What Is Vocabulary | The Lexical Decision Task — How Researchers Measure Word Recognition

LEXICAL DECISION TASK · WORD RECOGNITION · REACTION TIME · LEXICAL ACCESS · NONWORDS · WORD FREQUENCY · SEMANTIC PRIMING

The lexical decision task is one of psycholinguistics’ classic methods for studying vocabulary recognition. A participant sees or hears a stimulus and decides whether it is a real word. Researchers then analyse response speed and accuracy to infer how lexical properties such as frequency, familiarity, age of acquisition, semantic priming and neighbourhood structure affect access to the mental lexicon.

For vocabulary research, lexical decision is useful because it separates a simple recognition decision from richer tasks such as defining, recalling or using a word. A learner may recognise a word quickly without possessing deep productive knowledge, while a low-frequency but well-known specialist word may behave differently from a common everyday item.

This guide explains visual and auditory lexical decision, words and pseudowords, reaction time, accuracy, priming, frequency effects, lexical competition, bilingual recognition and experimental limitations. Existing eduKateSG Word Recognition, Lexical Access and Mental Lexicon owners remain untouched.

What the lexical decision task measures

In a lexical decision task, participants decide as quickly and accurately as possible whether a letter string or sound sequence is a real word. Reaction time and error rate provide indirect evidence about lexical recognition.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Words and nonwords

The task contrasts known lexical items with pronounceable or unpronounceable nonwords. Nonword design matters because an obviously impossible string can make the decision artificially easy.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Reaction time

Faster responses are usually interpreted as easier recognition under the tested conditions, but reaction time reflects perception, decision processes, motor response and experimental design as well as lexical access.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Accuracy

Error rates complement speed. A fast response is not useful evidence of efficient recognition if accuracy collapses.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Frequency effects

High-frequency words are generally recognised faster than low-frequency words because repeated exposure strengthens accessibility.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Age of acquisition

Words learned earlier often receive faster lexical decisions even after frequency is considered, though mechanisms and measurement remain debated.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Word length

Longer forms can increase perceptual or orthographic processing demands, but length effects interact with frequency, morphology and neighbourhood structure.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Neighbourhood effects

Similar-looking or similar-sounding words can facilitate activation and create competition. The exact effect depends on modality and task.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Semantic priming

A related prime can speed recognition of a target, revealing activation across semantic networks.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Morphological priming

Roots and affixes can facilitate recognition of related forms, providing evidence about morphological organisation.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Pseudoword effects

Wordlike nonwords are harder to reject than impossible strings because they activate plausible lexical candidates.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Visual lexical decision

Printed lexical decision probes orthographic recognition and its interaction with phonology and meaning.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Auditory lexical decision

Spoken lexical decision adds temporal unfolding: listeners cannot see the whole form at once and must update candidates as sound arrives.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Bilingual lexical decision

Bilingual tasks reveal cross-language activation, cognate facilitation and false-friend competition.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Task limitations

Lexical decision is an experimental proxy, not ordinary reading. Real comprehension rarely asks people to decide explicitly whether every string is a word.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

Teaching interpretation

Teachers can borrow the diagnostic logic—speed plus accuracy plus controlled contrasts—without pretending classroom drills are laboratory measures.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

AI-era interpretation

Computational systems can classify tokens instantly, but human lexical decision remains valuable because it reveals processing cost, competition and familiarity in biological language users.

The central interpretive rule is to avoid treating one response time as a direct reading of “how much vocabulary” a person has. Experimental effects emerge from comparisons across carefully matched conditions.

18. Casebook A — recognition and frequency

1. Lexical decision case — house

Baseline: present house among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare house with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from house. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede house with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use house. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises house rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

2. Lexical decision case — table

Baseline: present table among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare table with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from table. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede table with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use table. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises table rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

3. Lexical decision case — doctor

Baseline: present doctor among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare doctor with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from doctor. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede doctor with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use doctor. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises doctor rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

4. Lexical decision case — nurse

Baseline: present nurse among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare nurse with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from nurse. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede nurse with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use nurse. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises nurse rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

5. Lexical decision case — school

Baseline: present school among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare school with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from school. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede school with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use school. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises school rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

6. Lexical decision case — teacher

Baseline: present teacher among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare teacher with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from teacher. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede teacher with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use teacher. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises teacher rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

7. Lexical decision case — student

Baseline: present student among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare student with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from student. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede student with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use student. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises student rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

8. Lexical decision case — river

Baseline: present river among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare river with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from river. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede river with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use river. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises river rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

9. Lexical decision case — mountain

Baseline: present mountain among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare mountain with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from mountain. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede mountain with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use mountain. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises mountain rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

