If we compare these two definitions we see that here the same word has been used for two distinct subjects, the one of which includes the other. Without here going deeper into the intricacies of these definitions, I will, for the purpose of this discussion, accept the broader of the two.
The question what instruction in bibliography should contain is already answered in the above definition itself.
The study of bibliographical systems for classification of books presupposes the study of the theoretical systems of classification of knowledge and this presupposes the study of the history of the sciences.
Bibliographical rules govern the practical art of book description, what is technically known as cataloging. There are various codes of rules, more or less arbitrary, as they are more or less the outcome of a compromise. But under all arbitrariness one will discern some underlying theory as to what a description of a book should contain. Such theories are founded on the practice of printing and publishing: thus the intelligent study of bibliographical rules presupposes the study of the history of printing and publishing.
Bibliographical repertories contain the systematic records of printed documents and the study of these repertories is what is called bibliography in the narrow sense. While the branches of study previously referred to may by some be regarded as of less value to the librarian there is surely none who will deny the necessity of his being thoroughly familiar with the literature of bibliographical repertories. However, I do not think that I am alone in the contention that all the different branches of bibliography in the broader sense are of the utmost importance to the librarian.
Dr. Dziatzko has pointed out that in such an eminently practical occupation as that of the librarian it is particularly important not to neglect altogether some kind of theoretical studies. There can be no studies of greater importance to the librarian than those just enumerated, namely, history of literature—the word taken in its broadest sense—history of the book in all its phases, and the study of bibliographical literature.
The library schools have done much to encourage the professional spirit of librarians and to develop the technical side of their work. It is, however, felt that something more is needed, something that a professional school or a training class cannot give, namely, solid bibliographical scholarship. This can, in my opinion, not be acquired except at a university with a faculty of specialists and an extensive equipment of bibliographical literature as a part of a large university library.
A post-graduate school of bibliography, such as I have in mind should offer instruction to two classes of students. The one class would be students in the other branches of instruction who would select as a minor one of the subjects offered by the school, and who should be required to pursue in the school the bibliographical study of their main topics and the preparation of the bibliographies that should be required as a necessary accompaniment to every dissertation. The other class would consist of persons wishing to prepare themselves for the professional work of the librarian and bibliographer. They would choose as their majors the studies offered at this school, and could choose as a minor any other scholastic subject. It would be of great importance to the would-be librarian, could he, while pursuing his special studies, be allowed to do university work in some other subject of his choice, such as literary history, philosophy, American history, mathematics, or the like.
As thorough bibliographical knowledge is the foundation for the work of the librarian, the central subject of instruction in the school should be the study of bibliographical repertories and of the record literature. This study should include seminar work in the handling of literary tools, in hunting up references on special questions, and in the preparation of bibliographical lists. This leads to the study of bibliographical methods. The principles of book description should be discussed, the leading codes of rules studied comparatively, their merits and defects discussed, but none should be taught as the one to be absolutely followed.