Wednesday, June 29, 2011

Collecting vs. Curating

I just came across Frank Chimero's blog yesterday and I feel like I stumbled in on a divine secret.  Frank writes such great stuff about design.  His writing is so elegant—clear, focused, and totally current.  He wrote this fantastic post called Sorting a Mass, which is about how we collect, identify, and sort information on the internet.

We all bookmark, flag, collect, and classify items that are important to us.  Much of the content generated is the amassing of things we've seen elsewhere.  The web makes it possible to classify information like never before, but it's also an inefficient system.  What it seems to come down to, at least where personal content classification is concerned, is the difference between collecting and curating.

In Chimero's post, he sends out a query on the difference to Twitter and receives many responses.  Everyone seems to agree that:

Collection is additive. Curation is subtractive.
Collecting is for yourself, curating is for others.

A mass of items and an audience seem to be the commonalities, but it gets tricky: one can have a collection viewed by many (a Pintrest wall, a collection of Flickr favorites).  That collection has an audience, so is it curated?

Chimero posits that the delineation should lie with intent (italics mine):
I think the thing that separates the two is intent, not in the size of the pool of stuff or in the presence of an audience, but rather in the implied purpose of the gathering of content. To me, collecting is usually about propagation—the collection grows in quantity and diversity. Curating seems to be about illumination, and having that set of selected items, no matter the size, come to a point.

I’ve recently launched another blog called The Mavenist, which could be considered a collection, but it feels to be squarely in the realm of curation. What is the difference? It is in the arrangement. Collecting is having all the pearls lose on the table, curating is stringing them together into a necklace. (Of course, there is no better or worse between those two situations. One just has an extra pass of arrangement.)
I love this.  One of the reasons why I blog is because I like going through that process of selecting a design I like, processing through why I like it (or don't), and explaining that reasoning in this space.  Curating takes more time, more effort.  It is an art, very much like editing.

I remember that in a past writing class, we were instructed not to say "I like this" or "I don't like it."  Offering the reasons why something worked or didn't inside the story was the only feedback allowed.  This kept things objective, and taught us to examine why some things fit and some didn't.  That process has become very useful for curating.

It should be noted that I do also really enjoy collecting.  I have a group of Flickr favorites and a number of Pinterest pinboards.  I agree with Chimero that we are prone to collecting, rather than curating, and with the web muddying up how we do both, that presents some interesting challenges:
I think it’s important to note that while we may collect for ourselves, those collections are being published, creating incidental audiences. I collect my Flickr favorites for me, but they are accessible to you, so what obligations do I have to you when I consider to favorite something? What happens when sharing is the default rather than purposeful? It blurs the line between curation and collecting into something more of a gradient than a binary this-or-that. And that’s why those neat definitions in the tweets seem slightly off. The situation isn’t as tidy as it used to be.
He's right: it's not.  We used to just have museums with big collections and curated exhibits of those collections.  But now more information is accessible than ever was before, and what are we going to do with it all?  Especially since the mechanisms to analyze all of it are heavily skewed towards chronology and popularity.

Google has attempted to make it so that it indexes usefulness, and I suspect that is the ultimate goal; however, frequency of keywords can only do so much.  If there was a way to index not just whether or not someone "liked" a piece of information, but also could show that it helped them, or that they found it useful, that would be the golden ticket.


  1. Great post; I love it when I get good food for thought.

    Collecting is a necessary (but not sufficent) step for curation. Someone has to collect for anyone to curate. In the physical world, this happens on a strict 1:1 relationship. Any given collection can only be curated by a single entity. If someone else reviews the collection and prepares it, they will be altering the previous curation. It's possible to compare different curations over time, but it's difficult to do.

    What's interesting about digital information is that since it's infinitely replicable, suddenly the relationship between a collection and a curator jumps from one-to-one to one-to-many, and it's not only possible, but easy to compare one curation to another, examing what gets culled from one set versus another. It also means that collectors can "crowd-source" curation, making large bodies of work available "at large" for people to sift through, remix, and arrange for presentation.

  2. I had guessed that a key aspect of curation (and not collection) is subtraction, but you and Chimero say even more. Chimero writes that "curating is for others"; you talk about usefulness.

    Of those two insights, I find yours the richer. At its most hollow, curation is simply tastemaking, which is all to often a tool for social hierarchy.

    But if we think of our tastes, our judgment, our discernment as being USEFUL, we can begin to think about how to turn the ideas and preferences that we've amassed into productive agents for whatever changes we want to see in the world.

    Thanks for spurring my (still forming) thoughts on this.

  3. The way Google works is actually trying to measure usefulness. They don't primarily use keywords to rank, but rather which and how many other pages link to the content.

    They don't have an explicit way to explain to whom the content at the top of the list is most useful, but they do track click paths to see which searches result in clicks on which links. So they are far more than a keyword searcher. That's why we all use them.

    There's actually an interesting contrast in search which is with Yahoo. Yahoo always believed at some level in human curation as a way of usefully indexing the web. But it's not scalable, so they lost to the algorithm guys at Google.

  4. Thanks for the comments. Alexis, I think Chimero had the same thoughts on scalability as yours. It's gotten so huge that algorithms won out. And as a consequence, there is quite a lot of static.

    Katie: I have the same issues with "tastemaking". Just because a bunch of people "like" something, doesn't mean that it's good. I suspect that it works similarly to the way Alexis describes Google working: that if people that are popular on the internet like something, that that thing becomes seen by wider and wider circles. Hence the whole "influencer" thing.

    I think it's dumb. I get tired of seeing the same content recycled over and over in my RSS feed and Pintrest wall. And it has nothing to do with "taste" really. It has more to do with the way information is recycled.

    It's also why I think curation is important: if I have an active interest in curating, I might edit out the stuff I see over and over again, and whittle down to the good nuggets.