I. " On the Mechanical Performance of Logical Inference .
" By W. STANLEY JEVONS , M.A. ( Lond. ) , Professor of Logic &amp;c. in Owens College .
Communicated by Professor E. RoscoE .
Received October 16 , 1869 .
( Abstract .
) It is remarkable that from the earliest times mechanical assistance has been employed in mathematical computation .
The Ise of pebbles , of the fingers , and of the abacus of the Greeks and Romans may be adduced as examples .
Mathematicians have constantly delighted in devising mechanical modes of calculations , as in the case of Napier 's bones , mechanical globes , slide rules , &amp;c. Actual machines for performing difficult calculations have been designed or constructed at various times since the early part of the 17th century , by Pascal , Morland , Leibnitz , Gersten , Babbage , and Scheutz .
In logic , on the contrary , we meet with a total absence of any actual mechanism , although logical works abound with expressions implying the need of such aid .
The name of Aristotle 's logical treatises , the ' Organon , ' or Instrument , and many definitions of logic , clearly express this idea , which is also distinctly stated by Bacon in the second aphorism of his 'New Organon .
' This inability of logicians to realize their notions of a mechanical logic in a material form , analogous to the many kinds of calculating machines , can only be explained by the extreme incompleteness of their doctrines , It is the advance of logical science , chiefly due to the late Dr. Boole , Prof. De Morgan , and George Bentham , which now enables us to produce a truly mechanical logic .
Boole , in his celebrated work on the ' Laws of Thought , ' first put forth the problem of logical science in its complete generality:-Given certain logical premises or conditions , to determine the description of any class of objects under those conditions .
The ancient forms of logical deductions are but a few isolated cases of this general problem , which Boole solved in a complete but exceedingly obscure manner .
In my 'Pure Logic ' ( London , 1864 , Stanford ) and my 'Substitution of Similars ' ( London , 1869 , Macmillan ) , I have endeavoured to show that the mysterious mathematical form of Boole 's logical system is altogether superfluous , and that in one point of great importance he was deeply mistaken .
His logical views , when simplified and corrected , give us a method of indirect deduction of extreme generality and power , founded directly upon this most fundamental Law of Thought .
A proof of the truthfulness and power of this system is to be found in the fact that it can be embodied in a machine just as the Calculus of Differences is embodied in Mr. Babbage 's calculating machine .
To explain the nature of the logical machine alluded to , it may be pointed out that the third of the fundamental Laws of Thought allow us to affirm of any object one or the other of two contradictory attributes , and that we are thus enabled to develope a series of alternatives which must contain the description of a given class or object .
Thus , if we are considering the propositionsIron is metal , Metal is element , we can at once affirm of iron that it is included among the four alternatives : Metal , element , Metal , not element .
Not metal , element .
Not metal , not element .
But according to the second Law of Thought , nothing can combine contradictory attributes , and this law prevents us from supposing that iron can be not metal , while the first premise affirms that it is metal .
The second premise again prevents our supposing that the combination metal , not element can exist .
Hence the only combination of properties which the premises allow us to affirm of iron is metal , element .
In a similar manner a complete solution of any logical problem may be effected by forming the complete list of combination , in which the terms of the problem can manifest themselves , and then striking out such of the combinations as cannot exist in consistency with the conditions of the problem .
The logical machine actually constructed represents the combination , 16 in number , of four positive terms , denoted by A , B , C , D , and their corresponding negatives , a , 6 , c , d. The instrument is provided with eight keys , representing these terms when appearing in the subject of a proposition , with eight keys , placed to the right hand of the former , representing the terms when occurring in the predicate of a proposition , and with the certain operation keys denoting the copular of the proposition , thefull stop at the end of it , and the conjunction or , according as it occurs in the eulycet or tredicate .
There is also a key denoting the Jinis or end of an argument , which has the effect of obliterating any previous impressions , and making the machine a tabula nasa .
If , now , each of the letter terms A , B , C , D be taken to represent some logical term or noun , and propositions concerning them be , as it were , played upon the machine , as upon a telegraphic instrument , the machine effects thereby such a classification and selection of certain rods representing the 16 possible combinations of the terms , that only those combinations consistent with the propositions remain indicated by the machine at the end of the operations .
When once a series of propositions is thus impressed upon the machine , it is capable of exhibiting an answer to any question which may be put to it concerning the possible combinations which form any class .
The machine thus embodies almost all the powers of Boole 's logical system up to problems involving four distinct terms , and to represent problems of any complexity involving any number of terms only requires the multiplication of the parts of the machine .
The construction involves no mechanical difficulties , and depends upon a peculiar arrangement of pins and levers , which it would not be easy to explain without drawings .
In this arrangement of the parts the conditions of correct thinking are observed ; the representative rods are just as numerous as the laws of thought require , and no rod represents inconsistent attributes .
The representative rods are classified , selected , or rejected by the reading of a proposition in a manner exactly answering to that in which a reasoning mind should treat its ideas , and at every step in the progress of a problem the machine indicates the proper condition of a mind exempt from mistake .
It is believed that this logical machine may be usefully employed in the logical class-room to exhibit the complete analysis of any argument or logical problem ; and it is superior for this purpose to a more rudimentary contrivance , the logical abacus , constructed by me for the same purpose and previously described .
But by far the chief importance of the machine is in a theoretical point of view as demonstrating , in the simplest and most evident manner , the character and powers of a universal system of logical deduction , of which the first , although obscure solution , was given by Dr. Boole .

