Saricks and Brown make a strong argument that those who do RA work should develop an individual reading plan and create annotations for each work read. They require at least two written annotations per month from their staff. Annotation writing is seen as essential in developing and maintaining RA awareness and effectiveness, particularly in looking at books through the eyes of other readers. The immediate benefit is for the staff member and then for the public served.
These annotations could then be added to an institutional annotation file that could be made available to users in a variety of ways. If annotations are limited to the local collection, readers will not discover works in distant collections that may be borrowed or even purchased in a book store. However, it may be frustrating for library users to discover the annotations for several appealing books that are not locally available and then leave the library empty-handed.
The issue of quality--who does them well?--is also important. I am unaware of evidence that compares the quality/utility of library created annotations to those found in What do I Read Next? for example. One could make an argument that those who are full-time annotators might create a better product than those who do it now and then. An opposite point would be that those who work with readers on a daily basis are more likely to link books to readers and know what readers are looking for. Related to the quality issue is one of duplication of effort. Why should we annotate books when adequate annotations are available elsewhere?
Years ago, librarians decided that it was wasteful to duplicate original cataloging. Instead, cataloging records are shared and then may be adapted (not often in practice) to the local situation. Why then should hundreds of library staff members read Fortress of Owls and each independently prepare an annotation? Some sort of cooperative or centralized effort would be more efficient. In fact, NoveList represents such an effort. However, if annotating is an essential part of preparation for reader guidance, it would be valuable regardless of duplication.
Self-service RA work would use annotations prepared by others if copyright issues could be resolved and permissions obtained. A library or librarians with less interest in RA work would be likely to rely on annotations prepared by others. If the primary thrust of RA work is the use of externally prepared reading lists, then it seems reasonable to select the annotation format that users would most likely encounter in the library. For example, it may be helpful to use the format from What Do I Read Next? if that title is the major RA resource used locally.
An important issue here is ownership of intellectual property. Are annotations prepared by librarians on library time? If so, the annotations belong to the library and it is reasonable to require a library-wide standard. However, if annotations are prepared by librarians on their own time, and perhaps as part of personal professional growth, then the librarian should feel comfortable in annotating in whatever manner seems most appropriate. Further, the annotations would belong to the librarian and not the library.
If library staff are going to annotate, it may be helpful to create a selection policy. Will annotators select the books or will books be assigned? Will annotators volunteer for a genre or a subgenre and then select the books they wish to read? Would it be better to have those involved with guidance read in a variety of genres? Will annotators be given a genre or a subgenre and then a list of authors or publishers to read? There are at least two issues here: (1) should each reader's advisor be a subject specialist (like research library reference librarians)? If so, how are the subjects--genres and subgenres--allocated? (2) Is it likely that subject specialists might lose touch with ordinary readers or develop likes and dislikes that may inhibit appropriate guidance? Once annotators have a genre or a subgenre, then the question of which books to read becomes an issue. Should they attempt to read the genre best sellers (say from the Locus best seller list for SF and fantasy, works that have received critical attention, or simply what they wish to read, perhaps their favorite authors and publishers? This can be a critical problem in a smaller library with a limited staff. Only a few books can be read and the RA data base will be limited to those items. GIGO [garbage in, garbage out] could be a concern.
There is a tension between limiting annotations to a few attributes which makes annotating easier while creating relatively small files and using more attributes which makes annotating more difficult, creates larger files, but allows users more access points. Ideally, the reader's advisor will know which access points are most important to the local audience. If in doubt, use Saricks and Brown or What Do I Read Next? as a standard.
The bibliographic citation would normally include: author, title, publication date, and number of pages. Special features such as llustrations, including maps, would also be identified. Series identification is also appropriate.
Book attributes would normally include:
Are there other attributes that you would add to this list?
An annotation is NOT a book review. Rather, it is a limited list of attributes used to link books to readers. The plot summary is normally written in a standard English sentence, but other variable responses are usually limited to words and phrases. Strong verbs and qualifiers must be used. As much as possible, the annotation should be objective but also persuasive. Obviously, appeal element characterization must be somewhat subjective and users may interpret "densely written" or "elegant" as negative judgements. Still, we are attempting to list persuasive attributes and NOT comment on the goodness of the work.
