Last night, I headed down to the beautiful Bell House, in the Gowanus section of Brooklyn, for this month's Secret Science Club, featuring Dr Alexandra Horowitz, director of Barnard College's Dog Cognition Lab. Dr Horowitz' new book is Being a Dog: Following the Dog Into a World of Smell.
Dr Horowitz began her lecture by noting that the study of dog (n.b.: in this recap, dog refers to Canis lupus familiaris) cognition is a young field. Much of the study of animal cognition involves looking for play behavior, as play is useful for developing cognitive abilities, and the preferred animal subjects tended to be primates. Dr Horowitz recounted taking her dog to the park while casting about for subjects to study, when she had the inspiration to study dogs. It took her two-and-a-half years to convince her advisor that dogs were proper subjects- dogs are ubiquitous, and they were perceived as being 'changed' by humans. The dog cognition expert Dr. Ádám Miklósi noted that the natural place for the existence of dogs is an anthropogenic environment created by humans. The dog 'wilderness' is among humans.
To some extent, dogs have a sense of social cognition- dogs look at humans for cues, they follow the human gaze to a point of interest. Dr Horowitz cited a proverb attributed to Confucius: "When a wise man points at the moon, a fool looks at his finger." While a dog can locate food based on cues given to them by a human, a chimpanzee would be unable to do so. Dr Horowitz expressed this in a particularly beautiful manner: "Dogs can see the mind behind our eyes, they pay attention to our attention." In one experiment, dogs were presented with food but instructed not to get it, and the time it took before the dog disobeyed was measured- they tended to disobey most quickly when the testers turned their backs, taking more time when testers closed their eyes. Dogs have some perception of language- in one case, a border collie was trained to distinguish labels for over one thousand toys. Although dogs were known to respond to linguistic cues, their potential cognitive capacity was largely unknown. People weren't studying dogs- the prevailing attitude could be summed up as "we know about dogs, they share our beds". Dr Horowitz posed the question, "Are we right?"
People tend to anthropomorphize dog behavior, with one particular doggy look being interpreted as a guilty look. In a 2009 study, Dr Horowitz endeavored to decipher this 'guilty' look. Dogs were presented with food, but instructed not to eat it. Obedient dogs didn't eat the food, disobedient dogs ate the food. The dogs owners were instructed to greet the obedient dogs and to scold the disobedient dogs. In some cases, the owner was given wrong information, and the disobedient dogs were greeted while the obedient dogs were scolded. The guilty look was not determined by 'guilt', but by the owner's reaction- dogs which were scolded assumed the guilty look, while greeted dogs did not. The guilty look was most prevalent among obedient dogs which were scolded. The 'guilty' look (eyes averted, paw up, even belly up) is a submissive posture, behavior designed to avoid punishment.
In 2012, Dr Horowitz set up an experiment to determine if dogs had a sense of fairness. Can dogs be jealous? In this experiment, trainers could be fair or unfair- they could dole out equal rewards or unequal rewards (fair trainers tended to give one treat to each dog while unfair trainers gave one treat to the subject and three to the control) to the subject dogs and 'control' dogs. The subject dogs did not favor the 'fair' trainer over the unfair trainer, they affiliated with the unfair trainer because she had more food than the 'fair' trainer and could potentially give out a bigger reward- Dr Horowitz quipped that dogs are optimists.
Dr Horowitz referenced Philosopher Thomas Nagle's paper What Is it Like to Be a Bat?, which posits that humans will never know what a bat experiences. Dr Horowitz joked that this is the philosophical approach, which is why she went into science... such things are knowable.
Dogs live in an olfactory world, and any study of dog cognition must take into account the olfactory experience. Humans have about five million olfactory cells while dogs have two-hundred million to one billion olfactory cells. These cells are packed into the olfactory epithelium at the back of the nose. Doctor Horowitz characterized the dog nose as a two-part structure... The 'first nose' tends to be at the end of a long snout. It has a wet, 'pebbled' surface and nostrils which have a unique musculature and act independently- dogs have stereo olfaction. Dog nostrils have slits on the side which allow exhalation without disturbing scent-causing molecules in front of the dog. Dogs sniff 'better' than humans do, their nostrils provide an 'olfactory route' for air as well as a 'breathing route'. Humans only have the 'breathing route'. Dogs sniff five to times per second while humans sniff once ever one-and-a-half seconds- dogs get more 'odor pictures' than humans do. With the exhalation of air out of the side nostril slits, more molecules are stirred up with the expired air, ready to be inhaled. The long snouts of most dogs are lined with convoluted turbinate bones lined with mucous membranes which humidify inhaled air and remove irritants. Humans have three turbinate bones. The 'second nose' of a dog is a vomeronasal organ, or Jacobson's organ. The vomeronasal organ is vestigial in humans, but it allows dogs to detect pheromones, which are water-soluble, non-volatile, particles with low molecular weight. Dogs can detect the hormones which signal that a female is ready to mate, and the hormone which signals aggression, cortisol.
