Athabasca University | AU Student/Staff Login | Invited Guest Login
The future of academic literature
Critics suggest that the trend towards replacing academic peer reviewed papers with blogs and wikis that bypass that process is a dangerous trend, leading to shallow, unreliable and unsupportable beliefs supplanting rigorous research. Similarly, the relative reliability and accountability of traditional journalism is being replaced by unaccountable, inaccurate and biased reporting by amateurs. On the face of it, there may be some truth in these criticisms: at least, it is harder to distinguish the chaff than it used to be, though there are gains in diversity and timeliness. However, at least in many cases, this perspective is a result of a skeuomorphic failure to recognise that such posts only superficially resemble the publications that they replace. A blog post is not a paper, a wiki page is not a publication, despite their intentional resemblance to those archetypal forms.
Social media such as blog posts and wiki pages do not exist in isolation: that's what 'social' means. They are surrounded by a web of commentary, dialogue, ripostes and critiques that are as much a part of the 'publication' as the post itself. So, if we have cause to criticise an original post or page, so will plenty of others. In fact, we can add our critique as part of that process, and engage and learn more deeply as a result. The outcome is a co-created medium of which a single post is only a part, a dynamic system in which peer review is not the input to improve the original but a part of the content itself. When it works well, with sufficient input from sufficient people, it can be a far more enlightening, rigorous, multi-faceted medium than any traditional forms. Of course, the process can fail: too much input, too little input, too little filtering, too much filtering can make it far less wonderful. And it can fail if we treat it like the forms it replaces: you can easily miss 90% of the value of Wikipedia, for instance, if you don't read the discussion page that leads to the entry itself. But, when it works, it works brilliantly.
My favourite example of what happens when you rethink the process and move beyond skeuomorphs is the now venerable Slashdot site. It is built for and by passionate geeks so it is not a form that is readily replicable: you have to delve into the complex mechanics of the ingenious use of collaborative filtering, the distributed bottom-up reputation management system, the ingenious checks and balances on bias and mob stupidity, and the management of explciit filters to get the full benefit of the system. Only a geek or a very determined non-geek is going to do that. The reward for those willing to put in the effort (and, despite the barriers, there are many tens or even hundreds of thousands that do) is an emergent literary form co-created by its inhabitants that evolves into an extremely high quality and reliable knowledge source with a richness, depth, creativity and diversity that no single author could hope to match. Or, if you prefer, a shallow humorous take on technology. Or a place to support rabid and improbable beliefs or biases. It's up to you. Once you start to customise it, it is an extremely dynamic, extremely personalised, extremely diverse system fuelled by the crowd that can be many different things to many different people. Slashdot is not the defining academic literature or journalism of the 21st century, but it points the way towards something that is potentially far more powerful than the result of the tree-based technological constraints of yesteryear.
- Sets, nets and groups revisited
July 3, 2024 - 10:41pm
Jon Dron
Jon Dron
Tag cloud
Comments
Here is my jaundiced and cynical view on academic literature: I think the traditional vehicles no longer serve their intended purpose (I can't say to what extent they ever did, but I sense that we are in a season of decline). There is a great deal of cronyism and mutual back scratching, even in the hard sciences (witness Elsevier's collusion with drug companies). The number of people with PhDs who feel the need to publish even if they have little to say has increased exponentially. I have to agree with one of my dissertation advisors who said to me before the World Wide Web existed, "You have to wade through a lot of drek." A person has to exercise the same discrimination for peer-reviewed articles as for blogs and wikis. I think the cream still rises to the top, but maybe the time has come when online communities can do a better job than the traditional peer review process--the Internet is by nature more transparent and more thorough, and as Jon said, when it works, it works brilliantly--it has to offer something better than what academic publishing has come to in the traditional channels.
Skeuomorph: word of the day. Like the floppy disk icon that you click to save something, even though floppy disks are totally obsolete.
To get on-topic though, I think to frame the proliferating sphere(s) of public knowledge as the "replacement" of refereed research by citizen journalism is misleadingly zero-sum, when media have never really been a zero-sum situation. (Well, okay, maybe for floppy disks it was.)
And I think there's some very powerful and publicly valuable synergy developing between peer review and open access.
I agree with Mary - the system is deeply flawed and has to evolve. Mark's point about it not being zero sum is very true, though - as Kevin Kelly rather brilliantly observed in 2006, old technologies seldom if ever actually die, a theme that he expanded in his wonderful book 'What Technology Wants' and posted a great update on fairly recently. This has spawned a challenge (a very enjoyable and fascinating read that illustrates my point rather well) that I don't think has has been fully met to find a genuinely extinct technology - one or two technologies that turned out to unintentionally kill people may have mutated a bit and some may have been forgotten, but otherwise the principle holds up very surprisingly well. And, as W. Brian Arthur's great exposition on the nature and evolution of technology 'The Nature of Technology: what it is and how it evolves' shows, technologies almost always evolve by a process of assembly, so it is highly likely that we will see increasing numbers of hybrids and ways to improve on both paradigms by taking parts of the old and new. Early hybrids like JIME (still interesting but not quite as cool as it was in the early to mid 2000s) and newer journal-ish things like PLoS hint at what might be, though are not quite there yet. All of which means that the adjacent possible will, as ever, continue to grow.
