This paper responds to Exploring the dynamism between propositional complexity and error rate: a case study by Jordan Van Horn (this issue). In her study, Jordan Van Horn analyzes asynchronous email exchanges between a native and nonnative speaker of English utilizing three methods of analysis: error analysis, complexity analysis, and qualitative analysis. In triangulating the data, the author states that this mixed-methods approach can offer a better solution to limitations that arise as a result of employing only one method to analyze performance data. Indeed, interlanguage data often exhibits patterns of variability in complexity and accuracy in L2 learners (Spoelman & Verspoor, 2010). In order to better account for this variability and enhance our understanding of second language acquisition (SLA), the author incorporated Complex Dynamic Systems Theory (CDST) to guide her analysis. CDST facilitates the unification of the three methods of analysis since it posits L2 language as exhibiting the characteristics of complex systems and L2 language development as one that involves natural flux and variation (Verspoor, Lowie & van Dijk, 2008). A major contribution of this study was Van Horn’s finding of connected growers between complexity and accuracy. The CDST construct “connected growers” is defined by the author as linguistic subsystems that develop symbiotically as they interact. In other words, linguistic subsystems such as complexity and accuracy were found to have a positive correlation. This finding has major implications not only for our understanding of complex dynamic systems, but also, of SLA in particular, as during L2 linguistic performance, complexity and accuracy often compete with each other for cognitive resources and positive correlations between these two areas have been less common in the literature (Skehan, 2009).