Page:Mind (New Series) Volume 6.djvu/272

 256 CRITICAL NOTICES: quite sure of here is, what kind of unity it is that is supposed to- connect the different elements. Such a qualitative unity as holds together, e.g., the different species of triangle seems rather to be suggested. But there are other kinds of unity of which some seem more appropriate to the present case than the merely quali- tative kind just referred to which, whatever its importance, is certainly not directly applicable to every judgment. According to Bergmann we have Immediate Inference when from one judgment or proposition we pass to another having the same (or part of the same) matter (Sachverhalt) but a different way of looking at it (Auffassung). Thus, from All S is P, Some P is S is an Inference but Some S is P is not an Inference from All S is P. But is not this (and cf. other cases) an arbitrary distinction ? Must we not say that wherever there is an Inference of one proposition from another, one of the two propositions is true if the other is true, and there must be some difference between them? But, what difference ? What constitutes them two different propositions ? It seems to me that the only answer is, Any difference for it would appear that -some difference of thought, however slight, must correspond to every difference of expression so that we should get the definition : If any two propositions differ in any respect, and one of them is true if the other is true, the first is an inference from the second. It does not do to say simply that there is Unity in Difference between the two proposi- tions compared. There is also unity in difference between S and P of any Categorical Proposition, and between This pencil and That pencil, and these are only two out of the kinds of unity in difference which are possible. What is it that is really at the bottom of any Inference ? Let us take any case of an affirmative Categorical say All R is (some) Q. The speaker here has in mind an object or group of things having the characteristics signified by both R and Q which may be symbolised by W| or (^ . Q) ; it is the R's that are Q that he is referring to, and the things that are both JR and Q may be indicated as Q's which are R, or E's which are Q thus we justify conversion in the case of affirmatives. And any one of the R's is Q, since all are. It does not appear clear why this passage from A to I should be refused the name of Inference, which (by Bergmann) is allowed to the change called Conversion. There may be quite as truly a change of Auffassung in passing from All R is Q (1) to This R is Q (2), as in passing from (1) to Some Q is R (3) ; and in (3) also we assert concerning only a part of the Sachverhalt referred to in (1). It seems here that Identity of Application (in conjunction of course with the plurality and connectedness of attributes that