Hapgood

Mike Caulfield's latest web incarnation. Networked Learning, Open Education, and Online Digital Literacy

A Simple Proposal for Killing Comments with Annotated Links

I would be interested to see what would happen if someone reconfigured blog comments as follows.

  1. Don’t call them “comments” call it “related pages” or “link annotations” or “Community Links”
  2. People have 140 characters and a link box at the end of the page.
  3. The way link annotation works is this:
    • You, the reader, read the post
    • You either find a related page or you write a page in your own space related to the post
    • You plug in the URL to the URL box
    • You have 140 characters to explain how the linked page responds, contradicts, or expands on the post you are annotating

Ideally you’d also have a mechanism to encourage reusable pages, e. g. instead of linking to a page that says “Here’s why Post X on CompStat is Wrong”, you’d link to a page that doesn’t mention Post X explicitly, but itemizes the reasons why sociologists no longer take the Broken Windows Theory CompStat’s model was built on seriously.

This linked page could be referenced from many articles, on other subjects, with the 140 character text of the annotation providing the localized segue, e.g. “CompStat was adopted in the heyday of Broken Windows Theory, a theory since discredited. See [[Broken Windows Theory Broken]].” where Broken Windows Theory Broken linked to your page (or the page of someone else).

The way the resuse incentive could work is this — if multiple people link the page from multiple other pages, the annotation floats to the top and a visual indicator shows that this page is in general use, not just an extended ranty reply to the post. If multiple people link it, but all form this page, it shows it’s considered a useful reply, but maybe specific. If one person links it and only from this page, it’s maybe a comment.

You could build a better set of incentives, perhaps, but that would get you started.

What would happen? I don’t know. Maybe people would still route around restrictions and find ways to use it to comment instead of extend and expand on things.

On the other hand, maybe a host of things would change, especially if the commenting had central analytics. You’d be able to generate a set of suggested reading for users based off what they had read, essentially crowdsourced. People could rate annotation links for relevance, and the results would form something close to a semantic map of the web. The 140 characters wouldn’t give you the Semantic Web, but it’d provide more signal and less noise than current approaches to linking and authoring do.

I’m focussed on federated wiki right now, so I’m not working on this — this is really just a sliver of what fedwiki does.

But I’m curious if someone has tried this. It seems to me the same way that Tumblr revolutionized blog commenting by “post-in-your-own-stream” behavior this set of small restrictions and incentives could radically reinvent the comment as a annotated, semantically dense link, which has all sorts of implications for both discourse and analytics. Has anyone seen anything like this? What am I missing?

(See the earlier post on Reader as Link Author for why reader-produced links are important.)

UPDATE: And we’re on our way!

Screenshot 2015-08-18 at 6.00.21 PM

mikecaulfield's avatar

Posted by:

mikecaulfield

The infolit guy.

17 responses to “A Simple Proposal for Killing Comments with Annotated Links”

  1. CogDog Avatar
    CogDog

    Sort of like a micro-trackback?

    Speaking of discourse, I am battling my with a discourse.org install Some relevant features when reading someone else’s post is “reply as linked topic” so it creates the relationship via a link, and you start a new “topic” or thread.

    It also seems to track all other replies anywhere in the site and lists the links into a topic.

    Reply
    1. mikecaulfield Avatar
      mikecaulfield

      More like a reverse trackback — trackbacks say “this was mentioned somewhere else” this says “I’m mentioning this thing as relevant, if if there’s no specific callback”.

      Trackforwards, perhaps?

      Reply
      1. CogDog Avatar
        CogDog

        Gotchya. Slow brain today

  2. Derek Bruff (@derekbruff) Avatar
    Derek Bruff (@derekbruff)

    Sounds useful, but, I have to admit, I kind of like comments. I miss discussion among interested parties right below a text.

    Reply
  3. billseitz Avatar
    billseitz

    The IndieWeb folks are playing with varieties of links in all directions across personally-controlled bloggy sites.

    This TantekC post has a Reply button which uses Twitter to reply (to his tweet spreading his post), in which you could put your Comment plus your own page’s URL.
    http://tantek.com/2015/230/t1/npsf-trackish

    Reply
    1. billseitz Avatar
      billseitz

      Note that all these like-trackback approaches have Spam problems.

      Reply
      1. mikecaulfield Avatar
        mikecaulfield

        Yeah, but so do comments as you know. I think I have something like 13,000 spam comments in the dashboard of this blog since 2007.

  4. billseitz Avatar
    billseitz

    I’m pretty comfortable with just sticking Disqus or the equivalent at the bottom of every page, and let comment/linking happen there. If you (the original author/poster) find something useful in those comments/links, you end up updating your page anyway, so even if comment-service disappears in the future you’ve gleaned the best stuff.
    http://webseitz.fluxent.com/wiki/WikiWebDialogue

    Reply
    1. mikecaulfield Avatar
      mikecaulfield

      I guess it comes down to the fact that commenting is really really easy, and coming up with links is harder — people follow the path of least resistance. So I’m jut curious what would happen if you changed the resistance ratio,

      If you think of Twitter, for example, you could always have blogged on Blogger in 140 character increments, nothing stopped you. But restrictions caused people to think about what they were doing in a different way. Ideally it might create more WikiWebDialogue, right?

