A definition states how a term is to be understood within a particular context, domain or system.
Dictionary definitions, legal definitions, scientific definitions, operational definitions, technical specifications and classroom definitions can all describe the same word differently without necessarily contradicting one another. Their scope and purpose matter.
Quick answer: how should definitions be categorised?
- Purpose: descriptive, operational, legal, technical, educational, stipulative?
- Scope: universal, domain-specific, local, temporary?
- Authority: dictionary, regulator, standards body, institution, author?
- Precision: broad, bounded, formal, quantitative?
- Context: which discourse or system does it belong to?
- Inclusions: what clearly belongs?
- Exclusions: what clearly does not?
- Boundary cases: where does ambiguity remain?
- Version: which wording or edition applies?
- Change: how and why has the definition evolved?
This article applies How to Categorise Anything to definitions as controlled meaning statements.
1. Separate word from definition
A term can have several definitions across domains and time.
2. Separate definition from example
An example illustrates membership; it does not by itself state the boundary.
3. Separate definition from description
A description can add detail without determining what counts as the thing.
4. Lexical definitions report common usage
They describe how a language community typically uses a word.
5. Stipulative definitions assign meaning
An author or project may define a term for one bounded purpose.
6. Operational definitions connect terms to measurement
They specify how a concept will be observed, tested or counted.
7. Legal definitions derive force from authority
The meaning of a term in legislation may differ from everyday language and should be tied to jurisdiction and effective date.
8. Technical definitions support interoperability
Standards and engineering systems require stable meanings so independent actors can coordinate.
9. Scientific definitions can evolve
Improved theory, evidence and measurement can change classification boundaries.
10. Educational definitions can simplify
A Primary-level explanation may deliberately omit edge cases that matter at advanced levels.
11. Simplification should preserve direction
A beginner definition should not teach something that later must be unlearned entirely.
12. Scope must be explicit
“Energy”, “risk”, “intelligence” and “market” can mean different things in physics, finance, psychology and ordinary speech.
13. Domain-specific definitions are not universal definitions
Use labels such as “in this standard” or “for this study” where appropriate.
14. Authority should be recorded
Who defines the term determines where the wording has force.
15. Authority does not guarantee usefulness everywhere
A legally authoritative definition may be poor for scientific explanation outside that legal purpose.
16. Necessary conditions define requirements
If a condition must be present for category membership, it is necessary.
17. Sufficient conditions define guarantees
If a condition ensures membership, it is sufficient under the stated scheme.
18. Definitions can use prototypes
Some categories are understood through typical examples rather than one perfect checklist.
19. Definitions can use family resemblance
Members may share overlapping similarities without one feature common to all.
20. Definitions can be recursive
Some mathematical and computational objects are defined partly in terms of smaller instances of themselves.
21. Definitions can be relational
A role such as parent, owner or neighbour depends on a relationship rather than an intrinsic property.
22. Boundary cases should be preserved
Ambiguity is sometimes a feature of the concept, not a defect in the definition.
23. Overbroad definitions create false positives
They admit cases the concept was not intended to cover.
24. Overnarrow definitions create false negatives
They exclude valid cases because the boundary is too tight.
25. Circular definitions fail to add information
Defining a term only through itself or an equivalent synonym may not explain membership.
26. Negative definitions can be useful but limited
“Not X” identifies exclusion without necessarily explaining what the thing positively is.
27. Definitions need examples and counterexamples
Together they make the intended boundary more inspectable.
28. Definition conflicts need provenance
When authorities disagree, store each definition with source and context rather than forcing premature unification.
29. Crosswalks can map definitions
Exact, broader, narrower and overlapping mappings help different systems communicate.
30. Versioning matters
Historical documents should be interpreted under the definitions valid when they were produced.
31. Definition drift should be visible
Words can change meaning gradually through social, scientific or institutional use.
32. AI can extract candidate definitions
Models can identify definitional language but should preserve source, domain and authority.
33. AI should not collapse competing definitions automatically
Semantic similarity does not prove equivalence of scope or authority.
34. A practical definition record
- definition ID;
- term;
- definition text;
- purpose;
- domain;
- authority;
- scope;
- inclusions;
- exclusions;
- boundary cases;
- examples;
- counterexamples;
- effective date;
- version;
- related definitions.
35. Definition categories should improve interpretation
A useful scheme helps readers know which meaning applies and how strongly the wording should govern.
36. The deeper idea
Definitions are boundary tools. They turn language into a controlled decision about what counts.
To categorise a definition well is to preserve who defines the term, for what purpose, over what scope, with what precision and under which version.
Final answer
Categorise definitions by purpose, scope, authority, precision, context, domain, operationality, inclusions, exclusions, boundary cases and version. Keep definitions separate from examples and descriptions, and preserve competing definitions when different systems use the same term differently.