SCIENCE (and common SENSE ) 57 



SCIENTIFIC THEORIES 



Rejoicing in possession of myriad predictively powerful colligative 

 relations, we find diem less dian our heart's desire in two important 

 respects. First is the practical problem of r^m^mben'ng— remember- 

 ing not merely relations but also, for each, the auxiliary material 

 without which the relation has little value: the "stops" defining its 

 general range of applicability, the "feel" for the reliability of its pre- 

 dictions in various parts of that range, the grasp of the many alter- 

 nate denotations attaching to its conceptual terms. This first problem 

 arises presumably from the inadequacy of human memory; a second 

 problem arises surely from the intransigence of human aspiration. 

 We want to grasp the origin of scientific laws. We want to know how 

 or why they are as good as they are, and how or why they are de- 

 fective as they are. Whatever the predictive capacities they give us, 

 these relations leave us unsatisfied until we can "explain" them. 

 There is then the theoretical problem of understanding. Were we 

 able to solve this second problem, we would take a long step toward 

 solution of the first. We easily remember a rational statement involv- 

 ing some hundred-odd words; we find it far more difficult to memo- 

 rize a hundred-odd words in a nonsense sequence. We seek then to 

 see scientific laws as parts of a higher rational order. 



Were our minds and our intellectual heritage diflFerent, we might 

 find quite difiFerent orders rational. Being what we are, we find su- 

 premely rational the order of a postulational system, that particular 

 order of which Euclidean geometry presents us with a familiar ex- 

 ample. Such a system stipulates a small number of axioms, and cer- 

 tain rules for handling them. From the axioms, following the rules, 

 we then arrive deductively at a vast number of theorems which are 

 for us thereby "explained." Even today only a few physical theories 

 achieve some distant approximation to the strict form of a postula- 

 tional system; and this form is stilh anything but apparent in the 

 representative theories of, say, geology and biology. But scientific 

 theories have invariably the character of postulational systems: al- 

 ways they offer rational correlations in which, from a given set of 

 premises, we deduce an array of theorems we identify with colliga- 

 tive relations. The "deduction" and, even more, the "identification" 

 involve complexities to be examined in Chapter VIII. For the present 



