I totally support the idea of a Curator’s Code, which is basically the use of microsyntax to represent different kinds of attribution in posts and or tweets:
ᔥ Maria Popova
The system is based on two basic types of attribution, each connoted by a special unicode character, much like ™ for “trademark” and for © “copyright”:
ᔥ stands for “via” and signifies a direct link of discovery, to be used when you simply repost a piece of content you found elsewhere, with little or no modification or addition. This type of attribution looks something like this:
↬ stands for the common “HT” or “hat tip,” signifying an indirect link of discovery, to be used for content you significantly modify or expand upon compared to your source, for story leads, or for indirect inspiration encountered elsewhere that led you to create your own original content. For example:
But I think the folks behind this made a few mistakes, ones that will kill the adoption of this Code.
First, there are well-established textual ways to accomplish what is envisioned, like via and h/t. They might have been better off to simply throw their weight behind a more uniform usage of those terms.
Second, the characters they picked for the microsyntax are hard to use. The sideways S, for example, doesn’t appear on Apple’s special characters. I tried using their bookmarklet, but it wouldn’t insert into Tumblr’s editor. Wouldn’t it have been better to use some sort of a ‘v’ for via, like shift-option-V = ‘◊’? Or maybe some sort of arrow? Like ‘←’?
I am emotionally in favor of this, but as a microsyntax observer, I don’t think it will catch on.
I think we need a manual of style for Tumblr. I am a fan of bottom-up order, but at the same time a lot of serious work is being done in Tumblr as well as casual reposting of cute cats and unicorn hats.
Consider just one issue: attribution. There are a wide variety of techniques in use on Tumblr for attributing when quoting or reposting other people’s works. And some are less good because they break the thread of connection from a new post or repost back to the initial source.
Tumblr both helps and hurts this. On one hand, reposting (or reblogging) something that you see in your Tumblr stream is subject to automatic formatting and the creation of the chain of Tumblr notes attributing backward to the original source. But the formatting options aren’t settable: I can’t turn off automatic nesting of blockquotes in text reposts, for example, although I think it is the wrong way to do it.
Leaving aside the automatic issues, there is no consistency in how Tumblr authors make attribution.
Here’s a post that I published recently. [Note: you don’t need to focus on the quote, just look at the attribution at the bottom of the post’s image.]
You can see that I give attribution to azspot. I saw when reblogging that the original quote came from this link
which wasn’t obvious when looking at azspot’s blog post. [Note that I am not criticizing azspot, I am using him as an example for illustrative purposes only.]
azspot’s post was this:
Note that he attributes the quote to Wendell Berry, but if you click on the link embedded in Berry’s name, you come to the source blog post, and one that is not written by Berry. It turns out to be a non-Tumblr blog, which could explain some of it, but it should still be cited anyway.
Here’s that initial post, made by jdaviddark:
So I edited the text that azspot had used to for the link, to ‘Wendell Berry,The Poetry of William Carlos Williams of Rutherford’.
[Note there is an attribution problem embedded in this post, too, because the photo has no information associated with it. It turns out to be Williams, but it might just as well have been Berry.]
The end state is as you see in my post. The original quote is properly attributed to Berry’s book, jdaviddark gets credit as the original digital source, and azspot is credited as the curator that brought Berry’s quote to my attention.
And that’s perhaps the point of this long-winded discussion: Tumblr authors — either manually or by the mechanisms built into Tumblr — should be clearer about what sort of attribution is involved when reposting things.
For this reason, if no other, I hope that Tumblr finally gets around to making a break between the original material captured the first time someone creates a Tumblr post based on material outside of Tumblr, and the comments that people write when adding their two cents at the point of reposting. The fact that we have three things lumped together in a big mess:
I can imagine various ways to simplify this complexity, but the simplest course is to amplify the notes with an optional text region where people can add ‘recomments’ at the point of reposting, and to make the original source content uneditable, so the original post is conserved as it was created. After all, if someone wants to clarify the provenance of a post they have seen — as I did with the Wendell Berry quote — then can follow the link, and start over with another original post, with a manual nod to a curator, instead.
And if you look at the notes on the azspot post you see this
Which doesn’t make it very clear what has happened when I reposted and changed the attribution, at all.
IN CONCLUSION, the attribution problem is only one example of the need for a manual of style, or a Tumblr handbook, perhaps. I could tell you how and why I used stoweboydpix.tumblr.com as the repository for the images in this post, for example, but that is a story for another day, or chapter.
I’ve commented numerous times (and so have many, many others) that the approach taken by Tumblr to attribute content in reblogs is a mess. Well, they have fixed it, and have taken the innovative approach of creating attribution metadata for posts. This turns out to be something like what I recommended in Tumblr’s Reblogging Mess Solved, but not exactly.
Here’s what the Tumblr staff blogs says:
Fixing Content Attribution (Once and For All) | Tumblr Staff
Back when we launched two of Tumblr’s most unique features, reblogging and the Tumblr Bookmarklet, we devised automatic “via” links in post captions as a simple solution for attribution. Three years later, this solution has gotten us pretty far. But it’s easy to spot some real shortcomings:
• It’s hard. Even with the best intentions, it’s possible to mangle attribution when reblogging. Links get dropped, and credit gets buried under reblog links.
• It’s ugly. Reblog links pile up. Credit is formatted as any permutation of url, author, blog name, and page title.
• It doesn’t play well with content that doesn’t originate on Tumblr. If I share one of Jacob’s Flickr photos, the post notes attribute me as the original poster. And posters too frequently mistakenly attribute content to re-publishers (Digg, BuzzFeed, 9GAG) instead of creators.
We know we can do better. After weeks of testing, we’ve got an upgrade that fixes all of these issues:
Starting today, reblogging will no longer insert attribution into the content/caption of the post except to quote content added by the parent post.
In practice, here’s what it will look like to readers. Below I have to version of a post where I selected text from a Susan Orlean piece at the New Yorker. In the topmost, I deleted the automatically generated link in the text pointing to the New Yorker, so the new approach to attribution was used. In the bottom post, I left the text link to the post, and the automatic attribution does not show.
I tested removing the attribution in the text after the first edit, and the automatic attribution did NOT reappear, so this approach only works during the first edit, as far as I can see.
I logged in as Underpaid Genius and reblogged the Susan Orlean post above, and in the first edit I was presented with an in-text attribution, as I has always been the case, but also an attribution pointing back to the New Yorker and to Stowe Boyd that cannot be deleted. If I delete the text attribution to Stowe Boyd those attributions remain:
And here’s a random person who reblogged the porst from Underpaid Genius — two degrees away from the original post:
The only thing that seems to be missing — at least in the automatically injected version I have fooled with so far — is that the HTML for the attribution is not styled. I think Tumblr should wrap the attribution up in a ‘source’ or ‘attribution’ span so that the text can be styled. But I guess I will wait for David Sutoyo to gen up a new version of Schema (this template) with attribution built in.
If you notice the attribution on this post, at the end, you’ll see one glitch: I quoted the Tumblr blog and selected a hunk of its text near the top of this post. Tumblr automatically attributes that but didn’t insert it following the text, so the attribution is at the end of the post. Seems like it is crying out for a footnote model, doesn’t it?
We will have to see how these changes are received by the Tumblr community, but is seems a much more esthetically pleasing and reliable mechanism than what preceded it.