You're in a coffee shop. You've got a book open. And three tables away, someone is having a conversation about their upcoming vacation and you cannot un-hear it. You try to read the same sentence four times. And here's the strange thing — the harder you try to ignore them, the louder they get. It's not just that you're distracted. It's that your brain seems to be actively working against you.
That's the compensation trap. And it's one of the most frustrating experiences for people whose auditory filtering doesn't work the way it's supposed to. We've talked before about what this feels like, and we've covered the hardware — noise-canceling headphones, white noise machines, acoustic treatments. But today we're going deeper. We're opening the hood on the actual neural machinery. The auditory filtering stack from the cochlea all the way up to the prefrontal cortex. And we're asking a very specific question: where exactly does this break in ADHD, in autism, in auditory processing disorder?
Here's what we're going to do. We're going to walk through four layers of filtering. Layer one: cochlear gain control — the medial olivocochlear reflex, which is basically a volume knob inside your ear that turns down background noise before it even becomes a neural signal. Layer two: brainstem sensory gating — something called P50 suppression, which is your brain's way of saying "I already processed that, stop sending it." Layer three: thalamic gating via the reticular nucleus — a gatekeeper deep in the brain that decides what gets through to consciousness. And layer four: top-down corticofugal control — the attentional spotlight that your prefrontal cortex uses to say "listen to this voice, ignore that one.
The central question is this. When the filter fails — and you're sitting in that coffee shop unable to read — is it a gain problem? Meaning the signal arrives too loud, the volume knob doesn't turn down. Is it a gating problem? The filter exists but it never engages. Or is it a top-down attention problem? The filter works fine, but nobody's telling it what to suppress. Because those are three fundamentally different failures, and they require fundamentally different solutions.
The answer, it turns out, is not the same for everyone. ADHD looks like one kind of breakdown. Autism looks like another. Auditory processing disorder looks like a third. And a lot of people have more than one of these, which is where things get genuinely complicated.
Let's start at the bottom of the stack. This is where sound becomes electricity, and it's also where the first layer of filtering happens — before you're even conscious of hearing anything.
That filtering is actually solving a much bigger problem — one the whole stack is designed for. It's called the cocktail party problem, and it's been a central question in auditory neuroscience since the nineteen fifties. Colin Cherry described it first — how do you follow one conversation in a room full of them? And the key thing is, this is not the AI version of the problem. In machine learning, cocktail party means source separation — take a mixed recording and algorithmically pull apart the voices. The neuroscience version is different. It's about selective attention in real time, using cues the brain extracts before you're even aware of them.
The AI problem is post-hoc — you've already recorded the mess, now untangle it. The brain has to do this live, with no second pass, and it uses spatial location, pitch, timbre, and temporal structure — the rhythm of speech, the gaps between words — to lock onto one stream and suppress the others. And it does this in about a hundred milliseconds.
And that speed tells us something important. This isn't one mechanism. It's a cascade. By the time you consciously perceive a voice as a voice, it has already passed through multiple gates, each one making a decision about whether to pass it up the chain or damp it down. So the four-layer stack we're walking through today — this is the actual architecture. It's not a metaphor.
Walk me through the layers again, from bottom to top. I want the names and what each one does, in order.
Layer one: cochlear gain control, mediated by the medial olivocochlear reflex. This is the fast, automatic volume knob. It lives in the inner ear itself and it reduces the sensitivity of the cochlea to background noise before the signal even becomes a neural impulse.
It's pre-neural, essentially.
Pre-neural filtering, yes. The outer hair cells in the cochlea physically change shape to dampen the basilar membrane's vibration. We'll get into the mechanics. Layer two: brainstem sensory gating, measured by something called P50 suppression. This is a mid-latency evoked potential — your brain's electrical response to sound — and it shows whether the brainstem is successfully inhibiting redundant or irrelevant auditory input. Layer three: thalamic gating via the thalamic reticular nucleus. This is a thin sheet of inhibitory neurons wrapped around the thalamus like a net, and it acts as a gatekeeper for sensory information heading to the cortex.
Layer four is the one most people think of when they think about attention — top-down corticofugal control from the prefrontal cortex and the auditory cortex. This is the attentional spotlight. It's your brain saying "that voice, not this one.
