1: Introduction to dog behavior

Introduction to dog behavior

Julie Hecht1 and Alexandra Horowitz2

1 Department of Psychology, The Graduate Center, City University of New York, Horowitz Dog Cognition Lab, Barnard College, New York, USA

2 Department of Psychology, Barnard College, New York, USA

Domestic dog evolution and behavior

Dog evolutionary history

What is a dog? The answer can come in the form of a description of the dog’s characteristic behavior, physical description, or evolutionary history. We will begin with the latter. The domestic dog, Canis familiaris, is a member of the Canidae family, genus Canis, along with such territorial social carnivores as the gray wolf (Canis lupus), the coyote (Canis latrans), and the jackal (e.g., Canis aureus and Canis mesomelas). The dog is the only domesticated species of the genus: that is to say, the only canid for whom artificial selection (selective breeding) by humans has usurped natural selection as a prime mover of the species.

A debate rages about how long ago, and where, a distinct species of dog appeared, given conflicting evidence from archeological sites and genetic analyses. There is much more agreement on one point: that dogs descended from wolves. Canis lupus, the present-day gray wolf, is the domestic dogs’ closest living ancestor, as both species are descended from some proto-wolf some tens of thousands of years ago. Archeological evidence suggests that the divergence between wolf and dog began up to 50,000 years ago, with the advent of early human agricultural societies (Clutton-Brock 1999). Whether the divergence was a singular, one-time event or whether it happened at different times and multiple locations is still in debate (e.g., Boyko et al. 2009; Larson et al. 2012; Thalmann et al. 2013). Genetic evidence, from mitochondrial DNA, suggests that wolves and dogs began diverging much earlier, even 145,000 years ago (Vilà et al. 1997).

Dogs’ domestication probably began with a human interest in animals who were relatively docile, perhaps willing to approach—or at least not flee from or attack—humans. The social nature of canids contributes to their interest in others, as well as the proto-dogs’ flexibility in seeing humans as nonthreatening. This hypothesis was famously tested by the geneticist Dmitry Belyaev by creating a kind of “domesticated” fox out of a Siberian farm-fox population simply by selectively breeding only those who reacted without fear or aggression to human approach. Over 40 generations, he had created foxes which looked and acted in many ways like familiar domestic dogs (Belyaev 1979; Trut 1999).

For millennia, dogs were bred for use for tasks (e.g. guarding and hunting) or as companions. Quite recently, in the 19th century, artificial selection began to be driven by an interest in creating pure breed lines, for show and competition in dog “fancies,” dog shows. Thus, the diverse array of breeds seen today is a result of specific breeding over the last century and a half for physical traits and temperament which suited the newly formed breed “standards” (Garber 1996). While some current dog breeds resemble ancient representations of dogs in art, no breed can be traced to those ancient dogs. As we will discuss, the diversification into breeds, some with exaggerated physical features, has led to the rise of inherited diseases which can be painful or even fatal (Asher et al. 2009). Isolated populations of purebred dogs now serve as useful models for naturally occurring cancers and diseases found in both humans and dogs (Breen & Modiano 2008).

Dog behavior in an evolutionary context

The story of domestication is informative because it gives the observer of dog behavior the background with which to interpret what she sees. That is, the dog is by no means a wolf but will share some behaviors with present-day wolves. Present-day dogs are highly designed by humans, have many behavioral and physical traits as a direct consequence of this design, and the affiliation between dogs and people is long-standing. Dogs are veritably members of human society and families (Horowitz 2009c).

Knowledge of the behavior of dogs’ wild cousins, gray wolves, helps give clearer explanation for many common dog behaviors. For instance, viewed in the context of a human family home, a dog’s propensity to sniff at the genital area of visitors to the home may seem odd, intrusive, or even “impolite.” Viewed in the context of canid social interaction, though, it is clear that the dog’s sniffing is analogous to all canids’ olfactory investigation of the genital and anal areas of conspecifics (Sommerville & Broom 1998). These regions are rich with glandular secretions which carry information about the identification, and perhaps recent activities and health, of the individual. The dog in the human household is simply trying to find out about this human visitor (Filiatre et al. 1991).

Another dog behavior, the dog’s licking of an owner’s face upon the owner returning home, is commonly viewed as an expression of love. Indeed, many owners refer to this behavior as dog “kisses.” Looking at wolf behavior again clarifies the interpretation. Wolves, living in family packs, approach and greet any wolves who are returning to the pack after hunting. The packmates lick—“kiss”—his or her face. Their licks are prompts for him to regurgitate some of the kill that he has just ingested. Similarly, a dog’s “kiss” is a greeting, to be sure, but it is also a vestigial interest in whatever it was an owner might have consumed since leaving the house (Horowitz 2009c).

On the other hand, dogs’ artificial selection history is explanatory of important differences in the behavior of wolves and dogs. Foremost among them is the dog’s ability to (and desire to) look at the eyes of humans for information or to solve a problem. Since mutual gaze is a vital part of human communication, dog behavior which seemed to match this human behavior may have been preferred and selected (Horowitz & Bekoff 2007). Indeed, the modern dog’s eyes are more rounded and forward-facing than those of wolves (Clutton-Brock 1999), and their faces have many neotenous (baby-like) features which human adults are predisposed to find appealing and human-like (Hecht & Horowitz 2013). The dog’s eye-gaze enables much of the species’ success at tasks of social cognition, such as following a human’s gaze or pointing arm or hand to a source of food or interest (e.g., Agnetta et al. 2000; Soproni et al. 2001), something characteristic of human–human interaction but quite unusual in nonhuman animal populations, in which to stare at another’s eyes is a threat (Fox 1971).

An understanding of the development of different dog breeds, and each’s use and habitual behaviors, is also explanatory in looking at the “average” dog’s behavior. In early domestication, breeding would have been somewhat haphazard, but by the time of the Romans, there were physically distinct breeds bred for particular functions: as guard dogs, sheep dogs, and companion (lap) dogs (Clutton-Brock 1995). The kinds of breeds and the uses for breeds multiplied in the Middle Ages and through the present day extending to employing dogs as both herders and as guarders of livestock; as hunting dogs—tracking, pointing at, or retrieving game; as load-carriers (e.g., sled dogs); as assistance dogs (in guiding blind persons or aiding those with other physical disabilities); and as therapeutic companions. In some cases, successful job performance may require extensive breeding (sled dogs) or training (glycaemia alert dogs) (Huson et al. 2010; Rooney et al. 2013).

When selective breeding for physical traits and behavioral tendencies of specific, named purebreds began in earnest, in the late 19th century, modifications occurred which, while useful in carrying out the desired task, may be undesired in nonworking contexts. Moreover, given the degree of inbreeding, these behaviors are often intractable and tenacious (as described further in section “Breeds and behavior”). Even in mixed breeds, some degree of these behavioral tendencies may endure.

Dog interspecific social cognition

Among social species, dogs are unique: They have the potential to interact as smoothly with a separate species as with their own. Canis familiaris and Homo sapiens engage together in everything from the seemingly mundane—sitting side-by-side on a park bench—to the complex—running an agility course, working together to detect explosives or locate animal scat, or alerting a deaf person to a ringing telephone. Even village dogs, who often retreat when approached by humans, live in the vicinity of people (Ortolani et al. 2009).

