Interpreting evidence: the three skills
The three historians' skills
- To analyse here is to use historians' skills to interpret evidence, in order to show understanding of historical concepts.
- There are three skills, and only three:
- Close reading — attending to the exact words, including what is implied rather than stated.
- Comprehension — understanding what the source says on its own terms.
- Extracting meaning — drawing out what the source shows beyond what it says.
- Notice what is not on the list: prior knowledge of the topic. The sources supply the content. Having heard of the event before is beside the point.
- Mark the sources up: dates in the margin, a word for who made each one, a line under anything you intend to quote. Holding eight sources in your head is not possible; annotating them takes two minutes.
- Insight has an exact meaning here: 'reading between the lines' to draw conclusions that go beyond the immediately obvious and show a high degree of engagement with the evidence.
The historical concepts evidence is read through
- Each set of sources is interpreted through a named historical concept, and there are five that recur:
- perspectives — whose view the evidence gives, and what produced it
- cause and effect — what brought the event about, and what followed
- continuity and change — what shifted across the evidence, and what held
- reliability and usefulness — how far a source can be trusted, and what for
- significance — why the event mattered, to whom, and for how long
- Learn the concepts, not an order. Any of them can be applied to any body of evidence.
- Sources are grouped, but the groups are floors, not ceilings. Reaching to a source outside the group, where it genuinely adds something, is a mark of a strong answer.
Worked Example
A source in a set is a table headed "Households in receipt of relief, 1932", compiled by a borough council. A student writes:
"Source E shows that 1,840 households were receiving relief in 1932, so poverty was widespread."
What has been read, and what has not?
Step 1 — separate what the source records from what it is about.
The table records households the council registered as receiving relief. It does not record households in poverty. Those are different populations, and the gap between them is where the reading has to happen.
Step 2 — ask who counted, and why.
A borough council counted, in order to administer a relief scheme. So the figure covers only households that applied, qualified under the council's own rules, and were processed before the count was taken.
Step 3 — work out which direction the error runs.
Every one of those filters removes households. Families outside the qualifying rules, families who did not apply, and families in districts the scheme did not cover are all absent. The figure is therefore a floor, not a measure — poverty was at least this widespread, and probably more so.
Step 4 — say what would settle it.
Another source counting the same year by a different method — a charity's records, a census return, a school's roll of children fed — would show how far apart the two populations were. Where the set does not contain one, say so: naming the missing evidence is a finding about the set.
The rewritten reading: