University of Illinois (Urbana-Champaign campus)..

The role of classification in the modern American library : papers presented at an institute conducted by the University of Illinois Graduate School of Library Science, November 1-4, 1959 (Volume 1959) online

. (page 3 of 15)
Online LibraryUniversity of Illinois (Urbana-Champaign campus).The role of classification in the modern American library : papers presented at an institute conducted by the University of Illinois Graduate School of Library Science, November 1-4, 1959 (Volume 1959) → online text (page 3 of 15)
Font size
QR-code for this ebook

"human learning as preserved in books," a classification of the
working, developments and productivity of the mind of man, nay,
of the mind of the Creator Himself, so far as that mind is re-
vealed to us through the phenomena of the universe. To this
great task the loftiest intellects have at time applied themselves,
and still left the work imperfect. In Edwards' Memoirs of Li-
braries are cited thirty-two celebrated schemes of classifica-
tion, and among them those of Bacon, Bentham, Coleridge, Am-
pere, Leibnitz, D'Alembert, and Schleiermacker.

It is needless to say the plain business men, who compose the
present Board of Directors of the Mercantile Library, would not
presume to improve on what these philosophers have left im-
perfect. They were compelled to choose from among such
schemes as lay before them, and after much comparison of the
various systems, including those now in use in the Boston Public
Library, the Public Library and Mercantile Library of Cincin-
nati, the Mercantile Library of St. Louis and others, have
adopted, without hesitation, as being the most complete and


exhaustive of any that have fallen under their observations, the
Baconian System as elaborated by Mr. Wm. T. Harris of St.


The scheme was further expanded for use of the Peoria Library in
a second catalogue, published in 1899. 3T The librarian remarked at
that time that the scheme "continues to give excellent satisfaction as
a working system.'" This catalogue shows variations from the
scheme as used in St. Louis.

Melvil Dewey's classification came into existence three years
after Harris published his scheme in the Journal of Speculative Philo-
sophy, Like Harris, Dewey made no pretense of having produced an
original scheme. He said that he had been influenced by the reading
he had done and the schemes he had examined. He noted that he had
received many ideas from the scheme of Natale Battezati which was
used by the Italian publishers in 1871. He specifically denied the use
of the Harris scheme as a model.

The plan of the St. Louis Public School Library and that of the
Apprentices Library of New York, which in some respects re-
semble his own, were not seen till all the essential features
were decided upon, though not given to the public. 39

Certainly the order of subdivisions follows the Harris order, as Harris
followed Johnston in many cases. That it was Dewey's scheme, rather
than Harris' which became established in American libraries may be
attributed both to the more easily remembered notation of the Decimal
Classification and to the fact that Dewey was an active librarian who
appeared at library meetings and talked much about the advantages of
his scheme. The scheme of Dewey, the practical librarian, was ac-
cepted; that of Harris, the philosopher, is mentioned today only in
library school courses in classification.

When Charles Ammi Cutter became librarian of the Boston Athen-
aeum that library used fixed location for shelf arrangement. Cutter
did not attempt to change this until he had completed a dictionary cata-
logue. He had intended to use Dewey's classification as printed but
upon examination he decided to modify it by adopting a larger base
using the letters of the alphabet to designate classes, and by establish-
ing a system of book numbers based on author entry. He worked out
a local list for designating places that was later adopted for use in
connection with other classification schemes. He was convinced that
this and his other mnemonic devices were superior to Dewey as is
shown in one of his letters to Katharine Sharp:

I am not satisfied with one sentence [of your letter]. "It lacks
mnemonic features which are a help to some people." I should
have said that the E.G. has ten times as many mnemonic fea-
tures as the D.C., it has a good deal of alliterative mnemonics,


all of which the D.C. hasn't because it does not employ letters.
The local list is a good mnemonic assistant. 40

The scheme was designed as a practical means for shelving books,
but a logical outline of knowledge was not ignored. In describing his
classification Cutter said that he had tried:

... to provide a classification at once logical and practical;
it is not intended for a classification of knowledge, but of books.
I believe however, that the maker of a scheme for book arrange-
ment is most likely to produce a work of permanent value if he
keeps always before his mind a classification of knowledge. 41

