Blog

First Principles in Mathematics

15 min read

Mathematics foundations discussed with a few known propaedeutic principles.

Filed under MathematicsDesign

Mr. Mahesh Shastry · Sydney, NSW, Australia · Personal mshastry@ieee.org

Introduction

Mathematics is built upon foundational principles that govern all mathematical reasoning, structures, and operations. These first principles serve as the bedrock upon which all mathematical concepts and theorems are constructed.

Core Principles

The Principle of Identity

A mathematical object is identical to itself.

Example: The number 2 is always 2, regardless of representation.

The Principle of Non-Contradiction

A statement cannot be both true and false at the same time.

Example: A number cannot be both even and odd.

The Principle of the Excluded Middle

Any statement is either true or false; no third option exists.

Example: Either or , but not both.

The Principle of Invariance

Certain properties remain unchanged under transformations. Example: The sum of the internal angles of a triangle is always in Euclidean space.

The Principle of Duality

Mathematical structures often come in pairs where one can be transformed into another while preserving meaning. Example: Boolean algebra has the duality principle, where AND ( ) and OR ( ) operations can be interchanged if 0 and 1 are swapped.

The Principle of Mutability

Mathematical structures can evolve while preserving essential relationships. Example: Graph theory models dynamic networks where nodes and edges change over time.

The Principle of Sufficient Reason

Every mathematical truth must have a reason or justification. Example: The proof of the infinitude of prime numbers follows logically from first principles.

The Principle of Induction

If something is true for a base case and can be extended step-by-step, then it is true for all cases.

Example: The sum of the first natural numbers:

The Principle of Well-Ordering

Every non-empty set of natural numbers has a least element.

The Principle of Substitution

If two things are equal, one can replace the other in any valid expression.

The Principle of Symmetry

If , then .

The Principle of Transitivity

If and , then .

The Principle of Extensionality

Two sets are equal if they contain the same elements.

The Principle of Comprehension (Separation and Replacement)

Naively, one might attempt to define a set by any property that its elements satisfy. However, unrestricted comprehension leads to paradoxes such as Russell’s. In formal set theory, this principle is replaced by more restricted axioms.

(Separation). For any set and any formula , the subset

exists.

(Replacement) (optional). If a definable function maps every element of a set to some unique , then the image

is also a set.

Example. The set of prime numbers can be defined by the property of being divisible only by 1 and itself:

Russell’s Paradox

The unrestricted form of comprehension asserts that for any property , there exists a set

Let us now consider the property , that is, the property of a set not being a member of itself. Define

We now ask: does contain itself?

Case 1. Suppose . Then, by the definition of , it follows that .

Case 2. Suppose . Then, by the defining property of , it follows that .

In either case we arrive at a contradiction:

This contradiction demonstrates that the unrestricted comprehension principle cannot be consistently maintained. Consequently, modern set theory restricts set formation through axioms such as Separation and Replacement, avoiding self-referential definitions.

Remark. Russell’s Paradox motivated the development of axiomatic systems such as Zermelo–Fraenkel set theory (ZF), which provides a consistent foundation for modern mathematics by replacing unrestricted comprehension with carefully formulated axioms.

The Principle of Continuity

Small changes in input lead to small changes in output in a continuous function.

Geometry: Incidence; Order; Congruence; Continuity

Primitives and Language

Primitives: (points); (lines); (planes). Predicates: for point–line incidence; for point–plane incidence; for betweenness; for segment or angle congruence; for parallelism. Equality is identity of primitives.

Incidence

  • G1 Existence and non–degeneracy: points not all collinear; a line; a plane; with some point not incident to and some line not incident to . Breakdown: prevents collapse to a single object; guarantees each primitive sort is populated.

  • G2 Line determination: . Breakdown: uniqueness gives a well–defined collinearity relation; enables coordinate construction.

  • G3 Plane determination: non–collinear; with all incident to ; and . Breakdown: axiomatises coplanarity and stabilises line inclusion.

  • G4 Line–plane stability: If two distinct points of a line lie in a plane; then the entire line lies in that plane. Derived: within a plane, two distinct lines meet in at most one point; two distinct planes meet in at most one line.

Order

  • G5 Betweenness axioms (Hilbert): (i) If then , , and are distinct and collinear; (ii) Symmetry: ; (iii) Genuine trichotomy: for any three distinct and collinear points , , and , exactly one of the following holds:

Breakdown: this establishes a definite, orientation free order along any line.

Optional density axiom. For any distinct points and on a line, there exists a point such that . Breakdown: ensures that between any two points lies another, giving the line an infinite (dense) structure.

  • G6 Segment addition: If ; then where length is the congruence class under . Breakdown: defines concatenation of lengths without numbers.

  • G7 Pasch axiom: If a line intersects one side of a triangle; it intersects exactly one of the other two sides. Breakdown: ensures coherent planar order; crucial for constructions and proofs.

