When a Year 7 student physically reacted to a prompt of another student by anxiously drumming the desk with his ruler, exclaiming "uuuuhh", the initial thought of the observing researcher, Laura, was: "this is an interesting account". This started a reflective journey of first applying robust research methodologies to the episode the authors initially saw as a linear, logical system, then slowly developing understanding of the complexity of the interaction, as well as their position as complex observers of a complex entity. Thus, the journey evolved into a second-order complexity narrative (the complex entity being the first-order, the complex observer being the second order; Tsoukas & Hatch, 2001) that they tell here. In this article, the authors explain how exploring a complexity perspective helped them to make sense of a collaborative problem solving (CPS) situation that creates so much non-deterministic variation that it becomes irreducible to linear reasoning only. According to Bruner (1986), there are two modes of thinking: one that operates on logic to justify arguments and seek truths, and another that builds up a narrative and can inform about likely related accounts. The authors used both modes to analyse data that included the Year 7 student's reaction. They found out that seeing educational theories, as well as all the human and non-human agents involved as adding to the complexity of CPS research (Fenwick & Edwards, 2010) can provide tools to discuss intuitive interpretations and likely associations of what is present in the data.