Adult Accessibility to L2 Representational Features:
Evidence from the Spanish DP
Tiffany Judy,
1
Pedro Guijarro-Fuentes
2
and Jason Rothman
1
University of Iowa
1
and Plymouth University
2
1. Introduction
For over a decade the Minimalist Program (MP, Chomsky 1995, 2000, 2005) has revolutionized
both our understanding of the architecture of linguistic systems as well as how these systems come to
be acquired. According to current MP assumptions, parameter values obtain as a consequence of cross-
linguistic differences in formal feature and functional category composition. Universal Grammar (UG)
purportedly provides a superset of formal features and functional categories to all human beings. The
learning task of a child, therefore, involves choosing which functional categories and formal features
apply to his native language in light of the primary linguistic data at his disposal during the course of
grammatical development. Within such a conceptualization of the mental constitution of linguistic
knowledge, the instantiation (or lack thereof) of particular features is seen on the surface as cross-
linguistic typological differences, otherwise known as parametric differences. For example, languages
that lack an uninterpretable +wh feature will not have +wh movement, resulting in so-called wh-in situ
languages such as Chinese (and, as a result, principles such as Subjacency will not pertain to these
languages) while languages such as English whose input provides evidence of a +wh feature has, as a
syntactic consequence, obligatory +wh movement and must respect Subjacency conditions (see
Hawkins and Hattori 2006). The learning task of the adult L2 learner is not, a priori, unlike that of the
L1 child. That is to say, in order for an adult learner to converge on the target representation of any
given L2, he must have access to the features of UG that remained unspecified during the course of L1
acquisition. However, there are many more variables involved in adult language learning, not the least
of which is the fact that the initial state of acquisition is different than that of the child. It is reasonable
to suppose that L1 transfer occurs (consult Epstein, Flynn and Martohardjono 1996, 1998 for an
alternative view), and therefore, adult language acquisition does not begin with the tabula rasa with
which L1 acquisition commences.
Couched within MP notions, contemporary theorizing within generative second language
acquisition (L2A) has been able to offer more fine-grained proposals over the last decade that are in a
better position to adequately describe and explain the process and outcomes of adult language
acquisition. One of several central questions in the current research program of generative L2A
involves the continued role of UG after the so-called critical period and how answering this query can
shed some light on the asymmetrical patterns of pre- and post-critical period language acquisition.
Some have argued that UG, in the sense that UG contains a superset of all possible formal grammatical
features, is no longer available in the same way to adult language learners. Proponents of the
Representational Deficit Hypothesis (RDH) claim that [some] features not instantiated in the L1 are no
longer available to adult learners and thus L2 variability is represented within the narrow syntax (Beck
1998; Hawkins and Chan 1997; Franceschina 2001; Hawkins and Franceschina 2004; Hawkins and
Hattori 2006; Tsimpli and Mastropavlou 2008; Tsimpli and Dimitrakopoulou 2007).
1
If true, this
1
We note that there are differences between the theories we lump together here under the label RDH, inclusive of
the predictions in L2 behavior they make. For example, Beck’s (1998) model, which claims that functional
features remain inert in L2 grammars even when the L1 and L2 have the same value predicts L2 variability in all
© 2008 Tiffany Judy, Pedro Guijarro-Fuentes, and Jason Rothman. Selected Proceedings of the 2007 Second
Language Research Forum, ed. Melissa Bowles et al., 1-21. Somerville, MA: Cascadilla Proceedings Project.
means that L2 acquisition is destined to have representational deficits within the narrow syntax
whenever the target grammar requires the acquisition of new formal features, at least uninterpretable
ones. Accordingly, L2 variability is explained by claiming that an insurmountable deficit in syntactic
representation results in surface inconsistency. Since L2 learners can arrive at what appears on the
surface to be the same underlying representation via domain-general learning mechanisms, the fact that
L2 learners are able to seemingly do some things in their interlanguage that they should not be able to
do is not necessarily problematic for RDH approaches. For example, it is uncontroversial to claim that
Chinese learners of L2 English do form wh-questions with apparent movement; the claim, however, is
that the underlying syntax of Chinese L2 English wh-questions is necessarily different from that of L1
English. What should, however, be impossible for Chinese learners to demonstrate target knowledge of
(with negligible variation that is) are the conditions on movement related to Subjacency restrictions.
Interpretations of the available data on this are conflicting (see White 1989 for early debates and
Hawkins and Hattori 2006 versus Lardiere 2006 for more current data and analysis).
Full Access approaches (FA), which also embrace contemporary MP assumptions, contrast
sharply with RDH approaches in that they claim that adult learners have unabridged access to UG’s
full inventory, interpretable and uninterpretable features alike (e.g. White 2003; Schwartz and Sprouse
1996, 2000; Duffield and White 1999). Since they claim that new features can be acquired, parameter
resetting within the narrow syntax should be relatively unproblematic (although not inevitably so, cf.
Schwartz and Sprouse 1996). Support for this hypothesis comes from the so-called logical problem of
adult language learning whereby L2 acquirers have been shown to acquire poverty-of-the-stimulus
properties associated with the acquisition of new morphosyntactic features, which according to
Schwartz and Sprouse (2000) constitute unassailable evidence for adult UG accessibility (see
Slabakova 2006 for a review of much of this literature). Since RDH approaches predict that the
acquisition of L2 properties associated with a poverty of the stimulus that cannot be account for via L1
transfer should never occur, this evidence simultaneously constitutes counterevidence to RDH
predictions. FAs do not ignore L2 variability and optionality, but claim they do not ensue as a result of
inevitable deficits within the L2 narrow syntax. FAs claim that variability and optionality occur as a
result of any number of L1 transfer, general processing deficits and/or interface related problems for
adult L2A, related to, for example, the syntax-morphology interface (Hazdenar and Schwartz 1997;
Prévost and White 2000; Lardiere 1998, 2006); the syntax-prosodic/phonology interface (Goad and
White 2006); the syntax-pragmatics interface (Sorace 2000, 2004, 2005; Sorace and Filiaci 2006;
Valenzuela 2006).
Focusing on the narrow syntax of the Spanish DP, the present study tests the predictions of the
RDH approaches versus those of the FA approaches by examining the acquisition of gender and
number features of nouns, adjectives and determiners by L1 English learners of adult L2 Spanish.
Since these features determine DP internal word order (the adjective with respect to the noun),
determiner-noun-adjective overt morphological accord and are both interpretable (on the head noun)
and uninterpretable (on adjectives and determiners), the FDH predicts insurmountable variability for
grammatical gender as well as variability in adjective noun word order (see Section 2 for syntactic
details). Testing an intermediate and advanced group of adult English learners of L2 Spanish, we
demonstrate across two linguistic tasks that L2 learners can and do acquire new features as evidenced
by their knowledge of grammatical gender and the semantic construals of adjective placement in L2
Spanish. Although space limitations do not afford us the opportunity to discuss all the implications of
this,
2
task 2 demonstrates that L2 learners at the advanced level do acquire semantic nuances related to
domains regardless of L1/L2 pairings while Hawkins and Chan (1997), for example, would only predict inevitable
variability due to representational differences where the L1 and L2 have different feature compositions. Finally,
Hawkins and Hattori (2006) and Tsimpli and Dimitrakopoulou (2007) predict representational differences only
when the L2 requires the acquisition of new uninterpretable features. Despite these important differences, we are
able to lump these theories together under one general approach in light of what we investigate, namely, gender
and number features (and ensuing word order) since these features are both interpretable and uninterpretable in
Spanish depending on what they are associated with.
2
We do not intend to ignore other contemporary debates such as the notion of interface vulnerability as being the
central cause of L2 variability (e.g. Belletti, Bennati and Sorace 2007; Sorace 2000, 2005; Sorace and Filiaci
2
adjective position in Spanish, showing clear knowledge that the default position is postnominal and
that the prenominal position evokes semantic differences that are not easily induced from input alone
(see Anderson 2001, 2007 a and b, 2008 for similar results from L2 French). Overall, the data we
present are only consistent with FA approaches and provide counterevidence to RDH approaches.