10. Lexical decision case — garden

Baseline: present garden among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare garden with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from garden. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede garden with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use garden. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises garden rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

11. Lexical decision case — justice

Baseline: present justice among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare justice with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from justice. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede justice with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use justice. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises justice rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

12. Lexical decision case — freedom

Baseline: present freedom among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare freedom with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from freedom. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede freedom with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use freedom. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises freedom rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

13. Lexical decision case — evidence

Baseline: present evidence among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare evidence with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from evidence. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede evidence with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use evidence. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises evidence rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

14. Lexical decision case — method

Baseline: present method among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare method with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from method. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede method with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use method. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises method rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

15. Lexical decision case — theory

Baseline: present theory among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare theory with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from theory. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede theory with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use theory. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises theory rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

16. Lexical decision case — result

Baseline: present result among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare result with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from result. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede result with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use result. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises result rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

17. Lexical decision case — system

Baseline: present system among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare system with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from system. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede system with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use system. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises system rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

18. Lexical decision case — factor

Baseline: present factor among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare factor with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from factor. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede factor with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use factor. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises factor rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

19. Lexical decision case — model

Baseline: present model among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare model with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from model. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede model with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use model. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises model rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

20. Lexical decision case — change

Baseline: present change among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare change with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from change. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede change with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use change. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises change rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

19. Casebook B — competition and priming

21. Lexical decision case — algorithm

Baseline: present algorithm among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare algorithm with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from algorithm. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede algorithm with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use algorithm. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises algorithm rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

22. Lexical decision case — database

Baseline: present database among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare database with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from database. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede database with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use database. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises database rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

23. Lexical decision case — network

Baseline: present network among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare network with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from network. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede network with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use network. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises network rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

24. Lexical decision case — signal

Baseline: present signal among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare signal with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from signal. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede signal with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use signal. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises signal rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

25. Lexical decision case — energy

Baseline: present energy among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare energy with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from energy. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede energy with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use energy. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises energy rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

26. Lexical decision case — matter

Baseline: present matter among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare matter with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from matter. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede matter with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use matter. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises matter rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

27. Lexical decision case — cell

Baseline: present cell among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare cell with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from cell. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede cell with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use cell. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises cell rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

28. Lexical decision case — enzyme

Baseline: present enzyme among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare enzyme with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from enzyme. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede enzyme with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use enzyme. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises enzyme rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

29. Lexical decision case — species

Baseline: present species among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare species with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from species. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede species with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use species. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises species rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

30. Lexical decision case — climate

Baseline: present climate among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare climate with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from climate. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede climate with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use climate. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises climate rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

31. Lexical decision case — inflation

Baseline: present inflation among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare inflation with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from inflation. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede inflation with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use inflation. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises inflation rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

32. Lexical decision case — market

Baseline: present market among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare market with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from market. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede market with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use market. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises market rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

33. Lexical decision case — credit

Baseline: present credit among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare credit with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from credit. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede credit with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use credit. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises credit rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

34. Lexical decision case — policy

Baseline: present policy among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare policy with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from policy. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede policy with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use policy. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises policy rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

35. Lexical decision case — court

Baseline: present court among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare court with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from court. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede court with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use court. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises court rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

36. Lexical decision case — verdict

Baseline: present verdict among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare verdict with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from verdict. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede verdict with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use verdict. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises verdict rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

37. Lexical decision case — contract

Baseline: present contract among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare contract with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from contract. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede contract with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use contract. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises contract rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

38. Lexical decision case — music

Baseline: present music among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare music with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from music. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede music with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use music. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises music rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

39. Lexical decision case — rhythm

Baseline: present rhythm among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare rhythm with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from rhythm. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede rhythm with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use rhythm. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises rhythm rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

40. Lexical decision case — tempo

Baseline: present tempo among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare tempo with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from tempo. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede tempo with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use tempo. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises tempo rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

20. Casebook C — depth and transfer

41. Lexical decision case — novel

Baseline: present novel among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare novel with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from novel. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede novel with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use novel. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises novel rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

42. Lexical decision case — poetry

Baseline: present poetry among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare poetry with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from poetry. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede poetry with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use poetry. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises poetry rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

43. Lexical decision case — metaphor

Baseline: present metaphor among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare metaphor with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from metaphor. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede metaphor with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use metaphor. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises metaphor rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

44. Lexical decision case — argument

Baseline: present argument among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare argument with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from argument. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede argument with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use argument. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises argument rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

45. Lexical decision case — analysis

Baseline: present analysis among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare analysis with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from analysis. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede analysis with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use analysis. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises analysis rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

