WRT: Writer Response Theory

Do you know the difference between emoticons and kaomoji?

Smile  (emoticon)   :-)
  Bow  (kaomoji)  m(_ _)m

Emoticons are sequences of printable characters intended to represent faces. (In some software like instant messenger clients, they also include small graphic icons substituted for the same purpose, but here I’ll be talking strictly about digital text emoticons.)

Within digital character art, emoticons exist at the limit where characters stop representing word-sounds and begin representing images. They are an iconographic shorthand just over the border from text and into ASCII art.

Kaomoji, (顔文字; literally, “face mark”) are a japanese lexicon of emoticons. Unlike emoticons however, kaomoji are primarily vertical rather than horizontal, and based on a rich component grammar, and this richness brings them much closer to ASCII art depictions of faces than the iconic shorthand of a smiley. The variety of kaomoji is staggering, and many make use of mixed western characters, hiragana, katakana, dingbats and more.

The Wikipedia article on emoticons gives a fascinating historical overview with numerous examples - I won’t attempt to summarize it all here. Their key distinction is between “Western style,” which are read at a 90 degree angle, and “East Asian style,” (kaomoji) which are read vertically.

Because kaomoji are rotated 90 degrees from their western counterparts, some term kaomoji “verticons” - emoticons lie horizontally, while verticons stand upright. Yet the primary difference of verticons is that their new orientation is used to take greater advantage of horizontal space. Not only does using the horizontal line allow for greater articulation, as in this excited person waving with arms in the air

\(^o^)/

but it allows multiple representations to be strung together easily into coherent scenes, such as this one in which two people offer to hug a crying friend.

(>^_^)> ( ;_;) <(^_^<)

These examples come from a subset of kaomoji/verticons sometimes called “Anime style” - an adaptation of the techniques of kaomoji to the much more graphically limited single-byte encoding of Western alphabets. Different native character sets make for different character art possibilities and aesthetics. However, while important, alphabetic variation is only one part of the picture - many emoticons are non-alphabetic in nature, and most styles exhibit heavy use of simple punctuation marks used across disparate languages.

As general computing continues to move towards the pervasive availability of unicode, will emoticon languages become as rich as the breadth of unicode makes possible?

Perhaps, but my impression is that they will remain largely limited to the native language keyboard set for two reasons - firstly, because emoticons are a practice of typing, and secondly, because the aesthetic of emoticons is a certain kind of cool that involves repurposing. If you make of use characters that aren’t commonly available (i.e. dingbats), it seems like cheating.

What is the distinction between emoticons / kaomoji and the larger world of ASCII art?

For me, the crucial distinction is what works at the command line. All of these examples can be typed serially, and thus incorporated in a speech-like fashion into communications like streaming chat. By contrast, multi-line ASCII art arrangements occur at the bottom of email messages, but have a very high risk of failure when transmitted over chat: the images may be obscured before it can scan correctly, not only by font, offset, and wrapping, but also by the high likelihood of interruption. Rather than being composited two-dimensionally, like graphics, emoticons/kaomoji are written linearly, like text. A two-or-more line formulation becomes ASCII art by definition.

This is not just a technical limitation, but a symptom of practice. Even as multiline chat software becomes common, emoticons will probably remain single-line due to the way rapid typing occurs as a linear process of composition in time.

In Understanding Comics, Scott McCloud argues that iconic abstraction and comic representation occur at the midpoint in a spectrum of representation that ranges from photorealism on one hand to the printed word. I’m tempted to arrange arrange ASCII art, kaomoji, emoticons, 13375p34k and chat acronyms on a similar spectrum of digital text representations. One important difference would be that, with McCloud, the visual media (photography, painting, line art, typescript, calligraphy) changes at each step. Here, the medium (digital fonts) is invarient, with only arrangement and context changing the way interpretation occurs. Thus

brb my =^.^= r0xx0rz teh @-}— again =D

might be roughly translated as “Be right back - my cat is attacking the flowers again, haha!” ( although this kind of total mashup of text dialects is unlikely in real chat transcripts). Yet it is possible, and where an effect is possible, someone somewhere is busily turning it into art - if you consider text substitution filters to be art.

[kaomoji discovered via my.bicycle]