L2 English Articles and the Computation of
Mei Yang and Tania Ionin
Guangdong University of Foreign Studies and University of Illinois at Urbana-Champaign
There has been much recent work investigating second language (L2) acquisition of English
articles by speakers of article-less L1s (Huebner 1983, Master 1987, Parrish 1987, Thomas 1989,
Kaneko 1996, Leung 2001, Ionin, Ko and Wexler 2004, Hawkins et al. 2006, Ko, Ionin and Wexler
2006, Ionin, Zubizarreta and Bautista Maldonado 2008, Trenkic 2008, among others). Examining
L2-learners production of English articles, this research has largely focused on the reasons behind
L2-learners’ article misuse. In particular, the work of Ionin et al. (2004) and Ko et al. (2006) has shown
that article misuse in L2-English is not random but rather is constrained by the semantic features of
definiteness, specificity, and partitivity.
In the present paper, we take a different approach: first, we investigate comprehension instead of
production; second, we ask how L2-English learners whose L1s lack articles acquire the very concept
of definiteness. Specifically, we investigate whether Mandarin Chinese-speaking L2-English learners
are sensitive to the semantic concept of uniqueness that underlies definiteness. We also identify
pragmatic factors that contribute to the computation of uniqueness in L2-English. Based on our
experimental data, we argue that L2-English learners are sensitive to the importance of uniqueness in
determining definiteness, but also over-rely on the discourse factors of previous-mention and
association in determining whether the or a should be used.
In this section, we briefly discuss the semantics and pragmatics of definiteness, and review
previous studies of articles in L2-acquisition.
2.1. Semantics of definiteness
Definiteness is a discourse-related semantic feature: it is related to the knowledge state of the
speaker and hearer in the discourse. As shown by the informal definition in (1), based on Ionin et al.
(2004), the feature [+definite] reflects the knowledge state of both speaker and hearer (see Heim 1991
on the formal semantics of definiteness).
(1) If a Determiner Phrase (DP) of the form [D NP] is marked as [+definite], this indicates that the
speaker assumes that the hearer shares the speaker’s presupposition of the existence of a unique
individual in the set denoted by the NP.
In English, the feature [+definite] is morphologically encoded by the definite article the, as
illustrated by (2). Upon the first mention of a pen, the speaker does not have grounds for assuming that
© 2009 Mei Yang and Tania Ionin. Proceedings of the 3rd Conference on Generative Approaches to Language
Acquisition North America (GALANA 2008), ed. Jean Crawford et al., 325-335. Somerville, MA: Cascadilla
the hearer shares knowledge of a unique, contextually salient pen; therefore, use of the is infelicitous,
and a has to be used instead. Upon the second mention of the same pen, however, the hearer is already
acquainted with a salient pen, the conditions of definiteness have been met, and the is used.
(2) I have a pen in my bag. I will give you the pen.
As spelled out in (1), felicitous use of the with singular NPs has three components: the existence
presupposition must be met (at least one salient pen must exist in the discourse); the uniqueness
presupposition must be met (at most one salient pen must exist in the discourse); and these
presuppositions must be shared by speaker and hearer (i.e., knowledge of a salient pen on the speaker’s
part only is not sufficient). While many recent studies, reviewed in the next section, have examined
whether L2-English learners are sensitive to the “hearer knowledge” requirement, and, to a lesser
extent, learners’ knowledge of the existence presupposition, there has been little investigation into how
L2-English learners compute the concept of uniqueness. This is the focus of our study.
2.2. Previous studies on articles in L2-English
Many researchers (Huebner 1983, Master 1987, Thomas 1989, Ionin et al. 2004, among many
others) have found that L2-English learners often overuse the with indefinites and/or overuse a with
definites. A number of proposals have been made to account for these patterns of article misuse, with
the focus on the role of speaker vs. hearer knowledge.