The remainder of this article is set up in the following manner. Section 2 discusses the syntax of
grammatical gender and number in Spanish and how the acquisition of relevant features comes to bear
on adjective placement. Additionally, we cursorily discuss some of the semantic differences associated
with adjective placement in Spanish as it comes to bear on task 2. Section 3 presents a review of
previous studies and background literature. Sections 4 and 5 present the methodology and results while
Section 6 presents a discussion of our findings. Finally, Section 7 presents a conclusion of the present
study.
2. Syntax and semantics of the Spanish DP
2.1 The syntax of gender, number and resulting word order within the DP
In this section, we present a concise description of the structural aspects of the Spanish determiner
phrase (DP) compared to the structure of the English DP. Spanish, like the majority of Romance
languages, has both subject-verb agreement and nominal agreement. The verb, inflected for Φ-
features, agrees in person and number with the subject DP and the different components of the Spanish
DP (i.e. determiners, nouns and adjectives agree with each other in both gender and number) as
illustrated in (1):
(1) La niña mexicana juega con
The.FEM.SING. girl.FEM.SING. mexican.FEM.SING. play.PRES.3SG. with
los perros.
the.MASC.PLRL. dogs.MASC.PLRL.
‘The Mexican girl plays with the dogs.’
At the very descriptive level, gender and number in Spanish, masculine or feminine, singular or
plural, are both morphologically marked on most nouns (Bernstein 1993; Harris 1991, 1995). As
illustrated in (2), there is gender and number agreement of the head noun with all the constituents of
the noun phrase including determiners and adjectives, which are also inflected for gender and number.
(2)
Conversely, English has limited nominal agreement. While English lacks grammatical gender,
English has number inflection, although this is limited to the head noun. In general, English adjectives,
2006; White to appear). In fact, the larger project from which the data we present here is a subset focuses on the
syntax-semantics interface and related debates. Given space limitations, however, we discuss our data here in light
of the current debate on UG accessibility; that is, in terms of whether or not adults continue to have access to all or
only some features from UG and how this relates to underlying L2 syntactic representations.
3
unlike Spanish adjectives which usually appear postnominally, appear almost exclusively in the
prenominal position, a point to which we return in greater detail below.
Within the generative tradition, Abney (1987) put forward the DP Hypothesis, which essentially
established a parallelism between the syntactic structure of noun phrases and the one of the sentence.
The so-called DP Hypothesis asserts that determiners are the head of a functional projection, the
Determiner Phrase (DP), which selects Noun Phrases (NP) as complements. The function of the DP is
to attach reference to the complement noun phrase. However, a more fine-grained analysis (e.g.
Bernstein 1993; Picallo 1991; Carstens 1991; Ritter 1993; Valois 1991; Cinque 1994) of the DP
structure proposes additional functional categories between the DP and NP. Simplifying a bit for the
purposes of the present study, we follow Bernstein’s (1993) analysis according to which the functional
categories NumP and WMP (Word Marker Phrase) are located between DP and NP ([DP D [NumP
Num [WMP WM [NP N]]]) where formal gender ([masc/fem]) and number ([±plural]) features are
checked, valued and deleted. Features in the DP-internal functional categories are responsible for the
differences or similarities in nominal word order amongst languages. In Romance languages nominal
movement (noun raising) is obligatory whereas in Germanic languages, such as English, the noun
remains in situ since there is no feature that attracts the noun to raise.
3
If an adjective is present in a specifier position of the NP, noun movement to the head of NumP
results in the head noun moving past the adjective, yielding the canonical [N Adj] word order of
Spanish (Carstens 1991; Zagona 2002). In Germanic languages, on the other hand feature checking is
satisfied without movement yielding the opposite word order [Adj N] at spell out. Within this
structural configuration, number features within the nominal domain are valued and deleted, and the
same applies to gender agreement (Picallo 1991; Bernstein 1993) as shown in the structure in (3).
4
3
For our purposes herein, what exactly attracts the noun to raise in Spanish is inconsequential as it is observable
that it does in Romance language, but does not in English. Older accounts capture this difference by claiming
number features are strong in Spanish, but weak in English. Such a notion is not compatible with current
assumptions, and so it is argued that number features are checked via AGREE in English without raising while
such feature checking happens via overt movement in Spanish, presumably due to an additional feature (of the
EPP-type) that attracts the noun to raise.
4
However, Masullo and Depiante (2003) claim that gender is an inherent feature on nouns and only number gets
checked in a higher functional category and consequently WMP does not exist (or in Picallo's (1991) terms
Gen(der)P).
4
(3) DP
Spec D’
D NumP
(AP) NumP
A’ Num’
pobre Num WMP
(unfortunate)
hombre
i [μ num]
WM’
WM NP
t
i [μ gen]
AP NP
A’ N’
pobre N
o
(not rich)
t
i
Given the L2 acquisition theories under investigation in the present article (see Section 1), it is
crucial to understand how grammatical gender and number are accounted for within a current
minimalist account. Spanish has interpretable gender and number features on nouns whereas these
features are uninterpretable on determiners and adjectives. In Spanish, the uninterpretable features
enter into an agreement relationship with the interpretable features of the head noun whereby they are
valued and deleted during the course of the derivation, that is, upon obligatory nominal movement to
Word Marker Phrase (WMP) and NumP respectively. Since there is no noun raising in English, which
thus accounts for its stringent determiner-adjective-noun word order, English learners of L2 Spanish
must acquire whatever features result in noun raising. Although English has number features these
features do not result in movement. Additionally, English entirely lacks grammatical gender features
and the functional category where they are checked. Therefore, English L2 learners must acquire a
new functional category, namely WMP where gender features are checked, as well as the independent
features associated with WMP and NumP that require raising in order to converge on the target syntax
of the Spanish DP.
However, this linguistic analysis of the syntactic properties of adjectives is not sufficient in
describing the word-order possibilities in Spanish (and other Romance languages; see Zagona 2002).
This is the task that we undertake in the next section. Specifically, placement of adjectives in the
Spanish DP shows at least a three way distinction that correlates with the semantic properties of the
adjectives.
2.2 Three types of Spanish adjectives
On the basis of the analysis presented in the previous section, we can immediately see two very
distinctive groups of languages: a group of languages whose adjectives are prenominal and another
group whose adjective are (mainly) postnominal. Furthermore, in Spanish, unlike English which has
but one possibility, the position in which an adjective can appear depends on both the nature of the
adjective’s semantic interpretation in a given NP as well as the adjective itself (namely, the
5
subcategory of adjective). Adopting Zagona’s analysis (2002), there is a first group of adjectives that
are strictly prenominal such as mero ‘mere’ and futuro ‘future’ (namely, temporal modifiers) compare
(4a) and (4b):
(4) a. Un mero soldado
‘A mere soldier’
b. *Un soldado mero’
‘A mere soldier’
The second type of adjectives, so-called ethnic and relational adjectives (namely, mathematical,
theatrical and so on) which convey relations of various types, such as origin and material composition
always follow the noun in a postnominal position compare (5a) and (5b).
(5) a. Una camisa mexicana
‘A Mexican shirt’
b. *Una mexicana camisa
‘A Mexican shirt’
A third class of adjectives, the so-called qualitative adjectives, such as viejo ‘old,’ pobre ‘poor’
and so on can be either prenominal or postnominal. In prenominal position they are appositive (non-
restrictive), but in postnominal position they are restrictive (Jackendoff 1977; Luján 1980; Alexiadou
2001; Zagona 2002). This type of adjective displays additional semantic construals depending on
whether they appear to the left or right of the noun: such as focus, emphasis, subjectivity, and
sometimes causing a change in meaning when they merge in different positions within the DP
structure (compare (6a) and (6b)):
(6) a. Un hombre pobre
‘A poor (not rich) man’ (restrictive interpretation)
b. Un pobre hombre
‘A poor (unfortunate) man (appositive interpretation)
In order to explain the different collocation possibilities of Spanish adjectives, we follow
Bernstein’s (1993) analysis that claims that restrictive/intersective adjectives are essentially
prenominal in Spanish, but that their postnominal position is derived from noun movement over the
adjective as shown in (7):
6
(7) DP
Spec D’
D NumP
(AP) NumP
A’ Num’
A Num WMP
X
i
WM’
WM NP
t
i
AP NP
A’ N’
A N
o
t
i
Following Bernstein (among others, Cinque 1994; Picallo 1991), the head noun moves (as
represented by X in (7)) from N to the head of WMP and then to NumP.
