Ito Go says that he first realized the distinction between 'character' (kyarakuta) and 'character icon' (kyara) when he heard a comment of girl fan of Nana, a recent big hit in girls' manga, which has also recently been made in to a film. She pointed out, 'Characters are prominent in Nana but character icons don't stand out'. I think this very clearly illustrates the difference between the two terms.

 If we briefly point out the difference between these two, we can say that 'character' has a certain kind of embodiedness and is often depicted in a realistic form that has a sense of actual existence. In contrast, 'character icons' are drawn with simple lines, like Doraemon and Tamogocchi, and do not seem to be real or exist in actual life, though they do have a vivid sense of being very much alive. Reality in realism is about how closely reality can be depicted. The more details depicted of actual experience, as in an autobiographical novel, the more realistic the representation. The reality of 'character' is the same as reality in realism.

 Nana aims to capture the sense of reality of today's ordinary young girls. The readers of this magazine are a little different from the usual manga fans. They are the kind of fashion conscious girls we find a lot in Shibuya who do not usually read manga but do read Nana as an exception. They reject ordinary girls' manga because the character icons are too prominent and unrealistic. Character icons are realistic as fiction but not in the sense that they are familiar characters in real life. The reality of characters lies in their life-like nature. On the contrary, character icons such as Doraemon and Atomu support the reality of fiction.

 The manga critique, Miyamoto Hirohito, defines character as follows. Characters are unique and possess autonomous and pseudo reality. For example, they can change by growing up, and, like human beings, they are multi faceted, have emotional struggles, complexity, and opacity, giving the impression that they have secrets which have not all been revealed. They have qualities like people such as multi-layered psychologies. A realistic depiction of a person would usually mean this kind of representation.

 Ito Go skillfully points out the difference between characters and character icons by taking up the opening picture of a manga called 'Masachan no boken' (Masa's Adventure). This picture presents a realistic depiction of the real boy, Masachan, and Masachan as a character iron is drawn with simple lines in a back frame. This indeed illustrates the difference between character and character icon.

 A critique Yomota Inuhiko makes an interesting point. He says, 'The same face is not drawn twice in manga, but the same face can be drawn in limitless ways.' Once we become familiar with character icons such as Atomu or Doraemon, even if the same face cannot be represented each time, we can identify the character icon in any drawing. We might say that a particular feature of character icons is that it has this kind of versatile transformative nature as in the case of icons.

 According to Ito's definition, a character icon is basically drawn in relatively simple lines, and reveals its presence as a personality by being represented by a unique form or being expected to be represented by a unique form. It lacks the sense of embodiedness of human beings. It need not have a body. It also lacks an intimate relationship with a story. The existence of character is dependent on a story, but character icon does not have such intimate relationship with a story.

 In this sense, character icons exist independently so it is easy to create parodies of them. As long as we keep their characteristics in mind, one character icon is compatible with another and we can have them appearing in all sorts of stories.

 In this sense, Mickey Mouse, for example, is a character icon. So we cannot say that character icons are uniquely Japanese, but I think there is a strong character icon culture in Japan that creates character icons and consumes them on a mass scale.

 For instance, each government department has a symbolic character icon. NHK (Japan Broadcasting Corporation) has one, and the police have one. Character icons are created each time there are events such as local revival movements. This is what Miura Jun calls 'yurukyara' (vaguely cute characters). I don't know whether we can call this particularly Japanese, but I think there is a kind of cultural trait underlying this kind of phenomena. Incidentally, in terms of semiotics, characters are similar to symbols and icons, but at any rate they are strange signs that are not easy to classify.

 Character icon culture, which is very much related to manga culture in a sense, is also influencing various other media, including of course animation. For example, in the field of light novels, the author firstly sets up a character icon, and in some cases gets an illustrator to draw an image of it and then think about how the story will go. I think the fact that techniques originating from manga culture are used in all sorts of media shows that the flexibility particular to manga culture pervades and stimulates creativity in all sorts of other media.

　

 Next, let me talk about 'high context space'. Context means a situation. It refers to background information which determines what significance a received stimulus has. It has a kind of continuity and holism. For example, the meaning of a word is fixed when there is context behind it. The meaning of the word 'baka' (stupid), for instance, can become an insult or expression of love depending on the context. In fact, I think there is another important link between Japanese culture and manga culture here.

 The term 'high context' was introduced by an American anthropologist called Edward T. Hall. He was fond of Japanese culture and strongly proposed that Japanese culture is high context. America is a multi-ethnic country and hence American culture is low context in the sense that since there are few shared codes, communication is not possible unless you pass on information in a way that is so clearly coded and attached with well-defined meaning that it cannot be misunderstood. You won't be understood unless you say yes or no clearly.