Companion dogs are often described by owners as having clear constructed identities, particularly that they are “minded, creative, empathetic, and responsive” (Sanders 1993). Relationships with dogs run so deep that they are sometimes mentioned in obituaries along with other survivors of the departed (Wilson et al. 2013)—suggesting that for many, dogs are placed within the familial structure (Hart 1995).

Magic is not behind humans’ feelings of connectedness toward dogs. Instead, companion dogs display social behaviors that support and reinforce the relationship, such as sensitivity to human actions and attentional states, and acting in accordance with humans in coordinated and synchronized ways. For example, dogs unable to access a desired item will alternate their gaze between the item and a nearby person (i.e., the behavior dogs perform when a ball rolls under the couch and you ultimately get it for them) (Miklósi et al. 2000). Dogs readily respond to human communicative gestures, whether stemming from our hands, face (e.g., eyes), or other body parts (Reid 2009). Dogs take note of our attentional states, particularly eye contact as well as head and body orientation—a dog being more likely to remove a muffin from a countertop if an owner’s back is turned or eyes are closed than if the owner is sitting in a chair with eyes fixed on the dog (Schwab & Huber 2006). Dogs also attend to the tone of human voice and behave appropriately (according to humans) when spoken to in a cooperative or a forbidding tone (Pettersson et al. 2011).

While training can enhance a dog’s ability to perform in social interactions (e.g., guiding-eye dogs and detection dogs), there are everyday examples of dogs showing complex, synchronized social exchanges with people. Kerepesi et al. (2005) found that companion dogs—not specifically trained—were able to engage in a cooperative interaction with their human partners that allowed for the completion of a joint task. In this study, people asked their dog for blocks to help them build a tower, and dogs provided the blocks in a nonrandom fashion that indicated cooperation. Similarly, companion dogs show a great deal of social anticipation, which can enhance synchronization and feelings of mutual cooperation. Dogs even adopt new routines established by people, such as a short, pointless detour made by owners upon returning home after a walk (Kubinyi et al. 2003). Over time, dogs in this study even began to perform the pointless detour before their owner. Social coordination is also found in play, a common inter- and intraspecific activity. Play is essentially marked by coordinated movements and synchronized interactions. Dogs and humans attend to each other’s play signals, and a dog’s play bow—or a person’s play lunge—is responded to meaningfully (Rooney et al. 2001).

While popular media often spotlight breed differences relating to social behavior, trainability, or “intelligence” (Coren 2006), research is mixed as to how artificial selection affects companion dog performance in human-guided tasks. In one study, dogs bred for cooperative interactions outperformed those bred for independent work on a human-guided task to locate hidden food (Gácsi et al. 2009). At the same time, there can be substantial differences between dog lines still selected and maintained for the original function and members of the breed not under continued election for performance (i.e., the difference between show dogs versus field dogs). In another study, subject dogs’ ability to follow a human-demonstrated detour was independent of breed (Pongrácz et al. 2005). Udell et al. (2014) found that breed-specific predatory motor patterns predicted dog success in following human pointing gestures, with Border Collies and Terriers outperforming Anatolian Shepherds, a breed selected for behavioral inhibition. At the same time, Anatolian Shepherds significantly improved their performance with little training. On that score, Border Collies Betsy, Rico, and Chaser have been empirically shown to possess extraordinary facility with human language, but so too have Bailey (a Yorkshire Terrier) and Sofia (a mixed breed) (Hecht 2012).

Dog interspecific attachment

Another meaningful mechanism underlying the dog–human relationship is that of attachment, a concept initially introduced to describe the affectionate bond between a human infant and a caregiver (Bowlby 1958). Initial examination of attachment relied on the “Strange Situation Test” (SST), a behavioral experiment in a novel environment designed to investigate specific behaviors from the infant toward the mother as opposed to a stranger (Ainsworth & Bell 1970). Attachment is evidenced through infant “behavioral preferences” for a figure of attachment (e.g., mother), such as proximity maintenance, distress upon separation, as well as comfort and increased exploration in her presence.

Ethological studies suggest that attachments form in many species, not just humans. A modified version of the SST was conducted between dogs and their owners (Topál et al. 1998). Like infants, dogs showed activation of attachment systems when in the presence of a stranger versus their owner, as well as the “secure base effect” where dogs were more likely to explore their environment in the presence of the owner than a stranger (Horn et al. 2013).

Subsequent studies found that for dogs, attachments can form later in life and even multiple times. Shelter dogs participated in the modified SST with someone assigned the role of “stranger” and another person assigned the role of “owner” (designated by three short interactions with the dog). Shelter dogs showed similar attachment behavior toward the newly appointed “owner” (Gácsi et al. 2001). Service dogs, like guide dogs for the blind, experience numerous early-life relationships and show attachment behavior toward their subsequent blind owner, who they met later in life (Fallani et al. 2006; Valsecchi et al. 2010).

These studies appear to be in tension with the initial assumption that for human-directed attachments to develop, dogs should be brought into the new owner’s home at 8 weeks of age (Scott & Fuller 1965). Instead, while it is recognized that early-life exposure to humans is important for normal social development, dog attachment relationships can form later in life, multiple times, and toward multiple people.

Physiological mediators also underlie dog–human relationships. The peptide hormone oxytocin (OT) is involved in affectionate bonds and may help to mediate dog–human social behavior. For example, Kis et al. (2014) found an association between OT polymorphisms and human-directed social behavior in German Shepherds and Border Collies. Owners and dogs who engage in petting and light play both show OT increases (Odendaal and Meintjes 2003). While simply seeing a known person can raise dog OT levels, it is often the quality of the interaction that matters. Rehn et al. (2014) found that a familiar person engaging in “physical and verbal contact in a calm and friendly way” when greeting a dog was associated with a sustained increase in dog OT levels. In another study, owners who engaged in longer periods of gaze with their dog and reported a higher degree of satisfaction with their dog had increased OT levels over owners who did not report similar satisfaction and did not display high levels of gaze (Nagasawa et al. 2009). (Importantly, while owner OT levels increased, dog hormone levels were not examined, and it is plausible that what is enjoyable for people is not always the same for dogs, such as prolonged or persistent direct eye contact.) At the same time, Jakovcevic et al. (2012) found that dogs characterized as highly sociable gazed longer at an experimenter’s face, even when the behavior (gaze) was no longer being reinforced.

Dog relationships with conspecifics and other nonhuman species appear to differ from the relationships dogs form with humans. Behavior toward the dam and members of a litter are not customarily described as attachment relationships (Pettijohn et al. 1977). A study of older dogs living in the same house did not find behavioral indicators of an attachment bond between cohabitating dogs, although activation of the stress response was reduced when in the presence of the companion dog (Mariti et al. 2014). On the other hand, in a novel setting, shelter dogs showed diminished stress response, not in the presence of known kennelmates, but in the presence of a known person (Tuber et al. 1996). At the same time, when a companion dog dies, some owners report behavioral change on the part of the remaining dogs, such as change in appetite, sleeping, solicitation of affection, and use of space (Schultz et al. 1995; Walker et al. 2013).

Taken together, dogs have complex and long-standing relationships with members of their own and other species. They have preferred play partners (Ward et al. 2008) and engage in mutual resting and grooming with members of their own and other species—for the latter, particularly if the non-dog species was introduced early in the dog’s life (Fox 1969; Feuerstein & Terkel 2008). Dogs can have meaningful and successful lives within the human environment, and their potential for success starts from the very beginning of life.

