On acquisition, age and articles in multilingual Switzerland

24
Pfenninger, S.E. (2013) On acquisition, age and articles in multilingual Switzerland. In J. Mihaljević Djigunović and M. Medved Krajnović (eds) UZRT 2012: Empirical Studies in English Applied Linguistics (pp. 22–35). Zagreb: FF Press. On Acquisition, Age and Articles in Multilingual Switzerland Simone E. Pfenninger English Department, University of Zurich, Switzerland [email protected] Introduction The complex semantic conditions governing English article usage are one of typical examples of the difficulties that might arise for a learner when making the necessary form-meaning mappings underlying second language (L2) learning. This paper explores whether an earlier age of onset and consequently a longer period of instruction make a difference in the extent of the learning of the English article system by Swiss German learners. English and Swiss German have an almost identical article system; however, even though Swiss German learners need not internalize new syntactic rules, the learners are known to struggle with the few minor exceptions that constitute the difference between the L1 and the L2 article system. This study will support the argument that the learning of the English article system represents a lexico-grammatical (rather than a purely grammatical) requirement for Swiss German students and that the learners learn articles best as lexical items in

Transcript of On acquisition, age and articles in multilingual Switzerland

 

Pfenninger, S.E. (2013) On acquisition, age and articles in multilingual Switzerland. In J. Mihaljević Djigunović and M. Medved Krajnović (eds) UZRT 2012: Empirical Studies in English Applied Linguistics (pp. 22–35). Zagreb: FF Press.

On Acquisition, Age and Articles in Multilingual Switzerland

Simone E. Pfenninger

English Department, University of Zurich, Switzerland

[email protected]

Introduction

The complex semantic conditions governing English article usage are one of typical

examples of the difficulties that might arise for a learner when making the necessary

form-meaning mappings underlying second language (L2) learning. This paper

explores whether an earlier age of onset and consequently a longer period of

instruction make a difference in the extent of the learning of the English article

system by Swiss German learners. English and Swiss German have an almost

identical article system; however, even though Swiss German learners need not

internalize new syntactic rules, the learners are known to struggle with the few minor

exceptions that constitute the difference between the L1 and the L2 article system.

This study will support the argument that the learning of the English article system

represents a lexico-grammatical (rather than a purely grammatical) requirement for

Swiss German students and that the learners learn articles best as lexical items in

 

context (cf. Hakuta, 1976; Master, 1997). The question arises, then, as to whether

learners with an earlier onset of L2 learning, such as students whose L2 instruction

begins in primary school (early classroom learners or ECLs), are at an advantage in

this area of lexico-grammar compared to late classroom learners (henceforth LCLs),

considering the ECLs’ advance in vocabulary learning and rote-learning of

prefabricated speech, which represent the main focus of attention in the L2 primary

school classroom in Switzerland. Interestingly, while there is plenty of evidence of

the misuse or omission of articles by learners whose L1 contains formal equivalents

of the English articles, or whose L1 lacks them (e.g. Zdorenko and Paradis, 2008),

there are few studies to date relating to the use/non-use of articles by learners of

different age groups who share the same L1. As will be shown in this study, analyzing

the acquisition of the English article system by different-aged multilingual learners

can foster a deeper understanding of some intricate issues in the learning of an L2

system that displays considerable variation.

Background

On the difficulty of L2 article acquisition

It is widely reported in the literature, particularly in a great number of naturalistic

studies, that articles are by far the most frequently occurring items in the ESL/EFL

input, and even though they are particularly salient compared to other morphemes,

article accuracy is known to be notoriously slow in developing. Learners’ difficulties

in this area of morphology are traditionally described from a contrastive perspective,

focusing on cross-linguistic differences, and/or a developmental perspective, focusing

 

on the linguistic development. From a cross-linguistic perspective, the overuse or

omission of an article is ascribed to L1 interference, which numerous researchers have

found to play a significant role in L2 learners’ acquisition of English articles (e.g.

