Skeptics Online


For about a year now I’ve been an active member on the Skeptics forum within the Stack Exchange (SE) network. Stack Exchange bills itself as “a fast-growing network of 114 question and answer sites on diverse topics from software programming to cooking to photography and gaming.” Many of my readers will be familiar with Stack Overflow, the site for professional and enthusiast programmers. Stack Overflow shows up frequently in search results about various software- and programming-related queries. The other sites in the network are less popular by several degrees of magnitude, but they also have more of a community feel. On Skeptics, the core group is small enough to recognize its members by name.

My purpose here is to describe this little corner of the Internet, both as an ethnographic exercise and as a moment of self-reflection. At the very least, I hope to capture a snapshot of the quickly evolving life of an online forum. All Stack Exchange sites look and work the same way, using the same underlying software service. The idea of question and answer forums has been around since the early pre-Internet days of bulletin boards: you visit, write a post that asks a question, and hope someone answers. SE improved on that model by seeking not just an answer, but the definitive answer. Where general forums encourage open-ended discussion, SE is set up to finish the conversation. In a perfect world, a question should have one succinct answer. That is what makes SE so popular. Where, on other forums, the answer is hidden in a long string of replies, SE prominently features the definitive answer on top of the pile of responses.

Like many other social web sites, SE is heavily “gamified.” Active users get points for good questions/answers and badges for various achievements (like answering a particularly old question, for example). A registered user is able to vote on the quality of the post (in a binary way, either up or down), adding to the total count of the author’s reputation points. The end effect is a system of social filtration. Poorly-received posts “sink” to the bottom of the pile. Quality content “floats” to the top.

Points and badges (which are the essence of gamification) can feel infantilizing sometimes, but in this case the achievements are tied to real editorial privileges. It takes roughly 125 points (at 10 points per upvote) to be able to downvote someone, for example. At 2,000 points, a member can start editing all questions and answers (and not just her own). The ability to vote to delete posts from the site entirely kicks in at 3,000 points. 20,000 points grant further editorial privileges. Interestingly enough, the community moderators are elected in an open election that does not require a reputation threshold. Of course most moderators (who can do things like change the look and feel of the site) tend to be long-time contributors to the community. This model of governance rewards stable identities and active, high-quality participation. (The quality part is an important piece here. Other reward systems encourage quantity over quality, which can result in the frequent appearance of repeated “meme” content. At SE, such posts would be voted down and some effort is taken to remove duplicate content).

The Skeptics forum has high evidential standards. Questions must present a notable claim—something that appears in popular media, for example. Similar to the Wikipedia policy, SE answers should not contain popular research, relying rather on peer-reviewed scholarship and other reputable sources. When the answer is good, other members of the community may ask for further clarification, better source material, or offer other editorial suggestions. And, although it is not required, the person asking the question is encouraged to accept the correct answer, which brings a few extra points to the answerer.

SE sites tend to cluster around communities of expertise, like programming, physics, photography, and English language usage. The Skeptics community differs slightly from these in that it is a forum for applying the general principles of scientific skepticism. The site specializes in debunking notable bogus claims, popular misconceptions, pseudo science, and superstition of all kind. Medicine comprises the most popular category by far, with nutrition and history following close behind. My most popular answers on the site include “Do wild dogs use trains to commute to and from Moscow?” (yes they do),1 “Did only a handful of people in Europe know how to do division before the 13th century?” (no, long for division was widely known at the time),2 and “Did the Ancient Egyptians use twenty-sided dice?” (yes!).3

Writing these posts is time consuming, taking anywhere between a few hours (when the answer is limited to simple citation) and a few days (when it requires extensive synthesis). Why do I contribute?

