See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261626905

Complex systems and applied linguistics

Article in Southern African Linguistics and Applied Language Studies · April 2010
DOI: 10.2989/SALALS.2009.27.2.9.872

CITATIONS READS

3 4,210

1 author:

Albert Weideman
University of the Free State
121 PUBLICATIONS 1,206 CITATIONS

SEE PROFILE

All content following this page was uploaded by Albert Weideman on 18 April 2017.

The user has requested enhancement of the downloaded file.

Southern African Linguistics and Applied Language Studies 2009, 27(2): 229–233 Copyright © NISC Pty Ltd
Printed in South Africa — All rights reserved SOUTHERN AFRICAN LINGUISTICS
AND APPLIED LANGUAGE STUDIES
ISSN 1607–3614 EISSN 1727–9461
DOI: 10.2989/SALALS.2009.27.2.9.872

Book Review

Complex systems and applied linguistics

Diane Larsen-Freeman and Lynne Cameron
2008, Oxford University Press, Oxford
ISBN 978 0 19 442244 4
287 pages

Conventionality and fashionability

One seldom has the privilege of reading and reviewing a book that you simply know, from the
outset, is groundbreaking in almost every respect. My own interest in reading it is as a philosopher
of applied linguistics – as someone interested in the foundations and history of the field.
An interest in foundations means that one is intrigued by theories and paradigms that have
informed the discipline. It is an interest that is therefore closely connected with the history of the
field, since, as in all other disciplines, new paradigms come and go with regularity, and their histor-
ical influence expands and contracts as applied linguists pick up on new trends, or discard conven-
tional ways of doing their work as they embrace the new. What has always been a concern to me
is how applied linguists actually make choices among the various styles of doing it. Now that we
are faced with a potential new way of doing applied linguistics, that question becomes even more
urgent. How do we select the theories that underlie our designs? And if we don’t consciously and
deliberately do so, how can we then defend such an acceptance of a paradigm in an academi-
cally responsible way? Or how do we know that we are not merely victims of a conventional style of
doing applied linguistic designs? As McNamara (2008: 304) points out:
It is important to keep alive an understanding of the theoretical perspectives that have been
proposed in the past so that their enduring relevance is appreciated and we do not go
on reinventing the wheel … Historical amnesia is a persistent temptation in a practically
oriented intellectual field such as ours.
This concern is, in my opinion, therefore not an idle one at all. If we do not wish to be at the mercy
of what is new and fashionable, if we do not intend falling victim to the latest fad, there should be
some way in which we choose our theoretical foundation. In a much earlier analysis that I did more
than 20 years ago (Weideman, 1987), I pointed out that at the time, the choice for applied linguists
lay between technocratic and revolutionary ways of working within the discipline. This is one way of
categorising the choices for our design work; another might be to look at distinguishing, as I have
also done, between six (or now, with the rise of complexity theory, seven) different and succes-
sive generations or styles of doing applied linguistics, from the earliest, first generation work that
claimed to be a ‘scientific’ approach, to its antithesis in the penultimate, postmodern paradigm (cf.
Pennycook, 1999, 2000, 2004), that claims to be, well, anti-disciplinary:
Rather than viewing critical applied linguistics as a new form of interdisciplinary knowledge,
I prefer to view it as a form of anti-disciplinary knowledge (Pennycook, 2004: 801).
For a discussion of these various styles, I refer to Weideman (1999, 2003, 2007a, 2007b). The
main point is: it is not easy or unproblematic for an applied linguist to make an informed choice of a
theoretical starting point.
And it is exactly a new theoretical basis for applied linguistics that these authors (Larsen-Freeman
& Cameron, 2008) explore in their new book. Although variously labelled as complex systems
theory (as in the title) or as complexity theory, there is for me no doubt that what we have here is
an emerging new paradigm in the field of applied linguistics. Though there have been several other
smaller and related studies (e.g. Cameron & Larsen-Freeman, 2007; De Bot, Lowie & Verspoor,
2007; Kramsch, 2008), Larsen-Freeman and Cameron’s (2008) book is the most ambitious and
detailed exposition to date of the theory and what it means for our field. In discussing it, I shall make
230 Weideman

use below of a number of distinctions that, though from a different angle and for a different purpose,
I have made in another critical analysis of theoretical frameworks for applied linguistics (Weideman,
2009b).

What does taking a complex systems approach mean?

