Snape, N. (2005).

The certain uses of articles in L2-English by

Japanese and Spanish speakers. Durham and Newcastle Working
Papers in Linguistics. vol. 11, 155 - 168.

THE CERTAIN USES OF ARTICLES IN L2-ENGLISH BY JAPANESE AND SPANISH

SPEAKERS∗

NEAL SNAPE

Abstract

It is well documented in second language (L2) studies that learners of English have difficulty
acquiring articles. Previous studies of L2 learners’ use of articles in English, such as Huebner
(1985) and others have attempted to explain why the definite article the is overused in
indefinite article a contexts, employing Bickerton’s (1981) semantic system [±SR, ±HK] for
noun phrase reference. More recently a more sophisticated approach to article classification
has been offered by Ionin (2003a&b) and Ionin and Wexler (2003; henceforth I&W) using
two models; the de re de dicto distinction and the referentiality distinction. I&W and White
(2003) both assume that interlanguage grammars are UG-constrained, but I&W claim that L2
learners of English ‘fluctuate’ between parameter settings for articles.
The present paper reports a study with ten Japanese learners of English (hereafter
JLEs), ten Spanish learners of English (hereafter SLEs) and native control speakers. The
prediction is that errors will be found in article use reflecting feature specifications or
parameter settings that are allowed by UG, but are inappropriate for English. Data were
collected using a similar gap-filling written task as used in I&W (2003), along with a story re-
call oral production task. The same contexts are used plus two extra contexts for definite
anaphoric (see Lyons 1999). The results, so far, are consistent with the results of I&W. They
found, as predicted, that their L1 Russian learners overused the in indefinite referential de re
contexts, and only the JLEs in this study overuse the in this context in the gap-filling study.
While advanced learners in our study were more accurate in their article use than intermediate
learners, article omission errors still persist for the advanced JLEs. Furthermore, JLEs seem to
encounter problems with mass nouns due to the determiner-noun concord in English.
We will argue (unlike I&W) that L1 transfer partly explains the results for the
Japanese and Spanish learners’ performance. The SLEs are more accurate in article usage
overall (Spanish has an article system), but JLEs are less accurate because there is no article
system present in Japanese. The results of the two studies seem to support Prevost & White’s
(2000) Missing Surface Inflection Hypothesis (MSIH), but this does not fully explain

∗
Earlier versions of this paper were presented at the University of Cambridge Looking at Language Acquisition
V workshop (May 2004), the University of Essex Graduate Students Presentation Day 9 (May 2004) and the
Seventh Durham Postgraduate in Theoretical and Applied Linguistics (June 2004.) This research is being funded
by the ESRC, award no PTA-030-2003-01043. I would like to thank Roger Hawkins and Y-k. I. Leung for
useful comments and suggestions.

- 155 -
differences between the two L2 groups. However, we believe the results support Schwartz
and Sprouse’s (1996) Full Transfer/Full Access model.

1. Introduction

It is well documented in L2 studies that learners of English have difficulty acquiring

articles. Previous studies of L2 learners’ use of articles in English, such as Huebner (1985),
Parrish (1987) and Murphy (1997) have attempted to explain this difficulty by employing
Bickerton’s (1981) semantic system [±SR, ±HK] for noun phrase reference. Robertson (2000)
examined article acquisition by advanced Chinese learners of English and claimed that three
principles could account for optionality in the use of articles. 1 More recently a more
sophisticated approach to article classification has been offered by Ionin (2003a&b) and Ionin
and Wexler (2003) (henceforth I&W) using two models; the de re/de dicto2 distinction and
the referentiality distinction.

1.1. Definitions

1. Definiteness: a DP is definite iff its referent is known to both speaker and hearer, and is
unique in the contextually relevant domain. Otherwise, the DP is indefinite.
definite: I read a book. The book was interesting.
indefinite: I read a book yesterday.

2. Referentiality: an indefinite DP is referential iff the speaker has its referent ‘in mind’ and
intends to refer to it. Otherwise, the DP is quantificational.
referential indefinite: I read an interesting book, which my cousin gave me.
quantificational indefinite: Mary read a book (but I don’t know which one).

