How teachers should choose what to teach

Every curriculum decision is, eventually, a decision about what to leave out. There are more topics, themes, texts, and ideas in any discipline than a five-year secondary curriculum can carry. A specification helps but does not exhaust the question; even within a specification, a department chooses which texts to teach, which case studies to use, which historical periods to deepen, which substantive examples will carry a concept.

This is the part of curriculum design that is most often left implicit. The official documents tend to describe what is taught; they rarely describe what was considered and rejected. That hidden layer of choice (the conversations in department meetings, the things one teacher quietly drops, the topic that gets two lessons instead of four) shapes the curriculum more than the formal documents do.

This guide is aimed at heads of department and curriculum leads facing the broader question of what should and should not be in the curriculum, and it sketches a decision frame for working through that question honestly.

Two departments teaching the same specification can deliver very different educations. One weights its history KS3 toward British political and constitutional history; another weights toward global history with strong threads on empire, migration, and revolution. Both might be defensible. They are different curricula in everything but name.

The choice matters because the act of selecting (and the act of cutting) is where the curriculum reveals what it believes the subject is for. If a department cannot articulate what it left out, it usually cannot articulate what it stands for either. Christine Counsell's framing is useful here: A curriculum is a body of knowledge selected and sequenced with purpose. Borrowing the specification's content list wholesale is not selecting; it is deferring.

It is tempting to start with the topic list and decide topic by topic. That tends to produce a curriculum that is locally reasonable and globally incoherent.

A more useful order is to decide three things first. What is the subject for in this school? What kind of thinking do you want students to be doing by the end of KS4? What body of knowledge do you want them to leave with? Those three answers are the lens through which every topic decision should be evaluated. A paragraph each is plenty, but they do need to be agreed and written down, or the team will quietly default to different answers.

Once the lens is in place, the next step is a frame for evaluating any candidate topic, text, case study, or example. The frame below has worked well for departments we have spoken with. It is not a scoring rubric; it is a small set of questions that surface what the team actually thinks.

A useful distinction, drawn from Christine Counsell and others, is between substantive knowledge (the actual content: Events, facts, formulae, texts, processes) and disciplinary knowledge (how the discipline thinks: How historians weigh evidence, how scientists design experiments).

Many curriculum decisions involve trading substantive depth for disciplinary depth or the other way round. A history department can teach more periods lightly with less attention to interpretation, or fewer periods in depth with strong work on the historian's craft. Most departments find that a tilt toward disciplinary depth at KS3 pays off across KS4: Students who can think like a historian, scientist, or literary critic absorb new content faster than students who have only met it as isolated facts.

The specification was not designed for your school. It was designed as a defensible national assessment of a discipline at age 16, which is not the same as a curriculum. Treat it as a constraint rather than a syllabus: Within that constraint, there is usually more freedom than departments use.

This matters especially at KS3, where the temptation to pre-teach the GCSE specification is strongest. A KS3 that simply foreshadows the GCSE tends to be thin; a KS3 that builds the substantive and disciplinary foundations the GCSE relies on is much richer, even if the topic list does not look like a junior version of the specification.

Every unit involves choosing specific examples: A novel, a play, a war, a case study country, an experiment, a worked example. These choices look small but are where the curriculum's character shows.

Pick examples the team can teach with conviction; an enthused teacher carries more than the abstract best choice on paper. Pick examples students can return to across units, building familiarity. Pick examples that genuinely illustrate the disciplinary move you are teaching, not the ones easiest to resource. And occasionally pick examples that stretch the team or the cohort; a curriculum entirely made of safe choices is its own kind of failure.

Cutting is harder than adding, both intellectually and politically. Teams are attached to units they have taught for years, and senior leaders tend to read 'we are cutting two units' as 'we are doing less'.

Frame each cut as a trade rather than a loss; naming what the freed-up time will allow makes the proposition additive. Show the audit evidence behind the decision; cuts grounded in cohort underperformance or time overruns are easier to defend than cuts grounded in HoD preference. Cut deliberately rather than by attrition, and keep what was cut visible somewhere so the next refresh cycle can revisit it.

A few decisions sit awkwardly inside the frame and tend to need a separate conversation.

A good curriculum is not a fixed document; it is a set of decisions you periodically test. The choices you made three years ago might still be right; equally, the cohort, the specification, the team, or the discipline might have moved.

A practical cadence is to review topic-level choices every two or three years against the same lens questions. Most will hold; a few will not, and those are the candidates for the next refresh. This is where written rationales pay off: A department that documented why it picked this novel over that one in 2023 will have a far easier conversation in 2026 than one reconstructing the reasoning from memory.