How to use Shelfmark

Shelfmark answers two questions that UCLA’s own search leaves to you: which library holds the thing, and which shelf it is on once you are standing in the building.

Where things are

Shelfmark has four pages. The bar at the top of each one links to all four, so getting from any page to any other never costs more than a single click.

Everything starts in the one box

Earlier versions of this tool gave each kind of question its own box, which meant you had to know what sort of thing you were looking for before you could look for it. That is a bad assumption at a reference desk. There is one field now, and it decides what you meant:

Whatever it decides is named on screen, and one click sends the same text down a different path. The alternative would be a tool that guesses wrong and says nothing about it, which is a worse failure than simply asking you which you meant.

Call numbers and the shelf map

Type a call number as printed. Spaces and the Cutter dot are optional, so QL737.C22, QL 737 C22 and W1 JO600 all work. A W1 number needs the space after W1.

The locator then finds the shelf whose range contains that number. It compares the class letters first, then the class number, and then each Cutter as a decimal, which is why AM4733 sorts before AM477. Cutters being decimals rather than whole numbers is the part that catches people out, including people who have shelved for years.

Reading the map

• Each green column is a double-sided shelf, with an L and an R face.
• Black is the bottom row.
• A dash means a face nobody has surveyed yet.
• Where a range start equals its end, many volumes share one call number, so every matching shelf is returned and you check the spine.

Reference sits on floor 4 and Special Collections on floor 9. To search either one, switch the Biomed map pill on the map page. From the search box you can also type Hist Div in front of a call number and it will go to Special Collections on its own. Only copies in the Biomed stacks resolve to a shelf you can walk to. A copy at Powell or YRL is reported as being there and nothing more, because this site has no survey of those buildings.

Any shelf shown in a catalog result is a link. Click it and the map opens on that floor with the shelf already lit.

Books

Results are scoped to whichever library you say you are working at, and they widen to the rest of UCLA only when yours turns up nothing. All 21 libraries are supported.

UCLA’s catalog neither ranks results nor corrects spelling, so both of those happen here instead. A title that matches no record is checked against Primo’s dictionary first and against the catalog’s own vocabulary second. Any correction it makes is named on screen rather than applied quietly, which matters because a silent correction can send you looking for a book that was never the one you wanted. A typo or two will still find the book.

Filters and syntax

The Filters panel is shorthand for tokens you can type instead, and the Search syntax panel in the tool lists all of them. The ones worth knowing by heart:

Two words both have to appear. Quote a phrase to require them together.

Articles

Articles come from the same index UCLA’s own search uses, and by default only papers UCLA can actually get you. Narrow by year, kind, language, peer review, whether it is free to read, or to one journal picked off a result.

Every row answers one question: whether you can read the thing.

The button takes the best available route in and names which door it is using. Everything else the library offers for that paper sits behind “other ways in”. Someone who wants to read one paper should not have to choose between eight buttons to do it.

Paste an identifier instead of a title and the link resolver will answer who carries the full text, and for which years. Check those years before you walk someone over to a shelf. A provider can carry a journal without carrying the issue you need.

Databases A–Z

The databases page carries UCLA’s whole list, all 1,360 of them. It comes from the same LibGuides feed the library’s own A to Z page is built from and is refreshed once a day. The page exists for the searches this site cannot do itself: PsycINFO, Web of Science, CINAHL, a newspaper archive, a dataset. Shelfmark searches UCLA’s catalog and its article index, and everything beyond those two lives behind one of these entries.

Type in the filter box to narrow the list by name. It matches anywhere in the name, so psyc finds PsycINFO and PsycARTICLES, and pub finds PubMed. Filtering happens in your browser against the list already on screen, so it answers as you type and costs nothing.

Two checkboxes narrow it further:

Every entry opens in a new tab, at the same address the library’s own page would send you to. A database marked no login should open straight away. Everything else will ask you to sign in, whether you are on campus or off.

The feed carries a name, a link, and those two facts. It carries no descriptions. The library publishes them on its own page but does not put them through this feed, so this list leaves out a description column that would have been empty on all 1,360 rows. When you need to know what a database covers before opening it, the library’s A to Z page has the blurbs.

Hours, and where the buildings are

Opening hours come from LibCal and are read live, so they are as current as the library’s own page. Every UCLA location is listed, not just Biomed. Pick any date within a year back or two years forward; the day strip moves a week at a time.

That page answers when. Where is a different question, and it is usually the one asked at a desk with somebody standing in front of you. So the page also carries UCLA’s campus map of the nine library locations, which is the thing to hold up and point at, and a link to the library’s own Locations & Hours page for the same information written out. The map belongs to UCLA and is reproduced here so that it opens in one tap instead of three.

Pickup walks

The pickup walk sits at the bottom of the map page. Give it a list of call numbers, or photographs of them, and it returns them in the order you would actually walk the stacks rather than the order you happened to type. Photographs are read in your own browser using optical character recognition, and the images are never uploaded anywhere.