Analysing Child Language Data

The ability to analyse child language data — transcripts of children's speech, samples of children's writing, or multimodal texts — is the central assessed skill of this topic. Everything you have studied (phonological processes, lexical development, grammar, pragmatics, CDS, literacy, bilingualism, and the great nature/nurture debate) exists to be applied to a piece of data under timed conditions. This lesson is the synthesis: it provides a systematic, level-by-level method for analysing CLA data, a mapping from observed features to the theories that explain them, and worked examples showing how to write to the assessment objectives.

---

How CLA Data Appears in the Exam

Children's Language Development is assessed on Paper 1 — Language, the Individual and Society (2 hours 30 minutes, 100 marks, 40% of the A-Level). It is examined in Section B: Children's Language Development (0–11 years), which is worth 30 marks and takes the form of one discursive essay chosen from two questions, based on data which may be spoken, written or multimodal. The five assessment objectives are weighted as follows for this question:

AO	What it rewards	Marks
AO1	Apply appropriate linguistic terminology and methods accurately and coherently	26
AO2	Demonstrate critical understanding of concepts and issues relevant to language use	26
AO3	Analyse and evaluate how contextual factors (here, developmental stage, caregiver input, mode) shape language	23
AO4	Explore connections across texts/data (e.g. comparing two children or two modes)	15
AO5	Demonstrate expertise and creativity in your own writing — a controlled, well-organised academic response	10

The practical upshot of this weighting is decisive for how you write. AO1 and AO2 together dominate, so accurate terminology and conceptual understanding must run through every paragraph; AO3 means you must always relate features to the developmental and interactional context, not just label them; AO4 rewards genuine connections where the data invites comparison; and AO5 means the essay must be coherent and well-written in its own right. It is worth internalising that no single assessment objective is won by listing features: every objective is satisfied by the same analytical behaviour — taking a precisely named feature, quoting it, explaining what it shows about the child, and arguing its theoretical significance in context. Keeping the five objectives in mind as a checklist, rather than as separate hurdles, helps ensure that a response is balanced and that no objective is neglected — a common cause of otherwise strong answers stalling a band below their potential.

Key Definition: Language data analysis — the systematic examination of a language sample (spoken, written or multimodal) to identify linguistic features and patterns and to explain, with reference to theory, what they reveal about the child's competence and developmental stage. It is interpretive: the aim is always to move from "what feature is present" to "what this shows about the child".

You may meet: transcripts of speech (child–caregiver or child–child); samples of writing (a recount, story or card); or a combination — for instance two children of different ages, or the same child's speech alongside their writing — which is the scenario that most rewards AO4 connection-making.

---

The Foundational Method: Feature → Quotation → Significance → Theory

Every strong CLA paragraph follows the same four-move structure, and drilling it is the fastest route to marks:

1. Identify the feature with the correct technical label (AO1).
2. Quote the precise example from the data — never analyse in the abstract.
3. Explain the significance — what it shows about the child's developmental stage (AO2/AO3).
4. Link to theory — name the researcher or framework that explains or contests it (AO2).

A point that does only step 1 (spots and labels) caps low; a point that completes all four moves accesses the higher bands. Memorise this loop; it is the engine of the whole essay.

---

The Language Levels Framework

Work systematically through the language levels, identifying relevant features at each:

Language level	What to look for	Key terminology
Phonology	Phonological (simplification) processes; the fis phenomenon; prosody; intonation	Fronting, stopping, gliding, consonant cluster reduction, final consonant deletion, assimilation, reduplication
Lexis / Semantics	Vocabulary range; over- and under-extension; concrete vs abstract; first-word categories	Holophrase, overextension, underextension, hypernym/hyponym, labelling/packaging/network-building
Grammar (morphology)	Inflectional morphemes (plural, past tense, -ing, possessive); overgeneralisation; MLU	Overgeneralisation, virtuous error, MLU, inflection, regular/irregular forms
Grammar (syntax)	Word order; negation; question formation; coordination vs subordination	Telegraphic speech, SVO, subject–verb inversion, subordinate clause, coordination
Pragmatics	Turn-taking; topic management; speech functions; politeness; repair; deixis	Halliday's functions, deixis, adjacency pair, repair, face
Discourse	Narrative structure; cohesion and coherence; text organisation (writing)	Cohesion, coherence, temporal connectives, anaphoric reference
Graphology (writing only)	Letter formation; spelling strategies; punctuation; layout	Invented spelling, emergent writing, grapheme–phoneme correspondence

---

Step-by-Step Analysis Method

Step 1 — Read the data carefully

• Read the whole transcript or writing sample at least twice before writing.
• Note the child's age (if given): it lets you predict the expected stage and judge whether the child is ahead, typical or delayed.
• Note the context: who is the interlocutor, what activity frames the talk, and what is the purpose/genre of any writing?

