It is clearly a precondition of all useful discussion that the author and reader should be in accord with respect to the authority of the generalisations and definitions which supply the premisses for his reasonings.

Though this might perhaps to the reader appear an impractical ideal, I would propose here to attempt to reach it by explaining the logical method which I have set myself to follow.

Although I have from literary necessity employed in my text some of the verbal forms of dogmatism, I am very far from laying claim to any dogmatic authority.  More than that, I would desire categorically to repudiate such a claim.

For I do not conceal from myself that, if I took up such a position, I should wantonly be placing myself at the mercy of my reader.  For he could then, by merely refusing to see in me an authority, bring down the whole edifice of my argument like a house of cards.

Moreover I am not blind to what would happen if, after I claimed to be taken as an authority, the reader was indulgent enough still to go on to read what I have written.

He would in such a case, the moment he encountered a statement with which he disagreed, simply waive me on one side with the words, “So you say.”

And if he should encounter a statement with which he agreed, he would in his wisdom, censure me for neglecting to provide for that proposition a satisfactory logical foundation.

If it is far from my thoughts to claim a right of dictation, it is equally remote from them to take up the position that I have in my arguments furnished proof of the thesis which I set out to establish.

It would be culpable misuse of language to speak in such connexion of proof or disproof.

Proof by testimony, which is available in con-nexion with questions of fact, is unavailable in connexion with general truths; and logical proof is obtainable only in that comparatively narrow sphere where reasoning is based—­as in mathematics—­upon axioms, or—­as in certain really crucial experiments in the mathematic sciences—­upon quasi-axiomatic premisses.

Everywhere else we base our reasonings on premisses which are simply more or less probable; and accordingly the conclusions which we arrive at have in them always an element of insecurity.

It will be clear that in philosophy, in jurisprudence, in political economy and sociology, and in literary criticism and such like, we are dealing not with certainties but with propositions which are, for literary convenience, invested with the garb of certainties.

What kind of logical sanction is it, then, which can attach to reasonings such as are to be set out here?

They have in point of fact the sanction which attaches to reasonings based upon premisses arrived at by the method of diacritical judgment.

It is, I hasten to notify the reader, not the method, but only the name here assigned to it, which is unfamiliar.  As soon as I exhibit it in the working, the reader will identify it as that by which every generalisation and definition ought to be put to the proof.