Secondary 4 English vocabulary becomes more reliable when students can check how words are actually used. Grade 10 students searching for Secondary 4 vocabulary, Grade 10 vocabulary, corpus vocabulary, concordance lines, collocations, academic vocabulary, word frequency, real English examples, vocabulary in context, dictionary examples and correct word usage need a method for answering practical questions: Do these words normally occur together? Is this phrase common or merely possible? Which preposition usually follows the verb? Is the formal-looking synonym actually natural in academic writing?
A corpus is a large structured collection of real language data that can be searched for patterns. A concordance displays multiple examples of a search word or phrase in context so learners can compare repeated usage. Corpus evidence can reveal frequency, collocation, grammar patterns, register, phraseology and recurring multiword expressions. It does not replace dictionaries, teachers or judgment; it gives learners another form of evidence when intuition is uncertain.
This article is the Secondary 4 / Grade 10 owner for corpus-informed vocabulary learning under the Secondary 4 English Vocabulary apex. It connects to Independent Vocabulary Learning, Lexical Chunks and Multiword Expressions, and the Secondary 4 Collocations owner. Its distinct job is verification: how a Secondary 4 learner uses authentic language evidence to test word choice, phrase patterns and natural usage.
50-second router
- If a phrase sounds possible but strange, check several authentic examples.
- If a thesaurus suggests an unfamiliar synonym, compare frequency and context before using it.
- If two collocations compete, inspect which one appears naturally with the intended meaning.
- If a preposition is uncertain, search the full word pattern rather than the word alone.
- If an AI suggestion sounds fluent, verify it against a reliable dictionary and corpus evidence.
- If one concordance line is confusing, inspect several; single examples can be unusual.
- If a frequent pattern appears across many lines, record the phrase as a chunk.
- If the examples come from one narrow genre, do not generalise automatically to all English.
- If a word changes meaning by context, group concordance lines by sense.
- If data become overwhelming, ask one specific question before searching.
What a corpus can show
frequency;
recurring neighbours;
collocations;
grammar patterns;
prepositions;
word-family behaviour;
register;
genre differences;
technical versus everyday senses;
multiword expressions.
Corpus work turns vocabulary study into pattern investigation.
What a corpus cannot decide alone
A corpus does not automatically tell a student what is best for a particular essay.
Frequency does not equal quality.
Rare does not equal wrong.
Common does not equal appropriate.
Interpretation still needs context, audience and meaning.
Concordance lines
A concordance line shows a search term surrounded by words from authentic examples.
If the search term is evidence, lines may reveal patterns such as:
strong evidence
provide evidence
evidence suggests
evidence of
evidence that
The student observes repeated structure.
Pattern before rule
Data-driven learning often asks learners to notice patterns first.
Search.
Compare.
Hypothesise.
Verify.
Generalise cautiously.
This process develops learner autonomy because students become language investigators.
Collocation
Corpus research is especially useful for collocation.
strong evidence is common and natural.
powerful evidence may be understandable but is less conventional in many contexts.
A corpus makes repeated co-occurrence visible.
Grammar patterns
Search contribute.
The learner may notice contribute to.
Search prevent.
Patterns may show prevent someone from doing.
Search depend.
Depend on becomes visible.
Vocabulary and grammar meet in corpus evidence.
Register
A phrase may be frequent in conversation but rare in formal academic prose.
Corpus composition matters.
Students should ask:
What kind of texts are in this corpus?
Who produced them?
When?
For what purpose?
A pattern from social media should not automatically guide a formal essay.
Frequency
Frequency helps prioritise words and phrases.
High-frequency useful language deserves more attention.
But raw frequency can mislead if one corpus overrepresents a topic or genre.
Frequency is evidence, not a command.
Dispersion
A word may appear many times because one source repeats it.
Dispersion asks how widely the word appears across texts or sources.
For learners, broad recurrence can be more informative than raw count alone.
Keyness
Keyness compares a target corpus with a reference corpus to identify unusually characteristic vocabulary.
Students do not need advanced statistics to use the underlying idea: some words are especially typical of a topic, genre or community.
Concordance reading
Concordances are not normal paragraphs.
The learner reads across examples to identify what repeats.
