How Concepts Become Possible: II. When Evidence Takes Shape

An idea can be in the air before anyone has formulated it.

But sometimes the situation is different. The problem is not that a solution is becoming increasingly imaginable. Rather, a scattered collection of observations begins to exert pressure on one another until a particular structure starts to emerge.

The discovery of the structure of DNA is a remarkable example.

It is usually remembered as the discovery of the double helix by James Watson and Francis Crick in 1953. But the story is more complicated, and more interesting, than that.

The double helix was not simply imagined and then confirmed.

Nor was it simply deduced from a single decisive experiment.

It emerged from a field of mutually constraining relations.

A structure waiting to be found?

By the early 1950s, much was already known about DNA.

The molecule was known to be extraordinarily important biologically. Evidence had accumulated that DNA was the carrier of hereditary information. Its chemical constituents were known. Erwin Chargaff had established regularities in the proportions of its bases. X-ray diffraction studies, particularly those associated with Rosalind Franklin and Maurice Wilkins, were revealing something about its physical structure.

But these facts did not announce a double helix.

They were pieces of evidence belonging to different experimental and conceptual contexts.

The problem was to find a structure capable of bringing them into relation.

This is a different kind of intellectual possibility from the one we encountered with Darwin.

With Darwin, an explanatory principle was becoming increasingly available within an intellectual environment.

With DNA, the challenge was more like this:

What structure could make all these different observations true at once?

The distinction matters.

Constraints that constrain one another

Chargaff's observations suggested relationships among the bases.

The chemical properties of the nucleotides constrained the ways in which they could be arranged.

The X-ray diffraction evidence constrained the overall geometry of the molecule.

The known dimensions of the molecule constrained possible arrangements still further.

And the biological requirement for replication imposed another extraordinary condition: whatever structure was proposed had somehow to make heredity physically intelligible.

Each constraint changed the significance of the others.

A proposed structure that satisfied one piece of evidence could fail another.

A chemically plausible arrangement might conflict with the physical dimensions suggested by diffraction.

A geometrically elegant model might fail to explain the observed relationships among bases.

The problem therefore wasn't simply to accumulate facts.

It was to find a configuration in which the facts could constrain one another coherently.

The discovery was a structure that could survive the whole network of constraints.

The model as a possibility

This is where Watson and Crick's model-building becomes particularly revealing.

A model is not merely a picture of something that already exists. It is a way of proposing a set of relations and asking whether the world will permit them.

A successful model therefore occupies an unusual position between imagination and evidence.

It begins as a possibility.

Then the evidence either sustains that possibility or destroys it.

In this respect, scientific discovery is not simply a matter of looking harder at reality. It is often a matter of constructing possibilities that reality can constrain.

The double helix became compelling because it did more than resemble the available evidence. It made disparate constraints mutually intelligible.

The pairing of bases, the complementary strands and the geometry of the molecule were not independent facts. They became aspects of a single relational structure.

And this is why the discovery is more interesting than the phrase "Watson and Crick discovered the double helix" suggests.

What was discovered was not merely a shape.

It was a way in which many things could be related.

The importance of Franklin

The historical story also warns us against treating conceptual discovery as the achievement of an isolated mind.

The evidence from Rosalind Franklin's X-ray diffraction work was crucial to understanding the physical structure of DNA. Her famous diffraction image, together with other experimental results, placed strong constraints on the possible geometry of the molecule.

This doesn't turn the discovery into a simple collective achievement in which everyone contributed an interchangeable piece.

Different people were doing different kinds of intellectual work.

Some produced observations.

Some developed chemical knowledge.

Some identified mathematical or geometrical constraints.

Some constructed models.

Some recognised that a particular arrangement could satisfy several constraints simultaneously.

The discovery therefore illustrates something important about conceptual evolution:

A conceptual possibility can be distributed across a community before it becomes concentrated in a particular formulation.

No single investigator necessarily possesses the whole possibility.

The possibility exists in the relations among their activities.

When evidence takes shape

This may be the key difference from our first case.

With Darwin and Wallace, the intellectual environment was becoming fertile for a particular explanatory principle. Independent discovery revealed the strength of that convergence.

With DNA, the evidence itself was becoming structurally organised.

The distinction is subtle.

Evidence does not simply sit there waiting to be collected. What counts as a constraint depends partly on the questions being asked and the conceptual possibilities being entertained.

The X-ray pattern becomes evidence for molecular geometry because someone has a reason to ask what geometry could produce it.

Chargaff's ratios become structurally significant because someone asks how the bases might be related within a molecule.

The evidence therefore does not determine the conceptual possibility from outside.

It participates in its formation.

This suggests a more reciprocal picture:

Possibilities generate tests; tests reshape possibilities.

A proposed structure makes certain observations relevant.

Those observations constrain the structure.

The revised structure makes new relationships visible.

Those relationships generate further constraints.

Discovery becomes a movement between possibility and constraint.

Not deduction, not guesswork

This is why scientific discovery is difficult to describe as either deduction or inspiration.

Deduction begins with premises and derives what follows.

But the double helix was not simply deduced from the known facts. There were many possible arrangements, and the relevant structure had to be constructed.

Nor was it mere guesswork.

A guess becomes scientifically productive when it enters a field of constraints capable of selecting among possibilities.

The creative act is therefore not necessarily the production of something unconstrained.

It can be the production of a possibility that is sufficiently structured for the world to answer it.

This may be one of the deepest roles of scientific imagination.

Imagination opens a possibility-space.

Evidence closes some of its paths.

The interaction between the two can make a previously invisible structure visible.

The structure changes the questions

There is another important feature of the DNA case.

Once the double helix had been established, it did not merely answer the question of DNA's structure.

It created new questions.

If the strands are complementary, how are they separated?

How does each strand participate in replication?

How is information encoded in the sequence?

How does a sequence of bases relate to the production of proteins?

The new structure therefore became an affordance for further discoveries.

A concept or model does not merely occupy a possibility-space.

It can reshape the possibility-space that follows it.

The discovery of structure creates possibilities for explanation.

And those explanations create possibilities for new questions.

This is conceptual evolution in a particularly concrete form.

A second kind of possibility

We can now add a second form to our emerging picture.

Some conceptual possibilities become increasingly probable because the intellectual environment is converging upon them.

Natural selection is our first example.

Other possibilities emerge because multiple constraints can suddenly be satisfied within one relational structure.

The double helix is an example of this.

In the first case, the question is:

What explanatory principle is the intellectual environment beginning to afford?

In the second:

What structure could make these apparently separate constraints belong together?

These are not the same kind of discovery.

And neither is inevitable in quite the same sense.

The evidence may strongly constrain what can survive, without uniquely determining what will be proposed. The conceptual possibility still has to be constructed.

This leaves us with a small but important modification to our earlier picture.

An affordance does not merely make some ideas easier to think.

A constraint does not merely eliminate possibilities.

Together, affordances and constraints can participate in the emergence of a structure that was previously unavailable as an intelligible possibility.

The double helix was not simply waiting inside the data.

Nor was it invented independently of the data.

It emerged between possibility and constraint.

And that "between" may turn out to be where much of conceptual evolution takes place.

The next case will complicate matters further.

For Copernicus, the crucial move was not to discover a structure that satisfied a growing collection of constraints.

It was to ask whether the centre itself had been placed in the wrong place.