3. The de re/de dicto distinction: an indefinite DP is de re iff it is not in the scope of an

operator such as an intentional verb, a modal, or negation.3 Otherwise, the DP is de dicto.
de re indefinite (referential): I’d like to meet an actor – I really like his movies. (exists in the
world)
de re indefinite (non-referential): I’d like to meet an actor after this performance.
de dicto indefinite (non-referential): I’d like to meet an actor – any famous actor will do. (not
of our world at this moment).

(based on I&W 2003)

I&W suggest that there is an Article Choice Parameter determining the distribution of
articles and claim that L2 learners possibly pick the wrong value (or fluctuate) when they are

1
Also see Snape (2002) for an L2 study using J.A. Hawkins’ (1978), Robertson’s (2000) and Bickerton’s (1981)
article classification systems.
2
I&W do differ to the previous L2 studies in that they believe the de re/de dicto distinction does not play a role
in article choice, only referentiality does.
3
I&W extend the traditional use of the term de re to cover sentences with no intensional verbs, modals, or
negation, since in the absence of another operator, the DP necessarily takes widest scope.

- 156 -
learning an L2 like English. They argue that their study is the first attempt to try to tease apart
the with referentiality and the with a de re/de dicto reading.
Our study also examines a further difficulty with mass nouns in English. As Japanese
does not distinguish between count and mass it is predicted that the JLEs in our study will
have problems with certain types of mass nouns and be unsure whether definite anaphoric the
can be licensed by mass nouns in English.4
Furthermore, it’s not clear whether all L2 learners go through a temporary stage of
development in article choice as I&W claim (a representational problem) or whether L2
learners can select the correct parameter value as White (2003) claims in her study of a
Turkish speaker. Both assume interlanguage grammars are UG-constrained, but the difference
is that White (2003) argues that when articles are used they are used correctly and any
omissions are the result of a mapping problem from syntactic representations to phonological
forms. Therefore, this study tries to look for possible evidence that the L2 learners have
maybe chosen the wrong parameter option made available by UG and at possible evidence for
a mapping problem (MSIH). Furthermore, we consider what role the L1 may play in these
two domains.

2. Research hypotheses

H1 JLEs and SLEs will tend to overuse the in indefinite referential de re singular contexts
in both studies, rather than overuse Ø as they tend to associate the with referentiality.
H2 Definite the associated with mass nouns will be more problematic for JLEs because
they involve determiner-noun concord which differs between English and Japanese.
H3 Advanced JLEs will be more accurate in article use than intermediate JLEs, if mis-
setting of the Article Choice Parameter is a temporary property of development.
H4 Advanced JLEs will continue to omit articles if a mapping of representations of forms
is a persistent feature of interlanguage grammar.
H5 Advanced JLEs and SLEs will show the same level of omission of articles if the
MSIH applies to all L2 learners.

3. Empirical study 1: story re-call task

3.1. The participants

The participants in this study were ten Japanese learners of English, ten Spanish
learners of English and six native control speakers. The native controls were able to re-call
each story and supply the required article in the appropriate context.5 All participants’ ages
ranged from 23 to 40 (mean = 28) and have been studying in the UK for an average of six
months. Even though all the participants have taken TOEFL and have scores equivalent of
575 or above, they were asked to take the Oxford Placement Test in order to obtain their
current proficiency level. All participants were then placed into levels according to their
scores on the OPT.

4
Wakabayashi (1998) claims that the JLEs are more successful when the noun phrase is [singular countable]
because they believe the use of articles is obligatory, but optional when it is either [plural –s] or [uncountable].
5
This is discussed in more depth in the coding procedure because coding production data is more difficult. Ionin
(2003b) argues that one of the main disadvantages of collecting production data is that it does not easily allow
the researcher to control the contexts in which articles are produced.

