"On the other hand, it is contended, and in a sense rightly contended, that with men as they are, and society as it is, the dictates of absolute morality are impracticable. Legal control, which involves the infliction of pain, alike on those who are restrained and on those who pay the cost of restraining them, is proved by this fact to be not absolutely moral, seeing that absolute morality is the regulation of conduct in such way that pain shall not be inflicted. Wherefore, if it be admitted that legal control is at present indispensable, it must be admitted that these a priori rules cannot be immediately carried out. And hence it follows that we must adapt our laws and actions to the existing character of mankind—that we must estimate the good or evil resulting from this or that arrangement, and so reach a posteriori a code fitted for the time being. In short, we must fall back on expediency." Spencer then goes on to argue that an advanced penal code is as impossible to an early stage of civilization as is an advanced form of government; a bloody penal code is both a natural product of the time and a needful restraint for the time, and is also the only one which could be carried out by the existing administration.

The aim of morality is life, of absolute morality complete life; society is therefore justified in coercing the criminal who breaks through the conditions of life or constrains us to do so. Coercion is legitimate to the extent of compelling restitution, and preventing a repetition of aggressions; no further. Less bloody systems of punishment, wherever introduced, have borne excellent fruit. It may be deductively shown that the best of all systems must be that best calculated to reform the criminal; too severe punishment, instead of awakening a sense of guilt, prevents the same, begetting a sense of injustice towards the inflicting power, which causes resentment; so that, even if the criminal, on reëntering society, commits no further crime, he is restrained by the lowest of motives—fear. The industrial system applied in prisons must have the best results—counteracting habits of idleness, strengthening self-control, and educating the will.

The principle of freedom, which runs through all Spencer's works, is especially enounced again, in his essay, "The Man versus the State" (1884), in which he combats "the great political superstition" of so-called "paternal government." He says: "Reduced to its lowest terms, every proposal to interfere with citizens' activities further than by enforcing their mutual limitations, is a proposal to improve life by breaking through the fundamental conditions of life."[40]

In "The Data of Ethics" (published 1874), Mr. Spencer assumes a somewhat different standpoint from that of his earlier works bearing on morals. The course of reasoning contained in this book is as follows:—

The doctrine that correlatives imply one another has, for one of its common examples, the relation between the conceptions of whole and part. Beyond the primary truth that no idea of a whole can be framed without a nascent idea of parts constituting it, and that no idea of a part can be framed without a nascent idea of some whole to which it belongs, there is the secondary truth that there can be no correct idea of a part without a correct idea of the correlative whole. Still less, when part and whole are dynamically related, and least of all when the whole is organic, can the part be understood except by comprehension of the whole to which it belongs. This truth holds not only of material but also of immaterial aggregates.

Conduct is a whole and, in a sense, an organic whole, and Ethics, of which it is a part, cannot be understood except through the understanding of the whole of conduct.

A definition of conduct must exclude purposeless actions,—such, for instance, as those of an epileptic in a fit. Hence the definition emerges either: acts adjusted to ends; or, the adjustment of acts to ends; according as we contemplate the formed body of acts, or think of the form alone. And conduct, in its full acceptation, must be taken as comprehending all adjustments of acts to ends, from the simplest to the most complex, whatever their special natures and whether they are considered separately or in their totality.

A large part of conduct is non-ethical, indifferent; this passes, by small degrees and in countless ways, into conduct which is either moral or immoral.

The acts of all living creatures, as acts adjusted to ends, come within the definition of conduct; the conduct of the higher animals as compared with that of man, and of the lower animals as compared with the higher, differs mainly in that the adjustments of acts to ends is relatively simple and relatively incomplete. And as in other cases, so in this case, we must interpret the more developed by the less developed; human conduct as a part of the whole of the conduct of animate beings. And further: as, in order to understand the part of human conduct with which Ethics is concerned, we must study it as a part of human conduct as a whole, and in order to understand human conduct, we must again study it as a part of the whole of conduct exhibited in animate beings, so, in order to comprehend this too, we must regard it as an outcome of former, less developed conduct, out of which it has arisen. Our first step must be to study the evolution of conduct.

Morphology deals with physical structure, physiology with the processes carried on in the body. But we enter on the subject of conduct when we begin to study such combinations among the actions of sensory and motor-organs as are externally manifested.