Hakuta, 1976; Liu and Gleason, 2002; Zdorenko and Paradis, 2008). From a

developmental perspective, the learners’ difficulties are attributed to (a) the

phonological characteristics of articles – they tend to become phonologically fused

with surrounding material as “their frequent production leads to lenition processes

resulting in the loss and erosion of gestures” (Ellis, 2006, p. 170) – and (b) the

inherent featural complexity and multiple functions of the articles, which are difficult

for L2 learners (irrespective of their L1 background) and L1 acquirers (see Thomas,

1989). For instance, several studies (e.g. Hawkins et al., 2006; Ionin and Wexler,

2003) report that the definite article is commonly acquired before the indefinite article

because “definite articles in English need not take number and the count/mass

distinction into account” (Lardiere, 2004, p. 335).

In view of the fact that this study is going to test multilingual learners, it has to be

mentioned that recent works on the influence of previous knowledge of foreign

languages within a cognitive approach (e.g. Williams and Lovatt, 2003) find that

learners who know many languages are more likely to exploit similarities between the

items of new foreign languages and features in other languages that they know. For

instance, knowledge of other article systems could provide an abstract understanding

of the way to make arbitrary article-noun associations (see Nayak et al., 1990).

It also has to be mentioned (e.g. Ellis, 2006; Goldschneider and DeKeyser, 2001;

Pfenninger, 2012) that external factors, such as formal instruction and learning

 

settings, influence the developmental processes related to the acquisition of morpho-

phonological forms in L2 acquisition. It has been hypothesized that explicit

comparison processes might be required for the mediated form-meaning associations

to be learned: implicit learning processes are often doubted to be effective in

instructed contexts, mainly due to a lack of massive input, a focus on fluency at the

expense of accuracy, and/or cognitive disadvantages of learners beyond early

childhood (e.g. de Graaff and Housen, 2009; DeKeyser, 2000; Ellis and Larsen-

Freeman, 2006).

Age effects on the learning of the L2 article system

Many naturalistic studies (e.g. McDonald, 2006, 2008) found articles to be among the

structures that correlate highly with age of onset of L2 acquisition. In psycholinguistic

accounts, this has been explained in terms of maturational constraints that have been

linked to resource limitations that might lead to the inability (a) to store, access and

retrieve L2 knowledge (semantically-related difficulties) and/or (b) to detect

phonological discriminations in the input (phonologically-related difficulties) (e.g.

McDonald, 2006, 2008). Furthermore, there seems to be a general agreement in

naturalistic studies that late learners are particularly affected by L1 interference

(Zdorenko and Paradis, 2008). It has to be kept in mind, however, that some

researchers (e.g. Zdorenko and Paradis, 2008, p. 232) suggest that only if children

master the L2 determiner system in less than five years of exposure (in a naturalistic

setting), such an outcome would contrast with the existing findings for adults.

Similarly, in the few classroom studies (see Ekiert, 2004; Master, 1997) that exist on

 

this topic, there is the consensus view that there should be no advantage for an earlier

start as far as L2 morpho-syntax is concerned.

Differences in English and Swiss German article semantics

Both English and Standard German are languages with an article system, the use of

the articles being determined by the semantic function of the NP in discourse.

However, there are some “minor” discrepancies that are known to cause problems

even for advanced learners, illustrated by the following transfer errors:

(a) There are discrepancies in the perception of noun countability

(count/noncount) or noun class (concrete/abstract).

(1) They fights until to the Death! (Allan, late starter, written narrative

essay)

(2) They became food but: What a food!!! (Serena, early starter,

written narrative essay)

(b) There are discrepancies in the perception of the lexical category of the

noun.

(3) But if the people don’t like the winners, I think that’s Ø pitty.

(Gloria, late starter, written argumentative essay)

(c) There are discrepancies in the use of L1 and L2 articles as they occur

in specific contexts with specific vocabulary (collocations and

idiomatic expressions).

