Dogs, as a domesticated pet, generally tend to pay close attention to humans. Think about how your own dog might react to seeing you or a stranger approaching; how different tones of voice can affect their response to you; how their response to non-human animals is starkly different than a reaction to humans. Dogs are generally very aware of humans around them, whether that human is their owner or not. How abstract of a human shape will dogs respond to? Delanoeije et al. (2020) attempted to study how dogs respond to human movement and form in point-light figures.
Johansson (1973) developed point-light figures (PLFs), which consist of very few dots that track movement. PLFs abstract figures, yet important features such as natural-looking motion are preserved. Delanoeije et al. (2020) attempted to test if dogs preferentially look towards humans by testing to see how dogs reach to human and non-human PLFs.
Delanoeije et al. (2020) performed three separate experiments, each of which are described in this paper. In each, dogs were presented with a projection on a screen of a human PLF and a distorted PLF, and their eye movements were tracked with a camera to map what they preferred to look at. The distorted PLFs were made by scrambling the dots (meaning changing their positions and movement timings) and/or inverting the PLF upside down (which effectively reduces even human ability to recognize normal human movement and figures as seen in Hiris et al. 2005).
In experiment 1, the authors presented the dogs with front-facing human PLFs appearing to walk towards the viewer/dog alongside a scrambled and inverted PLF. After the dog settled down in the dark room, the authors played a noise (male human, female human, pigeon, or chicken) to get the dog’s attention as the PLF visual projection began (“sound phase”). The sound continued for 7 seconds and then only the PLF projection played for 17 seconds (“silent phase”). The authors found that the dogs responded very similarly no matter which of the sound options were played – they were all equally effective at getting the dog’s attention in the sound phase. The researchers found that the dogs preferred to look at the front-facing human PLF rather than the non-human one. This may be because dogs are socially taught to pay attention to humans looking or moving in their direction. This could also be because dogs might need to recognize “looming” behavior – any animal moving towards them, as that could be an incoming threat. However, the preference to look at the human is less obvious in the sound phase than the silent phase. The authors were unsure why – maybe the sound was distracting, or maybe the dog took some time to recognize the human figure.
In experiment 2, the authors only used the female voice sound to get the dogs’ attention and projected a human walking in profile (sideways) rather than facing the viewer/dog. Otherwise, the setup remained the same. This time, the dogs did not preferentially look at the human PLF. This could be because the figure is harder to recognize (for example, it is much less symmetrical) or because there is less social importance for a dog to pay attention to a human walking in profile.
In experiment 3, the researchers tried to test if the dogs respond to human motion or human shape. They did this by presenting the dogs with front-facing human PLFs as in experiment 1, and scrambled, but not inverted, PLFs. The purpose of this is to preserve some specific motion characteristics (such as, for example, recognizable arm movements) while disrupting the larger “global” human form. Essentially, when a PLF is scrambled and not inverted, parts of the human walking movement might be recognizable, but actual human form isn’t. In this experiment, the dogs still preferred to look at the human PLFs just as much as they did when the non-human PLF was also inverted in experiment 1. Thus, the larger global form of the PLF is important; the dogs don’t only notice natural-appearing movement, they actually notice human body features and respond to that.
Taken together, these three experiments show that dogs prefer to pay attention to human figures, even when those figures are highly abstracted. They tend to recognize (or at least respond to) front-facing human figures more than side-facing human figures. This supports the idea that dogs look at things that are socially relevant. These experiments also show that dogs respond more to human-specific movements rather than generally natural movements, as seen in experiment 3. The researchers also found that the dogs did not have different attentional reactions to the PLFs based on what sounds grabbed their attention (male human, female human, pigeon, or chicken), which the researchers did not expect.
Citations
Delanoeije, J., L. Gerencsér, & Á. Miklósi. 2020. Do dogs mind the dots? Investigating domestic
dogs’ (Canis familiaris) preferential looking at human‐shaped point‐light figures.
Ethology 126(6):637-650. https://doi.org/10.1111/eth.13016
Hiris, E., D. Humphrey, & A. Stout. 2005. Temporal properties in masking biological motion.
Perception & Psychophysics 67:435-443. https://doi.org/10.3758/BF03193322
Johansson, G. 1973. Visual perception of biological motion and a model for its analysis.
Perception & Psychophysics 14:201-211. https://doi.org/10.3758/BF03212378