5
Following this logic,
appositive/non-intersective adjectives are merged higher in the DP, either in Spec NumP or some other
functional projection higher above NumP, thus providing the possibility of Adj N word order even
after noun raising. In summary, Bernstein (1993) asserts that different adjectives emerge in different
structural positions: some of them being APs which adjoin to various XPs, whereas others being
similar to functional projections in which the head Adj takes a complement. These different structural
positions are related to the specific semantic characteristics of the adjectives as shown through
examples (4), (5) and (6) above. Because we are dealing with [Det N Adj] and [Det Adj N] orders, the
structure given in (7) reconciles these structures.
6
5
Within the Minimalist Framework (Chomsky 1995), the attribute adjectives are endowed with lexical categories
which are checked by the merger the adjectives in the nominal structure. That is, the merger of adjectives results
in the checking of [+Interpretable] features (in a way similar to θ-features). Thus, these features are checked
before Spell-Out, but they cannot be erased and they are interpretable.
6
English, on the other hand, exhibits covert movement at Logical Form (LF) only. More recent analyses,
however, induce the placement of the adjectives in terms of the existence of some extra categories besides the NP
and the DP, for example, [nP] and [FocP] which will be positioned above the nP (e.g. Demonte 2008). Taking the
English sentence ‘the poor man’ as an illustration for our argumentation, its structure would correspond to the one
given in (i):
(i) [
DP
a ] [
NumP
[
AP
poor] [
Num
+singular]] [
nP
[
AP
poor] [
N
man]]
↓ ↓
‘unfortunate’ ‘not rich’
Thus, a phrase like ‘a poor man’ is ambiguous because poor can be related underlyingly either to the position in
NumP (which is taken as non-intersective) or to the position in nP (which is taken as intersective). In Spanish, on
the other hand, N raises to NumP, so that any prenominal adjective must be non-intersective, and any postnominal
adjective must be intersective (see Demonte 2001, 2008). Thus, in Spanish the syntactic position of the adjectives
7
2.3 Summary and the task of the L2 learner
From a learnability point of view, taking into account the pure syntax (i.e. structural representation
of the DP and abstract feature(s) in NumP and WMP, see 2.1), L2 learners of Spanish must learn that
adjectives in Spanish are mostly postnominal, that nouns have grammatical gender and number
features and that there are number and gender agreement between nouns, determiners and adjectives
(which requires the acquisition of new uninterpretable features). On the other hand, L2 learners need to
acquire the semantic features of DPs in Spanish which deal with adjective placement. We have shown
that the distinctive semantic interpretations of the adjectives in Spanish are captured by different
syntactic word orders. We have outlined a syntactic and semantic analysis for different adjective
classes in Spanish. The placement of adjectives within the NP mirrors both their sub-categorization
and their semantic interpretations within the NP. In particular, what L2 learners of Spanish need to
acquire is (a) the fact that Spanish has grammatical gender (which does not exist in English), (b) that
Spanish also has grammatical number (which exists in English) and (c) the order of the noun and
modifier(s), which syntactically varies in Spanish compared to English. Table 1 highlights the aspects
of the nominal system that will need to be learned by English learners of L2 Spanish.
Table 1: Spanish and English nominals
Spanish English
Overt gender on the noun
(interpretable gender features)
+ -
Overt number on the noun
(interpretable gender features)
+ +
Overt gender concord
(uninterpretable gender features)
+ -
Overt/covert number concord
(uninterpretable number features)
+ +
Prenominal adjectives
(via no movement in English; after two types of movement in Spanish)
+ +
Postnominal adjectives
(instantiation of WMP and obligatory noun raising in Spanish only)
+ -
Recall that some of these features are underdetermined in the input and not syntactically
represented in the learners’ native language (English). In general terms, English learners of L2 Spanish
receive instruction that adjectives in Spanish are generally postnominal (e.g. adjectives expressing
origin, material composition such as estudiante francés ‘French student’ or mesa italiana ‘Italian
table’), but there are a few exceptions with some adjectives appearing prenominally (such as mero
soldado ‘mere soldier’) and still other adjectives whose meaning change or depend on their position
(e.g. largely qualitative adjectives such as viejo ‘old,’ pobre ‘poor/impoverished’ and so on). So,
explicit instruction received in the classroom does not always fully capture this variability in adjective
placement. Crucially, what English learners are not instructed on are the mutually exclusive
appositive/restrictive semantic interpretations that are strictly enforced by (fall out from) their syntactic
position. In other words, what English learners will have to ‘unlearn’ is the fact that prenominal
adjectives in Spanish are not ambiguous in meaning as they are in English precisely because adjectives
can only be overtly and not underlying linked to their syntactic position. And so, if these L2 learners
show knowledge of the semantic properties of qualitative adjective to a native-like degree depending
are conditioned by their semantic interpretations. Following Demonte (2001, 2008), the semantic interpretations
illustrated in examples (4)-(6) above are associated to three syntactic structures: adjectives with a restrictive
interpretation (that is, they refer to the whole set of properties on noun) occupy a postnominal position, adjectives
with a non-restrictive interpretation (that is, they refer to a subset of a class of objects denoted by the noun) tend to
appear prenominally and F in focus position. We point this out to be thorough, but such details remain outside the
scope of the present research and do not come to bear on what we investigate herein.
8
on their syntactic position this could be used as robust evidence to suggest that they have converged on
the target representation for the Spanish DP.
3. Previous Studies
7
There have been several studies on adult L2 acquisition of new functional features related to the
DP. Of particular interest are those studies that adopt the RDH, claiming that persisting variability in
morphological marking is evidence of “L2 learner’s use of different sets of representational primitives
to ‘break down’ the L2 input” (Franceschina 2005: 4), namely, accessibility to linguistic features
instantiated within their L1 only (e.g. Hawkins 1998, 2001; Franceschina 2001, 2005; Hawkins and
Franceschina 2004). Franceschina (2001, 2005) reports data from a case study on Martin, an English
learner of L2 Spanish who has supposedly reached a steady state grammar for L2 Spanish. She
reported that he was nearly natively accurate with gender assignment on nouns (where these features
are interpretable; see Hawkins and Hattori 2006 for implications); however, he demonstrated
considerable variability with gender assignment on determiners and adjectives, favoring a masculine
default. She also reported that Martin was accurate to a native level with number assignment across
nouns, determiners and adjectives. Since English has grammatical number, but not grammatical
gender she interprets this difference in morphological performance as evidence of the fact that new L2
features cannot be acquired. In doing so, Franceschina (2005) joins others (e.g. Hawkins 1998, 2001;
Franceschina and Hawkins (2004)) in arguing that the mental representation of grammatical gender for
L1 English adult learners of Spanish is necessarily different than native Spanish speakers. That is, such
evidence is interpreted by them as prima facie support for the notion that L1/L2 disparities result from
differences in accessibility to representation resources after the Critical Period.
Other research such as Bruhn de Garavito and White (2002), Fernández (1999), White,
Valenzuela, Koslowska-McGregor and Leung (2004) and Cabrelli, Iverson, Judy and Rothman (2008)
question the tenability of the claims made by RDH approaches using evidence within this domain.
Fernández (1999) demonstrates that L2 learners have less difficulty with gender assignment with
determiners than they do with adjectives. Bruhn de Garavito and White (2002) tested the acquisition of
grammatical gender in French L2 learners of Spanish in order to test between Beck’s (1998)
hypothesis of featureless values, irrespective of the L1, and Full Access. According to Beck’s
proposal, both English and French learners should demonstrate variability with gender assignment in
L2 Spanish despite the fact that the French learners’ L1 has grammatical gender features. Since these
features are argued to remain inert indefinitely, variability/optionality is not predicted to be ever
completely reached, even at the steady state, for either set of L2 Spanish learners. Bruhn de Garavito
and White (2002) showed that intermediate French learners had difficulty in assigning gender,
especially to adjectives (less so with determiners), similar to Fernández’s data for English L2 Spanish.
