It seems to me that there is another aspect of the question of whether science explains anything that deserves attention.
Perhaps before asking how completely science explains reality, we should clarify what we mean by a scientific explanation and what science actually works with.
Let me try to explain what I mean.
Suppose we want to calculate the motion of the Earth around the Sun. We construct a mathematical model, take into account the gravitational interaction between the Earth and the Sun, include the influence of the Moon and other planets where necessary, and choose the required precision of our calculations.
But the Earth is not subject only to the gravity of these celestial bodies.
There are other stars, distant galaxies, and an enormous range of physical interactions that we do not include individually in a particular calculation. Their influence is not necessarily exactly zero. It may simply be so small, relative to the question we are asking, that omitting it does not significantly affect the required accuracy.
Moreover, we can adjust the level of detail depending on what we want to know.
For one task, a relatively simple Newtonian model may be sufficient. For another, we may need a more sophisticated calculation that accounts for additional perturbations and relativistic effects.
This is not a weakness of the scientific method.
On the contrary, the ability to identify what matters, abstract away from what is less relevant, and formalize particular relationships is a major reason for science’s extraordinary success.
But this raises the question that interests me.
What exactly have we obtained when we construct an exceptionally accurate model of the Earth’s motion?
We have certainly learned something about reality—something objective, testable, and practically significant.
But have we captured reality itself in its entirety?
Clearly not.
We have constructed a conceptual and mathematical representation that allows us to describe, predict, and explain selected features of a real physical process within a particular range of accuracy and applicability.
And the selectivity of that representation does not mean that the causal relationships it identifies are merely imaginary or approximate in the same sense. The model may simplify the situation while nevertheless providing genuine knowledge of how the world works.
The representation refers to reality, but it is not identical to reality.
I think this distinction is fundamental.
When we say that the Earth follows a particular trajectory because of gravitational interactions, this is a genuine scientific explanation. It does not cease to be an explanation simply because we can ask another question: why does the physical world possess these properties in the first place?
Science does more than identify stable correlations. It can reveal real causal relationships and mechanisms. For example, we can explain in detail how a particular virus causes disease, even though there is still no single universally accepted scientific definition of life. Moreover, whether viruses themselves should be considered living entities remains a subject of debate. None of this diminishes the scientific explanation of the mechanisms of viral infection.
At the same time, explaining a mechanism does not necessarily exhaust what can be asked about the nature of the reality in which that mechanism exists.
This is where I think we need to distinguish two different claims.
The first is that science allows us to acquire reliable knowledge of the real world.
The second is that our scientific description of the world constitutes an exhaustive account of everything that exists.
I readily accept the first claim.
But I do not see why the second should follow from it.
Scientific models are constructed within particular conceptual frameworks. They isolate properties, quantities, relationships, and regularities that allow a question to be investigated in a manageable way.
In a certain sense, we must make aspects of reality intellectually tractable before we can construct their mathematical or experimental representation.
We decide what to measure, which parameters are relevant, which idealizations are acceptable, and under what conditions a result can be applied.
Of course, reality is under no obligation to conform to our expectations. That is precisely why models are tested against observation and experiment, refined, and sometimes rejected.
But even an extraordinarily successful model remains a particular way of representing reality.
There is another interesting aspect to this.
We develop mathematical languages, choose symbols, and formulate equations. But we do not create the fact that these equations can sometimes describe, with astonishing accuracy, patterns in a world that existed long before human beings appeared.
Why is such a correspondence possible at all between the abstract constructions we develop and a physical reality that exists independently of us?
That seems to me to be a question of a somewhat different kind.
We can investigate how a mathematical model is confirmed. We can explain why one formalism works better than another. We can refine our understanding of the domains in which particular equations are applicable.
But the question of why the world is intelligible at all, why it exhibits a stable objective order, and why that order admits mathematical representation is not necessarily exhausted by identifying further physical regularities.
Nor do I think we need to set science and philosophy against each other here.
Scientific inquiry involves assumptions and standards concerning explanation, measurement, causation, evidence, and the distinction between successful and unsuccessful models.
Those assumptions and standards can themselves become subjects of investigation and revision.
So philosophy does not necessarily begin where science ends. Sometimes it helps us understand what science has already established and what status our claims about those results should have.
And this brings us, perhaps, to the most important distinction.
It is one thing to say: “We have identified a regularity and understand how a particular phenomenon arises.”
It is quite another to say: “We have therefore established that reality as a whole consists of nothing beyond the entities, properties, and relationships included in our scientific descriptions.”
The first claim may be the result of successful scientific investigation.
The second requires an additional philosophical argument.
And it does not become a scientific conclusion merely because it draws upon an extraordinarily successful scientific picture of the world.
Perhaps this is why I am not entirely comfortable with either extreme: the claim that science explains nothing and merely describes, or the claim that scientific explanation must, in principle, exhaust everything meaningful that can be said about reality.
The first, I think, underestimates the genuine explanatory power of science.
The second risks confusing a method of investigation with a complete ontological account of what exists.
Science does not need to claim an exhaustive explanation of existence in order to remain one of humanity’s greatest intellectual achievements.
And philosophy does not need to deny scientific explanations in order to ask questions extending beyond particular scientific models.
But I would add one further distinction.
The fact that a scientific explanation is not exhaustive does not, by itself, establish the truth of any particular metaphysical explanation—materialist, Aristotelian, theistic, or otherwise.
Each must also justify its own starting assumptions.
So for me, the most interesting question is not really whether science explains anything. Of course it does.
What interests me more is this:
At what point do we move from speaking about what we have established through a scientific model to making claims about what reality itself ultimately is? And do we always notice when we make that move?
It seems to me that this is one of the most interesting boundaries between scientific explanation and the philosophical interpretation of its results.
