Ask almost anyone what a scientific model is, and the answer will probably sound something like this:
A model is a simplified representation of reality.
It is an appealing idea.
Maps represent landscapes. Scale models represent buildings. Diagrams represent machines. Scientific theories, we naturally suppose, represent the world.
The metaphor seems so obvious that we rarely stop to examine it.
Perhaps we should.
The history of science is littered with successful models that describe the same phenomena in remarkably different ways. Light has been modelled as particles, as waves, as electromagnetic fields, and as quantum systems. Matter has been described using atoms, fields, strings, and geometrical structures. Even space and time have been reconceived repeatedly.
If models are representations, which one represents reality correctly?
Perhaps none of them.
Or perhaps that question misunderstands what models are doing.
Consider a map.
A road map does not represent a mountain range in the same way as a geological survey. A weather map does not represent political boundaries. A nautical chart ignores features essential to a hiking map.
None of these maps is false.
Each is organised for a different purpose.
The map is not attempting to reproduce the landscape in miniature.
It is making particular distinctions useful.
Scientific models behave in much the same way.
A model does not present reality exactly as it is.
It organises possible observations in ways that make certain relations visible.
The familiar language of representation quietly encourages us to imagine that somewhere behind every model lies a complete reality waiting to be copied with increasing accuracy.
Yet scientific practice rarely proceeds in this manner.
Scientists build models because particular questions demand particular distinctions. Different models reveal different regularities. Some models become extraordinarily successful within one domain while proving almost useless in another.
This is not a failure of science.
It is precisely how science progresses.
The trouble begins when we mistake the usefulness of a model for evidence that it mirrors reality itself.
A subway map offers a simple illustration.
No one mistakes the coloured lines of a subway diagram for the city itself. Distances are distorted. Streets disappear. Rivers are simplified. Entire neighbourhoods may be omitted.
The map succeeds precisely because it leaves so much out.
Its power lies not in representing everything but in organising the distinctions relevant to travelling through a transport network.
Scientific models achieve something remarkably similar.
They foreground certain relations while backgrounding others.
The resulting organisation is not arbitrary.
It is disciplined by observation, experiment, mathematics, and continual empirical testing.
But none of this requires that the model exist as a miniature duplicate of reality.
The representational metaphor quietly invites another confusion.
If a model represents reality, then success is naturally interpreted as correspondence between the model and the world.
But what if successful science depends less upon correspondence than upon the capacity to generate fruitful construals?
A model allows us to ask new questions.
It allows us to make new predictions.
It allows us to recognise new phenomena.
Most importantly, it allows us to make distinctions that were previously unavailable.
Seen from this perspective, a scientific model is not a mirror held up to reality.
It is an instrument of construal.
This does not make reality subjective.
Reality continues to constrain every successful model. Experiments still fail. Predictions remain testable. Nature stubbornly refuses to cooperate with inadequate theories.
The world is not invented by our models.
But neither is it simply copied by them.
Between invention and imitation lies something more interesting.
Construal.
A model participates in the actualisation of phenomena by making particular relational distinctions available. It opens one way of seeing while necessarily closing others. Different models therefore reveal different aspects of the world's relational organisation without requiring that any one of them constitute reality itself.
This observation explains something curious about the history of science.
Again and again, revolutionary theories are not created by discovering new objects.
They emerge by reorganising existing relations.
The mathematics changes.
The distinctions change.
The phenomena become newly intelligible.
Reality itself has not changed.
Our construal of it has.
Nothing in scientific practice becomes less rigorous under this interpretation.
Models remain indispensable.
Prediction remains essential.
Experiment remains the final discipline upon speculation.
What changes is the role we assign to models.
They are no longer miniature replicas of an independently structured world.
They are disciplined instruments through which particular relational organisations become available to inquiry.
Models do not represent reality.
They participate in construal.
And perhaps that is why the history of science has been so extraordinarily creative—not because humanity has gradually assembled a perfect picture of reality, but because we have continually discovered new ways of making reality intelligible.
No comments:
Post a Comment