Biology Reference
In-Depth Information
powers seem to reduce to causal powers located at lower levels. For some they
even seem to 'drain away'. We think that this picture about both the organization
and the explanation of systemic properties and behaviors of complex systems is
not well taken. Indeed, it is important to avoid a simple transition from thinking
of levels of organization to levels of explanation, though the two are doubtlessly
connected. We will carefully distinguish these sorts of questions in what
follows.
We begin to draw a different sort of picture that seems to be more adequate,
or so we think: at the center of our two-dimensional approach is the notion of
different degrees of resolution. Instead of shifting levels, we first of all allow to
change the resolution (of description) at which complex systems are described
that exhibit systemic properties and behaviors. By doing so, we always keep the
focus on the whole system, we do not shift to lower levels. The characteristic
entities remain the same. Only the descriptions change. As a consequence, we
stay at the same level of organization.
Let us go into more detail. Mechanistic models aim at explaining the behaviors
or capacities of complex systems. For example, we may be interested in the
behavior or capacities of a person, a trade union, an organism, or a cell. These are
clearly different levels of organization and require very different explanations.
At whatever level, we start with a description of the behavior to be explained;
for example, the growth rate a bacterium exhibits in liquid culture, the ability
of a bacterium to digest lactose, or the behavioral symptoms characteristic of
schizophrenia. The description of the behavior should be precise enough to
identify the explanatory target within a certain degree of tolerance. Exactness is
not the issue, as it is not available. Thus
(1) We specify the behavior to be explained within a given degree of tolerance.
(2) We determine the grade of resolution for the description of the behavior to
be explained.
The questions are not independent. As we specify a behavior more precisely,
we will doubtless need a more exact description of the mechanism involved;
conversely, as we have more exact descriptions of alternative mechanisms, we
will need more precise descriptions of the behavior to be explained. The idea is
this: Explaining the behavior of a system will naturally involve some degree of
tolerance. So we might want to explain the rate at which an organism acquires
an adaptive response to an environmental stimulus. We will not require that the
explanation be exact. The degree to which we allow it to be inexact is the degree
of tolerance. The behavior to be explained can be described with varying degrees
of 'resolution'. So we may describe an adaptive response at the organismic
level, or in terms, say, of organ systems or of cells. At a resolution focused on
organisms, the adaptive response is a matter of learning. At a resolution focused
