Shedding Empirical Light on Complex Dynamic System Theory

We recently published Profiling Learner Language as a Dynamic System edited by ZhaoHong Han. In this post the editor explains why the book is important.

Profiling Learner Language as a Dynamic System was born out of an intense interest in contributing to the empirical basis in SLA of the new theoretical paradigm now known as Complex Dynamic System Theory (CDST) (de Bot, 2017; Larsen-Freeman, 2017; Lourdes & Han, 2017). Much of the work so far on CDST has remained rhetorical, and while a concerted effort has been made to push for empirical understandings, methodological insights are as yet incipient, though broad pointers are on the horizon. For example, the study needs to be longitudinal, and should focus on individual learners.

Many of the extant empirical studies have, however, tended to narrowly focus on one or a small number of linguistic elements, taking, a priori, each as part of a (sub)system, producing findings that are limited in scope and do not convincingly demonstrate, in one breath, the ‘complex,’ ‘dynamic,’ and ‘systemic’ nature of learner language.

This book seeks to help fill some of these gaps, by subjecting individuals’ systems to multiple lenses. Recognizing that revealing these properties necessitates a much larger undertaking than an individual study, the book has its five main chapters each target a particular aspect of interlanguage, traversing the domains of morpho-syntax, semantics, pragmatics, and discourse. The uniqueness of this approach lies also in employing the same longitudinal corpus involving two dyads interacting over a shared course requirement. The data analyses tracked both within-dyad and between-dyad similarities and differences, yielding both general patterns and idiosyncrasies. Together, the five sets of data analyses shed light on, and even go beyond, core claims of CDST.

For more information about this book please see our website.


If you found this interesting, you might also like Motivational Dynamics in Language Learning edited by Zoltán Dörnyei, Peter D. MacIntyre and Alastair Henry.

Discontinuity in Second Language Acquisition

Next month we are publishing Discontinuity in Second Language Acquisition by Stefano Rastelli. Here, Stefano introduces some of the main themes of his book and tells us more about discontinuity in language learning. 

9781783092468In this book I criticise the idea of piecemeal language acquisition. I propose instead that Second Language Acquisition (SLA) in adulthood abides by the law of dispersion/conservation of energy in nature. As such, it is a quantized and discontinuous process, made of leaps and instantaneous jumps from one level to another.

According to the discontinuity hypothesis, a learner’s competence is made of a superposition of coexisting levels of representation/processing. These levels are Statistical Learning and Grammatical Learning.

Statistical Learning is learning by chunks (unanalysed wholes). It is a bottom-up, frequency-driven process, which is based on a learner’s sensitivity to transition probabilities between adjacent words.

Grammatical Learning is learning by abstract categories (such as gender, number, aspect, tense etc) or labels that learners attach to frequently encountered concatenations of words.

Statistical Learning in adulthood paves the way for Grammatical Learning. Morphosyntactic items in adulthood are learned twice, first statistically and then grammatically. A learner’s competence is made of statistical representations coexisting with their grammatical counterparts. A steady-state condition is therefore achieved by learners when they become capable of using either representational/processing routes depending on the resources available.

Evidence of discontinuity comes from both the analysis of the changes in brain responses (Event-Related Potentials) occurring with learners’ increasing proficiency in longitudinal studies and from statistical modelling of learner corpora data.

The idea that SLA is quantized and discontinuous is hard to accommodate with both existing theories of language acquisition and with the common sense. Instead, the idea that SLA is a slow, gradual process conciliates common sense and scientific description because it fits people’s perception of the world around. Everything that grows (takes and releases energy) seems to do it gradually. Graduality always offers an all-internal explanation in terms of changes, history and evolution. On the contrary, the idea that changes take place suddenly has always been seen with scepticism and casts a shade of anti-Darwinism over those who claim it.

Discontinuity entails a deep criticism of the notion of ‘interlanguage’. Stages of interlanguage can be identified only if SLA is seen as a gradual phenomenon in which target-like forms are substituted the corresponding incorrect ones. Interlanguage ‘stadiation’ can be established only if one counts a learner’s errors, which in turn are seen as being mere transition states along the way to full competence.

While the idea that SLA is gradual and occurs by stages inherently commits the comparative fallacy, the discontinuity hypothesis is a more ecological view on the L2 developmental path in adulthood. According to the discontinuity hypothesis, shallow processing, uninflected, lexically computed wholes and also incorrect formulas and constructions do not represent a lower stage in the developmental path. They are not substituted by the correct counterparts once a learner’s competence advances. They represent a permanent level of a learner’s competence throughout.

cover Gabrys BarkerFor more information about the book please see our website or contact Stefano Rastello at If you found this interesting you might also be interested in Morphosyntactic Issues in Second Language Acquisition edited by Danuta Gabryś-Barker.