CHAPTER 3

RETRIEVAL TOOLS

T his chapter discusses retrieval tools, which are basic building blocks in the organization of recorded
information. It addresses the following questions: Why do we need retrieval tools? What are the
basic retrieval tools? What are their formats and their functions?

THE NEED FOR RETRIEVAL TOOLS

Retrieval tools are systems created for discovering information. They are designed to help users find,
identify, select, obtain, and explore information resources of all types. They assist users by retrieving
information directly (e.g., through a web search) or through relevant documents that are part of organized
collections (e.g., a catalog search). At this writing, retrieval tools typically contain records that act as
surrogates for resources. That is, each surrogate record (also called a description or metadata) provides enough
information, such as creator, title, subject, and date, so that it can serve as a short representation for and
facilitate access to an individual resource in a collection (e.g., a DVD in a library).
Surrogate records, however, may not be the only way to retrieve information. As Semantic Web
technologies advance, the emphasis on the surrogate record as a unit of description may wane. Instead, more
focus may be placed on creating individual metadata statements that are separately linked to a resource as a
form of identification and description. At this time, however, the surrogate record remains the primary unit of
description in most information environments and, thus, is still the focus of this chapter. Surrogate records are
arranged and retrieved by access points. An access point can be a name, title, or subject term chosen by an
indexer or cataloger. In online systems an access point can be almost any word in a record if keyword
searching of every word that is not a stopword is allowed.
Retrieval tools are essential as basic building blocks for a system that will organize as much of the world’s
recorded information as possible. A dream of being able to provide access to all recorded information has
existed since 1892, when Paul Otlet and Henri La Fontaine organized a conference to create Universal
Bibliographic Control (UBC). They wanted to create a central file that would include surrogate records
particularly for scientific articles in all the scientific journals of the world. The magnitude of the undertaking
meant that new techniques different from conventional library practices had to be developed. As UBC evolved
throughout the twentieth century, it came to mean that each country of the world would be responsible for
creating descriptions for its information resources and would share those records with all other countries. The
concept was extended also to authority control of names and titles used as access points. A program of the
International Federation of Library Associations and Institutions (IFLA) combined the ideals of UBC with
the concept of making the records machine readable. The many retrieval tools that have been developed as a
result of Otlet and La Fontaine’s dream have brought us closer to UBC.

THE BASIC RETRIEVAL TOOLS, THEIR

FORMATS, AND THEIR FUNCTIONS
The basic retrieval tools discussed in this chapter are

• bibliographies

74
 • catalogs
• indexes
• finding aids
• registers
• search engines

Databases and bibliographic networks have distinct roles in housing retrieval tools. They are discussed in the
next chapter, which addresses the topics of systems and system design.

Bibliographies
Fundamentally, bibliographies are lists of resources. They are essential to scholars and to those involved
professionally with books and other sources of information (e.g., collectors, dealers, librarians), and they are
also useful resources for all serious readers. They bring together lists of sources based on subject matter,
creator, time period, and the like (see the more detailed list of these below). Some bibliographies include
annotations, that is, brief reviews indicating the subject matter and/or commenting on the usefulness of the
information resource.
Bibliographies can be attached to a scholarly work and consist of the information resources that were
consulted by the author of the work, or they can be completely separate entities—works in their own right.
Each resource represented in the list has a short description referred to as a citation (not to be confused with
an annotation). A typical citation includes creator, title, edition or version, publisher, place, and date of
publication for a book or other such whole entity. For a part of a work (such as a journal article or a poem),
one typically includes creator, title, name of the larger work, volume and issue numbers (if applicable), date,
and page numbers or other part designation. Some citations also include physical characteristics.
In a bibliography each citation usually appears in only one place, most often under the creator (first creator
if more than one) of the work. This is an example of a retrieval tool that generally provides only one access
point. In a bibliography that is arranged only by creators’ names, for example, other attributes such as titles,
second or later creators, and subjects are not access points (i.e., users who only know the title, the name of a
second or later creator, or the subject of a work will not easily be able to retrieve information about it from the
bibliography). The citations may be constructed according to various styles, one of which is chosen by the
creator of the bibliography or by a publisher (for bibliographies attached to scholarly works). Examples of style
manuals and a citation from each are provided below:

• APA (American Psychological Association)1

Mitchell, T. R., & Larson, J. R., Jr. (1987). People in organizations: An introduction to
organizational behavior (3rd ed.). New York: McGraw-Hill.
• Chicago Manual of Style2
Mitchell, Terence R., and James R. Larson, Jr. People in Organizations: An Introduction to
Organizational Behavior. 3rd ed. New York: McGraw-Hill, 1987.
• MLA (Modern Language Association)3
Mitchell, Terence R., and James R. Larson, Jr. People in Organizations: An Introduction to
Organizational Behavior. 3rd ed. New York: McGraw-Hill, 1987. Print.
• Science (Scientific Style and Format)4
Mitchell TR, Larson JR, Jr. People in organizations: an introduction to organizational behavior.
3rd ed. New York (NY): McGraw-Hill; 1987.
• Turabian5
Mitchell, Terence R., and James R. Larson, Jr. People in Organizations: An Introduction to
Organizational Behavior. 3rd ed. New York: McGraw-Hill, 1987.
• Style Manual (U.S. Government Style Manual)6
Mitchell, Terence R., and James R. Larson, Jr. People in Organizations: An Introduction to
Organizational Behavior. 3rd ed. (New York: McGraw-Hill, 1987).

The above examples are for a book, but each style manual provides guidelines for citing all types of resources,
75
The above examples are for a book, but each style manual provides guidelines for citing all types of resources,
including journal articles, theses and dissertations, and websites, to name a few.
Each bibliography has a particular focus or arrangement. The most common include the following:

Focus Purpose Example

Subject Gathers together resources about a particular subject. A Bibliography of Female Economic Thought
Up to 19407
Creator Gathers all or some works of a particular creator and sometimes A Bibliography of Jane Austen8
including sources about the creator.
Language Gathers textual resources in which the text is in one or more An Extensive Bibliography of Studies in
particular languages. English, German, and French on Turkish
Foreign Policy, 1923–19979
Time Period Gathers all resources that came to light in a particular time period. Russian and Soviet Education 1731–198910
Locale Gathers resources created in a particular location. This could be a Area Bibliography of China11
large locale such as a whole continent, or a smaller locale such as a
country, region, state, city, or community. It could also be an University of Iowa College of Law’s Faculty
institution, such as a bibliography of all the works of the faculty in Bibliography12
a particular university.
Publisher Gathers all of the products of a particular publisher The Stinehour Press: Work of the First Fifty
Years13
Form Gathers information resources that appear in a certain form, Maps and Mapping of Africa: A Resource
format, or genre (e.g., videocassettes, electronic resources, poetry, Guide14
biographies). These are virtually always combined with one of the
other foci.

