Enjoyment and anxiety in second language communicationAn idiodynamic approach

Picture yourself trying to speak a second language and feeling your stomach flip as you search for a word. Then, in the very next breath, you hit a groove, and the whole thing starts to feel fun. That whiplash is the point. For a long time, research on second language learning treated emotion like a stable setting on a dial: some people are anxious, others enjoy it, and those traits explain outcomes. But emotions in the act of speaking aren't settings. They're sparks. They flare, fade, and sometimes cascade. That's the promise of a dynamic view: follow the flicker, not just the average. Complex dynamic systems theory gives us language for this. As Diane Larsen-Freeman and Lynne Cameron argued, emotions in communication are emergent; they organize themselves in real time with the task and the social context. Barbara Fredrickson's broaden-and-build idea adds a twist: positive emotions widen your field of view and help you build resources, while negative ones narrow and lock onto threats. The result isn't a single seesaw but two distinct emotions—enjoyment and anxiety—that can move together, drift apart, or trade places in an instant. Jean-Marc Dewaele and Peter MacIntyre showed at a broad level that people who report more enjoyment tend to report less anxiety, with a modest correlation around minus point three. But that's a snapshot. What happens while someone is actually speaking? Carmen Boudreau, Peter MacIntyre, and Jean-Marc Dewaele decided to watch emotion unfold second by second during live French speaking. Their sample was small on purpose: ten university students at Cape Breton University, all native English speakers using French as a second language. Most were in their second or third year, seven women and three men, and the mean age hovered around twenty-one. The point wasn't to generalize to everyone; it was to get a high-resolution look inside each person. They set up two tasks that push and pull in different ways. In one, the photo narrative, each student brought a picture they enjoyed and talked about it in French for three to five minutes. It's personal and free-form. In the other, an oral interview, they faced five prompted challenges—describe what you're wearing, count to one hundred by tens, explain Parliament's role—answering in French after hearing the question in English. The order was randomized so nobody always started with the same kind of pressure. Here's the clever part. After each task, the student watched the video of themselves and rated how they felt, second by second, using a software tool called Anion Variable Tester Version 2. The scale ran from minus five to plus five, where minus five meant very low enjoyment or anxiety and plus five meant very high. They did one emotion first, saw a graph of their ratings, talked with the researcher about the spikes and plateaus, and then repeated the whole process for the other emotion. Those conversations were recorded and later transcribed. So you don't just get a line that jitters—you get the story behind each jolt. To make sense of the patterns, the team computed, for each person and each task, the correlation between their enjoyment line and their anxiety line. That yielded twenty within-person correlations in all. They weren't trying to run population-level hypothesis tests; this is a descriptive lens. Still, as a practical yardstick, they used the so-called personality coefficient—about point three in magnitude—as a marker of a relationship that's not trivial. What emerged is a strong theme with room for nuance. Out of those twenty task-level correlations, eighteen were negative. Not a single one was strongly positive. Using that point three yardstick, twelve of the twenty hit substantial magnitude, and all twelve were negative. In plain language, most of the time, when anxiety rose, enjoyment fell. But the strength of that coupling was all over the map, which is where this gets interesting. Take Participant 1. During the oral interview, their enjoyment and anxiety were almost perfect opposites; the correlation was minus point eight seven. When anxiety went up, enjoyment dropped in lockstep. In the photo narrative, the pattern was still clearly negative, though less extreme, at minus point six five. This is the purest form of the seesaw—clean, strong, and consistent across two very different tasks. Now contrast that with Participant 5, who showed a pronounced unraveling over time. In the photo narrative, the correlation came in at minus point five six. In the oral interview, it was even stronger, at minus point seven two. You can hear the dynamic in their own words: it "got worse and worse and worse and I kept forgetting more and more. So my anxiety just kind of increased and I could feel that emotion and it just made things worse." That's a cascade. Once anxiety crosses a threshold, it starts to dominate, and the space for enjoyment collapses. But not everyone rides a single track. Participant 8 is the perfect counterexample. Overall, in the photo narrative, their enjoyment and anxiety were modestly negatively related, around minus point three three. Zoom in, though, and the relationship