The Ghost in the Machine: A Conceptual Autopsy of the Infinitesimal

The infinitesimal, a quantity conceived as being smaller than any assignable positive number yet stubbornly greater than zero, represents one of the most potent and controversial ideas in the history of science. Its trajectory—from an indispensable but logically suspect tool of discovery to an exiled ghost and finally to a fully resurrected citizen of a rigorous mathematical system—encapsulates the perennial tension between creative intuition and formal justification that drives intellectual progress. This monograph chapter argues that the history of the infinitesimal is the history of mathematics learning to tame its own most powerful intuitions, a process that redefined the very meaning of number, proof, and reality. The foundational postulates of the infinitesimal calculus, as wielded by its creators, were deceptively simple yet ontologically explosive: first, that such non-zero, evanescent quantities exist and can be manipulated algebraically; second, that the ratio of two infinitesimals can yield a finite, determinate value, representing a rate of change; and third, that higher-order infinitesimals can be legitimately discarded in calculations, an act of convenient neglect that would prove both miraculously effective and logically scandalous.

The crisis that necessitated the infinitesimal was born from the limits of classical geometry. The Greek method of exhaustion, perfected by Archimedes in the 3rd century BCE, could calculate areas and volumes with impeccable rigor but did so by laboriously trapping a desired value between inscribed and circumscribed polygons, a process that explicitly avoided confronting the infinite directly (Heath 1912). This deep-seated horror of the actual infinite, a legacy of Zeno's paradoxes, established a standard of rigor that the more intuitive methods of the 17th century would spectacularly violate. Precursors like Johannes Kepler and Bonaventura Cavalieri began treating curves as composed of an infinite number of "indivisibles"—lines made of points, or solids made of planes—a powerful heuristic that produced correct results for areas and volumes but lacked any logical defense (Boyer 1949). By the mid-1600s, a tension had reached its breaking point: the static, finite world of Euclidean geometry was incapable of describing motion, change, and curvature, precisely the problems at the heart of the new physics. The old regime could prove results about finished quantities but was mute on the subject of becoming.

The conceptual and evidentiary trajectory of the infinitesimal began in earnest with the near-simultaneous publications of Isaac Newton and Gottfried Wilhelm Leibniz in the late 17th century. Newton conceived of his "fluxions" as instantaneous speeds and his "fluents" as changing quantities, using infinitesimally small time increments, or "moments," denoted by 'o', to calculate them. His method involved introducing the moment, performing algebraic simplifications, and then causing the moment to "vanish" (Newton 1687, Principia Mathematica). Leibniz, working independently, developed a more elegant and enduring notation, explicitly positing infinitesimals 'dx' and 'dy' and treating them as the foundations of his differential and integral calculus (Leibniz 1684). Both methods produced a torrent of correct results, most spectacularly Newton's derivation of Kepler's laws of planetary motion, providing overwhelming instrumental evidence for their validity. Yet, this very success papered over a glaring logical inconsistency, which was ruthlessly exposed by Bishop George Berkeley in his 1734 treatise The Analyst. Berkeley famously derided infinitesimals as the "ghosts of departed quantities," pointing out the central contradiction: in calculating a derivative, such as that of $y=x^2$, one divides by 'dx' assuming it is non-zero, only to discard terms containing 'dx' in the final step by treating it as zero. This critique was logically unanswerable at the time; mathematicians had produced a machine that worked flawlessly but could not explain how, creating a foundational crisis that would shadow calculus for nearly two centuries. The mainstream reaction, led by figures like Jean le Rond d'Alembert, was to begin the slow process of reformulating calculus on the concept of the limit, aiming to eliminate the troubling infinitesimals entirely (Grabiner 1981).