Two or more of these foci are often combined in bibliographies. For example, the title above that illustrates
language is a combination of language, subject, and time period; the one illustrating time period is a
combination of subject, locale, and time period; and the one illustrating form is a combination of form and
locale.
There is a special kind of bibliography found in libraries that is truly meant to be a retrieval tool. They
were for many years called pathfinders, but now they are usually referred to as library guides, subject guides, or
research guides. These tools are subject bibliographies with a special function in the library world. While most
bibliographies do not indicate whether a location actually has a specific resource, research guides focus on the
resources in a defined subject area available in a particular setting. In addition to the list of relevant resources,
the guide may also include a list of locally accessible databases to search, specific instructions on how to search
the local catalog, and a list of specific reference sources related to the particular subject area. For example, the
guide may provide a list of relevant subject headings that may be used in the local catalog. Subject and
reference librarians have created these tools for decades. Originally pathfinders were printed documents
available only in the library, but now these guides are primarily found on a library’s website.15 In recent years,
software products have been developed for the creation and management of library research guides. These
range from proprietary software products (such as LibGuides)16 to open-source software (such as
SubjectsPlus).17 Another option is to adapt other web tools, such as wikis or blogs, to meet the purpose, as
they are somewhat inexpensive and easy to implement.

Catalogs
Catalogs provide access to individual items within collections of information resources (e.g., physical
entities such as books, DVDs, and maps in a library; artists’ works in an art museum; web pages). Each
resource is represented by a description that is somewhat longer than an entry in a bibliography. The
descriptions are assigned one or more access points. As mentioned earlier, an access point can be almost any
word in a record when keyword searching is used; however, the term access point is usually applied to a
particular name, title, or subject that is listed on the record separately from the description. An access point is
constructed in a certain order (e.g., surname followed by forename or forenames), and it is maintained under
authority control. Authority control is the process of bringing together all of the forms of name that apply to a
single entity; gathering together all the variant titles that apply to a single work; and relating all the synonyms,
related terms, broader terms, and narrower terms that apply to a particular subject heading.
The descriptions in a catalog are constructed according to a standard style selected by a particular
community (e.g., RDA: Resource Description & Access for libraries; Describing Archives: A Content Standard

76
(DACS) for archives in the United States; Cataloging Cultural Objects (CCO) for some art collections).
Several different standards for description are discussed in more detail in Chapter 7.

Purposes of Catalogs
Catalogs have traditionally served two main user groups. One group is the employees of the institution,
who need to retrieve information resources or who need to retrieve information about those resources (i.e.,
metadata) in order to work with the collection. For example, the catalog is used by collection development
librarians in their process of discovering what the library already owns or does not own before selecting new
materials; another such use is by the employee of a museum who is looking for objects to place in an exhibit.
The most commonly understood use for a catalog, though, is use by the patrons of the institution who
wish to borrow material or make use of it on the premises. If such users have a known resource in mind, they
may search for it in the catalog by creator or title (called known-item searching). Users might also try searching
by keyword if they only remember certain words of the title or some other attribute. If users know they want
resources by a particular creator, they may search under the creator’s name. If users do not know of a particular
resource but are searching for something on a particular topic, they may use a subject heading search, a subject
keyword search, or a general keyword search. In online catalogs, keyword searches are often appropriate for
helping a person find a record that looks like it might be on the user’s topic. Once a potentially useful record is
found, the user can identify in that record an authority-controlled subject heading or a classification notation
for the topic of interest. The user, then, may conduct a search for the subject heading or the classification
notation in the catalog (often by clicking on hyperlinks in the catalog). Alternatively, one may go to the
location of the classification notation in the stacks to determine if there are other pertinent works shelved with
the one that has been identified.
A number of attempts have been made through the years to identify purposes of catalogs. Charles A.
Cutter presented his “objects” (i.e., objectives) of a catalog in his Rules for a Printed Dictionary Catalog in 1876.
Cutter’s objectives are listed in Figure 3.1.18 If the language of the objectives were updated, they would now,
among other things, refer to resources, not just to books. The first object, a finding function, says that a catalog
should help a user retrieve a resource if one of its key access points is known and it is part of the collection.
The second object, a gathering function, states that a catalog should be able to display all resources in the
collection by a creator, on a specific subject, or in a particular genre or literary form. The third object, a
selecting function, is to help users to choose the resource they need based on bibliographic attributes (e.g.,
version, language, format, date) or the nature of the content (e.g., genre, subject matter). Cutter was speaking
only of library catalogs in which books were represented, but if these objectives are broadened to archives,
museums, and other collecting institutions, they still represent what catalogs and other retrieval tools are
supposed to do today.

77
Figure 3.1. Cutter’s Objects of the Catalog.

Seymour Lubetzky worked in the mid-twentieth century to simplify cataloging rules, and he posited that
the cataloging code should be reconstructed “in accordance with deliberately adopted objectives . . . and well
considered principles.”19 He then stated that:

The first objective is to enable the user of the catalog to determine readily whether or not the library
has the book he wants. . . . The second objective is to reveal to the user of the catalog, under one form
of the author’s name, what works the library has by a given author and what editions or translations of
a given work.20

Lubetzky’s work was the basis for the Paris Principles, adopted at the International Conference on
Cataloguing Principles in Paris in 1961. The second of the principles was for “Functions of the Catalogue”
and stated that the catalog should be an efficient instrument for ascertaining

2.1. whether the library contains a particular book specified by

(a) its author and title, or
(b) if the author is not named in the book, its title alone, or
(c) if author and title are inappropriate or insufficient for identification, a suitable substitute for
the title; and

2.2. (a) which works by a particular author and

(b) which editions of a particular work are in the library.21

In 1998 IFLA published Functional Requirements for Bibliographic Records (FRBR), which identified four
generic user tasks that the bibliographic record is intended to support. These four tasks have been built on the
work of Cutter, Lubetzky, and the Paris Principles, and essentially represent the four main functions of a
catalog:

• to find entities that correspond to a user’s search criteria

• to identify entities (e.g., persons, works, subjects)
• to select entities appropriate to the user’s needs
• to obtain access to the entities described22

