The Network That Decides What You Pay Attention To
You Have an Air Traffic Controller in Your Head
Right now, as you read this sentence, something extraordinary is happening between the front and back of your brain.
Your frontal cortex, the part right behind your forehead, just sent a signal to your parietal cortex, the region near the top and back of your skull. That signal said, essentially: "These words. Focus on these. Ignore the email notification that just popped up. Ignore the ambient noise. Ignore the slight hunger in your stomach. These words, right here, are what matters."
Your parietal cortex replied by adjusting the gain on your sensory inputs, turning up the volume on visual processing and turning down everything else. This entire exchange took about 150 milliseconds. You didn't feel it happen. You just... kept reading.
This is the frontoparietal network in action. And it's doing this thousands of times per day, every single day, without you ever noticing. It decides what gets through to your conscious awareness and what gets filtered out. It's the reason you can concentrate in a noisy coffee shop, track a conversation at a loud dinner party, or read a dense paragraph without getting derailed by every stray thought.
When it works well, you don't notice it at all. When it fails, you notice everything. Too much, in fact. Every sound, every visual flicker, every random thought becomes equally demanding of your attention. That experience has a name that about 10% of adults are familiar with: ADHD brain patterns.
But here's the thing that makes the frontoparietal network genuinely fascinating. It isn't one network. It's two. And they do opposite things.
The Two Networks That Run Your Attention
In 2002, a neuroscientist named Maurizio Corbetta at Washington University published a paper that would reshape how the entire field thinks about attention. Using fMRI, Corbetta and his colleague Gordon Shulman showed that the brain doesn't have a single "attention system." It has two, and they're organized in fundamentally different ways.
The Dorsal Attention Network: Your Voluntary Spotlight
The first system runs along the top (dorsal) surface of the brain. Picture a highway connecting two cities: one in the upper frontal cortex (an area called the frontal eye fields) and one in the upper parietal cortex (the intraparietal sulcus). This is the dorsal attention network, and it handles voluntary, goal-directed attention.
When you decide to focus on something, this network turns on. You're studying for an exam and deliberately concentrating on a textbook. You're a goalkeeper tracking a soccer ball in flight. You're a programmer scanning through code looking for a bug. In all these cases, the dorsal attention network is maintaining a top-down attentional set, an internal template that says "this is what I'm looking for, this is what matters right now."
The dorsal network is bilateral, meaning it operates in both hemispheres of the brain. It's active before a stimulus even appears. If someone tells you to watch the left side of a screen, your dorsal attention network starts biasing your visual cortex toward the left visual field before anything shows up. It's predictive, preparatory, and under your conscious control.
The Ventral Attention Network: Your Circuit Breaker
The second system runs along the right side of the brain, lower down. It connects the ventral frontal cortex (near the right temple) to the temporoparietal junction (where the temporal and parietal lobes meet). This is the ventral attention network, and it does something entirely different.
The ventral network is your circuit breaker. It interrupts whatever you're currently doing when something unexpected but potentially important happens.
You're deep in a book. Somewhere behind you, glass shatters. In about 200 milliseconds, your ventral attention network fires, overrides your dorsal network's top-down focus, and redirects your attention toward the sound. You didn't choose to pay attention to the glass breaking. Your ventral network made that decision for you.
Here's the critical detail: the ventral network doesn't respond to just anything unexpected. It responds to stimuli that are behaviorally relevant, things that your brain, based on context and past experience, flags as potentially important. A loud crash gets through. The subtle hum of an air conditioner does not. The ventral network is constantly evaluating the background for signals that warrant interrupting your current focus.
And here's the "I had no idea" moment: the ventral attention network is strongly right-lateralized. It operates predominantly in the right hemisphere of the brain. This is why damage to the right parietal cortex produces one of the most dramatic neurological conditions in all of medicine, but we'll get to that in a moment.
The dorsal and ventral attention networks aren't independent. They work together through a dynamic interplay. The dorsal network sets your attentional priorities ("focus on the road"). The ventral network serves as the interrupt ("a child just ran into the street"). When the ventral network fires, it temporarily suppresses the dorsal network, reorients your attention, and then hands control back. This constant negotiation between sustained focus and interrupt detection is the fundamental architecture of human attention.
