In my book, the author defines posets formally in the following way:

Let $P$ be a set, and let $\le$ be a relationship on $P$ so that,

$a$. $\le$ is reflective.

$b$. $\le$ is transitive.

$c$. $\le$ is antisymmetric.

Say for $a$, does this merely mean that if some element $x\in P$, $x$ should always have the same relation to itself? and for $b$ if $x$ has the relation to $y$ and $y$ has the relation to $z$, this implies that $x$ has the relation to $z$?

Moreover, when trying to determine if a something is a poset,do I just have to determine if such a relationship exists? And that relationship is not necessarily the usual meaning of "$\le$"


Yes, to each of your questions. A poset is a set $P$ and a partial-order relation defined on the set; when trying to determine if a relation and the set on which it is defined is a poset, you need to check each of the defining properties of a partial order relation with respect to the set. That is true when testing any relationship, not always denoted by $\leq$.

(You need to also show that the relation is antisymmetric, if it is to be a poset.)

In the following, we use the notation "$\,\leq\,$" to represent any partial order relation, not just the narrower/usual meaning of "$\,\leq\,$" ('is less than or equal to'), which also defines a poset on suitable sets. There are many relations which, together with the sets on which they are respectively defined, form a poset. You'll likely encounter some such examples in you studies and/or in problem sets.

A poset consists a set $P$ and a relation on the set. A relation on the set form a poset if and only if:

  • $\forall a \in P, \;a\leq a\;$ (reflexity);
  • $\forall a, b, c \in P,\;$ whenever $a \leq b$ and $b \leq c$, then this implies $a \leq c\;$ (transitivity);
  • $\forall a, b \in P,\;$ whenever $a \leq b$ and $b\leq a$, then this implies $a = b\;$ (antisymmetry).

So given a relation and a set on which it is defined, in order to determine whether you have a poset, you need to confirm that each of these three properties hold for all $a, b, c$ in the set.

Put differently, if there are no $a \in P$ for which reflexivity fails, then the relation reflexive; if there are no $a, b, c \in P$ such that transitivity fails, then the relation is transitive. And if there are no $a, b\in P$ in the set such that antisymmetry fails, then the relation is antisymmetric.

  • $\begingroup$ Thanks for the reassurance! $\endgroup$
    – MITjanitor
    Mar 10 '13 at 20:56
  • $\begingroup$ You're welcome! $\endgroup$
    – amWhy
    Mar 10 '13 at 20:59
  • $\begingroup$ Are you good with this now, Calc1DropOut? Any lingering questions? $\endgroup$
    – amWhy
    Mar 10 '13 at 21:32
  • $\begingroup$ Sorry, forgot to accept. I'm good now, thanks! $\endgroup$
    – MITjanitor
    Mar 10 '13 at 21:34
  • $\begingroup$ Great! good luck... $\endgroup$
    – amWhy
    Mar 10 '13 at 21:34

Your statements are correct.

On your last question: A poset is a pair ($P$, $R$), where $P$ is a set and $R$ is a relation on $P$ (which must have properties a, b and c). So the relation is a part of the poset, there is no freedom to choose it.

Hence to determine if something is a poset, you don't have to determine if a suitable order relation exists (it always does), but if the set together with the given relation is a poset. The relation should always be clear from the context, in particular if the relation symbol is not ''$\leq$''.

  • $\begingroup$ To emphasise the last point, one notes that $=$ always makes a set into a poset. $\endgroup$
    – Zhen Lin
    Mar 10 '13 at 21:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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