# General topology General topology — point-set topology — is the axiomatic study of nearness: a topology on a set declares which subsets are *open*, demanding only that arbitrary unions and finite intersections of open sets stay open. From that spare contract flow rigorous definitions of continuity, convergence, compactness, and connectedness — the working vocabulary of [[Mathematical_analysis]], the foundation on which [[Topology]] builds its global invariants, and the substrate every [[Manifold]], [[Phase_space]], and [[Dynamical_system]] stands on. Assembled by Fréchet (metric spaces, 1906), Hausdorff (neighborhood axioms, 1914), and Kuratowski (closure axioms, 1922), the subject is famous both for powerhouse theorems — Tychonoff's compactness of products, Urysohn's metrization — and for a counterexample culture that polices exactly which hypotheses each theorem needs, with excursions deep enough into [[Set_theory]] that some of its questions are provably unsettleable in ZFC, the standard axioms of [[Logic|mathematical logic]]. ## Nearness without numbers The open-set axioms distill what ε–δ arguments in [[Calculus]] actually use: not distances, but membership in neighborhoods. A function is continuous when preimages of open sets are open — no metric mentioned — and a homeomorphism, a continuous bijection with continuous inverse, makes two spaces [[Topology|topologically identical]]: the fabled coffee cup and doughnut. Different metrics can induce the same topology, which is the licensing insight for [[Systems_theory|model-independent]] reasoning: whatever survives re-metrization is structural. Topologies also arise with no metric in sight — the Zariski topology of [[Algebraic_geometry]], where open sets are complements of polynomial zero sets and no two of them are disjoint, or product topologies on infinite configuration sets like the state spaces of a [[Cellular_automaton]]. The axioms are permissive by design; the subject's craft lies in knowing which added hypotheses — countability, separation, compactness — purchase which theorems of [[Geometry]] and [[Mathematical_analysis|analysis]]. ## The vocabulary of limits In metric spaces, sequences detect everything: closures, continuity, compactness. In general spaces they do not, and the repair kit — Moore–Smith nets (1922) and Cartan's filters (1937) — generalizes "eventually" from ℕ to arbitrary directed index sets, restoring the equivalence between continuity and limit-preservation. Countability axioms calibrate how far sequences suffice; separability and second countability keep spaces small enough for constructive [[Algorithm|algorithmic]] handling and for the function spaces of [[Complex_analysis]] and [[Fourier_analysis]]. The payoff structure is visible in [[Probability_theory]]: weak convergence of [[Probability_distribution|probability measures]] is convergence in a specific topology on the space of measures, and Prokhorov's theorem (1956) — tightness equals relative compactness — is general topology doing statistics' heavy lifting, underwriting every central-limit argument that passes to a limit object rather than a limit number in [[Statistics]]. ## Compactness: the finiteness substitute Compactness — every open cover admits a finite subcover — is the axiom-grade replacement for finiteness, and most existence theorems in applied [[Mathematics]] route through it. Heine–Borel identifies the compact subsets of ℝⁿ as the closed and bounded ones; continuous functions on compact sets attain extrema, which is why [[Mathematical_optimization]] posits compact [[Operations_research|feasible sets]], why the direct method of the [[Calculus_of_variations]] and existence theory in [[Optimal_control]] hunt for compactness before minimizers, and why equilibrium proofs in [[Game_theory]] begin by compactifying strategy spaces. Tychonoff's theorem — arbitrary products of compact spaces are compact (1930; general form by Čech, 1937) — powers all of this at infinite scale, and Kelley proved (1950) it is equivalent to the axiom of choice, an exact price tag in [[Set_theory|set-theoretic]] currency. Attractors of a [[Dynamical_system]] are compact invariant sets; ω-limit sets of orbits bounded in [[Phase_space]] are nonempty *because* of compactness — remove