296 ALFEED SIDGWICK'S FALLACIES. Fallacies : A View of Logic from the Practical Side. By ALFRED SIDGWICK, B.A. Oxon., Berkeley Fellow of the Owens Col- lege, Manchester. London : Kegan Paul, Trench, 1883. Pp. xvi., 375. This is an ingenious and consistent attempt to work out a theory of Logic, or rather of so much of the substance of that science as falls within the writer's defined scope, from a rather original point of view. Eeaders of Mill's Logic will be familiar with a certain phraseology, enounced by him in regard to the import of propositions, about one thing or attribute being " a mark " of another. Mill's own treatise was too extensive, and, as I should express it, too objective in its general design and ap- plication, to make it centre round such a doctrine as this, but it is a possible view, and, for certain practical purposes, a useful one, as illustrated in the volume before us. As the title indicates, the essay is concerned with the detec- tion of Fallacy : " I hold that to combat Fallacy is the raison d'etre of Logic". It assumes that we have before us a de- finite thesis and may demand a reason for the same, and the inquiry is, What sorts of failure can there be in the application of the reason to the thesis? " The work before us is to survey, classify, explain, and illustrate the possible objections which can be brought against any belief, so soon as it is definite enough to take shape in language, and thereby to become a thesis for proof." Such a design involves considerable departure from the methods and phraseology of the common systems. To begin with : in- stead of the threefold division naturally yielded by the two terms of the proposition, and the copula, these two terms themselves presenting as subject and predicate a more or less decidedly re- cognised difference of character, we have a duality only, the mark and the thing marked, both of these presenting the com- paratively substantive character of a subject, and being symbolised accordingly by the letters S and 5>. The things which we may thus term " indicator " and " indicated " are to be understood of course with the widest possible latitude as to complexity, artifi- ciality, and time or place-relation. Common language adopts many and various expressions for this relation, but we can always, with some violence to usage, translate it into the desired form : " the case of death may indicate (or point to) poison, or my pulse at the present moment may indicate (or show) the absence of fever; or yesterday's panic in the city may indicate (or fore- shadow) a future increase of bankruptcy ; but it is undoubtedly clumsy to say that Bavius ' indicates the qualities of ' a fool : we habitually condense those four words into the one word ' is '." In an extended system of Logic, or examination of the prin- ciple of evidence in general, it would become necessary to discuss the foundations limits and ultimate warrant of this indication. This was what was intended in the above allusion to Mill. But in any detailed application of the principle, such as the discussion
