Finally, the time has come to talk about what I was actually doing in Krakow: my research project.

The first few weeks of my stay were spent on improving my original research proposal. The main idea remained the same, but the experimental design has undergone a number of changes. It was a little scary, watching as everything I had planned and prepared for so long was remolded by our endless meetings and discussions. But I was also so appreciative of the thorough feedback from everyone in the lab, and I think the study that shaped up was the best we could do given our time and resources. So, here’s it is:

The investigation is based on the adaptive control hypothesis (Green & Abutalebi, 2013), which basically proposes that different types of the bilingual experience have different effects on how our brains work. It is not only the matter of speaking multiple languages, but also the manner of doing so that affects a bilingual’s mental processes, or executive functions. There are 3 different contexts (and thus, types of language use) that are outlined by the hypothesis.

  1. Single-language. This is when the use of two languages is completely separated. One common example is a bilingual who speaks one language at home, and another one at work/school.
  2. Dual-language. In this context, the use of both languages is required within one setting and is dependent on the other speakers. For example, a bilingual has two colleagues at work, one of which only speaks English, and the other only Chinese. Thus, the bilingual speaker is forced to switch between languages depending on who she is addressing, sometimes all within the same conversation.
  3. Dense code-switching. Code-switching occurs only between bilingual speakers who are fluent in the same languages. In this context, the bilingual speaker uses words from both languages within one sentence. Since both languages are easily understood by the other speaker, it is a matter of convenience (ex: which word is easier to remember) that determines the switching.

In my study, I wanted to recreate these contexts in a controlled laboratory setting. It quickly became clear that the code-switching context will be nearly impossible to investigate within the scope of this study, considering that code-switching occurs only in some bilingual communities, and it would be difficult both to recruit participants and also to induce code-switching in the lab. Thus, we focused on the first two contexts: single- and dual- language.

We also focused on one aspect of cognition: inhibition. Supposedly, the dual-language context places greater demands on this process, as the bilingual speakers have to inhibit the language which is currently unneeded, depending on who they are addressing. We used two well-supported cognitive tasks to measure inhibition: Stroop and Stop-Signal. The participants completed these tasks in the EEG, so in addition to behavioral data (accuracy of pressing the keys, reaction time, etc.), we will also have a direct insight into their brain activity. Based on these measures, we hope to see that after the dual-language context the participants will have better inhibition because it was “trained” by that type of language use. The single-language context should also lead to better inhibition, to a smaller extent. The two contexts will be compared to the baseline performance.

The experiment was set up in this way: each participant came in for 3 sessions. During each session, they participated in language training for the first 2 hours, and then had a 2-hour EEG session. During the language training, the participant was asked to play a game with two other players. In order to create a more natural environment, the participant was led to believe that the other players were also participants in the study. In fact, they were trained volunteers who worked as our confederates.

The game was loosely based on the map task designed by Beatty-Martínez and Dussias (2017). In this task, we asked the participant to explain a series of slides with pictures on them to the two confederates. The confederates had slides with the same background, but their pictures were arranged in a different way. So, using the background as their reference, they had to work together to make sure that the confederates’ slides match the participant’s slides as closely as possible. They were encouraged to give as much detail as possible, ask for clarifications, and engage in conversation with each other—because, of course, this was the real goal of the game.

Sample participant’s slide

 

Sample confederate’s slide

As the game went on, the slides became more complicated, with more objects and backgrounds that are difficult to describe precisely. Here’s one example, which I’m a little proud of (should I submit it to an art exhibit, maybe?):

Making the slides was pretty fun at first, but then I started running out of ideas. Still, I made all of them myself—54 unique slides (108 with the confederates’ re-arranged copies). This provided enough materials for all 3 sessions, 6 hours of talking total. And I’m happy to say that the game seems to have worked! The participants were engaged, talking for the majority of the time, and really employing full use of their languages to remember specific vocab words and to give accurate, precise directions. Many would laugh at their mistakes and slip-ups, sigh with frustration at the most difficult slides, and smile proudly when they completed the game well. So, I hope, this time-consuming experiment was at least a bit fun both for the participants and for our volunteers. 🙂

In this study, we recruited young adults who were native speakers of Polish with high proficiency in English. We also had 16 volunteers, who were fluent in Polish, English, or both. During the game, the participant took turns in explaining the slides to the other two players. In the baseline condition, everyone spoke in their native Polish. In the single-language condition, the whole interaction took place in English (different from their day-to-day life, a separate environment with a separate language). Finally, in the dual-language condition, the participant spoke only Polish to one confederate, and only English to the other person. This was very demanding, as he/she had to continually switch back and forth.

After one of these sessions was over (either only Polish, only English, or dual), the participant was plugged into the EEG to measure any immediate training effect on the inhibition function. In this within-subjects design, all participants had to do all 3 sessions. The order of the sessions was counterbalanced across the participants, and so were the slides.

As you might imagine, this was a very involved project. We fully tested 17 participants, each coming in for 12 hours. That’s 204 hours of pure testing in just about a month. (We also started testing others, and the project will continue even past my stay). Of course, in addition to testing, there was recruitment and scheduling of participants, training of the volunteers, lots of paperwork, and set-up and clean-up for each session. The amount of time and effort that went into this is pretty incredible. I had to put in a lot of hours, working long days without weekends. So did Jonas, who was absolutely crucial in making this project happen, and who is heroically continuing testing even after my PIRE trip is over. And of course, we could not have done it without the volunteers, who gave their time even during the most stressful time of the semester, during the exam session.

It is impossible to say as of right now whether we got any results. Extensive data analysis will be needed before we can draw any conclusions. Certainly, there are plenty of things that could have gone wrong. Was the sample size large enough? Was the language training successful in recreating the different bilingual contexts? Were Stroop and Stop signal good tasks for capturing inhibition? But I am hopeful that some (statistically significant 😉 ) results will come from this research project. For now, all I can do is cross my fingers and hope for the best. I guess, in the end, science does contain a bit of superstition!