However, they demonstrated that this difficulty virtually disappears by the advanced stages. These
results suggest that Beck’s proposal and Hawkins and colleagues’ proposals are untenable since,
according to Beck, variability should remain and, according to Hawkins and colleagues, the French
learners should not have these problems, in that they should clearly not pattern like English learners of
L2 Spanish. As a result, Bruhn de Garavito and White (2002) conclude that the problem is a
morphological one, thus explaining why similar patterns between English and French learners of L2
Spanish obtain.
Further corroborating the claims of Bruhn de Garavito and White (2002), work by White et al.
(2004) and Cabrelli et al. (2008) have demonstrated that grammatical gender is acquired in L2
acquisition (and is a facilitative factor at the initial state of L3 see Cabrelli et al. (2008)). Both studies
show that N-drop is uncomplicatedly acquired by English speakers of non-native Spanish. Even under
7
Although it is our intention to review and include some of the most current and relevant studies on the
acquisition of DP properties in this section, we are aware of other types of work done on the topic using different
methodological techniques such as Event Related Potential (ERP) (e.g. Frenck-Mestre, McLaughlin, Osterhout
and Foucart (in press) and references therein), but due to space limitations we will not review that type of research
here.
9
the most updated RDH proposals, which claim a post-critical period failure of uninterpretable features
only (e.g. Hawkins 2005; Hawkins and Hattori 2006), this type of nominal ellipsis should not be
acquired by English learners of Spanish since it is contingent on gender features from adjectives and
determiners, which are uninterpretable.
There are few generative studies on the L2 acquisition of adjective placement and adjectival
semantics. Anderson (2001, 2007a, 2008), however, has demonstrated that L1 English adult learners of
L2 French do come to acquire proper canonical word order (determiner-noun-adjective) as well as the
semantics difference between the canonical word order and prenominal adjective placement in French.
However, Anderson (2007b) also demonstrates that an incongruity between pedagogical rules with
classroom input and texts written for and by native French speakers can result in particular non-target-
like performance even with highly proficient L1 English-speaking learners
in acceptability judgements
of contextualized French
sentences, a point upon which we return as we analyse the present data for L2
Spanish.
4. The Study
4.1 Participants
The data come from a corpus of participants from three distinct groups. The control group
consisted of 12 native speakers (NS) of Spanish. Their ages ranged from 25-54 and they were from
various Spanish-speaking countries. The advanced speaker group (AS) consisted of 12 participants that
had studied Spanish at the university level for a minimum of 5 years (8 years minimum of formal
instruction ranging from high school to university level) and had also studied abroad in a Spanish-
speaking country. Their ages ranged from 21-31. Finally, the intermediate speaker group (IS) consisted
of 29 undergraduate students that had completed the General Education program requirements (which
consist of a series of four university-level classes) at University of Iowa (or its equivalent). The age
range for this group was 19-23.
4.2 Methodology
Besides a linguistic background questionnaire, all participants completed two linguistic tasks: a
Grammaticality Judgment/Correction Task (GJCT) and a Context-based Collocation Task (CBCT).
The GJCT tested for knowledge of gender agreement between determiners and nouns and nouns and
adjectives. The purpose of the CBCT was to test for semantic knowledge of felicitous prenominal and
postnominal adjectives. It should be mentioned that vocabulary was controlled for and there was no
time limit. Participants were also instructed to ask the investigator if the directions or vocabulary were
unclear. This was done in an attempt to reduce error based on misunderstanding. Given space
limitations, we only provide examples of and report data on the relevant empirical sentence types,
which is to say, not the fillers.
4.2.1 Grammaticality Judgment/Correction (GJCT)
The GJCT consisted of a total of 80 tokens. Tokens containing ungrammatical determiner-noun
gender agreement (n=5) and ungrammatical determiner-noun number agreement (n=5) were
counterbalanced by tokens containing grammatical determiner-noun gender and number agreement
(n=5 each). Tokens containing ungrammatical noun-adjective gender agreement (n=5) and
ungrammatical noun-adjective number agreement (n=5) were counterbalanced by grammatical noun-
adjective gender and number agreement (n=5 each). It is important to note that the number of feminine
vs. masculine and singular vs. plural tokens were controlled for across all token types. Ungrammatical
prenominal (n=5) and postnominal adjective tokens (n=5) were counterbalanced by grammatical
prenominal and postnominal adjective tokens (n=5 each). Finally, fillers (n=20) were included to
disguise the purpose of the task. Participants were instructed to read each sentence and correct it if they
judged it to be ungrammatical. Examples (8a) and (8b) show grammatical noun-adjective agreement
10
and grammatical determiner-noun agreement, respectively. Examples (8c) and (8d) show
ungrammatical noun-adjective gender agreement and ungrammatical determiner-noun gender
agreement, respectively. The anticipated correction for (8c) is ‘famosos’ to agree with the masculine,
plural noun while for (8d) the anticipated correction is ‘la ropa.’ Example (8e) shows ungrammatical
noun-adjective number agreement. The anticipated correction is ‘su vida académica.’ Example (8f)
shows ungrammatical determiner-noun number agreement where the anticipated correction is ‘la
pulsera.’
(8) Knowledge of gender and number agreement
a. Ellas prefieren comida fresca. (grammatical Noun-Adj gender/number)
They prefer food FEM.SING. fresh FEM.SING.
‘They prefer fresh food.’
b. El chico siempre lee aquí. (grammatical Det-Noun gender/number agreement)
The MASC.SING. boy MASC.SING. always reads here
‘The boy always reads here.’
c. *Mis tíos de Los Ángeles son famosas. (ungrammatical Noun-Adj gender agreement – long
distance)
My uncles MASC.PLRL. from L.A. are famous FEM.PLRL.
‘My uncles from L.A. are famous.’
d. *Iván necesita lavar el ropa. (ungrammatical Det-Noun gender agreement)
Iván needs to wash the MASC.SING. clothes FEM.SING.
‘Iván needs to wash the clothes.’
e. *Su vida académicas es muy difícil. (ungrammatical Noun-Adj number agreement)
His life FEM.SING. academic FEM.PLRL.is very difficult
‘His academic life is very difficult.’
f. *Felipe me compró las pulsera. (ungrammatical Det-Noun number agreement)
Felipe me bought the FEM.PLRL. bracelet FEM.SING.
‘Felipe bought me the bracelet.’
As stated above, this task also tested for knowledge of syntactic positions of Spanish adjectives.
Examples (9a) and (9b) are examples of grammatical prenominal and postnominal adjectives,
respectively. Example (9c) is an example of an ungrammatical prenominal adjective. The noun and the
adjective should be inverted. Example (9d) is an example of an ungrammatical postnominal adjective.
The noun and the adjective should be inverted. The adjectives used in this portion of the GJCT were of
the type that are strictly prenominal or strictly postnominal, corresponding to the types of adjectives
exemplified in (4) and (5) of section 2.2.
(9) Knowledge of syntactic positions of adjectives
a. La policía está investigando el supuesto asesinato. (grammatical prenominal adjective)
‘The police are investigating the alleged assassination.’
b. Quiero comprar los pantalones azules. (grammatical postnominal adjective)
‘I want to buy the blue pants.’
c. *Ellos compraron una japonesa mesa. (ungrammatical postnominal adjective in
prenominal position)
‘They bought a Japanese table.’
d. *Juan es el agresor presunto del crimen. (ungrammatical prenominal adjective in
postnominal position)
‘Juan is the alleged aggressor of the crime.’
11
4.2.2 Context-based Collocation Task (CBCT)
The Context-based Collocation Task was designed to test for semantic knowledge of prenominal
and postnominal adjectives by means of a pseudo-production task (n=10).
8
Participants were instructed
to read a short contextualized situation and write the adjective found in bold at the end of the sentence
in the correct position based on the context. All adjectives used in this task could be either prenominal
or postnominal. The collocation, then, depended directly and entirely on the context. There were five
prenominal tokens and five postnominal tokens. The context in example (10) calls for a prenominal
adjective whereas that of example (11) calls for a postnominal adjective.
