The Web is a curious place. A land populated chiefly by cats and porn stars is also a land that helps to overthrow governments, spread awareness of important social and societal issues, and ensure that nobody has less than 200 "friends". It is also the land of text. For as much as some decry CN as outdated for its entirely text-based appearance and game play, the vast majority of communication on the web takes the form of text. Written language reigns supreme in the realm of interconnected microprocessors, speech is relegated to second place. However, this yields an important side effect: the !@#$%^& effect.
Humans by and large are risk averse. As a species, we tend to avoid risk because where there is risk, there may be death. And life must find a way, or whatever BS Jeff Goldblum spouted in that small-time sci-fi indie movie. But it wasn't all BS: life, as a general rule, prefers to continue on living, and those that are living will do almost anything to remain living. So of course living beings tend to avoid anything that can get in the way of this drive. (Don't worry, there will be a point soon, I promise.)
Given that as true, enter the !@#$%^& effect. The !@#$%^& effect states that we tend to, in the absence of contravening evidence, assume that others may not have the best of intentions. We may not necessarily assume they are #$%^&*!+, but we definitely don't assume they're angels either. We go with !@#$%^& over angel because we have nothing to fear from angels, it's the #@$*&(!+ we have to worry about, and if we wrongly assume that an !@#$%^& is an angel, we may be in for some serious trouble. The problem lies in where we find most of this contravening evidence: body language and tone of voice. The content of communication rarely offers enough insight to contravene the immediate assumption of less than beneficent intent. The largest exception to this is in fiction, where we are not generally interested in the author's intentions but in the intentions of the characters, and since they are in a whole constructed world, their intentions can be made clear even within the confines of text. In the real world though this usually is not the case.
The internet thus finds itself with a problem. Even in cases where another has good intentions or is being entirely genuine in what they are saying, we have a hard time assuming or accepting this because we don't have enough evidence to the contrary. We can't hear their tone and we can't see what they are doing while they say it. Hence the !@#$%^& effect runs rampart, which makes genuine communication over the internet very difficult. Since this game is played almost entirely through text, the !@#$%^& effect runs this world, for better or worse. We can attempt to combat it by asking ourselves what intentions we would have if we said what another has said. Run the test of reasonableness (ToR) and see if the intentions we are ascribing to this other person make sense in the context. We must assume not that the other has less than beneficent intentions but rather that most here are reasonable people and it is unlikely that their intentions in given contexts are all that different from ours. We must do this because while the !@#$%^& effect is actually helpful off the web (since we rarely lack contravening evidence where it exists, and thus when the !@#$%^& effect takes hold, we are likely correct in our assumptions of intent), it is nearly useless here because we are almost never at risk of losing anything significant. This is a place where it's usually OK to give another the benefit of the doubt.