Look left.
Look right.
Group patterns.
Notice grammar.
Notice meaning.
Do not try to read every line as a full story.
Sense grouping
A word such as issue may produce several meanings.
issue a statement
an important issue
the latest issue of a journal
Group lines by sense before drawing conclusions.
Lemmas and word families
Searching analyse may miss analysis or analytical depending on the tool.
Corpus searches can be designed around lemmas or families.
Students should understand what the search tool is actually counting.
Corpus evidence and dictionaries
Modern dictionaries are heavily informed by corpus evidence.
Dictionaries summarise patterns.
Corpora expose examples.
Use both.
A learner dictionary is efficient for a quick answer; concordance evidence is useful when the learner wants to inspect variation and natural phraseology.
Corpus evidence and AI
AI can propose phrases instantly.
Corpus evidence asks whether those phrases are attested, frequent, genre-appropriate and used with the intended meaning.
The best workflow is:
draft;
generate options;
verify;
choose;
reuse independently.
Corpus evidence and writing
Students can investigate a phrase from their own writing.
“big evidence”
Search evidence.
Observe strong evidence, compelling evidence, substantial evidence.
Revise.
The learner repairs a real error rather than memorising a random list.
Corpus evidence and reading
Concordances can provide multiple contexts for one word.
This supports depth.
The learner sees how meaning shifts while form remains stable.
Corpus evidence and subject vocabulary
A specialised corpus can reveal how a term behaves in Science, History, Economics or another discipline.
Technical phraseology can differ from general English.
This supports disciplinary literacy.
Corpus evidence and lexical chunks
Repeated sequences become visible.
as a result of
in the context of
the extent to which
evidence suggests that
These patterns can be learned by function and register.
Corpus evidence and independent learning
The goal is not to make every Secondary 4 student a corpus linguist.
The goal is to teach a verification habit.
When unsure, look for repeated evidence.
Ask a precise question.
Compare several examples.
Record the pattern.
Use it.
Research-informed foundations
Recent Cambridge research on corpus linguistics for language teaching describes data-driven learning as learner engagement with authentic corpus examples, often through concordance and collocation analysis. Cambridge work on collocations synthesises corpus-based research into second-language phrase learning. A 2026 Cambridge handbook chapter on dictionaries and lexicography describes how corpora support frequency analysis, collocations, syntactic patterns and example selection, while also noting that AI will not remove the need for corpus evidence and expert human feedback.
Sources and further reading
- Cambridge Core: Corpus Linguistics for Language Teaching and Learning
- Cambridge: Collocations, Corpora and Language Learning
- Cambridge: Learning from a Corpus
- Cambridge 2026: Dictionaries and Lexicography
The Grade 10 corpus workflow
- Start with a specific question.
- Search the word or phrase.
- Inspect several lines.
- Group recurring patterns.
- Separate different senses.
- Check register and genre.
- Compare with a learner dictionary.
- Record one or two useful patterns.
- Use them in a new sentence.
- Recheck later if uncertainty remains.
Grade 10 corpus-verification laboratory
These laboratories do not invent corpus counts. They teach the questions a learner can investigate and the patterns that should be checked against several authentic examples.
evidence: corpus question
Question: Which adjectives and verbs naturally combine with evidence?
Patterns to inspect: strong/compelling/substantial evidence; provide evidence; evidence suggests/supports. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare strength and grammar around the noun. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not infer that a common phrase automatically proves the underlying claim. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
decision: corpus question
Question: Which verbs commonly precede decision?
Patterns to inspect: make/reach/reverse/announce a decision. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare make and reach, then inspect surrounding contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Avoid translating a first-language verb pattern directly if English convention differs. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
conclusion: corpus question
Question: Which verbs form natural phrases with conclusion?
Patterns to inspect: draw/reach/come to a conclusion. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect whether the phrase describes reasoning or final judgment. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not choose a phrase only because every word is individually correct. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
impact: corpus question
Question: Which prepositions and modifiers appear around impact?
Patterns to inspect: impact on; significant/negative/long-term impact. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate noun patterns from the verb impact. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Register and discipline can affect preferred phrasing. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
effect: corpus question
Question: How does effect differ from impact in phraseology?
