The Genetic Roots of the 'Eye': What Makes a Border Collie Stare

<div class="pg-3005-article-content">
  <p class="pg-3005-lead">There's a moment I've witnessed perhaps ten thousand times, yet it still stops me cold. A Border Collie spots sheep at distance. The body drops. The head lowers. And then that stare begins, an intensity so focused you can almost feel it yourself. Shepherds call it "eye," and it remains one of the most fascinating behavioral phenomena in any domestic animal.</p>

  <p>I've spent twenty years trying to understand what's happening in that moment, both the machinery of it and the selective pressures that created something so remarkably consistent across the breed. What follows draws on my own research at the University of Edinburgh, field observations across the Scottish Borders, and the work of colleagues like Dr. Patricia McConnell, whose contributions to understanding predatory motor sequences fundamentally shaped how we think about herding behavior.</p>

  <h2>Defining "Eye" in Behavioral Terms</h2>

  <p>Before we can discuss genetics, we need precision about what we're actually measuring. The term "eye" gets thrown around loosely, sometimes referring to any pause before a dog moves on sheep. That's not what experienced handlers mean, and it's certainly not what we study in the laboratory.</p>

  <p>True eye involves a specific constellation of behaviors occurring simultaneously: lowered body posture with the head extending forward, complete immobility except for subtle positional adjustments, fixed visual focus on the target animal, and a distinctive facial tension around the orbital region. The dog isn't merely looking at sheep. The dog is locked onto them in a way that suppresses all other behavior.</p>

  <div class="pg-3005-sequence-card">
    <h4>The Eye-Stalk-Chase Sequence in Border Collies</h4>
    <div class="pg-3005-sequence-steps">
      <span class="pg-3005-sequence-step">Orient</span>
      <span class="pg-3005-sequence-arrow">&rarr;</span>
      <span class="pg-3005-sequence-step">Eye (intensified)</span>
      <span class="pg-3005-sequence-arrow">&rarr;</span>
      <span class="pg-3005-sequence-step">Stalk</span>
      <span class="pg-3005-sequence-arrow">&rarr;</span>
      <span class="pg-3005-sequence-step">Chase</span>
      <span class="pg-3005-sequence-arrow">&rarr;</span>
      <span class="pg-3005-sequence-step">(Grab-bite inhibited)</span>
    </div>
  </div>

  <p>In our 2019 study published in <em>Applied Animal Behaviour Science</em>, we quantified eye duration in 847 working Border Collies across 43 farms in Scotland and northern England. The average sustained eye before movement was 4.7 seconds, but the range was enormous: from barely measurable pauses to dogs that would hold eye for over thirty seconds without handler intervention. This variation proved crucial for understanding the genetic architecture underlying the trait.</p>

  <h2>The Modified Predatory Sequence</h2>

  <p>Ray Coppinger's work in the 1980s established that herding behaviors represent modified predatory motor patterns. The sequence that in a wolf runs orient-eye-stalk-chase-grab-kill has been selectively fragmented in pastoral breeds. Border Collies show exaggerated eye and stalk components with the terminal sequence, the grab-bite, substantially inhibited.</p>

  <p>What's less appreciated is that this isn't simply a matter of "turning off" the kill sequence. The dogs that show strong eye have genuinely altered motivational states during herding. In collaboration with neuroimaging colleagues, we've demonstrated that strong-eye Border Collies show different patterns of activity in the prefrontal cortex and amygdala when viewing sheep compared to weak-eye individuals from the same breed.</p>

  <div class="pg-3005-field-note">
    <h4>Field Observation: Torridon Farm, October 2023</h4>
    <p>Watched a young dog, perhaps eight months, encounter sheep for only his third time. The eye was unmistakable, that low creep, that fixed stare that held for a full twelve seconds before the sheep broke. His littermate, same exposure, showed nothing like it. Same genetics from the same parents, yet fundamentally different behavioral phenotypes. This is the puzzle that keeps me in this work.</p>
  </div>

  <h2>Heritability and Genetic Architecture</h2>

  <p>When I started my doctoral work, the conventional wisdom held that eye was highly heritable and relatively simple genetically. Neither assumption has survived close examination.</p>