- 157 -
3.2. The task

I&W (2003) conducted two studies involving a written translation task and an
elicitation task. This study differs to that of I&W because it is an oral production task.
However, it does focus on the same four relevant contexts: definite anaphoric singular,
indefinite referential de re singular, indefinite non-referential de re singular and
indefinite non-referential de dicto singular, plus two new contexts; definite anaphoric
plural and mass. The participants were asked to listen to thirteen short stories. Each story
was presented twice and prompts were given to the participants to assist them in the re-call of
the story. Once they had listened twice to one story they were then asked to re-call the story,
using the prompts in the order they appeared on the slide. Each story re-call was recorded
digitally using PolderbitS Sound Recorder and Editor. The stories contained certain contexts
where different articles could be used. Referential indefinite de re contexts did not involve
using modifiers like certain or specific. Nonreferential de re contexts involved simple first-
mention indefinites with no scope interactions, as used in I&W’s 2003 study; in (1) there is no
reason to think that the speaker intends to refer to a particular ‘young man’. Finally,
nonreferential de dicto indefinites all involved DPs in the scope of an intentional verb or a
modal operator.

(1) At Colchester North station, an elderly woman’s daughter watched a young man run
quickly down platform three to catch the next train to London. The daughter of the
elderly woman caught the same train, but took her time, strolling down the platform. `I
thought the train was leaving,' the young man said. `They can't find a driver,' the
elderly woman’s daughter replied.
Prompts: station, elderly woman, daughter, young man, train to London, driver

3.3. Coding procedure

Coding the results of each story is as follows. Once each story had been transcribed
using Express Scribe software, we compared the original stories to the transcribed versions
and scored them on their correct and incorrect usage of articles. If a different article appeared
to the one we were expecting in a particular context, we would score it if it was
unambiguously definite or indefinite.6 For example, in (1) (repeated below as (2)) ‘a driver’
has the reading non-referential, de dicto: a;

(2) “I thought the train was leaving,” the young man said. “They can't find a driver,” the
elderly woman’s daughter replied.

When re-calling the story, some of the participants would say ‘the driver’ in this
context. This is not the correct usage, but it is still grammatically correct to use a definite here.
The use of a definite changes the meaning of the sentence. ‘They can’t find a driver’ means
that they can’t find any driver for the train (no specific driver exists), whereas ‘They can’t
find the driver’ means there is a specific driver for this specific train. It is an associated use of

6
As explained by Ionin (2003b:205), “An unambiguously definite context is one in which L1-English speakers
consistently put the. An unambiguously indefinite context is one in which L1-English speakers consistently put a
(for singulars) or omit articles (for plurals)”.

- 158 -
the, as defined by J. Hawkins (1978). Therefore, if a definite was used here it would be scored
as correct, but placed in the new context of ‘definite association singular’.

3.4. Results

Story re-call results

Table 1. Japanese subjects (n=10)

context: target article
article used definite definite definite
anaphoric anaphoric anaphoric
singular: the plural: the mass: the
the 70.0% 27.3% 52.0%
a 6.0% 18.2% 8.0%
∗
Ø 24.0% 54.5% 40%

context: target article

article used referential de non- non-
re: a referential, de referential, de
re: a dicto: a
the 12.5% 10.7% 21.5%
a 63.9% 50.8% 67.7%
Ø 23.6% 38.5% 10.8%

Table 2. Spanish subjects (n=10)

context: target article
article used definite definite definite
anaphoric anaphoric anaphoric
singular: the plural: the mass: the
the 95.7% 83.3% 100%
a 0% 0% 0%
Ø 4.3% 16.7% 0%

context: target article

article used referential de non- non-
re: a referential, de referential, de
re: a dicto: a
the 8.5% 8.9% 8.1%
a 87.3% 79.8% 86.5%
Ø 4.2% 11.3% 5.4%

∗
This is a very high omission rate largely due to one item in the study. The item ‘strawberry’ may be correctly
analysed as countable, but it could be that some JLEs do not realize that a definite anaphoric the can be used
with it in context.

- 159 -
 Between-group comparisons (one-way ANOVA with post-hoc Tukey HSD tests) were
performed. Each group of L2 learners was compared against the English native speakers’
(NS) performance. Results revealed the following:

Definites
(i.) For all definite contexts there are significant differences between each group of L2
learners and the NS in the use of the in definite anaphoric plural/mass contexts
(F(2,25) = 7.904, p=0.002 for definite anaphoric plural contexts and F = 3.825,
p=0.03 for definite anaphoric mass contexts). Post-hoc tests reveal that there is a
significant difference (p=0.01) between SLEs and JLEs on the definite anaphoric
plural context.
(ii.) For overuse of indefinites in definite contexts no significant difference was found
between the three groups.
(iii.) For overuse of Ø in definite contexts, there is a significant difference between all
groups in the definite anaphoric singular and definite anaphoric mass contexts (F =
7.456 for definite anaphoric singular and F = 11.795 for the definite anaphoric
mass context; all contexts p<0.05). Post-hoc tests reveal there is a significant
difference between the SLEs and JLEs (p<0.05).7