(10) Prenominal Adjective
a. En general, la dinastía Ming era una de las más fuertes y exitosas en la historia del
mundo. Los
_______________ Ming______________ conquistaron muchas tierras. (aventurero)
‘In general, the Ming dynasty was one of the strongest and most successful dynasties in the
history of the world. The adventurous Ming conquered many lands.’
(11) Postnominal Adjective
a. Hay muchos padres en la audiencia del partido. Hay niños que juegan muy bien, pero los
que juegan mal son una vergüenza a sus padres. Los ______________padres_____________
aplauden y gritan por sus hijos. (orgulloso)
‘There are many parents in the audience at the game. There are children that play very well,
but those that play badly are an embarrassment to their parents. The proud parents applaud
and yell for their children.’
5. Results
This section presents the empirical results of the study. It is divided into two sections, one
corresponding to each of the two tasks. The statistical analysis was conducted in the following manner.
A one-way ANOVA was computed when comparing all three groups to each other. A two-sample t-
test was conducted when making an intergroup comparison (for example when comparing the NS
group to the AS group). The alpha was set at .05 for all tests conducted. Additionally, it is important to
note that post hoc tests were only done when the ANOVA indicated that it was necessary, so as to
avoid a type I error (this methodology is known as Fisher’s LSD).
5.1 Quantitative Analysis: GJCT
The first part of the GJCT tested for the acquisition of L2 gender features, which are
uninterpretable on determiners and adjectives and interpretable on nouns. Figure 1 below shows the
average number of tokens accepted (n=10) by each of the three groups for the following types of
sentences: (a) grammatical noun-adjective agreement, (b) ungrammatical noun-adjective agreement,
(c) grammatical determiner-noun agreement and (d) ungrammatical determiner-noun agreement. See
Section 4.2.1 for examples of each token.
8
Participants were also required to complete a third task (a Semantic Interpretation task) that is not reported here
due to space limitations. Included in this task was a mixture of prenominal, postnominal and predicate adjectives.
12
Figure 1: GJCT results
Note: NA-good = grammatical noun-adjective agreement; NA-bad = ungrammatical noun-adjective agreement;
DN-good = grammatical determiner-noun agreement; DN-bad = ungrammatical determiner-noun agreement
As is observable, the AS group performed quite similarly to the NS group with all four token
types. The same holds for the IS group performance with respect to the grammatical tokens; however,
the IS group accepted more ungrammatical noun-adjective tokens and more ungrammatical
determiner-noun tokens than the NS group. In fact, the IS group accepted more tokens, whether
grammatical or ungrammatical, than the AS group and the NS group did on average. In order to
determine if these differences were significant, further statistical analyses were conducted. First, a one-
way ANOVA that compared all three groups to each other was computed. In the case of noun-
adjective agreement, a statistically significant difference was found. Therefore, a two-sample t-test was
conducted to determine where the differences were. As can be seen from Table 2, upon comparing the
AS group to the NS group, it was found that the AS group did not differ statistically from the NS
group for either of the two token types in question. However, the IS group results were statistically
significantly different in regards to the grammatical and ungrammatical noun-adjective tokens. The IS
group also differed significantly from the AS group in regards to the ungrammatical noun-adjective
tokens. These differences are highlighted.
Table 2: GJCT (intergroup comparison)
N
A-
g
ood
N
A-
b
ad D
N
-
g
ood D
N
-
b
ad
t p df t p df t p df t p df
ANOVA 4.11 0.022 2 8.98 <0.00 2 0.73 0.485 2 0.78 0.463 2
NS v. AS 1.91 0.082 0.45 0.68 0.501 21 * * * * * *
NS v. IS 4.05 <0.001 0.78 4.66 <0.00 37 * * * * * *
AS v. IS 1.72 0.094 34 3.96 <0.00 38 * * * * * *
Note: NA-good = grammatical noun-adjective agreement; NA-bad = ungrammatical noun-adjective agreement;
DN-good = grammatical determiner-noun agreement; DN-bad = ungrammatical determiner-noun agreement
The second part of the GJCT tested for knowledge of syntactic positions of Spanish adjectives.
Recall that while there is only one position for adjectives in English (prenominal), adjectives in
Spanish can be either prenominal or postnominal depending on their type and their meaning within a
given context (see Section 2.2). Figure 2 below shows the average number of tokens accepted (n=5) by
the three groups for the following sentence types: (a) grammatical prenominal adjectives, (b)
13
ungrammatical prenominal adjectives, (c) grammatical postnominal adjectives and (d) ungrammatical
postnominal adjectives. Examples of each token can be found in Section 4.2.1.
Figure 2: GJCT results
Note: PreAdj-good = grammatical prenominal adjective; PreAdj-bad = ungrammatical prenominal adjective;
PosAdj-good = grammatical postnominal adjective; PosAdj-bad = ungrammatical postnominal adjective
From Figure 2, we observe that the NS group distinguished between the grammatical and
ungrammatical tokens. This can be seen by their high group average of acceptance for the grammatical
tokens and their low group average of acceptance for the ungrammatical tokens. The AS distinguished
between the two to a slightly lesser extent and the IS group made virtually no distinction between
grammatical and ungrammatical prenominal and postnominal adjectives. To determine if the apparent
differences were in fact significant, the same battery of statistical analyses discussed above were
conducted. The one-way ANOVA showed that there were in fact statistically significant differences
among the three groups. Thus, a two-sample t-test was conducted to determine where the differences
were. The results of this test showed that the AS group performance differed from that of the NS group
for two token types: grammatical prenominal adjectives and ungrammatical postnominal adjectives.
These results are not entirely surprising since all of the participants in the advanced speaker group
were instructed learners. One possible explanation for this result is the fact that instructed L2 learners
are traditionally taught that these two structures (grammatical prenominal and ungrammatical
postnominal adjectives) are ungrammatical (that is, that some adjectives can only appear prenominally,
whereas others can only appear postnominally). In these cases, pedagogical teachings may have
interfered with the participants’ assessment of the tokens. In contrast to the AS group, statistically
significant differences were found between the NS group and the IS group in all three categories in
question. This information is summarized below in Table 3 with the significant differences
highlighted.
14
Table 3: GJCT (intergroup comparison)
PreAd
j
-
g
ood PreAd
j
-
b
ad PosAd
j
-
g
ood PosAd
j
-
b
ad
t p df t p df t p df t p df
ANOVA 7.67 0.001 2 13.9 <0.001 2 0.46 0.63 2 41.75 <0.001 2
NS v. 2.78 0.018 11 1.48 0.166 0.38 * * * 3.79 0.002 14
NS v. IS 5.98 <0.001 32 6.22 <0.001 1.58 * * * 12.63 <0.001 28
AS v. IS 1.34 0.192 23 5.28 <0.001 34 * * * 2.98 0.010 14
Note: PreAdj-good = grammatical prenominal adjective; PreAdj-bad = ungrammatical prenominal adjective;
PosAdj-good = grammatical postnominal adjective; PosAdj-bad = ungrammatical postnominal adjective
While some of the AS and IS group averages were similar to those of the NS group, it was crucial
to determine if the L2 learner groups differentiated between grammatical and ungrammatical tokens.
This was done by comparing the average number of tokens accepted for each of two corresponding
structures for all three groups. For example, the average number of grammatical noun-adjective tokens
accepted by the AS group was compared to the average number of ungrammatical noun-adjective
tokens accepted by the AS group to determine if the AS group differentiated between grammatical and
ungrammatical gender on adjectives. A paired t-test was used to determine this. As can be seen from
Table 4, all three groups differentiated between the grammatical and ungrammatical structures in task
1.
Table 4: GJCT (intragroup comparison)
NA-good v. NA-bad DN-good v. DN-bad PreAdj-good v. PreAdj
bad
PosAdj-good v.
