Indexes and religion: reflections on
research in the history of indexes
Bella Hass Welnberg
A search of the Hebrew manuscripts and incunabula in the Vatican Library and other large Judaica
collections yielded many dictionaries and citation indexes but almost no subject indexes, although the
latter are common in Latin Christian manuscripts and incunabula. Reasons for this are suggested, and
the compilation of early indexes for religious purposes is discussed. General conclusions relate to the
methodological and terminological difficulties of conducting research on the earliest indexes, which
were compiled in languages and scripts that are little known. The original manuscripts of early indexes
are scattered in libraries throughout the world. Writings on the history of indexing are not concentrated
in a single discipline: Descriptions of early dictionaries, concordances, and indexes are often found in
religious publications rather than the literature of library-information science. The earliest reference tools
related to religion are presented in a chronological table.
Approach to the study
(distinct from Italy) in terms of the number of Hebrew
This is not the article I thought I would be writing after doing
research on early Hebrew indexes at the Vatican Library. I had
previously done a study of the earliest Hebrew citation indexes
— reference tools that indicate in which later works a given
document is cited (Weinberg 1997), and in summer 1998
manuscripts held (Israel 1968, Foreword).
In sum, had Hebrew subject indexes been compiled in
substantial numbers during the age of the codex and early printed
book, they should have been represented in the holdings of the
Vatican Library.
traveled to Italy, the cradle of Hebrew printing (Amram 1963, p.
9), in search of the earliest Hebrew subject indexes. Cognizant
of the special characteristics of this consonantal Semitic script,
Methodology and terminology
I expected early Hebrew subject indexes to differ structurally
How did I determine which early Hebrew works in the Vatican
from those in the Latin alphabet. I planned to analyze their
Library might contain indexes ? The descriptions of the Vatican' s
structure and write up the results.
incunabula in this language are concentrated in a special section
In Hans Wellisch's (1994a) study of indexes in incunabula
of the four-volume catalog of all the Library's incunabula
(books printed before 1500), no Hebrew works are cited,
(Biblioteca 1997). [Tishby's list of Hebrew incunabula in the
although Dr. Wellisch knows that language well. Perhaps
Vatican (1983, pp. 853-7) lacks several titles acquired recently.]
Hebrew incunabula were not well represented in the collections
I filled out a call slip for the Vatican's most complete copy of
that Wellisch sampled for his study. Estimates of the number of
each work because I was interested in examining not only
Hebrew incunabula range between 139 (Offenberg 1990, p. xiii)
indexes,
and 175 (Sheehan 1990, p. [365]) — a tiny percentage of the
pagination or foliation, running heads, and tables of contents.
but
also
relevant
typographic
features
such
as
estimated total number of incunabula in all languages, whether
The last-mentioned feature is considered the forerunner of
one accepts Sheehan's figure of 40,000 or the more conservative
indexes (Knight 1968, p. 14).
statistics cited by Wellisch: 36,000 or 27,000 (1994a, p. 3). The
Personal examination of older works for access features is
Vatican Library owns 46 Hebrew titles from this period, some
necessary because standard bibliographic descriptions of rare
in more than one copy. Tishby (1983, p. 853) notes that the
books frequently fail to note the inclusion of tables of contents and
excellent condition of the Vatican's Hebrew incunabula is
indexes. Dunkin's (1973) manual of rare book cataloging does not
unparalleled in any other collection in the world: The vast
require this, and neither term is found in the manual's detailed
majority of these books are complete; in particular, the first and
index. The catalog record for the Nuremberg Chronicle, which I
last leaves are present. (Indexes to early printed books were
examined at the National Library of Canada (call no. Reserve D
sometimes placed at the front. The index to the Nuremberg
17 S35 fol.), carefully notes missing blank leaves, but does not
Chronicle is a prime example, noted by Witty [1965], p. 43.)
identify the Registrum as a substantial index. {Registrum is an
in
ambiguous term: it is frequently found in incunabula at the head
manuscripts, reporting that the earliest ones were compiled by
of a list of signatures, letters or numbers indicating the sequence
Christians for works in Greek and Latin (p. 268). The Vatican
of quires or gatherings — groups of pages sewn together
Wellisch
(1994b)
has
also
written
about
indexes
codices
[McKerrow 1928, p. 25].) The terms tables ofcontents and indexes
(manuscripts in book form [singular codex], as opposed to
are lacking as well in a thesaurus entitled Printing and publishing
scrolls) and is ranked sixth among the countries of the world
evidence, issued by the Association of College and Research
possesses
more
than
800
Hebrew
The Indexer Vol. 21 No. 3 April 1999
manuscript
111
Indexes and religion: reflections on research in the history of indexes
Libraries (1986). Furthermore, this thesaurus treats as equivalent
the terms foliation, pagination, and columniation; the first and last
are lacking in the Vatican's copy. Personal examination of this
incunabulum revealed, however, that the detailed table of
terms mentioned are cross references to pagination in this
contents is defective, and there is no index.
controlled vocabulary. These three phenomena are very different
when the precision of index locators is being considered, and
should have been separate descriptors.
(literally, "that which shows the place"), which designates both
In his bibliography of Hebrew incunabula, Tishby (1983, pp.
810-11)
provided
more
detailed
In my study of citation indexes (Weinberg 1997, p. 325), I
explained the ambiguity of the Hebrew term mar'eh makom
descriptions
than
his
a reference to an earlier source and a citation in a later source.
