Category: law
By - Thursday, June 3, 2010 - 6:56pm
In 2009, Wolters Kluwer (WK) launched a web-based legal research system called IntelliConnect. The site incorporated a wide array of WK's print content from publishing brands such as CCH and Aspen. Within the law librarian community, the product launch generated a lot of negative opinions. Seeking to rehabilitate the product's image, WK invited several law librarian bloggers to the company's offices in New York a few weeks ago for a day of presentations and meetings. Full FTC disclosure: I accepted the invitation, with WK footing the bill for my flights, hotel room and a festive Cinco de Mayo lunch, as well as providing a per diem to cover other meals and transportation.
My exposure to IntelliConnect since its launch has been intermittent. This parallels my experience with the company's print products. Services like CCH's reporters are primarily practice materials, and as an academic librarian who does very little research in the tax and business arenas that WK specializes in, use of these materials is hardly a daily event. Nor weekly.
It's for that reason that I think my reaction to IntelliConnect upon launch was more positive than a lot of my colleagues. WK had placed its materials online prior to IntelliConnect in a manner that essentially matched its print products, thus you needed to know what content each title contained in order to use the materials. Much like the classic Westlaw and Lexis, you needed to know where something relevant would be found before looking for it. IntelliConnect, however, incorporated federated searching with faceted results, allowing users who didn't know with any specificity where the useful content for a query was to search the entire system at once and find matching content regardless of its location or print title.
While I found this to be a vast improvement over the old system, the federated search model, and its accompanying move away from print organization, upset a lot of longtime users. And therein lies the seemingly unsolvable problem for Wolters Kluwer: How do you make all types of users happy?
Power users of IntelliConnect's CCH/Aspen/etc content know what's there, where it is, and expect the ability to go straight to it with as few clicks as possible. New and less frequent users than that group, however, might prefer a system they can log into and navigate with ease, guiding them to the material they're looking for even when they don't know where it is.
The initial launch played more to the needs of the latter group, with content organized by type (case, explanation, etc.), rather than practice area. This provided a major source of complaints from power users who wanted materials organized by publication title. Since the launch, WK responded to these complaints by reorganizing its content by practice area, a change that should make a big difference in usability to those users. Of course, for the non-power users this means the "CCH for Dummies" interface is gone. Personally, I wish the company could find a way to integrate both organizational models without one getting in the way of the other. I'm not sure how this could best be accomplished, and without a concrete suggestion to offer, I have to agree with WK that pleasing the existing user base has to be a higher priority for the company right now.
WK added a number of other features since the launch, including navigational enhancements like "next document"/"previous document" buttons, book browsing and full document path information for every piece of content in the system. All of these features are being added in response to user suggestions and complaints, while indicative of real problems in the initial product launch, demonstrates that WK is unusually responsive for a high end legal information vendor.
Beyond what's already been added, the company has several enhancements slated for release this year. In July the front page of IntelliConnect, currently a mostly blank screen void of useful content other than a search box will be replaced with a number of links to the user's "favorite" materials and a number of support documents that are currently buried within the system. More importantly, the system's "browse tree" will be visible on the front page, allowing users to immediately browse the system's contents without having to click the small "Browse" command first.
Also on the way is a "Titles A-Z" list that provides users an easy way to view and find every title included in the current subscription. Coupled with a "Title Finder" search box, this will finally allow a user to know what they've subscribed to without needing to navigate a confusing backend interface.
From a content perspective, IntelliConnect is a valuable practitioner resource, providing searchable electronic access to CCH's goldmine of looseleaf publications, as well as a number of Aspen publications. Despite the confusing branding on IntelliConnect's homepage (the URL says "CCH," the browser title bar says "IntelliConnect," and the page banner says "Wolters Kluwer), the company says IntelliConnect is intended to be the online presence for all of WK's legal information content, including (eventually) LoisLaw, the primary law database recently purchased by the company.
With all of these positives to recommend IntelliConnect, there is one aspect of the system that gives me pause: user interface. The layout of the system is something akin to a 1990's CD-ROM product running Folio Views. IntelliConnect's browse tree is constructed with threaded menus in which a user has to click on a small plus sign to expand the next level of the menu. The more levels down a user drills, the farther to the right the links are indented. And because this navigation pane is constructed with HTML panels, that means the titles become hidden behind the main content pane, requiring users to either scroll with a horizontal scroll bar or grab the panel's border and change the width of the nav pane. Is this functional? Yes. But it's not optimal nor is it in line with current web design norms.
