He seeks to show not only that the difficulties in the systems of natural and revealed religion have counterparts in nature, but also that the facts of nature, far from being adverse to the principles of religion, are a distinct ground for inferring their probable truth. He endeavours to show that the balance of probability is entirely in favour of the scheme of religion, that this probability is the natural conclusion from an inspection of nature, and that, as religion is a matter of practice, we are bound to adopt the course of action which is even probably the right one. If, we may imagine him saying, the precepts of religion are entirely analogous in their partial obscurity and apparent difficulty to the ordinary course of nature disclosed to us by experience, then it is credible that these precepts are true; not only can no objections be drawn against them from experience, but the balance of probability is in their favour. This mode of reasoning from what is known of nature to the probable truth of what is contained in religion is the celebrated method of analogy.
Although Butler's work is peculiarly one of those which ought not to be exhibited in outline, for its strength lies in the organic completeness with which the details are wrought into the whole argument, yet a summary of his results will throw more light on the method than any description can.
Keeping clearly in view his premises - the existence of God and the limited nature of knowledge - Butler begins by inquiring into the fundamental pre-requisite of all natural religion - the immortality of the soul. Evidently the stress of the whole question is here. Were man not immortal, religion would be of little value. Now, Butler does not attempt to prove the truth of the doctrine; that proof comes from another quarter. The only questions he asks are - Does experience forbid us to admit immortality as a possibility? Does experience furnish any probable reason for inferring that immortality is a fact? To the first of these a negative, to the second an affirmative answer is returned. All the analogies of our life here lead us to conclude that we shall continue to live after death; and neither from experience nor from the reason of the thing can any argument against the possibility of this be drawn. Immortality, then, is not unreasonable; it is probable. If, he continues, we are to live after death, it is of importance for us to consider on what our future state may depend; for we may be either happy or miserable.
Now, whatever speculation may say as to God's purpose being necessarily universal benevolence, experience plainly shows us that our present happiness and misery depend upon our conduct, and are not distributed indiscriminately. Therefore no argument can be brought from experience against the possibility of our future happiness and misery likewise depending upon conduct. The whole analogy of nature is in favour of such a dispensation; it is therefore reasonable or probable. Further, we are not only under a government in which actions considered simply as such are rewarded and punished, but it is known from experience that virtue and vice are followed by their natural consequents - happiness and misery. And though the distribution of these rewards is not perfect, all hindrances are plainly temporary or accidental. It may therefore be concluded that the balance of probability is in favour of God's government in general being a moral scheme, where virtue and vice are respectively rewarded and punished. It need not be objected to the justice of this arrangement that men are sorely tempted, and may very easily be brought to neglect that on which their future welfare depends, for the very same holds good in nature.
Experience shows man to be in a state of trial so far as regards the present; it cannot, therefore, be unreasonable to suppose that we are in a similar state as regards the future. Finally, it can surely never be advanced as an argument against the truth of religion that there are many things in it which we do not comprehend, when experience exhibits to us such a copious stock of incomprehensibilities in the ordinary course and constitution of nature.
It cannot have escaped observation, that in the foregoing course of argument the conclusion is invariably from experience of the present order of things to the reasonableness or probability of some other system - of a future state. The inference in all cases passes beyond the field of experience; that it does so may be and has been advanced as a conclusive objection against it. See for example a passage in Hume, Works (ed. 1854), iv. 161-162, cf. p. 160, which says, in short, that no argument from experience can ever carry us beyond experience itself. However well grounded this reasoning may be, it altogether misses the point at which Butler aimed, and is indeed a misconception of the nature of analogical argument. Butler never attempts to prove that a future life regulated according to the requirements of ethical law is a reality; he only desires to show that the conception of such a life is not irreconcilable with what we know of the course of nature, and that consequently it is not unreasonable to suppose that there is such a life.
Hume readily grants this much, though he hints at a formidable difficulty which the plan of the Analogy prevented Butler from facing, the proof of the existence of God. Butler seems willing to rest satisfied with his opponents' admission that the being of God is proved by reason, but it would be hard to discover how, upon his own conception of the nature and limits of reason, such a proof could ever be given. It has been said that it is no flaw in Butler's argument that he has left atheism as a possible mode of viewing the universe, because his work was not directed against the atheists. It is, however, in some degree a defect; for his defence of religion against the deists rests on a view of reason which would for ever preclude a demonstrative proof of God's existence.