What is the difference between method overriding and method overloading, and why doesn't Python natively support overloading?
Overriding replaces a parent method; overloading by arity is unnecessary in a dynamically typed language
Overriding is when a subclass defines a method with the same name as an ancestor; the subclass version replaces the ancestor for that lookup path. Overloading traditionally means multiple methods with the same name but different signatures, resolved by argument count or types at compile time. Python does not resolve by signature because names are bound at runtime and functions accept any arguments. Instead, one function handles multiple forms using default arguments, *args/**kwargs, type checks, or singledispatch. Defining the same method twice in a class simply overwrites the first definition, which is a frequent surprise for engineers coming from Java or C++.
- 1
Use default arguments and **kwargs for simple arity variation.
- 2
Use functools.singledispatch or singledispatchmethod for type-based dispatch, which is the closest idiomatic Python gets to overloading.
- 3
Use typing.overload purely for static type checkers; it has no runtime effect.
- 4
Trade-off: runtime dispatch is flexible but hides branching. singledispatch makes dispatch explicit and extensible.
- 5
Common mistake: defining two methods with the same name in one class and expecting both to be callable. Only the last one exists.
- 6
Version note: singledispatch is in functools since 3.4; singledispatchmethod since 3.8. typing.overload is a type-checker construct, not runtime behavior.
Scenario Questions
0-2 years experience
- 1You define two methods named area with different parameters. Why does only one work?
- 2What is the simplest way to support an optional argument in a method?
2-5 years experience
- 1You need the same function name to handle int, str, and list differently. What is the idiomatic tool?
- 2A teammate uses typing.overload and expects it to dispatch at runtime. How do you correct them?
5-8 years experience
- 1You need a pluggable dispatch mechanism that third parties can extend without editing your module. How do you build it?
- 2You are porting Java-style overloaded APIs to Python. How do you preserve type-safety and readability?
8+ years experience
- 1Design a dispatch layer that handles both type-based and arity-based selection with good performance and introspectable error messages.
- 2Explain the trade-offs between singledispatch, explicit isinstance branching, and a registry pattern for extensible dispatch.
Follow-up Questions
- Why does defining the same method twice in a class not create an overload?
- When is singledispatch preferable to isinstance checks inside one function?