Thank you, Madam Chair.
I would like to thank the committee for inviting me to this meeting and giving me the opportunity to speak about hybrid proceedings.
Unfortunately, I cannot be with you, but believe me, I am in a rather good position to6 understand the additional workload that my virtual presence creates, especially for the interpreters. It is that additional workload that I would like to talk to you about today.
My name is Kilian Seeber. I am an associate professor at the Faculty of Translation and Interpreting at the University of Geneva. I have spent the last 15 years, give or take, trying to better understand the construct of cognitive load, especially cognitive load as it relates to simultaneous conference interpreting.
We do know that the human brain has an extraordinary capacity when it comes to the storage of long-term information, to the tune of roughly seven billion gigabytes, but when it comes to short-term storage, and particularly when it comes attention or cognitive control, it is unfortunately rather limited. Our working memory, which we believe to be the system that is responsible for short-term storage and manipulation of information, is finite.
This is where three important notions come in that I would like to cover before I tell you about the empirical studies that we have just concluded: cognitive capacity, being the processing resources that can be deployed by the system; cognitive load, being the processing demands that are imposed on the system; and cognitive effort, being the processing capacity that's actually allocated to a task. When the imposed load exceeds capacity, or when the invested effort doesn't meet task demands, the process will slow down and eventually break down. Interestingly, in cognitive terms, simultaneous interpretive training, rather than focusing on language training, aims at acquiring the skills required to strategically allocate resources to accommodate this increased task load.
As I was saying before, we very recently carried out some studies at the University of Geneva where we looked into the relationship between deteriorated sound and cognitive load in simultaneous interpreters.
In the first study, we observed interpreters and their psychophysiological response in the field to what I would call “frequently occurring salient triggers”, or events that you'll run into time and again when you have online meetings or hybrid meetings such as this one. Interpreters show significant psychophysiological responses. Their body responds to instances of bad sound, including echos, distortions, pops, clicks or background noises.
In the second study, we measured the interpreters’ psychophysiological response to deteriorated sound. We artificially deteriorated the sound by reducing the frequency response. We found that interpreters showed significant cognitive and emotional response during low-quality sound. There was a significant increase in the subjective load they perceived during low-quality sound. The increased load sets in as early as after the first 10 minutes on task. The experiment was designed in a way where they would be on task for 10 minutes, off for 15 and on for 10, in an iteration of four times. Interpretation quality, importantly, decreased significantly with low-quality sound. Of the three parameters analyzed, it was not style, not presentation, but content that suffered significantly.
In the third study, we then measured cognitive load as it changes with artificially deteriorated sound with, again, reduced frequency response. We found that cognitive load as measured with pupil dilation—again, psychophysiologically—didn't significantly change when interpreters just had to listen to that sound, but the cognitive load did significantly increase when they interpreted. Taken together—