The opposition schools that emerged were not offering an alternative calculus but were united by a demand for logical rigor that the infinitesimal framework could not supply. The earliest and most devastating critique came from the philosophical and theological program of George Berkeley, who saw the logical sloppiness of calculus as an affront to reason and a poor foundation for the mechanical philosophy that he felt was promoting religious skepticism (Berkeley 1734). His critique was not mathematical in the sense of proposing a new method, but meta-mathematical, challenging the very coherence of the discipline's foundational concepts. This line of criticism was tacitly accepted by the mathematical community, which spent the 18th century in a state of productive confusion, using the powerful but flawed tool while searching for a better justification. The true opposition program that crystallized was the "arithmetization of analysis," a movement that gained momentum throughout the 19th century. Its proponents, such as Augustin-Louis Cauchy and later Karl Weierstrass, sought to rebuild calculus from the ground up on the bedrock of arithmetic and the concept of the limit, explicitly banishing infinitesimals. Cauchy's work in the 1820s provided a less formal definition of the limit, but it was Weierstrass in the 1870s who delivered the final, decisive formulation with his static, logical (ϵ,δ) definition, a construction that made no appeal to motion or intuition (Boyer 1949). This school's rationale was that mathematics must be free of all contradiction and all appeals to vague physical or intuitive notions; its language must be that of inequalities and logical quantifiers alone.

The resolution of Berkeley's paradox appeared to be absolute and definitive with the triumph of the Weierstrassian (ϵ,δ) definition of a limit by the late 19th century. The infinitesimal was purged from formal mathematics, relegated to the status of a useful, albeit dangerous, historical heuristic for introductory pedagogy or a physicist's private scratchpad. The entire edifice of real and complex analysis was rebuilt on the limit concept, creating a period of unprecedented rigor and certainty. For nearly a century, the case was considered closed. The skeptical claims of Berkeley had been fully vindicated and subsequently addressed by expelling the offending concept. However, this resolution proved to be a long stalemate rather than a final victory. In the 1960s, the logician Abraham Robinson, using the sophisticated tools of model theory, demonstrated the existence of a consistent mathematical structure—the hyperreal number system—that included true, logically sound infinitesimals (Robinson 1966). His non-standard analysis proved that a number system could be formally constructed to contain numbers 'dx' that satisfy the intuitive properties Leibniz had ascribed to them. This stunning development showed that the flight from infinitesimals was a historical choice, a path taken to achieve rigor, but not the only possible one. The ghost had returned, now fully embodied in a rigorous logical framework, demonstrating that the original intuition of the founders was not, after all, incoherent.

The lexicon of the debate evolved dramatically, reflecting the conceptual shifts. Initially, Newton's 'fluxion' captured the dynamic essence of a rate of change, while his 'moment' (o) was a phantom-like, nascent increment of time. Leibniz's 'infinitesimal' (dx) and 'differential' were more abstract and ultimately more flexible. Berkeley's critique forced the introduction of the 'limit' as the central organizing concept, defined by Cauchy and then rigorously by Weierstrass using 'epsilon' (ϵ) and 'delta' (δ) to represent arbitrarily small but always finite, real numbers. The central thought experiment, implicit in Berkeley's critique, remains the most powerful illustration of the core problem. To find the slope of the tangent to the parabola $f(x)=x^2$ at a point x, the method requires evaluating the slope of a secant line through (x,x2) and a nearby point $(x+dx,(x+dx{)}^2)$. The slope is $\frac{(x+dx{)}^2-x^2}{dx}=\frac{2xdx+(dx{)}^2}{dx}=2x+dx$. To find the slope of the tangent, we must let the nearby point merge with the first, which means setting $dx=0$. The result is 2x. The mainstream reading of this paradox for two centuries was that the procedure was a flawed shorthand for a proper limit calculation. The rival reading, vindicated by Robinson, is that in a hyperreal system, 'dx' is a true infinitesimal, and the 'standard part' of the number $2x+dx$ is simply 2x, a well-defined operation that avoids the contradiction of treating 'dx' as both zero and non-zero. Robinson's work introduced a new lexicon, including 'hyperreal numbers', 'non-standard analysis', and the 'standard part' function, which formally maps a finite hyperreal number to its nearest real number, thereby legitimizing the final step of the old Leibnizian calculation.

