Table of Contents
Fetching ...

On noncontinuous bisymmetric strictly monotone operations

Gergely Kiss

TL;DR

The paper investigates regularity of bisymmetric, strictly increasing binary operations without assuming reflexivity. It constructs discontinuous, weighted-quasi-sum type operations $F(x,y)=f^{-1}(\alpha f(x)+\beta f(y))$ on an interval using a Cantor-type perfect set, showing that bisymmetry and monotonicity do not imply continuity outside the reflexive setting, and extends the construction to $n$-ary operations. A two-point reflexivity result is proved: if $F(a,a)=a$ and $F(b,b)=b$ for $a<b$, then $F$ is continuous on $[a,b]$ and has a quasi-arithmetic representation there. The work thus reveals a sharp dichotomy between local rigidity under two-point reflexivity and global irregularity in the non-reflexive regime, and it raises open problems for global continuity and multivariate characterization of bisymmetric, monotone operations.

Abstract

In this paper, we construct a bisymmetric, strictly increasing, symmetric binary operation $F$ on an interval which is not continuous. This answers a natural question in the study of bisymmetric and mean-type operations by showing that continuity may fail even for non-reflexive operations of the form \[ F(x,y)=f^{-1}\bigl(αf(x)+βf(y)\bigr), \] where $f$ is a bijection between an interval and a perfect, nowhere dense fractal-type set, and $α,β>0$ with $α+β\neq 1$. As a consequence, we also obtain a noncontinuous, associative, strictly increasing, symmetric operation on an interval. We generalize these constructions to the multivariate case. We also prove a complementary result: if a symmetric, bisymmetric, strictly increasing operation is reflexive at the endpoints of an interval, then it must be continuous and coincide with a quasi-arithmetic mean on that interval.

On noncontinuous bisymmetric strictly monotone operations

TL;DR

The paper investigates regularity of bisymmetric, strictly increasing binary operations without assuming reflexivity. It constructs discontinuous, weighted-quasi-sum type operations on an interval using a Cantor-type perfect set, showing that bisymmetry and monotonicity do not imply continuity outside the reflexive setting, and extends the construction to -ary operations. A two-point reflexivity result is proved: if and for , then is continuous on and has a quasi-arithmetic representation there. The work thus reveals a sharp dichotomy between local rigidity under two-point reflexivity and global irregularity in the non-reflexive regime, and it raises open problems for global continuity and multivariate characterization of bisymmetric, monotone operations.

Abstract

In this paper, we construct a bisymmetric, strictly increasing, symmetric binary operation on an interval which is not continuous. This answers a natural question in the study of bisymmetric and mean-type operations by showing that continuity may fail even for non-reflexive operations of the form where is a bijection between an interval and a perfect, nowhere dense fractal-type set, and with . As a consequence, we also obtain a noncontinuous, associative, strictly increasing, symmetric operation on an interval. We generalize these constructions to the multivariate case. We also prove a complementary result: if a symmetric, bisymmetric, strictly increasing operation is reflexive at the endpoints of an interval, then it must be continuous and coincide with a quasi-arithmetic mean on that interval.
Paper Structure (11 sections, 14 theorems, 92 equations)

This paper contains 11 sections, 14 theorems, 92 equations.

Key Result

Theorem 2.1

A function $F:I^2\to I$ is continuous, reflexive, partially strictly monotone, symmetric, and bisymmetric if and only if there exists a continuous strictly monotone function $f:I\to\mathbb{R}$ such that In this case $F$ is called a quasi–arithmetic mean generated by $f$.

Theorems & Definitions (28)

  • Theorem 2.1: Aczél, 1948; quasi–arithmetic means
  • Theorem 2.2: Aczél–Dhombres; quasi–sums
  • Theorem 2.3: Aczél
  • Theorem 2.4: Burai–Kiss–Szokol, 2021
  • Theorem 2.5: Burai–Kiss–Szokol, 2023
  • Lemma 3.1
  • proof
  • Lemma 3.2
  • proof
  • Theorem 3.3
  • ...and 18 more