Here's why the distinction matters. Those first three layers are largely automatic. They happen whether you're trying to pay attention or not. The fourth layer is volitional — it requires effort. And that distinction, between automatic and effortful filtering, is exactly where the compensation trap lives. If your automatic filters are broken, you try to compensate with layer four — top-down effort. But layer four isn't designed to do the job of layers one through three. It's designed to fine-tune a signal that's already been cleaned up.
You're asking the prefrontal cortex to do the janitorial work that should have been handled downstairs. And it's not good at that.
It's terrible at it. And the effort itself — the noradrenergic arousal that comes with trying harder — can actually make things worse by increasing gain across all channels. But that's jumping ahead. The framework I want listeners to hold onto is this: the breakdown question — gain versus gating versus top-down control — maps onto different layers of the stack. A gain problem is layer one, the cochlear volume knob. A gating problem could be layer two or three — brainstem or thalamus. A top-down problem is layer four. And different neurotypes seem to have failures at different layers.
Which means the same experience — can't filter background noise — could have three completely different neural causes. And the intervention that works for one might do nothing for another, or might even make things worse.
That's the thesis of this whole episode. So let's start at the bottom and work our way up.
The cochlea is not a passive microphone. That's the first thing to understand. It's an active, motorized system. Inside it, you've got two types of hair cells. The inner hair cells are the actual transducers — they convert mechanical vibration into electrical signals that travel up the auditory nerve. But the outer hair cells — these are the amplifiers. They change shape in response to sound. They physically lengthen and contract, a property called electromotility, and that movement amplifies the basilar membrane's vibration by something like forty to sixty decibels.
Your ear is literally pumping energy back into the sound signal before it becomes a neural signal. It's not just receiving, it's boosting.
And that boost is tunable. Which brings us to the medial olivocochlear reflex — the MOC reflex. This is layer one of our filtering stack. The superior olivary complex in the brainstem — one of the first stops for auditory information after it leaves the cochlea — sends a bundle of efferent fibers back down to the cochlea. Efferent meaning the signal is traveling away from the brain, toward the sensory organ. These MOC fibers synapse directly onto the outer hair cells, and when they fire, they release acetylcholine, which reduces the outer hair cells' electromotility.
The brain is sending a signal back down to the ear that says "turn down the amplification.
And this happens fast — we're talking tens of milliseconds. It's a reflex loop. Sound enters the cochlea, travels up to the brainstem, the superior olivary complex computes the overall intensity, and if it's loud or noisy, it fires the MOC efferents back down to dampen the outer hair cells. The basilar membrane moves less, the inner hair cells are less stimulated, and less signal goes up the auditory nerve.
It's like turning down the gain on a microphone before the signal hits the cable. The noise is still in the room, but the transducer is less sensitive to it.
Here's where it gets clinically relevant. You can measure MOC reflex strength non-invasively. You play a sound into one ear — a burst of noise — and you measure something called otoacoustic emissions in the other ear. Otoacoustic emissions are tiny sounds the cochlea itself produces as a byproduct of outer hair cell activity. When the MOC reflex activates, those emissions get suppressed. The amount of suppression tells you how strong the reflex is.
The strength of that reflex predicts how well someone can understand speech in a noisy room?
It does, and this is one of those findings that should be more famous than it is. Multiple studies have shown that individuals with stronger contralateral suppression of otoacoustic emissions — meaning a stronger MOC reflex — perform better on speech-in-noise tests. The effect is moderate but consistent. If your cochlear volume knob works well, you have an easier time pulling a voice out of background chatter.
Layer one is essentially a hardware-level automatic gain control. No conscious effort required. The brainstem hears noise and turns down the ear.
Before you're even aware there's noise. Now let's move up one level. Layer two: brainstem sensory gating, measured by P50 suppression. This is not about volume anymore. This is about redundancy filtering.
Explain the paired-click paradigm. I've read about this and it's one of the cleverest experimental designs in neuroscience.
It really is. Here's how it works. You put electrodes on someone's scalp — it's an EEG setup — and you play two identical clicks through headphones, separated by five hundred milliseconds. The first click triggers a cascade of electrical activity in the auditory pathway. About fifty milliseconds after the click, you get a positive voltage deflection in the EEG signal — that's the P50 wave. It reflects the brainstem and early thalamic response to the sound. Then, five hundred milliseconds later, you play the second click.
Same pitch, same volume.
