The experiment was simple. Thirty-six university students sat down in a quiet lab in Trondheim, Norway. A research assistant fitted each one with a soft cap holding 256 sensors pressed against the scalp. Words flashed on a screen, one at a time. The students wrote each word two ways. First by hand with a digital pen on a touchscreen. Then by typing the same word on a standard keyboard. The sensors recorded the brain’s electrical activity from start to finish.
The finding was hard to misread. When the students wrote by hand, large networks of brain regions lit up and talked to each other across visual, sensory, and motor areas in the exact frequency bands that earlier work has tied to memory and learning. When the same students typed the same words, those networks went quiet. The hands moved. The screen showed letters. The brain barely engaged.
The paper was published in Frontiers in Psychology on January 26, 2024. The lead author was Audrey van der Meer, a developmental neuroscientist who runs the Developmental Neuroscience Laboratory at the Norwegian University of Science and Technology (NTNU). She has been publishing on this question, in some form, since 2017. The pattern has held across studies of adults, university students, and 12-year-old children.
Who is Audrey van der Meer, and why has this taken so long to reach classrooms?
Van der Meer has spent her career at NTNU studying how human brains develop, with a focus on the link between motor action and perception. Her lab uses high-density EEG, the same family of equipment used in clinical neurology, to measure how thinking happens in real time. She works alongside her long-time research partner Ruud van der Weel, who shares authorship on most of the handwriting papers.
Her first formal handwriting-versus-typing EEG paper appeared in 2017. A larger 2020 study examined the same question with 12 university students and 12 twelve-year-old children, also at NTNU, published again in Frontiers in Psychology. The 2024 paper, with 36 university students and a richer sensor array, was the cleanest version of the experiment to date. The findings have not really changed across the three studies. The brain networks involved in handwriting are wider, more connected, and active in frequencies associated with learning. Typing does not produce the same pattern.
One reason the work has stayed mostly inside her field is that classroom policy moves slowly and education-technology companies have spent two decades arguing the opposite. Schools across Europe and the United States rolled out tablets and laptops on the assumption that screens would improve learning. That assumption has now been tested in the real world for long enough that several governments are reversing course.
What did the 2024 EEG study actually measure?
Here is what happened, in plain terms. Each of the 36 participants saw single words on a screen. Some words they wrote by hand using a stylus on a touch-sensitive surface, the closest digital approximation of writing on paper. Some words they typed on a standard keyboard. The order was randomized so practice and fatigue effects would cancel out.
During the writing or typing, the EEG cap recorded electrical activity at 256 points on the scalp. The researchers were not interested in which brain region lit up brightest. They were interested in connectivity, meaning how many distant brain regions were synchronized and exchanging signals in real time. Connectivity in the theta band (around 4 to 7 hertz) and the alpha band (around 8 to 12 hertz) is what other neuroscience research has tied to memory encoding, integration of new information, and learning.
The handwriting condition showed widespread connectivity across visual processing regions in the back of the brain, sensorimotor regions controlling the hand, and regions in the parietal cortex that handle the integration of sensory and motor information. The typing condition showed almost none of that connectivity. The motor cortex moved the fingers. The visual cortex saw the letters appear. The two systems mostly worked in isolation.
The plain-language summary van der Meer gave in interviews around the paper was direct. When you write a letter by hand, you have to plan the shape, control the small movements of your fingers, watch the result form on the page, and adjust as you go. That sequence forces the brain to integrate signals across many regions. When you type, the same keystroke produces the same letter every time. The motor planning required is minimal. There is no shape to construct. The integration step never has to happen.
The 2020 children’s study made the case worse, not better
If the 2024 study had stood alone, you could argue the brain pattern in university students says nothing about how children learn. But the 2020 paper looked at exactly this. The team ran the same EEG comparison on 12 twelve-year-old Norwegian schoolchildren and 12 young adults. The connectivity pattern in the children matched the adult pattern. Handwriting produced wide-network connectivity. Typing did not.
The 2020 paper went further. It compared cursive handwriting, where the letters connect, with print handwriting and with typing. Cursive produced the widest connectivity of all. The researchers concluded that the act of building each letter through controlled small motor movements is the part of writing that does the work. Cursive amplifies that. Typing skips it.
The Mueller and Oppenheimer note-taking study
The Norwegian work fits a broader body of research. The most famous behavioral study in this space is a 2014 Princeton and UCLA paper by Pam Mueller and Daniel Oppenheimer called The Pen Is Mightier Than the Keyboard. They had college students take notes during recorded lectures either by hand or on a laptop, then tested them later on factual recall and conceptual understanding.