The fourth user task in FRBR actually is an addition to the functions identified by Cutter, but it was
78
The fourth user task in FRBR actually is an addition to the functions identified by Cutter, but it was
recognized by several writers in the late twentieth century that a catalog differs from certain other retrieval
tools in that it facilitates physically locating the information resources that are represented by descriptions in
the catalog.23 In 2016 a fifth user task was proposed in the draft IFLA Library Reference Model (LRM)—a
document that attempts to harmonize FRBR and two other functional-requirements models established by
IFLA. The fifth task is to explore relationships between one resource and another.24 The LRM has not been
adopted at the time of this writing, but it is expected that this addition to the user tasks will be approved.
Another important purpose served by catalogs has traditionally been to act as an inventory of the
collection—that is, to provide a record of what is owned. Often a shelflist has been used to accomplish this
purpose. A shelflist includes one copy of every record in a catalog arranged in the order in which the
information resources, objects, and so forth, are arranged on the shelf. So, for example, one could use the
shelflist in the process known as shelf reading to determine whether resources from the collection were
misshelved. Originally, shelflists were literally in the order of items on the shelf, starting a new sequence with
each change in format, or change in collection location, or change in size. Later, shelflists often were arranged
by classification notation regardless of format or other categorization. This is the way the concept works in
online catalogs. The purpose of serving as an inventory is still there, but the mental image of a shelflist with
arrangement as it is on the shelf is lost.
A union catalog is a variation of the concept that a catalog represents just the holdings of one institution. A
union catalog represents the holdings of more than one institution or collection. For example, a union catalog
can show the resources held in a main library and also in one or more of its branch libraries. The location
information indicates where a resource is held. A union catalog of a consortium of institutions works the same
way: location information shows which resources are housed in which of the cooperating institutions. The
ultimate union catalog is the one maintained by the largest bibliographic network, OCLC (see description in
Chapter 4), where each information resource has one master record; associated with that record is a holdings
record that shows the holding symbol of each member of the network that has cataloged the resource through
OCLC or has asked that its symbol be added to the record.

Forms of Catalogs
Catalogs can appear in different forms and can have different arrangements. The following formats are
discussed:

• book catalogs
• card catalogs
• microform catalogs
• online catalogs

Book Catalogs. Book catalogs originally were just handwritten lists. After the invention of printing with
moveable type, book catalogs were printed, but not always in a discernible order. Eventually, entries were
printed in alphabetical or classified order, but book catalogs were very expensive to produce and were,
therefore, not reproduced and updated often. By the early 1900s, book catalogs were almost completely
replaced with card catalogs that could be updated as soon as the cards could be filed and were relatively
inexpensive to maintain and create. Book catalogs had a brief renaissance in the 1960s and 1970s when (1)
computers again made them easy and less expensive to produce, (2) large card catalogs became unwieldy, and
(3) rapid growth of new libraries and new branches made it desirable to have multiple copies of catalogs. But
in order to keep a book catalog up to date, supplements were usually produced, resulting in multiple look-ups
for one search. In addition, it was usually three to six months before supplements were produced, meaning
that new materials were not represented in the catalog during that time.
Online catalogs have replaced both book and card catalogs in most technologically advanced and
economically advantaged countries, but book catalogs are still used in some cases for rare materials, catalogs of
exhibits, artists’ works, and so forth. Book catalogs provide a way to make the contents of special collections
known to users in many locations. For example, book catalogs of historical societies are popular acquisitions
by collectors of genealogical materials. This use, though, may be replaced by availability of such catalogs
online. Book catalogs also still exist in some libraries and archives as the only retrieval tool available to access
older materials, because some institutions have chosen not to convert all of the catalog records for rarely used
resources into machine-readable form for the online catalog.

An advantage of book catalogs over online catalogs once was that book catalogs are compact and portable
79
 An advantage of book catalogs over online catalogs once was that book catalogs are compact and portable
and can be consulted anywhere they can be carried. However, with laptop computers, smart phones, and
wireless technology, online catalogs are now accessible from almost anywhere. Another advantage of book
catalogs is that glancing over a page of book catalog entries is relatively fast, and some people prefer this to
paging through screen after screen of online responses. An illustration of a book catalog is found in Figure
3.2.

80
Figure 3.2. A Representation of a Page from a Book Catalog.

Card Catalogs. In a card catalog, each bibliographic description is prepared on a standard 7.5 × 12.5 cm
card (roughly 3 × 5 in.), and interfiled into drawers full of cards in specially designed cabinets. The dimensions
of the cards were not always standard; in the early days of card catalogs, different sizes were in use (3 × 5 in., 1

81
1⁄2 × 5 in., 1 1⁄2 × 10 in., and 4 × 6 in.). Card catalogs were popularized in the United States by Library of
Congress (LC) cards, first made available for sale in 1901, and by H. W. Wilson cards, which began
production in 1938 in response to the needs of small libraries. (Both have now ceased card production.)
Technological advances encouraged further use of the card catalog. The descriptions on cards were at first
hand-written, using a standardized library hand type of writing developed by Melvil Dewey in the 1870s (See
Figure 3.3). Later, typewriters, one of the great technological innovations in libraries at the turn of the
twentieth century, made handwritten cards unnecessary. Offset printing was used by LC for its cards. Then
photocopying allowed the local creation of whole card sets from one master card. Finally, the advent of
computer printing made it possible to have customized cards made either locally or at a distant facility. When
created at a distance, the cards often were shipped to the receiver in boxes already alphabetized and ready to
file.
Because of the influence of LC cards, card catalogs and the order of information on cards in libraries were
standardized for many decades. Users of card catalogs could go from library to library, using catalogs with
confidence that they would be able to use distant catalogs with as much ease as their local ones, although the
filing arrangement of those catalogs often differed from library to library.

Figure 3.3. Example of an 1895 Catalog Card Written in Library Hand.

Online catalogs have replaced most card catalogs in the United States, but some libraries, archives, and
museums still have card catalogs, especially small institutions or those where there has been only minimal
conversion of data to machine-readable form. In many other countries there are more card catalogs, and in
some, card catalogs are still the predominant form. Figure 3.4 is an example of a more contemporary catalog
card.

82
Figure 3.4. A Representation of a Printed Catalog Card.

Microform Catalogs. The creation of computer output microform catalogs (or COM catalogs) became
possible in the 1960s. They are produced on either microfiche or microfilm and require a microform reader in
order to be able to use them. They are produced like book catalogs, but because they do not have to be
reproduced on paper and bound, they can be completely reproduced with new additions every three months or
so without having to go through the supplement stage. Due to unpopularity, microform catalogs were
replaced rather quickly by online catalogs. It has been found that users will use microfilm if that is the only
way to get the information they need, but most people find the readers hard to use and difficult to read.
Online Catalogs. Online catalogs, often referred to as online public access catalogs or OPACs, are the
predominant form of catalog in the United States and in a number of other countries today. In these catalogs,
records are stored on a local or remote server. The records are displayed only as needed. There is much
flexibility in the look of the displays. Figure 3.5 shows the basic search screen of the OPAC used at Simmons
College’s Beatley Library. Figure 3.6 shows the view of a single record in that library’s catalog.