Previous studies within the framework of Bickerton (1981) tried to explain article misuse by
investigating whether L2-English learners associate the definite article the with [+SR] (specific
referent) or with [+HK] (assumed hearer knowledge). The results, however, were somewhat
contradictory (see Thomas 1989 for findings that the is associated with the [+SR] feature, and for a
review of earlier studies). One of the problems for this approach, pointed out by Ionin (2003), is that
the feature specific referent was never clearly defined. Another problem, from our perspective, is that
the focus on hearer knowledge ignores the relevance of the uniqueness presupposition to the semantics
More recent work, by Ionin (2003) and Ionin et al. (2004, 2008) argued that L2-English learners
from article-less L1s associate the with the semantic feature of specificity, defined as “speaker intent to
refer”, with the informal definition in (3) (see Ionin 2003, Ionin et al. 2004 for the formal definition).
If a Determiner Phrase (DP) of the form [D NP] is [+specific], then the speaker intends to refer to
a unique individual in the set denoted by the NP, and considers this individual to possess some
Based on the observation that some languages, notably Samoan, morphologically encode
specificity instead of definiteness, Ionin et al. (2004) proposed that L2-English learners coming from
article-less L1s should have access to the semantic universals of both definiteness and specificity,
through Universal Grammar. According to Ionin et al.’s Fluctuation Hypothesis, L2-learners are
predicted to fluctuate between using English articles to mark definiteness and using them to mark
specificity, with the result that errors were predicted to occur only in the contexts of specific
indefinites (overuse of the) and non-specific definites (overuse of a).1 These predictions were largely
confirmed with adult L1-Russian and L1-Korean L2-English learners, who were found to often
1 See Ionin, Zubizarreta and Philippov (2009) for modifications to the original Fluctuation Hypothesis. Based on
new evidence from Samoan (Fuli 2007), Ionin et al. (2009) argue that only specificity-related errors with
indefinites, not specificity-related errors with definites, reflect L2-learners’ access to the semantic universal of
specificity; their revised proposal receives support from findings with L1-Russian L2-English children. This
revision has no direct bearing on the present paper, however.
associate the with specificity as speaker knowledge, rather than always associating it with definiteness
A separate issue is whether L2-English learners can correctly associate the with the presupposition
of uniqueness. This was the question investigated by Ko et al. (2006), who examined article use in
L2-English in the context of partitivity, whose definition is given in (4) (cf. Enç 1991 for a theoretical
account of partitivity).
If a DP of the form [D NP] is [+partitive], it denotes an individual that is a member of a set
introduced by previous discourse.
An example of a [+partitive] context is given in (5). The mention of three pens in the first sentence
establishes the existence presupposition, but not the uniqueness presupposition: a set of pens is
presupposed to exist, but it contains more than one element. Therefore, use of the with a singular NP in
the second sentence is infelicitous, and a pen (or an overt partitive, one of the pens) should be used
instead. However, Ko et al. (2006) found that L1-Korean L2-English learners frequently overused the
in contexts such as (5) (this finding was later replicated with L1-Serbo-Croatian learners of English, by
Ko, Perovic, Ionin and Wexler 2008; see also Kaneko 1996 for similar findings with L1-Japanese
(5) I have three pens in my bag. I will give you a pen / #the pen.
Ko et al. (2006) proposed that L2-English learners from article-less L1s fluctuate between the
semantic universals of definiteness, specificity, and partitivity, so that errors of the overuse can be tied
either to specificity (speaker knowledge without hearer knowledge) or partitivity (existence without
uniqueness). Similar findings of the associated with partitivity have been attested in L1-acquisition.
The studies of Maratsos (1976) as well as Schafer and de Villiers (2000) found that young
English-acquiring children overused the when a set of referents was established in the discourse. While
the original account of Maratsos (1976) attributed this error pattern to young children’s egocentricity,
Wexler (2003) proposed an alternative semantic explanation, and argued that young children optionally
associate the with the existence presupposition but not the uniqueness presupposition. The findings of
Ko et al. (2006, 2008) show that this association of the with existence exists in L2-acquisition as well.
