There are many nouns to describe assemblages of people:

Whyte discusses perhaps the most common type of crowd (at least for we who live in and about a city): The street crowd. His observations seem so obvious when we read them, yet we take them for granted whenever we interact with a crowd on the street. In a sense, crowds are based and depend upon, and indeed create (in a sense) a language. There is the language of the clothes people wear, the fashion (which we have already discussed). The public nature of the type of crowds Whyte studies creates a forum, or at least a stage where people can show off their own fashion and view that of others. More interactionally, there is the speech between people within and about a crowd. Then there is the language of the superstructure of the crowd: how does it move? what are the flow patterns? where are people assembling? where are the points of congestion and what is causing them? We become very adept at reading the crowd, and understand (usually) how to move through it or become part of it. The crowd clearly fascinates us, to which both Whyte's people watchers and Milgram's theoretical mob members (p. 265) attest.

Milgram's familiar stranger, though not necessarily a crowd phenomenon, does point to a fundamental social process that exists whenever people gather on a regular basis. A familiar stranger doesn't necessarily need a long time to gestate; even at a party or singular gathering, a person may take on this familiar role, becoming a meter by which one can judge the social situation.

Milgram's other work, in outlining a number of theories and frameworks for how and why crowds form, touches on ideas of swarming behavior, flow dynamics, and involuntary social reaction. Its most basic assertion is that crowds are a fundamental element of social interaction. Whenever a large enough number of people get together, certain group behaviors manifest themselves. Though there are still no definitive theories for why this should happen, the basic fact that large groups of people do exhibit crowding behavior--though not necessarily with all of the connotations the word "crowd" implies--is reason enough to explore how crowding manifests itself in the disembodied environment of the online world.

In the real world, interaction with a crowd, either within its bounds or beyond them looking in, is about filtering and sorting perceptual information and sorting through it. Interacting with a crowd means having a purpose for that interaction, whether its to find a friend in a large group of people, to listen to what a speaker is saying, or to get from one end of the room to another. Even just standing in the crowd means becoming part of it, giving yourself up a little to go along with the larger being of the crowd. Looking in, you might be interested in merely seeing who is there, what people are wearing, or just entertaining yourself by people watching. The stimuli that a crowd generates, be it visual, auditory, or olfactory on a personal scale, or the larger scale flow patterns and arrangement or people is enormous.

To interact or observe the crowd, we must filter most of this information to get at the small bits that are of use. Looking for a friend in a crowd, a visual search, causes you to scan peoples faces or body types quickly trying to find a match for a model you have in your head. You must move through the crowd efficiently while keeping your eye on everyone that's around you. Interestingly, if someone is present in a crowd of a searchable size, it becomes relatively easy to pick them out (as individual experimentation will demonstrate). Trying to guess the identity of a person while people-watching is an entirely different prospect. You give a particular person, someone who has caught your eye for some reason, a good deal of attention, and take in their entire visual representation, including what they're wearing, what they look like, how they talk (if they are talking), and how they move. In previous readings we have seen how people are surprisingly good at predicting personal information about a person based on these factors.

Online, we are at a great disadvantage. There are certainly large assemblages of people that would in real life be called a crowd. However, all of those well honed abilities that we have developed in RL rely on perceptual information that is unavailable online: view of the person, view of the crowd, the sounds the crowd generates, a smell that might permeate the group. The most we see is a large list of names, and only when we log onto a chat room. Nelson's work on visualizing a crowd on a website moves us a little closer to a visual representation for an online crowd, but we are still missing a truly large scale group. For example, when Steve Jobs delivers the keynote for MacWorld, there are tens of thousands of people that are simultaneously watching the live broadcast of his speech via webcast. Yet this group of people, which in real life would constitute a crowd, has no sense of its own structure or even existence as it would were all of the people in the same room.

There seem to be two fundamental components to the crowd: the perception of it, and the interaction with it. Perception, as discussed briefly in the previous paragraph, deals with a person's intake of stimuli generated by the crowd and the people within it. Fundamentally, this is absent online, since there is very little portrayal of others online. In some sense, this information is available, and has always been. The 'w' command on a multiuser system will generate a list of everyone that's logged on and their current activity. But knowledge of other individuals isn't enough. Milgram's work shows that there is a self awareness to the crowd. People know that that they are in a crowd and that crowd has certain macroscopic behaviors that define its existence. Perhaps a crowd knows itself because individuals understand their existence within the greater whole. Providing this locational information--both visually and psychologically--may be enough to give people a sense of being in a crowd online. Some work with web presence does this, but on too small a scale. If you were able to see yourself within the sea of visitors to yahoo!, there may be enough visual volume to give the impression of a crowd. The commonality that Milgram claims is fundamental is missing in this scenario, and perhaps crowds could only develop where people actually share something more tenuous than just typing in a web address at the same time.

Interaction is a more difficult proposition, since you must be able to not only control yourself in the crowd, but also allow the crowd influence you. This two-way interaction then causes the larger scale group behavior that we see in Whyte's and Milgram's pictures. Unfortunately, most of the influences on crowd behavior Whyte discusses have no direct analogue online. Further, the influence of a crowd on an individual in the real world is direct, whereas online it can only be indirect, such as moving an avatar. Even with haptic devices, the experience is only partial. In a real world crowd, the entire body interacts with the whole. Online, only the hand does, and that is disjoint from the eyes which see the effect on the screen. But perhaps this direct interaction isn't required, or at least only in some circumstances.

page last updated 3/27/2001