The students with laptops took more notes, almost twice as many, because typing is faster than handwriting. The students with notebooks took fewer notes but performed substantially better on conceptual questions. Mueller and Oppenheimer’s interpretation: handwriting forces you to summarize and translate the lecture into your own words because you cannot keep up with verbatim transcription. That summarization step is the learning. Laptop-takers had transcripts. Hand-takers had understanding.
The Norwegian EEG work explains why. The brain networks that turn an external input into something the learner owns require integration across regions. Handwriting forces that integration. Typing does not.
Why is Sweden pulling tablets out of classrooms?
Sweden went the other direction first. From around 2010 onward, Swedish schools rolled out tablets aggressively. By the early 2020s, many Swedish classrooms looked like Apple stores. Then the country’s reading comprehension scores started falling.
The Progress in International Reading Literacy Study (PIRLS) is an international benchmark for fourth-grade reading comprehension. Sweden’s PIRLS score dropped from the top tier in 2016 to “intermediate” by 2021. In 2022, Sweden’s Minister of Education Lotta Edholm announced the country would scale back digital devices in early grades and invest in printed books and handwriting practice instead.
The 2023 budget included 685 million kronor, roughly 63 million US dollars, for textbook purchases. The 2024 and 2025 budgets each added another 500 million kronor. Children under six in Swedish schools no longer use tablets during instruction. Older students got their printed textbooks back.
Sweden cited multiple causes for the score drop. Tablets are not the only one. But the policy reversal was direct. The government concluded that the assumption “more screens means better learning” had been wrong for their context. Norway, France, and the Netherlands have begun similar reviews.
What does this mean in the AI era?
The handwriting research was not designed with AI in mind. The 2017, 2020, and 2024 papers ask a narrower question: what happens in the brain when a person writes versus types? But the findings land at a particularly uncomfortable moment for education. Every major AI tool encourages typing. Claude, ChatGPT, and Gemini are designed for text input. Students are using these tools more, and they are using them by typing. AI-assisted writing accelerates a shift from pen to keyboard that has already been underway for two decades.
The risk is not that AI tools are bad. Used well, they extend what a learner or a professional can produce. The risk is that AI tools make the typing path so frictionless that students never develop the brain networks handwriting builds. A child who types their way through every assignment, with AI generating the first draft and the student lightly editing, is doing very little of the integration work that builds memory and conceptual understanding.
The same applies to adults. The professional who types notes during meetings, asks AI to summarize what was said, and lets AI draft their follow-up emails is offloading the very steps that turn external input into internal knowledge.
Where the human still matters
What AI cannot do here, no matter how good it gets:
- AI cannot make your brain build the neural connections handwriting builds. That requires your hand and your attention.
- AI cannot do the small integration work that turns a fact you read into a fact you actually remember.
- AI cannot generate the small surprise of your own handwriting revealing what you actually think, the way journaling does.
- AI cannot replace the focus that comes from slowing down enough to form each letter.
Use AI for the boring parts. Use your hand for the parts that change you.
What should you actually do?
If you are a parent
Keep handwriting in the routine. Even if your child’s school is heavy on tablets, build in a daily window where the child writes by hand. Journaling, copying out passages, working through math on paper, drafting a story. The format matters less than the act of forming letters by hand. The research suggests cursive is best, but print handwriting still produces far more connectivity than typing.
Do not be afraid of AI. Beginners in AI is built on the belief that AI tools can make life better when used well. But pair AI use with handwriting practice. Have your child write notes during a YouTube lesson and only then use AI to fill in what they missed. The order matters.
If you are a student
Take lecture notes by hand. The Mueller and Oppenheimer research is direct: laptop note-takers do worse on conceptual questions than hand note-takers. If you do use a laptop, use it for transcription only and write a handwritten summary at the end of the class. The summary is where the learning happens.
When you study with AI, use it to explain concepts you cannot grasp. Then close the chat and write the concept out by hand in your own words. If you cannot write it without looking, you did not learn it.
If you are a teacher
Make handwriting practice a non-negotiable part of the day for any subject that involves new vocabulary, new concepts, or memorization. Spelling, history, science terminology, foreign language vocabulary. The neuroscience is clear that the act of writing the word by hand is part of how it gets remembered.
