What Is Vocabulary | The Neighborhood Activation Model — How Similar-Sounding Words Compete in Listening

NEIGHBORHOOD ACTIVATION MODEL · NAM · SPOKEN WORD RECOGNITION · PHONOLOGICAL NEIGHBOURS · NEIGHBORHOOD DENSITY · LEXICAL COMPETITION

The Neighborhood Activation Model explains spoken-word recognition as a competition between the intended word and similar-sounding words in the mental lexicon. Recognition depends not only on how well the acoustic signal matches the target, but also on how much evidence supports competing neighbours.

This makes vocabulary structure consequential. A word in a dense phonological neighbourhood may face many competitors, while a sparse-neighbourhood word faces fewer. Target frequency, neighbour frequency, noise and phonological precision all affect which candidate wins.

This guide explains phonological neighbourhood density, neighbour frequency, target frequency, acoustic similarity, lexical competition, noisy listening, second-language recognition and lexical integration. Existing eduKateSG Phonological Neighbourhood Density and Word Recognition owners remain untouched.

Neighborhood Activation Model

The Neighborhood Activation Model, or NAM, explains spoken-word recognition through activation of a target and phonologically similar lexical neighbours.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Phonological neighbours

A neighbour is typically a word that can be formed by adding, deleting or substituting a phoneme, though operational definitions vary.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Neighborhood density

Dense neighbourhoods contain many similar-sounding words; sparse neighbourhoods contain relatively few.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Neighbour frequency

A high-frequency neighbour can be a powerful competitor even when the target itself is familiar.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Target frequency

Frequent targets generally enjoy stronger lexical evidence and can overcome competition more efficiently.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Similarity

Competition depends on degree of acoustic-phonetic similarity, not merely an abstract neighbour count.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Decision rule

Recognition reflects evidence for the target relative to the summed activation of competitors.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Noise

Degraded input increases uncertainty and can magnify neighbourhood competition because more candidates remain plausible.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Second-language listening

L2 listeners may experience enlarged or distorted neighbourhoods when phoneme contrasts are not robustly represented.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Accent

Accent variation changes acoustic similarity and can alter which lexical neighbours become active.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Word learning

New vocabulary can enter existing phonological neighbourhoods and eventually become a competitor for older words.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Lexical integration

A newly learned form is more convincingly integrated when it begins to participate in competition with established neighbours.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Auditory lexical decision

NAM predicts patterns in tasks where listeners decide whether a spoken sequence is a word.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Word identification

The model also applies to identification under noise, where target-to-neighbour evidence becomes especially important.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Density is not destiny

Dense neighbourhoods can sometimes support learning through phonotactic familiarity even while increasing recognition competition.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

Teaching

Learners benefit from explicit contrast among high-confusability neighbours, especially when one competitor is much more frequent.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

AI-era listening

Speech recognisers also rank competing hypotheses, but NAM is a cognitive account with its own assumptions and should not be collapsed into modern ASR.

The useful diagnostic question is relative evidence: how strongly does the signal support the target compared with the combined activation of its neighbours?

18. NAM casebook A

1. NAM case — cat

Neighbour set: generate words differing from cat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of cat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

2. NAM case — cap

Neighbour set: generate words differing from cap by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of cap, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

3. NAM case — cab

Neighbour set: generate words differing from cab by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of cab, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

4. NAM case — can

Neighbour set: generate words differing from can by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of can, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

5. NAM case — bat

Neighbour set: generate words differing from bat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

6. NAM case — bad

Neighbour set: generate words differing from bad by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bad, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

7. NAM case — bag

Neighbour set: generate words differing from bag by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bag, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

8. NAM case — bed

Neighbour set: generate words differing from bed by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bed, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

9. NAM case — bet

Neighbour set: generate words differing from bet by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bet, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

10. NAM case — bell

Neighbour set: generate words differing from bell by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bell, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

11. NAM case — coat

Neighbour set: generate words differing from coat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of coat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

12. NAM case — goat

Neighbour set: generate words differing from goat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of goat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

13. NAM case — boat

Neighbour set: generate words differing from boat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of boat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

14. NAM case — cold

Neighbour set: generate words differing from cold by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of cold, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

