Glacial Erratics

Social Software As Tool

May 26, 2003

The people at HeadShift have released a paper called Smarter, Simpler Social that's being discussed here and there. It's an overview of SocialSoftware that reviews what's different in that arena and what must be done. It starts with a review of the current state of affairs and points out some future directions.    (0000RR)

I was happy to see that it cited my ComputersAsTools paper, referring to the way in which future computing applications will need to augment human communication rather than automate processes. I agree with this and for the most part I think the paper is quite good, but I think their analysis of how we got where we are today misses some important points.    (0000RS)

The paper argues that it is the failure of existing software to adequately model human process that created the need for software that is more social. While this is certainly true in part, it misplaces the emphasis and repeats the common error of overemphasizing the technology over its users.    (0000RT)

There is an enormous mass of existing software automating many processes of data-transport and repeatable computation. That software is performing so well that, for the most part, we don't notice. Content (transported as raw data) makes its way from server to client and between peers in such a reliable fashion that when in the rare chance it doesn't work, it is a real disruption.    (0000RU)

Automation software has provided an infrastructure that lubricates communication. As communication becomes easier, we desire to improve that communication: make it work better and be more fulfilling. Now that we have automated the process of getting the data that represents communication from one place to another we need to augment how we are able to manipulate the communication.    (0000RV)

SocialSoftware is a new name on an existing concept. From ComputersAsTools:    (0000RW)

Landauer distinguishes between two phases of computer applications. Phase one applications automate tasks “replacing humans” for the performance of “almost any process that science, engineering, and statistics have captured in their theories”. Phase two applications, on the other hand, are applications that assist humans in tasks for which there is no established theory of action. Phase two applications include the very large body of office productivity applications, web browsers, and desktop operating systems; anything where the human uses the computer throughout the process. They are the applications we use to process information in flexible and potentially undefined ways.    (0000RX)

Phase one applications are close to ubiquitous and are fading into the background. We now have the infrastructure to work on phase two applications. Or, if you prefer, SocialSoftware: applications humans use to manipulate language, to communicate. Communication is not a process that can be fully captured in theories.    (0000RY)

Personal computing devices, especially devices that are highly available in both time and space (laptops, wireless PDAs, cell-phones, etc.), are promising tools to help achieve the view of computers as language representation manipulators provided by Winograd and Flores in Computers and Cognition. It is only very recently that the network and networked communication tools have become pervasive, reliable and cheap enough to enable social software. The communication tools run on top of a network that is automatic and transparent: it has faded into the background and for the most just works. It is stupid.    (0000RZ)

The HeadShift suggestion that social software be adaptable is appropriate but it is important to distinguish between human adaptability and software adaptability. Too often people want computers to act like people but they are not the same and it does us no good to hope for such a thing. Computers are tools and something to be utilized, worked and taken advantage of (lot's people seem to feel the same about humans; they can rot).    (0000S0)

If we think of computers as acting like people, and having intelligence like people, we grant it intention (from ComputersAsTools again):    (0000S1)

When the computer is viewed as having intention “the personification of the machine is reinforced” (Suchman). The interaction between the user and the computer is the locus of negotiation for performing the task. The computer takes a privileged stance, above the task. When in that stance we expect the computer to truly have, given the intention we have granted it, the intelligence, inferential power and adaptability that Suchman says we expect in social interaction. This is unfortunate because the computer is not intelligent; it cannot compare arbitrary and dynamic categories. It has no true and general inferential power; it cannot create links between categories. It is not truly adaptable; it can only create new classes of distinction according to a limited rule set. The expectation of intelligence sets up a poor mental model of the real situation.    (0000S2)

Humans, on the other hand, are very adept at comparison and linking. SocialSoftware should augment this behavior by providing tools that ease search, discovery, browsing and linking.    (0000S3)

Human adaptability comes from our ability to work with the truly unexpected and do something smart with it. Computers need a little more guidance. Therefore an adaptable piece of software is not one that is smarter as HeadShift suggests, but is simpler in the sense that it makes clear what it can do. Adaptable software is software that can be adapted by someone or a group, not that adapts itself. It is software that people may use in unintended ways. These adaptations are allowed by at least two factors:    (0000S4)

Clear and simple affordances    (0000S5)
Adaptable software makes it clear what it can do, what inputs it can accept, what outputs it can produce and provides handles to doing more.    (0000S6)
Replacability and loose coupling    (0000S7)
Adaptable software is a replaceable link in a chain of tools. Replacing the link can change the function or enable trials of different tools. Replacing a link does not break the rest of the chain. Loose coupling at the application level is good for the same reasons that loose coupling is good in programming: it encourages and enables the flexibility necessary to benefit from inevitable change; it allows unintended uses.    (0000S8)

These factors allow interoperability, which is a complicated word meaning talking.    (0000S9)

Social software is about talking. Two characteristics make adaptable software social software:    (0000SA)

People    (0000SB)
people, in groups loose or tight, use social software. It's as simple as that. It's not the computers that are being social.    (0000SC)
Reference, reuse and notification    (0000SD)
Social software allows people to talk about what people are talking about, reuse what people are talking about and tell people that people are talking about what they are talking about.    (0000SE)

Blogging and the tools it uses, especially those that allow RSS syndication and TrackBack, are good examples of social software. They help to create groups without a visible system of central arbitration and control. Groups are created by the participants rather than something they join that mediates their participation. Participants exist in the group with a sense of identity and ownership over their content while still allowing reuse and reference of what they have created.    (0000SF)

There's no doubt that the SocialSoftware phenomenon is an important development. It deserves some of the hype. However, care must be taken: the exciting part about social software is people not software.    (0000SG)

Comments

1/1
On May 26, 2003 09:13 AM Lee Bryant said:

Hello and thanks for linking to our draft paper (Smarter, Simpler, Social). I am grateful for the feedback because the paper was only circulated to people quoted in it for their feedback. I will try to find the time very soon to take on board all the feedback and produce a release version of the paper for circulation.    (0000T8)

I think we probably agree more than you realise, especially with the key statement "the exciting part about social software is people not software." Whilst our paper seeks to produce some technology recommendations (because we are looking for practical ways of influencing the work of our peers), we are far more interested in the application of social science to technology than the other way around. By way of illustration, you may have heard of an attempt to create a Social Software Alliance, which had its first main get-together at the Emerging Tech conference in April. If you ever see a transcript of that meeting then rest assured most of the swearing is from me - a reaction to what we regard as the absurdity of trying to establish technical standards for social software before we have even worked out what the hell social software is supposed to be!    (0000T9)

Sending...