Definition of Composition - What Is Composition? | OneNoughtOne

System Design (LLD)Composition vs Inheritance

What Is Composition?

LevelIntermediate

Duration50 mins

TopicComposition vs Inheritance

1 / 4

Definition of Composition

1

Building Blocks of Software Construction

Consider how architects design buildings. They don't create monolithic structures from a single piece of material. Instead, they assemble buildings from distinct components: steel beams provide structural support, concrete slabs form floors, glass panels create windows, and electrical systems power the interior. Each component has a specific purpose, can be sourced from different suppliers, and can often be replaced or upgraded independently.

Software composition follows the same principle. Rather than creating massive, monolithic objects that attempt to do everything, we construct complex software entities by combining simpler, focused objects—each with its own well-defined responsibility. This is the essence of composition.

What You Will Learn

By the end of this page, you will understand compositional thinking as a fundamental design philosophy. You'll learn the formal definition of composition, how it differs from other relationships between objects, and why it forms the backbone of flexible, maintainable software architecture. This understanding will transform how you approach system design.

1

What Is Composition?

Composition is an object-oriented design principle where complex objects are constructed by combining (or "composing") simpler objects, rather than inheriting behavior from parent classes. In composition, an object contains references to other objects that provide required functionality—delegating work to these contained objects rather than implementing everything itself.

Let's establish a precise technical definition:

The key insight of composition is that behavior comes from collaboration. Instead of one large object knowing how to do everything, multiple specialized objects work together—each contributing its expertise to achieve the larger goal.

The Core Distinction

Inheritance says: "I AM a type of that thing, so I inherit its behaviors."

Composition says: "I HAVE that thing, and I use it to perform certain behaviors."

This seemingly simple difference has profound implications for software flexibility, testability, and maintainability.

Consider this simple example to illustrate the distinction:

Inheritance approach: A Car class inherits from an Engine class to gain engine functionality.

Composition approach: A Car class contains an Engine object and delegates propulsion-related operations to it.

Intuitively, composition makes more sense here—a car is not a type of engine; a car has an engine. This semantic correctness extends to practical benefits: with composition, we can easily swap engines, test the car and engine independently, or reuse the engine design in motorcycles, boats, and generators.

2

Anatomy of Compositional Relationships

To understand composition deeply, we must examine the structure and semantics of compositional relationships. Every compositional relationship involves several key elements:

Elements of Compositional Relationships
Element	Description	Example
Container (Composite)	The object that holds references to other objects and coordinates their behavior	`Car` holds references to `Engine`, `Transmission`, `Wheels`
Component (Part)	The contained object that provides specific functionality to the container	`Engine` provides propulsion capability to `Car`
Interface/Contract	The abstraction that defines how the container interacts with components	`Propulsion` interface that `Engine` implements
Delegation	The mechanism by which the container forwards requests to components	`car.accelerate()` calls `engine.increasePower()`
Lifecycle Management	How the container manages creation, usage, and destruction of components	`Car` creates engine on construction, destroys on scrapping

Let's visualize this with a concrete code example:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
// Component: Provides specific functionality
class Engine {
    private power: number = 0;
    private maxPower: number;
    
    constructor(maxPower: number) {
        this.maxPower = maxPower;
    }
    
    increasePower(amount: number): void {
        this.power = Math.min(this.power + amount, this.maxPower);
    }
    
    decreasePower(amount: number): void {
        this.power = Math.max(this.power - amount, 0);
    }
    
    getCurrentPower(): number {
        return this.power;
    }
    
    getEfficiency(): number {
        return this.power / this.maxPower;
    }
}
 
// Container (Composite): Coordinates components to provide higher-level behavior
class Car {
    // Composition: Car CONTAINS an Engine
    private engine: Engine;
    private speed: number = 0;
    private powerToSpeedRatio: number = 2.5;
    
    constructor(engine: Engine) {
        // Car takes ownership of the engine (lifecycle management)
        this.engine = engine;
    }
    
    accelerate(): void {
        // Delegation: Car forwards request to Engine
        this.engine.increasePower(10);
        this.speed = this.engine.getCurrentPower() * this.powerToSpeedRatio;
    }
    
    brake(): void {
        // Delegation again
        this.engine.decreasePower(15);
        this.speed = this.engine.getCurrentPower() * this.powerToSpeedRatio;
    }
    
    getSpeed(): number {
        return this.speed;
    }
    
    getEngineEfficiency(): number {
        // Delegating capability query
        return this.engine.getEfficiency();
    }
}
 