This ambiguity complicated my search for citation indexes,
predecessors. Some of the data elements he added, such as
along with the use of the term concordance for these tools in
watermarks, are germane to the dating of books. Tishby records
English-language sources. In documenting the present paper, I
the beginning and end of each quire, which has implications for
encountered multiple meanings for the latter term as well. Lyons
tables of contents and indexes, but he does not have a field that
and Smith (1910, p. 186) discuss the interchangeability of the
explicitly notes the presence or absence of these features,
terms index, dictionary, and concordance. The identification of
although he has such a field for running heads.
both early subject indexes and citation indexes, and the analysis
Personal inspection of incunabula is also desirable because a
book without a printed index may include a handwritten index
of their structure, can thus be done only through examination of
the primary sources.
compiled by its owner (Wellisch 1986, p. 76). Tishby's brief
descriptions of the Vatican's incunabula report missing leaves
but do not record handwritten addenda.
It was possible for me to examine all of the Vatican's Hebrew
Findings
There were no "back-of-the-book" indexes in the Hebrew
incunabula; however, given the quota of manuscripts that
incunabula
researchers at the Vatican Library may request each day, as well
subsequently at the Lowy Collection of the National Library of
as the limited amount of time I had in Rome, I could not examine
Canada and the Rare Book Room of the Jewish Theological
all 800 Hebrew codices. I therefore selected from the printed
Seminary of America (JTS); the last mentioned has the largest
catalogs works that I thought were most likely to contain indexes,
collection in the world (129 titles). In examining these works I
such as those in the disciplines of astronomy and medicine. I
observed that Hebrew incunabula share many typographic
requested approximately 150 codices, many of which contain
features with their Latin counterparts, notably, the method of
multiple works bound together.
A major terminological problem for this research is that the
word index was used in the Latin catalogs of the Vatican
(Biblioteca 1756,1831,1956) — as recently as four decades ago
— with all the meanings enumerated by Wellisch (1996) for the
earliest occurrences of the term. Thus I examined many codices
for which the catalog description contains the word index, but
the work itself has only a table of contents or a list of books.
When I continued my research in New York, I found that the
catalog of the Hebrew science manuscripts at the Jewish
Theological Seminary (1990), whose language of description is
that
I
examined
at
the
Vatican
Library
or
numbering signatures — mutatis mutandis — Hebrew letters
were substituted for Latin ones as well as for arabic numerals.
(Witty [1973, p. 194] states that the use of Hebrew letters as
numerals is found in the Bible.) Some famous early Hebrew
printers were not Jewish, and those who were learned their craft
from Gentiles; thus the conventions of Latin incunabula were
transferred to Hebraica — only in the opposite direction (right
to left). The absence of subject indexes in early Hebrew books
can
thus
not be explained
by
isolation
of the
language
community or by lack of communication between Jews and
Christians.
English, uses the term index of contents for table of contents.
The only alphabetical Hebrew reference tools that I found in
Even the phrase index of subjects (in the description of ms. no.
manuscript form were dictionaries of biblical roots or Talmudic
2725) led to a list of section headings arranged under chapter
terms. These were also among the first printed Hebrew works.
numbers, i.e., a table of contents.
Sheehan (1990) lists two editions of Sefer ha-Shorashim ("The
The Latin catalogs of Hebrew manuscripts at the Vatican have
Book of roots") by David Kimhi as well as the Arukh, an early
more detailed descriptions than the Hebrew one (Israel 1968),
dictionary of the Talmud by Nathan ben Jehiel, among the
but the latter is more up to date, i.e., it covers more works. A
terminological trap in the Hebrew descriptions is the ambiguity
of that language's term for index — mafteah, from the root
meaning 'to open'. The noun mafteah also means 'key' (the
symbol of the American Society of Indexers), and works
containing this Hebrew word in the title are not necessarily
indexes. For example, Mafteah ha-nikud u-mafteah ha-dikduk
(catalog no. Vat. Ebr. 107-5), which might have meant "Index
Vatican's incunabula. Such dictionaries may be viewed as
indexes because they are arranged in a known order, are in a
different order from the text to which they refer (ANSI 1984, p.
7), and include locators in the form of source references. An
index of roots resembles a concept index because it brings
together semantically related words. The process of compiling a
root dictionary is, however, different from that of writing a
subject index. The former involves grammatical analysis of all
to [Hebrew] vocalization and grammar", is not arranged in a
the words in a text, while the latter entails considering which
known order. The title is therefore to be translated "Key to
words are significant in running text. The Vatican has medieval
[Hebrew] vocalization and grammar". The Latin word clavis
manuscripts of Hebrew dictionaries, but not the original ones.
(key) is used in the Vatican catalogs to translate mafteah in this
The latter are described below in the section on The Origins of
sense. Like the word index, mafteah is also applied to tables of
Alphabetical Indexing.
contents. Tishby (1983, p. 854) uses the term in this sense in
There were no subject (concept) indexes in the Hebrew
describing Jacob ben Asher's legal code. Entry no. 9 of Tishby's
manuscripts I examined at the Vatican — only citation indexes.
list states that four leaves of mafteah ha-dinim (index of laws)
Reasons for this are proposed in the following section.