First, I find it relaxing. There is great pleasure in using my research skills in areas which I don’t normally encounter in my professional life. Second, I believe in the cause of tough-minded skepticism. It is the sort of thing that often goes by the name of “critical thinking,” even though few are willing to apply it to all aspects of their belief. Third, I feel compelled to do it as a small measure of civic duty or citizen scholarship. I have easy access to university resources like PubMed and JSTOR, which are closed to the rest of the world. It takes me just a few minutes to answer questions like “Do girls mature more quickly than boys?”4 or “Is the value of a tree $193,250?”5 using fairly reliable, state-of-the-art sources. Finally, I find in Stack Exchange a powerful model for academic publishing (or publishing of any kind for that matter). Running a journal requires an enormous amount of work (by editors, managing editors, and reviewers). Most of this labor is invisible and, for the most part, unrewarded. We could learn a lot about streamlining the peer-review process from communities like SE. Imagine, for example, accruing reputation points for being an active reviewer (or being on time with your comments), and then trading those points for expanded editorial privileges or for faster turn-around times when submitting your own articles for publication.


Asocial Text

Writing is a fundamentally social activity. Even when you do it alone, in a locked room, wearing your new noise-canceling headphones, you are (hopefully) writing for someone. A private language, says Dr. Wittgenstein, is an impossibility. Reading is a social activity too, because at the very least, it is an encounter of two minds. But usually, there are many more minds involved: other texts, other writers, co-authors, co-readers, book clubs, literature professors, snooty bookstore employees, publishers, and book critics.

Yet, these are quiet social encounters. They require a measure of focus, solitude, and introspection. It would be a mistake then to envision the future of the book simply in terms of social media. Part of what makes a book a book is its ability to block a part of the present physical world in favor of atemporal virtual reality. The book literally blocks vision. It privileges mental constructs over immediate input of the senses. To be lost in a book is to project one’s sense of being into another world.

Let’s imagine then a better book, one that further protects the sanctity of mental life, at least for the duration of reading. Imagine a book which, when opened, literally surrounds its reader in a protective cocoon. Imagine a book that can balance the reader’s dopamine levels. Imagine a wearable winter coat book, a pillow and blanket book, an umbrella book, a climate-controlled book built like a house or a nuclear fallout shelter or a biodome.

Paper, as it turns out, is a pretty durable material—much more durable than, let’s say, silicon chips or copper circuit boards. It can also be used for insulation, it bends and burns better, and can make for versatile construction material (for the folding of paper planes, for example). I say this without irony and without fetishism or nostalgia. Whatever technology comes beyond the book, it should at the very least do all those things better than cloth and paper.

Bad Links


It is my intention here to convince you that links are bad. They are bad when it comes to writing for the web in general, bad for books, bad for long-form journalism, and even worse in academic publication. It is not that I am against the idea of links. As we will see here, the problem lies in the way links are used. That is also to say that we can do something about using links better. But first, why are links so bad?

To start with, links are opaque. The worst of lot are links like this and this. Of the two “thises,” the first leads us to Google and the second to Bing. But your readers would not know that just by looking at the text. The best they can do is “hover” over the word with their mouse cursor, relying on the browser interface to show them where the link is going. And once they get there, there are no easy ways to get back. The writer must have faith in the browser to “do the right thing” in guiding the reader through an intertextual maze. And that is not right when it comes to writing. In most situations, the author should architect that experience explicitly. If you think about it, the old-fashioned apparatus of quoting an external text is itself a type of linking. But rather than quoting the whole text, the author only quotes the relevant bits. Sending readers away to do that work on their own is lazy and irresponsible. Imagine a tour guide who tells his tourists to “just go over there and look at some stuff,” and “come back when you’re done.” Links can be that disorienting.

Links disrupt the reading experience, and that is the second reason for why links are bad. It is possible that you want the reader’s experience to be disrupted. But in many cases you don’t. And the reader is already distracted by the proliferation of parallel windows and devices that augment their reading in some way. Do we need to make that distraction easier? Should I link the Wikipedia article on media multitasking or is it enough for my purposes to simply mention Wikipedia, or to trust my reader to look something up later, in a reference source of their own choosing? Or better yet, should I help the reader along by summarizing the findings? It mentions that most folks already read with a second screen in tow. It is not that unusual to see someone look something up on their phone or tablet while reading a newspaper or an e-book. Why? Because they don’t want to leave the flow of the first screen. There is great pleasure in immersive, uninterrupted reading.