And in a neurotypical brain with healthy sensory gating, the P50 response to the second click is dramatically reduced — suppressed by roughly eighty percent compared to the first. The brain has already processed that sound. It's tagged it as redundant. And it actively inhibits the response to the repeat.
The gate closes on the second click. The brain says "I've heard this one, nothing new here, don't bother cortex with it.
The P50 suppression ratio is calculated by dividing the amplitude of the second response by the first. A ratio of zero point two or lower — meaning eighty percent suppression — is considered normal. Higher ratios indicate impaired gating. The generators of the P50 response are thought to be in the hippocampus and temporal lobe, but the actual gating mechanism involves the reticular formation in the brainstem and the thalamus.
This is still automatic. Nobody is consciously deciding to suppress the second click. It's a pre-attentive process.
Completely pre-attentive. Which is why it's such a useful measure. It tells you whether the basic sensory filtering machinery is working before attention even enters the picture. If your P50 suppression is impaired, you're getting flooded with redundant sensory information that a healthy brain would have already discarded.
Which brings us to layer three. The thalamic reticular nucleus. And this, to me, is the most architecturally elegant piece of the whole system.
The TRN is beautiful. Picture a thin sheet of neurons — about thirty thousand of them in humans — that wraps around the thalamus like a net or a shell. Every single one of those neurons is GABAergic. They're purely inhibitory. And here's the key: every axon that travels from the thalamus up to the cortex, and every axon that travels from the cortex back down to the thalamus, passes through the TRN and can be inhibited by it.
It's a gatekeeper that sits between the sensory world and conscious perception. Everything has to go through it, and it can block anything.
It gets input from both directions. The TRN receives excitatory input from the sensory pathways themselves — bottom-up information about what's coming in — and it also receives input from corticofugal fibers — top-down instructions about what to prioritize. So the TRN is where bottom-up salience meets top-down control. It's the integration point.
It's like a bouncer who takes instructions from both the crowd outside and the manager inside.
If a sound is loud or novel, bottom-up signals tell the TRN to let it through. But if the prefrontal cortex is focused on a specific conversation, top-down signals tell the TRN to suppress everything outside that frequency channel or spatial location. The TRN's GABAergic neurons inhibit the thalamocortical relay neurons, and that inhibition prevents the sensory information from ever reaching the cortex.
Which means you can have a sound hitting your eardrum, going through the cochlea, triggering a brainstem response — but never making it to conscious awareness because the thalamic gate said no.
And this happens constantly. Your thalamus is filtering out vast amounts of sensory information every second. You're only consciously aware of a tiny fraction of what your ears are actually picking up.
Now layer four. Top-down corticofugal control. This is the one people intuitively understand as attention.
Right, but the mechanism is more specific than most people realize. The auditory cortex and the prefrontal cortex send descending projections — corticofugal fibers — that travel all the way back down the auditory pathway. These fibers can synapse onto neurons in the thalamus, the brainstem, even the cochlea itself. And they can enhance or suppress specific frequency channels or spatial locations.
If you're trying to follow a conversation with someone who has a deep voice, your prefrontal cortex can send a signal down that says "boost everything in the one hundred to two hundred hertz range and suppress
everything above two thousand hertz, and ignore anything coming from the left."
That ability — to shape the sensory stream before it reaches conscious perception — is what allows a neurotypical person to follow one conversation in a noisy room. The prefrontal cortex holds a template of the target voice: its pitch, its timbre, its spatial location, its rhythm. That template gets sent down the corticofugal pathway, and the lower levels of the auditory system adjust their tuning to match.
We've walked through the four-layer stack. Cochlear gain control, brainstem gating, thalamic gating, top-down control. Now the question is: where does this break?
Let's start with ADHD, because the evidence here tells a surprisingly specific story. Multiple studies have looked at P50 suppression in people with ADHD, and the finding is consistent: it's normal. The paired-click paradigm shows the same roughly eighty percent suppression you see in neurotypical controls. The brainstem gate works.
Layer two is intact. The automatic redundancy filter is doing its job. Which means the problem has to be somewhere else.
If you look at later components of the evoked potential — specifically the P300, which reflects conscious attentional processing in cortex — that's where ADHD brains diverge. The P300 amplitude is reduced, and the gating of the P300 — the suppression of irrelevant stimuli that have already been consciously processed — is impaired. So the brainstem filters the clicks, but the cortex doesn't properly gate the attended versus unattended streams.
