What Makes Learning Second-Language

Grammar Difficult? A Review of Issues

Robert M. DeKeyser
University of Pittsburgh

If the ability to use language in the restricted sense (i.e., a

communication system characterized by double articulation) is
quintessentially human, then explaining this ability is a crucial
task for cognitive science, and explaining its acquisition is a
crucial task for developmental psychology. If the difficulty of
acquiring a second language (L2), at least at a later age, stands
in sharp contrast to the childs celebrated accomplishments, then
explaining this contrast is equally important for developing a
complete understanding of humans abilities to use and acquire
language. In fact, it can be argued that it is the enormous
contrast between the two phenomena that needs explaining,
rather than either of the two phenomena per se. One way to
tackle this problem is the social science approach of correlating
age and many other demographic variables with success in
acquisition: to disentangle design features of the species from
accidental characteristics of the environment. Another approach
consists of investigating what elements or characteristics of an
L2 are hard to acquire: to understand better how weaknesses in
the acquisition process interact with the design features of human
languages. And of course, one can look at the two variables in
interaction with each other: Which problematic elements of the
Correspondence concerning this article should be addressed to Robert
DeKeyser, Department of Linguistics, University of Pittsburgh, Pittsburgh
PA 15260, USA. Internet: [email protected]

What Makes Learning L2 Grammar Difficult?

language are an issue in L2 learning at any age, and which

are mostly a problem for later acquirers only? Or even better,
how do five different variables interact in L2 acquisition: the
characteristics of the L2, the influence of the first language (L1),
the role of age, the role of individual differences in cognitive and
affective aptitudes, and the role of learning context, be it the
native-speaking environment or the classroom, the latter
representing, of course, a wide variety of learning contexts
with different degrees of emphasis on form and meaning?
The focus of this introductory review article is on the characteristics of the L2 itself (and its differences from L1) that make
its acquisition difficult. Given the very broad nature of the topic,
I will touch on the issues of age, other individual differences, and
learning context only to the extent that they cannot be ignored
because nothing can be generalized without taking them into
account. I also restrict discussion to morphosyntax rather than
phonology or the lexicon and to the acquisition of competence
rather than processing, recognizing here too that all such
separations are to some extent superficial because the meaning
of morphemes and the distribution of their allomorphs cannot be
acquired without the phonological capacity to extricate them from
the flood of sounds in every sentence, and because competence is
only a (some would say fictional) abstraction of what humans do
when they understand or produce language, and acquiring this
competence necessarily happens through processing input
(cf., esp., Pienemann, 2003; Truscott & Sharwood Smith, 2004).
Finally, I emphasize publications from the past 5 years in keeping
with the criteria for selecting articles for the Best of Language
Learning series.

Broad Definitions of Difficulty

Even a cursory glance at some well-known discussions of what
is difficult in L2 acquisition shows how tricky this concept is.
Krashen (1982) and R. Ellis (1990), for instance, at first sight
appear to agree that one needs to make a distinction between

DeKeyser

formal and functional complexity. But the same structure, third

person s in English as a Second Language (ESL), was classified
by Krashen (1982) as easy to learn because it is simple, and by
Ellis (1990) as hard to learn because it is complex. One might
think that the reason for this discrepancy is that Krashen was
dealing with learning in the narrow sense here (as opposed to
implicit acquisition) and Ellis with a broader meaning of learning,
but when one looks at their reasons for classifying this structure
as easy or difficult, it is clear that they used different criteria for
deciding on the complexity of s: Krashen pointed to the simple
dichotomous choice between supplying this simple morpheme or
not, whereas Ellis, referring to Pienemann (1984), pointed to the
long-distance relationship between the grammatical number of
the subject and the presence or absence of s on the verb. Nor is the
disagreement due to a mere focus on formal complexity by Krashen
versus a broader look at the form-function relationship by Ellis:
The latter actually goes beyond Krashen by considering even the
form-function relationship for s to be simple (transparent); it is
only because of the processing operations required that Ellis
considered the structure to be complex (1990, p. 167).
It appears, then, that at least three factors are involved
in determining grammatical difficulty: complexity of form,
complexity of meaning, and complexity of the form-meaning
relationship. Even this picture, however, is not complete; it
actually leaves out the core psycholinguistic difficulty of
acquisition, that is, the difficulty of grasping the form-meaning
relationship while processing a sentence in the L2. Rather than
forms, meanings, or form-meaning relationships, it is the transparency of form-meaning relationships to a learner who is
processing language for meaning that determines the difficulty
of acquisition, at least for learners who are left to their own
resources instead of presented with a reasonably complete set
of rules about form-meaning relationships.
Part of what determines this transparency is the degree of
importance of a linguistic form for the meaning it expresses:
Certain morphemes are the one and only clue to the meaning

What Makes Learning L2 Grammar Difficult?

they express; others are largely or completely redundant, because

they mark grammatical agreement with meanings whose primary
representations are elsewhere in the sentence or the discourse.
VanPatten (e.g., 1990) has therefore emphasized the distinction
between meaningful and redundant for predicting what will be
easy or hard to acquire, especially in early stages of L2 development.
Some researchersmost notably Stockwell, Bowen, and
Martin (1965)have drawn up elaborate hierarchies of difficulty
of acquisition based on form-meaning mapping. But the work of
Stockwell, Bowen, and Martin (1965) focused primarily on
Spanish, put more emphasis on L1L2 surface differences than
is warranted by more recent research on the role of L1, and was
largely nonempirical. They also left out completely the notion of
salience in the input, presumably because they were thinking of
instructed learning contexts, in which salience is less of an issue
than in naturalistic L2 acquisition, a phenomenon that was not
yet a topic of research at that time.
Other researchers, recognizing the difficulty of defining
difficulty, have avoided a theoretical conceptualization altogether.
When structures needed to be classified according to difficulty,
they chose to ask teachers to rate L2 structures for the level of
difficulty they seemed to present intuitively (e.g., Robinson, 1996).
While such an approach may be a useful operationalization,
depending on the nature of the study, it still leaves us with the
question of what constitutes difficulty.
In what follows, I present various components of grammatical
difficulty, with at least a modest amount of empirical evidence for
their importance in acquisition or lack thereof and for how they fit
into the broader picture of interaction with each other and with
individual and contextual factors (the latter including L1 as well
as instruction). One way of isolating components of difficulty is to
look separately at problems of meaning, problems of form, and
problems of form-meaning mapping.