Where These Networks Live in the Brain
To understand how EEG captures frontoparietal network activity, you need to know the anatomy. These networks aren't vague clouds of activation. They're precise circuits with identifiable nodes.
| Network | Key Frontal Region | Key Parietal Region | Primary Function |
|---|---|---|---|
| Dorsal Attention Network | Frontal eye fields (FEF) | Intraparietal sulcus (IPS) / Superior parietal lobule | Voluntary, goal-directed focus |
| Ventral Attention Network | Ventral frontal cortex (right-lateralized) | Temporoparietal junction (TPJ, right-lateralized) | Detecting unexpected relevant stimuli |
| Central Executive Network | Dorsolateral prefrontal cortex (dlPFC) | Posterior parietal cortex | Working memory, decision-making, cognitive control |
| Frontoparietal Control Network | Lateral prefrontal cortex | Inferior parietal lobule | Switching between dorsal and default mode networks |
Notice something? All four of these networks share the same basic architecture: frontal regions talking to parietal regions. The frontal cortex handles the executive decisions (what to pay attention to, what rules to apply, what goals to pursue), and the parietal cortex handles the spatial and sensory implementation (where to look, what information to boost, what to filter out).
This frontal-parietal axis is so central to human cognition that some neuroscientists have called it the "multiple demand system," a flexible network that's recruited whenever your brain faces a challenging task, regardless of the specific domain. Whether you're solving a math problem, navigating a new city, or composing an email, you're running computations along the frontal-parietal axis.
What Are the EEG Signatures of Attention?
Here's where things get practical. If the frontoparietal network controls attention, and attention has specific neural signatures, then EEG should be able to capture those signatures in real time.
It can. And the signals are surprisingly informative.
Alpha Suppression: The Attention Spotlight in EEG
alpha brainwaves (8-13 Hz) are the most well-studied EEG marker of attention, and their behavior reveals something elegant about how the frontoparietal network works.
When you're not paying attention to anything in particular, alpha power is high across parietal and occipital cortex. Alpha, in this context, acts as an inhibitory signal. High alpha means "this area is idling." Think of it as a neurological "do not disturb" sign.
When the dorsal attention network activates and directs your focus to a specific location or stimulus, alpha power decreases over the parietal regions processing that stimulus. This is called alpha desynchronization or alpha suppression, and it's one of the most reliable EEG signatures in all of cognitive neuroscience.
But here's the elegant part. Alpha doesn't just decrease over the relevant regions. It increases over the irrelevant regions. If you're told to pay attention to the right side of a screen, alpha drops over the left parietal cortex (which processes the right visual field) and simultaneously rises over the right parietal cortex (which processes the left visual field, where you're not supposed to look).
The frontoparietal network, in other words, doesn't just enhance the signal. It actively suppresses the noise. Alpha is the mechanism it uses to do both.
Beta Coherence: The Conversation Between Frontal and Parietal
If alpha tells you where attention is being directed, beta tells you that frontal and parietal regions are communicating.
Beta oscillations (13-30 Hz) between frontal and parietal electrode sites show increased coherence during sustained attention tasks. Coherence measures how consistently two brain regions oscillate at the same frequency with a stable phase relationship. High frontal-parietal beta coherence means the frontal cortex and parietal cortex are talking to each other, maintaining the top-down attentional set that keeps you focused.
A 2015 study by Bastos and colleagues published in Neuron found that top-down attention signals from frontal cortex to visual areas travel predominantly in the beta frequency range. Meanwhile, bottom-up sensory information flowing from visual areas to frontal cortex travels in the gamma range. Beta is the frequency of control. It's the signal the frontal cortex uses to tell parietal and sensory regions what to pay attention to.
This has profound implications. When your frontal-parietal beta coherence drops, the "control channel" weakens. Your parietal cortex stops getting clear instructions about what to focus on. Sensory inputs start competing for attention without top-down guidance. The subjective experience of this? Distraction.
Frontal Theta: Executive Control Under Load
Frontal midline theta (4-8 Hz), generated primarily in the medial prefrontal cortex and anterior cingulate cortex, increases when the frontoparietal network faces a demanding cognitive control challenge.
Need to hold multiple items in working memory? Frontal theta goes up. Need to inhibit a prepotent response (like not looking at a flashing distractor)? Frontal theta goes up. Need to switch between two competing task sets? Frontal theta spikes.
Research by Cavanagh and Frank (2014) proposed that frontal theta serves as a "need for control" signal. When the brain detects conflict, uncertainty, or high cognitive load, theta power increases over frontal midline sites, signaling the frontoparietal network to tighten its grip on attention and working memory.
