Abstract: The ability to reliably detect what others are attending to seems important for social species to interact with their partners. Domestic horses (Equus caballus) have lived with humans for over five thousand years, hence they might have developed sensitivity to human attention. In the present study, we investigated whether horses would discriminate the situation in which a human experimenter could see them from the situation in which she could not. Specifically, we tested whether horses understand the role of eyes in human attentional states, produce more visual gestures when the experimenter can see their begging behaviors and produce more auditory or tactile gestures when she can not. We used with a slight modification the paradigm that previously yielded support for chimpanzee understanding of human attention (Hostetter et al. 2007). Twelve horses were offered food by the experimenter who showed various attentional states in front of them. We scored frequency of begging behaviors by the horses. In experiment 1, we set three kinds of condition: hand over the eyes, hand over the mouth and away. In the last condition there was only a food in front of horses, which was a control condition. The results showed that horses produced more auditory or tactile begging behaviors when the experimenter“s eyes were not visible than when her eyes were visible, but there was no difference in visual begging behaviors. In experiment 2, we set two kinds of condition: eyes closed and eyes open. The horses also produced more auditory or tactile begging behaviors when the experimenter”s eyes were closed than when they were open. However, there was no difference in visual begging behaviors. These results show that horses discriminate the situation in which humans can see from that in which humans can not. Of special interest, horses increased only auditory or tactile behaviors, not all types of communicative behaviors, when the experimenter could not see their begging behaviors. This result suggests that horses are sensitive to human attentional states. Moreover, horses may do recognize the eyes as an important indicator of whether or not humans will respond to their behavior and they may be able to behave flexibly depending upon human attentional states

Abstract: The cognitive abilities of social ungulates, in particular horses, have widely been neglected. Preliminary results suggest that horses are capable of social cognition which they acquire through social learning. They gain information from the observation of the interaction of a conspecific and a human experimenter, and adjust their own behaviour towards the experimenter with respect to the observed horse"s reaction and relative dominance status (Krueger and Heinze, 2007). Horses are a highly social species that still exists in different evolutionary stages: domestic horses, feral horses and wild horses (Przewalski horses). Additionally, domestic and wild horses differ in their individual social behaviour. For example, in social interactions Przewalski horses appear to act significantly more aggressively than domestic horses. Therefore studies on horses are particularly suitable to investigate whether convergent social evolution favours convergent cognitive evolution. By a comparative study concerning their reasoning abilities in a specific situation, we will attempt to determine the influence of domestication and feralisation on the evolution of social cognition and to investigate possible differences in their abilities to cope with stressful situations. We started to observe the behaviour of domestic and wild horses, in particular during the integration into new social groups, especially in relation with their knowledge of the social structure of new groups and their own relative social status. Selected agonistic interactions will be measured and statistically evaluated. Additionally, the stress level of the horses will be determined by an analysis of stress hormone levels, particularly cortisol metabolites, in plasma, saliva and faeces.

Abstract: Recent research has shown that domestic dogs are particularly good at reading human attentional cues, often outperforming chimpanzees and hand reared wolves [1, 2]. It has been suggested that the close evolutionary relationship between humans and dogs has led to the development of this ability, however very few other species have been studied [3]. We tested the ability of 24 domestic horses to discriminate between an attentive and inattentive person when choosing whom to approach for food. While the attentive person faced forwards, the inattentive person either stood with their body turned 180° away from the subject (body orientation condition), stood with their body facing forwards but their head facing away (head orientation condition) or stood facing forwards but with their eyes closed (eyes closed condition). A fourth, mixed condition was included where the attentive person stood with their body facing away from the subjects but their head turned towards the subject while the inattentive person stood with their body facing the subject but their head turned away. Horses chose the attentive person significantly more often using the body cue (n = 24, k = 19, p = 0.003), the head cue (n = 24, k = 18, p = 0.011), and the eye cue (n = 24, k = 19, p = 0.003) but not the mixed cue (n = 24, k = 13, p = 0.42). In an additional pilot study, horses were tested in an object choice task. A human experimenter cued one of two buckets by either tapping the bucket (tap condition), orienting their body towards the bucket and pointing (body and point condition), pointing while facing forwards (point condition) or orienting their body towards the bucket (body condition). If the subjects chose the correct bucket they were rewarded. Subjects were able to use the tap cue (n = 31, k = 21, p = 0.035), body + point cue (n= 31, k = 21, p = 0.035) and the point cue (n = 30, k = 21, p = 0.021) but not the body cue (n = 31, k = 11, p = 0.076). These results taken together suggest that domestic horses are also very sensitive to human attentional cues, including gaze.
Keywords:
social cognition, animal-human interaction, horses, attention attribution, domestication
1. Hare, B., Brown, M., Williamson, C., and Tomasello, M. (2002). The domestication of social cognition in dogs. Science 298, 1634-1636.
2. Gacsi, M., Miklosi, A., Varga, O., Topal, J., and Csanyi, V. (2004). Are readers of our face readers of our minds` Dogs (Canis familiaris) show situation-dependent recognition of human’s attention. Animal Cognition 7, 144-153.
3. Hare, B., and Tomasello, M. (2005). Human-like social skills in dogs? Trends Cogn. Sci. 9, 439-444.

