Neuroscientists at the Massachusetts Institute of Technology (MIT) have gained valuable insight into how the brain reads computer code. They found that rather than relying on the regions of the brain involved in language processing, the ‘multiple demand network’ is activated. This network is also what is responsible for complex cognitive tasks like mathematical problems.
It is easy to see why one would think the areas involved in language processing would be the same ones for computer programming, as the two tasks are similar in various ways. For example, computer coding also requires learning new symbols and terms, which then have to be organized so that the machine knows what to do. Computer code must also be extremely clear, allowing programmers to comprehend it.
Despite this, things get even more complex as reading computer code with the multiple demand network does not require the same cognitive demands as mathematics.
Anna Ivanova is an MIT graduate student and the lead author of the study.
“Understanding computer code seems to be its own thing. It’s not the same as language, and it’s not the same as math and logic,” Ivanova says.
The research also involved senior author Evelina Fedorenko, an Associate Professor of Neuroscience and a member of the McGovern Institute for Brain Research, as well as researchers from MIT’s Computer Science and Artificial Intelligence Laboratory and Tufts University.
The paper was published in eLife this month.
Fedorenko’s work focuses on the functions that rely on the brain’s language network, which is located in the left hemisphere of the brain. Some of her previous work demonstrated that the language network appears to not be activated by music and math.
“Here, we were interested in exploring the relationship between language and computer programming, partially because computer programming is such a new invention that we know that there couldn't be any hardwired mechanisms that make us good programmers,” Ivanova says.
Activated Multiple Demand Network
The researchers found very little response to code in these language regions, but they did find the multiple demand network activated throughout the frontal and parietal lobes. This network is relied on when tasks require large amounts of information to be held at once, which leads to our ability to perform a range of mental tasks.
“It does pretty much anything that's cognitively challenging, that makes you think hard,” Ivanova says.
The researchers did not pinpoint specific regions that are used exclusively for programming, but they believe individuals with substantial coding experience could develop specialized brain activity.
“It's possible that if you take people who are professional programmers, who have spent 30 or 40 years coding in a particular language, you may start seeing some specialization, or some crystallization of parts of the multiple demand system,” Fedorenko says. “In people who are familiar with coding and can efficiently do these tasks, but have had relatively limited experience, it just doesn't seem like you see any specialization yet.”
A team of researchers at Johns Hopkins University also had similar findings, specifically with the multiple demand network being activated for coding problems. According to the researchers, these findings mean coding can’t be restricted as a math-based skill or language-based skill as it could rely on both language and multiple demand systems.
“There have been claims from both camps — it has to be together with math, it has to be together with language,” Ivanova says. “But it looks like computer science educators will have to develop their own approaches for teaching code most effectively.”
- The Black Box Problem in LLMs: Challenges and Emerging Solutions
- Alex Ratner, CEO & Co-Founder of Snorkel AI – Interview Series
- Circleboom Review: The Best AI-Powered Social Media Tool?
- Stable Video Diffusion: Latent Video Diffusion Models to Large Datasets
- Donny White, CEO & Co-Founder of Satisfi Labs – Interview Series