This makes frontal theta a real-time indicator of how hard your executive attention system is working. Low frontal theta during a demanding task could mean either that the task has become easy (you've automated it) or that your frontoparietal network has disengaged (you've zoned out). The context tells you which.
| EEG Signature | Frequency Band | What It Reflects | Frontoparietal Connection |
|---|---|---|---|
| Parietal alpha suppression | 8-13 Hz (decrease) | Attention directed to a specific target | Dorsal network biasing sensory cortex |
| Contralateral alpha asymmetry | 8-13 Hz (lateralized) | Spatial attention to left or right | Dorsal network spatial priority map |
| Frontal-parietal beta coherence | 13-30 Hz | Top-down control signal strength | Communication between frontal and parietal nodes |
| Frontal midline theta | 4-8 Hz (increase) | Executive control demand, conflict monitoring | Cognitive control via medial frontal cortex |
| Parietal P300 amplitude | Event-related, 300ms | Stimulus evaluation, attention allocation | Ventral network reorienting response |
When the Frontoparietal Network Breaks: Neglect, ADHD, and the Right Hemisphere
The clearest proof that the frontoparietal network controls attention comes from what happens when it's damaged.
Hemispatial Neglect: Half the World Disappears
Remember the ventral attention network's strong right-hemisphere bias? This anatomical quirk produces one of the most striking conditions in neurology.
When stroke or injury damages the right parietal cortex, particularly the temporoparietal junction, patients develop hemispatial neglect. They don't go blind on the left side. Their eyes work fine. Instead, they lose the ability to attend to the left side of space. The difference is subtle but profound.
A patient with neglect will eat food from only the right side of their plate. They'll shave only the right side of their face. Ask them to draw a clock and they'll squeeze all 12 numbers onto the right half. Ask them to copy a drawing of a house and they'll reproduce only the right side. Some patients deny that their left arm belongs to them.
Why does right hemisphere damage cause left-side neglect, but left hemisphere damage rarely causes right-side neglect? Because of the asymmetry in how the two hemispheres handle attention. The right parietal cortex monitors both left and right space. The left parietal cortex monitors primarily right space. When the right hemisphere is damaged, the left hemisphere can still handle right space but nobody is covering the left. When the left hemisphere is damaged, the right hemisphere can cover both sides, so neglect rarely occurs.
This neurological asymmetry is visible in EEG. In healthy subjects, attention tasks produce alpha suppression that's slightly stronger over the right parietal cortex, reflecting its broader attentional role.
ADHD: The Frontal-Parietal Connection Keeps Dropping
ADHD is not a deficit of attention itself. People with ADHD can ADHD and flow state on tasks they find engaging for hours. The problem is attention regulation, the ability to direct, sustain, and switch attention according to goals rather than impulse.
EEG research has revealed specific frontoparietal signatures associated with ADHD. A meta-analysis by Barry and colleagues found that children and adults with ADHD show:
- Elevated theta power over frontal sites, reflecting underactivation of executive control circuits
- Reduced beta power over frontal and central sites, reflecting weaker top-down regulatory signals
- Decreased frontal-parietal coherence in the beta band during sustained attention tasks, suggesting the communication channel between the frontal "commander" and the parietal "implementer" is unreliable
Think about what this means functionally. The frontal cortex is sending a weaker control signal (less beta). The parietal cortex, lacking clear top-down instructions, defaults to responding to whatever stimulus is most salient in the moment (reduced alpha suppression of distractors). Meanwhile, the executive control system runs at higher load even during routine tasks (elevated frontal theta).
The frontoparietal network in ADHD isn't broken. It's intermittent. It works fine sometimes, which is why hyperfocus is possible. But it drops the connection at unpredictable moments, which is why sustained attention on non-preferred tasks feels like trying to hold a phone call with terrible cell service.
The frontoparietal network operates largely beneath conscious awareness. You can't "will" your beta coherence to increase any more than you can will your heart rate to change by staring at a heart rate monitor. But neurofeedback, which provides real-time information about these signals, gives your brain the feedback loop it needs to learn self-regulation. This is why neurofeedback for attention works: not because it teaches willpower, but because it trains the specific neural circuits that attention depends on.
The Frontoparietal Network and the Default Mode Network: The Great Seesaw
If you've read about the default mode network, you already know about the brain's most famous anticorrelation. When the default mode network (DMN) is active, the frontoparietal network quiets down, and vice versa.
But the real story is more nuanced than a simple seesaw.
Research by Vincent and colleagues (2008) identified a third network, sometimes called the frontoparietal control network, that sits between the dorsal attention network and the DMN. This control network doesn't pick sides. Instead, it acts as a mediator, coupling with the dorsal attention network during focused external tasks and coupling with the DMN during internally directed thought.
