is improving simply aren’t part of the conversation. When he finally noted that I had lost noticeable weight, he asked how I had been doing it. When I told him by repairing my metabolism, he looked at me with genuine confusion. The concept was simply outside his framework.

Metabolic restoration isn’t part of mainstream medical training or standard clinical practice — and there are structural reasons for that beyond simple ignorance. Metabolic restoration is slow, nonlinear, and hard to measure with standard clinical tools. Medicine works on appointment cycles and billable interventions. A process that takes months or years of gradual dietary and lifestyle change doesn’t fit the clinical model well, even when it works. Physicians have also seen enough patients fail to sustain lifestyle changes that recommending them without pharmaceutical backup feels, to many doctors, like setting patients up for another failure. And recommending something outside the standard of care, even when it’s better, exposes a physician to liability if something goes wrong. None of these reasons make the standard approach right. But they explain why it persists.

A handful of physicians have built practices explicitly around metabolic restoration — Jason Fung, a Canadian nephrologist whose work on intermittent fasting and insulin resistance has reached millions through books like The Obesity Code; Peter Attia, whose practice focuses on longevity through metabolic health; Eric Westman, who has spent decades demonstrating that type 2 diabetes can be put into remission through dietary intervention. All of them work largely outside the mainstream clinical framework, and all have faced institutional resistance for doing so. That they exist at all is evidence that the approach is possible.

A growing body of researchers argues that the medical establishment has the causality backwards — that metabolic dysfunction, particularly chronic insulin elevation, drives fat accumulation rather than the reverse, and that the accompanying conditions are a result of the same dysfunction rather than a result of the fat itself. The body doesn’t just accumulate fat in response to metabolic dysfunction; it actively resists losing it, because the same dysfunctional signals that drove the accumulation are still running. If correct, this reframes every intervention that targets fat as the primary problem: you’re treating a symptom while the underlying cause continues unchecked. The medical establishment has largely treated excess body fat as the cause of the insulin resistance, cardiovascular risk, and inflammation that accompany it. But if metabolic dysfunction is the root cause and fat accumulation is a result of it, then removing the fat through restriction, surgery, or drugs while leaving the metabolic dysfunction in place is precisely the error at the center of every failed intervention. Fix the system, and the fat resolves as a downstream consequence.

None of this has meaningfully changed what medicine actually prescribes. The dietary advice that has emerged from medicine over the past fifty years remains calibrated to a population that eats extraordinarily badly — which means the recommendations reflect what’s achievable as an improvement from that catastrophic baseline rather than what’s actually optimal. When doctors recommend eating less red meat and more whole grains, they rarely consider that those same changes reduce intake of B12, omega-3s, zinc, and protein. The advice is calibrated to what the average patient is likely to actually do — modest modifications to a bad diet, not a wholesale rethinking of what and how they eat. Small improvements to a catastrophically bad baseline just move you from catastrophically bad to not quite as bad. It’s still not optimal, or even good.

Now for the diet industry, which arrived to solve the problem the food industry and the medical establishment had helped create — and made things considerably worse in its own distinctive way.

The slogan ‘eat less, move more’ captures the model in four words. Both halves are wrong in the same way — they treat the body as a simple input/output machine and say nothing about the specific substances doing the damage or the specific intervention that actually changes the metabolic picture. Not all movement changes metabolism equally. Walking more doesn’t rebuild a damaged regulatory system. Cardio burns calories in the short term, but studies on exercise-only interventions for weight loss consistently show disappointing results, partly because exercise increases appetite, often more than it burns. The approaches that actually work — specifically, resistance training — build the muscle that raises resting metabolic rate and improves insulin sensitivity. ‘Move more’ steers people toward the treadmill and away from the weight room. It’s the wrong direction dressed up as common sense.

And caloric restriction, the foundation of virtually every commercial diet program, is also built on a faulty premise. The calories in, calories out model borrows from the First Law of Thermodynamics, which holds that energy is conserved in a closed system. The body is not a closed system. It is an open, adaptive system that responds to reduced intake by reducing output. When you cut calories, the body adapts: metabolic rate drops, thyroid output decreases, non-exercise activity thermogenesis — the unconscious movement that accounts for a surprising portion of daily calorie burn, things like fidgeting and postural adjustments — quietly decreases, and hunger hormones increase. The body defends its weight because, to a system that evolved during periods of genuine food scarcity, sustained caloric deficit looks exactly like famine. Calories in, calories out is a tautology dressed as an effective intervention. Of course energy balance determines weight over time — that’s just physics. The question is what determines energy balance, and the answer is a complex adaptive system that actively resists the simple manipulations the diet industry is selling.

The Biggest Loser, which premiered on NBC in 2004, was the first television show to turn obesity into entertainment — and it ran for seventeen seasons. Morbidly obese contestants competed to lose the most weight through extreme caloric restriction and hours of daily supervised exercise, with a cash prize for whoever lost the highest percentage of body weight — and a big helping of humiliation for those who didn’t. A 2016 study published in the journal Obesity followed fourteen contestants six years after filming. Their resting metabolic rates remained suppressed at roughly the same level measured at the end of the competition — about 500 calories per day lower than expected for people their size and age. The body adapted to the extreme restriction and didn’t unadapt when the restriction ended. It is estimated that only one in five people who lose ten percent or more of their body weight successfully maintain that loss for even a year. The diet industry interprets this as proof that people can’t stick to diets. A more honest interpretation is that the diet itself is what’s broken — not the people trying to follow it.

Weight loss and fat loss are not the same thing, a distinction the diet industry never makes clearly. Weight loss through caloric restriction includes muscle loss, water loss, and potentially bone density loss alongside fat loss. A person who loses twenty pounds through severe restriction may have lost as much muscle as fat, ending up lighter but metabolically worse off — lower resting metabolic rate, less functional capacity, a body more prone to future fat gain. The physician who sees the lower number on the scale congratulates the patient without knowing what was actually lost. I know this firsthand. At one point I was so desperate to lose weight that I was eating only six pepperoncini or an apple as my entire daily food intake. My doctor, seeing the weight drop, congratulated me on my healthy weight loss. Meanwhile, my hair was falling out. My body was cannibalizing itself. The scale said success.

The diet industry’s external rules model — caloric restriction, eat every few hours, track every bite, remove all temptation from the house — replaced patients’ internal signals with an external program. But suppressing the output of a damaged system doesn’t repair the system. A person who has been trained to follow the program instead of listening to their body is no better equipped to regulate their eating when the program ends than they were when they started. The internal signals were what needed to be restored, not replaced by external rules. The model is also unfalsifiable by design. When the program works, the program gets the credit. When it fails — which it does, for most people, most of the time — the patient gets the blame. The ideal customer for the diet industry is one who fails, blames themselves, and comes back to buy again.

There’s a contradiction at the heart of diet culture that rarely gets named. Tracking and counting food requires making food the center of your mental life. The same programs that prescribe this also promise to reduce your obsession with food — and tell you that your preoccupation with food is pathological, a symptom of the disorder that needs correcting. These things cannot all be true simultaneously. The apparatus of control is the apparatus of obsession. The disorder being diagnosed is largely the one being created.

Diet ranking organizations publish annual lists of popular approaches. Paleo and ketogenic eating consistently end up near the bottom — not because they’re ineffective or nutritionally inferior, but because they’re judged too difficult for most people to sustain. The difficulty is real, but its source is worth understanding: these approaches require departing significantly from the standard American diet, which is the catastrophically bad baseline most people are starting from. But that distance from the baseline is exactly what’s needed to produce meaningful metabolic improvement. Modest modifications to a terrible diet produce modest results. The approaches that actually work get penalized in the rankings precisely because they work. Think about what this ranking system actually measures. We would never rank universities by how easy they are to graduate from, putting Harvard at the bottom because its standards are too demanding. That’s exactly what these rankings do.

Women got a particularly bad deal from the cultural framework surrounding diet culture. In the same decades that restriction and cardio were being prescribed as the path to health, resistance training was coded as masculine. It was so unusual for a woman to have visible muscle that those who did were invited onto daytime talk shows as curiosities. When Linda Hamilton appeared in Terminator 2 in 1991, visibly strong in a way female bodies simply weren’t supposed to be on screen, the reaction was a mixture of admiration and unease. Women were steered toward aerobics, step classes, the elliptical, because these served the aesthetic goal of smallness. The women’s gyms that proliferated in this era often used resistance machines with insufficient weight to build actual muscle — marketed as resistance training, functioning essentially as cardio, calibrated to the aesthetic goal of smallness rather than metabolic health. Resistance training — the one intervention most likely to improve body composition, raise metabolic rate, and produce lasting change — was culturally off-limits for exactly the people who needed it most. The dietary framework said eat less. The cultural framework said don’t build muscle. Between them, they steered women away from every effective intervention simultaneously.

I was a member of Curves — one of exactly the kind of women-only gyms just described — for years, eating 2–3 meals a day, which was, by the conventional dietary wisdom of the time, too few meals. One of the employees at my gym, an overweight woman in her 60s, was genuinely concerned about me because I didn’t eat frequently enough. She herself ate at least five small meals a day, as the standard advice prescribed. She was also on statins for her cardiovascular risk. She died while I was still a member, I believe from a heart attack. She was a nice person and it was sad to lose her. She had done everything she’d been told. The framework failed her, not the other way around.

The entertainment industry contributed its own damage by turning the struggle into spectacle. Corporate television, whose profit depends on drama and ratings, found a reliable formula in fat-struggle programming. The Biggest Loser set the template — public weigh-ins, minimal clothing, suffering presented as inspiration, scale weight as the only measure of success. The genre that followed, including shows like Heavy and My 600 Lb. Life, operated on the same logic, presenting the most extreme interventions and the most dramatic visual transformations because that’s what the format required. Individual creators on YouTube and other platforms have built a genuinely different kind of content — slow, undramatic body recomposition documented honestly over months, without suffering or spectacle. The difference reflects different incentive structures. Corporate television needs drama because drama generates profit. Individual creators who build audiences by being useful have different motives entirely. But the television format shaped what millions of viewers absorbed as the template for what weight loss was supposed to look like — and that template was wrong in almost every particular.

Some GLP-1 drugs — semaglutide under the brand name Wegovy being the primary example — are now explicitly FDA-approved for weight loss, having originally been developed as diabetes treatments. Others in the class are still widely prescribed off-label for weight loss. Either way, the long-term safety data for weight loss use specifically is thinner than it might appear, because the trials that established safety were designed around diabetes management, not indefinite weight loss use in otherwise healthy people. These drugs produce real results: reduced appetite, improved blood sugar regulation, weight loss. They’re also a temporary solution, not a permanent one. The effects last exactly as long as the prescription continues. Most patients regain weight after stopping. The metabolic effects these drugs produce can be achieved through lifestyle intervention — whole foods, a compressed eating window, resistance training — with one critical difference: the lifestyle effects persist, because the underlying system is being restored rather than temporarily overridden. Muscle loss is already documented as a side effect, leaving patients lighter but metabolically worse off — the same outcome as severe caloric restriction, achieved more efficiently. And the long-term side effects of indefinite use remain genuinely unknown.

Bariatric surgery is the most extreme version of the override logic: permanently modifying a healthy organ because the regulatory system it’s part of has been damaged. The surgery works in the narrow sense that it makes overeating physically impossible or immediately punishing. It doesn’t address the food environment that damaged the system. It doesn’t touch the psychological relationship with food. It modifies the container while leaving the conditions intact. And the stomach, elastic by design, accommodates over time — stretching back toward its original capacity under consistent pressure. The complications are permanent: bowel dysfunction, nutrient absorption problems, lifelong supplementation requirements. The patient trades a permanent cost for a conditional and temporary benefit, then goes home to the same food environment that produced the problem.

Running beneath all of this is a profit motive that rarely gets named plainly. Chronic disease is a revenue stream. Wellness is not. A person with a functional metabolism who eats real food and trains for strength is a terrible customer for the food industry, the diet industry, the medical establishment, and the pharmaceutical industry simultaneously. None of those revenue streams survive if the problem gets solved upstream. The system isn’t failing. For the institutions that built it, it’s working exactly as intended.

And yet, in the middle of all this failure, there were approaches that worked. The ancestral health movement — writers and researchers publishing outside institutional channels in the mid-2000s — offered a framework grounded in how humans actually evolved to eat: whole foods, adequate protein, appropriate fat, movement that builds strength. Mark Sisson, whose Primal Blueprint framework and Mark’s Daily Apple platform reached millions outside institutional channels, helped build the movement into something people could actually follow. Gary Taubes, whose books Good Calories, Bad Calories and Why We Get Fat brought the insulin hypothesis to a general audience, provided much of the intellectual foundation — including the argument, developed further by researchers since, that metabolic dysfunction drives fat accumulation rather than the reverse. Robb Wolf popularized the paleo framework through The Paleo Solution. And Nina Teicholz, in The Big Fat Surprise, documented the institutional corruption behind the low-fat guidelines in forensic detail. The resistance this movement met wasn’t purely scientific. It was economic — it threatened every institutional revenue stream simultaneously, and there’s no ongoing profit in eating real food. It was cultural — the approach requires a wholesale departure from everything people have been taught to think of as normal eating, from the processed food, the refined carbohydrates, the seed oils, the breakfast cereals, the fast food that passes for normal eating in America. And it was institutional — recommending something that far outside the standard of care carries liability. All three layers of resistance got dressed up as scientific skepticism.

Before we move to what actually works for people, it’s worth sitting for a moment with what all of this has cost — not in economic terms, but in human ones. We are talking about people who have spent their entire lives fighting their own bodies. Put on a first diet in childhood. Decades of restriction, failure, shame, and obsession — and the obsession itself pathologized by the same culture that created it. Fat shaming from strangers, from employers, from physicians who should know better — a steady drumbeat of public judgment that drives people away from healthcare, compounds the psychological damage, and operates on the fiction that humiliation is a health intervention. Body dysmorphia so deeply instilled that the mirror stops being a reliable source of information. Eating disorders that exist on a continuum with the behaviors diet culture celebrates as discipline, invisible in fat bodies because the diagnostic framework requires a certain silhouette. Plastic surgery as a last resort when the diet fails again — liposuction, tummy tucks, procedures to remove or rearrange what the diet industry couldn’t eliminate — still not fixing the system, still treating the body as the problem. When physicians focus on obesity as the cause of every medical complaint, actual problems go uninvestigated and real diagnoses get missed. When they signal discomfort or frustration with large bodies, patients stop seeking care entirely. I stopped getting pap smears after a physician made clear during a pelvic exam that she found examining a large body frustrating. The discomfort of one appointment translated into years of foregone preventive care. And through all of it, the mental bandwidth consumed: the calorie counting, the food tracking, the meal planning, the guilt, the weighing, the measuring, the negotiating with yourself over every bite. That is bandwidth that could have gone toward anything else. Years of a finite life, spent fighting a problem that was never the person’s fault to begin with, using tools that were guaranteed not to fix it.

Here is the logical problem at the center of every institutional intervention we have described: you cannot restore a system by permanently overriding it. A smoke alarm going off in a burning building is not the problem. Disconnecting it doesn’t put out the fire. The hunger signals, the dysregulated appetite, the cravings — these are outputs of a damaged system. Every intervention that overrides the output while leaving the system damage intact is guaranteed to produce the same result: temporary suppression, followed by return of the signal, because the underlying cause was never addressed. The body is not a passive machine that simply responds to inputs. It is an active, adaptive system with a goal — survival — and it responds to caloric restriction by defending its weight, because that is what kept humans alive during famines. The interventions are fighting the body’s own intelligence. And the body’s intelligence is older and more sophisticated than any diet program ever devised.

The only intervention that works is the one that addresses the cause.

What the Failure Cost: The Individual Damage

The institutional failures didn’t land in the abstract. They landed in bodies — mostly women’s bodies, since women have borne the majority of the cultural weight of the thinness imperative — and produced specific, concrete damage.

The subtractive logic of diet culture — organized entirely around what to remove, never around what to provide — created a phenomenon that remains poorly recognized: the person who is simultaneously overweight and malnourished. You can carry significant excess body fat while being deficient in protein, B vitamins, magnesium, zinc, and essentially every micronutrient the body needs to function. The calories are there. The nutrition isn’t. The rice cake — a puffed disc of air and rice with essentially no nutritional value — was the apotheosis of diet culture. The best dieter was the one who had successfully eliminated food while still technically eating. When the body is undernourished despite adequate caloric intake, hunger persists — not because the person is eating too little, but because the body is still waiting for nutrients it hasn’t received. This is hunger as nutritional incompleteness, and no amount of caloric restriction addresses it.

The school lunch that removes fat from the milk while keeping the chicken nuggets and the refined carbohydrates illustrates the same logic at institutional scale — a subtractive intervention on a peripheral variable while the actual nutritional problem goes untouched. Removing fat from milk actually makes it nutritionally worse — stripped of the fat-soluble vitamins A, D, E, and K, and the fat that made the lactose metabolically manageable. When advocates push for genuinely better school nutrition, the standard institutional response is that children won’t eat it — ‘kids won’t eat healthy food’ — a circular argument that conveniently closes off any serious reform. It’s also an argument that gets applied most aggressively to the lowest-income children, for whom school lunch may be the only meaningful nutritional intervention available. Palate is formed by exposure. Children eat what they’re given when that’s what there is. The National School Lunch Program serves roughly 30 million children daily. What those children eat matters enormously. Institutional gestures toward health that leave the actual problem untouched are not a neutral choice at that scale.

Protein deserves particular attention here, because it is probably the most consequential and most systematically underconsumed macronutrient in the American diet. Muscle protein synthesis becomes less efficient with age, meaning older adults need more protein per pound of body weight than younger ones — not less, as the persistent ‘high protein is bad for your kidneys’ belief suggests. That belief derives from research on people with pre-existing kidney disease and does not apply to healthy adults. The dietary advice that has spent decades steering people away from red meat and eggs toward grains systematically reduces intake of the most bioavailable protein sources while increasing refined carbohydrate load. The result is a population that is simultaneously overfed and protein-deficient — eating plenty of calories, getting insufficient nutrition, and wondering why they’re still hungry.