Patterns to inspect: effect on; have an effect; adverse/positive effect. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare surrounding verbs and prepositions. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Semantic overlap does not make collocations identical. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
cause: corpus question
Question: How do noun and verb uses differ?
Patterns to inspect: cause of; cause concern; cause damage. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group examples by part of speech and sense. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not confuse cause with factor or trigger. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
factor: corpus question
Question: Which modifiers signal importance?
Patterns to inspect: key/major/contributing factor; factor in. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect whether factor is singular cause or one influence among several. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Avoid turning factor into sole cause. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
risk: corpus question
Question: Which verbs occur with risk?
Patterns to inspect: pose/reduce/increase/mitigate risk; risk of. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group prevention, exposure and probability contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Risk phrases often encode degree and domain. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
issue: corpus question
Question: How many senses appear?
Patterns to inspect: important issue; latest issue; issue a statement. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group concordance lines by sense before recording patterns. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: One headword can hide several unrelated study needs. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
problem: corpus question
Question: Which verbs describe response?
Patterns to inspect: address/tackle/solve/identify a problem. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare address versus solve. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Frequency cannot erase semantic strength differences. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
challenge: corpus question
Question: Which phrases sound natural?
Patterns to inspect: pose/face/overcome a challenge; significant challenge. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect whether challenge is literal difficulty or positive framing. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not replace every problem with challenge. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
significant: corpus question
Question: What nouns follow significant?
Patterns to inspect: significant factor/difference/increase/impact. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare general academic and technical statistical contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: A corpus may mix senses; sort them. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
substantial: corpus question
Question: Which nouns attract substantial?
Patterns to inspect: substantial evidence/increase/amount/support. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare formal academic use with everyday contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Substantial is not merely a longer huge. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
marginal: corpus question
Question: How does marginal behave?
Patterns to inspect: marginal improvement/effect/increase. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare technical and general senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not confuse with marginalised. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
feasible: corpus question
Question: Which modifiers and nouns appear?
Patterns to inspect: financially/technically feasible; feasible option/solution. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect constraint language nearby. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Feasible is not identical to possible. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
possible: corpus question
Question: Which structures follow possible?
Patterns to inspect: possible explanation/outcome; it is possible that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with plausible and feasible. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Different grammatical frames can reveal different uses. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
plausible: corpus question
Question: Which nouns commonly follow?
Patterns to inspect: plausible explanation/account/scenario. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare evidential contexts with possible. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Plausible adds reasonable believability. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
effective: corpus question
Question: Which patterns occur?
Patterns to inspect: effective strategy/method; effective at/in. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with efficient. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Always identify intended outcome. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
efficient: corpus question
Question: Which nouns and modifiers recur?
Patterns to inspect: efficient system/process; energy-efficient; highly efficient. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare resource-use contexts with effective. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Positive evaluation does not make the words interchangeable. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
equitable: corpus question
Question: Which domains use equitable?
Patterns to inspect: equitable access/distribution/outcomes. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect fairness contexts and compare equal. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Corpus evidence can show domain concentration. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
sustainable: corpus question
Question: Which dimensions appear?
Patterns to inspect: environmentally/financially sustainable; sustainable practice/growth. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group examples by environmental, financial and general use. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Always identify what must be sustained. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
proportionate: corpus question
Question: Which preposition appears?
Patterns to inspect: proportionate to; proportionate response. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect legal/policy versus general contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not reduce to equal. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
credible: corpus question
Question: Which nouns collocate?
Patterns to inspect: credible source/explanation/account. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with reliable. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Believability and consistency differ. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
reliable: corpus question
Question: Which nouns and verbs appear?
Patterns to inspect: reliable source/measure/data; reliably predict. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate adjective and adverb patterns. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Reliability is not validity. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
valid: corpus question
Question: What senses occur?
Patterns to inspect: valid argument/reason/ticket/measure. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group logical, administrative and methodological senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: One frequent sense may not fit school analysis. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
relevant: corpus question
Question: Which preposition is common?
Patterns to inspect: relevant to; relevant evidence/information. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect what relevance connects to. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Interesting does not equal relevant. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
sufficient: corpus question
Question: Which structures occur?
