Exam text analysis: analysing and comparing unseen texts using the discourse framework, building a comparative argument, and writing to time.

What this dot point is asking

Edexcel wants you to analyse and compare unseen texts under timed conditions, using the discourse framework to build a comparative argument that integrates context and theory and always reaches the effect on the reader. On Paper 1 you meet unseen data (often a spoken transcript alongside a written text); on Paper 2 you meet texts illustrating language change or children's language. In every case the marked skill is the same: select the most productive language levels, write a sustained argument rather than a feature list, and ground every observation in the effect it has on a real audience.

The answer

The discourse framework

Mode is rarely binary. Many exam texts are blended: a text message is written but carries spoken features (ellipsis, non-standard spelling, paralinguistic emoji); a scripted broadcast is spoken but planned and polished. David Crystal's term "texting" and the broader category of computer-mediated communication (CMC) describe these hybrids, where the written channel borrows the immediacy and informality of speech. Naming a text as blended, and explaining which spoken or written features it imports and why, is itself a high-AO2 move.

Field shapes lexis. A medical leaflet draws on a semantic field of clinical terminology; a sports report draws on metaphor and hyperbole. Tenor governs register: the power relationship (equal, asymmetric), the social distance (intimate, formal) and the resulting choices in address, modality and politeness. Penelope Brown and Stephen Levinson's politeness theory (positive politeness building solidarity, negative politeness respecting autonomy and avoiding imposition) and Paul Grice's cooperative principle (the maxims of quantity, quality, relation and manner, and the meaning generated when speakers flout them) give you the vocabulary to analyse tenor precisely.

Selecting language levels under time

You will not use every level on every text, and the exam does not reward you for trying. Read each text twice: once to grasp genre, audience and purpose, and once to mark the four or five features that most clearly serve that purpose. For a persuasive advert, graphology and lexis may dominate; for a conversation transcript, pragmatics and discourse structure will. Selecting the relevant levels is an analytical skill, and a planned answer built on the strongest evidence beats a rushed sweep through all six.

Building the comparative argument

The strongest exam answers read as a sustained argument: a thesis, points that compare both texts with precise evidence and metalanguage, and consistent attention to purpose and effect. A comparative paragraph follows a clear shape: a topic sentence making a comparative claim ("both texts construct authority, but Text A does so through institutional register while Text B relies on personal anecdote"), evidence from each text, named features, and a closing sentence on the differing effect. Connectives of comparison and contrast (whereas, similarly, by contrast, conversely) keep the comparison live on the page so the examiner never has to infer it.

Writing to time

Plan for two or three minutes, write for the rest. A workable split for a 20-mark comparison is a short framing of both texts' discourse parameters, three to four comparative points built on the most productive levels, and a one-line synthesis. Resist the urge to transcribe long quotations; embed short, precise evidence and spend your words on the effect. If you run short of time, a clear final comparative point is worth far more than an unfinished paragraph drifting back into feature-spotting.

Examples in context

Example 1. A spoken transcript paired with a written text. When Paper 1 pairs a conversation transcript with a written article, the highest-scoring answers exploit mode as the engine of comparison. The transcript will show adjacency pairs (question and answer), overlaps and interruptions (analysable through Zimmerman and West's dominance work), fillers and false starts that mark spontaneity. The written text will show planning: subordination, cohesion, edited lexis. A strong candidate does not list these separately; they argue that the transcript's interactivity and the article's deliberateness produce different relationships with the audience, citing Grice (how speakers flout maxims for effect in the transcript) and Brown and Levinson (face-work in both).

Example 2. Two texts on the same topic, different audiences. Given a tabloid and a broadsheet report of one event, the discourse framework isolates the variable: field is shared (same event), so the comparison lives in tenor and lexis. The tabloid's monosyllabic, emotive lexis, short declaratives and direct address build proximity with a mass readership; the broadsheet's nominalisation, hypotactic syntax and hedged modality build authority with an informed one. Naming nominalisation (turning processes into nouns, "the killing" rather than "x killed y") and explaining how it backgrounds agency is a precise, high-AO2 observation that separates a sustained argument from a list.

Try this

Q1. What do mode, field and tenor each describe in the discourse framework? [3 marks]

• Cue. Mode is spoken, written or blended and how the channel shapes the text; field is the subject matter and its specialist lexis; tenor is the relationship between participants and the resulting register.

Q2. How should you structure a comparison of two unseen texts? [2 marks]

• Cue. Build a comparative thesis and analyse both texts together around shared points (such as how each constructs authority or addresses its audience), not one text then the other.

Q3. Why does naming a feature without explaining its effect score poorly? [2 marks]

• Cue. AO2 marks reward analysis of why a choice was made and what it does to the reader in context; identification alone is feature-spotting.