0 The development of lexical proficiency in L2 speaking and writing tasks by Dutch-speaking learners of French in Brussels Bram Bulté & Alex Housen ACQUILANG.

1

The development of lexical proficiency in L2 speaking and

writing tasks by Dutch-speaking learners of French in

Brussels

Bram Bulté & Alex Housen

ACQUILANG (Centre for Studies on Second Language Learning & Teaching)

TBLT 2009, Lancaster

.

• Research project on the development of L2

proficiency in French, English and Dutch in

different educational contexts.

• Theoretical, conceptual and methodological

issues and empirical research.

• Empirical research:• Longitudinal

• Learner background data

• Spoken and written L2 production data.

Research background

.

1. Comparison between the oral and written task

modality.

2. Framework for the analysis of lexical L2

development.

3. Empirical study of the lexical development of

Dutch-speaking learners of French.

Outline presentation

.

PART I: Comparison between the oral and

written task modality

Outline

.

Oral and written modes

Influence of mode on lexical performance

• Oral L2 production is considered to give evidence of the

learner’s implicit knowledge (Towell et al., 1996); written

production L2 allows for the use of explicit knowledge.

• Writing is 5 to 8 times slower than speaking in the same

individual (Fayol, 1997).

• Difficult to separate effect of (esp. online) planning from

effect of mode.

.

Oral and written modes

Lexical differences between speaking and writing:

• Disfluency markers: lubricators, interjections, fillers,

modifiers, …

• Repetition and paraphrase, false starts.

• Clause linking: small range of connectors.

• Words with vague semantics.

• Low lexical density.

.

PART II: Framework for the analysis of lexical

L2 development

Outline

.

Lexical L2 competence

• Word = ‘lexical entry’ (Jiang 2000).

• Knowing vs. using a word -> ability to use the relevant

lexical information in a wide range of contexts when the

need arises (McCarthy, 1990).

• Lexical competence = lexical knowledge and ability to

apply that knowledge (procedural).

• Lexical proficiency = the concrete manifestation of lexical

competence

.

Measuring lexical L2 competence• What do we want to measure? Extent of lexical competence.

• Lexical competence = lexical knowledge + procedural

knowledge

• Lexical knowledge can be characterized by its size, width

and depth.

• Size refers to the number of lexical entries in memory.

• Width and depth refer to the quality and degree of

elaboration of the knowledge of the lexical entries in

memory.

• Procedural knowledge is a matter of control / skill / ability.

.

Measuring lexical L2 competence

Different options:

• Purpose-built tests vs. ‘free’ language production.

• Subjective rating vs. ‘objective’ measures.

Methodological concerns:

• Which quantitative measures should be used to

assess lexical competence?

.

Measuring lexical L2 competenceProposed quantitative measures:

• Number of different (content) words or lemmas

=> Lexical productivity

• TTR and transformations (Guiraud, Uber, Herdan, D)

=> Lexical diversity

• Proportion of lexical / function words

=> Lexical density

• Frequency based measures (LFP, Advanced G)

=> Lexical sophistication

• Temporal measures (words / time unit)

=> Lexical fluency

• Error analysis

=> Lexical accuracy

.

Measuring lexical L2 competence• What do we want the measures to measure?

How do they relate to the theoretical view on lexical competence?

• 3 levels of analysis:• Theoretical level of cognitive constructs

• Observational level of behavioral constructs

• Operational level of statistical constructs

.

Measuring lexical L2 competenceRelations between different levels of analysis

THEORETICAL LEVELOF

COGNITIVE CONSTRUCTS:LEXICAL COMPETENCE

OBSERVATIONAL LEVELOF

BEHAVIOURAL CONSTRUCTS:

LEXICAL PROFICIENCY ASMANIFESTED IN L2 PROCESSING

(esp. PRODUCTION)

OPERATIONAL LEVELOF

STATISTICAL CONSTRUCTS:

SCORES, MEASURES &MEASUREMENTS

SIZEOF KNOWLEDGE

WIDTHOF KNOWLEDGE

DEPTHOF KNOWLEDGE

PROCEDURALKNOWLEDGE

LEXICALDIVERSITY

LEXICALSOPHISTICATION

LEXICALCOMPLEXITY

LEXICALPRODUCTIVITY

LEXICALFLUENCY

GENERALTTRÕs

(andtransformations)

FREQUENCY-BASED LEXICAL

MEASURES

(Advanced TTRÕs)

LEXICALCLOZE TEST

(or gap-fillingtest)

NUMBER OF(CONTENT)

WORD TYPES

(or lemmas)

TEMPORALMEASURES (e.g.

content words / time unit)& LEXICAL

DECISION TASKS

.

