An important question in the study of canine cognition is how dogs understand us humans, given that they show impressive abilities for interacting and communicating with us, which may include an understanding of seeing and knowing (Huber 2016). The new three-year project funded by the Austrian Science Fund FWF (Principal Investigator: Prof. Ludwig Huber, starting in January 2024) aims to test two main hypotheses about dogs' ability to take perspectives (Huber & Lonardo 2023) using new experimental and methodological approaches:

1. Dogs are able to use previous experiences from their own perspective to understand what others might be able to see and therefore know and intend to do.

2. Dogs can distinguish between false and true beliefs of others, they can flexibly use this ability in implicit and explicit tasks to predict the actions of others, and they can attribute beliefs based on information they have received through different sensory modalities (visual and auditory).

The following tests have been conducted so far in the Clever Dog Lab

The following tests have been planned or are ongoing:

Investigating perspective-taking in dogs with implicit methods: eye-tracking

In this study (ongoing) we apply the eye-tracker technology (Huber et al. 2023). Dogs’ implicit sensitivity to an agent’s false belief will be inferred from their anticipatory looks, pupil sizes (Völter & Huber 2021) and looking times as they observe videos on a computer screen. The videos will show, in two initial familiarisation trials, a conspecific, the agent dog, approach one of two tents where he has seen another dog, the hiding dog, went to hide. During test trials we will employ a change of location task (Lonardo et al. 2021), in all conditions of which the hiding dog will first go to hide in tent A, then move to tent B and finally go back into tent A (side of A and B counterbalanced across dogs and sessions). We will manipulate the agent dog’s mental state (false belief, ignorance, true belief) about where the hiding dog is hidden at the end of the hiding sequence. If subjects are implicitly sensitive to others’ mental states and this task is suitable to show it, we predict that subjects will expect that the agent dog will search for the hiding dog in the mental state-congruent tent more often than in the mental state-incongruent tent.

Human cognition is involuntarily influenced by the presence of others and it has therefore been defined “altercentric” (Kampis & Southgate 2020). Dogs have been proposed as comparative models of human social cognition as their upbringing in human households as well as their closely interconnected evolutionary history with humans might have shaped similarities in the socio-cognitive mechanisms of the two species (Huber 2016). In the proposed experiment, we aim at investigating whether (8- to 10-month-old) infants’ and adult dogs’ pupil sizes can be used as a measure of an “altercentric bias”, the tendency to unconsciously take on others’ perspective. To this end, we plan on conducting a comparative eye-tracking study.

A central issue when testing animals’ ToM with non-verbal tasks is the difficulty of finding measures that tap into the animals’ ability to go beyond the mere observation of directly visible cues and that actually require the additional inference of an invisible mental state to be passed (Heyes, 1998). One methodology to test this ability is the so-called “experience-projection test” (Heyes, 1998), which requires the participants to project their own previous experience with novel visual barriers (i.e., the possibility or impossibility to see through them) onto the experimenter. The only two studies adapting this methodology for dogs have failed to provide evidence for such an ability in this species (Lonardo et al. 2024; West-Brownbill et al. 2024). In this study we will install the experience-projection procedure into the non-verbal change of location task which has been already successfully employed by us in an earlier study (Lonardo et al. 2021).

Previous studies investigating canine perspective-taking in the Clever Dog Lab

In the study published in the journal Proceedings of the Royal Society B (Lonardo et al. 2021), we investigated whether dogs distinguish between human true (TB) and false beliefs (FB). In three experiments with a pre-registered change of location task, dogs (n = 260) could retrieve food from one of two opaque buckets after witnessing a misleading suggestion by a human informant (the ‘communicator’) who held either a TB or a FB about the location of food. Dogs in both the TB and FB group witnessed the initial hiding of food, its subsequent displacement by a second experimenter, and finally, the misleading suggestion to the empty bucket by the communicator. On average, dogs chose the suggested container significantly more often in the FB group than in the TB group and hence were sensitive to the experimental manipulation. Terriers were the only group of breeds that behaved like human infants and apes tested in previous studies with a similar paradigm, by following the communicator’s suggestion more often in the TB than in the FB group. We discussed the results in terms of processing of goals and beliefs. Overall, we provide evidence that pet dogs distinguish between TB and FB scenarios, suggesting that the mechanisms underlying sensitivity to others’ beliefs have not evolved uniquely in the primate lineage.

In this study (Lonardo et al. 2024), we tested pet dogs (in total, N = 92) on adaptations of the ‘goggles test’ (Heyes 1998) previously used with human infants (Meltzoff & Brooks 2008) and great apes (Vonk & Povinelli 2011; Karg et al. 2015). In both a cooperative and a competitive task, dogs were given direct experience with the properties of novel screens (one opaque, the other transparent) inserted into identical, but differently coloured, tunnels. Dogs learned and remembered the properties of the screens even when, later on, these were no longer directly visible to them. Nevertheless, they were not more likely to follow the experimenter’s gaze to a target object when the experimenter could see it through the transparent screen. Further, they did not prefer to steal a forbidden treat first in a location obstructed from the experimenter’s view by the opaque screen. Therefore, dogs did not show perspective-taking abilities in this study in which the only available cue to infer others’ visual access consisted of the subjects’ own previous experience with novel visual barriers. We conclude that the behaviour of our dogs, unlike that of infants and apes in previous studies, does not show evidence of experience-projection abilities.