Do Dogs Understand the Consequences of Their Own Actions?

New research shows they do, as do other ecologically valid studies.

Posted Feb 21, 2021

For the past few days, my email inbox has been ringing with references to a recent study by a team of Hungarian researchers called "Dogs (Canis familiaris) recognize their own body as a physical obstacle" published in Scientific Reports and available online.1 To conduct their novel and clever research, Rita Lenkei and her colleagues at Eötvös Loránd University in Budapest tested 32 dogs who were faced with two different situations and tested in what they call the "body as an obstacle" task. Dogs standing on a mat were asked to pick up and give an object to their humans in two different situations. In the test situation, the object was attached to the mat and a dog had to leave the mat to be able to lift the object and give it to their human. In the other test condition, the control, the object was attached to the ground and the dog didn't have to leave the mat to give it to their human. 

 "Dogs (Canis familiaris) recognize their own body as a physical obstacle."/Open access
The schematic outlay of the experimental setup. The dog was asked to pick up the object (dog toy) and pass it to the hands of the owner, who stood in front of the dog, along the short side of the mat.
Source: "Dogs (Canis familiaris) recognize their own body as a physical obstacle."/Open access

The researchers learned, "Dogs came off the mat more frequently and sooner in the test condition than in the main control condition, where the object was attached to the ground." Based on these data they concluded, "This is the first convincing evidence of body awareness through the understanding of the consequence of own actions in a species where previously no higher-order self-representation capacity was found. We urge for an ecologically valid approach, and following of bottom-up methods, in studying modularly constructed self-representation."

I also was interested in this study because I've also written about what I call "body-ness," referring to situations when an individual knows the placement in space of parts of their body as they run, jump, perform acrobatics, or move as a coordinated hunting unit or flock without running into one another.

I'm a fan of simple and innovative research and this is a great example. While I have no problem with how the study was conducted, with the quality of the data, or how the information was analyzed, I question whether this really "is the first convincing evidence of body awareness through the understanding of the consequence of [a dog's] own actions..." The people who took the time to write to me were also unconvinced. And 8 of the 10 people, including five researchers, pointed to a host of other examples that easily could be explained as "evidence of body awareness through the understanding of the consequence of [a dog's] own actions." Maggie commented, "This is a great study, but isn't this what dog training is all about?"

Here are a few examples. None comes from a controlled experiment, but they all come from extensive observations of dogs (and other species) in a number of different social contexts. Let's consider the general area called "social learning," during which an individual learns, among other things, what their own body movements mean and how they acquire knowledge about the consequences of their actions. There are countless examples of this in the animal behavior/ethology literature.2 

Simply put, individuals learn how to interact with other individuals to accomplish certain goals in activities including greeting, courtship, mating, threatening and fighting (agonistic encounters), feeding, and playing. As they do this, over time they learn the consequences of their own actions and how to modify their movements during the course of these encounters and learn what works and what doesn't work. 

In addition to what students of animal behavior call "proximate" conditions, what's happening right now and who's involved, there's also an evolutionary story involving how different movements become social signals. This process is called "ritualization" and it involves how movements become, evolve into, clear and unambiguous signals. A good working definition of ritualization is, "the evolutionary process by which an action or behavior pattern in an animal loses its original function but is retained for its role in display or other social interaction."3

Body awareness and social play

 Marc Bekoff
Ruby (right) bowing to Scone.
Source: Marc Bekoff

An excellent example of social learning and ritualization is the use of bows by dogs and other animals to ask other individuals to play and to maintain the play mood. Many animals use different movements to announce that they want to play and not fight or mate. Canids punctuate play sequences using a bow to solicit play, crouching on their forelimbs while standing on their hind legs. Bows are used almost exclusively during play and are highly stereotyped—that is, they always look the same—so the message “Come play with me” or “I still want to play” is clear. Play bows are ritualized honest signals, a sign of trust. They provide a foundation of fairness for social play and are used by dogs to share their intentions to play.

