Rebuilding Mathematics Education: From Number and Shape to Modern Math

The goal of mathematics education should not be merely to make students calculate faster. Nor should it be to march them through algebra, geometry, calculus and a sequence of course titles. The real goal is to help a person build a complete map of mathematics.

Calculation is part of mathematics, but it is only one basic skill within mathematical activity. Genuine study also means recognizing mathematical objects, understanding why definitions take the form they do, discovering relationships among objects, seeing why the conditions of a theorem matter, constructing proofs and counterexamples, translating among algebraic, geometric and analytic languages, abstracting real problems into mathematical structures, and then using those structures to understand reality again.

The map must show the main fields of mathematics. It must also show how they grow from intuitions about number and shape, how they separate, and how modern mathematics brings them together again.

1. Three levels of mathematical mastery

To prevent "I learned it" from meaning only "I once saw the formula," concepts and theorems can be assigned three levels of mastery.

Proof-level mastery means stating a definition or theorem accurately, reconstructing its proof independently, explaining why its conditions are necessary, constructing counterexamples and applying it to unfamiliar problems.

Structural mastery means explaining the main idea of a theorem, seeing what the proof depends on, knowing which problem it solves and understanding its links to other fields.

Horizon-level mastery means knowing what the theorem says, having seen representative examples, understanding its place in the larger architecture and knowing where deeper study leads.

Not everything must be mastered in the same way. Foundational theorems must truly be owned. More advanced results can first be understood structurally. Frontier subjects can begin by opening the horizon.

2. Primary school: build the first intuitions of number and shape

Arithmetic and elementary geometry should remain the center of primary mathematics. The point is not mechanical completion of exercises. It is to make quantity and space into objects a child can grasp directly.

Arithmetic

Primary arithmetic should include natural numbers, integers and place value; addition, subtraction, multiplication, division and their inverse relationships; factors, multiples, parity and divisibility; fractions, decimals and percentages; ratio, proportion and scaling; estimation and orders of magnitude; units, measurement and conversion; simple sequences and patterns; averages and intuitive organization of data.

Students should know more than how to calculate. They should know what an operation changes. Why can multiplication describe repetition, an array or scaling? Why can division describe both sharing and measurement? Why is a fraction both a part-to-whole relationship and a ratio of integers? Why can one quantity be written as a fraction, decimal or percentage?

Elementary geometry

Primary geometry should include length, angle, perimeter, area and volume; triangles, quadrilaterals, circles and basic polygons; cubes, rectangular prisms, cylinders and other elementary solids; symmetry, translation, rotation and scaling; simple coordinates and direction; cutting, recombining and equal-area transformations; and understanding solid structure from nets.

This stage does not require rigorous axiomatic proof, but it should cultivate the habit of explanation. A student should gradually answer not only "What is the result?" but also: Why can it be calculated this way? Is there another partition? Why do these figures have equal area? Does the conclusion hold in every case?

By the end of primary school, a student should have reliable senses of number, quantity, proportion and space, and should begin to distinguish observation, conjecture and explanation.

3. Secondary school: six years of algebra, geometry and analysis in parallel

Secondary mathematics should no longer be treated as one vague subject. It should develop through three long-running strands. Algebra studies structure and relation. Geometry studies space and invariants. Analysis studies change, approximation and infinite processes.

The three strands grow separately while meeting repeatedly through functions, transformations, symmetry, coordinates and models.

4. Secondary algebra: from operations on numbers to operations on structures

The aim of algebra is to show that mathematics does not only calculate with particular numbers. It studies general objects and the operations those objects permit.

Stage one: variables, relations and elementary structures

The first two years cover sets, elements, subsets and basic logic; maps, functions, composition and inverses; relationships among natural, integer, rational and real numbers; variables, expressions and identities; linear and quadratic equations; systems of equations and basic inequalities; polynomial arithmetic and factorization; exponents and logarithms; domains, ranges, monotonicity and composition of functions.

Proof-level results should include Euclidean division, the Euclidean algorithm, Bézout's identity, the fundamental theorem of arithmetic, polynomial division, the remainder and factor theorems, the binomial theorem, relations between the roots and coefficients of a quadratic, and the basic laws of exponents and logarithms.

Euclidean algorithm -> greatest common divisor -> Bézout's identity -> coprimality and divisibility structure

Stage two: linear algebra and discrete structure

The middle two years introduce vectors and linear combinations; linear dependence and independence; bases, dimension and coordinates; matrices and linear transformations; Gaussian elimination; systems of linear equations; determinants; rank, kernel and image; inner products, orthogonality and projection; basic counting; permutations and combinations; the pigeonhole and inclusion-exclusion principles; recurrences; and introductory graph theory.

