VECTOR AUTHORITY SPECIFICATION v1.0
PURPOSE
Vector Authority defines the universal mathematical foundation for structured data across all domains. It specifies how to decompose any domain into orthogonal axes, address points in the resulting coordinate space, and project views onto subspaces.
This specification is domain-agnostic. Construction, marketing, legal, medical, and any other vertical inherit these rules while defining their own axes.
CORE PRINCIPLES
- Data exists without interpretation — Coordinates are addresses, not meanings
- Axes are orthogonal — No axis can be derived from another
- Coordinates are unique — Each point in the space has exactly one address
- Projections are views — Modules/features are slices of the space, not separate data
DEFINITIONS
| Term | Definition |
|---|---|
| Axis | An independent, measurable dimension of the data space |
| Coordinate | A specific point in the space, defined by values on each axis |
| Projection | A view of the space that collapses or filters axes |
| Sparsity | The property that most coordinates in the space are empty |
| Orthogonal | Two axes are orthogonal if knowing one tells you nothing about the other |
| Derived | A value calculated from coordinates, not a true axis |
AXIS REQUIREMENTS
For a dimension to qualify as a true axis:
| Requirement | Test |
|---|---|
| Independent | Cannot be calculated from other axes |
| Measurable | Has discrete or continuous values |
| Orthogonal | Does not correlate with other axes |
| Universal | Applies to all entities in the domain |
| Stable | Values do not change based on observer |
True Axis Examples
| Domain | True Axes |
|---|---|
| Construction | Thickness, Substrate, Rating, Density, Method |
| Marketing | Contact, Campaign, Channel, Time, Outcome |
| Legal | Party, Document, Clause, Status, Date |
Non-Axis Examples (Derived/Calculated)
| Value | Why Not an Axis |
|---|---|
| Segment | Derived from Contact attributes |
| Score | Calculated from multiple axes |
| Total | Aggregation across coordinates |
| Rank | Relative to other coordinates |
COORDINATE NOTATION
Format
[L1].[L2].[L3].[L4].[L5]
Structure
| Level | Name | Function | Cardinality |
|---|---|---|---|
| L1 | Category | Industry vertical | Low (10s) |
| L2 | Domain | Functional area | Medium (100s) |
| L3 | Entity | Discrete object | High (1000s) |
| L4 | Attribute | Property of entity | Medium (100s) |
| L5 | State | Validation status | Low (10s) |
Depth Limit
Maximum: 5 levels. No exceptions.
Rationale: Beyond 5 levels adds navigation burden without semantic value. If more specificity is needed, add axes (width) not levels (depth).
Examples
Construction.Roofing.Assembly.RValue.Validated
Construction.Roofing.Assembly.Thickness.Pending
Marketing.Contacts.Lead.Score.Qualified
Marketing.Campaigns.Email.Status.Sent
Legal.Contracts.NDA.Term.Executed
Medical.Patients.Diagnosis.ICD10.Confirmed
COORDINATE SPACE
Definition
The coordinate space is the Cartesian product of all axes:
Space = A1 × A2 × A3 × ... × An
Where each Ai is the set of valid values for axis i.
Dimensionality
| Domain | Typical Axes | Space Size |
|---|---|---|
| Construction Materials | 20 | 10^20 theoretical |
| Marketing Events | 8 | 10^12 theoretical |
| Legal Documents | 12 | 10^15 theoretical |
Sparsity
Real data occupies a tiny fraction of the theoretical space:
Density = Occupied Coordinates / Total Possible Coordinates
Typical density: 10^-6 to 10^-12 (extremely sparse)
This sparsity enables efficient storage via sparse matrix representations.