// Usage: Constructing the composite object
const sportsEngine = new Engine(500);
const sportsCar = new Car(sportsEngine);
 
sportsCar.accelerate();
sportsCar.accelerate();
console.log(`Speed: ${sportsCar.getSpeed()} mph`);
console.log(`Engine Efficiency: ${sportsCar.getEngineEfficiency() * 100}%`);

Observe the Structure

Notice how Car doesn't know HOW the engine works internally—it only knows that the engine can increase/decrease power and report its state. This separation of concerns is a hallmark of good composition. The Car is not coupled to engine implementation details.

3

Composition vs Aggregation: Understanding the Spectrum

In object-oriented design, "composition" is sometimes used loosely to describe any situation where one object contains another. However, there is a more precise distinction between composition and aggregation—two related but semantically different relationships.

Composition (Strong Ownership)

•Parts cannot exist independently of the whole
•Whole controls the lifecycle of parts
•Destroying the whole destroys the parts
•Parts typically internal/private to the whole
•Example: Body and Heart — heart can't exist without body

Aggregation (Weak Ownership)

•Parts can exist independently of the whole
•Whole does not control part lifecycle
•Parts may be shared across multiple wholes
•Parts may outlive the whole
•Example: Team and Player — player exists without team

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
// COMPOSITION: Strong ownership, lifecycle control
// The Document creates and owns its Paragraphs
class Document {
    private paragraphs: Paragraph[] = [];
    
    addSection(text: string): void {
        // Document creates the paragraph (lifecycle control)
        const paragraph = new Paragraph(text);
        this.paragraphs.push(paragraph);
    }
    
    // When Document is destroyed, Paragraphs are destroyed too
    // Paragraphs have no independent existence
}
 
// AGGREGATION: Weak ownership, independent lifecycles
// The Team references Players, but doesn't control their existence
class Team {
    private players: Player[] = [];
    
    addPlayer(player: Player): void {
        // Team receives an existing player (no lifecycle control)
        this.players.push(player);
    }
    
    removePlayer(player: Player): void {
        const index = this.players.indexOf(player);
        if (index !== -1) {
            this.players.splice(index, 1);
            // Player continues to exist after removal
        }
    }
    
    // When Team is dissolved, Players continue to exist
    // Players can join other teams
}

Why does this distinction matter?

The semantic difference affects how you reason about your design:

Memory management: In composition, destroying the container destroys components. In aggregation, you must track component lifecycles separately.
Sharing: Aggregated parts can belong to multiple containers simultaneously. Composed parts are exclusive to one container.
Testing: Aggregated relationships are often easier to test because you can create parts independently and inject them into containers.
Semantic clarity: Choosing the right relationship expresses design intent. A House "composes" Rooms (rooms are meaningless without the house). A Library "aggregates" Books (books exist independently).

Practical Note on Terminology

In everyday discussion, many developers use "composition" to mean any HAS-A relationship, whether technically composition or aggregation. This is acceptable in casual conversation, but when designing systems, understanding the ownership semantics matters greatly for lifecycle management and resource cleanup.

4

The Philosophical Foundation of Composition

Composition is not merely a technical pattern—it reflects a fundamental philosophy about how complex systems should be constructed. This philosophy, rooted in both software engineering wisdom and broader systems thinking, has profound implications for how we approach design.