Indefinites
(i.) For all indefinite contexts there are significant differences between each group of
L2 learners and the NS in the use of indefinite referential/non-referential de
re/non-referential de dicto contexts (F = 7.425 for indefinite referential de re
singular, F = 14.978 for indefinite non-referential de re singular and F = 9.699 for
indefinite non-referential de dicto singular contexts; all contexts p<0.05). Post-hoc
tests reveal that there is a difference between the NS and the two L2 groups on all
contexts. However, the SLEs performed better than the JLEs in the indefinite non-
referential de re singular context as there was a significant difference (p=0.00)
between the SLEs and JLEs.
(ii.) For overuse of definites in indefinite contexts, a significant difference was found
between the NS and the JLEs (F = 3.624, p=0.04) in the indefinite non-referential
de dicto singular context.
(iii.) For overuse of Ø in indefinite contexts there is a significant difference between NS
and each group of L2 learners in the indefinite referential de re singular and
indefinite non-referential de re singular contexts (F = 5.489 for indefinite
referential de re singular and F = 24.633 for the indefinite non-referential de re
singular context; all contexts p<0.05). Post-hoc tests reveal that there is a
significant difference between the SLEs and JLEs (p<0.05). JLEs produced more
omissions in these contexts.

4. Empirical study 2: gap-filling task

7
No significant difference was found in the definite anaphoric plural context because the NS produced 14.3% of
omissions in this context.

- 160 -
 The goal of this second study was similar to that of I&W’s written translation task. We
wanted to test article use in various definite and indefinite contexts in a written task, looking
specifically at referentiality and the de re/de dicto distinction. We also used relative clause
(RC) modification for the referential de re context,8 as did I&W in their elicitation study.

4.1. The task

The participants were asked to read short dialogues in which there is a gap. Following
each dialogue there are four possible items that could fill the gap. They were asked to choose
the item that they felt was most appropriate to fill the gap and put a circle around it. There
were four short dialogues for each context and the six contexts were randomised:

definite anaphoric singular definite anaphoric plural

A: Come on! We’ve been in this shop A: Hi, Jimmy! How was school?
for hours. B: We had two chemistry tests.
B: I can’t make up my mind. Which A: Did you find ____ tests difficult?
shirt do you like best?
C: I prefer ____ shirt with stripes.

the a an Ø an Ø the a

definite anaphoric mass referential de re with RC-modification: a

A: Jason just asked me for some more A: Excuse me.

cash! B: How can I help?
B: I don’t understand why he is always A: I would like to buy ___ CD that I
so short. have been trying to find for ages.
A: He says he will use ____ cash to pay
off his credit card bill.

an Ø the a Ø an a the

non-referential, de re: a non-referential, de dicto: a

A: Rose is happy. A: It is my birthday next week.

B: Why? B: Do you expect to get many gifts?
A: She got ____ car for her birthday. I A: No, but I have told my brother that I
wonder what it looks like? want ____ watch.

the a an Ø an Ø the a

8
According to Fodor and Sag (1982), RC-modification biases an indefinite in favour of the referential reading.

- 161 -
4.2 Results

Gap-filling results

Table 3. Japanese subjects (n=10)

context: target article
article used definite definite definite
anaphoric anaphoric anaphoric
singular: the plural: the mass: the
the 90% 95% 50%
a 7.5% 2.5% 5%
Ø 2.5% 2.5% 45%

context: target article

article used referential de non- non-
re with RC- referential, de referential, de
modification: re: a dicto: a
a
the 29.5% 10% 4.5%
a 68% 90% 88%
Ø 2.5% - 7.5%

Table 4. Spanish subjects (n=10)

context: target article
article used definite definite definite
anaphoric anaphoric anaphoric
singular: the plural: the mass: the
the 95% 95% 85%
a 5% - 5%
Ø - 5% 10%

context: target article

article used referential de non- non-
re with RC- referential, de referential, de
modification: re: a dicto: a
a
the 7% 2.5% 5%
a 93% 95% 95%
Ø - 2.5% -

