representone kind of model (theoretical models), and that all models are simplifi-
cations, as indeed are all representations. Thus a model aeroplane represents reality
(‘A model airplane should look like a real airplane’), but this kind of model cannot,
by itself, explain how aeroplanes manage to stay airborne. By contrast, a theoretical
model might explain why aeroplanes fly but would not look like an airplane.
‘Explanatory power is gained by moving away from “reality”’.^28
Waltz was not saying that models are not real in themselves.^29 Maps are models
too. They provisionally substitute one reality for another, in the process becoming
an integral feature of the world they are said to represent. A theoretical model is a
map representing/simplifying ‘some part’ of the world – ‘a picture, mentally formed,
of a bounded realm or domain of activity’^30 – on which the theorist has marked a
sequence of events that will necessarily take place. Necessity arises not from the
world as such, but from the choices the theorist has made in representing the world.
Amodel aeroplane may fly if the plans for it (map) provide for all the right
components – motor, wings, controls, etc. – connected in the right order.
Drawing maps, making choices on what events to connect, putting things in the
right order are imaginative acts, involving guesswork but harnessed to experience.
Does the theorist’s model work – does it seem to explain events in the real world
that it purports to? Can it be made to work better? These questions follow from the
familiar model of modern science, itself institutionalized though general acceptance
of the procedural rules (the scientific method) for checking theoretical models
(models stipulating necessary relations) against evidence taken to represent some
feature of ‘the real world’ and refining those models accordingly.
Thirty years ago, Waltz endorsed this model of science. I see no reason to think
he would repudiate it now. He acknowledged the importance it assigns to imagi-
nation (in his words, creativity and intuition): ‘To form a theory requires envisioning
a pattern where none is visible to the naked eye’ – or, for that matter, to any of the
senses, no matter how much instrumental assistance we give them.^31 By implication,
there are many patterns visible as such, if not to all observers, then to those observers
with some idea where to look.
For most observers, rules, institutions, and agents variously constitute visible
patterns (constitutein both passive and active senses). The invisible patterns that
models propose are structures, and structures are theoretically relevant if they
function either as causes or as limits on causal processes.^32 Different models assign
causal significance to different structures and direct observers to look for visible
patterns (including especially frequent if not constant conjunctions). Rather
contentiously, Waltz’s theory invokes a restrictive conception of structure – so
restrictive it eliminates most visible patterns from consideration.^33
Waltz’s restrictive conception of structure discounts the visible presence of
institutions plausibly enabling Waltz’s structure of international politics to function
as the model requires. Other scholars propose models defining structure less
restrictively and, like Waltz, ask how they comport with some part of the world.^34
In the pursuit of reliable knowledge about the world, scholars oriented to science
tack between their theoretical models and the best evidence they can muster. Doing
Structure? What structure? 93