A-Level Revision Aid: This resource is designed to support your revision. Content may contain errors. If you find a discrepancy with your class teaching, your teacher is correct — please let us know at alevelrevise@scott.scottrix.co.uk.
Language Frameworks and Methods
Year 1 / ASYear 2 / A-LevelAQA, OCR, WJEC/Eduqas
A* revision of the AQA 7702 and OCR H470 language levels - lexis, semantics, grammar, pragmatics, phonology and graphology - and the method for deploying them as systematic analysis of unseen texts for AO1, AO2 and AO3.
🎯 Learning Objectives
Classify word classes (open vs closed), phrase, clause and sentence types, and mood, and justify why each choice matters in an unseen text (AO1).
Apply semantic concepts - connotation, semantic fields, figurative language - to build representation claims rather than label devices (AO1/AO2).
Use Grice's maxims, speech act theory and face/politeness to explain inferred meanings and interpersonal stance (AO1/AO2).
Analyse phonemic, prosodic and graphological/multimodal features appropriate to the mode of the text (AO1/AO3).
Push every point through the chain quotation - terminology - effect - context, selecting features rather than listing them (AO1/AO3).
Evaluate frameworks critically: name the theorist, apply the concept, then judge its explanatory power for the text (AO2 - AQA 'critical understanding'; OCR 'critical understanding of concepts and issues').
📋 Key Definitions and Core Concepts
Word class (part of speech): The category of a word: open classes (nouns, lexical verbs, adjectives, adverbs) are productive and take new members; closed classes (determiners, pronouns, primary and modal auxiliaries, prepositions, conjunctions) form a fixed set.
Semantic field: A group of lexemes sharing a domain of meaning (e.g. a field of war, of disease); identifying the field matters only alongside what its selection represents.
Connotation: The associations a word carries beyond its denotation; 'slender' and 'scrawny' denote the same body type but connote approval and disorder respectively.
Transitivity: Halliday's system of process types - material, mental, relational, verbal, behavioural, existential - which analyses who or what the grammar casts as actor and what kind of action is represented.
Mood: The grammatical system of declarative, interrogative, imperative and exclamative; it is interpersonal, matching sentence form to the discourse function of the text.
Conversational implicature: Grice's term for meaning generated when a speaker openly flouts a maxim of the Cooperative Principle, leaving the hearer to compute the unstated meaning.
Illocutionary force: In speech act theory (Austin, Searle), what an utterance does - warn, promise, apologise, command - as distinct from its locutionary form (the words) and perlocutionary effect (the impact on the hearer).
Face-threatening act (FTA): An act (request, criticism, refusal) that damages the addressee's positive face (want to be approved of) or negative face (want to be unimpeded); politeness strategies mitigate FTAs (Brown & Levinson).
Prosodics: Vocal effects beyond individual segmental sounds: intonation, pitch, stress, tempo, volume, pausing and rhythm; central to speech analysis alongside paralinguistics and non-fluency features.
Multimodality: The co-occurrence of semiotic modes - words, image, colour, typography, layout - within one text (Kress & van Leeuwen, 2006); analysed compositionally through information value, salience and framing.
📌 Key Points
AQA 7702 examines frameworks through Paper 1 Section A (70 marks: Q1 25, Q2 25, Q3 20) and Paper 2 Q3 (40 marks, two-text discourses analysis); OCR's most direct framework test is H470/01 Section A 'Language under the microscope' (2 x 10 marks).
AO1 (both boards) = apply appropriate methods with associated terminology and coherent expression: feature-spotting without effect caps the mark.
AO2 = critical understanding of concepts and issues: name the framework (Grice, Austin/Searle, Brown & Levinson, Halliday, Kress & van Leeuwen), apply it, then evaluate its explanatory power.
AO3 = analyse and evaluate how contextual factors and language features construct meaning: mode determines which levels matter - prosodics for speech, graphology for print and electronic, multimodality for screen texts.
Selection beats coverage: four features pushed deep through quotation, terminology, effect and context outscore twelve named and dropped.