The Expansive Classification consisted of seven classification
schemes, the first designed for a library of 100 volumes, the seventh
sufficiently minute to accomodate a library of ten million volumes.
The scheme can be said to date from 1879 since the first accounts of
it appeared at that time. The fifth expansion was published in 1882
and the sixth was completed between 1891 and 1893. The work on the
seventh expansion had not been completed at the time of Cutter's
death in 1903 and no complete index for the seventh expansion was

The Expansive Classification ranked next to the Decimal Classifi-
cation in acceptance by libraries but it is impossible to estimate the
number of libraries that adopted the scheme. We have five sources
which give us some information: the Kephart report presented at
the 1893 meeting of the American Library Association; 42 the figures
assembled by the A.L.A. survey of 1920-22; 43 a general statement in
a biography of C.A. Cutter which was published in 1931; 44 a survey of
college and university libraries made in 1953; 45 and a survey of public
libraries made in 1955. 46

At the time of the 1893 report, eighteen of the 127 large libraries
(libraries with collections of 25,000 volumes or more) were using
Cutter's classification. Several of the reporting libraries were in
the process of adopting the sixth classification which had just been
completed. Others reported that they had adopted the scheme as
planned for the Boston Athenaeum and had been using it for a number
of years. Only one library expressed dissatisfaction with the scheme;
the Peabody (Massachusetts) Institute of Technology would have pre-
ferred a simpler scheme. Unfortunately there is no record of the
number of libraries with less than 25,000 volumes that were using the
Expansive Classification, but some early experimental applications
were made in the public library of Winchester, Massachusetts, and it
is believed that other small libraries in that state adopted the scheme.
The A.L.A. survey of 1920-22 reported that twenty of the 1243 public
and semi-public libraries included in the survey used Cutter's classi-
fication. The same survey stated that only four of the 261 college and
university libraries had adopted this scheme. These figures are


obviously incomplete since replies to the questionnaire used in the
1953 survey accounted for at least thirteen college and university
libraries which were still using Expansive Classification as late as
1925. 45 In a biography of his uncle, which W.P. Cutter published in
1931, is a statement that a total of at least one hundred libraries wer
using the scheme at the date of writing the biography. The libraries
are not listed but we assume that this is an approximately correct
figure for the period. Since libraries were unlikely to change from
another scheme to Cutter's in the period between 1924 and 1931 it
must be assumed that the twenty-four public and academic libraries
reported by the A.L.A. survey must represent incomplete returns.
The 1953 survey of college and university libraries found the Expan-
sive Classification in only four of the 744 libraries reporting. The
1955 survey of public libraries, with collections of 25,000 volumes 01
more, found Cutter's scheme used, in whole or in part, by fifteen of
the 863 libraries. As there was no record for libraries with less
than 25,000 volumes in 1893 so there is no record for the smaller
public libraries sixty years later.

The use of the Expansive Classification was not limited to New
England although it was probably used more extensively in Massachu
setts than elsewhere. The Library Society of Charleston, South
Carolina, and various libraries in Texas used it. It found its way to
Montana, and although both the Montana School of Mines and Western
Montana College of Education have reclassified, the Helena, Montana.
Public Library is still a Cutter library unless a change has been mad
in the last few years. A number of colleges in Wisconsin used Cut-
ter's scheme but most of them reclassified a number of years ago.
J.C.M. Hanson introduced Cutter into the University of Wisconsin
where it remained until the very recent reclassification project. To-
day, so far as I have been able to ascertain, Cutter's Expansive
Classification can be found in only three colleges: Lake Forest in
Illinois, Wesleyan, and Mount Holyoke. It is gradually slipping out ol
those public libraries, probably less than twenty at this time, which
have retained it. If there is a tone of regret in my voice, it is only
what you hear in the voices of many classifiers. Expansive Classifi-
cation was a good classification, a classifier's classification it is tru
but easy for patrons to use. It has been called Cutter's best work,
but he will probably be known to future generations for his Rules for
the Dictionary Catalogue* 1 and for his Alfabetic Author Tables* 6

Frederick Beecher Perkins was one of the many brilliant men
drawn into librarianship in the early years. He was deeply intereste
in the profession and articles by him appeared often in library perio(
icals. He also took over the task of preparing catalogues for librarie
His catalogue of the Fall River, Massachusetts, Public Library is an
excellent example of a dictionary catalogue with classified sections.*
He began work as a school teacher in New York City in 1849, moved
on to Newark in 1850, and in 1851 he became assistant in the Boston
Public Library. From 1879 to 1887 he was librarian of the San Fran
cisco Public Library.