83
Figure 3.5. Basic Search Screen in an OPAC. (Source: Beatley Library, Simmons College, Boston, Massachusetts)

Online catalogs have not been standardized, although in two or more institutions that have purchased the
same system, the displays may look somewhat similar. Writers in the field have long called for
standardization, so that patrons can move from catalog to catalog or search multiple catalogs from the same
location and find records displayed in the same manner, but this has yet to happen.25 In recent years,
dissatisfaction with online catalogs has been expressed by many in the library and information professions.
This frustration—that the online catalog is still difficult to use—is intensified by comparisons to Internet
search engines such as Google and Bing that are relatively easy to use. At times, this discontent has led some
to question the long-term prospects for the online catalog. If catalog vendors respond by improving
arrangements of displayed responses (e.g., by emphasizing relationships among resources), online catalogs may
again be seen as providing something not available in search engines and then may survive. Some believe that
when the Semantic Web is realized and all types of resources (bibliographic and otherwise) are connected
using linked data techniques, then perhaps general search engines will replace the OPACs that we use today
to find information in libraries. This, however, may be a long time coming.

84
Figure 3.6. A Bibliographic Record in an OPAC. (Source: Beatley Library, Simmons College, Boston, Massachusetts,
http://library.simmons.edu/record=b1055231~S0)

Arrangement and Displays within Catalogs

The records within catalogs must be arranged in some fashion or they are unusable. In card catalogs,
records are arranged by being filed in a certain order. In book and microform catalogs, records are arranged by
being printed one after another in a particular order. Records in OPACs, however, are not set in one place;
there is flexibility because the records are arranged internally within the database either in sequence of order of
entry into the system or in random order. This allows for the customization of the display of search results in
online catalogs. So the arrangement discussed here applies to the arrangement of the displayed responses to
search queries in the case of online catalogs and not to the internal order within the database. (For more
information about arrangement of metadata displays, see Appendix C.)
In general, there were two basic arrangements that made sense in printed catalogs (i.e., all the formats of
catalogs occurring before the online version). Catalog records were either in classified order or in alphabetical
order. These arrangements are possible in the online catalog, although they are no longer the only choices for
display. In today’s OPACs, many options for displaying search results are offered; users may choose the
arrangement that works best for their particular searches from a list of options. In the next sections, the two
traditional approaches to the arrangement of printed catalogs are discussed; a brief section then follows on the
choices for display in the online catalog.
Classified. Classified catalogs usually have more than one section. In what is considered to be the “main”
section of the catalog, the arrangement or display is in the order of the classification scheme used in that

85
institution. That is, this section is arranged in subject order, where the subject is represented by a classification
notation rather than by subject terminology. For example, a book of recipes for use in a slow cooker is
represented and arranged by the Dewey Decimal Classification (DDC) notation 641.5884 rather than by its
analogous subject heading, Electric cooking, Slow. There can be as many classification notations assigned to a
single record as there are subject concepts in the information resource.
Classified catalogs have the advantage that users can look at records for broader and narrower concepts at
the same place they are looking for records on a specific concept. In a way, it is similar to browsing in the
stacks with a classification notation, except that in the case of the classified catalog, each information resource
is represented by several notations signifying all of its concepts, not just one notation as is true of resources in
the stacks. An example of a page from a classified printed catalog is found in Figure 3.7; in it one can see the
entries arranged by the call number associated with each resource.
As it is nearly impossible for anyone to know all the classification notations relating to a subject, there
should be a section of the catalog that lists verbal representations of all topics and gives the notation for each
topic (e.g., Electric cooking, Slow = 641.5884). In some situations, this need could be met by placing a copy
of the classification scheme at the catalog, because most classification schemes include an index of topics in
alphabetical order. And, of course, there are users who want to search for authors and/or titles, so there are
other sections of the catalog for these. In printed catalogs, these subject-term, author, and title sections are
arranged in alphabetical order. In OPACs they are word-searchable and the order of displays varies.

86
Figure 3.7. A Representation of a Page from a Classified Book Catalog.

Classified catalogs have traditionally been used in European and other countries where several languages
are spoken and represented in the setting. The classified subject section of the catalog can include records for
every language represented, with several languages being interfiled at the same classification notation. The

87
sections of the catalog that give access to subject terms, authors, and titles can be in the language(s)
appropriate to the clientele. The United States has not traditionally felt the need for this approach. Among
the reasons the concept is still relevant today is that there is now global access to online catalogs, and a
classified catalog can hold and display records in any language with classification notations that are universal.
If indexes to a classified catalog are made in many languages, access can be gained through any one of the
many languages. In addition, it makes broadening and narrowing of searches, as well as browsing, much
easier.
Alphabetical. Early American catalogs were arranged by broad subject categories in alphabetical order.
With a collection consisting of a few books, there was little need for elaborate classification or arrangement.
As catalogs grew, the broad categories needed to be subdivided, so somewhat narrower categories were created
and placed alphabetically within each broad category. For example, if the broad category were Domestic
animals, the sub-categories under it might be Cats, Cows, Dogs, Horses, Mules, and so forth. These were
called alphabetico-classed catalogs. As subject categories multiplied, it became more difficult to predict the
subject category and where it would be found. It began to make sense to place all categories and sub-
categories in alphabetical order. Cutter was instrumental in the development of what he called the dictionary
catalog. He recommended alphabetical arrangement with authors, titles, and subjects all interfiled in the same
catalog. An example of a printed dictionary catalog page is found earlier in this chapter in Figure 3.2. The
page in that figure contains the following mix of entries:

1. a title entry for Jottings from Jail

2. an author entry for Joseph Joubert’s notebooks
3. an author entry for Daniel N. Joudrey for a book on cataloging
4. another author entry for Joudrey for an earlier edition of this book, The Organization of Information
5. a title entry for a book of French poems, Les Joues en Feu
6. a subject entry for a biographical children’s book about Jack Jouett
7. an entry for Alain Jouffroy, an editor
8. an entry for a handbook edited by Louis Joughin
9. an author entry for Joughin’s book on Sacco and Vanzetti
10. a title entry for a book about the film Le Jour se Lève
11. a personal name cross-reference (from Jouve to Ward Jouve)

The interfiling of the various types of entries is what defines this type of arrangement as a dictionary catalog.
The dictionary card catalog was the standard for the first half of the twentieth century. Later, it became
quite complicated to file new cards into large catalogs because of the complexity of filing rules that seemed to
grow as the catalogs grew. Early filing rules had reflected the influence of the classified catalog and presorted
catalog entries into categories. Cutter recommended alphabetical filing, but names and titles beginning with a
word that could also be a topical subject (e.g., Glass) were supposed to be filed in the order: personal name,
geographic entity, corporate body, subject heading, title. This is seen in the following example:

Glass, Richard [personal name]

Glass Mountains (Tex.) [geographic entity]
Glass Art Society [corporate body]
Glass [subject heading]
Glass art [subject heading]
Glass menagerie [title]

It can be seen that these are not alphabetical and it is easy to see how complicated such an arrangement was in
large catalogs.
Attempts to break up the large files resulted in divided catalogs. Divided catalogs were alphabetical catalogs
that were partitioned into two or possibly three sections. If there were two, they were often divided so that
authors and titles were in one part and subject headings were in the other. In a catalog with three sections, the

88
division was usually authors, titles, and subjects. In this tripartite arrangement, records for resources about a
person were usually filed in the subject section, while records for works created by that person were filed in the
author section. However, it was considered useful to keep records for works by and about a person together, so
sometimes all names were placed in one part of the catalog whether it was for the name as a creator or for the
name as a subject. In a divided card catalog, there might be separate cabinets (or sets of cabinets) used for the
different sections of the catalog (see Figure 3.8). The divided catalog was easier and less expensive for the
keepers of the catalog, but it was assumed that users knew the differences among author, title, and subject
entries, which was not always the case.

Figure 3.8. A Divided Alphabetical Card Catalog.

With the development of OPACs in the 1980s, the divided catalog was moved online. One usually had to
search either by author or by title or by subject (although, later, more options for searching became available).
In this case, though, it was seldom possible to retrieve works both by and about a person in the same search;
this was not possible until general keyword searching became available in the online catalog, and even then it
was not terribly precise because the name might appear in parts of the record other than a creator or subject
field. General keyword searches are similar to the dictionary catalog approach to arranging a catalog, because
one does not have to choose the type of search one wishes to conduct; search terms are retrieved no matter
where they appear in the catalog record. The classified arrangement is also available in today’s OPAC; it is
usually achieved by performing a call number search, if that type of search is available (which is not always the
case). In some ways, the sophistication that had been built into the design of card catalogs through elaborate
intellectual filing rules was abandoned in exchange for the convenience of automated searching.
Other Arrangements. Displays of search results today are not always in alphabetical or classified order.
The specific OPAC’s programming determines which types of arrangements of search results are possible and
when each arrangement is used as the default display. In some catalogs, searches for creators or titles may
bring up an alphabetical display of authority-controlled access points for creators or a list of titles in
alphabetical order, but other systems have a default setting that is used for almost every type of search:
relevancy ranking. Arrangement by relevance can be mysterious to users and information professionals alike.
The idea is that the resources that are most closely connected to the search terms are ranked higher in the
search results. Relevancy is determined algorithmically, and it is defined differently by each system. It is based
on a combination of factors (e.g., placement, field weighting, term frequency) that may be adjusted somewhat
according to an institution’s preferences.
The choice of display arrangements and the choice of searches available vary from catalog to catalog.
Some, for example, do not provide users with the ability to conduct a classification or call number search, but
they may allow title or creator search results to be sorted in call number order. Some OPACs do not allow

89
browsing of authority-controlled lists of names, instead focusing primarily on keyword searching. Most online
catalogs now allow users a choice of the order in which search results are to be displayed. Catalogs may allow
sorting by
• relevance ranking,
• creator’s name (alphabetical by last name),
• title (alphabetical by first non-article word),
• call number,
• ascending chronological order (publication date from earliest to latest),
• descending chronological order (from latest to earliest),
• most popular resources first, or
• format or material type.

The choices offered by catalogs may vary from just two or three of these sorting options to, perhaps, many or
most of the types listed. In some systems, these arrangements can also include options for a variety of sub-
arrangements under the primary sort key (e.g., sorting first by creator but then choosing sub-arrangement by
title or publication date).

Indexes
Indexes are retrieval tools that provide access to the analyzed contents of information resources (e.g.,
articles in journals, short stories in collections, papers in conference proceedings). Although back-of-the-book
indexes provide access to the analyzed contents of one work, they are not retrieval tools in the sense defined in
this chapter; they are prepared at the time of publication of the work, not at a later time in an effort to provide
bibliographic control. They do, however, aid with retrieval of the information found in the text at hand. Some
websites also have indexes that function very much like back-of-the-book indexes, but instead of page locator
references, each entry links directly to an HTML anchor within the site being indexed. Online these are
called A–Z Indexes (to distinguish them from search trees and search engines; see discussion below). A–Z
indexing employs specialized software (such as HTML Indexer and XRefHT32).26 Like back-of-the-book
indexes, an A–Z web index should contain properly identified second-level entries, variant entries, and cross-
references.
Indexes that are retrieval tools in the sense discussed in this chapter are separate from the information
resources being analyzed. Database indexes (also called journal indexes or periodical indexes) are the longest-lived
examples of indexes as retrieval tools. Such indexes often do not have authority control of names. A name
such as Lois Mai Chan can be entered into the same index at different times as Chan, L.; Chan, LM; Chan,
L. M.; Chan, Lois M; Chan, Lois Mai; and even Mai Chan, Lois. They do, however, usually maintain
authority control over subjects. Indexers tend to use hierarchical controlled vocabularies (called thesauri) for
the topical terms they wish to bring out in the index. A thesaurus contains systematized language that
subsumes narrower terms under broader terms and provides a structure of relationships between related terms.
(See Chapter 10 for more information on thesauri and other controlled vocabularies.)
Unlike catalogs, indexes are not limited to what is available in a local setting or a particular collection, and
they do not usually give location information as such. They do give, as part of the description, the larger
resource in which the smaller work can be found. Especially with print indexes, it may then be necessary for a
user of the index to search a catalog for the location of the larger work. In some online indexes, there exists
the capacity to link to online catalogs to allow users to see quickly if their library owns the larger work. If the
larger work is not found in the local catalog, the user may have to search a union catalog or make a request
through interlibrary loan (ILL). Increasingly, online indexes have links to full text versions of articles.
Indexing can be carried out by people or by machines or by a combination of both. Database indexing is
still mostly completed by specially trained indexers, although they may have some machine assistance. For
many decades, indexes were available in print form, but today they are primarily online tools. Some print
versions were arranged in alphabetical dictionary fashion, with entries for authors, titles, and subjects
interfiled. Others reflected a divided arrangement, having an author/title part separate from the subject index.
Online indexes have interfaces that dictate how they are searched and how results are displayed. As with
OPACs, there is no standardization from index to index. OPACs, at least, have the standardization that

90
comes from using RDA (or another set of descriptive rules) whereas there is no such standard commonly used
by every index. The National Information Standards Organization (NISO) tried to update Z39.4–1984,
“Basic Criteria for Indexes,”27 but committees could not come to agreement with the American Society for
Indexing, and so Z39.4 was withdrawn in 1996. A standard from the International Organization for
Standardization (ISO), which the British and others have adopted, was published in 1996.28
Another difference between catalogs and indexes is that indexes tend to be created by for-profit
organizations, such as H. W. Wilson, or professional societies, such as the American Chemical Society. Often
there is a charge for access to the online versions. Libraries pay for the right to allow their patrons to use the
indexes online without cost. Some libraries have both print and online access, although there seems to be a
trend to continue only with the online versions. In any case, the index often is cataloged so that the whole
index can be found through the catalog.
The sample index entry in Figure 3.9 illustrates the types of information included in an index for scholarly
journal articles. It begins with the title of the article and then lists various characteristics of the article,
including who wrote it, contact information, and affiliations. There is also information about the journal,
including its title, type, and an identifier, usually the International Standard Serial Number (ISSN). Other
information includes the publisher’s name, the volume and issue in which the article is found, the page
numbers, and the article’s length. There is also an indicative (descriptive) abstract and associated subjects.
Most of the indexing already discussed is performed by humans. However, as early as the 1950s people
began working on automatic indexing that would rely on the power of computers to perform repetitive tasks at
high speed. The first such method, introduced by Hans Peter Luhn in 1958, became known as KWIC (Key
Word In Context) indexing. In a KWIC index, titles are arranged so that each of the words in a title appears
once in alphabetical order in the center of a page, with all other words to the left or right of the center word
printed in the order in which they appear in the title (so that the alphabetized word is in context).

91
Figure 3.9. A Representation of an Index Entry for a Scholarly Journal Article.

KWIC
Introduction to Cataloging and Classification.
Introduction to Cataloging and Classification.
Introduction to Cataloging and Classification.

An adaptation of the KWIC method is known as KWOC (Key Word Out of Context), which prints the
significant words in the left-hand margin instead of in the middle of the page, and the title is printed to the
right or beneath the keyword.

92
 KWOC
Cataloging Introduction to Cataloging and Classification.
Classification Introduction to Cataloging and Classification.
Introduction Introduction to Cataloging and Classification.

Two additional types both identified as KWAC are used somewhat less frequently. The first, Key Word And
Context is a combination of KWIC and KWOC, in which the significant word appears in both the heading
position and in context with the rest of the title. The second, Key Word Augmented in Context, allows the
indexer to augment the title with additional words to provide more meaning and to enhance access to the
resource.

KWAC: Key Word And Context

Cataloging and Classification. Introduction to

Classification. Introduction to Cataloging and
Introduction to Cataloging and Classification.

KWAC: Key Word Augmented in Context

Cataloging and Classification [in Libraries]. Introduction to

Classification [in Libraries]. Introduction to Cataloging and
Introduction to Cataloging and Classification [in Libraries].
Libraries. Introduction to Cataloging and Classification

The current application of automatic indexing is in search engine indexing. A search engine is sometimes
referred to as an index but, more accurately, a search engine creates lists of terms that are referred to as
indexes. A search engine is a computer program that searches web documents for specified keywords via a
program called a spider. The list of documents is returned to an internal database and placed in order by the
indexing program according to the words or concepts contained in the document. With the development of
microdata and similar annotating schemes, general markup vocabularies (such as Schema.org), and the use of
other encoded metadata tags, search engine indexing could become more accurate for users. Search engines are
described more fully below.

Finding Aids
Finding aids (sometimes called inventories) are descriptions of archival collections, and they tend to be
longer than typical catalog records. Archives usually maintain control over entire collections (not individual
pieces) of archival materials from personal, corporate, or institutional sources. Thus a finding aid describes a
collection, not a single letter or record within that collection. While some finding aids are not publicly
accessible, this is becoming increasingly rare as archives add more finding aids to the web. The finding aid,
itself a retrieval tool, is often linked to a summary description of the collection in a catalog record. Summary
records provide name, title, and subject access points for archival collections in the catalog and also provide a
means to point users to more detailed finding aids. Describing Archives: A Content Standard (DACS) is used in
the United States to create both finding aids and MARC-based catalog records. An example of a brief finding
aid is presented in Figure 3.10.
A finding aid may be a brief or quite lengthy description, depending upon the size of the archival
collection, the complexity of its organization, and the level of granularity desired in the description. Because
the materials in archives are different from the types of information resources found in other institutions, the
types of metadata recorded in a finding aid are considerably different from that found in catalogs or indexes.
The following types of information are found in a typical finding aid:

• Provenance—including creator, title, dates, and bio-graphical or administrative history

• Physical extent and condition—including the amount (number, volume, and linear or cubic feet)
and a statement of the collection’s condition
• Scope and content notes—including a narrative descrip-tion of the nature of the collection, what is
included, the extent and depth of the collection, gaps in the collection, types of materials, and
access points for names and subjects

93
94
95
Figure 3.10. An Example of a Finding Aid. (Source: Beatley Library Collection: CC30, Simmons College Archives, Boston, MA, USA)

• Order and structure—including a description of how the collection is arranged and organized, and
detailed container lists
• Administrative information—including location, conditions of use, restrictions, processing,
citation, acquisition and accession, repository, and conservation information

Registers and Other Museum Databases

Registers constitute the primary collections control tool for museums. A register may also be called an
accession log. It functions like a catalog, although it has additional kinds of access points. The process of
registration in a museum is much like the process of cataloging in a library. During the process, the registrar
identifies the object, the donor, any associations (e.g., having belonged to a particular person), any
information needed for insurance purposes, and so forth. An identification number is assigned. This process

96
of documentation is internal and seldom made available to the public. The accession record becomes the basis
for one or more files that help provide organization and control of the museum’s content.
Additionally, museums sometimes provide databases for public access online. For example, the Pitt Rivers
Museum in Oxford, England, has an objects database and a photos database. One can search each database by
filling in fields in a search record. The objects database (also called the Objects Collection Catalogue) has search
boxes for such things as name of the object, country the object is from (a pull-down menu provides the current
United Nations list of country names), cultural group from which the object comes, material made from,
when collected, and so forth. In addition to the same fields as for objects, the photo database (also called the
Photographs Collection Catalogue) has search boxes for such things as activity depicted, event depicted, person
shown, photographer, photographic process, image dimensions, date of photograph, and so forth. It is
suggested that users should not try to search more than three separate fields in the database at one time.29
This approach is very factual, with a good deal of information provided by the databases, but it lacks a stylized
presentation of the collection that may be found at larger museums.
An example of a more comprehensive and stylized approach is that found at the National Gallery of Art
(NGA) in Washington, DC. The entire collection (drawings, paintings, photographs, prints, and sculptures)
is searchable in at least two distinct methods through their website. The first interface is marked “Search the
Collection.” It allows the public to search the collection’s “object records” by artist name, keywords in title or
in the object description, donors and owners, accession number, exhibition history, and a few other fields.30
After an initial search is complete, users may filter their search results by medium, nationality, time span,
styles, and whether an object is on exhibit. A digital image for many, but not all, works is provided. In
addition, some artworks have extensive descriptions. For example, the search results for Edward Hopper’s
1939 painting Cape Cod Evening begin with some brief information:

