Lexis and structure: six items, one candidate, four broken rules
A candidate's answers to six lexis and structure items are returned with the choices he made:
- Neither the principal nor the teachers ___ present. (was / were) - he chose 'was'.
- A number of students ___ absent. (was / were) - he chose 'was'.
- The number of applicants ___ risen sharply. (has / have) - he chose 'have'.
- He is one of the boys who ___ won a prize. (has / have) - he chose 'has'.
- Hardly had he sat down ___ the bell rang. (when / than) - he chose 'than'.
- She is capable ___ doing it. (of / to) - he chose 'to'.
His teacher notes that he is quick at vocabulary items but loses marks on items with more than one noun in them, and that he revises by memorising lists of prepositions without using them in sentences. He has four days before the examination and wants to know which single area to revise.
1. Group his six errors by the rule each one breaks, and name the one rule that would repair the most marks in four days.
Read the worked answer
Concord (subject-verb agreement): items 1-4. That is four of six marks, and they have a single underlying skill - locating the real subject instead of grabbing the nearest noun.
- 'Neither ... nor' takes a verb agreeing with the nearer subject: 'teachers' → were.
- 'A number of' + plural noun is plural → were.
- 'The number of' + plural noun is singular (the number itself) → has.
- 'One of the boys who ...' - 'who' refers to 'boys', so → have. This one needs the extra step of identifying the relative pronoun's antecedent.
Collocation/idiom: items 5-6. 5. 'Hardly ... when', not 'than'. 'Than' belongs to comparatives ('no sooner ... than'). 6. 'Capable of', not 'capable to'.
So: concord first. It is one decision procedure applied four ways, it is testable in minutes, and it accounts for two thirds of his losses here. Preposition lists come second, and should be learned in sentences, since items 5 and 6 are both failures of collocation rather than of meaning.
2. Explain items 2 and 3 together, and say why a candidate who knows the rule still gets them wrong.
Read the worked answer
Items 2 and 3 are a matched pair designed to trap. The strings 'a number of students' and 'the number of applicants' differ by one word, and the verb flips.
- 'A number of students were absent': here 'a number of' is a quantifier meaning 'several', so the head is 'students' → plural.
- 'The number of applicants has risen': here 'the number' is the head, a single figure, and 'of applicants' merely describes it → singular.
The candidate knows both rules in isolation. He fails because he matches the verb to the noun nearest the gap ('students', 'applicants'), and both sentences put a plural noun immediately before the verb slot. The surface is identical; only the article and the structure of the subject differ.
Repair method: bracket the subject before choosing. '[A number of students] were'; '[The number of applicants] has'. Say the bracket aloud. This converts a memory task into a parsing task, which is what the item is testing.
3. Design a four-day revision routine that would have prevented these errors, and justify each element.
Read the worked answer
Day 1 - Concord, one rule at a time: 20 items built from the four patterns above, with the subject bracketed in pencil before answering. Justification: the errors are not four separate facts but one procedure, and bracketing makes the procedure visible.
Day 2 - Adversarial pairs. Rewrite each of his wrong items into a correct/incorrect twin ('A number of students was' / 'The number of students were'). Justification: recognition of near-identical twins is exactly the discrimination the examination tests; the candidate must produce the contrast, not read it.
Day 3 - Prepositions in sentences, not lists. Ten verbs/adjectives with their prepositions, each written into an original sentence about school life ('capable of', 'insist on', 'prefer to'). Justification: he already recognises the words; what fails is production under time, and lists do not train production.
Day 4 - Timed mixed set of 30 items in 20 minutes, then mark and write the rule beside each error. Justification: mixed, timed conditions approximate the paper and expose whether the bracket habit survives pressure.
Excluded deliberately: reading through a grammar textbook. Four days is too short for coverage; it is long enough for one procedure to become automatic.