Toot toot! Hop on the Herbert Spencer cluetrain!
Assuming it to be in other respects satisfactory, a rule, principle, or axiom, is valuable only in so far as the words in which it is expressed have a definite meaning. The terms used must be universally accepted in the same sense, otherwise the proposition will be liable to such various constructions, as to lose all claim to the title—a rule. We must therefore take it for granted that when he announced “the greatest happiness to the greatest number” as the canon of social morality, its originator supposed mankind to be unanimous in their definition of “greatest happiness.”
This was a most unfortunate assumption, for no fact is more palpable than that the standard of happiness is infinitely variable. In all ages—amongst every people—by each class—do we find different notions of it entertained. …
Generalizing such facts, we see that the standard of “greatest happiness” possesses as little fixity as the other exponents of human nature. Between nations the differences of opinion are conspicuous enough. On contrasting the Hebrew patriarchs with their existing descendants, we observe that even in the same race the beau ideal of existence changes. The members of each community disagree upon the question. Neither, if we compare the wishes of the gluttonous school-boy with those of the earth-scorning transcendentalist into whom he may afterwards grow, do we find any constancy in the individual. So we may say, not only that every epoch and every people has its peculiar conceptions of happiness, but that no two men have like conceptions; and further, that in each man the conception is not the same at any two periods of life.
The rationale of this is simple enough. Happiness signifies a gratified state of all the faculties. The gratification of a faculty is produced by its exercise. To be agreeable that exercise must be proportionate to the power of the faculty; if it is insufficient discontent arises, and its excess produces weariness. Hence, to have complete felicity is to have all the faculties exerted in the ratio of their several developments; and an ideal arrangement of circumstances calculated to secure this constitutes the standard of “greatest happiness;” but the minds of no two individuals contain the same combination of elements. Duplicate men are not to be found. There is in each a different balance of desires. Therefore the conditions adapted for the highest enjoyment of one, would not perfectly compass the same end for any other. And consequently the notion of happiness must vary with the disposition and character; that is, must vary indefinitely.
Whereby we are also led to the inevitable conclusion that a true conception of what human life should be, is possible only to the ideal man. We may make approximate estimates, but he only in whom the component feelings exist in their normal proportions is capable of a perfect aspiration. And as the world yet contains none such, it follows that a specific idea of “greatest happiness” is for the present unattainable. It is not then to be wondered at, if Paleys and Benthams make vain attempts at a definition. The question involves one of those mysteries which men are ever trying to penetrate and ever failing. It is the insoluble riddle which Care, Sphinx-like, puts to each new comer, and in default of answer devours him. And as yet there is no Œdipus, nor any sign of one.
It's worth emphasizing that this is not for Spencer an anti-utilitarian argument. Spencer is a utilitarian. But, fascinatingly, Spencer is a pluralist about both the composition of happiness, and about conceptions of the composition of happiness. His own thick conception of happiness—that it is the gratification produced by the maximal exercise of the several faculties enabled by their degrees of development—accomodates variability across persons in the capacity of faculties and their development. In a separate argument, Spencer notes that there may be tradeoffs in the development in faculties (developing one more may require developing another less), and maintains that there is no adequate general principle for determining the relative value of the (evidently qualitatively different) gratification of different faculties. Spencer also notes that even were the nature of happiness unitary, epistemically transparent, and uncontested, individual variation would in any case pose an intractable knowledge problem for a benevolent utilitarian policy czar. The upshot of Spencer's pluralism about happiness is the same as the upshot of pluralism about value in general. The best bet politically is a general, neutral framework of rights that enable harmonious social cooperation in pursuit of one's good, however one conceives it. As far as I can tell from my amateur Spencer scholarship is that this argument is pivotal for Spencer's general view about the congruence of rights and utility.