The philosophy can be summarized in three core tenets:

The Three Tenets of Compositional Design

•Specialization over Generalization: Each component should do one thing exceptionally well, rather than many things adequately. A specialized Logger, Validator, or Calculator component can be perfected in isolation and reused broadly.
•Collaboration over Hierarchy: Complex behavior emerges from the interaction of simple parts, not from elaborate inheritance trees. Objects collaborate as peers, each contributing its expertise.
•Loose Coupling over Tight Binding: Components connect through well-defined interfaces, not through internal knowledge. This allows parts to be replaced, upgraded, or tested independently.

Historical Context: The Evolution of Compositional Thinking

The preference for composition over inheritance emerged from decades of collective experience with object-oriented systems. Early OOP adoption (1980s-1990s) heavily favored inheritance—it seemed elegant to model real-world taxonomies in code. However, practitioners discovered that inheritance hierarchies tended to:

Become rigid and difficult to change as requirements evolved
Couple code in unexpected ways (changing a base class affects all descendants)
Force uncomfortable design choices (where should FlyingCar fit in the hierarchy?)
Make testing more difficult (testing a derived class requires the entire ancestor chain)

By the late 1990s, influential voices in the design community—including the Gang of Four in their seminal Design Patterns book—began advocating: "Favor object composition over class inheritance."

This wasn't a rejection of inheritance, but a recognition that composition offers greater flexibility for most design problems.

From the Gang of Four

"Favor object composition over class inheritance... Inheritance breaks encapsulation and can make designs fragile. Good designs use inheritance for what it's good at—modeling is-a relationships and enabling polymorphism—and composition for everything else." — Design Patterns, 1994

5

Composition in Formal Terms

For those who appreciate precision, let's express composition in more formal terms. Understanding these formal properties helps when reasoning about designs at a theoretical level.

Key Properties of Composition:

Formal Properties of Compositional Relationships
Property	Definition	Implication
Transitivity	If A composes B, and B composes C, then A transitively composes C	Complex systems can be built from deeply nested compositions
Substitutability	Any component satisfying the interface contract can replace another	Enables runtime flexibility and testability (e.g., mock injection)
Information Hiding	The container need not expose its components to external users	Implementation details can change without affecting clients
Multiplicity	A container may hold zero, one, or many components of a type	Enables collections, optional dependencies, and flexible cardinality
Directionality	The HAS-A relationship flows from container to component	Clarifies ownership and responsibility chains

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
// Demonstrating compositional properties
 
// 1. SUBSTITUTABILITY: Interface-based composition
interface PaymentProcessor {
    process(amount: number): Promise<boolean>;
}
 
class StripeProcessor implements PaymentProcessor {
    async process(amount: number): Promise<boolean> {
        // Stripe-specific implementation
        console.log(`Processing $${amount} via Stripe`);
        return true;
    }
}
 
class PayPalProcessor implements PaymentProcessor {
    async process(amount: number): Promise<boolean> {
        // PayPal-specific implementation
        console.log(`Processing $${amount} via PayPal`);
        return true;
    }
}
 
class Checkout {
    private processor: PaymentProcessor;
    
    // Any PaymentProcessor implementation can be composed here
    constructor(processor: PaymentProcessor) {
        this.processor = processor;
    }
    
    async complete(amount: number): Promise<boolean> {
        return this.processor.process(amount);
    }
}
 
// 2. TRANSITIVITY: Nested composition
class ShoppingCart {
    private checkout: Checkout;
    
    constructor(checkout: Checkout) {
        // Cart composes Checkout, which composes PaymentProcessor
        // Cart transitively composes PaymentProcessor
        this.checkout = checkout;
    }
}
 
// 3. MULTIPLICITY: Multiple components
class Orchestra {
    private instruments: Instrument[] = []; // Many components
    private conductor: Conductor;            // Single component
    
    constructor(conductor: Conductor) {
        this.conductor = conductor;
    }
    
    addInstrument(instrument: Instrument): void {
        this.instruments.push(instrument);
    }
}

6

Why Composition Should Be Your Default Choice

Experienced software architects generally adopt a composition-first mindset. When facing a design problem involving code reuse or polymorphic behavior, they consider composition before inheritance. This isn't dogma—it's hard-won wisdom from building and maintaining large systems.

