Degrees of Relationship¶
ERD Guide — Part 2 of 6
The number of entity types that participate in a relationship.
What is a Degree?¶
The degree of a relationship refers to how many entity types participate in that relationship.
1. Unary Relationship (Degree 1)¶
Only one entity type is involved — the entity relates to itself.
erDiagram
EMPLOYEE ||--o{ EMPLOYEE : "manages"Example: An employee manages other employees. The "manages" relationship involves only the EMPLOYEE entity.
Common unary relationships:
| Entity | Relationship | Description |
|---|---|---|
| EMPLOYEE | manages / is managed by | Supervisor hierarchy |
| PERSON | is married to | Marriage |
| PART | contains | Bill of materials |
Note
Unary relationships can also be many-to-many! Think of a parts hierarchy where one part contains many sub-parts, and one sub-part can be used in many parent parts.
2. Binary Relationship (Degree 2)¶
Two entity types participate in the relationship. This is the most common type.
erDiagram
CUSTOMER }o--o{ PRODUCT : "purchases"Example: Customers purchase products. Two entities involved: CUSTOMER and PRODUCT.
3. Ternary Relationship (Degree 3)¶
Three entity types participate in a single relationship.
erDiagram
WAREHOUSE ||--o{ SHIPMENT : "ships from"
PRODUCT ||--o{ SHIPMENT : "contains"
CUSTOMER ||--o{ SHIPMENT : "ships to"
SHIPMENT {
int ShipmentID PK
int WarehouseID FK
int ProductID FK
int CustomerID FK
}Example: A shipment involves a WAREHOUSE (where it ships from), a PRODUCT (what is shipped), and a CUSTOMER (who receives it). All three are needed to describe the relationship.
4. N-ary Relationship (Degree 4+)¶
Any relationship involving more than three entity types. These are rare in practice and are usually decomposed into simpler relationships.
Summary¶
| Degree | Name | Entities Involved | Example |
|---|---|---|---|
| 1 | Unary | 1 (self-referencing) | Employee manages Employee |
| 2 | Binary | 2 | Customer purchases Product |
| 3 | Ternary | 3 | Warehouse ships Product to Customer |
| 4+ | N-ary | 4+ | Rare — usually decomposed |
Design Tip
When you encounter a ternary or higher relationship, ask yourself: "Can I break this into multiple binary relationships?" Often you can, and it simplifies the model.