 

(4) At home she makes the breakfast or the lunch. (Jenny, late starter,

written narrative essay)

(5) But the Swiss television shoots the bird. (Peter, late starter, written

argumentative essay) [from (Swiss) German den Vogel

abschiessen ‘take the cake’]

The examples in (a) and (b) refer to rule-governed uses of articles, where the learners

must come to realize that the choice of articles depends upon the nature of the

accompanying noun. As concerns (a), Yoon (1993) suggests that if the L1 and the L2

article systems differ in this way, this indicates the existence of differing perceptions

of nouns by the speakers and therefore proposes that the systematicity in the article

system is governed by a misperception of countability. Another problem is that most

English nouns can be used in either countable or uncountable ways depending on the

context (Goto Butler, 2002). Since several researchers (e.g. Goto Butler, 2002;

Klages-Kubitzki, 1995) suggest that the re-conceptualization of NPs might take

substantial practice, the ECLs in this study could thus be at an advantage due to their

extended period of L2 learning. Concerning point (b), Parrish (1987) found that

systematicity in the article system is also governed by the lexical categories of NPs

and attempts to keep linguistically related forms consistent with one another. Finally,

(c) exemplifies memory-based uses, where the learners have to form associations

between familiar items that otherwise do not occur together in the input, i.e., article

use in fixed multi-word sequences that are memorized as whole chunks; they can take

the form of collocational prepositional phrases, such as make breakfast in (4), as well

as idioms, as in (5). While a failure to learn idiomatic expressions might not be

 

specific to L1 Swiss German learners of English, it is nevertheless relevant to a

discussion of the impact of different ages of onset and length of instruction.

The present study

The following research questions (RQs) are addressed in the present study with 200

Swiss German middle school students (Grade 7) as participants:

(1) Are accuracy rates higher with the in definite contexts than they are with

a in indefinite contexts or the zero article in zero contexts? (RQ1)

(2) Are the accuracy scores significantly different for the ECLs than the

LCLs after 6 months at middle school? (RQ2)

(3) What are the most common error types? Are there any group-specific

patterns? (RQ3)

Concerning RQ1, it is anticipated that the use of the is more accurate than the use of

a, as pointed out earlier. As concerns RQ2, if the LCLs perform significantly better

than the ECLs, it follows that a 5-year difference in the amount of input is not

sufficient to make a difference in the extent of learning an L2 system that displays as

much variation as the English article system. As described in RQ3, the present study

also searches for group-specific patterns caused by the linguistic context of the three

articles, i.e., it will be analyzed whether there is a difference in the error types

produced by the two learner groups.

Participants

 

100 ECLs (52 females and 48 males) and 100 LCLs (51 females and 49 males) of

Grade 7 participated in this study, six months after they began middle school. They

differed in age of onset of acquisition (ECLs: 8; LCLs: 13) and length of instruction

(ECLs: 5;6, LCLs: 0;6), but they had the same biological age at testing (13;6), which

ensured that neither group was at a cognitive advantage. In primary school, the ECLs

received on average 90 minutes of Early English per week.

The learners speak the Zurich standard dialect, which is one of the largest in

Switzerland. Besides Standard German as the first L2, all of the participants had had

French as a school subject for 2.5 years, with two years in primary school (two 45-

minute classes per week) and six months in middle school (three 45-minute classes

per week). This means that for the ECLs, English represents the second foreign

language (or L3) to be learned at school, while for the LCLs, it is the third foreign

language (or L4). Since the participants all attended the same school in the state

system, we can safely assume that the type and amount of English input the groups

received did not differ significantly. The classes of the participants were intact but not

mixed, since at the school where the learners were tested, ECLs and LCLs did not

come together in the same L2 class. However, the two groups were instructed by the

same teachers and followed the same L2 curriculum.

The control group included 20 native speakers that attended the same school. This

control group was not included to compare the learners with native speakers (see

Larson-Hall, 2008 for a discussion of this), but to ensure the validity of the tests.

 

As concerns the teaching/learning of the L2 article system in Swiss classrooms, some

rule-governed uses of articles are explicitly taught in the initial stages of acquisition in

secondary school, that is, the use of articles with count vs. uncount nouns (singular

and plural) and the use of the allomorph an before nouns with word-initial vowels.

However, there is often little classroom time devoted to discussing and practicing

them. The learners are usually not taught explicitly which articles to combine with

concrete vs. abstract nouns; the expectation is that they will learn the suppliance rule

inductively. Memory-based uses, that is, idiomatic expressions and collocations, are

taught both in primary school and in secondary school. Whereas in secondary school,

the approach is largely explicit in that chunks are an integral part of vocabulary

learning, the primary school students memorize exemplars without explicit instruction

on the suppliance rule, mainly because the Swiss Conference of the Education

Directors promotes an implicit learning (CLIL) approach at primary school level,

leaving formal grammatical instruction to teaching at secondary school level.

