Bilingualism

What are the key types and terms related to bilingualism?

• Early bilinguals: exposed to two languages in childhood

  • Simultaneous: learn both at the same time

  • Early sequential: learn one first, then the second soon after

• Late learners: learn a second language after mastering their first

• Balanced bilinguals: equally skilled in both languages

• Multilinguals: speak more than two languages

• L1 (first language): mother tongue; not always the dominant one

• Dominance: the language used most at a given time; can change

• L2 (second language): the additional language learned

What are the key terms in bilingualism?

• Age of acquisition: when a word/language is learned

  • Related to, but different from, frequency

• Frequency: how often a word is used

• Proficiency (Competence): overall ability in a language

  • Levels: native, near-native, learner

• Fluency: ease, speed, and accuracy of language production

  • Can be fast but error-prone

• Dominance: most-used or most automatic language at a given time

  • Reflected in ease of use or even dreaming

  • Changes over time; independent of proficiency/fluency

How do bilinguals manage two language systems in one mind?

• Must keep two language systems separate to avoid:

  • Mixing syntax from one language with words from another

  • Unintentional code-mixing within a sentence

• Key questions in bilingualism:

  • How are the two language systems represented and related in the mind?

  • How are they kept separate during:

    • Speaking

    • Listening/reading

• Broader issue:

  • How do dialects compare to full languages?

How does translating between languages work in the mind?

• Assumes a single conceptual store for meaning

• Both L1 and L2 connect to this shared meaning system

• Alternative outlook: translation from L1 to L2 for understanding?

• Concept mediation:

  • Translation goes through meaning

  • Requires accessing the concept before translating

• Word association:

  • Direct link between L1 and L2 words

  • No need to access meaning explicitly

  • More like word-to-word translation

How can we tell if L2 speakers have direct conceptual access?

• Compare L1 → L2 translation vs L2 picture naming

• In L1:

  • Word reading is ~250ms faster than picture naming

  • Extra time = accessing concepts

• Predictions:

  • Word association model:

    • Translation faster (no concept access needed)

  • Concept mediation model:

    • Same speed (both require concept access)

• Study: Potter et al.

  • Tested fluent Chinese–English bilinguals

  • Found no speed difference between tasks

• Conclusion:

  • Supports concept mediation (direct access to meaning)

Get smarter responses, upload files and images, and more.

Log in

Sign up for free

What is concept mediation in bilingual language processing?

• Concept mediation: both L1 and L2 access a shared conceptual system

• Picture naming (3 steps):

  • Picture → concept → word (L1 or L2)

• Translation (3 steps):

  • L1 word → concept → L2 word

• Key idea:

  • Concepts have bidirectional links to both languages

• Note on processing:

  • L1 may access concepts more easily, helping L2 retrieve meaning

What is the word association model in bilingualism?

• Word association: L2 accessed via direct links to L1 (not concepts)

• Picture naming (4 steps):

  • Picture → concept → L1 → L2

• Translation (2 steps):

  • L1 word → L2 word (direct link, no concept needed)

• Prediction:

  • Translation should be faster than picture naming

• Evidence:

  • Potter et al. found no speed advantage

• Conclusion:

  • Model not supported

How is bilingual processing related to fluency?

• Study: Kroll & Curley (late L2 learners, varying experience)

• < 2 years learning L2:

  • Faster at translation than picture naming

  • Suggests reliance on word association (L1 → L2 links)

• > 2 years learning L2:

  • No speed difference

  • Suggests concept mediation (direct access to meaning)

• Conclusion (developmental hypothesis):

  • Early stage: build word-to-word links between languages

  • Later stage: develop direct links between L2 and concepts

What is the developmental hypothesis in bilingualism?

• Low proficiency:

  • Rely on L1 to access meaning (L2 → L1 → concepts)

• Higher proficiency:

  • Build direct links between L2 and concepts

  • Reduced reliance on L1

• System structure:

  • Bidirectional links between concepts, L1, and L2

  • Shared semantic system across both languages

  • Same meaning for words across languages (e.g., “cat” / “gatto”)

• Key issue:

  • How bilinguals avoid interference between the two languages

You’re now using our basic model

Do bilinguals use selective or non-selective language activation?

• Key issue: how bilinguals avoid interference from the non-used language

• Selective activation hypothesis:

  • Only the target language is activated

  • Other language is fully suppressed

• Non-selective activation hypothesis:

  • Both languages are automatically activated

  • Requires later selection of the correct language

• Core debate: whether bilingual control happens by blocking one system or managing competition between both

Is the non-target language active when bilinguals speak?