Lexis and semantics: word classes, semantic fields, figurative language
Word classes are the starting point, but spotting them is not analysis - a top answer selects the class choices that carry the representational or interpersonal weight. Open classes are productive and take new members; closed classes form a fixed set.
Class
Subtypes to name precisely
What it signals in a text
Nouns (open)
Proper, common, abstract, collective, count and mass.
Abstract noun density ('freedom', 'loss') signals emotive or ideological writing; proper nouns anchor representation to specific people and places.
Verbs (open)
Dynamic vs stative; transitive vs intransitive; main vs auxiliary; modals (epistemic = possibility, deontic = obligation).
A shift from stative to dynamic verbs energises narrative; modal auxiliaries mark hedging in expert or political texts - epistemic 'may' vs deontic 'must' is a precise AO1 point.
Adjectives and adverbs (open)
Attributive vs predicative; gradable, comparative, superlative; adverbs of manner, time, place, degree.
Evaluative adjectives are the engine of representation ('unspoilt', 'shambolic'); intensifiers ('so', 'utterly') mark register and stance.
Pronoun choice maps audience and power (inclusive 'we', addressing 'you'); subordinating conjunctions build the hierarchy of complex prose.
Morphology feeds AO2: inflectional morphemes (plural -s, past -ed, comparative -er) vs derivational (un-, -ness, -ise) show how word class is created, and neologisms, blends and compounds are evidence for the Language Change strand (7702/2; H470/02 Section C). Semantics is where lexis earns its effect:
Denotation vs connotation: 'slender' and 'scrawny' denote thinness but connote approval and disorder - connotation carries representation.
Semantic fields: groups of lexemes sharing a domain (a field of war in a sports report, of disease in a tabloid scare story); name the field, then what its selection represents.
Lexical relations: synonymy, antonymy, hyponymy and hypernymy; euphemism vs dysphemism ('passed away' vs 'croaked') - register and attitude in a single choice.
Figurative language: metaphor (including extended and conceptual metaphor), simile, personification, metonymy ('the Crown'), hyperbole and irony. One extended metaphor doing representational work beats five scattered similes named without effect.
Applied lexis - the chain in miniature: take a hypothetical tabloid headline, 'Benefit scroungers drain the system again'. The lexis ('scroungers', 'drain') builds a semantic field of parasitism, with the history of derogation riding on 'scrounger'; the adverb 'again' is a presupposition trigger - earlier drains are assumed, no evidence given; and the noun phrase positions benefits claimants as agents of harm. Three levels, one sentence, one representation claim. That density - selection plus effect - is what separates AO1 analysis from word-spotting, and it is the same chain you run in every 25-mark answer (7702/1) and every OCR microscope item (H470/01).
Grammar: phrases, clauses, sentences, mood and transitivity
Grammar is the highest-yield level for AO1 in written texts because structure and effect can be named precisely. Build outward: phrases (noun, verb, adjective, adverb and prepositional phrases, each built around a head with pre- and post-modification) combine into clauses, which combine into sentences.
Clause elements: subject, verb, object (direct/indirect), complement and adverbial - SVOCA. Missing elements are analytically significant: verbless minor sentences ('No chance.') foreground and mimic speech.
Clause types: main vs subordinate (adverbial, relative and nominal clauses); embedded clauses pack information into modification - dense subordination is a hallmark of planned, written mode.
Sentence types: simple, compound, complex and compound-complex. Simple sentences give impact or clarity; complexes build qualification and hierarchy; a short sentence after a long one manipulates rhythm. Coordination presents ideas as equal; subordination ranks them.
Mood: declarative, interrogative, imperative, exclamative. Mood is interpersonal: interrogative headlines promise revelation, imperatives command the reader in adverts and instructions, exclamatives simulate excitement. Distinguish mood from function - 'Can you sit down?' is interrogative in form but directive in force, which is where pragmatics takes over.
Voice and transitivity are the representation levers:
Active vs passive: the passive can omit or demote the agent ('mistakes were made') - ask who the grammar lets off the hook and why.