PosAdj-bad
t p t p t p t p
NS 41.87 <0.001 26.63 <0.001 59.00 <0.001 15.40 <0.001
AS 22.82 <0.001 38.95 <0.001 8.25 <0.001 4.29 0.001
IS 14.19 <0.001 20.30 <0.001 3.55 0.001 2.39 0.024
Note: NA-good = grammatical noun-adjective agreement; NA-bad = ungrammatical noun-adjective agreement;
DN-good = grammatical determiner-noun agreement; DN-bad = ungrammatical determiner-noun agreement;
PreAdj-good = grammatical prenominal adjective; PreAdj-bad = ungrammatical prenominal adjective; PosAdj-
good = grammatical postnominal adjective; PosAdj-bad = ungrammatical postnominal adjective
5.2 Quantitative Analysis: CBCT
The second task tested for knowledge of the syntactic-semantic collocation of prenominal and
postnominal adjectives by means of a production task. Figure 3 below shows the average number of
correct collocations (n=5) by each of the three groups for two types of sentences: (a) sentences that
require a prenominal adjective and (b) sentences that require a postnominal adjective. Section 4.2.2
provides examples of each sentence type.
15
Figure 3: CBCT results
Note: PreAdj-Col = grammatical prenominal adjective; PosAdj-Col = grammatical postnominal
adjective
As can be seen from Figure 3, the AS group average number of correct collocations is very similar
to that of the NS group for both prenominal and postnominal adjectives. However, the IS group
average number of correct collocations was similar to the NS group for postnominal adjectives only.
The IS group average number of correct collocations is below that of the NS group for prenominal
adjectives. In order to determine if either the AS group or the IS group performed in a native-like
manner, two additional statistical analyses were conducted. A one-way ANOVA was computed to
compare all three groups to each other. The results of this test showed that there was a statistically
significant difference among the groups in regards to the prenominal adjective tokens only. Therefore,
a two-sample t-test was conducted to determine among which groups these differences were. The two-
sample t-test revealed that the IS group differed statistically from the NS and AS groups. No
statistically significant differences were found between the performance of the NS group and the AS
group. This is summarized below in Table 5.
9
9
Assuming pedagogical influence to be the cause of the lower performance on prenominal adjectives
demonstrated by the L2 groups, one must still explain why the NS group also demonstrated more variation with
prenominal adjectives than with postnominal adjectives. One possibility is that a couple tokens were ambiguous
even for NS. Nevertheless, the AS group comparatively performed in a native-like manner in regards to the
prenominal tokens at both the group and individual level.
16
Table 5: CBCT (intergroup comparison)
PreAdj-Col PosAdj-Col
t p df t p df
ANOVA 8.36 0.001 2 1.91 0.159 2
NS v. AS 0.19 0.850 16 * * *
NS v. IS 4.48 <0.001 29 * * *
AS v. IS 2.57 0.021 15 * * *
Note: PreAdj-Col = grammatical prenominal adjective; PosAdj-Col = grammatical postnominal
adjective
A paired t-test revealed that there was a statistically significant difference in the number of correct
collocations for each of the three groups. This result was expected and is due to the significantly higher
number of correct postnominal collocations as compared to the number of correct prenominal
collocations. Table 6 summarizes this information.
Table 6: CBCT (intragroup comparison)
PreAdj-Col v. PosAdj-Col
t p
NS 8.04 <0.001
AS 4.78 0.001
IS 8.07 <0.001
Note: PreAdj-Col = grammatical prenominal adjective; PosAdj-Col = grammatical postnominal
adjective
6. Discussion
In this section, we bring together the important findings of both tasks in an effort to see how these
data come to bear on the theoretical significance of the overall study. Crucially, the GJCT
tested for
the acquisition of uninterpretable gender features (via their morphological reflexes) on determiners and
adjectives as well as knowledge of the various syntactic positions of Spanish adjectives delimited
strictly by their subcategorization. The intergroup comparisons conducted revealed that there were
significant difference between the NS and IS group performance for grammatical and ungrammatical
noun-adjective agreement as well as grammatical and ungrammatical prenominal adjectives and
ungrammatical postnominal adjectives. Therefore, the IS group performed in a native-like manner only
with respect to grammatical and ungrammatical determiner-noun agreement and postnominal
adjectives. From these results alone, it is difficult to determine whether the IS group, at this point in
interlanguage development, has in fact fixed the narrow syntactic properties of the Spanish DP and
thus acquired new L2 features (and instantiated a new functional category, WMP).
10
These findings
correlate to previous research of others, suggesting that adjectival agreement is more problematic than
determiner gender agreement (Fernández 1999; Bruhn de Garavito and White 2002) and that this is
also true when the L1 has similar gender features. With others, we take the position that it is a
morphological and not a narrow syntax problem, since, conversely, one would also expect equal
variability with determiner-noun agreement, yet this is not observed, and optionality in this respect
should extend to the advanced group as well. However, when one examines the results of the AS
group, this predication is not supported. No significant differences were found between the AS group
10
As stated above, the results from the two tasks reported on in this article are part of a larger study that includes
other linguistic tasks which test for semantic interpretations assumed to fall out from the acquisition of target
syntax. The results of these tasks allow us to determine with more accuracy whether or not both the AS and the IS
groups have in fact acquired the uninterpretable features found on determiners and adjectives in Spanish.
17
and the NS group when considering the gender agreement between nouns and adjectives or
determiners and nouns, suggesting that although there were significant differences found among these
two groups (with respect to grammatical prenominal adjectives, ungrammatical postnominal and the
differences between grammatical and ungrammatical prenominal and postnominal adjectives
contextualized judgements), we join Anderson (2007b) in claiming that this is likely to result from
pedagogical explanations received in the classroom. The generalization that prenominal adjectives are
always ungrammatical and that postnominal adjectives are always grammatical is explicitly taught to
instructed learners. Therefore, it is possible that explicit instruction interfered with the AS group’s
ability to accurately judge these tokens, especially in light of the results from the second task. This
might also explain, in part, the IS weaker performance (coupling this with some type of morphological
mapping problem; see Lardiere 2006 and works cited within) since the data reveal that they clearly
demonstrate a trend towards native-like performance. That is, as opposed to focusing on how different
they are from the controls, one could highlight the fact that they show a tendency towards
convergence, which suggests they have acquired or are in the process of acquiring new Spanish DP-
properties. Furthermore, because the main variables between the IS group and the AS group are the
amount and type of exposure to Spanish (i.e. classroom vs. a combination of classroom and
naturalistic), the fact that the AS group pattern, for the most part, like the native controls indicates that
with continued exposure to the target language the IS group participants will also come to fully acquire
new DP-features and overcome morphological competence deficits.
Results from the second task, the Context-based Collocation Task,
tested for knowledge of
adjectival placement based on a specific context. The relevant intergroup comparison conducted
revealed that no statistically significant differences were found between the three groups when
comparing the number of correct collocations of postnominal adjectives, suggesting that the syntax of
adjectival placement in L2 Spanish is target-like even at the intermediate level. Nevertheless, a
statistically significant difference was found between the IS group’s number of correct prenominal
collocations and that of the NS group. When this result is considered in isolation, we are faced with a
similar difficulty that was encountered in task 1. However, based on a coupling of all the data for this
group, it is reasonable to argue that they have acquired the target narrow syntactic properties of the
Spanish DP but that the syntax-semantics interface properties which regulate semantic interpretations
are delayed. Putting aside possible implications this could have for recent theorizing regarding
interface vulnerabilities (see e.g. Sorace 2000, 2005; White to appear), it is not entirely unexpected
that these L2 learners demonstrate greater difficulty with prenominal adjectival semantic construals
even when the target syntax is acquired since in addition to acquiring the DP features that result in
noun raising they need to realize that Spanish adjectives that are in a prenominal position cannot be
underlying associated with a lower syntactic position as they can in English.
11
Since no significant
differences were found among the AS group and the NS with respect to both prenominal and
postnominal collocation, it is clear that this difficulty is overcome through later stages of L2
interlanguage development. Crucially, the holistic picture that emerges from combining all the data
favor the position that new uninterpretable features not instantiated in the L1 can be and are acquired
in adult SLA. In an effort to address the better performance on task 2 as opposed to task 1, specifically
with respect to adjectives, we join Anderson (2007b) in suggesting that effects of different types of
input inclusive of explicit pedagogical rules could be the source of [some] task modality differences in
terms of adjectival placement and judgments. However, such problems seemingly lessen (and in some
instances disappear) with a methodology that requires contextualized production as opposed to
judgments. This can be explained by considering the differences in processing type required by the
linguistic tasks. Taken together, the results show that pedagogical effects are a realistic consideration,
but they do not affect competence and overall performance. Thus, these results nicely support
Rothman’s (2008) suggestion of the possibility that instructed L2 learners have two competing systems
that are especially active at the level of grammaticality judgment: a grammatical competence (their
11
Space limitations and the scope of this article do not permit us to explore the implications of this. Future
research from reporting data of additional tests within the same population will provide a more grounded and fine-
grained analysis that focuses on the implications for interface vulnerability hypotheses. See footnote 2.