flips. In a twenty-second window near a minute and a half, the two moved in opposite directions very strongly—minus point six six. Later, about three minutes in, they rose and fell together, producing a positive correlation of point three eight. And earlier, they drifted almost independently, near zero. Same person, same task, different micro-episodes. That's what dynamic coupling looks like when you can actually see it. There were also cases where the lines barely talked to each other at all. Five of the twenty tasks produced near-zero correlations. Participant 2 on the oral interview sat around minus point zero nine, and Participant 6, also on the interview, around minus point zero three—essentially flat. In contrast, Participants 10 and 3 bounced around a lot emotionally, but their overall coupling stayed small: Participant 10's correlation on the interview hovered at point zero nine, and Participant 3's at minus point zero seven. The small p-values in the notes—point zero six and point twenty-five—aren't the headline here. They just remind us this is within-person description, not a search for population-level significance. The interviews fill in the human texture behind those numbers. Participant 10 described a kind of relief—once the rust shook off and the first mistakes were out, the task felt manageable. Participant 3 said that even with some anxiety humming in the background, speaking French was still enjoyable. These are small reframes, but they matter. They point to strategies or mindsets that can buffer the spiral we heard from Participant 5, keeping challenge inside the window where it fuels engagement instead of crowding it out. Step back, and a pattern comes into focus. Enjoyment and anxiety are not opposite ends of one stick. They're distinct emotions that can interact in multiple ways, moment by moment, depending on what the task is asking of you and how your language system is holding up right then. That sounds like theory-speak until you picture the concrete events driving those jolts—searching for a verb and coming up empty, then nailing a phrase you'd been practicing; realizing the topic just shifted to politics; noticing the camera; feeling heard by the interviewer. Complex dynamic systems theory predicts this kind of event sensitivity, and broaden-and-build gives us a reason why pockets of enjoyment can widen the field just enough to keep going, even when the stakes feel sharp. It's also a reminder for teaching. If there's a threshold where surging anxiety starts to suppress the felt presence of positive emotion, then the timing of support matters. It's not just "less anxiety, more enjoyment" in the abstract. It's when and where. A hint, a well-timed prompt, or a moment to breathe could keep the balance from tipping into a runaway cycle. There are caveats. This was a lab, not a café or a classroom. Some students said the camera itself made them tense, and being recorded can amplify self-consciousness. The rating tool required a lot of clicking; not everyone used it with the same frequency, which affects data density. The team even suggested that rating every two seconds might be kinder on the hands and still capture the gist. And the group was specific: native English speakers using classroom French. Whether the same patterns would pop in more immersive settings, or with speakers at different proficiency levels, remains to be seen. For context, this idiodynamic approach traces back to work by MacIntyre in 2012 and has since become a go-to for capturing second-by-second affect, but it's still a choice with trade-offs. You get exquisite within-person resolution. You give up claims about averages across broad populations. One more touchstone helps locate these findings. At the trait level, as Dewaele and MacIntyre reported a decade ago, enjoyment and anxiety are modestly negatively related—people who like using a language more tend to worry less. That association, around minus point three four, sets a baseline expectation. What Boudreau and colleagues show is that inside any one person, the real-time relationship often tilts negative too, and sometimes very strongly, but it can bend, reverse, or loosen, even within a single four-minute narrative. That's not a contradiction. It's a sign that averages blur what matters in the moment. Where does this go next? Two ideas feel especially powerful. First, adding physiological traces—heart rate or heart rate variability—could anchor those self-reported spikes in a parallel stream of arousal. Not to replace the story people tell about what they felt, but to triangulate it in the wild, during actual classroom debates or hallway chats. Second, widening the lens on context—different languages, proficiency bands, and more naturalistic tasks—would test how robust these micro-dynamics are when the walls of the lab fall away. But even without the gadgets, there's a simple takeaway to carry into the next conversation you have in another language or the next class you teach. Emotions in speaking are fast and plural. The space between "this is scary" and "this is fun" can be just a breath, and it moves. The trick is seeing that movement—and sometimes, nudging it—before the moment passes.