Receptive Versus Productive

Grammar Knowledge
NATSUKO SHINTANI

­Framing the Issue

Theories of language differ in terms of whether they view grammatical knowl-

edge as comprising a single unified store or separate stores responsible for com-
prehension and production. Examples of unified store models are universal
grammar (UG), information-processing models, and connectionism. Skill-learning
theory assumes that comprehension and production are distinct skills drawing on
separate knowledge stores.
Both receptive and productive grammar knowledge can be either implicit or
explicit. Implicit linguistic knowledge is knowledge of language. It is intuitive,
automatized, and used without awareness. Fluent language use draws on implicit
knowledge for both comprehension and production. Explicit knowledge is knowl-
edge about language. It is held consciously and language users are able to tell what
they know. It is available only through controlled processing and for this reason is
more easily accessed in reading and writing than in listening or speaking. Theories
of L2 acquisition research are primarily concerned with how learners develop
implicit knowledge. For this reason, the following sections focus on implicit lin-
guistic knowledge to consider how various theoretical positions view receptive
and productive grammar knowledge.

Unified Store Models

According to generative theories of grammar such as universal grammar (UG) the
acquisition of the grammar of a particular language draws on innate knowledge of
the abstract principles that underlie the form of the grammar of any language.
Generative theories distinguish competence and performance but aim to explain
only the former. In other words they assume that there is a single knowledge
store––competence––that is drawn on for both reception and production.

The TESOL Encyclopedia of English Language Teaching, First Edition.

Edited by John I. Liontas (Project Editor: Margo DelliCarpini).
© 2018 John Wiley & Sons, Inc. Published 2018 by John Wiley & Sons, Inc.
DOI: 10.1002/9781118784235.eelt0095
2 Receptive Versus Productive Grammar Knowledge

Apperceived input – Comprehended input – Intake – Integration – Output

Figure 1  A model of second language acquisition. Reproduced with permission from Gass
(1997, p. 3) © Taylor and Francis Group, LLC, a division of Informa plc.

Differences in the processes involved in comprehension and production are

accounted for by a separate theory of performance.
According to information-processing models grammatical knowledge is acquired
when learners comprehend input, notice exemplars of specific grammatical features
in working memory (i.e., intake them), and then integrate them into long-term mem-
ory. Learners then draw on their internalized system for production. While there are
some differences among information-processing theories, researchers generally agree
with the L2 acquisition model proposed by Gass (1997) as shown in Figure 1.
Some positions view comprehending input by itself as sufficient for intake and
integration to take place while others claim that some level of conscious noticing is
also needed. Irrespective of these differences, information-processing theories see
acquisition as primarily input-driven with output as the manifestation of the stored
system. However, a case has also been made for a formative role for production in
the developing linguistic system. Swain (1985), for example, proposed that produc-
tion pushes syntactical processing, gives learners the opportunity to experiment
with their L2 grammar, and elicits corrective feedback on their use of it. Common
to all information-processing theories is that there is a single grammar store that
­develops through processing input (and in some models also through ­production)
and that is utilized in both the comprehension and production of the L2.
Whereas information-processing models view grammar as a system of rules,
connectionist models view grammar as a network which develops as neural
­connections are made, strengthened, and modified, leading not to rule-based
behavior but to rule-like behavior. Acquisition of what linguists call “grammar”
takes place through the cumulative experience of using the language meaning-
fully (mostly through comprehension of the language). Acquisition of implicit
knowledge is usage-driven. Learners store ready-made chunks from the input
they are exposed to which they gradually and unconsciously analyze over time.
Gradually the connectionist network grows as chunks give way to “constructions”
that provide templates for combining linguistic elements. For example, a child
might hear his/her mother saying “Put it on the table” several times, which is
initially stored as whole. Later the child hears utterances such as “Put it on the
bed” and “Put it on the floor,” again stored as wholes. At some point the child
realizes that these chunks are comprised of a fixed component (“Put it on the ___”)
and a variable component (“table,” “bed,” and “floor”) and arrives at a simple
construction allowing new words (e.g., “chair”) to be inserted in the empty slot.
Later, the learner discovers that the initial word in the chunks (“put”) is also
replaceable (e.g., by “place” or “throw”) and also that “on” can replace “in.” In this
way, the construction becomes increasingly schematic and is activated when both
comprehending input and producing output.
All these positions see grammar knowledge as one construct, viewing it as either
an innate system (in UG), a stored system (in information-processing models), or
 Receptive Versus Productive Grammar Knowledge 3

a neural network (connectionist model), which can be used in both the reception
and production of language.

