Reading every kind of source in the booklet
The booklet always mixes types, and it does so deliberately
- Every resource booklet 2022–2025 contained both written and visual sources, and both primary and secondary ones.
- The 2026 specification states it: the booklet contains "a variety of historical sources, primary and secondary."
- The mixing is not decoration. A question that could be answered from one type of source would test comprehension; a question spanning types tests whether you can extract meaning from evidence that does not announce its meaning.
- The types that appear every year: encyclopaedia and reference entries, historians' writing, newspaper reports of the day, letters to the editor, oral history extracts, photographs, cartoons, advertisements, statistics and tables.
- Candidates handle written prose well and the rest badly, which is why the rest is worth preparing.
Reading a photograph
- A photograph is evidence of three different things, and separating them is most of the work.
- What is in the frame — the subject. Usually the least informative.
- That the photograph was taken — someone judged the scene worth recording, which is evidence about what was considered notable.
- The caption — usually written by the publisher, not the photographer, and often the most loaded element in the source.
- Ask who is absent. A photograph of a public occasion showing only one kind of person is evidence about who attended, who was invited, or who the photographer pointed the camera at — and distinguishing between those is a genuine analytical move.
- Do not describe the photograph. The photograph shows a queue of women is comprehension. What the queue's existence implies is the analysis.
Reading a cartoon
- A cartoon argues. It is closer to an editorial than to a record, and it is evidence of a position rather than of events.
- Three things to identify:
- The target — who or what is being criticised.
- The assumed knowledge — a cartoon only works if readers recognise its references, so what it does not explain tells you what readers were expected to know.
- The shared value it appeals to. A cartoon persuades by invoking something the reader already accepts. That value is the most useful thing in the source, because it is evidence of a widely held assumption.
- A cartoon is unreliable as a record and excellent evidence of attitudes — which is exactly the redescription move, and the 2026 context is about attitudes.
Reading statistics and tables
- Read the column headings before the numbers. What is being counted, and by whom?
- Three questions that are almost always productive:
- Who was doing the counting, and why? A figure collected for rating purposes measures something different from one collected for a census.
- What unit is used? Households and individuals give very different pictures of the same population.
- What is not counted? Categories excluded from a table are excluded by a decision.
- A change in a series may be a change in the world or a change in how the counting was done. Saying which is a Level 3 move, and the standard's specific and generalised concept is about exactly this.
Reading oral history
- An oral history extract is a primary source recorded late. It has two dates — when the events happened and when the account was given — and both matter.
- Its characteristic strength: it records what someone experienced and how they understood it, which official records almost never contain.
- Its characteristic weakness: it is shaped by everything that happened in between, including knowing how things turned out.
- The productive question is not "is this accurate?" but "what is this accurate about?" A recollection may be unreliable about dates and reliable about how something felt at the time — and for a paper about attitudes, the second is what is wanted.
Reading a secondary source in a booklet full of primary ones
- A historian's extract is somebody's conclusion, not a record.
- It does three jobs in the booklet:
- Supplies context the primary sources assume.
- Names a category you can then look for in the primary sources — the 2025 report's Excellence candidates "applied conceptual frameworks", and a secondary source often hands you one.
- Offers a claim you can test against the primary sources, which is the strongest use and the least common.
- A secondary source that agrees with your reading is worth one sentence. One you can test against a primary source is worth a paragraph.
Worked Example
Worked example
Note: this uses the INVENTED Borough of Northmere set. Northmere, its council, its newspaper and all its sources do not exist.
A question asks how useful Source G — a photograph described as showing a queue outside the baths on a women's day, with the caption written by the newspaper — would be to a historian studying attitudes.
Two candidates write about it. Compare them.
Candidate 1:
Candidate 2:
Answer:
Step 1 — identify what Candidate 1 has done.
Described the frame, then drawn the obvious inference. A queue means demand. This is comprehension, and it is accurate.
Against the schedule it is one observation about one source — no generalisation, no cross-reference, no reading beyond the immediately obvious. Achieved.
Step 2 — identify Candidate 2's first move.
They separate the three things a photograph evidences and pick the two that bear on the stated purpose:
- not the subject — a queue is ordinary
- but the act of photographing it — evidence of what was considered notable
- and the caption's authorship — evidence of whose framing this is
Step 3 — note the precision of the conclusion.
Direct evidence of how the newspaper represented women's use, and only indirect evidence about the women themselves. This is a bounded claim, and it is more accurate than Candidate 1's, which attributes to the women a motive the image cannot show.
Step 4 — note the cross-reference, which is what lifts it.
Candidate 2 reads G against D, a different source of a different type, and finds them making the same claim by different means. That is a claim about two sources as a group, which is the schedule's unit.
Step 5 — note the restraint at the end, which is the Excellence marker.
Whether the newspaper intended that support is not recoverable from the image — the candidate states what the source cannot establish — but the paper printed both within the same month, which salvages a weaker claim the evidence does support.
That is exactly "reading between the lines to draw conclusions that go beyond the immediately obvious", followed by an honest bound on how far the reading reaches.