Over the last years, some reputed researchers I greatly admire have developed a new research architecture, that of non-linear dynamic systems theory applied to the analysis of how we learn a foreign language, a complex system by definition. In tune with that new approach Hiver & Al-Hoorie (2015) introduce their “dynamic ensemble”, a template of methodological considerations for doing or testing empirical second language development (SLD) within a complexity/dynamic systems theory (CDST) framework.
They start underlying the importance of transferring complex research designs to SLD using (Larsen-Freeman & Cameron, 2008) as springboard; then they move into the analysis of complexity tools trying to clarify some of the key issues in CDST (complexity/dynamic systems theory). They attempt to settle a basic model for empirical CDST research, an extremely complex endeavour given the very nature of the system to categorize. They use Spoelman & Verspoor (2010) as the exemplary study to illustrate the proposal; the study analizes “the development of accuracy rates and complexity measures in the learner language of a Dutch adult learner of Finnish”.
There are a number of considerations to take into account with such a model, mainly regarding the very nature of the SLD research objectives. When providing a controlled environment for the insight we allow for a high control of the sample at the expense of limiting the external validity of the outcomes. Hiver & Al-Hoorie state nine considerations to guide empirical research in practice and conclude introducing some research methods in tune with CDST basic operating principles.
The paper definitely provides food for thought regarding our efforts to attain reliable outcomes in SLD providing an insight into basic guidelines for complex dynamic systems analysis. The complexity meta-theory theoretical construct proposed by Larsen-Freeman (2015) combines a number of research principles that are “capable of informing all theories” (Morin, 1992, p.271), inviting us to use an integrated procedures system to inquiry about any issue in SLD. However, as the authors state:
The real and more exciting contribution the complexity perspective has made is not purely in the realm of methods of instrumentation and analysis, but instead in ontological and epistemological considerations of how we think about the world, considerations that are linked with the issue of how we engage in scientific inquiry (Ellis & Larsen–Freeman, 2006; Ortega, 2013) (Hiver & Al-Hoorie, 2016, p.743)
My concern with the open enthusiasm in the field is that of a real implementation of the procedure; the invitation to reflect on the effectiveness of our research methods and the attempt to blueprint an overarching protocol are great news; however, if we try to extrapolate the basics of the Dynamic System Theory (De Bot, Lowie, & Verspoor, 2007) we may find some complex issues to overcome; there we have: “sensitive dependence on initial conditions, complete interconnectedness of subsystems, the emergence of attractor states in development over time and variation both in and among individuals” (p. 1).
Taking into account as many factors as possible in SLA research is a staple diet of every attempt to conduct a serious enquiry process. I can’t imagine a researcher designing a research procedure on let’s say self-esteem, and not taking into account motivation, or anxiety or… initial conditions, on the other hand, the triangulation of methods and techniques in SLA research is a must since Seliger & Shohamy (1989). Of an extraordinary importance is the Hiver & Al-Hoorie’s statement:
“constellations (e.g., goals, interest) and abstract phenomena (e.g., L2 proficiency, L2 motivation) differ from systems because they do not produce an outcome by themselves, and must first be located within an agent who experiences and acts on them” (p. 745)
We then get the initial conditions of the agents-systems, take into account during the whole process the different factors that may individually, socially, cognitively and affectively influence her and her learning process and we have it. I guess in theoretical terms, the mathematical equation may work, what I find a bit steep is the development of studies with such a high level of control and the necessary N to allow generalizations for the learning population. Take for instance Spoelman & Verspoor (2010) study, they limit their analysis to three related measures of complexity (i.e., at the word, NP, and
sentence level). The research procedure in terms of internal structure and validity is reliable; unfortunately, there are many reasons to play havoc with its external validity.
Hiver & Al-Hoorie state:
“However, settling on precise levels of detail in data collection and analysis, as Spoelman and Verspoor (2010) do, for instance, in limiting their analysis to three related measures of complexity (i.e., at the word, NP, and sentence level), can contribute to understanding complex phenomena without knowing the entire hierarchy of nested levels and timescales” (p. 745)
However, de Bot, Lowie and Verspoor (2007) when settling the basics of the DST ask for a more interdependent approach:
Dynamic systems are characterized by what is called COMPLETE INTERCONNECTEDNESS; all variables are interrelated, and therefore changes in one variable will have an impact on all other variables that are part of the system. (p.8)
As Hiver & Al-Hoorie argue later (p.476) that contextual factors are determinant of how the complex systems behave and Spoelman and Verspoor (2010) are not analysing their full potential incidence on the final outcome. They are in fact also missing to describe the networked relationships among systems missing the interconnected factor determinant in DST.
Especially interesting in the paper is the section devoted to taking those considerations into practice. Linear analysis of the cuasi-experimental nature we have been doing so far does not fit into DST paradigm of research; to understand a complex system as a language learner the authors propose different methods that might be used in a coordinated fashion. Qualitative comparative analysis, process-tracing, concept-mapping, social network methods and agent-based modelling are considered in tune with the requirements of DST.
In our endeavour as ever-learning researchers we have faced revolutionary changes in approaches, methods and techniques in the realm of SLA research, we have been longing for a validating overarching system for years, it seems we have here the first moves and a possible way of overcoming the complications of the project would be the creation of an interinstitutional, transnational network where data might be accessed for research purposes. Come prepared! there is a long way to go.
Larsen–Freeman, D., & Cameron, L. (2008a). Complex systems and applied linguistics. 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. http://doi.org/10.1017/S1366728906002732
Hiver, P., & Al-Hoorie, A. H. (2016). A Dynamic Ensemble for Second Language Research: Putting Complexity Theory Into Practice. Modern Language Journal. http://doi.org/10.1111/modl.12347
Larsen-Freeman, D. (2015). Complexity theory. In Theories in second language acquisition: An introduction. http://doi.org/10.1080/01425690701737481
Larsen-Freeman, D., & Cameron, L. (2008). Research methodology on language development from a complex systems perspective. The Modern Language Journal, 92, 200–213. http://doi.org/10.1111/j.1540-4781.2008.00714.x
Morin, E. (1992). From the concept of system to the paradigm of complexity. Journal of Social and Evolutionary Systems, 15(4), 371–385. http://doi.org/10.1016/1061-7361(92)90024-8
Seliger, H. W., & Shohamy, E. (1989). Second language research methods. The Oxford Applied Linguistics (Vol. 13).
Spoelman, M., & Verspoor, M. (2010). Dynamic patterns in development of accuracy and complexity: A longitudinal case study in the acquisition of finnish. Applied Linguistics. http://doi.org/10.1093/applin/amq001