Dog development and behavior (early and late life)

Unlike precocial species (e.g., zebras, sheep, and some birds), born capable of moving around and caring for themselves soon after birth, altricial species (e.g., canids and humans) require substantial dependent care while they pass through a number of developmental stages in their first months of life. This time is marked by physiological maturation and the growth of sensory abilities that facilitate structured motor patterns and, ultimately, the presentation of adult dog behavior. During this time of intense physiological and sensory development, dogs are most malleable. They are essentially sponges, taking in information and readily updating and changing their behavior.

While the natural ecological niche for dogs is the human environment (Miklósi 2007), within this general environment, dogs are exposed to a wide diversity of anthropogenic settings. For example, there are an estimated one billion dogs on the planet, and the majority live as stray or village dogs (Lord et al. 2013): They live on the streets, scavenge from human refuse sites, and move and interact with conspecifics and other species on their own accords. In other parts of the world, dogs have entirely different surroundings and different roles to play. Dogs live in over one-third of US homes (AVMA 2012); many sleep in a bed with a person at night and are expected to stay home, possibly alone, during workdays (Horowitz 2014). Companion dogs are often expected to be leashed, urinate, and defecate in specified locations and interact (in a “civilized” manner) with a changing array of conspecifics and people. As mentioned, dogs can also perform a wide variety of working functions, and some dogs serve as subjects in medical labs. What is expected of dogs varies considerably based on the specific human environment in which the dog finds himself. Early-life experiences are instrumental to successful environmental integration.

In these early months, young puppies need considerable social support and stimulation—both from conspecifics and from humans—in preparation for the expectations that will be applied to them. The support and environmental inputs that puppies do or do not receive affects their developing personality and later behavior. A 20-year study at the Jackson Laboratory in Bar Harbor, Maine, set out to explore the behavioral and genetic underpinnings of behavior. The researchers found that “critical” or “sensitive” periods of development—specific weeks or months in which dogs develop particular abilities—along with early-life environmental inputs, were instrumental to normal development (Scott & Fuller 1965).

While developmental periods have a clear progression (a dog will not play bow before it has opened its eyes), transitions between each stage are more gradual than initially thought (Bateson 1979). The following periods are instead guidelines—without hard-and-fast beginning and end points—and individual dogs will move quicker or slower from one phase to the next. Rates of development (heterochrony) can differ between breeds as well as between individuals.

Neonatal period: birth to approximately week 2

Dogs enter the world unable to survive on their own. Direct contact with the mother, the dam—who provides food and initiates elimination by tactile stimulation—allows pups to proceed with physical and neurological development. Neonatal pups are without vision, hearing, or coordination and rely on tactile and simple olfactory sensations (Scott & Fuller 1965; Lord 2013). Unable to self-regulate temperature, newborns spend most of their time sleeping and in physical proximity with the dam and littermates. Although most elements of their sensorium are underdeveloped, neonatal pups appear responsive to olfactory cues. Wells and Hepper (2006) found that neonatal pups (tested at 15 min and 24 h after birth) preferred water with the flavor aniseed when the dam had consumed aniseed during the pregnancy. Puppies did not show similar preference for vanilla, a different novel scent that the dam had not been exposed to—suggesting that gestational exposure (which has also been found in other mammals) is behind this neonatal preference.

While the majority of the neonatal period is spent prostrate (in a flat, pancake-like pose), newborn pups show behaviors associated with attaining food: “kneading” or “swimming” behavior directed at the teat or milk source. They also show discomfort: If isolated, pups display distress vocalizations, high-pitched calls—whines or yelps—that are frequently described as care-soliciting behavior (Elliot & Scott 1961). These early vocalizations later transform into other vocalizations that are contextually similar. For example, adult dogs produce high-pitched, high-frequency “alone barks” that may also elicit attention (Yin & McCowan 2004; Pongrácz et al. 2006).

Transitional period: week 2 to week 3

The maturation process of the first few weeks of life becomes more evident at 14–21 days, when puppies spend less time in a flat, pancake state and more time moving toward presenting typical dog-like behavior. Pup eyes and ears open, allowing for a startle response (Scott 1958). Motor patterns and social behaviors like walking and tail wagging begin, as do rudimentary elements of play. Because of dog’s increased sensorium, now is the time to start introducing novel items, and “exposing puppies to normal household sounds, smells, and sights; daily handling; petting; and gentle brushing” (Case 2005).

Sensitive or Socialization period: week 3 to weeks 12–14

This is a period of considerable growth (particularly of species-specific social behaviors) and many experiential and learning opportunities. Socialization is described as the process of adopting “behavior patterns appropriate to the social environment in which [an individual will] live, allowing them to coexist/interact with other individuals” (Blackwell 2010). Attention to a dog’s individual experiences during this period, particularly a dog destined for companionship, is essential.

Motor patterns develop and adult-like behaviors are expressed in a more coordinated manner. Social behaviors like approach and avoidance emerge, as do tail wagging, growling, and additional play behaviors (Bekoff 1974). Vocalizations become more complex and are incorporated into social situations. Adler and Adler (1977) suggest that as soon as puppies have the physical capacity to recognize conspecifics, social learning is possible. Puppies who watched their mother perform in narcotics detection during this developmental period were more likely to work in narcotics detection themselves (Slabbert & Rasa 1997). Pups also show attention to and interest in humans which includes affiliative, social behaviors like approach and tail wagging. Dog propensity to follow human gaze or pointing cues increases as dogs age (Riedel et al. 2008; Dorey et al. 2010).

Dogs are weaned in the first part of this period, between approximately weeks 4 and 8, though there are considerable individual differences in weaning behavior even within breed (Rheingold 1963). A study of the weaning of German Shepherd puppies and their dams found that when puppies attempted to nurse, dams responded with “inhibited bites” or growls, mouthed threats, nibbles, and licks (Trivers 1974). In response, pups showed social behaviors, such as withdrawal and passive submission (Schenkel 1967). Dams also began to show “inhibited bites” toward puppies during play. Such social experiences are important for later social exchanges, see Appendix A.7.

This period is commonly referred to as a “sensitive” social period because pups can notice and interact with other species and novelty without hesitation—particularly before 5 weeks of age. Dogs show considerable exploratory behavior and approach novelty without hesitancy between 3 and approximately 5 weeks. As they grow, they can show hesitation to novel stimuli, and at about 8–10 weeks, this change magnifies, and some puppies display decreased comfort with new stimuli, like people, sounds, objects, and contexts (Case 2005). This presentation of fear could be modulated by both genetics and early-life experiences (Freedman et al. 1961; Uhde et al. 1992), and caution should be taken against exposure to noxious stimuli and situations, particularly during weeks 8–10.

Socialization in dogs

Socialization from week 3 to about week 14 is paramount. The American Veterinary Society of Animal Behavior recently issued a Position Statement recommending puppies start socialization classes early as 7–8 weeks and with a minimum of one set of vaccines (AVSAB 2008). As in other social mammals, early-life restrictions—both environmental and experiential—hinder later-in-life behavior and coping strategies and are associated with fear and anxiety (Scott & Fuller 1965). For example, puppies exposed to premature maternal separation were found to show higher prevalence of “destruction of objects, excessive barking, fearfulness on walks, fear of noises, possessiveness of toys, attention seeking, aversion towards people of unusual appearance, play biting, tail chasing, pica, possessiveness of food, aggression towards unfamiliar people, and house soiling” than control dogs who remained with dams until 2 months of age, that is, through weaning (Pierantoni & Verga 2007).