Transitivity (Halliday): the process types - material ('do' actions), mental (perception, cognition, affect), relational (being, having), verbal (saying), behavioural and existential. Casting a participant in material processes ('plunge', 'crash') represents it as an agent of harm; relational processes ('is weak') represent it as a state. This is the standard bridge from grammar to representation on AQA 7702/1 Questions 1-3 and OCR H470/01 Section C.
Exam habit: pair every structural observation with its interpersonal or representational consequence - 'the passive withholds the agent, so responsibility is diffuse' is a point; 'there is a passive' is not.
Applied grammar - the chain in miniature: compare two hypothetical lead sentences. 'Police cracked down on protesters.' gives agency to the police - material process, active voice, transitive verb - and positions protesters as objects of control; 'Protesters were met with a police crackdown' demotes the protesters and nominalises the action ('crackdown'), diffusing agency across an event rather than an actor. Same facts, competing constructions - and the representation claim follows from the grammar alone. In complex prose, track the subordination: a periodic sentence with three embedded clauses signals a planned, specialist register, while parataxis (coordination with 'and', 'but') mimics the rhythm of speech.
Pragmatics: implicature, speech acts and face
Pragmatics explains meaning beyond the words: what is inferred, what is assumed, and what the utterance does.
Grice's Cooperative Principle and maxims: cooperation operates through four maxims - quantity (be as informative as required, not more), quality (be truthful), relation (be relevant), manner (be clear, avoid obscurity and ambiguity). Observing maxims cooperates; flouting them openly generates a conversational implicature - 'Lovely weather!' during a storm flouts quality to implicate irony. Violating (covert lying) is different: discovered, it destroys trust. Exam applications: advertising flouts quality through puffery and hyperbole; headlines flout manner for economy; data-dense texts flout quantity to perform authority.
Presupposition vs entailment: a presupposition is an assumption built into wording that survives negation ('she managed to finish' presupposes she tried; 'another delay' presupposes earlier ones); an entailment follows logically and does not survive negation. Triggers - definite descriptions, 'again', 'even', factive verbs ('realise', 'know') - are the fastest pragmatic point in headlines and political language.
Speech act theory (Austin, Searle): every utterance has a locutionary act (the words), an illocutionary force (what it does - warn, promise, apologise) and a perlocutionary effect (its impact on the hearer). Direct speech acts match form to function; indirect acts do not. Searle's classes: assertives (representatives), directives, commissives, expressives and declarations - a useful checklist for classifying what a text's sentences do.
Face and politeness (Goffman; Brown & Levinson): positive face is the want to be approved of; negative face is the want to be unimpeded. Face-threatening acts (requests, criticisms, refusals) damage one or both, and strategies mitigate them: bald on-record, positive politeness (solidarity), negative politeness (hedging, indirectness), off-record hints. This is the tenor engine: in asymmetrical encounters (doctor-patient, manager-employee) mitigation patterns with power - the bridge to the Language and Power strand.
Pragmatics earns AO2 when the framework is evaluated, not just applied: 'Grice assumes cooperation; the advert exploits that expectation commercially' is critical understanding.
Pragmatics in written texts: irony, presupposition and implicature are not speech-only. A headline flouting the maxim of quality through hyperbole ('Housing crisis: Britain at breaking point') implicates a scale it cannot evidence; a standfirst presupposing a problem ('How to fix our broken exams') positions the reader to accept the premise before any argument is made; and synthetic personalisation ('you') plus direct address collapses the distance between mass text and individual reader. Evaluate the frameworks as well: Grice's model assumes cooperation, but persuasive texts weaponise the expectation of cooperation - that judgement is the difference between applying a theory and understanding it (AO2).
Phonology, graphology and the method: applying frameworks to unseen texts
Phonology covers sound at and above the level of the phoneme; prosodics covers vocal effects beyond individual sounds; graphology covers the visual layer. Match the level to the mode - analysing prosodics in a printed leaflet wastes marks.
