Essay · Imaginary Tool
Spatial Optics: Let Geometry Do the Computing
The lens does not calculate. The physical arrangement is the operation.
A lens does not calculate. We calculate what a lens will do, but the lens itself does nothing of the kind.
Light arrives at a physical arrangement, and the result happens.
The mathematics is our description of that behaviour. It is not an instruction being executed inside the glass.
We draw diagrams, trace rays and use equations to predict what a lens will do.
But once the lens exists, none of that is required for the lens itself.
The light does not wait for an answer.
Nature behaves.
The optical transformation happens directly.
Let geometry do the computing.
Not because geometry is secretly performing arithmetic. Quite the opposite.
The physical arrangement makes the arithmetic unnecessary.
The lens does not simulate the transformation.
It physically performs it.
Binoculars make this easy to see.
Light comes from the world, passes through the optical system and continues directly into our eyes.
There may be no sensor, no image processor and no software reconstruction.
Yet the optical transformation has already happened.
A camera merely places a recording surface where we want to keep the result.
The sensor records what the optics have already done.
The arrangement is the operation
A conventional computer represents something and then operates on the representation.
A lens does not need that intermediate step.
The physical arrangement itself performs the transformation.
That suggests a different question.
Instead of asking only:
How do we calculate the answer?
Ask:
Can we arrange the physical situation so that the answer happens?
And this is not only an optical idea.
A soap film settles into a shape.
Flow sorts material.
Resonance selects some patterns and suppresses others.
We can describe all of these mathematically.
But the behaviour does not wait for our mathematics.
It happens anyway.
Let Nature Do the Computing
This gives us a portable thinking tool.
Before beginning a calculation, simulation, reconstruction or optimisation, ask:
Is the system already producing the result through its own behaviour?
If it is, perhaps the problem is not how to calculate the answer.
Perhaps the problem is how to use the answer that is already appearing.
That may mean placing a detector in the right place.
It may mean arranging the material differently.
It may mean allowing a process to settle, separate, filter or select.
Sometimes the difficult problem is not producing the answer.
It is finding where the answer becomes available.
A banana does not need to be calculated
A banana is an ordinary object.
That is precisely why it is useful.
Almost nobody looks at a banana and thinks about how difficult it would be to manufacture one from scratch.
The plant has to move water, sugars and minerals. It has to grow cells, build tissue, regulate chemistry, respond to light, maintain structure and gradually form the fruit.
An engineer could describe many of these processes mathematically.
The banana plant does not need the description.
It grows.
The useful point is not that nature has a hidden calculator somewhere inside the plant.
It is that the behaviour is already built into the living system.
Water moves because the physical conditions allow it to move. Cells divide because the biological machinery already exists. Chemical gradients form. Tissues respond. Growth changes the conditions for further growth.
The banana appears through the process.
Am I trying to calculate something that the system could simply be allowed to produce?
A banana makes the question almost ridiculous.
Of course we do not calculate the banana.
We grow it.
This is not an argument against mathematics or computers.
Calculation is indispensable when the natural process is inaccessible, too slow, too dangerous, too noisy, or simply does not produce what we need.
The discipline is only this:
Do not calculate twice.
Before reproducing a transformation numerically, check whether it is already being performed.
A portable question
Carry three questions with you:
Is the system already producing the result through its own behaviour?
Can I arrange the conditions so that the useful result appears directly?
Where do I need to stand, measure or act in order to use that result?
Let Geometry Do the Computing began with a lens.
But the lens was only the easiest place to see the principle.
Do not begin by asking how to calculate the answer.
First ask whether nature is already doing the work.