Dorothy Broderick says that good annotations should not contain more than one long sentence. I am not certain that is possible if the annotation is to be adequately informative. Still, the annotation needs to be ruthlessly edited until each word "must" be there.
Sharon Baker says that the annotation should tantalize prospective readers, entice them, and provide enough information so that they can quickly determine if this book is "right" for them. The annotation must "save the time of the reader."
In an institutional setting, it is important to agree not just on list of attributes, but also on appropriate values. For example, a controlled vocabulary is needed for geographical and time period values, subject headings, and appeal. Without being too restrictive, there should be a reasonable length limit to the plot summary.
Perhaps the easiest attribute of all. Complications occur when the author writes under pen names and there is a need to link them here. I will assume that the annotator will use the name on the title page rather than the "real name" although the real name should be noted if similar work is written under that name.
Another easy one. However, there may be confusion over the title when the work title is part of a series and the series title appears to be the main title and the work title appears to be a subtitle. Since readers often confuse series title--Lord of the Rings and work title, some care may be needed. Would it be appropriate to mention here if alternative versions of the title are held in the library, i.e. audio or video versions? Do note that titles cannot be copyrighted and are thus duplicated from time to time. You may have two quite different works by different authors but with the same title.
I wonder if copyright date might be better since publication date may vary with edition--cloth, trade paper, mass market paper. Too, many popular genre fiction titles are reprinted and the publication date might be decades after the original date of publication cation. Besides helping those who wish to read newer works, date might also help in resolving similar or exact titles.
I will assume that we would total the number of front and back matter pages as well as the pages in the main body. I wonder if this attribute might easily be expanded to include special feature mention. For example, maps, chronologies, illustrations, list of characters, and the like. Some readers might appreciate being able to identify works with special features.
We will assume an adopted list of geographical descriptors. There may be some subjectivity in selecting a primary place name if the work takes place in several places or several place names may be allowed. Only place names with a substantial presence would be mentioned. A decision would need to be made about fictional places, e.g. Narnia (fictional place), or places that are somewhat like a real place.
We will assume an adopted list of chronological descriptors and the decisions will be similar to those mentioned for place. Saricks and Brown suggest using the descriptors from the Fiction Catalog. A few descriptors may need to be added to accommodate fantasy and SF which have fictional time periods, e.g. Third Age (fictional time).
Emphasize author's series, including prequels and sequels, and some publisher's series. Not all authors set out to write a six volume series and tell you about that in the beginning. Works may be written as stand-alones, but feature the same character or universe. It may require some effort to link works that function as a de factoseries, but do not share a series title. Too, some series are named by reviewers or enthusiasts rather than the author or publisher. Saricks and Brown emphasize the importance of recording main character names here since readers often ask for more works involving Honor Harrington, for example.
Mary K. Chelton says that the annotation is based on six elements:
The summary needs to be long enough to indicate the nature of the story, identify main characters, and briefly characterize the plot. The central character of the work and her attributes receives special attention. As Broderick says, the central character's challenge must be identified, e.g. "to find the lost sword" since that drives the plot. Notable secondary characters who interact with the main character are mentioned briefly. Saricks and Brown say that the summary should entice the reader without giving away who did it or some such. An enticing summary would seem to need a hook, and would be different than a summary written for other library staff. A good place to begin the enticing summary is the plot summary on the front flap of the book (if there is one). Flap copy varies in descriptive coverage, but it certainly attempts to entice. It also identifies some likely appeal elements. This would be unethical in a book reviewwhere content is to be original, but seems reasonable here either as a starting point or as a check to see if you might have missed something important. I would try for less that 125 words.
Saricks and Brown suggest that this is a good place to add style and plot elements as well as special features that might not otherwise appear in a summary, i.e. "wonderful introduction to 12th Century monastic life" or "appealing color illustrations."