The paradigm of calculus, built on infinitesimals, defined its boundaries by its unparalleled success in describing the physical world, supplanting the cumbersome geometric methods of the past. Its primary intellectual competitor was not an alternative calculus but the demand for rigor that it itself could not satisfy. The Weierstrassian limit-based formulation emerged not as a competitor but as a successor foundation, co-opting the entire edifice of results from the original calculus and placing it on what was considered solid ground. This new paradigm so completely dominated mathematical thought that it redefined the boundaries of acceptable argument within analysis. The emergence of Abraham Robinson's non-standard analysis in the 1960s did not overthrow the limit-based approach but established itself as a co-equal, alternative foundation. It demonstrated that the intuitive power of the infinitesimal, which had been a key driver of skepticism, could be fully reconciled with modern standards of logical rigor. Successor theories have not sought to replace calculus but to generalize it, such as in the fields of stochastic calculus, which deals with random processes, or to explore alternative logical foundations, as seen in constructive analysis, which rejects proofs relying on the law of the excluded middle, a critique that echoes Berkeley's demand for tangible meaning in mathematics.

The central controversy was a battle between pragmatic utility and logical purity, personified by the titanic figures of Newton and Leibniz on one side and the sharp philosophical critique of Berkeley on the other. For nearly two centuries, the utility of calculus in physics and engineering was so overwhelming that the foundational questions were effectively bracketed; the machine worked, and its output was too valuable to discard over philosophical unease. Consensus was forged not through logical resolution but through overwhelming instrumental success. The social dynamics were shaped by nationalistic fervor, particularly the bitter priority dispute between British followers of Newton and continental mathematicians who adopted Leibniz's superior notation. Later, the institutionalization of the (ϵ,δ) method within the elite German university system of the late 19th century, driven by the immense prestige of Weierstrass and his students like Cantor and Dedekind, cemented its status as the sole legitimate foundation for analysis worldwide (Grattan-Guinness 2000). This consensus was so complete that when Robinson developed non-standard analysis, it was met initially not with hostility but with a degree of indifference and skepticism from an analytic community for whom the infinitesimal was a long-settled historical error.

The invention of calculus triggered a profound ontological rupture. Before, the number line was Archimedean and intuitively complete; the world of mathematics was static. The infinitesimal introduced a new kind of object, a number that was in a state of 'becoming zero', shattering the static ontology of classical mathematics and enabling the description of a dynamic, changing world. This created an epistemological crisis articulated by Berkeley: how can we trust knowledge built on a foundation of logical contradiction? The Weierstrassian revolution represented a counter-rupture, a restoration of a static, Archimedean ontology where the continuum was constructed from sets of discrete rational numbers. Epistemologically, it provided a secure, if complex, foundation for knowledge based on the logic of quantifiers, trading intuitive clarity for logical certainty. The dissenters' alternative ontology, that of a world containing actual infinitesimals, was dismissed as incoherent until Robinson's work demonstrated that a non-Archimedean field (one containing infinitesimals) could be just as consistent as the real numbers. This re-introduced a Leibnizian ontology as a viable, rigorous alternative, demonstrating that our conception of the number line itself is a consequence of our chosen logical axioms.

Methodologically, the early use of infinitesimals was a tool of unprecedented heuristic power, allowing for rapid discovery and calculation in physics and geometry by prioritizing algorithmic procedure over deductive proof. The subsequent shift to the (ϵ,δ) framework, born from the critique of these methods, forced the development of a new, meticulous style of mathematical proof. The "arithmetization of analysis" made rigor itself a central object of mathematical practice, shifting the focus from simply getting the right answer to constructing an unassailable logical path to it. This new methodology spread from analysis to all other areas of mathematics, defining the modern standard of proof. The re-introduction of infinitesimals through non-standard analysis provided a new methodological toolkit. Many proofs that are notoriously complex using (ϵ,δ) arguments—such as the intermediate value theorem or the existence of solutions to differential equations—become strikingly simple and intuitive when translated into the language of infinitesimals (Keisler 1976). This created a new methodological debate over which foundation offers greater pedagogical clarity and is a more powerful engine for future discovery.

