Scientists have discovered that individuals who are particularly good at learning patterns and sequences tend to struggle with tasks requiring active thinking and decision-making. Their new research, published in npj Science of Learning, found a negative correlation between statistical learning and executive functions, suggesting that as one strengthens, the other may diminish slightly. This finding provides valuable insights into the competitive interactions that underpin our cognitive skills.
The motivation behind the study originated from a desire to deepen our understanding of how different cognitive systems within the brain interact, and potentially interfere with one another. Central to this investigation are two fundamental cognitive processes: implicit statistical learning and executive functions.
Implicit statistical learning is a crucial cognitive skill that allows individuals to unconsciously detect patterns and regularities in the environment, underpinning abilities in areas ranging from language acquisition to social interactions. On the other hand, executive functions are high-level cognitive processes that are essential for planning, decision-making, error correction, and adapting to new and complex situations, primarily managed by the prefrontal cortex.
This research was driven by the hypothesis of a competitive interaction between these systems, known as the “competition hypothesis,” which posits that reliance on one cognitive system could diminish the efficacy or engagement of the other. Previous studies provided preliminary evidence suggesting such interactions, but they were limited by small sample sizes and narrow assessments of cognitive abilities. The researchers aimed to build on this foundation to provide clearer insights into how these cognitive processes coexist or conflict within the brain.
“Our brain is a complex ecosystem. Different neurocognitive processes are constantly interacting with each other. This interaction can be cooperative, but what is very exciting and interesting is that these interactions can also be competitive,” said study author Dezső Németh of the Lyon Neuroscience Research Center at INSERM in France.
“So there is competition between different neurocognitive processes in the brain. This is something I have been researching for years. In this paper, we have shown that skill learning and statistical learning underlying predictive processes are negatively correlated with prefrontal lobe-related functions such as executive functions or control functions.”
The researchers conducted two experiments. Study 1 involved 186 young adults from France who underwent a two-day session where they first completed the Alternating Serial Reaction Time (ASRT) task to measure statistical learning. In this task, participants responded to visual stimuli (arrows pointing in one of four directions) by pressing corresponding buttons on a response box. Unbeknownst to participants, these stimuli followed a structured sequence interspersed with random elements, allowing researchers to measure how quickly and accurately individuals could learn and predict these patterns without explicit instructions.
The following day, a series of neuropsychological tests assessed various executive functions, including cognitive flexibility, inhibition, and working memory. Criteria for participation included being right-handed, under 35 years old, and having minimal musical training, factors known to influence cognitive processing.
Similarly, Study 2 replicated the structure of Study 1 but included 157 university students from Hungary, with slight variations to adapt to local resources and contexts. The ASRT task in this study used images of a dog’s head and keyboard responses, and it was self-paced, allowing researchers to test the consistency of the learning measure under different procedural conditions. Like in Study 1, the second session involved testing executive functions through a comparable set of tasks, adapted slightly for local execution.
Across both studies, a consistent negative correlation was observed between statistical learning and most measures of executive functioning. This suggests that individuals who excel in tasks requiring high levels of executive control, such as complex problem-solving and decision-making, may find it more challenging to engage in or benefit from implicit learning processes that rely on the subconscious detection of patterns and regularities in the environment.
“It’s very surprising to see this competition in the background of the skills learning,” Németh said.
The researchers employed factor analysis techniques to delve deeper into the data, revealing that certain aspects of executive functioning — specifically tasks that measured verbal fluency and complex working memory — were most strongly associated with these negative correlations. The researchers posited that this competition might arise because these executive tasks require active control and manipulation of information, processes that could interfere with the passive, automatic pattern recognition that characterizes implicit statistical learning.
The findings challenge the traditional view of cognitive abilities as isolated skills, instead highlighting the interactive and potentially competitive nature of different cognitive systems within the brain.
“Humans have multiple learning and memory processes and systems,” Németh told PsyPost. “So, there is no such thing as ‘the learning’ and ‘the memory’ system. Instead, there are learnings (i.e. multiple learning processes) and memory systems (multiple memory systems). When I want to learn a brand new thing, a new pattern, or a brand new sequence from an environment that I’ve never seen before, I can do it better if the prefrontal lobe functions (executive functions) in my brain are less efficient.”
“In other words, if you want to learn a new skill, like playing a new musical instrument, it’s very good if the functions associated with the prefrontal networks are weaker. It’s quite counter-intuitive. In many school performances, we see the opposite: If you need to understand a history or biology lesson, it is good and optimal if the prefrontal function is strong.”
However, the effect sizes were modest, indicating that while the relationships are statistically significant, they might not be strong. This suggests that other factors not measured in this study may also play significant roles in cognitive performance. Nevertheless, the findings “are very important for fundamental research,” Németh said. “They tell us a lot about how our brains work. The question is whether these results can be translated into practice.”
“These results are among the first in this field,” Németh added. “However, it is important to recognize that executive functions (prefrontal functions) and statistical learning (predictive processes) are not monolithic constructs; rather, they encompass a multitude of executive functions and various facets of statistical learning.”
“The critical inquiry pertains to which specific executive functions and elements of statistical learning exhibit positive or negative correlations with one another. When do they compete, and when do they cooperate? This question extends to the cerebral level as well. My objective is to unravel the brain mechanisms underlying these interactions.”
The study, “Evidence for a competitive relationship between executive functions and statistical learning,” was authored by Felipe Pedraza, Bence C. Farkas, Teodóra Vékony, Frederic Haesebaert, Romane Phelipon, Imola Mihalecz, Karolina Janacsek, Royce Anders, Barbara Tillmann, Gaën Plancher, and Dezső Németh.