Bibliographic Software Adding New Features, Becoming Web Savvy

Web Savvy Hard times for science are turning out to be good times for publishers of personal bibliographic software. As scientists feel increasing pressure to apply for grants from several agencies and submit articles to multiple journals, the value of bibliographic software rises. These programs, which store detailed reference information and export them in a wide variety of formats, are ubiquitous in scientific offices and laboratories (F. Hoke, The Scientist, Jan. 11, 1993, page 18; June 27, 1994, page 18). According to the people who write and market these software packages and the scientists who use them, the programs have been undergoing a kind of convergent evolution in recent years. Virtually all offer important central features needed by every scientist. For example, the ability to import data from a variety of sources and to export references in the formats required by dozens of journals are features in most programs. These bibliographic applications also are being adapted to different operating systems. Software that once was available only on the Macintosh can now be run on IBM-compatible machines, and companies that sold only DOS-based programs are creating versions for Windows and the Macintosh. While the major features of the programs are converging, they differ greatly in their details and styles of operation. Some programs are beginning to exploit at least one new niche-the World Wide Web. But although there are some differences in detail, all bibliographic programs share one feature: They are major time-savers. No longer are scientists responsible for the tedious job of checking every reference to see that the periods, commas, and semicolons are in the right place, and that the name of each journal is abbreviated correctly. William R. Woodward, a professor of neurology at Oregon Health Sciences University in Portland, remembers the days before bibliographic software. "In the mid-'80s, I put together a grant that had about 80 citations," he recalls. "It took me two days to put the bibliography together and get it all squared away. The following year I put together another grant with about the same number of citations, but this time I had them in the Reference Manager software [a product from Research Information Systems Inc. of Carlsbad, Calif., a division of the Philadelphia-based Institute for Scientific Information]." With the help of the software, Woodward enthuses: "It took me all of two or three minutes to put together the bibliography." Bibliographic software is particularly useful when a journal's format calls for sequential numbering of in-text citations. "I write papers which go through a referee process," states Stephen Pickup, an assistant professor of radiology at the University of Missouri, Columbia. "Upon review they typically will ask you to eliminate or add a few paragraphs, which generally include a few references. Now, if I have three or four citations in every paragraph, just renumbering things could be an all-day affair. It becomes a nightmare. A real strength in bibliographic software in general is that it will renumber everything." Another reason to use bibliographic software is that it can improve the overall reliability of scientific literature, making it less likely that erroneous citations will see print. "This has always been a big beef of mine," explains Woodward. "You read a paper, and you decide to look up one of the citations in it. But you go to the library, and you can't find it. There's an error in the citation. Either the [name of the] journal is wrong, the pages are wrong, or it's the year or the volume or something. Once the references are in a database and are correct, they're going to be correct every time you use them." MASTERING CITATIONS: The Library Master program from Balboa Software of Willowdale, Ontario, Canada. Among the leading programs in the field are: Bookends Pro ($129) and Bookends Web ($179) from Westing Software Inc. in Corte Madera, Calif.; EndNote Plus ($299) from Niles and Associates Inc. in Berkeley, Calif.; Library Master ($250) from Balboa Software in Willowdale, Ontario, Canada; Papyrus ($99) from Research Software Design in Portland, Ore.; ProCite ($345) from Personal Bibliographic Software Inc. in Ann Arbor, Mich., which was recently acquired by the Institute for Scientific Information; RefBase ($195) from DataChip Corp. in Omaha, Neb.; and Reference Manager (Windows and DOS $349, Macintosh $249) from Research Information Systems Inc. (All prices mentioned in this article are single-user list prices.) VENERABLE PRODUCT: Bookends, first written for the Apple II, has been around since 1983. Many companies provide educational, promotional, or volume discounts as well as licenses for multiuser networks. Price comparisons are further complicated by the fact that some programs come with an "import module" for transferring references obtained from disparate sources into the database, whereas users of other programs must purchase import modules separately. "There's a lot of good software out there," maintains Jonathan Ashwell, the author of Bookends and a pioneer in the field. Ashwell wrote the first version of Bookends in 1983 for the venerable Apple II, and has been modifying it ever since during time off from his day job at the National Cancer Institute, where he is chief of the Laboratory of Immune Cell Biology. "All the programs pretty much do the same things now, and what differentiates them is the features around the edges that meet particular needs," he says. The first step in using a bibliographic program is to input the references into the database. The simplest, but most time-consuming, way to do this is to type them in manually. Most of the top programs provide shortcuts that make this tedious job go somewhat faster. For example, if the author's name has been typed in once, most programs will let the user select it from a list of all authors previously entered in the database thereafter. Pointing at a name and clicking on it with a mouse can be easier-and more accurate-than typing the name anew each time. "Bookends maintains lists in four different fields for you: author, editor, keywords, and journal," explains Ashwell. "And if