The instrumental power of calculus was immediate and world-altering. Newton's ability to derive the elliptical orbits of planets from his inverse-square law of gravitation using his fluxional calculus was the paradigm's first and greatest validation, effectively launching modern theoretical physics (Westfall 1980). Throughout the 18th and 19th centuries, it became the indispensable language of engineering, fluid dynamics, electromagnetism, and thermodynamics. The demand for rigor from the opposition did not directly generate these applications, but it secured their logical underpinnings, ensuring the stability of the rapidly growing technological world built upon them. After its formalization, non-standard analysis also found applications, particularly in providing simpler models for phenomena in stochastic processes, hydrodynamics, and quantum physics, where the concept of an infinitesimal change provides a more natural physical model than the elaborate logical machinery of limits.

Situated within the Enlightenment, the invention of calculus was seen as a triumphant validation of the power of human reason to decode the universe. The ability to tame the infinite and precisely describe the laws of motion fed a cultural narrative of progress and mastery over nature. Berkeley's theological critique was a direct countercurrent, a warning against the hubris of a purely mechanical worldview built on what he saw as logically flawed foundations. The eventual victory of the rigorous, abstract (ϵ,δ) approach in the late 19th century reflected a broader cultural shift in science towards formalism and a suspicion of intuition, partly driven by the discovery of counterintuitive objects like space-filling curves. The infinitesimal has persisted in popular culture as a metaphor for something immeasurably small, often misunderstood as simply "very tiny" rather than as a distinct ontological class. Ethically, the debate highlights the tension between practical utility and intellectual honesty—a dilemma that recurs in science when powerful new techniques outpace their conceptual justification.

The contemporary research frontier is no longer concerned with the logical validity of infinitesimals, which is now settled, but with their pedagogical and heuristic value. A significant open problem revolves around education: could a curriculum based on the intuitive framework of non-standard analysis, as advocated by mathematicians like H. Jerome Keisler, provide students with a more direct and less conceptually difficult path into calculus than the traditional (ϵ,δ) approach (Keisler 1976)? Experimental curricula have shown promise but have not displaced the established pedagogy. On the research frontier, infinitesimals are used as a tool in mathematical physics and probability theory to model systems with multiple scales, where separating standard and non-standard parts can simplify complex problems. The debate has thus shifted from the existential to the pragmatic: now that we know infinitesimals can exist without contradiction, where and how can they be most effectively deployed to advance mathematical understanding and scientific discovery?

In synthesis, the trajectory of the infinitesimal is a dramatic three-act play: a period of heroic, intuitive, and logically flawed discovery; a long exile enforced by a new regime of rigor; and a surprising modern rehabilitation through the power of formal logic. The infinitesimal's enduring legacy is its demonstration that mathematical truth is not monolithic; different, equally consistent foundations can support the same magnificent structure. It reveals that concepts dismissed as incoherent may simply be awaiting the development of a logical language rich enough to express them. The core insight is that the tension between powerful intuition and the demand for rigorous proof is the engine of mathematical progress, a dialectic that forces the discipline to continuously reinvent its understanding of its most fundamental objects. This long historical arc leaves us with one profound, lingering question: does the formal equivalence of the limit and infinitesimal approaches imply they are mere linguistic variants, or does the resurrected infinitesimal offer a genuinely different, and perhaps more fundamental, insight into the ultimate nature of the continuous?

References

• Berkeley, G. (1734). The Analyst; or, a Discourse Addressed to an Infidel Mathematician.

• Boyer, C. B. (1949). The History of the Calculus and Its Conceptual Development. Dover Publications.

• Grabiner, J. V. (1981). The Origins of Cauchy's Rigorous Calculus. MIT Press.

• Grattan-Guinness, I. (2000). The Rainbow of Mathematics: A History of the Mathematical Sciences. W. W. Norton & Company.

• Heath, T. L. (1912). The Method of Archimedes, Recently Discovered by Heiberg. Cambridge University Press.

• Keisler, H. J. (1976). Elementary Calculus: An Infinitesimal Approach. Prindle, Weber & Schmidt.

• Leibniz, G. W. (1684). "Nova Methodus pro Maximis et Minimis, itemque Tangentibus..." in Acta Eruditorum.

• Newton, I. (1687). Philosophiæ Naturalis Principia Mathematica.

• Robinson, A. (1966). Non-standard Analysis. North-Holland Publishing Company.

• Westfall, R. S. (1980). Never at Rest: A Biography of Isaac Newton. Cambridge University Press.