Materials and procedure

In this paper, I will examine written production as well as grammaticality judgment

(henceforth GJ) data. The written production tasks consisted of two types of essays

that elicited different semantic contexts: an argumentative essay on the pros and cons

of casting shows and a narrative re-telling task, encouraging students to narrate “what

happened in the silent movie ‘The Triplets of Belleville’”, which they had watched in

the previous session without the experimenter. The re-telling task has proven to be a

reliable instrument ensuring that the learners are forced to use a fixed set of noun

phrase environments.

 

The GJ task was chosen for the following reasons: (a) since essays involve the risk of

avoidance of uncertain uses of articles, it is important to consult more controlled data

sources; (b) the GJ task is a response task designed to measure the (subconscious)

knowledge of the linguistic rules that constitute the learner’s internal grammar; and

(c) it has been suggested in naturalistic studies that late L2 learners experience more

difficulty in their grammaticality judgments than early learners, since “memory

capacity, decoding ability and processing speed are deficient in late L2 learners”

(McDonald, 2006, p. 383). The GJ task was presented in written form to the learners.

The students had a maximum of 15 minutes to make their judgments (approx. 10

seconds per sentence), which, for their age, was necessary in order to read and

understand the sentences. Of the 82 items of the GJ task, 6 were relevant to the

discussion of the mastery of the L2 article system (cf. Appendix). They were partly

adopted from McDonald’s (2006) GJ task; sentences were made ungrammatical in

that articles were either omitted or added, and all the article environments and the

main error types outlined above were covered.

Depending on the size of the classes, at least 5 testing sessions of 45 minutes each

were conducted with each class. The tasks were administered in a controlled setting

(i.e., in the learners’ regular classrooms, incl. a supervisor). In the production tasks,

where the students had 45 minutes per essay, every subject produced between 450 and

900 written words, of which the first 200 of each essay were selected, thus amounting

to 400 written words per student.

Method and data scoring

 

In the GJ task, the proportion of learners’ correct acceptances of grammatical targets

and correct rejections of ungrammatical targets were calculated. In order to analyze

how the participants used, misused or failed to use articles in the production tasks, the

data was scrutinized for non-/occurrence of a, the and Ø (zero article) used with

nouns in indefinite, definite and zero environments (see RQ1 and RQ2 above).

Articles were coded according to their appropriateness in the three contexts as

follows:

(a) the in indefinite contexts (~ overuse of the)

(b) the in zero contexts (~ overuse of the)

(c) a in definite contexts (~ overuse of a)

(d) a in zero contexts (~ overuse of a)

(e) Ø in definite contexts (~ omission of the)

(f) Ø in indefinite contexts (~ omission of a)

(g) Others (e.g. the use of the definite article in “possessive” environments)

The score values were then added applying Pica’s (1983) suppliance in obligatory

context (SOC) analysis, according to which learners can score 1 point for the correct

form; 0.5 point for a morpheme misformation (a instead of an before vowels) and 0

points for no morpheme, i.e., omitted articles.

Besides this broad distinction between different article contexts, a more detailed

analysis of the sources of the different error types was conducted (see RQ2 and RQ3).

More specifically, a distinction is made between developmental errors vs. transfer

 

errors, and rule-governed use vs. memory-based use of articles. Transfer errors are

errors due to L1 interference; it is important to keep in mind that the absence of a

comparison group whose L1 has no article system makes it difficult to identify errors

due to L1 interference for certain.

Note that the omission of the variant an before nouns with a word-initial vowel sound

was counted as a developmental error (a) because it is not due to L1 interference (the

Swiss German indefinite article also takes an -n before vowels to aid in

pronunciation) and (b) with increasing proficiency, processing resources are freed up

to attend to non-salient phonological details (see VanPatten, 1996). Needless to say

that the distinction between a and an is trivial in terms of knowing the semantics of

the articles, since they are allomorphs of the same grammatical form. However, the

incorrect suppliance of an reveals problems with the detection of phonological

discriminations in the input (see above) and is therefore included in the analysis.