‘Fat but fit’ exists and is documented — metabolically healthy obesity, significant excess weight combined with normal blood pressure, normal insulin sensitivity, good cardiovascular fitness, favorable inflammatory markers. Cardiorespiratory fitness is a stronger predictor of all-cause mortality than BMI. An overweight person with high fitness has better health outcomes than a normal-weight person with low fitness. This is inconvenient for the moral architecture of the standard narrative. If a fat person can be healthier than a thin person, weight isn’t the variable that matters for health — which means the entire apparatus of fat shaming and thinness-as-virtue can’t be justified on health grounds. The shaming framework can’t accommodate this finding, so it largely ignores it.

Fat shaming rests on an epistemic error almost nobody points out: seeing a fat person is a snapshot. It tells you nothing about trajectory. A month ago that person might have been twenty pounds heavier. They might be in the middle of the most significant health transformation of their lives, doing everything right, and simply not there yet. The shamer has no way to know. The judgment is rendered on a single data point that carries no information about direction or momentum. The shamer may be shaming someone actively losing weight. And that’s before the bodily autonomy argument, which is simpler: a person’s weight is their own decision. Morbid obesity is hard on the body. So is skydiving, motorcycle riding, and drinking alcohol. We don’t have skydiving shaming trolls because those risks aren’t visible on the outside of the body in a way that violates an aesthetic standard. The selective moral concern for fat people’s health reveals itself, under any scrutiny, as concern for appearance wearing the costume of medicine.

The worthiness framework women have internalized runs deeper than aesthetics. Thinness, in the cultural logic that has dominated the last half century, signals discipline, self-control, and virtue. Fat signals their absence. Once that equation is internalized — and most women in this era absorbed it before they had any critical framework to resist it — it shapes everything: what a woman believes she deserves, how she presents herself, what she thinks she’s entitled to ask for professionally and personally. The insecurity produced by this framework isn’t a side effect of diet culture. It’s the product. The insecurity is the market.

The behaviors that define clinical eating disorders — severe restriction, obsessive food monitoring, elimination of food categories, constant body scrutiny — are indistinguishable from what diet culture prescribes and celebrates as discipline. The diagnostic threshold for anorexia has historically required visible emaciation. The same eating pattern in a fat body — the same severe restriction, the same hair loss, the same nutritional collapse — gets read as finally doing something right. The disorder only becomes visible when the body looks like what we imagine an anorexic body looks like. Fat people engaging in identical behaviors receive congratulation rather than concern.

Diet culture removed the pleasure from food and replaced it with two distinct and equally damaging things: guilt and deprivation anxiety.

Guilt operates through moral categorization. Food gets sorted into allowed and forbidden, every choice a small referendum on character. The forbidden category accumulates a psychological charge that has nothing to do with hunger. That charge is what makes forbidden food feel irresistible — not desire exactly, but the specific electricity of transgression. Remove the moral category and the charge goes with it. People who have genuinely removed moral judgment from food are often surprised to find that the compulsion that seemed overwhelming simply isn’t there anymore. Not because they have more willpower. Because the fuel that powered it is gone.

Deprivation anxiety is different, and for many people the more powerful driver. It’s not guilt over eating something forbidden — it’s fear that the thing won’t be there when you want it. Scarcity, real or perceived, triggers a biological imperative to consume while the resource is available. For anyone with a history of actual food scarcity — a significant portion of the population diet culture targets — this response isn’t irrational. It’s a learned survival mechanism, now misfiring in an environment of abundance. Diet culture’s standard prescription — remove all tempting food from the house — directly triggers this response in the people most likely to have it. The intervention creates the compulsion it claims to prevent. The person who cannot have something becomes preoccupied with it. The person who can have it whenever they want, and knows they can, often finds they mostly don’t want it.

The Path Out: Restoration, Not Override

The argument here is not that you should simply trust your body. Damaged systems give disordered signals. A person deep in metabolic damage, with hunger and satiety signals genuinely dysregulated by years of wrong intervention, cannot just listen to their body and expect accurate information. The heroin addict wants heroin. The compass has to be recalibrated before you can navigate by it.

The argument is that the goal is restoration — not permanent override of a broken system, which is what every institutional intervention has attempted, and not naive signal-following, which ignores the real damage that’s been done. Restoration means repairing the regulatory system until it can be trusted again, and then trusting it. This is what the body is designed to do, given the right conditions. The metabolic effects that GLP-1 drugs produce pharmacologically can be produced by the body itself when the system is being repaired rather than temporarily overridden. The drug is a workaround for a problem the body already knows how to solve.

So what does a restored metabolic system actually look like? It’s worth being specific, because restoration is the destination, and knowing what you’re aiming for matters.

Insulin sensitivity restored means cells respond appropriately to insulin signals. Blood sugar is managed efficiently without chronic elevation. The body can access stored fat for fuel between meals rather than being locked in perpetual storage mode. This is the foundational marker — everything else improves downstream of it.

Leptin sensitivity restored means the satiety signal works again. Leptin is produced by fat cells to tell the brain that energy stores are adequate. In a dysregulated system — leptin resistance — the signal is being sent but not received. Restoration means it gets through. You feel full when you have enough. You stop eating when you’re satisfied, not when the food is gone.

Ghrelin normalized means hunger arrives at predictable intervals, is proportionate to actual nutritional need, and goes away when that need is met. Ghrelin is the primary hunger hormone. In a chronically restricted or dysregulated system, it stays elevated and doesn’t drop appropriately after eating — producing the constant, unrelenting hunger that diet culture attributes to lack of willpower. When ghrelin normalizes, hunger becomes manageable and intermittent rather than constant and overwhelming.

Cortisol patterns normalized means the body is producing cortisol in appropriate rhythms — elevated in the morning to support waking, declining through the day — rather than chronically elevated from sustained restriction, metabolic stress, or both. Chronic cortisol elevation drives fat storage particularly around the abdomen, disrupts sleep, and perpetuates insulin resistance. When it normalizes, sleep improves, inflammation decreases, and the body stops working against itself.

A healthy gut microbiome is both a marker of metabolic recovery and an active contributor to it. A diverse gut microbiome improves nutrient absorption, produces short-chain fatty acids that support insulin sensitivity, helps regulate appetite hormones, and influences systemic inflammation. Ultra-processed food, seed oils, and low-fiber diets damage microbiome diversity. Whole foods, particularly fermented foods and vegetables high in prebiotic fiber, support its restoration.

Restorative sleep is another sign of metabolic improvement — and a necessary condition for resistance training to produce its full results. The majority of Human Growth Hormone secretion happens during deep sleep, particularly in the first few sleep cycles. HGH drives muscle protein synthesis, fat metabolism, and tissue repair. Chronic sleep deprivation suppresses HGH release, which means someone doing resistance training but sleeping poorly is leaving most of their gains from training on the table. When metabolic restoration is progressing, sleep quality typically improves as a downstream benefit — and that improved sleep then accelerates the restoration further.

Body composition moving in the right direction — fat decreasing, muscle holding or increasing — as measured by body fat percentage and circumference measurements, not the scale.

Hunger signals you can actually trust. When you’re hungry, you’re nutritionally hungry — not anxious, not bored, not withdrawing from engineered palatability. When you’re full, you know it and you stop. The compulsive quality is gone. Food has become what it always should have been: a tool that does its job and gets out of the way.

A Roadmap for Individuals

What follows is not a diet. These are changes that work together and reinforce each other — the more of them you can make simultaneously, the better. Some may need to be phased in based on your circumstances. But the goal is the whole picture, not a rigid sequence.

Ditch the scale and make food morally neutral, at the same time. These are the same problem wearing different clothes. Both the scale and diet culture’s food morality are external judgment systems that generate anxiety, distort your relationship with your body, and drown out the internal signals you’re trying to restore. The scale doesn’t measure fat — it measures the combined weight of fat, muscle, bone, and water, a number that can move in any direction for reasons entirely unrelated to metabolic health. Gaining muscle while losing fat, which is exactly what you want, can register as weight gain. I learned this the hard way: the first time I joined a gym, I was about thirty pounds overweight. In the first weeks of resistance training I gained eight pounds on the scale, interpreted it as failure, and quit. That decision, made on the basis of a number that was measuring muscle gain and water retention rather than fat, led eventually to being over a hundred pounds overweight. The scale didn’t just fail to help. It actively steered me toward the worst possible choice.

At the doctor’s office, where the weigh-in is treated as mandatory, it isn’t. ‘I’m declining the weigh-in today’ is a complete sentence. It requires no explanation, no negotiation, no apology. The evaluation of your health belongs to you, not to a number on a machine.

Removing moral categories from food means exactly that — no forbidden foods, no allowed foods, no good or bad. Food is a tool. A tool isn’t good or bad — it’s suited or unsuited to the job at hand. A hammer isn’t morally inferior to a screwdriver; it’s just the wrong tool for a screw. Some foods are better suited to the job of nourishing a functional metabolic system than others. None of them are moral categories and none of them should carry a charge. The guilt that diet culture has attached to eating certain things isn’t a health intervention. It’s a psychological burden that makes the regulatory system’s job harder. Set it down.

Replace the scale with metrics that actually tell you something. Body fat percentage, measured by an InBody analysis, DEXA scan, or tape measure calculation, distinguishes between fat and muscle in a way the scale never can. Watching body fat percentage drop while scale weight stays stable or rises is confirmation that body recomposition is happening — exactly what you want to see. Circumference measurements — waist, hips, thighs — capture fat loss where it matters most metabolically. Visceral fat, the dangerous kind that accumulates around the organs, shows up in a shrinking waist measurement before it shows up anywhere else. Functional metrics tell you what the body can actually do: how much you can lift, how you recover from exertion, your resting heart rate over time, your energy levels and mental clarity. And clothing fit — imprecise but honest. Clothes don’t lie the way scales do.

Stop eating the substances that disorder the system. This is the heroin logic applied directly. The goal isn’t to eat less of the engineered food-like substances — it’s to stop eating them. Not perfect elimination forever, but understanding the category for what it is — products engineered to make stopping difficult, containing little or nothing the body actually needs — and removing them from the center of your diet. Ultra-processed products with ingredient lists that read like chemistry experiments, products designed to generate craving without generating nourishment: these are the substances that damaged the regulatory system. Reducing them isn’t a diet. It’s removing the thing that’s causing the problem.

Eat food that actually satisfies the body’s nutritional requirements. Whole, nutrient-dense food — meat, fish, eggs, vegetables, fruit, legumes. And keep in mind that vegetables and fruit are not the same thing nutritionally, despite being grouped together in virtually every piece of dietary advice ever issued. Vegetables are generally more nutrient-dense, higher in fiber, and lower in sugar. Fruit, particularly modern commercially grown fruit selectively bred for sweetness, delivers a sugar load that vegetables don’t. Eat whole fruit rather than drinking fruit juice — juice strips out the fiber that slows sugar absorption, leaving concentrated fructose without the nutritional context that makes whole fruit manageable. A glass of orange juice delivers the sugar of four to five oranges without the fiber of even one. Both fruits and vegetables have a place in a whole-foods diet, but they’re not interchangeable. Lean heavily on vegetables. Be more selective with fruit.

On grains: the standard recommendation to make whole grains a dietary staple deserves scrutiny. Whole is better than refined in one narrow sense — more fiber, more micronutrients. But from an ancestral health standpoint, grains as a dietary staple are themselves worth questioning, whole or not. Lectins, phytic acid, and their effect on blood sugar are concerns regardless of processing. And glyphosate contamination, which concentrates in the outer bran layer, makes the whole grain recommendation more complicated than it appears. If you tolerate grains and choose to include them, whole is better than refined. But the more fundamental question — whether grains should be at the center of the diet at all — is one the guidelines have never seriously engaged with.

On the carbohydrate math: the official USDA guidelines recommend that 45–65% of daily calories come from carbohydrates. But consider how that works in practice for someone doing resistance training. A 150-pound woman targeting 0.85 grams of protein per pound of body weight needs 128 grams of protein — 512 calories. At 1,600 calories per day — a typical target for a modestly active woman maintaining a caloric deficit — with 50% carbohydrate, that leaves only 288 calories for fat, or 18% of total intake — well below what most researchers consider adequate for hormonal function. And 200 grams of carbohydrate is enough to keep insulin chronically elevated throughout the day. An ancestral health framework suggests protein at 25–35% of calories, fat at 40–60%, and carbohydrates at 10–30% skewed toward vegetables, some fruit, and tubers rather than grains. These aren’t just different numbers. They reflect entirely different models of what the body runs best on.

On protein: the general standard for active adults is roughly 0.7 to 1.0 grams per pound of body weight. For vegetarians who include eggs and dairy, hitting adequate protein targets is achievable with careful planning. For strict vegans doing serious resistance training while maintaining a calorie deficit, it is genuinely difficult without using processed protein supplements.

Cooking your own food from whole ingredients is the most direct path to nutritional completeness — and it’s almost always cheaper than the alternative, despite what the ‘healthy eating is expensive’ narrative claims.

Skip meals when you’re not hungry. Never eat to keep yourself from being hungry later. This runs counter to decades of advice about eating frequently to ‘keep your metabolism going’ — advice that, as discussed, keeps insulin chronically elevated and prevents the body from accessing stored fat between meals. Genuine hunger is information. Eating in the absence of it is noise. As the system restores, hunger signals become more reliable and intermittent. Trust them.

Expect discomfort during the transition and recognize it for what it is. Moving from a diet centered on ultra-processed products to one based on whole foods often feels genuinely terrible for a week or two — fatigue, brain fog, irritability, intense cravings. Diet culture interprets this as the body saying stop. It’s actually the body withdrawing from substances that reorganized its reward and energy systems in much the same way addictive drugs do. The discomfort is the recalibration, not evidence that the intervention is wrong. Most people stop here because they’ve been told to listen to their body, the body is saying stop very loudly, and nobody has given them the framework to understand that this particular signal is withdrawal, not warning. Getting through it is the cost of reaching the other side, where the signals start being trustworthy again.

Adopt intermittent fasting, at least initially. This is a recognized and well-researched approach with multiple viable schedules — the most common being 16:8, eating within an 8-hour window and fasting for 16. But 5:2, alternate day fasting, and other variations work on the same principle: giving the insulin system genuine rest between eating periods.

Frequent eating keeps insulin chronically elevated, which perpetuates the insulin resistance at the root of the metabolic problem. Giving the system genuine rest between meals is one of the most direct ways to begin restoring insulin sensitivity. As that sensitivity improves, hunger signals become more reliable — and the eating pattern that emerges naturally tends to compress further on its own. The specific schedule matters less than the consistency. The best one is the one that fits your life well enough to sustain.

Add resistance training. Muscle is metabolically active tissue — it raises the rate at which the body burns fuel at rest, improves insulin sensitivity, supports bone density, and produces body composition changes that cardio alone cannot achieve. For larger bodies or those with physical limitations, machine-based resistance training is often the most accessible starting point — it allows precise load control and external support without requiring your bodyweight to work against you. Start wherever you are. The adaptation begins immediately and the metabolic benefits accumulate in ways the scale won’t capture but your body absolutely will.

Prioritize sleep. This is not a peripheral recommendation. The majority of Human Growth Hormone secretion happens during deep sleep, in the first few sleep cycles of the night. HGH drives muscle protein synthesis, fat metabolism, and tissue repair — which means the resistance training you did during the day gets converted into actual muscle during the night. Chronic sleep deprivation suppresses HGH release and keeps cortisol elevated, which is catabolic — it actively breaks down the muscle tissue you’re working to build.

Someone doing everything else right but sleeping poorly is leaving most of their gains from training unrealized. As metabolic restoration progresses, sleep quality typically improves on its own. But if it isn’t improving, it’s worth addressing directly — sleep hygiene, light exposure, meal timing relative to bedtime — because without restorative sleep, the rest of the program is working at a significant disadvantage.

A Roadmap for Society

Individual restoration is possible, but it’s harder inside a broken environment. The structural causes of the obesity crisis require structural responses. Individual change, while necessary and achievable, is swimming against a current that policy and industry have spent decades creating.

Redirect agricultural subsidies. The United States spends roughly twenty billion dollars annually subsidizing commodity crops — primarily corn, wheat, and soybeans — which makes the ultra-processed food built from them artificially cheap, and whole food relatively expensive by comparison. The economics of the food environment aren’t natural. They’re policy choices. Ending subsidies for commodity crops that primarily feed the processed food industry and redirecting support toward regenerative farming — which produces nutritionally dense food, restores soil health, and builds a food system that doesn’t depend on industrial inputs — would change the foundational economics of what people can afford to eat. The agricultural lobby is powerful. But the current structure is a policy choice, and policy choices can be changed.

Return cooking education to schools. Embodied knowledge — how to break down a whole chicken, how to build a meal from ingredients rather than from package instructions, what real food actually looks and smells and tastes like — cannot be transmitted through a pamphlet. It requires practice, instruction, and the freedom to make mistakes. The generation that lost this knowledge passed the loss on. Recovering it requires deliberate transmission, starting young, before dependency on processed food is established. Cooking class isn’t a luxury elective. It’s one of the most practically consequential skills a person can have.

Reform food labeling. The serving size listed on most processed food packaging bears no relationship to the amount any human being actually consumes in one sitting. This isn’t an accident. It’s a mechanism by which the food industry misrepresents the caloric and nutritional content of its products. Mandatory realistic serving sizes, based on documented consumption patterns, would make the actual nutritional picture visible. People can’t make informed decisions with systematically distorted information.

Overhaul school lunches substantively. Not subtractive gestures while underlying food quality stays the same. The lowest-income children, for whom school lunch may be the primary nutritional intervention available, deserve better than institutional performances of health that leave the actual problem untouched.

Hold the food industry accountable. The science behind the engineering of products designed to maximize profits and override satiety signals is documented — companies have internal research showing that specific ingredient combinations produce consumption patterns no naturally occurring food produces. This has an uncomfortable precedent: the tobacco industry spent decades manipulating nicotine levels to maximize addictiveness while publicly denying any such thing. Internal documents eventually proved otherwise. The food industry has its own internal research showing that specific combinations of salt, sugar, and fat produce consumption patterns that no naturally occurring food produces — and has marketed the resulting products as food while publicly attributing overconsumption to individual weakness. The parallel is not rhetorical. It is structural. Accountability for that engineering — through regulation, labeling requirements, or liability — would change the incentive structure that currently rewards maximizing consumption over nutritional value.