It was during the time that he served as Librarian of the San Fran-
cisco Public Library that his classification scheme was published.
The first edition appeared in 1881 and the revision in 1882. 50 He ex-
plained the origin of the scheme thus:

The present classification originated in that drawn up a good
many years ago, substantially on the basis of Brunet's or the
"Paris system" by Mr. S. Hastings Grant, long the courteous
and efficient Librarian of the New York Mercantile Library.
The catalogues of that library consisted of two parts viz., an
alphabet by authors' names (and anonymous titles), and the clas-
sification under topics. There were thus no title- entries proper
at all. This scheme had more merit for practical purposes than
has been attributed to it. I revised this work of Mr. Grant's
twice over, for successive catalogue issues of that library in
1866, 1869 and 1872, each time enlarging the number of topics
or ultimate sections. If I had prepared another catalogue for
that library, I meant to make the classification such as I have
now made it. 51

However, a comparison of the Grant and Perkins outlines shows some
differences. 52 It is noted that in main classes Perkins' scheme fol-
lowed Brunet more closely than Grant's did. In his later revision
Perkins continued the Brunet form.

In explaining the changes between the revised edition of his scheme
and the original edition Perkins said:

A few sections or topics have been added, some of them from
the well-considered classification recently published by Mr.
Lloyd P, Smith, Librarian of the Philadelphia Library Company,
and some other minor alterations and additions have been
made. 53

The State Library of Iowa adopted Perkins' scheme about 1883, and
the State Library of Nebraska was also using it in 1893. San Fran-
cisco did not use it but continued to use Dewey although Mr. Perkins
said that the more he used Dewey's scheme the less he liked it. 54

The Library Company of Philadelphia, founded in 1731, had a
Baconian catalogue in 1789, 55 but the scheme of the French book-
sellers was adopted by George Campbell when he prepared the clas-
sified catalogue which was published in 1807. The books, however,
remained on the shelves in size groups by order of accession. At the
1853 Librarian's Conference, Lloyd P. Smith, at that time librarian of
the Library Company of Philadelphia, mentioned the difficulty of lo-
cating books arranged in this manner. When the Loganian Library
was moved to the Ridgeway Branch in 1878 it was decided to adopt a
classified arrangement of books on the shelves. Smith developed a
new scheme, still using Brunet as a base. He used the five traditional


classes of Religion, Jurisprudence, Science and Arts, Belles Lettres,
and History, and made a sixth class for Bibliography and History of
Literature. The scheme was published in 1882 with an alphabetic in-
dex which provided notation for "Mr. Dui's system of classification"
as well as for Smith's scheme. 56

Smith was much more flexible in his approach to classification
than most makers of classification schemes. Usually the author is
anxious that his scheme be not tampered with in any way, but Smith
stated that while he thought of his six main classes and their sub-
classes as permanent, he felt that the subdivisions could be adjusted
as a librarian wished. He further suggested that anyone who wanted
to expand his scheme could do so by consulting Brunet for examples.
It may well be that the scheme was adopted more widely than our
present records show. Its simplicity (except in notation of sub-divi-
sions) and the possibility of adapting to local needs might have ap-
pealed to librarians who heard it discussed at library meetings.
Central College in Fayette, Missouri, used it and the Kansas State
Library, at Topeka, was using it, with additions, in 1893.

John Edmands, librarian of the Mercantile Library of Philadelphia,
was rescued from virtual oblivion by Verner Clapp's article, "A.L.A.
Member Number 13: A First Glance at John Edmands." 5 Edmands
served as assistant librarian to the Brothers of Unity Society at Yale
in 1845. After graduation he taught school for a year and then returned
to the Yale Divinity School. After that graduation he became assistant
in the Yale College library and helped develop a classification scheme
for use there. In April of 1856 Edmands accepted a temporary ap-
pointment to prepare a supplement to the catalogue of the Mercantile
Library of Philadelphia. The librarian resigned shortly after he
arrived and Edmands was appointed to his position. He retired in 1901
at the age of 80, but remained as librarian emeritus until his death in

