This is to certify that Sugandha Kaur has prepared the thesis entitled "Lexical Representation and Processing in Bodo−Assamese Bilingual" for the degree of Doctor of Philosophy at the Department of Humanities and Social Sciences, Indian Institute of Technology Guwahati, India. Special thanks goes to Deepangshu Goyal for writing the program for the experiments of the pilot study.
L IST OF T ABLES
Prime Type and Prime Language in Experiment 11A 308 Table 5.36 Mean RTs (ms) and Percentage of errors for related and non-. Prime Type and Prime Language in Experiment 11B 312 Table 5.38 Mean RTs (ms) and Percentage of errors for related and non-.
L IST OF F IGURES
Bilingual group and relationship type in experiment 6A 203 Figure 4.15 Mean RTs (ms) of the correct translation pairs as a function of. Bilingual group and relationship type in experiment 6B 211 Figure 4.16 Size of the interference effect (ms) in the four types.
A BBREVIATIONS
I NTRODUCTION
T HEORETICAL B ACKGROUND
- Introduction
- Chapter Overview
- Definition of Bilingualism
- Classifying Bilinguals
- Psycholinguistics and Bilingualism
- Models of Bilingual Language Representation and Processing
Most people in the world are able to speak two or more languages. In the current decade, interesting trends have appeared for the description of bilingual lexical organization with the application of the connectionist framework.
The Bilingual Interactive Activation (Plus) Model(s) (BIA and BIA+). The Bilingual Interactive Activation (BIA) model of lexical access
- Models of Speech Production
- The Language Mode Hypothesis
- Rationale for the Present Study
- Goals, Research Questions and Hypotheses
- Overview of Methodology
- Thesis Structure
- Chapter Summary
According to the assumptions of the selective view, lexical access in bilingual word production is fundamentally a selective process. On the other hand, lexical access in bilingual word production is fundamentally a non-selective process.
L ITERATURE R EVIEW
Introduction
This chapter consists of a review of the literature regarding some areas of bilingual lexical representation and processing that are considered controversial but deserve mention in the discussion of bilingual word recognition and production. These studies addressed the issue of bilingual lexical representation and processing using different paradigms in both word recognition and production, and the results emerging from the different experiments strongly suggest that the contradictory results are due to different factors.
Chapter Overview
The previous chapter had tried to introduce the research and give an overview of the structure of the thesis in general. The issues mentioned will appear in later discussions concerning proposed theoretical models of bilingual lexical representation and processing, and we will speak in terms of evidence in favor of (or against) these bilingual theories.
Research on Representation
However, evidence from some studies has reinforced the idea that the two languages of a bilingual are represented in two separate mental lexicons. The first evidence supporting the idea that there are no separate lexicons comes from the study by Cristonffanini, Kirsner and Milech (1986) in which the critical stimuli were related and unrelated translations. The results of the study showed priming effect only for cognates that was similar to within-language repetition priming.
The findings of this study showed that only relatives and homographs benefited from exposure in the first block. There are many models in the literature that assume separate lexicons for L1 and L2 and assume that these lexicons do not influence each other in performance. In addition, some earlier hierarchical models of word recognition, such as the word association model, the concept mediation model, and the revised hierarchical model, postulated separate independent lexicons for the two bilinguals, but, like Paradis, argued for the functional autonomy of L1 and L2 words.
Research on Processing
- Lexical Access: Selective or Non-selective?
- The Cognate Facilitation Effect
- The Influence of Age of Acquisition and Proficiency
- The Role of Script
- Asymmetry in Language Processing
- Processing Route
- The Word Type Effect
One of the leading models explaining language representation and processing in bilinguals is the Revised Hierarchical Model (RHM), which proposes that bilinguals have a language control mechanism (selective access). However, unrelated translation equivalents tend to show much less (or no) priming in the bilingual speakers (De Groot & Nas, 1991;). According to the morphological account, the cognate facilitation effect reflects an underlying structural difference in the way cognate and unrelated translations work. translation equivalents are represented in the bilingual lexicon.
In the following section, we address this question of whether differences in second language age of acquisition and proficiency may affect the representation and processing of language in bilinguals. In other words, the majority of previous studies have examined bilinguals who differ in their second language age of acquisition. For example, using a stimulus-onset asynchrony (SOA) of 200 ms, Altarriba (1992) observed only translation priming effects in the L1–L2 direction.
Research on Bilingualism in India
Chapter Summary
R ESEARCH D ESIGN AND M ETHODOLOGY
Introduction
Although the methods used in bilingual research borrow lavishly from available disciplines of study, namely psychology and others, the choice of methodology to be used will depend, however, on the central topic of the question the research aims to address (Wei & Moyer, 2008). Most bilingual studies by psycholinguists are experimental studies in which researchers study human participants in a laboratory setting and where the researchers can control exactly what information is presented to the subjects. Also, they can exchange information with similar or different participants in order to test specific hypotheses.
The next section describes step by step what was done in carrying out the research.
Chapter Overview
The Bodo−Assamese Bilinguals
- The Bodo−Assamese Contact Situation
- Linguistic Features of Bodo and Assamese
The Hinduization of the Bodo tribe, ever since the period of consolidation of Ahom rule in Assam and their own adoption of Hinduism, may have been a contributing factor to the strong Indo-Aryan influence on the Bodo speech ( Basumatary, 2015). A strong tendency of the Bodo people to switch to contact with the Indo-Aryan languages and a widespread and far-reaching bilingualism among the Bodo people had previously been reported (Grierson, 1903). It is the easternmost member of the Indo-Aryan branch of the Indo-European language family.