Congruence

  • G8 Segment transport: For any segment and any ray ; there exists a unique on that ray with ; segment congruence is an equivalence relation. Breakdown: defines lengths up to rigid motion.

  • G9 Angle transport: Given oriented rays; there exists a unique transported angle congruent to a given one on a prescribed side; angle congruence is an equivalence relation. Breakdown: supports angle arithmetic independent of measurement.

  • G10 SAS axiom: If two sides and the included angle of one triangle equal those of another; the triangles are congruent; consequently corresponding parts are equal. Derived: triangle inequality; base–angles theorem; isometries exist.

Continuity and Parallels

  • G11 Archimedean principle: On any ray; repeated laying–off of a fixed segment eventually surpasses any given point. Breakdown: prohibits infinitesimals at the geometric level.

  • G13 Parallel axiom; scheme: Euclidean choice: through a point not on a line; exactly one parallel exists; hyperbolic and elliptic adopt their replacements. Breakdown: chooses geometry class.

Numbers: Algebraic; Order; Completeness

Peano Arithmetic

  • N1 Peano primitives: ; successor ; axioms: ; injective; . Breakdown: defines the discrete progression of natural numbers.

  • N2 Induction schema: For any formula ; if and ; then . Breakdown: proves properties for all naturals via basis and step.

Fields and Ordered Fields

  • N3 Field axioms: commutative; associative; distributive; additive inverses for all elements; multiplicative inverses for non–zero elements; . Breakdown: algebraic laws for addition and multiplication.

  • N4 Order compatibility: is a total order with translation and product monotonicity: ; and . Breakdown: order respects the field structure.

Completeness and Constructions

  • N5 Least upper bound axiom: Every non–empty bounded above has . Equivalent: nested intervals; monotone convergence; Heine–Borel in .

  • N6 Cauchy completeness: Every Cauchy sequence converges in . Breakdown: identifies as the completion of under the metric .

Calculus: Differential–Integral Structure

Fix an ordered field and an –algebra of real–valued functions on a domain .

Differentiation

  • C1 Derivation: A map is –linear; ; and satisfies Leibniz: .

  • C2 Normalisation: For identity on ; ; and for constants ; .

  • C3 Chain rule schema: For with appropriate domains; .

  • C4 Higher derivatives: Define recursively; product and chain rules extend by induction. In particular, the th derivative of a product satisfies the general Leibniz rule:

For example,

Integration

  • C5 Integral operator: For intervals ; an operator assigns to a suitable class of functions; linearly: .

  • C6 Additivity: If ; then whenever the terms exist.

  • C7 Order: If almost everywhere on ; then .

Bridging and Regularity

  • C8 Fundamental theorem schema: For a regularity class ; define ; then ; moreover for .

  • C9 Minimal regularity: Polynomials; continuous piecewise– functions on compact intervals; and their uniform limits belong to the integral domain of .

Specialisations: Riemann; Henstock–Kurzweil; Lebesgue; obtained by strengthening existence and convergence axioms; e.g.; monotone convergence and dominated convergence as schemas in the Lebesgue setting.

Topology: Open Sets; Separation; Countability; Closure

Topological Space

  • T1 Open–set axioms: A topology with ; arbitrary unions in ; finite intersections in .

  • T2 Continuity: continuous iff for all .

Separation; Countability; Compactness

  • T3 : Kolmogorov; points are closed; Hausdorff: distinct points admit disjoint neighbourhoods.

  • T4 Regular and normal: Regular: points and closed sets can be separated by neighbourhoods; Normal: disjoint closed sets can be separated by disjoint open sets.

  • T5 Countability: First countable: each point has a countable local base; Second countable: there is a countable base for .

  • T6 Compactness: Every open cover has a finite subcover; in Hausdorff spaces; compact sets are closed and continuous images of compact sets are compact.

Closure; Interior; Boundary

  • T7 Kuratowski closure axioms: A map with ; ; ; ; defines a unique topology in which is closure.

  • T8 Dual interior: ; boundary .

Derived: Urysohn lemma schema for normal spaces; Tietze extension for normal; second countable Hausdorff implies metrisable under additional axioms.

Spectral Theory: Operators; Spectra; Banach and –Algebras

Banach Spaces and Bounded Operators

  • S1 Banach space: A normed complex vector space that is complete under its norm.

  • S2 Bounded operators: is the algebra of bounded linear maps with operator norm ; is a unital Banach algebra.

Spectrum in a Unital Complex Banach Algebra

  • S3 Spectrum: For ; is non–empty; compact; and .

  • S4 Resolvent: is open; is analytic on ; and satisfies the resolvent identity

  • S5 Spectral radius: .

  • S6 Spectral mapping: For any polynomial ; .

Operator–theoretic Breakdown

  • S7 Spectral types on : For ; decompose where point; continuous; residual spectra are defined via injectivity; range density; and surjectivity properties of .