Results

Research questions 1 and 2

Table 1 presents the percentage correct in obligatory contexts by token counts, while

Table 2 presents the same by type counts.

Table 1 Comparison of SOC percentage scores (tokens) for the two groups

Morpheme ECLs

(n=100)

LCLs

(n=100)

Percentage

difference

Control

group (n=20)

 

Definite article .95 .94 +1 1.00

Indefinite article .91 .87 +4 1.00

Zero article .85 .86 –1 1.00

Table 2 Comparison of SOC percentage scores (types) for the two groups

Morpheme ECLs

(n=100)

LCLs

(n=100)

Percentage

difference

Control

group (n=20)

Definite article .98 .97 +1 1.00

Indefinite article .94 .92 +2 1.00

Zero article .87 .87 0 1.00

For both groups, accuracy scores by context were higher with the than they were with

a and with Ø, which is in line with previous findings. Defining mastery of structure as

over 90% accuracy (Zdorenko and Paradis, 2008), all structures but one (the zero

article) were mastered by the two learner groups. Differences in percentage points

between groups are small on all the morphemes.

Table 3 shows the mean numbers of the, a and Ø in their inappropriate contexts in the

written production data. I conducted independent sample t-tests on the learners’ mean

proportion of the different error types across all rounds, divided by context and

starting age group.

Table 3 Mean numbers of the, a and Ø in their inappropriate contexts

Age group Mean SD t-value p-value

Incorrect the in zero context (overuse of the):

ECLs .56 1.02

 

LCLs .71 1.11 0.995 .321

Incorrect the in indefinite context (overuse of the):

ECLs .05 .22

LCLs .11 .42 1.266 .207

Incorrect a in definite context (overuse of a):

ECLs .03 .16

LCLs .04 .25 0.337 .737

Incorrect a in zero context (overuse of a):

ECLs .38 .54

LCLs .12 .43 3.767 >.01*

Incorrect Ø in definite context (omission of the):

ECLs .23 .54

LCLs .28 .63 0.603 .548

Incorrect Ø in indefinite context (omission of a):

ECLs .18 .47

LCLs .43 .84 2.597 .01*

Other

ECLs .08 .27

LCLs .10 .34 0.461 .646

* Significant at a 99% confidence level

The results demonstrate that overuse of the definite article in zero environments was

clearly the dominant error type for both LCLs and ECLs (with no significant

differences). The only significant group differences are for indefinite article use in

zero contexts, where the ECLs produced a higher proportion of errors, and for

omissions of the indefinite articles, where the ECLs outperformed the LCLs. Table 4

displays the grammaticality judgments of the two groups:

Table 4 % of sentences judged correctly by ECLs vs. LCLs

 

Sent. no. ECLs

(n=100)

LCLs

(n=100)

t-value p-value Control group

(n=20)

58/65 (a cat) .84 1.00 n.s. n.s. 1.00

73/81 (at the hotel) .87 1.00 n.s. n.s. 1.00

64/81 (by car) .87 1.00 n.s. n.s. 1.00

*69/*69 (in Ø pool) .17 .19 n.s. n.s. 1.00

*36/*34 (an information) .26 .53** 2.825 <.01 1.00

*79/*80 (on the TV) .29 .50** 2.007 <.05 1.00

** Significant at a 99% confidence level

bold types = significantly higher score (i.e. stronger performance)

In general, the accuracy rates were highest across the grammatical sentences and

lowest across the ungrammatical items. Furthermore, the results show that the LCLs’

judgment of the grammatical sentences was 100% correct across all items, while the

ECLs had slightly – but not significantly – lower accuracy scores. By contrast, the

LCLs showed significantly superior abilities to detect the ungrammatical omission

and overuse of the definite article in items *36/*34 and *79/*80 respectively, both of

which include transfer errors.

In the following, the different error types will be analyzed; it will be of particular

interest why omission of the indefinite article seems to be an error type specific to the

LCLs, whereas the ECLs have more difficulty with the zero article, overusing a in

zero contexts to a significantly higher extent than the LCLs.

Research question 3

 

Tables 5 and 6 shed light on the sources of overuse and omissions of definite articles

and indefinite articles respectively by making a distinction between transfer errors and

developmental errors.