DeKeyser

Problems of Meaning
Regardless of the form used to express a meaning, the meaning
itself can constitute a source of difficulty, because of novelty,
abstractness, or a combination of both. Articles, classifiers,
grammatical gender, and verbal aspect are notoriously hard to
acquire for native speakers of L1s that do not have them or that
use a very different system (for articles in ESL, see, e.g., Jarvis,
2002; Liu & Gleason, 2002; Robertson, 2000; Tarone & Parrish,
1988; Thomas, 1989; Young, 1996; cf. also Celce-Murcia & LarsenFreeman, 1999, chap. 15; for classifiers in Japanese and Chinese,
see, e.g., Hansen & Chen, 2001; for grammatical gender in a
variety of languages, see, e.g., Carroll, this volume; Kempe &
Brooks, this volume; Taraban, 2004; Williams & Lovatt, this
volume; for aspect in Romance or Germanic languages, see, e.g.,
Andersen & Shirai, 1994, 1996; Bardovi-Harlig, 1998, 1999, 2000;
Collins, 2002; Dietrich, Klein, & Noyau, 1995; Lee, 2001; Montrul
& Slabakova, 2003; Salaberry, 2000).
These elements of grammar are even strongly resistant to
instructional treatments (for aspect see, e.g., Ayoun, 2004; Ishida,
2004; for gender see, e.g., Leeman, 2003; for articles see, e.g.,
Butler, 2002; Master, 1997). What they all have in common is
that they express highly abstract notions that are extremely hard
to infer, implicitly or explicitly, from the input. Where the
semantic system of the L1 is different from that of the L2, as is
very often the case for aspect, or where equivalent notions do not
get expressed overtly in L1, except through discourse patterns, as
may be the case for ESL articles for native speakers of most Slavic
languages or Chinese, Japanese, or Korean, the learning problem
is serious and long-lasting.

Problems of Form
Difficulty of language form is largely an issue of complexity.
Assuming the learner knows exactly the meanings that need to be
expressed, difficulty of form could be described as the number of

What Makes Learning L2 Grammar Difficult?

choices involved in picking all the right morphemes and

allomorphs to express these meanings and putting them in the
right place. Clearly, this problem is most complex in richly
inflected languages, whether they be agglutinative, polysynthetic,
or inflectional in the narrow sense. Everything else (such as
semantic difficulty) being the same, the more that needs to be
expressed overtly, the more choices need to be made about
morphemes, allomorphs, and their position. Morphology in L2 is
hard: Basic word order is typically nonproblematic past the initial
stages of acquisition, but even the most basic morphology is often
lacking from the speech of untutored immigrants (see, e.g., Klein
& Dittmar, 1979) and of classroom learners who are not able to
monitor themselves effectively (see, e.g., Krashen & Pon, 1975;
Tarone, 1985). Morphology is even shakily represented in
learners intuitions, even after many years of exposure to the L2
(e.g., DeKeyser, 2000; Johnson & Newport, 1989).
Much ink has been spent discussing whether continuing
failure to supply these morphemes systematically is truly a
problem of competence or one of mere performance. Given the
poor scores of adult immigrants on grammaticality judgment tests
on this point (in ESL, e.g., third person s, articles, or plurals in
DeKeyser, 2000; Johnson & Newport, 1989; Yeni-Komshian,
Robbins, & Flege, 2001) and the failure of intermediate Englishspeaking foreign language students even to take into account the
meaning of elementary morphology in order to come to the correct
understanding of a sentence (see, esp., VanPatten, 2004; cf. also
MacWhinney, 2001, in press), it seems safe to conclude that more
than processing is at stake. Jiang (2004), in particular, showed
convincingly that errors of verb agreement with complex noun
phrases in ESL were due to lack of sensitivity to plural marking
on the noun, not to problems with processing agreement.
This problem of L2 users failing to use morphology, even in
comprehension, is so fundamental that it has by itself spawned
entire bodies of literature. The research on processing instruction
has showed that students benefit from intensive training in paying
attention to elements of morphology for comprehension, because

DeKeyser

without such practice they tend to gloss over the morphology

(especially students of a morphology-poor language like English
acquiring a relatively morphology-rich language like Spanish).
The interpretation of various aspects of the processing-instruction
literature has been controversial (see, e.g., DeKeyser, Salaberry,
Robinson, & Harrington, 2002; VanPatten & Wong, 2002). But
nobody doubts that L2 students need to have their attention
drawn to morphology while processing input, because otherwise
they tend to ignore the morphological cues to sentence meaning.
The research within the framework of the competition model,
on the other hand, has shown repeatedly, with speakers of a
variety of L1s and L2s, that morphology is a weak cue in initial
stages of language learning, at least for English L1 speakers, and
that if it becomes stronger over time, this only happens in a very
slow and gradual fashion (cf., esp., MacWhinney, 2001, in press;
see, e.g., Hertel, 2003; Kempe & MacWhinney, 1998; McDonald,
1987).
More formal approaches to morphosyntax, however different
they may be in other respects, coincide in singling out morphology
as hard to acquire in comparison with syntax. Lardiere (1998),
Prevost and White (2000), and Sprouse (1998) provide evidence
that morphological and syntactic features that are closely linked
in syntactic theory (verb raising and inflection) are not acquired
together. One way out of this problem, from the point of view of a
theory in which such a link is seen as crucial, is the view that
learners acquire the syntactic features easily but continue to have
problems with their morphological instantiation (cf. Sorace, 2003,
and Lardieres summary of several of her own articles in Long,
2003).

Problems of Form-Meaning Mapping

Even assuming that neither form nor meaning is particularly
problematic according to the criteria mentioned in the previous
two sections, acquiring the form-meaning mapping can still be
difficult if the link between form and meaning is not transparent.