As Larsen-Freeman and Cameron make clear quite early on (2008: x, 5), complex systems thinking
finds its roots in biology. Key concepts therefore echo organic terms, and include the adapta-
bility (2008: 33) and potential of systems, especially their ability to self-organise (2008: 62). There
is an emphasis throughout on ‘the organic nature of change’ within those systems (2008: 1, 17).
The phenomenon of change, which is interpreted as constant, dynamic, ongoing change, is one
that is related in the first instance not to a physical, but to a biotic understanding of things: ‘… an
organism’s ongoing activity continuously changes its neural states, just as growth changes the
physical dimensions of the body’, the authors remark (2008: 17, see too 29, 32, 72). In a complex
systems approach, the emphasis is therefore on dynamics, which requires ‘us to look for change
and for processes that lead to change, rather than for static, unchanging entities’ (2008: 16, 26). We
should specifically note that the emphasis is not as much on analogical physical concepts, such as
dynamic effect or power, as on analogical biotic conceptualisations of phenomena. Phrased differ-
ently: the flux or change that is the focus of the approach is interpreted in an organically dynamic
way.
A second important and related tenet of the new approach is the non-linearity of the processes
of change that it focuses on. A complex system is defined as one whose behaviour is not predict-
able in terms of a single dimension, or set of dimensions, but whose operation emerges from the
interactions of its components (2008: 2). So another important implication of adopting a complexity
approach is the acknowledgement that the behaviour of a system cannot be predicted in a
linear fashion (2008: 72). Therefore causal explanations, so typical of modernist explanations of
phenomena, are no longer sufficient. Instead, since all the various components of a complex system
(including what was previously sometimes inappropriately sidelined as ‘context’) are in continuous
interaction, there is ‘reciprocal causality’ (2008: 7, 60). The processes of change can therefore be
described as a ‘movement in a trajectory across a “state space” or “phase space”’ (2008: 20, 43).
The change process, if drawn towards a sufficiently powerful ‘attractor’ (2008: 20, Chapter 3), can
come to a provisional stability.
How do these insights relate to language and the design of language interventions? What linguists
and applied linguists, including some ethnographers, have conventionally labelled ‘context’, in this
perspective becomes part of the whole interacting system – not an entirely new insight, perhaps,
but in my view a necessary correction. Another example of how this perspective can be applied can
be found in the temporary stability that certain emergent patterns of language use find, in becoming
momentarily stable around the strong attractor of the notion of a standard language (2008: 81).
It should be noted, however, that the emergent patterns that a complex systems approach finds
are but relative stabilities. Complex systems theory acknowledges that no system is wholly free of
change, since the many interacting components of systems each bring with them their own measure
of instability and thus unpredictability. So even small changes that are introduced into a system –
and teaching or instruction in a language can certainly constitute more than small perturbations in
a learner’s language growth trajectory – can have dramatic effects, spreading ‘through the system,
diluting the determinism and rendering the outcome of system activity unpredictable’ (2008: 75).

Complex systems theory, language and learning

In order to design their instructional interventions, applied linguists of course need to discover
and identify regular patterns, both in language and in learning. From a complexity perspec-
tive, the grammatical subsystem of a language is just one of many interacting subsystems and
components which have such recurrently regular patterns (2008: 84). Language develops, from
a complex systems perspective, in a process of co-adaptation, which gives rise to an alignment
of patterns between, for example, learner and interlocutor (2008: 127). One must note again the
biotic terminology: even the relationship between accelerated lexical growth and grammatical
Southern African Linguistics and Applied Language Studies 2009, 27(2): 229–233 231

development is described in organic concepts, as two subsystems that are connected growers
(2008: 149). So, too, learning a language is seen as language development rather than as acquisi-
tion, as a process of dynamic adaptation (2008: 157) rather than as something that, once learned,
is ‘possessed’ for all time. From a complexity point of view, language can never be in an entirely
stable state, so it cannot be ‘acquired’ once and for all.
As regards other lingual systems and subsystems, discourse and discourse types are seen as
multiply interconnected complex systems, and the language using patterns that each discourse type
yields are a resource or language potential that is actualised in each instance of talk or text creation
within such a system (2008: 174). Since humans, as language using agents, ‘soft-assemble’
language from the resources or potential at their disposal, the expectations that derive from
previous experiences of co-creating and aligning oneself with the latent meaning potential of others
through discourse constitute an important resource for them. Linguists familiar with the identification
and treatment of lingual expectations by, for example, conversation analysis, will immediately see
how a complex systems perspective links up with this earlier kind of ethnomethodological analysis.
The lingual expectations that have been identified for interlocutors by conversation analysis become
‘attractors’, or locations of relative stability for language use in discourse creating contexts (2008:
179, 193). A complexity perspective also takes a view on the language using resources needed
to make written texts. These can be viewed as co-constructed, yet asynchronous, collabora-
tive compositions, in which writers imaginatively engage with, and thus interact with, prospective
readers (2008: 188).

