Ammerlaan Tom

These pages contain
abstract , an outline in chapters and a list of some publications on first language attrition of speech processes in the Dutch mother tongue of emigrants to Australia in the fifties.
You can also read the Dutch version of the abstract.

T. Ammerlaan, 1996


Exploring bilingual lexical retrieval processes in the context of first language attrition

ABSTRACT: This dissertation examines the nature of variables influencing first language (L1) attrition. The study concentrates on the lexicon of long-term Dutch expatriates living in Victoria (Australia). It was investigated whether particular word types were more susceptible to language attrition than other types as suggested by emigrant Dutch. The aims were to determine whether subjects' ideas on having "forgotten their mother tongue" are empirically verifiable, which types of Dutch words tended to be difficult to remember and whether specific aspects of speech production played a role. In particular, the effect of various types of lexical similarity on access and retrieval processes was investigated. These effects were studied in a picture-naming experiment involving emigrants who no longer used their L1 in Australia.
Many long-term emigrants had, during interviews about the quality and use of Dutch in Australia, claimed that they had lost their native tongue as a result of disuse. Language shift to English is very widespread among the first-generation Dutch emigrants in Australia. Unlike other non-English communities in Australia, second generation Dutch-Australians tends to be monolingual speakers of English. Variables which influenced the reluctance in the emigrant population to use Dutch were investigated in a questionnaire.

To test whether first language knowledge had 'attrited' or whether retrieval processes in Dutch had become 'rusty' I devised a picture naming and recognition experiment. If the pictures could not be named, the pictures were represented again in an unexpected recognition session, now accompanied by 5 distracters and the target name. By comparing the results of the two sessions it was tested whether Dutch words were merely difficult to retrieve or also could no longer be recognised.
The pictures used in the experiment were taken from Snodgrass and Vanderwart's set, and pretested further before being digitized into images and pretested again. The pictures represented concrete words for familiar objects common to both Dutch and Australian cultures.
To test where in the speech processing stages problems occurred, the picture names varied in form similarity to English in several aspects: single-stem versus compound, and within the single-stem names, degrees of phonological and morphological similarity to the English picture names. It was attempted to keep the single-stem categories equal for frequency and word length in both languages. Examples of the single-stem word categories are: arm-arm, tafel-table, paard-horse, asperge-asparagus and buro-desk.
Within the compound words I looked at the effect of meaning transparency and stem-agreement. Examples of the multi-stem categories are: vingerhoedje- thimble, aardbei-strawberry, gieter-watering can and stinkdier-skunk.

Subjects were 76 Dutch emigrants who had been dobbed in by Dutch friends and relatives to participate because they were unaffiliated to Dutch societies and did not use the Dutch language.
Questionnaire data on the use and usage of Dutch and English showed Dutch was rarely used by these emigrants, even privately in endogamous marriages. Subjects had spent a minimal 6 years in the Netherlands prior to emigrating. If Dutch was used this was in its spoken form, usually involving listening. Dutch was avoided for a variety of reasons, some of which involved a negative appreciation of the quality of Dutch. Proficiency tests (Cloze, Editing and Fluency tests) showed greater fluency in English than in Dutch, and considerable evidence of being 'rusty' in their old L1.
Statistical analyses on the background data showed that the principle factor influencing the subjects' current level of proficiency was the age at which they had departed the Netherlands. Subsequent contact with Dutch proved to be a function of this variable.
No reaction time analysis could be conducted on the data: subjects were very slow and made many comments during the experiment.
The experimental data showed more errors with poorer levels of L1 proficiency, and differences in word category and task effects depending on proficiency.
Analyses of the error data suggested that lower proficient subjects had relatively more access problems because the proportion of recognition errors to naming errors was relatively high. High proficiency subjects had predominant retrieval problems as they tended to recognise all initially unrecalled names.
Error analysis between the word categories shows that morphologically distinct words are difficult to recall and recognise, and that phonological similarity helps. But words that are partially similar (phonologically similar but morphologically distinct: asperge-asparagus) cause more problems than the words that are distinct in both aspects (buro-desk). This effect depends, however, on the level of L1 proficiency: high proficiency subjects are least affected, poor proficient subjects are hindered by phonological similarity.

Analyses of the recall and recognition strategies and retrospective comments provide some insight into the reasons why cross-linguistic similarity to English as used in the recall of Dutch. Processing was very rusty in Dutch but extremely automated in English, and the code-mixing was considered communicatively appropriate in Australia. The latter could be a reason why dissimilarity did not have a cumulative effect in the single-stem categories.