What Makes Learning L2 Grammar Difficult?

Such lack of transparency can be due to at least three factors:

redundancy, optionality, or opacity.
Redundancy means that the form at issue is not semantically necessary because its meaning is also expressed by at least
one other element of the sentence; for example, a verb ending
can be redundant because the subject is explicit, whether it be a
full noun phrase or a pronoun, which makes person and number
information redundant, and because adverbs or other lexical
items make information such as tense or aspect redundant
(cf., e.g., VanPatten, 1990). When the redundant element is also
abstract and novel, then the learning problem is particularly
severe: Robinson (2002), for instance, showed that learners of
Samoan L2 had more trouble with the ergative marker than
with the locative or noun incorporation, because the ergative
marker (both novel for learners of L1 English and abstract) was
semantically redundant, which the other two structures were not.
Optionality of certain elements, such as null subjects in
Spanish or Italian (see, e.g., Herschensohn, 2000; Liceras, 1989)
or case marking in Korean, only makes matters worse. Not only
does the optional character of the case marking or the overt
subject pronoun suggest it is redundant, but its alternating
presence or absence in the presence of the same meaning,
except for subtle aspects of pragmatics, makes the formmeaning link even harder to establish. (This optionality in the
L2 as a cause of acquisition problems is not to be confused with
optionality in the sense of interlanguage variability, which is a
consequence of a variety of problems with the acquisition of
features that are not variable in the L2; cf. Papp, 2000; Prevost
& White, 2000; Robertson, 2000; Sorace, 2000, 2003.)
Opacity is a complex form of the problem of low form-meaning
correlation. When a morpheme has different allomorphs, and at
the same time it is homophonous with other grammatical
morphemes, then the correlation between form and meaning
becomes very hard to detect: Different forms stand for the same
meaning, and the same form stands for different meanings. This,
of course, is exactly the case for s in English, which can be the

DeKeyser

third-person singular of the verb, the plural of the noun, or the

genitive of the noun and in each case has the same three
allomorphs. Where third-person singular is concerned, this
problem of low correlation between form and meaning is further
compounded by the so-called morphemes really being a morph (a
chunk of sound isolated in morphological analysis, but without a
one-to-one mapping with any meaning), in this case conflating
three different meanings (singular, third person, present tense),
all of which are expressed by separate morphemes in many
languages, and all of which have to be present at the same time
for s to appear, and all of which are rather abstract and therefore
difficult by themselves. More generally, instances of morphological
irregularity, such as irregular plurals and irregular past tenses, all
fall into this category: The problem is not so much a problem of
form (an irregular is not necessarily more formally complex); it is
the form-meaning mapping that becomes more opaque and/or
complex.
Other examples of opaque form-meaning mappings can be
found in syntax, such as the relationship between the order of
subject and verb in Spanish, on the one hand, and the lexical
semantics of the verb and the discourse functions of subject-verb
inversion, on the other hand. Hertel (2003) showed how learners
of L2 Spanish do not use much verb-subject (VS) order till they
are quite advanced in their L2 proficiency and that even
the advanced learners do not make any distinction between
unergatives and unaccusatives when it comes to VS order, even
though native speakers clearly do. The optional nature of the VS
order makes it even harder to acquire its correlation with the
abstract semantic elements that favor its appearance. DeKeyser
(2005) also documents a virtual total absence of VS order for
any kind of verb in declarative sentences among intermediate
Spanish learners during a 6-week stay in Argentina. Jungs
(2004) finding that English speakers did not acquire topic
prominence in Korean L2 till the advanced level, in spite of the
hypothetical universal nature of topic prominence in early interlanguage, can be seen as another example of how difficult it is to

10

What Makes Learning L2 Grammar Difficult?

map a variable phenomenon to abstract discourse-dependent

semantics. To the learner, many cases of opacity probably appear
to be instances of optionality.
The acquisition problem is compounded even further when
optionality and discourse-motivated preferences for one of the
options interact with arbitrary or semantically obscure subcategorization restrictions, such as, in Japanese L2, the restriction of
goal PPs to directed-motion verbs (Inagaki, 2001) or the restriction
of quantifier floating to unaccusatives (Sorace & Shomura, 2001),
and in English L2, the restriction of agentive use to manner-ofmotion verbs as opposed to change-of-state verbs (Montrul, 2001),
the restrictions on dative alternation (see esp. Inagaki, 1997;
Whong-Barr & Schwartz, 2002), and those on locative alternation
(see esp. Bley-Vroman & Joo, 2001; Joo, 2003; Juffs, 1996). The
possibility of alternation seems to be motivated by semantic
criteria, but according to Pinker (1989), at least 14 semantically
defined verb classes need to be distinguished for locative
alternation and 10 for dative alternation. When, on top of that,
relevant input is very limited in ESL materials (Juffs, 1998) and
probably even in more natural input, it is clear that acquisition is
a challenge, to say the least. Even simpler subcategorization
restrictions, such as which English verbs take an infinitive and
which a gerund, appear to be problematic, even after decades of
exposure to the language (DeKeyser, 2000; Flege, Yeni-Komshian,
& Liu, 1999; Johnson & Newport, 1989; McDonald, 2000).
Finally, an important factor that helps determine ease or
difficulty of learning form-meaning mappings is, of course,
frequency. N. Ellis (2002, 2003) provided evidence from a variety
of sources on the role of frequency in L2 learning. He argued that
the typical route of acquisition of grammar structures is from
formulae through low-scope patterns to constructions and that
the abstraction of regularities within these constructions is
frequency-based. In principle, the importance of frequency is
independent of semantic transparency, but how important
frequency is depends to some extent on the transparency of the
mapping. If the mapping is very clear, minimal exposure may be

DeKeyser

11

enough for acquisition; if it is very obscure, the structure may well

never be acquired by adults. If the transparency of the mapping is
in between these two extremes, frequency may largely determine
whether the mapping is acquired or not, in further interaction, of
course, with the learners aptitude and (methods of) instruction.
Hansen and Chen (2001), for instance, provided a clear example of
a set of morphemes in which frequency played a very important
role: classifiers in Chinese and Japanese L2. As the structure is
fairly salient and its function is clear, but the choice of the right
form is based on semantic criteria that are hard to define, the level
of difficulty is such that the role of frequency is maximized.

