A Book about Unlawful Animal Behavior Strictly Obeying the Laws of Behavior

Roach, M. (2021). Fuzz: When Nature breaks the law. Norton. 

This image has an empty alt attribute; its file name is rumbustious3.jpg

The trial began with great fanfare as the defendant, listed in the record as Bailly, was paraded before an expectant crowd and into the courtroom. Although a court-appointed official argued vigorously on her behalf, the testimony of nine witnesses sealed the verdict. The defendant was found guilty of having caused an infant’s death, and the judge declared that she was to be, “hanged and strangled on a gibbet of wood near and adjoinant to the gallows.” On the appointed day, an audience of her peers turned out to watch the sentence carried out.

This might be any other murder trial in 14th Century France except for one small detail: The defendant was a pig. Strange as that may sound, criminal prosecution of animals was not unheard of in Medieval Europe, with defendants including rats, roosters, snakes, and insects. In this case, presumably as a deterrent to other potential offenders, the audience for Bailly’s hanging was other pigs.

Operant Punishment is Lawful. Legal Punishment Not So Much.

This image has an empty alt attribute; its file name is pigtrial.jpg
Bailly the Pig takes the stand in a 14th Century trial. Public domain image from Wikimedia Commons.

Let’s skip over the dubious status of animal trials within the contemporary legal standard of culpability, and jump right to a bigger issue: Legal punishment and punishment-the-behavior-process are two very different things. Had Bailly the Pig experienced effective behavioral punishment immediately upon accosting the child (and not been executed of course) she might have lived a long and aggression-free life thereafter. Maybe (see below). But executing her put a stop to all learning. Switching to a penalty other than death likely wouldn’t have helped because judicial sentences occur too long after the behavior to function as punishment. As Skinner observed long ago in Science and Human Behavior, that’s one key problem (among many) with all legal punishment.

This is not some bleeding-heart screed against the entire concept of legal punishment, only an observation, mirroring Skinner’s, that legal punishment is exceptionally badly constructed. The legal system is founded on a partial truth, that unpleasant outcomes, experienced personally or vicariously following undesirable behavior, reduce the future probability of that behavior. The scientific perspective is of course more nuanced. Operant punishment effects are real, but prevail only under limited conditions: Consequences must be contingent, and must follow immediately and with high probability after the behavior. Because those conditions are virtually never met in legal punishment, legal proceedings against animals make only slightly less sense scientifically than legal proceedings against humans. For example, according to the National Institute of Justice, consistent with a behavior-principles perspective, there is no evidence that the death penalty deters crime, and non-capital legal punishment does not seem to reduce the odds of a repeat offense. Consistent with the principles of behavior, there is, however, fairly good evidence that the biggest deterrent of criminal behavior is not the threat of specific penalties but rather the probability of being caught.

For the record, here’s an operant-inspired take on punishment in the legal system by a legal scholar, and another one by two behavior analysts. You’ll see that they’re not that different, both converging on the notion that just because some threatened penalty sounds pretty awful doesn’t automatically mean it will suppress behavior. So I’m retreading points here that are widely known and hardly new. The reason why is to set the stage for Mary Roach’s enjoyable Fuzz: When Nature Breaks the Law, which examines how humans respond when Nature impinges upon human safety and convenience. As the book’s dust jacket proclaims:

What’s to be done about a jaywalking moose? A bear caught breaking and entering? A murderous tree? [Centuries] ago, animals that broke the law would be assigned legal representation and put on trial. These days, as New York Times best-selling author Mary roach discovers, the answers are best found not in jurisprudence, but in science: the curious science of human-wildlife conflict, a discipline at the crossroads of human behavior and wildlife biology.

This is a book you can squeeze into a busy schedule. Each chapter, focusing on a different kind of unruly wildlife (rats, bears, elephants, and more) has a stand-alone quality, so the book can be consumed in installments without losing coherence. Roach’s writing is clear, informative, and often fun.

But woven through the book, perhaps not fully intended, are many reminders of how fraught that “crossroads of human behavior and wildlife biology” can be. And the reason is to a large extent because humans concerned with animal misbehavior have recapitulated the approach taken by humans who to try to eradicate human misbehavior. When viewed through this lens, Fuzz is an object lesson about our species: Specifically, as Skinner taught us, that without expert guidance from a science of behavior, we can be really bad at behavior control.

As a result, a surprising — or maybe unsurprising — number of human-concocted strategies to control wildlife fail. I’ll let you dig into the book for yourself to learn about strategies ranging from dynamite to laser light shows, but I will briefly discuss my favorite, known as bear hazing.

A Bear-ly Coherent Strategy

This image has an empty alt attribute; its file name is cafeteria-scaled.jpeg
Cafeteria’s open.

The trouble with bears is that they are smart and they like to eat as much as possible while working as little as possible for the privelege. Bears thereby are drawn to human enclaves where easy meal pickins tend to be plentiful. After all, it’s a lot less work to pop open a cooler full of camping grub than to dig up roots or bring down a surly moose. And neither of those comes with a side of Twinkies.