Another illustration and some further examples

As we have noted above in respect of the design of language learning opportunities in the classroom,
a complex systems approach emphasises that even small interventions can make a big difference
(2008: 200). What can the new approach actually contribute towards instructional design for language
teaching? A full chapter of Larsen-Freeman and Cameron’s (2008) book deals exactly with this.
To take one illustration of a complexity perspective on a how a classroom task was accomplished,
we may look at their analysis and interpretation (2008: 204f.) of the variation on a language task,
both in terms of language use and the potential to grow the potential resources at the disposal of
learners. The task in the example required of English-language learning Norwegian students to
nominate and then discuss an arctic animal in the target language. By making use of the concept
of an interaction differential, they measure the difference between the demand from the teacher
(from open requests, the most demanding, to closed questions, the least demanding) and the
responses from the learners, that range from minimal (the least desirable) to more expansive (the
most desirable) offerings.
The interactive talk between the teacher and individual learners is then carefully plotted, turn by
turn, in terms of this differential. The analysis shows that the trajectory of the interaction goes from
a high (and therefore pedagogically meaningful) differential towards the powerful, but pedagogically
less helpful attractor of a low differential. The reason for this pedagogically less desirable trajectory is,
perhaps not entirely surprisingly, to be found in the teacher’s efforts to instil some sense of progress,
where there actually is very little. The teacher leads a process of co-adaptation that will ensure at least
some measure of success on this language task. In one significant case, however, the trajectory does
not slide into this less desirable stable condition. This is where a learner chooses an animal of which
he indeed has some knowledge, and on his own initiative increases the interaction differential.
This interpretation is significant for the design of language tasks. It is evident that, in tasks like
these, if the teacher beforehand ensures that the learners’ content knowledge is at a higher level,
the interaction differential may increase, and so much more optimally stretch the potential or latent
language resources of the learners. As the authors remark (Larsen-Freeman & Cameron, 2008:
212), activities in the language class would enhance the potential of learning and growth of learners’
language if they can be designed ‘to challenge learners to exploit the meaning potential of their
developing systems in new ways’.
If this sounds familiar to teachers, no one needs to be surprised; not all of an approach, even one
that is undeniably new, can ever be entirely novel. There are a number of continuities with existing
232 Weideman

insights into language teaching and its design.

To take another example that ties in with the above: a complex systems theory enables new
interpretations to be given of past observations on language fossilisation. A learner’s stage of acquisi-
tion of an additional language has long been seen as an interlanguage (Selinker,1972), a kind of
waystage on the path to the desirable end of (almost) native-like competence in the target language.
Rather than taking a cross-sectional, and therefore static view as this, a complex systems approach
can plot the path of individual growth and variation across a timescale (Larsen-Freeman & Cameron,
2008: 245). Indeed, since ‘every organism is changing and determining what is important in its world
— creating and remaking the world in which it lives’ (2008: 143), the explanations for fossilisation
must yield, in a complex systems perspective, to an acknowledgement of the ‘boundless’ potential to
grow one’s language resources, and not stop at the powerful attractor that is ‘the neural commitment
of the first language, and the ensuing entrenchment, [that] may lead to a deep valley or well’ (2008:
142), which normally constitutes a trough in the trajectory of how an additional language develops.
Our final example derives from an answer given in a complexity perspective to an age-old puzzle
for teachers. How is it, one often hears language teachers ask, that learners sometimes ‘unlearn’ or
simply fail to learn elements of the target language that they should, by linear expectations, already
have possessed or acquired? In a complexity perspective, teachers and language course designers
can now find an explanation not only for individual variation, but also of apparent lapses in language
learning. In two telling examples, Larsen-Freeman and Cameron (2008: 135f.) show how individual
growth may vary in terms of fluency, vocabulary complexity and grammatical complexity – three
critically important language resources – for a number of language learners, and how, in a single
learner, there may demonstrably be both growth and lapses. Learners are seen as organisms that
are free to explore new behaviours (2008: 148), and since language growth does not follow a linear
path, a complexity theory explanation can readily provide an interpretation for a phenomenon that
many teachers will attest to.
In a complex systems perspective, the many interacting subsystems of language, and the
abilities of learners in terms of components of this complex whole – such as discourse practices
and structure, grammatical patterns, vocabulary resources, as well as the various other interacting
dimensions, like interventions and language demands in specific contexts – provide an explana-
tion for non-linear growth in learners’ language, since any variation in one or more parts of these
interacting systems may change the state of development of a language learner.
From a design angle, this means that a complex systems approach would make us more
sensitive, as applied linguists, to the varying demands and levels of learning that can sometimes be
found in a single classroom. As these authors put it (2008: 226):
… language resources of individuals exist only as latent potential to engage in appropriate
patterns of interaction until realized in specific discourse environments … The challenge is
for interaction, tasks and tests to be designed, planned, and managed so as to push and
stretch an individual’s language resources to the edge of their current potential.

