Science advances neither by simple linear accumulation nor by pure historical arbitrariness. It progresses under constraint. That constraint is reality. A theoretical framework can organise, simplify, predict, unify. But it cannot make what resists it disappear indefinitely. At a certain point, anomalies accumulate, patches multiply, the explanatory cost becomes too high. Then what occurs is not a pure and simple destruction of prior knowledge, but a reconfiguration of perspective.
This is how the great scientific ruptures must be understood. They are neither intellectual whims nor mere fashions. They occur when a framework ceases to be a good operator of visibility of the real, and another becomes capable of doing better.
A theory is not a copy of the world. It is a lens
Perhaps the best way to understand a scientific theory is to conceive it as an optical device. A theory does not reproduce reality identically. It selects, hierarchises, makes certain phenomena visible, and relegates others to the status of background noise. It also decides, implicitly, what counts as an anomaly.
The Ptolemaic system was long powerful because it allowed calculation. But as it grew heavy with epicycles, its conceptual fertility diminished. Copernicus, then especially Kepler and Galileo, did not merely change a few equations. They changed the lens itself.
Similarly, Einstein did not merely "correct" Newton. He displaced the very plane of physical intelligibility, replacing the idea of a force acting in absolute space and time with a dynamic geometry of spacetime.
A living theory simplifies the field of the visible while increasing its explanatory power. A theory that rigidifies must constantly add external corrections to save appearances.
Science is historical. It is not relativist
Recognising that science is historically situated does not lead to saying that everything is equal. There is a real asymmetry between frameworks. Some survive tests better than others. Some predict more, with greater precision, across a greater number of domains. Some unify without arbitrarily masking counter-examples.
This is why the historicity of science does not abolish the real. On the contrary, it is the concrete form of access to it. Scientific knowledge is not absolute, but it is not floating. It converges through successive rectifications toward more robust invariants. This can be called, without naivety, a convergent or structural realism.
Ruptures do not annul. They encompass
The great scientific revolutions do not always throw old frameworks in the bin. They reinscribe them within a more restricted domain of validity. Newton remains extraordinarily effective in the regime of weak fields and velocities small compared to that of light. General relativity encompasses it as a limiting case.
In other words, the change of perspective is not necessarily a destruction. It is often a hierarchical reorganisation of the thinkable.
The real role of great thinkers
Great thinkers are historical pivots, not demiurges. They formulate, crystallise, render coherent a shift that has become possible because a set of conditions matured simultaneously. Anomalies must accumulate, mathematical tools must be available, instruments must be capable of deciding, a community must be able to understand and test the proposition.
Individual discovery still exists, but the stabilisation of truth is now almost always collective.
Quantum gravity: a crisis that is finally becoming testable
For decades, it has been known that something is stuck between general relativity and quantum mechanics. Singularities signal internal limits of the classical relativistic framework. For a long time, this crisis had an almost exclusively conceptual status.
Yet it is precisely this point that is beginning to shift. In recent years, and more distinctly in 2025-2026, the problem of quantum gravity is no longer merely a matter of formal coherence. At low energy, it is becoming an arena for experimental testing.
The GQuEST project, for example, aims to test signatures of spacetime fluctuations using an ultrasensitive tabletop interferometer. In parallel, a whole family of experiments inspired by the QGEM protocol seeks to determine whether a gravitational interaction between mesoscopic masses can produce quantum entanglement.
What would a positive detection actually prove?
This is where the situation becomes philosophically fascinating. In 2025, a paper published in Nature argued that certain classical theories of gravity could also produce entanglement in such devices. Jonathan Oppenheim's post-quantum gravity programme has continued to structure an alternative where gravity remains classical but coupled to quantum matter in a stochastic and testable way.
In other words, a positive experiment would not mechanically suffice to conclude "gravity is quantum" without additional conceptual work.
Two scenarios
First scenario: gravity proves indeed capable of mediating irreducibly quantum correlations. We would then likely be dealing with a new encompassment. General relativity would appear as a macroscopic limit of a deeper structure.
Second scenario: experiments favour an image where spacetime remains fundamentally classical, but standard quantum mechanics must be modified. This would be an even more radical rupture, because it would displace not only our conception of gravity, but also that of quantum theory itself.
In both cases, what matters is less the anticipated victory of one school than the transformation of the very regime of debate. A theoretical crisis becomes an experimental crisis. And it is always at that moment that science regains its strongest form.
What perspective shifts truly teach us
Perspective shifts in science do not show that truth is arbitrary. They show, on the contrary, that scientific truth is accessible only at the price of constant revision of the conceptual instruments through which we look at the world.
A theory is worth what it allows us to see, what it allows us to predict, what it renders coherent, and its capacity to long withstand the friction of the world without taking refuge in indefinite accommodations.
SYNTHESIS
The great scientific revolutions are less negations of the past than expansions of perspective. They displace the frontier of the visible. They reorder the field of anomalies and evidences. They reinscribe the old within the new, while opening a vaster space of intelligibility.