This is a top-down failure. The filter exists, it's physically capable of working, but it's not being told what to suppress.
The prefrontal cortex in ADHD is underactive at baseline. It has difficulty generating and maintaining that attentional template — the representation of "this is the voice I'm listening to, everything else is noise." Without a clear template, the corticofugal system doesn't know what to enhance and what to inhibit. So it does neither effectively.
Which brings us to the compensation trap. Because the intuitive response — and the advice most people get — is "try harder to focus.
This is where the neurochemistry gets brutal. Trying harder engages the prefrontal cortex more intensely. That increases corticofugal output. But in a system where the attentional template is weak or absent, that increased output isn't selective — it amplifies everything. All auditory streams get boosted. The noise gets louder along with the signal.
Simultaneously, the effort triggers noradrenergic arousal from the locus coeruleus. Which has a U-shaped dose-response curve for signal-to-noise ratio.
This is one of the most important curves in all of psychopharmacology, and it's underappreciated. Too little norepinephrine, and you're drowsy, unfocused — the signal is weak. Too much, and you're anxious, overaroused — the noise is amplified and the signal degrades. There's a sweet spot in the middle where signal-to-noise ratio is optimal. Effort pushes you past the peak of that curve.
Trying harder does two things simultaneously. It engages a top-down control system that can't properly target its output, and it floods the system with norepinephrine that amplifies everything indiscriminately. The result is that the very act of trying to filter makes the noise more intrusive.
That's the compensation trap. And it explains why so many people with ADHD describe the experience of focusing in a noisy environment as actively painful. It's not just that they can't filter — it's that the effort of trying makes the sensory experience worse.
You said earlier the evidence points to a different failure mode — a bottom-up gain problem.
And the key finding here comes from measuring the MOC reflex. Multiple studies have found that MOC reflex strength — measured by contralateral suppression of otoacoustic emissions — is reduced in many autistic individuals. Sometimes significantly reduced. The cochlear volume knob doesn't turn down as effectively.
Layer one is compromised. The signal arrives at the brainstem already louder than it should be.
It gets worse at layer two. Unlike ADHD, where P50 suppression is normal, studies in autism often show reduced P50 suppression. The brainstem gate doesn't close properly on redundant input. So you've got a double hit: the cochlea isn't dampening the signal, and the brainstem isn't filtering the repeats.
Which means the sensory information reaching the thalamus and cortex is already overwhelming. Before attention even enters the picture.
This is consistent with the broader sensory hypersensitivity phenotype in autism. It's not that autistic people have super-hearing in the sense of lower detection thresholds — many have perfectly normal hearing acuity. The problem is that the gain is set too high and the early filters aren't engaging. The world literally arrives louder.
This is primarily a GABAergic story. The inhibitory neurotransmission that drives both the MOC reflex and P50 suppression relies on GABAergic interneurons. In autism, there's converging evidence — from genetics, from postmortem studies, from animal models — of GABAergic dysfunction. Particularly in the parvalbumin-positive fast-spiking interneurons.
Parvalbumin-positive interneurons are the workhorses of inhibition. They account for roughly forty percent of all GABAergic neurons in cortex, and they're critical for generating gamma-frequency oscillations — the brain rhythms that coordinate sensory processing. When these interneurons aren't functioning properly, inhibition is weakened across multiple levels of the auditory pathway. The MOC reflex is weaker. P50 suppression is reduced. The thalamic gate is leakier.
In autism, the core problem is that the inhibitory brakes are worn down. The signal arrives too hot and the filters don't close.
That brings us to APD — auditory processing disorder — which seems to fall somewhere in the middle. The evidence here is less settled, but the emerging picture suggests a thalamic gating problem. The signal arrives at normal volume, the brainstem gate works, but the thalamic reticular nucleus fails to properly inhibit irrelevant streams.
Layer three is the primary failure point. The integration hub where bottom-up salience meets top-down control isn't doing its job.
The mechanism is likely GABAergic interneuron dysfunction within the TRN itself. The TRN is a sheet of pure inhibition — every neuron in it is GABAergic. If those neurons aren't firing properly, the gate stays open. Everything gets through to cortex. And because the TRN is where top-down control interfaces with bottom-up filtering, even a strong attentional template from the prefrontal cortex can't close the gate if the TRN can't execute the instruction.
