You are here
Homequantale
Primary tabs
quantale
A quantale $Q$ is a set with three binary operations on it: $\wedge,\vee$, and $\cdot$, such that
1. $(Q,\wedge,\vee)$ is a complete lattice (with $0$ as the bottom and $1$ as the top), and
2. 3. $\cdot$ distributes over arbitrary joins; that is, for any $a\in Q$ and any subset $S\subseteq Q$,
$a\cdot\big(\bigvee S\big)=\bigvee\{a\cdot s\mid s\in S\}\quad\mbox{ and }\quad% \big(\bigvee S\big)\cdot a=\bigvee\{s\cdot a\mid s\in S\}.$
It is sometimes convenient to drop the multiplication symbol, when there is no confusion. So instead of writing $a\cdot b$, we write $ab$.
The most obvious example of a quantale comes from ring theory. Let $R$ be a commutative ring with $1$. Then $L(R)$, the lattice of ideals of $R$, is a quantale.
Proof.
In addition to being a (complete) lattice, $L(R)$ has an inherent multiplication operation induced by the multiplication on $R$, namely,
$IJ:=\{\sum_{{i=1}}^{n}r_{i}s_{i}\mid r_{i}\in I\mbox{ and }s_{i}\in J\mbox{, }% n\in\mathbb{N}\},$ 
making it into a semigroup under the multiplication.
Now, let $S=\{I_{i}\mid i\in N\}$ be a set of ideals of $R$ and let $I=\bigvee S$. If $J$ is any ideal of $R$, we want to show that $IJ=\bigvee\{I_{i}J\mid i\in N\}$ and, since $R$ is commutative, we would have the other equality $JI=\bigvee\{JI_{i}\mid i\in N\}$. To see this, let $a\in IJ$. Then $a=\sum r_{i}s_{i}$ with $r_{i}\in I$ and $s_{i}\in J$. Since each $r_{i}$ is a finite sum of elements of $\bigcup S$, $r_{i}s_{i}$ is a finite sum of elements of $\bigcup\{I_{i}J\mid i\in N\}$, so $a\in\bigvee\{I_{i}J\mid i\in N\}$. This shows $IJ\subseteq\bigvee\{I_{i}J\mid i\in N\}$. Conversely, if $a\in\bigvee\{I_{i}J\mid i\in N\}$, then $a$ can be written as a finite sum of elements of $\bigcup\{I_{i}J\mid i\in N\}$. In turn, each of these additive components is a finite sum of products of the form $r_{k}s_{k}$, where $r_{k}\in I_{i}$ for some $i$, and $s_{k}\in J$. As a result, $a$ is a finite sum of elements of the form $r_{k}s_{k}$, so $a\in IJ$ and we have the other inclusion $\bigvee\{I_{i}J\mid i\in N\}\subseteq IJ$.
Finally, we observe that $R$ is the multiplicative identity in $L(R)$, as $IR=RI=I$ for all $I\in L(R)$. This completes the proof. ∎
Remark. In the above example, notice that $IJ\leq I$ and $IJ\leq J$, and we actually have $IJ\leq I\wedge J$. In particular, $I^{2}\leq I$. With an added condition, this fact can be characterized in an arbitrary quantale (see below).
Properties. Let $Q$ be a quantale.
1. Multiplication is monotone in each argument. This means that if $a,b\in Q$, then $a\leq b$ implies that $ac\leq bc$ and $ca\leq cb$ for all $c\in Q$. This is easily verified. For example, if $a\leq b$, then $ac\vee bc=(a\vee b)c=bc$, so $ac\leq bc$. So a quantale is a partially ordered semigroup, and in fact, an lmonoid (an lsemigroup and a monoid at the same time).
2. If $1=1^{{\prime}}$, then $ab\leq a\wedge b$: since $a\leq 1$, then $ab\leq a1=a1^{{\prime}}=a$; similarly, $b\leq ab$. In particular, the bottom $0$ is also the multiplicative zero: $a0\leq a\wedge 0=0$, and $0a=0$ similarly.
3. Actually, $a0=0a=0$ is true even without $1=1^{{\prime}}$: since $a\varnothing=\{ab\mid b\in\varnothing\}=\varnothing$ and $0:=\bigvee\varnothing$, we have $a0=a\bigvee\varnothing=\bigvee a\varnothing=\bigvee\varnothing=0$. Similarly $0a=0$. So a quantale is a semiring, if $\vee$ is identified as $+$ (with $0$ as the additive identity), and $\cdot$ is again $\cdot$ (with $1^{{\prime}}$ the multiplicative identity).
4. Viewing quantale $Q$ now as a semiring, we see in fact that $Q$ is an idempotent semiring, since $a+a=a\vee a=a$.
5. Now, view $Q$ as an isemiring. For each $a\in Q$, let $S=\{1^{{\prime}},a,a^{2},\ldots\}$ and define $a^{*}=\bigvee S$. We observe some basic properties

