Two Accounts of the 1906 Earthquake
In a fourth-grade unit on the 1906 San Francisco earthquake, students read two sources. Source A is the front page of the San Francisco Chronicle from April 19, 1906, written within hours of the quake; it reports fires spreading and gives death counts that later proved far too low. Source B is a memoir published in 1955 by a woman who was eight years old in 1906; she recalls the ground 'rolling like a wave,' sleeping in Golden Gate Park, and her mother's calm voice. During discussion, Marcus says Source B is more trustworthy 'because she was actually there,' while Priya says Source A is better 'because it's a real newspaper, not just one person.' The teacher wants students to move beyond ranking the sources as simply true or false. The class will later write a paragraph answering: 'What was it like for children in San Francisco in the days after the earthquake?'
1. How should the teacher respond to Marcus and Priya, and what does each source actually support?
Read the worked answer
Both students are making a global trustworthiness judgment, which is the habit the teacher needs to disrupt. Historical reasoning asks: trustworthy for which question? Source A has proximity and institutional access, but it was produced under deadline with incomplete information; casualty figures were provisional and, in 1906, likely shaped by officials wanting to protect the city's reputation. It is strong evidence for what was known and published on April 19, how information circulated, and the scale of destruction. It is weak for precise death tolls or children's feelings.
Source B has eyewitness experience and sensory detail, which is valuable for the experience of children and families. But it was written 49 years later; memory is reconstructed, shaped by later telling, and selective. It is weak for chronology and exact numbers. The teacher should validate each criterion (proximity, eyewitness) and then ask: 'For which question is each one useful?' The paragraph on children's experience should lean on B for experience and on A for context such as fires and park encampments.
2. Design a task that requires corroboration rather than ranking.
Read the worked answer
Give pairs a three-column chart: claim, source, status (confirmed, not confirmed, contradicted). Students pull three details from the memoir, such as sleeping in Golden Gate Park, the ground's motion, and family members separated. For each, they search Source A or a third source (a photograph, a relief-agency notice) for support. If the newspaper reports refugees in parks, the park detail is corroborated; the 'rolling like a wave' detail is not something a newspaper would record the same way, so it stays as one person's perception, not 'false.' Students then write a sentence that names the source and the question it answers, for example: 'The Chronicle shows that thousands went to parks; the memoir shows what one child did there.' The chart makes corroboration a procedure, not a vote.
3. A student says the low death count proves the newspaper lied. How should the teacher respond?
Read the worked answer
Separate intent from accuracy. A newspaper can be inaccurate without lying. Within 24 hours of a major earthquake and fire, records were destroyed, bodies were uncounted, communications were down, and officials had incentives to minimize the disaster. Later estimates rose substantially. The teacher can ask: 'What could a reporter actually know on April 19? Who counted? Who benefited from a lower number?' This turns the error into a lesson about provisional evidence and about how power shapes what gets published.
Then have students compare the first-day report with a later account and note what changed and why. The takeaway is not cynicism about newspapers but a disciplined question: what is this source in a position to know, and what would I need to check?