Chimpanzees can detect when others can and cannot see them. With one another, for example, they use visually based gestures mostly when the potential recipient is attending to them already. Indeed, if the potential recipient is not attending to them, they will sometimes walk around in front of her before gesturing. In support of these findings, Liebal et al. found in an experimental study that when a human who was facing a chimpanzee and giving him food then turned his back to the chimpanzee, the chimpanzee subject walked around him to face him again before gesturing. Povinelli and Eddy found that chimpanzees do not preferentially beg food from a human with uncovered eyes over begging it from one wearing a blindfold. This suggests that they do not know when others can and cannot see them. But Kaminski et al. modified this paradigm slightly to reflect more natural communicative situations, and they found different results. Their modification was that the chimpanzees did not have to choose between two human communicators – a very unnatural situation – but were always faced with only one communicator who was oriented in different ways in different experimental conditions. For example, in one condition the human faced the subject. In another, his back and head were both facing away. In yet another, his body was turned away as he looked back over his shoulder at the subject. The main finding was that chimpanzees gestured differently to the human depending on whether the human's face was oriented toward them, but only if the human was facing them bodily as well. Kaminski et al. argued, therefore, that body orientation and face orientation indicate two different things to an ape when it begs food from a human. Whereas body orientation indicates the human's disposition to give the subject food (i.e., when he is oriented so as to transfer food effectively), face orientation indicates whether the human is able to see the subject's begging gesture. This two-factor account helps to explain Povinelli and Eddy's negative findings. Interestingly and significantly, Kaminski et al. found that as long as the human was facing them, chimpanzees did not differentiate between the human's eyes being open and closed. This accords with the findings of Tomasello et al. that in gaze-following situations chimpanzees follow mainly the direction of the head as a whole, and to a much lesser extent the direction of the eyes.