The use of frames raises a larger issue than just the ease of navigation, which is overall browser functionality. One of the reason HTML frames fell out of favor in the last decade is that they render the URL visible in the browser's address bar unusable. That URL reflects the address of the page containing the frames, not the addresses of the frames currently loaded within that page. As a result, a user cannot copy and paste the visible URL into an email or another browser window or post a link to it on an intranet page so other users can access the precise material being viewed at any given time. Instead, if I email a colleague a link to the case I'm reading, when he or she tries to open it, it will load the IntelliConnect home page. In most situations, IntelliConnect's built-in email mechanism provides a workaround, but the functionality remains inconsistent that provided by most internet sites. The one place where there is no workaround is the browser's refresh button. If, as sometimes happens in IntelliConnect, a page doesn't load properly or freezes, users expect the refresh button to reload that page. Because frames are used, however, clicking the refresh button reloads the IntelliConnect home page, taking the user back to the beginning of the research trail. Frames can also break the browser's "Back" button (a common glitch in classic Westlaw), though this is less of a problem now than it used to be.
Another problematic issue with the interface is that it isn't compatible with all browsers. In fact, if a user loads IntelliConnect in anything other than Internet Explorer, a warning appears informing the user that he should use IE. While the system seems to work okay in Firefox despite the warning, it is effectively broken in both Safari and Chrome, with important buttons rendered unclickable or even invisible in those browsers. While many legal information providers rely on the accepted wisdom that all law firms are Windows shops that force employees to use only Internet Explorer, this ignores the realities of attorneys who prefer a different operating system or browser and ignore firm IT mandates when possible. And given what I see in the law school environment, where about 50% of students are now Mac users, as the current generation of students and young lawyers gain seniority in firms, the Windows-only mandates will evaporate. When questioned on the issue of browser compatibility, the folks at WK assured us they are aware of the problem and want IntelliConnect to be cross-browser compatible, no details as to what they're doing to fix the problem were offered nor a target date for when a fix would be in place.
Apart from technical concerns, the UI simply isn't consistent with design norms currently used across the web. This is a vague complaint to elaborate on, so I'll use an example within WK itself: AspenLaw.com. The colors, fonts, nav structure and browsing experience on this site are consistent with both contemporary concepts of what's "pretty" and what internet users have been conditioned to expect from popular sites like Facebook, Google or CNN. Upon selecting a publication series from the Student Central menu, the product listings are displayed in a two column layout, with facet navigation on the left and results on the right. Neither of these columns are frames. Each of the search facets list a few of the most popular categories by default, which tells the user what the facet title actually means, with a link to expand the list further if necessary. Selecting a specific item from the results loads that document in its own page with a permanent URL. To get back to the search results, a user need only click the browser's back button. No special training is required to navigate the site because it operates exactly the way most of popular sites do. AspenLaw.com illustrates that WK has excellent web designers (a distinct role that is separate from the web developers who actually build sites) at its disposal to design user interfaces for its products. I hope at some point the company asks them to tackle the IntelliConnect UI.
These UI complaints do nothing to detract from the high quality of IntelliConnect's content or the ways in which it organizes the content in response to user needs and suggestions. But with an updated look and feel that incorporates the ways users already navigate the web, the rich WK content provided in IntelliConnect could become not just functional but intuitive.
No comments
By - Friday, January 29, 2010 - 5:46pm
As I mentioned in my previous post, I traveled to Eagan, Minnesota (on Thomson Reuters' dime) earlier this week along with several other writers to get a sneak peek at Thomson Reuters' new legal research product, WestlawNext. I've already participated in a video roundtable discussion on the system, but I wanted to go into a little more detail in writing. Several others have provided comprehensive reviews of the product already, so I don't want to duplicate too much of their information. I do, however, want to look at things from the perspective of legal research instruction, since that's a big part of what I do in my job each fall, and because I've written in the past about Lexis and Westlaw's usability from that perspective over at VoxPopuLII.
The simplest assessment is this: Once law students gain access to WestlawNext (and "New Lexis," launching later this year), legal research instructors will have some 'splaining (or at least some adapting) to do.