The presence of big, scary bears unnerves humans because, should a bear and a human ever clash, things will go poorly for the human.  As noted in the U.S. Fish and Wildlife Service’s Grizzly Bear Hazing Guidelines,

Grizzly bears can pose a threat to human safety and should be discouraged from using areas near homes and other human-occupied areas. Hazing discourages undesirable behavior in wildlife, and when properly conducted, does not create a likelihood of injury to grizzly bears to such an extent as to significantly disrupt normal behavioral patterns.

Hazing involves creating circumstances that bears find aversive, including with non-lethal projectiles and water cannons (see the Postscript for a complete list of government-approved hazing strategies). The underlying theory, of course, is that if approaching human-controlled areas leads to these consequences, bears will choose not to approach.

The reality is that hazing may drive bears away temporarily, but they can come back. One study found recidivism in nearly half of food-stealing bears who experienced hazing. The reason? Because, as every student of behavior analysis knows, punishment, even if applied exactly according to the behavior-science script, does not eliminate reinforcement for the same behavior. Once bears tap into those tasty human food sources, it’s hard to discourage them from returning.

The same study did find that hazing is quite effective if applied before errant bears can gobble down human treats. Unfortunately, as in all cases of unwanted behavior, would-be human behavior-control agents are neither omnipresent nor omniscient. A lot of ursine munchie runs can go unnoticed before someone gets around to hazing. And the instant bear tongue meets human food, a punishment-based strategy faces an uphill climb.

Perhaps recognizing that hazing has its limitations, wildlife experts sometimes employ an alternative strategy for making bears go away: They take them away by anesthetizing them and transporting them to a more remote area. This is getting harder to accomplish as humans occupy more and more of the land that was once bear stomping grounds, but even when translocation is possible it’s often not effective. Roach reports that bears have ambled their way back to their point of capture from up to 142 miles away. One bear swam 6 miles across the ocean in order to return. Oh, and by the way, bears that don’t return to the original scene of the crime often manufacture new crime scenes by simply finding new human settlements to raid.

This image has an empty alt attribute; its file name is multiplechoice.jpg
Formative assessment: When bears and humans come into conflict, what is the most likely outcome?

That’s enough to give you the gist of Fuzz in my reading: Human-wildlife cohabitation can be uneasy… the stakes for resolving this problem can be high… and solutions that are naive about behavior principles are doomed to fail. Behavior analysts might therefore have a lot to offer toward solving a problem that’s increasing in prevalence every day as humans encroach on wildlife habitats. However, Fuzz makes clear that there are situational and species-specific complexities that no solution can ignore, and this is not stuff about which we learn about in our behavior analysis graduate programs. As the dust jacket of Fuzz noted, this as a true interdisciplinary problem, and we behavior analysts are unlikely to generate effective solutions without first gaining some interdisciplinary expertise. Fuzz is at least an engaging introduction to that challenge.


(1) Carl Cheney was a pioneer in using behavior principles to understand the behavior of wild and grazing animals. Here are a few of his papers:

(2) “Acceptable bear hazing strategies.” Reproduced from the U.S. Fish & Wildlife Service’s (2020) Grizzly Bear Hazing Guidelines.

  • Non-Projectile Auditory Deterrents: Yelling, clapping, banging pots or other objects, air horns, vehicle horns, vehicle sirens, and P.A. systems (vehicle-based or hand-held) are often sufficient to move bears short distances, and should be employed before other methods are tried.
  • Visual Stimuli/Deterrents: Sometimes simple visual stimuli, such as spotlights or flashing lights can frighten a bear away. Use in combination with yelling so the bear associates the stimuli with human presence.
  • Vehicle Threat Pressure: Vehicles can be an effective hazing tool. By driving vehicles slowly towards bears (without hitting them), it can apply enough threat pressure to get bears to leave the immediate vicinity (200 yards). The effectiveness of vehicle pressure can be enhanced by using it in combination with auditory deterrents such as sirens or horns. Caution must be taken that no contact is made between the vehicle and the bear. As soon as the bear is clear of the immediate vicinity, pressure and noise should be discontinued.
  • Dogs: Dogs, such as Karelian Bear Dogs or trained guard dogs used on a leash can be an effective deterrence tool.
  • Water: The use of water in a large steady stream has proven effective at temporarily displacing some bears from human use areas. Hoses with a nozzle that gives a high pressure, long-range stream, such as a fire hose, can be an effective deterrence tool. As with all deterrence techniques, use of water for hazing should be accompanied by yelling so the bear associates the experience with human presence.
  • Stones or marbles: Can be either thrown or sent out of a slingshot. Wooden balls are also available for use with a sling shot. They should NOT be aimed at the face due to the danger of hitting an eye, but rather aimed at the rump of a bear. The maximum size of projectile should be the size of a golf ball. Range: 15-200+ ft. Paintballs: Paintballs can also be used as a deterrent, avoiding injuring to the bear (when shot away from the face). An alternative to paintballs are rubber balls. Rubber balls are extremely cost effective and can be used repeatedly. Range: 30-150 ft. Note: Bears are attracted to paintball residue, therefore the area must be cleaned up after the use of paintballs.
  • Noise-making Projectiles: This category includes projectiles fired from a weapon that explode, creating a sudden loud noise, such as bangers and 12-gauge crackers. Range: 75-90 ft. (bangers); 180-250 ft. (cracker shells). The explosive noise of cracker shell or banger must occur between the shooter and the bear.