Second Language Learning Notes

Problems in Output and the Cognitive Processes They Generate

  • This paper discusses how producing a second language (L2) can make learners aware of linguistic problems.
  • Noticing these problems can motivate learners to modify their output.
  • This modification may require more syntactic processing compared to comprehension.
  • Output initiates a 'noticing' process, which then triggers mental processes leading to modified output, contributing to second language learning.

Introduction

  • Output's role in second language (SL) proficiency development has been largely ignored.
  • This paper argues for the roles of output in second language learning.
  • It presents data showing how problems during L2 production can trigger cognitive processes involved in second language learning.
  • The paper is structured to provide background, discuss output functions, and present data from adolescent French learners.
  • The aim is not to minimize comprehension but to show how output can uniquely facilitate second language learning.

Background

  • Immersion programs in Canada and the US integrate second language teaching with content teaching.
  • These programs aim for both academic achievement and high L2 proficiency.
  • Research indicates immersion students achieve native-like listening and reading comprehension by the end of elementary school.
  • However, their speaking and writing skills still reveal them as non-native speakers.
  • Despite continued progress, their interlanguage remains noticeably 'off-target'.
  • This challenges theories claiming comprehensible input as the sole source of second language acquisition.
  • Immersion students become fluent and can communicate effectively but their L2 development slows over time.
  • The focus is now on helping these learners achieve more accurate, target-like French.
  • A classroom study revealed limited sustained talk in French among students.
  • Teacher feedback on errors appeared unsystematic.

Functions of Output

  • The limited sustained talk in French prompted a re-evaluation of output's role in L2 learning.
  • Producing L2 makes learners aware of linguistic problems through external or internal feedback.
  • Noticing a problem encourages learners to modify their output.
  • This can lead to syntactic processing, unlike the semantic processing common in comprehension.
  • Output can trigger 'noticing' and stimulate linguistic analysis.
  • Schmidt and Frota's (1986) 'notice the gap principle': A learner acquires a target-like form only if it's in comprehended input and consciously noticed.
  • The hypothesis is that output is a trigger for noticing, prompting recognition of what they don't know.
  • Producing the target language may make learners consciously recognize their linguistic problems.
  • The 'output hypothesis' suggests that learners may notice gaps in their knowledge while producing L2, even without external feedback.
  • Communication strategy literature shows learners notice problems and attempt to address them.
  • The paper examines whether learners focus on morphology and syntax when they notice a problem.
  • Other research indicates learners modify output in response to clarification requests or confirmation checks during meaning negotiation.
  • Pica et al. (1989) found that over one-third of learners' responses were modified semantically or morphosyntactically in response to clarification and confirmation requests.
  • It is assumed that modification contributes to second language acquisition, although it has not been directly proven that reprocessed responses are maintained in the learner's interlanguage.
  • Nobuyoshi and Ellis (1993) suggest 'pushing' learners to improve accuracy results in both immediate and long-term accuracy gains, with a small sample size.
  • They distinguish between acquisition as internalizing new forms and increasing control over existing forms.
  • Modifying output leads to increased ability to use existing grammatical knowledge more accurately and learners 'test hypotheses about the second language, experiment with new structures and forms, and expand and exploit their interlanguage resources in creative ways'.
  • Modified output represents the leading edge of the learner's interlanguage.
  • Tarone (1993) relates variation theory to second language acquisition, suggesting that communication strategies stretch a learner's interlanguage system.
  • Tarone (1993) A learner's use of communication strategies can function to stretch an IL system beyond its current limits, resulting in free variation as the learner tests new hypotheses in the search for an appropriate word or structure.
  • Liu (1991)'s study of a Chinese child showed more complex linguistic structures in interactions with a researcher, suggesting interlanguage stretching.
  • Structures used with the researcher appeared later in other settings, suggesting the researcher-learner context fosters development.
  • Tarone (1993) suggests that the need to produce output that the interlanguage system cannot handle pushes the learner to expand the system.
  • Accumulating evidence supports the claim that pushing learners beyond their current level enhances performance, representing new knowledge internalization or existing knowledge consolidation.
  • This paper examines mental processes learners use in moving from original to modified output, from encountering a linguistic problem to developing a solution.
  • Producing language involves different processes than comprehending it.
  • Van Dijk and Kintsch (1983) showed that comprehension relies on strategies rather than strict grammatical rules.
  • Comprehension may involve guessing based on syntactic or semantic clues and using content words alone.
  • Krashen (1982) suggests syntax is often unnecessary for understanding as we often get the message with a combination of vocabulary and extra-linguistic information.
  • Gary and Gary (1981) state that speaking requires linguistically more complex tasks than comprehension.
  • Lightbown and Halter (1993) suggest focused instruction and corrective feedback are needed to fill gaps in comprehension-based language learning.
  • Swain (1985, 1988) and Sharwood Smith (1986) argue that processing for comprehension differs from that required for production and acquisition.
  • Cook (1991) argues that decoding language differs from 'code breaking', or discovering linguistic systems.
  • One output function is to force learners from semantic processing (comprehension) to syntactic processing (production).
  • Producing language forces learners to recognize what they don't know, triggering syntactic analysis of input or analysis of internal linguistic resources.
  • Cumming (1990) found that decision-making processes used by adult second language learners while composing may have potential for second language learning. This will return to later in the notes.