Here are the fundamental reasons composition should be your default:

Advantages of Composition-First Design

•Flexibility at Runtime: Composed components can be swapped at runtime. Change behavior by injecting a different collaborator, not by restructuring class hierarchies.
•Separation of Concerns: Each component handles one aspect of behavior. The container orchestrates. This separation makes code easier to understand, test, and modify.
•Easier Testing: Components can be tested in isolation. Containers can be tested with mock/stub components. No need to instantiate complex inheritance chains.
•Reduced Coupling: Changes to a component's internals don't affect its users, as long as the interface contract holds. This limits the blast radius of changes.
•Multi-faceted Behavior: A class can compose multiple components, gaining diverse capabilities. Inheritance limits you to one parent (in most languages).
•Clearer Intent: Composition expresses "uses" or "has-a" relationships clearly. These are often more accurate models of the problem domain than is-a hierarchies.

The practical reality:

In real-world systems, most relationships between concepts are collaborative (HAS-A), not hierarchical (IS-A). A User has a Profile, has Preferences, has Permissions. An Order has LineItems, has a PaymentMethod, has a ShippingAddress. These are not inheritance relationships—they are compositions.

When you default to composition, your designs naturally mirror these real relationships, leading to more intuitive and maintainable code.

The Litmus Test

When deciding between composition and inheritance, ask: "Would I be comfortable saying that X IS-A Y in all contexts, now and in the future?" If there's any hesitation, composition is probably the right choice. Mistaken inheritance is hard to undo; mistaken composition is easy to refactor.

7

Common Misconceptions About Composition

Before we proceed deeper into compositional design, let's address some frequent misconceptions that can lead developers astray:

Misconceptions to Avoid

•"Composition is always better than inheritance" — Not true. Inheritance is the right tool for genuine IS-A relationships where substitutability and polymorphism are required. The principle is to FAVOR composition, not to ELIMINATE inheritance.
•"Composition means more code" — Sometimes initially, but compositional designs tend to have less code overall. Reusable components eliminate duplication. Clear interfaces reduce documentation needs. Easier testing reduces debugging time.
•"Composition is just about delegation" — Delegation is one mechanism, but composition is a design philosophy. It's about building systems from focused, collaborating parts—whether through delegation, configuration, assembly, or other patterns.
•"Composition can't achieve polymorphism" — False. Composition achieves polymorphism through interfaces. A container that depends on an interface can work with any implementation—this is polymorphism without inheritance.
•"Composition is a modern trend" — Compositional thinking has existed since the early days of structured programming. The Gang of Four advocated it in 1994. It's not a trend—it's established wisdom.

The Right Mindset

Think of composition and inheritance as tools in a toolbox. A skilled craftsperson knows when each tool is appropriate. Composition is the general-purpose tool you'll reach for most often; inheritance is the specialized tool for specific situations.

8

Summary: Understanding Composition

We've established the foundational understanding of composition as a design principle. Let's consolidate the key concepts:

Key Takeaways

•Composition is a HAS-A relationship — Objects contain references to other objects and delegate work to them, rather than inheriting behavior.
•The structure involves containers and components — Containers coordinate; components specialize. Each part has a defined role.
•Composition and aggregation differ in ownership — Composition implies lifecycle control; aggregation allows independent existence.
•The philosophy emphasizes collaboration — Complex behavior emerges from simple parts working together through well-defined interfaces.
•Composition should be your default — It offers flexibility, testability, and clear separation of concerns that inheritance often cannot match.
•Both composition and inheritance have roles — The goal is appropriate choice, not dogmatic exclusion of either technique.

What's next:

With the definition and philosophy of composition established, we'll next explore the practical mechanics of composing objects. You'll learn how to structure objects that contain other objects, how to manage the relationships between containers and components, and how to design compositions that remain flexible as requirements evolve.

Page Complete

You now understand composition as a fundamental design principle—not just a coding technique, but a philosophy for building maintainable, flexible software. Next, we'll examine how to put these principles into practice by exploring the mechanics of objects containing other objects.

1 / 4