15. NAM case — gold

Neighbour set: generate words differing from gold by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of gold, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

16. NAM case — fold

Neighbour set: generate words differing from fold by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of fold, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

17. NAM case — light

Neighbour set: generate words differing from light by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of light, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

18. NAM case — right

Neighbour set: generate words differing from right by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of right, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

19. NAM case — night

Neighbour set: generate words differing from night by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of night, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

20. NAM case — fight

Neighbour set: generate words differing from fight by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of fight, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

19. NAM casebook B

21. NAM case — pin

Neighbour set: generate words differing from pin by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of pin, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

22. NAM case — bin

Neighbour set: generate words differing from bin by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of bin, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

23. NAM case — tin

Neighbour set: generate words differing from tin by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of tin, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

24. NAM case — win

Neighbour set: generate words differing from win by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of win, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

25. NAM case — ship

Neighbour set: generate words differing from ship by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of ship, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

26. NAM case — shop

Neighbour set: generate words differing from shop by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of shop, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

27. NAM case — shot

Neighbour set: generate words differing from shot by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of shot, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

28. NAM case — shut

Neighbour set: generate words differing from shut by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of shut, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

29. NAM case — book

Neighbour set: generate words differing from book by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of book, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

30. NAM case — cook

Neighbour set: generate words differing from cook by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of cook, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

31. NAM case — look

Neighbour set: generate words differing from look by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of look, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

32. NAM case — hook

Neighbour set: generate words differing from hook by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of hook, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

33. NAM case — seat

Neighbour set: generate words differing from seat by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of seat, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

34. NAM case — seed

Neighbour set: generate words differing from seed by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of seed, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

35. NAM case — seek

Neighbour set: generate words differing from seek by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of seek, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

36. NAM case — seem

Neighbour set: generate words differing from seem by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of seem, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

37. NAM case — park

Neighbour set: generate words differing from park by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of park, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

38. NAM case — part

Neighbour set: generate words differing from part by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of part, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

39. NAM case — mark

Neighbour set: generate words differing from mark by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of mark, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

40. NAM case — dark

Neighbour set: generate words differing from dark by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of dark, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

20. NAM casebook C

41. NAM case — train

Neighbour set: generate words differing from train by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of train, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

42. NAM case — drain

Neighbour set: generate words differing from drain by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of drain, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

43. NAM case — brain

Neighbour set: generate words differing from brain by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of brain, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

44. NAM case — grain

Neighbour set: generate words differing from grain by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of grain, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

45. NAM case — play

Neighbour set: generate words differing from play by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of play, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

46. NAM case — clay

Neighbour set: generate words differing from clay by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of clay, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

47. NAM case — glow

Neighbour set: generate words differing from glow by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of glow, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

48. NAM case — grow

Neighbour set: generate words differing from grow by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of grow, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

49. NAM case — stone

Neighbour set: generate words differing from stone by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of stone, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

50. NAM case — phone

Neighbour set: generate words differing from phone by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of phone, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

51. NAM case — mouse

Neighbour set: generate words differing from mouse by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of mouse, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

52. NAM case — mouth

Neighbour set: generate words differing from mouth by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of mouth, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

53. NAM case — house

Neighbour set: generate words differing from house by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of house, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

54. NAM case — horse

Neighbour set: generate words differing from horse by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of horse, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

55. NAM case — teacher

Neighbour set: generate words differing from teacher by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of teacher, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

56. NAM case — teaser

Neighbour set: generate words differing from teaser by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of teaser, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

57. NAM case — reason

Neighbour set: generate words differing from reason by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of reason, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

58. NAM case — season

Neighbour set: generate words differing from season by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of season, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

59. NAM case — signal

Neighbour set: generate words differing from signal by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of signal, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

60. NAM case — single

Neighbour set: generate words differing from single by one phoneme through substitution, addition or deletion. Count only real lexical items known to the learner when interpreting personal competition.

Density: classify the target as relatively dense or sparse within the chosen lexicon. A raw dictionary count is not identical to the active neighbourhood of a child, L2 learner or specialist.

Frequency competition: identify the highest-frequency neighbour. Predict whether that competitor could attract recognition when the acoustic signal is noisy or the distinguishing phoneme is weak.