• Study: Costa et al. (2000)

• Method: compared picture naming in bilinguals vs monolinguals

  • Focus: cognates vs non-cognates

• Cognates: words that sound and mean similar across languages

  • e.g., cat–“Katze”, bread–“Brot”, hair–“Haar”

• Findings:

  • Bilinguals are faster for cognates

  • Shows cross-language activation during speech

• Conclusion:

  • Supports non-selective activation

  • Both languages are automatically active even when speaking one language

What did Costa et al. (2000) show about bilingual language activation?

• Key idea: shared sounds (cognates) get double activation in bilinguals

• Findings:

  • Bilinguals named cognates faster than non-cognates

  • Monolinguals showed no cognate advantage

• Interpretation:

  • In bilinguals, both languages activate phonology

  • Shared sound patterns receive combined activation boost

• Conclusion:

  • Supports non-selective activation

  • Both language systems are active and influence speech production at the phonological level

Is the non-target language active when bilinguals read?

• Study: Dijkstra et al.

• Method: visual lexical decision task using interlingual homographs

  • Words spelled the same but different meanings across languages

  • e.g., “coin” (French: corner), “room” (Dutch: cream), “brand” (Dutch: fire)

• Findings:

  • Slower responses for homographs vs non-homographs

  • Suggests both language meanings are activated simultaneously

• Interpretation:

  • Non-target language meaning interferes during reading

  • Leads to uncertainty in word recognition

• Conclusion:

  • Supports non-selective activation in reading

  • Both languages are active during word recognition processes

Is the non-target language active when bilinguals listen?

• Study: Marian & Spivey (2003)

• Method: eye-tracking during spoken language comprehension

  • Example: “Click on the plug” (L2 English) while L1 Russian contains similar-sounding items

• Findings (when hearing L2):

  • Within-language competition (L2 words compete with each other)

  • Between-language competition (L1 also interferes)

  • → L1 is active and influences L2 processing

• Findings (when hearing L1):

  • Only within-language competition

  • L2 does not strongly interfere with L1

• Conclusion:

  • Supports non-selective activation in listening

  • Influence is asymmetric and depends on language dominance (L1 > L2 effect)

How do bilinguals prevent interference between two active languages?

• Problem: both lexicons are often simultaneously active

• Language non-specific account:

  • Both languages activate together

  • Non-target language is inhibited

  • Inhibition strength depends on dominance

    • L1 is harder to suppress than L2 (unless L1 is rarely used)

  • Some cross-language competition remains if inhibition is incomplete

• Language-specific account:

  • Both systems may be active, but only target language is considered

  • Non-target language is effectively ignored, not inhibited

  • Selection mechanism restricts competition to one language system only

  • Predicts no cross-language interference

Is the non-target language competing during L2 speech production?

• Study: Hermans et al. (1998)

• Method: picture-word interference task in L2 naming

  • Participants name pictures while ignoring distractor words

• Distractor types:

  • Phonological (L2): e.g., “mouth”

  • Semantic: e.g., “valley”

  • Phonological (L1): e.g., “bench”

  • Unrelated: e.g., “table”

• Key test:

  • If L1 is active, L1 phonological distractors should affect L2 naming

• Findings:

  • L1 phonological distractors do influence L2 naming

  • Shows cross-language activation

• Conclusion:

  • Supports non-target language competition

  • L1 remains active during L2 production and can interfere at the phonological level

What did Hermans et al. find about non-target language interference?

• Study: Hermans et al. (1998)

• Finding:

  • L2 picture naming (“mountain”) was slower when paired with L1 distractor “bench” vs unrelated words

• Interpretation:

  • L1 word “bench” activates its translation equivalent (“Berg”)

  • This increases competition during L2 naming

• Conclusion:

  • Supports non-target language activation

  • Non-target language words can interfere by strengthening competing lexical candidates

How do bilinguals separate two active languages during speaking?

• Both languages are simultaneously active during speech

• Evidence of interaction:

  • Cognate facilitation (similar words across languages are faster)

  • Cross-language interference (non-target words can slow naming)

  • Indirect semantic/phonological competition affects picture naming

• Mixed effects:

  • Translations can speed naming

  • But competitors can also slow it down

• Overall interpretation:

  • Non-target language is automatically activated

  • Then rapid inhibition helps select the correct language

  • Language interaction is dynamic, not purely inhibitory or purely separate

What is the evidence for inhibition in bilingual language switching?