Which creates a particularly frustrating experience. You're trying to focus, you know what you want to listen to, but the thalamus just won't stop forwarding everything.
Here's the clinical complication. These three failure modes — bottom-up gain in autism, mid-level gating in APD, top-down control in ADHD — they're not mutually exclusive. Many people have more than one of these conditions. When that happens, the failures stack.
Someone with both ADHD and autism, which is not uncommon, could have reduced MOC reflex strength, reduced P50 suppression, and impaired top-down control. The entire stack is compromised.
That's why broad labels like "sensory processing disorder" can be unhelpful. They collapse distinct mechanisms into a single bucket. The treatment for a gain problem is different from the treatment for a gating problem is different from the treatment for a top-down control problem.
If the problem is primarily bottom-up gain — the autism pattern — then environmental modifications make the most sense. Acoustic treatment, noise-canceling headphones, controlling the input rather than trying to filter it after the fact.
If the problem is primarily top-down — the ADHD pattern — then stimulant medication that boosts prefrontal dopamine and norepinephrine can help. It strengthens the attentional template, which gives the corticofugal system a clearer target. Methylphenidate and amphetamine both increase dopamine and norepinephrine in the prefrontal cortex, and there's evidence they improve P300 gating specifically.
If the problem is mid-level thalamic gating — the APD pattern — then auditory training protocols that target temporal processing might help by strengthening the TRN's ability to discriminate between streams based on timing cues
Those APD protocols are interesting because they don't target attention directly. They target the temporal resolution of the auditory system — the brain's ability to track rapid changes in sound. If the TRN can better discriminate between streams based on millisecond-level timing differences, the gate works more effectively even if the attentional template is fine.
The intervention logic follows the failure mode. Gain problem: control the input. Gating problem: train the temporal filter. Top-down problem: strengthen the attentional template pharmacologically.
Which brings us to the central question the prompt was driving at. Is this one mechanism with different presentations, or different failure modes?
The evidence, from everything we've walked through, leans toward different failure modes.
The neurophysiological signatures are distinct. Normal P50 suppression with impaired P300 gating in ADHD. Reduced MOC reflex and reduced P50 suppression in autism. Intact peripheral function with suspected TRN dysfunction in APD. These aren't different expressions of the same underlying problem — they're different breaks in different parts of the cascade.
Though I want to push back slightly on the neatness of that picture. The heterogeneity within each neurotype is enormous. Not everyone with ADHD has the same P300 profile. Not every autistic person shows reduced MOC reflex. The categories themselves are behavioral diagnoses, not neurophysiological ones.
That's a crucial caveat. The diagnostic labels are clusters of observable traits, not mechanism-level descriptions. Two people with the same ADHD diagnosis might have very different underlying filter failures. One might have a clean top-down problem. Another might have subtle brainstem involvement that standard P50 paradigms miss.
Which is why "try harder" fails across the board, but for different reasons depending on the mechanism. If you have a top-down failure, trying harder amplifies noise because your attentional template is too weak to direct the gain. If you have a bottom-up gain failure, trying harder is irrelevant — the problem is happening before effort even enters the system. And if you have a thalamic gating failure, trying harder can't close a gate that's physically stuck open.
The broader lesson is that understanding mechanism changes how you evaluate interventions. If someone tells you to try a white noise machine, the question isn't "does white noise help with sensory issues" — the question is "which layer is failing, and does white noise address that specific mechanism?
White noise might help a gain problem by providing a steady baseline that reduces the contrast between signal and noise. But for a top-down control problem, it might just add another stream to the unfiltered pile.
For a thalamic gating problem, it depends on whether the white noise helps the TRN establish a rhythm or just gives it more to gate. The point is that interventions aren't universally good or bad — they're matched or mismatched to mechanism.
There's a humility in this framework that I think is valuable. It moves away from "here's what works for sensory issues" toward "here's what might work given where your filter is breaking.
It reframes the experience itself. The person who can't filter background noise in a café isn't failing to try hard enough. They're not weak-willed or undisciplined. Their auditory filtering stack has a specific failure mode — possibly more than one — and the effort they're expending is counterproductive.
Which is not just validating. It's actionable. If you know your problem is primarily top-down, you might experiment with stimulant medication or with strategies that reduce the cognitive load on the prefrontal cortex — like breaking work into shorter bursts where the attentional template doesn't have to be maintained as long.