Where to Look for Direct Evidence on What Is Difficult?

I have presented a number of factors that make the
acquisition of L2 grammar difficult, attempting to organize
evidence from a wide variety of studies which addressed all
kinds of narrowly focused questions about the acquisition of
grammar, but in the process I have also documented aspects
of the difficulty of specific language structures. Relatively few
studies have actually attempted a systematic empirical investigation of difficulty by comparing acquisition for a broad range of
language structures. The vast majority of studies either have a
narrow linguistic focus or a very wide linguistic scope with little
or no interest in comparing acquisition of different structures
systematically. In principle, there are several research areas in
which one could look for systematic evidencedocumentation of
fossilization, ultimate attainment studies with adult learners, and
research on order of acquisitionbut not all of these areas have
yielded much that can help answer the question of what makes an
L2 structure difficult.
Fossilization is a concept that has been around for decades,
but the concept has remained rather vague. If one uses a narrow
definition, such as that of Long (2003), requiring that learners
have been fully exposed to the L2 for at least 10 years and that
lack of change for a specific structure has been systematically

12

What Makes Learning L2 Grammar Difficult?

documented for at least 5 years, then it turns out that only

a handful of studies have met this criterion. Long (2003) cited
Han (2000), Lardiere (1998, 2000), and Long (1997), which have
documented fossilization of, respectively, nontarget unaccusative verbs, missing verb morphology, and lexically determined
inflection on nouns and verbs.
Ultimate attainment studies of adult learnerswith the goal
of establishing age effectshave become popular in the last decade
or so, but again not many have engaged in a systematic comparison
of how well different aspects of grammar have been acquired. The
only studies in the morphosyntax category that have explored
differential age effects for different language elements are
Birdsong (1992), DeKeyser (2000), DeKeyser, Ravid, and AlfiShabtay (2005), Flege et al. (1999), Johnson and Newport (1989,
1991), and McDonald (2000).
Both Birdsong (1992) and Johnson and Newport (1989, 1991)
made a distinction between structures within and outside of the
realm of universal grammar. Birdsong made a direct comparison
within his study and found that both categories of structures were
equally sensitive to age effects. Johnson and Newport carried out
two studies, one with structures outside of Universal Grammar
(UG; 1989), the other with structures exemplifying the subjacency
principle, assumed to be an element of UG (1991). They found
essentially the same strong age effect in both studies. It should be
pointed out, however, that the UG/non-UG classification in
Birdsong (1992) was tentative, as the author indicated himself,
and that the classification of subjacency among the principles of
UG is not uncontroversial either (Huang, 1982; Pesetsky, 1987;
Rizzi, 1982).
Flege et al. (1999) made a different kind of distinction. They
administered an ESL grammaticality judgment test, mostly
drawn from Johnson and Newport (1989), to speakers of Korean
L1 and found overall age-proficiency correlations of
.71 for age
on arrival (AoA) < 15 and
.23 for AoA > 15. The researchers
showed, through analysis of a series of subsamples matched for
AoA, that education was a significant predictor for performance

DeKeyser

13

on rule-based items and use of English was a significant predictor

for performance on lexically based items.
Johnson and Newport (1989), DeKeyser (2000), and McDonald
(2000), all using variants of the same grammaticality judgment test
for English L2, found a tendency for a few elements of morphosyntax to be resistant to age effects, especially basic word order and
yes/no questions. Moreover, both Flege et al. (1999) and McDonald
(2000) found that auxiliaries, subcategorization and wh-questions
were more sensitive to length of residence or other measures of
usage than AoA. On the basis of such patterns and further post
hoc analyses of his own data (for instance, successful acquisition of
subject-verb inversion in yes-no questions, but not in wh-questions),
DeKeyser (2000) hypothesized a strong role of salience, in the sense
that this variable becomes increasingly important with age.
One more study to mention here is Yeni-Komshian et al.
(2001). Among the Korean speakers of L2 English in this study,
performance was markedly worse for plural s than for third-person
s, and this difference became gradually larger with increasing AoA.
The authors of this study attributed the difference in correctness
between noun phrase (NP) and verb phrase (VP) to the higher
salience of the VP in Korean L1 and to a transfer of processing
strategies in this respect.
In contrast to ultimate attainment studies, research on
acquisition order, of course, has made explaining difficulty into
an explicit goal, at least if one accepts that difficulty can be
operationalized as order of acquisition. This type of research
was prominent from the early 1970s to the early 1980s, but while
researchers eventually agreed on a more or less universal order of
acquisition of grammatical morphemes in ESL, they could not
agree on an explanation for that order (cf., e.g., Gass & Selinker,
2001; Larsen-Freeman & Long, 1991; Long & Sato, 1984). This
lack of explanatory adequacy was probably one of the main reasons
why the order-of-acquisition question for morphemes faded from
the published literature in the mid-1980s. Parallel work on
acquisition order of syntactic patterns such as interrogative
structures (e.g., Eckman, Moravcsik, & Wirth, 1989) or relative

14

What Makes Learning L2 Grammar Difficult?

