Can a book its author later dismissed as ‘foolish’ still hold the key to understanding how we know space itself? Bertrand Russell’s An Essay on the Foundations of Geometry—written when he was just 25—attempts exactly that: to reshape how we understand space, knowledge, and the very limits of human experience.
Overview
An Essay on the Foundations of Geometry (1897) is Bertrand Russell’s first philosophical book—written when he was just 25 years old. It attempts to answer a deceptively simple question: Which parts of geometry are necessarily true, and which are just accidents of our particular universe?
Russell’s central argument is that Kant was half-right. Some axioms of geometry are indeed a priori—they are necessary conditions for any possible experience of an external world. But the specifically Euclidean axioms—especially the parallel postulate—are not among them.
They are empirical, discoverable only through measurement and observation.
The book is not a beginner’s introduction to geometry. It assumes familiarity with nineteenth-century mathematical developments (non-Euclidean geometry, projective geometry, Riemannian manifolds) and engages deeply with Kant’s Critique of Pure Reason, Hegelian idealism, and the work of mathematicians like Gauss, Lobatchewsky, Bolyai, Riemann, Helmholtz, and Cayley.
Suitable readers
An Essay on the Foundations of Geometry is best for:
- Philosophy students interested in the foundations of mathematics and the Kantian tradition
- Mathematicians curious about the philosophical implications of non-Euclidean geometry
- Russell scholars tracing the development of his thought
- Readers who enjoy dense, argument-driven philosophical prose with historical depth
An Essay on the Foundations of Geometry is not a good choice for:
- Casual readers looking for an easy introduction to Russell or geometry
- Those without some background in Kantian philosophy or nineteenth-century mathematics
- Readers who prefer contemporary, empirically-focused philosophy of science
Table of Contents
1. Introduction
“This was one of the great events of my life, as dazzling as first love. I had not imagined there was anything so delicious in the world.”
That is how Bertrand Russell described his first encounter with Euclid’s Elements at age eleven. Less than fifteen years later, he would write a book that systematically dismantled the certainty of that Euclidean world—and then spend the rest of his career trying to live down the embarrassment.
An Essay on the Foundations of Geometry was first published in 1897, when Russell was 25 years old. It was based on his fellowship dissertation for Trinity College, Cambridge, submitted in 1895. The book marks his first major foray into analytic philosophy—a movement of which Russell would become a founding figure.
But here is the irony: Russell later dismissed the book as “somewhat foolish” and declared that “I do not think that there is anything valid in this early book”. The geometry in Einstein’s General Theory of Relativity, he admitted, was “such as I had declared to be impossible”.
So why should anyone read a book its author disowned?
Because the book’s failures are as instructive as its successes. It captures a brilliant young mind wrestling with the deepest questions about space, knowledge, and reality—at the precise moment when nineteenth-century mathematics was exploding the certainties of the eighteenth.
And because, as Michael Potter notes in his foreword to the 2023 Routledge Classics edition, “there may be more of his project worth rescuing than he here recognized”.
2. Background
To understand Russell’s project, you need to understand the intellectual earthquake that preceded it.
For over two thousand years, Euclid’s Elements stood as the model of certainty. Its axioms seemed self-evident; its theorems followed with logical necessity. Geometry was the impregnable fortress of the idealists—the proof that certain knowledge, independent of experience, was possible about the real world.
Then came the non-Euclidean revolution.
In the 1820s and 1830s, Nikolai Lobatchewsky and János Bolyai independently developed geometries that denied Euclid’s parallel postulate—and found them logically consistent. If you assume that through a point outside a line, more than one parallel can be drawn, you get hyperbolic geometry. If you assume none can be drawn, you get elliptic geometry.
The philosophical implications were devastating. If multiple geometries were logically possible, then Euclidean geometry could not be a priori necessary in the way Kant had claimed. As Russell put it: “Metageometry has destroyed the legitimacy of the argument from Geometry to space; we can no longer affirm, on purely geometrical grounds, the apodeictic certainty of Euclid”.
But Russell was not content to simply declare Kant wrong. He wanted to save what could be saved.
3. An Essay on the Foundations of Geometry Summary
Chapter 1: A Short History of Metageometry
Russell begins with a masterful survey of the development of non-Euclidean geometry, divided into three periods.
First Period (Lobatchewsky, Bolyai, Gauss): The pioneers who denied the parallel postulate and developed hyperbolic geometry. Their goal was indirect—to prove that the parallel postulate was logically independent of the other axioms.
As Russell notes, “If, by denying the axiom of parallels while retaining the rest, we can obtain a system free from logical contradictions, it follows that the axiom of parallels cannot be implicitly contained in the others”.
Second Period (Riemann, Helmholtz, Beltrami): This period was “largely philosophical in its aims and constructive in its methods”. Riemann introduced the concept of a manifold and the measure of curvature—a purely analytical expression that determines whether space is Euclidean (zero curvature), spherical (positive), or hyperbolic (negative). Russell notes that Helmholtz’s work, though influential, was “slightly unprofessional” mathematically.
