2 
But in reality the innovations which mark the long run change 
are themselves determined or influenced by the preceding 
economic process. This is so because new know-how can not 
develop in abstract space but only in close association with 
the process of production and investment (learning by doing). 
It is a joint product of the actual economic process. By this 
earthbound character of technological evolution (which also 
applies to the scientific evolution which is as much a product 
as a source of the former) it is closely related to accumulation 
which it brings about as well as being brought about by it. 
How can we separate out what is exogenous from the technological 
development (evolution of know-how)? This is difficult because 
the know-how is always bound up with accumulation, output 
and productivity, and seems to be measureable only in this 
form. What then shall we put on the right hand side of the 
differential equation? It can not be the innovation, a full 
grown result of a stage of evolution. But if we try to go back 
to earlier stages in the development of the know-how (invention? 
research results?) we shall get no further because we can hardly 
quantify it. 
In a non-committal way we can represent the exogenous element 
simply as a random shock. (This I did in my paper on Concepts 
and Ideas of Growth). But this looks rather like an evasive