Ruby and Scone are a good example of the ways in which dogs learn what their movements mean—in this case, bows and other play signals—and the nature of their play changed over time as they got to know one another. I've seen scenes like this numerous times in dog parks and among free-ranging dogs. From these observations, we also make inferences about dogs having a Theory of Mind (ToM), 

In a research project I conducted with my students many years ago, we learned that young and adult dogs, young coyotes, and young wolves modify their body movements in terms of how long they bow and how they bow—what bows look like—depending on where they are performed—either at the beginning of a play bout or while they're playing. We learned that the mean duration of bows performed during play bouts was on average 0.03 to 0.07 seconds shorter than the mean duration of bows at the beginning of play sequences. In addition, for coyotes, beagles, and adult dogs, bows performed during an interaction showed significantly higher variability in duration than bows performed at the beginning of sequences. Also, bows performed to ask another individual to play were more stereotyped—and perhaps more recognizable—than bows used during play sequences. 

There are a few explanations for the differences in bows that occurred at the beginning of a play and those that occurred during a play session. The greater variation in duration for bows performed during a sequence can be explained by the fact that these bows were preceded by a variety of different acts from which the individual went into the bow. On the other hand, the bows that occurred at the beginning of sequences almost always took place after the individual had been standing upright for a few seconds or as part of an approach.

The highly stereotyped bow—what they looked like—when they were used to initiate play can be explained by a greater need for an individual to tell another individual they want to play rather than fight, mate, or have them for a meal. It might be less important to do so during a play session after the play mood has been established. 

 Joan Baez, used with permission
Source: Joan Baez, used with permission

Ecological validity

In line with Dr. Lenkei and her colleagues calling for more ecological valid or ecologically relevant studies, analyses of play bows and other movements show evidence of body awareness through the understanding of the consequence of an individual's own actions. 

I was pleased to learn about their simple yet extremely important study on a topic that needs much more research. It raises many salient questions that call for further study.

I look forward to the further discussion of body awareness in diverse nonhumans. Development studies would be especially important for learning about how body awareness emerges over time in different sorts of social encounters.

Comparative evolutionary studies also are needed. Some researchers believe that this sort of self-awareness might be more prevalent in highly social species such as domestic dogs, but we really don't know much about the evolution of body-ness. It wouldn't be surprising to see that it has evolved in many different species for various reasons, and that's what makes studies of body/self-awareness so exciting.

References

Notes

1) The abstract for this study reads: Mental representations of one’s own body provide useful reference when negotiating physical environmental challenges. Body-awareness is a neuro-ontogenetic precursor for higher order self-representation, but there is a lack of an ecologically valid experimental approach to it among nonhuman species. We tested dogs (N = 32) in the ‘body as an obstacle’ task. They had to pick up and give an object to their owner, whilst standing on a small mat. In the test condition we attached the object to the mat, thus the dogs had to leave the mat because otherwise they could not lift the object. Dogs came off the mat more frequently and sooner in the test condition, than in the main control condition, where the object was attached to the ground. This is the first convincing evidence of body awareness through the understanding of the consequence of own actions in a species where previously no higher-order self-representation capacity was found. We urge for an ecologically valid approach, and following of bottom-up methods, in studying modularly constructed self-representation.

Mass media coverage of this study can be found here and here.

2) Numerous references can be found here

3) For more information on the process of ritualization click here

Bekoff, Marc. When Dogs Play, They Follow the Golden Rules of Fairness. (Fair play requires dogs to stick to mutually agreed upon codes of conduct.)

_____. Social Communication in Canids: Evidence for the Evolution of a Stereotyped Mammalian Display. Science, 197, 1097-1099, 1977. 

_____. Play signals as punctuation: The structure of social play in canidsBehaviour 132 419-429, 1995. 

_____. Dogs at Play: Feeling Safe, Having Fun, and Playing Fair

_____. Dogs at Play: What They Do, Know, Think, and Feel.

_____. Theory of Mind and Play: Ape Exceptionalism Is Too Narrow. (A narrow focus on great apes excludes the possibilities of ToM in other animals.)

_____. Do Animals Know Who they Are? (There are degrees of self-awareness in animals.)

_____. Awareness: Animal reflections. Nature 419, 255, 2002.

_____. Consciousness and Self in Animals: Some Reflections. Zygon, 229-245. 2003.

_____. Dog Training: Blending Science With Individual Personalities.

_____ and Paul Sherman. Reflections on Animal Selves. Trends in ecology & evolution 19(4), 176-180, 2004.