During the first quarter of the 20th century, the Bodo writers used Assamese script for the composition of books and periodicals. The evolution of the Assamese script can be divided into three different stages: Old Assamese script or the Kamrupi script (4th/5th to 13th century), the origin of which in turn could be traced back to the Gupta Brahmi script; Medieval Assamese script (13th to early part of 19th century) and Modern Assamese script from early part of nineteenth century. Next to grammar, vocabulary was the biggest contribution of the tribal groups to the formation of the Assamese language, among which the Bodo tribe stands out as the biggest contributor (Phangcho, 2006).
The Present Study
The Assamese language has been deeply influenced by various tribal languages in terms of grammar.
Phase One: Pilot Study
- Data Collection
- Method
- Data Analysis
- Results of Unmasked Translation Priming from L1−L2 (Experiment 1)
- Results of Unmasked Translation Priming from L2−L1 (Experiment 2)
- Results of Masked Translation Priming from L1−L2 (Experiment 3)
- Results of Masked Translation Priming from L2−L1 (Experiment 4)
- Findings
- Validity and Reliability
On the other hand, in a masked version of the translation priming paradigm, a series of symbols visually mask the translation prime. None of the unrelated prime–target word pairs were morphologically or semantically related (see Appendix F). On all participants, two parts of the study were conducted in the following order: the language background study and the priming experiments.
In contrast to some previous findings, the current experiment did produce a priming effect in the L2−L1, although this was smaller and not significant. This results in some information about the prime reaching consciousness, that is, processing of the target is facilitated by a briefly presented translation prime. One of the central questions of this study was whether the magnitude of the translation priming effect was symmetric in both directions (L1−L2 and L2−L1).
Phase Two: Main Fieldwork
- Data Collection
- Method
- Data Analysis
- Data Trimming
- Statistical Procedures
The questionnaire was first constructed in English and then translated into Bodo keeping in mind that the same was the mother tongue of the study participants. However, for the course of the study, the focus is on prime numbers that are written words or in picture form. However, as will be evident in the rest of the thesis, this issue is much more complex.
Group was used to examine the role of age of acquisition and proficiency of the bilingual. For the present study, a Sennheiser microphone was used to record the vocal response time of the participants in the production experiments. The following section aims to map the statistical techniques used during the study.
Chapter summary
C OMPREHENSION S TUDIES
W ORD R ECOGNITION
Introduction
Understanding what people say and write (ie, understanding language) is more complicated than it might first appear. An initial step in understanding any message is word recognition, and therefore researchers in cognitive science have a long history of answering questions about the nature of mental processes by examining word recognition. In the field of word recognition research, there has been a debate for several decades about exactly how cross-language effects relate to how words from different languages are stored and processed.
In short, most of the evidence on word recognition studies has argued against the idea of separate lexical representations and suggests that lexical access is non-selective with respect to language, that is, word representations of both languages become active in parallel during recognition.
Chapter Overview
Finally, we will discuss two translation recognition experiments to examine the nature of lexical-conceptual connections in the bilingual lexicon.
Visual Lexical Decision
- Translation Priming
- The Present Study
- Unmasked Translation Priming from L1−L2 (Experiment 1A)
- Unmasked Translation Priming from L2−L1 (Experiment 1B)
- Masked Translation Priming from L1−L2 (Experiment 1C)
- Masked Translation Priming from L2−L1 (Experiment 1D)
- General Discussion
- Semantic Priming
- The Present Study
- Unmasked Semantic Priming from L1−L2 (Experiment 2A)
- Unmasked Semantic Priming from L2−L1 (Experiment 2B)
- Masked Semantic Priming from L1−L2 (Experiment 2C)
- Masked Semantic Priming from L2−L1 (Experiment 2D)
- General Discussion
- Associative Priming
- The Present Study
- Unmasked Associative Priming from L1−L2 (Experiment 3A)
- Unmasked Associative Priming from L2−L1 (Experiment 3B)
- Masked Associative Priming from L1−L2 (Experiment 3C)
- Masked Associative Priming from L2−L1 (Experiment 3D)
- General Discussion
- Phonological Priming
- The Present Study
- Unmasked Phonological Priming from L1−L2 (Experiment 4A)
Word length (i.e., number of letters) for Assamese targets was matched for related and unrelated prime conditions. Results of the individual primary effects for the three groups of bilinguals showed a no. In the next section, we explore one of the most studied effects in psycholinguistics—the so-called
Planned comparisons conducted on the individual priming effects for the three groups of bilinguals revealed that the reaction times of the early high proficient bilinguals (740 ms) were the shortest compared to the late high proficient (816 ms) and late low proficient bilinguals (870 ms). In contrast to Experiments 2A and 2B, reaction times for the early high-proficiency bilinguals (889 ms) were longest compared to the late high-proficiency (824 ms) and the late-low proficiency bilinguals (730 ms). The results of our study are consistent with the predictions of the RHM—the pattern of cross-language semantic priming effect from L1-L2 was strong and consistent, but the effect from L2-L1 was null.
Similar to Experiment 3A, the reaction times of the high-early-proficiency bilinguals (957 ms) were significantly longer than the reaction times of the high-late-proficiency (762 ms) and low-late-proficiency bilinguals (681 ms). In the mixed-effects analysis of the error data, the main effect of Prime Type did not approach significance [F < 1]. The main effect of Similar Status was significant [F p = .000]. Participants recognized non-related targets more accurately than related targets.