18 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



GENERAL METHODOLOGICAL CONSIDERATIONS. 



The problem of studying the consequences of any experimental inter- 

 ference with a living organism is fundamentally a problem of scientific 

 method, both general and specific. Both the experimenter and his 

 critical reader should have clearly in mind the available canons of 

 investigation, and the degree of accuracy that may be legitimately 

 expected in the results, as well as the specific lines of investigation 

 and the specific techniques that can be relied on to yield adequate 

 quantitative data. Since in our case the question at issue is the nature 

 of the neuro-muscular consequences of the ingestion of alcohol, the 

 logical problem is strictly causal. It is our task to isolate from the 

 complex phenomena that may follow the experimental ingestion of 

 alcohol the uniform and necessary consequences. 



It may not be amiss to emphasize at the beginning that the basic 

 experimental method of difference in its true form is inapplicable in 

 such experiments as these. It is obviously impossible to isolate a 

 single experimental circumstance in man. The living human organism 

 includes too many complex variables. It is subject to too many 

 rhythmic and arrhythmic changes, which make it, at any moment of 

 time, different from what it has ever been before. After the intro- 

 duction of our experimental circumstance into this complex of ever- 

 changing conditions, we could not be sure that even notable variations 

 in the measurements of selected processes were not conditioned in 

 whole or in part by organic changes which were quite unrelated to 

 our experiment. Our excuse for appearing to insist on the obvious 

 is that in past experiments on the physiological effects of drugs the 

 obvious has not always been noticed. The importance of distinguish- 

 ing between the accidental and the necessary by carefully planned 

 series of control experiments is a relatively recent product. It is nqt 

 always realized even in current studies. Still more frequently do 

 experiments on the effects of drugs on animals as well as on man fail 

 to provide for an adequate statistical elaboration of their results. 

 This is a particularly difficult matter in operative techniques. But, 

 in man at least, it is never a satisfactory procedure to regard succeed- 

 ing changes in a measured phenomenon as the effect of an experi- 

 mental change, merely because one is consequent to the other. 



If all the organic rhythms and accidental changes were adequately 

 known we might arrive at the quantitative results of our experiment 

 by a process of subduction. Unfortunately, with our present knowl- 

 edge this can not be done directly. With a few notable exceptions, 

 such as the work of Lombard, 1 and of Grabfield and Martin, 2 we know 

 altogether too little about the daily rhythms of even the simplest neuro- 

 muscular processes. We have still scantier quantitative data of the 



Bombard, Journ. Physiol., 1892, 13, p. 25. 



2 Grabfield and Martin, Am. Journ. Physiol., 1912-13, 31, p. 300. 