Stop treating fat shaming as legitimate concern. It operates on incomplete information, violates bodily autonomy, drives people away from healthcare, and has never been shown to produce better health outcomes. A culture that shames people for their weight while simultaneously engineering the food environment that produced it, selling them failed solutions, and profiting from the resulting chronic disease has no standing to moralize. The shame is part of the system. Withdrawing it is a small but meaningful act of structural change.

Insulate dietary guidelines from industry influence. The history of guidelines shaped by the grain lobby, the sugar lobby, and the dairy industry, delivered as settled science, is a history that has cost people their metabolic health and in some cases their lives. Meaningful conflict-of-interest standards for the committees that design these guidelines would not guarantee perfect guidance, but would remove the most direct mechanism by which industry interests have historically overridden nutritional ones.

Conclusion

The institutions described in this essay are not going to reform themselves. The food industry will not voluntarily stop engineering products designed to maximize profits and override satiety. The diet industry will not abandon a business model that depends on failure. The medical establishment will not suddenly develop meaningful nutrition training or abandon the metrics it has built its practice around. The pharmaceutical industry will not walk away from a drug class generating billions in annual revenue. These are not pessimistic conclusions — they are observations about how institutions with entrenched financial interests actually behave. Waiting for systemic change before taking personal action is waiting for something that may not arrive in your lifetime.

What individuals can achieve is a different story.

I can tell you what this looks like in practice, because I’m living it — as a work in progress. I don’t count calories. I don’t measure portions or track what I eat. I don’t stress over how many calories I’m burning. I cook from scratch, mostly. My kitchen includes cookies, ice cream, and other foods that diet culture would call forbidden — and I throw most of them away after six months or so because I simply haven’t wanted them. I eat whole, nutrient-dense food when I’m hungry, and I stop when I’m not — In practice, I’ve found a one meal a day approach suits me best. I do resistance training on a consistent schedule. And I have gone from a size 28 to a size 18 — not by fighting my body, but by giving it the conditions it needed to do what it already knew how to do. I am not done. But I am moving in the right direction, and for the first time in my life, my body is working with me rather than against me.

For many people, the damage done by decades of wrong advice is not permanent — the body retains more capacity for restoration than the institutions that caused the damage would have you believe. Blood sugar normalizes. Insulin sensitivity returns. Fatty liver disease reverses. Cardiovascular markers improve. Chronic inflammation decreases. Body composition changes in ways the scale won’t capture but the body absolutely registers — more energy, better sleep, stronger, leaner, more capable than before. For others, particularly those whose pancreatic function has been significantly compromised, the window has narrowed. This is one more reason not to wait. The right conditions can be approximated even inside the broken food environment we’ve built. It requires deliberate choices, real sacrifice, and the willingness to go against advice that feels authoritative because it comes from institutions that have been wrong for fifty years and show no signs of correcting themselves.

The body was never the problem. Give it the right conditions, and it will demonstrate that.

What Actually Changes With Age

To make this argument honestly, it has to begin with an acknowledgment: aging is real. Some things genuinely change in the human body over time, independent of lifestyle. Processing speed — the raw speed at which the nervous system operates — slows with age and is not fully recoverable. Presbyopia, the age-related loss of the ability to focus on close objects visually, is nearly universal after age 45 and is caused by hardening of the lens. High-frequency hearing loss has a genuine biological component, though noise exposure across a lifetime dramatically accelerates it. Menopause is a biological event with real downstream consequences. Arterial stiffness increases with age due to changes in vessel wall composition. The immune system’s response to novel threats weakens.

These changes are real. They are also merely a small fraction of what gets attributed to aging in the typical American. The rest — the muscle loss, the joint pain, the metabolic deterioration, the cognitive decline, the cardiovascular disease, the frailty, the loss of functional independence — is not primarily biological. It is primarily lifestyle-driven. One you can’t change. The other you can — if you’re willing to.

Before going further, it is worth understanding why the medical system’s picture of aging looks the way it does. Medicine tracks what it can measure easily and act on quickly: cholesterol numbers, blood pressure readings, blood glucose levels. These biomarkers can be obtained in a standard appointment, compared to a reference range, and addressed with a prescription — without requiring the clinician to develop a comprehensive picture of how the patient actually lives, moves, eats, and functions. The holistic view is harder, slower, and less reimbursable. The biomarker is easier. And so the biomarker becomes the goal, and the outcomes it was supposed to predict become secondary. Much of what follows in this essay is a consequence of that substitution.

The Lifestyle Drivers of Decline

Sarcopenia — the loss of muscle mass with age — is the most visible and consequential of the declines most people associate with aging, and it is the most preventable. Sedentary adults lose roughly 3–8% of muscle mass per decade after age 30, accelerating after age 60. Resistance-trained older adults lose significantly less, and some studies show net gains even in their 80s. Competitive bodybuilders in their 80s and 90s — people who have maintained progressive resistance training and adequate protein intake across decades — demonstrate that the capacity for significant muscle mass doesn’t disappear with age. What disappears is the physical work the body was doing, and the adequate protein required to maintain muscle.

The same pattern holds across nearly every system in the body. Aerobic capacity declines roughly 10% per decade after age 30 in sedentary adults but only 5% in consistently trained ones — half the rate. Masters athletes in their 70s routinely have cardiovascular profiles comparable to sedentary 40-year-olds. Blood pressure rises with age in Western populations but shows virtually no age-related increase in the Yanomami people of the Amazon, whose diet and activity patterns differ dramatically from the American baseline. Insulin sensitivity declines with age in sedentary people eating the standard American diet and is well-preserved in active people eating differently. Bone density loss accelerates after menopause but is significantly attenuated by resistance training, which has been shown to maintain and even increase bone density at any age. Balance, flexibility, and the functional capacity to perform basic daily movements — getting up from the floor, climbing stairs, carrying loads — decline almost entirely from disuse rather than from any biological imperative.

The cognitive picture is similar. Processing speed genuinely slows with age, but memory, executive function, and dementia risk are all heavily modifiable by lifestyle. About a third of Alzheimer’s cases are estimated to be attributable to modifiable risk factors — physical inactivity, hypertension, obesity, diabetes, smoking, depression, social isolation, and hearing loss. The same amyloid plaques that characterize Alzheimer’s disease are found in the brains of people who never developed dementia, suggesting that cognitive reserve built through a lifetime of engagement provides significant protection against the clinical expression of the pathology. The brain, like the muscle, responds to use.

Chronic pain — the vague joint aches, the stiffness, the low-level discomfort that becomes a constant companion for so many older Americans — is perhaps the most universally accepted feature of aging and one of the least examined. Weak muscles can’t adequately support and stabilize joints, so the joints bear load they weren’t designed to carry alone, producing chronic irritation and inflammation. Sedentary behavior reduces circulation to connective tissue, which depends on movement for nutrient delivery. The chronic systemic inflammation produced by the standard American diet creates an environment in which joints and soft tissue are persistently irritated. The pain then produces its own feedback loop: it hurts, so you move less; moving less accelerates the muscle loss and reduces circulation; the reduced circulation and muscle loss increase the pain. The pain gets attributed to aging rather than to the sedentary response to pain, and the accommodation deepens. Knee and hip replacements are among the most common surgeries performed in the United States, with the average age of recipients falling as sedentary lifestyles produce joint damage earlier in life. They are the surgical accommodation to a problem that is substantially preventable. Chronic pain in older adults is not a normal feature of aging. It is a normal feature of how most older adults in Western societies live.

The Functional Independence Question

Loss of functional independence — the ability to perform basic activities of daily living without assistance — is the primary driver of assisted living and nursing home placement. And the physical capacity most directly implicated is lower body strength: specifically, the ability to perform sit-to-stand movements independently. Getting on and off the toilet, getting up from a chair, rising from the floor — these require significant quadriceps and glute strength, hip mobility, and balance. They are, in functional terms, variations of a squat.

The sitting-rising test, developed by Brazilian physician Claudio Gil Araújo and published in 2012, is not a standard clinical test but gained significant attention when research showed it predicted all-cause mortality with surprising accuracy. The test asks a simple question: can you sit down cross-legged on the floor and get back up without using your hands, knees, or arms for support? The score correlates remarkably strongly with mortality in middle-aged and older adults — not because the test itself matters, but because it measures the functional lower body strength and mobility that determines independence. That capacity is almost entirely trainable through resistance exercise. Most older adults simply never train it.

What gets normalized is the accommodation rather than the capacity. When an older person can’t get off the floor, the response is to remove floor-level activities from their life rather than address the underlying strength deficit. The accommodation confirms and deepens the limitation. In contrast, cultures where floor sitting is a normal daily activity — for meals, prayer, socializing — maintain the functional capacity to do so into old age because they never stop doing it. The capacity doesn’t disappear with age; it disappears with disuse, and the disuse gets labeled as aging.

The kyphosis — the hunched forward posture — common in older adults is partly bone density loss in the vertebrae, but largely the result of a lifetime of weak posterior chain muscles that can no longer hold the spine upright, combined with chronically shortened chest muscles from decades of sitting with rounded shoulders. It is not primarily an age effect. It is a disuse effect that age makes visible because the accumulated deficit finally exceeds the body’s compensatory capacity.

Diet, Inflammation, and a Broken Baseline

Much of what gets attributed to aging in Western populations is substantially driven by the chronic low-grade inflammation produced by the standard American diet. Ultra-processed foods are pro-inflammatory through multiple mechanisms: refined seed oils high in omega-6 fatty acids shift the ratio toward inflammation; refined carbohydrates spike insulin and promote visceral fat accumulation; emulsifiers and additives disrupt the gut microbiome; and the absence of fiber, polyphenols, and micronutrients removes the dietary elements that actively suppress inflammation. The result is a state of chronic systemic inflammation that underlies joint pain, cardiovascular disease, cognitive decline, immune dysfunction, and cancer risk simultaneously — and that gets attributed to aging because it is nearly universal in the population being studied.

Sugar deserves specific attention — not simply because of its caloric content, but because of what it does metabolically. Fructose is metabolized almost exclusively in the liver, and chronic high intake is a significant driver of non-alcoholic fatty liver disease regardless of whether it puts a person in a caloric surplus. Glucose spikes insulin, and chronic insulin elevation drives visceral fat accumulation, insulin resistance, and inflammation independent of total calorie intake. A calorie of sugar is not equivalent to a calorie of protein or fat — the metabolic consequences are different, and more damaging.

The definition of healthy eating has been calibrated to a population that eats extraordinarily badly, so the recommendations reflect what’s achievable as an improvement from that baseline rather than what’s actually optimal. When doctors recommend eating less red meat and more whole grains to improve cholesterol numbers, they rarely consider that those same changes reduce intake of B12, omega-3s, zinc, and protein. The advice is also calibrated to what the average patient is likely to actually do — modest modifications to a bad diet, not a wholesale rethinking of what and how they eat. The result is guidance designed for the median American’s willingness to change, which is not the same thing as guidance designed to produce optimal health.

Nutritional advice invariably groups fruits and vegetables together as though they are equivalent — “eat more fruits and vegetables” is one of the most repeated dietary recommendations in medicine. They are not equivalent. Vegetables deliver fiber, polyphenols, and micronutrients without significant sugar load. Modern commercial fruit has been selectively bred for sweetness to the point where it delivers a sugar load disproportionate to its nutritional benefits. The grouping obscures a meaningful distinction.

Glyphosate — the herbicide used as a pre-harvest desiccant on grain crops — has been found in the majority of commercial bread brands tested by the Florida Department of Health in 2026, with whole grain varieties showing the highest residue levels. The bran layer that makes whole grain nutritionally superior also concentrates pesticide residues. But bread is only the most visible example. Glyphosate contamination is in the grain itself, meaning any product made from conventionally grown grain — pasta, cereal, crackers, baked goods, anything marketed as made with “healthy whole grains” — carries the same concern. Whether the levels found pose a direct health risk is contested; what is not contested is that the regulatory limits were set based on single-exposure toxicology rather than on lifetime cumulative exposure from daily consumption. The organic varieties tested were not exempt from contamination, suggesting that glyphosate has become pervasive enough in the agricultural environment that deliberate avoidance through organic purchasing may be insufficient.

Protein is the most consequential macronutrient for the aging body and the most systematically underconsumed. Muscle protein synthesis becomes less efficient with age, meaning older adults need more protein per pound of body weight than younger adults — not less, as the persistent “high protein is bad” belief suggests. That belief derives from research on people with pre-existing kidney disease and does not apply to healthy adults. The dietary advice that steers people away from red meat and eggs toward grains and plant-based alternatives systematically reduces intake of the most bioavailable protein sources while increasing refined carbohydrate load. For people trying to build muscle while maintaining a caloric deficit — the goal of body recomposition — animal protein sources are not optional. The math of getting adequate protein from plant sources while keeping calories in deficit is nearly impossible in practice.

GLP-1 receptor agonists like semaglutide have become widely used for weight loss, producing real metabolic improvement while the drug is taken. But they don’t address the dietary and lifestyle patterns that created the dysfunction, so when people stop taking them, the weight returns and the metabolic markers follow. A diet of whole foods, high in protein and fat, low in refined carbohydrates, eaten in a compressed time window, produces many of the same effects through dietary means — without the dependency or the significant lean mass loss that GLP-1s produce. The drug is doing the work that lifestyle should be doing.

Caloric restriction approaches to weight loss — the dominant model in both popular diet culture and the medical system — address the symptom the medical system has decided to treat without addressing nutritional adequacy. Hunger is not simply a response to caloric deficit; it is a response to nutritional incompleteness. A body that has consumed adequate calories for its energy needs while missing protein, essential fats, or key micronutrients will continue signaling hunger. Long-term caloric restriction also dysregulates the hormones most directly responsible for satiety and hunger, producing a physiological drive to overeat that persists even after reaching a target weight. The person who has lost weight through restriction and still can’t stop eating when full is not failing through lack of willpower; they are experiencing the predictable hormonal consequence of how the weight was lost.

Nutritional Deficiencies Masquerading as Aging

Several nutritional deficiencies are nearly universal in older adults eating the standard American diet, produce symptoms that are routinely attributed to aging, and are straightforward to address once identified. The irony is that the dietary advice most commonly given by doctors systematically reduces intake of the foods highest in the nutrients most commonly deficient.

B12 deficiency develops slowly — the body stores several years’ worth in the liver — which means it can accumulate undetected for a long time before symptoms appear. Those symptoms include fatigue, cognitive fog, memory problems, depression, balance problems, and peripheral neuropathy. Stomach acid production declines naturally with age, impairing the ability to cleave B12 from the food proteins it’s bound to — meaning even people eating adequate amounts of B12-rich food may absorb less of it as they get older. Peripheral neuropathy in particular is worth flagging: it is a known complication of diabetes, but it is also a symptom of B12 deficiency, and metformin — the most commonly prescribed drug for type 2 diabetes — depletes B12. A diabetic patient on metformin who develops peripheral neuropathy may be experiencing a drug-induced nutritional deficiency rather than a diabetic complication. The distinction matters because one has a straightforward fix — before the neuropathy becomes permanent. It is rarely investigated because the more familiar explanation is already available. The standard serum B12 test is inadequate for detecting functional deficiency — ask for methylmalonic acid, a compound that accumulates when B12 is functionally insufficient, which provides a more accurate picture of B12 status than the serum test alone.

Vitamin D deficiency is probably the most widespread. Modern indoor lifestyles, sunscreen use, and limited dietary sources mean most people are not getting adequate vitamin D. Deficiency affects immune function, muscle strength, bone density, mood, and cardiovascular health. The standard sufficiency threshold of 20 ng/mL is contested — many researchers argue optimal function requires levels closer to 40–60 ng/mL, meaning people who test sufficient by current standards may still be functionally suboptimal.

Magnesium deficiency is nearly universal in people eating processed food and is actively depleted by stress, alcohol, diuretics, and acid-suppressing medications — both prescription and over-the-counter. It produces muscle cramps, poor sleep, anxiety, and elevated blood pressure. Serum magnesium is a poor indicator of total body magnesium since most is intracellular, so levels can look normal while cellular stores are depleted. Omega-3 deficiency drives chronic inflammation when the diet is high in seed oils and low in fatty fish. Zinc deficiency — worsened by high grain consumption, since grains contain phytates that bind zinc and block absorption — impairs immune function, wound healing, taste and smell, and cognitive function. Choline, found primarily in eggs and liver — both foods specifically discouraged by mainstream dietary advice — is essential for liver function, brain health, and cell membrane integrity, and deficiency is associated with non-alcoholic fatty liver disease and cognitive decline.

Weight vs. Fitness: A Conflation With Consequences

Body weight and physical fitness are not the same variable, and the medical establishment’s treatment of weight as the primary health metric produces predictable errors. The research on what is sometimes called “fat but fit” is consistent: cardiorespiratory fitness is a stronger predictor of mortality than BMI. An overweight person with high fitness has better outcomes than a normal-weight person with low fitness. The “skinny fat” phenomenon — normal weight but high body fat percentage and low muscle mass — illustrates the problem clearly: a person who looks unremarkable on a scale may have poor metabolic health, since high body fat, particularly visceral fat, drives insulin resistance and systemic inflammation, while low muscle mass reduces resting metabolic rate and functional capacity. The scale captures none of this.

Weight loss and fat loss are not the same thing, and most physicians don’t discriminate between the two. Weight loss through caloric restriction includes muscle loss, water loss, and potentially bone density loss alongside fat loss — the scale goes down, but what has been lost matters enormously. A person who loses 20 pounds through severe caloric restriction may have lost as much muscle as fat, ending up lighter but metabolically worse off: lower resting metabolic rate, less functional capacity, and a body that is more prone to future fat gain. The physician who sees the lower number on the scale congratulates the patient without knowing what was actually lost.

The conflation has practical consequences beyond the clinical. The exercises most commonly recommended for overweight people — walking, bodyweight exercises — are precisely the ones that are hardest and most uncomfortable for larger bodies, not in the productive way that leads to adaptation but because of the mechanical disadvantage of carrying extra weight through those specific movement patterns. Machine-based resistance training, where the load is external and adjustable, the body is often supported, and the movement is controlled, is both more accessible and more effective for larger bodies, older adults, and people with physical limitations. The person who finds walking painful, push-ups impossible, or conventional exercises mechanically awkward due to body size or disability is not failing at fitness. They have been pointed toward the wrong tools.