112
The Indexer Vol. 21 No. 3 April 1999
Indexes and religion: reflections on research in the history of indexes
controversial for its lack of citations to Biblical and Talmudic
Subject indexes vs. citation indexes
One day I came to our Rome hotel from the Vatican Library and
asked my husband (Gerard Weinberg), "Why do you think the
Jews produced citation indexes while the Christians compiled
subject indexes — because the Jews had layers of text such as
the Bible, Talmud, and codes?"
"No," he responded, "It's because Christianity was a rejection
of legalism, while law was central in Judaism." This explanation
made sense, as the citation indexes I had explored included many
with a legal purpose. For example, 'En mishpat (Fount of
sources, and was banned for a while (Ben-Sasson 1972). (See
also the section on Religious Censorship and Indexes, below.)
I am convinced that citation indexing began with Hebrew texts
because of the "layer" theory suggested above. Gaster believed
that citation indexes originated as access tools to single works,
but his article preceded Fishman's description of a Biblical
citation index to multiple works.
Justice), the citation index embedded in the Talmud since 1546,
The origins of alphabetical indexing
shows where a law from that compilation is cited in the code of
Rouse and Rouse report (p. 11) that subject indexing was
Maimonides (1135-1204), who lived about 1000 years after the
invented in Paris in the thirteenth century. (This is ironic in light
Talmud was recorded. An article sent to me by a colleague
of the absence of indexes in many modern French works.) It is
(David Levy of Enoch Pratt Library) after my return from Rome
an interesting coincidence that the citation index described by
includes an explicit statement on the Christian rejection of law:
Gaster was produced in the same century as the first subject
"The end of the Law is Christ" (Romans
10:4, cited by
indexes, and in the same country — in the city of Avignon.
Grundmann 1968, p. 103; the King James Authorized Version
Wellisch (1994b, p. 268) suggests that subject indexing began in
has the variant reading "For Christ is the end of the law").
the 4th century with the Apothegmata, a compilation of sayings
Another colleague, Dr. David Gilner of Hebrew Union
of the Greek Church Fathers. Witty describes this as an
College, later pointed out that the Christian rejection of law was
alphabetically arranged tool rather than a subject index to a
short-lived. As Christianity became an organized religion, it
narrative text (1973, p. 196). Richardson (1939, p. 844) states
developed a body of law. Phillimore's (1910) survey of canon
that the earliest Biblical dictionary was the Onomasticon by
law shows how early this occurred. As the corpus of Christian
Eusebius (264-340 C.E.), but it was not in alphabetical order.
laws grew, it required an index. Rouse and Rouse (1979, p. 5)
Bacher (1912) notes that a Greek dictionary of Biblical proper
describe "a primitive subject index" to the canonical collection
names is ascribed to Philo Judaeus, who lived in Alexandria from
of Cardinal Deusdedit, written between 1083 and 1087.
20B.C.E.to40C.E.
According to the Rouses (p. 4), indexing began in the Latin
community, and Hebrew indexes came later. They suggest that
the first Christian citation index, Tabula septem custodiarum
super bibliam (which they call "a concordance to the incidental
passages in the writings of the fathers") appeared in the
mid-thirteenth century (pp. 18-19). Professors of theology with
whom
I
had previously
discussed citation
indexes
were
unfamiliar with this genre, which apparently is not found
frequently in the early works of Christianity.
Fishman (1935) described a Hebrew citation index written in
the late 1100s, arranged in alphabetical order by the quoted
verse. At the Vatican I examined a citation index arranged in
canonical order, i.e., the order of the Biblical text, bound with a
Hebrew translation of Maimonides' Guide for the Perplexed
(originally written in Arabic), copied in the year 1205 (catalog
no. Vat. Ebr. 263). Both the form and structure of this index are
of interest: (a) the headwords (the names of the cited Biblical
books) are bolded and run in, i.e., they are not at the margin; (b)
the locators (the section number in Maimonides' code) precede
the subheadings (the Biblical verses). This is the earliest citation
index I have seen to a single book, and the genre was
undoubtedly much older, as Gaster (1929, pp. 280, 281) states
in describing a late 13th-century citation index.
In the Jewish tradition, Biblical verses served as concept
symbols,
and
citation
indexes
were
effective
for
several
purposes: (a) homiletics (preaching), an application emphasized
by Gaster, (b) to show the development of a law, and (c) for
vertical legitimation (Weinreich
1980, p. 207) — to lend
credence to a text through citation of a classic authority. I believe
that
the
last-mentioned
reason
was
particularly
true
for
The Rouses state that dictionaries and concordances were the
antecedents of subject indexes (pp. 7, 34). MacFadyen's (1910)
survey of concordances begins with 13th-century Latin tools, but
Busa (1971, p. 595) says that "Biblical concordances were very
probably in existence in the seventh and eighth centuries". One
of the sources he cites, Leclercq (1948), describes a Greek
concordance of geographic names dating from the 6th-7th
century, but it was not in alphabetical order.
Wellisch considers the Masoretes, who standardized the text of
the Hebrew Bible, "forerunners of the concordance makers"
(1985, p. 72). In his survey of Hebrew Biblical concordances,
Wellisch (ibid.) notes that Hebrew word lists were compiled by
the Masoretes, who began their work in the 4th or 5th century. (Dr.
Seth Jerchower of the Jewish Theological Seminary says that
Masoretic activity was not that early; the 7th century would be a
more accurate date [personal communication].) I examined several
Masoretic lists in early Biblical manuscripts at the Vatican.