it and the guarantees evaporate. ## Separation axioms and the counterexample zoo Between "any topology" and "metric space" runs a graded ladder of separation axioms. Hausdorff's T₂ — distinct points own disjoint neighborhoods — is what makes [[Mathematical_analysis|limits]] unique; the line with a doubled origin satisfies everything a [[Manifold]] needs *except* T₂, which is precisely why manifolds are defined as Hausdorff plus second countable. Urysohn's lemma and metrization theorem (1925) climb the ladder: regular plus second countable suffices for a metric. The subject patrols these hypotheses with a bestiary curated in Steen and Seebach's *Counterexamples in Topology* (1970): the Cantor set — compact, uncountable, totally disconnected, the ur-[[Fractal]] and the address space of [[Chaos_theory|symbolic dynamics]]; the topologist's sine curve — connected but not path-connected; the Sorgenfrey line and long line, each detonating one plausible implication. Deeper still, [[First-order_logic|independence]] results bite: whether every normal Moore space is metrizable cannot be settled in ZFC alone, and Suslin's problem on ordered continua was proved independent (Solovay–Tennenbaum, 1971) — topology's questions escalating into [[Set_theory]] itself. ## Fixed points and equilibria General topology's most exported theorem family says: under the right compactness and convexity, something must stay put. Brouwer (1911): every continuous self-map of a closed ball fixes a point. Kakutani (1941) extended it to set-valued maps, and Nash's 1950 existence proof for equilibria in finite games is Kakutani applied to best-response correspondences — the theorem beneath modern [[Game_theory]] and, via Arrow–Debreu (1954), beneath general equilibrium in [[Economics]]. The metric-space cousin, Banach's contraction principle (1922), is constructive: iterate any contraction and converge geometrically to the unique fixed point — the engine of Picard existence for [[Ordinary_differential_equation|ODEs]], of [[Algorithm|value-iteration]] convergence in [[Dynamic_programming]] (Bellman's operator is a sup-norm contraction, the fact [[Richard_Bellman]]'s method banks on), and of arguments that a Markov chain settles into a stationary [[Probability_distribution|distribution]], bread and butter of [[Probability_theory]]. Schauder (1930) carried fixed points into infinite dimensions, where existence theory for [[Partial_differential_equation|PDEs]] still routes through compactness. ## Topology under dynamics Dynamics is topology in motion. Conjugacy — a homeomorphism intertwining two maps — is the field's notion of "same [[Dynamical_system|system]]," preserving periodic orbits, transitivity, and topological [[Entropy|entropy]] while discarding coordinates; structural stability asks when small perturbations stay conjugate. The now-standard definition of [[Chaos_theory|chaos]] is point-set through and through: topological transitivity plus dense periodic points, from which sensitive dependence follows automatically (Banks et al., 1992). Symbolic dynamics runs on the product topology of sequence spaces — compact, totally disconnected, Cantor-like — where the Curtis–Hedlund–Lyndon theorem (1969) characterizes every [[Cellular_automaton]] as exactly a continuous, shift-commuting map: an entire computational universe specified by two topological adjectives. Poincaré recurrence — [[Henri_Poincaré|Poincaré's]] 1890 discovery that [[Liouville's_theorem_(Hamiltonian)|volume-preserving]] flows in bounded [[Phase_space]] return arbitrarily near their start — needs compactness again, and through it topology reaches [[Statistical_mechanics]], where the return times are longer than the age of the universe but the theorem is airtight. **On the spine:** [[Topology]] · [[Set_theory]] · [[Mathematical_analysis]] · [[Manifold]] · [[Dynamical_system]]. ## Wikipedia : Wikitube **Strict pair:** [Wikipedia](https://en.wikipedia.org/wiki/General_topology) : [Wikitube](https://en.wikitube.io/wiki/General_topology) ## Previous hub tags Hubs: `Systems`. Portals: [[PORTAL_Dynamical_system]], [[PORTAL_Graph_theory]], [[PORTAL_Information_theory]], [[PORTAL_Control_theory]], [[PORTAL_Operations_research]]. --- *Repopulated 2026-08-12 · redlink fill · 0 deletions.*