Step 2 — Estimate the developmental stage

Use the observed features (and any stated age) to locate the child. Treat the ages as approximate ranges — there is wide normal variation, and you should say so rather than asserting precise milestones:

Age (approx.)	Expected stage (speaking)	Expected stage (writing)
12–18 months	Holophrastic (one-word) stage	Pre-writing (scribbling)
18–24 months	Two-word stage; telegraphic speech	Mark-making
24–36 months	Multi-word utterances; emerging grammar; early overgeneralisation	Mock writing / mock letters
36–48 months	Complex sentences emerging; questions and negatives developing	Letter formation; invented spelling
48–60 months	Most grammatical structures present; narratives developing	Conventional writing emerging; Kroll's consolidation
5–7 years	Near-adult grammar; growing pragmatic sophistication	Kroll's consolidation/differentiation; Rothery's recount/narrative
7–11 years	Refined pragmatics; metalinguistic awareness	Kroll's differentiation toward integration; genre awareness

Step 3 — Analyse systematically by language level

Apply the four-move loop (feature → quotation → significance → theory) at each relevant level. You do not have to cover every level exhaustively; cover the levels where the data is richest, and go deep (analyse and evaluate) rather than merely wide.

Step 4 — Map features to theory

This mapping is the heart of AO2. Learn it so that, on seeing a feature, the relevant theorist comes automatically to mind:

Feature observed	Relevant theory / researcher
Overgeneralisation ("goed", "mouses")	Nativist (Chomsky) — rule application, not imitation; Berko's wug test (1958) demonstrates productive rule use
Child imitating the caregiver	Behaviourist (Skinner) — imitation and reinforcement (note its limits — it cannot explain virtuous errors)
Caregiver using expansions, recasts, CDS	Social interactionist (Bruner — LASS, scaffolding, formats; Snow — CDS, proto-conversations; Saxton — recasts as negative evidence)
Language tied to cognitive level (e.g. object permanence)	Cognitive (Piaget) — language reflects prior cognitive development
Seven communicative functions	Halliday (1975) — instrumental, regulatory, interactional, personal, heuristic, imaginative, representational
Phonological simplification processes	Ingram (1976) — phonological process analysis; the fis phenomenon (Berko and Brown, 1960) shows perception outstrips production
Turn-taking, routines, joint attention	Bruner (1983) — formats/ritualised interaction; Snow (1977) — proto-conversations
First-word categories / noun bias	Nelson (1973) — referential vs expressive first-word categories
Over-/under-extension; word-meaning growth	Eve Clark (1973) — semantic feature hypothesis; Aitchison — labelling, packaging, network-building
Stages of negation and question formation	Bellugi — staged development of negatives and interrogatives
MLU and order of morpheme acquisition	Brown (1973) — stages and the order in which grammatical morphemes are acquired
Reading features	Frith (1985) — logographic/alphabetic/orthographic; Chall (1983) — stages; the phonics debate
Writing features	Kroll (1981) — phases; Rothery (1984) — genres; Perera (1984) — coordination → subordination
Code-switching; bilingual development	Cummins — BICS/CALP, threshold, CUP; Lambert — additive/subtractive

---

Common Features to Look For

In speech data

• Phonological processes — deletion, substitution, assimilation, reduplication (always name the specific process).
• Telegraphic speech — omission of function words (articles, prepositions, auxiliaries).
• Overgeneralisation — regular rules applied to irregular forms ("goed", "mouses").
• Holophrases — single words carrying sentence-level meaning (interpret via intonation/context).
• Questions — note the stage (rising intonation only → wh-word without inversion → correct inversion).
• Negatives — note the stage (sentence-external "no" → internal without auxiliary → correct auxiliary).
• CDS features in any caregiver speech — expansions, recasts, questions, repetition, raised pitch.
• Turn-taking and topic management — does the child hold turns and sustain topics?
• Halliday's functions — which functions is the child deploying?
• Deixis — including developmental deictic errors such as pronoun reversal.

In writing data

• Spelling — invented spellings and the phonetic strategies behind them.
• Grammar — sentence type, coordination vs subordination (Perera), tense consistency.
• Vocabulary — range; concrete vs abstract; repetition.
• Genre — classify it (Rothery) and judge it against the genre's conventions.
• Relationship to speech — Kroll's consolidation (speech-like) vs differentiation.
• Graphology — letter formation, spacing, case-mixing, reversals, emerging punctuation.