$1^{{\prime}}+aa^{*}=a^{*}$: since $1^{{\prime}}\vee(a\bigvee S)=1^{{\prime}}\vee(\bigvee\{a1^{{\prime}},aa,aa^{2}% ,\ldots\})=\bigvee\{1^{{\prime}},a,a^{2},\ldots\}=\bigvee S=a^{*}$

$1^{{\prime}}+a^{*}a=a^{*}$ as well

if $ab\leq b$, then $a^{*}b\leq b$: by induction on $n$, we have $a^{n}b\leq b$ whenever $a\leq b$, so that $a^{*}b=\bigvee\{a^{n}b\mid n\in\mathbb{N}\cup\{0\}\}\leq b$.

similarly, if $ba\leq b$, then $ba^{*}\leq b$
All of the above properties satisfy the conditions for an isemiring to be a Kleene algebra. For this reason, a quantale is sometimes called a standard Kleene algebra.

6. Call the multiplication idempotent if each element is an idempotent with respect to the multiplication: $aa=a$ for any $a\in Q$. If $\cdot$ is idempotent and $1=1^{{\prime}}$, then $\cdot=\wedge$. In other words, $ab=a\wedge b$.
Proof.
As we have seen, $ab\leq a\wedge b$ in the 2 above. Now, suppose $c\leq a\wedge b$. Then $c\leq a$ and $c\leq b$, so $c=c^{2}\leq cb\leq ab$. So $ab$ is the greatest lower bound of $a$ and $b$, i.e., $ab=a\wedge b$. This also means that $ba=b\wedge a=a\wedge b=ab$. ∎
7. In fact, a locale is a quantale if we define $\cdot:=\wedge$. Conversely, a quantale where $\cdot$ is idempotent and $1=1^{{\prime}}$ is a locale.
Proof.
If $Q$ is a locale with $\cdot=\wedge$, then $aa=a\wedge a=a$ and $a1=a\wedge 1=a=1\wedge a=1a$, implying $1=1^{{\prime}}$. The infinite distributivity of $\cdot$ over $\vee$ is just a restatement of the infinite distributivity of $\wedge$ over $\vee$ in a locale. Conversely, if $\cdot$ is idempotent and $1=1^{{\prime}}$, then $\cdot=\wedge$ as shown previously, so $a\wedge(\bigvee S)=a(\bigvee S)=\bigvee\{as\mid s\in S\}=\bigvee\{a\wedge s% \mid s\in S\}$. Similarly $(\bigvee S)\wedge a=\bigvee\{s\wedge a\mid s\in S\}$. Therefore, $Q$ is a locale. ∎
Remark. A quantale homomorphism between two quantales is a complete lattice homomorphism and a monoid homomorphism at the same time.
References
 1 S. Vickers, Topology via Logic, Cambridge University Press, Cambridge (1989).
Mathematics Subject Classification
06F07 no label found Forums
 Planetary Bugs
 HS/Secondary
 University/Tertiary
 Graduate/Advanced
 Industry/Practice
 Research Topics
 LaTeX help
 Math Comptetitions
 Math History
 Math Humor
 PlanetMath Comments
 PlanetMath System Updates and News
 PlanetMath help
 PlanetMath.ORG
 Strategic Communications Development
 The Math Pub
 Testing messages (ignore)
 Other useful stuff
Recent Activity
new question: Prove a formula is part of the Gentzen System by LadyAnne
Mar 30
new question: A problem about Euler's totient function by mbhatia
new problem: Problem: Show that phi(a^n1), (where phi is the Euler totient function), is divisible by n for any natural number n and any natural number a >1. by mbhatia
new problem: MSC browser just displays "No articles found. Up to ." by jaimeglz
Mar 26
new correction: Misspelled name by DavidSteinsaltz
Mar 21
new correction: underlinetypo by Filipe
Mar 19
new correction: cocycle pro cocyle by pahio
Mar 7
new image: plot W(t) = P(waiting time <= t) (2nd attempt) by robert_dodier
new image: expected waiting time by robert_dodier
new image: plot W(t) = P(waiting time <= t) by robert_dodier