Apeil 14, 1911] 



SCIENCE 



581 



From this it follows as a third canon of 

 thought that all conclusions need an independ- 

 ent verification. If reasoning from generali- 

 zations were permissible and thought could 

 legitimately move from an acknowledged uni- 

 versal to a particular, verifications would not 

 be necessary. It is interesting to see the many 

 ways in which thinkers try to avoid the need 

 of verification. From this temptation scien- 

 tists are no freer than other thinkers and they 

 have furnished many notable examples of such 

 errors. Verifications are, however, always 

 necessary and they must be based on fresh 

 data. Reasoning merely points out where these 

 data exist and what data are pertinent. The 

 work of getting at the truth is only half done 

 when the conclusions drawn from premises 

 are shown to be valid. 



Thought is the connection between two ob- 

 jects or ideas brought about by the similarity 

 or dissimilarity of their inherent elements. 

 Progress in thought consists in passing from 

 indefinite marks of this identity or difference 

 to those capable of definite description and 

 measurement. This fourth rule of good think- 

 ing brings out the relation between observa- 

 tion and measurement. Verifications are 

 improved when observations are verified by 

 experiment and experiment by observation. 

 Only in this way can we be sure that the data 

 of the verification are independent of those of 

 the premises. There is a still further im- 

 provement when enunciated principles are 

 based on observation and their verification 

 consists of data arising from experiment and 

 measurement. Such proofs are the most 

 stable science can offer. It is a goal that can 

 not always be reached, but it should always be 

 striven for. Principles are most readily ob- 

 tained from observation ; their proof, however, 

 is complete only when measurable data af- 

 ford them a verification. 



A final rule of thought is that no law is to 

 be regarded general unless it is capable of in- 

 dependent statement and verification in many 

 fields of investigation. This is the doctrine of 

 multiple verification. It is often assumed that 

 the way to prove a law is to get more data in 

 some one field. Such a proof is less satisfac- 



tory than independent verifications coming 

 from data derived from other sciences. It is 

 the extension and restatements of a law in 

 other fields and by independent investigators 

 that raise its validity above local generaliza- 

 tions which have both time and space limita- 

 tions. All observations and experiments are 

 of this local character. They need multiple 

 verifications to make them worthy of general 

 acceptance. It is a corollary of this that 

 thought is improved not by additional erudi- 

 tion in a given field, but by the movement of 

 thinkers from field to field. There is a strong 

 tendency to resent such a movement and the 

 intruder is likely to receive rough treatment 

 by his new colleagues. Yet this has been the 

 way in which the greatest victories of thought 

 are won. If Pasteur had not passed from 

 chemistry to medicine his work might have 

 been scientific, but it would not have been 

 effective. Narrow specialization tends to such 

 complete isolation of a group that its activity 

 becomes socially valueless. The mobility of 

 thinkers is the only safeguard against these 

 evils. It makes trouble but it brings results. 



Two conclusions follow from the preceding 

 discussion. The first is that there is but one 

 true method of reasoning in the use of which 

 all are equally liable to error. The second is 

 that there are two kinds of data, observations 

 and experiments, both of which must form a 

 part of any complete verification. These gen- 

 eral statements would be of little use if they 

 were not applied to the problems that sepa- 

 rate social from physical science. While the 

 opposition is general it is focused upon the 

 controversies about man and his relations to 

 the animal world. The one group use as their 

 data the experimental knowledge of animals 

 and then sweepingly apply this knowledge to 

 man. " What is true of dogs is true of men " 

 is a dictum coming from deductive medicine 

 which illustrates the methods of all biologic 

 sociologists. On the other side, there are so- 

 cial laws established by observation which 

 have been accepted by the mass of mankind as 

 rules of conduct : Which of these is right both 

 in method and fact ? 



In the first place, it should be recognized 