Bacher (1912) says that the first Hebrew lexicon was to the
Talmud, the compilation of Oral Law which was written down
much later than the Bible. This lexicon was compiled in the last
quarter of the ninth century by Gaon Zemah ben Paltoi of
Pumbedita. Lyons & Smith (1910, p. 196) state that Zemah's
dictionary was in alphabetical order. Brisman reports that this
work was more of a glossary than a dictionary. The manuscript
did not survive, but scholars have proven that it existed. The
definitions may have been in a language other than Hebrew, but
in that alphabet (Brisman, forthcoming).
The Rouses claim that the earliest reference works, Biblical
distinctions, were compiled in the middle of the 12th century (p.
Maimonides' Guide to the Perplexed, a controversial work in the
8). In their words, "Distinction collections provide one with the
Jewish community: The Biblical references lent the work
various figurative or symbolic meanings of a noun found in the
authority.
scriptures, illustrating each meaning with a scriptural passage"
Maimonides1
legal
code,
The Indexer Vol. 21 No. 3 April 1999
Mishneh
Torah,
was
113
Indexes and religion: reflections on research in the history of indexes
(p. 7). Hebrew dictionaries of Biblical roots antedate these.
According to Bacher and other sources, the first Hebrew lexicon
of the Bible was compiled by Saadia Gaon in the year 913. Like
distinctions, it included Biblical passages; thus it is a type of
index. Saadia's lexicon, entitled Egron, covered not just the
Bible, but the entire Hebrew literature of his era (Brisman,
forthcoming). Menahem ben Jacob Ibn Saruq compiled a
Hebrew dictionary called Mahberet [Notebook] in the 10th
century, and Jonah Ibn Janah wrote Sefer ha-Shorashim [The
Book of roots] in the 1 lth century — all antedating distinctions.
Richardson (1939) points out that alphabetical order is found in
the Hebrew Bible itself: in Psalm 119 the initial letters of the
verses are in the order of the Hebrew alphabet. (Psalms 34 and
145 also have such an arrangement [S. Brisman, personal
communication].) The Psalms are attributed to King David, who
lived from 907 to 837 B.C.E. (Kantor 1989, pp. 41, 48);
Encyclopaedia Judaica gives the dates of his reign as 1010-970
B.C.E. (Oded 1972, col. 1318). Witty (1973, p. 194) says that
the order of the Hebrew alphabet goes back to the second
millennium B.C.E.
Early Hebrew dictionaries serve to demonstrate that the
concept of alphabetical arrangement was known in that language
community at the time when Latin indexes were being compiled.
Table 1 summarizes the statements made by various authorities
on the earliest indexes and lexicons related to religion.
Table 1. Chronology of alphabetically arranged texts, dictionaries, concordances, and indexes
related to religion
'Brisman (forthcoming) cites sources indicating that Mordecai Nathan is the author.
114
The Indexer Vol. 21 No. 3 April 1999
Indexes and religion: reflections on research in the history of indexes
censored. This codex is of interest to bibliographers, as it shows
Religion vs. science
It emerges from all the cited works that early indexes were
compiled primarily for the advancement of religion. Wellisch
(1994a, p. 4) expressed surprise that more theological works than
medical books from the incunabula period contain indexes. The
same pattern obtains in the Hebraica I examined: The medical
treatises generally lack subject indexes, while citation indexes
to religious texts are common.
Shortly
before
I
flew
to
Rome,
Robert
Singerman,
a
well-known Judaica bibliographer, sent me a note about an index
to a Hebrew medical work printed in 1491; the index exists in
manuscript in the Parma Library (Bloch 1976, p. 134n58). Parma
is quite far from Rome, and I did not have time to make the
pilgrimage to see this early index. It may be the exception that
proves the rule: The medical work itself was printed, but not the
index; the indexer, a physician named Israel Nathan Soncino,
probably felt the need for an access tool. Since he influenced his
son to found a Hebrew printing press (Habermann 1972, col.
141), it is likely that Dr. Soncino was aware of indexes in Latin
books. In New York I reviewed the entire science manuscripts
catalog of the Jewish Theological Seminary; no indexes were
noted
in
the
descriptions
of
13th-15th
century
Hebrew
manuscripts in the pure or applied sciences, including medicine.
As Taylor (1966, p. 31) put it, "theology was ... the queen of
the sciences" in the Middle Ages. The Rouses suggest that
subject indexes were compiled primarily to aid preachers. In
Wellisch's table of indexes by subject (1994a, p. 4), Sermons has
rank 1, and Theology rank 2. There are two other categories in
Wellisch's table that relate to religion — Bible commentaries
and Lives of saints. If we add up the percentages of all these
categories, it emerges that 59% of incunabula with subject
indexes are in the realm of religion. Wellisch (p. 10) admits that
the number would have been even higher had his sample not been
biased: he selected more herbals and medical works (60% of the
total), given his interest in this genre (Wellisch 1978). I saw
many alphabetically arranged Hebrew/multiscript herbals at the
Vatican and JTS. Such reference tools are not indexes, however,
because they do not refer to a text in a different order.
It is surprising that Wellisch (1994a, p. 10) wrote that "Bibles
... did not need any indexes", as most early reference tools
focused on this sacred text. The basic concepts of early
Christianity could not be adequately retrieved through a word
index or concordance to the Bible, and therefore subject indexes
were necessary.
which Hebrew works were known at the time, often citing
specific editions. But the indexer's perspective is of particular
concern here: The process of compiling this reference work was
surely almost identical to that of writing a subject index.