Definites
(i.) For all definite contexts there is a significant difference between each group of L2
learners and the NS in the use of the in the definite anaphoric mass context (F(2,29)

- 162 -
 = 12.696, p=0.00). The JLEs used the far less than the SLEs in this context with
post-hoc tests revealing a significant difference (p=0.00).
(ii.) For overuse of indefinites in definite contexts no significant difference was found
between all three groups.
(iii.) For overuse of Ø in definite contexts there is a significant difference between all
three groups in the definite anaphoric mass context (F = 17.229, p=0.00). Post-hoc
tests reveal that there is a significant difference between the SLEs and JLEs
(p<0.05).

Indefinites
(i.) For all indefinite contexts there is a significant difference in the use of a in the
indefinite referential de re singular context between each group of L2 learners and
the NS (F = 4.918, p=0.01). The JLEs used a far less than the SLEs in this context
with post-hoc tests showing this as significant (p=0.03).
(ii.) For overuse of definites in indefinite contexts a significant difference was found
between all three groups in the indefinite referential de re singular context (F =
4.096, p=0.02) Post-hoc tests reveal that there is no significant difference between
the two L2 groups.9
(iii.) For overuse of Ø in indefinite contexts there is a significant difference between the
NS and the L2 groups in the indefinite non-referential de dicto singular context (F
= 3.857, p=0.03).

5. Discussion

In I&W’s (2003) study they found their L2-learners overused the in indefinite
referential contexts involving wide scope and/or RC-modification and in our studies only the
JLEs (not SLEs) tended to overuse the in this context more often in the gap-filling study.
Again in the story re-call study, only the JLEs overused the, but in the indefinite non-
referential de dicto singular context.10 This is not predicted to happen according to I&W.
Therefore, H1 is only partly supported:

H1 JLEs and SLEs will tend to overuse the in indefinite referential de re singular contexts
in both studies, rather than overuse Ø as they tend to associate the with referentiality.

The overuse of Ø appeared in the story re-call and gap-filling studies for definite anaphoric
mass contexts, thus supporting H2:

H2 Definite the associated with mass nouns will be more problematic for JLEs because
they involve determiner-noun concord which differs between English and Japanese.

9
This was almost significant p=0.05.
10
A possible explanation for the overuse of the is due to certain items in the task and how they have been
interpreted by the JLEs and coded by the researcher.

- 163 -
Overall, more omission of articles were made by the JLEs in the story re-call task. No overuse
of a was found in definite contexts for both studies.11

H3 and H4 are also supported:

H3 Advanced JLEs will be more accurate in article use than intermediate JLEs, if mis-
setting of the Article Choice Parameter is a temporary property of development.

H4 Advanced JLEs will continue to omit articles if a mapping of representations of

forms is a persistent feature of interlanguage grammar.
All the advanced and intermediate learners’ results have been collapsed as have
I&W’s (2003) for the tables of results in each study. In order to have a clearer understanding
of what the two proficiency groups are doing, we produced scatterplots:

Fig 1. Story re-call study: individual results for all groups on the correct usage (%) of
definites and indefinites supplied in the six contexts.
100 100
definite anaphoric singular context (%)

90 90
definite anaphoric plural context (%)

80 80

70 70

60 60

50 50

40 40
30 30
20 20
Spanish Spanish
10 10
Japanese Japanese
0 0
-10 NS -10 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

100 100
indefinite referential singular context (%)

90 90
definite anaphoric mass context (%)

80 80

70 70

60 60

50 50

40 40
30 30
20 20
Spanish Spanish
10 10
Japanese Japanese
0 0
-10 NS -10 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

11
Leung (2001) found an overuse of indefinite a in her elicited written production study. The L2-learners were
less accurate with the definite article.