Patterns to inspect: sufficient evidence/time; sufficient to; sufficiently clear. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Look for threshold language. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Ask sufficient for what purpose. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
representative: corpus question
Question: Which patterns appear?
Patterns to inspect: representative sample; representative of. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare statistical and symbolic senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Context controls interpretation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
corroborate: corpus question
Question: What objects follow the verb?
Patterns to inspect: corroborate an account/claim/evidence. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect source-evaluation contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Corroboration requires additional support, not repetition. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
contradict: corpus question
Question: Which subjects and objects recur?
Patterns to inspect: evidence contradicts claim; contradict a statement. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare logical conflict with simple disagreement. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Two different views are not always contradictory. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
claim: corpus question
Question: How do noun and verb uses differ?
Patterns to inspect: make/support a claim; claim that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect stance around reporting. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Claim need not imply falsehood. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
argument: corpus question
Question: Which adjectives and verbs occur?
Patterns to inspect: strong/weak argument; make/develop/support an argument. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare argument with claim. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: A single claim is not necessarily a full argument. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
assumption: corpus question
Question: Which verbs occur?
Patterns to inspect: make/challenge/question an assumption; underlying assumption. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Look for reasoning contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not treat assumption as fact. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
criterion: corpus question
Question: Which verbs and adjectives recur?
Patterns to inspect: meet/apply a criterion; key criterion. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Check singular criterion versus plural criteria. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Corpus examples can reinforce irregular plural use. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
implication: corpus question
Question: Which modifiers occur?
Patterns to inspect: practical/wider/important implication; implication for. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare result with implication. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Implication often means significance, not direct result. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
perspective: corpus question
Question: Which frames occur?
Patterns to inspect: from the perspective of; alternative perspective. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect stakeholder and analytical contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Perspective is not simply opinion. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
bias: corpus question
Question: Which compounds recur?
Patterns to inspect: selection/reporting/confirmation bias; reduce bias. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group by mechanism. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not use bias as a generic insult. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
constraint: corpus question
Question: Which modifiers recur?
Patterns to inspect: budget/time/resource constraint; constraint on. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect practical and technical contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Constraint narrows options but does not mean impossibility. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
trade-off: corpus question
Question: Which preposition appears?
Patterns to inspect: trade-off between; make/consider trade-offs. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect two-sided benefit/cost structures. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Name both sides of the exchange. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
capacity: corpus question
Question: Which patterns occur?
Patterns to inspect: capacity to; production/storage capacity; limited capacity. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group ability and amount senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not flatten human and system senses. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
allocate: corpus question
Question: Which objects follow?
Patterns to inspect: allocate funds/resources/time; allocate to. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect distribution contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Record verb pattern, not definition alone. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
implement: corpus question
Question: Which nouns occur?
Patterns to inspect: implement a policy/plan/system; implementation of. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare proposal versus action. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Announcement is not implementation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
monitor: corpus question
Question: Which objects occur?
Patterns to inspect: monitor progress/performance/conditions. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect systematic repeated observation. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Monitor is more structured than watch in many formal contexts. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
refine: corpus question
Question: Which objects occur?
Patterns to inspect: refine a method/argument/model. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with improve and revise. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Refine often signals careful adjustment. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
resilient: corpus question
Question: Which nouns occur?
Patterns to inspect: resilient system/community/economy. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect recovery/adaptation contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not equate with simply strong. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
ambiguous: corpus question
Question: Which nouns attract the adjective?
Patterns to inspect: ambiguous wording/statement/meaning. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare unclear. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Ambiguity means more than one plausible interpretation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
explicit: corpus question
Question: Which nouns and patterns occur?
Patterns to inspect: explicit instruction/reference/statement; explicitly state. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with clear. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Explicit concerns direct statement. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
implicit: corpus question
Question: Which nouns recur?
Patterns to inspect: implicit assumption/meaning/bias. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with hidden or unstated. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Infer from context; do not assume. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
coherent: corpus question
Question: Which nouns appear?
