A constraint is a limit on what can be done, chosen, built, changed or achieved.
Budget, time, law, physics, staffing, capacity, safety, policy, geography, compatibility and ethics can all constrain action. But they do not all behave the same way. Some are absolute, some negotiable, some temporary, some inherited, some local and some only appear because of another dependency.
Quick answer: how should constraints be categorised?
- Source: physical, legal, financial, technical, human, ethical, organisational?
- Hardness: hard, soft, preferred, assumed?
- Scope: which tasks, actors, systems or jurisdictions are affected?
- Time: permanent, temporary, deadline-bound, conditional?
- Dependency: does the constraint come from another resource, rule or decision?
- Negotiability: can it be relaxed, traded or escalated?
- Cost: what does relaxing it require?
- Authority: who can change it?
- Uncertainty: how well is the limit known?
- Consequence: what happens if it is violated?
This article applies How to Categorise Anything to constraints as boundaries on possible action.
1. Separate constraint from goal
A goal describes the desired future; a constraint limits which routes toward that future are allowed or feasible.
2. Separate constraint from preference
“Must cost below $10,000” differs from “prefer the cheapest option”. Preferences can usually be traded more freely.
3. Hard constraints define feasibility
Violating a hard constraint makes an option invalid, impossible or unacceptable.
4. Soft constraints define desirability
They influence ranking but may be relaxed when trade-offs justify it.
5. Assumed constraints may be false
Teams often inherit limits that were once true but are no longer binding.
6. Physical constraints come from reality
Mass, geometry, thermodynamics, distance and material properties cannot be negotiated by policy.
7. Technical constraints come from system design
Interface compatibility, software versions and architecture choices may constrain implementation.
8. Legal constraints come from authority
Laws, licences, contracts and regulations can limit action within a defined jurisdiction.
9. Policy constraints are organisational
Internal rules may be changeable by a different authority than external law.
10. Financial constraints limit spend or exposure
Budget, cash flow, capital and borrowing limits constrain different parts of a plan.
11. Capacity constraints limit throughput
Staff, machines, classrooms, bandwidth and storage can become bottlenecks.
12. Time constraints limit sequence
Deadlines, lead times and fixed windows can eliminate options that are otherwise feasible.
13. Human constraints include capability and attention
Skill, fatigue, availability and cognitive load can shape what can be executed reliably.
14. Ethical constraints protect values
Some options may be technically and legally possible yet rejected because they violate ethical principles.
15. Safety constraints protect acceptable risk
These may be derived from engineering limits, law, professional practice or organisational tolerance.
16. Scope should be explicit
A constraint may apply only to one component, one project, one country or one user group.
17. Local constraints can look global
A bottleneck in one department may not constrain the whole organisation if work can be rerouted.
18. Temporary constraints need expiry
Emergency rules, maintenance windows and temporary shortages should not silently become permanent assumptions.
19. Conditional constraints activate only in some states
A limit may apply only above a risk level, after a threshold or during one operating mode.
20. Dependencies create inherited constraints
If Task B depends on Task A, Task A’s delay becomes an indirect time constraint on B.
21. Resource constraints are often dynamic
Capacity can change through hiring, maintenance, procurement or substitution.
22. Negotiability should be recorded
Some limits can be relaxed by paying more, escalating authority or changing scope.
23. Relaxation has cost
A constraint can be technically negotiable but prohibitively expensive or slow to change.
24. Authority determines who can relax a constraint
A project manager may change sequence but not law; a regulator may change rules but not physics.
25. Constraints can conflict
A speed target and a safety minimum can collide, forcing trade-offs or redesign.
26. Constraint interaction can create infeasibility
Each limit may be reasonable alone while the combination leaves no valid solution.
27. Constraint satisfaction should be tested as a set
Evaluate candidate solutions against all binding constraints, not one at a time.
28. Bottleneck constraints dominate systems
Improving non-binding constraints may add complexity without increasing total capability.
29. Slack measures distance from a limit
A constraint with large spare capacity behaves differently from one that is almost binding.
30. Shadow cost reveals constraint value
In optimisation, the value of relaxing a constraint can be estimated by how much the objective improves.
31. Uncertain constraints need ranges
Unknown lead times, budgets or capacities should not be represented as falsely exact limits.
32. Scenario constraints help planning
Different future conditions can produce different feasible sets.
33. AI can extract constraints from text
Models can identify deadlines, budgets and prohibitions, but should preserve source and authority.
34. AI can confuse preferences with constraints
Language such as “ideally”, “preferably” and “must” should not be collapsed into one severity.
35. Constraint taxonomies drift
New tools, regulations and capabilities can remove old limits and create new ones.
36. Constraints should connect to goals and decisions
Their purpose is operational: they shape what futures and choices remain feasible.
37. A practical constraint record
- constraint ID;
- source;
- authority;
- scope;
- hardness;
- condition;
- limit;
- valid time;
- dependencies;
- negotiability;
- relaxation cost;
- slack;
- uncertainty;
- violation consequence;
- version.
38. Constraint categories should improve design
A useful classification reveals what is impossible, what is expensive, what is negotiable and who can change the boundary.
39. Constraints shape solution space
They are not merely obstacles; they define the geometry within which valid solutions exist.
40. The deeper idea
Every plan lives inside a boundary of possibility. Categorising constraints makes that boundary visible.
To categorise a constraint well is to know what creates the limit, how binding it is, who can change it, how long it lasts and what becomes possible if it moves.
Final answer
Categorise constraints by source, hardness, scope, time, dependency, negotiability, relaxation cost, authority, uncertainty and consequence. Keep hard limits separate from preferences and assumptions, test interacting constraints as a set, and version temporary or changeable limits so plans remain explainable.