Besides being disruptive, links are ugly. They are ugly together, as in when many links conspire to produce a tangled mess. And they are also ugly when naked on their own, like this: That string of characters is not meant for human consumption! The period at the end kills me entirely. Meaningless punctuation inside of links coupled with regular punctuation ruins the sentence and the paragraph. Of course, I could just tell you to read something on Google Docs. That looks much better, but then we are making the opaqueness problem worse by hiding the address behind words that may or may not be related to the destination. It seems that we are stuck compromising on either transparency, reading flow, or visual impact.

Links aren’t very secure to begin with, but hiding links behind words further compromises security. You’ve probably heard of link-baiting: the purposefully malicious attempts to trick a reader into revealing personal information when following a link that masquerades as a legitimate destination. You can visit my site to learn more about link-baiting. You shouldn’t have clicked that! (Don’t worry, that was the real Google login page.) But even if one means well, viruses and browser exploits can inject bad links into your otherwise legitimate ones. A common technique is to install a browser script along with some seemingly useful “search bar” that will redirect all legitimate links to a site that makes money by advertising. Worse yet, you could end up on a site that attempts to further compromise your computer. Links are not secure because in linking, we outsource the relationship between reader and content to the browser.

Links are opaque, disruptive, ugly, unsafe, and they rot. Links don’t last because the content at the address is dynamic. It is not guaranteed to be there decades, months, minutes after your initial visit. In that case, why even bother? The link works best for ephemeral output (like a tweet). We must think of something much more robust for any “serious” writing that hopes to survive to the end of the week. And for the really good stuff, the kind of stuff that is the purview of librarians, we need to cultivate sustainable, long-lasting, responsible practices of online citation. It should work as well, if not better, than the familiar bibliographic citation in print. This practice should combat digital decay, not aid it. We need to think about the ways our links can be accessed, mined, and preserved with the archive-grade zeal of the rare book librarian.

Finally, links are terrible for accessibility. It is bad enough that clicking on a small word like this is difficult for people with any sort of fine-motor control problems. Being a little older in itself can make the online reading experience painful. Things are much worse for those with Parkinson’s or for the blind. Sina Bahram, a blind usability expert (who is himself blind) reports that some sites contain thousands (!) of links in advance of actual content. Screen readers for the blind must read each one of them out loud. For the screen reader, there is no difference between garbage links and useful content. If you thought looking at links is disruptive, imagine listening to a robotic voice that pronounces every slash and every useless number in: And that is why Sina Bahram listens to his reader at 950 words per minute.

Any one of these issues alone should give us pause. Together, they are a cause of grave concern. How did we get here? And what can we do to make links good again?

How did we get here is not an easy question. A part of the story is surely the excitement we once felt about hypertext. Links were supposed to break the hegemony of linear narrative, ushering in a new interconnected world. To some extent the dream came true. But links also brought with them such things as Search Engine Optimization (SEO). Google’s PageRank algorithm tracks, among other things, the number of incoming and outgoing links. This bias for connectivity encourages “link farms”: sites that attempt to game the system by aggregating links or cross-linking their own content. A sure sign of a vacuous SEO-driven piece of writing is a certain cynical and strategic use of links to other popular sources. How long until the SEO logic infects poetry, fiction, or investigative journalism?

What can we do to make links better? There are a few things we all can do now. First, let’s use links sparingly. Think smartly about whether you need to link or whether you can make do with a good, old-fashioned quote or citation.  Don’t link just because you can. Second, link explicitly: is better than this. Third, realize that online content is dynamic. It makes no sense to link a dynamic resource when the intent is to create a link to a static version of a document. Tools like the Save Page Now service, hosted by the Internet Archive, do just that. You can find this essay at but its earlier draft is best captured in a snapshot here: Finally, do not neglect the humble footnote. Footnotes provide a nice blend between usability, transparency, and good knowledge design.1