• Studies: Allport & Meuter (1999) and Costa & Santesteban (2004)

• Method: language switching task

  • Participants name pictures in either of two languages

• Task setup:

  • Colour cues indicate language (e.g., green = English, red = Spanish)

  • Rapid switching between languages required

• Key idea:

  • Switching performance reflects inhibitory control of the non-target language

• Conclusion:

  • Provides evidence that bilinguals use inhibition to manage competing languages during production

What is evidence for asymmetrical inhibition in bilingual language switching?

• Studies: Allport & Meuter (1999) and Costa & Santesteban (2004)

• Method: language switching tasks (naming pictures in L1 or L2)

• Key findings (late learners):

  • Faster on non-switch trials than switch trials (switch cost)

  • Asymmetrical switch costs:

    • Slower to switch into L1 from L2

    • Faster to switch into L2 from L1

• Interpretation:

  • L1 is strongly inhibited to allow L2 use

  • L2 requires less inhibition when L1 is active

• Balanced bilinguals:

  • No asymmetry in switch costs

  • Suggests similar inhibition levels for both languages

What is the Inhibitory Control Model of bilingual language production?

• Model: Kroll, Bobb, & Wodneicka, 2006

• Core idea: lexical selection is top-down

  • Concepts activate candidate words (lemmas) in both languages

• Processing:

  • A cohort of lemmas from both languages becomes activated

• Control mechanism:

  • Language cue/tag guides selection (e.g., “English” or “Spanish”)

  • Cue inhibits non-target language representations

• Outcome:

  • Correct language is selected by suppressing competing language systems

  • Enables controlled bilingual speech production

What are the cognitive costs and benefits of being bilingual?

• Past view: bilingualism seen as confusing or harmful to development

  • Concern about “foreignness” and delayed language growth

• Early development:

  • Possible slight initial delay in language acquisition

  • But children typically catch up before school age

• Cognitive demands:

  • Each language gets reduced practice/frequency of use

  • Requires strong executive control, attention, and switching ability

• Key idea:

  • Managing two languages trains control systems in the brain

• Current findings:

  • Evidence of cognitive benefits from bilingual experience

  • Especially in attention and task switching

What are the cognitive costs of being bilingual?

• Research (Gollan and colleagues): bilinguals show slower word retrieval than monolinguals

• Key findings:

  • Slower naming of common objects, even in dominant language

  • Fewer words produced in free recall tasks

  • More tip-of-the-tongue (TOT) states

• Possible explanations:

  • Competition between two phonological systems

  • Weaker Links Hypothesis: weaker connections between concepts and words due to divided use

• Supporting evidence:

  • CODAs (children of deaf adults) also show more TOTs

  • Suggests issue is weaker links, not direct language competition

What does generational change in dialect attitudes show (Buckie study)?

• Study: dialect attitudes in Buckie

• School students (15–17 years):

  • Mostly lifelong residents

  • Some report strong regional accent

  • Almost all would change accent in different settings

  • Motivation: be better understood / more professional

• Older community members (42–84 years):

  • Mix of lifelong and non-lifelong residents

  • All report strong regional accent

  • Most would change accent for clarity only

  • No mention of professionalism as a reason

• Conclusion:

  • Younger speakers show more social awareness and flexibility in accent use

  • Suggests generational shift in attitudes toward dialect and identity

What are the differences between languages and dialects in lexical processing?

• Study 1: Costa & Caramazza (1999)

  • Compared words across different languages (e.g., TABLE–MESA)

  • Measured reaction times for: same meaning, related meaning, unrelated pairs

  • Tested within- vs between-language processing

• Study 2: Melinger (2018)

  • Compared words across dialects (e.g., PITCHER–JUG)

  • Same design: semantic relatedness and reaction times

  • Tested within- vs between-dialect processing

• Key comparison:

  • Both languages and dialects show similar patterns of semantic activation and interference

• Conclusion:

  • Cognitive processing of dialects is similar to languages

  • Supports idea that distinction between language and dialect is graded rather than absolute

What is the summary of bilingual language processing?

• Bilingual system is similar to monolingual system but with two languages, increasing potential for error and interference

  • Yet bilinguals do not show proportionally more difficulties

• Late learners:

  • Build L2 on top of L1

  • Gradually develop direct L2–concept links, reducing reliance on L1

• Language activation:

  • Both lexica are automatically active in speaking and comprehension

  • Language dominance reduces interference from the non-target language

• Dialects and accents:

  • Growing area of research with parallels to bilingualism

  • Accent/dialect cues do not affect early lexical competition during listening