We’re forgoing our usual format this week to do a deeper dive on signals in a space that’s very close
Addressable ideas & the connective tissue of the web
We’re forgoing our usual format this week to do a deeper dive on signals in a space that’s very close to our hearts (and work) — reading and writing on the internet. We look at the foundational affordances for writing and linking on the web, and explore how some nascent alternatives might change the way we share ideas and have conversations.
—Alexis & Matt
Back in 2015, in our explorations at The New York Times R&D Lab, Alexis wrote an essay entitled “The future of news is not an article”. In it, she discussed how we have conflated the artifacts of news with their content. The article as a format is primarily the result of constraints of print journalism, and has serious constraints as a method for codifying, sharing, and discussing ideas. Instead of articles, she suggested that we think about “particles” — the concepts, statements, and ideas contained within those narratives. How might we semantically structure, point to, and remix those particles to create new knowledge or conversations?
But let’s step back: these assumptions don’t just apply to news, but to the internet as a whole. Just as articles have been the atomic unit of journalism, pages have been the atomic unit of the web, and those pages are interlinked with urls. Those links form the connective tissue by which we start a conversation, build on each others ideas, reference inspiration, or have a debate. These two structures — pages and links — form the basis for our assumptions about what the internet is, how it works, and how we engage with it.
Particles & bonds
Azlen Elza frames writing as two things: distilled thought, where writing forces you to make your ideas more specific and concrete; and conversation, where you can work through ideas in discussion with others. “Through conversation you can think thoughts which might not have been possible to think through a different method of thought, and vice-versa.” The best version of the web affords both: ideas that can be distilled through the process of writing, and then further explored by being interrogated and remixed in conversation with others. In his “Proposal for near-future blogging megastructures”, Brendan Schlagel says: “Sure we have hyperlinks, and even some esoteric magic with the likes of webmentions. But I want big, simple, legible ways to link blog discussions together.”
In order to afford this richness of conversation, we need two things: we need a way to point to particles, to address ideas at a more granular level than the page, and we need those pointers to provide the right amount of context about the connection between the two ideas or spaces. If particles are the fragments of ideas found in larger texts, then bonds are the links between them given weight, strength, and meaning.
Many of the concepts we’re discussing aren’t all that new; as far back as 1960 Ted Nelson theorized a bi-directional, non-linear hypertext scheme called Project Xanadu that would allow ideas to be embedded into works, to reference sources, and have sources know about the works in which they had been cited. It’s a far more robust and informative view of hypertext than the one-way linking of the web, but it’s also far more complicated to build. (An early, incomplete implementation was released in 1998, and a “working deliverable” in 2014. Neither had much of an impact.) A somewhat more visible (and far more useful) experiment was quietly released on The New York Times website in late 2010; Michael Donohoe, a developer at The Times, deployed an open source library called Emphasis that let users link directly to a particular piece of text and highlight it when that link was followed, turning a single page into an infinite collection of idea particles.
Future signals: Text fragments, quotebacks, and more
Identifying the particles of ideas embedded in long text is difficult, but can be solved with advanced text parsing, more deliberate location definitions within text, or a combination. Adding information and context to the hyperlink — strengthening the bond — is a harder problem, but one that has recently seen some compelling innovation.
Much of that innovation has been happening in the context of personal software for note-taking and creativity. Obsidian will create a knowledge graph from specially-marked text files on your computer, and Coda and Notion allow for both simple documents and complex databases to be constructed with no code and shared among small groups, replete with links out to documents and texts. Recent experiments using Roam propose to go a step further, allowing you to see two related documents side by side with their interconnections highlighted in the space between, showing how one idea leads to its refinement or elaboration.
Adding on collaboration gives us new possibilities, using spatial rather than semantic linking to facilitate remix and discussion. MakeSpace is still under development but promises to combine video collaboration with live snippets of websites, allowing participants to embed themselves in the space of ideas for real-time ideation, and the ability to link ideas to the spaces around them. Preserving those interactions as visual-spatial artifacts also facilitates recall, and maps to how our brains make associations.
Evolving the written web
These are just weak signals at the moment, and may continue to be a side conversation in the larger trajectory of the web. But if they gain momentum, there is compelling potential here. These approaches codify the value that synthesis and conversation bring to ideas, providing the context for assemblage. They point to a future web beyond pages, where ideas themselves are addressable, remixable, and relational, and the connections between ideas carry meaning and context. The network can become a graph of concepts and conversation, a space for exploration that provides new ways to see and understand.
Did you enjoy this issue?