Self-recognition in video

of a discourse with a trustworthy caretaker. The evoked mental images in such con-texts include properties of the real. The word “cake” does not get your mouth wet, but adding a few visualisations the concept “cake” might. A photograph of a cake, in a fully differentiated mode, would also not whet your appetite, but adding a few associations it might. In a less differentiated mode it might indeed make your mouth water, and in a full reality mode you might even bite into it. The whole concept of gorillas flooded Sherman when he found those photographs in a context where he was about to go for an outing. Happening upon a lexigram, or a stuffed toy gorilla, might not have had the same effect.

out that video seem to not only work as a “magical window,” but also as a “magical mirror.” One can just as well say that Sherman and Austin found out how video works, at least closed-circuit video. But how pictorial is closed-circuit video? There need not be anything “magical” about it at all since there is no need neither for dif-ferentiation nor reference when the feedback between the world, or body, and tele-vised image is instant. Their comprehension of recorded video is unclear. It is only reported that they easily differentiated between live and recorded imagery, and that they were more interested in the former. Exciting work remains to be done in this field.

Hirata (2007) studied self-recognition on televised media in chimpanzees. Out of the ten Kyoto chimpanzees observed only two displayed self-exploratory behaviours while watching their own image on live video. One of these, Chloe again, also ma-nipulated objects with a clear focus on viewing these actions in the monitor. It thus seems that more experience than the one provided by a single experiment is needed to foster an interest in video guided exploration in most subjects. This support the findings with Sherman and Austin who required quite some experience before they started to use video in this manner. Perhaps Hirata’s or Menzel et al.’s study would be even more interesting if live video would have been contrasted with delayed video. When the instantaneous feedback from the video is removed reality mode processing is potentially compromised.¹⁰⁹ In order to make sense of the video feed-back and use the image for exploration one must consider the image to be separated from one’s immediate actions, but still depict them. However, this would not only be a pictorial challenge, it would also tax one’s attention to the line of events that one is currently involved in. One must both anticipate where one’s change in movement will take one’s limbs in the spatial layout of the screen, and then update this anticipation with the actual video feedback. Only attending to either part of the task will result in aimless movement.

The depicting nature of delayed video can for example be tested as in Povinelli and Simon (1998). When they are viewing a video recording of events that took place three minutes earlier, 4-year-old human children, but less than half of 3-year-olds, have been found to reach for a sticker that had been covertly placed on their heads during filming of the event in question.

Law and Lock (1994) tested the spontaneous reaction of four gorillas towards their live video, as well as video of themselves recorded at an earlier occasion, and video of unfamiliar gorillas. The gorillas did not seem to show social behaviours to-wards video images of other gorillas with the exception of one possible social gesture.

One response noted during the delayed playback condition seemed to indicate that the recorded video was mistaken for live video. The viewing gorilla turned around and looked behind himself when he saw himself on screen approaching the spot where he was currently sitting. He could either have made sure that there was no gorilla approaching him from behind, or he might suddenly have recognised his en-closure in the video and looked around to check something. In either case it suggests

109 It remains an empirical question.

that he perceived the monitor more as a mirror than a window of sorts, but the evi-dence is inconclusive.¹¹⁰

The response to live video feedback was markedly different from the previous two conditions. Besides avid attention several gorillas showed signs of image testing and face exploration. Also stretching for objects without taking the eyes of the video screen was evident. The image testing gave the impression of exploring the “strange”

match between one’s own and the video image’s movements. Facial exploration con-sisted either of looking into one’s mouth, a potential that was discovered seemingly by accident during yawning, or feeling along one’s brow ridge which is an area that one normally do not have visual access to.¹¹¹

When watching a chimpanzee movie, Premack’s Sarah, suddenly after 30 minutes of calm, gets very excited by the capturing of a young chimpanzee in a net. Sarah hoots and throws paper towards the screen “seemingly aimed at the animal’s captors”

(Premack, 1976, p. 346). When the trainer who was present touched the captured animal onscreen “Sarah shuddered and turned a wildly startled face to the trainer.”

Rightfully, Premack evokes the problem of knowing whether the animal confuses reality and image at such occasions. But he concludes that rather than confusion, what is displayed is similar to when humans sob when watching a sad movie or ca-resses the photograph of a loved one. He furthermore suggests that one should view the depth of such participation as a measure of intelligence. Only an intelligent spe-cies can treat a scene as if it was real, and still not show all the behaviours that would accompany a real scene. Sarah did for example not flee the room although the scene seemed scary. Outward signs of participation are indications of internal participa-tion, he maintains. This is of course true for confusion events as well and Premack adds that an individual must also show that it can use pictures referentially, e.g. in communication, before confusion can be ruled out.

Premack might be right, but I think that the specific case that he describes is still a case of seeing television as a kind of reality. Sarah for example throws objects at the screen, which contradicts Premack’s suggestion that Sarah inhibits her reality based responses (all she is not doing is leaving the room). He also implies that Sarah knows that the television image cannot harm her, but still she reacts strongly when the caretaker touches the screen. The scene on television is thus seen to be somewhat differentiated from reality in that, while Sarah is very upset, she does not act towards it in an improper way, such as charging the screen. She rather takes into considera-tion that a television is a television and only throws paper at it. Perhaps what is lack-ing is the referential knowledge that the scene is far removed in time and space.

However, Premack would probably maintain that Sarah would continue to respond in the manner she did because her reason for acting was never confusion but in-volvement. Furthermore, her inability, or lack of motivation, to inhibit her re-sponses would be the same no matter what.

110 That primates look behind themselves when they see something in a mirror is not uncommon (see section 9.4).

111 A question for future research is whether apparent self-interest in front of mirrors is a result of image testing or vanity.

If Premack is right a prediction would be that Sarah could throw objects towards the screen and react with fear when someone gets too close to it also when a cartoon is played. Given, of course, that Sarah can comprehend cartoons. She would also be similarly excited when an upsetting story is read to her, given she comprehends lan-guage. I believe that Sarah’s reaction was not motivated by a need of hers to express her attitude towards certain acts in principle, but by the perceived realness of the situation and what was directly there, in front of her eyes. The reason humans cry when watching sad movies is not because they allow themselves this luxury despite that it is not real, but because parts of them did never make the difference in the first place. “Involvement” is when this part takes over. Reality mode gets a revival in such circumstances. So does Sarah watch television as the typical human, only that she is expressing her involvement in a chimpanzee manner, or is her way different? I agree with Premack (1976) that free response television viewing alone cannot answer this question, but further tests are needed to map Sarah’s pictorial competence.