PROJECTION RULES
Projections reduce dimensionality by:
- Fixing — Hold one axis constant
- Collapsing — Ignore an axis entirely
- Aggregating — Sum/count across an axis
Projection Notation
π[axes](Space) = Projected subspace
Examples:
π[Contact](MarketingSpace) = All contacts (Module M01)
π[Campaign](MarketingSpace) = All campaigns (Module M06)
π[Time, Outcome](MarketingSpace) = Time series of outcomes (Reports)
Module-to-Projection Mapping
| Module | Projection | Axes Shown |
|---|---|---|
| All Contacts | π[Contact] | Contact only |
| Campaigns | π[Campaign, Status] | Campaign + Status |
| Reports | π[Time, Outcome, Value] | Aggregated metrics |
| Revenue | π[Contact, Value] | Attribution |
STORAGE MODEL
Key-Value Representation
Each occupied coordinate is a key-value pair:
KEY: [L1].[L2].[L3].[L4].[L5]
VALUE: { payload }
Lookup Complexity
| Operation | Complexity | Notes |
|---|---|---|
| Point lookup | O(1) | Direct address |
| Axis scan | O(n) | n = values on axis |
| Projection | O(k) | k = occupied coordinates |
| Full scan | O(k) | k << total space |
Storage Formats
| Format | Extension | Use Case |
|---|---|---|
| JSON | .va.json | Data interchange |
| Binary | .va.bin | High performance |
| Markdown | .va.md | Documentation |
SCHEMA DEFINITION
Each domain must declare its axes:
{
"schema": {
"domain": "Construction",
"version": "1.0",
"axes": [
{
"id": "L1",
"name": "Category",
"type": "enum",
"values": ["Roofing", "Waterproofing", "Fireproofing", "Insulation"]
},
{
"id": "L2",
"name": "Domain",
"type": "enum",
"values": ["Assembly", "Component", "Accessory", "System"]
},
{
"id": "L3",
"name": "Entity",
"type": "string",
"pattern": "^[A-Z][a-zA-Z0-9]+$"
},
{
"id": "L4",
"name": "Attribute",
"type": "enum",
"values": ["Thickness", "RValue", "Rating", "Density", "Method"]
},
{
"id": "L5",
"name": "State",
"type": "enum",
"values": ["Draft", "Pending", "Validated", "Rejected", "Archived"]
}
],
"depth_limit": 5
}
}
VALIDATION RULES
| Code | Rule |
|---|---|
| VA.01 | Coordinate depth MUST NOT exceed schema depth_limit |
| VA.02 | Each level value MUST conform to axis type/pattern |
| VA.03 | Axes MUST be orthogonal (no derivable axes) |
| VA.04 | Coordinates MUST be unique within the space |
| VA.05 | Projections MUST reference valid axes |
| VA.06 | Schema MUST declare all axes before use |
INVARIANTS
| Code | Invariant |
|---|---|
| IV.01 | Axis count is unlimited (width) |
| IV.02 | Coordinate depth is limited to 5 (depth) |
| IV.03 | Empty coordinates are not stored |
| IV.04 | Coordinates are case-sensitive |
| IV.05 | Schema is immutable after declaration |
| IV.06 | Projections do not modify underlying data |
DOMAIN INHERITANCE
All domain kernels inherit Vector Authority rules:
VECTOR AUTHORITY (This Spec)
│
├── Construction Kernel
│ └── 20 material axes
│
├── Marketing Kernel
│ └── 8 event axes
│
├── Legal Kernel
│ └── 12 document axes
│
└── [Future Domains]
└── Domain-specific axes
Each domain:
- Declares its axes per this spec
- Follows coordinate notation
- Respects depth limit
- Implements projections for modules
CROSS-DOMAIN QUERIES
When data spans domains, prefix with domain identifier:
Construction:Roofing.Assembly.RValue.Validated
Marketing:Contact.Campaign.Email.Opened
The colon separates domain from coordinate.
FILE FORMAT
Document Structure
VECTOR AUTHORITY DOCUMENT
│
├── HEADER
│ ├── version
│ ├── domain
│ ├── created_by
│ └── signature
│
├── SCHEMA
│ └── axes[]
│
├── DATA
│ └── coordinates[]
│
├── PROJECTIONS
│ └── views[]
│
└── AUDIT
└── event_log[]
Extensions
| Extension | Description |
|---|---|
| .va.json | Vector Authority data document |
| .va.md | Vector Authority specification |
| .va.html | Vector Authority rendered view |
REFERENCES
- SuccinctAuthority.com — Philosophical foundation
- ValidKernel.com — Validation infrastructure
- CanonicalKernel.com — Frozen vocabulary
- LDSKernel.com — Lefebvre Data Standard
VERSION HISTORY
| Version | Date | Changes |
|---|---|---|
| 1.0 | 2026-01-18 | Initial specification |
Issued by: Armand Lefebvre
Date: 2026-01-18
Authority: SuccinctAuthority.com / VectorAuthority.com