  • S8 Approximate point spectrum: ; always non–empty; contains the boundary of .

  • S9 Riesz projection: If is a rectifiable curve enclosing an isolated spectral set; define

then is a bounded projection commuting with ; with range the corresponding spectral subspace.

–Algebras and Normal Operators

  • S10 –axiom: A complex Banach algebra with involution satisfying ; ; positivity: .

  • S11 Spectral constraints: If then ; if is unitary; .

  • S12 Gelfand duality; commutative case: Every unital commutative –algebra is –isomorphic and isometric to for a compact Hausdorff ; spectrum corresponds to the Gelfand transform.

  • S13 Hilbert–space spectral theorem; schema: For normal ; there exists a unique projection–valued measure on such that

for self–adjoint the integral runs over .

Compact and Fredholm Operators

  • S14 Compact operators: If is compact on ; then is at most countable with only possible accumulation at ; non–zero spectral points are eigenvalues of finite multiplicity.

  • S15 Fredholm theory; outline: is Fredholm if and are finite; index is stable under compact perturbations.

Integration guidance: Each item now states primitives; laws; and immediate consequences; tags G1–G13; N1–N6; C1–C9; T1–T8; S1–S15 remain stable for cross–referencing; swap in your preferred theorem environments as needed.

Turing Completeness: Computational Universality

Definition

A formal system or programming language is said to be Turing Complete if it can simulate a universal Turing machine; that is, it can express any computation that can be described algorithmically, provided sufficient time and memory.

Axioms

  • TC1 There exists a set of states , a tape alphabet including a blank symbol , a transition function , a start state , and a halting state .

  • TC2 A system is Turing Complete if there exists a computable encoding between its syntactic constructs and the configurations of such a universal machine.

  • TC3 For any partial recursive function , there exists a program in the system such that halts with output whenever is defined.

Remarks

Languages such as Lambda Calculus, Post Systems, and Brainf*ck are all Turing Complete. Systems lacking unbounded recursion, loops, or state representation (e.g. finite automata) are not.

NP Completeness: Computational Hardness

Definition

A decision problem is NP–Complete if:

  • NPC1 : there exists a deterministic Turing machine that verifies any certificate of a solution in polynomial time.

  • NPC2 For every , there exists a polynomial–time reduction such that .

Canonical Examples

  • Boolean satisfiability problem (SAT)

  • 3–SAT; CLIQUE; VERTEX–COVER; HAMILTONIAN–CYCLE

  • Travelling Salesman (decision version)

Complexity Classes

  • NPC3 : set of problems solvable in polynomial time.

  • NPC4 : set of problems verifiable in polynomial time.

  • NPC5 : hardest problems in NP, to which every NP problem reduces.

  • NPC6 : problems at least as hard as NP, not necessarily in NP.

Remarks

If any NP–Complete problem has a polynomial–time algorithm, then . This remains one of the central open problems in theoretical computer science.

Incomplete and Incomputability

Definition

A formal system is non–complete with respect to a given domain of discourse if there exist statements within that domain that are neither provable nor refutable within the system.

Gödelian Framework

  • NC1 Any consistent, effectively axiomatized theory that can encode arithmetic is incomplete.

  • NC2 There exists a true arithmetic statement such that neither nor is provable in .

  • NC3 The consistency of cannot be proven within itself if is consistent.

Computational Non Completeness

A system is computationally non–complete if it cannot express all Turing–computable functions. Examples include:

  • Finite automata; regular expressions without back–references.

  • Context–free grammars with out recursion.

  • Domain–specific languages with restricted control flow.

Remarks

Non–completeness may arise by design to ensure safety, predictability, or termination guarantees, such as in total functional programming languages (e.g. Agda, Coq).

Asymptotic Complexity: Growth and Resource Bounds

Big–O Notation

Let be positive functions. Then

Other Asymptotic Classes

Time and Space Complexity

  • Time Complexity: denotes the number of primitive steps executed by an algorithm on input size .

  • Space Complexity: denotes the maximum amount of memory used.

  • Deterministic vs. Non–deterministic complexity classes are denoted and respectively.

Hierarchy and Limits

  • OC1 Time Hierarchy Theorem: If and are time-constructible functions and

then

  • OC2 Space Hierarchy Theorem: If and are space-constructible functions and

then

  • OC3 Polynomial Hierarchy (PH): Generalises via alternating quantifiers; each level and corresponds to bounded alternation of existential and universal quantifiers over polynomial-time predicates.

Summary: Turing completeness expresses the reach of computation; NP–completeness measures computational hardness; non–completeness marks theoretical limitation; O–complexity classifies algorithmic efficiency.

Conclusion

These principles encompass the core structural ideas behind mathematics, including invariance, duality, and mutability, along with logic, proofs, and the nature of mathematical objects.