Phonemic features: consonant classes - plosives (/p b t d k g/), fricatives (/f v s z ʃ θ ð h/), affricates (/tʃ dʒ/), nasals (/m n ŋ/), approximants including liquids (/l r/) and glides (/w j/) - plus voicing. Sound patterning: alliteration, sibilance, assonance, consonance, onomatopoeia, rhyme and rhythm. Sound symbolism: plosive clusters feel percussive or aggressive; sibilance can hiss or soothe depending on the lexis it wraps. Use in transcripts (dialect phonology), headlines, slogans and literary prose.
Prosodics: intonation (rising = tentativeness or question, falling = assertion), pitch, stress - especially contrastive stress ('I didn't say HE stole it') - tempo, volume, pausing and rhythm; paralinguistic effects (laughter, voice quality) and non-fluency features (fillers, false starts, repair, micropauses). AQA's transcription key marks micropauses (.) and timed pauses (2.0) - analyse where they fall, not just that they exist.
Graphology: layout (columns, sections, white space), typography (font, weight, size, case), colour, images with captions, bulleting and headings. In multimodal texts (Kress & van Leeuwen, 2006) analyse compositionally: information value (left/right = given/new, top/bottom = ideal/real), salience (what the eye is drawn to first) and framing (what is connected or separated). A captioned photograph anchors an image's meaning; bold weight gives a claim authority before it is read.
Applying the frameworks - the method (AO1/AO2/AO3):
Read twice: first for genre, audience, purpose and what is represented or presented; second for quotations banked against each level.
Plan by angle, not by level: representational angle for AQA 7702/1 Q1/Q2 (25 marks each), point of comparison for Q3 (20) and OCR H470/01 Section C (36).
Push every point through the chain: quotation → terminology → effect → context. An effect without a quotation is assertion; a quotation without effect is description; context bolted on at the end caps AO3.
Let mode filter the levels: speech → prosodics and turn-taking; print → grammar and graphology; electronic → multimodality and engagement features.
For AO2, evaluate concepts: name the framework (Grice, Austin/Searle, Brown & Levinson, Halliday, Kress & van Leeuwen), apply it, then judge what it reveals and what it misses in this text.
Timing: AQA Paper 1 Section A is 70 marks in roughly 90 minutes - about 25 per text question and 20 for the comparison, after reading time. OCR H470/01 gives two hours for 80 marks: roughly 15 minutes per microscope item (Section A), 35 for the directed writing (Section B) and 55 for the comparison (Section C). Selection beats coverage every time.
Common errors that cap marks: (1) catalogue mode - listing devices with no representation claim; (2) context dump - a paragraph of production facts with no link to a linguistic choice; (3) theorist name-dropping - 'Grice' appended to a point the theory does not explain; (4) cross-mode comparison error - comparing prosodics against a printed text. For OCR's 'Language under the microscope' (H470/01 Section A, 2 x 10 marks), compress the chain: the item names the levels to analyse, so lead with the most significant feature, quote precisely, explain the effect in context, and stop - depth in ten marks, not coverage.