Daily tactile contact is important, and there are benefits to starting even earlier than the third week. Daily gentle tactile stimulation and handling of puppies’ bodies between days 3 and 21 was associated with more exploratory behavior when alone, and such puppies were less quick to vocalize than puppies that were not handled (Gazzano et al. 2008). Daily engagement of the senses promoted dogs who were more active, sociable, and less neophobic than puppies not handled as such (Fox & Stelzner 1966).

Careful, early exposure to potentially noxious stimuli could help with later-in-life coping. Newborn rats handled and exposed to mild stressors showed less stress activation and more exploratory behavior than unhandled rats when exposed to novelty as adults (Núñez et al. 1996). Pluijmakers et al. (2010) found that exposing puppies to audiovisual playback—consisting of animate and inanimate objects and noises at normal volume—between 3 and 5 weeks of age was associated with decreased fear to novel objects and unfamiliar settings. Puppies without exposure to the audiovisual condition show increased crouching, increased arousal—as indicated by rapid tail wagging—and increased locomotion, all of which are associated with stress or fear (Beerda et al. 1997). This early-life exposure is aimed to combat the fear response that can develop after 5 weeks. Still, socialization should not be performed by throwing dogs off the deep-end and into overstimulating situations, such as street fairs or lengthy social gatherings. Small doses of successful and enjoyable experiences are key, and dog behavior should be continually monitored for low-level indications of discomfort and distress (see section “Patterns of communication”). Classical and operant techniques can be used to increase comfort during socialization.

Because of the importance of inter- and intraspecific interactions and exposure to stimuli and social experiences, shelters with puppies under their care should prioritize early-life socialization or find appropriate housing outside the shelter that can.

While restricted early-life environments can elicit profound behavioral changes in dogs, there is room for later-in-life behavioral flexibility. A recent study found that dogs who had lived in commercial breeding establishments, commonly referred to as “puppy mills” or “puppy farms,” were described by subsequent owners as displaying higher rates of “fear, house-soiling and compulsive staring” than a matched sample of dogs (McMillan et al. 2011). In 2013, the American Society for the Prevention of Cruelty to Animals (ASPCA) began a study investigating whether exposing fearful dogs to in-shelter counter-conditioning, habituation, and desensitization training plans could effectively mitigate dogs’ fear response before being placed into homes (ASPCA 2013). The ongoing success of the programs is a reminder that while experiences during early life are important to later-in-life behavior, dogs are malleable even beyond the sensitive period of socialization.

Aging dogs

The behavior and cognition of aging dogs is not typically considered part of the stages of dog behavioral development, but the realities of aging can be incredibly important to dog well-being. Just as young dogs undergo notable changes early in life, so do they experience changes later in life. Since adult and aged dogs are members of the shelter population (Shore & Girrens 2001), their unique position in life, as it relates to normal, successful aging versus cognitive dysfunction, merits consideration.

Considering age-related changes in dogs, researchers are attempting to discriminate the normal aging process from canine cognitive dysfunction. Some describe the behavioral changes resulting from normal aging as a “rate of cognitive deterioration that does not affect the day-to-day functioning of the individual” (Salvin et al. 2011). Owners of dogs 8 years and older describe certain trends associated with normal aging, such as deterioration of “play levels and response to commands” and increase in “fears and phobias.” Older dogs showed less enthusiasm “for eating and chewing” and an increase in water consumption, most likely as a function of age-related health factors like teeth and mouth diseases, as well as renal problems.

Cognitive dysfunction, on the other hand, is characterized by behavior changes relating to deterioration of cognitive functioning and recognition, and the acronym DISHA describes changes like “Disorientation, altered Interactions with people or other pets, Sleep–wake cycle alterations, House-soiling and altered Activity level” (Landsberg et al. 2003). These challenges can play out in increased destructive behavior, house soiling, and increased vocalizations, unrelated to earlier-life behavior (Chapman & Voith 1990). These changes, particularly relating to memory, have made dogs models for human aging and dementia (Cummings et al. 1996). As in humans, therapeutic products are being tested and developed to treat cognitive dysfunction in senior dogs, some with validated efficacy (Landsberg 2005).

Normal dog behavior

Listen to people talk about companion dogs, and you are apt to hear descriptors like “crazy” or “bonkers.” While anyone who has ever lived with a dog might commiserate and find these labels at times appropriate, the labels do not offer much insight into what the dog is actually doing. Is the dog heating up a frying pan and preparing brunch for the family? That would be “crazy.” When the doorbell rings, does the dog assume the role of Olympic runner and high-jumper, taking laps around the living room and finishing the routine by jumping on entering guests? This is less “crazy” and more normal dog behavior performed in a context not always appreciated by humans.

What is behavior?

Dutch ethologist Niko Tinbergen—cowinner of the 1973 Nobel Prize in Physiology or Medicine—proposed an integrated approach to the study of behavior, characterizing two kinds of questions that researchers may ask and attempt to answer. “Why Questions,” commonly described as questions relating to ultimate causes of behavior, explore evolutionary forces behind behavior; “How Questions,” or questions relating to proximate causes of behavior, focus on a behavior’s immediate prompts, in both mechanistic and developmental terms (Tinbergen 1963). This approach, accepted by most researchers as a sound guide, expects that an individual’s behavior is a product of an individual’s life experiences (proximate explanations) and evolutionary history (ultimate explanations).

Thus, dog behavior can be framed first in the context of their species-specific characteristics: a gregarious, social canid with behaviors that support both inter- and intraspecific communication, as well as a species affected by recent artificial selection on the part of humans. Additionally, proximate factors such as dog individual life experiences and individual development are relevant for dog behavior.

Dogs, like all species, come with a “normal” repertoire1 of things they do, that is, possible behaviors. To name a few, dogs have the potential to play growl, sniff, and run in circles, but they cannot fly or sleep underwater. Even when a dog witnesses a bird flying, he cannot learn to perform that behavior. Underlying the concept of “normal” behavior is the concept of “behavior” itself. Behavior does not have a universally accepted definition (Levitis et al. 2009), although Tinbergen offers that behavior is “the total movements made by the intact animal” (Tinbergen 1951). This definition does not ignore that physiological processes (neuronal firing, hormone secretion, etc.) underlie behavior, but it does highlight that behavior is observable and measurable, which makes the study of dog behavior within reach.

There are many ways to scientifically describe dog behavior (see Miklósi 2007 for review). Species-specific behavior can be split into different categories, often determined by the topic of interest (Altmann 1974; Martin & Bateson 2007). For example, a dog could be described as engaging in “locomotion” to describe any type of lateral or vertical movement, or movement could be described based on quality—such as walking, running, or trotting. Behaviors can be examined separately, a “yawn” or a “paw raise,” or pooled together to describe behavioral states such as “play” or “aggression.” Behavior can also be described by its sequence as well as frequency, duration, and intensity. Dogs mainly engage in visual, acoustic, and olfactory communication, and each plays an important role in inter- and intraspecific communication.

