Functions and Norms 97
the perspective of their contribution to some performance of the containing system, and
this is a fundamentally technological perspective. The structure under consideration is one
that fulfi lls certain tasks, without which the system would itself not perform as it should.
If we say that the valves are a particular means to blood circulation, we imply that what
they do explains why they are there in the fi rst place: it is the end to which they are
means.
When the means-end relation is viewed as a chain, it can be iterated at will: A is a means
to B, which is a means to C, which is a means to D, and so on. The cause of an effect can
be viewed as a means to an end, and to the extent that the effect is desired (is an end),
some appropriate means to it will be desired as well. The intermediate effects, too, will
become ends—but only relative ends, namely relative to the end-character of that end to
which they in turn are means. But this potential regress of means to the ends that ground
them always stops at some point that legitimates the series; and there is in the formulation
itself an expectation that there will be some fi nal end that anchors the whole series of
means and ends.
There would seem to be three basic ways of stopping (anchoring) such a regress of
functions. Assuming that A is good for (doing) B, and B is good for (doing) C, and that
the regress stops at C, then
- C is something I happen to value or to be interested in. Free agents can set goals,
and the function of an item is relative to such goals. The regress of means to ends can be
arbitrarily broken off by the decision of a free agent. - C is something that I am (or should be) interested in for good reasons. There are
reasonable goals that agents should take into account. The regress of means and ends can
be ended for good reasons. There is some argument that can be adduced such that the
regress of means and ends is not just arbitrarily ended but rightly ended. For instance, an
item may have many different effects, only one of which fi ts well with the kind of thing
we take it to be. The escapement mechanism of a clock adds to the weight of the clock,
but weighing down is not what the hierarchical organization of the clock is directed at; it
is directed at keeping time. The pumping, but not the thumping, of the heart plays a rele-
vant role in the life of living creatures. - C is something that lies in the nature of things (or in human nature). The regress
can have a natural end. C is something that can benefi t from B. C has a good, which may
or may not involve intentionality.
The fi rst of these alternatives ends the regress of functions by appealing to our inter-
ests—cognitive or otherwise. The second appeals to warranted or justifi able conscious
interests. Neither of these seems to be very problematical if one acknowledges the exis-
tence of intentional agents. It is the third alternative that is likely to cause us diffi culties.
This third kind of functional regress corresponds to a standard narrative technique in
children’s stories. For instance, the old farmer Pettersson uses a fi shing rod to get the key