There is also a protective effect of moderate overweight in older women that the clinical focus on weight loss consistently ignores. For women over age 65–70, being moderately overweight is associated with lower all-cause mortality than being normal weight or underweight — likely due to metabolic reserve during illness, greater bone density, and the fact that unintentional weight loss in older adults is frequently a marker of underlying disease. The clinical guidance doesn’t reflect this because the framework was built on data from younger populations and because the cultural bias toward thinness as health is so deeply embedded that contradictory evidence struggles to penetrate practice.

The Medical Establishment’s Role

Medicine excels at acute care, surgery, and infectious disease. It is considerably less reliable as a guide to nutrition, body composition, and the optimization of physical function across a lifetime. The dietary advice that has come from medicine over the past fifty years has coincided with an obesity epidemic, a type 2 diabetes epidemic, and a metabolic health crisis of historic proportions. The system optimizes for biomarkers — LDL cholesterol, blood pressure, blood glucose — rather than for the outcomes those biomarkers are supposed to predict, and when the biomarker and the outcome diverge, the biomarker wins in clinical practice.

The pharmaceutical burden in older adults is a specific and underexamined problem. The average American over age 65 takes multiple prescription medications, each evaluated for safety and efficacy largely in isolation, in clinical trials that typically excluded older adults, people with multiple conditions, and people taking other medications. The interactions between drugs, and the cumulative side effect burden of multiple drugs taken together, are poorly understood and poorly tracked. Many common medications produce side effects in older adults that look exactly like aging symptoms: statins cause muscle weakness and pain; blood pressure medications cause fatigue and cognitive fog; acid reflux medications impair absorption of B12, magnesium, and calcium; antihistamines accumulate anticholinergic burden that is increasingly recognized as a dementia risk factor. Each symptom that emerges gets evaluated as a new problem rather than as a drug effect, and frequently gets treated with another drug.

Over-the-counter medications add to this burden invisibly — they are not tracked, not coordinated with prescribing physicians, and carry the implicit message that they are safe because they don’t require a prescription. NSAIDs taken regularly for the chronic joint pain that is itself largely a lifestyle artifact damage the gut lining, increase cardiovascular risk, and raise the risk of gastrointestinal bleeding. Antacids and proton pump inhibitors — both widely available without a prescription — suppress stomach acid, which is required to absorb B12 from food. People who take them regularly for acid reflux (itself often caused by poor diet) may be quietly depleting the B12 their aging stomach was already struggling to absorb. Antihistamines taken regularly for sleep accumulate anticholinergic burden. The liver bears further load from regular OTC analgesic use and alcohol — a combination that is common and rarely examined as a whole.

Medicine also tends to steer older adults toward cardio rather than resistance training — and it is resistance training that has the broader and more consequential impact on aging outcomes. Cardio improves the cardiovascular biomarkers medicine tracks; resistance training builds the functional capacity and body composition that actually determines quality of life and independence. The advice follows the easy metric rather than the outcome. Similarly, the weight bias in medicine produces harmful guidance for overweight patients: the clinical system consistently pushes them toward caloric restriction that reduces muscle mass, avoidance of physical activity until weight is lost, and dietary changes that reduce intake of the most nutritionally dense foods. Following that guidance conscientiously for decades and failing is not a personal failure. It is the predictable result of advice that was wrong for the goal.

If you want to know how to lose fat and build muscle, the honest answer is that you should not ask a doctor. Medicine’s expertise is in disease management, not in optimizing body composition or physical performance. The knowledge base for body recomposition lives almost entirely outside medicine, in the strength and conditioning world and in the bodybuilding community, which has been running human experiments on body composition for decades with far more practical sophistication than clinical research has produced. The expert on building functional physical capacity in an aging body may be a 70-year-old competitive powerlifter. Recognizing that requires evaluating sources by their results rather than their credentials.

The Cultural Forces That Maintain the Decline

The expectation of decline doesn’t just normalize what happens — it helps cause it. Research by Becca Levy at Yale shows that people who internalize negative stereotypes about aging earlier in life show measurably worse physical and cognitive outcomes decades later, independent of their baseline health. The expectation shapes behavior, and the behavior produces the outcome the expectation predicted. When a family member worries that you’ll have trouble with the stairs, when a doctor tells you to take it easy, when the culture signals that vigor in older age is surprising rather than normal — these are pressures toward accommodation that feel like care but function as acceleration of decline. The stairs, incidentally, are not the problem. For someone with adequate lower body strength and balance, they are a daily resistance training stimulus. The concern that prompts someone to avoid them removes exactly the activity that would keep them able to climb them safely.

The normalization of poor nutrition in older adults is one of the most invisible aspects of this pattern. We know that aging animals need high-quality, nutrient-dense food — the pet food industry has built an entire category of premium senior formulations around this fact. Yet when it comes to aging humans, the expectation quietly reverses: smaller portions, blander food, less effort, less protein. My grandmother never ate particularly well. She lived through the Great Depression and was always trying to economize on food, even when she could have afforded higher quality food. When she got into her 80s and 90s, her diet deteriorated further: peanut butter, bananas, breakfast cereal, the occasional canned soup. She also stopped hydrating adequately. She then began having atrial fibrillation and small strokes, which led to falls and fractures. Each symptom was attributed to age. But atrial fibrillation has strong associations with magnesium and potassium deficiency. Cognitive decline and small strokes are linked to B12 deficiency, inadequate protein, and chronic dehydration. It still upsets me to think about how the false economies she was making contributed to her death.

American attitudes toward aging are unusually grim, and the grimness is not accidental. A youth-obsessed culture, a medical system that pathologizes aging, a retirement cliff that removes purpose and identity at a fixed date, geographic and social isolation of older adults, and the near-absence of multigenerational living combine to create a cultural environment that is probably as responsible for the decline as any dietary or activity pattern. Retirement as currently practiced removes the structure of daily life, identity, social connection, and cognitive challenge simultaneously — and the health consequences are significant and well-documented. Most people derive purpose from external sources: their job, their obligations, their role. When those end, there is often nothing underneath, because internal purpose — the kind that is self-generated and doesn’t depend on anyone else’s needs — was never developed. Passive technology consumption fills the time without building anything and accelerates exactly the cognitive decline it distracts from.

Ageism compounds this. At its mild end it is dismissiveness; at its worst it is a cultural narrative in which older people are net consumers of resources rather than contributors, whose appropriate role is to exit gracefully and promptly. An older person who is vigorous, capable, and independent doesn’t just defy the expectation — they implicitly challenge the narrative that justifies the resentment. The social pressure toward accommodation is persistent, and it is enforced by people who mean well. That makes it harder to push back against, because pushing back looks like ingratitude or recklessness rather than accurate self-assessment.

What Other Cultures Show

The clearest evidence that the American aging pattern is not inevitable comes from populations where it doesn’t happen. The Yanomami people of the Amazon show virtually no age-related blood pressure increase — a finding that would be impossible if arterial stiffening with age were purely biological. The Tsimane people of Bolivia show minimal cardiovascular disease and cognitive decline with age despite no access to modern medicine, attributable to high physical activity, low processed food consumption, and strong social integration. Traditional Okinawans — before the Westernization of their diet — had exceptional longevity and functional capacity, low rates of dementia, and remained physically active and socially integrated throughout life.

These populations are part of a broader pattern documented in what researchers call Blue Zones — geographically defined areas with unusually high concentrations of centenarians and dramatically lower rates of the chronic diseases that define American aging. The five identified Blue Zones are Okinawa, Japan; Sardinia, Italy; Ikaria, Greece; the Nicoya Peninsula in Costa Rica; and Loma Linda, California. Sardinia has the world’s highest concentration of centenarian men. Ikaria has some of the world’s lowest rates of middle-age mortality and dementia. Nicoya has the world’s lowest rate of middle-age mortality. Loma Linda is perhaps the most instructive example for an American audience: it is a community of Seventh-day Adventists living in California who consistently outlive their American neighbors by significant margins — not because of geography or genetics, but because of how they eat, move, and structure their social lives. The pattern isn’t confined to remote or exotic cultures. It exists within the United States, among people making different choices in the same environment everyone else inhabits.

Cultures where floor sitting is a normal daily activity — for meals, prayer, socializing — maintain the functional capacity to rise from the floor into old age because they never stop doing it. Cultures without a sharp retirement cliff, where older people remain integrated into productive community life and continue contributing meaningfully, show dramatically better health outcomes across nearly every metric. The biological aging is the same. The pattern is different. What differs is the lifestyle, the social integration, and the expectation.

The lesson is not that Americans should adopt any particular culture’s practices wholesale. It is that the American pattern of aging is a specific cultural and lifestyle product, not a biological universal — and that the components of other cultures’ better outcomes are identifiable, understandable, and largely replicable by individuals willing to make different choices.

Taking Back Control

None of what has been described in this essay has to be your story. But it will be, by default, if you leave the decisions to others — to a medical system that treats weight as the primary metric of health, to a food industry that has normalized a diet that makes you sick, to a culture that expects you to decline and accommodates that expectation until it becomes reality. The alternative is taking those decisions back. The following are where to start.

Resistance training. If there is one intervention with the broadest impact on the aging process, this is it. Resistance training builds and preserves muscle mass, increases bone density, improves insulin sensitivity, raises resting metabolic rate, reduces chronic inflammation, improves balance and functional capacity, and directly trains the movements — getting up from the floor, carrying loads, climbing stairs — that determine whether you remain independent in later life. It is the primary tool for body recomposition: simultaneously reducing fat mass and increasing muscle mass in ways that cardio alone cannot achieve. The goal is not aesthetics, though those follow. The goal is a body that can do things — and keeps being able to do them.

The principle underlying effective resistance training is progressive overload — gradually increasing the weight or resistance over time so the body continues to adapt rather than plateau. Without progression, the training maintains what you have but doesn’t build on it. Machine-based resistance training is often the most accessible and effective starting point, particularly for larger bodies, older adults, and people with physical limitations or disabilities. Unlike walking, bodyweight exercises, or conventional gym movements, machines allow you to control the load precisely, support the body externally, and isolate specific muscle groups without requiring your bodyweight to work against you. The person who finds walking painful, push-ups impossible, or conventional exercises mechanically awkward due to body size, injury, or disability is not failing at fitness — they have been pointed toward the wrong tools. Machine-based resistance training meets the body where it is.

Eat enough protein. Muscle protein synthesis becomes less efficient with age, meaning older adults need more dietary protein per pound of body weight than younger adults — not less, as the persistent “high protein is bad” belief suggests. That belief derives from research on people with pre-existing kidney disease and does not apply to healthy adults. A useful target for older adults engaged in resistance training is roughly 0.7 to 1 gram of protein per pound of bodyweight per day — significantly more than the RDA, which was set to prevent deficiency in sedentary people, not to optimize muscle retention in active ones. Prioritize complete protein sources: meat, fish, eggs, and dairy contain all essential amino acids in the ratios the body needs. Plant proteins are incomplete and less bioavailable, and for people trying to build muscle while maintaining a caloric deficit, the math of getting adequate protein from plant sources without exceeding caloric targets is nearly impossible in practice.

Eat for nutritional density, not calories. The standard American diet is engineered for palatability and shelf life, not for what the body actually needs. Shifting away from it doesn’t require a rigid program — it requires a different organizing principle. Instead of asking “how many calories is this?” ask “what does this give my body per calorie?” Meat, fish, eggs, and dairy deliver high-quality complete protein, essential fats, and dense micronutrients. Vegetables deliver fiber, polyphenols, and micronutrients without significant sugar load. These are the foods that have sustained human health across cultures and millennia. Ditch the grains, even the whole ones, and limit fruit consumption — modern commercial fruit has been selectively bred for sweetness to the point where it delivers a sugar load disproportionate to its nutritional benefits. Ultra-processed foods, refined seed oils, refined carbohydrates, and added sugars deliver calories with minimal nutritional return and significant inflammatory cost.

A concrete example: swap the avocado toast for tinned sardines packed in olive oil, not seed oils. Avocado toast is culturally coded as healthy. Sardines are not glamorous. But sardines deliver complete protein, omega-3 fatty acids, calcium from the bones, vitamin D, and B12. The avocado toast delivers refined carbohydrates and monounsaturated fat with minimal protein and no omega-3s. The choice between them is a small illustration of the larger principle: eat what actually does something, not what looks like it should. Be skeptical of dietary advice calibrated to a population eating catastrophically badly and optimized for biomarkers rather than outcomes. That advice has coincided with fifty years of worsening metabolic health. You can do better than the baseline it was designed for.

Cook your own food. The most direct way to control what goes into your body is to prepare it yourself. Restaurant food, takeout, and processed convenience food are prepared with cheap seed oils, excess sodium, added sugars, and ingredients chosen for shelf life and palatability rather than nutritional value. Cooking at home means you know what you’re eating — what oils were used, what the protein source actually is, what has been added and what hasn’t. It also means you can apply the nutritional density principle directly: choosing ingredients that do something rather than ones that merely fill space. Cooking is also, not incidentally, a skill worth having — purposeful, creative, and cognitively engaging in ways that passive consumption is not.

Take sleep seriously. The belief that older people need less sleep is an observation masquerading as a fact — older people often get less sleep, but that reflects the lifestyle factors degrading their sleep quality, not a reduced biological requirement. The need for deep, restorative sleep doesn’t diminish with age. If anything, it increases when you’re placing real physical demands on your body through resistance training — slow-wave deep sleep is where muscle protein synthesis and growth hormone release happen, where the adaptation from training actually occurs. Shortchanging sleep shortchanges the training. Investigate what’s disrupting your sleep rather than accepting fragmentation as normal. Common and addressable culprits include untreated sleep apnea, alcohol, excess light exposure at night, insufficient physical demand during the day, and medication side effects. Sleep apnea in particular is dramatically underdiagnosed in older adults and produces cognitive and cardiovascular damage that gets attributed to aging.

Know what your doctor is good at — and what they aren’t. Medicine excels at acute care, surgery, and infectious disease. It is considerably less reliable as a guide to nutrition, body composition, and the optimization of physical function in aging. Push back on weight as the primary metric of your health. Weight is easy to measure but is a poor proxy for individual health — fitness, functional capacity, muscle mass, and metabolic health are the variables that actually predict how well you age and how long you remain independent.

Review your medication burden, including over-the-counter medications taken regularly. Many common drugs produce side effects in older adults that look exactly like aging symptoms. Ask your doctor periodically whether each medication you take is still necessary, what it is actually doing for your all-cause mortality rather than just your biomarkers, and what its known effects on muscle, cognition, and nutritional status are. Also ask about interactions — older adults frequently take multiple medications simultaneously, and the combined effects are poorly understood and rarely systematically evaluated. Each drug was tested largely in isolation; what happens when five or six of them are taken together is often unknown, and the symptoms that result get attributed to aging rather than to the pharmaceutical burden as a whole.

Ask specifically for nutritional testing that routine care doesn’t include: B12 and methylmalonic acid, vitamin D, magnesium, omega-3 index, zinc. These deficiencies are common, produce symptoms that get attributed to aging, are straightforward to address, and are almost never checked unless you request them.

Address nutritional deficiencies proactively. Several deficiencies are nearly universal in older adults eating the standard American diet, produce symptoms routinely mistaken for aging, and are straightforward to correct once identified. The irony is that the dietary advice most commonly given by doctors systematically reduces intake of the foods highest in the nutrients most commonly deficient. Vitamin D, B12, magnesium, omega-3s, zinc, and choline are the most consequential. Don’t wait for symptoms to appear — by the time B12 deficiency is symptomatic, it has been accumulating for years. Get tested, supplement where indicated, and address the dietary patterns that created the deficiency in the first place.

Resist the cultural expectation of decline. The expectation that aging means slowing down, giving things up, and being careful is not a neutral observation — it is an instruction, and following it produces exactly the outcome it predicts. When a family member worries about the stairs, when a doctor tells you to take it easy, when the culture signals that vigor in older age is surprising — these are pressures toward accommodation that feel like care but function as acceleration of decline. You are allowed to ignore them. Set high expectations for what your body can do, and assume that falling short of them still leaves you better off than having started from lower ones.

Develop internal purpose before you need it. Most people derive purpose from external sources — their job, their obligations, their role as a parent. When those end, there is often nothing underneath, because internal purpose — the kind that is self-generated and doesn’t depend on anyone else’s needs — was never developed. The activities that sustain people through the later decades of life are the ones that are intrinsically motivated: pursued because they matter to you, not because anyone needs you to pursue them. They can’t be retired from. The time to build them is not after the external sources of purpose disappear. Passive technology consumption fills time without building anything and accelerates exactly the cognitive decline it distracts from.

Stay socially connected. The evidence on social isolation and health outcomes is as strong as almost anything in medicine — chronic isolation carries mortality risk comparable to smoking. Social connection in later life requires active maintenance in a culture that segregates older adults and removes the workplace social world at retirement. Invest in relationships that don’t depend on shared employment or shared obligation. Seek out people who are intellectually alive and physically engaged. The social environment shapes expectations and behavior as surely as any individual choice.

The Choice

The decline described in this essay is not inevitable. It is a choice — usually not a conscious one, but a choice nonetheless, made by default when the decisions get left to others. The people who age differently from the American norm are not genetic outliers or the unusually fortunate. They are people who made different choices, held different expectations, and refused to accept the cultural script that told them what their later decades were supposed to look like. Some of them are in populations that never adopted the American lifestyle. Some of them are in your gym, lifting weights at age 75, moving without pain, sleeping well, eating food that does something. They are not remarkable. They are what is possible when the default assumptions get questioned.

The assisted living brochure and the cremation brochure will arrive in your mailbox regardless — that’s how mass mailing works. What doesn’t have to arrive on schedule is the reality they assume. But it will, by default, if the decisions get left to others — to a system that has normalized the decline it was supposed to prevent, to a culture that has made frailty and chronic disease the expected shape of old age, to the quiet accumulation of accommodations that each seem reasonable and together constitute a life diminished before its time. The alternative is available. It requires effort, skepticism of comfortable advice, and the willingness to hold higher expectations than the people around you think are warranted. It is worth it.