The above gives a simplified picture of the study. In the project itself, a large range of linguistic, sociolinguistic and psycholinguistic aspects of first language attrition were investigated, far more than were warranted by the experiment (for details, see outline below). This was done primarily because at the time of testing multiculturalism was beginning to become more and more accepted by the Dutch emigrants. As a result of a relative revival of Dutch since 1988 the population of emigrants was expected to change in nature. A comprehensive investigation was considered a good starting point for a future investigation into the effects of the "ethnic revival" on the subjects. For further details about language attrition and bilingual processing I refer to the dissertation itself (ISBN 90-9008973-X).
Back to top

DISSERTATION OUTLINE IN CHAPTERS Chapter 1 is a theoretical exposition of the approach taken towards investigating lexical L1 attrition. An outline is provided of previous sociolinguistic (par. 1.2) and linguistic (par. 1.3) research on lexical variables in language attrition research in general. This approach may account for phenomena which appear to suggest selective attrition of Dutch in Australia. Integration of the various findings is provided through psychological research on forgetting (par. 1.4) that not only allows specification of the term `attrition' and the nature of important variables, but furthermore suggests that retrieval processes play a crucial role in selective attrition. It is therefore suggested that psycholinguistic research is better able to determine if the emigrants had not "lost" their Dutch as some had claimed or if their observations about selective lexical attrition could result from interference by similar Australian-English (AE) words which affects their access and retrieval processes in Dutch.

Chapter 2 provides a review of psycholinguistic studies on the nature of retrieval processes. After a brief characterization of the language processes identified in monolingual research and the test paradigms used (par. 2.2), a review is given of research on what is known about bilingual lexical processing (par. 2.3). Specific attention is given to bilingual speech production, particularly bilingual lexical retrieval processes. Insights about these processes are derived from research on bilingual code-mixing and processing and interpreted through Levelt's framework of monolingual speech processing (1989). Findings from various types of bilingual studies on the effects of types of cross-linguistic similarity are used to predict how this characteristic affects the retention of types of words. Such research on lexical processing suggests that cross-linguistic interference depended on the perceived nature (e.g. similarity, complexity) of the items, the type of task and the linguistic background of the subjects (e.g. learning context, fluency, proficiency). These variables influence how a word is accessed and retrieved, and thus whether it can be successfully recalled or recognised with or without the influence of Australian-English.

Chapter 3 lists the implications of the literature reviews in Chapters 1 and 2 for an investigation of the "loss of Dutch" and specifies the paradigms of the project on Dutch-Australian emigrants (par. 3.2). The impact of cross-linguistic similarity in L1 attrition was investigated in an experiment involving emigrants varying in background variables. To calculate the effect of similarity on lexical attrition within this group of Dutch native speakers in Australia, the experiment was complemented by tasks that assessed the biolinguistic background information of the Dutch emigrants. These tasks are described in par. 3.3.

Chapter 4 outlines the linguistic situation of these emigrants as based on a questionnaire that established when, where, and why they used which language (par. 4.1) and on objective and self-assessment of their proficiency and fluency (par. 4.2). Interrelations between the various data types are given in par. 4.3.

The picture-naming and identification experiment is described in Chapter 5. This experiment investigated whether certain types of Dutch words were more easily retrieved than others by measuring the effect on bilingual processing of variables such as word length, frequency of occurrence, and form-similarity in conjunction with various biolinguistic background characteristics. More specifically, the experiment aimed at determining the relationship between bilingual recall and recognition of certain Dutch words ranging in similarity in lemma and lexeme to their English translation equivalents. Whether retrieval processes had affected non-cognates more than cognates, low-frequency words more than high-frequency words, multi-stem more than single-stem words, and morphologically dissimilar words more than morphologically similar, and phonologically dissimilar words more than phonologically similar words was investigated by comparing naming and identification sessions of a picture-naming experiment in Dutch. If subjects were unable to provide the target name, they were subsequently presented with the same picture accompanied by six written names. If subjects were unable to identify the L1 correct name, this suggested "loss". Chapter 5 also provides the results of a pretest of the experiment conducted by monolinguals.

The findings of the bilingual experiment are discussed in Chapters 6 and 7. Chapter 6 gives the results of the experiment on the so-called `dormant' Dutch-Australian bilinguals, starting with the strategies used (par. 6.2). Determined next is the measure to be used for proficiency effects (par. 6.3). This is followed by analyses of the error distribution in the naming and identification sessions (par. 6.4) and the distracters used (par. 6.5) in conjunction with the extent of the emigrants' residual proficiency in Dutch.

Chapter 7 summarizes the various findings and discusses their relevance in combination to earlier research on L1 attrition (par. 7.2). In addition to discussing the implications of the reported research for bilingual lexical processing, this chapter offers suggestions for further study of L1 attrition and Strutch (par. 7.3) from various perspectives. A list of the abbreviations, the references used, and the appendices to each chapter are given in the final part of this dissertation.

"You get a bit Wobbly", PhD thesis, Dept. of Appied Linguistics, 1996, CopyPrint 2000, Enschede. Back to top

A selection of related publications: Ammerlaan, A. (1984a). , A process-oriented approach to lexical strategies in interlanguage communication. Paper presented at AILA - 7th World Congress of Applied Linguistics, Brussels: Vrije Universiteit.