Visual communication

The initial step to visual communication is knowing which parts of the body convey meaningful information. For example, unlike peacocks in which eyespots or train length could affect mate choice (Petrie et al. 1991; Hale et al. 2009), a dog’s piebald facial coloration is apparently not an informative detail in dog–dog visual communication but is instead a by-product of domestication (Trut 1999). Instead, other body parts and visual signals are meaningful in canine communication (Figure 1.1).


Figure 1.1 Know where to look for clues.

Reproduced with permission of Natalya Zahn. © Natalya Zahn.

Behavior not morphology

The body parts that contribute to visual communication merit discussion because research finds that people do not readily look at actual dog behavior. Instead, dog physical appearance, not behavior, often captures people’s attention. Physical appearance has been associated with dog adoption rates (Weiss et al. 2012), and physical appearance has been shown to be responsible for personality attributions. One study found that an image of a yellow-coated dog was rated as more agreeable, conscientious, and possessing emotional stability than an image of the same dog with a black coat (Fratkin & Baker 2013). In another study, attributions to dogs differed based on who the dog was with. Place a pit bull-type dog with an elderly woman or child, and people offered more positive ratings of the dog than if the dog was with a “rough” male (Gunter 2013). Furthermore, Horowitz and Bekoff (2007) suggest that people are attracted to dogs that exhibit seemingly human-like characteristics, such as flexuous facial features like raising the eyebrows or appearing to smile, both of which have been supported by recent studies (Hecht & Horowitz 2013; Waller et al. 2013). Overall, people construct meaning out of the way dogs look, often to the neglect of the way dogs behave.


If there is a body part people do take note of, it is the tail (Tami & Gallagher 2009). Charles Darwin points out that it is hard for a human to ignore a tail held high or one that is tucked deep beneath (Darwin 1872). Tails are mobile and can assume a range of heights and positions or swing at different speeds, each providing different information. At the same time, recent research finds that tail use might be even more complex and nuanced than initially thought.

Tails hold important information, especially in dog–dog communication. Simply the absence of the tail can affect communication, as can docked tails. Researchers who designed a mechanical dog outfitted with tails of different lengths (long or short) which were able to move or be still found that dogs were more likely to approach the robot dog when the tail was long and wagging as opposed to when it was long and still (Leaver & Reimchen 2008). Absent any other communicative cues, a wagging tail in this context appears to be interpreted by dogs as “friendly.” On the other hand, a short tail, whether still or wagging, was approached similarly, suggesting that short tails might be harder for dogs to interpret.2

The direction of a tail wag is also an informative detail. Tails that wag more to the right or left side of the dog’s body are called “lateralized” and may be connected to the dog’s emotional state. Typically, movements on the left side of the body correspond to right-hemisphere brain activation, and movements on the right side of the body correspond to left-hemisphere brain activity. Generally speaking, these hemispheres of the brain are associated with different behavioral outputs—approach (behavior on the right-/left-hemisphere activation) or avoidance (behavior on the left-/right-hemisphere activation) (Rogers 2009). For example, chicks forage for food with their right eye (left-hemisphere activation, i.e., approach) and look for predators with their left eye (right-hemisphere activation, i.e., avoidance) (Rogers 2000).

Dogs presented with stimuli of positive valence, such as an owner, wag more to the right side, or left-hemisphere activation (i.e., approach), whereas an unknown dog prompts more left-bias wags, or right-hemisphere activation (i.e., avoidance) (Quaranta et al. 2007). While this research has been extended to suggest that dogs can even attend to the side of another dog’s tail wag (Siniscalchi et al. 2013), it remains unclear whether dogs in real-life settings are picking up on these subtleties.3

Dog tails vary in flexibility and expressiveness, and some are not easily seen, either because of breeding or other human interventions (Bennett & Perini 2003). Other tails have a normal position that is curled, tucked, or naturally falling to one side. Because of their variable physical appearance, tail movement is studied from the base, not the tip, and tail-wagging musculature moves the rump more than the tail.

The base of the tail, closest to the dog’s rump, gives details as to whether the tail is being carried along the midline or is raised or tucked. Relaxed tails are commonly held in a neutral position, extending from or dropped below the midline, although the “neutral position” will vary from dog to dog. Generally speaking, a high tail indicates excitement or arousal, and a high tail can be seen in a variety of approach-oriented behaviors, ranging from greeting and playing to fighting and threatening (Kiley-Worthington 1976). Tucked tails, on the other hand, indicate some degree of fear, submission, or appeasement. Tails can be held in a stiff, still position at all heights and could be the dog’s natural tail or a postural display. Stillness is common in dog interactions: For example, play incorporates many pauses interspersed within fluid movements and play signals. But a still tail without such indicators could suggest fear or aggression.

Probably, the most noticeable and heavily generalized part of the tail relates to movement. “A wagging tail indicates a happy dog,” it is often stated. If only it were so simple (or true). A tail wagging wholeheartedly, fluidly, and generously from side to side (usually at the level of the midline) is most readily associated with greeting or excitement. This is the “happy” tail we are so familiar with, and it might be accompanied by jumping, licking, running in circles, or other behaviors of arousal.

A tail wagging low and quickly indicates nervousness or timidity. Again, the tail is wagging, but its position and rate could indicate fear, submissiveness, or a dog in conflict—sometimes referred to as a “mixed motivational state.” Dogs who perform a low wag upon being approached, and then flip over to expose their underside, are displaying passive submission; for other dogs, a low wag and a low body posture are part of their normal greetings and are part of active submission (Schenkel 1967). A low wag should not be considered in isolation because its meaning takes shape only in the context of the dog’s entire body. Low wags should always be considered within the dog’s environmental context and behavior as a whole. High, fast wags indicate arousal, but they should also be viewed with some caution. Arousal can take different forms, such as general excitement, interest in interacting, or even aggression. There are many individual variations of tail wag—circling; going more counterclockwise than clockwise; banging—but whose meaning or significance has not been studied (and should not be assumed).

Overall, tail behavior should be considered in relation to the tail’s normal, relaxed position, which will differ from dog to dog. For dogs in a shelter, watching the tail and its postural changes over time can provide a better estimation of the “neutral” tail position for that dog. The nuances of dog tails are important to learn and convey to the general public.


Piloerection is a physical response akin to getting goose bumps. Hackles tend to raise (i.e., hair tends to stand up) in areas from the base of the tail to the shoulders and down the spine. While it can be a meaningful indicator that a dog is excited (either happily or in alarm), this behavior is not within the animal’s control (London 2012).

Research has not specifically investigated whether raised hackles is associated with different emotional states, although it is often associated with aggression or fear. The location of the raised hackles may be informative about an underlying emotional state: Some suggest that hackles raised near the base of the tail could be associated with “a high level of confidence” and a dog “more likely to go on the offensive” (London 2012). On the other hand, piloerection around the shoulder region may suggest that the dog is fearful, and hackles raised by both the shoulders and the base of the tail could indicate “an ambivalent emotional state and feeling conflicted” (London 2012).

Because raised hackles indicate arousal generally, the presence and location of piloerection should be considered in conjunction with ear, tail, and mouth position and overall body leaning and posture to assess the specifics of that aroused state.


Like the tail, ears are incredibly nuanced in natural presentation and carriage. Some are permanently pricked, while others droop to the side. Like tails, ear carriage is evidenced by looking at the base of the ear. Ears can flatten to varying degrees toward the head, and even in long-eared breeds like Basset Hounds, “ears back” can be noted by paying attention to the base. Ears pressed back are generally associated with greater levels of fear, submission, retreat, or even defensive aggression. Ears forward are the opposite, suggesting interest, attention, alert, and approach as opposed to withdrawal.