46. Lexical decision case — compare

Baseline: present compare among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare compare with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from compare. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede compare with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use compare. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises compare rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

47. Lexical decision case — evaluate

Baseline: present evaluate among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare evaluate with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from evaluate. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede evaluate with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use evaluate. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises evaluate rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

48. Lexical decision case — infer

Baseline: present infer among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare infer with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from infer. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede infer with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use infer. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises infer rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

49. Lexical decision case — predict

Baseline: present predict among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare predict with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from predict. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede predict with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use predict. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises predict rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

50. Lexical decision case — explain

Baseline: present explain among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare explain with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from explain. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede explain with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use explain. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises explain rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

51. Lexical decision case — mitigate

Baseline: present mitigate among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare mitigate with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from mitigate. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede mitigate with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use mitigate. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises mitigate rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

52. Lexical decision case — plausible

Baseline: present plausible among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare plausible with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from plausible. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede plausible with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use plausible. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises plausible rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

53. Lexical decision case — reluctant

Baseline: present reluctant among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare reluctant with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from reluctant. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede reluctant with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use reluctant. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises reluctant rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

54. Lexical decision case — ubiquitous

Baseline: present ubiquitous among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare ubiquitous with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from ubiquitous. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede ubiquitous with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use ubiquitous. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises ubiquitous rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

55. Lexical decision case — meticulous

Baseline: present meticulous among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare meticulous with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from meticulous. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede meticulous with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use meticulous. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises meticulous rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

56. Lexical decision case — resilient

Baseline: present resilient among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare resilient with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from resilient. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede resilient with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use resilient. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises resilient rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

57. Lexical decision case — ambiguous

Baseline: present ambiguous among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare ambiguous with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from ambiguous. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede ambiguous with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use ambiguous. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises ambiguous rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

58. Lexical decision case — coherent

Baseline: present coherent among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare coherent with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from coherent. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede coherent with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use coherent. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises coherent rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

59. Lexical decision case — precise

Baseline: present precise among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare precise with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from precise. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede precise with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use precise. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises precise rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

60. Lexical decision case — significant

Baseline: present significant among matched real words and plausible nonwords. Record both reaction time and accuracy. A single trial is noisy; the useful evidence comes from repeated observations and condition-level comparisons.

Frequency control: compare significant with another item of similar length and spelling pattern but different corpus frequency. If the more frequent item is recognised faster across participants, the contrast supports a frequency effect rather than a simple length effect.

Neighbourhood control: construct a plausible nonword differing by one letter or phoneme from significant. Harder rejection indicates that wordlike competitors activate lexical knowledge and make the decision less trivial.

Priming test: precede significant with a semantically related prime and an unrelated prime. A reliable speed advantage after the related prime provides evidence that activation spreads through semantic relationships.

Depth warning: correct lexical decision does not prove the learner can define, spell, pronounce, collocate or use significant. Follow recognition with meaning and production tasks before calling the word mastered.

Classroom translation: use timed recognition only as one diagnostic layer. If the learner recognises significant rapidly but cannot use it, move instruction toward retrieval and contextual production rather than more recognition exposure.

21. Final model

Lexical decision measures the efficiency of deciding that a form belongs to the lexicon under controlled conditions. Its power comes from contrasts: word versus nonword, frequent versus rare, primed versus unprimed, dense versus sparse neighbourhoods and familiar versus unfamiliar forms.

Its limitation is equally important. Fast recognition is one component of vocabulary knowledge, not the whole construct. Deep vocabulary includes meaning, form, grammar, collocation, register and productive access.

22. eduKateSG routes

Word Recognition · Lexical Access · Mental Lexicon · Vocabulary Master Router

23. Extended experimental design lab

1. Design extension — house

Build four matched conditions around house: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

2. Design extension — table

Build four matched conditions around table: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

3. Design extension — doctor

Build four matched conditions around doctor: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

4. Design extension — nurse

Build four matched conditions around nurse: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

5. Design extension — school

Build four matched conditions around school: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

6. Design extension — teacher

Build four matched conditions around teacher: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

7. Design extension — student

Build four matched conditions around student: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

8. Design extension — river

Build four matched conditions around river: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

9. Design extension — mountain

Build four matched conditions around mountain: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

10. Design extension — garden

Build four matched conditions around garden: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

11. Design extension — justice

Build four matched conditions around justice: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

12. Design extension — freedom

Build four matched conditions around freedom: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

13. Design extension — evidence

Build four matched conditions around evidence: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

14. Design extension — method

Build four matched conditions around method: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

15. Design extension — theory

Build four matched conditions around theory: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

