Tom Coates writes about "social software":
I love working in it but I'm scared of the way people are talking it up and I wish people would build more brilliant things rather than talking about it. [...] It's not like it was with my other baby. Weblogging grew gradually and properly and organically through the interactions of real people. This one's being increasingly owned by the wrong people.
I suspect (well, I hope) that Tom didn't mean that last bit about the "wrong people" as a circling of the wagons among the early adopters. Who are the "right" people to be talking about all this? (I'm not sure anybody should be "owning" it, really.)
I do understand Tom's fears about the way the current discussions on "social software" (I keep using quotes because it still feels like too amorphous a concept to be a solid term just yet) reflect the pre-bubble hype about the web.
On the other hand, most of the hype I see right now is not about social software, but about weblogs. And just as the recent Pew survey shows what a small percentage of internet users currently use (or as Clay pointed out on a list I'm on, recognize that they're using) weblogs, a far smaller number of people are actually talking about social software these days. It was pretty easy to build my neighborhood list, because the arena's far from saturated with meaningful voices right now. Perhaps it's that "echo chamber" effect that makes it seem to Tom like the volume's been turned up on the discussion. But I'm not convinced.
In his earlier essay on The Excesses of Social Software, Tom wrote that "There seems to be a bizarre lack of history to the whole enterprise - a desire to claim a territory as unexplored when it's patently not."
I guess I'm not seeing the blindness to history that's worrying him. Instead, I'm seeing the opportunity for those of with knowledge of the history of (and, I might add, the research into--which is not nonexistent) CMC ("computer mediated communication") and social contexts. This is a field that I've been watching since the late 1980s. It's still a pretty small niche, but there are a reasonable number of smart people who think, write, and teach about CMC (which, "back in the day," was the term used for what's now being called social software).
The people I see most involved in the discussions right now are adding a lot to the conversations. The folks in my earlier "neighborhood" post are all having an impact on my curriculum development, for example. And the end goal of that curriculum is to turn out people who can build the "brilliant things" that Tom wants.
What we shouldn't have, in my not-so-humble opinion, is a lot of product development happening in a vacuum. Many of my students have the technical skills to build amazingly cool things. But they don't understand the context in which those things need to operate. I want them to read these conversations, I want them to participate in them. I want them to ask questions, and see the questions other people are asking. I want them to learn the history, see the mistakes and successes that have already happened. And then I want them to build the brilliant things that answer questions and solve problems.
So I hope Tom does write his promised "huge tract about social software - about the good things and the bad things." And I hope he doesn't let the fact that these conversations are becoming more visible and more participatory scare him away from the process. His voice is worth a lot.
I suspect Tom's "wrong people" are the people who think more often than they make stuff. His comment that he wishes they'd build more things surely must reasonate with you: can you name any product that's come out of the CMC efforts that's been popular, easy to use, and effective?
Ok, you probably can name a couple of things. But compare that with the efforts of video game designers, who've pretty much leapfrogged the academic design world in the last few years with their ideas and actual products.
What impact do you think the CMC field's had on the designers of "Grand Theft Auto: Vice City"? Nada. What impact do you think that game will have on those who think about social software or CMC? Huge. It's the difference between a genuinely creative stance and one which is (often) merely critical and responsive.
You also say: "Many of my students have the technical skills to build amazingly cool things. But they don�t understand the context in which those things need to operate."
Is there much of a difference between building a few crappy pieces of software for four years before figuring out how to do it well, and spending four years in school learning history, asking questions, and considering issues?
I don't see it as an "either/or" proposition. The fact that we have historians doesn't mean that nobody makes history. The fact that we have sociologists doesn't mean that nobody makes social connections. To have people who think about and do research on CMC and social software doesn't mean that nobody will make it. But they can learn from each other.
As to your last question, I'd certainly like to think that the educational process has some value in teaching problem-solving and contextual skills, as well as technical skills. As with any field of study, I think there are people who can teach themselves "on the job," and do. But in general, no--I don't think that the quality of work done by somebody who's spent fouryears "building a few crappy pieces of software" is likely to be as good as the person who's spent those same four years building a foundation for understanding what makes _good_ software.