Abstract: For ages horses and donkeys constituted the most important domesticated animals. Even though equids were often bred and kept in close proximity to humans, surprisingly little is known about their cognitive abilities. Traditionally, horses are not regarded as clever animals; common beliefs maintain that the horses“ behaviour is merely driven by conditioned-responses. Additionally, from an anthropomorphic point of view donkeys are believed to be ”stupid" animals.
Our study investigates the equids“ ability to recover a hidden object. For this purpose the animal has to create and maintain a representation of the object and its location in space, from the moment in which it disappears from direct perception, till the moment in which it reappears. The knowledge about objects being entities that continue to exist even when they are no longer available for direct perception is refered to as the well known concept of ”object permanence".
We primarly assessed the ability of Esperia's pony and donkeys to solve a Detour problem while employing an opaque “U-shaped” barrier. Each animal observed a food bucket moving and disappearing behind the barrier. Immediately after the object"s disappearance, the animal was released to search for the object. If it solved the task by detouring the barrier it was positively reinforced.
The ability to retain in memory the hidden object as well as its spatial location was subsequently tested in the presence of two, rather than one, screens (Working Memory testing phase). The food bucket was made to move and hidden behind one of the two identical screens, while the animal was watching it. Following a pre-established delayed period of 10 sec, the animal was set free to look for the food. In such a test the detour problem is combined with the classical delayed-response task, which is in use for the comparison of memory duration in different species.
In order to recover the hidden objects, animals must encode, maintain and correctly regain from their working memory the existence of the no longer visible object and its location from their working memory.
Both donkeys and ponies performed the Detour task showing to grasp the fact that an object which is no longer perceivable still continues to exist and can be regained. They also were able to correctly retrieve the goal object after a delay of 10 s in the Working Memory tasks, showing that they had encoded, maintained and correclty retrieved from their working memory the spatial location of the hidden object as well as its existence.
Nevertheless, when Standardbreds, raised in traditional stables, were tested in identical conditions to those describe for the ponies and donkeys, they could not succeed in the Detour tasks. The reason for such differences needs to be studied Even though it would be interesting to focus on handling differences, i.e., the ponies had been living in an environment rich in natural stimuli, while the Standardbreds had lived in a man-controlled environment since birth. It is also noteworthy that, like donkeys, the Esperia's pony have a reputation for being hard to handle.
Several considerations could arise from our preliminary investigations, and we will have the pleasure to leave them open for discussion.

Abstract: A number of studies have demonstrated that horses have chromatic vision and are able to distinguish colours from grey, independent of brightness. Anatomical and physiological data support the view that they are dichromats. In two experiments we provide behavioural evidence for their dichromacy. In the first we measured photopic spectral sensitivity. Using a spatial two alternative forced choice task, two horses were required to discriminate a coloured circular patch on an achromatic background from a blank field over a range of wavelengths . The obtained spectral sensitivity function was consistent with the presence of two cone classes with predicted spectral peaks at 429 nm and 545 nm, respectively. In the second experiment we obtained wavelength discrimination functions. The same animals were required to make discriminations between a series of standard wavelengths and a series of comparison wavelengths across the spectrum. The resultant wavelength discrimination function showed a single minimum at approximately 480 nm, which is also consistent with the view that horses are dichromats. Key Words: Horse, colour vision, spectral sensitivity, wavelength discrimination