-2. 
The second background is internal to photogrammetry and 
relates to the ultimate theoretical capacity and limitations of 
analytical instruments and to the practical engineering analysis of 
existing devices. The unprecedented universality of these instruments 
which is based on two powerful concepts, that is on digital photogram- 
metric data processing and on the feedback control of these processes, 
clearly indicate the possibility of extending their use to the domain 
of activities formerly confined within the boundaries of human intelli- 
gence. Notwithstanding the persisting limitations in automatic pattern 
recognition and especially pattern interpretation, the successful 
automation of processes requiring lower levels of intelligence, support 
the credibility of these indications about the theoretical capacity of 
analytical instruments. 
Accepting the fact that for the time being the interpretation 
of semantic information remains almost entirely the task of the human, 
the engineering analysis of analytical instruments has to deal predomi- 
nantly with the processing of metric information. The processing of 
semantic information, after it has been defined by the human, is 
carried out via the metric information as its carrier. Namely, identi- 
fiers that describe in a coded form a certain meaning that is to be 
attached to the metric information are in some way tied to sets of 
measured coordinates. Consequently the analysis of the significance of 
analytical instruments at the present stage of their development is 
confined to four main groups of operations in photogrammetric proces- 
sing: transformation of photo-coordinates of selected sets of points 
into a desired reference system of coordinates; determination of the 
parameters that are necessary to perform that transformation; merging 
of proper identifiers with a corresponding set of coordinates so that 
the assigned semantic information remains with the same metric informa- 
tion throughout its transformation, its storage and its retrieval, and 
presentation of the results of these operations. For any particular 
application these operations may be assessed with respect to accuracy, 
flexibility, speed, convenience and the labor saving that is offered by 
a particular procedure and instrumentation. Nevertheless, the signifi- 
cance of analytical instruments should not be looked upon merely as a 
clear-cut matter of economy but also as a matter of intellectual 
interest, since an unanticipated discovery may be inspired directly by 
the operation of these instruments. 
The matter of intellectual interest brings about the question 
of the definition of analytical instruments. Despite some limited 
efforts in this direction a good definition is lacking. Even the 
general notion of analytical instruments is still more or less an 
intuitive and a casual one. This kind of notion inevitably leads to 
misunderstandings. On the surface a definition is just a convention 
but intellectually its acceptance may have a much more active role. It 
is then not surprising that finding a good and acceptable definition is 
a major effort in creative scientific work. The idea about an instru- 
ment or a system is in distinct contrast with the idea of a theory. 
For instance, a physical theory is a generalisation about some aspects 
of the behavior of objects in the physical world. If the predictions 
that come from the theory are not confirmed then the theory is to be 
criticized and modified. Contrary to this the idea of an instrument or 
a physical system is a specification how physical objects ought to work. 
If they do not work according to that idea, then the behavior of the