Course contentsShow
← TypeScript
Lesson 152 of 1,7964. Type InferenceFree lesson

Inference Limitations with Empty Values

Why empty arrays, objects, and null require annotations or will infer overly general types.

Inference Limitations with Empty Values

You'll learn: Why TypeScript can't always infer useful types from empty arrays, objects, and null values, and what to do about it.

When you give TypeScript an empty value—like [], {}, or null—the compiler faces a dilemma: what type should it infer? There's no data to analyze, so it must make a conservative guess.

The Empty Array Problem

let items = [];  // TypeScript infers: any[]

Without any elements to examine, TypeScript defaults to any[]—an array that can hold anything. This defeats the purpose of type safety! If you later push numbers and strings into it, TypeScript won't complain:

items.push(42);
items.push("hello");  // Both allowed with any[]

The Empty Object Problem

let config = {};  // TypeScript infers: {}

An empty object literal infers as {}, which means "an object with no known properties." You can't safely add properties later without type errors.

The Null Problem

let data = null;  // TypeScript infers: null

Inferring null as the type means you can only assign null to this variable—it's permanently stuck!

The Solution: Add Type Annotations

When starting with empty values, help TypeScript by providing explicit annotations:

let numbers: number[] = [];
let config: { theme: string; debug: boolean } = {} as any;  // (will learn better patterns later)
let data: string | null = null;

Think of it like telling someone to prepare an empty container: you need to specify what kind of container (shoe box? tool chest?) before they can gather the right one.

Key Takeaway: Empty arrays, objects, and null lack context for inference—TypeScript defaults to overly general or overly narrow types. Always provide explicit type annotations when initializing with empty values.