Edmands found a .classified catalogue, arranged under thirty-four
main headings, which was unsatisfactory to him; he proceeded to
develop a new scheme. 58 His aim was to arrange books on the shelves
so that they could be found without using a catalogue. He reduced the
classes to twenty-three; twenty-two of them designated by letters of
the alphabet, omitting I, Q, U, and Z. Prose fiction, the twenty-third
class, was left without a notational symbol. An examination of the
scheme shows nine of the twenty-three classes devoted to history,
and three to literature. Edmands also developed an author notation
using figures from 1 to 9,999. No author initial was required with
this scheme. A similar scheme is found today in Benyon's Law
schedule. 59 H. J. Carr, librarian of the Cedar Rapids, Iowa, Public
Library, seems to have adopted the Edmands scheme in 1884. Min-
neapolis adopted it about 1889 under Herbert Putnam. Certain changes
were made in Minneapolis both in the order of classes and in the author
numbers. 60 Biography and music still follow the Edmands scheme in
the Minneapolis library, according to information provided on a ques-


William Frederick Poole, one of the contributors to the 1876 report
on Public Libraries in the United States, was actively engaged in li-
brary work from his undergraduate college days until his death. At
Yale he was the assistant to John Edmands in the Brothers of Unity
Library. In 1851 he became an assistant in the Boston Athenaeum,
and eleven months later became the librarian of the Boston Mercantile
Library Association. In 1856, when the librarian of the Boston Athen-
aeum retired, he was appointed to the position but he resigned sudden-
ly in 1869 and for a period of two years was a consultant in the organ-
ization of libraries. He became librarian of the new Cincinnati Public
Library in 1871 after serving as consultant for the library. In 1873 he
became the first librarian of the new Chicago Public Library, and in
1887 accepted the task of organizing a new reference library being
formed in Chicago. He remained at the Newberry Library until his
death in 1894.

Poole emphasized the dictionary catalogue as a means of finding
material in the library and used a rather loose shelf arrangement
for books. He has been credited with originating the dictionary cata-
logue which was later developed by Cutter,

.... the modern dictionary catalogue combining authors and
subjects in one alphabet which it is to the credit of Mr. Poole
to have invented, and of Messers. Cutter, Noyes, and others to
have developed. 61

But if Poole has been deprived of honor due him for the dictionary
catalogue, perhaps it can be balanced by the undeserved credit, which
properly belongs to Edmands, given him for originating the index to
periodical articles. 62 His shelf arrangement was a practical means
of assembling books. Letters which were assigned to the classes stood
for cases, and a block of numbers, thought to be sufficient for probable
titles in that class, was assigned to the letters. If an unexpected num-
ber of books was added a new block of numbers was begun.

This very flexible scheme was applied to the Enoch Pratt Free Li-
brary, the Chicago Public Library, and the Newberry Library. The
scheme used in the Indianapolis Public Library was basically the
same scheme although there were more classes and fewer numbers
assigned to individual classes. Omaha used the same scheme as
Indianapolis. Since these schemes were developed by individual li-
braries, within the framework set up by Poole, they are often thought
of as local schemes.

Josephus Larned, who is known to all librarians as an historian,
is one of several librarians who used the Decimal Classification of
Melvil Dewey but developed a scheme of his own which presumably
pleased him better. Larned had been a bookkeeper and a newspaper
reporter before he was elected superintendent of public instruction
in Buffalo, New York, in 1871. In 1877 he was appointed superintendent
of the Buffalo Young Men's Christian Association Library with the


understanding that he was to reorganize the collection. The books
were not classified and after studying various available schemes he
selected Decimal Classification for his use. It has been claimed that
this was the first library to adopt the Decimal Classification. Larned's
own scheme was developed in 1884. It was an interesting scheme,
consisting of a series of tables which could be coupled together to
represent more minute divisions of subjects. 63

The classification which J.C. Rowell developed for use in the Uni-
versity of California Library in 1892 was based on the curriculum of
the University and. was prepared with the cooperation of certain
members of the faculty. 64 The arrangement of mathematics, for ex-
ample, was a slightly modified version of one prepared by Professor
Irving Stringman. Rowell wanted the shortest possible notation and in
order to decide how much space he would need for each class he
counted the books in various classes. He used A for Bibliography, B
for Encyclopedias, and C for Periodicals and built his series of sub-
ject areas on a base of 999. The outline resembled Dewey, with
certain variations, but there were no mnenomic features and the
notation was brief, although lower case letters were added to mark

