EXPLANATORY SCIENCE 201 



is the expression of an assertion in a more acceptable and 

 satisfactory form." x He then goes on to show that a state- 

 ment is more acceptable and satisfactory when it expresses 

 ideas with which we are more familiar, i.e., "it invokes a 

 definite response in our minds which we describe by saying 

 that we understand the statement." Presumably one does 

 not explain, therefore, unless he knows more about the un- 

 derived propositions than he does about the derived proposi- 

 tions; one always explains the unknown in terms of the 

 known. 



There is a measure of truth in this contention, and a 

 proper understanding of the logic of science demands that 

 the notion be clarified. What, precisely, is meant by saying 

 that in explanation one must know more about the ex- 

 planatory entity than about that which is to be explained? 

 Clearly there is a sense in which one knows more about that 

 which is to be explained, since this is given while the ex- 

 planatory entity is hypothetical and conjectural, i.e., the 

 datum is clearly given while the hypothesis is obscurely 

 given. Yet if the explanatory entity is to explain one must 

 know more about it than he does about that which he is 

 attempting to explain. This apparent paradox can be re- 

 solved if one recognizes the precise character of explanation. 

 The datum is always more familiar in the sense that it is 

 given observationally, but the hypothesis is always more 

 familiar in the sense that it is determined imaginatively. 

 Now the observational and the imaginative may differ both 

 quantitatively and qualitatively. Quantitatively, the hy- 

 pothesis contains less than the observationally given in 

 one sense, and more in another sense. No hypothesis at- 

 tempts to explain more than a selection from the total given, 

 hence every hypothesis asserts less than is contained in the 

 given; but every hypothesis attempts to make predictions 

 about the given, hence asserts more than is contained in 

 the given. For example, there are many thermal properties 

 of objects (the occurrence of melting and boiling at one 



. l What Is Science?, p. 77. 