---

Analytical Writing Technique

A strong Section B essay will:

1. Open with a focused overview — state the child's approximate stage and the most striking features, framing the discursive line you will take.
2. Work through the richest language levels — analysing, not just listing, with accurate terminology (AO1/AO2).
3. Quote precisely — every claim anchored to the data (AO3).
4. Explain significance — say what each feature shows about development, not merely that it occurs.
5. Engage theory and debate — weigh competing explanations (nativist vs interactionist vs cognitive vs behaviourist) rather than name-dropping (AO2).
6. Make connections where invited — across two children, two ages, or two modes (AO4).
7. Write coherently — a controlled, well-structured academic register throughout (AO5).

Example analytical paragraph (phonology)

"The child shows several phonological simplification processes typical of the early stages. In 'tat' for 'cat' the child uses fronting — replacing the velar plosive /k/ with the alveolar plosive /t/ — a common process described by Ingram (1976) that typically resolves by around age 3 to 3.5. The production 'poon' for 'spoon' shows consonant cluster reduction, the /s/ of the /sp/ cluster being deleted. Together these suggest a child who has secured early-acquired plosives and nasals but not yet fricatives or clusters, consistent with the broadly attested order of phoneme acquisition. Notably, the fis phenomenon (Berko and Brown, 1960) reminds us that such substitutions are likely production limitations rather than perceptual ones — the child probably hears the adult target accurately even while unable to produce it."

This paragraph completes all four moves — names the feature, quotes the data, explains the developmental significance, and links to (and even evaluates with) named research.

---

Worked Example: Speech Transcript

Consider an extract between a child of approximately 2;6 (two years, six months) and their mother:

Speaker	Utterance
Child	"Mummy, look — big doggy!"
Mother	"Yes, it is a big dog, isn't it?"
Child	"Doggy runned away"
Mother	"The dog ran away? Where did it go?"
Child	"It goed over there"

Analysis points:

• Lexis: the diminutive "doggy" and concrete nouns are consistent with Nelson's (1973) referential first-word tendency, in which naming words for salient objects dominate early vocabulary.
• Morphology: "runned" and "goed" are overgeneralisations — the regular past-tense -ed rule applied to irregular verbs. As virtuous errors these support the nativist case (the child cannot have heard these forms), and they parallel Berko's (1958) wug finding that children apply morphological rules productively.
• Syntax: multi-word utterances with emerging structure (subject + verb + adverbial in "It goed over there") indicate movement beyond the two-word stage into early grammatical combination (broadly Brown's middle stages).
• Pragmatics: the child uses Halliday's (1975) personal function (expressing reaction — "look!") and representational function (conveying information about the dog).
• CDS / interaction: the mother recasts ("runned" → "ran") without overt correction and poses a question ("Where did it go?"), scaffolding further production — strong support for Bruner's interactionism and, via Saxton, for recasts as a source of corrective evidence.
• Pragmatics: clear emerging turn-taking — the child initiates, the mother responds, the child answers — illustrating the conversational competence Snow traced back to proto-conversations.

A discursive answer would not merely list these; it would weigh them — for example arguing that the overgeneralisations tilt toward a nativist reading while the mother's recasting tilts toward interactionism, and concluding that the data is best explained by the two working together.

---

Reading Transcription Conventions

Spoken data is presented as a transcript, and you must be able to read its notation so you can comment on features of interaction (AO3), not just on words. Common conventions include:

Convention	Meaning	Why it matters analytically
(.)	A micropause (a very brief pause)	Pauses after a caregiver's question can be deliberate "turn slots" inviting the child to respond
(2.0)	A timed pause, here two seconds	Long child pauses may reflect processing load; long caregiver pauses may be scaffolding wait-time
= (latching)	One turn follows another with no gap	Smooth latching shows developing turn-taking competence
[ ] (overlap)	Speech spoken simultaneously by both	Overlap can show enthusiasm/engagement or still-developing turn-taking
CAPITALS	Speech noticeably louder/emphatic	Caregiver stress on a key content word is a CDS feature aiding segmentation
**:: ** (e.g. "no::")	A sound is elongated	Exaggerated vowel lengthening is typical of CDS prosody
(( ))	Non-verbal/contextual information, e.g. ((points at dog))	Vital for interpreting holophrases and deixis — meaning often lives in the action

Two practical points. First, do not "correct" the transcript: features like pauses, repetitions and overlaps are data, and commenting on them earns AO3 credit. Second, use the contextual notes in double brackets: a child saying "that!" is meaningless until ((points at the moon)) reveals it as a labelling request — and it is impossible to discuss deixis or holophrastic meaning without them.

---

Worked Example: A Writing Sample

The method applies equally to written data. Consider a piece by a child of about 6;5, reproduced with original spelling:

"On satdy we wendt to the farm. I sor a cow and a pig. the pig was pinck. I likd the baby goats best. it woz a fun day."

A four-move analysis across the relevant levels: