That "things link to other things" is a golden rule of the internet, and accounts for the "web" in "world wide web." It's easy to forget that the humble hyperlink was once a revolutionary concept, and created this writhing, evolving organism. Because of this, when you hear about a dark "web," you might imagine that it's an interconnected and active community, like the web you're used to.
But, according to new research from the Massachusetts Institute of Technology's SMART lab in Singapore, posted to the arXiv preprint server this week, the dark web isn't a web at all. Instead of being made up of sites that link to each other, allowing people to easily jump from page to page, the dark web is more like a collection of self-contained silos.
It's just one more example of how the dark web as it exists in the popular imagination—a bustling underground market in some night-shrouded dystopia (and there definitely is some very nefarious activity going on there)—isn't quite what it seems. It's actually a sparse and isolated place, perhaps because of how dark web users approach their privacy.
Read More: The Motherboard Guide to VPNs
"It is an interesting finding about how people create bonds when there is no trust, but fear," Carlo Ratti, director of MIT's Senseable City Lab and co-author of the paper, wrote me in an email from Singapore. "As a result, however, there is an impact on how people use the [dark web], as they cannot surf freely from one site to the next."
After crawling the dark web, Ratti and his colleagues found that 87 percent of sites on the dark web don't have a single link out to another site. Of the sites that did, a third of the time it was just a single page linking out. "All together, the onionweb [a name for the Tor network that makes up the dark web] is a sparse hub-and-spoke place," the authors wrote.
"The term 'dark web' is commonplace, but based on our analysis, the 'web' is a misnomer," they continued in the paper. "It is more accurate to view it as a set of dark silos."
Interestingly, the researchers also hypothesize that this sparseness isn't due to any technical limitation, like the fact that the garbled URLs on the dark web are notoriously impermanent. Instead, they surmise, the reason must be a social one having to do with the mindset of people who use the dark web. This makes some intuitive sense, since the technology that underpins it, the Tor network, is designed to protect users' privacy, so the people who decide to use it might be a bit more paranoid about building connections.
The current paper is just a start for the work, Ratti wrote, and future research might shed more light on why the dark web isn't much of a web.
"As next step," he said, "we are planning to develop a model to explain how a network develops when nodes do not trust each other."
Subscribe to Science Solved It , Motherboard's new show about the greatest mysteries that were solved by science.