iment on a particular point of
conduct has been consciously raised, and the change is fully realised,
it seldom happens that it afterwards recedes, though the automatic or
semi-conscious adaptations of society to new needs and circumstances,
when regarded from a more general point of view, are not infrequently
found to be regressive as well as progressive. Thus, though we may
imagine the distinctions between the different classes of society
becoming more numerous or more accentuated (as I believe to have
actually occurred in England during the present century), or the evasion
of taxation becoming more general than it at present is, we can hardly
conceive a recurrence to slavery, or a needless increase in the severity
of punishments, or a revival of the hard-drinking habits of the last
century. When society is fully aware of its moral gains, it is not
likely knowingly to surrender them. Hence, allowing for occasional
oscillations and for possible exceptions in certain departments of
conduct, morality, as a whole, almost necessarily advances with the
general progress of intelligence.

It is not altogether easy to adjust the respective claims of society at
large and of the individual thinker in the constitution of moral theory,
or, in other words, to determine the limits within which the speculative
moralist may legitimately endeavour to reform the existing moral
sentiment. It is plain that it must be open to the moralist, and, in
fact, to every intelligent citizen, to criticize the current morality,
or else moral progress, even if it took place at all, would, on many
points of conduct, be exceedingly slow. But, on the other hand, it is
equally plain that a constant discussion of the accepted rules of
conduct would weaken the moral sentiment, lessen the sense of
obligation, and suggest a general uncertainty as to the validity of the
maxims which, in their relations to one another, men usually take for
granted. Hence, though it would be almost fatal to moral progress to
discourage speculation on moral topics, the moralist must always bear in
mind that his task is one which is not lightly to be undertaken, and
that, with an exception to be noticed presently, the presumption should
always be in favour of existing rules of conduct. If for no other
reason, this presumption ought to be made on the practical ground that a
disturbance of the moral sentiment on one point is likely to weaken its
force generally, and, before we expos