At last month's ESVCE conference in Avignon, France, there was heated discussion about the concepts of hierarchy and dominance, made all the more complex by the conference being held in English but with mainly French, German, Belgian, Spanish and Italian participants. However semantic and esoteric this debate becomes, it can never be forgotten that these concepts can be decisive in the interpretation of problematic behaviour in pet animals. Getting the underlying idea wrong can have real consequences for the welfare of individual animals.
Much of the smoke—even fire—has come from disagreements as to how the social behaviour of wolves and domestic dogs should be interpreted. Yet similar (unresolved) differences of opinion remain among cat behaviourists. Some authorities use the concept freely, while others, myself included, prefer to think of cats as territorial animals that adapt to living in a crowd through a variety of strategies, that can include co-operation as well as avoidance and, in extreme cases, outright tyranny.
Some of the differences of opinion revolve around the interpretation of individual animals' behaviour: for example, what the submissive/affiliative posture of wolves means, likewise head-to-head rubbing in cats. However, misunderstandings also arise because of the use of the same term to describe subtly different, if related, phenomena, and none more so than the use of the word "dominance".
Wolf - submissive/affiliative behaviour
Dominance is both a word in everyday use, with connotations of importance, power and influence, and also a technical term in animal behaviour; misunderstandings about which meaning is being used can lead to confusion. In ethology, dominance is used to describe an asymmetric relationship between two animals, in which one animal has more influence over the behaviour of the other than vice-versa. Traditionally, this has been conceived of—and measured—as the result of repeated aggressive encounters, which one animal usually wins, and the other loses. "Intelligent" animals like dogs and cats may then learn to modify their behaviour when dealing with the same animal over and over again: an animal that has lost in the past can signal its intention not to escalate the contest, either using a species-typical signal or simply by getting out of the way, while the winner needs only to announce its intentions with a threat. Both animals are thought to benefit from this de-escalation, since neither has to run the risk of being injured in an actual fight.
In a group of three or more animals, the pair-by-pair dominance relationships can often be arranged in a more-or-less linear social hierarchy. The term "hierarchy" is normally reserved for groups, since it describes the additional level of complexity that emerges from combining together several (at least three) dominance relationships. Even in the rare situation that each of these relationships is clear-cut, the hierarchy may not be; in real groups of animals, circular and situation-specific hierarchies turn out to be the rule rather than the exception.
In the ethological literature, dominance and hierarchy are (or at least should be) used as properties of pairs and groups of animals respectively. They are used to quantify relationships between animals in a way that can be compared with other measures—reproductive success, for example. While lifetime reproductive success is a key idea in evolutionary biology, it is generally accepted that this is a concept available only to the human observer, and that the animals themselves, while struggling to leave as many descendants as possible, are themselves unaware of how successful they are being, or how successful their immediate rivals are by comparison with themselves.
Likewise, dominance is a property that emerges from encounters between animals; there is no need for the animals themselves to be aware of their relationships. All they have to do is to be able to modify their behaviour when they identify an animal they've encountered before. It's possible to build robots that establish very convincing and stable dominance hierarchies, based on straightforward stimulus-and-response rules, and no "awareness" whatsoever (for a detailed discussion of the differences between dogs and robots, see David McFarland's "Guilty Robots, Happy Dogs").
So far, cognitive science has failed to demonstrate "theory of mind" in dogs (or cats), although dogs at least do seem to have some comprehension of third-party social relationships. Therefore it has to be left as a moot point whether dogs (or cats) actually understand the relationships that they have with the animals around them (including humans), or whether they are simply responding to combinations of stimuli that have gained meaning through previous encounters.
And it's yet another step to the way that "dominance" is used by some dog trainers as a motivation, something that dogs are trying to gain for its own sake. Biologists generally think of dominance as something observable that emerges out of competitions for key resources—food and mating opportunities are the two most obvious examples. Somehow this concept seems to have been turned upside down, that competition between dog and owner is the result of the dog's need to achieve dominance for its own sake.