While the mouth and muzzle are not often described in behavioral studies, these body parts are explicitly attended to during shelter behavior assessments of dogs (see ASPCA SAFER Glossary). The position of the mouth holds valuable information about what a dog might do next. Open versus shut is the first consideration, and further qualitative elements provide more detail. An open, relaxed mouth indicates a comfortable dog, while a tight mouth could indicate discomfort, fear, or simply a neutral position. The corners of the mouth, or labial commissure, is also important. What is sometimes described as a “long lip,” where the commissure pulls back toward the ear, is often seen in fear, stress, or appeasement displays. In a submissive grin, the lips are retracted and the teeth are visible, but the eyes may be squinty and the forehead smooth. A “short lip” is pushed forward, forming a tight “c” shape of the mouth, as if a wind source behind the dog is pushing the facial features forward. This is part of an aggressive display, and the top of the muzzle is wrinkled, and the eyes are open and hard.


Dog tongues are known to hang generously out of mouths during play, but they can also serve as indicators of discomfort. A tongue extended and retracted quickly is a tongue flick: Like the raised hackles, it may be a reflexive response to discomfort. Dogs also use tongues socially, to investigate substrates and surfaces (urine on the street, you after a run), as well as in greeting where dogs are apt to lick the mouth of both dogs and people.


As previously mentioned, physical appearance can strongly relate to dog personality attributions. This is relevant for dog faces where eyebrows, depending on color and flexibility, can make a dog appear “angry” or “elated” without much concern for actual behavior. Dog eyes demand our attention, particularly when they take the form of what the ASPCA SAFER Glossary defines as “hard eye: dog’s eye is large and the whites are likely observable.” This hard, direct, unwavering appearance indicates threat, and the whale eye (with white sclera visible) can indicate discomfort or nervousness. A stiff, unwavering body posture often accompanies this type of eye presentation, and caution should be taken. Eyes can also assume a soft, squinty, more almond-shaped appearance leading McConnell (2007) to title sections of a book “Wrinkles Are Good,” and “Warm Eyes, Warm Heart.”


Like tails, paws do a lot of social “talking” although paws are much less noticed than tails. People who interact with companion dogs often take note of paws for parlor tricks like “high five” or “give paw.” These gestures bear no social meaning for dogs, apart from the possible resulting food reward or social praise. Instead, for dogs, “offering a paw” is a submissive or appeasing display (Lorenz 1954). Watch a dog respond to an upset owner (e.g., “Guilty Look” videos on YouTube), to see a paw raise used appropriately in a social context. Raising a paw is part of many social exchanges, see Appendices A.4, A.7, and A.9.

Body weight distribution and distance management

In social interaction, dog behaviors can be characterized as those associated with “coming closer” (distance between animals decreasing) or “backing up” (distance increasing). A dog’s body weight distribution offers subtle, yet important information. A dog with weight shifted forward, upper body pressed over the front legs, shows forward momentum, interest, confidence, or alertness. If a dog leans forward toward another dog—and the receiver leans back, looks away, or moves away—the second is engaging in conflict avoidance.

Similarly, “submissive” displays in canid social behavior aid in the prevention or reduction of fighting, aggression, or conflict. Submissive postures involve a reduction in perceived size, through lowered body and tail, pressing ears back, and, possibly, exposure of the inguinal region (Schenkel 1967), see Appendices A.4 and A.5. A dog being attacked in these postures is rare. A dog who continues to be approached could respond in defensive aggression if their initial tactic—leaning back, decreasing size, turning head—did not stop another’s advance, see Appendix A.6. Unfortunately, if dog signals go unheeded, dogs can learn to increase the use of defensive aggression over time and even fade out the use of distance-increasing signals.

Challenges to visual communication

Given the extreme morphological diversity of dogs, not all dog body parts will be visible all the time, nor are all body postures physically possible for all dogs (Price 1999). For instance, the hair or fur of some dogs prevents visible piloerection. Other dogs, particularly brachycephalic dogs, lack the highly flexible or expressive face of a German Shepherd–type dog (Bloom & Friedman 2013). Some dogs may thus be unable to signal, or their signal may not be noted. This diminishment of social signaling capacity is noteworthy because communication, as well as interpretation of communicative signals, is integral to modulating social interactions. As a result, individual dog behaviors should be considered in light of what is possible for that dog. It might be that something as trivial as shifting one’s weight back, or turning one’s head, is highly outwardly expressive for a particular dog.

Acoustic communication

Social animals tend to have more vocal nuances than those that are asocial, and dogs make a lot more noise than other canids, both in quality and quantity. Dogs whine, yelp, growl, howl, and bark (Tembrock 1976; Pongrácz et al. 2010), in addition to other less-described vocalizations such as laughing and grunting, to name a few (Simonet et al. 2001; Lord et al. 2009).

Barks and howls

Barks and howls are loud and noisy and can garner considerable attention. Howls carry for long distances, while barks are used for shorter-range communication (Feddersen-Petersen 2000). Howls and barks can be socially facilitating and can attract attention and participation from other dogs (Adams & Johnson 1994). Although, some dogs appear to bark more than others, even in the presence of the same stimulus.

Barks vary in duration and acoustic properties, but each bark is repetitive and loud. The acoustic properties of barks differ between contexts so barks performed in a disturbance (“stranger approaching”), isolation, or play context will sound different from one another (Yin & McCowan 2004). As a result, human listeners are able to characterize barks and describe tonal and high-pitched barks as indicating “fear” or “desperation” (e.g., “alone” bark), while low-pitched barks that are harsher with little amplitude modulation are described as “aggressive” (e.g., “stranger approaching” bark) (Pongrácz et al. 2006). For dogs, like other vocal mammals, vocalizations associated with affiliation and approach (high-pitched and tonal) sound different from those associated with withdrawal (low-pitched and atonal) (Morton 1977).

These acoustic rules can be applied to successful communication between humans and dogs. McConnell (1990) found that short, rapidly repeating notes were more successful in provoking dog movement than long, descending notes. This research can be put into practice in shelter settings, and volunteers should consider that tone and pitch can be more meaningful to dogs than the actual meaning of uttered words (ASPCA Webinar 2013).

Dog barks are one of the lesser-appreciated vocalizations and are associated with dog relinquishment and “misbehavior” (Senn & Lewin 1975; Wells & Hepper 2000). Owner problems with barking can stem from bark quantity (frequency) or quality (style or context) (Pongrácz et al. 2010). While barking has contextually specific acoustic properties—“meanings”—barking is a behavior that can be put under operant control, depending on the consequences that follow from the behavior. Applied Behavior Analyst, Susan Friedman, PhD, explains, “Once this idea is [understood], it opens the door to changing the duration, intensity and frequency of the behavior by changing the consequences” (Hecht 2013). Understanding that barking can be a learned behavior—and increased or decreased in particular contexts—allows people living with dogs to work with them to modulate barking when necessary (Juarbe-Díaz 1997). At shelters, everyone might benefit if dogs could learn to be quieter (see section “Shelter environment”).


