366 BELL SYSTEM TECHNICAL JOURNAL 



2. Object 



The object of this paper is the presentation of a scientific basis for 

 interpreting the significance of chance variations in quality of product 

 and for eliminating causes of variability which need not be left to 

 chance, making possible more uniform quality and thereby effecting 

 certain economies. 



3. Nature of Control 



Let us consider a very simple example of our inability to do exactly 

 what we want to do and thereby illustrate two characteristics of a 

 controlled product. 



Write the letter a on a piece of paper. Now make another a just like 

 the first one; then another and another until you have a series of a's, 

 a, a, a, a, . . . . You try to make all the a's alike but you don't; you 

 can't. You are willing to accept this as an empirically established 

 fact. But what of it? Let us see just what this means in respect to 

 control. Why can we not do a simple thing like making all the a's just 

 alike? Your answer leads to a generalization which all of us are per- 

 haps willing to accept. It is that there are many causes of variability 

 among the a's: the paper was not smooth, the lead in the pencil was not 

 uniform and the unavoidable variability in your external surroundings 

 reacted upon you to introduce variations in the a's. But are these the 

 only causes of variability in the a's? Probably not. 



We accept our human limitations and say that likely there are many 

 other factors. If we could but name all the reasons why we cannot 

 make the a's alike, we would most assuredly have a better understand- 

 ing of a certain part of nature than we now have. Of course this 

 conception of what it means to be able to do what we want to do is not 

 new; it does not belong exclusively to any one field of human thought; 

 it is a commonly accepted conception. 



The point to be made in this simple illustration is that we are limited 

 in doing what we want to do ; that to do what we set out to do, even in so 

 simple a thing as making a's that are alike requires almost infinite 

 knowledge compared with that which we now possess. It follows, 

 therefore, since we are thus willing to accept as axiomatic that we 

 cannot do what we want to do and that we cannot hope to understand 

 why we cannot, that we must also accept as axiomatic that a controlled 

 quality will not be a constant quality. Instead a controlled quality 

 must be a variable quality. This is the first characteristic. 



But let us go back to the results of the experiment on the a's and we 

 shall find out something more about control. Your a's are different 

 from my a's; there is something about your a's which makes them yours 