clauses (e.g., Eckman, Bell, & Nelson, 1988; Gass, 1979), however,
pointed to markedness as a potentially important factor in
determining order of acquisition/difficulty. Bardovi-Harlig (1987),
on the other hand, showed that salience prevailed over
markedness by comparing the acquisition of pied piping and
preposition stranding in ESL. Goldschneider and DeKeyser (this
volume) returned to the issue of morpheme acquisition order and
found that saliencebroadly construed as a combination of phonological salience, semantic complexity, morphological regularity, and
frequencyaccounted for a large percentage of the variance in the
order of L2 acquisition. They found it impossible to tease out the
contribution of the various components of salience, however,
because these factors are strongly intercorrelated in English
morphology.
Given the evidence for the importance of salience in both the
ultimate-attainment and the order-of-acquisition literatures, on
the one hand, and the lack of systematic research on the role of
salience in ultimate attainment by adult learners, on the other
hand, DeKeyser, Ravid, and Alfi-Shabtay (2005) decided to
investigate the role of salience in the acquisition of Hebrew
morphology by adult immigrants. They found not only a strong
effect of salience in determining difficulty (as measured by
a grammaticality judgment test), but also a significant interaction
with age in the sense that the role of salience grew more important
with increasing age of acquisition. Further analysis of the data
(DeKeyser, Alfi-Shabtay, Ravid, & Shi, 2005) showed that several
components of salience played an independent role in determining
difficulty for all learners: phonological salience (length in phones
and /
syllabic character of the morpheme) and /
homonymy
with other morphemes. Two other components of salience did
not show a main effect but interacted with age in the sense
that they were an important predictor of learning for older
learners only: distance (between morphemes in agreement
patterns) and stress.
In conclusion, there is increasing evidence from both the
order-of-acquisition literature and the ultimate-attainment

DeKeyser

15

literature that lack of salience plays an important role in

acquisition difficulty. Much remains to be done in this area,
however. On the one hand, it is much harder to agree on an
operationalizaton of salience in syntax as opposed to morphology
or phonology. On the other hand, even for morphology, there is still
a lack of systematic, let alone cross-linguistic, research comparing
acquisition difficulty for morphemes with different degrees of
salience or different degrees of other factors mentioned above, for
that matter, such as redundancy, optionality, and opacity of formmeaning mapping, novelty and abstractness of meaning, and sheer
complexity of form. As research on ultimate attainment effects is
accumulating, there is hope, however, that a meta-analysis of that
literature may soon be able to help determine the characteristics of
consistently poorly learned L2 structures.

Mitigating Factors
Meanwhile, additional insights on what is difficult and why
come from studies that have investigated the interaction
between characteristics of the L2 structures being learned and
individual learner or contextual factors. While a substantial
literature exists on individual differences, not much work has
addressed the question of the differential impact of factors such
as aptitude and motivation on specific elements within
morphology and syntax, in other words, on elements characterized by specific types of difficulty. Two recent studies that
stand out in this area, however, are Willliams (1999) and
Williams and Lovatt (this volume).
The findings of both studies are complex, but Williams (1999)
showed that meaningful form-function mapping (in Italian L2)
resulted from conceptually driven, explicit learning (a function of
aptitude in the sense of grammatical sensitivity), whereas semantically redundant agreement rules were largely the result of datadriven, implicit learning (a function of memory). DeKeyser (2003)
argued that this is because the agreement rules amounted to
concrete sound-sound correspondences (even euphony), whereas

16

What Makes Learning L2 Grammar Difficult?

the meaningful form-function mappings required associating an

element of the noun phrase with an element of the verb phrase,
each element taking a very different concrete form. Associating
nonmeaningful co-occurrence of concrete elements logically draws
more on memory, whereas establishing meaningful relationships
between abstract entities draws more on insight (cf. Gomez, 1997;
Perruchet & Pacteau, 1990, 1991; Reed & Johnson, 1998). Williams
and Lovatt (this volume) found that even for more abstract
patterns of morpheme agreement, not involving euphony, phonological short-term memory played an important role; there was also
evidence, however, that explicit processes were strongly involved in
learning these patterns.
It is interesting to compare these findings to those of
Taraban (2004), who studied induction of gender-like categories
in miniature linguistic systems. He did not take individual
difference measures but found that learning was greatly
facilitated either by providing explicit instruction or by drawing
learners attention to the correlated sets of grammatical
morphemes by means of blocking trials as a function of noun
category. When time is limited and the pattern is made salient,
explicit learning (presumably drawing on aptitude) is clearly
important; when there is more time and when the pattern is far
less salient, as would be the case in naturalistic language
acquisition, the role of more implicit learning, relying more
heavily on mere associative memory, is likely to increase.
Few researchers so far seem to have ventured further and
investigated how linguistic characteristics of the patterns to
be acquired interact with both individual differences and
instructional conditions at the same time. Robinson (2002), for
instance, provided interesting data on the interaction between
aptitudes, on the one hand, and linguistic characteristics of the
structures to be learned, instructional conditions, or testing
conditions, on the other hand, but not on the three-way
interaction between linguistic characteristics, instructional
conditions, and aptitudes. Robinson (1996), on the other hand,
documented an interaction between linguistic characteristics

DeKeyser

17

and instructional conditions (i.e., instructed learners outperformed

others in learning simple rules), but he did not investigate an
interaction with aptitude. Robinson (1997), however, did
document a three-way interaction between structure, aptitude,
and instructional condition, in the sense that the learning of easy
structures was predicted by grammatical sensitivity but not by
memory and the learning of complex structures by memory but
not by grammatical sensitivity, for instance, and that this pattern
only obtained for one of the four instructional conditions (rule
search, i.e., inductive explicit learning).

Implications for Instruction

The nature and degree of difficulty of individual structures
have a range of potential implications for instructional decision
making. At the most basic level, one can argue that instruction is
not necessary for the easiest structures and doomed to failure for
the hardest, in particular where focus on form is concerned
(cf. DeKeyser, 2003). Within form-focused instruction, however,
whether it be characterized by traditional focus on forms or more
narrowly defined focus on form (cf. Long & Robinson, 1998),
different activities are likely to have a differential impact on
different structures characterized by different learning
problems. Larsen-Freeman (2003, pp. 117120), for instance,
made distinctions among three kinds of activities, aimed at
association (e.g., through phrase combination tasks), frequent
use, and choice (e.g., in a fill-in-the blanks format). She
associated these three kinds of activities with the problems of
learning the meaning of a grammar structure, the form itself, or
its use, respectively. Within the terminological framework I have
used in the present article, one could argue that association
activities are particularly useful when the learning issue is one
of form-meaning mapping, that frequent use should be the goal
when the problem is one of semantically redundant form-form
mapping, and that choice among forms should be the focus of the
activity when novel meanings are at issue.

18

What Makes Learning L2 Grammar Difficult?