💡 Worked Example
Exam-style: Analyse how language in Text A creates meanings and representations. [25 marks] (AQA 7702/1 Question 1 or 2 style; OCR equivalent: H470/01 Section C, 36 marks). Shorter-form version: using appropriate linguistic terminology, analyse the lexis, grammar and graphology of this short charity appeal. [10 marks] (OCR H470/01 Section A, 'Language under the microscope' style)
Mark-scheme logic: Mark-scheme logic (AO1 + AO2 + AO3, per the AQA 7702/1 levels-based mark scheme; OCR rewards the same through 'appropriate linguistic concepts and methods'):<br><br><strong>Opening:</strong> a thesis - what the text represents (e.g. the crisis as urgent but solvable; the donor as the hero) and through which genre, audience and purpose (charity appeal; sympathetic mass audience; to persuade and collect).<br><br><strong>AO1 - systematic methods (the analysis engine):</strong> (1) <strong>Lexis/semantics:</strong> a semantic field of need ('struggling', 'crisis') set against a field of hope ('change', 'future'); connotation carrying the representation; euphemism or dysphemism framing the cause. (2) <strong>Grammar:</strong> second-person pronouns and synthetic personalisation (Fairclough, 1989); imperative calls to action ('Donate now'); modal auxiliaries ('can', 'will') building epistemic confidence; transitivity casting the donor in material processes and the beneficiary in relational ones. (3) <strong>Pragmatics:</strong> presupposition that donation is expected ('your gift will...'); flouted quantity via statistics performing authority; indirect speech acts softening the ask. (4) <strong>Graphology/multimodality:</strong> bold totals, captioned images, white space and salience directing the eye to the ask (Kress & van Leeuwen's information value and salience).<br><br><strong>AO2 - critical understanding:</strong> frameworks named and evaluated - e.g. Grice's maxims assumed then commercially exploited; Brown & Levinson's positive politeness ('you, together with us') building solidarity before the FTA of the request.<br><br><strong>AO3 - context woven:</strong> the promotional mode, the mass/non-specialist audience constraining jargon, and the purpose (influence, not just inform) driving every choice - woven through, never listed.<br><br><strong>Top band:</strong> perceptive, conceptualised analysis where every claim runs quotation → terminology → effect → context, and the representation thesis is advanced in every paragraph.
❓ Practice Questions & Mark Schemes
Using appropriate linguistic terminology, analyse the lexis and semantics of this short advertising copy. [10 marks] (OCR H470/01 Section A, 'Language under the microscope' style)
Answer: Mark hits (AO1 10):<br><br>(1) <strong>Connotation:</strong> paired words denoting similar qualities but carrying opposite evaluation - the copy's stance rides on connotative choice. (2) <strong>Semantic field:</strong> identify the domain (e.g. luxury, science, nature) and what its selection represents - the field constructs the product's positioning. (3) <strong>Figurative language:</strong> metaphor or personification transferring desirable qualities to the product - explain the transfer, don't just name the trope. (4) <strong>Evaluative and intensifying lexis:</strong> gradable adjectives and intensifiers ('simply', 'utterly') performing persuasion. (5) <strong>Morphology/coinage:</strong> any blend or derivation ('skin-fusion') signals marketing novelty.<br><br>A 10-mark microscope answer compresses the chain: feature identified, precise quotation, effect explained. Naming five devices without effect scores half; four features with full effect score top band.
Analyse how language in this transcript of a doctor-patient consultation creates meanings and representations. [25 marks] (AQA 7702/1 style - spoken text)
Answer: Mark-scheme logic (AO1 + AO3):<br><br><strong>Opening:</strong> thesis - the consultation represents the doctor as the expert gatekeeper and the patient as a compliant supplicant, through the genre (institutional consultation), audience (the professional on the other side of the exchange) and purpose (diagnosis, not conversation).<br><br><strong>Body:</strong> (1) <strong>Prosodics:</strong> falling intonation on the doctor's directives performing authority; micropauses and timed pauses before the patient discloses symptoms - analyse their placement. (2) <strong>Turn-taking and structure:</strong> IRF sequences (initiation - response - feedback) where the doctor initiates and evaluates; who controls topic and allocation of turns. (3) <strong>Pragmatics:</strong> indirect directives ('Can you just roll up your sleeve?') - interrogative form, directive force; face-threatening acts and the patient's negative politeness (hedged requests); deixis tied to the shared clinical context. (4) <strong>Lexis:</strong> occupational jargon marking expert status (link to Language and Power); register shift when the doctor moves to plain English. (5) <strong>Grammar:</strong> interrogatives from the doctor, declaratives from the patient - mood patterning with institutional asymmetry.<br><br><strong>AO3:</strong> the spoken, ephemeral, co-constructed mode; the unequal institutional tenor (Wareing's instrumental power); the purpose that constrains what either party can say - woven through.<br><br><strong>Top band:</strong> co-construction of the power relationship analysed in real time, anchored to transcription evidence, with Grice, Austin/Searle or Brown & Levinson named and evaluated.
