From Robbins and Langton (2001)...
Literature has shown that it's possible to convey the full gamut of emotions in text. The real problem is that it takes a long time and a lot of talent to do this. Consequently, it's not that text doesn't have emotional clues, but it's so difficult to put them in. To that end, with text, paralinguistic clues are
/me crieson InternetRelayChat)
The choice of medium used for communication dictates how many and of what breadth the paralinguistic cues are. Face-to-face communication is has the most ChannelRichness?, getting worse as we progress through telephone conversations, e-mail, memos, bulletins and post-it notes. Since ParaLanguage is necessary to determine intent in ambiguous circumstances (was he joking? being sarcastic?), increasing the amount of ParaLanguage is crucial to bridging the emotional gap between people on an online community. In this way, FormOverContent is important.
Some (Rutter, 1987; Sproull & Kiesler, 1988) have argued that due to the impoverishment of text and audio communications compared to face-to-face communication, the absence of social cues will depersonalize the communication and increase antigonization. This could be a major cause of FlameWars, which is why it is imperative to AssumeGoodFaith in all textual exchanges.
Also see the related BodyLanguage.
Robbins, S. and Langton, N. (2001) Organizational Behaviour: Concepts, Controversies, Applications (2nd Canadian ed.). Upper Saddle River, NJ: Prentice-Hall.
Rutter, D.R. (1987) Communicating by telephone. Oxford: Pergamon Press.
Sproull, L. and Kiesler, S. (1988) Reducing social context cues: electronic mail in organisational communication. In Computer Supported Co-operative Work: a Book of Readings. San Mateo, CA: Morgan Kaufmann.
I have been a moderator/host/forum leader of bulletin boards and other collaborative discussion system since the early 80's, on Compuserve, GEnie, AOL, on IETF mailing lists, and in a number large web forums and wikis, so I have have considerable experience with both moderating flames between folks in our community, and dealing with flames directed toward me.
One piece of advice that I give is in regards to how emotions are amplified in the online text medium, causing a escalating cycle that can result in flames.
When I was doing some research on CSCW (Computer Supported Collaborative Work -- an academic term that predates the term groupware) in the mid-80s I ran into a study that talked about the emotional content of email. Their research showed that people misinterpreted the emotional content, typically strengthening it. For instance, if a message was written to expressed a mild pleasure by the author, the reader would interpret it as a stronger delight. If someone wrote expressing annoyance, the reader would think the expression was of anger. My personal belief as to the source of this escalation is that in absense of visual (face and body language) and audio (tone and intensity) feedback, our deeper emotional mind over-compensates. The evolutionary psychologist in me says that interpreting emotion accurately in groups was a survival trait for 100,000 years, but our emotional brain didn't evolve to accurately understand the emotions in the written word.
Some other studies in the psychological field showed that people respond to emotional content in communication not specifically in kind, but at level of intensity. If a conversation starts out unemotional, others will respond in kind. If someone breaks the wall and raises emotional language, other people will allow emotional language of that intensity to enter the conversation. The nature of the emotion (pleasure, anger, friendlyness) was not always echoed, but the if mild words were used, mild words would be answered, if more intense words were used it would give permission for others to respond at that level of intensity.
The combination of these these to, in my opinion, is an underlying cause for why flames and other forms of conflict seem to be so much more a problem on the internet then they are in personal groups. Basically, a vicious cycle gets started. For instance, someone posts something that has a little humor in it. This is misinterpreted as having some irony in it, and this allows the reader to respond with irony. This in turn is thought to be be sarcasm, not irony, and allows readers to respond at that level of emotion. Sarcasm is then misinterpreted as mild-insult, and then to insult, and then to anger. Now the emotional intensity is quite high, and we find ourselves wanting to respond at that level of intensity. Thus the flame.
I've had good luck in the last 15 years since I learned these concepts in using them as a meme to explain to faciliators and moderators who work for me what is going on. I teach them that certain words have more emotional content then you may intend. For instance "assume", "didn't", "should" and "forgot" are commonly used words that has a lot of emotional baggage. Sometimes even you should start with a "yes" rather then a "no" because "no" can have more emotional content when presented first. I try avoid "blame" or "responsbility" phrases in my words. I try to refer specifically to my own emotional state, whether as simply as using smilies, or with an explicit statement "I'm not angry, but just a little concerned". I try to make sure my staffers understand that they are responsible for their own internal emotional state, not that of the people they moderate.
This tips have helped considerably.
(Refactored partially into DampenEmotions.)
One interesting effect of ParaLanguageon? is that radio and television journalists have to modulate their voices in such a way that when recounting a report, they appear to remain objective and distant from the story. They speak wholly unnaturally, quite unlike a storyteller when telling a tale. However, they don't speak in monotone, since the human brain has difficulty differentiating the phonemes monotone speech. More to the point, it's boring to listen to monoton. Rather, they oscillate' the intonation of words up and down in ways that are not present in natural speech, but in what is considered to be journalistese. Listen carefully next time you hear a journalist on the radio or television, and you will hear what I mean. The intonational ParaLanguage that conveys emotional intent is like colour that connects chunks'' of words and ideas together (that is, it's higher level than clauses). This requires consistency and smoothness. The oscillation breaks the smooth flow, and thus disrupts the brain's ability to detect emotional quality, yet it is better than monotone since words come out clearly distinguished from one another. Quite a mindhack. -- SunirShah
--- Has anyone tried to completely codify paralanguage? This would be very useful in dealing with people of autistic natures. It could help explain alot of things to people. If I were to do it although, I wouldn't really know where to start --MahyarMcDonald?