14
H. KAesER, and J. A. BURNS
essential basis for our understanding. In fact the quantitative formulation
contains within it the structural information but not vice versa. The set of
simultaneous differential equations which describes the whole system is
the only means of answering the functional questions which we can pose.
System Limits and Master Reactions
The carrying into practice of this desirable prescription, however, meets
with two formidable difficulties. The one concerns the method of obtaining
the necessary quantitative information, the second the problem of solving
the equations which this information makes possible.
Our first concern must be to define our system. This is a necessary first
step both to determine what measurements to make and what equations to
set up. In the example in Fig. 2 we have arbitrarily defined it by a small
number of elements. In any real organism, however, this represents only
a part and is connected, perhaps in many ways, to other elements. Must we
include these too? And if we do, must we not include further elements
which, in turn, are connected to the bigger system? Does this reasoning
not lead us to the conclusion that we must consider the whole organism
even if we are to ask questions about only a part of it? This is a fundamental
methodological problem and we must enquire whether there are any a priori
or empirical reasons which allow us to escape from this dilemma. Before
attempting some kind of answer I would like to discuss an aspect of
measurement which is basic to this problem.
As was mentioned before, the investigator can choose his elements of
analysis or, put in another way, the measurements he wishes to make.
Furthermore he can make measurements both on the system and on chosen
parts of it. In doing so, he will find that certain classes of measurement can
be made in both situations, on the system and on part of it. For example, he
can weigh an organism, then dissect it~if necessary down to the molecular
level~and weigh each part. (If, in addition, he knows the law of additivity
of mass, he will come out with the same answer). But the main point is:
he can determine weight on any part of the system.
There is, however, another class of measurement, possible on the system
but not on the parts. These are properties which arise only because of the
interactions of the parts and which therefore naturally 'disappear' in the
process of elementary analysis. In this class are most aspects of growth and
many of the homeostatic properties, variously described as buffering,
canalisation, integration, etc. Certain aspects of morphogenesis~the
development of specific forms in growing organisms~also fall in this
category.
The ability of an organism to maintain, say, constant internal temperature against external fluctuation is not measurable when the analysis is
H. KAesER, and J. A. BURNS
essential basis for our understanding. In fact the quantitative formulation
contains within it the structural information but not vice versa. The set of
simultaneous differential equations which describes the whole system is
the only means of answering the functional questions which we can pose.
System Limits and Master Reactions
The carrying into practice of this desirable prescription, however, meets
with two formidable difficulties. The one concerns the method of obtaining
the necessary quantitative information, the second the problem of solving
the equations which this information makes possible.
Our first concern must be to define our system. This is a necessary first
step both to determine what measurements to make and what equations to
set up. In the example in Fig. 2 we have arbitrarily defined it by a small
number of elements. In any real organism, however, this represents only
a part and is connected, perhaps in many ways, to other elements. Must we
include these too? And if we do, must we not include further elements
which, in turn, are connected to the bigger system? Does this reasoning
not lead us to the conclusion that we must consider the whole organism
even if we are to ask questions about only a part of it? This is a fundamental
methodological problem and we must enquire whether there are any a priori
or empirical reasons which allow us to escape from this dilemma. Before
attempting some kind of answer I would like to discuss an aspect of
measurement which is basic to this problem.
As was mentioned before, the investigator can choose his elements of
analysis or, put in another way, the measurements he wishes to make.
Furthermore he can make measurements both on the system and on chosen
parts of it. In doing so, he will find that certain classes of measurement can
be made in both situations, on the system and on part of it. For example, he
can weigh an organism, then dissect it~if necessary down to the molecular
level~and weigh each part. (If, in addition, he knows the law of additivity
of mass, he will come out with the same answer). But the main point is:
he can determine weight on any part of the system.
There is, however, another class of measurement, possible on the system
but not on the parts. These are properties which arise only because of the
interactions of the parts and which therefore naturally 'disappear' in the
process of elementary analysis. In this class are most aspects of growth and
many of the homeostatic properties, variously described as buffering,
canalisation, integration, etc. Certain aspects of morphogenesis~the
development of specific forms in growing organisms~also fall in this
category.
The ability of an organism to maintain, say, constant internal temperature against external fluctuation is not measurable when the analysis is