Certainty in the fundamental rules of nature is attainable, but only through the careful induction from observed phenomena and their mathematical explication. We know the laws of motion and universal gravitation not by conjecture, but by demonstrating their necessity to explain the orbits of the planets and the fall of an apple alike. Anything beyond this rigorous deduction, any speculation unfettered by experimental proof, is mere hypothesis and has no place in natural philosophy.
We can never know the fundamental rules of nature with absolute, final certainty. Instead, our understanding advances through a dynamic process of conjecture, proof, and, crucially, refutation. Consider the story of Euler's formula, $V - E + F = 2$, for polyhedra. Initially, it seemed a universal truth, demonstrable by proofs. Yet, it was only through the construction of counterexamples – objects that satisfied the intuitive definition of a polyhedron but violated the formula – that mathematicians were forced to refine their definitions and proofs. This ongoing cycle of 'proofs and refutations' is the engine of scientific progress, not the attainment of static certainty.
Newton’s emphasis on induction and experimental proof is undoubtedly vital; it forms the 'hard core' of any viable research program. However, to demand *certainty* based solely on past observations risks ossification. The history of science is replete with examples where established 'laws,' once thought certain, were later shown to be approximations or special cases within broader, more encompassing theories. The transition from Newtonian mechanics to relativity, for instance, did not disprove Newton but rather demonstrated its limitations and indicated a more sophisticated, albeit less intuitively 'certain,' framework. Certainty is a phantom, a seductive illusion that hinders the very progress we seek. Our strength lies not in claiming certainty, but in our capacity for self-correction and the continuous refinement of our conjectures.
Certainty regarding the fundamental rules of nature is a notion best relegated to the realm of theological dogma, not scientific inquiry. We operate within axiomatic systems, constructing models that, at any given moment, offer the most coherent and predictive explanation of observed phenomena. The power of these models, be they quantum mechanics or classical mechanics, lies in their internal consistency and their utility, not in any supposed apprehension of absolute truth. We don't *know* the rules; we develop extremely effective formalisms that *behave* as if they are the rules.
Lakatos speaks of proofs and refutations, a valuable process. However, the *certainty* he fears is a phantom. What we achieve is a high degree of confidence within a defined mathematical framework, a framework we can rigorously test. When a new observation arises that contradicts our current model – for instance, if a subatomic particle demonstrably violated the probabilistic predictions of quantum field theory – it signifies a breakdown in the *model*, not necessarily a fundamental error in the *process* of model-building. We then refine or replace the model. This is not a refutation of certainty, but a demonstration of the ongoing utility of our formalized, probabilistic approach. Newton’s ambition for absolute, demonstrable certainty is admirable, but it overlooks the inherent probabilistic nature of the universe as described by modern physics.
The notion of "certainty" concerning the fundamental rules of nature is, I submit, a linguistic and philosophical muddle. What we achieve is not certainty, but rather a high degree of probability coupled with a sophisticated understanding of the logical relationships between propositions. We can state with considerable confidence that *if* our axioms describing physical phenomena hold, *then* certain consequences follow. To speak of knowing these rules with *certainty* implies a direct, unmediated apprehension of ultimate reality, a claim unsupported by logic and, I would venture, by the very nature of knowledge itself.
Newton’s formulation, while brilliant, rests on the assumption that the observed regularities represent the entirety of truth. He sought a bedrock of absolute proof, which, as Lakatos rightly observes, is perpetually elusive. We progress not by discovering immutable truths, but by constructing increasingly robust, predictive frameworks. When a framework fails, as it inevitably will if pressed beyond its logical scope, it signals not the demise of certainty, but the refinement of our models. Von Neumann's point about axiomatic systems is pertinent here; our confidence lies in the internal coherence and predictive power of these systems, not in any mystical attainment of final truth.
The certainty I speak of is not that of absolute, immutable decree, but of demonstrable necessity derived from the phenomena themselves. Consider the inverse-square law of gravitation. It is not a conjecture offered for acceptance, but a mathematical consequence demonstrably required to reconcile Kepler’s meticulous observations of planetary motion with the very principles of motion I laid down. This is a certainty born of empirical validation and logical deduction, a stark contrast to the probabilistic models and self-correcting conjectures described by Lakatos and von Neumann, which acknowledge a persistent uncertainty at their core. My pursuit is to reveal the established order, not to fashion it anew through a perpetual process of revision.
Newton insists on a certainty derived from empirical necessity. But this is precisely where the illusion of certainty arises. Take the very example of the inverse-square law. When we observe the motions of planets, we are observing a complex interplay of forces and positions. The "necessity" of the inverse-square law is a product of our mathematical model imposed upon these observations, a model that successfully explains the data *within its intended scope*. The history of physics is a testament to how these seemingly necessary explanations are often refined or superseded by more comprehensive ones, like general relativity, which then reinterprets the "necessity" of the earlier law as an approximation valid under specific conditions. This is not an attack on Newton's genius, but a recognition that what appears as bedrock certainty is often a well-supported, yet fallible, conjecture within a powerful research programme.
Newton’s “demonstrable necessity” is precisely the sort of comforting illusion I find problematic. The inverse-square law of gravitation, for instance, is a superb mathematical description of planetary motion *within the framework of classical mechanics*. But this framework, as Lakatos implied with relativity, has its limits. When we encounter phenomena that defy that description – say, anomalies in Mercury’s perihelion – we don't declare the universe *fundamentally* uncertain. We recognize that our *model* is insufficient and seek a new formalism. The certainty, if it exists, is not in the rules themselves, but in our capacity to construct and refine *consistent and predictive formal systems*. The universe doesn't owe us ultimate truth; it offers data for us to organize into ever more elegant and functional mathematical structures.
The very language employed by Newton, speaking of "demonstrable necessity," betrays the very uncertainty he seems to disavow. This "necessity" is not an inherent property of nature itself, but a property of our *models* of nature. When we assert the inverse-square law, we are not stating a fact about the universe *tout court*, but rather a proposition within a specific logical system that has, to date, been remarkably successful in predicting observed phenomena. The "certainty" resides in the logical entailment: *if* these axioms are true, *then* this law follows. The failure of these axioms in new regimes, as pointed out by Lakatos and von Neumann, does not mean the universe is capricious, but that our current axiomatic framework is incomplete, requiring revision, not a crisis of ultimate unknowability.