An Introduction to Applied Linguistics

particular form is ungrammatical) might be sufficient to have learners reset the 
parameters of UG principles in order to reflect the differences between the native 
language and target language grammars (White, 1987). Others felt that explicit 
grammar teaching had a role (Norris and Ortega, 2000), with some claiming 
that explicit attention to grammar was essential for older language learners 
whose ability to acquire language implicitly, much as children learn their native 
language, was no longer possible, or at least no longer efficient.
Second language acquisition (SLA) research in both naturalistic and classroom 
environments has informed modern perspectives of grammar learning (see Chapter 
7, Second Language Acquisition). SLA research tells us that an analysis of the language 
that learners use, their ‘interlanguage’, reveals that grammar is not acquired in a 
linear fashion, one structure being mastered after another. Further, with regard to 
any one structure, learners use a lot of intermediate forms before conforming to 
what is accurate in the target language. It can easily be seen that many learners’ 
utterances are overgeneralizations. For example, learners of English produce 
‘eated’ for ‘ate’, interpreted by some researchers as evidence for the process of rule 
formation in SLA. Learners also use forms that do not resemble target forms, and 
they do so consistently, such as using pre-verbal negation during early English 
language acquisition (for example, ‘no want’), regardless of the native language 
of learners. This behaviour explains why it has been said that the interlanguage is 
systematic, that is, learners operate consistently within a system, albeit one that is 
not consonant with the target language. New structures are not simply assimilated 
one by one, but rather as a new structure makes its appearance into a learner’s 
interlanguage, the learner’s system begins to shift. Thus, learning does not add 
knowledge to an unchanging system – it changes the system (Feldman, 2006).
It is also clear, however, that rule formation does not account for all of grammar 
learning. Indeed, some would argue that it has no role in SLA at all. Connectionist 
or neural network models support such a conclusion (Ellis, 1998). Repeated 
exposure to target language forms contributes to the strengthening of connections 
in neural network models. The models simulate rule-like grammatical behaviour 
even though no rules or algorithms are used in constructing the model. Instead, 
patterns are abstracted from the way structures are statistically distributed in 
massive amounts of input data. With the use of connectionism to simulate the 
way that neural networks in the brain function (see Chapter 7, Second Language 
Acquisition), new ways of conceptualizing grammar learning are coming to the 
forefront.
One method that is receiving a great deal of attention is emergentism (Ellis 
and Larsen-Freeman, 2006; Larsen-Freeman, 2006). Emergentists believe that 
rather than speakers’ performance being managed by a ‘top-down’ rule-governed 
system, learners’ interlanguage emerges from repeated encounters with structures 
and with opportunities to use them. In this way, it could be said that language 
learning is an iterative process, revisiting the same or similar territory again and 
again (Larsen-Freeman and Cameron, 2008). Thus, grammar learning is facilitated 
by the frequency of use of the forms in the language to which the learner is 
exposed. The Zipfian profile of language, in which certain forms are used very 

---

28 An Introduction to Applied Linguistics
frequently while others are used far less so, facilitates the process of abstracting 
the patterns (Ellis and Larsen-Freeman, 2009). The fact that frequently-occurring 
constructions are often semantically concrete and short in length aids the learning 
process (Goldberg, 2006).
Regardless of which type (or types, as is more likely the case) of process is 
responsible for learning, SLA research makes clear to most researchers that some 
attention must be given to grammar by second language learners. However, 
it is also clear that the attention to form should not come through the use of 
decontextualized drills or isolated grammar exercises. Learners will be able 
to complete the exercises satisfactorily when their attention is focused on the 
grammar, but when their attention shifts to a more communicative interaction, 
the grammar will be forgotten. In order for learners to be able to transfer what 
they have learned in the classroom to more communicative contexts outside 
it, pedagogical activities have to be psychologically authentic, where there is 
alignment between the conditions of learning and the conditions of subsequent 
use (Segalowitz, 2003).
Further, for new forms to be incorporated into the intermediate language, or 
‘interlanguage’, that learners speak, it is thought that students must first notice 
what it is they are to learn (Schmidt, 1990). Until they do, the target form may 
merely remain as part of the ‘noise’ in the input. Then too, even when they are 
able to produce grammatical structures accurately, students still need to learn 
what they mean and when they are used. In other words, learning grammar does 
not merely entail learning form.
In fact, as we noted earlier, what needs to be learnt about grammar can be 
characterized by three dimensions: form, meaning and use. We have seen in Figure 
2.1 that the dimensions are interconnected, but nonetheless can be described 
discretely. For example, in learning the rule of English subject–verb agreement 
discussed above, students would have to learn the form that an ‘s’ is added to the 
verb stem and that the orthographic ‘s’ may be realized in pronunciation as one 
of three allomorphs /s/, /z/ or /
∂z/. (The slashes indicate sounds; see Chapter 9,