Interpret corpus data and historical language
Read an n-gram graph, corpus word table and historical text together, then use their evidence in a measured analysis.
Section Paper 3 · Language Analysis
Exam scope and assessment notes
These original lessons support study of the official syllabus; they are not Cambridge materials. Practice is self-checked and does not award Cambridge marks or grades. Extended writing and analytical essays need qualified subject-teacher review. Full timed papers and examiner-marked learner responses are not provided here. Interpret language choices in context and qualify claims about audience response.
Before you begin
Work on paper. Try each question before revealing help, and explain why your method works.
Quick readiness check
A word table lists 'school' near a target word. What does this support?
What you will learn
- Read graph axes, scales and labels before interpreting a trend.
- Explain what corpus tables and collocates do and do not show.
- Synthesize sources without overclaiming.
Check what each data source measures
Paper 3 Section A combines a written text from around 1500 to the present with an n-gram graph and a corpus word table. A graph may track relative frequency of a phrase over time; a table may show collocates or synonyms.
Check corpus, period, normalisation, labels and scale. A trend records use in the selected dataset; it does not identify a cause or represent every speaker. A collocate list shows nearby words in that corpus, not a complete definition.
Synthesize instead of listing
Use the historical text to examine a feature in context, the graph to describe a pattern and the table to explore associations or alternatives. Explain where the sources support one another and where they answer different questions.
Write 'the graph records an increase in this corpus' rather than 'everyone began saying it'. Combine data with relevant research and explain limitations.
Sources for this lesson
Pause and explain
A graph rises across two periods. Which conclusion best fits?
Worked example
An invented graph records 2, 5 and 9 occurrences per million words in successive periods. A table lists 'public' and 'service' nearby. What can you infer?
Show the worked solution
- Describe the rise in recorded frequency.
- Use collocates to suggest an association, not a fixed meaning.
- Note that corpus selection and other evidence are needed to explain the trend.
Answer The sampled corpus records a rise across the periods. The collocates suggest an association with public services in that dataset, but neither source alone proves why the trend occurred or how all speakers used the expression.
Try a topic test
Choose Easy, Normal or Difficult. Save your answers and open worked review after finishing.
Test this topic →From guided practice to a new situation
Write your answer and reasoning first. Use a hint only if you are stuck. The model solution is for self-checking; your written work is not automatically marked.
Invented frequencies per million: 1800s = 1, 1900s = 4, 2000s = 7. Give one observation and one limitation.
Give me a hint
Describe the measured pattern and keep the claim within the dataset.
Compare my reasoning
- State the increase across periods.
- Name it as the selected corpus's record.
- Say the graph does not identify cause or represent every speaker.
The selected corpus records an increase from 1 to 7 instances per million words across the periods shown. This does not by itself explain why the increase occurred or show how every English speaker used the expression.
Look for these in your work
- Uses the data accurately
- Notes corpus scope
- Avoids causation
A table lists 'rapid', 'slow' and 'careful' as near-synonyms of a historical word. How would you use it?
Give me a hint
Treat the entries as candidates, then test them against context and method.
Compare my reasoning
- Consider the possible semantic range.
- Check the historical word's sentence and surrounding use.
- Note that the table's method may affect its suggestions.
Use the list to consider a semantic range involving speed or manner, then test it against the historical text. The table may reflect how the corpus tool generated synonyms, so it cannot replace close reading or other evidence.
Look for these in your work
- Uses the table as evidence
- Checks context
- Notes a limitation
A historical text uses an older form, its graph declines, and a table shows a newer alternative. Write a synthesis point.
Give me a hint
Connect compatible evidence without claiming complete replacement.
Compare my reasoning
- Describe the older form in its text.
- State the graph's trend and table association.
- Frame replacement as a possible interpretation needing more evidence.
The text exemplifies the older form, while the graph records declining frequency and the table associates a newer alternative with related contexts. Together they are consistent with changing preferences, but do not prove the older form disappeared from all use.
Look for these in your work
- Integrates sources
- Distinguishes occurrence from trend
- Qualifies the interpretation
Common mistakes
- Ignoring axes, units or corpus scope
- Treating frequency as proof of cause
- Assuming collocates are exact synonyms
What can you explain now?
What does an n-gram graph directly show?
Compare with the explanation
A pattern in recorded frequency within its dataset
Interpretation depends on the graph and how its corpus was built.
After trying it yourself, choose your next review. This is your self-assessment.
Your review choice appears on Today. Sign in to sync it across devices.
Make it stick
Why is 'the corpus records an increase' more defensible than 'people began using the form'?
If you needed the solution, close it and solve the problem again from a blank page. A correct answer is stronger when you can explain the reason for each step.