18
actual grammar for the target L2) and a pedagogical competence (a system of conscious, explicit
rules). The fact that the AS group demonstrates clear knowledge of the syntactic-semantic positions of
adjectives provides evidence that supports FAs and evidence against the RDH.
Overall, our results support our general claim that L2 development is guided by Universal
Grammar, that UG’s features are entirely accessible to adult L2 learners and true acquisition takes
place after the purported offset of the critical period, as opposed to adults defaulting to linear learning
alone. However, while we have provided explanations that can account for divergent performances
it
is important to note that the burden of proof given the hypotheses under investigation (RDH vs. FA
approaches) the crucial question is whether or not L2 learners can acquire the relevant features, not
whether they perform in a native like way. Since both L2 groups demonstrate relevant contrasts in both
tasks, our data suggest that they are, in fact, able to acquire new features, as well as interpretive
contrasts depending on word order. Notwithstanding, the fact that the AS participant group performs
like the NS participant group on task 2 provides especially strong evidence against RDHs, since the
semantic interpretations of prenominal adjectives in Spanish would be predicted to demonstrate
interpretive optionality if indeed noun raising was never acquired and L2ers make use of other
conventions to derive the typical noun adjective word order as RDH approaches assume.
7. Conclusions
This study examined the acquisition of properties related to the internal structure of the Spanish
DP of intermediate and advanced English native learners of Spanish. Testing the predictions of Full
Access approaches vs. Representational Deficits Hypothesis approaches, we demonstrated that the
present data can only support Full Access approaches. Reporting on data from two linguistic tasks
testing for the acquisition of new DP-features, we demonstrated that intermediate English learners of
Spanish demonstrate evidence that they have acquired new grammatical features and that where they
differ from NS controls patterns like French L2 learners of Spanish differences from NS controls at the
same level (see Bruhn de Garavito and White 2002). Results from the second task confirm that
intermediate learners have acquired the syntax of adjectival placement, but have yet to fully acquire
the syntax-semantic interpretative differences, that is, unlearning the interpretive ambiguity of
prenominal adjectives. The advanced learners, however, performed more or less (see Sections 5 and 6)
like the NS control on both tasks, suggesting that new grammatical features, as evidenced by their
semantic reflexes, can be fully acquired after the so-called critical period. This means that the
hypothesis that the Critical Period amounts to differences in L1/L2 accessibility to representational
features (i.e. UG-features) is not a tenable one. It is prudent to state, however, that such claims in no
way diminish or seek to invalidate the observation that L1/L2 acquisitions are qualitatively and
quantitatively different. It simply suggests that accessibility to features is not the source of such
differences.
References
Abney, Steven. (1987). The English noun phrase in its sentential aspect. Ph.D, Massachusetts Institute of
Technology, Cambridge, MA.
Alexiadou, Artemis. (2001). Adjective syntax and noun raising: word order asymmetries in the DP as the result of
adjective distribution. Studia Linguistica, 55, 217-248.
Anderson, Bruce. (2001). Adjective Position and Interpretation in L2 French. In J. Camps & C-R. Wiltshire
(Eds.), Romance Syntax, Semantics and L2 Acquisition (pp. 27-42). Amsterdam: John Benjamins.
Anderson, Bruce. (2007a). Learnability and Parametric Change in the Nominal System of L2 French. Language
Acquisition, 14(2), 165–214.
Anderson, Bruce. (2007b). Pedagogical Rules and their Relationship to Frequency in the Input: Observational and
Empirical Data from L2 French. Applied Linguistics, 28(2), 286–308.
Anderson, Bruce. (2008). Forms of evidence and grammatical development in the acquisition of adjective position
in L2 French. Studies in Second Language Acquisition, 30, 1–30.
Beck, Maria-Louise. (1998). L2 acquisition and obligatory head movement: English speaking learners of German
and the Local Impairment Hypothesis. Studies in Second Language Acquisition, 20, 311−348.
19
Belletti, Adriana, Bennati, Elisa & Sorace, Antonella (2007). Theoretical and developmental issues in the syntax
of subjects: evidence from near-native Italian. Natural Language and Linguistic Theory, 2, 657-689.
Bernstein, Judy. (1993). Topics in the syntax of nominal structure across Romance. Ph.D, Graduate Center. City
University of New York.
Bruhn de Garavito, Joyce & White, Lydia. (2002). The L2 acquisition of Spanish DPs. The status of grammatical
features. In A.T. Pérez-Leroux & J. Liceras (Eds.), The Acquisition of Spanish Morphosyntax: The L1/L2
Connection (pp. 151-176). Dordrecht: Kluwer.
Cabrelli, Jennifer, Iverson, Michael, Judy, Tiffany & Rothman, Jason (2008). What the start of L3 tells us about
the end of L2: N drop in L2 and L3 Portuguese. In H. Chan, E. Kapia & H. Jacob (Eds.), A Supplement to the
Proceedings of the 32nd Boston University Conference on Language Development:
http://128.197.86.186/posters/32.
Carstens, Vicki. (1991). The morphology and syntax of determiner phrase in Kiswahili. Ph.D, University of
California, LA.
Carstens, Vicki. (2000). Concord in Minimalist theory. Linguistic Inquiry, 31, 319−355.
Chomsky, Noam. (1995). The minimalist program. Cambridge, MA: MIT Press.
Chomsky, Noam. (2000). Minimalist Inquiries: The Framework. In R. Martin, D. Michaels, & J. Uriagereka
(Eds.), Step by step: essays on minimalist syntax in honor of Howard Lasnik (pp. 31-52). Cambridge: MIT
Press.
Chomsky, Noam. (2005). On phrases. Unpublished manuscript. MIT.
Cinque, Guglielmo. (1994). On the evidence for Partial N Movement in the Romance DP. In G. Cinque, J. Koster,
J-Y., Pollock, L. Rizzi & R. Zanuttini (Eds.), Paths Towards Universal Grammar. Washington DC:
Georgetown University Press.
Demonte, Violeta. (2001). Merging and agreement of adjectives in DP: a feature based analysis. Ms. Universidad
Autónoma de Madrid.
Demonte, Violeta. (2008). Meaning-Form correlations and adjective position in Spanish. In C. Kennedy and L.
Mcnally (Eds.), The semantics of adjectives and adverbs. Oxford University Press.
Duffield, Nigel & White, Lydia. (1999). Assessing L2 knowledge of Spanish clitic placement: converging
methodologies. Second Language Research, 15, 133−160.
Epstein, Samuel, Flynn, Suzanne, & Martohardjono, Gita. (1996). Second language acquisition: Theoretical and
experimental issues in contemporary research. Brain and Behavioral Sciences, 19, 677-714.
Epstein, Samuel, Flynn, Suzanne & Martohardjono, Gita. (1998). The strong continuity hypothesis in adult L2
acquisition. In S. Flynn, G. Martohardjono & W. O’Neil (Eds.), The generative study of second language
acquisition (pp. 61-79). Mahwah: Lawrence Erlbaum Associates.
Fernández, Marisol. (1999). Patterns in gender agreement in the speech of second language learners. In J.
Gutiérrez-Rexach & F. Martínex-Gil (Eds.) Advances in Hispanic Linguistics (pp. 3-15). Sommerville MA:
Cascadilla Press.
Franceschina, Florencia. (2001). Morphological or syntactic deficit in near-native speakers? An assessment of
some current proposals. Second Language Research, 17, 213−147.
Franceschina, Florencia. (2005). Fossilized Second Language Grammars: The Acquisition of grammatical
gender. Amsterdam: John Benjamins.
Frenck-Mestre, C., McLaughlin, J., Osterhout, L. & Foucart, A. (in press). The effect of phonological realization
of inflectional morphology on verbal agreement in French. Evidence from ERPs, Acta Psychologica.
