discovery, as in science, originality of creation, as in art, are for most men equally impossible; but originality of compilation, the power to read facts, to analyze, collate, combine them, to give them promptly such direction as shall unfold their significance, according to each writer's view of the significance, is a mastery hardly to be shirked by any educated man. To energize knowledge is the office of persuasion; but to realize knowledge comes first, and that is the office of exposition. In college and in the world the man that can give out with most clearness and largest suggestion of relations what he has taken in shows most grasp. That is the aim of

exposition.

56. Specifically, then, exposition means the succinct and orderly setting forth of some piece of knowledge, whether the knowledge come from one's own observation, from the regular course of his studies, or from the reading of his predilection. It means a luminous report of conditions in track athletics, debating, local or national or foreign politics, rapid transit, the contemporary stage; a concise explanation of Greek tragedy, the turbine wheel, the Roman legion, the feudal system; a criticism of a book, a tendency in art, a salient character or period in history. The range is wide; the aim and use are single. For the aim of all exposition is, without waste of words, clearness; its use is to develop, in analysis and synthesis, grasp.

III. METHOD

a. Theoretical

I. Division

57. Grasp here means power in grouping facts, in seeing their bearings, in expressing their gist. To

take the bearings of a subject is essentially what is known in logic as division, separation into component parts. For scientific treatment and for argument, division must be complete; i.e. must take account of all the parts. Division in this sense is scientific classification. Thus Gray divides surgical anatomy into osteology, articulations, muscles and fasciæ, arteries, veins, nervous system, etc. No conclusions concerning surgical anatomy could be based on a division that failed to take account of all its components. No conclusions concerning the nervous system as a whole could ignore any part of that system. But ordinary exposition, especially in brief essays, does not aim at scientific completeness, nor at scientific conclusions. Its very subjects, being commonly those that bear on human nature or human life, are not readily susceptible of scientific treatment. And even when the subject may be so treated, no one shall say that it must be.

Religious "conversion " has been analyzed by the science of psychology. The statistics thus collected and the resulting division into groups according to temperament cannot be ignored by any subsequent discussion that aims at scientific conclusions. But that division does not preclude another division for another purpose. Wishing to explain that "conversion" is something, whether fact or idea, ingrained in the American mind, I may divide my treatment into: (1) the familiarly recurrent mention of it in talk, (2) its cruder manifestations among negroes, and how far these are essentially like those among educated whites, (3) whether it is usually regarded as a test, and of what, (4) why I never heard of it in France, (5) etc. My essay will not be scientific. Neither will it be unscientific; it will not ignore any phase that seems to bear on my purpose; for my purpose my

division is complete. Science hardly enters; division in this aspect is purely a matter of rhetoric.

Thus the kind of division sought by popular exposition, that is by brief essays for ordinary readers, is such as shall present a subject, or any definite part of a subject, most clearly to a particular audience. This kind of division is complete when it omits nothing that affects its purpose.

58. Two other rules of logic concerning division enter more largely into popular exposition; for they are both safeguards against confusion. The first is that the parts into which a subject is divided must be mutually exclusive, must not overlap. The second is that the division must be according to one principle, that there must be no "cross-division."

The importance of these will appear in two very obvious violations. Suppose I divide disease into (1) diseases of the nervous system, (2) fevers, (3) diseases of the intestines, etc. Here I at once muddle the matter instead of clarifying it; for typhoid might be put under both (2) and (3) and perhaps also under (1). Suppose again I divide the American people into Republicans, Democrats, Socialists, Imperialists, Catholics, etc. Here I start to divide by party, proceed to divide by political theory, and then by religion. No educated man is likely to lapse so far; but many less obvious cases of confusion are due to violation of these same rules.

Logic says, It is prudent to begin, if possible, by a twofold division; i.e. by separating into two mutually exclusive parts, each afterwards to be subdivided: rhetoric says, Scrutinize every division to make sure that it really divides; and remember that the object of exposition is to simplify.

2. Definition

59. The other fundamental process of exposition is definition. Essentially definition is what we all learned of it in geometry; i.e. the giving of genus and differentia, the class to which a thing belongs and the peculiarities that separate it from other members of the same class.

(a) A cube is a (genus) polyedron (differentia) bounded by six equal squares.

(b) A cathedral is a (genus) church (differentia) in which a bishop has his seat.

(c) Faith is (genus) certitude (differentia) with respect to matters in which verification is unattainable.

Long or short, every definition has these two parts. 60. Thus it is plain, especially from example (c), that definition is the result of a process of generalization. After much consideration faith is assigned to the class certitude. So "second-sight" might be relegated by some expositor to the class hallucination, or Kalmucks to the class Tartars; and another expositor might reach a different generalization, might read the reported facts differently. Two considerations follow:

(1) Definition and division go together; it is hard to generalize without classifying, or to classify without generalizing.

(2) Generalizing leads often to argument; for concerning the interpretation of facts there is often difference of opinion. Therefore no one may insist on his generalization without proving it; but anybody, of course, may suggest one.

61. A definition can rarely be left with a single proposition, however luminous. A single proposition

may, usually should, sum up the whole result; but it will not make the result clear by itself. Either before or after this summary, or both before and after, there must be enough development and iteration to impress it clearly on the particular audience. This is the process outlined at § 16 and exemplified in the corresponding section of the Appendix. In a wider sense, then, definition implies almost all the processes of exposition.

b. Practical (Compilation)

1. Taking Notes

62. Division and definition comprise the logic, or theory, of exposition. Practically, luminous division and definition are reached, more often than in any other way, through skilful compilation, compilation that engages, not only the industry of the compiler, but also his intelligence. The object set forth at the beginning of this chapter is to gain grasp in grouping and in interpretation. Whatever writing involves this is composition, and therefore profitable; whatever writing does not involve grouping is beneath the attention of educated men. Practically this means that every exposition drawn from reading must, in order to be worth while, have more than one source of information. An essay drawn from one source is at worst mere reproduction, like the dictation in primary schools; at best it is a digest, summarizing point by point. Summary is good elementary exercise, but it is not composition. How to compose material from several books so that the result shall be original and the composition valuable as practice, is of common, everyday concern; in other words, how to compile.