The Study

  • This study replicates Cumming's (1990) study, but with adolescent French immersion students instead of adult learners.
  • The study investigates if young learners identify problems while producing the target language and how they attempt to solve them.
  • It examines if reported solutions relate to second language learning processes.
  • The study explores whether learners engage in grammatical/syntactic analyses.
  • Research Questions:
    • Do young adolescent learners become aware of gaps in their linguistic knowledge while producing their L2?
    • If so, what cognitive processes are triggered that are involved in second language acquisition?
    • Do learners engage in grammatical/syntactic analyses to solve linguistic problems?

Methodology

  • Subjects: 18 grade 8 early French immersion students from a lower-middle to middle class background.
  • The students, whose average age was 13, had a wide range of academic abilities.
  • They had participated in an early French immersion program since kindergarten.
  • Immersion teachers use learner-centered approaches and a whole-language approach to language teaching which emphasizes process writing.
  • Instruction was entirely in French until grade 3; later, English was introduced for up to 50% of the day.
  • In grade 8, they studied history, geography, mathematics, and French language arts in French.
  • Formal grammar instruction is often an important feature of the French immersion classroom.
  • The students had been exposed to an eclectic second language teaching approach.
  • Of the 18 students, 9 were selected for data analysis.
  • The sample had the teacher rank students according to their overall proficiency in French and chose the top two, and bottom two, students.
  • 5 additional students were randomly selected from among the middle group.
  • The final sample consisted of 6 girls and 3 boys.
  • Procedure:
    • The wnting task was developed in consultation with the students' homeroom teacher.
    • The students were given a theme that they had covered in class.
    • Each student met individually with the researcher in a small, quiet room.
    • Procedures suggested by Ericsson and Simon (1993) for eliciting think-alouds were followed.
    • Students received a brief introduction in English outlining the task which they would be expected to do.
    • They were told to think aloud while writing a short article in French on an assigned topic, and that the researcher was interested in knowing what they were thinking as they were wnting.
    • The researcher provided an example of thinking aloud.
    • The specific wnting task: write an article as a reporter for your local newspaper about an environmental problem facing your community and offer solutions.
    • Students were told to write in French and think aloud in either French or English.
    • They were asked not to erase mistakes, but cross them out and continue.
    • If they stopped talking for very long, the researcher prompted them to think aloud.
    • Whenever students made a change without commenting on it, they were prompted to verbalize what they were thinking with the question 'what are you thinking9'.
    • Students could not use dictionaries or get help from the researcher.
    • Upon completion of a draft, students edited their work using a red pen to record any of their changes, and to think aloud while they were editing.
    • A tape-recorder was placed on the table where the student was working and the research sat nearby.
    • Each session lasted approximately one hour.

Language-Related Episodes

  • Definition:
    • A language-related episode (LRE) is a segment where a learner discusses a language problem encountered while writing and either solves it correctly or incorrectly.
    • It also includes instances where a learner solves a problem without explicitly identifying it.
    • Example 1: Student struggles with how to say 'kill all the fish', demonstrating a sophisticated understanding of transitivity.
    • Example 2: Student attempts to produce 'la destruction', influenced by both English and French, creating new knowledge through a search of existing knowledge.
    • Example 3: Student recognizes the need for 'arreter' to have a complement and inserts one, correcting an error common among immersion students.
  • Each episode relates to a problem in target language production, involving both gist (meaning) and language use.
  • Some episodes focused only on language use, particularly the written form, representing the consolidation of linguistic knowledge.
    • Example 4: Student corrects accent usage based on grammatical rule.
  • These episodes differ from those focused on generating ideas before composing.
    • Example 5: Student generates ideas about cutting down trees.
  • Each language-related episode dealt only with one linguistic item, even if episodes overlapped.
    • Example 6: Student grapples with tense, gender, and subject during the written passage.
  • Reliability: Four researchers independently identified and discussed language-related episodes, reaching consensus on the definition and identification process.

