till be found to be a
factor in the government, if only we can sweep the whole field and
measure the caste in its true degree of power, direct or represented, in
its potentiality of harm to the higher castes, and in its identification
with them for some important purposes, however deeply hidden from
ordinary view. No slaves, not the worst abused of all, but help to form
the government. They are an interest group within it.

Tested by the interest groups that function through them, legislatures
are of two general types. First are those which represent one class or
set of classes in the government as opposed to some other class, which
is usually represented in a monarch. Second are those which are not the
exclusive stronghold of one class or set of classes, but are instead the
channel for the functioning of all groupings of the population. The
borders between the two types are of course indistinct, but they
approximate closely to the borders between a society with class
organization and one with classes broken down into freer and more
changeable group interests.

Neither the number of chambers in the legislative body nor the
constitutional relations of the legislature to the executive can serve
to define the two types. The several chambers may represent several
classes, or again the double-chamber system may be in fact merely a
technical division, with the same interests present in both chambers.
The executive may be a class representative, or merely a co-ordinate
organ, dividing with the legislature the labor of providing channels
through which the same lot of manifold interest groups can work.

It lies almost on the surface that a legislature which is a class agency
will produce results in accordance with the class pressure behind it.
Its existence has been established by struggle, and its life is a
continual struggle against the representatives of the opposite class. Of
course there will be an immense deal of argument to be heard on both
sides, and the argument will involve the setting forth of "reasons" in
limitless number. It is indeed because of the advantages (in group
terms, of course) of such argument as a technical means of adjustment
that the legislative bodies survive. Argument under certain conditions
is a greater labor-saver than blows, and in it the group interests more
fully unfold themselves. But beneath all the argument lies the strength.
The arguments go no farther than the strength goes. What the new