Change is a natural part of life so resisting it has always seemed futile to me. My hair falls out and turns grey and I prefer to just accept that it has happened rather than to wear a wig or dye my hair. But sometimes change happens and we feel regret for what is lost in the process.
The move from print to electronic journals has been embraced enthusiastically by scientists. They love the speed and convenience of instant desktop (and even mobile) access, and the flexibility of searching and linking. But this week two separate people have commented to me what a shame it is that they can no longer flick through print journals and chance upon unexpectedly interesting papers.
So, here’s a challenge – how could we (we being libraries, publishers or readers) recapture that feature of serendipity? In an earlier post I noted how lecturers are frustrated that their students do not read more widely, and one comment suggested that academics in general have a narrow range of reading. Specialisation and the growth in the volume of literature have combined to make a wide reading habit a challenge for all but the speediest of speed readers. The poor browsability of electronic journals and consequent lack of serendipity accentuate this trend.
There is nothing new about this. The SuperJournal project 15 years ago noted the importance of serendipity when browsing print journals. A 2002 study into the effect of journal format on use stated
current electronic formats do not facilitate all types of uses and thus may be changing learning patterns as well
noting that
these data also have implications for publishers and educators;
An editorial in Nature Genetics about research creativity reported a scientist saying
we should not lose paper journals for browsing; [they have] an effect on serendipity
Printed scientific journals are a thing of the past; I cannot envisage the change from P to E going into reverse. So, what can we do to ameliorate these alleged consequences of reduced serendipity and impoverished reading patterns?
Some ideas have a ring of job-creation-for-librarians, and harking back to print. The blog post,
Seven Ways to Add Serendipity to Your Research in the Digital Age, suggests browsing library stacks, browsing in the library catalogue, or asking a librarian, along with recommendations to use browse functions in databases. I like the title of another blog post: Add a little more random to your product. We want a reading list to be in between boring predictability and total chaos, it suggests, citing the iPod shuffle as a useful model.
I wonder whether services like Faculty of 1000 and Mendeley have a role, but they are both based on peer recommendations, rather than off-the-wall suggestions. I haven’t used StumbleUpon but perhaps a science version of that would be the answer. Maybe it already exists? For myself, I have found that Twitter is quite a good way of stumbling across things of interest, though there is a danger of straying into chaos.
Do you think that technical solutions (improved online browsability or greater search randomness) or social solutions (more incentives to wider reading) will work? Is the semantic web (semantic journal, semantic article) a solution? Is the whole thing a non-problem?
I’d be interested to hear your views.
I wish I had an answer for you. However, I set foot in my medical school library for the first time since I joined this institution eight months ago. I wanted to know if medical theses were consultable. The librarian enthusiastically explained that to a certain extent, beyond titles, they were. First happy point. I love resuscitating old case reports for the rare disease I’m interested in.
Second happy point, I followed up, using purely electronic tools, an interesting case report dating back to an Italian journal from 1940. I learned about the life of the illustrious pathologist who wrote it. Our library quite naturally did not have this volume. I asked if I could get a photocopy through interlibrary loan. It was on my desk not 24 hours later. The publication itself cites interesting cases dating back through the 19th century. It looks like there is a lot of fun ahead for me.
Third mention – I don’t know if it qualifies as a happy point. I always go out of my way to include, when I write a review article, at least one obscure and slightly off-topic, if not ancient and no longer cited, article. I hope to perhaps thereby, incite someone to go look it up someday and find it as enriching as I had. I’m lucky to be one of those speed-readers so I think it incumbs to me to maintain the ecological diversity of publications in my field.
Finally, I have less of a bleak outlook. TOC in my inbox often incite me to click through to a serendipitous article in the same journal, because it’s so easy to consult. Citations, likewise, when they are accessible (and link out to the actual article if possible, rather than to the PubMed citation). Many publishers now have TOC for “related journals” or “related articles” in their own stable, which is a start and sometimes leads to nice findings as well.
I noticed when I wrote “consultable” re: theses, I meant, “could I search with keywords and in what”? Apparently supplied keywords or abstracts are also searchable using computing tools, which might mean I could strike it lucky beyond whatever title these works had been given.