Growls, too, have received scientific scrutiny. Once described simply as an “aggressive or distance-increasing call” (Houpt 2011), growls are more nuanced than initially thought (Yeon 2007). For example, growls can provide information about the growler’s size (Faragó et al. 2010a; Taylor et al. 2010), and they are performed in not just agonistic but also play contexts. In one study (Faragó et al. 2010b), growls were recorded in three different contexts: guarding a bone, growling at an approaching stranger, and during play. Growls were then played to dogs as they approached a bone that had been placed in front of concealed speakers. Dogs responded differently toward the bone depending on the growl played, suggesting that growl acoustic properties are meaningful for dogs. Dogs were more likely to retreat for a “my bone” growl than when hearing growls associated with a threatening stranger.

Olfactory communication

Dogs are known for their noses, and with good reason. Compared with relatively anosmic or “poor-smelling” animals like humans, dogs have the ability to detect and discriminate a huge number of odors (Horowitz 2009c) due to physiological structures that prioritize smelling. Scent particles enter the nose by both sniffing and regular breathing (Neuhaus 1981). These particles then enter the nasal cavity where a mucus lining covers the olfactory epithelium and mediates olfaction—smelling (Furton & Myers 2001). Considerably more genes code for olfactory receptors in dogs than in humans (Quignon et al. 2003).

The dog’s nose is a powerful tool readily harnessed for detection, discrimination, and identification (Gadbois & Reeve 2014). To name a few, dogs can be trained to identify cancerous from noncancerous tissue samples, scat of particular species, and even whether a now-absent dead body had been lying on a carpet (Willis et al. 2004; Long et al. 2007; Oesterhelweg et al. 2008). In a study of dog ability to detect the direction of a track, German Shepherd dogs inspected a small number of footprints for 3–5 s and used this information to follow the track in the right direction (Thesen et al. 1993). This ability suggests that a dog’s nose attends to minute differences in scent molecules that ultimately provide information on which footprint was laid more or less recently. Research in this area continues to grow, particularly studies investigating which training methods foster faster detection and scent learning (Hall et al. 2013).

Although dogs hold the potential for great olfactory acuity and discrimination, dogs are not necessarily relying on their sense of smell all the time (Horowitz et al. 2013). Factors such as dehydration and increased temperatures—that increase panting—can impair detection (Gazit & Terkel 2003). Additionally, differences between dogs with respect to the position of the olfactory lobe could affect dog olfaction (Roberts et al. 2010).

While dogs might enjoy engaging their noses to serve human purposes, dogs have species-specific uses for olfaction. Dogs have a secondary molecule-detection organ, the vomeronasal organ (VNO), which is directly involved in social communication and assessment of pheromones (Adams & Wiekamp 1984). Distinct from the main olfactory epithelium, the VNO is located below the nasal cavity, and its receptors also carry information to the olfactory bulb. This chemosensory organ is ordinarily viewed as responsible for pheromone detection in urine, feces, and saliva, as well as glands in the anogenital region, mouth, and face. Olfaction plays an important role in intra- and interspecific social encounters, discussed further in section “Real-world interactions.

Olfaction is essential to the dog umwelt or perceptual world (Horowitz 2009c). The job of humans, as their caretakers and observers, is to know that the dog’s nose is in play, regardless of whether we can see the nostrils twitching ever so slightly.

Patterns of communication

When interacting with dogs, people need to be aware of dog visual, acoustic, and olfactory communication. The following patterns of dog communication are particularly relevant for shelter and foster-care settings.


To live is to encounter “stressors.” Widely discussed since the early 1900s, endocrinologist Hans Selye defined stress as “the nonspecific response of [an] organism to a noxious stimulus” (Mariti et al. 2012). While stress can be considered deleterious, “stress is an environmental effect on an individual which over-taxes its control systems and reduces its fitness” (Broom 1988) stress is also functional. It serves to activate the body for protection and action. If a zebra did not perceive and respond immediately to a stressor (a lion approaching), the zebra could be killed (Sapolsky 2004). At its core, stress can promote survival.

That being said, prolonged or repeated activation of the stress response—chronic stress—can have adverse consequences. Research has found relationships between stress and increasing levels of arousal, fear, and aggression (Mills 2002; Dreschel & Granger 2005); decreased immune functioning (Glaser & Kiecolt-Glaser 2005); and decreased life span (Dreschel 2010). Attending to the physiological and behavioral aspects of stress can help ameliorate or prevent stress in the future. At the same time, there are challenges to stress identification, such as individual variability in physiological and behavioral responses, as well as a lack of correlation between behavioral and physiological stress measures (Rooney et al. 2007; Hekman et al. 2012).

Stress response

Assessment of a stressor prompts immediate physiological changes. The fight-or-flight response prepares the body for immediate action: Pupils dilate, respiration and heart rate increase, and blood moves into limbs readying the body for immediate action. Stress also produces a hormonal response—effectively preparing the individual for sustained exertion—characterized by a cascade of hormonal responses resulting in the production of the glucocorticoid, cortisol which is the most common in mammals. Cortisol levels elevate during times of stress, regardless of whether it is eustress—“good” stress, as when playing—or distress, “bad” stress. Cortisol measures—customarily collected from saliva, blood, and urine, but also feces and hair—along with behavior can offer insights into a being’s assessment of a situation. When the stressor is removed or perceived to be removed, normal bodily functions—such as food digestion, regular breathing and heart rate—return. Unfortunately, if an individual lives in a continual state of change and stressors (or perceives as much), levels could remain elevated and indicate chronic stress (Beerda et al. 1997, 1998).

Stress behaviors

Dog owners frequently refer to overt changes in dog behavior as indicators of stress, such as piloerection, trembling, and panting (Mariti et al. 2012). Research suggests that behavioral indicators of stress are less than straightforward and can vary between individuals. Thus, there is no definitive list of signs of stress (Rooney et al. 2009). Generally speaking, stress-related behaviors overlap with those associated with fear, anxiety, appeasement, and conflict. They can take on the appearance of behaviors associated with flight, freezing, or even fight.

Starting from the dog’s head, oral behaviors could include subtle snout/lip licking, yawning, and panting. Dogs may avoid eye contact or look away. Trembling and body shaking are often indicators of high psychological stress and could be accompanied by a lowered body posture, cowering, and hiding (Rooney et al. 2009). Dogs paw-lift in both asocial and social contexts, when alone and distressed, and also during social (inter- or intraspecific) conflict, confusion, or fear (for instance, of punishment) (Schilder & van der Borg 2004; Rooney et al. 2009). Periods of continual barking, whining, and howling suggest frustration or distress, although vocalization could also be socially mediated (Rooney et al. 2009).

Displacement behaviors are also important to attend to as they constitute normal behaviors performed in an “inappropriate” context (Falk 1977). Displacement behaviors are often associated with motivational conflict or frustration and could have crossover with stress-related behavior. For example, the appearance of another dog outside a dog’s run might increase yawning, a behavior not typical for dog–dog greetings.

Veterinarians mention lack of urination or defecation, or even dry mouth, as associated with stress, and one study even described “a characteristic breath odor” in distressed dogs (Mills et al. 2006). Human anxiety is associated with increased production of volatile sulfur compounds (Calil & Marcondes 2006), and persistent panting and/or drooling in dogs could alter the smell of dog breath.