2.3. Ways of establishing uniqueness
All of the studies discussed above focus on L2-English learners’ sensitivity to article semantics. A
separate, albeit closely related, research question is whether L2-English learners are sensitive to the
pragmatic conditions on article use, in particular, to the different pragmatic ways of establishing of
uniqueness. As shown by Hawkins (1978, 1991), the central notion of uniqueness can be established in
many different ways.
According to Hawkins (1978), definiteness is related to the identifiability of referents in
discourse. A definite referent is one that can be uniquely identified in the discourse. Hawkins (1978)
identifies a total of eight types of non-generic uses of the, some of which are exemplified below.
Multiple ways of establishing uniqueness
a. Anaphorically (through previous mention):
I bought a blue cup. The blue cup broke.
We went to a wedding. The bride was very tall.
Through entailment (via PPs, adjectives, or modifying clauses):
The movie that we are going to watch is the most popular one.
The roof of our house is leaking.
d. In the visible or immediate situation:
Pass me the salt, please.
Don’t go there, the dog will bite you.
e. Through world knowledge:
sun is shining.
We can see from (6) that while the uniqueness of the referent in (6a-c) is established through the
discourse context, the uniqueness of the referent is established exclusively through knowledge of the
world and/or the situation in (6d-e). The focus of our study is on uniqueness established through the
discourse context, so we will restrict our attention to the three subtypes of definites in (6a-c).
We furthermore examine whether L2-learners are able to distinguish the effects of previous
mention and association from the effects of uniqueness, i.e., whether they know that a rather than the
should be used in the contexts in (7). The contexts in (7) establish the presupposition of existence (the
relevant set of students or wedding guests is presupposed to exist), but not the presupposition of
uniqueness (since there are three students and multiple guests). Thus, those contexts correspond to the
partitive indefinite contexts tested by Ko et al. (2006, 2008); examining them in our study allows us to
investigate the effects of partitivity (set membership without uniqueness) in comprehension.
(7) a. Previous mention without uniqueness:
Carl had dinner with three students and two professors yesterday. A/#the
student brought the wine they drank.
Association without uniqueness:
I went to a wedding yesterday. A/#the guest gave a speech.
3. Research questions and hypotheses
Following the proposals of Ionin et al. (2004) and Ko et al. (2006), we hypothesize that
L2-English learners are able to access the semantic universal of definiteness, as stated in (8).
(8) Hypothesis 1
Since L2-English learners have access to the semantic universal of definiteness, with the
semantics of uniqueness, they should prefer the to a when uniqueness is established in the
discourse context, as in (6a-c).
We further hypothesize that L2-English learners may have difficulty computing uniqueness across
a variety of discourse contexts: while they should be sensitive to the central concept of uniqueness,
their ability to compute uniqueness may depend on the discourse context. This leads to the hypothesis
(9) Hypothesis 2
L2-English learners will not be equally successful at computing the uniqueness of the referent in
the context types in (6a-c), and will therefore show variation in their responses to these contexts,
in contrast to native English speakers.
Finally, we consider the findings of Ko et al. (2006, 2008) that in production, L2-English learners
from article-less L1s overuse the with partitive indefinites, in contexts similar to those in (7); we
therefore expect to see the same phenomenon in comprehension – i.e., for learners to associate the with
previous mention and association even when the uniqueness presupposition has not been met. This
leads to our third hypothesis, in (10).
L2-English learners will allow the in contexts involving previous mention and association even
when the uniqueness presupposition has not been established.
The participants in this study were 65 L1-Mandarin Chinese L2-English speakers and 27 adult
L1-English controls. Mandarin Chinese is an article-less language, and speakers of this language have
been shown to exhibit the effects of specificity in the production of English articles (Trenkic 2008)2;
this population has not been previously tested for effects of partitivity.
All of the L2-English learners were students at the University of Shanghai, aged between 18 to 23
(mean age 19.3). At the time of testing, all of the learners had studied English for at least 6 years, and
had achieved a score of between 111 and 130 on the National Matriculation English Test (NMET) in
China, which has the maximum possible score of 150. Their average self-rated proficiency score in
English was 3.2 on a scale from 1 (beginner) to 5 (advanced). Based on the NMET scores and the
self-ratings, these learners were classified as having intermediate L2-proficiency.3 They had been
exposed to English articles in their schooling, but were not expected to be fully target-like. None of the
learners had ever lived in an English speaking country.