• Material/Technique: oil on canvas

• Dimensions:
O Overall: 76.2 × 101.6 cm (30 x 40 in.)
O Framed: 106.7 × 132.1 cm (42 x 52 in.)
• Collection: John Hay Whitney Collection
• Accession number: 1982.76.6

Following this, there is, among other things, an essay that describes the history of the work along with
bibliographic references, images of initial sketches used to create the painting, a technical summary and notes,
descriptions of the work’s provenance, its exhibition history, and resources where the work has appeared.31
Some artworks in the collection have significantly more information provided than others.
The other interface available is through the NGA Images repository. This separate retrieval tool allows
users to search by keyword anywhere in the description of the artwork or in the title, by artist’s name, and
accession number. It provides search filters for nationality, classification (the type of art: painting, sculpture,
media art), medium, and dates. The search results provide downloadable digital images for many of the works,
a link to the object record (mentioned above), and a brief description:

• Artist Luis Meléndez

• Artist Info Spanish, 1715–1780
• Title Still Life with Figs and Bread
• Dated c. 1770
• Medium oil on canvas
• Classification Painting
• Dimensions
O Overall: 47.6 × 34 cm (18 3/4 × 13 3/8 in.)
O Framed: 65.6 × 49.2 × 4.4 cm (25 13/16 × 19 3/8 × 1 3/4 in.)
• Credit Patrons’ Permanent Fund
• Accession No. 2000.6.1
• Digitization Direct Digital Capture

97
 • Image Use open access32

In addition to searching for particular artworks, the National Gallery of Art also provides a searchable index of
artist’s names, which can be filtered by nationality, dates, the first letter of the last name, and so on. This is
one particular museum approach to providing some information about their collections to the general public.
Other cultural heritage institutions approach it in other ways. As time goes along, we may or may not see
more consistency among presentations of retrieval tools on museum websites.

Search Engines
For the majority of information seekers, search engines are probably the most familiar of all the retrieval
tools discussed in this chapter. Search engines are tools developed for computer systems, particularly the
Internet, to find instances of requested words or phrases that can be found in the documents covered by the
scope of the tool. They were developed for the purpose of searching full text documents (or indexes of those
documents) for particular words or phrases. Some search engines are used to search only a single particular
resource, such as the website for an organization such as OCLC. Other search engines are for searching the
entire Internet (although only a small percentage of the Internet is actually searched by each one). Using a
program called a spider (also known as a crawler or robot), each search engine automatically collects
information from web resources and places it into a database of records or in a full-text index that is similar to
a concordance. The spider is typically programmed to go onto the web to retrieve and download copies of web
pages and everything linked to them, everything linked to the links, and so on. Massive collective indexes—
created from the spider’s database—are stored in data centers around the world. These indexes (some
comprising over 100 million gigabytes of data) are what are searched when users enter terms into the search
box.
Obviously, not every web resource has been indexed. Some resources may be indexed by only a few search
engines—thus, a variety of responses can be had when searching different engines for the same concept. In
addition, not every search engine indexes every type of material available online. It has only been in more
recent years that resources in PDF and PowerPoint format have become searchable online. In some cases,
resources may be password protected, and as a result, they may not be indexed in a search engine because they
are not accessible to the crawler. In other cases, some resources may be indexed and returned as search results,
despite being restricted—a fact that the searcher may discover only after trying to access the information. In
general, only the surface web is included in search engine indexes; the so-called deep web (or hidden web)
includes many things the general public would not want to see indexed in search engines (e.g., email
correspondence, financial transactions, dating profiles, nefarious dealings of all kinds).
Unlike catalogs and finding aids, search engines have become a ubiquitous part of users’ routine
interactions with the world (at least for those advantaged enough to have consistent and reliable Internet
access). Due to a general familiarity with search engines and their relative ease of use (just throw some
keywords into the search box!), other retrieval tools often look antiquated and complex in comparison.
Although search engines may or may not provide results that are as intellectually satisfactory as the results
from other retrieval tools, users often report satisfaction because they find something related to what they were
searching for, and find it fast. But most users do not know if what they found is authentic, authoritative, or
the best that is available on their topic. Searches for known items or for specific names are sometimes less
satisfactory than searches for topical information. Some users do not always realize that search engines may
push advertisements and sponsored links to the top of the results page, or that more in-depth information
may be found much further down in the list of results.
There is no doubt that search engines have become more advanced in recent years. Google33 is one of the
most sophisticated both in giving searching assistance (e.g., asking, “Did you really mean to search for . . .?” in
response to a misspelled word) and in display of results, but there are still drawbacks. At the time of this
writing, there is still no real distinction among homographs. For example, on the first page of search results
for the term bridge, Google brings back a mixture of resources about card games, structures spanning
geographic spaces or gaps, specific named highway structures, and some local organizations with the word
bridge in their names. The handling of synonyms and related terms is somewhat mysterious. In Google,
searching funny pets brings back videos about funny pets, but the top search results on the page are for funny
dogs (where the word pets is nowhere to be found on the page). Although humans understand that dogs are a
type of pet, machines do not and cannot know this; they have to be programmed to connect these strings of
characters and to treat them as related concepts. Although Google acknowledges that it has a “synonyms
system” of which they are very proud and that it involves analyzing “petabytes of web documents and historical

