← Latest papers
🔢 mathematics

Cobham's theorem for the Gaussian integers

Assuming the four exponentials conjecture is not required, this paper proves Hansel and Safer's conjecture that any subset of Gaussian integers recognizable in two multiplicatively independent bases (where at least one is not a root of an integer) must be eventually periodic, thereby generalizing the Cobham-Semenov theorem to Gaussian numerations.

Original authors: Álvaro Bustos-Gajardo, Robbert Fokkink, Reem Yassawi

Published 2026-08-11
📖 5 min read🧠 Deep dive

Original authors: Álvaro Bustos-Gajardo, Robbert Fokkink, Reem Yassawi

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Secret Language of Numbers

Imagine you are trying to teach a robot how to count. You give it a set of rules, like "write down the number 5 as '101'." This is how our brains and computers work: we use a numeration system, a way of turning big numbers into strings of smaller symbols (digits). Usually, we use base-10 (digits 0 through 9), but you could use base-2 (just 0 and 1) or even base-12.

Now, imagine a robot that doesn't just read numbers, but predicts what comes next. If you feed it the string for a number, it spits out a color or a sound. If the robot is simple enough—using a limited number of "states" or "moods" to decide its output—we call the pattern it creates an automatic sequence. These patterns are fascinating because they are complex enough to look random, yet simple enough to be built by a tiny machine.

For a long time, mathematicians have been playing a game with these patterns. They asked: "If a pattern can be generated by two different counting systems (say, base-2 and base-3), does that mean the pattern is actually just a boring, repeating loop?" In the world of regular whole numbers, the answer is a resounding yes. This is a famous rule called Cobham's Theorem. It says that if a pattern is "automatic" in two different bases that don't share a simple relationship, the pattern must be eventually periodic—meaning it settles down into a predictable, repeating rhythm after a while.

But what happens if we leave the straight line of whole numbers and step into a more complex world? What if our numbers aren't just 1, 2, 3, but include imaginary parts, like 1+i1+i or 2i2-i? These are called Gaussian integers. They live on a flat grid (the complex plane) rather than a single line. The big question was: Does Cobham's Theorem still hold here? If a pattern on this grid looks simple in two different "imaginary" counting systems, is it still just a repeating loop?

The Paper's Discovery: Taming the Grid

This paper, titled "Cobham's Theorem for the Gaussian Integers" by Álvaro Bustos-Gajardo, Robbert Fokkink, and Reem Yassawi, answers that question with a definitive yes, but with a few important caveats. The authors prove that if you have a pattern on the grid of Gaussian integers that can be generated by two different "imaginary" counting systems (bases), and those bases are "multiplicatively independent" (meaning one isn't just a power of the other), then the pattern must be eventually periodic.

To understand why this is a big deal, think of the Gaussian integers as a vast, infinite checkerboard. Usually, patterns on this board can be wild and chaotic. The authors show that if you try to force a pattern to be "simple" (automatic) using two different, unrelated ways of counting on this board, the universe forces the pattern to collapse into a neat, repeating tile. It's as if the grid has a hidden law: you can't have a truly complex, non-repeating pattern that satisfies two different simple rules at the same time.

However, the paper also draws a sharp line in the sand. The rule only works if at least one of the counting bases is not a "root of an integer."

  • The Exception: If the base is a root of an integer (like a number that, when multiplied by itself a few times, becomes a normal whole number), then the rule breaks. In this specific case, you can create patterns that are simple in two different bases but never settle into a repeating loop. The authors prove that these "non-repeating" patterns exist and are unavoidable if you pick these special bases.
  • The Proof: The authors didn't just guess or simulate this; they provided a rigorous mathematical proof. They used a clever trick involving "pumping lemmas" (a tool from computer science that finds loops in machines) and "Dirichlet approximation" (a way of finding numbers that are very close to each other) to show that if the bases aren't special roots, the pattern must repeat.

Why It Matters (Without the Jargon)

Before this paper, mathematicians suspected this rule was true for Gaussian integers, but they needed a massive, unproven assumption from deep number theory (called the "four exponentials conjecture") to make the math work. That assumption was like a bridge made of clouds; it might hold, but no one was sure.

This paper's main achievement is that it removed the need for that shaky bridge. They proved the result using only solid, established math. They showed that the "cloud bridge" wasn't necessary after all. They also settled a specific conjecture made by Hansel and Safer, confirming that for the most common types of Gaussian counting systems (those using natural numbers as digits), the pattern is always eventually periodic.

In short, the paper tells us that the chaotic world of imaginary numbers has a hidden order. If you try to describe a pattern on this grid using two different, unrelated counting languages, the pattern will inevitably reveal its true nature: a simple, repeating dance. The only time this dance gets messy is if you choose very specific, "special" counting bases, which the authors have now fully identified and categorized.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →