Why should we care about lexical neighborhoods? What is CLEARPOND? What does CLEARPOND reveal about...

Preview:

Citation preview

• Why should we care about lexical neighborhoods?

• What is CLEARPOND?• What does CLEARPOND reveal about

languages’ structures?• How can I use CLEARPOND?– http://clearpond.northwestern.edu

Orthographic Neighbors

Addition

Substitution

Deletion

Phonological Neighbors

Addition

Substitution

Deletion

Neighborhoods Effects onLanguage Processing

• Lexical decision (Andrews, 1992; Luce & Pisoni, 1998; Vitevitch & Luce, 1999; Yates, Locker, & Simpson, 2004; Ziegler, Muneaux, & Grainger, 2003)

• Word learning (Frisch et al., 2000; Luce & Large, 2001; Thorn &

Frankish, 2005; Roodenrys & Hinton, 2002; Storkel et al., 2006)

Differences across languages

• Verbal picture naming– Dense phonological neighborhoods in English or

Dutch speed up picture naming (Groot, Borgwaldt, Bos, & Vandeneijnden, 2002; Marian & Blumenfeld, 2006; Vitevitch & Stamer, 2006)

– Dense phonological neighborhoods in Spanish slow down picture naming (Vitevitch & Stamer, 2006)

face2 substitution neighbors (fact, race)

1 deletion neighbor (ace)

face3 substitutions

1 deletion1 addition

21

face5 Spanish substitutions

311

• SUBTLEX-NL (Dutch)• SUBTLEX-US (English)• LEXIQUE (French)• SUBTLEX-DE (German)• SUBTLEX-ESP (Spanish)

• Phonological transcriptions: eSpeak(http://espeak.sourceforge.net/)

Introducing:

Dutch English French German Spanish

Dutch English French German Spanish

27,751 Words in Each Language

Dutch English French German Spanish

Dutch English French German Spanish

Comparing neighborhoods across languages

Homophones (light)Unique words (dark)

mer /mɛʁ/

ver /vɛʁ/vers /vɛʁ/vert /vɛʁ/verre /vɛʁ/

Recap

• English has the largest orthographic neighborhood density

• French has the largest phonological neighborhood density– Driven by homophones

• Substitution neighbors are more common than additions or deletions

•Most words are 6-8 letters long, but shorter words have larger neighborhoods.•Average word length in phonemes is language-specific

Foreign Neighbors

75% English Neighbors25% French Neighbors

Orthographic Within-Language and Foreign Neighbors

Phonological Within-Language and Foreign Neighbors

Vitevitch, 2011

How to use CLEARPOND in your research

http://clearpond.northwestern.edu

1. Multi-Word Input2. Rich Output

Selection3. Frequency/length

Filters to generate word lists

4. Neighborhood filters across languages

Example Stimuli Creation:Bilingual lexical decision task

• Generate four lists:– English ‘bridge’ words: large English and French

neighborhoods– English-specific words: large English

neighborhoods, but no French neighbors– French ‘bridge’ words: large French and English

neighborhoods– French-specific words: large French

neighborhoods, but no English neighborhoods

• Built on comparable word lists across Dutch, English, French, German, and Spanish

• Reveals common neighborhood patterns as well as language-specific trends

• Calculates neighborhoods across languages for cross-linguistic and bilingual research

• Free, online tool (http://clearpond.northwestern.edu)

– Look up neighborhoods for a list of words– Generate a list based on desired characteristics

Acknowledgments• Bilingualism and Psycholinguistics Lab

– Tony Lam– Xin Wang– Scott Schroeder

• Funding sources– NICHD RO1 HD059858-01A

Visit http://clearpond.northwestern.edu

And see our manuscript in PLoS ONE (Marian, Bartolotti, Chabal, and Shook, 2012)