Proof-level mastery should include the structure of solutions to linear systems, basic theorems on basis and dimension, rank-nullity, equivalent conditions for matrix invertibility, the relationship between determinants and linear transformations, the Cauchy-Schwarz inequality, the pigeonhole and inclusion-exclusion principles, combinatorial and algebraic interpretations of binomial coefficients, the handshake lemma and basic properties of trees.

Students should understand structurally why a determinant describes area, volume and oriented scaling; why a matrix is not a table of numbers but the coordinate expression of a linear map; why changing the basis changes the matrix but not the map; and how a recurrence can become a matrix iteration or generating function.

Stage three: complex numbers, eigenstructure and an introduction to abstract algebra

The final two years cover complex numbers and their geometric representation, polar form, De Moivre's formula, roots of polynomials over the complex numbers, eigenvalues and eigenvectors, diagonalization, quadratic forms, symmetric matrices, groups, subgroups, cosets, homomorphisms and isomorphisms, rings, ideals, quotient rings, fields and the motivation for number-field extensions, permutation groups, symmetry groups and introductory generating functions.

Proof-level results should include De Moivre's formula and the structure of roots of unity, Lagrange's theorem for finite groups, the relationship between eigenvalues and invariant directions, linear independence of eigenvectors for distinct eigenvalues, the reality of eigenvalues of real symmetric matrices, and orthogonal diagonalization of real symmetric matrices.

The Cayley-Hamilton theorem, minimal polynomials and conditions for diagonalization, quotient structures, the first isomorphism theorem, and the link between quadratic forms and classification of conics should reach structural mastery. The fundamental theorem of algebra, field extensions and solvability of equations, and the Galois-theoretic answer to the general quintic can remain at horizon level.

The end point of the algebra strand is a habit: when facing a mathematical object, first ask which operations it permits, which laws those operations satisfy and which properties survive transformation.

5. Secondary geometry: from proofs about figures to the structure of space

Geometry should not be a collection of tricks for drawing auxiliary lines. It should establish ideas of space, transformation and invariance.

Stage one: Euclidean geometry and proof

The first two years cover points, lines, planes and angles; definitions, axioms, propositions and proofs; congruence and similarity; parallel lines; polygons; circles, chords, tangents and inscribed angles; area and volume; straightedge and compass construction; loci; contradiction and construction methods.

Proof-level theorems include tests for congruence and similarity, the Pythagorean theorem and its converse, the midpoint theorem, the angle bisector theorem, the inscribed angle theorem, the tangent-chord theorem, intersecting chord and secant theorems, power of a point, Ceva's theorem and Menelaus' theorem.

These should not be learned as isolated tricks. They should be organized into major proof routes through congruence, similarity, area, circles and power, transformation and coordinates.

Stage two: analytic, vector and transformation geometry

The middle years cover coordinates in the plane and space; vectors, dot and cross products; lines, planes and their relative positions; distance, angles and projection; conics; parametric equations; translation, rotation, reflection and scaling; isometries, similarities and affine transformations; matrix representations of geometric transformations; and quadratic forms used to classify conics.

Proof-level mastery includes the relationship between dot products, length and angle; orthogonal projection; equations of lines and planes; focal properties of ellipses, hyperbolas and parabolas; basic invariants of rigid motion; and the basic classification of planar isometries.

Structural mastery includes matrix expressions of quadratic curves, simplification of quadratic forms by coordinate translation and orthogonal transformation, how eigenvalues determine the shapes of quadratic curves and surfaces, and what affine transformations preserve or fail to preserve.

geometric object <-> equation <-> matrix <-> transformation

The same object can appear as a figure, an equation, a matrix or a transformation. These are not four unrelated subjects. They are four languages for one structure.

Stage three: projective ideas, curves, surfaces and topological intuition

The final years introduce the projective plane, points at infinity, homogeneous coordinates, cross-ratios, duality and projective transformations; parametric curves and surfaces; tangent and normal vectors; an intuitive definition of curvature; polyhedra and Euler characteristic; connectedness, boundary, holes and continuous deformation; and a first intuition for manifolds.

Invariance of the cross-ratio under projective transformations, Desargues' theorem, Pappus' theorem, Euler's polyhedron formula, tangent lines to regular curves and elementary curvature formulas should reach proof or structural mastery.