Every person carries in their mind an internal model of reality — a picture of how the world works, what other people are like, what is true and what isn’t, what is likely and what isn’t. This model is not optional. It is the lens through which all decisions are made, all situations are read, all other people are assessed. You cannot act effectively without it. An inaccurate model produces bad decisions — and the costs of those decisions accumulate over time.

There is a skill that predicts life outcomes better than almost anything else we typically measure — better than optimism, better than credentials, arguably better than intelligence. It isn’t glamorous, and it doesn’t feel good in the short term. It is simply the ability to see yourself, other people, and situations as they actually are, rather than as you would prefer them to be. Accurate reality-modeling — asking yourself, given how reality works, what is most likely true in this situation, and letting the answer correct your picture of the world — produces better outcomes over a lifetime than choosing the more comfortable distortion. Not because it feels better. Often it doesn’t. Because it lets you act on what’s actually true.

One clarification before going further: clear perception is a cognitive tool, not a moral virtue. A knife can be used by a surgeon or a murderer. The precision of the blade is independent of what it’s aimed at, and so is this. Accuracy aimed at harm still serves the person wielding it.

What Poor Reality-Modeling Actually Looks Like

Distorted thinking takes many forms. The following are examples — not an exhaustive catalog, but a representative sample of the ways people routinely substitute a more comfortable or convenient picture for an accurate one.

Overestimating one’s own competence. People who lack a skill are often also the least aware of the gap — they don’t know enough to know what they don’t know. The harm is practical: it prevents seeking help, taking corrective action, or even recognizing that a problem exists.

Attributing other people’s behavior to character rather than circumstance, or circumstance rather than character. Someone is late and you conclude they’re inconsiderate rather than considering they hit traffic — but the reverse error is equally costly: excusing genuinely inconsiderate behavior as situational when it’s actually a reliable pattern. Wrong conclusions about people produce wrong decisions about how to deal with them.

Confusing correlation with causation. The stock went up after you bought it, therefore your analysis was correct. This produces false confidence in a reasoning process that may have had nothing to do with the outcome, leading to larger bets on a flawed model.

Sunk cost reasoning. Continuing a bad investment, relationship, or course of action because of what’s already been invested rather than what’s actually ahead. Past losses are unrecoverable regardless of what you do next — letting them drive future decisions means compounding a mistake instead of cutting it off.

Availability bias. People attach too much weight to vivid or recent examples when estimating probability. Plane crashes are feared more than car accidents because crashes are dramatic and memorable, despite the actual odds running the other way. Decisions get made on the basis of what comes easily to mind rather than what’s actually likely.

Motivated reasoning. Starting with the conclusion you want and working backward to justify it, rather than following the evidence forward. It looks like thinking — it uses the machinery of logic and evidence — but it’s aimed at confirming what you already believe rather than finding out what’s true. It’s the hardest of these errors to catch in yourself because it feels indistinguishable from genuine reasoning.

Assuming that because something bad hasn’t happened yet, it won’t. The odds are indifferent to your personal history with them. Nobody gets in the accident expecting to.

Not knowing what things should cost. Knowing current market prices — for groceries, for services, for anything you buy regularly — gives you a baseline against which to evaluate any given transaction. Without it, you have no way to recognize a good price or a bad one, and you end up paying whatever you’re asked.

The Compounding Argument

The reason seeing clearly produces better outcomes isn’t that it guarantees success on any single decision. It doesn’t. The reason is that it produces a small but consistent edge that accumulates over many decisions — and over a lifetime, that accumulation is decisive.

Consider how a casino operates. The house doesn’t win every hand. It wins reliably over time because the odds are marginally in its favor on every hand. The edge is small enough that any individual player can win on any given night. But extend the game long enough, and the math is inexorable. The same logic applies to decisions made under accurate versus distorted thinking. Hold everything else constant and run the same person down two branches — one where they see situations clearly, one where they consistently choose the more comfortable picture — and the accurate branch outperforms the distorted branch. Not on every decision. Over a lifetime of decisions.

This is closer to a logical claim than an empirical one, and it should be stated as such. It doesn’t require a longitudinal study to establish.

What the Research Says, and What It Actually Shows

There is a body of research on positive illusions — the tendency most people have to see themselves, their prospects, and their control over outcomes in a somewhat more favorable light than the evidence warrants. This research is real and shouldn’t be dismissed. Positive illusions do correlate with better subjective well-being. People who hold them report feeling better, coping better with adversity, maintaining more optimism under stress.

The research does show that positive illusions produce better subjective well-being. What it does not show is better objective outcomes. The distinction matters enormously, and it tends to get lost. Feeling better about your situation is not the same as improving it. The aspiring dancer who invests years pursuing a career without the requisite talent feels better about those years, perhaps, than she would if she’d seen clearly earlier. The gambler who overestimates his ability to influence a chance outcome feels more in control. Neither of them is better off for the illusion — they are simply more comfortable on the way to a worse outcome.

There is no reason clear perception and feeling good are mutually exclusive over the long term. You can have a genuinely good marriage and feel great about it. You can build a genuinely successful career and take real satisfaction in it. What positive illusions actually deliver is short-term comfort — and that’s not nothing. Feeling better now is a real benefit. But it comes at a cost that compounds in the other direction: the marriage that’s bad and feels fine, the career that’s failing and looks acceptable, the situation that needed addressing six months ago and still hasn’t been. The illusion is consolation, not achievement. It only becomes relevant once you’ve already lost on the objective dimension, and it tends to keep you there.

The Depressive Realism Literature, Corrected

A related piece of research gets invoked as a counterexample to the value of clear perception: depressive realism. The original finding, from Alloy and Abramson, was that depressed people sometimes judge their actual degree of control over outcomes more accurately than non-depressed people do. The researchers interpreted this as depression causing accurate perception. But the causal arrow is possibly backwards.

Accurate perception of a genuinely bad situation can produce depression, because the situation is genuinely bad. That’s not a malfunction — that’s the alarm going off. If you lose your job, that is a genuinely depressing thing. The accurate response is to feel the weight of it and act on what you know. The distorted response is denial: refusing to fully accept the situation, delaying telling people, avoiding updating the resume, telling yourself it might work out.

The critical distinction that the research tends to collapse is between situational depression — an appropriate response to circumstances that are actually bad — and pathological depression, which persists independent of circumstances and distorts perception in its own right. Seeing clearly that your situation is hard is not a symptom. It is correct information. The discomfort is the message, not the malfunction.

It is worth pausing on the interpretive failure itself. The researchers found accurate perception correlated with depression and concluded that depression causes accurate perception — without apparently considering the more obvious sequence: that seeing a bad situation clearly might be what produces the depression. The interpretation they landed on was also the more convenient one for the field. If accurate perception is a feature of depression, it’s a clinical symptom to be managed. If depression is sometimes the correct response to a genuinely bad situation, then the situation needs addressing — and that’s a harder problem. In interpreting their own findings, the researchers demonstrated the very failure they were studying.

Why Seeing Clearly Is Harder Now Than It Used to Be

For most of human history, the consequences of distorted thinking were direct, immediate, and personal. Poor judgment about physical risk produced injury. Financial miscalculation produced loss. Social misjudgment in a small community — where inaccurate claims didn’t survive contact with someone who knew you — produced damaged relationships and damaged reputation. These consequences were unpleasant. They were also a feedback loop. They forced the development of clear perception, and they revealed failures quickly enough to allow correction.

That loop has been substantially weakened by the complexity of the modern world — its technology, its institutional layering, and the resulting lack of transparency between causes and their consequences. Insurance spreads the financial cost of bad judgment across a pool of people, introducing what economists call moral hazard: when the party making the judgment is no longer the party bearing the full cost if the judgment is wrong, the incentive to judge accurately weakens. Financial systems socialize investment risk. Social programs buffer the consequences of a wide range of failures. The result is that when things go wrong, the cause is usually diffuse enough that no single person’s action is clearly implicated.

The result is not just that failures go uncorrected. It’s that the skill fails to develop in the first place. A feedback loop that never runs with real stakes doesn’t just fail to correct errors — it fails to build the capacity to avoid them. No-child-left-behind style education policy made this explicit: the signal “you don’t actually know this” was deliberately removed from the system, on the theory that the signal was harmful to children. What was actually removed was the mechanism by which children learn that their picture of their own competence needs to be checked against reality. The teacher who demands accurate performance becomes the problem. The standard that expects it is “too much.”

The Normalization of Distorted Thinking

This is no longer just the removal of consequence. It has become the active normalization of distorted thinking — the social enforcement of the idea that demanding accurate assessment is itself the offense. Religion is the most entrenched example: a set of beliefs about the nature of reality that is structurally insulated from correction, where the insulation is not merely tolerated but enforced, and where criticism of the beliefs — or of the insulation itself — is socially penalized. The normalization isn’t passive. It requires active maintenance.

The “alternative facts” phenomenon took this a step further — asserting not merely that certain beliefs should be protected from criticism, but that reality itself is a matter of preference. The laws of physics, however, are indifferent to what anyone prefers. A structural flaw in a bridge does not suspend its consequences because the engineer believed otherwise. A disease does not alter its course to accommodate a patient’s conviction that it isn’t serious. Reality operates independently of the belief systems people construct around it, and the consequences of that independence accumulate whether or not anyone acknowledges them.

The Cost of Comfort Mistaken for Kindness

When distorted thinking produces distress — as it often does, eventually — the socially rewarded response is frequently to address the distress rather than the distortion. To change how someone feels about a bad situation rather than helping them change the situation. This is not kindness. If you saw smoke coming from your neighbor’s house, you wouldn’t bring them a fan to blow it away — you would call 911 and get them out. Comfort that leaves the source of the problem intact is the fan. It makes the immediate moment more bearable and the underlying situation worse.

This pattern has been institutionalized. Mental health treatment has built comfort-restoration into its standard toolkit — not universally, and not without exception, but as a default strong enough that a clinician’s first response to expressed distress is frequently to ask how to reduce it rather than what it’s accurately signaling. The distinction between situational depression, which is appropriate distress in response to real circumstances, and pathological depression, which requires different intervention entirely, often gets lost in the process. Adjusting the perception rather than addressing the situation is the clinical version of the same error individuals make when they choose the comfortable distortion.

At the civilizational scale, the pattern is visible in every context where performative empathy gets rewarded over actual help. “Thoughts and prayers” is the canonical example: a response to tragedy that is entirely oriented toward the feelings of the responder and the comfort of the recipient, with no mechanism for changing the situation that produced the tragedy. It is socially rewarded. It produces nothing. The gap between the reward and the outcome is a measure of how thoroughly comfort has been institutionalized as a substitute for clarity.

What You Can Do About It

Clear perception is a skill, not a fixed trait. It can be developed. The following are some directions worth moving in — most of them never covered in formal education, which spends considerable effort teaching people how to earn a living and very little teaching them how to navigate the world they’ll be living in.

Understanding basic science. You don’t need to be a scientist, but a basic grasp of how the physical and natural world operates gives you a framework for evaluating what’s actually possible. It provides a check on claims that violate scientific principles — perpetual motion machines, miracle cures that work by “energy,” products that promise to do things physics says they can’t. It’s also a defense against organized misinformation that depends on scientific illiteracy: flat earth theories, young earth creationism, climate denial, and vaccine skepticism all require ignoring or contradicting well-established principles of physics, biology, and earth science. If you don’t know what those principles actually say, you have no basis for recognizing when someone is asking you to believe something that contradicts them.

Understanding basic economics and financial principles. A working knowledge of how markets, prices, interest, and risk operate gives you tools that most people never acquire formally. Knowing what things should cost — and why — lets you recognize when you’re being overcharged, when a deal is genuinely good, and when a financial product is designed to benefit the seller rather than you. Most formal education, to the extent it addresses financial life at all, focuses exclusively on earning — getting the credentials that lead to the job that produces the income. How to keep that income, grow it, and protect it is treated as someone else’s problem, or not addressed at all, unless you are training for a career in finance. Understanding compound interest, the difference between assets and liabilities, how insurance works, what diversification actually does, and how to evaluate risk are not advanced topics — they are basic tools for navigating adult life, and the majority of people reach adulthood without them. The absence of this knowledge doesn’t stay neutral; it gets exploited.

Knowing at least a little about a lot of domains. Specialized knowledge is valuable, but exclusive focus on one area leaves you dependent on other people for everything outside it — and dependent means vulnerable. The danger isn’t only from people trying to take advantage of your ignorance; it’s at least as often from well-meaning people who are themselves poorly informed and don’t know it. A working familiarity with medicine, law, home maintenance, nutrition, technology, and other domains that affect daily life lets you ask better questions, recognize bad advice regardless of the intentions behind it, and maintain enough independence to make your own informed decisions. You don’t need to be an expert in everything. You need to know enough to know when something doesn’t add up.

Keeping up with current events. An accurate picture of the world requires current information. The world changes — governments, markets, technologies, social norms, scientific understanding — and a picture of reality built on outdated information produces outdated decisions. This doesn’t mean consuming news compulsively or following every story; it means maintaining enough awareness of what’s happening in the domains that affect your life to make informed decisions. It also means being selective about sources — not all information presented as news is accurate, and part of keeping up with current events is developing the ability to evaluate the reliability of what you’re reading.

Eliminating magical thinking. Magical thinking is the belief that the world operates according to wishes, intentions, or narratives rather than physical and probabilistic laws. It shows up in obvious places — superstition, astrology, certain religious beliefs — but also in subtler ones: the conviction that wanting something badly enough will make it happen, that a run of bad luck must eventually “even out,” or that putting something “out to the universe” will cause it to accommodate you. The universe is indifferent to your wants and needs. Recognizing and eliminating magical thinking in yourself is one of the more useful things you can do.

Seeking out disconfirming evidence. Most people, most of the time, consume information that confirms what they already believe. It feels better. It’s also a reliable path to an increasingly distorted picture of reality. The problem has been compounded by the algorithms that drive social media and news feeds, which are designed to show you more of what you’ve already engaged with — meaning the longer you use them, the more your information environment narrows around your existing beliefs. The corrective is deliberately seeking out credible sources that challenge your existing views — not to be contrary, but because the information you’re most likely to be missing is the information that doesn’t fit your current picture. If you can’t accurately state the strongest case against your own position, you probably don’t understand the issue as well as you think you do.

Being willing to update your views. Changing your mind when presented with good evidence means admitting you were wrong — and that’s exactly what it should mean. Being wrong about a fact or an assessment is not a moral failure; it’s simply having operated on incomplete or inaccurate information. The problem is that many people treat being wrong as a reflection of character rather than a data point, which makes updating feel like a threat to identity rather than an improvement to their picture of the world. The reframe that helps is straightforward: a changed position means you now see more clearly than you did before. That’s not a concession. It’s the point.

Being able to say “I don’t know.” The reach for a confident answer where none is warranted is one of the most common and least examined habits people have. It feels better to have an answer than to sit with uncertainty — but a confident wrong answer is worse than an acknowledged gap, because it closes off the inquiry that might have produced an accurate one. “I don’t know” is not a failure; it’s an accurate description of your current state of information. It’s also the only honest starting point for finding out.

Slowing down on consequential decisions. Many of the reasoning errors described earlier — availability bias, motivated reasoning, sunk cost reasoning — are more likely to operate when you’re moving fast and not examining your own thinking. The corrective is simple in principle and difficult in practice: on decisions that matter, slow down. Ask what you might be missing. Ask who benefits if you decide quickly. Ask whether you’d make the same decision if you had another week to think about it. Urgency is frequently manufactured by people who want you to decide before you’ve had time to think clearly, and recognizing that pressure for what it is — a manipulation tactic, not a genuine constraint — is itself a form of clear thinking.

Learning to weight future consequences. Most people have a natural bias toward the present — immediate costs and benefits feel more real than distant ones, even when the distant ones are larger. This bias is exploited constantly: by lenders who emphasize the monthly payment rather than the total cost, by politicians who promise immediate relief with long-term price tags, by our own preferences when we choose comfort now over effort that would pay off later. The corrective is developing the habit of asking, for any significant decision, what this will look like in five or ten years — not just what it solves or costs right now. The compounding argument at the heart of this essay depends on this capacity: the edge that clear perception produces only accumulates if you’re playing a long game.

Improving emotional intelligence. This operates in two directions. Inward: the ability to accurately read your own emotional states and recognize how they’re affecting your thinking. Fear makes threats look larger. Desire makes risks look smaller. Anger narrows focus in ways that exclude relevant information. Someone who can recognize “I’m afraid right now, and that’s affecting how I’m reading this situation” is in a much better position to correct for the distortion than someone who mistakes their emotional state for an objective assessment of reality. Outward: the ability to accurately read what other people are actually feeling, wanting, and motivated by — as distinct from what they’re saying. Most people are not fully transparent about their motives, sometimes because they don’t know them themselves. The ability to read beneath the surface of what someone is telling you is one of the more practically useful skills a person can develop.

Paying attention to results rather than intentions. It is tempting to evaluate decisions by what you meant to accomplish rather than what they actually produced. Intentions are comfortable to assess — you know what you meant, and you meant well. Results are harder, because they exist independently of what you wanted. The same error applies when evaluating other people: someone who damages you with good intentions has still damaged you. The damage is real regardless of what motivated it, and excusing it on the basis of intentions rather than outcomes is a form of motivated reasoning that leaves the underlying problem unaddressed. The habit of honestly tracking what decisions actually produce — yours and other people’s — rather than what they were intended to produce, is one of the more reliable ways to improve the accuracy of your thinking over time. This extends to how you read people in general: what someone does is more reliable information than what they say. Words are easy to produce and easy to shape toward what the listener wants to hear. Actions are harder to fake consistently over time. If someone’s actions and words are in conflict, the actions are the more accurate data.

Treating procrastination as a signal. Procrastination is commonly attributed to laziness, but that’s rarely what’s actually driving it. More often, procrastination is a signal that there’s something about the task you’re unwilling to engage with directly — a difficult conversation you’re avoiding, a decision whose implications you don’t want to face, a piece of work that will require confronting how much you don’t know. Recognizing procrastination as avoidance rather than laziness changes what you do about it. The question isn’t “how do I make myself do this?” It’s “what am I unwilling to look at clearly, and why?”