- 164 -
 100 100
indefinite non-referential singular context (%)

indefinite de-dicto singular context (%)

90 90

80 80

70 70

60 60

50 50

40 40
30 30
20 20
Spanish Spanish
10 10
Japanese Japanese
0 0
-10 NS -10 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

Fig 2. Gap-filling study: individual results for all groups on the correct usage (%) of
definites and indefinites supplied in the six contexts.
100 100
definite anaphoric singular context (%)

90 90
definite anaphoric plural context (%)
80 80

70 70

60 60

50 50

40 40

30 30

20 Spanish Spanish
20

10 Japanese Japanese
10
0 NS 0 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

100 100
indefinite referential singular context (%)

90 90
definite anaphoric mass context (%)

80 80

70 70

60 60

50 50

40 40
30 30
20 20
Spanish Spanish
10 10
Japanese Japanese
0 0
-10 NS -10 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

- 165 -
 100 100
indefinite non-referential singular context (%)

indefinite de-dicto singular context (%)

90 90

80 80

70 70

60 60

50 50

40 40

30 30

20 Spanish Spanish
20

10 Japanese Japanese
10
0 NS 0 NS
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

OPT scores OPT scores

Overall proficiency (as measured by the OPT) had little effect on article usage. While
advanced learners were more accurate than the upper/lower intermediate learners, errors of
overuse of Ø persisted for the advanced JLEs group.12
Finally, H5 is not supported because the SLEs performance in both studies
is far better than the JLEs:

H5 Advanced JLEs and SLEs will show the same level of omission of articles if the
MSIH applies to all L2 learners.

6. Conclusion
The present studies examined similar contexts to those in I&W’s (2003) studies. They
assume that the definiteness and referentiality semantic features are two settings of the Article
Choice Parameter. They claim that their Russian L2-learners of English fluctuate between the
two settings using a and the interchangeably with specific indefinites. They conclude that L2-
learners have full access to UG principles and possible parameter settings, but have problems
generalizing from the input. Thus, setting the relevant parameters remains a difficulty even
for advanced L2-learners of English and that the triggers related to article choice may be too
subtle.
Our results suggest that there is a difference between L2 learners in the use of the, a
and Ø. Kuribara (1999) argues that JLEs have difficulty acquiring the functional head
Determiner in English because Japanese lacks a D head, therefore parameter-resetting is not
possible. We argue JLEs can acquire the D head in English, as the ± definiteness distinction
is represented in their grammars, as White (2003) found with her Turkish subject. But, it
seems that at advanced levels of English, JLEs still tend to overuse the in referential/non-
referential de re and de dicto singular contexts and continue to produce high omission errors,
as found in this study. This may be because JLEs set the inappropriate parameter value13, but
at very advanced levels do select the right values for English. Additional problems for JLEs
seem to be plural and mass nouns because they involve determiner-noun concord which
differs between English and Japanese.

12
Between-group comparisons and post-hoc tests reveal significant differences (p<0.05) between the advanced
SLEs and advanced JLEs and the intermediate SLEs and intermediate JLEs for some contexts.
13
Hawkins (p.c) suggests that JLEs may simply go through a transient phase in acquisition of articles.

- 166 -
 Franceschina (2002) and White et al. (2004) propose that Spanish has a D head as
Spanish has definite and indefinite articles like English. If this is the case the SLEs are not
acquiring a D head from ‘scratch’. 14 Therefore, if the JLEs’ and SLEs’ L1s differ, it is
expected L1 transfer will take place to a certain degree when learning English.15
To conclude, JLEs’ interlanguage grammars seem to allow more optionality within the
Determiner Phrase than SLEs. We argue that the MSIH proposed by Prévost & White (2000)
seems to be a correct characterization of speakers with L1s that lack articles, but not for
speakers with L1s that have articles. This suggests that the MSIH is not universal to all L2
speakers. Rather, it is L1 dependant. However, we believe the results do support Full
Transfer/Full Access16 as proposed by Schwartz & Sprouse (1996).

References

Bickerton, D. (1981). Roots of language. Ann Arbor: Karoma Press.