You can absolutely use AI in your classroom. The handwriting research does not argue against AI. It argues against typing as the only mode of student-content interaction. AI can plan a lesson, generate worksheets, grade essays, and draft individualized feedback. Use it for those things. Then have the students write.
If you are a working professional
Keep a paper notebook for meetings. Type the action items into your tool stack later if you have to, but capture in the moment by hand. The integration work happens at the point of capture.
Journal by hand. Five minutes in the morning or five minutes at night. The research suggests that writing about something is closer to thinking about it than typing about it. If you want to think clearly, write clearly. By hand.
The Beginners in AI position
We are pro-technology. We exist to help non-technical readers use AI well. We believe AI will make life better for people who learn to work with it.
We are also pro-human first. AI is at its best when it enhances what a person can do, not when it replaces the parts of a person’s work that build the person. Handwriting is one of those parts. So is conversation. So is sitting with a problem instead of asking an LLM to solve it. So is reading a book end to end. The research on handwriting and the brain is one of the cleanest pieces of evidence we have for the broader principle.
Use AI. Use a pen too.
Frequently asked questions
Does it matter whether I use a digital pen or a real pen on paper?
In the 2024 study, participants used a digital pen on a touch-sensitive surface, not paper. The wide-network connectivity still appeared. The act of forming the letters is what matters more than the surface. That said, paper has the advantage of zero distractions and no notifications. If your goal is focused learning, paper is the safer choice.
Is cursive really better than print?
The 2020 study suggested cursive produces the widest connectivity. The likely reason is that cursive demands continuous fine motor control as letters connect into each other. But print handwriting still produces far more connectivity than typing. Any handwriting beats no handwriting.
Does this mean we should ban screens in classrooms?
Van der Meer’s own statements have been more careful than that. She advocates for keeping handwriting in the curriculum, especially for young learners, while not opposing digital tools entirely. Sweden’s reversal is also partial, not a full ban. The realistic position is balance: handwriting for vocabulary, new concepts, note-taking, and journaling; digital tools for research, collaboration, and tasks where speed and editing matter.
Will AI tools eventually generate handwriting practice for kids?
They already do. Tools can generate cursive worksheets, custom letter-tracing pages, and personalized handwriting drills. This is a good example of AI used to enhance the human work, not replace it. The student still has to pick up the pen.
If I am an adult who hasn’t written by hand in years, is it too late?
No. The brain stays plastic into late adulthood. The networks that activate during handwriting are not skills you lose after childhood. They activate every time you write. Start with five minutes a day. A morning journal entry, a meeting note, a grocery list.
Where can I read the original 2024 paper?
The full paper is open access at Frontiers in Psychology. The citation is Van der Weel & Van der Meer, “Handwriting but not typewriting leads to widespread brain connectivity: a high-density EEG study with implications for the classroom,” Frontiers in Psychology, Vol 14, 1219945 (January 26, 2024). Link in Sources below.
Sources
- Van der Weel, F. R., & Van der Meer, A. L. H. (2024). Handwriting but not typewriting leads to widespread brain connectivity: a high-density EEG study with implications for the classroom. Frontiers in Psychology, 14, 1219945.
- Ose Askvik, E., Van der Weel, F. R., & Van der Meer, A. L. H. (2020). The Importance of Cursive Handwriting Over Typewriting for Learning in the Classroom: A High-Density EEG Study of 12-Year-Old Children and Young Adults. Frontiers in Psychology, 11, 1810.
- Mueller, P. A., & Oppenheimer, D. M. (2014). The Pen Is Mightier Than the Keyboard: Advantages of Longhand Over Laptop Note Taking. Psychological Science, 25(6), 1159-1168.
- Audrey van der Meer faculty profile, Norwegian University of Science and Technology.
- Why Writing by Hand Is Better for Memory and Learning, Scientific American, February 2024.
- Writing by hand may increase brain connectivity more than typing on a keyboard, EurekAlert / Frontiers press release, January 2024.
- Sweden Ministry of Education statements on PIRLS results and 2023 digital scale-back; coverage in major outlets including Undark and Think Academy.
You might also like
- AI for Education in 2026: A Beginner’s Overview
- AI for Teachers: Practical Uses for K-12 and Higher Ed
- AI for Parents: How to Help Your Child Use AI Well
- AI for Homeschooling: Lesson Planning and Beyond
- AI for Students: Studying Smarter Without Cheating Yourself
- Claude for Teachers: The Beginner’s Setup Guide
- Common AI Myths, Debunked
- Beginners in AI Special Reports