In response to my article on "What a Dog's Tail Wags Really Mean" I received a note about a Labrador Retriever named Molly. It read in part:
After her accident [involving a collision with a motorcycle] the vet had to amputate her tail leaving a stump of only around two inches. She recovered okay and still has a personality that is good with people and kids (just like before the accident) but something has changed in her dealings with dogs. We always took her to an off-leash dog park here in Vancouver and she always had good relations with the other dogs. Since her tail was cut off though, other dogs seem to be suspicious of her and they don't come up to her the same way they used to. A few have actually snapped or growled at her, which never happened before. Could this be because her tail is gone so that the other dogs can't understand her tail communications anymore?
Given the importance of a dog's tail for signaling this is a likely conclusion. This idea that tail length might influence canine communication worried me particularly in the context of tail docking, which is where a dog's tail is deliberately cut short in certain breeds of dogs, such as Rottweilers, Doberman Pinchers, Cocker Spaniels, Boxers and so forth. My concern has always been that docking significantly limits the use of tail signals and thus reduces the effectiveness of a major channel of communication in dogs. Although this is an important issue I could find no data in the scientific literature which addressed the question of whether dogs with short or missing tails had signaling problems. To shed some light on the matter I carried out an observational experiment and published the data in my book "How to Speak Dog." In our study we observed dogs interacting in a confined city park area where dogs were allowed to be off leash. We tallied 431 encounters between dogs. Most of these (382 or 88 percent) were typical canine greeting behaviors, often followed by play behaviors including the usual chase games. The remaining 49 encounters contained an aggressive element on the part of one or more of the dogs involved. These could be as mild as a snarl and a snap with no physical contact or as severe as an actual physical assault drawing blood. The dogs involved were coded simply on the basis of whether they were tailless (most likely docked) or with a tail (undocked or only partially docked). To be classified as being tailless the dog had to have a tail approximately six inches or less in length (we eliminated small toy dogs from the sample confining our observations to dogs that stood approximately 18 inches at the shoulder or more). The proportion of dogs with tails was considerably higher in this population, and amounted to 76 percent, as opposed to 24 percent of dogs without tails. However, when we looked at the dogs involved in aggressive incidents 26 (or 53 percent) of these confrontations included dogs without tails. On the basis of the number of dogs with and without tails, we would have expected only 12 aggressive incidents (24 percent) to involve tailless dogs. The surplus number of confrontational incidents involving tailless dogs is highly statistically significant when we do the appropriate calculations. Thus our results show that dogs with short or absent tails are twice as likely to have aggressive encounters as dogs with longer, and hence more visible, tails. One cannot help but wonder if the increase in aggressive encounters in short tailed dogs might not have to do with the ambiguity or absence of appropriate visible tail signals that could have indicated a social versus a hostile attitude on the part of the docked tail dogs and thus avoided a potential conflict.