The adult L1-English controls were all students at the University of Illinois at Urbana-Champaign,
and were aged 18 to 23 (mean age 20.1).
The test instrument was an Acceptability Judgment Task (AJT). The participants were instructed
to read pairs of sentences and judge whether the second sentence was acceptable as a continuation of
the first sentence, using a scale from 1 (unacceptable) to 4 (acceptable).
Altogether there were 88 sentence pairs, belonging to 11 different categories of 8 items each: 4
items containing the and 4 otherwise identical items containing a. Of the 11 categories, 5 categories
(40 items) are of relevance to the present study (see Yang and Ionin, in preparation, for results of some
other categories, which investigated effects of specificity). The relevant categories are illustrated in (11)
through (15), with two examples for each type, one with the and the other with a.4 Half of the items in
each category had the target DP in subject position, and half in object position. While the stimuli are
presented in pairs here, for ease of reference, they were presented individually in the actual test (e.g.,
(11a) and (11b) were presented at different points in the test).
2However, Trenkic (2008) gives a different explanation to these findings, arguing for the role of cognitive
strategies rather than semantic universals. See Ionin et al. (2009) for more discussion of the competing theories of
3The 65 participants were chosen from a larger sample of 93 based on their proficiency scores; the remaining 28
participants were either of low proficiency (NMET score below 111) or advanced proficiency (NMET score above
130), or else failed to report their proficiency. Focusing on the intermediate learners allows us to control for
4We originally included a sixth category, non-unique possessive definites, as in “Please don't sit in this chair. I
accidentally broke the leg of this chair yesterday”, to see whether L2-English learners would allow the in this
context. However, not only L2-English learners but also native English speakers allowed the in this context,
apparently treating it is as definite. We therefore do not include this category in our analysis. See Barker (2004)
for an account of the unusual behavior of these possessive constructions, which Barker calls “weak definites”.
(11) Category 1: uniqueness established through previous mention (target: the)
I saw a cat. I stroked the cat.
I saw a cat. I stroked a cat.
(12) Category 2: uniqueness established through association (target: the)
I went to a wedding yesterday. The bride was very beautiful.
I went to a wedding yesterday. A bride was very beautiful.
(13) Category 3: uniqueness established through the possession relation, with a genitive of-PP (target:
I bought a house. The roof of my house is grey.
I bought a house. A roof of my house is grey.
(14) Category 4: non-uniqueness with previous mention (target: a)
Dennis has many interesting books. His cousin borrowed the book from him yesterday.
Dennis has many interesting books. His cousin borrowed a book from him yesterday.
(15) Category 5: non-uniqueness with association (target: a)
I went to a wedding yesterday. The guest gave a speech.
I went to a wedding yesterday. A guest gave a speech.
Under Hypothesis 1 in (8), we predict that L2-English learners will rate sentences with the
higher than sentences with a in Categories 1, 2, and 3 above. In other words, we expect that
L2-learners will link the to the semantic concept of uniqueness. Under Hypothesis 2 in (9), we predict
that learners’ responses across these three categories will not be uniform, in contrast to native speakers’
responses. Finally, under Hypothesis 3 in (10), we predict that L2-English learners will rate sentences
with the highly in previous mention and association contexts even when uniqueness is not established,
in Categories 4 and 5 above, due to an association of the with the presupposition of existence.
The AJT as well as a language background questionnaire were placed online using a
university-based online survey system. Participants in both China and the U.S. were provided with the
url for the test, and completed the test on their own, on their home or school computers. The AJT was
prefaced with three examples, and the task instructions were repeated on each page of the task. The
task took about an hour for L2-English learners and half an hour for native English speakers to
complete. Both the L2-English learners and the L1-English controls received extra credit for their
participation in the study.
