Humans and dogs (Canis familiaris) share a special bond, as any dog owner would be keen to point out. This has been true over the history of our interspecies relationship, it’s even been suggested that dogs could have been the catalyst for human civilization (Hare, Woods 2013). Worsley and O’Hara’s study set out to definitively show that dogs can and do use cross-species referential signaling in their communication with humans (2017). This is noteworthy for multiple reasons, the foremost being that these signaling gestures in non-human species are extremely uncommon, especially in the wild (Vail, Manica, Bshary 2013).
Secondly, while it has been shown that dogs can understand referential signals from humans, there has been little research into dogs’ ability to communicate with humans in the same manner. Prior research into this was focused on dogs ‘showing’ ability, a behavior in which they orient their body towards a goal object while alternating their gaze between the human whom they are trying to signal and the goal object (Miklosi, Polgardi, Topal, and Csanyi, 2000).
The idea behind this study is to further examine and document the ability of dogs to provide referential signals to their human cohabitants. Understanding the depth of the ability of dogs to communicate with us has the potential to not only shed light on the history of our coevolution but also to make us more in tune with our companions in the present.
Owners of 37 domestic dogs were recruited for this study, in total there were 16 females, and 21 males and their ages ranged from 1.
5 to 15 years old. The dogs must have lived with the same owner for a minimum of 5 months before the study was conducted. Breed types were varied and the sources from which the dogs were obtained ranged from breeders to rescues. It was also noted how many other humans lived in each dog’s household, aside from the primary owner.
This study was conducted using the citizen science method. The researchers asked dog owners to film their dogs performing communicative behaviors using their cell phones. They were instructed to do so each time a behavior occurred. The owners were shown some example footage so they understood what they should be looking for and were told that there was no limit to the amount of footage they could submit. Owners were encouraged to record every bout of communication they saw so that the same type of communication would be documented multiple times. The downside of this sort of data collection is that it can be unreliable, as it rests on the owners being conscientious, observant, and dedicated to a study for which they likely would receive no reward for participating.
After the footage had been collected, the researchers analyzed the footage and labeled each behavior and the perceived goal. Gestures were categorized by their apparent satisfactory outcome (ASO). ASOS is the endpoint of a communicative bout, where the end goal was achieved and the dog got what it wanted. The four ASOs in the study include, “Scratch me!,” “Give me water/food!,” “Open the door!,” and “Get my toy!” By limiting communication documented to these 4 goals the researchers could make their data more concise and focused, but at the same time, it’s extremely limiting. Speaking from experience, dogs can communicate about many, many more outcomes than just these four, and I was hoping that this study would go a bit further in-depth.
Researchers narrowed down the footage to identify 19 referential gestures which fit in the definition of a referential gesture, namely that the behavior is directed towards an object or subject, references another object or subject, receives a voluntary response, is intentionally produced, and on its own, the behavior is mechanically ineffective. 1016 instances of these 19 gestures were recorded from 242 bouts of communication. It was noted that different or even multiple gestures could be used to indicate the same ASO. For example, 14 referential gestures were recorded relating to the “Scratch me!” ASO, the most common, from affectionate pawing the owner to rolling over and exposing the belly. Conversely, the most common gesture was gaze alternation, which was recorded 381 times and was used for each of the four ASOs the researchers settled upon. Interestingly, dogs that lived with more humans in the household demonstrated a greater repertoire of gestures as compared to those dogs which lived with only one owner in the household.
This study shows one of the hallmarks that has made the human and dog relationship so special throughout our history and has been a great complementary addition to previous research that’s shown how adept dogs are at understanding human referential signals. That being said, as a dog owner, reading this study was akin to me reading a study showing that humans are capable of communicating with spoken language. While I can acknowledge that it’s important to have documented this behavior in a formal study, the results were decidedly unremarkable.
As noted by the researchers themselves, this study only scratches the surface of the depth with which humans and dogs can communicate using referential signaling. Initially, 47 potential referential gestures had been identified, but because Worsley and O’Hara had been unable to conclusively determine that those gestures fit in with the definition of referential signaling, they disregarded 28 of them. Possible reasons that they would have been left out include not receiving a voluntary response, so it’s very possible that the owners were just clueless. Furthermore, if the owners were clueless, likely, many bouts of communication were not documented in the first place.
I would be tempted to perform a similar study to this, but be a bit more selective with participants, focusing on owners who spend a lot more time with their dogs than the average. These could include military and police dogs and their handlers, service animals, or other working dogs and their owners. I think the data collected from such a study might be more illuminating on the extent of the potential for interspecies communication and possibly more relevant for putting our coevolution success into context.
There are a couple of points in the study which seem incorrect and indicate to me that the researchers aren’t fully aware of the extent of codependency in a dog/human relationship. They point out that the four ASOs identified only benefit the dog and not the owner. I think any ASO for any creature in the world has some self-serving intent, but the authors including this line in the study imply that there’s sort of a one-sided benefit to this dog/human relationship. I would argue that all the ASOs in this study are mutually beneficial. For example, it’s been documented that petting or scratching a dog has numerous positive effects on both the dog and the human. It’s been shown that even a few minutes of interaction with a pet dog provides an increased reduction in stress compared to the same amount of time spent with a human friend (Allen, 2002). As for the “give me water/food!” ASO, it’s pretty obvious that if the owner values the companionship of the dog, it’s in both of their best interests to keep it fed and provide it with water. The communicative behaviors indicate that ASO facilitates that process and logically would be a benefit to both owner and dog.