This foundational 1988 paper by Lehmann et al. challenges the traditional view of brain waves in cognitive neuroscience. Instead of treating the electroencephalogram (EEG) as a continuous stream of oscillating voltages, the authors propose that the brain operates in a series of discrete, stable "microstates"—the literal "atoms of thought."
The Problem: Missing the Spatial Big Picture
In the late 1980s, most EEG research focused on "time series analysis"—tracking the wiggly lines of voltage over time at specific, isolated points on the scalp. While this approach is useful for identifying general states like sleep or wakefulness, it fundamentally misses the spatial "big picture" of brain activity.
The authors argue that the brain’s functional state is much better represented by the spatial configuration of its global electromagnetic field. If the map of the electric field changes, it indicates that a different population of neurons has become active. The challenge for researchers was to move beyond single-channel frequencies and develop a robust way to segment the brain's global, spatial activity into meaningful functional units.
Methodology: Mapping the Brain's Electricity
To capture these spatial shifts, the researchers employed a technique called adaptive segmentation to analyze EEG data across three distinct experimental setups: resting state (eyes closed), a reaction time task to acoustic stimuli, and a visual perception task involving illusionary shapes.
- Spatial Mapping: Instead of looking at one electrode, they used arrays of 16 to 21 electrodes to create "potential distribution maps"—essentially topographic weather maps of the brain's electrical activity.
- Extrema Tracking: They defined the "landscape" of each map by locating its maximum (positive) and minimum (negative) voltage points.
- Segmentation: By monitoring these landscapes over time, they found that as long as the "peaks" and "valleys" stayed within a certain spatial window, the brain was in a single, stable segment. When the landscape shifted abruptly, a new segment began.
To validate whether these segments actually mattered for behavior, they measured subjects' reaction times to acoustic stimuli. They checked if the specific "map class" (the shape of the field) active at the exact moment a sound arrived influenced how fast the subject pressed a button.
Key Insights: The "Atoms of Thought"
- Quantized Consciousness: The authors hypothesize that these stable segments are the basic building blocks of information processing. Just as a movie is made of discrete frames, consciousness and cognition might be "quantized" into these microstates.
- State-Dependent Processing: The brain is not a passive receiver. Its internal state at the moment of stimulus arrival acts as a constraint. The study proved that if the brain is in "Map State A," it might process a sound faster than if it were in "Map State B."
- The Orienting Reaction: The paper frames the "orienting reaction" (the "What is that?" reflex) as a rapid re-adjustment of the brain's functional state to deal with new or important information.
- Shared Neural Resources: By comparing maps generated during attention tasks and figure perception tasks, they found that perceiving a complex shape uses similar neural "landscapes" as voluntary attention. This suggests these different cognitive tasks draw on the same underlying neural machinery.
Core Results
- Temporal Microstructure: Spontaneous brain activity is not a continuous blur; it consists of stable epochs with a mean duration of 210 milliseconds.
- Behavioral Impact: There were significant differences in reaction times depending on which map class was active when a stimulus was presented (ANOVA p < 0.02). This proves these spatial maps aren't just noise—they represent distinct functional states.
- Polarity Reversal vs. Spatial Stability: Interestingly, while the polarity of the field (positive vs. negative) flips every 100ms or so (driven by the alpha rhythm), the actual spatial configuration (the orientation of the field) remains stable for much longer.
Limitations and Legacy
While groundbreaking, the study was constrained by the technology of its time. Using only 16–21 electrodes provided a very coarse view of brain activity, and the authors acknowledged that their "dipole model" was a vast simplification of millions of active neurons. Furthermore, while they proved that states differ and impact behavior, they couldn't yet map specific functions to specific map configurations (e.g., identifying a "language map" versus a "spatial reasoning map").
Despite these limitations, this work paved the way for modern "EEG microstate analysis." Today, this field continues to investigate how these sub-second bursts of organized neural activity relate to the nature of the conscious stream and mental health disorders like schizophrenia.
Summary updated