Goad, Heather & White, Lydia. (2006). Ultimate attainment in interlanguage grammars: a prosodic approach.
Second Language Research, 22(3), 243-268.
Harris, James. (1991). The exponence of gender in Spanish. Linguistic Inquiry, 22, 27-67.
Harris, James. (1995). The syntax and morphology of class marker suppression in Spanish. Ms. MIT.
Hawkins, Roger. (1998). Explaining the difficulty of gender attribution for speakers of English. Paper presented at
EUROSLA, Paris.
Hawkins, Roger. (2001). Second Language Syntax. Malden NJ: Blackwell.
Hawkins, Roger. (2005). Revisiting Wh-movement: The availability of an uninterpretable [wh] feature in
interlanguage grammars. In L. Dekydtspotter et al (Eds.), Proceedings of the 7th Generative Approaches to
Second Language Acquisition Conference (GASLA-7) (pp. 124−137). Somerville, MA: Cascadilla.
Hawkins, Roger & Chan, Cecilia. (1997). The partial accessibility of Universal Grammar in second language
acquisition: The failed functional features hypothesis. Second Language Research, 13, 187−226.
Hawkins, Roger & Franceschina, Florencia. (2004). Explaining the Acquisition and Non-Acquisition of
Determiner-Noun Concord in French and Spanish. In P. Prévost & J. Paradis (Eds.), The Acquisition of
French in Different Contexts: Focus on Functional Categories (pp. 175-205). Amsterdam: John Benjamins.
20
Hawkins, Roger & Hattori, Hajime. (2006). Interpretation of English multiple wh-questions by Japanese speakers,
a missing uninterpretable feature account. Second Language Research, 22, 269−301.
Haznedar, Belma & Schwartz, Bonnie D. (1997). Are there optional infinitives in child L2 acquisition? In E.
Hughes, M. Hughes & A. Greenhill (Eds.), Proceedings of the 21st Annual BUCLD (pp. 257-268).
Somerville, MA: Cascadilla Press.
Jackendoff, Ray. (1977). X-bar Syntax: A study of phrase structure. Cambridge: MIT Press.
Lardiere, Donna. (1998). Dissociating syntax from morphology in a divergent L2 end-state grammar. Second
Language Research, 14, 359–375.
Lardiere, Donna. (2006).Ultimate attainment in second language acquisition: a case study. Mahwah: LEA.
Luján, Marta. (1980). Sintaxis y Semántica del adjetivo. S. A. Madrid: Cátedra.
Masullo, P.J. & Depiante, M.A. (2003). Gender is in the lexicon, Number in the syntax: Evidence from nominal
ellipsis in Spanish. Paper presented at LSRL XXXIII at Indiana University.
Picallo, M.C. (1991). Nominals and nominaliziation in Catalan. Probus, 3, 279–316.
Prévost, Philippe & White, Lydia. (2000). Missing surface inflection or impairment in second language
acquisition? Evidence from tense and agreement. Second Language Research, 16, 103–133.
Ritter, Elizabeth. (1993). Where is gender? Linguistic Inquiry, 24, 795-803.
Rothman, Jason. (2008). Aspectual Morphology use in Adult L2 Spanish and the Competing Systems Hypothesis:
When Pedagogical and Linguistics Rules Conflict. Languages in Contact, 8(1), 74-106.
Schwartz, Bonnie & Sprouse, Rex. (1996). L2 cognitive states and the Full Transfer/Full Access model. Second
Language Research, 12, 40−72.
Schwartz, Bonnie & Sprouse, Rex. ( 2000). When syntactic theories evolve: consequences for L2 acquisition
research. In J. Archibald (Ed.), Second language acquisition and linguistic theory (pp.157–86). Blackwell.
Slabakova, Roumyana. (2006). Is there a Critical Period for Semantics? Second Language Research, 22, 302-338.
Sorace, Antonella. (2000). Syntactic optionality in non-native grammars. Second Language Research, 16, 93-102.
Sorace, Antonella. (2004). Native language attrition and developmental instability at the syntax-discourse
interface: Data, interpretations and methods. Bilingualism: Language and Cognition, 7, 143-145.
Sorace, Antonella. (2005). Selective optionality in language development. In L. Cornips & K. Corrigan (Eds.),
Biolinguistic and sociolinguistic accounts of syntactic variation (pp. 55–80). Amsterdam: John Benjamins.
Sorace, Antonella & Filiaci, Francesca. (2006). Anaphora resolution in near-native speakers of Italian. Second
Language Research, 22, 339-368.
Tsimpli, Ianthi Maria & Dimitrakopoulou, Maria. (2007). The Interpretability Hypothesis: Evidence from Wh-
Interrogatives in Second Language Acquisition. Second Language Research, 23, 215−242.
Tsimpli, Ianthi Maria & Mastropavlou, Maria. (2008). Feature interpretability in L2 acquisition and SLI: Greek
clitics and determiners. In J. Liceras, H. Zobl & H. Goodluck (Eds.), The role of formal features in second
language acquisition, (pp. 142-183). Mahwah, NJ: Lawrence Erlbaum.
Valenzuela, Elena. (2006). L2 end state grammars and incomplete acquisition of Spanish CLLD constructions. In
R. Slabakova, S. Montrul & P. Prévost (Eds.), Inquiries in linguistic development: in honor of Lydia White
(pp. 283-304). Amsterdam: John Benjamins.
Valois, Daniel. (1991). The internal Syntax of DP. PhD dissertation, UCLA.
White, Lydia (1989). Universal Grammar and Second Language Acquisition. Amsterdam: John Benjamins
White, Lydia. (2003). Second language acquisition and Universal Grammar. New York: Cambridge University
Press.
White, Lydia. (to appear). Grammatical Theory: Interfaces and L2 knowledge. In W. Ritchie and T. K. Bhatia
(Eds.). Handbook of Second Language Acquisition. Blackwell.
White, Lydia, Valenzuela, Elena, Koslowska-MacGregor, Martyna, & Leung,Yan-Kit Ingrid. (2004). Gender and
number agreement in nonnative Spanish. Applied Psycholinguistics, 25, 105−133.
Zagona, Karen. (2002). The syntax of Spanish. Cambridge: Cambridge University Press.
21
Selected Proceedings of the
2007 Second Language Research Forum
edited by Melissa Bowles, Rebecca Foote,
Silvia Perpiñán, and Rakesh Bhatt
Cascadilla Proceedings Project Somerville, MA 2008
Copyright information
Selected Proceedings of the 2007 Second Language Research Forum
© 2008 Cascadilla Proceedings Project, Somerville, MA. All rights reserved
ISBN 978-1-57473-425-6 library binding
A copyright notice for each paper is located at the bottom of the first page of the paper.
Reprints for course packs can be authorized by Cascadilla Proceedings Project.
Ordering information
Orders for the library binding edition are handled by Cascadilla Press.
To place an order, go to www.lingref.com or contact:
Cascadilla Press, P.O. Box 440355, Somerville, MA 02144, USA
phone: 1-617-776-2370, fax: 1-617-776-2271, e-mail: [email protected]
Web access and citation information
This entire proceedings can also be viewed on the web at www.lingref.com. Each paper has a unique document #
which can be added to citations to facilitate access. The document # should not replace the full citation.
This paper can be cited as:
Judy, Tiffany, Pedro Guijarro-Fuentes, and Jason Rothman. 2008. Adult Accessibility to L2 Representational
Features: Evidence from the Spanish DP. In Selected Proceedings of the 2007 Second Language Research Forum,
ed. Melissa Bowles, Rebecca Foote, Silvia Perpiñán, and Rakesh Bhatt, 1-21. Somerville, MA: Cascadilla
Proceedings Project.
or:
Judy, Tiffany, Pedro Guijarro-Fuentes, and Jason Rothman. 2008. Adult Accessibility to L2 Representational
Features: Evidence from the Spanish DP. In Selected Proceedings of the 2007 Second Language Research Forum,
ed. Melissa Bowles, Rebecca Foote, Silvia Perpiñán, and Rakesh Bhatt, 1-21. Somerville, MA: Cascadilla
Proceedings Project. www.lingref.com, document #1731.