We now turn to the results. The group mean ratings across categories are summarized in (16);
recall that the rating scale was from 1 (unacceptable) to 4 (acceptable). The cells corresponding to the
target response for each category are highlighted.
5.1. Computing uniqueness
Let us first take a look at the L2-English learners’ and control group’s performance in Categories 1,
2, and 3. As (16) shows, L2-English learners rated sentences with the significantly higher than
sentences with a in Categories 1, 2, and 3, which provides evidence in support of Hypothesis 1 (in (8)),
according to which L2-learners are able to distinguish between the and a on the basis of uniqueness. At
the same time, (16) shows that L2-learners are not uniform across the three categories, consistent with
Hypothesis 2 (in (9)): the difference between the and a is greater for previous-mention (Category 1)
and association (Category 2) definites than for possessive definites (Category 3). In contrast, native
speakers made distinctions of equal magnitude across the three categories. This contrast is represented
pictorially in Figure 1, which gives the means as well as standard deviations for each category.
(16) Mean ratings by category (1=unacceptable, 4=acceptable)
L1-Chinese L2-English learners (N=65)
L1-English speakers (N=27)
mean rating of
mean rating of
mean rating of
mean rating of
sentences with the
sentences with a
sentences with the
sentences with a
1. unique through
2. unique through
3. unique through
4. non-unique with
5. non-unique with
*difference in mean rating of the vs. a is significant at p < .05
We conducted a repeated-measures ANOVA for each group, with article (a vs. the) and category
(1, 2, and 3) as within-subjects variables. In the case of L2-learners, the ANOVA yielded significant
effects of both article (F(1,64)=53, p<.001) and category (F(2,128)=34, p<.001), as well as a
significant interaction between the two (F(2,128)=23, p<.001). In contrast, the ANOVA on native
speaker results yielded significant effects of both article (F(1,26)=123, p<.001) and category
(F(2,52)=3.4, p<.05), but no interaction between the two (F(2,52)=.31, p=.74). While both groups
treated the three categories differently (as seen in Figure 1, some categories tended to receive higher
ratings than others, perhaps because of the felicity of individual sentences), native speakers made an
equal-magnitude distinction between the and a across the three categories. In contrast, L2-learners
made a much stronger distinction in previous-mention and association contexts than with possessives.
5.2. Uniqueness vs. existence
Next, we consider the participants’ performance on Categories 1 and 2 vs. 4 and 5, in order to see
whether L2-learners are able to distinguish the effects of previous mention/association (present in all
four categories) from the effect of uniqueness (categories 1 and 2 only). As (16) shows, although
L2-learners rated the higher than a in the presence of uniqueness (Categories 1 and 2), and rated a
higher in the absence of uniqueness (Categories 4 and 5), they also tended to rate the quite highly in
Categories 4 and 5: in the case of Category 5 in particular, learners failed to distinguish between the
and a, accepting both equally. For native English speakers, however, there is a strong distinction
between a and the in the absence of uniqueness.
Figure 2 represents this pictorially: Categories 1 and 2 are combined under the heading “unique”
while Categories 4 and 5 are combined under the heading “non-unique”, and the means and standard
deviations are reported. The data on Categories 4 and 5 support our Hypothesis 3 (in (10)) by showing
that L2-English learners overaccept the when existence has been established, but uniqueness has not.
Figure 1: comparing performance across Categories 1, 2 and 3.