 On the other hand, Japanese culture is high context so many cultural codes are implicitly shared. Therefore, everyone shares the background information by tacit understanding and insiders' knowledge so to speak, and a great amount of information can be transmitted by just a little hint and tone, like knowing ten things if you are told one. Hall praises Japanese culture from the point of view of efficiency of information exchange which is possible due to its high context culture.

 However, there are opinions against this view and the film director, Itami Juzo, argues the opposite. He employs the same theory to point out the problems of Japanese cinema. Hollywood movies, he says, must apply codes that can be understood by everyone, so the representations are clear and easy to understand, and capture international fans. He says that Japanese cinema relies on high context culture, so it appears as incomprehensible and impoverished representation to foreigners. For example, it takes for granted that a particular feeling is shared when a four and half tatami-sized room appears on the screen. This argument seems reasonable to some extent but I do have my reservations. In fact, I think high contextuality can also be universal. The problem rather is that high context films can more easily fall into stereotypes and Japanese cinema contains too much of such stereotypical patterns.

 Leaving aside arguments about cinema, I think there is an order in context representation according to genre, as follows: animation, manga, television, cinema, photography. Photography is at the extreme end of low context and as you move towards manga it becomes more high context. For example, somehow something is communicated when someone says, 'I saw anime yesterday' But if someone says, 'I saw photographs yesterday' the quantity of information communicated is extremely small and you would not know what photographs the person is talking about. We can say that the more mass oriented the representation, the more likely that it will be high context. The genre and composer/author of popular music or manga can be guessed by just hearing or seeing a fragment of it, but it would be more difficult to do so in the case of classical music or art.

 There is higher contextuality when there is smaller amount of information per picture. To give a rather extreme example, the information content in animation can be so little that a scene with just one line drawn in a cell picture can exist by itself. This is possible because the picture is part of a well defined and detailed context of a particular animation. There is a general tendency to value photographs as being of higher quality when there is a rich information load, though of course there are exceptions like Sugimito Hiroshi's photographs.

 We can say that McLuhan's term 'cool media' refers to media with a small amount of information to communicate. Cool media tends to be very high context. We can see this when we compare Japanese manga and American comics. American comics are indeed low context representations. A single frame of an American comic contains densely drawn pictures and can be presented as a piece of work. We can know what is being depicted by just seeing a single picture. In contrast, we cannot understand Japanese manga by just looking at one frame. A single frame of Japanese manga can hardly be an independent representation in itself, though there are some exceptions in recent years. But we can guess the genre and the author by just looking at one frame, so this is indeed an extreme example of high contextuality.

 As a result, high contextuality of manga enables it to be read very fast and makes it possible for the content and pictures to be very closely connected. For instance, sudden changes between comical and serious representations have become a feature particular to Japanese manga and commonly used after the 1980s. Since the context is already defined in detail, even if the main character has a serious face in one scene and is suddenly depicted in the next as a large headed character with the size of the head as big as the rest of the body, we can read it without any sense of it being out of place. This representation is indeed unique to manga. Animation also applies this technique. But this is impossible in cinema, let alone novels. This kind of representation is indeed possible due to high contextuality of manga.

 I think the greatest contribution of manga representation is that it allows the reader to get totally immersed in the work and facilitates the transmission of emotions rather than logic. In the term used by the psychiatrist Luc Ciompi, there is a kind of 'affect-logic' at work. This logic differs from the ordinary chronos type logic and perhaps it is an aspect that no one has analyzed sufficiently. I think we can understand this kind of emotional logic by analyzing grammar of manga expression in detail.

 There is an interesting manga called 'Saru de mo egakeru manga kyoshitsu' (Manga class for Monkeys) that imitates a manga introduction book. I often refer to this manga to explain the high contextuality of manga. This manga parodies the different representations and styles of various manga magazines. I think those of you who are familiar with boys' magazines would understand this humor.

 For example, if the story of 'Momotaro' (boy born from a peach in a popular Japanese folktale) were to be depicted as manga in the magazine Shonen Jump, it would become 'Saiba senshi momo' (Momo the Cyber Warrior) with a muscular hero like Kenshiro thrashing powerful enemies one after another. In Shonen Magazine, the story would be sporty with a touch of love comedy, in Sunday it would have a light sophisticated touch in animation style, and in Champion it would be a juvenile delinquent and yakuza manga entitled 'Bad boys, momotaro gang'.

 Thus this is what I mean by manga's high contextuality. As you can see, it is overloaded with context. Just by looking at the picture for a moment, you can guess the magazine that publishes it, its genre, or even how the story will turn out and the name of the author. We can deduce all this information in an instant by looking at just one frame. This would be extremely difficult in other media of representation, apart from writers and film directors who have very idiosyncratic styles. To repeat, the particularity of manga representation is not in the combination of pictures and words but in the high contextuality that arise as a result of chaotic combination of pictures and words.