Third Period (Cayley, Klein): The projective period, which reduced metrical properties to projective ones. Russell is skeptical of this reduction, arguing that it is “purely technical, and is not philosophically valid”. The use of imaginary points, he insists, is “wholly destitute of philosophical importance”.
Chapter 2: Critical Account of Previous Philosophical Theories
Russell examines four major philosophical positions:
Kant: Russell argues that Kant’s transcendental aesthetic proves too much. The argument from geometry “is upset by Metageometry, at least so far as those properties are concerned, which belong to Euclid but not to non-Euclidean spaces”. But Kant’s other arguments—those that prove space is a priori from the possibility of experience—remain valid, provided we substitute “any form of externality” for “Euclidean space”.
Riemann: Russell criticizes Riemann for assuming that space can be treated as a magnitude from the start. “A knowledge of the essential properties of space can never be obtained from judgments of quantity, which neglect these properties, while they yet presuppose them”.
Helmholtz: Russell rejects Helmholtz’s claim that geometry depends on mechanics and rigid bodies. “Mechanics logically presupposes Geometry,” he insists. Without a prior ability to measure space, physics itself would be impossible.
Lotze: Russell defends metageometry against Lotze’s criticisms, arguing that Lotze “has been misled by insufficient mathematical knowledge of the subject”.
Chapter 3: The Axioms of Geometry
This is the heart of the book. Russell divides geometry into two parts:
Section A: Projective Geometry
Russell argues that projective geometry is wholly a priori. Its axioms—homogeneity, dimensions, and the straight line—are “deducible from the conception of a form of externality, and, since some such form is necessary to experience, were all declared à priori”.
The key concept here is the relativity of position. As Russell puts it: “Externality is an essentially relative conception—nothing can be external to itself. To be external to something is to be another with some relation to that thing”.
From this relativity, Russell derives:
- The homogeneity of space—all positions are qualitatively similar
- The infinite divisibility of space
- The finite integral number of dimensions
Section B: Metrical Geometry
Here Russell distinguishes between a priori and empirical axioms:
A Priori Axioms (common to all geometries):
- Free Mobility—figures can be moved without distortion
- Finite integral number of dimensions
- Distance—two points determine a unique quantitative relation
Empirical Axioms (peculiar to Euclid):
- The three dimensions of our actual space
- The parallel postulate
- The axiom that two straight lines cannot enclose a space
Russell concludes: “The remaining axioms of Euclid—the axiom of three dimensions, the axiom that two straight lines can never enclose a space, and the axiom of parallels—were regarded as empirical laws, derived from the investigation and measurement of our actual space”.
Chapter 4: Philosophical Consequences
Russell addresses two fundamental questions:
First: How can a purely logical proof apply to experienced space?
Russell’s answer is transcendental. Some form of externality is necessary for experience—without it, we could not distinguish between different things. As he writes: “Consciousness of a world of mutually external things demands, in presentations, a cognitive but non-inferential element leading to the discrimination of the objects presented”.
Second: How do we resolve the contradictions in space?
Space gives rise to paradoxes: it is infinitely divisible yet has no actual elements (points are “zero quanta of space”); it is purely relational yet appears as something more.
Russell’s solution is to introduce matter—not the matter of physics, but an “abstract kind of matter, which is not regarded as possessing any causal qualities”. Spatial relations are relations between material atoms; space itself is merely “the bare possibility of relations”.
4. An Essay on the Foundations of Geometry Analysis
What Works
The Kantian Rescue Operation
Russell’s most original move is to split the difference between Kant and the metageometers. Kant was right that some geometrical knowledge is a priori; he was wrong to think Euclidean geometry in particular is necessary.
As Russell puts it: “Those axioms, which Euclid and Metageometry have in common, coincide with those properties of any form of externality which are deducible… from the possibility of experience of an external world”.
This is a genuinely sophisticated philosophical position—one that anticipates later work in the philosophy of science. As Michael Potter notes, Russell’s project raises “the question of whether the proposition that the angles of a triangle add up to 180 degrees is open to a kind of doubt to which the homogeneity of space is not susceptible”.
The Historical Survey
Russell’s history of metageometry is exceptionally clear for a book written in 1897.
He distinguishes the three periods with precision, identifies the key mathematical ideas, and explains their philosophical significance. Even today, it remains a useful introduction to the development of non-Euclidean geometry.
The Axiomatic Approach
Russell’s method—identifying the axioms common to all geometries and separating them from those unique to Euclid—is methodologically sound. It anticipates the axiomatic turn in mathematics that would culminate in Hilbert’s Foundations of Geometry (1899).
