Published undergraduate dissertation. Universita` di Bologna

Fluency in native and non-native speakers

Introduction

Who is a fluent speaker? What characteristics should he display to be considered as such? For a start, let us examine two general definitions in the Collins Cobuild Dictionary (1995): 1) Someone whose speech, reading, or writing is fluent speaks, reads, or writes easily, smoothly and clearly with no hesitation or mistakes. 2) Someone who is fluent in a particular language, or who speaks fluent Spanish, French, Russian, etc can speak or write the language easily and correctly, with no hesitation or inaccuracy.

The first definition suggests that fluency applies virtually to all language abilities, ranging across mode (spoken/written) and domains of processing involved (productive/receptive). The second definition is somewhat narrower, as it focuses on the particular ability to speak/write a second language. Like the first, however, it encompasses different linguistic skills that may require very different scale criteria for assessment. For example, measuring hesitation can be useful in speech but is hardly meaningful for written language, where other features will be given prominence (e.g. smooth phrasing, wide vocabulary usage). In addition, it should be noted that -according to the above definitions- the characteristics that make a person fluent do not necessarily go hand in hand: one can carry out a task `easily` and `smoothly` and at the same time be pretty inaccurate. What these remarks suggest is that a linguist setting out to devise operational criteria for assessing fluency meets with not inconsiderable problems. On the one hand, fluency features need to be amenable to testing practices and procedures, and as such must be restricted. On the other, a study on fluency has to capture the essence of fluent behaviour that the dictionary definitions imply, namely the ability to actually do something rather than just knowing something. In a nutshell, the focus has to be on skills, or the capacity to put knowledge into practice.

Chapter 1 presents the cognitive approach to fluency that has been taken in this study and discusses a number of theories drawn from the fields of experimental psychology and linguistics that account for the mechanisms underlying fluent behaviour and its development.

Chapter 2 moves on to establish a small array of experimental parameters to proceduralise the notion of fluency in the light of the psycholinguistics literature on hesitation phenomena.

Finally, having provided a theoretical and experimental paradigm a case study is presented in chapter 3, examining a group of eight adult speakers who are supposedly balanced bilinguals. The aim of the enquiry is to investigate the ultimate attainment of these speakers in terms of fluency in their L1 and L2 in a particular recall task. There are good reasons for doing so: firstly, Second Language Acquisition research has traditionally been more concerned with lower levels of linguistic ability, at the expense of investigating near-natives. Secondly, it is theoretically stimulating to understand what degree of success people can ultimately achieve in specific cognitive operations. Recent evidence suggests that "second language acquisition often - presumably in the majority of cases - leads to a state of ultimate attainment which is unlike that of native speakers of the target language" (Hyltenstam 1992: 351). This study specifically investigates this idea.

Finally, chapter 5 discusses the results spelled out in Chapter 4 in the light of the theoretical conceptualisations originally put forward and suggests further research.

A terminological remark should be made at the outset. This study is concerned with providing theoretical support and defining operative criteria for the evaluation of fluency in cross-linguistic settings. In order to do so, it makes use of the terms native and non-native speaker. It must be stressed that no quantitative nor qualitative distinction is a priori alluded to or implied by such a dichotomy. Native simply means that a speaker was grown and educated with one language (L1). Conversely, non-native means that a speaker has learned a second language (L2) in adolescence or later. A balanced bilingual is defined as a person whose ability in speaking is reputed to be virtually as good in his/her native as his/her non-native language in all domains of language use (Segalowitz 1986). In theory, s/he may also be better in L2, but, by definition, will remain a non-native speaker. I am aware that issues in bilingualism are controversial and that these distinctions are not watertight. Many people are bilinguals from early childhood; in some cases, the addition of a second language leads to gradual erosion of competence in the first language (Ellis 1994: 694). Indeed, the truly balanced bilingual may be nearly nonexistent (Kilborn 1994: 934) and there seems to be a continuum of degrees of nativeness in the outcomes of second language acquisition (Hylthenstam 1992: 351). However, our aim is not to have the last word in these matters, and the study is restricted to late L2 learners.