The "gold standard" of what I'm talking about is a place like the MIT Media Lab--where theory and practice intertwine, and real, tangible, useful products emerge.
God, I am having an awful day on the web. Twice now I've lost something I was writing. I'm at a house with a bad connection.
In reaction to this post I just wrote a nice reply (I thought it was nice) but when I hit submit the modem screwed up somehow and I got a 404. Hitting the backbutton didn't bring back my text, only an empty form.
This brings up, I think, a very important issue for social software: reliabiltiy. Especially as that relates to this little HTML form that I am filling out as I write these words. At some point the major browsers, IE, Netscape, Mozilla, etc, are going to have to realize that people are typing stuff in these boxes that are quite important. Therefore these boxes need to have the auto-save and auto-protect features we've all gotten used to in word proccessors like Word.
I'm not exaggerating to say that every week I lose something I write, at least once, because there is screw up when I got to submit. Either I've lost a connection on a phone line, or the server at the other end is busy and the browser times out.
For now, when I'm awake and clear headed, I remember to Select All, and then Copy, to make a quick copy before I submit. But if I'm tired I often forget to do that.
Okay, lets see if this will post now.
You're right--the reliability issue is _huge_ with new and non-technical users. It only takes one really bad experience to turn someone off a new technology for a long, long time.
Poor reliability and bad documentation are two things that geeks tend to tolerate well, but that create a huge barrier for new users and those who want tools rather than toys.
"But they can learn from each other."
Sure, no doubt. Although the academic world seems to have a very hard time getting through to the game-designers (for example) than the other way round.
"fouryears �building a few crappy pieces of software� is likely to be as good as the person who�s spent those same four years building a foundation for understanding what makes good software."
I'd hope not, too. But the problem is--I don't see it in the real world. I can't think of any software designers whose expertise comes primarily from an academic background. Mostly they spend a while sucking, then learn enough to get good. (The market seems to encourage this anyway.) I'd be happy to change my opinion based on really only one thing: what's the very best peice of software you've seen come out of a research or academic source, and what's been its impact on more "popular" products? (No fair going back 30 years to PARC or the early Unix world!)
BTW, I'm not down on graduate education, and certainly not on historical awareness; I have an MA in Art History that I (tell myself) I use all the time as an interaction designer. :-)
It's cause we're all commenting to this post at once (this Saturday afternoon). :-)
Idea for social software: are multiple users writing to the same blog post (nearly) simultaneously? Shunt them into a modified IM session (one which encourages comment-length statements rather than IM-speak), and log the conversation as a set of comments to the original post.
I agree with Andrew that there is much coming out of academic circles that doesn't have much real value. Designers and programmers should do what is fun and that will lead to something. I'm not suggesting that is the only development model, but it is one that will never die and always produce good work - the programmer who invents something to "scratch an itch", as Stallman put it, because he/she needed it or liked it and then, what do you know, it turned out to be useful to others.
Also, some theoreticians give opposite advice - Nielson insists that interfaces should be kept as simple as possible, whereas Tufte says you should trust your users with a lot of information, and always assume they are intelligent.
Obviously the best thing to come out of an academic setting was Google. There are a few hits, just not many.
"Idea for social software: are multiple users writing to the same blog post (nearly) simultaneously?"
PostNuke already does this, or something very close to it. The monkeyclaus team (of which I'm a part) hopes to have our software do this too by the end of the month.
I don't think there are very many programs out there yet that are turning out software designers for today's world. That's something we're trying to do at RIT, but we're a pretty small producer in a very large world.
Once his power is back on, perhaps my colleague Andy Phelps will weigh in with more authority on the gaming side of this.
But as to things of value coming out of academia...I'm more interested in web-based applications than "traditional software," so I'll name a few that have come from people who began with academic training, and then spun things out into the "real world"...
* Blogdex; by an MIT Media Lab grad student; in the process of impacting a variety of blog-related software products
* Firefly Networks; also came from the Media Lab; hugely influential in the whole "recommendation networks" software field
* TurnItIn; developed as part of a PhD dissertation at (damn, can't find it now); significant impact on students and educators
That's just a start. There are more. It's also the case that most of the people who have brought us innovative net-related tools--from the WWW itself to Mosaic to Yahoo! to Google--come from an academic context, not from a few years of building crappy software. :-)
Ooooo...I _love_ the comments-->IM-->comments idea. Definitely one to give a grad student. They're lining up outside my door looking for good projects these days. :-) (Have I mentioned that I love my job?)