Heather – Great that you had a good expereince with your institutional library. I think ‘service’ and ‘helping’ are still our lynchpins in the library world. If we fail on those there is no hope.
Great post, Frank, and you raise some excellent points. I for one, just 10 y ago used to spend my lunch (quarter) hour as a post-doc flipping through a stack of journals. Despite looking at papers with little or no perceived relevance to my own work, I found that every day or two I would find a paper that supplied new ideas or techniques that could be incorporated into my own studies. So you are absolutely correct that the digital age seems to pose a real problem for the serendipitous reading that used to occur.
Today, I think I have found my own solution: I review 4 cycles (of about 13 grants per cycle) for the National Institutes of Health and the American Heart Association. Frequently, the proposals are way outside of my little niche of expertise, so I am constantly learning and finding myself digging up literature that may not be random, but serves the purpose well.
Another possibility is the use of Highwire Press, which has a PDF-searchable function that will go past the abstract and look through an entire paper. So a researcher can plug in a keyword that may not appear in the abstract, and may only be “minor” to the paper itself, but this type of search frequently elicits papers that I am unfamiliar with and may be of interest–but are peripheral to my own research.
As for aging and changing, I can’t resist an awful pun: You can’t sweep aging “under the rug”…
(just in case: rug=wig in colloquial American)
I knew I was going to rug-ret mentioning my hair.
I am one of those people who don’t use the table of contents and like to go through textbooks page by page. It certainly takes more time to find things this way, but I find a lot of interesting and irrelevant things and it makes the searching process a lot more fun! I especially love text books with boxes that explain a certain term or contain a little history about an experiment.
Reading papers as pdf files does not feel wholesome at all.The catch that pubcrawler brings into my inbox every week are articles within my narrow area of interest . So every now and then I go to the website of a journal and scroll through issues and just go through every article in the issue, just like flipping through the pages of a printed journal. It is not the same, but is close enough. Doing this once every 1-2 months makes me feel quite satisfied.
You bring up valid points about ‘reduced serendipity and impoverished reading patterns’. Imho it is definitely not a non-problem, but something that could increase the incidence of tunnel-vision among scientists. Magazines like ‘The Scientist’ or ‘Scientific American’ are quite helpful as articles in them give information about interesting new findings in diverse fields and usually many such articles also contain links to the original papers.
From the UK we also have New Scientist. I also used to enjoy looking at Perspectives in Biology & Medicine but we had to cancel it a few years back.
Thank you, Frank.
I am already rewarded with these articles in New scientist
http://www.newscientist.com/article/mg21028122.400-the-grand-delusion-egotist-moi.html and
http://www.newscientist.com/special/the-grand-delusion
And all the other articles under The Grand Delusion..
@MGG: If you like those, may I also make a plug for my personal favorite, American Scientist? Unfortunately they are saddled with a name that hardly distinguishes them from the two you have cited above. But as a working scientist, I love the review articles, particularly in physics or math theory, and the columnists for computer science and engineering are out of this world, while the biology fields are always very diverse. I still remember a great article about the genetics of domestication.
Thank you for the suggestion, I will try American Scientist. I love the reviews too, esp those in areas that are unrelated to what I work on.
Really nice post and an interesting problem. @Heather I too use TOC to broaden my reading within a particular journal or site. I also rely on blogs and Twitter (if you follow the right people). I do wonder though how often the serendipitous journal find from the glory days of print actually occurred! I believe that using the e-resources correctly can actually maximize the finding of useful articles that might be missed. After all, flicking through a recent paper journal isn’t that much different to skimming your favourite blogs and sites.
Nice post, Frank, and set me thinking. Time was when I was looking for a science story I’d go to the Science Reference Library and browse the journals. Now, though, the Mountain comes to Mahomet, as I subscribe to a lot of table-of-contents alerts. These, I admit, are journals in the areas that I cover for Your Favourite Weekly Professional Etcetera, so I don’t have that random crazy factor that yielded a few memorable leads in years past.