V - E + F = 2

Students should also understand, at horizon level, how projective geometry unifies conics, why topology studies invariants under continuous deformation, why Gaussian curvature is intrinsic to a surface, and how manifolds generalize local Euclidean space.

The end point of the geometry strand is the ability to choose and move among synthetic geometry, coordinates, vectors, matrices and transformations instead of being trapped in one method.

6. Secondary analysis: from an intuition of change to rigorous control of infinity

Analysis should run through the whole secondary curriculum. Its purpose is not early drilling in differentiation and integration. It should gradually answer two questions: what is continuous change, and how can infinite approximation produce a definite result?

Stage one: functions, sequences and approximation

The first two years cover functions and graphs; composition and inverses; sequences, recurrences and iteration; monotonicity and boundedness; rates of change; secants and tangents; partition and approximation of area; infinite decimals and irrational numbers; intuitive limits; numerical error and approximate calculation.

The questions should begin concretely. Does an iterative process settle down? How does a polygon approximate the area of a circle? How does instantaneous velocity arise from average velocity? Why can an infinite decimal represent a definite real number?

Stage two: rigorous limits, differentiation and integration

The middle years introduce completeness of the real numbers, epsilon-N definitions of sequence limits, epsilon-delta definitions of function limits, continuity and uniform continuity, derivatives, mean value theorems, Taylor expansion, the Riemann integral and the fundamental theorem of calculus.

Proof-level results include uniqueness and algebraic laws of limits, convergence of bounded monotone sequences, the Cauchy criterion, nested intervals, Bolzano-Weierstrass, Heine-Borel, the intermediate value theorem, extreme values on a closed interval, Rolle's theorem, the Lagrange and Cauchy mean value theorems, Taylor's theorem with remainder, Riemann integrability of continuous functions and the fundamental theorem of calculus.

Students must understand the role of completeness. Many limit theorems are not consequences of algebra alone. They depend on the real numbers having no gaps.

Stage three: series, multivariable analysis and differential equations

The final years cover infinite series, absolute and conditional convergence, tests for positive series, power series, termwise differentiation and integration, pointwise and uniform convergence, multivariable functions, partial and directional derivatives, total differentials, gradients, Jacobian and Hessian matrices, multivariable Taylor formulas, extrema and Lagrange multipliers, multiple integrals, line and surface integrals, ordinary differential equations, phase space and introductory stability, and basic metric spaces.

Proof-level mastery includes comparison, ratio, root and integral tests; properties of the radius of convergence; conditions under which uniform convergence preserves continuity; the multivariable chain rule; the relationship between gradients and directional derivatives; and the solution and structure of first-order and constant-coefficient linear differential equations.

The inverse and implicit function theorems, Picard-Lindelöf existence and uniqueness, Green's theorem, the divergence theorem, Stokes' theorem and the basic idea of Fourier series should reach structural mastery.

The end point of analysis is an understanding of how limits make infinite processes rigorous, derivatives describe local linearization, integrals describe continuous accumulation, and differential equations turn local laws of change into global evolution.

7. The three strands must meet repeatedly

Algebra, geometry and analysis cannot become separate railways. Every stage of secondary school should include themes that cross fields.

Quadratic functions and quadratic forms connect polynomials and equations, parabolas, derivatives, extrema and integrals, matrices and optimization.

Complex numbers connect extensions of number systems, polynomial roots, rotations of the plane, trigonometric and exponential functions, and complex analysis. Euler's formula is a model of the meeting of algebra, geometry and analysis:

e^(i theta) = cos theta + i sin theta

Symmetry connects groups and group actions, geometric transformation, invariants, polynomials, crystal structures and physical laws.

Exponentials and logarithms connect algebraic laws, continuous growth, differential equations, complex rotation, probability distributions, and models of population, finance and physics.

Eigenvalues and eigenvectors connect matrices and linear maps, principal axes in geometry, stability of differential equations, modes of vibration, Markov processes, dimensionality reduction and principal component analysis.

Fourier ideas connect trigonometric functions, orthogonal expansion, function spaces, sound and images, partial differential equations, signal processing and quantum mechanics.

Through recurring themes like these, a student sees not a catalogue of topics but multiple appearances of the same structures.

8. University: four central intersections of modern mathematics

University should not repeat elementary calculus, basic matrix computation and routine analytic geometry. With those foundations already in place, it can enter modern mathematics directly.