It would be interesting to see a plug-in designed for this. Would it be a knowledge management tool driven by end-user selections, or a customizable recommender system that is driven by a process (eg. peer review process).
It's loosely part of where I'm trying to steer the Landing, but it's much more than a single plugin, I think. There are numerous existing and planned plugins that we could use to deal with (amongst other things) reputation-based recommenders, content recommenders (using collaborative filtering algorithms), simple visit counters, ratings, rich tagging and so on, all of which will make it easier to navigate to things of perceived value, I hope. But to be able to deal with the diverse richness and complexity of the different and overlapping networks, interest sets, groups and subsets on the Landing is quite tricky, without falling into a semi-formal process like the hybrid journals or the unfeasibly high cognitive complexity of Slashdot.
Having said that Slashdot, as it has done since the 90s, is innovating at least a decade ahead of the rest. Its default public views are now pretty simple to interact with yet still offer higher quality content, on the whole, than simpler low dimensional recommender-based discussion/sharing sites like Digg and Reddit. However it's still not fantastic without the customisation and personalisation aspects. Apart from the karma-point-based reputation management which they invented (which means human editors emerge through a bottom-up approach and so the quality is higher than an uncurated system) it really needs that end-user complexity do its stuff properly. The principles it uses rely heavily on intelligent use of parcellation and the capacity for nuanced recommendations within that parcellated space, which means it winds up with a richly speciated ecology of knowledge that has, literally and in a strong Darwinian sense, evolved, but only within that particular environmental niche. The trouble with trying to display the results of that to an unknown public is that it has to fall back on a simpler popular/unpopular and highly rated/low rated metric like the rest of them, leaving only the power of karma and a large and enthusiastic user base to generate great content. It's like putting sparrows from the Galapagos Islands on the African Veldt, the adaptations get lost in the expanse.
Maybe one answer is a really smart interaction design that makes the customisation easier? Or perhaps a meta-layer of adaptation that mines that from implicit preferences? Tricky.
From my perspective (in the back room, tinkering with the nuts&bolts) I don't see academic papers as particularly distinct from the rest of the information tsunami, though I appreciate their particular quality, the role they play, and the problems they face.
I find the lack of access poignant. Truly.
But to my specific: I just tweeted this to Juan Pablo Alperin (@juancommander), researcher with the Public Knowledge Project and PhD student under John Willinsky:
Has @PKP addressed the problem of linking to anchors? I can on my own pages, or in YouTube, but not generally. I mean this:
On one of my own pages it's very easy to place an internal anchor very precisely. It's a no brainer.
I can link to a precise moment in a YouTube video by specifying #t=2m42s ... spot on.
Has PKP come up with a way of linking to a specific point (page / paragraph / sentence) in a journal article? That would be //wonderful//!
The corpus is magnificent. It deserves our best methods.
cheers
--@bentrem
- Ben Tremblay
That would indeed be very useful but I'm not aware of anyone doing anything serious to make it happen. The Web is increasingly evolving towards Ted Nelson's original concept of Xanadu and the more recent but apparently moribund transliterary standard (http://transliterature.org) that would easily handle such things, but there is always a tension between the simplicity, flexibility and forgiveness of HTML (which is why it took off in the first place) and the richness and complexity we really need.
There are quite a number of interesting initiatives. I mean innovation in annotation. (DAV ... does anybody even use that accro anymore?) But none have the mass/gravity that would actually reduce fragmentation. (I'm particularly fond of the system used for feedback on gplv3-draft-4, but it was orphaned from the day it was born.)
What I see as intermediate / work-around ... and I think this would run into copyright and fair-use problems ... is something like Scribd that would host substantial extracts, so the anchor point could be seen in context, allowing the reader to then access the full document.
BTW while TimBL was creating WWW I'd come up with a system of hyper-docs for avionics R&D, using WordPerfect 4.2 (I say I was ham-strung by not being on the net. Of course there's more to it, like him being genius. <G>)
I look to folk like PKP and PLoS to push this forward, while I work on what I call "the pivot point".
p.s. wanted to quote Steven Harnard; found the right paper and was going to use it as example of how we can't link to a specific section. Alas, alas ... openscholarship.org EnablingOpenScholarship (EOS) "an organisation for universities and research institutions worldwide. The organisation is both an information service and ..." is offline. Sad. #BitRot in spades!
- Ben Tremblay
Addendum: Harnad's Open Access is still online.
- Ben Tremblay