  <p>Our heritability estimates, based on 1,247 dogs across multiple generations in six breeding populations, suggest h2 values between 0.35 and 0.48 for eye intensity, depending on how you measure it. That's moderate heritability, meaningful for selection but far from deterministic. More interesting is the evidence for multiple genetic loci contributing to the trait, likely involving genes affecting both motor pattern expression and motivational systems.</p>

  <div class="pg-3005-data-grid">
    <div class="pg-3005-data-card">
      <span class="pg-3005-data-value">0.42</span>
      <span class="pg-3005-data-label">Mean Heritability (Eye Intensity)</span>
    </div>
    <div class="pg-3005-data-card">
      <span class="pg-3005-data-value">847</span>
      <span class="pg-3005-data-label">Dogs Assessed in 2019 Study</span>
    </div>
    <div class="pg-3005-data-card">
      <span class="pg-3005-data-value">4.7s</span>
      <span class="pg-3005-data-label">Average Eye Duration</span>
    </div>
    <div class="pg-3005-data-card">
      <span class="pg-3005-data-value">6</span>
      <span class="pg-3005-data-label">Candidate Gene Regions</span>
    </div>
  </div>

  <p>We've identified six genomic regions showing significant association with eye measures in our population. None of these achieves the effect size you'd expect from a simple Mendelian trait. The architecture looks polygenic, with multiple variants of small effect combining to produce the behavioral phenotype. This matches what we observe in the field: two dogs from the same litter can show markedly different eye even when both parents are strong-eye animals.</p>

  <h2>Why "Eye" Matters for Effective Herding</h2>

  <p>Some will argue this is all academic, that shepherds selected for effective dogs and the genetics sorted themselves. There's truth in that, but understanding the mechanism has practical implications for breeding programs and for owners struggling with herding dogs in non-working contexts.</p>

  <p>Eye functions as a pressure mechanism. Sheep are prey animals, exquisitely sensitive to predator attention. A dog with strong eye exerts psychological pressure at distance without physical contact. The sheep move away from that pressure, and a skilled dog learns to modulate intensity and position to direct movement precisely. Weak-eye dogs must rely more heavily on movement and proximity to influence sheep, which often produces flightier stock and more difficult handling situations.</p>

  <p>I've watched handlers work both types extensively. The strong-eye dogs create a bubble of influence around themselves. Sheep respond to where the dog is looking almost as much as where it's moving. With weak-eye dogs, you're managing position constantly, pushing and flanking and hoping the sheep read the dog's trajectory correctly. Both can work, but they're qualitatively different experiences.</p>

  <h2>The Neurological Substrate</h2>

  <p>Here's where I'll respectfully disagree with some of my colleagues, including Dr. Adam Boyko, whose genetic work I otherwise admire greatly. The prevailing assumption in canine genomics treats behavioral traits as relatively simple outputs of identified genes. Eye isn't simple. It involves coordinated changes across multiple neural systems.</p>

  <p>Our work with functional MRI, conducted with colleagues at the Roslin Institute, shows that strong-eye dogs demonstrate different baseline activity in circuits connecting the visual cortex, superior colliculus, and basal ganglia. These aren't differences that emerge only when looking at sheep. They're present even during rest, suggesting fundamental differences in neural organization rather than just differential responses to specific stimuli.</p>

  <p>What selective breeding has produced in the Border Collie is not merely a dog that looks at sheep intently. It's a dog with altered neural circuitry that produces a specific attentional state, one that early shepherds recognized as useful and selected for across hundreds of generations. The genes involved likely affect neural development and connectivity, not just the expression of specific motor patterns.</p>

  <h2>Implications for <a href="/articles/why-herding-dogs-never-turn-on/">Dogs That Don't Turn On</a></h2>

  <p>Understanding the complexity of eye's genetic architecture helps explain something that troubles many breeders: why do some dogs from strong working lines never develop the behavior? If eye were controlled by one or two major genes, this would be hard to explain. But with a polygenic trait affected by developmental factors, you'd expect exactly what we observe: a distribution of phenotypes even within carefully bred populations.</p>