Patterns to inspect: coherent argument/explanation/account. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with clear and cohesive. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Coherence is logical unity, not grammar alone. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
concise: corpus question
Question: Which nouns occur?
Patterns to inspect: concise summary/explanation/description. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with brief. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Concise implies economy without unnecessary loss. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
precise: corpus question
Question: Which nouns occur?
Patterns to inspect: precise language/measurement/definition. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with accurate. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Precision concerns specificity or repeatability depending on domain. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
accurate: corpus question
Question: Which nouns appear?
Patterns to inspect: accurate information/measurement/prediction. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with precise. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Accuracy concerns correctness relative to truth/reference. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
approximate: corpus question
Question: Which noun/adverb uses occur?
Patterns to inspect: approximate value/number; approximately. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect estimation contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Approximate signals closeness, not exactness. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
prevalent: corpus question
Question: Which preposition or frame appears?
Patterns to inspect: prevalent among/in; highly prevalent. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with popular/common. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Widespread does not mean liked. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
conventional: corpus question
Question: Which nouns recur?
Patterns to inspect: conventional method/view/approach. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with common and traditional. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Frequency and conformity are different. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
novel: corpus question
Question: Which academic phrases appear?
Patterns to inspect: novel approach/method/finding. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate adjective sense from noun novel. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Novel can mean new/original in formal contexts. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
obsolete: corpus question
Question: Which nouns occur?
Patterns to inspect: obsolete technology/system/practice. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with old/outdated. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Not every old item is obsolete. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
robust: corpus question
Question: Which nouns recur?
Patterns to inspect: robust evidence/method/system. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect research and engineering contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Meaning often involves strength under variation or challenge. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
tentative: corpus question
Question: Which nouns occur?
Patterns to inspect: tentative conclusion/plan/estimate. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with weak. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Tentative signals caution or provisional status. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
subsequent: corpus question
Question: Which nouns occur?
Patterns to inspect: subsequent analysis/events/study. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with next/later. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Formal sequence language has genre patterns. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
preceding: corpus question
Question: Which nouns occur?
Patterns to inspect: preceding paragraph/section/events. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with previous. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Inspect written academic register. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
derive: corpus question
Question: Which preposition appears?
Patterns to inspect: derive from; derive benefit/meaning. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group origin and obtain senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Verb patterns reveal sense. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
synthesise: corpus question
Question: Which objects recur?
Patterns to inspect: synthesise information/evidence/findings. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with summarise. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Synthesis combines, not merely shortens. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
differentiate: corpus question
Question: Which prepositions occur?
Patterns to inspect: differentiate between; differentiate X from Y. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Record both frames. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Vocabulary learning includes syntactic patterning. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
prioritise: corpus question
Question: Which objects occur?
Patterns to inspect: prioritise tasks/resources/safety. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect decision contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Priority implies ranking. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
adapt: corpus question
Question: Which frames occur?
Patterns to inspect: adapt to; adapt a strategy/material. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate intransitive and transitive uses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Pattern affects grammar. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
transfer: corpus question
Question: Which nouns and prepositions recur?
Patterns to inspect: transfer knowledge/skills; transfer to/across. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect learning and physical senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Context selects domain. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
retrieve: corpus question
Question: Which objects occur?
Patterns to inspect: retrieve information/word/data. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare memory and physical retrieval. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Technical learning sense is common in education research. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
retain: corpus question
Question: Which objects recur?
Patterns to inspect: retain information/knowledge/control. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect memory and possession senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Do not reduce to remember. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
recall: corpus question
Question: Which noun/verb uses occur?
Patterns to inspect: recall a word; free recall; recall that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with recognise. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Generation and identification differ. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
recognise: corpus question
Question: Which objects and patterns appear?
Patterns to inspect: recognise a word/pattern; recognise that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with recall. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Recognition supplies a cue. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
context: corpus question
Question: Which compounds recur?
Patterns to inspect: context clue; wider context; in context. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect linguistic and situational uses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Context is not one thing. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
collocation: corpus question
Question: Which modifiers and verbs occur?
Patterns to inspect: strong collocation; collocates with; collocation pattern. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect linguistic terminology. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Corpus study can teach its own metalanguage. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
register: corpus question
Question: Which phrases occur?
