This year the Donders Institute is celebrating the 200th birthday of Franciscus Donders, founder of cognitive brain research. On the occasion of this anniversary, language psychologist Ardi Roelofs repeated the groundbreaking experiment with which Donders showed 150 years ago that thinking takes time. He used the original stimulus lists from the archives of the Utrecht University Museum.
“Finally I had a good reason to dive into the history of Franciscus Donders during working hours,” says Ardi Roelofs, who is affiliated with the Donders Institute as professor of Language and Attention at Radboud University. This Nijmegen research institute for brains, cognition and behavior is named after the doctor and physiologist who was born in Tilburg in 1818. Although Donders gained his worldwide fame with research into the eye and eye disorders, he also stood at the cradle of a field that was still unexplored at the time:cognitive brain science.
“At the time, it was thought that nerve conduction was faster than light. So not measurable. The same applied to the speed of thinking”, says Roelofs. "But Donders' friend Hermann von Helmholtz showed in 1850 that this was not true." The German physicist chopped off a frog's leg and embedded it in an electrical circuit. He stimulated the nerve pathway at different distances from the muscle and watched how quickly the muscle contracted. Based on his measurements, he estimated the nerve speed at thirty meters per second.
Also in the 19 e century it was unethical to cut off a leg for experimental purposes, so von Helmholtz had to adapt his method slightly to investigate how this worked in humans. He gave his subjects an electrical stimulus at various places on the leg and asked them to press a button as quickly as possible. Because the foot is further away than the thigh, you could also estimate the nerve speed here. Roelofs:“Von Helmholtz, however, found a lot of variation between his test subjects. According to him, that difficult brain was the cause. Donders found it interesting that the signal ran through the brain. He repeated the experiment with his daughter Marie, among others, and realized that the brain ensured that the measurement was indirect.” He then devised another experiment – with which his name ended up on the outside facade of the Nijmegen research institute in 2002.
Thinking takes time
In this experiment, the subject performs three nearly identical tasks. In task A, someone keeps saying 'ki' and the subject has to repeat it as quickly as possible. In task B, the subject also repeats what is said, but now it can be 'ka', 'ke', 'ki', 'ko' or 'ku'. The subject's brain must therefore distinguish the sounds from each other and quickly select the corresponding articulation program. In task C, the subject hears the same sounds as in B, but only has to repeat them when he hears 'ki'. His brain can therefore already select the articulation program, but the test subject has to listen carefully to distinguish the sounds from each other.
Featured by the editors
MedicineWhat are the microplastics doing in my sunscreen?!
AstronomySun, sea and science
BiologyExpedition to melting land
Donders registered the different reaction times and subtracted them from each other to determine how much time the extra 'thinking step' took. By subtracting the response time of task A from that of task C, he determined that distinguishing between the sounds takes about 36 milliseconds. Selecting the right articulation program takes approximately 47 milliseconds, as was apparent from the difference in reaction times between task B and C. Roelofs:“He showed, for example, that mental processes take time. A startling finding, which he immediately published in various languages.”
Register thinking times
Donders' reaction time experiments were very innovative. Of course, he didn't have fast computers yet, so how could he measure so accurately? “For this he used a phonautograph, a patented device by the Frenchman Leon Scott de Martinville,” says Roelofs. “It is actually a kind of half gramophone, with which you can only record, not play. When you speak into a horn, a membrane vibrates. That membrane uses a small needle to create a wave on sooty paper, which rotates on a cylinder. Edison later thought that if you replace the sooted paper with wax, you could also reconstruct the sound.”
Donders had the horn's mouth widened a bit, so that he could properly register the sounds of two people next to each other. He also put a tuning fork on the device:the even sound wave of the pure tone served as an indication of time. Et voilà, the noematachograph ('thought speed writer') was born.
Remarkable discovery
The same noematachograph is now in the possession of the Utrecht University Museum – and was on display in the Anatomical Museum of the Radboudumc on the occasion of this anniversary. The Utrecht University Museum has more valuable pieces from Donders' past, as Roelofs discovered during his search. “The museum also has all of Donders' original, handwritten stimulus lists,” he says. The lists contain the syllables that were read for the various tasks, including the results. “For example, I saw that on Monday, August 21, 1865, at 7 o'clock in the evening, Donders completed nine lists with two of his students, Hamer and Stark. They kept changing roles in the experiment, but Donders only published his own reaction times. That was customary at the time.”
Historical parallel
Roelofs decided to analyze the reaction times of the students and made a remarkable discovery. While the reaction time for task B was clearly longer for Donders than for task C, these differed little or not at all in his students. Roelofs:“Donders already mentioned in his article that people had difficulty with task C, that they made mistakes because they reacted too quickly.” But Donders' contemporary Wilhelm Wundt saw a different problem. This founder of experimental psychology noted that although you no longer have to select an articulation program there, you still have to choose whether it should be pronounced or not. It is a 'go/don't go' task, in which the choice (in addition to distinguishing) also takes time.
Because this issue had never been resolved, Roelofs started working on it himself last spring. With his daughter Sterre – 'to create a beautiful historical parallel'. He typed out Donders' lists, had his daughter read them aloud and measured his own reaction times. “The averages over the nine lists show the same pattern as with Donders, although I was a bit slower than him overall,” he says. “Moreover, the difference between task B and C is smaller with me than with him. I also had more trouble with the 'go/don't go' task.”
Extra thinking step
He then tested the point of contention using the computational model for language production, which he developed over the past 25 years with colleagues at the neighboring Max Planck Institute. This model incorporates all the stages necessary for conducting the experiment:hearing, discerning, choosing, articulating. Roelofs:“I get the classic Donders pattern from that. But if I build in a 'go/don't go' decision and let that take time, I get exactly the pattern of Donders' students. That is an additional support for Wundt's idea that that extra thinking step takes time.” So it seems that Donders himself was extremely good at his own job.
For Peter Hagoort, director of the Donders Institute, it was clear from the start that the research institute should be named after the founder of cognitive brain research. On the institute's site, he writes:“Donders was a well-known name because of his subtraction method, on which contemporary imaging techniques are based. I wanted a name that would be associated with neuroscience all over the world .” With the replication study by Donders researcher Roelofs, the circle is now nicely round.