Table 5 Error types and sources (definite articles)

Age group Mean SD t-value p-value

Overuse I: Transfer

ECLs .56 .91

LCLs .72 .86 1.278 .203

Overuse II: Developmental

ECLs .05 .22

LCLs .10 .34 1.235 .218

Omissions I: Transfer

ECLs .13 .41

LCLs .21 .60 1.101 .272

Omissions II: Developmental

ECLs .10 .38

LCLs .07 .19 0.706 .481

Table 6 Error types and sources (indefinite articles)

Errors per age group Mean SD t-value p-value

Overuse I: Transfer

ECLs .38 .54

LCLs .12 .43 3.767 >.01*

Overuse II: Developmental

ECLs - - - -

LCLs - - - -

Omissions I: Transfer

 

ECLs .08 .27

LCLs .11 .35 0.678 .498

Omissions II: Developmental

ECLs .10 .43

LCLs .32 .61 2.948 >.01*

* Significant at a 99% confidence level

Table 6 shows that the main source of the overuse of the definite article (which is the

predominant error type for both groups) is L1 interference. The residue of

developmental errors as concerns the overuse of the points to a problem with

referentiality, since they occur solely in combination with referential (first-mention)

NPs in indefinite contexts, which has also been observed in several studies of L1

acquisition (e.g. Thomas, 1989). The results indicate that the learners occasionally

assumed that the events or objects they described were part of the teacher’s (or

experimenter’s) knowledge, even though they had been informed beforehand that the

experimenter had not seen the film. However, even though the ECLs and LCLs do not

always mark first-mention nouns with indefiniteness, they use the appropriate article

in most non-referential environments, with an error rate of only 2.1% for the LCLs

and 2.3% for the ECLs.

Focusing on the areas where the performances of the two groups differ significantly,

we can make two observations: on the one hand, transfer does not seem to be a

problem exclusive to late learners, as often suggested in the literature; the ECLs’

weaker performance in the use of the indefinite article (overusing a significantly more

often) can be fully ascribed to L1 interference, with not one instance of any other

error type in this area. Most of the errors (93%) produced by the ECLs occur as a

 

result of the localized contexts of cultural article usage, i.e., they produce errors on

memory-based idioms and collocations. By contrast, incorrect rule-governed use of

the indefinite article is quite rare, particularly in combination with concrete,

uncountable nouns (as in (2) above), which only occurs 5 times in the data.

The LCL’s higher number of omissions of the indefinite article is not due to L1

interference but due to phonological difficulties: a closer analysis of the most

frequently occurring errors reveals that the higher number of omissions are due to a

lack of use of the variant an, which is obligatory before a vowel and counted as 0.5

errors if it was not supplied.

Discussion

The analysis presented above for research question 1 shows that the LCLs and ECLs

are more accurate on definites than on indefinites and zero articles. Hence, the

common finding that the target article is used more often in definite contexts than in

indefinite contexts – a robust pattern across learners with different L1 backgrounds

and sometimes called “a general property of English L2 acquisition” (Zdorenko and

Paradis, 2008, p. 246) – could be confirmed for both groups. The weaker performance

on zero articles is due to the overuse of the in zero environments, which constituted

the most prominent error type for both groups. It is hypothesized in the research

literature (cf., e.g., Master, 1997) that Ø-usage and accuracy continue to increase

slowly with increasing proficiency. However, the 5-year Early English program did

not seem to long enough (or intensive enough) for the ECLs to learn more than they

did.

 

With respect to research questions 2 and 3, I found that age of onset of learning and

length of instruction were not associated with differential accuracy rates for article

use, neither in the production tasks nor in the response task. Notably, the hypothesis

that L1 interference is less pronounced in early L2 acquisition than in late L2

acquisition can thus not be supported in this study. Interestingly, the ELCs’

difficulties seem to be reduced to some idiomatic use of articles and how they differ

in the two languages, i.e., their transfer errors are predominantly memory-based

(rather than rule-governed), which points to a failure to learn vocabulary (chunks).

This is interesting insofar as the Early English program in Switzerland is designed in

such a way that the ECLs should be able to assemble and memorize a range of

idiomatic expressions and collocations including various article-noun combinations.