Before the four central fields, the first stage of university needs a modern language upgrade. This is not remediation. It builds a shared language: groups, rings, fields, modules and tensor products; point-set topology; metric and topological spaces; measure and Lebesgue integration; smooth manifolds; fundamental groups and the ideas of homology and cohomology; and introductory language of categories, functors and natural transformations.

9. Algebraic geometry: from polynomial equations to geometric objects

Algebraic geometry should begin with affine varieties and eventually reach schemes.

The first layer covers polynomial rings; ideals, radicals and prime ideals; affine algebraic sets; the Zariski topology; coordinate rings; regular functions; morphisms of varieties; irreducibility, dimension, singularities and tangent spaces; projective space; homogeneous ideals and projective varieties; intersection and multiplicity.

Proof-level mastery includes Hilbert's basis theorem, Zariski's lemma, Hilbert's Nullstellensatz, the correspondence between ideals and algebraic sets, the contravariant correspondence between affine varieties and finitely generated reduced algebras, and basic properties of projective space.

Dimension theorems, elimination theory, Bézout's theorem, algebraic tests for singularity, and the way projectivization restores intersections at infinity should reach structural mastery.

The second layer introduces Spec A, prime spectra, structure sheaves, local rings, morphisms and gluing of schemes, fiber products, divisors, line bundles, and the basic ideas of sheaves and sheaf cohomology.

By graduation, students should be able to work fluently with affine schemes, translate between algebraic and geometric objects, understand localization geometrically, diagnose simple singularities with tangent spaces, see how schemes record multiplicity, degeneration and arithmetic information, and know where Riemann-Roch and cohomology belong.

The undergraduate end point is the ability to enter a modern algebraic geometry text and understand the central idea that a geometric object can be reconstructed in reverse from its ring of functions.

10. Differential geometry: from curves and surfaces to manifolds

Differential geometry should begin with smooth manifolds rather than stop with computations on curves and surfaces in three-dimensional space.

Its main subjects include manifolds and charts; tangent and cotangent spaces; vector fields and flows; tensors, differential forms, exterior derivatives and Lie brackets; submanifolds, immersions and embeddings; Riemannian metrics; the Levi-Civita connection; geodesics; curvature, Ricci curvature and scalar curvature; basic Lie groups; and de Rham cohomology.

Proof-level mastery includes applications of the inverse and implicit function theorems to manifolds, the regular value theorem, existence of partitions of unity, existence and uniqueness of the Levi-Civita connection, Stokes' theorem on manifolds, and Gauss-Bonnet for surfaces.

Sard's theorem, Frobenius, Hopf-Rinow, de Rham's theorem, the general idea of Gauss-Bonnet and the control of local geometry and global topology by curvature should reach structural mastery.

The undergraduate end point is a coordinate-free geometric language of manifolds, connections and curvature, with an understanding of how local differential structure creates global properties of space.

11. Complex analysis: the rigidity created by analyticity

Complex analysis should not be reduced to a course in residue calculations. It should show why complex differentiability is so much stronger than real differentiability.

The main topics are complex differentiability and holomorphic functions, the Cauchy-Riemann equations, power series, complex integration, Cauchy's integral theorem and formula, analytic continuation, isolated singularities, Laurent series, residues, harmonic functions, conformal maps, Möbius transformations, normal families and an introduction to Riemann surfaces.

Proof-level mastery includes Cauchy's theorem and formula, power series expansion of holomorphic functions, Liouville's theorem, the fundamental theorem of algebra, the maximum modulus principle, identity and open mapping theorems, the argument principle, Rouché's theorem, the residue theorem and Schwarz's lemma.

Analytic continuation and uniqueness, Montel's theorem, the Riemann mapping theorem, uniformization and meromorphic functions and differentials on Riemann surfaces should reach structural mastery.

The undergraduate end point is to understand how local differentiability in the complex plane automatically creates infinite differentiability, power series expansions and powerful global constraints.

12. Functional analysis: functions as points in a space

Functional analysis extends finite-dimensional linear algebra to infinite-dimensional spaces. It is a shared language of analysis, equations, probability and physics.

The main topics include normed, Banach and Hilbert spaces; bounded linear operators and dual spaces; orthogonal and complete orthogonal systems; weak convergence and weak topology; compact and self-adjoint operators; spectra; L^p spaces; distributions; Sobolev spaces; and introductory operator semigroups.

Proof-level mastery includes the Banach fixed point theorem, Hahn-Banach, uniform boundedness, open mapping and closed graph theorems, the projection theorem for Hilbert spaces, the Riesz representation theorem, the spectral theorem for compact self-adjoint operators and basic properties of L^p spaces.