98
search data to build an intricate understanding of what words can mean in different contexts,”34 Google is not
particularly forthcoming about how synonym control works. Disappointingly, its synonyms system is not
something a user can explore directly.
With most search engines, displays of search results are usually arranged according to relevance, which can
be calculated in various ways. A search engine may calculate relevance by giving different weightings to factors
such as how many search terms are found in each web page, how often each term is found in a page, whether
the terms are in proximity or dispersed, whether the terms are in the head of the document or buried further
down in its body, and how common (or rare) a particular search term is. Search engines continue to strive for
more sophisticated methods of calculating relevance and for displaying results. For example, Google uses a
formula that examines over 200 different factors in determining relevance. Google uses common factors, such
as word placement and word frequency, in its formula, but also less-expected criteria such as site quality,
geographic region, search history, recency, and site popularity (i.e., how often a page is linked to by other web
pages) as elements in the ranking of search results.35 Google, Bing, and other search engines are continuously
updating, refining, and testing their algorithms and formulas in their quests to improve the Internet searching
experience.
When metadata was first being developed to describe web documents, most search engines did not take
user-supplied metadata into consideration in their relevance ranking due to metadata mischief and
malfeasance (e.g., including popular terms that have no relationship to the site being described). With the
development of linked data, however, the major search engines are now embracing the inclusion of more
structured data as part of the underlying textual markup of web resources. Encoding approaches such as
JSON-LD (Java Script Object Notation for Linked Data), microdata, and RDFa (Resource Description
Framework in Attributes)—in combination with specialized vocabularies (such as the categories established in
Schema.org) to identify necessary semantics—may be used to add more contextual metadata directly into web
resources. For example, when searching recipes on a culinary website, it would be very helpful if the search
engine could distinguish between dietary restrictions associated with the recipe and a recipe’s ingredients.
With support for linked data and for processing contextual metadata found encoded within web documents,
search engines are expected to continue to improve over the coming decades.

Directories
In the 1990s and early 2000s, there were still some alternatives for finding web resources other than
through the use of search engines; these alternatives were directories and subject gateways. As discussed in
Chapter 1, directories were organized collections of links to websites. The directory created by Yahoo!36 was
one of the largest and the best known. In its heyday, Yahoo! employed human indexers to create and maintain
its directory. Humans added subject terms and categorized the records into a hierarchical subject tree index.
Searching was done by browsing (or “drilling down”) through the categories or by keyword searching. A small
number of Internet directories were organized according to traditional library classification, but most had
organization schemes that were created by the persons who devised the directory, usually with 12 to 16 top-
level categories (e.g., Arts, Business, Computers, Finance, Games).
Directories and subject gateways, once the first place to look for websites, have waned considerably in
popularity due to the ascendancy of the search engine and also to the tremendous efforts and resources needed
to maintain extensive lists of ever-changing web resources. Most of the long-standing Internet directories and
subject gateways have ceased operations, but there are a few that remain open. To get a sense of how these
directories were organized, and what searching for information on the web was like in the previous
millennium, please visit one of the last large-scale Internet directories: the Best of the Web directory.37

THE NEED FOR RETRIEVAL TOOLS REVISITED

In today’s heavily web-focused information environment, it is often tempting to think that nothing more
than a search engine is ever needed. Some feel that if a resource cannot be found through Google’s web,
scholar, or book search, then that resource may not be worth finding. This, however, could not be further
from the truth. It can be easy to forget that much of the world’s information is not online at this time and that
much of that information probably never will be. In addition, a great deal of what is available in electronic
form is not necessarily freely available to all. Information is still a commodity. Most e-books are not free; they

99
are sold. A needed article may be behind a paywall. There can be costs associated with information access
online.
Although this may change sometime in the future, individual retrieval tools (beyond search engines) still
have vital roles to play in information gathering. Catalogs, indexes, and other retrieval tools, however, are
often compared negatively to search engines. Students and scholars alike have said, “Why is the catalog still
hard to use?”38 and “Why can’t the catalog be more like Google [or Amazon.com or Facebook]?” The desire
for straightforward and user-friendly tools is understandable and pervasive, and there are information
professionals who are working to rectify the situation. However, we must be careful to remember that this is
not a contest. Not everything is as simple as putting keywords into a single Google search box. There are
times that users and information professionals need more advanced types of searching that are currently not
available in the search engine approach (e.g., an author search rather than a general keyword search).
Different types of retrieval tools have been designed to describe different kinds of materials for various
communities. Information professionals must understand where and how to access the best information for
each particular situation. For example, one must access a finding aid if one wants detailed information about
an archival collection—relying on the brief catalog record for the collection usually will not suffice. Using
Google to get to the Wikipedia page for Eleanor Roosevelt might be a fine starting point for a school report,
but it will not provide comprehensive information about her life and accomplishments; published print
materials, such as biographies and histories, will provide a greater depth of information—and a catalog will
easily retrieve those materials (as long as they are in that collection). As psychologist Abraham Maslow said, “I
suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.”39 We
must be very cautious about overreliance on a single type of retrieval tool. We must not let the web search
engine become our hammer. The trick is to make sure that all the tools in the toolbox are understood, and
that they are used when necessary and appropriate.

CONCLUSION
A century ago Charles Cutter stated that catalogs should enable people to find books for which titles,
authors, or subjects were known; should show what was available in a library by a particular author, on a
particular subject, or in a kind of literature; and should assist in the choice of a book by its edition or
character. A century later, Cutter’s “objects” are still quite appropriate, except that they have been expanded to
all kinds of information resources beyond just books in libraries and to a number of different kinds of retrieval
tools beyond catalogs.
This chapter has discussed the major retrieval tools used in the organization and retrieval of recorded
information. The surrogate records that make up a retrieval tool must be created and encoded, either by
humans or automatically by software programs created by humans. The individual retrieval tools discussed in
this chapter are types of information systems. In the next chapter, we provide an overview of some of the
issues and problems in organizing information as they relate to information systems and system design.
Chapters on metadata and encoding then follow, before settling into a multi-chapter discussion of the
descriptive metadata found in retrieval tools.

100
NOTES
All URLs accessed June 2017.
1. Publication Manual of the American Psychological Association, 6th ed. (Washington, DC: American
Psychological Association, 2010).
2. Chicago Manual of Style, 16th ed. (Chicago: University of Chicago Press, 2010).
3. MLA Handbook, 8th ed. (New York: Modern Language Association of America, 2016).
4. Council of Science Editors, Scientific Style and Format: The CSE Manual for Authors, Editors, and Publishers,
8th ed. (Chicago: Council of Science Editors in cooperation with the University of Chicago Press, 2014).
5. Kate L. Turabian, A Manual for Writers of Research Papers, Theses, and Dissertations, 8th ed., revised by
Wayne C. Booth, Gregory Colomb, Joseph M. Williams, and the University of Chicago Press editorial staff
(Chicago: University of Chicago Press, 2013).
6. U.S. Government Printing Office, Style Manual, 30th ed. (Washington, DC: U.S. Government Printing
Office, 2008). Also available at: https://www.gpo.gov/fdsys/pkg/GPO-STYLEMANUAL-2008/pdf/GPO-
STYLEMANUAL-2008.pdf.
7. Kirsten Madden, Michèle Pujol, and Janet Seiz, A Bibliography of Female Economic Thought Up to 1940
(London: Routledge, 2012).

101