Research instruction is marginalized and splintered at most law schools as it is. If law librarians -- the most qualified research instructors within any law school -- are part of a required 1L research curriculum, the length of that instruction is often brief. At Loyola Law School, for example, we have five weeks in the fall semester to cover the basics. At other schools, librarians might play no role. Instead, a legal writing program might rely on second-year students to teach the material. At the far end of the spectrum, some schools have no required research curriculum, offering only advanced research electives. Regardless of these requirements, electronic research instruction is often farmed out to vendor representatives, with Westlaw reps teaching Westlaw and Lexis reps teaching Lexis. The reliability of rep training varies, and a common complaint among librarians is that these reps emphasize full-text searching of primary law at the expense of secondary sources and other analytical materials. This leads to a somewhat disjointed program of instruction, even without the ground shifting beneath us.
When using the current versions of Westlaw and Lexis, a researcher needs to know exactly where relevant information is located before running a search. Therefore, when searching for primary law, she has to already know if she's looking for cases or statutes or regulations or some other type of document. In a familiar area of law, that's not a big hurdle, but when researching a new topic, the first task in research is often figuring out what type of law governs. Only then can the researcher move on to primary materials.
Using WestlawNext, however, a researcher no longer needs to select a source database before running a search. Instead, searches are most often limited only by jurisdiction, using a pop-up page overlay (NOT a pop-up browser window/tab) that allows precise selection of both state and federal jurisdictions. Then, upon running a search, the system provides results from all types of sources -- primary and secondary -- in that jurisdiction: cases, statutes, regulations, secondary sources, briefs, etc. The overview page shows just the first one or two results in each category, but the left sidebar of the page lists all the types of documents available with a count of the number of results in each of these facets. To see complete results for a particular type of resource, a researcher need only click the link for that category. When viewing these faceted results, more limiting options appear in the sidebar, such as jurisdiction, date, topic, and publication name depending on the type of materials being viewed.
Generally speaking, WestlawNext eliminates the need for researchers to know where to look for legal documents before running their searches. Don't know whether your clients' issue involves statutory or regulatory law? Just run the search and find out from the results.
To be fair, the current version of Westlaw already allows researchers to search multiple databases simultaneously, but the implementation is poor. Selecting the databases you want is a tedious process, the results display in one lump-sum list of results with no limiting facets, and basic functionality like tables of contents are nowhere to be found when viewing documents from your results. These problems are all dealbreakers, and they are corrected in WestlawNext.
While the relevant sources of law can be gleaned more easily from search results, it presents a challenge to legal research instructors. As legal professionals who have used the old systems (and print resources) for years, we already understand what the various sources of law are and how they work together. To even use Westlaw and Lexis, law students needed to learn this foundation. With WestlawNext (and most likely "New Lexis") providing Google-esque search with faceted results, students can and will run searches without an understanding of legal sources and yet not feel confused by the results. Well, not at first.
Given this likelihood, research instructors will have to provide a solid overview of the sources of law to their students. Hopefully, we already do this. But until now we could rely on Westlaw's database selection requirement to force students into learning at least a little bit about these sources before running a search that provided meaningful results. Not anymore. A student need not understand anything about sources in order to retrieve a wide swath of relevant material, and many will have the illusory feeling that the research process has been simplified enough to eliminate any need for foundational training. After all, if the fact pattern mentions "unlawful sexual intercourse" and "California" (yes, I used a Roman Polanski hypothetical in my class), a student can search those terms and retrieve a California case that seems to be on point and perhaps believe they've performed due diligence. Of course, the governing law might actually be statutory. Or there might be an split among appellate courts in California on the specific issue. Or a higher court may have decided an issue a little closer to the one in the fact pattern, making the case in hand irrelevant.
This isn't a criticism of WestlawNext. Assuming one understands sources of law, the search experience in WLN is more efficient and more likely to provide relevant materials in results. From a single search a researcher can pull in a wide array of materials and browse them easily using the provided facets. A lawyer unfamiliar with California real estate law need not know of the existence of "Miller & Starr California Real Estate" ahead of time in order to easily find that source's information in WestlawNext.
Foundational source issues are already covered in legal research instruction. The problem isn't that we don't teach them. The problem is that students might be less likely to listen once research begins to seem deceptively easy. This makes it all the more important for us to spend substantial time on instruction and assignments that cover the sources of law, independent of the specific research tools and mechanics. The good news is that if the WestlawNext interface is the future of legal research, we'll be able to spend less time in the classroom teaching Westlaw and Lexis navigation, providing extra time for foundational information.
