foundation24 min studyLesson 115 in Mathematics

Partial orders: compare what can be compared

Separate antisymmetry from symmetry, draw cover relations, and distinguish maximal from greatest elements.

On this page

What you’ll understand

  • Verify a partial order and locate incomparable elements.
  • Read a finite Hasse diagram from its cover relations.
  • Distinguish minimal, maximal, least, and greatest elements.

A project has tasks that must precede other tasks, yet two independent tasks may have no prescribed order. A partial order records these constraints without inventing comparisons. The same language describes subset inclusion and divisibility. Its central subtlety is that being impossible to move upward from an element is weaker than being above everything: a partially ordered collection can have several maximal elements while having no greatest element at all.

Antisymmetry allows one direction but forbids distinct ties

A partial order is reflexive, antisymmetric, and transitive. Antisymmetry means that if a precedes b and b precedes a, then they must be the same object. It does not mean that reversing a comparison is always false, since reflexivity requires each object to relate to itself. It also does not require every pair to be comparable. A total order adds that additional requirement, as ordinary numerical order does for real numbers.

Subset inclusion is a model example: every set contains itself, two containments in opposite directions give equality, and nested containments compose. Two distinct singleton sets are incomparable because neither contains the other. Divisibility on positive integers is another partial order, but restricting to positive integers matters: on all nonzero integers, both 2 divides −2 and −2 divides 2 despite unequal elements. A domain choice can determine whether an axiom holds.

a⪯a;(a⪯b ∧ b⪯a)⇒a=b;(a⪯b ∧ b⪯c)⇒a⪯ca\preceq a;\qquad (a\preceq b\ \land\ b\preceq a)\Rightarrow a=b;\qquad (a\preceq b\ \land\ b\preceq c)\Rightarrow a\preceq c
These are reflexivity, antisymmetry, and transitivity. They are required for all elements in the declared set.↔ Scroll wide equations horizontally.

A Hasse diagram keeps only essential upward steps

For a finite partial order, b covers a when a is strictly below b and no element lies strictly between them. A Hasse diagram draws just these cover edges, with larger elements higher on the page. It omits self-loops and comparisons already forced by an upward path. You recover the full order by allowing a zero-length path for equality and an upward path for every strict comparison. Edges are not arbitrary visual connections.

Consider subsets of a two-element set ordered by inclusion. The empty set is at the bottom, the two singletons are in the middle, and the whole set is at the top. There is no middle edge between the singletons because they are incomparable. There is also no direct bottom-to-top edge because a singleton lies in between. The missing direct edge does not mean the bottom and top are incomparable; the upward paths encode their comparison.

a≺b is a cover ⟺ a≺b and there is no c with a≺c≺ba\prec b\text{ is a cover}\ \Longleftrightarrow\ a\prec b\ \text{and there is no }c\text{ with }a\prec c\prec b
The strict relation means a precedes b and a is unequal to b. Finite Hasse diagrams retain these covers and recover other comparisons by transitivity.↔ Scroll wide equations horizontally.

Local extremality differs from a universal bound

An element is maximal if no distinct element lies above it. It is greatest if every element lies below it. A greatest element is maximal and is unique by antisymmetry, but a maximal element need not dominate incomparable neighbors. Minimal and least reverse the directions. Among the subsets of a two-element set after deleting the whole set, both singletons are maximal, while neither is greatest because the other is not below it.

Every nonempty finite partially ordered set has a maximal element: keep moving strictly upward whenever possible, and finiteness prevents an endless chain of distinct elements. The argument does not prove there is only one. Nor does it apply unchanged to infinite orders: the natural numbers have no maximal element. When interpreting prerequisites or resource dependencies, a drawing's height is only a layout choice; the relation and its paths establish the mathematical comparisons.

Divisibility among four integers

Order {1,2,3,6} by divisibility. Identify cover pairs, incomparable pairs, and the least and greatest elements.

  1. Every element divides itself. For distinct elements, the divisibility comparisons are 1 below 2, 3, and 6, and both 2 and 3 below 6.

  2. Remove the comparison 1 below 6 from the edge list because 2 or 3 lies between. The four covers are 1–2, 1–3, 2–6, and 3–6, drawn upward.

  3. Neither 2 divides 3 nor 3 divides 2, so they are incomparable. The element 1 divides every member and is least; every member divides 6, so 6 is greatest.

The result

The Hasse diagram is a diamond with bottom 1 and top 6. The two middle elements are incomparable despite sharing both bounds.

Explore the model

Interactive lab

Interactive lab · reliability

A monotone count is weaker than a full subset order

Fix replica count n and increase unavailable count k. Predict the available-count decrease and the majority threshold crossing. If unavailable replicas form nested subsets, inclusion forces this count trend. The lab tracks counts only, so equal-size different subsets are indistinguishable; it is not a Hasse diagram and cannot identify incomparable failure sets or prove consistency.

A monotone count is weaker than a full subset orderA replica count and a majority threshold, without a consensus protocol. Playback highlights fixed replica states. Having a majority alone does not establish linearizability, durability or failure independence. Current readings: Available replicas: 4 of 5; Majority needed: 3; Majority available?: Yes. All parameters are adjustable with the labeled controls below.Majority available · need 3 of 5Playback highlights fixed statesR1 · upR2 · upR3 · upR4 · upR5 · downCyan mark: majority threshold
AvailableUnavailableMajority threshold
Available replicas
4 of 5
Majority needed
3
Majority available?
Yes
t = 0 s

A replica count and a majority threshold, without a consensus protocol. Playback highlights fixed replica states. Having a majority alone does not establish linearizability, durability or failure independence. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forA numerical summary can preserve order implications while discarding which objects were ordered.

Check your understanding

A finite nonempty partial order has two distinct maximal elements. What follows?

Try it yourself

In {2,3,6} ordered by divisibility, find all minimal elements, all maximal elements, and whether a least or greatest element exists.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

For minimality, look only within this restricted set; the absent integer 1 cannot serve as a lower element.

Compare with the solution

Both 2 and 3 are minimal because neither has a distinct divisor in the set. They are incomparable, so neither is below every element and no least element exists. The only maximal element is 6, and it is also greatest because both 2 and 3 divide it. Minimality and leastness differ because one tests for a local predecessor and the other compares against the entire set.

Take it with you

  • Partial orders allow incomparable pairs and require antisymmetry.
  • A finite Hasse diagram draws covers; paths restore transitive comparisons.
  • Maximal means nothing is strictly above; greatest means everything is below.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.