You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
When people speak naturally, they do not produce words as isolated, carefully articulated units. Speech flows in a continuous stream in which sounds influence, overlap and modify one another, and word boundaries blur. This natural, fluid speech is called connected speech, and the systematic modifications within it are connected speech processes. Understanding them is essential for AQA A-Level English Language (7702) because they reveal how spoken language actually works — and because they are among the richest features to analyse in transcripts of real talk. The skill serves chiefly AO1 (naming and classifying the processes precisely) while also feeding AO3, since the density of connected-speech features is a strong index of register, formality and the relationship between speakers. The reference accent throughout is RP / Standard Southern British English (SSBE), used as a documented baseline, not a standard of correctness.
The citation form of a word is its pronunciation in isolation — how you would say it reading carefully from a list. In real conversation, speakers constantly depart from citation forms for efficiency and ease of articulation. This is emphatically not "lazy" or "sloppy" speech: it is a universal, systematic, rule-governed property of all spoken languages, and even careful, formal speech contains some of it.
Connected speech processes are driven by the principle of least effort (sometimes called ease of articulation): speakers economise on articulatory movement while staying intelligible. Listeners decode connected speech effortlessly because they are attuned to these processes and compensate automatically — which is also why a transcript that "tidies" speech into citation forms misrepresents the data.
Key Definition: Connected speech — the natural, continuous flow of spoken language in which words are linked and sounds are modified by their phonetic environment, producing systematic changes such as assimilation, elision, liaison and reduction.
Assimilation occurs when a sound changes to become more like a neighbouring sound, because the articulators "anticipate" or "carry over" the gesture needed for an adjacent sound. It is one of the most common processes in English.
| Type | Direction | Example |
|---|---|---|
| Regressive (anticipatory) | A sound changes to match the following sound — the most common type in English | "ten boys" /ten bɔɪz/ → [tem bɔɪz]: /n/ becomes [m] before bilabial /b/ |
| Progressive (perseverative) | A sound changes to match the preceding sound — less common | The plural morpheme stays voiced /z/ after voiced /g/ in "dogs" /dɒgz/ |
| Coalescent | Two adjacent sounds merge into a new sound | "don't you" /dəʊnt juː/ → [ˈdəʊntʃuː]: /t/ + /j/ → /tʃ/ |
| Feature | Example |
|---|---|
| Place of articulation | "ten pens" /ten penz/ → [tem penz] (alveolar /n/ → bilabial [m] before /p/); "ten cars" → [teŋ kɑːz] (/n/ → velar [ŋ] before /k/) |
| Voicing | "have to" /hæv tuː/ → [hæf tə] (voiced /v/ → voiceless [f] before voiceless /t/) |
| Manner of articulation | Rarer, but possible in rapid speech where one manner shifts toward a neighbour |
Naming the direction and the feature is what makes assimilation analysis precise — e.g. "regressive place assimilation of /n/ to [m]".
A particularly important case for A-Level is yod coalescence, where an alveolar consonant fuses with the palatal approximant /j/ (the "yod") to make a post-alveolar affricate or fricative:
| Sequence | Result | Example |
|---|---|---|
| /t/ + /j/ | /tʃ/ | "don't you" → /ˈdəʊntʃuː/; "what you" → /ˈwɒtʃuː/ |
| /d/ + /j/ | /dʒ/ | "would you" → /ˈwʊdʒuː/; "did you" → /ˈdɪdʒuː/ |
| /s/ + /j/ | /ʃ/ | "this year" → /ˈðɪʃɪə/; "miss you" → /ˈmɪʃuː/ |
| /z/ + /j/ | /ʒ/ | "as you" → /ˈæʒuː/ |
This is so widespread that it has become lexicalised in some words for many speakers: "Tuesday" /ˈtʃuːzdeɪ/ (historically /ˈtjuːzdeɪ/) and "tune" /tʃuːn/ no longer feel like connected-speech outputs at all — a neat example of a process turning into established change (AO3).
Key Definition: Assimilation — a process in which a sound changes one or more features (place, voicing or manner) to become more like a neighbouring sound, driven by articulatory economy. Directions: regressive, progressive, coalescent.
Elision is the omission (deletion) of a sound or syllable present in the citation form. Like assimilation, it increases efficiency and is fully systematic.
| Type | Description | Examples |
|---|---|---|
| Consonant-cluster simplification | In a run of consonants (often across a word boundary), a middle consonant is dropped | "last night" → /lɑːs naɪt/ (/t/ elided); "handbag" → /ˈhænbæg/ (/d/ elided); "next please" → /neks pliːz/ |
| Schwa elision (syncope) | An unstressed vowel (usually schwa) is deleted, losing a syllable | "library" /ˈlaɪbrəri/ → /ˈlaɪbri/; "temperature" → /ˈtemprətʃə/; "chocolate" → /ˈtʃɒklət/ |
| Initial unstressed syllable loss (aphaeresis) | The first unstressed syllable is dropped | "because" → 'cause /kəz/; "about" → 'bout /baʊt/ |
| /h/-elision in weak function words | /h/ is deleted in unstressed grammatical words | "give him" → /gɪv ɪm/; "tell her" → /tel ə/ |
| Word-final /t/ or /d/ deletion | Dropped especially before a following consonant | "left side" → /lef saɪd/; "old man" → /əʊl mæn/ |
Elision is natural, but some instances are socially evaluated while others pass unnoticed. Schwa elision in "chocolate" /ˈtʃɒklət/ is universal and invisible, whereas /h/-elision in stressed lexical words ("I 'ate 'im") is stigmatised as non-standard. Note the distinction from Lesson 6's accent-level H-dropping: dropping /h/ in unstressed function words (him, her, has) is universal and unremarkable; dropping it in stressed content words is the stigmatised accent feature. In analysis, describe elision objectively and treat its social loading as a separate, AO3 question.