The Cost of the Alternative

The habits and knowledge described in this section are not a program with a finish line. They are directions to move in, and moving in them is a lifelong practice. It is also, at times, a genuinely hard one — and that difficulty is worth naming directly.

Seeing clearly doesn’t just require perception. It sometimes requires acting on what you see when the action is costly in the present and the reward is distant. I stayed in a job I disliked for six additional years because, at the end of it, I would have a lifetime pension, paid health insurance, and the ability to retire at 56. When I mentioned this calculation to my therapist, she was surprised. I was surprised that she was surprised — it seemed like an obvious trade. She told me most people wouldn’t make it. They would do what was easier in the present and pay for it later.

That gap — between the calculation and the willingness to act on it — is where clear perception either pays off or doesn’t. Seeing clearly isn’t enough if you’re not willing to accept the present cost of what you see. And a society that has systematically removed consequences, normalized distortion, and rewarded comfort over clarity makes that willingness harder to develop and easier to abandon.

The choice, in the end, is straightforward even when it isn’t easy. Seeing clearly requires deferring the comfort of the flattering picture in exchange for better outcomes over time. The alternative is the short-term ease of the comfortable distortion — and a lifetime of paying its compounding costs.

Say that religion causes harm, and prepare for the reaction. Not disagreement — outrage. The feeling that you have reached into something sacred and pulled it apart with your bare hands. That you have attacked not just an idea but a grandmother’s faith, a childhood Sunday morning, a deathbed comfort, something that has held families together through grief and poverty and war. The charge of harm lands like an accusation against the most intimate parts of a person’s life, and the response is rarely calm.

You will be told that you are confusing religion with organized religion — that the church has corrupted the pure teachings, that the Inquisition and the Crusades were institutional failures, not theological ones. You will be reminded that Jesus said to love your neighbor. You will be asked what you have against comfort, community, and hope.

But religion is not benign. The teachings themselves — not merely their institutional corruptions — are the source of harm: a Bible that has justified slavery, genocide, and the mistreatment of women in its own words; a definition of faith that structurally undermines the reasoning capacity of its adherents; a system that asks people to treat the suppression of evidence as a virtue. These teachings are not harmless, no matter how warmly and sincerely they are practiced.

The Mythology of Christian America

Before examining the harms produced by religious teaching, it is worth clearing away some of the mythology that makes those teachings feel more foundational to American life than they actually are.

The United States was not founded as a Christian nation. The Constitution — the foundational legal document of the republic — does not mention God, Jesus, or Christianity. This was not an oversight. The founders were explicit about their intent to create a secular government, and several of the most prominent among them were not Christians in any orthodox sense. Thomas Jefferson, the author of “all men are created equal,” was a deist who found the supernatural claims of Christianity unpersuasive enough that he produced his own edited version of the Gospels, cutting out the miracles and the resurrection. Jefferson also owned more than six hundred enslaved people over his lifetime and had a decades-long sexual relationship with Sally Hemings, an enslaved woman who could not consent. The gap between Jefferson’s stated principles and his lived practice is not incidental — it is a preview of the selective reading that has characterized American Christianity ever since.

Benjamin Franklin was a deist. James Madison, the primary architect of the Constitution, was deeply skeptical of organized religion and explicitly opposed to government entanglement with it. The founding generation was shaped by Enlightenment political philosophy — Locke, Montesquieu, Rousseau — far more than by Christian theology. The “Christian nation” narrative erases this history in favor of a mythology that serves contemporary political purposes.

Two of the most commonly cited pieces of evidence for America’s Christian identity — “In God We Trust” on currency and “one nation under God” in the Pledge of Allegiance — were both added in the 1950s, during the Cold War, as explicit contrasts to godless communism. Neither is a founding principle. The Pledge itself was written in 1892 by a Baptist minister and did not include “under God” for its first sixty-two years of existence. These facts are not obscure. They are a matter of public record. That so many Americans believe otherwise is a demonstration that faith communities are not reliable custodians of accurate history.

The Big Book of Multiple Choice

The Bible is, among other things, a document that contains its own contradictions — and those contradictions are a feature, not a bug. They are the feature that makes the text infinitely useful as a justification for whatever the reader already wants to do. Call it what it is: the Big Book of Multiple Choice. Whatever command you can find in its pages, there is probably another that countermands it. Whatever cruelty you want to commit, there is a passage that sanctions it. Whatever compassion you want to practice, there is a passage that supports it. The book is not a moral framework. It is a moral menu.

Consider the Ten Commandments, which American Christians have fought to display in courthouses and public schools as the foundation of Western morality. Most people invoking the Ten Commandments are unaware that there is more than one version of them. The familiar version appears in Exodus 20 and is repeated in Deuteronomy 5, with differences. But there is a third version, in Exodus 34, that God himself explicitly labels the Ten Commandments — and that version includes commands like “do not boil a young goat in its mother’s milk.” This is not a minor textual footnote. It is the text. The version being cited in constitutional arguments is not the one the Bible calls by that name.

When confronted with the Bible’s more troubling passages — the ones commanding slavery, genocide, and the mistreatment of women and children — the standard deflection is that these are Old Testament commands, superseded by the New Testament. This deflection collapses immediately when examined. The Ten Commandments appear only in the Old Testament. If the Old Testament is superseded, the Ten Commandments have no authority. If the Ten Commandments have authority, the Old Testament is still in play — including Leviticus’s sanction of slavery, Numbers’ commands to slaughter entire peoples, Deuteronomy’s instructions for the treatment of women, and 2 Kings 2:23–24, in which God sends two bears to maul forty-two children for mocking a prophet’s baldness.

The Bible explicitly endorses slavery. Leviticus 25:44–46 permits the ownership of enslaved people from neighboring nations and instructs that they can be passed on as inherited property. Ephesians 6:5 instructs enslaved people to obey their masters. Exodus 21 provides rules for the treatment of enslaved people, including how long a beating can last before it counts as a crime. American slaveholders did not pervert scripture to justify their practices. They read it and implemented it.

The Bible commands the mistreatment of women. 1 Timothy 2:12 instructs that women should not teach or have authority over men. 1 Corinthians 14:34 commands women to be silent in church. Ephesians 5:22 instructs wives to submit to their husbands. These are not obscure passages. They have been cited from pulpits to deny women education, professional authority, and bodily autonomy for centuries.

The Bible sanctions homophobia. Leviticus 20:13 calls for the execution of men who have sex with men. Romans 1:26–27 describes same-sex relations as shameful and unnatural. These passages have been used to justify violence, family rejection, conversion therapy, and legislative discrimination against LGBTQ+ people. The people wielding these passages are not misreading the text.

The Bible commands genocide. Numbers 31 records God’s command to kill the Midianite men and boys and take the women and girls as spoils. Deuteronomy 7 commands the Israelites to destroy seven nations completely, showing them no mercy. 1 Samuel 15 records God’s command to kill every Amalekite, including infants.

The standard response to these passages is that they must be understood in their historical and cultural context. But consider this — if God’s commands were contingent on historical context, they were not universal moral truths. And if they were not universal moral truths, the entire basis for using them to legislate today collapses. But more than that: the belief that there exists a context in which slavery is not harmful, in which genocide is not harmful, in which the systematic mistreatment of women is not harmful — that belief is itself one of the most insidious harms religion produces. It does not require anyone to own a slave. It requires only the conviction that owning a slave could, under the right circumstances, be righteous.

Magical Thinking and Its Spillover

Faith is defined, within most Christian traditions, as belief without requiring evidence — and this is treated as a virtue. Hebrews 11:1 defines faith as “the substance of things hoped for, the evidence of things not seen.” To believe without evidence is presented not as an epistemic failure but as a spiritual achievement.

The harm of this is not limited to the specific beliefs it produces. A habit of mind does not respect domain boundaries. Someone who has been trained to accept claims without evidence, and to treat that acceptance as virtuous, has weakened their evidence-evaluation instincts across every domain of their life. The same cognitive move that works for the resurrection works equally well for homeopathy, for conspiracy theories, for the claim that a particular political candidate has been chosen by God. The mechanism is identical. The training transfers.

There is a specific and underexamined version of this harm that deserves attention: the belief that God is actively guiding events, that there is a divine plan unfolding, that God is present and in control of outcomes. This belief functions like being a passenger in a car whose driver you trust completely — you stop watching the road, stop developing your own navigation skills, stop taking responsibility for where you end up. If God is steering, your own judgment is secondary. The problem is that there may be no driver. And a passenger who believes they are being cared for, in a vehicle with no one at the wheel, is careening toward whatever the road delivers — not guided, but unguarded.

The spillover from magical thinking into other domains is well documented even if the religious root is rarely named. Anti-vaccine movements rely on the same willingness to dismiss expert consensus in favor of personally compelling narrative. Conspiracy theories require the same capacity to maintain belief in the face of contradicting evidence and to dismiss that evidence as part of the conspiracy. Alternative medicine operates on the same preference for belief over mechanism. These are not separate phenomena. They are the same habit of mind, trained in one domain and expressed in others.

Comfort Over Clarity

Religion offers comfort. This is frequently cited as its primary social benefit, and it is not a negligible one. The comfort of believing in a benevolent God, in an afterlife, in the meaningfulness of suffering — these are real psychological resources, and dismissing them entirely misunderstands what they provide.

But there is a distinction worth making between comfort that helps someone cope with an accurately perceived reality, and comfort that requires distorting reality to be sustained. The first kind is genuinely useful. The second kind is a trade: you receive the feeling of comfort in exchange for giving up your capacity to see clearly. Religion, as a system, offers the second kind. It asks its adherents to accept as true things that are not demonstrably true, and to treat the willingness to do so as a virtue. The price of the comfort is epistemic harm.

This makes religion unique among major human institutions. Other comforting beliefs get corrected by consequences — if you believe you are a better driver than you are, reality eventually provides feedback. Religious belief is specifically structured to resist that correction. Doubt is reframed as sin or weakness. Evidence that contradicts the framework is dismissed as a test of faith or the work of the adversary. The feedback loop between belief and reality is deliberately broken. No other institution asks its members to treat that breakage as a sign of spiritual health.

Human Exceptionalism and Its Costs

Genesis places humans at the apex of creation, made in God’s image, granted dominion over every living thing. This is not a minor theological detail. It is the foundational claim that positions humanity as categorically different from and superior to the rest of the natural world — not merely more complex, but differently constituted, possessing an immortal soul that no other creature shares.

The scientific costs of this claim have been significant. The most consequential is the resistance to evolutionary biology, which directly contradicts the biblical account of human origins and, more fundamentally, the claim of human exceptionalism itself. Evolution does not place humans at the apex of creation. It places humans as one branch among millions, continuous with rather than categorically above the rest of life on earth. This is not a comfortable finding for a tradition built on the claim of divine special creation, which is why resistance to it has been so persistent.

The medical consequences of evolution resistance are not abstract. Modern medicine depends on evolutionary biology at every level. Understanding how pathogens evolve resistance to antibiotics requires evolutionary thinking. Vaccine development requires understanding how viruses mutate. Cancer treatment increasingly relies on understanding how tumor cells evolve. Genetic medicine is built on the understanding of shared ancestry across species. A population systematically taught to distrust evolutionary biology is a population poorly equipped to understand, evaluate, or advocate for modern medicine.

The Genesis mandate of dominion has functioned as theological permission to treat the natural world as raw material for human use. The consequences — environmental destruction, factory farming, the extinction of species — are not a misapplication of the teaching. They are the teaching working as intended.

Suffering as Spiritual Failure

The Prosperity Gospel — the theological position that financial wealth and physical health are signs of God’s favor, and that poverty and illness are signs of insufficient faith — is the most explicit and commercially lucrative version of a broader assumption that runs through much of Christian thinking: that suffering is meaningful, that it is sent or permitted by God, and that its presence in a person’s life says something about their spiritual state.

The harms of this framework are multiple and compounding. It produces shame and self-blame in people who are already struggling — not just the suffering of poverty or illness, but the additional burden of believing that suffering is their fault, a sign of spiritual inadequacy. It discourages systemic analysis of the conditions that produce poverty and suffering, because if those conditions are divinely ordained or spiritually caused, human intervention is at best beside the point and at worst presumptuous. It undermines empathy for people who are struggling, since their difficulty can be attributed to their own relationship with God rather than to circumstances that deserve collective response. And it has been used explicitly to argue against social safety nets, on the grounds that alleviating poverty interferes with divine will or removes the spiritual incentive for people to improve their faith.

The Prosperity Gospel is, in the most precise sense, a spiritual Ponzi scheme: congregants give money believing divine favor will return it multiplied, the wealth flows upward to the pastor, and the theology explains why it keeps working for some and not others — insufficient faith. This is not a corruption of the teaching. It is the teaching operating as designed.

Where the Harm Gets Concrete

The epistemological harms described above would be serious enough if they stayed abstract. They do not. They produce concrete, documented, measurable damage in specific domains.

The influence of Rapture theology on environmental policy is not theoretical. James Watt, Ronald Reagan’s Secretary of the Interior from 1981 to 1983, was a self-described dispensationalist — a theological tradition centered on the imminent Second Coming of Christ and the Rapture. In congressional testimony in 1981, Watt said: “I do not know how many future generations we can count on until the Lord returns.” During his tenure he opened nearly all of America’s coastal waters to oil and gas drilling, widened coal access on federal lands, and eased restrictions on strip-mining. Watt was not an anomaly. Senator James Inhofe of Oklahoma, one of the most influential opponents of climate legislation in Congress, has explicitly cited his belief that God controls the climate as grounds for dismissing climate science. Multiple peer-reviewed studies have found that end-times theology is a statistically significant predictor of Americans opposing government action to protect the environment. This is the natural endpoint of human exceptionalism combined with Rapture theology: if the world is temporary and its end divinely scheduled, stewardship is beside the point.

Stem cell research has been significantly slowed by religious objections to research on embryos, with direct consequences for the development of treatments for Parkinson’s disease, diabetes, spinal cord injuries, and other conditions. People are living with these conditions, and dying from them, in part because the research that might help them has been constrained by theological objections.

Reproductive technologies including IVF have faced religious opposition that has produced real legal vulnerability. The Alabama Supreme Court’s 2024 ruling that frozen embryos are children — a ruling grounded explicitly in Christian theology — briefly shut down IVF services in the state, affecting people who needed those services to have children.

Abstinence-only sex education, mandated in many states by religious influence on public policy, produces measurably higher rates of teen pregnancy and sexually transmitted infections in the populations where it is taught. This is not a theoretical harm. It is documented in the epidemiological literature. The consequences of religiously motivated ignorance about human reproduction extend far beyond the classroom. In 2012, Missouri Congressman Todd Akin, a member of the House Science Committee, claimed that in cases of “legitimate rape,” the female body has ways to “shut that whole thing down” to prevent pregnancy — a belief with no basis in biology, deployed to argue against rape exceptions in abortion law.

The President’s Emergency Plan for AIDS Relief, known as PEPFAR, launched in 2003 as one of the largest international health initiatives in history, directing billions of dollars toward HIV prevention and treatment in sub-Saharan Africa and other severely affected regions. Religious conservatives successfully lobbied to require that a significant portion of PEPFAR funding go toward abstinence-only programs rather than condom distribution. Organizations receiving PEPFAR funds were also required to sign an anti-prostitution pledge that limited their ability to work with sex workers, among the highest-risk populations for HIV transmission. The United States spent more than $1.4 billion on these abstinence programs over twelve years. A 2016 Stanford University study examining nearly 500,000 individuals across twenty-two sub-Saharan African countries found no evidence that the abstinence funding produced any reduction in high-risk sexual behavior. The consequence was not theological disagreement. It was preventable HIV transmission and preventable death, at scale, in some of the world’s most vulnerable populations.

Religious communities have historically discouraged or actively opposed psychiatric treatment, with serious consequences for people with treatable conditions. Clinical researchers across multiple countries have documented cases of patients with schizophrenia and other serious mental illnesses abandoning psychiatric medication in favor of exorcism and prayer, delaying or forgoing treatment with predictable consequences. The case of Andrea Yates — who drowned her five children in 2001 while experiencing severe postpartum psychosis that had gone inadequately treated, in part due to her family’s religious environment discouraging psychiatric intervention — illustrates what untreated serious mental illness can produce. Prayer, exorcism, and spiritual counsel have been substituted for medication and therapy, sometimes with fatal results.

Religious opposition to assisted dying forces people to endure suffering they would choose to end. In states and countries without legal end-of-life options, terminally ill people die in prolonged pain because religious objections have prevented the legal frameworks that would allow them to choose otherwise. This is not a matter of theological disagreement. It is the imposition of one group’s beliefs about suffering and death onto people who do not share those beliefs, at the moment of their greatest vulnerability.

Imposing Belief on Everyone Else

The political harm of religion is distinct from the epistemological harms, though it flows from them. It is not about bad reasoning producing bad outcomes. It is about one group using institutional religious authority to restrict the rights of people who do not share their beliefs — including people who actively reject them.

Reproductive rights have been the most visible arena. The overturning of Roe v. Wade was accomplished by a legal movement explicitly rooted in religious conviction about when life begins — a theological position, not a scientific one, and one that a significant portion of the American public does not share. The result is that people who do not hold that theological position are now subject to laws derived from it, with documented fatal consequences. Amber Nicole Thurman, 28, died in Georgia on August 19, 2022 after doctors delayed the care she needed because of the state’s abortion restrictions — afraid to act until her condition became legally unambiguous. Josseli Barnica died in Texas after waiting forty hours for miscarriage care while doctors repeatedly checked for a fetal heartbeat but would not intervene. These are not abstractions. They are named women with families, killed by the application of theology to medicine. Ironically, the Bible is actually silent on abortion — except for one instance in Numbers 5, where a priestly ritual prescribes a potion to terminate a pregnancy as a remedy for suspected infidelity.

LGBTQ+ rights have been systematically restricted through legislation rooted in religious doctrine. Marriage equality was opposed for decades on explicitly religious grounds. Conversion therapy — a practice that produces documented psychological harm, particularly in minors — has been defended and in some jurisdictions protected on religious liberty grounds. Employment and housing discrimination against LGBTQ+ people has been defended as religious freedom.