It could be that the Semantic Web might offer a solution, of the ‘If You Liked That, Try This’ variety. It occasionally comes to pass at Your Favourite Weekley etcetera etcetera that I receive a manuscript and it reminds me of another manuscript I’ve seen, though I can’t remember the title, the author or exactly when I’ve received it … that’s when I call up some magic software that makes all sorts of semantic comparisons between the manuscript in front of me and the archive of submissions. Quite often the results don’t produce what I am looking for, but they do often act as an aide-memoire.
I think the re-creation of serendipitous discovery is going to be a really big thing in shifting us finally away from paper for all the reasons that you discuss. There are lots of potential ways to create serendipitous discovery – many of which will be a lot better than “stuff between the same covers” – but which ones will work the best, and in particular avoid just reinforcement effects is something we certainly don’t have an answer yet. My guess would be careful tuning of some social filtering and connection parameters that bring things from further and further away in your social network into your inbox (which might well look more like the Flipboard iPad app) with some probability distribution that tails off with network distance. Or it might be subject or index based distance, which would be closer to the “between covers” kind of serendipity…
I second Cameron’s vote for “Flipboard for Journals”. I strongly suspect that future (premium) versions of Flipboard or alternative Flipboard-like products will help to address this problem of diversity in reading/serendipity.
Thanks all for the kind and interesting comments. A quick summary thus far:
TOCs services (tables of contents) do lead to serendipitous discovery (Heather, Henry and LeadingEdge).
Browsing journal websites and reading everything in journal issues is still possible (MGG)
‘Related articles’ links are useful for this too. Article reference lists too. (Heather)
Full-text search can generate unexpected but useful results (Steve)
Just reading everything in a book/article (MGG)
Reading broad-based magazines (MGG, Heather)
Henry (Cromacrox) has some magic semantic software
Some activities (e.g. reviewing grants) force you to read material outside your narrow subject area (Steve)
Tools like Flipboard (for the iPad) have shown how to create a better browsing experience
My boss commented to me that “it’s a state of mind as much as anything. If you are the kind of person who likes to read around a subject and even outside it, you will do so.” I think there is a lot of truth in that. I remember years ago seeing an article recommending a good reading habit for researchers as being to read one textbook in your broad discipline every year, regularly read review articles on the fringes of your field, and of course read the primary literature of your field.
I liked the comment from LeadingEdge too, being sceptical of the ‘good old days’ of print serendipity. Both MGG and Heather say that TOCs in the inbox can deliver surprising discoveries. Perhaps we shouldn’t make too much of this – it may be a matter of tweaking interfaces rather than radically changing anything. More Flipboard-like products would be nice to see though.
I like the idea of more semantic links peppered around journal articles, so you can follow links madly. But will it ever come? Will it take 5 years or 25 years? Meanwhile, searching is getting better and better. We have full-text searching in Google Scholar and in UKPubMedCentral. I think searching can be serendipitous in a different way than browsing.
Meanwhile, I rather liked Heather’s ‘guerilla citation’ idea, of including at least one obscure and slightly off-topic article in every review that she writes. I shall have to have a look at her past reviews.
I hope you can forgive me for being a little anti-social and writing my rambling, ill-coordinated thoughts on this on my own blog (linked above). Great post, Frank – good to think about.
Grant – thanks and thanks too for your response on Code for Life. I will comment there in a bit.
I’m a bit late to this party, but I had a conversation with Mrs. Ricardipus, who is a schoolteacher, yesterday – on the topic of libraries. Seems that a number of schools are dispensing with them altogether as a cost-saving measure (and the librarian too, of course). A big driver of this is E-books – publishers will provide licenses for typically a single hardcopy, and a bunch of concurrent-use licenses for the book. They’re accessed online by anyone who has the login and password for that school (the password presumably is changed from time to time, perhaps annually, but maybe not).
As a cost-saving model, this has some advantages, and as it is internet-based it avoids the problem of dedicated hardware readers (Kindle or similar). One major downside (apart from the lack of a librarian to help students locate resources) is inclusivity – it presumes every student will have access to an internet-enabled computer or mobile device, which even here in big-city Canadian suburbia, is probably not the case. Hence the need for the hardcopy, I guess.