Separate Store Models

The skill-learning theory distinguishes between declarative knowledge (i.e., the
representation of facts) and procedural knowledge (i.e., the presentation of actions
in particular situations). Skill development involves the proceduralization of
knowledge that exists initially in declarative form. Subsequently, through practice,
procedural knowledge is automatized at which point conscious awareness of a
grammatical form may be lost. A key prediction of skill-learning theory is that the
development of language is skill-specific, that is declarative knowledge is proce-
duralized and automatized through domain-specific practice (i.e., either receptive
or productive practice) (DeKeyser, 1997). A corollary of this position is that there
are separate receptive and productive stores for the grammar of a language.

­Making the Case

To investigate the different positions, two questions need to be answered: (a) what
evidence is there to show that there is a single store drawn on both receptively and
productively, and (b) what evidence is there to show that comprehension and pro-
duction draw on different knowledge stores?
Evidence in support of a single store comes from studies investigating input
processing (VanPatten, 1996). Input processing involves the cognitive mechanisms
responsible for obtaining intake from input by establishing connections between
linguistic forms and their meanings. VanPatten proposed that learners operate in
accordance with default input-processing principles that explain how they allo-
cate their attention during online processing. These principles prevent learners
from attending to certain grammatical forms (e.g., verb -ed in English) in the input
with the result that they fail to establish form-meaning mappings (e.g., the V-ed
signals a completed action in the past). Thus learners need to abandon these inhib-
iting processing strategies in order for acquisition to take place. VanPatten sug-
gested that this can be accomplished by means of processing instruction. Processing
instruction builds on Krashen’s idea that acquisition is entirely input-driven (i.e.,
learners do not need to produce a grammatical form to learn it) but assumes that
comprehensible input by itself is not sufficient to ensure learners will automati-
cally acquire the full target language grammar and that input needs to be specially
structured to direct attention at difficult-to-acquire grammatical forms. Processing
instruction also optionally includes explicit instruction consisting of grammar
explanation and input processing strategy training (information about default
input processing principle and how to avoid it).
A number of studies have investigated the effects of processing instruction.
These studies measured the learning by means of tests both of learners’ ability to
comprehend and to produce the grammatical structures targeted by the instruc-
tion. Overall, the results show that structured input activities lead to gains in both
4 Receptive Versus Productive Grammar Knowledge

comprehension and production (see Shintani, 2015a, for a review). In other words,
the studies provide evidence that input-based instruction results in the internali-
zation of grammatical knowledge that can be used for both comprehension and
production as shown in Gass’s model.
However, there is also evidence for the alternative view, namely that receptive
and productive knowledge are skill-specific. DeKeyser and Sokalski (1996) inves-
tigated the effects of comprehension and production practice on the acquisition of
two structures that differed in terms of grammatical complexity. The immediate
test results showed that the input group significantly outperformed the control
group in the comprehension tests for both structures, while the output group out-
performed the control group in the production tests for both structures and the
comprehension test for the complex structure. However, the significant differences
disappeared in the delayed post-test conducted one week later. DeKeyser and
Sokalski claimed that the results largely support the predictions of skill-learning
theory but that “both testing time and the morphosyntactic nature of the structure
in question favor one skill or the other” (p. 615).
A close look at the results of the processing instruction studies also lends some
support for skill-specificity. In a meta-analysis of 42 experiments that had investi-
gated effects of structured input and production-based instruction on the compre-
hension and production of different target structures, Shintani (2015a) reported
that overall structured input led to higher scores in the receptive tests but similar
scores in the productive tests. However, when both groups received the same
explicit instruction, the production-based instruction resulted in higher scores on
the production test than structured input instruction. DeKeyser and Prieto Botana
(2015) argued that the results of processing instruction studies are best explained
by skill-learning theory. That is, in general input-based practice leads to the proce-
duralization of receptive grammatical knowledge and output-based practice helps
to proceduralize productive knowledge.
Most of the processing instruction studies investigated the effects of instruction
on grammatical structures that the learners already had partial knowledge of (i.e.,
they were not completely new structures). Shintani (2015b), however, compared
the effects of input-based and production-based instruction on the incidental
acquisition of a structure (English plural -s) by young children who had no prior
knowledge of this structure. Those learners who experienced production-based
instruction developed neither receptive nor productive knowledge of plural -s.
Those learners who experienced the input-based instruction developed receptive
knowledge but with a few exceptions they did not acquire productive knowl-
edge. The exceptions were two students who had freely produced plural nouns
during the lessons. An interpretation of these results is that knowledge of a new
grammatical form is initially receptive only and, as predicted by skill-learning
theory, is developed through comprehension-based rather than production-based
instruction. However, once receptive knowledge is available, opportunities to
produce the new form help productive knowledge to develop.
Studies investigating the neurobiological bases of comprehension and produc-
tion also suggest a degree of disassociation. People who suffer from Broca’s
 Receptive Versus Productive Grammar Knowledge 5

