The way of thinking that breaks a task into steps, finds the pattern and writes down exactly what happens. Outside computing, its value lies in the precision it forces.
State the process precisely enough that something else could carry it out
Wing described computational thinking as a fundamental skill for everyone, not only for computer scientists: decomposition, pattern recognition, abstraction and the design of step-by-step procedures. Brennan and Resnick separate its concepts, practices and perspectives. Its claims to transfer beyond computing are contested, and Denning has argued they were overstated. It is strongest where a process can be written down and has not been. It is weakest where the thing that matters cannot be made precise.
In ordinary life
You give a friend your recipe. It fails. You watched yourself cook it a hundred times, and it turns out half of it was never written down: the pan you always use, the point where you turn the heat down because it smells right. The recipe was a summary. The process was in your hands.
Computational thinking is the habit of writing the process down precisely enough that someone, or something, else could carry it out. Doing so exposes every step you were doing without noticing. Papert saw this decades before the term existed. Children programming a turtle to draw a shape had to make their own thinking explicit, and debugging it taught them something about thinking.
Abstraction is the harder half. Deciding which details of a process matter, and which can be hidden, is a judgement about what the process is for. A timetable, a marking scheme and a referral procedure are all abstractions of this kind. Most people only notice them when one fails.
The educational experience
Every page follows the same Singapore secondary school on the same kind of day. The school wants to give Secondary 1 one afternoon a week for a project that crosses Geography, Science and English Language, and the proposal is due for a decision. The five people and the situation stay fixed. Only the way of thinking changes.
All five, interconnected
The five threads as they were written, interleaved, so the hand-offs between them are visible. Pick a person to follow their day on its own. The people are illustrative, and the same five appear on every way’s page.
07:00
Mr Hafiz, before flag-raising
The proposal has been nearly ready for six weeks. His explanation is that Science and English Language are reluctant. He asks the teacher who runs the infocomm club to have a Secondary 1 group map what one afternoon would actually involve, expecting the map to confirm it.
09:15
Arjun and the infocomm club teacher
They start with one question. What has to happen, in order, for three Secondary 1 classes, about 120 students, to spend one afternoon on the project? Arjun suggests one box per step, and no step skipped because it seems obvious.
10:30
Ms Chen, being interviewed
She walks them through it: set groups across the three classes, book the lab and the library, check the afternoon against CCA, find cover for the English Language teacher who has a Secondary 3 class that period, agree how each subject will assess the work. Arjun counts fourteen steps before a single student starts.
11:45
Arjun, drawing the flowchart
He notices a loop. Each department signs off the outcomes for each group, and any change to the groups sends the plan back for sign-off. Every time a room falls through, the groups change. He marks it in red.
12:30
Mdm Kavitha, reading an email
A reply about her kitchen offer finally arrives, asking for a risk assessment and a named member of staff. The flowchart shows why it took so long. Off-site visits wait for the committee that approves them, which meets once a term, and it met the week after she wrote.
14:00
Mr Hafiz, seeing the flowchart
He had a story about reluctance. The flowchart shows a lost room changing the groups, new groups reopening sign-off and sign-off waiting on three department meetings. Science was not reluctant. It was being asked to approve the same plan for the fourth time.
15:20
Dr Ong, on a call with the school
She reads the flowchart and recognises her own guidance in the sign-off step. It says every cross-subject project must map its work to each subject’s syllabus. It was written so projects would not crowd out content. It was never written as a procedure, so nobody checked what it did when the groups kept changing.
16:00
Same call
They agree a change to one rule. Departments sign off outcomes for the whole level, not for each set of groups. The mapping stays and the loop goes. Cover for the English Language teacher is still unsolved, and nobody has written a step for it.
19:30
Arjun and Mdm Kavitha
Arjun shows Mdm Kavitha the flowchart with her kitchen in it, waiting at step nine. She is pleased to find the school had not ignored her. He is pleased the bug was in the process, and less pleased that the process still has one.
The explanation everyone had was about departments’ attitudes. Writing the afternoon down step by step, with nothing skipped for being obvious, found a loop in a well-meant rule. Fixing it cleared one blockage and exposed the next, cover, which no rule can write a teacher into. The computational part was the insistence on precision about what actually happens next.
Implications
The same five dimensions for all four ways. This is the part that can be tabulated. The two narratives above are the part that can be argued with.
Steps that were being done without anyone writing them down, and the conditions that send a process back to the start. It notices repetition that could become a rule, and the point at which an informal routine becomes too complicated to carry in anyone’s head.
Anything that loses its meaning when made precise. A flowchart of a good lesson captures the sequence and misses the judgement. Computational thinking tends to treat what cannot be specified as noise. In schools, that is often where the important thing is.
It can also see a process without seeing its purpose. A perfectly efficient sign-off procedure is only good if sign-off is the right thing to be checking.
Precision, and patience with the gap between what you meant and what you said. Debugging is the core experience: the process did exactly what you wrote, which was not what you intended. Grover and Pea note that schools often teach the concepts without the practices, so students learn what a loop is without ever debugging one.
Good work writes a process someone else could follow and shows where it breaks. Weak work produces a flowchart of the obvious steps and hides the difficult one inside a box labelled "process".
In the abstraction. Deciding which details a procedure can safely ignore is deciding what does not matter. O’Neil’s account of scoring systems in schools and hiring shows what happens when that decision is made carelessly and then automated. A rule applied precisely at scale carries its values further and faster than any person could.
The hollow version is a coding club renamed as a thinking curriculum. It teaches syntax, calls it thinking and claims that programming in a block language makes students better at everything. Denning’s critique is that the transfer claims were never well evidenced. The strong version teaches precision about processes, which is useful long after the syntax is forgotten.
Reading
Published works only. Start with the first on the list.
Wing, J. M. (2006). Computational thinking. Communications of the ACM, 49(3), 33–35.
The short essay that made the case for computational thinking as a skill for everyone.
Papert, S. (1980). Mindstorms: Children, Computers, and Powerful Ideas. Basic Books.
Children programming as a way of making their own thinking explicit, long before the term.
Brennan, K., & Resnick, M. (2012). New frameworks for studying and assessing the development of computational thinking. Paper presented at the annual meeting of the American Educational Research Association, Vancouver.
Concepts, practices and perspectives as three separate things to teach and assess.
Grover, S., & Pea, R. (2013). Computational thinking in K–12: A review of the state of the field. Educational Researcher, 42(1), 38–43.
What schools are teaching under the name, and the gaps.
Denning, P. J. (2017). Remaining trouble spots with computational thinking. Communications of the ACM, 60(6), 33–39.
The critique of overstated claims that computational thinking transfers everywhere.
O’Neil, C. (2016). Weapons of Math Destruction. Crown.
What happens to the values inside a procedure once it is automated at scale.