But, after all, this progress is one towards the normal. Attention to details in the case of the poor man is forced upon him. Except in rare cases, he does not really care to shovel his own snow; he would prefer to hire a man to do it, and as soon as he can he does do so. So long as his sidewalk is properly cleared he is willing to leave the details to the man who clears it. He does not care whether that man begins at the north or the south end, or whether his shovelfuls are small or large. Here, if we examine, we shall find a common characteristic of those kinds of work where laymen are in control--the persons for whom the work is done care very much about results; they are careless of methods so long as those results are attained. And in a very large number of cases the persons for whom the work is done will be found to be the public, or so large a section of it that it is practically a group of laymen so far as the particular work in question may be concerned. A lay board of directors or a lay departmental head, then, is simply and properly a representative of a greater lay body that is particularly anxious for results and not particularly anxious about methods. Lay control is thus not illogical, but is the outcome of a regular and very proper development. But, as has been said, it is not the only method of controlling a great institution. An institution may be managed by a graded body of experts. So were the old guilds of craftsmen managed. So are many ecclesiastical bodies, notably the Roman Catholic Church. We may call this method of control hierarchical. It has some advantages over lay control and some disadvantages. We may imagine such a system applied to libraries. All the libraries in a state, we will say, would then be managed by the state librarian, and all these officers would be subject to the orders of the librarian of the national library, who would be supreme and accountable to no one. Without going into detailed discussion of this extremely supposititious case, we may say that the objection to it would be that the persons who are especially interested in the results of the work done are not represented in the controlling hierarchy. Where the persons interested are all experts, as in a guild of craftsmen, there can perhaps be no objection to control by experts; though even in this case we are leaving out of consideration the persons, generally laymen, for whom the craftsmen do their work.