"As with any field of study, I think there are people who can teach themselves “on the job,” and do. But in general, no—I don’t think that the quality of work done by somebody [who learned on their own] is likely to be as good as the person who’s spent those same four years building a foundation"
There is a spectrum of learning skills that allow people to learn either on their own or in an institution. Since elementary school I've been identified as being on the extreme end of the spectrum - barely able to learn in schools but learning more than my peers on my own.
As such, I've spent a lifetime piling up observations about the end results of those different learning styles. I could write a long post about that, but in brief, the self taught tend to have an informal an more eclectic skill set than the formally taught. Whether formality translates to better software is doubtful in my mind but it is hard to do an apples to apples comparison.
To make a reasonable comparison, we'd want to compare the best to the best - the best work of the self-taught, compared with the best work of school taught. We'd want to ask ourselves where the formality of the school-taught is an advantage and where the informality of the self-taught is an advantage. Only then could we make reasonable comparrisons.
The easiest way to do the comment->IM->comment system is use people's names as a kind of login, and have a page that autorefreshes, and that page should contain all the text, and each time it autorefreshes it should embed the user name in it.
Of course one can build something fancy - a Java thing that does chat. There are a ton of those. But I'm using just PHP and HTML, and it works well enough.
I've got a load of faith in the value of working in the wild, outside of business contexts, and of the value of contributing to debates and discussions by putting ones ideas out in the open in a build and working form. Build something that's a development in the right direction and watch as other people adopt it or rebel against it. I'm not a programmer by trade so I find these things very difficult to do myself - I hate that aspect of my work in fact - that I'm not ABLE to flesh out my ideas and put them into the world. That's why I'm so honoured to have a collaborator like Cal Henderson who I can do work with on occasion (although it does limit me to only doing things that I can convince him are worth doing).
I'm an academic by training, so in many ways I can't believe I'm saying this, but there are aspects to the understanding of how people interact with each other online that can only be understood by being part of a social group or building spaces for those groups to operate and seeing what the people within them feel they need and what the consequences are of giving them what they need. That's a very long and scarily ungrammatical sentence, but essentially I'm just saying that there's a really big limit in what you can learn through research and a lot that must be learnt by implementation and experience. There's a pretension around our work that says that we're scientists - but mostly we're not - we're artisans. We build things for people to use. We build things that extent the abilities of individuals in one way or another. As such people who work in this field should be doing apprenticeships as much as they should be doing research. They should be managing a community, understanding the tensions and the collapses, noticing the problems and the benefits, seeing where people get stuck and where they need to get stuck - where they need structure and where that structure will kill them.
[Quick aside - I hear myself speaking here and I know again that I'm talking about communities and not social software. I'm talking message boards and bulletin boards, Usenet and mailing lists. I'm talking about the simple flawed tools we've been using for decades and how important it is to know where their strengths and failings are before we push forward - as well as how people operate within them. There's much more out there, some of it's impenetrable, some of it's intensely personal and private - areas that you can't intervene or surveil or control. But I maintain as a matter of course - as a first principle - that there are lessons to be learnt in building and running a community site that apply subsequently to everything that follows in social software...]
This aspect of building things for people to use and seeing how they use them is really important to me, and has taught me more than I can express about the practical running of web-based software that connects people together socially. This is not the limit of research that needs to be undertaken, but I have little respect for people who undertake this work without having run such a community.
These are the people who shouldn't be claiming the nascent territory - the ones with the fiery passion but little sense of the random behaviours that you have to deal with - no sense of the relationship between code and behaviour, UI and structure, interaction and collaboration... These are the people with the sense of the wonder of the big picture who don't see how no actual real person will ever fall in line with their dreams. I've had debates with people about the very value of structuring the kinds of interactions that people might have through social software - respectable intelligent people with powerful positions in large organisations telling me that the very sense that one should 'structure' interactions smacked of big brother, nanny-states, and control obsessions - that the future would be heirarchy-free, uncontrolled, value-free utopian spaces. Well it's just not true! It just won't happen (even if an unstructured space was even possible)!