The frontoparietal control network appears to be the brain's "switch operator," the system that decides when to transition between externally focused attention and internally focused reflection. People who switch between these modes fluidly tend to score higher on measures of cognitive flexibility, creative thinking, and general intelligence. People who get stuck in one mode, either trapped in focus or trapped in mind-wandering, tend to struggle.
In EEG, this switching dynamic shows up as transient changes in frontal-parietal connectivity patterns. When the switch from default mode to task-positive attention is clean, you see a rapid shift: alpha suppression over parietal cortex, increased frontal-parietal beta coherence, and a brief burst of frontal theta as the control network engages. When the switch is sloppy, these signatures are muddied, with incomplete alpha suppression and unstable beta coherence, and the subjective experience is that foggy, half-distracted state where you're trying to focus but your mind keeps drifting.
Measuring Frontoparietal Attention with Consumer EEG
For most of the history of attention neuroscience, studying the frontoparietal network required lab-grade EEG systems with 64 or 128 electrodes, an EEG cap that takes 30 minutes to set up, and a research team to run the experiment.
That barrier is collapsing.
The key insight is that you don't need to map every node of the frontoparietal network to capture its functional state. You need electrodes over two critical zones: frontal cortex and parietal cortex. If you can measure the activity at both endpoints and the relationship between them, you can track the network's operation in real time.
The Neurosity Crown places its 8 EEG channels at CP3, C3, F5, PO3, PO4, F6, C4, and CP4. Look at that electrode montage through the lens of the frontoparietal network and something clicks. F5 and F6 sit over the lateral frontal cortex, near the dorsolateral prefrontal cortex and frontal eye fields. CP3 and CP4 sit over centroparietal cortex. PO3 and PO4 sit over parieto-occipital cortex. The Crown literally spans the frontal-parietal axis.
This means the Crown can capture:
- Alpha power at parietal sites (CP3, CP4, PO3, PO4): the attention spotlight's footprint
- Beta power at frontal sites (F5, F6): top-down control signal strength
- Frontal-parietal coherence (F5/F6 to CP3/CP4 or PO3/PO4): communication between the network's command center and its implementation sites
- Frontal theta (F5, F6): executive control demand
The Crown's 256Hz sampling rate provides more than enough resolution for all of these measurements. Alpha (8-13 Hz) and beta (13-30 Hz) need only about 60Hz of sampling to capture accurately (by the Nyquist theorem, you need at least twice the frequency of interest). At 256Hz, the Crown can resolve oscillations up to 128Hz, well into the gamma range.
The on-device N3 chipset processes power spectral density and power-by-band data in real time. Through the JavaScript or Python SDK, developers can access raw EEG data from all 8 channels and compute custom coherence and connectivity metrics that map directly onto the frontoparietal network signatures discussed in this guide.
Using the Neurosity SDK, you could build a real-time frontoparietal attention monitor by combining three signals: parietal alpha power from CP3/CP4/PO3/PO4 (lower alpha means more engaged attention), frontal beta power from F5/F6 (higher beta means stronger top-down control), and beta coherence between frontal and parietal electrode pairs (higher coherence means the network is connected and communicating). Together, these three measures capture the core functional state of the dorsal attention network. Through MCP integration with AI tools like Claude, you could create a system that adjusts your work environment based on the real-time state of your frontoparietal network.
What This Means for How You Think About Attention
The frontoparietal network story changes how you think about focus.
Attention isn't a single thing you either have or don't have. It's at least four things working together: a voluntary spotlight (dorsal attention network), an interrupt detector (ventral attention network), a switch operator (frontoparietal control network), and an executive controller (central executive network). Each has distinct anatomy, distinct EEG signatures, and distinct failure modes.
This means "I can't focus" is never the full story. Maybe your dorsal attention network is fine but your ventral network is overreactive, pulling your attention toward every notification and noise. Maybe your executive control is intact but the switching mechanism is sticky, leaving you trapped in default mode rumination when you need to be in task-positive mode. Maybe the frontal-parietal communication channel is intermittent, like in ADHD, so your parietal cortex keeps losing the signal from your frontal cortex about what to focus on.
Each of these failure modes looks different in EEG. And that's the profound implication of everything we've covered. The frontoparietal network isn't a black box. It's a circuit with measurable signals at every node. For the first time, you can watch this circuit operate in your own brain, in real time, without a research lab.
Your brain has been running this attention network since the day you were born. Every book you've read, every conversation you've followed, every problem you've solved, all of it depended on signals bouncing between the front and back of your brain at the speed of thought. The signals were always there. The electrodes just weren't in the right place to catch them.
Now they are.