Interestingly, water consumption could be an indicator of enhanced coping, as one study found that dogs who consumed water on the first day at a shelter had lower cortisol levels than dogs not observed to drink water (Hiby et al. 2006). In another study, dogs who were quicker to rest had lower cortisol levels than those who were more active (Batt et al. 2009). At the same time, dogs experiencing stress could be anywhere from shut down and inactive to highly active (Hiby et al. 2006). Sociability could be another indicator as dogs more sociable with humans had lower cortisol levels than those described as less sociable (De Palma et al. 2005). Taken together, dogs who are inactive but showing overt or subtle social avoidance should also be considered as possibly experiencing increased stress levels.

Challenges associated with stress

Even for people living with dogs, subtle dog behaviors are not necessarily attended to, and global body movements and vocalizations may be easier to recognize (Tami & Gallagher 2009; Mariti et al. 2012). Because of the overlap between stress, fear, and aggressive behaviors, subtle indicators of stress are important to observe. Dogs often behave in a graded fashion and a lip lick, head turn, avoid gaze, and freeze may come prior to a bite. Unfortunately, by not attending to these subtle behavioral indicators, an aggressive display might seem to come “out of nowhere.”

Another major challenge in attending to stress in dogs is that there is intense variation in perception of stressors, as people living in multi-dog households may know. One dog might find loud noises terrifying, while another lounges on the couch during fireworks. From an early age, dogs appear to display individually distinct coping strategies (Riemer et al. 2013), “characterized by consistent behavioral and neuroendocrine characteristics” (Koolhaas et al. 1999). Coping strategies are often described as “proactive” and “reactive,” the former characterized by boldness exploration, and fight-or-flight in response to stressors, while reactive individuals tend toward freezing when encountering aversiveness.

Ultimately, individual monitoring and attention to individual coping strategies is useful to detect a stress response. Researchers concerned with the welfare of dogs have noted the importance of “[paying] attention to individual dogs and [noting] any changes in their behavior” (Rooney et al. 2009).

Stereotypic behavior

Stereotypic behavior has traditionally been defined as behavior patterns that are repetitive, unvarying, and seemingly functionless (Mason 1991) and that manifest differently between species. Behaviors could include repetitive spinning, jumping, pacing, licking, and self-biting, among others. Abnormal behaviors can develop as a coping mechanism to poor environments and can maintain even in the face of environmental improvement. As a general matter, they may indicate poor welfare, but on an individual level, these behaviors could offer individuals a type of “do-it-yourself” enrichment, and nonstereotyping individuals in poor environment could be in a worse state than stereotyping individuals (Mason & Latham 2004). Studies of the conditions under which repetitive behaviors are performed by kenneled (and shelter) dogs can give further insight into their meaning and indicators for welfare (Denham et al. 2014). As a result of their complicated presentation and meaning, simply thwarting such behaviors could increase distress or the frequency of new deleterious behaviors. Repetitive, unvarying behaviors necessitate attention.

Fear and aggression

The outward appearance of aggression—loud noises, teeth bared and flashing—is hard to miss. But the precursors to aggression are many, and given the novelty of shelter environments for dogs, fear, and fear-related aggression, are noteworthy.

Fear is an emotional response evident in both physiologic and behavioral responses when something is perceived as frightening or indicative of danger (Boissy 1995). Fear-related behaviors at the veterinary clinic have been described as “fixed stare, lowered or tucked tail, crouched body posture, hiding, pressing into owner, attempt to jump off table” (Döring et al. 2009). Whereas confident or calm dogs have a high or mid-length tail and raised or neutral posture, fearful dogs are marked by low tail, depressed posture, and ears back (Darwin 1872) (see Appendices A.3 and A.6).

Fear and aggression are often connected. If pressed, dogs exhibiting fearful postures may freeze, continue to withdraw, or even flip onto their backs in a display of passive submission (Schenkel 1967). But others with a more “reactive” coping strategy may display a defensive attack. This posture differs from an offensive aggressive display in that the defensive dog’s posture is pulled back, with ears back and tail tucked; while he might bark, bare teeth, and lunge forward, ultimately the dog is retreating, attempting to escape or decrease proximity.

Dogs displaying more offensive aggression may lean forward with a fixed stare, raised tail, and stiff or frozen body, giving a “hard eye” with a closed mouth or offensive pucker: in a sense making themselves appear bigger.

Fear and aggressive behavior, like stress-related behaviors, can be a functional response to try to increase distance from a feared object or animal. But animals experiencing unrelenting—chronic—fearful or aggressive states can have decreased well-being. There can be a strong relationship between fear and stress, as dogs who crouch when exposed to frightening stimuli have higher cortisol levels than those who maintain an upright posture (King et al. 2003).

Dogs exhibiting continual fear, anxiety, and frustration might have increased arousal (whether subtle or overt) and have a lower threshold for aggression (Panksepp 1998). Sadly, dogs living with chronic stress or fear may have negatively impacted health and decreased length of life (Dreschel 2010).


The term “dominance” is readily used by the general public and applied to everything from dogs being “disobedient” (jumping up, stealing food, etc.) and scuffles between dogs and dogs appearing to show aggressive or assertive behaviors. The term is readily applied without consideration for contextual learning or preceding behaviors.

Unfortunately, these “definitions” of dominance do not have scientific merit. When used in animal behavior contexts, “dominance” is not an attribute of an individual; instead, it is commonly used to describe a dyadic relationship (Drews 1993). “Dominance relationships”—in which one individual is more assertive and the other is more submissive—are not set in stone and are malleable. For example, motivations for particular resources (e.g., resource-holding potential) differ between individuals and affect outcomes (Bradshaw et al. 2009).

Unfortunately, the idea of dominance has been widely overstated and oversimplified as it relates to the dog–human relationship. Owners and dog handlers sometimes use forceful methods to deter dogs from “asserting dominance.” Unfortunately, confrontational methods such as the “alpha roll” and “dominance down” can be associated with an aggressive response on the part of the dog (Herron et al. 2009).

When investigating the complex interplay between life experience and individual dog behavior, it becomes apparent that while a concept like dominance may enter into the social behavior of dogs, their individual behaviors are not defined by it. For example, a dog who is described as “dominant” because he guards food could, likely, learn to stop guarding food (Wood 2011). Thus, the utility of this label is questionable.


As any observer of dogs knows, dogs play—a lot. Young dogs may spend up to one-third of their awake life in object play, social play, or running, locomotor play; and among dogs, play continues, albeit at a reduced rate, into adulthood—a rare and perhaps singular phenomenon among animals (Horowitz 2002). While “play” may seem to be trivial, play behavior is an integral part of social and physical development for dogs (Rooney & Bradshaw 2014). Dogs not only play with other dogs, but readily, and often, with humans and even other species. While play might be seen as “just something that dogs do,” it has unique characteristics that could offer a snapshot into the dog’s mind. Researchers breaking down the nuances of play find that it is marked by “a dizzying series of synchronous behaviors, active role swapping, variations on communicative displays, flexible adaptation to others’ attention, and rapid movement between highly diverse play acts” (Horowitz 2009c). The patterns of behaviors in play indicate that dogs have some rudimentary understanding of the minds and perspectives of other dogs (Horowitz 2009a) (Figure 1.2).


Figure 1.2 A dog play bow.

Reproduced with permission of Natalya Zahn. © Natalya Zahn.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 7, 2017 | Posted by in GENERAL | Comments Off on 1: Introduction to dog behavior

Full access? Get Clinical Tree

Get Clinical Tree app for offline access