In thinking about a pair of presentations I last spring, one at the 2008 Libraries in the Digital Age Conference, and one at ALA Annual, RUSA's pre-conference, "Reinvented Reference IV", I came to decide that the notion of reference service as an exchange of a question and answer is a false construction. Oh, it's an incredibly useful false construction. It gives us a vocabulary to discuss our work and tells us when to make hash marks, and sometimes it even happens.

I think a broader (and paradoxically, more accurate) characterization of reference service is that people have stories, and reference is what happens when people decide to involve a librarian in their story. Dave Lankes told us that librarians reference was a 'sea of chaos' and that librarians thrived in it. We are the best at answering questions in context.

One of the promises of a statewide, multi-type collaborative reference service is that a vast network of subject experts will be able to provide patrons the best answers about any topic at any time.

If only because we never challenged the idea, it has always made sense that our chat service provide general reference (and lets face it, homework help) and that our e-mail service provide the subject expertise.

E-mail reference has always been a part of our service, in part because the recommendations in the 2002 evaluation of Oregon's implementation of LSTA specified that we not replace the library-to-library referral service that has existed in Oregon at least since 1994.

But, we've never quite lived up to our potential as subject expert utopia.

In part, we've always emphasized our live chat service - we're online when you're online, chat live 24/7. Patrons, and especially kids, have responded, asking us over 21,000 questions last year, of which just over 95% were in our live chat service.

We are switching software in a few days. On September 29, we'll go from using OCLC QuestionPoint for live chat and e-mail reference to using a customized implementation (in partnership with Ohio's KnowItNow and the Northeast Ohio Regional Library System's AfterDark services) of an enterprise Instant messaging solution for live chat and ... what were we supposed to do about our small, but important e-mail reference service?

Five years ago, We started out using RefTracker from Altarama. I thought the software was quite nice for a single library to use to manage all of it's electronic communication, but at the time, it wasn't a good fit for us. The main issue was that the referral system demanded we choose a single person to send a question to, and we wanted to refer questions to libraries. The workaround we found was to pretend each library was a person.

When we switched to OCLC QuestionPoint, we got our wish of being able to refer to a library. Every librarian got a login, which were each assigned to a library, but it meant that only librarians from the same institutions were able to collaborate on questions. We could refer questions to other libraries, but except for local history/genealogy, law and health, it was really difficult to decide where to send them.

Thinking about it some more, if we did know intrinsically what libraries might have expertise in what subjects, our collective value and expertise is actually greater if we aggregate our individual, rather than institutional, expertise.

One of the strongest ideas I've heard and expressed lately about how libraries can and should collaborate is that our librarians work to provide a common service but generally work in isolation. Referrals are rare, and there have never been opportunities for librarians from different institutions to work on the same question. In 2007, one of our excellent law librarians, Laura Orr, suggested to me that we should seek systems that didn't have to be one-on-one. Dave Lankes echoed this thought at the 2008 Oregon Virtual Reference Summit.

This is a really long story, ok? Ok.

Last spring, I decided to try to use this website, and the content management framework it rode in on, to provide e-mail and other asynchronous reference services. The idea was that we wouldn't have to do any programming to build our system, only cobble together bits and pieces of other people's contributions to open source software.

Fair enough, though its somewhat of a lie. We did build an e-mail / asynchronous reference service on this website, using Drupal's built in content management tools, spam filtering and taxonomy framework, along with contributed modules for free tagging and e-mail processing. But I have done a fair amount of coding too, much of it probably unnecessary, and eventually I had to make some decisions about how the software should work.

What I absolutely didn't want to do was to make decisions about how librarians should work, though, even in saying that, I kind of did.

Here are the basic criteria for a multi-type collaborative reference service I came up with:

• encourage use of subject expertise
• be able to connect a patron with a specific library
• encourage collaboration between librarians

Which is to say, we're crowdsourcing reference.

Crowdsourcing is the title of a recent book by Jeff Howe that I found enormously boring, but generally it means using a big group of people to accomplish a task that a single person or small group usually does. Like write encyclopedia articles, create software, or answer reference questions.

You might be familiar with sites like Yahoo! Answers, where anyone can ask a question and anyone can answer. The askers are the answerers. It's really neat because all kinds of people do really know the darnedest things, and because all kinds of people really do like helping all kinds of other people, but it's not what I'm talking about. Remember, reference is what happens when a person decides to involve the library in the information seeking process.

One of the things Howe talks about in his book is that isn't enough to have a crowd perform a task - you have to have the right crowd. I believe that we are the right crowd for providing reference service.

So here is what our asynchronous virtual reference software does:

Patrons can now e-mail us directly, fill out our online form, or have a live chat referred to e-mail for referral.

Librarians are notified of new questions by visiting our site, checking RSS feeds, or by e-mail.

Once we get the question, librarians can then organize questions (by tagging), discuss them (by commenting) and respond to patrons (by sending a clarifying question or answer). If you haven't been to one of our trainings lately, you'll get to see what it looks like when I can find time to writing up some documentation.

That's pretty much it. The rest is up to you.

And while this post itself may seem like it will be longer than the concatenated text of the 1,000 odd e-mail reference questions we get this year, there are two really good reasons why I think we'll use this system a lot:

First, with no one backing us up all day, patrons will be given the opportunity to 'leave a message'. I think a lot of messages will get left.

Second, text messaging is hugely popular, has great potential for reference service, and is basically an asynchronous service. It is more like e-mail than it is like chat. When we add text messaging to L-net, it will be on top of this system.

And third - I know I said there were two reasons - third, our "e-mail" reference service is a website. Patrons can e-mail us, but generally, they fill out a form our website, we fill out a form on our website, and they check their e-mail to read what we wrote. A much more organic way of doing it would be to let the patron simply read the answer on our website.

Web 2.0! I hear it's going to be big!

I'm excited, and I hope you are too.