Phonotactics
Phonotactics is the branch of phonology concerned with the rules governing permissible sequences and combinations of phonemes in a given language. Every language has its own phonotactic constraints — they determine which sound sequences are possible words and which are not.
English Phonotactic Constraints
Onset Constraints
| Cluster size | Rule | Legal examples | Illegal examples |
|---|---|---|---|
| CC | /s/ + plosive, plosive/fricative + approximant | /sp/ spin, /tr/ tree, /fl/ fly | */tl-/, */pw-/, */sr-/ |
| CCC | Must be /s/ + voiceless plosive + approximant | /spl/ split, /str/ strong, /skw/ square | */stl-/, */spr̩-/ |
Specific two-consonant onset constraints:
- /ŋ/ never appears in onsets
- /ʒ/ never appears in onsets (in native English words)
- /h/ only appears in onsets, never in codas
- */tl-/ and */dl-/ are banned (though they exist in other languages)
Coda Constraints
Codas allow up to four consonants, typically created by inflectional morphology:
- /teksts/ texts, /sɪksθs/ sixths, /streŋkθs/ strengths
Constraints include:
- /h/ and /w/ never appear in codas
- /j/ only in onsets
- /r/ in codas only in rhotic accents
Phoneme Sequence Constraints
Beyond onset/coda positions, English restricts sequences within morphemes:
- /ŋ/ never precedes /p b m f v/ within a morpheme
- Long vowels and diphthongs do not precede /ŋ/
- /ʒ/ does not follow /ɪ æ ʌ/ in native words
The Notion of Possible vs Actual Words
Phonotactics defines three categories:
- Actual words: string, blank, jump — exist in the lexicon
- Possible but non-existent words: blick /blɪk/, strim /strɪm/ — obey English phonotactics but happen not to be words (yet)
- Impossible words: */ŋæt/, */bnɪk/, */pftɔː/ — violate English phonotactic rules
This distinction is psychologically real: native speakers accept blick as a plausible English word but reject ŋlick immediately.
Cross-Linguistic Phonotactics
Languages differ dramatically in what they permit:
| Language | Typical syllable template | Notable restrictions |
|---|---|---|
| English | (C)(C)(C)V(C)(C)(C)(C) | Complex clusters permitted |
| Vietnamese | (C)V(C) | No onset clusters; codas limited to /p t k m n ŋ/ |
| Japanese | (C)V or /n/ | Only /n/ in coda; no clusters |
| Hawaiian | (C)V | No codas at all; no clusters |
| Georgian | Up to 6-consonant onsets | Extremely permissive |
| Arabic | CV(C)(C) | No onset clusters |
Phonotactics and L2 Learning
L1 phonotactic constraints are among the strongest predictors of L2 pronunciation difficulty. Learners unconsciously apply their L1 phonotactic rules to L2 input, resulting in systematic errors:
Repair Strategies
When learners encounter L2 sequences that violate their L1 phonotactics, they employ predictable repair strategies:
| Strategy | Description | Example (Vietnamese L1 → English) |
|---|---|---|
| Epenthesis | Insert a vowel to break up an illegal cluster | street → [sətəriːt] |
| Deletion | Delete a consonant to simplify the cluster | asked → [ɑːst] |
| Prothesis | Add a vowel before the word | school → [əskuːl] (Arabic speakers) |
| Substitution | Replace an illegal sound with a legal one | /θ/ → [t] or [s] |
Perception and Phonotactics
L1 phonotactics affect not just production but perception. Japanese listeners perceive an illusory vowel between consonants in clusters — they "hear" ebzo as ebuzo because their L1 phonotactics require CV structure. This perceptual repair happens automatically and is very difficult to override.
Teaching Implications
- Diagnose the learner's L1 syllable structure constraints before targeting pronunciation — it predicts most of their difficulties.
- Teach consonant clusters in order of increasing complexity.
- Address repair strategies explicitly: help learners notice when they are inserting or deleting sounds.
- Work on perception as well as production — if learners cannot perceive the target cluster, they cannot produce it.
- Phonotactic awareness can be raised through word-recognition tasks: which of these could be English words? (blick, bnick, brick).