Some questions
However familiar the above may sound to practising teachers, there is no doubt that complexity
theory makes a new contribution to our understanding of how to design language tasks. The
main reason for its ability to do so probably derives from its alternative perspective on language.
One would be justified in feeling, therefore, that a complex systems approach does not in the first
instance give us an alternative design for our teaching, but rather offers a novel view of language.
It is the effects of taking this perspective seriously, I would predict, that will have an impact on
designing language teaching in the future.
By taking a transdisciplinary approach, complex systems theory reaches across the boundaries
of cognitive psychology and sociolinguistics. By borrowing methods and concepts from studies as
diverse as those of finger movements (2008: 208), or by reinforcing and giving fresh interpretations
to some of the more conventional current approaches such as conversation analysis, discourse
studies, ethnographic description (2008: 242) and action research (2008: 244), the complex
systems view described by Larsen-Freeman and Cameron also illustrates its link both with other
Southern African Linguistics and Applied Language Studies 2009, 27(2): 229–233 233

disciplines and with what has gone before in linguistic and applied linguistic thinking.
Does it take us beyond postmodernism into a truly new, seventh generation of applied linguistics?
I am not altogether sure, but at the same time I cannot help feeling excited about the prospects of
advancing beyond postmodernism. Will it be just another -ism that we shall add to the many -isms
that have preceded it? Probably (I already feel tempted to use the term ‘emergentism’ in relation to
its views on language), but, at the same time, I must express admiration for its serious confrontation
with and critique of reductionism. For the moment, complexity theory has the benefit of the doubt,
and I salute the gracious acknowledgment of the current authors that complexity theory needs to
be complemented by other theories (2008: 14). Let us hope that the disciples, who will now surely
emerge, maintain that level of humility; the history of applied linguistics is unequivocal on one point:
new styles of accomplishing our designs eventually yield to others.

References
De Bot K, Lowie W & Verspoor M. 2007. A dynamic systems theory approach to second language
acquisition. Bilingualism: Language and Cognition 10(1): 7–21.
Cameron L & Larsen-Freeman D. 2007. Preview article: Complex systems and applied linguistics.
International Journal of Applied Linguistics 17(2): 226–240.
Larsen-Freeman D & L Cameron. 2008. Complex systems and applied linguistics. Oxford: Oxford
University Press.
Kramsch C. 2008. Ecological perspectives on foreign language education. Language Teaching
41(3): 389–408.
McNamara T. 2008. Mapping the scope of theory in TESOL. Contribution to symposium: Theory in
TESOL. TESOL Quarterly 42(2): 302–305.
Pennycook A. 1999. Introduction: critical approaches to TESOL. TESOL Quarterly, 33(3):
329–347.
Pennycook A. 2000. The social politics and the cultural politics of language classrooms. In Hall
JK & Eggington WG (eds) The sociopolitics of English language teaching. Clevedon: Multilingual
Matters, pp 89–103.
Pennycook A. 2004. Critical applied linguistics. In Davies A & Elder C (eds) The handbook of
applied linguistics. Oxford: Blackwell, pp 784–807.
Selinker L. 1972. Interlanguage. International Review of Applied Linguistics 10: 209–231.
Weideman A. 1987. Applied linguistics as a discipline of design: a foundational study. Unpublished
PhD thesis, University of the Free State.
Weideman A. 1999. Five generations of applied linguistics: some framework issues. Acta
Academica 31(1): 77–98.
Weideman A. 2003. Towards accountability: a point of orientation for post-modern applied
linguistics in the third millennium. Literator 24(1): 1–20.
Weideman A. 2007a. The redefinition of applied linguistics: Modernist and postmodernist views.
Southern African Linguistics and Applied Language Studies 25(4): 589–605.
Weideman A. 2007b. Towards a responsible agenda for applied linguistics: confessions of a
philosopher. Per linguam 23(2): 29–53.
Weideman A. 2009a. Constitutive and regulative conditions for the assessment of academic
literacy. Paper prepared for a symposium at AILA 2008 (Essen) on Assessing the academic
literacy of additional language students, 26 August. Submitted to a special edition of Southern
African Linguistics and Applied Language Studies.
Weideman A. 2009b. Uncharted territory: emerging paradigms and the foundations of applied
linguistics. Submitted to Per linguam.

Albert Weideman
Department of English, University of the Free State
View publication stats