Chaudron, C. & Parker, K. (1990). Discourse markedness and structural markedness: the
acquisition of English noun phrases. Studies in Second Language Acquisition 12. 43-
64.
Chierchia, G. (1998). Reference to kinds across languages. Natural Language Semantics 6.
339-405.
Fodor, J & Sag, I.. (1982). Referential and quantificational indefinites. Linguistics and
Philosophy 5. 355–398.
Franceschina, F. (2002). Case and phi-feature agreement in advanced L2 Spanish grammars.
In Foster-Cohen, S. (ed.), EUROSLA Yearbook, vol. 2. Amsterdam: John Benjamins.
71-86.
Hawkins, J.A. (1978). Definiteness and indefiniteness. Cambridge: Trinity Hall.
Hawkins, R, & Chan, Y. (1997). The partial availability of UG in SLA: the failed functional
features hypothesis. Second Language Research 13. 187–226.
Huebner, T. (1985). System and variability in interlanguage syntax. Language Learning 35.
141-163.
Ionin, T. (2003a) The interpretation of ‘the’: A new look at articles in L2-English. In
Beachley, B. et al. (eds.), Proceedings of the 27th Boston University Conference on
Language Development. Somerville, MA: Cascadilla Press. 346-357.
Ionin, T. (2003b). Article semantics in second language acquisition. PhD dissertation, MIT.
Distributed by MIT Working Papers in Linguistics.
Ionin, T & Wexler, K. (2003). The certain uses of the in L2-English. In Liceras, J. M. et al.
(eds.), Proceedings of the 6th Generative Approaches to Second Language Acquisition
Conference. Somerville, MA: Cascadilla Press. 150-160.
Kuribara, C. (1999). Resettability of a syntactic parameter: Acquisition of category D in
English by adult Japanese speakers. Representation and process: Proceedings of the
3rd Pacific Second Language Research Forum. 13-22.

14
Wakabayashi (1997) claims that Japanese speakers must acquire the specification of English D from ‘scratch’
15
See Chierchia’s (1998) proposal for a nominal mapping parameter with a three-way distinction.
16
It is not clear whether the failure to acquire certain features within DP are due to what Lardiere (2000) refers
to as a ‘mapping problem’ or it is a failure in the selection of parameterized formal features (uninterpretable
features), as proposed by Hawkins and Chan (1997).

- 167 -
Lardiere, D. (2000). Mapping features to forms in SLA. In Archibald, J. (ed), Second
language acquisition and linguistic theory. Oxford: Blackwell. 102–29.
Leung, Y-k. I. (2001). The initial state of L3A: Full Transfer and Failed Features? In Bonch-
Bruevich, X., Crawford, W.J., Hellermann, J., Higgins, C. & Nguyen, H. (eds.), The
past, present and future of second language research: Selected proceedings of the 4th
Second Language Research Forum. Somerville, MA: Cascadilla Press. 55-75.
Lyons, C. (1999). Definiteness. Cambridge: Cambridge University Press.
Murphy, S. (1997). Knowledge and production of English articles by advanced second
language learners. PhD. dissertation, University of Texas at Austin.
Parrish, B. (1987). A new look at methodologies in the study of article acquisition for learners
of ESL. Language Learning 37. 361-83.
Prévost, P. & White, L. (2000). Missing surface inflection or impairment? Evidence from
tense and agreement. Second Language Research 16. 103-133.
Robertson, D. (2000). Variability in the use of the English article system by Chinese learners
of English. Second Language Research 16: 2. 135-172.
Schwartz, B.D. & Sprouse, R.A. (1996). L2 cognitive states and the Full Transfer/Full Access
model. Second Language Research 12:1. 40–72.
Snape, N. (2002). Variability in the use of the English article system by Japanese learners of
English. Unpublished MA dissertation. University of Essex.
Thomas, M. (1989). The acquisition of English articles by first- and second-language learners.
Applied Psycholinguistics 10. 335-355.
Wakabayashi, S. (1997). The acquisition of functional categories by learners of English. PhD
dissertation, University of Cambridge.
Wakabayashi, S. (1998). Systematicity in the use of the definite article by Japanese learners
of English. Gunma kenritsu joshi daigaku kiyo 19.
White, L. (2003). Fossilization in steady state L2 grammars: implications of persistent
problems with inflectional morphology. Bilingualism: Language and cognition 6:2.
128-141.
White, L., E. Valenzuela, M. K. Macgregor and Y-k. I. Leung (2004). Gender and number
agreement in non-native Spanish. Applied Psycholinguistics, 25:1. 105-33.

Neal Snape
University of Essex
Department of Language & Linguistics
Wivenhoe Park CO4 3SQ
UK

[email protected]

- 168 -