If your problem is primarily bottom-up gain, you might invest in acoustic treatment for your workspace or in high-quality noise-canceling headphones, and stop beating yourself up for not being able to "focus through" noise that is literally arriving at your cortex at a higher volume than it does for other people.
If your problem is mid-level gating, you might look into auditory training protocols or FM systems that deliver the target signal directly to your ears, bypassing some of the acoustic competition that overwhelms the TRN.
None of this requires a formal diagnosis with electrode arrays in your brainstem. The patterns are often discernible from behavior and response to different interventions. If stimulants help your noise sensitivity, that points toward a top-down component. If they don't touch it, that points toward a gain or gating problem.
The field is moving toward mechanism-specific profiling. There are labs working on quick behavioral assays — things like speech-in-noise thresholds under different conditions — that can hint at which layer is involved without needing an EEG cap.
That's where the open question sits. Can we develop something simple enough to be clinically useful that distinguishes between these failure modes in an individual? Because right now, most people with noise sensitivity get the same generic advice regardless of mechanism. And for a significant number of them, that advice is not just useless — it's making things worse.
The compensation trap in action. The advice itself becomes part of the problem.
The MEG Lab at UC San Francisco is already working on something like this. They're developing closed-loop auditory stimulation that reads EEG markers of gating in real time and adjusts gain adaptively. The system detects when your P50 suppression is weakening — meaning the gate is starting to fail — and delivers a precisely timed auditory pulse that effectively resets the filter.
An external prosthetic for the gating mechanism itself. Not noise-canceling headphones that block sound, but a device that helps your brain do the blocking.
That's the vision. It's speculative, and it's early-stage, but the principle is sound. If we can measure when the gate is failing, we can intervene at the right moment to help it close. The timing has to be exquisitely precise — we're talking millisecond-level delivery locked to the phase of ongoing neural oscillations — but the technology exists to do it.
It's a different philosophy entirely from the coping tools we have now. Current approaches either change the environment or change your conscious strategy. This would change the filter itself, directly, without you having to do anything.
That's where I think the field is heading. Away from broad diagnostic categories and toward mechanism-specific profiles with matched interventions. The ADHD label tells you something about behavior. It doesn't tell you whether your auditory filtering problem is a layer-four failure, a subtle layer-two involvement, or something else entirely.
The question the prompt was circling — and honestly, the question that sits underneath a lot of people's frustration with this — is "why can't I just try harder and fix this?" And the answer, after walking through the entire stack, is that trying harder engages the wrong mechanism for most of the failure modes.
If your filter is broken at layer one or layer two, effort never reaches the problem. If it's broken at layer three, effort can't close a gate that's physically stuck. If it's broken at layer four, effort actively makes it worse by amplifying everything. There's no failure mode where "try harder" is the correct intervention.
The experience of being unable to filter noise isn't a character flaw. It's not a lack of discipline. It's a specific neural failure mode — possibly more than one — and understanding which mode is yours is the first step toward actually solving it.
That reframing matters. The shame people carry about this — the sense that they're somehow weak or lazy because they can't concentrate in a noisy room — that shame is built on a misunderstanding of what's actually happening in their auditory system.
You wouldn't tell someone with a broken leg to try harder to walk. You'd fix the leg. The auditory filtering stack is no different — it's physical circuitry, and when it breaks, the solution is to identify the break and address it, not to moralize the failure.
To land this: the evidence points toward different failure modes across neurotypes. ADHD is primarily a top-down control failure. Autism is primarily a bottom-up gain problem. APD is primarily a mid-level gating problem. They can stack, they're heterogeneous within each category, and the right intervention depends on knowing which layer is involved.
The open question that'll be fascinating to watch is whether we can build something — a test, a device, a protocol — that maps an individual's specific filter profile quickly and cheaply enough to be clinically useful. Because right now, most people are guessing.
Hilbert Flumingtop produced this episode, and we're grateful as always for his work.
This has been My Weird Prompts. Find us at myweirdprompts.com or wherever you get your podcasts.
And now: Hilbert's daily fun fact.
Hilbert: In seventeen eighty-two, a buzkashi tournament on the island of Vanuatu was narrowly averted when the cargo ship carrying the championship goat carcass ran aground on a reef three miles from shore, and the players — unable to agree on whether a waterlogged goat was regulation — spent the next four days arguing maritime salvage law instead.
...right.
We'll be back next time.