you want, you can turn on 'data validation.' As you type in references, Bookends checks [these fields], and if it's an item it hasn't seen before it'll flag it, bring up the existing list, and find the closest approximation to what it thinks you typed in. It's useful for consistency, especially if it's not you entering the references-if you have a secretary or a summer student doing the data entry. It's a validation check on the data." FEATURES: Bookends Pro offers "data validation" and "cite by content" capabilities A much more efficient way to get references into the database is to import the results from a bibliographic CD-ROM or an online service such as MEDLINE, from the National Library of Medicine in Bethesda, Md. This sounds simple, but the difficulty comes because there are dozens or hundreds of different sources of bibliographic data, and each one presents that information in a different format. For personal bibliographic software to make sense of these disparate data, it must have an import module. Some programs, such as Papyrus and Bookends, include this module as part of the basic program. Reasoning that not every user needs this feature, other companies sell their import modules as optional add-ons. For example, EndLink ($99) is EndNote's import module, and Reference Manager uses Capture (Windows and DOS $149, Macintosh $99). Whether citations are typed manually or imported wholesale, most programs assign either a sequential or a random "reference number" to each citation. RefBase, in contrast, allows the researcher to use his or her own scheme for identifying the citations. Charles (Chip) Mackenzie III, president of DataChip, which makes RefBase, explains that "the disadvantage [of preassigned reference numbers] is: How do you correlate what's in your bibliographic database with what's in your [office] file cabinet?" For example, RefBase's flexibility allows a pesticide scientist of Mackenzie's acquaintance to assign his liver-function articles reference numbers in the form hepata.1, hepata.2, and so on; he calls articles on toxic chemicals toxin.1, toxin.2, and so on. Since this is the way they're stored in his file cabinet, RefBase keeps them easy to find, and it also provides somewhat of a mnemonic device when he composes a paper. Most programs require users to insert reference numbers in braces (curly brackets) at the point of reference as they write a paper using their word-processing application. Later the program will scan the document, replace the reference numbers with citations formatted to the specifications of an individual journal, and add a full bibliography at the end. COMPATIBLE: Niles and Associates' EndNote program integrates well with Microsoft Word. The most typical way of inserting citations requires the bibliographic program and the word-processing application to be active concurrently. The writer copies the reference from the bibliographic program to the operating system's clipboard, switches to the word-processing program, and pastes the citation into the document. Bookends Pro gives users another option, called "cite by content." If one remembers only that "Smith" was one of the authors of a cited reference and the word "calcium" was in the title, enclosing those two items in braces may be all that's necessary. Later, when Bookends scans the document, it simply inserts the citation if those two terms are sufficient to identify one reference uniquely. If more than one reference contains "Smith" and "calcium," the program presents a menu of choices. And if no references contain the two items, it displays an error message. A recent trend in bibliographic software is ever closer integration with specific word-processing programs. The latest version of EndNote, for example, works especially well with Microsoft Word. It adds several items to Word's "Tools" menu that allow users to access their bibliographic database from within the word-processing program. Some programs let the user attach a great deal of information to a bibliographic citation. Papyrus, for example, has a feature called "Notecard" that may be especially useful to a graduate student writing a thesis or to a researcher working on a review article. Dave Goldman, president of Research Software Design, explains that the Notecard feature is "the equivalent of having a stack of 3 x 5 index cards attached to any reference in the database. You can take extensive notes or include quotations from the work. For example, if you were summarizing a paper, you could have one notecard on methods, one notecard on results, and one on the discussion. If you were summarizing a book, you could have one notecard on each chapter. Each notecard has its own comment field, plus its own title and its own set of keywords. There's no limit to the number of notecards you can attach to any given reference." Bookends goes even further. Ashwell notes that the program "lets you attach any kind of [computer file] to a reference. If the file is a picture or a movie or a sound, for example, Bookends can open it for display. You can use it for chemical structures. You can scan in the abstract to a paper. This isn't directly related to reference management in the narrow sense, but it makes Bookends a kind of central information-management environment." Of course one of the most critical components of any bibliographic management program is its ability to export data into a wide variety of journal formats. All the top programs have dozens of export formats, and their creators add more and update the old ones continually. Selected Suppliers of Bibliographic Software Balboa Software (Library Master) DataChip Corp. (RefBase) Niles and Associates Inc. (EndNote Plus) Personal Bibliographic Software Inc. (ProCite) Research Information Systems Inc. (Reference Manager) Research Software Design (Papyrus) SeaChange Corp. (BookWhere Pro) Westing Software Inc. (Bookends Pro) While this is one of the most straightforward aspects of any bibliographic software, Niles and Associates president Richard Niles cautions that some programs may not export names properly. "Some software puts out the author pretty much the way you type it into your database, and if you need the author name to be reversed or the initials to appear without