Someone had to read through the complete text of numerous
works and identify the passages that were objectionable from a
Christian theological perspective. This is concept indexing, not
a simple process of word extraction. Precise locators (e.g., "at
the head of the page", "at the bottom of the page") were
developed to assist censors in finding the passages to be blacked
out. One can only imagine how much time the compilation of
this reference work took.
Although not identified or indexed as such in the catalog of
Vatican Hebrew manuscripts (Israel 1968, p. 85, entry 728), this
work is apparently an expanded edition of Sefer ha-zikuk or
Index
expurgatorius,
by
Domenico
Irasolimitano
(1596).
Porges1 (1903) description of and excerpts from Domenico's
work do not include the precise locators of the 1626 edition, nor
does
the copy of the original
Theological Seminary. The
manuscript at the Jewish
1596 edition is not arranged
alphabetically by title, nor does it have an index of titles. The
1626 edition has both of these features.
We may compare such censors' tools to stoplists, which have
been called negative thesauri. These "indexes" required censors
to be on the lookout for certain terms, and either black them out
or replace them by euphemisms. Hebrew terms for Gentiles
(goyim [literally 'nations'] or nokhrim [literally 'strangers'])
were frequently to be replaced by the acronym for 'idol
worshipers' —akum. Not all occurrences of these terms were to
be changed, however. The introduction to the 1596 edition of
Index expurgatorius allows the word Gentile to remain in
discussions of Jewish law, where the use of the term is not seen
as derogatory. This involves an indexer's judgment: substitute
"derogatory mentions" for "significant mentions".
Sonne
(1976,
pp.
204-5)
discusses
the
Hebrew
terms
associated with censorship, including several which connote
"thorough examination", the hallmark of indexing. He makes
another observation that is germane to the subject at hand
(pp. 220-22): Censorship of Hebrew books in the 16th century
led to the disappearance of indexes that had been found in earlier
editions of these works. It is sobering that index-like tools
created for censors had the result of limiting access to religious
texts. Although Intner's article on censorship in indexing
(1984) does not discuss religion, it includes statements that
Religious censorship and indexes
The most interesting and comprehensive (albeit not the earliest)
are applicable in this context:"... indexes are tools of discovery
... that is why it is so hard to think of an index as being used to
withhold information" (p. 105).
Hebrew indexes that I saw at the Vatican Library were compiled
The purpose of censorship in the Catholic Church was to
for Christian religious purposes. A Hebrew incunabulum of the
"guard against heretical writings" (Boudinhon 1910, p. 374). I
Bible dated 1488 (Sheehan 1990, entry no. 3) contains a
handwritten index (undated) that notes all the occurrences in the
recall reading that the first indexer of the Bible was burned at the
stake. Rev. Norman Hillyer, former editor of the "Indexes
text of the Hebrew word 'almah. The translation of this word
Reviewed" column of this journal, pointed me to a source that
('young woman' or 'virgin') is crucial to the Christian tenet of
documents this anecdote: "John Marbeck is responsible for the
immaculate conception. Rabnett (1982, p. 88) notes that such
first concordance of the entire English Bible. Marbeck, a church
indexes were prompted by the need for quick reference to proof
musician, was sentenced to the stake for heresy in 1544. Bishop
texts during disputations (religious debates).
Stephen Gardiner, who was fond of his music, interceded for
Most amazing, however, is a manuscript Hebrew bibliography
him. Marbeck's life was spared ...." (Danker 1993, p. 12).
from the Orientalia collection of the Barberini popes (Barb. Or.
Some suggest that Marbeck's index was objectionable because
no. 53 1626) listing the passages in each work that are to be
(a) it detracted from the authority of Christian clergy by making
The Indexer Vol. 21 No. 3 April 1999
115
Indexes and religion: reflections on research in the history of indexes
theological information directly accessible to the laity, or (b) that
in early and medieval Christianity priests were to serve as
intermediaries
between
the
individual
and
God,
and
as
interpreters of religious truths. Fr. Jean-Pierre Ruiz (a professor
of theology at St. John's University) informed me, however, that
it is not true that Christianity discourages individual study of the
Bible. In his view, the clergy objected to Marbeck's concordance
"for fear that Divine Revelation would be reduced to human
proportions and that the canonical shape of the Bible would be
challenged" (personal communication). The idea that indexes to
sacred works detract from their holiness is fascinating.
In modern Judaism, one encounters negative attitudes to
reference works,
including indexes,
in part because they
minimize the effort required to master religious texts. A key
concern is that laypersons may use indexes to decide matters of
Jewish law by themselves, as opposed to asking a rabbi, who will
take their personal situation into account before rendering a
decision (Weinberg 1990).
Conclusions and implications for further
research
The primary thesis of this essay is that it is no accident that
citation indexes predominated in early Hebraica, i.e., Jewish
texts,
while
subject
indexes
are
common
in
Christian
manuscripts and incunabula. This suggests that it may be
worthwhile to explore the reasons for the frequency or lack of
indexes in other cultures. Why do so many French works of
nonfiction have no indexes? Lack of awareness of indexes, or
lack of knowledge on how to compile indexes, is probably not
the reason. The explanation for the omission of indexes in French
books may be the well-known scholarly attitude that indexes
discourage reading.
