In approaching the actual study of primitive libraries it is necessary to touch a little on definition and method. Both these matters, essential to the approach of any topic scientifically, doubly need some attention at this point, because library history has heretofore not troubled itself much about primitive libraries at all or indeed about libraries for the first two thousand years after they had left their more primitive stages. The very method, therefore, lies chiefly outside the experience of library history, being gathered mainly from primitive art and anthropology, and definition must needs consider what the essential nature of these primitive libraries is that links them with the great libraries of modern times. Discussion of definition is the more necessary in that the already contradictory usage has been still farther confused in the matter of the earlier historical libraries by those who, wishing to distinguish the collection of purely business records, public or private, from the collection of purely literary works by calling the former an archive, have yet applied the term archive, incorrectly, under their own definition, to mixed collections of business and other records.

Many answers have been given to this question: What is a library? All of these imply a book or books, a place of keeping and somebody to do the keeping—books, building and librarian—but some definitions emphasize the books, some the place and some the keeping. Far the commonest words used have been the Greek bibliotheke and the Latin libraria and their derivatives. The one rather emphasizes the place and the other the books but both were used sometimes for both library and bookshop. When modern languages succeeded to the Latin the Romance languages kept bibliotheca for library and libraria for bookshop. Germanic languages on the other hand kept both words for library, although in the course of time German has nearly dropped librerei for bibliothek, and English has quite deserted bibliotheke for library. Both English and German call “book shop”, or “book business”, what French, Italian and Spanish call “library”.

Library is thus the common modern word in English for a certain something which the German calls Bibliothek, the Frenchman bibliothÈque and the Italian, Spaniard, Scandinavian and Slav call by some similar name. This something in its last analysis is a book or books kept for use rather than kept for sale or for the paper mill. A library is thus a book or books kept for use.

Among the many definitions of the library which do not recognize use as the library’s chief distinction, the commonest are perhaps those which adopt plurality or collection as the distinguishing factor. Many however adopt the building as chief factor. Typically, of course, the modern library does include many books, a whole separate building and a librarian, but even if the books are few, the place only a room, a chest, a bookcase, or a single shelf, and even if it is only the owner who is at the same time the keeper, it is still recognized to be a library if the books are kept for use and not for sale. Quantity does not matter: the point which divides is the matter of use or sale. Even a one book library is, in fact, a library just as much as a one cell plant is plant or a one cell animal is animal. A one book library is a very insignificant affair compared with the New York Public Library with its many books and many branches, but it is just as truly a library—or else you must find some other word. In point of fact “library” in English, or some derivative of bibliotheca in most other languages, is the word which in practice stands to the book-for-use as the word animal or plant does in biology for the living thing whether it is a single cell or a cell complex.

Some definitions again try to limit the library to printed books or bound books or literary works as distinguished from official or business documents, and these definitions have, as before said, sometimes led to a good deal of misunderstanding. Even if “archive” is assumed to be the right name for a collection of business documents, still such a collection is simply one kind of a library. Every one recognizes this when the collection of business documents is one of printed and bound public documents (U. S. public documents e.g.), and if the documents are tablets, rolls or folded documents, the case does not differ. If books are kept for use it makes no difference whether they are of wood, stone, metal, clay, vellum, or paper, whether they are folded documents, rolls or codexes, whether they are literary works, government or business documents: if intended for use they form a something for which some word must be found which will apply equally to all kinds of records for use and to a one-book-for-use library as well as to the New York Public Library. The right word in the English language seems to be this word “library”. The “business documents” in active current use in the registry or the counting house are perhaps the farthest away from the “library” of common speech but they are equally far from “archives” in the scientific sense, and curiously these have retained one of the very simplest and oldest names of the true library, “the books”, and of librarianship “book keeping”.

But the definition of a library as a book or books kept for use only brings us up against the farther question, What is a book? To this it may be answered that a book is any record of thought in words. Here again neither size, form, nor material matters; even a one word record may be a book and that book a library. This leads again however to still another question: What is a word? Without stopping to elaborate or to discuss definitions in detail, we may take the next step and define a word as “any sign for any thing”, and again explain the sign as anything which points to something other than itself. This is not an arbitrary definition but one founded in modern psychology and philology and to be found in sundry stout volumes by Marty, Leroy, Wundt, Dittrich, van Ginneken, Gabelentz, and others. The sign may be a sound, a color, a gesture, a mark or an object. In some stenographic systems a single dot stands for a whole word.

The most insignificant object, therefore, kept to suggest something not itself may be a library. A single word book is of course a very insignificant book indeed, and the single letter, single word, single book library a still more insignificant library, but, unless you invent other words for them, they are truly book and library, and there is no more reason to invent another word for book or library in this case, than another word for animal when it is intended to include both the amoeba and man. The very simplest library consists therefore of a single recorded sign kept for use. It is the feeble faint beginning of a library but just as much a library as the New York Public Library, the Library of Congress, the British Museum, or the BibliothÈque Nationale—and the beginning of library wisdom is to seek out diligently the nature of these rudimentary libraries.