Noise test: mask the critical phoneme and ask listeners to identify the word. Increased neighbour responses illustrate how uncertainty redistributes activation across the lexical neighbourhood.

Learning test: teach a new pseudoword neighbour of single, consolidate it, then test whether recognition of the established word slows. Competition provides evidence that the new form has entered the lexical network.

Teaching transfer: practise the target against its strongest real competitors using varied speakers, rates and sentence contexts. The goal is robust discrimination, not memorisation of one clean recording.

21. Final model

NAM makes one central point vivid: a spoken word is recognised against a competitive lexical background. The same acoustic evidence can be easy or difficult depending on how many strong neighbours it activates.

For vocabulary teaching, phonological precision matters because every imprecise contrast changes the competitive neighbourhood the learner experiences.

22. eduKateSG routes

Phonological Neighbourhoods · TRACE Model · Word Recognition · Vocabulary Master Router

23. Extended neighbourhood laboratory

1. cat

Build a target-plus-five-neighbour set for cat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

2. cap

Build a target-plus-five-neighbour set for cap. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

3. cab

Build a target-plus-five-neighbour set for cab. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

4. can

Build a target-plus-five-neighbour set for can. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

5. bat

Build a target-plus-five-neighbour set for bat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

6. bad

Build a target-plus-five-neighbour set for bad. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

7. bag

Build a target-plus-five-neighbour set for bag. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

8. bed

Build a target-plus-five-neighbour set for bed. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

9. bet

Build a target-plus-five-neighbour set for bet. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

10. bell

Build a target-plus-five-neighbour set for bell. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

11. coat

Build a target-plus-five-neighbour set for coat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

12. goat

Build a target-plus-five-neighbour set for goat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

13. boat

Build a target-plus-five-neighbour set for boat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

14. cold

Build a target-plus-five-neighbour set for cold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

15. gold

Build a target-plus-five-neighbour set for gold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

16. fold

Build a target-plus-five-neighbour set for fold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

17. light

Build a target-plus-five-neighbour set for light. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

18. right

Build a target-plus-five-neighbour set for right. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

19. night

Build a target-plus-five-neighbour set for night. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

20. fight

Build a target-plus-five-neighbour set for fight. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

21. pin

Build a target-plus-five-neighbour set for pin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

22. bin

Build a target-plus-five-neighbour set for bin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

23. tin

Build a target-plus-five-neighbour set for tin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

24. win

Build a target-plus-five-neighbour set for win. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

25. ship

Build a target-plus-five-neighbour set for ship. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

26. shop

Build a target-plus-five-neighbour set for shop. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

27. shot

Build a target-plus-five-neighbour set for shot. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

28. shut

Build a target-plus-five-neighbour set for shut. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

29. book

Build a target-plus-five-neighbour set for book. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

30. cook

Build a target-plus-five-neighbour set for cook. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

31. look

Build a target-plus-five-neighbour set for look. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

32. hook

Build a target-plus-five-neighbour set for hook. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

33. seat

Build a target-plus-five-neighbour set for seat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

34. seed

Build a target-plus-five-neighbour set for seed. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

35. seek

Build a target-plus-five-neighbour set for seek. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

36. seem

Build a target-plus-five-neighbour set for seem. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

37. park

Build a target-plus-five-neighbour set for park. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

38. part

Build a target-plus-five-neighbour set for part. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

39. mark

Build a target-plus-five-neighbour set for mark. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

40. dark

Build a target-plus-five-neighbour set for dark. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

41. train

Build a target-plus-five-neighbour set for train. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

42. drain

Build a target-plus-five-neighbour set for drain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

43. brain

Build a target-plus-five-neighbour set for brain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

44. grain

Build a target-plus-five-neighbour set for grain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

45. play

Build a target-plus-five-neighbour set for play. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

46. clay

Build a target-plus-five-neighbour set for clay. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

47. glow

Build a target-plus-five-neighbour set for glow. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

48. grow

Build a target-plus-five-neighbour set for grow. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

49. stone

Build a target-plus-five-neighbour set for stone. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

50. phone

Build a target-plus-five-neighbour set for phone. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

51. mouse

Build a target-plus-five-neighbour set for mouse. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