Explore the similarities and differences in the ways that Text A and Text B use language to represent the same event. [20 marks] (AQA 7702/1 Q3 style)
Answer: Mark-scheme logic (AO1 + AO3 + AO4):<br><br><strong>Method:</strong> plan by point of comparison, never text-by-text. Each paragraph: comparative claim → Text A evidence → Text B evidence → explicit link ('whereas', 'similarly', 'both').<br><br><strong>Model points:</strong> (1) <strong>Agency:</strong> Text A's transitive verbs and active voice cast the authorities as actors; Text B's passives and nominalisations withhold agency - same event, different representation. (2) <strong>Semantic fields:</strong> a field of disorder in Text A vs a field of resilience in Text B. (3) <strong>Pragmatics:</strong> presuppositions framing the event as exceptional in one text and routine in the other. (4) <strong>Mode and graphology:</strong> electronic layout (short paragraphs, bold subheads, captioned images) vs print conventions (long-form paragraphs, columns) - the fastest AO3 route. (5) <strong>Audience:</strong> synthetic personalisation in the mass text vs a niche text assuming shared knowledge.<br><br><strong>AO3:</strong> attribute each difference to context - mode, era, audience, production - explained, not observed.<br><br><strong>AO4:</strong> the connection is the frame: same event, competing constructions - and similarities matter too (a shared convention across eras signals enduring genre norms).<br><br><strong>Top band:</strong> new comparison points (not recycled Q1/Q2 features), integrated structure, context explaining every difference.
Evaluate the idea that some language levels are more useful than others when analysing written texts. [30 marks] (AQA 7702/2 Section A evaluation style)
Answer: Mark-scheme logic (AO1 10 + AO2 20, per the 7702/2 Section A mark scheme):<br><br><strong>Step 1 - define the claim:</strong> 'useful' must mean which levels generate the most explanatory analysis of meaning, and 'written texts' must be qualified - planned, permanent, often multimodal.<br><br><strong>Step 2 - support:</strong> grammar and lexis/semantics are the highest-yield levels for written mode: clause types, transitivity and voice map agency; semantic fields and connotation carry representation - each with a worked example. Graphology and multimodality are essential where layout, typography and image share the meaning-making (Kress & van Leeuwen) - any online or printed text analysed without its visual layer is half analysed.<br><br><strong>Step 3 - challenge:</strong> the claim assumes levels compete, but they are complementary - prosodics is inapplicable to print but indispensable to transcripts; pragmatics (Grice, speech acts, face) operates in <em>any</em> mode because inference is not mode-bound; dismissing a level by mode is a category error, since the levels are one framework with different windows.<br><br><strong>Step 4 - judge:</strong> usefulness is mode-relative and purpose-relative, not intrinsic: the strongest analysis selects the levels the text's construction demands - a verdict, not a compromise.<br><br><strong>AO2:</strong> frameworks named and evaluated (Halliday, Kress & van Leeuwen, Grice); AO1: accurate terminology and precise examples throughout - description without evaluation caps the mark at mid-band.
Using appropriate linguistic concepts, analyse how graphological and multimodal features construct meanings in this social media post. [10 marks] (OCR H470/01 Section A, 'Language under the microscope' style)
Answer: Mark hits (AO1 + AO2):<br><br>(1) <strong>Layout and information value:</strong> what is placed first (the hook) and last (the ask) - left/right and top/bottom composition (Kress & van Leeuwen, 2006) order the reading path. (2) <strong>Salience:</strong> the image, emoji or bold text the eye lands on first carries the evaluative payload. (3) <strong>Typography and colour:</strong> capitals for shouting, emoji as evaluative gloss, hashtags aggregating the topic for search. (4) <strong>Interactivity:</strong> affordances of the platform - comments, likes, shares - make reception visible, so the post designs for engagement: second-person address, questions. (5) <strong>Synthetic personalisation (Fairclough):</strong> a mass post addressing each follower as an individual ('you').<br><br>A 10-mark answer needs the explanatory concept first, then compact illustrations from the post - not a feature inventory. Where a concept is named, its effect on the audience must be explicit.
amazon.co.uk