Banach-Alaoglu, weak compactness and reflexivity, the spectral theorem for general bounded self-adjoint operators, basic problems of unbounded operators, Sobolev embedding, variational methods and weak solutions of partial differential equations should reach structural mastery.

The undergraduate end point is to understand functions as points in infinite-dimensional spaces, derivatives and integrals as operators, differential equations as operator equations, and Fourier expansion as orthogonal decomposition in a Hilbert space.

13. Probability and statistics: a mathematics of randomness built on analysis

At university, probability becomes a mathematical theory founded on measure.

Probability should cover probability spaces, random variables and distributions, expectation and integration, independence, conditional expectation, characteristic functions, modes of convergence, laws of large numbers, the central limit theorem, martingales, Markov chains and introductory stochastic processes.

Core results include monotone convergence, Fatou's lemma, dominated convergence, the Borel-Cantelli lemmas, weak and strong laws of large numbers, the central limit theorem, existence and projection interpretations of conditional expectation, and basic forms of martingale convergence.

Statistics should include parametric and nonparametric models, likelihood, sufficient statistics, point and interval estimation, hypothesis testing, regression, Bayesian inference, asymptotic theory and the basic structure of causal inference.

The point is no longer to consult a table. It is to understand the random mechanism that produced the data, the assumptions built into the model, the sensitivity of conclusions to those assumptions, the measurement of uncertainty, and the differences among correlation, prediction and causation.

14. Six actions for every mathematical subject

No important topic should end with "explain the formula, imitate the example, repeat the exercise." Each should require six actions.

1. Identify the object. What is it, and why do we need to define it?

2. Build examples. What are the typical cases? What is the simplest nontrivial one?

3. Find counterexamples. Which similar objects fail the definition? If a condition is removed, does the theorem still hold?

4. Study structure. Which operations, maps and transformations are permitted? Which properties remain invariant?

5. Prove theorems. Which conclusions are necessary, why are they true, and which deeper structures does the proof use?

6. Make connections. How does the object appear in algebra, geometry, analysis, probability and models of the world?

Learning eigenvalues, for example, cannot stop with computing a characteristic polynomial. Algebraically, an eigenvalue marks an invariant subspace. Geometrically, it describes a principal direction of a transformation. In analysis, it controls growth and decay in a differential equation. In physics, it describes modes of vibration and observables. In data analysis, it yields principal components. In probability, it governs the long-run behavior of a Markov process.

That is what complete mastery of a mathematical concept looks like.

15. Calculation, proof, connection and modeling belong side by side

This curriculum does not abolish calculation. It changes its place.

Calculation is the ability to handle a particular object accurately.

Proof explains why a conclusion must be true.

Connection translates among mathematical languages.

Modeling extracts structure from reality and explains the boundary of the resulting model.

Calculation lets us operate on mathematical objects. Proof makes conclusions trustworthy. Connection reveals mathematics as a whole. Modeling makes it a method for understanding the world. Remove any one of the four and mathematical education becomes partial.

16. What graduation really completes

Under this system, a university degree does not mean that a student has finished mathematics. It means the student has acquired the ability to enter it.

The graduate should be able to read abstract definitions and construct examples and counterexamples; read and reconstruct medium-length proofs independently; identify the conditions, conclusion and dependencies of a theorem; translate among algebraic, geometric, analytic and probabilistic languages; understand relationships between local and global, finite and infinite, discrete and continuous, deterministic and random; use mathematical software for experiments without mistaking calculation for proof; abstract real problems into mathematical structures; judge what a model explains and what it ignores; and enter an unfamiliar field by building a knowledge structure from definitions, theorems and examples.

Faced with mathematics, this person no longer asks only, "How do I calculate this answer?" The questions become:

What is the actual object? Which relationships are structural? What remains invariant under transformation? Can local information determine the whole? Why does an infinite process converge? How do determinism and randomness meet? Can the same problem be expressed in another mathematical language?

That is the way of thinking mathematics education should finally establish.

Primary school introduces number and shape. Secondary school develops structure, space and change. University enters the modern mathematics formed where those fields meet. Calculation is the foundation, proof supplies reliability, connection creates an integrated understanding, and modeling turns mathematical thought into a tool for understanding reality.

Graduation is only the true beginning of mathematics because it is the first moment when the student is no longer standing before a few textbooks. The student stands at the entrance to a vast building, able to see its main structures, internal passages and directions for further travel.