English literary scholars are familiar with Jonathan Swift's
negative remarks on indexes, expressed in A tale of a tub, first
published in 1704. In this work (p. 145 in the 1958 edition) the
metaphor for the use of indexes is learning through the "back
door".
Swift was a minister [although this
fact is
well
documented, I learned it from Prof. Michael Gelber of St. John's
University], and A tale of a tub is described as a satire on "the
Indexes and memory
numerous and gross conceptions in religion and learning" (Swift
Indexes and related reference works serve as auxiliary memory.
made from a religious perspective (Janet Shuter, personal
It is interesting to contrast these with alphabetically arranged
texts (acrostics), which have a mnemonic (memory-aiding)
function (Witty 1973, p. 195), i.e., helping people memorize a
text. Boorstin (1983, p. 485) observes that when books had
1909, pp. x-xi), but his observation on indexes may not have been
communication). In checking the references to Swift in The
Indexer, which were kindly provided to me by Hazel Bell
(former editor), I found that Vickers (1990) quoted Swift's
statement: "consulting Indexes ... is to read Books Hebraically
indexes, "the only essential feat of Memory was to remember
and begin where others usually end" (1721, p. 14; emphasis in
the order of the alphabet".
original) — an observation that is surely relevant to this paper.
Indexes appear when memorization declines. Wellisch (1985,
p. 70) has pointed out that medieval Jews did not need
concordances, because they "knew the Scriptures by heart".
Latin concordances preceded Hebrew ones by two centuries.
[Many people do not realize that the original language of the Old
Testament is Hebrew; translations of the Bible to any other
language are subject to revision, rendering concordances to older
editions of the translated text obsolete. Hebrew Biblical
concordances can be organized in different ways, but the text
from which the entries are derived remains stable.]
The Talmud, known as the Oral Law of Judaism, was also once
memorized; it was written down in recognition of "the decline
of the generations". The language of the Talmud is Aramaic, the
vernacular of the Jews from approximately 300 B.C.E. to 1100
C.E. (Kutscher 1972). The earliest dictionary of the Talmud was
compiled at a time when Aramaic was being displaced by other
languages in the Jewish community. The fact that the language
of definition
of Saadia
Gaon's
Hebrew
dictionary
was
Judeo-Arabic suggests that lexicons may be compiled when the
use
or
knowledge
of
a
language
diminishes.
Brisman
(forthcoming) points out that Saadia's Egron was designed to
assist Hebrew poets, because the language was no longer spoken.
While Swift felt that important books should be read and not
indexed, the obverse of this position has been encountered as
well. James Anderson often states at conferences that not every
document should receive human indexing; the vast majority of
documents have little value and merit only automatic indexing.
(This opinion is recorded in his forthcoming book.) Works that
have been cited and/or given
favorable reviews deserve
time-consuming human analysis, in Anderson's view. When we
examine the earliest indexes, we find that works considered
important enough
to receive careful
human analysis
are
primarily in the domain of religion.
Because the earliest indexes were associated with Judaism and
Christianity, articles describing them are often found not in
information science periodicals, but rather in journals in the field
of religion. The footnotes to Rouse and Rouse point to many
descriptions of early concordances and indexes that were
published in theological journals, often in foreign languages.
With respect to the earliest Hebrew citation indexes, I have
already discussed the lack of links between the Hebrew literature
describing these tools and the largely English literature of
information science (Weinberg 1997, p. 328).
Works in non-Roman scripts that are closely associated with
the world's major religions have been excluded from several
Brisman suggests that Jews did not need [alphabetical subject]
studies of the history of books and indexing. Taylor's history of
"indexes because the Mishnah and the Talmuds were already
indexes (1966) admits to this in the preface (p. 7). Katz's recent
arranged by subject" (letter of Jan. 10, 1999). The knowledge
(1998) history of reference books has no index entries for
required to use these classified arrangements, as well as the
Hebrew or Judaism, although Jews and Jewish reference books
canonical sequence of Biblical citation indexes, and early
are alluded to in several endnotes (p. 16nlO, p. 127n78). In
Hebrew dictionaries arranged by the number of letters in a root
another of his unindexed endnotes (p. 16nl 1) Katz admits that
is the subject of a separate paper that I am scheduled to present
Arabic sources are excluded from the study. Kilgour's (1998)
at the Association of Jewish Libraries Convention in June 1999.
history of the book has substantial chapters on the role of Islam
116
The Indexer Vol. 21 No. 3 April 1999
Indexes and religion: reflections on research in the history of indexes
and Christendom but no discussion of Judaism or Hebrew. In the
dictionaries and indexes was an expression of thanks to God at
Golden Age of Spain, there was close cooperation between
the conclusion of the work — sometimes even at the end of each
Islamic and Jewish philologists, and innovations in dictionary
letter of the alphabet, e.g., in the dictionary of Jonah ibn Janah,
design can no doubt be discovered from the study of early Arabic
entitled Sefer ha-Shorashim, copied in 1294 (Vatican catalog no.
Urb. 54). Contemporary indexers no doubt frequently say
and Hebrew reference works.
Few Western scholars are fluent in Latin, Greek, Hebrew, and
"Thank God it's over" or "Hallelujah" when they complete an
Arabic, as well as knowledgeable regarding indexing. This
indexing project, but probably without the religious fervor of
explains the conflicting statements about the earliest indexes.
their medieval counterparts!
