You know how they say that to someone with a hammer, all problems start looking like something that can be solved by pounding?

We librarians run the same risk in our own thinking. As wonderfully flexible as our minds are, we still often tend to head one place, and one place only to answer questions--our reference collections. That's what we know well, after all, and we're used to making the translation: that's a "what is it?" question, about math, so I need a math dictionary or encyclopedia. Which is indeed just fine as long as we have a math dictionary or encyclopedia. What if we don't?

This is the point at which some of us will head to the net, on the assumption that there are a great many specialized dictionaries and encyclopedias online, so why not one on math? And indeed, that strategy is going to work.

But sometimes the answers we are looking for are in our own collection and we just don't realize it. Does anybody but a gov docs librarian ever remember to use the gov docs collection for reference? Does anybody but a children's librarian think to use the children's collection (which is an unparalleled source for answering "what does it look like?" questions)?

Probably only a serials librarian think of periodicals as a wonderful source of statistics (Survey of Current Business, for instance), schedules of events (New Yorker, Chronicle of Higher Education), documents (e.g., the Pope Speaks), directories of suppliers (e.g., Library Journal), the pros and cons of public issues (CQ Researcher or Congressional Digest, for example).

But if what our patron needs is just not anyplace to be found in our library, and we can't find it on the web, we need to think of someplace that logically would collect that information and refer them. We might send them to a city Information and Referral service for social services information, or to a local labor union office or food bank for information about the last economic downturn, or to a historical society for local history materials we don't keep.

We hate telling our patrons "we don't have anything." Unless, of course, we can at the same time tell them, "here are some people who might." Which means thinking beyond our particular little box, the walls of our library.

How do we build this kind of flexibility into our minds? Here's something I do to keep me from falling into the "one place and one place only" trap (and yes, of course I have this problem myself). Each week, I go back to two or three reference questions I answered perfectly satisfactorily, and try to find the answer in a different way.

If I found it on the net, I'll try to find it in a reference book or periodical, and vice versa. If I found it in one database, I'll try the search in some other databases and see what i come up with there. I might look in the government pages or yellow pages to see if there's some local agency or organization that would have the expertise or resources to answer the question. Try it. It's a good way to force yourself to stretch your reference muscles and discover the reference sources you didn't realize you had.

---

PAYING ATTENTION TO OUR USERS

In some libraries, the person who does the computer design and technical stuff doesn't have time to work the reference desk anymore -- or maybe has never worked a reference desk. Which is a pity, because it separates the person who's designing the interface from the users of the system.

Some things I have learned about our students, by watching the way they look for stuff, have important service implications for us.

• A student came up to me once after using Social Work Abstracts, nearly screaming, "I keep clicking and clicking on this article and it won't come up! What's WRONG with your machine?" Of course there was nothing wrong with the machine; she was just in a database that only gave citations and abstracts. But the internet, and various full-text journal databases, have led people to the conclusion that what they want is always there, full-text and clickable.

  That means, whether we want to or not (and there are good and logical reasons why we might not), we are going to be forced to consider adding full-text databases to our services.

• Students seem to prefer to work in pairs or groups. That has implications in terms of both space allocation and noise.

• Watch people as they approach the web or any database, and you will realize that many of them do NOT want to read text, and do NOT want to scroll down the screen. All they want to do is type what they want in a box and get it. That has obvious implications for the way we design search interfaces. Above all, it means we are going to want to put a searchbox in plain site and make the whole site keyword searchable.

• Most people have no conceptual grasp of the difference between databases and the internet. An interface that offers a whole cluster of icons to access different services is likely to confuse users who have no idea what the differences between them are.

• Many users don't realize that when they don't find something on their subject, it's because they're doing a subject heading search with a word that is not a descriptor. We can increase people's chances of finding things if we make the search screens default to keyword searches.

These are things I have observed with our patrons, who are mostly college students. Your own users may behave differently. But we need to pay attention to them, because the closer our search systems match the ways our users look for things, the more likely they will be to find what they want. And the more likely they are to think our library is a wonderful place.