QuestionsWhat is happiness?Happiness, an ideal of the imagination
What is happiness?No lasting happiness

Happiness, an ideal of the imagination

The indeterminate concept of happiness; "an ideal not of reason but of imagination"

Immanuel Kant 1724 · 1804Set the limits of reason; act from duty.
See profile
What it claims

Happiness is an ideal of imagination, not of reason, and morality cannot rest on it.

Why

Happiness demands an absolute whole, yet its elements come from experience: nobody can say precisely what they want.

The same idea through history
Immanuel Kant1785Groundwork of the Metaphysics of Morals, section II. Kant does not say that happiness is bad nor that being happy is impossible: he says one cannot say precisely what it consists in, hence cannot draw a law from it.
John Stuart Mill1861Mill's utilitarianism takes the opposite road: it defines happiness (pleasure, absence of pain) and draws morality from it. It is the main answer to Kant on this point.
In short

"Happiness, an ideal of the imagination" is an argument by Immanuel Kant (Groundwork of the Metaphysics of Morals (1785), section II (4:417-419, 4:441), 1785) on the question "What is happiness?" (answer: no lasting happiness). Its claim: Happiness is an ideal of imagination, not of reason, and morality cannot rest on it. According to AskPhilosophy's logic check (Z3 solver), the argument is valid only if one grants a premise the author leaves unstated (P5). AskPhilosophy presents 3 objections, 0 of them with a documented reply, and the original text behind each premise, checked word for word.

Groundwork of the Metaphysics of Morals (1785), section II (4:417-419, 4:441)
Valid with an implicit premise

Everyone wants to be happy, says Kant, but no one really knows what that means. Do you want wealth? It brings envy and worries. Knowledge? It may show you more clearly evils you cannot avoid. A long life? It may be a long suffering. Health? It may let you fall into excess. To know what would really make you happy, you would have to know everything about the future. So happiness is not an ideal of reason, but of imagination. Consequence: one can give advice for being happy, not commands. And morality, which must hold for everyone, cannot rest on it.

P1All the elements of the concept of happiness are empirical: they must be drawn from experience.
P2Yet the idea of happiness requires an absolute whole: a maximum of well-being in my present state and in my whole future condition.
P3No finite being, however clear-sighted and powerful, can form a definite concept of what it really wills here. It would need omniscience.
P4So, although every man wishes to be happy, no one can say precisely and consistently what he wishes, and no sure rule for being happy is possible, only advice drawn from experience.
P5 · implicit premiseAn object from which reason cannot draw any sure rule, and which rests only on principles drawn from experience, is not an ideal of reason but of imagination.
P6Principles drawn from experience can never ground a moral law, because a moral law must hold without exception for all rational beings, with a necessity that depends neither on human nature in particular nor on contingent circumstances.
Therefore happiness is an ideal of imagination, not of reason, and morality cannot rest on it.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
Without the implicit premise: Without P5, the computer finds a world where happiness is too indeterminate to give a sure rule, but can still be counted among the ideals of reason (for example because reason could determine it some other way). The premises are true, the conclusion is not.
Valid with P5. P1 to P4 establish that no sure rule for being happy is possible; P5 links this to Kant's phrase ("ideal of imagination"), and P6 to morality. Without P5, we have half the conclusion. The most debated part is P1 to P4: P2 puts a very strong demand on happiness (a maximum, for the whole of life). If it is defined otherwise, more modestly, the proof no longer holds. Note also that P4 does not claim that being happy is impossible, only that it cannot be made a rule.
Analogy. Someone tells you: "Draw me the house of your dreams, with everything you will like in twenty years." You can imagine a beautiful house, but you cannot really draw it, because you do not know who you will be, or what you will lack. Kant says: a picture, not a plan. And you do not build a law on a picture.

Objections

3

Each objection is a conversation: click to see who replies and where the debate stands.

Aristotle 330 BC · attacks premise 2

Aristotle affirms on the contrary that one can say what happiness is. It is the highest end, the one desired for itself and never for the sake of something else, and it is defined by looking for what is proper to man. The good of man is activity of the soul in accordance with virtue. This definition does not depend on a calculation of all the pleasures to come.

no reply
We have not yet found a direct reply by Kant in the texts.
The debate goes on (our summary) In the Groundwork, Kant does not reply to Aristotle on this point. A Kantian would say that Aristotle defines happiness by virtue, that is, by something other than well-being: it is no longer the idea Kant criticises (a maximum satisfaction of inclinations). An Aristotelian would answer that this is precisely the right idea of happiness, and that Kant attacks an impoverished version. Both refuse, at bottom, to make mere pleasure the measure of the good.
John Stuart Mill 1861 · attacks premise 2

For Mill, happiness is pleasure and the absence of pain, and it need not be an absolute maximum. The philosophers who make happiness the aim of life do not mean a life of continuous rapture: they mean an existence in which pains are few and brief, and pleasures many and varied. This idea is precise and attainable, and many people realise it for a good part of their lives.

no reply
No reply possible: objection raised in 1861, after the author's death (1804).
The debate goes on (our summary) Kant died in 1804 and could not answer Mill. Kantians reply that, even defined as "many pleasures, few pains", happiness depends on each person's taste and circumstances: it cannot supply a law valid for all (P6). Mill's defenders reply that utilitarianism seeks not personal happiness but the happiness of all, and that this starting question (what is good for all?) is precisely what a morality must deal with. See the argument "Better to be Socrates dissatisfied".
John Stuart Mill 1861 · attacks premise 3

For Mill, one need not be omniscient to know what leads to happiness. Humanity has had, to learn it, the whole length of its history: it has learned by experience the effects of actions, and on this experience depend prudence as well as the morality of life. He is answering here the objection that there would be no time to calculate before acting, not Kant himself.

no reply
No reply possible: objection raised in 1861, after the author's death (1804).
The debate goes on (our summary) Mill answers a different objection from Kant's: that of time, not of certainty. A Kantian would note that humanity's experience yields general rules, not the certainty a moral law demands (P6), and that this gives back exactly "counsels" (Kant) and not "commands". A defender of Mill would answer that no one asks for absolute certainty, in morals as in medicine, and that solid rules learned by experience are enough.