Much empirical work needs to be done in order to test

hypotheses such as these, but they should provide a useful
starting point for research that goes beyond the simplistic
question of whether explicit grammar teaching and systematic
practice are useful for L2 grammar learning.

The Articles in This Volume

The four articles that follow were all published in Language
Learning between 1999 and 2003. They are all about the
acquisition of L2 morphology. Three are about the acquisition
of gender in particular: Kempe and Brooks demonstrate how
diminutives facilitate gender acquisition in Russian L2 by
eliminating nontransparent morphophonological marking. Carroll
shows that gender acquisition in French L2 is largely
determined by the semantic distinctions the learner makes on
the basis of previous linguistic experience. Williams and Lovatt
document the role of individual differences, in particular,
phonological short-term memory, in the acquisition of gender
in a semiartificial miniature linguistic system.
These three studies illustrate with different methodologies
and with different languages how factors such as consistency of
form-meaning mapping, semantically driven insights derived
from prior linguistic knowledge, and the learners phonological
short-term memory all play an important role in solving a
difficult problem in acquisition: form-meaning mapping, in which
the form is complex and the meaning redundant, abstract, and
novel. Together with Goldschneider and DeKeysers meta-analysis
of a wider range of morphological elements, these studies show
how success in L2 acquisition is strongly influenced by (a) how
transparent the form-meaning link is to the learner, either
because of the salience of the linguistic structure itself or
because its apperception is facilitated by the structure or the
frequency of the input, or (b) the learners aptitudes and
previous linguistic experience.

DeKeyser

19

References
Andersen, R. W., & Shirai, Y. (1994). Discourse motivations for some
cognitive acquisition principles. Studies in Second Language Acquisition,
16, 133156.
Andersen, R. W., & Shirai, Y. (1996). The primacy of aspect in first and second
language acquisition: The pidgin-creole connection. In W. C. Ritchie &
T. K. Bhatia (Eds.), Handbook of second language acquisition (pp.
527570). San Diego, CA: Academic Press.
Ayoun, D. (2004). The effectiveness of written recasts in the second
language acquisition of aspectual distinctions in French: A follow-up
study. Modern Language Journal, 88, 3155.
Bardovi-Harlig, K. (1987). Markedness and salience in second-language
acquisition. Language Learning, 37, 385407.
Bardovi-Harlig, K. (1998). Narrative structure and lexical aspect: Conspiring
factors in second language acquisition of tense-aspect morphology.
Studies in Second Language Acquisition, 20, 471508.
Bardovi-Harlig, K. (1999). From morpheme studies to temporal semantics:
Tense-aspect research in SLA. Studies in Second Language Acquisition,
21, 341382.
Bardovi-Harlig, K. (2000). Tense and aspect in second language acquisition:
Form, meaning, and use. Malden, MA: Blackwell.
Bley-Vroman, R., & Joo, H.-R. (2001). The acquisition and interpretation of
English locative constructions by native speakers of Korean. Studies in
Second Language Acquisition, 23, 207219.
Birdsong, D. (1992). Ultimate attainment in second language acquisition.
Language, 68, 706755.
Butler, Y. G. (2002). Second language learners theories on the use of
English articles: An analysis of the metalinguistic knowledge used by
Japanese students in acquiring the English article system. Studies in
Second Language Acquisition, 24, 451480.
Celce-Murcia, M., & Larsen-Freeman, D. (1999). The grammar book: An
ESL/EFL teachers course. Boston: Heinle & Heinle.
Collins, L. (2002). The roles of L1 influence and lexical aspect in the
acquisition of temporal morphology. Language Learning, 52, 4394.
DeKeyser, R. M. (2000). The robustness of critical period effects in second
language acquisition. Studies in Second Language Acquisition, 22, 499533.
DeKeyser, R. M. (2003). Implicit and explicit learning. In C. Doughty &
M. Long (Eds.), Handbook of second language acquisition (pp. 313348).
Oxford, UK: Blackwell.

20

What Makes Learning L2 Grammar Difficult?

DeKeyser, R. M. (2005). Monitoring processes in Spanish as a second

language during a study abroad program. Manuscript in preparation.
DeKeyser, R. M., Alfi-Shabtay, I., Ravid, D., & Shi, M. (2005). The role of
salience in the acquisition of Hebrew as a second language. Manuscript
in preparation.
DeKeyser, R. M., Ravid, D., & Alfi-Shabtay, I. (2005). Cross-linguistic
evidence for the nature of age effects in second language acquisition.
Manuscript in preparation.
DeKeyser, R. M., Salaberry, R., Robinson, P., & Harrington, M. (2002). What
gets processed in processing instruction? A commentary on Bill VanPattens
Processing Instruction: An Update. Language Learning, 52, 805823.
Dietrich, R., Klein, W., & Noyau, C. (1995). The acquisition of temporality
in a second language. Amsterdam: Benjamins.
Eckman, F. R., Bell, L., & Nelson, D. (1988). On the generalization of relative
clause instruction in the acquisition of English as a second language.
Applied Linguistics, 9, 120.
Eckman, F. R., Moravcsik, E. A., & Wirth, J. R. (1989). Implicational
universals and interrogative structures in the interlanguage of ESL
learners. Language Learning, 39, 173205.
Ellis, N. (2002). Frequency effects in language processing: A review with
implications for theories of implicit and explicit language acquisition.
Studies in Second Language Acquisition, 24, 143188.
Ellis, N. (2003). Constructions, chunking, and connectionism: The emergence of
second language structure. In C. J. Doughty & M. H. Long (Eds.), Handbook
of second language acquisition (pp. 63103). Oxford, UK: Blackwell.
Ellis, R. (1990). Instructed second language acquisition. Oxford, UK:
Blackwell.
Flege, J. E., Yeni-Komshian, G. H., & Liu, S. (1999). Age constraints on secondlanguage acquisition. Journal of Memory and Language, 41, 78104.
Gass, S. M. (1979). Language transfer and universal grammatical relations.
Language Learning, 29, 327344.
Gass, S. M., & Selinker, L. (2001). Second language acquisition. An introductory course. Mahwah, NJ: Erlbaum.
Gomez, R. L. (1997). Transfer and complexity in artificial grammar learning.
Cognitive Psychology, 33, 154207.
Han, Z. (2000). Persistence of the implicit influence of NL: The case of the
pseudo-passive. Applied Linguistics, 21(1), 5582.
Hansen, L., & Chen, Y. L. (2001). What counts in the acquisition and
attribution of numeral classifiers? JALT Journal, 23, 90110.
Herschensohn, J. (2000). The second time around: Minimalism and L2
acquisition. Amsterdam: Benjamins.

