Twenty thousand conversationsBMJ 2002; 324 doi: http://dx.doi.org/10.1136/bmj.324.7347.1171 (Published 18 May 2002) Cite this as: BMJ 2002;324:1171
Rapid responses suggest new models of knowledge creation
Conversation is a meeting of minds with different memories and habits. When minds meet, they don't just exchange facts: they transform them, reshape them, draw different implications from them, and engage in new trains of thought. Conversation doesn't just reshuffle the cards; it creates new cards.1
Four years ago we added a feature to bmj.com, which we called “rapid responses.”2 The feature allows readers to respond to articles directly via the website as they are reading them. We don't regard them as second class letters: they are just as eligible for inclusion in the paper journal as letters received in other forms. In fact, most of the letters to the editor that we now print in the paper journal started life as rapid responses.
Our original intention was to post all but the libellous, gratuitously rude, trivial, irrelevant, or incomprehensible on the website within 72 hours.We hoped that, at the very least, it would solve the problem of receiving far more letters to the editor than we could possibly print in the paper journal. We wondered whether rapid responses marked the most democratic step that the journal had ever taken.
So how has it gone? With 20 000 rapid responses now published on the website we consider the experiment a success. Particularly gratifying has been the participation of readers from outside the United Kingdom; in fact, those furthest from the centres of medical power have been the most appreciative. We've begun to capture the opinions and experience of patients—“the dimension missing from medical discourse for millennia” 2—but we have much further to go.
We've made a few changes over time. Each response is now searchable by author's name and topic. We've speeded up posting—now all but the most troublesome responses are posted within 24 hours. And there are few of these: we publish just about anything that isn't libellous or doesn't breach patient confidentiality. There are losses in having such a low threshold for acceptance, but we think the gains outweigh them. We side with Harry Truman, who argued that we need to fear the suppression of ideas, not their expression.
Not everyone is as enthusiastic about rapid responses as we are. Democracy, it turns out, cuts both ways: while everyone is now allowed a place at the table, high status loses its currency. Some authors have been surprised that their eminence hasn't protected them from a mauling at the hands of “the mad, the bad, and the misinformed.”3 Others have been piqued at not being given the chance to negotiate over criticisms of their work before publication.
Rapid responses allow both groups a remedy: authors can use them to respond to the responses. Either they can mount a robust defence of their position, or else they can't—publicly. Users can sign up for our Citetrack service, which automatically alerts them when responses are posted to a particular article (see option 3 in the figure). To help corresponding authors keep abreast of these we sign them up to this service as a default.
It's hard to pin down exactly what sort of thing rapid responses are. While they're still recognisably letters to the editor, they share the characteristics of email, being intimate, direct, and not always impeccably spelt and punctuated. In content, they often resemble the discussions that occur at meetings after presentations. “Conversation” is probably the closest approximation we can get.
The best analogy for rapid responses is therefore a massive party, with a multitude of conversations going on. By moving about—as at a party—it's easy to spot the important issues of the day by the number of people voicing their opinions, and the passion with which they're expressed. Poet John Milton reassures us about the level of noise: “Where there is much desire to learn there of necessity will be much arguing, much writing, many opinions; for opinion in good men is but knowledge in the making.”
Like all big parties ours has its percentage of bores, but they are easier to sidestep on the web than in real life. No one forces you to read their postings, and you are free to move on to something more interesting. We've discovered that our bores are the same as everyone else's bores, with the same needs for attention and dominance. 4 5 But an online poll and discussion we've had on the topic showed no great appetite for introducing sanctions, or even club rules, and we recognise—as have organisers of other sites open to all comers—that “problem children” have their uses. 6 7
1665 and all that
To get an idea of where rapid responses might be heading, it's instructive to know where we've come from. Before journals, letters were the main means for scholars to communicate ideas. Official scientific societies gave a great spur to this traffic in letters. At their meetings, letters were read out from the societies' correspondents, and these filled many pages of the first scientific journals. 8 9 In these early days of scientific publishing, the notion that a community's musings were discrete from journal “content” was an alien one—both were forms of conversation.
Fast forwarding to today, we find conversations once again counting as important content—rapid responses now make up 40% of the searchable content of bmj.com; within three years there will be more rapid responses than articles.
If the internet returns us to the pre-journal model of knowledge creation by conversation then the conversations are likely to be qualitatively different from what has gone before. Participants need not be in the same room at the same time, and there is no limit to their number. Mario Morino, who studies the internet and social change, believes that tapping into the collective intelligence of thousands of people using the web could provide a new model for problem solving: “The power of bringing together the right minds around a subject in an online dialogue, well facilitated, well deliberated, has enormous potential to help us get through issues that we've never solved before.”10 He cites the success of the open source movement in software, where the freely given efforts of hundreds of software developers have produced solutions unobtainable by the usual methods of problem solving.11
Our experiment began with moving letters to the editor to bmj.com. Should it matter that we don't know how it will end up? Levy counsels us against expecting the digital world to be neat and tidy so soon: we are just at the beginning of figuring it all out. “Moving existing genres online doesn't mean that they'll be stable,” he writes. “Indeed, it pretty much guarantees that they won't be.”12