Money is worth having because of what it does. Health is worth having partly for what it lets you do and partly just to have. Contentment is not worth having for anything — asking what it is for is a confusion.

That is the distinction. Instrumental value is value as a means. Intrinsic value is value as an end. Aristotle opens the Nicomachean Ethics with it, and the reason it survives as more than a classification is the argument he builds on it next.

The regress argument

Take anything you want and ask what you want it for. The answer names something else. Ask again. Either the chain runs forever, which would mean nothing is ever actually worth anything and the whole structure floats; or it loops, which is the same problem in a smaller space; or it stops.

Aristotle argues it must stop, at something wanted for itself and never for anything further. His candidate is eudaimonia — a life going well as a whole, which is the only thing nobody sensibly asks the point of.

You can reject his answer and keep the method, and the method is what makes this practical. The test is running the chain on something specific and seeing where it terminates.

What running it actually finds

Most people, asked to do this honestly, find one of three things.

A chain that ends somewhere they would defend. The work is for the money, the money is for security, the security is for being able to raise children without dread. That terminates, and it terminates somewhere a person will stand behind. Nothing further is needed.

A chain that ends in a comparison. The promotion is for the title, the title is for the standing, the standing is for not being behind. That is not an end, it is a treadmill, and a comparison-based terminus never delivers because the comparison keeps moving. See status games and the hedonic treadmill.

A chain that stops at an unexamined inheritance. The answer is because that is what you do, or because my father did, or nothing at all. That is not an error, necessarily — plenty of inherited ends are good ones — but it is worth knowing that it has never been looked at.

The common failure it catches is means that have quietly become ends. Money is the usual example and not the only one: a career pursued so long for what it would buy that the buying stopped being the point, credentials collected past any use for them, a business built for freedom and run in a way that produces none.

The problem with one final end

Aristotle needs the chain to terminate in a single thing, because his argument requires a unified target for a life to aim at. It is not obvious that it does.

The alternative — value pluralism — holds that there are several things worth wanting for themselves, that they do not reduce to a common currency, and that they can genuinely conflict. Isaiah Berlin's version of this is the influential modern one, and its consequence is unwelcome: if ends are plural and incommensurable, some choices involve a real loss rather than a miscalculation. Choosing the career and the city means a version of the life is gone, and no correct answer would have avoided that.

That is either a defect in the theory or an accurate description of what it is like to decide things. It is probably the second.

Why the distinction is worth the effort

It makes "and then what?" a usable question. Most of what looks like a values problem is a chain that was never run past the second link.

It explains a specific failure that looks like success. Someone who has obtained everything instrumental and finds it thin has not failed at the pursuit; they have completed a chain that did not terminate anywhere.

It is the only correction for optimizing a proxy. Any measurable target stands in for something unmeasured, and pressure on the measure degrades the thing it was standing in for — Goodhart's law at the level of a life. Knowing which of your targets are instruments is what makes it possible to notice when the instrument has drifted from the end.

The practical version is not a philosophy seminar. It is one question, asked about whatever you are currently spending your time on, repeated until it stops.