DeKeyser

21

Hertel, T. J. (2003). Lexical and discourse factors in the second language

acquisition of Spanish word order. Second Language Research, 19, 273304.
Huang, C.-T. (1982). Logical relations in Chinese and the theory of
grammar. Unpublished doctoral dissertation, Massachusetts Institute
of Technology, Cambridge, MA.
Inagaki, S. (1997). Japanese and Chinese learners acquisition of the
narrow-range rules for the dative alternation in English. Language
Learning, 47, 637669.
Inagaki, S. (2001). Motion verbs with goal PPs in the L2 acquisition of English
and Japanese. Studies in Second Language Acquisition, 23, 153170.
Ishida, M. (2004). Effects of recasts on the acquisition of the aspectual form
-te i-(ru) by learners of Japanese as a foreign language. Language
Learning, 54, 311394.
Jarvis, S. (2002). Topic continuity in L2 English article use. Studies in
Second Language Acquisition, 24, 387418.
Jiang, N. (2004). Morphological insensitivity in second language processing.
Applied Psycholinguistics, 25, 603634.
Johnson, J. S., & Newport, E. L. (1989). Critical period effects in second
language learning: The influence of maturational state on the acquisition
of English as a second language. Cognitive Psychology, 20, 6099.
Johnson, J. S., & Newport, E. L. (1991). Critical period effects on universal
properties of language: The status of subjacency in the acquisition of a
second language. Cognition, 39, 215258.
Joo, H.-R. (2003). Second language learnability and the acquisition of
English locative verbs by Korean speakers. Second Language Research,
19, 305328.
Juffs, A. (1996). Semantics-syntax correspondences in second language
acquisition. Second Language Research, 12, 177221.
Juffs, A. (1998). The acquisition of semantics-syntax correspondences and verb
frequencies in EL materials. Language Teaching Research, 2, 93123.
Jung, E. H. (2004). Topic and subject prominence in interlanguage
development. Language Learning, 54, 713738.
Kempe, V., & MacWhinney, B. (1998). The acquisition of case marking by
adult learners of Russian and German. Studies in Second Language
Acquisition, 20, 543587.
Klein, W., & Dittmar, N. (1979). Developing grammars. Berlin: Springer.
Krashen, S. D. (1982). Principles and practice in second language
acquisition. Englewood Cliffs, NJ: Prentice-Hall.
Krashen, S. D., & Pon, P. (1975). An error analysis of and advanced ESL
learner. Working Papers in Bilingualism, 7, 125129.

22

What Makes Learning L2 Grammar Difficult?

Lardiere, D. (1998). Case and tense in the fossilized steady-state. Second

Language Research, 14, 126.
Lardiere, D. (2000). Mapping features to forms in second language
acquisition. In J. Archibald (Ed.), Second language acquisition and
linguistic theory (pp. 102129). Oxford, UK: Blackwell.
Larsen-Freeman, D. (2003). Teaching language: From grammar to
grammaring. Boston: Heinle.
Larsen-Freeman, D., & Long, M. H. (1991). An introduction to second
language acquisition research. New York: Longman.
Lee, E.-J. (2001). Interlanguage development by two Korean speakers of
English with a focus on temporality. Language Learning, 51, 591633.
Leeman, J. (2003). Recasts and second language development: Beyond
negative evidence. Studies in Second Language Acquisition, 25, 3763.
Liceras, J. M. (1989). On some properties of the pro-drop parameter: Looking
for missing subjects in nonnative Spanish. In S. Gass & J. Schachter (Eds.),
Linguistic perspectives on second language acquisition (pp. 109133). New
York: Cambridge University Press.
Liu, D., & Gleason, J. L. (2002). Acquisition of the article the by nonnative
speakers of English: An analysis of four nongeneric uses. Studies in
Second Language Acquisition, 24, 126.
Long, M. H. (1997). Fossilization: Rigor mortis in living linguistic systems?
Plenary address to the EUROSLA 97 conference, Universitat Pompeu
Fabra, Barcelona, Spain.
Long, M. H. (2003). Stabilization and fossilization in interlanguage
development. In C. J. Doughty & M. H. Long (Eds.), Handbook of second
language acquisition (pp. 487535). Oxford, UK: Blackwell.
Long, M. H., & Robinson, P. (1998). Focus on form: Theory, research, and
practice. In C. Doughty & J. Williams (Eds.), Focus on form in classroom
second language acquisition (pp. 1541). New York: Cambridge University
Press.
Long, M. H., & Sato, C. (1984). Methodological issues in interlanguage
studies: An interactionist perspective. In A. Davies, C. Criper, & A. P. R.
Howatt (Eds.), Interlanguage (pp. 253279). Edinburgh, UK: Edinburgh
University Press.
MacWhinney, B. (2001). The competition model: The input, the context, and
the brain. In P. Robinson (Ed.), Cognition and second language instruction
(pp. 6990). New York: Cambridge University Press.
MacWhinney, B. (in press). A unified model of language acquisition. In
J. F. Kroll & A. M. B. de Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches. Oxford, UK: Oxford University Press.