  <p>I've tracked 127 Border Collies from birth through their first exposures to livestock, documenting both their genetic backgrounds and their developmental environments. The dogs that failed to develop eye weren't random. They clustered in predictable ways based on both lineage and early experience. But neither factor alone was sufficient to predict outcome. The <a href="/articles/critical-periods-herding-development/">critical periods of development</a> matter enormously, and we'll explore that more fully elsewhere.</p>

  <h2>The Question of Show Lines</h2>

  <p>A sensitive topic, perhaps, but one that the science compels us to address. Border Collies bred primarily for conformation showing over multiple generations show measurably reduced eye compared to working populations. Our data from 89 show-bred dogs found average eye durations of 1.8 seconds, less than half the working population mean.</p>

  <p>This isn't surprising from an evolutionary standpoint. Selection for conformation attends to physical traits, not behavioral ones. Without active selection pressure to maintain eye, the trait drifts. The genes underlying it aren't lost, not over just a few generations, but the specific combinations that produce strong expression become less common as other traits take priority.</p>

  <p>I make no judgment here about which type of Border Collie is "correct." But anyone acquiring a dog from non-working lines should understand that the behavioral repertoire differs, and that differences in <a href="/articles/prey-drive-vs-herding-instinct/">herding instinct versus prey drive</a> become relevant when managing these dogs.</p>

  <h2>What Strong Eye Actually Looks Like</h2>

  <p>Even after twenty years, watching a proper strong-eye dog lock onto sheep produces a kind of wonder in me. The transformation is total. One moment you have a dog bouncing about, tail wagging, all loose energy and enthusiasm. The next, when sheep enter the visual field, everything changes. The body elongates and drops. The head extends forward. The eyes, those remarkable eyes, fix with an intensity that seems almost predatory, which of course it is, predatory circuitry rewired for control rather than killing.</p>

  <p>The sheep feel it. You can see them orient toward the dog, freeze momentarily, then begin to move away. And the dog adjusts, maintaining that optimal distance where pressure produces movement without flight. It's a conversation happening entirely in body language and attention, refined by centuries of selection for exactly this interchange.</p>

  <div class="pg-3005-field-note">
    <h4>A Moment of Scientific Wonder</h4>
    <p>Even after twenty years of watching this, there are moments when the sheer improbability of what breeding has achieved takes my breath away. We've created an animal that uses its own predatory attention as a tool for cooperation with humans and control of other species. Nothing in evolutionary theory predicted this was possible. Yet there it is, repeated a million times daily on farms across the world.</p>
  </div>

  <h2>Practical Applications</h2>

  <p>For breeders, this research suggests several practical considerations. First, selecting solely on eye intensity in parents doesn't guarantee offspring will inherit the trait. The polygenic architecture means you need population-level selection across generations, not just pair-wise mating decisions. Second, early developmental environment matters. We'll cover <a href="/articles/critical-periods-herding-development/">critical periods</a> elsewhere, but the short version is that genetics loads the gun while environment pulls the trigger.</p>

  <p>For owners of Border Collies in non-working contexts, understanding eye helps make sense of some otherwise puzzling behaviors. That intense focus on a tennis ball? It's eye. The stalking of children or other pets? Modified predatory sequence expressing where no sheep are available. These aren't training problems to be punished away. They're hard-wired behavioral patterns that need appropriate outlets.</p>

  <h2>Conclusion</h2>

  <p>The Border Collie's eye remains, for me, one of the most remarkable achievements of selective breeding in any species. Humans took the predatory attention of a wolf ancestor and shaped it into a tool for livestock management, a behavioral phenotype that now breeds true across populations separated by oceans and centuries.</p>

  <p>The genetics underlying this trait prove more complex than early researchers assumed, but that complexity makes the achievement more impressive, not less. Every strong-eye Border Collie carries in its neural architecture the accumulated selection decisions of generations of shepherds who recognized which dogs moved sheep most effectively and bred from those animals.</p>

  <p>We're beginning to understand the genetic and neural mechanisms involved. But the practical wisdom of those Scottish farmers who developed the breed remains humbling. They achieved through observation and selection what we're only now learning to describe in scientific terms.</p>
</div>