Librarians may differ on the need to identify or characterize the degree and nature of language or sexual and violent activity. Since the plot summary is intended to help readers to select the "right book," it seems reasonable to tell them when these elements are substantial and are likely to limit a work's appeal. The problem is "how much is too much" or "how much is enough" to warrant mention or characterization? It is also important to discover the right word or phrase to use. It might be helpful for reader's advisors to decide upon some appropriate labels to insure more reliable treatment.
Years ago, I read that LC normally gave less than two subject headings for most nonfiction while indexing and abstracting services often gave three or four times that many. While LC has begun to add more subject added entries for fiction, there is still a gap. Using Fiction Catalog or NoveList subject headings can simplify the subject heading step. You may also wish to use genre and subgenre tags as subject headings, although these might be unwieldily in a larger collection. Identifying the primary subject can be frustrating and the more subject headings the more likely that a user with a particular subject interest will find the work. I will assume an institutional agreement on the number of subject headings a typical work might receive and a standard list of descriptors. In Saricks and Brown, I wonder about the utility of a MURDER subject heading since it seems broad and vague. It is clear that there could be considerable variability here. Some traditional lists of subject headings do not adequately represent contemporary life and culture. Using the same subject headings as NoveList, for example, is a good idea if you users will also be using it. Variant headings for the same subject will frustrate users.
Appeal elements have received attention elsewhere. Too, most of the comments made about subject headings would seem to apply here. If appeal elements are to be search terms, controlled vocabulary would be most helpful. If appeal elements are to be used by end users, I wonder how clear and helpful they must be. Do we need to provide a supplementary work that briefly defines such words as "elegant"?
This attribute, and the series information for reading more by the same author, is crucially important since "read alikes" are the heart of RA work. This is also the most difficult to respond to. It is not only that "similar" is inevitably subjective, but that we need to deal with similar in what regard. Is the plot similar? The plucky female heroine similar? Is the style and manner of writing similar? Considerable personal reading experience in a genre as well as much user feedback would be necessary to feel comfortable about this one. Thus, this may be where librarians would rely more on prepared lists such as NoveList and What Do I read next?
In the digital age, it seems best to prepare annotations as digital files that can easily be added to databases, displayed on web sites, and printed out on demand. Obviously, selecting the appropriate software is an important decision. Today, we should assume that the database will be used by end users and likely accessed via a web site as well from library work stations.
There are at least three approaches here. One would be to pull annotations from the database to create an annotated list of books on a particular topic or theme. Since intellectual content is recycled, time and effort (cost) would be minimized. A second approach would be to the same thing, but edit the annotations to better meet a particular need. A third approach, which Saricks and Brown advocate, is to develop the list from scratch--selecting books, writing annotations, and creating an exhibit to accompany the list. They cover this process well in our text, but it seems unnecessary to me at a time when readers could create their own annotated bibliography by querying the existing annotation database (local or distant) and then printing out an annotated list of works that met specific criteria.
The rationale relating demand to repeated questions for help with a certain type of reading is similar to what reference librarians do. When a continuing number of requests for a certain type of information is received, many reference departments will create a "pathfinder" or "library guide" that allows the user to meet her information needs by following the logical steps in the guide and allowing reference staff to focus on less common questions. Here, rather than suggesting similar works to several different users, annotated reading lists could substitute for the repeated query. These could be printed and distributed in the library or mounted on a library web page (likely more cost effective).
Saricks and Brown emphasize that the card or database annotation differs from the annotation in the annotated book list. The latter, perhaps because it is printed, is much shorter and includes more "hook," a limited plot summary, and few--if any--of the other variables discussed above. I wonder if the difference in annotation format results from the different purpose--18 to 24 annotations to catch the reader's eye need to be short and zippy--or is imposed by the cost of printed handouts. It still seems inappropriate to develop these annotated lists from scratch when a database of fuller annotations may be readily available. If the problem is a lack of annotations related to a new subgenre or theme, it might be better to create the fuller annotations which can be reused than creating a single purpose and much more limited annotation. If the original annotation database is properly constructed, it should be possible to create annotated book lists from it. Since Saricks and Brown are experts, I may be wrong here.
Return to 590 Page