I'm probably revealing the depths of my ignorance by an exchange like this - that my narrow experience of one kind of working is blocking off many alternative valuable paths and approaches - that I'm not working at a large-enough scale or that I'm not thinking far-enough ahead. All of these things are almost certainly true. But it seems to me that almost everything of value that's happened online hasn't been researched to death first - it's been building and playing and sharing and arguing and making newer and newer things.
That's not to say that I don't see the value of research - but that (as you say) research and building intertwined is by far the best approach to this kind of thing.
Interesting comments, here's mine.
I think social software is less about creative tool development, and more about creative use of fairly simple tools (wiki, for example). I've considered that a fundamental of social software thinking, though nobody's written the manifesto yet (thank god).
Another fundamental: there's power in combinations of simple tools, as in the happening, which combines telephone conference with chat and wiki with useful results.
My preference is always to keep the tools in the background and focus on the conversation.
The process of putting the tools in the background, though, is often a function of designing better interfaces to the tools...which is also a tool developer's job.
The tools we have now have too many barriers (thus the surge of interest when I posted about a beginner's guide to movable type). We have too many tools built by technical people for technical people, and not enough built for people who don't want to know anything about the underlying issues.
Then maybe we can move the conversation beyond the tools themselves. :-)
I suppose the reason I'm so convinced that there's value in an academic perspective on all of this is that it helps get us away from anecdotal evidence, which tends to be the driving force in software development. "I asked the guy sitting next to me in the lab, and *he* thought it was easy to use..."
Liz wrote: "I suppose the reason I�m so convinced that there�s value in an academic perspective on all of this is that it helps get us away from anecdotal evidence, which tends to be the driving force in software development. �I asked the guy sitting next to me in the lab, and he thought it was easy to use��
This is a serious issue, but it's a red herring here. It's very easy to get conscientious desigers all het up by bringing up coder-driven development. That seems really not the issue here. Sure, academics (and many many non-academics) believe in the value of user-centered design methods, including testing.
But knowing or even applying them doesn't mitigate what Tom describes above as "the fiery passion but little sense of the random behaviours that you have to deal with - no sense of the relationship between code and behaviour, UI and structure, interaction and collaboration�"
Agreed. But one of the things that CMC has done a fairly good job of is describing those relationships. In fact, isn't that really the whole point of social science research? To allow understanding of experiences that can't necessarily be shared directly by everyone?
Back in 1993, Howard Rheingold wrote this:
I think it's still true...
I agree - to an extent - although I also think that the speed and essentially low-cost nature of the internet also means that the market, the heady memetic soup of the internet itself and the altruism of certain web people combined with natural selection and copying will probably do a more effective process of designing new interactions that take hold than research. That's not to say I don't value research, but that I suspect that often it has to take the secondary position - of helping to make sense of the things people build - working out WHY they're good...
I think the academics have a hugely important contribution to make - it's exactly as Liz says, the ability to rise above purely anecdotal evidence. Some beautiful collaborations of academics and practioners have occurred in some industries, or professions. Look at the business world - you've had people like Drucker and Dempsey making important contributions, and business people picking up their ideas and putting them to use. But what's important is that the best academics stayed close to the practioners, eyes wide open, seeing where their theories failed, never trying to defend a theory that's been proven wrong by real life.
There is much going on in social software development right now that no one can explain - the success of Ben and Mena Trott, for example. I just raised this issue on my own blog, because it confuses me greatly. If network effects usually cause first-movers to end up with monopoly positions, how is that Ben and Mena Trott were able to come into a field that already had Dave Winer, and tons of other good CMS's, and pick up so much market share? I'm stumped.
I'm stumped, but if we had a conversation about it, it would be fascinating to hear the opinions of both academics and practitioners (like Ben and Mena Trott themselves). Hearing from only one group would be less interesting than hearing from both.
There's a really interesting piece in the book "Linked: The New Science of Networks" about how particularly virulent ideas or certain individual nodes get more links in a network and rise to prominence quickly. The first to start only wins if all things are equal. If they are not all equal, then the first only has an advantage...