It is not only in posing or appearance that emotions or moods are presented graphically. Sometimes the emotions take shape external to the character, and are projected in diegetic space. The elements we will discuss in this chapter belong to another set of highly coded elements. They differ from other visual thought projections in diegetic space because of their standardization as symbols. We find them both in- and outside of thought balloons, often near character’s heads. Examples are: sweat drops flying from the head of an embarrassed character, stars from a hurt character, steam clouds from an angry character or a light bulb from a character with an epiphany (check the entire Gaston comic for many examples). Obviously, these elements are not words, but they are not true images either.

Metz names five facets in which images differ from words:
- ‘Film images are […] infinite in number’;
- ‘They are in principle the invention of the speaker’;
- ‘They yield to the receiver a quantity of indefinite information’;
- ‘They are actualized units’;
- ‘Only to a small degree do they assume their meanings in paradigmatic opposition to other images that could have appeared at the same point along the filmic chain’ (26).

Now, light bulbs and stars have their variations, but are definitely not infinite in number. They are not inventions of the speaker, are only halfway actualized (they are not truly ‘statements’, like normal images), and are heavily dependant on context for their meaning. That is why they are symbols, somewhere between words and images. In Japan, these emotive symbols developed independently, up to the point where some are actually unrecognisable for readers only familiar with European and American conventions – stressing the importance of context for their understanding. A bleeding nose in manga usually means a character is sexually aroused, for example. Like thought clouds or speech balloons, neither the bleeding nose nor the stars and the sweat drops exist in diegetic space for other characters to see (although many several meta-textual or, if you will, postmodern jokes refute this basic assumption). Emotive symbols can fill the entire background of an image, for example with flames of anger, or lines of shock.

The source of emotive symbols is a single character, and they stem from his or her feelings, so they are examples of subjective narration. The fact that they are on the border of word and image makes it hard to say whether they are direct thought and speech (they occur in speech clouds too) or emotive focalisation. What is more, we can also argue that they fall in the same category as exaggerated posing and appearance. The narrator does not express subjectivity through them because, far from being individualised and personal, emotive symbols are a standardized set, highly bound by conventions. Other characters do not see the excessive sweat drops on Gaston’s head, but neither does the character himself experience his anxiety like that. Thus, we could also conclude that the sweat drops and other emotive symbols are inferred by the narrator, to express vividly what the character should be feeling. Recognising and analysing symbols remains a complex matter.

It is important here to note again that the conventions of posing and emotions are not only typical of comics, nor are they only possible in comics. Many of the observations made here hold true for animation as well. Even the most outrageous examples of deforming characters in screwball manga are literally used in similar anime, for example.

Conclusion

Like film in its early, mute days, comics had to include written language in their medium. Trying to compete with writing’s richness was nigh impossible, as it would make comics overly verbose. However, comics developed its own richness with written language, by allowing the letters to integrate with the images. The result is a highly coded medium with high expressive potential. The marriage between written language and images may not define comics as a medium per se, but it has become one of its most interesting and widely recognized features. Subjective narration in this medium can only be understood from a thorough understanding of comics’ conventions. The thought cloud is the most well-known form of subjective narration in comics. Trying to study just the image or just the written language is not enough to understand devices like it. Even looking at comics with the two approaches of word and image combined is not enough, as the sum amounts to more than its parts.

The integration of word and image is not solely a shift of the first towards the latter. Some images have been symbols, and become bound by conventions as representations of representations in a language-like way. The human figure itself becomes a symbol for that character’s attitude, or mood. We are so accustomed to many of these that we tend to forget we really read comics, and that we had to learn this before we could do so. A first-time look at manga’s alien emotive symbols is the best affirmation of this assumption. Standardization and conventions did not lead to an evolutionary dead end; continuous meta-textual play and evolution actually opened a wealth of possibilities. The word-like quality of the image gives comics a certain kinetic energy and dynamism. The grotesque or simplified look of characters may seem an instance of subjective narration, but it is usually not. Only when the image is explicitly attributed to a character’s point of view can we take an appearance as subjective narration. Otherwise, this appearance is the view of the narrator on a particular diegetic world. The case with emotive symbols is harder to crack, but we will hold them as instances of focalisation. Note that we have only dealt here with the parts of the image that have attained a linguistic sense, usually in the way characters are depicted. Besides that, the image is used to convey subjective narration a lot, as we will see when we come to discuss transitions from image to image.