Chemistry Reference
In-Depth Information
Table 2.1 Specification of the components of the structuralist conception of theories, cf. Kuipers
(
2007
)
Component
Description
M
p
The potential models, defined as structures of the type
h
D
1
,
...
,
D
k
,
n
1
,
...
n
p
,
t
1
,
...
t
q
i
or
h
D
1
,
...
,
D
k
,
x
1
...
x
p
+
q
i
M
pp
The partial potential models
h
D
1
,
...
,
D
k
,
n
1
...
n
p
i
M
M
p
The models of the theory, which satisfy all the laws of the theory
r
: M
p
!
M
pp
The
'
restriction
'
relation which connects the potential models to the partial
potential models
M
p
The
constraint
relation (which will be taken as implicitly present in most of
C
P
'
'
what follows)
r
(M)
The projected models, i.e. the restriction of the models to the level of partial
potential models
K
The theory
core
, defined as
h
M
p
,
M
,
M
pp
,
r
,
C
i
'
'
I r
(
M
)
Weak empirical claim (note that constraints are implicitly assumed)
I
¼
r
(
M
)
Strong empirical claim (constraints are implicitly assumed)
its use. This similarity between conceptual spaces and potential models led Kuhn to
argue that the structuralist approach was to a high degree compatible with his notion
of a paradigm (see Kuhn (
1976
)) and also forms part of the discussion in Kuipers
(
2007
) on research programmes.
2.3.2 Belief Revision as Regimentation of Reduction
As I have outlined in Hettema (
2012a
), the notion of reduction in the structuralist
conception of theories is extraordinarily weak, a weakness which can be turned to
strength in cases where we wish to consider the sort of liberal interpretation of
Nagel required in the reduction of chemistry and physics. The leading idea in the
structuralist conception of reduction is a notion of isomorphism between structures.
Beyond that, a number of additional conditions may be imposed, as discussed in
Balzer et al. (
1987
), but the majority of these depend on the
kind
of reduction
relation that one wants to defend.
For the purposes of this paper, I will focus on the relationship between
AGM theory and the development of conceptual spaces as discussed by G¨rdenfors
and Zenker (
2011
).
8
G¨rdenfors and Zenker note that the framework of conceptual
spaces allows for four types of theory change:
8
In addition, belief revision has been introduced into the structuralist model by Enqvist (
2011
).
Enqvist develops a highly specific alternative to the notion of
'
reduction postulates
'
qua
'
linking
commitments
'
which I developed in Hettema (
2012a
). Enqvist
'
s construction relies on a construc-
tion of specialisation theory nets, to which he applies the AGM belief revision strategies. Enqvist
does not fully develop the AGM theory in a structuralist model, and ignores the stratification
between theoretical / non-theoretical levels of the theory. In general, developing complex notions
in the stratified model adds complications which are usually ignored in the first
of the
development of such models, see for instance the development of truthlikeness developed by
Kuipers (
1992
).
step
'
'