aphasia have difficulty in producing grammatically complex sentences. Their

speech is telegraphic and lacks function words and morphological markers. In
contrast, patients suffering from Wernicke’s aphasia, where there is damage to a
different part of the brain, lose the ability to understand language although they
can speak clearly. However, neuro-imaging studies (see, for example, Gernsbacher
& Kaschak, 2003) paint a more complicated picture. They show that multiple areas
of the brain in both the left and right hemispheres are involved in both comprehen-
sion and production and that there are many-to-many mappings of structure to
function and that no specific areas of the brain are exclusively involved in compre-
hension and production. Neuroimaging studies (see Abutalebi, Cappa, & Perani,
2001) also indicate that the neural machinery involved in both comprehension and
production varies according to the proficiency level of the learner. Disassociation
is more clearly evident in native speakers and in learners with high L2 proficiency.
In low proficiency learners, a less well-defined network is activated for both com-
prehension and production. Thus the neurobiological research raises the intrigu-
ing possibility that the relatively specialized regions of the brain involved in
comprehension and production develop as proficiency develops. It is clear, how-
ever, that even in fully developed linguistic systems, there is considerable overlap-
ping in the brain structures utilized for comprehension and production.

Summary
The evidence points to a degree of separation between the grammar systems
responsible for reception and production, as claimed by skill-learning theory, but
the research also shows that both comprehension-based and production-based
instruction can benefit both reception and production ability at least for grammati-
cal features that have already been partially internalized. Perhaps the best conclu-
sion––supported by the neurobiological research––is that there is a degree of
separation in the knowledge stores drawn on in comprehension and production
but that these stores are connected and interface with each other.

­Pedagogical Implications

If one assumes that grammatical knowledge resides in a single store feeding both
comprehension and production and that this store develops primarily through
exposure to input, as assumed in both information-processing and connectionist
models, then it follows that grammar can be most effectively taught via input
rather than output. This is the position that VanPatten adopts. Comprehensible
input is necessary but not sufficient. Structured input is needed to induce learners
to process those grammatical features they find difficult to acquire “naturally.”
Once they have acquired them they will be able to use them in production although
they may need to practice doing so to achieve automaticity. If one assumes that
comprehension and production draw on separate stores of grammatical knowl-
edge––as skill-learning theory does––then it will be necessary to provide both
6 Receptive Versus Productive Grammar Knowledge

input-based and production-based practice activities to ensure automatized skill

in both comprehension and production.
By and large, grammar teaching materials are production-oriented, reflecting an
underlying belief that, unless learners can produce a grammatical structure, they do
not “know” it. There is no basis for such a belief. It is clear that learners can acquire
a grammatical structure without having to produce it. An approach that is arguably
compatible with both single store and dual store models is as follows: (a) providing
input-based instruction that can help to develop receptive knowledge of completely
new structures; and (b) once a structure is established receptively, employing a mix-
ture of input-based instruction and production-based instruction to automatize
receptive knowledge and to develop the learner’s production of grammar.
Low-proficiency learners in general may benefit more from input-based grammar
instruction as it is less taxing on their limited capacity of working memory. More
advanced-level learners will benefit more from being “pushed” to produce those
grammatical structures they know but cannot yet use correctly. Advanced-level
learners may continue to need practice in processing already-acquired structures
receptively in real-time input.

SEE ALSO: Explicit Versus Implicit Grammar Knowledge; Incidental Learning

of Vocabulary; Skills-Based Training; Universal Grammar

­References

Abutalebi, J., Cappa, S. F., & Perani, D. (2001). The bilingual brain as revealed by functional
neuroimaging. Bilingualism: Language and Cognition, 4(2), 179–90.
DeKeyser, R. M. (1997). Beyond explicit rule learning: Automatizing second language
morphosyntax. Studies in Second Language Acquisition, 19, 195–221.
DeKeyser, R. M., & Botana, G. P. (2015). The effectiveness of processing instruction in L2
grammar acquisition: A narrative review. Applied Linguistics, 36(3). doi:10.1093/applin/
amu071
DeKeyser, R. M., & Sokalski, K. J. (1996). The differential role of comprehension and
production practice. Language Learning, 46(4), 613–42.
Gass, S. M. (1997). Input, interaction, and the second language learner. Mahwah, NJ: Lawrence
Erlbaum.
Gernsbacher, M. A., & Kaschak, M. P. (2003). Neuroimaging studies of language production
and comprehension. Annual Review of Psychology, 54, 91–114. doi:10.1146/annurev.
psych.54.101601.145128
Shintani, N. (2015a). The effectiveness of processing instruction on L2 grammar acquisition:
A meta-analysis. Applied Linguistics, 36(3), 306–25.
Shintani, N. (2015b). The incidental grammar acquisition in focus on form and focus on
forms instruction for young beginner learners. TESOL Quarterly. doi:10.1002/tesq.166
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and
comprehensible output in its development. In S. Gass & C. Madden (Eds.), Input in second
language acquisition (pp. 235–56). Rowley, MA: Newbury House.
VanPatten, B. (1996). Input processing and grammar instruction: Theory and research. Norwood,
NJ: Ablex.