SCIENCE (and common SENSE ) 61 



ing, we find this deduction easily confirmable, as are indeed a good 

 many others of its kind drawn from the kinetic theory. 



Theories and nets. In the first stage of the organization of experi- 

 ence flocks (in the sense of locks of wool or hair) of hypothetical 

 "atomic stimuli" are formed into immediate, individual constructs— 

 e.g., "objects" that exist stably in an otherwise chaotic welter of stim- 

 uli. In the second stage classes of objects and of the qualities of 

 objects are developed. Concepts are formed; from selected groups of 

 fibers are spun multiple fine threads which, in the third stage, are 

 intertwined to form colligative relations. At this level the maxims 

 of common sense, the rules of the craftsman, and many of the laws 

 of the scientist appear together. Only in science do we push on fur- 

 ther, to a fourth stage, in a fully self-conscious way. We weave the 

 relations together in postulational systems, theories. Says Popper: 



Theories are nets cast to catch what we call "the world": to ra- 

 tionalize, to explain, and to master it. We endeavor to make the mesh 

 ever finer and finer. 



He who realizes the existence of such a conceptual fabric, and is 

 capable of lifting it, carries with it all its cords, all the colligative 

 relations it accommodates. 



This metaphor grossly oversimplifies: each "stage" is not simply 

 produced from, but also helps to produce, those that precede it. 

 Moreover, in describing the functions of scientific theories, the case 

 I have sketched is too "ideal." Qualitatively the elements of gain are 

 correctly represented. Quantitatively the gains may be less impres- 

 sive, because fully to master a scientific theory is a task far more 

 difficult than I have indicated. For example, from the theoretical 

 postulates the manifold colligative relations descend by deductive 

 chains that normally involve, beyond conscious assumptions and ap- 

 proximations we have to learn to use as appropriate, further assump- 

 tions and approximations the experienced scientist makes automati- 

 cally, without full awareness of what he is doing. To master the 

 theory, and so to command the relations it holds in chains, one then 

 requires— beyond premises and rules for their manipulation— a "feel" 

 for the system. Immensely powerful tools, scientific theories are also 

 imperfect tools, handled with full effectiveness only by those who 

 have that subtle art of which Toulmin writes: 