Contraception access, end-of-life options, and the content of public school curricula have all been shaped by religious lobbying that seeks to impose sectarian beliefs on a pluralistic society. The harm here is not merely practical. It is the subordination of the principle that people should be free to live according to their own values — a principle that is the foundation of a free society — to the theological preferences of a particular religious community.

When the Data Speaks

The argument that religion is harmful is not made only from the evidence of what religious teachings say and what religious institutions have done. It is also made by the data on how societies organized around secular principles compare to those organized around religious ones — and the comparison is stark.

The World Happiness Report, which has ranked national wellbeing annually since 2012, has consistently placed the same countries at the top: Finland, Denmark, Iceland, Norway, Sweden. These are among the most secular countries in the world. Church attendance across the Nordic countries is among the lowest on earth. Formal religious affiliation has declined so dramatically that some Nordic churches have begun removing members automatically unless they actively opt in. And yet these countries consistently outperform every other nation not just on reported happiness but on every associated measure: social trust, gender equality, low corruption, democratic quality, Human Development Index scores, felt safety, and income equality. The World Happiness Report’s own analysis concludes that the most prominent explanations for Nordic exceptionalism include quality of institutions, reliable welfare benefits, low corruption, and well-functioning democracy — not religious frameworks.

The inverse relationship holds as well. An analysis using the Social Progress Index found that in every category — basic human needs, wellbeing, and opportunity — the most religious countries consistently did the worst and the least religious countries outperformed them. A similar analysis within the United States found a negative correlation between state-level religiosity and the Human Development Index when adjusted for income inequality. A study spanning three time periods between 2000 and 2015 found that higher national levels of religiosity were consistently associated with lower national levels of science and mathematics achievement.

The pattern is consistent enough to constitute evidence, not coincidence. The societies that have most thoroughly replaced religious frameworks with secular, evidence-based institutions — in governance, education, healthcare, and social welfare — are measurably better places to live by every outcome that matters. The argument that religion is harmful is not made only from scripture and history. It is made by the data on how people actually live when religion recedes and secular institutions take its place.

The Comfort and the Harm Are the Same Thing

Return to where we began. You will be told that the harm can be attributed to organized religion, not to the teachings. You will be told that the Bible’s troubling passages must be read in context. You will be told that true Christianity is about love.

The teachings themselves endorse slavery, in their own words. The teachings themselves command the mistreatment of women, in their own words. The teachings themselves define faith as the suspension of evidence-based reasoning and call that suspension virtuous. The teachings themselves place humans at the apex of creation in a way that has cost the natural world and generated resistance to the science that most accurately describes our place in it. These are not institutional corruptions of pure doctrine.

They are the doctrine.

A system that structurally undermines the reasoning capacity of its adherents is not made harmless by the warmth of its community or the sincerity of its practice. A book that can be used to justify slavery as readily as it can be used to justify loving your neighbor is not a reliable moral guide. A tradition that asks people to treat the suppression of doubt as a spiritual virtue is not neutral in its effects on how those people navigate a world that requires clear thinking to survive.

Amber Nicole Thurman was 28 years old when she died in a Georgia hospital, septic, waiting for doctors too afraid of prosecution to perform the procedure that would have saved her life. In the same hospital system, in the same state, in the same tradition, she would have been told that God loves her. That he has a plan for her. That her suffering has meaning. That she is not alone. This is the insidiousness that makes religion not merely harmful but uniquely so: it administers the poison and the comfort with the same hand. The faith that held her as a child is the same faith that led to her death. The God who was supposed to be with her in her darkest moment was, in practice, the reason the doctors would not act. The warmth is not separate from the harm. It is how the harm is delivered.

In the year 1077, the Holy Roman Emperor Henry IV stood barefoot in the snow outside the castle of Canossa, waiting for Pope Gregory VII to lift his excommunication. The most powerful secular ruler in Europe, humiliated before a churchman, because without the church’s sanction his authority to rule was void. This was not a story about one man’s crisis of faith. It was a demonstration of the most complete system of social control in recorded history.

The medieval Catholic Church controlled what people believed, what they were permitted to know, who held political authority, how they married and died, and what happened to their souls in eternity. No single institution before or since has achieved that simultaneity of grip — metaphysical, political, legal, social, and interior, all at once, all reinforcing each other.

This essay is not a history of the medieval church. It is a warning. The mechanisms that made the church’s control so durable — information monopoly, existential stakes, the sanctioning of hatred toward enemies, the suppression of interior dissent — are not historical curiosities. They are being reassembled, in distributed and digitally amplified form, in the present. The system is not yet complete. That is precisely why this is the moment to name it.

The Medieval Model: Centralized Total Control

The medieval church’s power rested on several interlocking mechanisms, each making the others stronger. The first was an information monopoly. The church controlled literacy. Monasteries were the universities, theology was the curriculum, and the clergy were virtually the only people who could read, write, and interpret authoritative texts. To control the texts was to control reality itself — what the cosmos was, what history meant, what God required.

The second mechanism was the fusion of religious and secular authority. Secular rulers did not represent a competing power center — they derived their legitimacy from the church. Henry IV at Canossa was not an anomaly; he was the system working as designed. The king ruled by divine sanction, which the church could grant or withdraw. This meant there was no outside position from which to challenge the church’s authority. Political resistance to the church was simultaneously heresy.

The third mechanism was existential stakes. Secular control can punish the body. The church controlled what happened to your soul after death — an eternity of salvation or damnation that made any earthly consequence feel trivial by comparison. Excommunication was not merely a social sanction; it was, in the sincere belief of the time, a death sentence for the immortal self. This leverage over the interior life, over private fear, was something no army or tax collector could replicate.

The fourth mechanism was the identification and sanctioning of enemies. The church provided not just an authority to obey but a target to hate. Jews, heretics, Muslims, and assorted infidels served as the outgroup against which Christian identity was defined. This was not incidental — it was structural. The enemy cohered the community, redirected anxieties that might otherwise be directed at the institution itself, and provided theological permission for cruelty that would otherwise require justification. The Crusades, the Inquisition, and centuries of antisemitic persecution were not aberrations. They were the enemy mechanism operating at scale.

The fifth mechanism was inner surveillance. The confessional system made the institution present inside the individual’s own mind. The penitent was required to monitor their own thoughts, to identify and report their own deviations from orthodoxy. Heresy was not just a public act — it was an interior state to be ferreted out and confessed. The church colonized the self.

Who benefited from all of this? The answer is not mysterious. The papacy, the bishoprics, and the monastic orders accumulated enormous wealth and land. The secular rulers who cooperated with the church maintained their thrones with divine backing. The merchant and peasant classes paid taxes to both. Women were systematically excluded from institutional power. The benefits flowed visibly upward through a hierarchy that did not require conspiracy to maintain — only structure.

The Current Model: Distributed Control

The current system has no Pope at its head. It has no single address, no hierarchy with a visible apex, no institution you can identify and resist. This is not evidence of its absence. It is the feature that makes it harder to resist than the medieval model.

The information environment has been restructured by social media platforms whose algorithms optimize for engagement regardless of truth. Outrage travels further and faster than accuracy. Misinformation that confirms existing beliefs spreads through networks that were built to monetize attention, not to inform. The platforms are not neutral infrastructure — they are platforms whose design rewards engagement over accuracy, and they are owned by people who benefit from the distortion.

Alongside the platforms, a right-wing media ecosystem has constructed a parallel reality with its own authoritative sources, its own vocabulary, and its own account of what is happening in the world. This ecosystem is not simply biased — it is structurally closed. It provides enough internal reinforcement that its consumers need not encounter contradicting information. The information monopoly does not require controlling all information; it requires making the in-group’s information feel complete and sufficient.

The podcast and influencer ecosystem has added a dimension the medieval church could not have imagined: parasocial authority. Figures like Joe Rogan or Andrew Tate are trusted not as institutional representatives but as authentic individuals — men who speak plainly, who resist the establishment, who tell you what the mainstream won’t. The manosphere, a loose network of content creators whose audience is primarily young men, operates through this mechanism, providing identity, community, grievance, and an enemy, all delivered through the intimacy of earbuds. The authority is real even though the institution is invisible.

QAnon is the clearest demonstration of what the distributed model can produce without any formal structure at all. Beginning as anonymous posts on internet forums, it assembled a complete cosmology: a hidden enemy (satanic elites), a heroic resistance (Q and Trump), an apocalyptic resolution (the Storm), and a community of believers who identified each other through shared knowledge. It had heresy mechanisms — those who questioned the narrative were shills or cowards. It had inner surveillance — believers were expected to “do their own research” and continuously recommit to the framework. It spread globally without a headquarters, a budget, or a founding institution. The medieval church built its epistemic infrastructure over centuries. QAnon built a functional equivalent in years.

This argument does not require secret coordination between actors who have never met, and it is important to be clear about why. A conspiracy theory posits hidden actors working in concert toward a shared goal. What is described here is something different: several separate systems — profit-seeking platforms, ideologically motivated media, opportunistic political actors, and foreign interference operations — each pursuing their own interests, producing overlapping effects when they interact. The platforms amplify outrage because outrage is profitable, not because they are coordinating with political movements. The political movements exploit the platforms because the opportunity exists, not because they control them. The result looks coordinated because the incentives align, not because there is a plan. Follow the money and the power: the beneficiaries are visible and their interests are a matter of public record.

Why Control Cannot Succeed Without Religion

Secular ideologies can control behavior. They can reward compliance and punish dissent. What they cannot do is control the interior — the private self that persists beneath performed conformity. Religion can, because religion provides what no secular system can match: existential stakes.

When your eternal soul is at stake, the calculus of compliance changes entirely. The medieval peasant who might have quietly resented the church’s demands faced not just social consequences for resistance but damnation. This is a leverage point that no tax collector or soldier can replicate. And it produces self-regulation — the believer monitors their own thoughts, confesses their own deviations, polices their own interior. The institution does not have to be everywhere if it has convinced the individual to be its agent inside their own mind.

Religion also provides epistemically closed frameworks. Revealed truth — knowledge that comes from God rather than from human inquiry — is by definition not subject to ordinary standards of evidence and argument. Questioning it is not just intellectual dissent; it is a spiritual failure, a sign of pride or corruption. This makes the belief system self-sealing. Evidence that contradicts the framework can be dismissed not on its merits but on the grounds that the questioner’s motives are impure.

Religion provides pre-built trust networks. A congregation is a community of people who have already established that they share foundational commitments. Information passed through that network carries the endorsement of the shared identity — it is trusted not because it has been evaluated but because of who is passing it. This makes religious communities exceptionally susceptible to coordinated manipulation. Intelligence community analyses of Russian interference operations have documented specific targeting of evangelical communities precisely because of this structural vulnerability.

And religion sanctions the identification of an enemy and legitimizes hatred toward them. This is not incidental to religion’s role in control systems — it is essential. An enemy coheres the community, redirects anxiety and grievance away from the institution, and provides theological permission for cruelty. The medieval church’s antisemitism was not a corruption of its theology; it was a feature of the control mechanism. The current movements carry the same mechanism in updated form.

Christian Nationalism fuses religious and political authority in the most direct way: America is a Christian nation, its laws should reflect Christian values, and dissent from the political program is simultaneously dissent from God. The authority of the state and the authority of God become mutually reinforcing, and the space for secular opposition narrows.

Dominionism and its most organized expression, the New Apostolic Reformation (NAR), provide the explicit theological framework. The NAR holds that Christians are called to take dominion over seven domains of society — government, education, media, arts, business, family, and religion. This is not a metaphor for personal virtue. It is a program for institutional control, stated openly, with strategies and timelines. The medieval church achieved dominion gradually and incompletely over centuries; the NAR is attempting something similar with modern organizational tools.

The Prosperity Gospel operates on a different lever: economic discipline. By framing wealth as divine reward and poverty as spiritual failure, it naturalizes inequality and discourages systemic critique. The poor are not victims of an unjust system; they are people whose faith is insufficient. This theology serves the interests of the wealthy donors who fund the movement with remarkable directness.

Quiverfull operates on the demographic and reproductive lever. Named after a Psalm, it frames large families as spiritual warfare — Christian demographic growth as a strategy for political dominance. The movement reinforces patriarchal control over women’s bodies and reproductive choices, frames that control in theological language, and produces the next generation inside a closed epistemic environment. It is simultaneously about reproduction, education, and the subordination of women — three control mechanisms operating as one.

Parallel Mechanisms Across Both Models

The structural parallels between the medieval model and the current one are not coincidental. They reflect the recurring logic of control systems, which reach for the same mechanisms because those mechanisms work.

Education. The medieval church controlled literacy and learning because controlling what the next generation knows is foundational to controlling what they believe. The monasteries were not incidentally educational institutions — epistemic formation was the point. The contemporary homeschooling movement, particularly in its Quiverfull and Christian Nationalist variants, operates on the same logic. Removing children from public education is not primarily about academic quality. It is about who controls the formation of belief.

The treatment of women. The subordination of women is a consistent feature of both models, and it is not accidental. Control over reproduction is control over the next generation. Control over domestic life is control over the primary site of early belief formation. The medieval church’s theological subordination of women and the contemporary movements’ insistence on patriarchal authority are different expressions of the same structural requirement.

The heresy mechanism. Every control system requires a way to identify and punish those who deviate from the authorized framework. The medieval church had formal processes — inquisitors, trials, defined categories of heresy. The current model has informal equivalents that are no less effective. Political purity tests within a movement function like heresy tribunals — a RINO is someone whose orthodoxy has been found insufficient. Online pile-ons and cancellation campaigns serve the same social function as excommunication, cutting the target off from community and livelihood. The mechanism does not require an institution. It requires only that the community enforce conformity.

Inner surveillance. The confessing penitent, monitoring their own thoughts for sin and reporting them to an institutional representative, was the medieval church’s most sophisticated achievement: making the institution present inside the self. The social media user who self-censors — who considers, before posting, how a statement will be received, whether it can be clipped and taken out of context, what the pile-on will look like — is performing a structurally identical function. The audience has been internalized. The surveillance is conducted by the self, on the self, without any external actor needing to be present.

Individual Experience Under Both Systems

What does it feel like, from inside, to dissent from a totalizing control system? The medieval answer and the contemporary answer share the same essential shape, even if the mechanisms differ.

Medieval dissent carried legible consequences. The heretic knew what they were risking — trial, torture, execution, or the spiritual terror of excommunication. The punishment was public and institutional. This made resistance psychologically possible in a way: you could identify the threat, weigh it against the conviction, and make a choice. The martyrs of the Reformation knew what they were walking into.

Contemporary dissent carries decentralized consequences that are in some ways harder to navigate. The person who speaks truthfully against movement orthodoxy — who corrects a factual error in a political context, who refuses to endorse the consensus reality of their community — faces no single identifiable threat. Instead, they face the coordinated action of people they may have never met: death threats delivered anonymously, home addresses published to invite harassment, employers contacted with fabricated accusations, years of social media harassment that can follow a person across platforms and contexts. There is no inquisitor to appeal to, no institutional process that can be challenged. The punishment is distributed and deniable.

The chilling effect is the point. Most people do not need to be directly punished for dissent to be suppressed. They need only to observe what happens to those who dissent. The visible examples — the election worker who required security, the journalist who left public life, the academic who was hounded out of their position — communicate the cost of speaking without anyone having to explicitly threaten every potential speaker. The mechanism is efficient precisely because it is distributed.

A System Still Consolidating

The medieval church’s total control was not established overnight. It consolidated over centuries, absorbing the political structures of the Roman Empire, surviving the chaos of the early medieval period, defeating or co-opting competitors, and gradually building the institutional infrastructure that made the Canossa moment possible. The current system is not yet at Canossa.

Don’t mistake unfinished for unlikely. The trajectory matters more than the current state. The movements examined in this essay — Christian Nationalism, Dominionism, NAR, Quiverfull — are building institutional reach. Their donors are funding judicial appointments, school board elections, and legislative campaigns with a patience and coordination that has few historical precedents in democratic politics. The algorithmic tools are still being refined. The parasocial authority structures are still expanding their reach. The fragmentation of shared reality is still deepening. And unfortunately, law has not caught up to technology. There is no comprehensive regulatory framework for algorithmic amplification, for the data harvesting that enables targeted manipulation, for foreign interference operating through domestic platforms. The infrastructure of distributed control has developed faster than the legal infrastructure that might constrain it.

What the fully developed version looks like is not a mystery. A society in which religious and political authority are fused, in which the information environment is controlled by epistemically closed systems, in which women’s reproductive autonomy is subordinated to demographic and theological imperatives, in which dissent is punished by distributed and deniable harassment, in which the next generation is formed inside closed epistemic environments — that society exists in historical examples and in partial contemporary form in several countries. We can see where this leads.

The window for resistance remains open. This is not a counsel of despair. The current system’s incompleteness is also its vulnerability. But windows do not stay open indefinitely.

What Individuals Can Do

The mechanisms described in this essay operate most effectively on people who have not developed the habit of examining their own beliefs. The primary individual act of resistance is therefore not political — it is epistemic. It is the cultivation of critical thinking as a genuine practice, not as a flattering self-description.

What this means practically: learn to distinguish between a source’s claim and the evidence for that claim. These are not the same thing, and treating them as the same is how misinformation propagates. A headline is not evidence. A viral video is not evidence. An authority figure asserting something is not evidence.

Do not merely avoid information that challenges your beliefs. Learn to evaluate it. The instinct to dismiss uncomfortable information — to find a reason it doesn’t count, to note the source’s bias, to retreat to the comfort of the in-group’s consensus — is the instinct the control systems rely on. Engage with the best version of arguments you disagree with. If you cannot accurately represent the opposing position, you have not evaluated it.

Develop a tolerance for uncertainty. Epistemically closed systems offer the comfort of complete answers. The world is more complicated than any framework can fully capture, and a person who needs immediate, comfortable answers is more susceptible to systems that offer false certainty than someone who has learned to sit with open questions while the evidence is still being gathered.