PART III: Empirical study of the lexical

development of Dutch-speaking

learners of French

Outline

.

1. How does the oral and written lexical performance in the FFL production of Dutch-speaking L2 learners develop over time?

2. Is there a difference in scores for written and spoken tasks? (group comparison)

3. Are learners’ lexical proficiency scores similar for written and oral tasks? (intra-individual comparison)

4. Is the lexical development of learners comparable for oral and written tasks? (inter-individual comparison)

RESEARCH QUESTIONS

.

“Lexical Diversity in Writing and Speaking Task Performances”

• “First study” comparing lexical diversity of spoken and written discourses produced by the same participants.

• Lexical diversity (D) of writing and speaking performances approximately at the same level.

• Lexical diversity (D) of compositions and interviews significantly correlated (r = 0.448).

YU (2009)

.

Subjects:

• 15 pupils, Dutch native speakers, 15-17y old, 3rd-5th grade, Dutch-speaking schools in Brussels.

Tasks:

• 1 oral task: retelling of a wordless picture story (frog story)

• 2 written tasks:

o Complaint letter

o Argument for or against a statement

Data collection:

• Longitudinal, 3 test times, 1y intervals – corpus-based.

RESEARCH DESIGN

.

Data processing:

• Recorded oral tasks and written tasks transcribed in CHAT-format.

• Spelling mistakes in written tasks corrected.

• Non-French words and interlanguage words tagged (@il).

• Hesitations, self-correction and repetitions coded in oral transcriptions.

• Excluded from analysis: interjections & recasts.

• ‘Chunks’ treated as one word (parce+que, à+côté).

• Words were lemmatized.

• Lexical words tagged (|lex).

PRODUCTIVE LANGUAGE CORPUS

.

Quantitative measures:

• Productivity: # tokens, # types, # lexical types.

• Diversity: D, G and U (all words), G and U (lexical words).

• Density: % of lexical words (lexical words / all words).

• Sophistication: # ‘advanced’ types, ‘advanced’ G and U (advanced types / V all tokens), % of advanced types (advanced types / all types).

• Combination: D, G Lex and G Advanced combined.

Statistical analyses:

• Correlations.

• Repeated measures ANOVA, with pair-wise comparisons.

DATA ANALYSIS

.

Combined measures: D, G Lex and G Adv.

Rescaling scores: Average score = 100

=> y1 = y * (100 / ȳ)

Formula: (D*(100/AvgD)+Glex*(100/AvgGlex)+Gadv*(100/AvgGadv))/3

DATA ANALYSIS

.

RESULTS

All types

{} 1 2 3

Written + + + +

Spoken + + / +

W vs. S /

W x S .43

% Lexical words

{} 1 2 3

Written / / / /

Spoken / / / /

W vs. S +

W x S /

.

RESULTS

D (all words)

{} 1 2 3

Written + / / +

Spoken + + / +

W vs. S +

W x S .45

G (lexical words)

{} 1 2 3

Written + + / +

Spoken + + / +

W vs. S +

W x S .52

.

RESULTS

Advanced types

{} 1 2 3

Written + + / +

Spoken + + / +

W vs. S /

W x S .32

% Advanced types

{} 1 2 3

Written + / / +

Spoken + / / +

W vs. S +

W x S /

.

RESULTS

Advanced G

{} 1 2 3

Written + + / +

Spoken + + / +

W vs. S +

W x S /

Combined

{} 1 2 3

Written + + / +

Spoken + + / +

W vs. S +

W x S .41

.

RESULTS

Similar individual development on written and spoken tasks?

• Gain scores for different measures and different modes.

• Not 1 significant correlation found between the gain scores of learners on the same measures for the 2 different modes.

=> Seems like progress on both modes is not related.

.

CONCLUSIONS

1. Development of lexical proficiency:

2. Written versus spoken tasks:

3. Intra-learner task correlation:

4. Similar development on ≠ tasks:

Summary

.

CLOSING REMARKS

1. Lexical proficiency in writing and speaking tasks.

• Increase, both on written and spoken.

• No parallel development on written and spoken tasks.

• Higher scores for writing tasks.

• Moderately high correlation between speaking and writing scores.

2. Limitations.

3. Directions for future research.

.

[email protected]

[email protected]

Thank you!