Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $S$ be the abelian monoid with elements $a_{n,m}$ where $n \in \mathbb{N}$ and $$\begin{cases} m=0 & \text{ if } n=0 \text{ or }1,\\ m \in \mathbb{Z} & \text{ if } n=2,\\ m \in \mathbb{Z}/2\mathbb{Z} & \text{ if } n \geq 3. \end{cases} $$

The semigroup operation is given by $$a_{n,m}+a_{n',m'} = a_{n+n',m+m'}$$ where $m+m'$ is to be computed in $\mathbb{Z}$ if $n+n' \le 2$ and in $\mathbb{Z}/2\mathbb{Z}$ if $n+n'\ge 3$ (if $n=2$ and $n' \ge 1$ for example, then $m$ is to be interpreted mod 2.)

I am trying to calculate the Grothendieck group of this monoid (this is Ex 1.1.7 in Rosenberg's book on Algebraic K-theory)

It seems the best approach to take is the following: The Grothendieck group $G(S)$ is the equivalence classes of of pairs $(x,y)$ with $x,y \in S$ where $(x,y) \sim (u,v)$ if and only if there is some $t \in S$ such that $$x+v+t = u+y+t \text{ in } S$$

Thus suppose we have two pairs $$(a_{m,n},a_{m',n'}) \sim (a_{x,y},a_{x',y'})$$ This means there is some $a_{\alpha,\beta}$ such that $$a_{m,n}+a_{x',y'}+a_{\alpha,\beta} = a_{m',n'}+a_{x,y}+a_{\alpha,\beta}$$

Now using the semigroup operation on the first subscript gives immediately that $m+x'=m'+x$. The second seems to be a bit tricker. If everything ends up less than $2$ or if everything ends up greater than 2 then it is ok, but it seems like a fair bit of casework to consider all the possibilities in between.

Am I missing something obvious here, or is it just a bit of grunt work, do the calculations type question? Is this the right way to approach the question (basically I was trying to determine all the isomoprhism classes).

share|cite|improve this question
Is $\mathbb{Z}/2$ the integers modulo $2$ (in which case, $\mathbb{Z}/(2)$ or $\mathbb{Z}/2\mathbb{Z}$ is better), or the half-integers ($\frac{a}{2}$ with $a\in\mathbb{Z}$, in which case $\mathbb{Z}[\frac{1}{2}]$ is better)? – Arturo Magidin Jan 19 '12 at 17:58
Your addition is not well defined; you miscopied the definition, which has "if $n\geq 3$" in the third option. – Arturo Magidin Jan 19 '12 at 18:00
@Arturo - Thanks, I have edited the question – Juan S Jan 19 '12 at 21:15
@Jack - opps missed that - I had missed another $\ge 3$ as well. Thanks for the edit. – Juan S Jan 20 '12 at 0:14
up vote 2 down vote accepted

There is a two-step process to compute the Grothendieck group that is usually a bit faster. Let $S$ be any abelian monoid, and define an equivalence relation $\approx$ on $S$ by $$ x \approx y \quad\Leftrightarrow\quad (\exists z\in S)(x+z = y+z). $$ Then $\approx$ is a congruence relation, so the quotient $S/{\approx}$ is a monoid. It is not hard to prove the following theorem:

Theorem. The quotient $S/{\approx}$ is isomorphic to the image of $S$ in $G(S)$.

It follows from this theorem that $G(S)$ is isomorphic to $G(S/{\approx})$, but the latter is usually easier to compute. In particular, if you can find any embedding of $S/{\approx}$ into an abelian group, then $G(S/{\approx})$ is isomorphic to the subgroup generated by the elements of $S/{\approx}$.

In the example you gave, the equivalence relation $\approx$ is obviously defined by $$ a_{n,m} \approx a_{n',m'} \quad\Leftrightarrow\quad n=n'\text{ and }m \equiv m'\:(\mathrm{mod}\;2) $$ Then the quotient consists of elements $A_{n,m}$, where $n\in\mathbb{N}$ and $$ \begin{cases}m=0 & \text{if }n = 0\text{ or }1, \\ m\in\mathbb{Z}/2\mathbb{Z} & \text{if } n\geq 2.\end{cases} $$ with operation defined by $$ A_{n,m} + A_{n',m'} = A_{n+n',m+m'} $$ where $m+m'$ is always computed in $\mathbb{Z}/2\mathbb{Z}$. This is obviously isomorphic to the submonoid of $\mathbb{Z} \oplus \mathbb{Z}/2\mathbb{Z}$ consisting of the elements $\{(0,0),(1,0),(2,0),(2,1),(3,0),(3,1),\ldots\}$. These elements generate $\mathbb{Z}\oplus\mathbb{Z}/2\mathbb{Z}$, so $G(S)$ is isomorphic to $\mathbb{Z}\oplus\mathbb{Z}/2\mathbb{Z}$.

share|cite|improve this answer
Thanks Jim - this was really bugging me! – Juan S Feb 3 '12 at 1:51

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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