Mr. Rowell did not know whether any libraries had adopted his
scheme. He replied to Miss Sharp's question in 1896 by saying:

No attempt to introduce the classification into other libraries has
been made; and I can not tell if it has been adopted elsewhere,
although from the very frequent calls for it, I believe it has
been, at least in modified form. Mr. Fletcher of Amherst has
thought very kindly of it, and perhaps knows of particular li-
braries using it. You might inquire also of the University of
Minnesota Library. 55

The only record of use of the scheme that has come to my attention
was at Haverford College, Pennsylvania, where it was used until re-
classification in 1923.

William I. Fletcher, librarian of Amherst had published the first
draft of a proposed classification scheme in the Library Journal as
early as 1889. He stated that his scheme was designed:

To offer a way of escape for those who shrink from the intri-
cacies and difficulties of elaborate systems, and to substitute
for painstaking analytical classification a simple arrangement
that is better adapted to be practically useful in a library while
doing away with most of the work involved in carrying out one
of those schemes. 66

A revised version of the scheme was included in his Public Libraries
in America 61 in 1894, and it was issued separately with alterations,
additions and index later the same year. 68 In the letter which he sent


to Miss Sharp in response to her request for copies and in answer to
her question about adoption of the scheme by libraries he could only

I am sorry that I cannot refer you to any library using my clas-
sification. I have paid no attention whatever to the question of
its use in any place and do not suppose it has been adopted in
many. 69

This superficial summary shows us that the 19th century was a
period of intense interest in classification. The leaders in the library
world were concerned with this aspect of librarianship. The majority
of these men were well known as librarians, scholars, historians,
authors, etc.; almost all of them appear in the Dictionary of American
Biography. Seth Hastings Grant and Lloyd P. Smith were active at the
1853 conference of librarians. Dewey, Cutter, Edmands, Poole, Smith,
and Lamed were charter members of the American Library Associa-
tion. Poole, Cutter, Dewey, Fletcher, and Larned were presidents of
A.L.A. In that period men of stature were interested in classification
and individual schemes were the rule. So long as books remained on
the shelves in fixed location, new schemes could be adopted for the
printed catalogues. This allowed a degree of experimentation that is
impossible in today's large libraries, which shelve their books in re-
lative location.

When Kephart sent out his questionnaire to the larger libraries in
1893, 70 he asked what classification scheme was used by the library.
He listed the schemes of Cutter, Dewey, Edmands, Fletcher, Harvard,
Larned, Perkins, Schwartz, and Smith, but learned later that the Har-
vard system had not been printed in full and that the Larned classifi-
cation had not been completed. Of the 127 libraries which replied,
eight were using Cutter, thirty-seven were using Dewey, two reported
Edmands, two Perkins, one Schwartz, two Smith. No libraries in-
dicated the use of Fletcher, Harvard, or Larned, but two reported the
use of Shurtleff, five Poole, 71 and two Harris. Sixty-one of the 127 large
libraries chose one of nine different published classification schemes;
sixty-six libraries were using local schemes, or a system of fixed
location. There are no records available for the libraries with less
than 25,000 volumes.

Sixty years later when I attempted to secure information used in
about 2,000 libraries, including all college and university libraries
and all public libraries with collections of 25,000 volumes or more, I
found a very different picture. Instead of nine classification schemes
there were two major ones; local scheme were rare. Of the public
libraries answering the questionnaire 93% used the Decimal Classifi-
cation. Of the college and university libraries, 84.6% used Decimal
Classification and 13.8% used Library of Congress Classification. A
survey taken today would alter the percentages slightly; one by one
the Cutter libraries convert to one of the two common schemes; a


certain amount of changing from D.C. to L.C. continues in the college
and university field. With the passing of the 19th century the interest

1 3 5 6 7 8 9 10 11 12 13 14 15

Online LibraryUniversity of Illinois (Urbana-Champaign campus).The role of classification in the modern American library : papers presented at an institute conducted by the University of Illinois Graduate School of Library Science, November 1-4, 1959 (Volume 1959) → online text (page 3 of 15)