Patterns to inspect: formal/informal register; register variation. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate language-style sense from administrative register. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Sense grouping is essential. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
connotation: corpus question
Question: Which verbs/nouns recur?
Patterns to inspect: negative/positive connotation; carry a connotation. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with denotation. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Association differs from core meaning. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
denotation: corpus question
Question: Which contrast phrases appear?
Patterns to inspect: denotation and connotation; literal denotation. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect semantics contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Likely low frequency outside language study. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
polysemy: corpus question
Question: Which academic contexts use the term?
Patterns to inspect: polysemy; polysemous word/sense. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect linguistic register. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Technical terms may be rare but locally essential. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
morphology: corpus question
Question: Which phrases recur?
Patterns to inspect: morphological awareness/structure/analysis. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare educational and linguistic contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Technical phraseology matters. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
paraphrase: corpus question
Question: Which verbs/nouns occur?
Patterns to inspect: paraphrase a passage; accurate paraphrase. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect education/academic writing contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Corpus examples can show grammar around the task word. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
summary: corpus question
Question: Which collocations recur?
Patterns to inspect: concise/brief summary; provide/write a summary. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect academic and general use. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Concise and brief differ slightly. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
qualify: corpus question
Question: Which academic sense occurs?
Patterns to inspect: qualify a claim/statement. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Separate from meet requirements sense. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Polysemy must be grouped. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
concede: corpus question
Question: Which objects/clauses occur?
Patterns to inspect: concede a point; concede that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect argumentative contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Concede often implies reluctant acknowledgment. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
rebut: corpus question
Question: Which objects occur?
Patterns to inspect: rebut an argument/claim/criticism. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with refute and disagree. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Register is formal. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
correlate: corpus question
Question: Which preposition appears?
Patterns to inspect: correlate with; strongly correlated. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect statistical contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Correlation does not prove causation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
predict: corpus question
Question: Which objects/clauses occur?
Patterns to inspect: predict outcome; predict that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect forecast/science contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Prediction can vary in certainty. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
observe: corpus question
Question: Which senses appear?
Patterns to inspect: observe a pattern; observe behaviour; observe that. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group perception and reporting uses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Sense grouping prevents overgeneralisation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
hypothesis: corpus question
Question: Which verbs occur?
Patterns to inspect: form/test/support a hypothesis. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect scientific register. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: A hypothesis is not a random guess. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
variable: corpus question
Question: Which compounds recur?
Patterns to inspect: independent/dependent/control variable. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare everyday adjective/noun senses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Technical collocations reveal subject meaning. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
trend: corpus question
Question: Which modifiers and verbs occur?
Patterns to inspect: upward/downward trend; trend shows/continues. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect data contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Trend requires more than one observation. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
proportion: corpus question
Question: Which prepositions occur?
Patterns to inspect: proportion of; large/small proportion. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect quantitative contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Compare with percentage. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
estimate: corpus question
Question: Which adjective/verb patterns occur?
Patterns to inspect: rough estimate; estimate that; estimate cost. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group noun and verb uses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Approximation is built into the meaning. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
assess: corpus question
Question: Which objects recur?
Patterns to inspect: assess performance/risk/impact. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare with evaluate. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Different domains prefer different objects. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
diagnose: corpus question
Question: Which objects occur?
Patterns to inspect: diagnose a condition/problem/weakness. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Compare clinical and educational uses. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Metaphorical extension should be appropriate. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
intervention: corpus question
Question: Which modifiers recur?
Patterns to inspect: targeted/educational/medical intervention. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Group domains. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Meaning varies with discipline. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
progress: corpus question
Question: Which verbs occur?
Patterns to inspect: make/monitor/show progress. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect learning/project contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: Progress is usually uncountable in these uses. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
mastery: corpus question
Question: Which phrases occur?
Patterns to inspect: vocabulary mastery; mastery of. These are useful candidates to compare, not fabricated frequency results.
Interpretation task: Inspect educational contexts. Group concordance lines by sense, grammar or genre before drawing a generalisation.