Just because one can search an entire jurisdiction's worth of material doesn't mean that Thomson-Reuters Legal has eliminated the ability to browse and search more specific information. The currently buried Westlaw Directory once again moves front and center in WestlawNext (albeit in a redesigned form). Rather than running a search from the front page, a researcher can browse via simplified tabs on the front page (which thankfully contain no search boxes or checkboxes). From the "State Materials" tab, if a researcher selects California, a clean list of California materials displays, and the search box at the top of the page now limits itself to just a search of California materials. Clicking "All California Secondary Sources" displays a list of state-specific secondary sources and a search box that now searches only these resources. Selecting a specific source, say "California Jurisprudence," displays the table of contents for that source and allows searching of only that title.
It's been said elsewhere, but it bears repeating: Boolean search still functions. Given that many doubters have yet to acknowledge any of my colleagues' statements on this topic, let me repeat that: Boolean search still functions. There may still be some glitches here and there in that functionality, but Thomson-Reuters Legal has made it clear they intend for those operators to work as expected in WestlawNext.
WestlawNext is not without its problems, however, and the biggest concern I have is a big one for legal research instruction: secondary sources. No matter how many times instructors tell students to begin research with a secondary source, many students will still insist on running full-text searches of cases and statutory codes. (I was one of them.) Part of the blame lies at the feet of database reps who consistently overemphasize full-text searching of primary law in their training sessions, but anytime you ask students to find a case, the logical inclination is to, well, search cases.
I won't quibble with the order of the search facets in WestlawNext. Cases, statutes and other primary materials do belong at the top of the food chain because that's what lawyers and law students should be citing as precedent. The problem is how the results within the secondary sources facet display. In these results, very little specific information about each source displays. Most of the time what a researcher sees is the title of the specific section, the title of the source, and a couple of text snippets in which the search terms are highlighted. Little to no information about where within that source the section appears shows in these results. For example, when a section of Witkin and Epstein's California Criminal Law title simply "Generally" appears, that's the extent of the citation information displayed. The precise article within Witkin & Epstein in which this section appears is nowhere to be found, so researchers won't know until viewing the document itself whether it's from an article about sex crimes, robbery or murder. The threaded information available in search results for statutory sections (Title, Chapter, Subpart, etc.) needs to be included in secondary sources as well, or else the results are confusing at best and unusable at worst.
The search algorithms for secondary sources are also problematic. When I select California as my jurisdiction, if there are California-specific resources available on my topic, those need to show up first in my results. Period. When I ran a search for the term "rape," the first 12 results were from "National" sources such as Am.Jur. Trials, Am.Jur. Proof of Facts, and several law reviews. More alarming, none of these were general overviews of rape law, but instead provided information on specific aspects of rape law in various jurisdictions. I discussed this problem with Mike Dahn, vice president of new product development at Thomson Reuters Legal, he assured me that this was not by design and that they were working to make sure jurisdiction-specific resources for a user's selected jurisdiction float to the top of search results, but as of now this isn't working correctly.
Apart from these concerns, I'm still excited by the overall direction of WestlawNext. This really is a significant step forward in electronic research. I hesitate to call WestlawNext a "game-changer" because I think the game changed for all online search providers awhile back. Westlaw and Lexis are simply catching up, finally pushing legal search into the 21st century.
One final thing to note about the progress made here by Thomson Reuters (and presumably by LexisNexis later this year) is that it furthers the gap between the haves and have-nots in legal information. Once these new products are pushed out to law students, the more comfortable they become searching jurisdictions instead of sources, the harder it will be for them to use source-oriented tools (especially print materials). Again, this presents a challenge for those of us who teach legal research to make sure we provide them the foundation necessary to perform research on any platform.
I'm hardly the only person writing about WestlawNext this week. Here's a list of the reviews already in circulation:
Robert Ambrogi, A First Look at WestlawNext
David Bilinski, Dave’s Top 10 List about WestlawNext
Laura Bergus, WestlawNext: It’s About Time
Simon Chester, The Future of WestLaw – A First Glimpse
Jason Eiseman, 5 random thoughts about WestlawNext
Carolyn Elefant, My Trip Out [to] West: A Preview of WestlawNext
Greg Lambert, WestlawNext - A Study in Applying Knowledge Management & Crowdsourcing
Betsy McKenzie, Westlaw Next
Lisa Solomon, WestlawNext Preview: Product and Pricing
Jason Wilson, WestlawNext Review: Ending the tyranny of the keyword?