uniqueness, with a
uniqueness, with the
uniqueness, with a
uniqueness, with the
entailment with 'of',
uniqueness, with a
entailment with 'of',
uniqueness, with the
Error bars: +/- 1.00 SD
Figure 2: comparing performance across Categories 1 and 2 vs. 4 and 5
Error bars: +/- 1.00 SD
The data reported in the previous section provides support for our hypotheses. We have observed
that L2-English learners are able to link the with the semantic concept of uniqueness. This suggests
that, as we predicted, L2-English learners have access to the semantic universal of definiteness, using
uniqueness to distinguish between the and a. At the same time, we have also observed that although
L2-English learners know that the semantic concept of uniqueness is at the core of definiteness, they
are not equally sensitive to different ways of establishing uniqueness. They over-rely on the
discourse-related factors of previous mention and association in their computation of uniqueness and
are not as good at using the possession relation to establish uniqueness. An open question at this point
is why this is the case: what makes previous mention and association stronger cues for uniqueness in
L2-English than the possession relationship? One possibility is that this is due to the greater syntactic
complexity of definite possessives: in order to compute uniqueness in a definite DP such as the roof of
my house, L2-learners need to understand the syntax of possession and the role of the genitive of-PP. In
contrast, in computing uniqueness for less syntactically complex DPs, such as the book or the bride,
learners need to consider the discourse situation only. Yet another possibility is that learners encounter
previous-mention and association contexts more frequently than possessive DPs in their input and/or
their classroom instruction, a possibility that requires further investigation. We leave the question open
for the time being.
Finally, we found that L2-learners overaccepted the in the context of previous mention or
association even when uniqueness is not established. This finding from comprehension is consistent
with Ko et al.’s (2006, 2008) findings from production data that L2-English learners overuse the when
the existence presupposition, but not the uniqueness presupposition, has been satisfied in the discourse.
Previous mention and association are both ways of establishing the existence presupposition, through
mention of a relevant set either explicitly (previous mention: three friends
a friend) or implicitly
(via association: the wedding
a wedding guest). Ko et al. similarly found that both explicit and
implicit set membership led to overuse of the with indefinites.
Our findings offer a possible explanation of why L2-learners sometimes associate the with
partitivity rather than definiteness: by over-relying on the discourse factors of previous mention and
association, learners may conclude that the encodes membership in a previously mentioned set (i.e.,
partitivity) rather than uniqueness. Support for this idea comes from our finding that when set
membership is not involved, as in the case of possessive definites (the roof of my house), L2-learners
show greater overuse of a. However, this finding contradicts the production findings of Ionin et al.
(2004) and Ko et al. (2006) that L2-learners produced the correctly with definites such as the owner of
this store, where no set membership is involved (no owners are previously mentioned, explicitly or
implicitly). It is not clear whether this difference between the studies is due to methodology
(production vs. comprehension), the learners’ native language, or the type of construction (note that in
the roof of my house, the roof belongs to the house, whereas in the owner of this store, the store
belongs to the owner, not the owner to the store). Thus, more work remains to be done.
In this paper, we have examined L2-English learners’ sensitivity to the semantics of uniqueness as
well as to the different ways of establishing uniqueness in the discourse. Our findings contribute to the
investigation of L2-English articles in two ways: first, by presenting comprehension data in a domain
that is typically studied only with respect to production; and second, by showing that L2-learners’
sensitivity to the semantics of definiteness is closely tied to the pragmatic context.
A number of open questions remain for further research. First, why do L2-learners over-rely on
previous mention and association in their computation of uniqueness, and have more difficulty with
possessive definites? Second, why do L2-learners overaccept the when the existence presupposition
has been satisfied but the uniqueness presupposition has not, and is this directly related to their
over-reliance on previous mention and association? Third, will there be any differences in learners’
production vs. comprehension of English articles, if both are tested with the same population of
learners? And finally, how do L2-learners overcome their over-reliance on discourse-related factors
and achieve target-like use and comprehension of English articles? More research is needed with
L2-English learners from different L1s, using a variety of methodologies as well as a variety of
Barker, Chris. 2004. Possessive weak definites. In J. Kim, Y. Lander and B. Partee, (eds.), Possessives and Beyond:
Semantics and Syntax. University of Massachusetts Occasional Papers in Linguistics, vol. 29. Amherst, MA:
Bickerton, Derek. 1981. Roots of Language. Ann Arbor, MI: Karoma.
Enç, Mürvet. 1991. The semantics of specificity. Linguistic Inquiry 22, 1-25.