Serendipity and "footnote chasing" were major factors in
documenting this paper. The Rouses' book (1979) came to my
attention through a reference in Kilgour (1998), which was on
display at my university library. I did a sequential search of the
unindexed bibliographic references in the endnotes to Katz
(1998), also a new book on display, but did not find Rouse cited.
(Wellisch [1994a, p. 12] cites a 1976 article by Rouse, but not
the book.) The sole topical subject heading assigned to the
Rouses' book by the National Library of Canada (in its
Cataloging in Publication Data on the verso of the title page) is
"Preaching—History—Middle Ages, 600-1500"; thus a search
on "Indexes — History" would not retrieve it.
Several of the cited works on the history of indexing and
reference books are poorly indexed. Missing index entries from
the Katz book have been noted. The "General Index" to the
Rouses' book lacks entries for indexes, personal indexes,
running headlines, and cross references, although all these
concepts are discussed by the authors in these terms. The index
to Kilgour's work lacks entries for religion, tables of contents,
Acknowledgments
My trip to the Vatican Library was supported by a grant from the
Eugene Garfield Foundation, and travel to the National Library
of Canada was funded by the H. W. Wilson Foundation.
Research support from St. John's University is also gratefully
acknowledged, including the contributions of my Research
Assistant, Gayle Payne. The excellent collections of the St.
John's University Libraries enhanced the documentation of this
paper. The cooperation of rare book and manuscript librarians at
the Vatican, the National Library of Canada, and the Jewish
Theological Seminary of America is sincerely appreciated. Dr.
Hans Wellisch read a draft of the article and suggested important
emendations. Dr. Shimeon Brisman, an expert on Judaica
bibliography, also read the draft and kindly shared with me
portions of the manuscript of his forthcoming history of Hebrew
dictionaries. Thanks are due as well to all the individuals cited
for personal communications.
subject indexes, and arabic numerals, all of which are discussed
(on p. 81). McKerrow's classic Introduction to bibliography
(1928) also has no index entries for indexes and tables of
contents, despite their relevance to the physical production of a
References
Amram, David Werner (1963). The makers of Hebrew books in Italy: Being
chapters in the history of the Hebrew printing press. London: The Holland
book, as described in the text; nor does he post all discussions of
these terms to the Yielding preliminaries (pp. 26, 188,190,217).
The catalog of Vatican Hebrew manuscripts (Israel 1968) has
only broad categories (bibliography, religion) in its subject
index,
rendering
it difficult
to
identify
citation
Press.
Anderson, James D. (forthcoming). Information retrieval database design.
Section 2.7.4: Growing role of automatic analysis and indexing.
[ANSI] American National Standards Institute (1984). American national
standard for library and information sciences and related publishing
indexes,
practices — Basic criteria fur indexes. New York: ANSI. (Z39.4-1984).
censorship bibliographies, etc. Time-consuming sequential
Association of College and Research Libraries. Rare Books and Manuscripts
Section. Standards Committee (1986). Printing and publishing evidence:
searches thus had to be done to document the history of indexes,
Thesauri for use in rare book and special collections cataloguing. Chicago:
which are, ironically, designed to obviate such searches.
This paper describes research done in three countries. The
American Library Association.
Bacher, W. (1912). "Dictionaries, Hebrew." The Jewish encyclopedia vol. 4,
Rouses' book includes a lengthy catalog of manuscripts (pp.
311-407) arranged by the city of the repository; this indicates the
wide dispersion of early reference works and indexes. The
Rouses did not travel to all the repositories they listed; in some
cases they checked microfilms (p. 311). This is not an ideal
medium for the study of indexes, as in some cases relevant
579-85. First published in 1901.
Barb. Or. 53 (1626). Index vanitatum multas expurgundasa libris Hebreorum
... I Manuscript, Vatican Library. In Hebrew.)
Ben-Sasson, Haim Hillel (1972). Maimonidean controversy. Encyclopaedia
Judaica vol. 11, cols. 745-754.
Bibliotcca Apostolica Vaticana (1756). Bibliothecae Apostolicae Vaticanae
codicum manuscriptorum catalogus. In tres partes distributus...
features such as foliation were written in after the creation of the
Stephanus Evodius Assemanus ... et Joseph Siinonius Asseinanus. Partis
original manuscript, and this is hard to detect on a microfilm.
primae tamus primus complectens codices Ebraicos et Samaritanos.
Romae: Ex Typographia Linguarum Orientalium Angeli Rotilii.
The position of an index (front or back of a codex) may have
MDCCLV1. Caption title: Codices Ebraici manuscripti Palatino-Vaticani.
been changed in the binding process, and only examination of
the original artifact allows one to determine this.
Reprint: Paris: Librairie Orientate et Amcricaine, 1926.
—
Romae: Typis Vaticanis, MDCCCXXXI. Hebrew Mss.: second section,
The scatter of the relevant documents combined with the fact
that standard catalog descriptions for rare books do not note the
inclusion of indexes complicates research on the history of this
pp. 83-93.
—
feature. It is clear, however, that citation and subject indexes
were invented at least 800 years ago, and alphabetical word lists
many centuries before that — all for the glory of God.
A common feature of early Hebrew and Latin (Christian)
The Indexer Vol. 21 No. 3 April 1999
(1956). Bybliothecae Apostolicae Vaticanae Codices Manv Scripti ...
