Summary: We all possess skills that are hard to explain — the precise pressure that keeps a bicycle upright or the instinctive judgment a specialist applies when viewing a complex image. These abilities are examples of tacit knowledge: practical know-how that is difficult to verbalize. New research from MIT shows that some of this unconscious expertise can be revealed through patterns of eye movements and brain activity, and that exposing those patterns to learners can dramatically speed up skill acquisition.
By recording where participants looked and how their brain activity synchronized with visual cues during a challenging image-classification task, researchers discovered that people gradually and unconsciously focused on the most informative parts of each image. Crucially, when participants were shown visual maps of their own attention and brain-response patterns, their accuracy improved substantially — suggesting a practical method for converting tacit knowledge into teachable guidance.
Key Facts
- The “Hidden” Shift: As participants learned the task, both gaze direction and EEG signals converged on the most informative region of each image, even though participants reported scanning the entire image.
- Biofeedback Boost: Showing participants maps of their own unconscious focus produced a measurable increase in classification accuracy.
- Visual Attention as a Proxy: This study provides direct evidence that patterns of visual attention can serve as indicators of tacit knowledge during perceptual classification tasks.
- Broad Applications: The method could accelerate training in areas that rely on perceptual expertise, including medical image reading, professional sports, and skilled crafts.
- Validating Polanyi: The work lends empirical support to Michael Polanyi’s long-standing idea that people “know more than they can tell.”
Source: MIT
Expertise is difficult to pass along. A seasoned cyclist can explain basic steps to a beginner, like posture and how to get started, yet many subtler elements — the exact force and timing that keep balance — remain intuitive and hard to describe. That kind of implicit know-how, known as tacit knowledge, usually requires practice and time to acquire.
A team of engineers at MIT asked whether an expert’s unconscious knowledge could be detected and then turned into explicit guidance to accelerate learning. Their results indicate the answer is yes, at least for a class of visual learning tasks that depend on subtle perceptual cues.
Published in the Journal of Neural Engineering, the study asked volunteers to classify images containing two shapes and a mix of colors and patterns. While participants viewed a sequence of more than 120 images, researchers recorded eye-tracking data and electroencephalography (EEG) to capture both overt gaze and covert neural attention. Each image was constructed so that only one side contained information relevant to correct classification; the other side functioned as distractor noise.
At first, participants guessed and performed like novices. Over time they improved, reaching a level the researchers classified as expert performance. Importantly, even as accuracy grew, volunteers typically reported that they had been looking across the whole image. Objective measurements told a different story: gaze and EEG-derived attention closely followed the informative region, revealing a tacit strategy that subjects could not consciously describe.
The team generated individualized attention maps showing where each participant’s gaze and neural signals concentrated during their early and later trials. Those maps exposed a clear transition: learners moved from broadly scanning images to repeatedly attending to the task-relevant half. When researchers then presented participants with visualizations of their own evolving attention — effectively making the tacit knowledge explicit — participants used that information and improved their classification accuracy further.
Lead author Alexandre Armengol-Urpi, a research scientist in MIT’s Department of Mechanical Engineering, explains that humans carry both explicit knowledge (readily written or explained) and tacit knowledge (internalized, difficult to verbalize). “If we can make that knowledge explicit,” he says, “we can transfer it more easily, accelerating education and training.”
Co-authors include Andrés F. Salazar-Gomez (MIT Media Lab), Pawan Sinha (Department of Brain and Cognitive Sciences), and Sanjay Sarma (Professor of Mechanical Engineering). The project was supported in part by Takeda Pharmaceutical Company.
Hidden gaze
The philosophical notion of tacit knowledge traces back to Michael Polanyi, who observed that people often act on knowledge they cannot fully explain. Subsequent empirical work has suggested tacit elements in many expert activities, from medical image diagnosis to subtle visual discriminations. Building on this background, the MIT team explored whether physiological signals could reveal such hidden expertise.
The shape of knowledge
In the experiment, participants viewed images containing combinations of simple geometric shapes — squares, triangles, circles — each with varying colors and patterns. The discrimination rule depended on a complex, hidden spatial asymmetry: only one side of each image contained the diagnostic information. The researchers measured where people looked with eye-tracking cameras, and where attention landed with EEG by encoding each shape to flicker at different, imperceptible frequencies and identifying which flicker entrained the participant’s brain waves.
Maps produced from these data showed that novice participants initially scanned both sides of images. As learning progressed, both gaze and EEG attention biased strongly toward the informative side. Despite this measurable shift, participants typically denied focusing on any particular region when asked, revealing the tacit character of the acquired strategy.
After exposing participants to visual summaries of their own attention patterns, the researchers observed further improvements in classification, demonstrating a practical route for converting implicit competence into explicit instructional cues.
Armengol-Urpi notes that the team is extending the approach to other domains where tacit knowledge matters, including skilled crafts like glassblowing, sports such as table tennis, and medical image interpretation. The central idea is to capture implicit expertise using physiological signals and then reinforce it through targeted biofeedback to speed training and enhance performance.
Funding: This research received partial support from Takeda Pharmaceutical Company.
Key Questions Answered:
A: Tacit knowledge is practical know-how that people use without being able to fully describe it. An expert can often perform a task reliably without being able to explain the subtle perceptual cues and internal heuristics that guide their decisions.
A: They combined eye-tracking with EEG. Eye tracking measured where observers looked, and EEG detected which visual element entrained the brain’s rhythmic response. Together, these signals revealed attention patterns that predicted accurate judgments before participants could consciously report them.
A: Potentially yes. The study showed that presenting learners with visual feedback of their own attention helped them perform better. Future biofeedback tools could make learning more efficient in perceptual and skill-based domains.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by staff.
About this neuroscience research news
Author: Abby Abazorius
Source: MIT
Contact: Abby Abazorius – MIT
Image: Image credited to Neuroscience News
Original Research: Closed access. “Cognitive reinforcement: capturing tacit knowledge and enhancing expertise with a biofeedback interface for visual attention” by Alexandre Armengol-Urpi, Andrés F. Salazar-Gomez, Pawan Sinha and Sanjay E. Sarma. Journal of Neural Engineering. DOI: 10.1088/1741-2552/ae3eb8
Abstract
Cognitive reinforcement: capturing tacit knowledge and enhancing expertise with a biofeedback interface for visual attention
Objective.
Tacit knowledge refers to the practical know-how that experts use but often cannot fully articulate or transfer explicitly. This implicit expertise creates challenges for teaching and training in domains that depend on perceptual skill, including apprenticeships, craftsmanship, sports, and medical image interpretation. The study investigates whether tacit expertise can be captured via EEG and eye-tracking and then leveraged through a gaze-informed biofeedback interface to improve skill transfer.
Approach.
Researchers designed an image-classification task with an embedded spatial asymmetry that participants could learn implicitly. They tracked overt gaze and covert attention using eye-tracking and EEG, respectively, and tested whether explicit feedback derived from these physiological signals could further boost performance beyond ordinary training.
Main results.
As participants trained, both gaze and EEG signals increasingly emphasized task-relevant image regions, indicating an unconscious internalization of the diagnostic asymmetry. Participants who received explicit, gaze-informed feedback based on their own attention patterns achieved greater gains in classification accuracy than an equally trained control group.
Significance.
These findings point to new possibilities for biofeedback-enabled training and expertise transfer. By extracting and visualizing tacit attention patterns, researchers can make implicit expertise accessible to novices and trained learners alike, a process the authors term cognitive reinforcement.