Skip to content
Foundations

Foundations

The classical core of computational thinking — the concepts every other section of this knowledge base builds on.

When Jeannette Wing argued in 2006 that computational thinking is a fundamental skill for everyone, not just computer scientists, she was describing a set of mental tools, not a programming curriculum. Two decades later those tools have outlived several generations of languages, frameworks, and now AI assistants — precisely because they operate at the level of how you think about problems, not how you type solutions.

This section covers the five capabilities that form that core, plus the framing that holds them together:

PageWhat you will learn
What Is Computational Thinking?The definition, the Wing 2006 lineage, how the concept evolved to 2026, and why CT is not the same thing as coding
DecompositionBreaking large problems into tractable parts — functional, data-flow, and stakeholder decomposition, with a worked example
Pattern RecognitionSpotting regularities you can exploit, and how this pillar connects directly to machine learning
AbstractionLayers, interfaces, and information hiding — plus the law of leaky abstractions and when to peek beneath them
Algorithm DesignTurning steps into algorithms: correctness, conceptual big-O intuition, and pseudocode you can actually use
Evaluation & DebuggingTesting, debugging as hypothesis-driven science, and judging solutions on correctness, efficiency, and robustness

How the pieces fit

The pillars are not a linear recipe. In real work you cycle through them: decomposition exposes patterns, patterns suggest abstractions, abstractions make algorithms expressible, and evaluation sends you back to re-decompose whatever you got wrong. Treat the six pages as one loop — read in order the first time, revisited in any order afterward.

New to CT? Read the pages in sequence — each assumes vocabulary introduced by the previous one. Experienced developer? Skim What Is Computational Thinking? for the historical framing, then jump to Evaluation & Debugging — the pillar most people underinvest in, and the one that matters most in the AI era.