periods, they don't always do that," he explains. In other words, if one types in the author's name as John W. Smith, will the program be able to export it in all its possible variations-such as Smith, J. W.; J.W. Smith; and Smith JW-depending on a journal's format? "These are the details, and they make or break a program for a lot of users," warns Niles. With the explosion of the World Wide Web, many publishers of bibliographic software are adding the ability to export citations in the Web's formatting codes, hypertext markup language (HTML). At the simplest level, programs such as EndNote just add the HTML codes for italics and the like to the bibliographic citations, which then can be incorporated easily into a Web page. ProCite is somewhat more ambitious. It will add HTML codes to the top and bottom of the bibliography, turning it into its own fully formatted Web page. Users of ProCite can purchase an optional product called Internet Enabler ($99) that helps on the import side. Internet Enabler has two components: NetCite, a utility that works within ProCite, and BookWhere Pro, a separate program from SeaChange Corp. of Mississauga, Ontario, Canada. NetCite allows a researcher to create a citation for a Web page by grabbing its address and title from within a concurrently active Web browser. And BookWhere Pro works on the Internet, searching any of hundreds of library catalogs and information providers at the user's command, automatically importing the results of its searches to ProCite. But for the moment, Westing Software is making the most ambitious use of the Web. Its newest product, Bookends Web, includes all the functions of Bookends Pro. Moreover, it allows users to publish their entire databases on the Web for others to search and format. "It's a way of sharing your data with anyone on any platform," explains Ashwell. Bookends Web and its database must reside on a Macintosh, and that Macintosh must be connected directly to the Web. But anyone using any computer anywhere in the world can log on through the Web and search the bibliographic information provided. The resulting bibliography is exported as a Web page and displayed in the user's browser. Citations can even include hypertext links to article abstracts. "Cross-platform compatibility," which Bookends Web provides, is becoming one of the most critical new buzzwords in the field. More and more software publishers are realizing that they must create versions for all the various incarnations of Windows, in addition to their DOS and Macintosh versions. NOTEWORTHY: Papyrus' "Notecard" function enables users to take notes on a cited article or include quotations from the cited text. This is exemplified by Papyrus. With its $99 price tag, it is clearly the most economical of the top bibliographic programs, particularly since its import module is included. But it has one clear drawback: It is available only for DOS. Although Papyrus is not a Windows program, "it works very well under Windows 3.x [Version 3.0 or later], Windows NT, and Windows 95," maintains Goldman. "It knows how to use the Windows clipboard to cut and paste. It's compatible with all the Windows word processors. Obviously, there are people these days who prefer a full Windows program because Windows programs have more consistent interfaces and you can have several windows open at once. We're working on a full Windows version of Papyrus, but we don't really know when it will be done." Mackenzie's company sells RefBase, which is available in both DOS and Macintosh versions, with a Windows version due later this year. "It's obviously not a practical product unless I migrate it into Windows," he acknowledges. "I can't keep up with the word processors as a DOS product. The Windows word processors are changing their formats, so they have to be manipulated within Windows. You can't do it very readily out of DOS." Many life scientists use Apple Macintosh computers; software publisher Niles estimates that the portion may be as high as one-third. Bookends is exclusively a Macintosh product (although its Web version does provide a means for cross-platform sharing of data). RefBase, Reference Manager, EndNote, and ProCite all have Macintosh versions available. These will soon be joined by Papyrus. "Macintoshes are alive and well, especially in universities," notes Goldman. "Nineteen ninety-six is the year in which Papyrus for Macintosh is finally going to come out." This summer the company plans to begin free distribution of preliminary test versions on its Web site (http://www.rsd.com/~rsd/). "In the past, there used to be more difference between the products," notes Woodward. "Some of them wouldn't do certain [important] things. My understanding now is [that] the ones that remain are pretty good. I tend to go on the recommendation of colleagues who have products that I can go take a look at." Sean S. Duffey, a professor of entomology at the University of California, Davis, who has about 10,000 references in his EndNote database, advises scientists to pay attention to the "aesthetic aspects" of a program. "It's important whether it looks good to you and whether it's intuitive to your eye," he explains. "Where people run into glitches is in the high detail of each program. How easy is it to go into the citation styles and alter them or format your own?" Moreover, Duffey notes: "Printout is a big bugaboo in some of these programs. They will create references for you, but it can be very difficult to get a printout directly from your database [instead of having to use your word-processing program]. You have to know your needs." Many companies distribute free demo programs-either through their Web sites or via regular mail-that allow prospective purchasers to determine if the "features around the edges" meet their needs. It's important to actually try a program before buying it, emphasizes Niles: "If you're going to look at any of these programs, you look at how they're going to feel to you. If you just look at a list of features, the programs really seem the same." Robert Finn, a freelance science writer based in Long Beach, Calif., is online at finn@nasw.org.

Bibliographic Software Adding New Features, Becoming Web Savvy

Interested in reading more?

Become a Member of