A friend of mine once apologised to me for a biscuit.
We were in her kitchen, she took one, and she said — genuinely, unprompted, to another adult who had not asked and did not care — “ooh, I shouldn’t.” Then she laughed in that slightly strained way people do, and ate it, and I could see her deciding what this meant about her day.
I’ve thought about that biscuit a lot since. Not because it’s unusual — it’s the most ordinary thing in the world, and I’d bet you’ve heard three versions of it this week — but because of what it reveals about the vocabulary we’ve built around eating. She wasn’t describing a nutritional event. She was confessing.
Somewhere in the last forty years, food acquired a moral dimension, and I think it’s done more damage to people’s health than most of the things we blame.
The vocabulary
Look at the words we use, laid out plainly, and they’re extraordinary.
Food is clean or it isn’t. It’s good or bad. Some of it is sinful, some is a guilty pleasure, some is naughty. We have cheat meals and cheat days, which implies a contract we’re breaking. We earn dinner and we burn off lunch. We’re being good this week. We’ve been naughty at the weekend and need to detox, or get back on track, or be strict.
Every one of those words comes from somewhere else. Sin, confession, purity, cheating, discipline, punishment — that’s the language of religious and legal morality, imported wholesale and applied to a plate of pasta. Nothing in the biology of food requires any of it. Chicken is not virtuous. Cake has no ethical valence. A biscuit is a small amount of flour, fat and sugar that will be metabolised without any reference whatsoever to how you felt about it.
And yet the framework is so completely installed that most people can’t hear it as strange. That’s the tell of a genuinely successful piece of cultural programming.
Where it came from
Partly, this is old. There’s a long tradition, particularly in Anglo-American culture, of treating appetite as a thing to be conquered and thinness as evidence of self-mastery. That’s centuries old and predates any of the modern industry.
But the current version got supercharged by a fairly simple commercial dynamic: guilt sells and satisfaction doesn’t.
A person who feels fine about their eating buys nothing. A person who feels vaguely, chronically wrong about it — never terribly, just enough — is a spectacular customer. They’ll buy the programme, and then the next programme when it stops working, and then the supplement, and then the book about the villain ingredient, and then the app. The industry doesn’t need you to fail dramatically. It needs you to feel slightly guilty, indefinitely.
“Clean eating” is my favourite example, because the term has never had a definition. Not once, in twenty years, has anyone produced a coherent account of which foods are clean and which are dirty that survives contact with actual nutrition science. It’s a purity framework with a health costume on, and its main function is to divide food into categories that let you feel a certain way about yourself depending on which you ate.
Why it backfires
You could argue that if guilt made people eat better, we could live with the framing. It doesn’t. It reliably does the opposite, and there are a couple of reasons.
The first is the restraint-and-rebound pattern, which is one of the more consistent findings in this whole area. Categorise a food as forbidden and you don’t reduce your desire for it — you increase its psychological weight enormously. Then, when you do eat it, you’re not eating a biscuit. You’re breaking a rule. And because the rule is already broken, the day is already ruined, and the sensible response is to eat everything now and start again on Monday.
Anyone who has ever done this knows the specific quality of that thought. It doesn’t feel like a loss of control. It feels almost logical. It’s the direct output of a system that has only two states, on-plan and off-plan, with nothing in between.
The second reason is that guilt is just a poor motivator generally. There’s decent evidence that self-criticism after a lapse predicts worse subsequent behaviour, not better, and that self-compassion predicts recovery. This is deeply counterintuitive to people who believe they need to be hard on themselves to stay in line — but the harshness is usually what produces the cycle, not what interrupts it.
And there’s a third cost that’s harder to measure. The moral framing makes eating anxious, and eating is meant to be one of the reliable pleasures of being alive. Trading that for a permanent low hum of dietary self-assessment is a genuinely bad exchange, whatever it does or doesn’t do to your body.
The “earned it” trap
There’s a specific version of this worth pulling out, because it corrupts exercise as well as food.
It’s the habit of pricing meals in physical activity. I ran this morning so I can have this. That’s a two-hour walk right there. I’ll need to work this off.
It sounds harmless and it does two damaging things at once.
It turns exercise into a punishment — into the cost of eating rather than a thing you’d do anyway. Once training is the invoice for dinner, you’ve made it something you owe, and people are very bad at sustaining things they owe.
And it makes food conditional on having deserved it, which means anyone who hasn’t trained is, in some background way, eating without permission. That’s a genuinely unpleasant thing to install in your own head, and it’s especially rough on rest days, injuries, and illness — the exact times when eating properly matters most.
The arithmetic doesn’t even work. The energy costs of exercise are much smaller than people assume, and the estimates most of us use are wrong by a wide margin. But the arithmetic is beside the point. The framing is the problem.
What good actually looks like
Here’s what I’ve noticed about people who have a genuinely easy relationship with food and are, incidentally, in good shape.
They don’t talk about it much. That’s the loudest signal. Food isn’t a topic requiring commentary or justification. They eat something, and then they eat something else later, and no narration accompanies it.
Their eating is unremarkable most of the time and completely unremarkable the rest of the time. There’s no on-plan and off-plan. There’s just how they eat, which is reasonably sensible by default and occasionally isn’t, and neither state generates a feeling.
They can eat at someone else’s table. This one matters enormously and gets ignored. A way of eating that stops you accepting hospitality, or makes a dinner party into a logistics problem, or requires you to explain yourself to a host, has taken something real from you in exchange for benefits that are almost always marginal.
And they can cook a handful of things. Not elaborately — half a dozen dishes they can make without thinking, using ingredients they usually have. I’m increasingly convinced this is the single most underrated health skill there is. It’s not knowledge of nutrition; almost everyone already knows roughly what’s good for them. It’s the practical capacity to produce something decent on a Tuesday when you’re tired, which is the exact moment every dietary intention in history has collapsed.
If someone asked me for one intervention, it wouldn’t be a plan or an app or a rule. It’d be: learn to make five things properly. It solves more than any of the rest of it.
Changing the language
Practically, this is mostly about noticing and then stopping.
Stop saying “I shouldn’t” out loud. Stop announcing that you’ve been good or bad. Stop describing a meal as a cheat — you’re not cheating anyone, there is no contract. Stop pricing food in exercise. Stop the small performative self-criticism that we do in front of other people, partly out of habit and partly as a strange social offering.