Caution: High competence is stronger than familiarity. One line is an example; repeated distribution across relevant texts is stronger evidence.
Production task: Choose one pattern that matches the intended meaning and write a fresh Grade 10 sentence. Then replace one component and decide whether the phrase remains natural, changes meaning or becomes unusual.
Verification loop: Compare the pattern with a learner dictionary, then revisit corpus examples. If dictionary and corpus evidence disagree, inspect sense, register, date and corpus composition before deciding.
Delayed transfer: Reuse the verified pattern in another topic a week later without reopening the corpus. The goal is independent ownership, not permanent dependence on search.
Corpus-learning workflows
Concordance left-right scan
Place the search word in the centre and inspect three to five words on each side. Mark repeated neighbours before reading full sentences.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Sense sorting
Print or copy ten concordance lines and group them by meaning. Label each sense in simple language.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Collocation hunt
Search a target noun and list recurring adjective, verb and preposition partners. Verify the most useful ones in a dictionary.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Preposition hunt
Search the verb plus nearby prepositions. Compare contribute to, depend on and distinguish between patterns.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Register comparison
Compare the same phrase in academic, news and conversational corpora if available. Note distribution rather than assuming one genre represents all English.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Thesaurus verification
Take three thesaurus alternatives and inspect authentic contexts before adopting any of them.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
AI verification
Ask AI for alternatives only after drafting. Search suspicious phrases in a reliable corpus or dictionary before reuse.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Error repair
Take one phrase from a student’s writing, search the headword, identify natural patterns, revise and then reuse after delay.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Phrase-family study
Search evidence suggests, evidence indicates and evidence demonstrates. Compare the strength and contexts of the verbs.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Word-family search
Compare analyse, analysis and analytical in surrounding grammar. Record typical noun/verb/adjective roles.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Technical sense search
Search function or significant inside subject-specific examples and compare with general English uses.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Concordance-to-rule
After inspecting examples, write a tentative rule. Then test the rule against more lines and revise it if exceptions appear.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Rule-to-concordance
Start with a textbook rule, search examples and see how actual usage supports or complicates it.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Chunk mining
Identify multiword sequences that recur around a high-utility academic word and record them by function.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Frequency caution
Compare raw count with spread across documents where the tool permits. Avoid treating one repeated source as broad usage.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Keyword exploration
Compare a specialised set of texts with a general reference set to find characteristic vocabulary, then interpret why the words are distinctive.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Narrow-reading concordance
Collect examples from one topic so the same vocabulary recurs in related contexts.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Cross-topic concordance
Compare the same word across unrelated topics to test how stable the meaning and phraseology remain.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Source-quality audit
Ask who produced the corpus, what genres it contains, when texts were created and how representative it is for the learner’s question.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Notebook conversion
Turn corpus findings into a compact entry: phrase, meaning, register, one example and one retrieval cue.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Delayed reuse
Use a verified collocation one week later without reopening the corpus. Corpus investigation should end in independent language use.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Classroom mini-corpus
Collect anonymous class writing around one task and inspect repeated lexical errors or successful phrase patterns.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Subject mini-corpus
Collect several school Science or Humanities passages and identify recurring academic verbs and collocations.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Exam-style corpus
Compare several public specimen or practice texts where copyright permits and identify repeated task vocabulary rather than reproducing passages.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Corpus-plus-dictionary triangulation
Check a phrase in corpus evidence and a learner dictionary; treat agreement as stronger evidence than either source alone.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Corpus-plus-reading triangulation
Notice a phrase in real reading, search wider usage, then return to the original text with a stronger pattern model.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Pattern confidence scale
Label findings as strong repeated pattern, plausible pattern, rare/specialised pattern or insufficient evidence.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Avoiding cherry-picking
Do not select only examples that support the phrase you wanted to use. Inspect a broad enough sample to see competing patterns.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Comparing near-synonyms
Search two near-synonyms and compare their noun partners, register and semantic environments.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Independent corpus question
Require the learner to formulate a precise usage question before opening the search tool.
Success standard: The learner can state the question, inspect multiple examples, avoid cherry-picking, distinguish sense and register, formulate a cautious pattern, and turn evidence into independent language use.