The somatosensory cortex processes the input from the various sensory organs in an organism. Different animals have different levels of commitment to different body parts. Dr Horowitz used the example of the star-nosed mole, which has a set of fleshy appendages on its snout which are used as tactile sense organs. She characterized the star-nosed mole's sensory cortex as being 'overcommited' to these appendages. The human sensory cortex is traditionally illustrated by a figure known as a sensory homonculous, which depicts human body parts with sizes relative to sensitivity- the fingertips, lips, and genitals are the most sensitive parts of the human body. An analogous depiction of a dog would feature an extra large nose, representing the 'overcommitment' to olfaction.
Tracking dogs can detect different concentrations of odor-producing molecules between the first and fifth footprints of their quarry, typically representing a two second difference between steps- thus they are able to determine direction of movement through scent. Trained explosive-sniffing dogs can detect a picogram of TNT. Dr Horowitz displayed a photograph depicting the path a dog takes to track a dragged pheasant to illustrate the dog's use of scent to determine directionality. Dogs can be trained to sniff out a myriad of items- explosives, drugs, insect pests... in one particular case, a dog in the Seattle metro area has been trained to sniff out orca feces. Dogs being trained to sniff out victims of building collapses learn their tasks in centers in which subjects hide in large pipes so the dogs can locate them. The ability of dogs to sniff out cancer is currently being studied.
Studying the cognitive abilities of dogs necessitates looking at the experience of a dog qua dog, it shouldn't conform to our human, primarily visual, senses. In the case of quantity discrimination, though, dogs don't distinguish quantity by smell alone- cues from a dog's owner can influence a dog's choices, though by visual means, a dog can determine that three hotdogs are preferable to one. Owner enthusiasm can influence a dog to choose a smaller quantity- though dogs tend to be attracted to a larger plate, they act on what their owners present them. Being smell-oriented creatures, environments which are too clean can 'turn off' dogs olfactorily- dogs in an odor-deficient environment will follow their owner's cues more than their own noses. Recently, an organization called K9NoseWork has provided olfactory play-spaces for dogs.
Dr Horowitz then brought up the subject of mirror self-recognition among dogs. Chimpanzees, bottlenose dolphins, and some Asian elephants can come to recognize themselves in a mirror- at first, they interact with a mirror as if it were another individual, eventually they will learn to use the mirror for self-examination. Dogs don't pass the mirror self-recognition test, do they lack a sense of self, or do they not 'care' about marks on their bodies? In order to test self-recognition among dogs, an olfactory component needs to be introduced. The right test needs to be implemented- do dogs recognize themselves with odor as they recognize others? In one experiment, urine was collected from several dogs, some urine was altered olfactorily. The experimenters then determined how long a particular dog would smell various samples- the urine of other dogs was typically smelled for a longer period of time. Dr Horowitz noted that, when one walks a dog, the dog doesn't typically turn back to sniff its own pee. While not exactly parallel to the mirror self-recognition test, dogs did tend to recognize their own urine. Dogs engage in chemical communication- they sniff each others' rears where anal glands and tail glands are located. Many dog behaviors can be interpreted olfactorily- dogs wag their tails to distribute scent from their glands and a dog drying itself off by shaking can disseminate a whole lot of odor-producing molecules.
In a particularly poetic turn of phrase, Dr Horowitz noted that dogs smell the passage of time- weaker odors are older odors, the future is on the breeze. Who will come is on the air. She then digressed, talking about her efforts to emulate her dogs- she sniffs what they sniff, she underwent olfactory training, studying under such scent senseis (scentseis!) as 'smell-mapping' artist Kate McLean and Secret Science Club alumn Dr Leslie Vosshall. She studied perfumery, wine odorants, animal tracking. She advised us, be more like a dog, master your sense of smell.
The lecture was followed by a Q&A session, which Dr Horowitz prefaced with a funny quip- a lot of questions in her Q&As are prefaced with 'my dog does this...' One dogless bastard in the audience asked about the funny, higher pitched 'dog voice' which was used even by serious scientists- has the role of auditory cues in dog cognition been studied? Dr Horowitz noted that the primary reason for speaking at a higher pitch is to signal to the dog that it should pay attention. Another question involved the olfactory abilities of brachycephalic dogs, and Dr Horowitz joked that there are very few bomb-sniffing pugs. Why do dogs roll in dead things? It's possibly to camouflage one's scent, it's also an attention getter. Perhaps the best question was, does anything smell bad to a dog? Dr Horowitz noted that humans have a binary attitude towards odors- there are good ones, there are bad ones. To a dog, though, all odors are information.
Dr Horowitz' lecture was entertaining and informative, and her love of her subjects was expressed throughout the talk. Once again, the Secret Science Club dished out a fantastic lecture- high fives to Dr Horowitz, Dorian and Margaret, and the staff of the beautiful Bell House.
Here's a nice, informative cartoon based on Doctor Horowitz' research:
Dr Horowitz showed this cartoon, teaching viewers the art of smelling:
Here's a longer video on the subject of dog cognition with the good doctor:
Crack open a beverage, if you're lucky enough to have a dog, give it a good ear scritch, and soak in some SCIENCE.