Ammerlaan, A. (1984b). , A process-oriented approach to strategies in referential communication. MA Thesis, Dept. of English, Nijmegen: KUN.

Ammerlaan A. (1985a). , Testing lexical retrieval in speech production of dormant bilinguals. Paper presented at the ALS Annual Conference, Brisbane: Griffith University.

Ammerlaan A. (1985b). , Exploring lexical retrieval in bilingual speech production. Paper presented at the 5th Australian Language and Speech Conference, Melbourne: Monash University.

Ammerlaan, A. (1987a). , Lexical retrieval in dormant bilinguals. Paper presented at AILA - 8th World Congress of Applied Linguistics, Sydney: The University of Sydney.

Ammerlaan, A. (1987b). , Attrition and interlanguage development. Paper presented at the Explaining Interlanguage Development Workshop, Melbourne: La Trobe University.

Ammerlaan A. (1987c). , Strategies employed by dormant Dutch bilinguals. Paper presented at The Dutch Community in Australia: Victims of Assimilationist Policies, Melbourne: Monash University.

Ammerlaan, A. (1989). , Dutchman down-under. Dutch Courier, May 1989 6. Reprint of 'First impressions', Paper presented at the 'Dutch Roots' seminar, April 9, 1989, the Associated Netherlands Societies in Victoria, Melbourne: Abel Tasman Club.

Ammerlaan, A. (1989). , The teaching of the Dutch language. Dutch Courier, 13 Nov. 8.

Ammerlaan, A. (1990a). , Linguistic variables in L1 attrition. Paper presented at the Combined ALS & ALAA Conference, September 1990, Sydney: Macquarie University.

Ammerlaan, A. (1990b). , Reasons for language loss amongst the Dutch I. Dutch Courier, October issue, 12-13.

Ammerlaan, A. (1990c). , Reasons for language loss amongst the Dutch II. Dutch Courier, November issue, 16-17.

Ammerlaan, A. (1990d). , Levelt's model and explanations in language acquisition. Impromptu paper presented at the Combined ALS & ALAA Conference, September 1990, Sydney: Macquarie University.

Ammerlaan, A. (1991). , Testing linguistic variables in first language attrition. in F. Lovejoy (Ed.), Studies in language loss: Theoretical, intragenerational, pathological, Sydney: Holborn Press, 20-41.

Ammerlaan, A. (1992). , Lexical retrieval and first language attrition. Poster session at the Second International Conference on Maintenance and Loss of Minority Languages, September 1992, Noordwijkerhout.

Ammerlaan, A. (1993a). , Bilingual lexical retrieval in dormant Dutch bilinguals (Australia). Centre for Pacific Studies, Oceania Newsletter 11-12, 41-44.

Ammerlaan, A. (1993b). , Taalverlies onder Nederlandse emigranten. [Language loss among Dutch emigrants] Newsletter Australia 3.3, 3-5.

Ammerlaan, A. (1994a). , Problemen van Nederlandse Emigrantenvrouwen Down-Under. [Problems of Dutch emigrant wives down-under] Raffia 6.1, 25-26.

Ammerlaan, A. (1994b). , Oudere Nederlanders in Australie. Aanpak Vereenzaming vergt Erkenning als Minderheid. [Elderly Dutch people in Australia. Tackling loneliness requires recognition] LeefTijd 6., 27-31.

Ammerlaan, A. (1994c). , Oudere Nederlanders in Australie. [Elderly Dutch people in Australia] Tijdschriften Overzicht '94 6/7, 15-16.

Ammerlaan, A. (1994d). , The First Dutch Australian Community Conference. Newsletter Australia 4.2, 4-8.

Ammerlaan, A. (1994e). , Terug naar Holland om te sterven. [Returning to the Netherlands to die] Dutch Courier, June, 31.

Ammerlaan, A. (1994f). , Reacties op reacties van artikelen over oudere Nederlanders in Australie. [Reaction to responses to articles about elderly Dutch people in Australia] Dutch Courier, September, 2-3.

Ammerlaan, A. (1994h). , Foute Informatie Wekt Verontrusting. Leeftijd 9, 19-20.

Ammerlaan, A. (1995a). , Luisteren naar Nederlandse tijdschriften en banden. [Listening to Dutch magazines and tapes] Dutch Courier, Feb., 8.

Ammerlaan, A. (1995b). , Naturalisatie. [Naturalisation] Dutch Courier, Feb., 11.

Ammerlaan, A. (1995c). , Aspects of reactivation of a lost mother tongue. in B. Gruter & J. Stracke, (Eds.), Dutch Australians taking stock. Melbourne: Maroondah publishing, .

Ammerlaan, A., Hoeks, J., Van Horick T. & Bentlage, A., (1982). , Ik kan het niet beschrijven. [I cannot describe it] MA Thesis, ITT, Nijmegen: KUN.

Back to top

General Information on AUSTRALIA: - Australian Web Sites

- Lonely Planet travel information

- Pictures on Australia .

GO BACK TO Tom's Homepage