Transfer Types In Second Language Acquisition Of English

Table of Contents

1 Introduction

2 Theoretical Background
2.1 Transfer
2.2 Negative and positive transfer, avoidance and over-use

3 Investigated Languages

4 Transfer Types in the Studies

5 Discussion

6 Conclusion

Reference Section

1 Introduction

Language transfer constitutes one of the main research areas of second language acquisition. The process of learning a second language (L2) automatically includes the application of pre-existing knowledge of linguistic structures and rules from the first language (L1). Since every language is governed by its own rules and regularities, each L1 presents an individual basis for new languages to be built on. However, some aspects seem to be shared between languages, and certain behavioral patterns can be witnessed across a variety of L1s.

Generally, the idea is that languages of distant origins, such as English and many Asian languages, do not share many features and structures as they did not develop from the same roots. The present paper aims to analyze the transfer that occurs especially in those language combinations. Three studies are examined regarding the L1s of the subjects and the transfer types that can be observed. The main research question is the following: What types of transfer can be observed with English as a second language (ESL) learners whose L1s are not closely related to English as the target language?

Another point of research will be whether learners with Asian L1s only exhibit negative transfer since the L1 and the L2 are of rather distant origin. Or whether one can perhaps still find positive transfer, even though the languages do not seem to have too much in common.

Due to the great distance between languages from the Asian language region and English, one could expect to find only negative transfer, where the L1 knowledge interferes with the process of acquisition and produces ungrammaticality, or ‘errors’.

It will furthermore be asked whether the Spanish speaking ESL learners on the other hand exhibit more positive transfer since their L1 is more closely related to English, or whether they still also exhibit negative transfer.

To allow a detailed investigation of the topic, a definition of transfer will be given first, before taking a closer look at the languages investigated in the individual studies, which will be followed by pointing out the different types of transfer that can be found in the research.

2 Theoretical Background

2.1 Transfer

In recent years, language transfer has become an area of increasing interest for research in second language acquisition (SLA), since the key aspect of learning a second language is the fact that at least one other language has previously been acquired, the L1 or mother tongue, and that language can provide a basis for acquiring the rules and structures of a new language. The theory behind linguistic transfer was first developed in the field of behaviorist psychology in the 1930s. It has since been subject to discussion, undergoing changes concerning its definition and the aspects it covers. In 1989, Odlin formulated a working definition that has since been widely accepted in the field and holds as the basis of the terminology used in this paper:

Transfer is the influence resulting from the similarities and the differences between the target language and any other languages that have been previously (and perhaps imperfectly) acquired. (Odlin 1989, in Ellis 1994: 301)

In other words: Transfer – or “crosslinguistic reference”, as Sharwood Smith and Kellerman (1986, in Ellis 1994: 301) suggest as a more specific term – describes the process of applying knowledge from one language to another.

The most common process is transfer from a person’s L1 (or other L2s that they have [partially] acquired) to the target language, but transfer can in fact be bidirectional, meaning that an L2 can always also influence the L1 (or other L2s) (Kwon & Han 2008: 306). This development can be a conscious process; however, usually it occurs subconsciously. This lack of awareness on the learner’s side can make the presence of transfer exceptionally hard to detect, as it sometimes might present itself in an apparent manner, but usually it is too closely interconnected with other developmental factors, especially in child SLA, as stated by Kwon and Han (2008: 306).

1.1 Negative and positive transfer, avoidance and over-use

For a long time, the only aspect investigated was negative transfer, often also called errors. Negative transfer occurs when learners unsuccessfully apply structures from their L1 to the target language even though the latter does not have corresponding structures, thus creating ungrammatical expressions. Negative transfer interferes with language learning and can therefore cause learning difficulties and slow down the L2 development (Ellis 1994: 300).

However, transfer can also be successful, videlicet when elements are transferred that do exist in the target language. This is called positive transfer or facilitation, as the existing knowledge of the L1 can facilitate and even accelerate the process of L2 acquisition. Ellis (1994: 303) points out that:

Facilitation is evident not so much in the total absence of certain errors […] but rather in a reduced number of errors and, also, in the rate of learning. Since positive transfer is often hard to detect, most studies are concerned with the effects of negative transfer.

Furthermore, Ellis (1994: 302 and 304-306) suggests two additional types: Avoidance, when learners “avoid using linguistic structures which they find difficult because of differences between their native language and the target language” (Elllis 1994: 304), and the opposite, over-use, which can occur as the result of avoidance. In that case, because learners avoid certain structures that they do not feel comfortable with, they overgeneralize other structures instead.

However, as these two types are even harder to detect and appear far less frequently in research, the present paper will focus on negative and positive transfer as they occur in the investigated studies.