Classification

  • The data was classified into categories based on how learners solved linguistic difficulties.
  • The goal was to identify the mental processes reflected in the changes students made to their output.
  • The categories were data-dependent and related to cognitive processes in SLA literature.
  • Seven descriptive categories were identified:
    1. Sounds right/doesn't sound right:
      • Lexical: Student changes word because it doesn't sound right.
        • Example 7: Student changes 'cette' to 'cet' because it doesn't go very well'. cette + ozone
      • Grammatical: Student changes grammatical structure because it doesn't sound right.
        • Example 8: Student changes 'les droits uhm d'animaux' to 'les droits des animaux' because it sounds better.
    2. Makes more sense/doesn't make sense: Student adds or changes words to make the sentence more logical.
      • Example 9: Student adding 'les' to make the sentence more logical. Puis il reconstruit (Then it rebuilds) \Rightarrow Puis il les reconstruit (Then it rebuilds them)
    3. Applied a grammatical rule: Student uses grammatical knowledge to correct or modify output.
      • Example 10: Student changes 'du' to 'des' based on the plural noun.
        Porter du (To wear some) \Rightarrow Porter des vetements (some clothes).
    4. Lexical search: Student searches for the correct word.
      • Via English: Student use use English to try and find the word in mind.
        • Example 11: Student trying to translate the word can \Rightarrow Uhm, contenu
      • Via French: Student uses French to try and find the word in mind.
        • Example 13: Student changes 'qui sont' to 'qui creent' to better fit the context.
        • Via Both: Student uses both French and English to find the word in mind.
        • Example 13: Student combines English and French to find the correct word.
    5. Translation (phrase or greater): Attempting to translate a phrase from English to French.
      • Example 15: Student translates 'The problem of acid rain is generally becoming worse' into French.
    6. Stylistic: Adjusting the wording for style or formality.
      • Example 16: Student avoids using 'ca' in writing.
    7. Spelling: Focusing on correct spelling.
      • Example 17: Student checks the spelling of 'faible'.
  • Reliability: High agreement among researchers in classifying episodes.

Comparison with Cumming's Categories

  • Cumming (1990) categorized decision-making processes used by adult second language learners while composing.
  • His categories were:
    1. Searching out and assessing improved phrasing.
    2. Comparing cross-linguistic equivalents.
    3. Reasoning about linguistic choices.
  • Cumming's first category overlaps with the lexical search category.
  • His second category is incorporated within the lexical search category and translation.
  • Cumming's third category is reflected in applying a linguistic rule and stylistic categories.
  • The 'sounds right/doesn't sound right' and 'makes more sense/doesn't make sense' categories are specific to this study.
  • Younger learners may rely more on sound and 'sense' due to age and experiential learning.

Findings

  • Young adolescent second language learners are aware of gaps in their linguistic knowledge during L2 production.

  • They engage in thought processes that may play a role in second language learning.

  • This process occurs even without external feedback.

  • The identified thought processes align with those hypothesized by other theorists, like:

    • Extending first language knowledge to second language contexts,
    • Extending second language knowledge to new target-language contexts, and
    • Formulating and testing hypotheses about linguistic forms and functions.

  • Learners engage in grammatical analysis, essential for accurate production.

  • Faulty reasoning can lead to incorrect hypotheses, indicating that feedback could play a crucial role in advancing their second language learning.

  • Quantification of thoughts: 113 language-related episodes noted during the drafting phase and 77 during the editing phase, totaling 190 episodes.

  • These episodes represent occasions where students consciously recognized and addressed a linguistic problem.

  • Lexical searches were more frequent during drafting, while attention to sound, sense, and grammar increased during editing.

  • The table summarizes the distribution of language-related episodes across categories during drafting and editing phases.

  • The drafting phase showed more emphasis on lexical searches.

  • The editing phase showed more attention to whether things makes sense, sounds right, and whether it is grammatical.

  • The high-proficiency students produced more language-related think-aloud episodes than the least proficient students.

  • In the editing phase, these seemed to rely less on grammar.

  • The data reveals two observations:

    1. During the drafting phase, more language related think-aloud episodes are created by high-proficiency students than low-proficiency students.

    2. During the editing phase, high-proficiency students pay more grammatical attention than low-proficiency ones.

      • The conscious knowledge of rules is involved in great L2 accuracy.
      • Hawkins and Towell (1992) say that plausible Explanation is that the conscious knowledge of the rules facilitates in some fashion the development of subconscious knowledge.

  • Three general categories of cognitive processes were extracted:

    1. Generating alternatives.
    2. Assessing those alternatives.
    3. Applying the resulting knowledge (See Table 3 for examples).

  • Gaps in Table 3 relate to the task and attentional focus of students at their proficiency level.

  • Figure 1 illustrates second language learning from an output perspective, and that the communicative need engendered by the task forced the learners into thinking about the form of their linguistic output. That is, it moved learners from semantic to grammatical processing.

  • In short, under some conditions, learners notice a problem and conduct analyses leading to modified output, this is a part of the L2 learning process.