Fodor, Janet and Ivan Sag. 1982. Referential and quantificational indefinites. Linguistics and Philosophy 5,
Fuli, Lotu Titi. 2007. Definiteness vs. specificity: an investigation into the terms used to describe articles in
Gagana Samoa. Master’s thesis, the University of Auckland.
Hawkins, John. 1978. Definiteness and Indefiniteness. London: Croom Helm.
Hawkins, John. 1991. On (in)definite articles: implicatures and (un)grammaticality prediction. Journal of
Linguistics 27, 405-442.
Roger Hawkins, Saleh Al-Eid, Ibrahim Almahboob, Panos Athanasopoulos, Rangsiya Chaengchenkit, James Hu,
Mohammad Rezai, Carol Jaensch, Yunju Jeon, Amy Jiang, Yan-kit Ingrid Leung, Keiko Matsunaga, Martha
Ortega, Ghisseh Sarko, Neal Snape, Kalinka Velasco-Zárate. 2006. Accounting for English article
interpretation by L2 speakers. In S. Foster-Cohen, M. Medved Krajnovic and J. Mihaljević Djigunović (eds.),
EUROSLA Yearbook 6, 7-25. Amsterdam: John Benjamins.
Heim, Irene. 1991. Articles and definiteness. Published in German as “Artikel und Definitheit,” in A. v. Stechow
& D. Wunderlich (eds.), Semantics: An international handbook of contemporary research. Berlin: de Gruyter.
Huebner, Thomas. 1983. A Longitudinal Analysis of the Acquisition of English. Ann Arbor: Karoma.
Ionin, Tania. 2003. Article semantics in second language acquisition. PhD dissertation, MIT
Ionin, Tania, Heejeong Ko and Ken Wexler. 2004. Article semantics in L2-acquisition: the role of specificity.
Language Acquisition 12, 3-69.
Ionin, Tania, Maria Luisa Zubizarreta and Salvador Bautista Maldonado. 2008. Sources of linguistic knowledge in
the second language acquisition of English articles. Lingua 118, 554-576.
Ionin, Tania, Maria Luisa Zubizarreta and Vadim Philippov. 2009. Acquisition of article semantics by child and
adult L2-English learners. Bilingualism: Language and Cognition 12, 337-361.
Kaneko, Yumina. 1996. Knowledge of the English article system in second language learning: To “the” or not to
“the”. Undergraduate thesis, Smith College.
Ko, Heejeong, Tania Ionin and Ken Wexler. 2006. Adult L2-learners lack the Maximality presupposition, too! In
K.U. Deen, J. Nomura, B. Schulz and B.D. Schwartz (eds.), Proceedings of the Inaugural Conference on
Generative Approaches to Language Acquisition–North America, Honolulu, HI, 171-182. Storrs, CT:
University of Connecticut Occasional Papers in Linguistics, vol. 4.
Ko, Heejeong, Alexandra Perovic, Tania Ionin and Ken Wexler. (2008). Semantic universals and variation in L2
article choice. In R. Slabakova, J. Rothman, P. Kempchinsky and E. Gavruseva (eds.), Proceedings of the 9th
Generative Approaches to Second Language Acquisition Conference, 118-129. Somerville, MA: Cascadilla
Leung, Yan-kit Ingrid. 2001. The Initial State of L3A: Full Transfer and Failed Features? In X. Bonch-Bruevich et
al. (eds.), The Past, Present and Future of Second Language Research: Selected Proceedings of the 2000
Second Language Research Forum, 55-75. Somerville, MA: Cascadilla Press.
Maratsos, Michael. 1976. The Use of Definite and Indefinite Reference in Young Children. Cambridge: Cambridge
Master, Peter. 1987. A cross-linguistic interlanguage analysis of the acquisition of the English article system. PhD
Parrish, Betsy. 1987. A new look at methodologies in the study of article acquisition for learners of ESL.
Language Learning 37, 361-383.
Schafer, Robin and Jill de Villiers. 2000. Imagining articles: What a and the can tell us about the emergence of DP.
In S. C. Howell et al. (eds.), BUCLD 24 Proceedings. Somerville, MA: Cascadilla Press.