A Hierarchy of Geometric Constructions
MohammadJavad Maarefvand
TL;DR
The paper investigates the limits of finite geometric constructions and introduces a hierarchical ladder—from ruler-and-compass to origami, neusis, conic sections, mechanical linkages, and select transcendental curves—that expands constructible reach while acknowledging fundamental bounds. Using field theory, it shows that constructible numbers under ruler-and-compass live in towers of quadratic extensions of $Q$ with degrees $[K_n:Q]=2^m$, that origami/neusis allow degrees dividing $2^m 3^n$, and that Kempe’s Universality via mechanical linkages yields all algebraic numbers. It further argues that finite sets of transcendental curves can construct specific constants like $pi$ and $e$, but cannot cover all computable numbers, leaving uncomputable reals beyond reach. The resulting hierarchy clarifies the boundaries between constructible, computable, and uncomputable numbers, highlighting how each added tool augments capability but cannot override fundamental theoretical limits.
Abstract
This article explores the limits of geometric construction using various tools, both classical and modern. Starting with ruler and compass constructions, we examine how adding methods such as origami, marked rulers (neusis), conic sections, mechanical linkages, and certain transcendental curves expands the range of constructible numbers. These methods allow the construction of increasingly complex numbers from square roots, to cube roots, to all algebraic numbers, and in some cases to specific transcendental constants like pi and e. We explain how field theory gives a precise way to understand these constructions, and how computability theory shows that no finite geometric method can produce every computable number. In particular, no construction process that can be described step by step can reach uncomputable numbers. The article concludes by presenting a hierarchy of geometric methods, showing how each step increases what is possible, while still leaving strict theoretical limits in place.