Evidence discipline: Do not report an exact frequency unless the tool actually provides it and the corpus scope is understood. Use phrases such as recurring pattern, common in this sample or attested example when that is all the evidence supports.
Twelve-week corpus-informed vocabulary programme
Week 1 teaches what corpora and concordances are. Week 2 practises left-right scanning. Week 3 groups senses. Week 4 investigates collocations. Week 5 investigates grammar patterns and prepositions. Week 6 compares near-synonyms. Week 7 compares register and genre. Week 8 investigates word families. Week 9 verifies student writing. Week 10 compares corpus evidence with AI and dictionaries. Week 11 uses subject-specific mini-corpora. Week 12 asks students to formulate and answer their own usage questions.
Final principle
Corpus evidence gives Secondary 4 learners a disciplined way to ask whether language is merely possible or genuinely attested and patterned. The strongest learner does not worship frequency or copy concordance lines. The learner asks a precise question, compares enough evidence, checks meaning and register, records the useful pattern, and then closes the tool and uses English independently.
Corpus evidence-quality clinic
Attested versus typical
A phrase appearing once proves only that it occurred. It does not show that the phrase is typical. Compare several lines and look for recurrence across independent texts before treating the pattern as conventional.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Corpus composition
A corpus of newspapers, student essays, social media posts or scientific articles will produce different vocabulary patterns. Always ask what texts are included before generalising from the results.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Genre bias
A phrase that appears often in legal writing may sound unnatural in a personal narrative. Record genre alongside the pattern so frequency is interpreted within context.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Dispersion
Repeated examples from one document can inflate apparent frequency. When possible, inspect whether a pattern appears across many texts or only one source.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Sense contamination
A search for issue may mix topic, publication and verb senses. Group the lines by sense before counting or interpreting collocations.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Lemma decisions
Searches may treat analyse, analysed, analysing and analysis differently depending on settings. Know whether the tool is searching one form, a lemma or a family.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Case and punctuation
Capitalisation, punctuation and search settings can change results. A careful learner records what was actually searched rather than treating the output as transparent.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Rare but correct
Some precise phrases are rare because the situation itself is rare. Low frequency does not automatically mean wrong. Check meaning, genre and authoritative usage.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Common but unsuitable
A very frequent conversational phrase may still be inappropriate for a formal report. Frequency and register answer different questions.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
AI suggestion audit
If AI proposes an unfamiliar phrase, first inspect dictionary evidence, then several authentic corpus examples. If the phrase is attested only in an unusual sense or genre, do not transfer it blindly.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Dictionary-corpus disagreement
A dictionary may prioritise general usage while a specialised corpus shows a technical pattern. Identify the learner’s actual target genre before deciding which evidence matters most.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Small sample caution
Ten concordance lines can reveal a useful pattern but cannot justify sweeping statistical claims. Phrase conclusions cautiously: ‘the examples suggest’ rather than ‘English always uses’.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Negative evidence caution
Failure to find a phrase does not prove impossibility. The corpus may be too small, too specialised or poorly matched to the query.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Collocation strength caution
Two words may co-occur because the topic puts them nearby, not because they form a conventional phrase. Read enough context to distinguish topical association from lexical collocation.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Corpus-to-production transfer
After verifying a phrase, close the tool and write a fresh sentence in a new topic. If the learner can only reuse the exact concordance sentence, the pattern is not yet independently owned.
Student decision: State exactly what the evidence supports and what it does not support. Separate observed examples from claims about frequency, naturalness, register and correctness.
Verification habit: When the question matters for graded writing, combine corpus evidence with a learner dictionary or another reliable reference. Agreement across sources strengthens confidence; disagreement should trigger closer inspection of sense and genre.
Transfer: Apply the same evidence-quality check to a different word or phrase. The long-term skill is not memorising one corpus result but learning how to evaluate language data responsibly.
Final corpus-verification checklist
Before adopting a new phrase, ask: Is it attested? Does it recur? Does it occur across more than one source? Which sense is active? Which genre is represented? Is the collocation genuinely conventional? Does a learner dictionary support it? Can I use it independently in a new sentence? If the answer chain is incomplete, keep the phrase in the verification queue rather than treating it as mastered.