Is This (Becoming) a Theory of Second Language Acquisition?: A Commentary on “Complexity and Difficulty in Second Language Acquisition: A Theoretical and Methodological Overview”

Language acquisition is a process whereby an individual develops a language in interaction with the environment (Tomasello, 2005). Language acquisition research studies this process with the aim of describing, explaining, and predicting it.

In view of the above, a question for this commentary is how the overview of and guidelines for complexity and difficulty presented by Bulté, Housen, and Pallotti (henceforth BHP) contribute to research in language acquisition, specifically in second language acquisition (SLA). I argue that the authors’ greater focus on difficulty leads to an increased contribution of this line of work to the SLA research agenda, but that the concept itself is not helpful.

Moreover, somewhat paradoxically, BHP's ambition to be conceptually clear and to distinguish between constructs necessarily leads to a broadening of their scope, which in turn leads to the need for other clarifications. In fact, what starts out as an overview and guidelines ends up being “a research program” at the end of the paper with the aim of answering “some fundamental questions of SLA research,” a program that could be the beginnings of a new theory of SLA.

The aims of the paper are “primarily to bring conceptual and terminological clarity firstly by proposing a key conceptual and terminological distinction between ‘complexity’ and ‘difficulty’ and, secondly, by discussing how these two constructs relate to ‘development’ and ‘proficiency’”. These laudable aims align well with the authors’ previous work, where, albeit not jointly, they have repeatedly discussed construct definitions, operationalization, and measurement of complexity (e.g., Bulté & Housen, 2012; Pallotti, 2015). The authors take these issues seriously, and it is one of the merits of their work.

Within SLA, language complexity has in the past been conceived of mainly as a dependent variable that, through the development of various metrics, has been used to describe progress in second language (L2) proficiency or performance in, for example, studies on task or modality effects and learning contexts.

BHP now argue that complexity can also contribute to a “property theory” of SLA, a term sometimes used by generativist scholars to characterize the abstract knowledge or the competence of the L2 learner. However, this becomes confusing since BHP are using this term in a very different way from, for example, Gregg (2003), who is quoted in this context. BHP are only concerned with the complexity of “forms” or “units” and “with describing the complexity and difficulty realized in actual texts,” that is, something closer to Saussurian parole than to langue. At this stage and in the absence of reference to any (linguistic) theory, it is difficult to see how this approach can account for “non-trivial aspects of the underlying interlanguage systems.”

The authors’ operationalization of difficulty draws on mainstream SLA literature. The resulting (nonexhaustive) list contains several well-known variables, relating either to the structures themselves (e.g., saliency, transparency, productivity) or to their use in discourse (e.g., frequency).

This raises three fundamental questions. First, if SLA research has already identified, defined, and tested more specific constructs and variables, what is the point of grouping them together under the descriptive umbrella of difficulty? Difficulty per se has no explanatory value. Isn't there a risk that such a wide concept will meet the same fate as complexity when it comes to conceptual ambiguity? This would in fact run counter to the central aims of the paper.

A second point is that all learner-related variables of difficulty are excluded from BHP's conceptualization. Possible age effects or individual differences all fall outside of their scope. It is also hard to see where transfer or crosslinguistic influence would fit in (surprisingly not discussed by BHP). The exclusion of other sources of difficulty than those related to forms and units is unfortunate since recent acquisition research has started to look at the interaction between learner-related and structure-related properties. This is the case, for example, in studies on bilingualism in early childhood. Ågren et al. (2014) investigated how three different structures in French (finiteness, object clitics, and subject–verb agreement) interact with age and frequency properties. They found that, whereas the development of finiteness is more related to age of onset, subject–verb agreement is more dependent on frequency effects. This type of research requires a focus on a single or a few well-understood linguistic structures, and despite what BHP seem to think, some L2 researchers are actually (still) interested in this.

Third, BHP's view of difficulty is “largely grounded on theories seeing language acquisition and use in terms of cognitive processes and skills,” such as skill acquisition theory (DeKeyser, 2020) and emergentism (O'Grady, 2022). It is unfortunate that the authors do not develop further how they view the relationship between these transition theories and their own conceptualizations of a property theory. For example, in some of the operationalization of complexity, BHP rely on syntactic structures, hierarchical dependencies, and relationships between nodes. The use of these notions suggests that BHP take the position that syntactic representations do exist. This seems to be different from an emergentist view without syntactically mediated mapping between form and meaning (O'Grady, 2022, pp. 17–19).

BHP end by reiterating the important point that complexity, difficulty, and development should not be confused and should be given clear definitions. They are optimistic that if this is done, it will allow researchers “to answer some fundamental questions about SLA” such as “how interlanguage systems develop over time” or “how language acquisition and use are affected by a number of internal and external factors.” These are indeed fundamental questions that a theory of SLA should be able to address, but the authors prefer to talk about a “research program,” at least at this stage.

In any case, with the promising focus on difficulty over complexity, the contribution of this work for SLA research is to move away from a situation where complexity was conceived of as a dependent variable to measure L2 progress. However, with the increased ambitions, the requirements will also increase, including the need for a more developed thinking about a property theory and its relation to a transition theory. I also think that if the “fundamental questions” are to be successfully addressed, the program needs to expand beyond the exclusive focus on structure-related complexity and difficulty as they appear in texts. In fact, the definition of language acquisition proposed at the beginning of this commentary boils down to five theoretical constructs that language acquisition research needs to consider: the individual (learner), development, language, interaction, and environment. Ultimately, we also need to understand how they interact with each other, no matter how daunting that task may seem.