Chemistry Reference
In-Depth Information
aspect of how the part-whole relation is applied in chemistry. In that respect its
interpretation as a branch of set theory is not rich enough for the job.
Taking a one membered set as an individual in this system a comprehensive
mereology can be devised provided that the sets involved do not differ one from
another by internal structures. For example Lewis ties sets to parts with the
principle
'
One class is part of another if and only if the first is a subclass of the
second
. So, in chemical terms, an atom is a single membered subset of the set of
atoms that is the relevant molecule, and so on. However Lewis notices that there is a
kind of disparity between a singleton as a one-membered set and as a distinct
individual. He asks how does a single item form a class when the notion of
'
class
'
is introduced as the elements that are
gathered together
as a class. If the concept
is entirely formal this issue seems irrelevant. Taken formally mereology does not
deal with unique individuals, that is any one for which there
could
not be another
like it.
'
8.4 Limits to Part-Whole and Whole-Part inferences
Mereological reasoning, so central to the formation of well-ordered chemical
discourses, is vulnerable in certain circumstances to its own brand of fallacies.
There are two fallacies involving inferences between claims about wholes and their
parts. It is a fallacy to ascribe an attribute the meaning of which is determined by its
use for a whole to a part of that whole (though not in every case). It is a fallacy to
take the parts of a whole to be constituents of the unanalyzed whole from which
they came (though not in every case). Is there a systematic way of deciding in
which cases there is the threat of a mereological fallacy and in which cases
projecting attributes from wholes to their parts and treating products of the analysis
of wholes as constituents of those wholes is acceptable?
Distinguishing between primary and secondary qualities, as Boyle and Locke
did, is one principled way to make this distinction, but is there a more general rule
for licensing and forbidding transfers of properties from parts to whole and whole to
parts? A clue to how this question could be answered can be found in the way
mereological fallacies haunt psychology, particularly neuroscience.
The first mereological fallacy is exemplified by a prevalent error in neuropsy-
chology (Bennett and Hacker
2003
). It is an error to ascribe an attribute which gets
its meaning from its use to characterize a whole human being to a part of that human
being. Thus, to say
the frontal lobes make decisions
is a mereological fallacy
'
'
because the word
gets its meaning as something a whole person does.
A drag racer can accelerate at
a
metres per second per second, but none of its
detached parts can. The acceleration they are subject to as parts of the assembled
racer can be attributed to each of them only in so far as they are integrated into
the car.
decide
'
'
is a concept that gets its meaning in this context from its use
to describe the motion of a whole car.
Acceleration
'
'