      Reply
  5. Jake Sandlund Avatar
    Jake Sandlund

    This is similar to the Chrome extension I showed you a while back, but I like your description of how it works better. I was thinking it would be more an impersonal wiki of links at the bottom of the page, sort of like a subreddit just for that webpage. I like the more personal tweet-like feel, plus having 140 characters gives you more room to explain why a link is relevant (more room than a link title).

    Reply
  6. D'Arcy Norman Avatar
    D’Arcy Norman

    Manual trackbacks. People don’t even use automated trackbacks anymore…

    Reply
    1. mikecaulfield Avatar
      mikecaulfield

      Trackbacks I think are fairly different.

      * Trackbacks are most often used to show the conversation you are in.
      * You follow the trackback and you continue the conversation
      * I’m sick of conversation. I’m sick of read these ten things to understand my thing.
      * What I’m describing is not what trackbacks were, but what ordinary links were originally imagined as. Not a way to tie together conversations, or to “share” things, but a way of tying together ideas, and as a way for READERS to spot those connections.
      * And that’s the piece that most people miss, because the idea of a web based on related meanings and significance has become so overwhelmed with people using it as a CB radio, which in my more despairing moments seems an endless narcissistic pit or eternal now-ness. And in my less-despairing moments it seems like its aggressively mining only one thing that it can do, only exploiting a fraction of its capability.

      Reply
      1. Chris Lott (@fncll) Avatar
        Chris Lott (@fncll)

        I’m missing it because I still can’t picture how it would actually work in a way that truly differs from what we have. It would be interesting to have a mockup of what this would look like in practice, because that matters a great deal (and I am constantly seeing references to “what links were meant to be” with little or no explanation of how, beyond the abstract description, it might actually work in practice).

        You speak of “tying together conversations” and “tying together ideas” as if they are obviously different things. Perhaps they are, sometimes, but often they are not. Similarly, I’m not sure that “READERS spot[ting] these connections” is obviously different from the way readers often do that now through links and comments. Decentering the conversation = cool sounding idea, but not much to go on in terms of what happens when fingers are at keyboards and eyes are on screens.

    2. Chris Lott (@fncll) Avatar
      Chris Lott (@fncll)

      I hear that. The keys are incentives and friction. What’s the “reward” for coming up with a link? What’s the mechanism by which I can—in as close to one step as possible—create the page-y, comment-y thing on my site and get that in the stream of the original piece of content?

      The invocation of hypothes.is resonates: I routinely think I should highlight and comment with it, but to be useful as a “comment” I have to then link back…and the 2-4 step dance generally means it doesn’t happen. I’m a slacker.

      Reply
      1. mikecaulfield Avatar
        mikecaulfield

        I guess I’m coming out of a year and a half of doing this for myself, it’s very clear to me what the feeling is because I get it everyday now. But the process was just kind of developed by Ward and I and others in the happenings over time, and you are right — what does the interface look like where this is evident right away instead of learned over a series of months?

        As far as linking ideas and conversations, the two are not mutually exclusive, but there are some pretty simple distinctions. Conversations have their pieces grounded in the utterance, the time, place, and previous sequence of statements up until now. A good conversation will often try to heighten that element (and in linguistics we even had markers like anaphoric reference to see how densely a statement was tied to the surrounding speech context.

        When I say linking ideas, what I mean is writing consciously in a way that you know will be arrived at from many different directions, some of which you don’t anticipate. These differences are often coded in the literature around genre as being the difference between a rhetorical/expressive mode (focused on the connection between interlocutors and the desired aims of that) versus a more expository mode (focused on the relationship between, in the words of Jakobson, message and referent).

        Again, I think this is hard for me, because I know it to be a very different sort of thing, but I’m not sure how to get the technology to point here more directly. We don’t need a mockup for this sort of thing as much as a translation. The question in a sense is “what does it look like” as much as “how can it be incentivized”. But maybe that’s your point?

        I do think one way to frame these discussion is to think of non-obvious ways (like this) that hypothesis could be used to run comment sections.

  7. What is not being said and where is that elephant? | doublemirror Avatar
    What is not being said and where is that elephant? | doublemirror

    […] you and challenge your ideas more easily, or even dislike you and  find your ideas inspiring? Mike also talks about ways to re-invent comments to avoid stereotypical thoughtless commentary. May be a shift to commenting on one’s own […]

    Reply
  8. Frank blogged on December 16, 2018 at 10:23PM | Notes From The Cave

    […] A Simple Proposal for Killing Comments with Annotated Links – Hapgood […]

    Reply

Leave a comment