4 INTRODUCTION 



or even about cause and effect relationships between different events. 

 It is in this abihty to see through a mass oi data and construct a reason- 

 able hypothesis to explain their relationships that scientists differ most. 



The role of a hypothesis is to penetrate beyond the immediate data 

 and place it into a ne^v, larger context, so that we can interpret the 

 unknown in terms of the known. There is no sharp distinction between 

 the usage of the words "hypothesis" and "theory," but the latter has, 

 in general, the connotation of greater certainty than a hypothesis. A 

 theory is a conceptual scheme which tries to explain the observed 

 phenomena and the relationships between them, so as to bring into 

 one structme the observations and hypotheses of several different fields. 

 The theory of evolution, for example, provides a conceptual scheme into 

 which fit a host of observations and hypotheses from paleontology, 

 anatomy, physiology, biochemistry and other sciences. 



A good theory correlates many previously separate facts into a logi- 

 cal, easily understood framework. The theory, by arranging the facts 

 properly, suggests new relationships between the individual facts, and 

 suggests further experiments or observations which might be made to 

 test these relationships. It may predict new phenomena that ^vill be 

 observed under certain circumstances and finally may provide the solu- 

 tion for practical problems. A good theory should be simple, and should 

 not require a separate proviso to explain each fact; it should be flexible, 

 able to grow and undergo modifications in the light of new data. A 

 theory is not discarded because of the existence of some isolated fact 

 which contradicts it, but only because some other theory is better able 

 to explain all of the known data. 



Once a hypothesis has been established, the rules of formal logic 

 can be applied to deduce certain consequences. In physics, and to a 

 lesser extent in the biological sciences, the hypotheses and deductions can 

 be stated in mathematical terms, and far-reaching conclusions may be 

 deduced. From these inferences, one can predict the results of other 

 observations and experiments. Each hypothesis is ultimately kept, 

 amended or discarded on the basis of its ability to make valid predic- 

 tions. A hypothesis must be subject to some sort of experimental test— 

 i.e., it must make a prediction that can be verified in some way— or it is 

 mere speculation. Conversely, unless a prediction follows as the logical 

 outgrowth of some theory it is no more than a guess. 



The finding of results contrary to those predicted by the hypothesis 

 causes the investigator, after he has assured himself of the validity of 

 his observation, either to discard the hypothesis or to change it to 

 account for both the original data and the new data. Hypotheses are 

 constantly being refined and elaborated. There are lew scientists who 

 \\'ould regard any hypothesis, no matter ho^v many times it may have 

 been tested, as a statement of absolute and universal truth. It is rather 

 regarded as the best available approximation to the truth for some 

 finite range of circumstances. For example, the Law of the Conservation 

 of Matter was widelv adhered to until the work of Einstein showed 

 that it had to be modified to allow for the possible interconversion of 

 matter and energ)'. 



