At almost the same time, Peng Hung-chih began attaching a pinhead-sized camera to the head of a dog, to enable people to see the world from a dog's perspective. "Siao-Pai" (1999) is the first work in this series. In fact, it is not only a matter of showing the world from the perspective of a dog (we will leave aside the epistemological question of how the world perceived with the eyes of a dog can be represented, because that is not what the work primarily involves). In fact, the so-called "dog's perspective" simulates the dog's field of vision, which means that the camera is able to follow the dog's attention and movements and record things accordingly. At the beginning of the video, the dog's owner does everything possible to attract the animal's attention. In the background we hear him constantly calling "Siao-Pai, Siao-Pai. Isn't that what you love best?" The pictures alternate sometimes slowly, sometimes quickly, sometimes the movement seems to be frozen. What entices viewers is that they sometimes recognize a content in the pictures the camera records, but sometimes they simply have to guess what it is that attracts the dog's attention, what he sees or thinks. Here the so-called content of the pictures produces different, highly ambivalent meanings. Content can be anything that the camera captures, but it can also refer to the intention of the photographer. Of course, this ambivalence is also present in pictures taken by human beings, but it is much more strongly expressed in scenes recorded by a dog, because people can better guess the intention behind certain pictures due to their capacity for empathy and the scenic language based on a common visual experience. In comparison, the empathy between humans and dogs is not especially well developed (although, if we think of the relationship human - rat or with other creatures, then the discrepancy is naturally even greater). In any case, the language of pictures is still a very alien world for a dog - at least it still is today. (When will it be possible for dogs to obtain a camera and capture that which appears important to them in pictures?) Therefore, Siao-Pai naturally does not know that he is capturing something in pictures. In other words, the aforementioned "intention of the filmer" practically does not exist in these images. Thus the intention that we think we can guess is basically Siao-Pai's intention as "the one taking action" and not Siao-Pai's intention as "photographer". What is particularly interesting here is that if we want to recognize Siao-Pai's intention as actor, then a relatively certain way to do so is by observing what he is just doing or which expression he assumes. If we have spent some time with the dog and are familiar with him, this is a relatively simple matter. Here, though, since the camera is just mounted on the dog's head and we cannot see the dog himself, we are confronted with the second level of meaning: what is Siao-Pai doing now, what is he thinking now, what expression does he have? And consequently the first level of meaning must change accordingly: what relationship is there between what the lens depicts and the intention of the "actor" guiding the camera?