16. Design extension — result

Build four matched conditions around result: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

17. Design extension — system

Build four matched conditions around system: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

18. Design extension — factor

Build four matched conditions around factor: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

19. Design extension — model

Build four matched conditions around model: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

20. Design extension — change

Build four matched conditions around change: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

21. Design extension — algorithm

Build four matched conditions around algorithm: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

22. Design extension — database

Build four matched conditions around database: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

23. Design extension — network

Build four matched conditions around network: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

24. Design extension — signal

Build four matched conditions around signal: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

25. Design extension — energy

Build four matched conditions around energy: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

26. Design extension — matter

Build four matched conditions around matter: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

27. Design extension — cell

Build four matched conditions around cell: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

28. Design extension — enzyme

Build four matched conditions around enzyme: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

29. Design extension — species

Build four matched conditions around species: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

30. Design extension — climate

Build four matched conditions around climate: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

31. Design extension — inflation

Build four matched conditions around inflation: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

32. Design extension — market

Build four matched conditions around market: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

33. Design extension — credit

Build four matched conditions around credit: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

34. Design extension — policy

Build four matched conditions around policy: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

35. Design extension — court

Build four matched conditions around court: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

36. Design extension — verdict

Build four matched conditions around verdict: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

37. Design extension — contract

Build four matched conditions around contract: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

38. Design extension — music

Build four matched conditions around music: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

39. Design extension — rhythm

Build four matched conditions around rhythm: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

40. Design extension — tempo

Build four matched conditions around tempo: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

41. Design extension — novel

Build four matched conditions around novel: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

42. Design extension — poetry

Build four matched conditions around poetry: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

43. Design extension — metaphor

Build four matched conditions around metaphor: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

44. Design extension — argument

Build four matched conditions around argument: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

45. Design extension — analysis

Build four matched conditions around analysis: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

46. Design extension — compare

Build four matched conditions around compare: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

47. Design extension — evaluate

Build four matched conditions around evaluate: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

48. Design extension — infer

Build four matched conditions around infer: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

49. Design extension — predict

Build four matched conditions around predict: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

50. Design extension — explain

Build four matched conditions around explain: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

51. Design extension — mitigate

Build four matched conditions around mitigate: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

52. Design extension — plausible

Build four matched conditions around plausible: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

53. Design extension — reluctant

Build four matched conditions around reluctant: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

54. Design extension — ubiquitous

Build four matched conditions around ubiquitous: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

55. Design extension — meticulous

Build four matched conditions around meticulous: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

56. Design extension — resilient

Build four matched conditions around resilient: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

57. Design extension — ambiguous

Build four matched conditions around ambiguous: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

58. Design extension — coherent

Build four matched conditions around coherent: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

59. Design extension — precise

Build four matched conditions around precise: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

60. Design extension — significant

Build four matched conditions around significant: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

61. Design extension — house

Build four matched conditions around house: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

62. Design extension — table

Build four matched conditions around table: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

63. Design extension — doctor

Build four matched conditions around doctor: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

64. Design extension — nurse

Build four matched conditions around nurse: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

65. Design extension — school

Build four matched conditions around school: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

66. Design extension — teacher

Build four matched conditions around teacher: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

67. Design extension — student

Build four matched conditions around student: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

68. Design extension — river

Build four matched conditions around river: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

69. Design extension — mountain

Build four matched conditions around mountain: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

70. Design extension — garden

Build four matched conditions around garden: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

71. Design extension — justice

Build four matched conditions around justice: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

72. Design extension — freedom

Build four matched conditions around freedom: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

73. Design extension — evidence

Build four matched conditions around evidence: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

74. Design extension — method

Build four matched conditions around method: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

75. Design extension — theory

Build four matched conditions around theory: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

76. Design extension — result

Build four matched conditions around result: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

77. Design extension — system

Build four matched conditions around system: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

78. Design extension — factor

Build four matched conditions around factor: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

79. Design extension — model

Build four matched conditions around model: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

80. Design extension — change

Build four matched conditions around change: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

81. Design extension — algorithm

Build four matched conditions around algorithm: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

82. Design extension — database

Build four matched conditions around database: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

83. Design extension — network

Build four matched conditions around network: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.

84. Design extension — signal

Build four matched conditions around signal: high-frequency real word, low-frequency real word, wordlike pseudoword and orthographically implausible nonword. Keep length and response mapping balanced. Predict which contrast should affect lexical activation and which may reflect only perceptual ease.

Then repeat the design auditorily. Because spoken words unfold through time, identify the point at which competitors diverge and explain why the auditory decision may differ from the visual decision. Finish by testing meaning separately so recognition speed is not mistaken for complete lexical knowledge.