Let a (rational) Puiseux monoid be a non-trivial submonoid of the non-negative rational numbers under (the usual operation of) addition. It is not difficult to show that, if $f \colon H \to K$ is a (monoid) homomorphism between Puiseux monoids, then there exists a rational number $r \ge 0$ such that $f(x) = r x$ for every $x \in H$.

Proof. Set $H^+ := H \setminus \{0\}$ and pick $p, q \in H^+$ (since $f$ maps $0$ to $0$, we can focus on the positive elements). There then exist positive integers $a$, $b$, $c$, and $d$ such that $p = a/b$ and $q = m/n$. It follows that $$ mb f(p) = f(mbp) = f(ma) = f(naq) = na f(q), $$ and hence $f(p)/p = f(q)/q$. That is, the function $H^+ \to \mathbb Q_{\ge 0} \colon x \mapsto f(x)/x$ is constant. []

Question. The result, along with a slighly more complicated proof, appears as Proposition 3.2(1) in [F. Gotti, Puiseux monoids and transfer homomorphisms, J. Algebra 516 (2018), 95–114], but I feel it must be older. Do you know of any reference that supports my feeling?

Note that the case of numerical monoids (that is, submonoids of $(\mathbb N, +)$ with finite complement in $\mathbb N$) is covered by Theorem 3.2 in [J.C. Higgins, Representing $N$-semigroups, Bull. Austral. Math. Soc. 1 (1969), 115–125], as I learned from Benjamin Steinberg's answer in another thread some time ago.


1 Answer 1


This seems to be Corollary 4 of the paper Morio Sasaki, Takayuki Tamura, Positive rational semigroups and commutative power joined cancellative semigroups without idempotent Czechoslovak Mathematical Journal, Vol. 21 (1971), No. 4, 567–576 that I mentioned in the comments to your previous question.

  • $\begingroup$ Thanks! Corollary 4 in Sasaki and Tamura's paper does indeed imply the result in the OP, after noticing that a monoid homomorphism between Puiseux monoids must either be identically zero or map positive rationals to positive rationals (and hence restrict, in the terminology of Sasaki and Tamura, to a semigroup homomorphism of positive rational semigroups). Before asking here, I had the term "rational semigroup" in mind but couldn't remember why... $\endgroup$ Commented Feb 13 at 3:29
  • $\begingroup$ I am confused. I think what they mean by positive rational semigroup is exactly puiseux monoids (except without the identity) and Cor 4 says every homomorphism between them is given by multiplication by a rational $\endgroup$ Commented Feb 13 at 3:41
  • $\begingroup$ I think that, by the term "positive rational semigroup", Sasaki and Tamura mean a (non-empty) subsemigroup of the positive rational numbers under addition. Corollary 4 in their paper says that any homomorphism between positive rational semigroups is in fact an isomorphism (of the form $x \mapsto rx$, where $r$ is a positive rational number). This is not the case with Puiseux monoids, but only because it is missing the zero homomorphism. $\endgroup$ Commented Feb 13 at 3:47
  • $\begingroup$ So, to recover the result in the OP from Sasaki and Tamura's corollary, I think one should first argue (as trivial as it may be) that a monoid hom $f \colon H \to K$ between Puiseux monoids restricts to a sgrp hom between the "positive cones" of $H$ and $K$, unless $f$ is identically zero. $\endgroup$ Commented Feb 13 at 3:56
  • $\begingroup$ @SalvoTringali, yes. This is easy to see though since any two elements have a common multiple and so if one element goes to 0 they all do. $\endgroup$ Commented Feb 13 at 4:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.