DeKeyser

23

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

pedagogy. System, 25, 215232.
McDonald, J. L. (1987). Sentence interpretation in bilingual speakers of
English and Dutch. Applied Psycholinguistics, 8, 379414.
McDonald, J. L. (2000). Grammaticality judgments in a second language:
Influences of age of acquisition and native language. Applied Psycholinguistics, 21, 395423.
Montrul, S. (2001). Agentive verbs of manner of motion in Spanish and
English as second languages. Studies in Second Language Acquisition,
23, 171206.
Montrul, S., & Slabakova, R. (2003). Competence similarities between native
and near-native speakers: An investigation of the preterite-imperfect
contrast in Spanish. Studies in Second Language Acquisition, 25, 351398.
Papp, S. (2000). Stable and developmental optionality in native and non-native
Hungarian grammars. Second Language Research, 16, 173200.
Perruchet, P., & Pacteau, C. (1990). Synthetic grammar learning: Implicit rule
abstraction or explicit fragmentary knowledge? Journal of Experimental
Psychology: General, 119, 264275.
Perruchet, P., & Pacteau, C. (1991). Implicit acquisition of abstract knowledge
about artificial grammar: Some methodological and conceptual issues.
Journal of Experimental Psychology: General, 120, 112116.
Pesetsky, D. (1987). Wh-in-situ: Movement and unselective binding. In
E. Reuland & A. G. B. ter Meulen (Eds.), The representation of indefiniteness (pp. 98129). Cambridge, MA: MIT Press.
Pienemann, M. (1984). Psychological constraints on the teachability of
languages. Studies in Second Language Acquisition, 6, 186214.
Pienemann, M. (2003). Language processing capacity. In C. J. Doughty & M.
H. Long (Eds.), Handbook of second language acquisition (pp. 679714).
Oxford, UK: Blackwell.
Pinker, S. (1989). Learnability and cognition: The acquisition of argument
structure. Cambridge, MA: MIT Press.
Prevost, P., & White, L. (2000). Missing surface inflection or impairment in
second language acquisition? Evidence from tense and agreement. Second
Language Research, 16, 103133.
Reed, J. M., & Johnson, P. J. (1998). Implicit learning: Methodological issues and
evidence of unique characteristics. In M. A. Stadler & P. A. Frensch (Eds.),
Handbook of implicit learning (pp. 261294). Thousand Oaks, CA: Sage.
Rizzi, L. (1982). Issues in Italian syntax. Dordrecht, The Netherlands: Foris.
Robertson, D. (2000). Variability in the use of the English article system by
Chinese learners of English. Second Language Research, 16, 135172.

24

What Makes Learning L2 Grammar Difficult?

Robinson, P. (1996). Learning simple and complex second language rules

under implicit, incidental, rule-search, and instructed conditions.
Studies in Second Language Acquisition, 18, 2767.
Robinson, P. (1997). Individual differences and the fundamental similarity
of implicit and explicit adult second language learning. Language
Learning, 47(1), 4599.
Robinson, P. (2002). Effects of individual differences in intelligence,
aptitude, and working memory on adult incidental SLA: A replication
and extension of Reber, Walkenfield, and Hernstadt (1991). In
P. Robinson (Ed.), Individual differences and instructed language
learning (pp. 211265). Amsterdam/Philadelphia: Benjamins.
Salaberry, R. (2000). The development of past tense morphology in L2
Spanish. Philadelphia: Benjamins.
Sorace, A. (2000). Syntactic optionality in non-native grammars. Second
Language Research, 16, 93102.
Sorace, A. (2003). Near-nativeness. In C. J. Doughty & M. H. Long (Eds.),
Handbook of second language acquisition (pp. 130151). Oxford, UK:
Blackwell.
Sorace, A., & Shomura, Y. (2001). Lexical constraints on the acquisition of
split intransitivity: Evidence from L2 Japanese. Studies in Second
Language Acquisition, 23, 247278.
Sprouse, R. (1998). Some notes on the relationship between inflectional
morphology and parameter setting in first and second language
acquisition. In M. Beck (Ed.), Morphology and its interfaces in second
language knowledge (pp. 4167). Amsterdam: Benjamins.
Stockwell, R. P., Bowen, J. D., & Martin, J. W. (1965). The grammatical
structures of English and Spanish. Chicago: University of Chicago Press.
Taraban, R. (2004). Drawing learners attention to syntactic context aids genderlike category induction. Journal of Memory and Language, 51, 202216.
Tarone, E. E. (1985). Variability in interlanguage use: A study of style-shifting
in morphology and syntax. Language Learning, 35, 373404.
Tarone, E. E., & Parrish, B. (1988). Task-related variation in interlanguage: The case of articles. Language Learning, 38, 2144.
Thomas, M. (1989). The acquisition of English articles by first- and secondlanguage learners. Applied Psycholinguistics, 10, 335355.
Truscott, J., & Sharwood Smith, M. (2004). Acquisition by processing: A
modular perspective on language development. Bilingualism: Language
and Cognition, 7, 120.
VanPatten, B. (1990). Attending to form and content in the input: An
experiment in consciousness. Studies in Second Language Acquisition,
12, 287301.

DeKeyser

25

VanPatten, B. (2002a). Processing instruction: An update. Language Learning,

52, 755803.
VanPatten, B. (2002b). Processing the content of input-processing and
processing instruction research: A response to DeKeyser, Salaberry,
Robinson, and Harrington. Language Learning, 52, 825831.
VanPatten, B. (2004). Input processing in second language acquisition. In
B. VanPatten (Ed.), Processing instruction: Theory, research, and
commentary (pp. 531). Mahwah, NJ: Erlbaum.
Whong-Barr, M., & Schwartz, B. D. (2002). Morphological and syntactic
transfer in child L2 acquisition of the English dative alternation. Studies
in Second Language Acquisition, 24, 579616.
Williams, J. N. (1999). Memory, attention, and inductive learning. Studies
in Second Language Acquisition, 21, 148.
Wong, W. (2001). Modality and attention to meaning and form in the input.
Studies in Second Language Acquisition, 23, 345368.
Yeni-Komshian, G., Robbins, M., & Flege, J. E. (2001). Effects of word class
differences on L2 pronunciation accuracy. Applied Psycholinguistics, 22,
283299.
Young, R. (1996). Form-function relationships in articles in English interlanguage. In R. Bayley & D. Preston (Eds.), Second language acquisition
and linguistic variation (pp. 135175). Amsterdam: Benjamins.