Key Definition: Elision — the omission of a sound or syllable in connected speech to reduce effort; common types include consonant-cluster simplification, schwa deletion (syncope) and weak-form /h/-elision.
Liaison is the insertion or maintenance of a sound between words to smooth a vowel-to-vowel transition. In English, the headline phenomena involve /r/, which behaves distinctively in non-rhotic accents like RP (where /r/ is not pronounced before a pause or consonant).
When a word spelled with a final "r" is followed by a word beginning with a vowel, the historically present /r/ resurfaces to link them:
Intrusive /r/ inserts an /r/ where there is no "r" in the spelling, after a non-high vowel (/ɑː/, /ɔː/, /ɜː/, /ə/) and before a following vowel:
Intrusive /r/ is sometimes stigmatised as "incorrect", but linguistically it is simply the linking-/r/ rule generalised to all such vowel–vowel junctures, regardless of spelling — the speaker is applying one consistent phonological rule. Both linking and intrusive /r/ are valuable diagnostics in data because they prove an accent is non-rhotic.
| Device | Description | Example |
|---|---|---|
| Linking /j/ | A [j] glide bridges a close front vowel and a following vowel | "see it" → [siː jɪt]; "I agree" → [aɪ jəˈgriː] |
| Linking /w/ | A [w] glide bridges a close back rounded vowel and a following vowel | "do it" → [duː wɪt]; "go on" → [gəʊ wɒn] |
A related effect is catenation — the re-syllabification of a final consonant onto a following vowel-initial word, so "an apple" is heard as [ə ˈnæpl] and "pick it up" as [pɪ ˈkɪ tʌp]. This boundary blurring is part of why word divisions in connected speech are so much less clear-cut than in writing.
Key Definition: Liaison — the insertion or maintenance of a consonant between words to smooth a vowel-to-vowel transition. The key English forms are linking /r/ (spelling has "r") and intrusive /r/ (it does not); related effects include linking /j/, linking /w/ and catenation.
Many high-frequency function words have two pronunciations: a strong form (stressed or in isolation) and a weak form (unstressed in connected speech). Weak forms typically reduce the vowel to schwa /ə/ and may drop consonants.
| Word | Strong form | Weak form(s) |
|---|---|---|
| and | /ænd/ | /ənd/, /ən/, /n/ |
| to | /tuː/ | /tə/, /tʊ/ |
| for | /fɔː/ | /fə/ |
| from | /frɒm/ | /frəm/ |
| of | /ɒv/ | /əv/, /ə/ |
| the | /ðiː/ | /ðə/, /ði/ (before vowels) |
| a | /eɪ/ | /ə/ |
| was | /wɒz/ | /wəz/ |
| have | /hæv/ | /həv/, /əv/, /v/ |
| can | /kæn/ | /kən/, /kn/ |
| but | /bʌt/ | /bət/ |
| are | /ɑː/ | /ə/ |
| at | /æt/ | /ət/ |
| that (conjunction) | /ðæt/ | /ðət/ |
Weak forms dominate natural speech. Using strong forms where weak forms are expected sounds unnatural, over-careful or emphatic. Speakers revert to strong forms for emphasis or contrast ("I said TO, not FROM"), and utterance-finally ("What are you looking at?" keeps "at" strong, /æt/). The proportion of weak to strong forms is itself a register cue: dense weak-form use signals casual, fluent talk; many strong forms signal careful or formal speech (AO3). One caution for transcription: "that" reduces to /ðət/ only as a conjunction or relativiser; the demonstrative "that" ("that book") keeps its strong form /ðæt/. The same word class sensitivity applies elsewhere — prepositional "to" reduces freely, but the same letters in an infinitive marker also reduce, so context, not spelling, decides whether a weak form is appropriate.
The schwa is the linchpin of connected speech, because it is the typical product of vowel reduction — the process whereby full vowels in unstressed positions collapse to the short, neutral mid-central /ə/. Schwa accounts for roughly one in three vowel sounds in fluent English.
A command of schwa underpins three things at once:
Indeed, the deeper point is that all these processes serve the same master: English is stress-timed, and assimilation, elision, weak forms and schwa reduction all work to compress the material between stressed syllables so that stresses can fall at roughly even intervals.
Because connected speech blurs word boundaries, the placement of boundaries can itself become ambiguous — a phenomenon known as juncture. The classic minimal pairs are phrases that share the same string of phonemes but differ in where the boundary (and its phonetic cues) falls:
The phonetic cues to juncture include slight differences in the aspiration of plosives (a word-initial /t/ is aspirated; a word-medial one may not be), the length of consonants and vowels, and the presence or absence of glottal reinforcement before a vowel-initial word. Juncture is a neat demonstration that speech is not simply "words with spaces removed" — the boundary information has to be carried by fine phonetic detail, and where that detail is weak, genuine ambiguity (and comedy, in mishearings and mondegreens) results.
Key Definition: Juncture — the phonetic features (timing, aspiration, length, glottal cues) that mark boundaries between words or morphemes in connected speech; where these cues are weak, the same sequence of phonemes can be parsed in more than one way ("an aim" vs "a name").
Because assimilation is so heavily examined, it is worth drilling the analytical routine until it is automatic. For any assimilation, answer three questions in order:
Run that routine and you can describe any assimilation as, for example, "regressive place assimilation of alveolar /n/ to bilabial [m] under the influence of the following bilabial /b/" — a sentence that demonstrates complete control. A worked table:
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.