52. mouth

Build a target-plus-five-neighbour set for mouth. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

53. house

Build a target-plus-five-neighbour set for house. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

54. horse

Build a target-plus-five-neighbour set for horse. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

55. teacher

Build a target-plus-five-neighbour set for teacher. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

56. teaser

Build a target-plus-five-neighbour set for teaser. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

57. reason

Build a target-plus-five-neighbour set for reason. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

58. season

Build a target-plus-five-neighbour set for season. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

59. signal

Build a target-plus-five-neighbour set for signal. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

60. single

Build a target-plus-five-neighbour set for single. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

61. cat

Build a target-plus-five-neighbour set for cat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

62. cap

Build a target-plus-five-neighbour set for cap. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

63. cab

Build a target-plus-five-neighbour set for cab. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

64. can

Build a target-plus-five-neighbour set for can. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

65. bat

Build a target-plus-five-neighbour set for bat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

66. bad

Build a target-plus-five-neighbour set for bad. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

67. bag

Build a target-plus-five-neighbour set for bag. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

68. bed

Build a target-plus-five-neighbour set for bed. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

69. bet

Build a target-plus-five-neighbour set for bet. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

70. bell

Build a target-plus-five-neighbour set for bell. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

71. coat

Build a target-plus-five-neighbour set for coat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

72. goat

Build a target-plus-five-neighbour set for goat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

73. boat

Build a target-plus-five-neighbour set for boat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

74. cold

Build a target-plus-five-neighbour set for cold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

75. gold

Build a target-plus-five-neighbour set for gold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

76. fold

Build a target-plus-five-neighbour set for fold. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

77. light

Build a target-plus-five-neighbour set for light. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

78. right

Build a target-plus-five-neighbour set for right. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

79. night

Build a target-plus-five-neighbour set for night. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

80. fight

Build a target-plus-five-neighbour set for fight. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

81. pin

Build a target-plus-five-neighbour set for pin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

82. bin

Build a target-plus-five-neighbour set for bin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

83. tin

Build a target-plus-five-neighbour set for tin. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

84. win

Build a target-plus-five-neighbour set for win. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

85. ship

Build a target-plus-five-neighbour set for ship. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

86. shop

Build a target-plus-five-neighbour set for shop. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

87. shot

Build a target-plus-five-neighbour set for shot. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

88. shut

Build a target-plus-five-neighbour set for shut. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

89. book

Build a target-plus-five-neighbour set for book. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

90. cook

Build a target-plus-five-neighbour set for cook. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

91. look

Build a target-plus-five-neighbour set for look. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

92. hook

Build a target-plus-five-neighbour set for hook. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

93. seat

Build a target-plus-five-neighbour set for seat. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

94. seed

Build a target-plus-five-neighbour set for seed. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

95. seek

Build a target-plus-five-neighbour set for seek. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

96. seem

Build a target-plus-five-neighbour set for seem. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

97. park

Build a target-plus-five-neighbour set for park. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

98. part

Build a target-plus-five-neighbour set for part. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

99. mark

Build a target-plus-five-neighbour set for mark. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

100. dark

Build a target-plus-five-neighbour set for dark. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

101. train

Build a target-plus-five-neighbour set for train. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

102. drain

Build a target-plus-five-neighbour set for drain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

103. brain

Build a target-plus-five-neighbour set for brain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

104. grain

Build a target-plus-five-neighbour set for grain. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

105. play

Build a target-plus-five-neighbour set for play. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

106. clay

Build a target-plus-five-neighbour set for clay. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

107. glow

Build a target-plus-five-neighbour set for glow. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

108. grow

Build a target-plus-five-neighbour set for grow. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

109. stone

Build a target-plus-five-neighbour set for stone. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.

110. phone

Build a target-plus-five-neighbour set for phone. Assign each item a rough familiarity or frequency rank. Then imagine the critical phoneme is degraded and predict which competitor should become most dangerous under NAM’s relative-evidence logic.

Repeat after adding a newly learned neighbour. Ask whether the learner’s effective neighbourhood has changed even though the acoustic target has not. Finish by testing in a sentence that strongly favours the target and explain how contextual support interacts with, but does not erase, phonological competition.