What Doesn’t Work
The Hegelian Baggage
Russell was still under the influence of Hegelian idealism when he wrote this book. As Michael Potter observes, “The Essay is marked throughout by Russell’s general belief (characteristic of the Hegelianism to which he then adhered) that abstract theories such as geometry inevitably contain contradictions, which can only be resolved by advancing to a more elaborate theory”.
This leads to some tortured arguments.
Russell spends much of Chapter 4 trying to resolve contradictions that, from a contemporary perspective, are simply not there. The “contradiction” of the point—that it is spatial yet contains no space—is only a problem if you insist on treating space as a substance rather than a structure.
The Dismissal of Projective Geometry
Russell’s claim that Cayley’s reduction of metrical to projective properties is “purely technical and philosophically irrelevant” is simply wrong.
Projective geometry has profound philosophical implications—it shows that metrical properties can be derived from purely qualitative ones. Russell’s dismissal of this insight reflects his metrical bias, not a weakness in projective geometry.
The Empiricism About Three Dimensions
Russell’s claim that the number of dimensions is “wholly the work of experience” is hard to defend. Why should three dimensions be empirical while two or four are not? Russell himself admits that “small errors are here impossible”—but if small errors are impossible, in what sense is the knowledge empirical?
The Physics Problem
Russell’s sharp distinction between geometry and physics was overtaken by events. As Potter notes, “The geometry in Einstein’s General Theory of Relativity is such as I had declared to be impossible”. In general relativity, geometry and physics are inseparable—the curvature of space is determined by the distribution of matter and energy.
5. Strengths and Weaknesses
Strengths
Intellectual ambition: Russell takes on Kant, Riemann, Helmholtz, and Lotze—and has something original to say about all of them.
Clarity of exposition: Despite the difficulty of the subject, Russell writes with remarkable clarity. His prose is precise without being pedantic.
Historical awareness: Russell’s history of metageometry is still worth reading today. He understands the mathematical developments and explains their philosophical significance.
The Kantian modification: Russell’s argument that some axioms are a priori while others are empirical is philosophically nuanced. It avoids the extremes of both Kantian apriorism and radical empiricism.
Weaknesses
The Hegelian framework: Russell’s attempt to resolve contradictions through dialectical progression feels forced and unnecessary.
The dismissal of projective geometry: Russell underestimates the philosophical significance of Cayley and Klein’s work.
The physics problem: Russell’s sharp distinction between geometry and physics cannot survive general relativity.
Russell’s own repudiation: When the author himself says “I do not think that there is anything valid in this early book,” it’s hard to ignore.
6. Comparison with Similar Works
Immanuel Kant, Critique of Pure Reason (1781)
Russell is responding directly to Kant’s transcendental aesthetic. Where Kant argued that Euclidean space is the necessary form of outer intuition, Russell argues that some form of externality is necessary—but not Euclidean space in particular. Russell’s book is essentially Kant updated for the age of non-Euclidean geometry.
Henri Poincaré, Science and Hypothesis (1902)
Poincaré reviewed Russell’s book in 1899 and disagreed fundamentally. Where Russell argued that some axioms are a priori, Poincaré argued that all geometrical axioms are conventions—“definitions in disguise”.
The choice between Euclidean and non-Euclidean geometry, for Poincaré, is a matter of convenience, not truth. Russell rejected this view, insisting that the axioms common to all geometries are genuinely necessary.
Hermann Helmholtz, On the Origin and Significance of the Axioms of Geometry (1870)
Helmholtz argued that all geometrical axioms are empirical, derived from experience with rigid bodies. Russell rejected this, arguing that “Mechanics logically presupposes Geometry”—without a prior ability to measure space, physics itself would be impossible.
G.E. Moore’s Critical Notice (1899)
Moore’s review of Russell’s book was devastating. Moore argued that Russell’s transcendental arguments were invalid and that his distinction between a priori and empirical axioms was unsound. Russell later admitted that “on all important points I agreed with it”.
7. Conclusion
Should You Read This Book?
Yes—if you are interested in:
- The history of philosophy of mathematics
- The Kantian tradition and its modifications
- The philosophical implications of non-Euclidean geometry
- The development of Russell’s thought
No—if you are looking for:
- A contemporary introduction to geometry
- An easy read
- A book that Russell himself would stand behind
Final Verdict
An Essay on the Foundations of Geometry is a brilliant failure.
It is brilliant because it captures a young philosopher at the height of his powers, wrestling with the deepest questions about space, knowledge, and reality. Russell’s historical survey is masterful, his arguments are precise, and his Kantian modification is philosophically sophisticated.
It is a failure because the project cannot work. Russell’s sharp distinction between geometry and physics is untenable after Einstein. His Hegelian framework is unnecessary. And his dismissal of projective geometry is simply wrong.
But as Michael Potter suggests, “there may be more of his project worth rescuing than he here recognized”. The question Russell raises—which propositions are so fundamental that we cannot coherently doubt them?—remains central to philosophy. And his attempt to answer it, however flawed, is still worth engaging with.
