In an era of change, it’s a little embarrassing to admit my Luddite tendencies. I resisted ATMs, answering machines, DVDs and cell phones longer than most of my friends. I hold on to already purchased technologies longer than their anticipated life span, upgrading my computer only when it does nothing but sputter and whir when I attempt to load the new version of a program that I absolutely must use. I only recently joined Facebook.
As an editor, I’ve bemoaned the change in reading practices among the current crop of graduate students, who don’t pick up journals anymore but rather identify what they want to read by eTOCs, RSS feeds and directed searches. What happened to browsing, to stumbling upon that interesting paper that had nothing to do with what you work on but changed your project in a fundamental and important way? Occasionally I consider guerrilla measures to force people to browse online, such as providing the wrong PDF every now and then so that their eyes will rest on something unexpected and potentially transformative.
I only have to look at my 9-year-old son, Jacob, to realize that I’m thinking about this in the wrong way. Jacob is ravenous for information and he finds it, not surprisingly, on the internet. His searches take him in all sorts of directions and peak his curiosity in different ways. Browsing definitely happens online, but the tools for browsing are different, and evolving.
An editor’s job includes both helping readers to find the information that they need and bringing different members of a community together to share ideas. We do this by acting in several roles: as pre-filters, sifting through a number of contributions for those of value to our defined readership; as initiators, inviting contributions about particular topics from particular authors; and, to some degree, as post-filters, highlighting contributions and assessing their value. At DMM, the pre-filter function applies to the research papers, whereas the initiating and post-filter functions apply to the front runners section of the journal. Chris Anderson, the Editor-in-Chief of Wired magazine and the author of The Long Tail, a book about ‘how technology is turning mass markets into millions of niches,’ has this to say about being an editor:
“As the editor of a magazine with a finite number of pages, I’m a classic pre-filter. I indulge in all sorts of brutal discrimination and guesswork to decide which articles to run. But Wired also does a lot of product reviews, and in that respect, we’re a post-filter. We look at the universe of what’s already out there and bring the best stuff to our readers’ attention.”
“As long as there’s a market for a pre-filtered package in the deliciously finite medium of bound glossy paper, I suspect there will continue to be demand for my old-fashioned discriminatory side. But the day when people like me decide what makes it to market and what doesn’t is fading. Soon everything will make it to market and the real opportunity will be in sorting it all out.” The (Longer) Long Tail (2008), p. 124, Hyperion.
The ‘long tail’ describes the graph of the likelihood that consumers will purchase items in an online environment that aren’t generally available in traditional stores because they are not popular enough to justify the shelf space (the right hand side of Fig. 1). Prior to this availability online, consumers generally purchased ‘hits’, with a small fraction of the available items accounting for most of the sales. Online, however, almost all available items (and there are many more of them) find their way to consumers, who now populate ‘niches’. As a result, ‘hits’ become less popular, and consumers often discover products that are better suited to their interests and tastes.
The long tail. In a physical store, the popularity of items determines whether stores will stock them. Anything to the right of the red area of the curve is not popular enough to justify the cost of the shelf space. In contrast, because ‘shelf space’ is essentially free in an online marketplace, many more items are available and a large fraction of them (as high as 98%, according to Chris Anderson) are purchased in any given quarter. Consumer behavior shifts from buying only current ‘hits’ to buying a bit of everything. Note: this is a public domain image from Wikimedia.
The long tail. In a physical store, the popularity of items determines whether stores will stock them. Anything to the right of the red area of the curve is not popular enough to justify the cost of the shelf space. In contrast, because ‘shelf space’ is essentially free in an online marketplace, many more items are available and a large fraction of them (as high as 98%, according to Chris Anderson) are purchased in any given quarter. Consumer behavior shifts from buying only current ‘hits’ to buying a bit of everything. Note: this is a public domain image from Wikimedia.
Thinking about the future, I’ve been trying to apply ideas developed in The Long Tail to the world of scientific communication. As readers of scientific literature, we generally scan our top journals, which provide the ‘hits’ of scientific research. But these journals are unlikely to give us all the articles we need for our research. We perform directed searches hoping to find the rest, but we are still likely to miss potentially time-saving, or even project-transforming, studies. In scientific communication, the long tail has not yet appeared. Why not?
Anderson defines three forces that enable long tails to emerge:
Democratized production: the personal computer and the software developed for it enables individuals to do for themselves what in the past only a few professionals could do, swelling the ranks of producers by many thousand-fold.
Democratized distribution: the internet makes it possible for essentially all products to be available for purchase from an individual online ‘store’ (think Amazon) without the need to maintain massive local inventories.
A connection between supply and demand: in order for all this increase in available content to reach consumers, the ‘search cost’ of finding the right niche content needs to be low. Mechanisms include search engines such as Google, blogs, recommendations, ‘best-seller’ lists and other forms of collaborative post-filtering.
Looking at this list, it seems that at least two things are holding back the long tail of scientific communication. As a culture, we have grown dependent on the pre-filter function of journals to define and reward research impact. As a result, we resist democratizing production and distribution, even though the tools are in place for authors to write and publish their own articles; compounding the problem, aggregators such as PubMed incorporate articles from only a small subset of journals and will not even consider self-published articles, making them more difficult to find. Furthermore, current efforts to post-filter the literature have fallen short of the mark, keeping the ‘search cost’ high. Science blogs exist, many journals provide ‘top download’ lists, and a few journals have enabled post-publication comments or other annotations of their own papers. These features, however, have not yet delivered on their promise: to help the end user absorb the largest amount of relevant information in the smallest amount of time.
Although the Luddite in me wants to protect the status quo and to ensure the success of DMM and other journals, I can’t convince myself that scholarly journal publishing is different from book publishing – a world transformed by Amazon. Change is coming in the communications arena, and I want to see it and contribute to it, rather than dig in my heels. But to do that, I need the help of people with different technological tendencies; therefore, I invited my friend and colleague Chris Patil (currently a researcher studying aging at the Buck Institute for Age Research, as well as a science blogger in his ‘copious’ spare time) to think through these issues with me. The editorial that follows is the first product of that conversation, carried out in large part on Facebook.
