Following on from my post the other day about Systems Builders, in which I touched on these themes: synthesis, analysis, visonaries, implementers. Some interesting trackbacks occured out of this. Let me first mention Jon Udell's post this morning, because his discussion of "sound bites" is particularly relevant to the points I want to make here. Paul Graham made a speech at Oscon 2004 that caused ripples of controversy around the Web. Ironically I haven't heard that particular speech yet, but I read Graham's 'Great Hackers' essay and listened to an earlier interview he did with Doug Kaye. Here's what Jon said about Graham's Oscon speech:
"Consider Paul Graham's remark. I suspect that most who commented on it did not actually hear it, but instead read it, or read about it. How much of its impact is conveyed by the text, and how much by the delivery? Whatever that ratio, access to the primary source -- the words as actually spoken -- is bound to affect the perception of the remark."
It's all about context. According to Jon's quote above, how you take Graham's remarks will depend largely on whether you heard them in the original audio or in text form (transcription, synthesis, extracts, etc). I'd go further and say that how you received Graham's remarks also depends on whether you listened to just an extract of the speech, or the whole thing. The most reliable context is listening to all of the original audio.
Jon Udell goes on to say:
"In the realm of public discourse, it's easy to imagine what this could mean. The presentation and analysis of sound bites has been almost entirely at the discretion of the broadcast media. Think how different it will be when we the media can choose the sound bites that we want to discuss."
Jon is putting a positive spin on the situation - every Joe and Jane Bloggs can now put things into their own contexts. We don't rely on broadcast media to do that so much now.
But... there's a flip side to that coin. Before I get to that, here's a bit more from Jon:
"Think about how we "write up" meetings today. Some people try to transcribe, and fail to synthesize. Others synthesize, at the risk of revising history. A collective synthesis rooted in the audio transcript seems like the best of both worlds."
It's true that a "collective synthesis" is very democratic and has wider breadth, because it's not just a product of a broadcasting elite (i.e. journalists). But let's not overlook the corollary of that: the more people you have transcribing, analyzing and synthesizing audio and text on the Web, the more things get taken out of their original context. For example, something that makes a great deal of sense within the context of the original source file, can take on a totally different meaning if you take a snippet of the original file and put it into your own post which is on a different subject.
Paul Graham wrote a number of controversial things in his 'Great Hackers' essay. For example, this paragraph:
"Hackers like to work for people with high standards. But it's not enough just to be exacting. You have to insist on the right things. Which usually means that you have to be a hacker yourself. I've seen occasional articles about how to manage programmers. Really there should be two articles: one about what to do if you are yourself a programmer, and one about what to do if you're not. And the second could probably be condensed into two words: give up."
When I read that in the original essay, I understood the point he was trying to make: that to manage hackers you need to understand their spirit, to be in the same headspace. That theme was recurrent throughout his essay and therefore it strongly resonated with me. But when you take that paragraph out of the context of the rest of his essay (as I've done just now), it becomes much more blunt and the meaning changes. In fact when that paragraph is isolated from the rest of the 'Great Hackers' essay, as in Andrew's post yesterday, I now find I disagree with what Graham says. I don't agree that only programmers can manage other programmers - that's just plain wrong. In my view a visionary may not be a programmer, yet he or she can certainly lead a team of programmers in the implementation of his or her vision. Examples are Mitch Kapor and Marc Canter.
So you see my point? I had two different reactions to Paul Graham's paragraph on managing programmers - I agreed with him in the context of his original essay, but I disagreed with him when I read it again in Andrew's post.
Incidentally, at the end of his post Jon Udell mentioned Glenn Gould's The Idea of North (did he get that link from me, via my link to him?). The form of audio splicing that Gould did in The Idea of North is one method of putting things people say into new contexts and creating new meaning out of that. That was re-contextualizing as art, but what's happening now on the Web is context-morphing on a mass scale.
Microcontent in the form of sound bites, links and text extracts are the lingua franca of the Web. They enable us to bootstrap the Web of Ideas. But context on the Web is much more fluid and it morphs very easily. So when we link to something (a piece of audio or text) but give it a different meaning - what are the moral and ethical implications of that?
]]>But I see two spins:
One is that content is content, and doesn't it have the freedom to move from place to place as it sees fit? "Information wants to be free."
Wouldn't you feel restricted if you were only allowed to be seen in your country of origin? Don't you think that content/data would feel similarly restricted?
The other being that this is precisely why we NEED the semantic web - with more precisely defined semantics, the context becomes highly codified in a rigorous manner and thus automated tools would prevent the reinterpretation within inappropriate context. But that would mean that the automated tools would need to have semantic data on every word, auditory and visual nuance, and so on. We're not there yet - unless we hand-code such data - but then the coding is only as good as we provide it - and if we code data that corresponds to stuff not from us, we introduce our context, not the content originator's context. So self-coding of non-self-data is not a viable option. So this spin would mean that everyone should provide ultra-semantic markup for every word and every nuance of everything they put in the web.... Not likely to happen soon.
Neither of these are likely wise spins. And I'm misquoting the "Information wants to be free" thing for laughs. ;)
]]>It occured to me after I published this (and listening to Dave Winer's audio post today on bloggers as journalists) that we have the same issue with journalists taking things out of context. I guess my spin on that is: sure, but with the Web you have 1000 content creators to every 1 journo. So context-morphing has just reached another level.
Let's not drag the Semantic Web into this :-), I want this to be a civilized discussion ;-)
]]>Great comments, thanks!
I agree that it's bad to sacrifice context. But with finite time and infinite sources of information, we do have to skim, and we have to rely on intermediaries.
Where possible, I'd just like the intermediaries to make primary sources easily accessible -- at various levels of granularity. Hearing a widely-discussed sound bite is more information that just reading about it, albeit less than listening to the whole thing. And hearing it might provoke you to listen to the whole thing, when otherwise you wouldn't.
"For example, something that makes a great deal of sense within the context of the original source file, can take on a totally different meaning if you take a snippet of the original file and put it into your own post which is on a different subject."
Agreed. But I think we're operating in an increasingly self-correcting medium. If I quote something in a way that does violence to the source, somebody will call me on it -- and in a very public way that tends to ripple to all interested parties.
It *was* you the reminded me of Glenn Gould, plus I'd recently seen a great documentary about him.
]]>