Pay attention to your own emotional responses to information. Strong emotion — outrage, contempt, the satisfying confirmation of something you already believed — is often a signal that you are being manipulated rather than informed. This is not an argument for emotional detachment. It is an argument for using emotional response as a diagnostic: when something makes you feel certain and righteous, that is the moment to slow down, not to share.

Be aware of how algorithmic systems are shaping what you encounter. The information that reaches you is not a neutral sample of what exists. It has been filtered and amplified by systems optimized for engagement. Deliberately seeking out sources and perspectives outside your algorithmic environment is not a luxury — it is a basic requirement for epistemic health. For those willing to take a more direct step: opt out of social media entirely. You cannot be algorithmically manipulated by a system you do not use. This is easier said than done for many people, and it is worth being honest about why. The platforms have deliberately engineered social dependency — coordinating family events, maintaining community ties, staying connected to friends — because dependency is profitable. Making users feel that participation is socially necessary is not a side effect of the platforms’ design; it is part of it. For people whose social lives run through these platforms, opting out carries real costs that should not be minimized. But for those whose social connections exist outside these systems, opting out is both possible and worth serious consideration. The epistemic benefits are significant: without algorithmic curation, you encounter information on your own terms rather than through a filter designed to maximize your engagement and emotional arousal. The information environment you construct independently — through books, long-form journalism, direct conversation — is simply less distorted than the one the platforms construct for you.

And be willing to bear the social cost of dissent. This is the hardest item on this list. The mechanisms described in this essay are specifically designed to make dissent costly. The person who corrects a falsehood in their social environment, who refuses to perform certainty they do not have, who asks inconvenient questions about the consensus — that person will encounter friction. The friction is the control system working as designed. Bearing it anyway is the individual act that the system is least equipped to handle.

Dismantling the Machinery

The medieval church eventually lost its total grip. The printing press broke the information monopoly. The Reformation fractured the theological consensus. The rise of secular states created competing centers of authority. The process took centuries and was violent and incomplete, but the machinery was dismantled — or rather, was forced to yield enough ground that pluralism became possible.

What dismantled it was not the spontaneous awakening of the population. It was the appearance of competing information sources, competing authority structures, and individuals willing to bear the cost of dissent — Luther nailing theses, printers running presses, scholars reading texts the church would have preferred they not read. The mechanism of change was epistemic before it was political.

The specific theology or ideology of a control system matters less than its mechanisms. Information monopoly, existential stakes, enemy identification, the suppression of interior dissent, the fusion of religious and political authority — these recur because they work. The medieval church perfected them in one form. The current distributed system is perfecting them in another.

The machinery is not new. Recognizing it is the first condition of resisting it.

And the Monarchs are not alone. Since 1970, North America has lost roughly three billion birds — nearly one third of the entire bird population. Most songbirds, even species we think of as seed eaters, depend on insects to feed their young during breeding season. Insects are not a side issue in bird conservation. They are the foundation of it. The decline of insects and the decline of birds are the same story.

So when California gardeners ask what they can do to help, they get a clear and confident answer: plant native plants. Native insects co-evolved with native plants. Native insects feed native birds. Therefore, planting natives is among the most important things a gardener can do. And whatever you do, don’t plant tropical milkweed — Asclepias curassavica — because it will confuse the Monarchs’ migration patterns, promote disease, and do more harm than good.

The advice feels actionable. It gives concerned people something concrete to do. And it is, at best, a distraction from the actual problem.

The western Monarch population has been declining for decades. The data documenting this decline comes from one of the longest running insect monitoring studies in North America — Dr. Art Shapiro’s continuous butterfly population surveys across Central California, running since 1972, visiting ten fixed sites at approximately two-week intervals for over fifty years. What that data shows is sobering: the decline is long-term, consistent, and driven by the accumulated weight of land use change, agricultural pesticide use, habitat fragmentation, and climate change. As one analysis of the data noted, no single recent event can be blamed — it is the whole of environmental degradation, incrementally accumulating over a century, that is responsible.

The tropical milkweed in your garden did not cause this. Removing it will not fix it.

There is an old saying about rearranging deck chairs on the Titanic — the image of passengers busily tidying the furniture while the ship goes down around them. This is what that looks like in conservation: directing genuine concern and real effort toward something visible and controllable, while the actual drivers of decline continue undisturbed. The ship is sinking. We are arguing about the furniture.

The Science Is More Complicated Than the Advice

The claim that tropical milkweed disrupts Monarch migration is not invented — there is research suggesting that exposure to it can interfere with the hormonal state that triggers migration and encourage year-round breeding in coastal California. The Xerces Society, a respected conservation organization, has promoted the advice to remove it based on this concern, also citing elevated rates of a parasitic infection called OE in Monarchs that breed on persistent non-native milkweed.

But the science is considerably more contested than the confident advice given to gardeners suggests.

Dr. David James, a longtime butterfly researcher, has argued that the impact of tropical milkweed on Monarchs has been “vastly overblown,” and that the presence of milkweed — native or non-native — is not a primary cue for the suspension of migratory behavior. A 2021 study found that Monarchs developing on non-native milkweed in urban San Francisco Bay Area gardens showed evidence of dispersing northward in spring, suggesting they may rejoin the migratory cohort rather than becoming permanent residents. That finding has itself been contested. The science, in short, is unsettled — which is precisely what you would not know from the certainty with which the advice is delivered.

What is not contested is what is actually driving the long-term decline. Habitat destruction. Pesticide use. Climate change. The milkweed species that Monarchs depend on across their range have themselves been declining for decades — not because gardeners planted the wrong variety, but because of agricultural practices and land use change operating at a scale no backyard can address. The same forces are driving the loss of the insects that feed the birds people love. The problem is not in our gardens. It is in our fields, our policies, and our food systems.

None of this means gardeners should not plant milkweed. It means the confident prescription about which milkweed to plant, delivered as if it were settled science, is doing something other than what it claims. It is giving people the feeling of action in the place of action itself. It is substituting performance for progress.

When Activism Outpaces Evidence

The native plant movement in the United States has no more influential figure than Doug Tallamy, professor of Entomology and Wildlife Ecology at the University of Delaware. His 2007 book Bringing Nature Home and its successors have inspired millions of gardeners to rethink their plant choices, and his advocacy has made him the closest thing the movement has to a scientific authority. Tallamy’s argument connects directly to bird conservation — native insects feed native birds, native plants support native insects, therefore planting natives directly supports bird populations. It is a chain of reasoning that sounds compelling.

But Tallamy is not primarily a scientist in the traditional sense. He is an advocate — a passionate, effective, and influential advocate for a particular vision of what gardens and landscapes should look like. His research has been conducted in service of that vision, and his public role has been to build and sustain a movement. These are not disqualifying attributes, but they are relevant ones when evaluating the certainty with which his claims are repeated.

There is a scientist who has been studying exactly this question in California for over fifty years, and he reached different conclusions — not as an advocate, but as a researcher whose only agenda was following the data wherever it led.

Dr. Arthur Shapiro is Distinguished Professor Emeritus of Evolution and Ecology at UC Davis. Since 1972, he has walked transects across ten fixed sites spanning central California — from the Sacramento River delta through the Sierra Nevada to the high desert — at approximately two-week intervals, every year, for over half a century. He did not set out to support or refute any particular conservation ideology. He set out to understand what was actually happening to butterfly populations in California, and he has been recording the results ever since. His dataset is one of the longest running insect monitoring studies in North America, used by researchers across the continent to understand butterfly population trends, climate change impacts, and insect-plant relationships.

When a prominent native plant activist named Jake Sigg published the claim that “90 percent of insects can only eat the native plant species with which they co-evolved,” Shapiro’s response was unambiguous: “No, I didn’t know 90% of insects can only eat the native plants with which they’ve co-evolved. I’ve only been studying insect-plant relationships and teaching about them for 50 years and that’s news to me, especially since on a global basis we don’t know what the vast majority of insects eat, period.”

This is not a minor quibble. The 90% claim is foundational to the native plant movement’s argument for why non-native plants are ecologically worthless. It is stated as fact in garden talks, conservation newsletters, and planting guides across the country. It has no data behind it. Shapiro, who has more relevant data than almost anyone alive, had never encountered it in fifty years of research.

Shapiro’s own data tells a different story than Tallamy’s framework predicts. Of the 82 butterfly species in California, 26 — nearly one third — are now using non-native plant species as hosts, and many of these are not generalists casually sampling available plants. They have made specific transitions to specific non-native hosts, demonstrating exactly the kind of adaptability the native-only framework denies. In California, low-elevation butterflies are heavily dependent on non-native plants. Tallamy’s framework, developed from research conducted in the mid-Atlantic United States, simply does not translate to California ecology — and Shapiro has been documenting that mismatch for decades, walking the same routes, recording the same data, following the evidence.

When confronted with Shapiro’s findings, Tallamy dismissed them as “anecdotal.” This is a remarkable response. Calling fifty years of systematic data collected at fixed sites on a regular schedule “anecdotal” is not a scientific argument. It is what an advocate does when the data threatens the framework — reach for a dismissal that sounds methodological without engaging with the actual evidence.

The problem runs deeper than the Shapiro dispute. Tallamy’s first study — the research that launched his career as the movement’s leading scientific voice — disproved his own hypothesis. The study found no significant preference by insects for native over non-native plants. A scientist following the data would have revised the hypothesis. Instead, the movement was already building, and subsequent studies were designed and interpreted to support the conclusion the first study had failed to reach.

This is the difference between science and advocacy. Science follows the data. Advocacy finds the data that supports the conclusion already reached. Tallamy has done important work in raising awareness about the relationship between plants and insects. But the certainty with which his framework has been adopted — and the certainty with which it is transmitted to gardeners — has outpaced the evidence that actually supports it. Shapiro’s fifty years of data in California suggests the framework is, at minimum, considerably more complicated than its advocates acknowledge.

A Framework Built on Shaky Foundations

Even setting aside the conflict with Shapiro’s data, Tallamy’s framework has fundamental problems that his advocates rarely examine.

The first is the baseline problem. The native plant movement operates on the assumption that there is a correct, pre-disturbance ecosystem that we should be trying to restore or approximate. In California, this baseline is generally understood to mean the ecosystem that existed before European contact. But ecosystems have never been static. Plants and animals have always moved, gone extinct, colonized new territory, and adapted to changing conditions. The California that existed before 1769 was itself the product of thousands of years of change, including significant modification by indigenous peoples through burning and cultivation practices. The choice to treat that particular snapshot in time as the ecological standard is a cultural and aesthetic preference, not a scientific one.

The second is the microclimate problem, which is especially relevant on the Central Coast. “California native” is a category that encompasses an enormous range of climates, from the wet north coast to the Mojave Desert. A plant native to the redwood belt is not native to a Mediterranean coastal climate. Pacific Coast Iris is widely promoted as a California native worth planting in gardens across the state, but it evolved in wetter conditions than the Central Coast provides, which is why gardeners here reliably struggle to grow it. The label “California native” is doing far more work than the biology can support.

The third problem is that the biodiversity argument defeats itself. The central claim of the native plant movement is that biodiversity is valuable and worth protecting. This is correct. But if biodiversity is intrinsically valuable, then a garden containing both native and naturalized non-native plants — supporting a wider range of plant-insect interactions, and therefore a wider range of food sources for birds, than a garden restricted to natives — should represent more biodiversity, not less. The movement’s answer to this is that only co-evolved relationships count, which is precisely the claim that Shapiro’s data calls into question.

The fourth problem is the European honeybee. No insect is more important to California agriculture and food production than Apis mellifera — and it is not native to North America. It was introduced from Europe in the 17th century and is now so thoroughly integrated into the ecosystem, and so essential to the pollination of crops that feed millions of people, that no serious conservationist proposes removing it. The native plant framework has no principled answer to the European honeybee. Its presence is simply ignored because the human welfare consequences of applying the framework consistently are too obvious to contemplate.

There is also the matter of the Gulf Fritillary. This spectacular tropical butterfly, with its vivid orange wings, has no native Passionflower species to feed on in California — there are none. Yet it has established thriving breeding populations in California gardens, its caterpillars feeding enthusiastically on cultivated Passionflower vines, sometimes in numbers sufficient to completely defoliate the plant. The Gulf Fritillary did not read the native plant guidelines. It found a chemically suitable host and adapted. So did the 26 California butterfly species now using non-native host plants identified in Shapiro’s data. Those butterflies and their caterpillars feed birds. A garden that supports them is doing exactly what bird conservationists should want.

And here is a point that deserves to be made plainly: human welfare is always the actual operating principle behind conservation decisions, whether or not it is acknowledged. We protect what benefits us — economically, aesthetically, culturally. Monarch butterflies receive enormous conservation attention because they are beautiful, charismatic, and culturally significant. Birds receive attention for the same reasons. An obscure beetle dependent on a single plant in a single county receives lip service at best. The native plant movement frames its preferences in ecological language, but the choices about what to prioritize and what to ignore are driven by the same human values and aesthetics as any other conservation movement. Acknowledging this honestly would produce better conservation outcomes than pretending otherwise.

The Tallamy framework asks gardeners to organize their planting choices around preserving hyper-specialized dependencies — essentially to design gardens around evolutionary dead ends. Shapiro’s data suggests this misunderstands what insects actually need, which is not a specific plant frozen in time, but a sufficient range of chemically suitable hosts. What survives and thrives, it turns out, is adaptability — not the rigid co-evolutionary dependencies the framework treats as worth preserving at significant effort and cost.

Garden Choices Aren’t the Problem or the Solution

To understand why individual garden choices are insufficient, it helps to understand what is actually driving insect and bird decline at scale — and what it replaced.

For most of human agricultural history, farming was inherently diverse. Fields were smaller, crops were rotated, hedgerows and field margins provided habitat corridors, and mixed animal and crop farming created the kind of varied landscape that supports dense insect and bird populations. Regenerative agriculture — a growing movement among farmers today — is largely a return to these principles, using modern understanding to design farming systems that work with natural processes rather than against them.

Industrial monoculture is the opposite of this. A monoculture is exactly what it sounds like — a vast acreage planted with a single crop, maintained in chemical uniformity from planting to harvest. Most of the agricultural land in the United States is managed this way. Where there were once diverse landscapes with hedgerows, field margins, wildflower strips, and varied plant communities supporting dense insect populations, there are now horizon-to-horizon fields of corn, soybeans, or wheat. These fields are hostile to insect life almost by design — the herbicides that maintain them eliminate the flowering plants insects need, the insecticides kill insects directly, and the monoculture structure eliminates the habitat diversity that supports complex insect communities.

Birds follow insects. Where insect populations collapse, bird populations follow. The three billion birds lost from North America since 1970 did not disappear because of garden plant choices. They disappeared because the agricultural landscape that once supported them was converted to chemical monoculture.

This system exists because it is heavily subsidized. The United States federal government provides billions of dollars annually in support to farmers growing corn, soybeans, and wheat — the raw ingredients of industrial food production. These subsidies make monoculture farming artificially economical, underpinning a food system built on cheap, chemically intensive commodity crops. Without those subsidies, the economics of monoculture farming would look considerably different.

The pesticide picture is more complicated than most people realize. Glyphosate — sold commercially as Roundup — is familiar to many gardeners as a weed killer. What is less widely known is that glyphosate is also used as a pre-harvest desiccant: sprayed directly onto wheat, oats, and other grain crops shortly before harvest to dry the crop uniformly and make mechanical harvesting more efficient. This means that the glyphosate exposure associated with eating conventionally produced bread, pasta, oatmeal, and cereal is not an accidental byproduct of weed management. It is a deliberate application to a food crop. The homeowner who applies glyphosate once or twice a year to eliminate weeds in a gravel pathway, and who is reproached by concerned neighbors for doing so, is applying vanishingly small quantities of the same chemical that is deliberately applied at industrial scale to crops eaten daily by millions of people. The concern is being directed at the wrong place.

The same food system that drives this pesticide use also drives the habitat destruction that is collapsing insect and bird populations. The ultra-processed foods that dominate supermarket shelves are manufactured from those subsidized commodity crops — corn syrup, soy derivatives, refined wheat flour — grown on land that was once diverse habitat. Every purchase of those products is a vote for the system that is doing the actual damage.

What Would Actually Help

If the goal is genuinely to support insect and bird populations, rather than to feel good about garden choices, the evidence points toward a different set of priorities. Here is a concrete list of actions that connect to the actual drivers of decline:

• Reform agricultural subsidies. Advocate for ending or redirecting federal subsidies away from monoculture commodity crops and toward diversified, regenerative farming practices. This is a policy issue, which means it requires political engagement — contacting elected representatives, supporting organizations that lobby for agricultural reform, and voting for candidates who prioritize it.
• Change purchasing habits. Buy food from local farms using diversified and regenerative practices where possible. Reduce or eliminate purchases of ultra-processed foods manufactured from commodity crops. Every dollar spent on industrially produced food supports the system driving insect and bird decline.
• Grow your own food. A vegetable garden, even a small one, reduces demand for industrially produced food and creates diverse habitat that supports insects and the birds that eat them. Permaculture principles — designing productive gardens that mimic diverse natural systems — are particularly effective.
• Support genuine habitat conservation. Donate to and advocate for organizations working to protect and restore actual habitat — not garden plant selection programs, but land conservation, habitat corridor creation, and restoration of agricultural margins and hedgerows.
• Reduce consumption generally. Every consumer product has a supply chain with environmental costs. Reducing consumption reduces pressure on the systems driving habitat loss and pollution.
• Have fewer children. This is the point conservation conversations almost universally avoid, because it is uncomfortable to say directly. The single greatest driver of habitat destruction, agricultural expansion, pesticide use, and climate change is human population growth. Every additional person requires food, shelter, and resources. Every additional person increases the pressure on the ecosystems we are trying to protect. Talking about which Asclepias to plant while avoiding the population question is the most elaborate deck chair rearrangement of all.

None of this is an argument against gardens. Gardens are good. Planting things that feed insects is good. The Gulf Fritillary caterpillars that periodically strip a Passionflower vine bare — native or non-native — are not a problem. They are exactly what a functioning ecosystem looks like. But the idea that individual plant choices are a meaningful response to insect and bird decline at the scale it is actually occurring is a comfortable illusion.

The ship requires more than furniture rearrangement. The insects and the birds need us to engage with the hard problems, not the easy ones.