The finding that the ECLs are more accurate in the use of the variant an in its required

environment can be explained in terms of the activities in the Early English

classroom, which are designed specifically to improve the students’ connected speech

with all its modifications and sound assimilations. Also, this finding might be

interpreted in terms of phonological ability, since the an variant is a structure “in

which fine phonological discrimination is necessary for one to perform accurately”

(McDonald, 2008, p. 983). Phonological ability, however, is known to decrease with

age, which might account for the LCLs’ less accurate performance.

As concerns the GJ task, the LCLs do not seem to have fewer problems with their

intuitive judgment of the grammaticality of sentences containing grammatical and

ungrammatical uses of articles; on the contrary, the LCLs’ judgments were superior to

those of the ECLs when it comes to detecting ungrammaticalities, notably (memory-

 

based as well as rule-based) transfer errors. These results are interesting insofar as it

has been suggested in naturalistic studies that late L2 learners should experience more

difficulty in their grammaticality judgments than early learners. In general, the GJ

results revealed that all the learners overwhelmingly accepted the forms with correct

article suppliance as grammatical, whereas the rejection rate is never over 50% in the

ungrammatical sentences. Similar to Jiang’s (2004) results, the learners in this study

seem to have knowledge about the tested morphemes, as can be seen in their better

performance on the production tasks and their nearly perfect accuracy scores on the

acceptance of grammatical sentences in the GJ task. However, this knowledge is not

automatically activated in their judgment of ungrammatical items, which is why Jiang

suggests that “such knowledge is not an integrated part” (p. 624) of the learners’ L2

competence. While Jiang restricted herself to an analysis of the plural morpheme, the

morphological analysis in this study was able to show that her observations can be

extended to other morphemes as well. I will further discuss this issue in Pfenninger

(in preparation).

Conclusion

In sum, we can thus conclude that in an instructed setting, age 8 does not seem to be

early enough of an age at which the performance on the English article system differs

from learners with a starting age in their early adolescence, and the 5-year difference

in instruction had no significant effect on the learning outcome. Indirectly, this might

also support the hypothesis that the implicit teaching approach, as practiced in the

Swiss Early English classroom, does not necessarily lead to the ECLs’ superiority as

far as morpho-syntactic accuracy is concerned (see Ellis, 2006; Goldschneider and

 

DeKeyser, 2001; Pfenninger, 2011; Pfenninger, 2012). Another explanation can be

provided in terms of the language background of the two groups: the LCLs’

superiority can be seen as a direct influence of their previous knowledge of French.

Since the LCLs studied English as an L4, they might have profited from L2 learning

strategies and cognitive learning mechanisms they developed while learning L3

French. Finally, it needs to be analyzed if the two starting age groups described in this

paper differ in terms of their motivational intensity and the structure of their

motivational profile, which would indicate that the crucial factors must lie in the kind

of input the two populations received in the past, i.e., different learning experiences

(curricular vs. extracurricular L2 acquisition).

Appendix

GJ items concerned with article semantics:

(1) I have never had a cat, but I have two dogs. (indefinite article a)

(2) I saw you at the hotel, but you didn’t see me. (definite article the)

(3) Let’s go by car! (zero article)

(4) *I jumped in pool and started swimming very quickly. (omission of the)

(5) *I need an information. (overuse of a; transfer error, rule-governed)

(6) *There was a nice show on the TV yesterday. (overuse of the; transfer error,

memory-based)

 

References

De Graaff, R., & Housen, A. (2009). Investigating effects and effectiveness of L2

instruction. In M. Long, & C. Doughty (Eds.), The Handbook of Language

Teaching (pp. 726-753). Oxford: Wiley-Blackwell.

DeKeyser R. M. (2000). The robustness of critical period effects in second language

acquisition. Studies in Second Language Acquisition, 22, 499-533.

Ekiert, M. (2004). Acquisition of the English Article System by Speakers of Polish in

ESL and EFL Settings. Working Papers in TESOL & Applied Linguistics,

4(1), 1-23.