3 Investigated Languages

A popular approach in ESL research is comparing L2 learners with L1s of varying proximity to English. Languages from the Asian language region are of special interest, since they stem from entirely different linguistic backgrounds than English. Therefore, one would assume that the languages do not have much in common that could be used for transfer. In many cases, such as Japanese or Korean, this idea proves to be true, as can be seen in the studies investigated in the present paper (Helms-Park 2003, Kondo 2005, Whong-Barr & Schwartz 2002). Nonetheless, two of the studies found at least some aspects of positive transfer where analogous features occurred despite the linguistic distance (Helms-Park 2003, Whong-Barr & Schwartz 2002).

However, even languages that go back to the same language family and therefore share some basic features, such as English and Spanish that both originate in the Indo-European family, might still differ in important areas. Precisely these differences are likely to cause difficulties in the process of second language acquisition that become apparent as negative transfer. What makes Spanish so interesting in comparison to Asian languages in ESL research is that it is both close (due to similar roots in the Indo-European language family) and distant (due to different development and influences over many centuries) to English at the same time. Due to this fact, L1 Spanish ESL learners are likely to experience both negative and positive transfer in the same contexts (Kondo 2005).

Results of transfer can best be demonstrated by comparing learners of different L1s, preferably L1s that exhibit varying degrees of similarity in comparison to English. These discrepancies create distinct starting points for L2 learners of different L1s, and it is exactly these differing areas where divergence in transfer is to be expected (Whong-Barr & Schwartz 2002: 593). Otherwise, the results could also occur due to developmental, intralingual or other cognitive processes (Ellis 1994: 302-303).

The first study by Helms-Park (2003) picks up this idea of comparing two non-related L1s, namely how Vietnamese ESL learners use serial verb constructions in English compared to Hindi-Urdu ESL learners. Vietnamese, as a language with roots far from English, exhibits structures that cannot be found in English. For instance, in Vietnamese, several verbs that describe the same action can be used in one serial verb construction:

(1) Tôi ném cốc vỡ«.

I throw glass break

“I threw the glass and it broke.”

(Mikami 1981: 115 in Helms-Park 2003: 216)

However, in Hindi-Urdu, as well as in English, verb serialization does not exist (Helms-Park 2003: 211). Hence, the Hindi-Urdu ESL learners served as a neutral control group to compare the results of the Vietnamese group to. Consequently, Helms-Park postulated that Hindi-Urdu speaking ESL learners will not be creating any serial verb constructions, as they neither know them from their L1 nor will learn them through input. Vietnamese learners on the other hand are most likely going to use verb serialization even though it is not an accepted form in English (Helms-Park 2003: 222).

Secondly, Kondo (2005) compared Japanese and Spanish ESL learners regarding their overpassivization errors of unaccusative verbs. L2 learners often use these verbs in passive voice, even if that is not correct:

L2 learners of English of some L1s overgeneralize the morphosyntactic reflex of the passive be + en to unaccusative verbs like break when used intransitively and also to unaccusatives without transitive counterparts like happen:

(13) a. * The window was broken when the ball hit it.

(for the intended ‘The window broke when the ball hit it’).

b. * Why was this happened?

(for the intended ‘Why did this happen?)

(Kondo 2005: 135)

Kondo investigates these two languages, since unlike English, they both use morphological marking of unaccusatives, but with different structures. It is therefore assumed that learners of both L1s will have difficulties when it comes to unaccusative verb structures in English. Due to the different knowledge of linguistic structures and verb forms provided by the two L1s, however, the learners should behave differently, exhibiting diverging behavior, distinct both from each other and from the native speaker control group (Kondo 2005: 143).

Likewise, the last study by Whong-Barr and Schwartz (2002) on English dative alternations presents a similar reasoning. English distinguishes two types of datives: to -datives and for -datives:

It is well known that there are restrictions on the verbs that can enter into the da- tive alternation – for example, you can show the results to someone and show some one the results; and you can demonstrate the results to someone but you cannot * demonstrate someone the results.

(Whong-Barr & Schwartz 2002: 579-580)

The study investigates the behavior of Japanese and Korean native speakers, whether they overgeneralize the double-object version like L1 English children do or whether they are able to transfer grammar from their mother tongues into English and use the forms correctly (Whong-Barr & Schwartz 2002: 595). The latter should only happen to the Korean children since Korean allows double-accusative constructions for analogues of for -datives whereas Japanese does not. Neither of the two languages allow the structures for to -datives, which suggests that both groups will acquire the structure like L1 English children, initially disallowing it and later overgeneralizing it (Whong-Barr & Schwartz 2002: 594-595).

[...]