And of course, be sure to check out Jason Eiseman's video roundtable with Greg Lambert, Jason Wilson and me:
Video: discussion of WestlawNext
By - Thursday, September 3, 2009 - 3:57pm
Being on the west coast, and a late riser to boot, I often wake to discover interesting debates about library issues well underway — or even winding down — among my law librarian colleagues on Twitter. Today's topic (initiated by @montserratlj) was a common source of discussion: collecting and maintaining print journals that are available electronically. This is a sticky issue for a number of reasons.
First, it's expensive to maintain journal titles in both print and electronic formats. As time passes, more students simply expect articles to be available online. This preference for electronic materials is often so strong that they give up a line of research when informed that it's only available in print. Conversely, many faculty still strongly prefer print materials, whether due to personal preference or because a title's electronic version sometimes doesn't include everything found in the print copy (e.g., photos, graphs, etc.). Making both groups happy means devoting institutional resources to redundant collections, which sacrifices a library's ability to collect a wider scope of materials — in any format.
Second, given that electronic access to law journals, even those edited by a law school's own students, is usually provided only by commercial vendors on a subscription basis, there is little guarantee of permanent electronic access to a title. With print, however, so long as the library takes care of its periodical collection, a title will be available in perpetuity, regardless of whether the publisher (or even the publishing law school) folds or the subscription is cancelled.
Library directors at 12 leading law schools raised these concerns when drafting the Durham Statement on Open Access to Legal Scholarship, "which calls for all law schools to stop publishing their journals in print format and to rely instead on electronic publication coupled with a commitment to keep the electronic versions available in stable, open, digital formats." As should probably be the case in such a call to action, the Durham Statement leaves open the question of what format is best suited for creating "stable" and "open" repositories, though given economic pressures and digital solutions that vary from school to school depending upon the preferences and expertise of an institution's IT staff, a lack of guidance on this issue could lead to delayed progress or incompatible systems.
In recent years, many law reviews have already begun posting PDFs of their articles online for free. This is hardly a universal movement, and such open availability can vary wildly even among publications produced at the same school. Access is often limited to browsing tables of contents, with no search functionality to be found. In most cases, these efforts are taken on by the journals themselves. While the initiative of such student staffers deserves our praise, there are certainly limits to what they can realistically accomplish. For example, given the transitory nature of law review staffs, there is little incentive to look beyond the digitization of the current volume, let alone establish a consistent system for subsequent years or plan a long term effort to digitize previous volumes. In addition, the easy solution is PDF-only, which is hardly a well-structured data format for any purpose other than sending a document to the nearest printer. Even when the time is taken to OCR these documents, or when the PDF itself is generated through a word processor (eliminating the need for OCR), one's ability to find anything is still nearly impossible without robust metadata. One needs to look no further than the latest Google Books kerfuffle to understand the importance of metadata for online research.
If metadata, structure, and permanence are vital to the success of the Durham Statement's desired action, librarians must do more than simply ask their institutions to create digital access systems for law reviews. What the Durham Statement asks schools to create are digital libraries. If librarians willingly cede the design, implementation and maintenance of these new libraries to IT staff and law students, it's a troubling statement about the future of our profession. Far too many commercial research vendors have bypassed librarian expertise in the creation of their online systems, and we end up with expensive databases that are difficult to use and fail to take into consideration how the content itself is utilized in research. By allowing this creative process to spread within our own walls, we are effectively putting our stamp of approval on it and relegating ourselves to increasing irrelevance in a digital world. In a world where "cutting out the middle man" is desireable enough to researchers that it becomes a marketing point, digital library design becomes a far more important aspect of our profession.
Many will be tempted to look to institutional repositories as the great hope for open electronic access to law reviews, but they're not. IRs serve a vital importance as preservers of a law school's scholarly and cultural output, and the IR efforts of institutions like Duke and Georgia should be models for every other law school in the country. IRs don't address reputational economies, nor should they. What defines an article's inclusion in an IR has nothing to do with who publishes it. It's about who wrote it. Thus, in an IR articles from Yale Law Journal should arguably have equal standing to working papers that never receive publication elsewhere.
