Can a node have more than two children?

Yes. Ordinal trees do not impose a binary branching limit.

Is an ordinal tree sorted by value?

Not necessarily. Child order can reflect document sequence or another application-defined arrangement.

Can two children have the same label?

Yes, if the application permits it. Position and node identity can distinguish them even when their displayed labels match.

AI Glossary / Data engineering and architecture / Intermediate

Ordinal tree

Definition

An ordinal tree is a rooted tree in which each node's children have a defined order. A node can have any finite number of children, and its first, second or later child is identified by position in that sequence.

What it means

A tree describes a hierarchy. It has one root, and every other node has one parent. Adding child order makes the sequence among siblings part of the structure, rather than an incidental detail of a drawing.

Consider a document with three sections: introduction, instructions and conclusion. Their shared parent identifies the document, while their order identifies how they appear. Reordering the sections changes the ordered hierarchy even if the same section objects remain present.

Ordinal here does not mean that labels must be numbers. It also does not mean binary search tree. A node may have more than two children, and their positions need not follow a comparison of their stored values.

How it works

A basic implementation stores each node's children as a sequence. Navigation can identify a parent, a particular child, a preceding or following sibling and the descendants of a node. The representation must preserve both parent relationships and sibling order.

Traversal rules provide predictable ways to visit the hierarchy. In preorder, visit a node before its children, visiting the children in their defined sequence. In postorder, finish the children before visiting the parent. These rules produce different sequences from the same tree.

Large trees can use specialized compact representations. Research on succinct trees studies how to encode structure with little space while retaining efficient navigation. The topology and labels are separate concerns: a compact structural representation does not automatically store all document text or other values.

Updates need to preserve the intended order. Inserting a new first child changes sibling positions, so a position is not necessarily a stable identifier for the same object over time.

Three practical examples

A document editor preserves the order of chapters and paragraphs within a hierarchy.
A parser represents a structured expression whose child sequence affects how the expression is interpreted.
A navigation tool presents an explicitly arranged hierarchy of topics, preserving the editor's chosen sibling order instead of sorting labels alphabetically.

Think of it this way

Think of a family of nested folders in a binder, with a specified sequence of dividers at each level. Nesting tells you which section contains a page; sequence tells you what comes before or after it.

The analogy is limited because ordinary computer folders may display different sorts without changing their underlying relationships. An ordinal tree makes the chosen sibling order part of the represented structure itself.

Worked example

Following two traversals

Let root R have children A, B and C in that order. Let B have children D and E in that order. A, C, D and E have no children. There are six nodes in total.

A preorder visit produces R, A, B, D, E, C: visit the root, complete A, then B and its descendants, then C. A postorder visit produces A, D, E, B, C, R because every node follows its children.

Now exchange A and C in R's child sequence while leaving B between them. The parent relationships are unchanged, but preorder becomes R, C, B, D, E, A. That difference is meaningful in an ordinal tree.

If a new child X is inserted before B, B's position shifts. A reference meaning “the second child” may now identify X rather than B. Applications that need lasting identity should distinguish identifiers from changing positions.

Illustrative case study

A fictional publishing team builds an AI-assisted guide editor. Its storage export records every section's parent but omits sibling positions. On import, the editor alphabetically sorts the children, placing the conclusion before the introduction.

The team adds an explicit ordering representation and tests nested rearrangements. Stable section identifiers continue to refer to the same content, while a separate order field or sequence determines presentation.

The AI can suggest a rearrangement, but the approved document structure must retain the intended sequence. The incident shows that preserving the set of sections and their parents is insufficient when readers depend on an ordered narrative.

Common misconceptions and mistakes

An ordinal tree has numeric labels. Labels can be words or other values; ordinal refers to ordered children.
Every tree drawing fixes sibling order. Some applications treat drawings as interchangeable because order is irrelevant; ordinal trees do not.
Child position is permanent identity. Insertions and moves can change a node's rank among siblings.

Business use

Ordered hierarchies occur in document workflows, interfaces and structured content. They are relevant whenever changing sequence changes meaning, presentation or subsequent processing.

Teams integrating AI-generated content should preserve the hierarchy and order expected by the destination format. A summary that reorders sections may be acceptable as a new summary, but it is not necessarily a faithful reconstruction of the original document.

For very large structures, representation choices can influence memory and navigation costs. Start with the operations the application needs, such as subtree movement or finding a particular child. A compact static representation may require different update tradeoffs from an ordinary editable sequence of nodes.

Check your understanding

Two document trees have identical parent relationships, but one reverses the children of a chapter. Are they necessarily the same ordinal tree?

Answer: No. Sibling order is part of the structure. Reversing it can change the represented ordered hierarchy.

Related terms

Parent node
Rooted tree
ordered tree
sibling
preorder traversal
postorder traversal
succinct data structure
document object model

Learn next

Read tree traversal and succinct data structure. Draw a six-node hierarchy with one node having three children, list its preorder and postorder sequences, then predict what changes after one sibling is moved.

Data structure

Sources

Davoodi, Raman and Satti, On Succinct Representations of Binary Trees: https://arxiv.org/abs/1410.4963. Primary research explicitly defines ordinal trees as rooted trees with ordered children.

Navarro and Sadakane, Fully-Functional Static and Dynamic Succinct Trees: https://arxiv.org/abs/0905.0768. Primary research on compact representations and navigation.

Meng He, Succinct Indexes, Chapter 5: https://cs.uwaterloo.ca/~imunro/He.pdf. Primary indexed discussion of ordinal-tree representation. The traversal example and publishing case are original illustrations.

Previous: Ordinal scale Next: Ordinal utility function