Ellis, N. C. (2006). Selective attention and transfer phenomena in L2 acquisition:

Contingency, cue competition, salience, interference, overshadowing,

blocking, and perceptual learning. Applied Linguistics, 27(2), 164-194.

Ellis, N. C., & Larsen-Freeman, D. (2006). Language emergence: implications for

applied linguistics. Introduction to the special issue. Applied Linguistics,

27(4), 558-589.

Goldschneider, J., & DeKeyser, R. (2001). Explaining the ‘natural order of L2

morpheme acquisition’ in English: A meta-analysis of multiple determinants.

Language Learning, 51(1), 1-50.

Goto Butler, Y. (2002). Second language learners theories on the use of English

articles: An analysis of the metalinguistic knowledge used by Japanese

students in acquiring the English article system. Studies in Second Language

Acquisition, 24(3), 451-480.

Hakuta, K. (1976). A case study of a Japanese child learning English as a second

language. Language Learning, 26, 321-351.

 

Hawkins, R., Al-Eid, S., Almahboob, I. & Athanasopoulos, P. (2006). Accounting for

English article interpretation by L2 speakers. In S. H. Foster-Cohen, M.

Medved Krajnovic, & J. Mihaljevíc Djigunovíc (Eds.), EUROSLA Yearbook,

6, 7-25.

Ionin, T., & Wexler, K. (2002). Why is ‘is’ easier than ‘-s’?: acquisition of

tense/agreement morphology by child second language learners of English.

Second Language Research, 18(2), 95-136.

Jiang, N. (2004). Morphological insensitivity in second language processing. Applied

Psycholinguistics, 25, 603-634.

Klages-Kubitzki, M. (1995). Article usage in English. A computer-based self-teaching

programme on the basis of a functional theory of reference. Frankfurt/M.,

Berlin, Bern, New York, Paris, Wien: Peter Lang.

Lardiere, D. (2004). Knowledge of definiteness despite variable article omission. In

A. Brugos, L. Micciulla, & C. E. Smith (Eds.), BUCLD 28 Proceedings.

Somerville, MA: Cascadilla Press, 328-39.

Liu, D. & Gleason, J. L. (2002). Acquisition of the article the by nonnative speakers

of English. Studies in Second Language Acquisition 24, 1-26.

McDonald, J. L. (2006). Alternatives to the critical period hypothesis: Processing-

based explanations for poor grammaticality judgment performance by late

second language learners. Journal of Memory and Language, 55, 381-401.

McDonald, J. L. (2008). Grammaticality judgments in children: The role of age,

working memory and phonological ability. Journal of Child Language, 35,

247-268.

Master, P. (1997). The English article system: Acquisition, function, and pedagogy.

System, 25, 215-232.

 

Nayak, N., Hansen, N., Krueger, N., & McLaughlin, B. (1990). Language-learning

strategies in monolingual and multilingual adults. Language Learning, 40,

221-244.

Parrish, B. (1987). A new look at methodologies in the study of article acquisition for

learners of ESL. Language Learning, 37, 361-383.

Pfenninger, S. E. (2011). Age effects on the acquisition of nominal and verbal

inflections in an instructed setting. Studies in Second Language Learning and

Teaching, 1(3), 401-420.

Pfenninger, S. E. (2012). On the effectiveness of early implicit classroom learning.

Evidence from morphology. ELT Research 26.

Pfenninger, S. E. (in prep). The earlier the better? On the benefit question of Early

English instruction in Switzerland.

Pica, T. (1983). Adult acquisition of English as a second language under

differentconditions of exposure. Language Learning, 33, 465-497.

Thomas M. (1989). The acquisition of English articles by first- and second-language

learners. Applied Psycholinguistics, 10, 335-355.

VanPatten, B. (1996). Input processing and grammar instruction in second language

acquisition. Norwood, NJ: Ablex.

Williams, J. N. & Lovatt, P. (2003). Phonological memory and rule learning.

Language Learning, 53(1), 67-121.

Yoon, K. K. (1993). Challenging prototype descriptions: Perception of noun

countability and indefinite vs. zero article use. IRAL, 31, 269-298.

Zdorenko, T. & Paradis, J. (2008). The acquisition of articles in child second

language English: fluctuation, transfer or both? Second Language Research,

24(2), 227-250.