Scientific theories are full of things we can't see. Think about electrons, quarks, or the curvature of spacetime. These concepts are essential for modern physics, but no one has ever directly observed one. This raises a fundamental question: to accept a scientific theory, do we have to believe that these unobservable things are literally real?

Scientific realists would say yes. They argue that the goal of science is to provide a true description of the world, both the parts we can see and the parts we can't. If a theory works, it's because it's accurately describing reality.

But there's another perspective. Bas van Fraassen, a prominent philosopher of science, proposed an alternative called constructive empiricism. It offers a different view on what it means for a scientific theory to be successful.

This simple statement contains the core of the idea. Constructive empiricism suggests that science doesn't need to aim for absolute truth. Instead, it aims for theories that are "empirically adequate." This introduces a crucial distinction.

What does it mean for a theory to be empirically adequate? It means that the theory correctly describes the observable parts of the world. It "saves the phenomena." If a theory’s predictions about what we can see, measure, and detect hold up, then it is empirically adequate.

A theory being true, on the other hand, is a much stronger claim. It means the theory is correct about everything it describes, including the unobservable entities and structures it proposes. For example, a theory of particle physics is empirically adequate if its predictions match the results we see in particle accelerators. The theory is true only if quarks and gluons actually exist and behave exactly as the theory describes.

A constructive empiricist argues that we don't need to make the extra leap of faith. Accepting a theory simply means believing that it is empirically adequate. We can use the concept of an electron to build computers and predict electrical currents without needing to believe that tiny, negatively charged particles are literally orbiting atomic nuclei.

This brings up a tricky question: what counts as observable? The line isn't as clear as "visible to the naked eye." Van Fraassen defines "observable" relative to our epistemic community, which is humanity. Something is observable if a human, under the right conditions and without the need for complex theoretical interpretation, could observe it.

Moons of Jupiter seen through a telescope? Observable. The vapor trail of a particle in a cloud chamber? Observable.

The particle itself that caused the trail? Unobservable. Its existence is inferred based on the theory that explains the trail.

By focusing on empirical adequacy, constructive empiricism avoids arguments about the reality of things we can never see. It positions science as a pragmatic and powerful discipline focused on modeling the observable world, without getting bogged down in metaphysics. The goal is not to have a complete and true picture of reality, but to have theories that work where it counts: in the world we can experience and measure.