However, in a world where electronic access to legal scholarship is limited to institutional repositories, where there is no measure for the prestige of an article, a tenure track faculty's worst nightmare is realized. A common criticism of electronic only publication is that it is inherently less prestigious than print. Non-tenured faculty are often advised to shun blogging for this very reason. But this argument only applies to electronic publications that lack a rigorous editorial process. If Harvard Law Review changes nothing other than its publication medium, keeping intact all its selection and editing processes, will the articles it publishes lose their prestige simply because they're electronic? That seems rather arbitrary.
(In a perfect world that makes me drool at the possibilities, each law review could simply be a UI layer that pulls its articles from the author's original institutional repository, but that's a tangent for another post. As is a discussion of why data standards are essential to both.)
When it comes to the Durham Statement, let's do more than sign it. Let's put our money where our mouths are. Assuming that permanent, open access to electronic law reviews is a desirable goal, librarians are the best professionals to make that happen.
By - Tuesday, August 25, 2009 - 11:17pm
Yesterday saw a lot of outrage from the law librarian community, thanks to a new (albeit short-lived) promotion from Thomson West that read in part:
Are you on a first name basis with the librarian? If so, chances are, you're spending too much time at the library.
(Click here to view the full ad)
Let me say up-front that I don't think this was a smart marketing strategy for West. Law librarians certainly play a role in throwing a sizeable chunk of change West's way, even if we aren't their biggest audience. Yet, while the ad isn't as disparaging on its face as some have complained (e.g., it doesn't really say anything negative about librarians), singling out a customer constituency in this way would be ill-advised for any company.
With that out of the way, let's look at this from a legal researcher's perspective. I tend to agree with the premise in West's ad that self-service research from one's own office is preferable to any situation that requires help from a librarian. Why do patrons talk to librarians? Because they can't find what they're looking for. As Terry Martin -- a law librarian -- said at a 1996 AALL "Town Meeting:"
Inevitably, there will be instances when people have got to ask a librarian for assistance. Now, I always think that this is a systems failure to some degree. We trained them improperly; we haven't designed a good catalog; and we haven't arranged materials well so that they can't help themselves.
Reprinted in Towards a Renaissance in Law Librarianship (Richard A. Danner, ed. 1997).
In a perfect (though certainly non-existent) world, if everyone involved in creating the system (law librarians, legal publishers, research database vendors) does his or her job right, a legal researcher would never need to ask a librarian for help.
That hardly means law librarians become irrelevant, but to remain a vital part of the system, we need to put as much effort in continuously re-designing and re-implementing that system as we do in helping lost researchers, because reducing those systems failures will save researchers time and frustration. Expressing outrage towards West for suggesting that self-service research is a good thing seems to indicate more concern on our part for the collective law librarian ego than for our patrons' needs.
You want to be angry at West for that ad? Be indignant at the suggestion that their online systems are comprehensive enough or designed intuitively enough to eliminate the need for librarian (or West's own customer support) assistance. Be outraged that they charge so much for their services that, no matter how well designed, our library budgets are the only thing that makes it possible for researchers to use them.
But never forget that if West, Lexis or some other legal database vendor ever fixed these problems, the need for our expertise, at least in its current form, would be reduced dramatically.
If West rethought its pricing structures and realized it could make a lot more money selling, to name one example, a Black's Law Dictionary iPhone app for $10 than it does for $50, a lot more patrons would buy their own research tools instead of relying on our deep pockets.
And if West put the energy and UI expertise responsible for its Black's Law Dictionary Digital software to work redesigning the Westlaw interface from the ground up (rather than simply dumping voluminous full text resources into its system with little regard for how they're used), there'd be a lot fewer questions at the reference desk about how to find something on Westlaw.
Are either of these things likely to happen anytime soon? History suggests they won't. But as I said yesterday on Twitter, I don't like relying on someone else's incompetence in order to stay relevant.
Truth is, West's questionable promotional piece advertised a world that, I believe, most researchers find appealing. Not because librarians are bad or unhelpful, but because research becomes far more efficient when someone finds what they're looking for without needing to ask for help at all. As Martin said, some system failure is inevitable. But that doesn't mean it can't be reduced. If an ad suggesting that such a reduction is positive scares librarians so much, my only question is, "Why?"