Codices Vuticani Hebraici. Codices 1-115. Recensvit Humbert us Cassuto.
access tool and renders it difficult to make definitive statements
on the earliest appearance of a certain type of index or indexing
(1831). Catalogus codicum Bibliothecae Vaticanae, edente Angelo Maio.
|Rome): In Bybliotheca Vaticana, MCMLVI. (First fascicle dated 1936.)
—
(1997). Bibliothecae Apostolicae Vaticanae Incunabula, ed. William J.
Sheehan. Cittadel Vaticano. Vol. II. D-O. Hebraica, pp. 651-67.
Bloch, Joshua (1976). Hebrew printing in Naples. In Hebrew printing and
bibliography, ed. Charles Berlin. New York: The New York Public Library
and Ktav, 113-38. Reprinted from Bulletin of The New York Public Library
46 (6). June 1942,489-514.
117
Indexes and religion: reflections on research in the history of indexes
Boorstin, Daniel J. (1983). The discoverers. New York: Random House.
Boudinhon, Auguste (1910). Index librorum prohibitorum. Encyclopaedia
Britannica. 11th ed., 374-5.
Brisman, Shimeon (forthcoming). A history and guide to Judaic dictionaries
and concordances. Hoboken, NJ: Ktav, 1999? (His Jewish Research
in some incunabula. Cataloging & Classification Quarterly 2 (3/4), 87-102.
Richardson, E. C. (1939). Dictionaries. In The international standard Bible
encyclopaedia, ed. James Orr. Vol. II, 843-8. (Originally published 1929.)
Rouse, Richard H. and Mary A. Rouse (1979). Preachers, florilegia and
sermons: Studies on the 'Manipulusflorum' ofTlwmas of Ireland.
Literature, vol. HI, part I).
Busa, Roberto (1971). Concordances. Encyclopedia of Library and Information
Science 5.592-604.
Toronto: Pontifical Institute of Mediaeval Studies. {Studies and texts, 47).
Sheehan, William (1990). Hebrew incunabula in the collections of the Vatican
Library. Miscellanea Bibliothecae Apostolicae Vaticanae IV; Studi e Testi,
Danker, Frederick W. (1993). Concordances. Chapter one of his Multipurpose
tools for Bible study. Rev. ed. Minneapolis: Fortress Press, 1-21.
Domenico Irosolimitano (1596). Sefer ha-zikuk. Mantua. (Ms. In Hebrew.)
no. 338, [365]-83.
Sonne, Isaiah (1976). Expurgation of Hebrew books: The work of Jewish
scholars: A contribution to the history of the censorship of Hebrew books
Original in Vatican. Copy in the Jewish Theological Seminary (ins. 2519),
in Italy in the sixteenth century. In Hebrew printing and bibliography, ed.
written in 1890.
Charles Berlin. New York: The New York Public Library and Ktav,
Dunkin, Paul Shaner (1973). How to catalog a rare book. 2nd ed. Chicago:
199-241. Originally published in Bulletin of The New York Public Library
American Library Association.
46 (12) December 1942,975-1015; reprinted as a separate with additions,
Fishman, Judah Leib Hacohen (1935). [Index of liomiletics by Maimonides:
from a unique manuscript in the world.] In J.L. Fishman (Ed.) Rabenu
Mosheh ben Maimun: Hayav, sefarav. pe'ulotav ve-de'otav. Part One —
1943.
Swift, Jonathan (1704). A tale of a tub. Editions cited: (1) A tale of a tub and
other satires, intro. by Lewis Melville. London: Dent, 1909, reprinted
special section. Jerusalem: ha-Merkaz ha-'olaini shel ha-Mizrahi. (In
1963. (Everyman's Library, no. 347). (2) A tale of a tub, to which is added
Hebrew.)
The battle of the books..., ed. by A. C. Guthkelch and D. Nichol Smith.
Gaster, Moses (1929). Eliezer Crescas and his Beth Zcbul: The Bible references
2nd ed. Oxford: Clarendon Press, 1958. (The passage is quoted in Tfte
in Talmud and Midrash. Hebrew Union College Annual 6,277-95.
Indexer 12 (4)0ct. 1981, 214, where the date given is 1701. Everyman's
Grundmann, Walter (1968). The Teacher of Righteousness of Qumran and the
Library edition [1909, p. x] reports that the book was written in 1697 and
question of justification by faith in the theology of the Apostle Paul. In
Paul and Qumran: Studies in New Testament exegesis, ed. by Jerome
Murphy-O'Connor. Chicago: Priory Press, 85-114.
Habermann, Abraham Meir (1972). Soncino. Encyclopaedia Judaica vol. 15,
cols. 140-42.
Intner, Sheila S. (1984). Censorship in indexing. The Indexer 14(2) Oct. 1984,
circulated in manuscript.)
—
University of Pennsylvania Press.
Tishby, P. (1983). Hebrew incunabula. Kiryat Sefer: Bibliographical Quarterly
of the Jewish National and University Library, Jerusalem. 58 (4), 808-57.
105-8.
(First installment, which includes a list of the Vatican's holdings. In
Israel (1968). Misrad ha-Hinukh veha-Tarbut. ha-Makhon le-khitve ha-yad
ha-*lvriyim. Reslu'mat tatslume kitve ha-yad lia- 'riyim ba-Makhon. Helek
3: Kitve ha-yad shebe-sifriyat ha-Vatikan. Added title page: List of
Hebrew.)
Vat. Ebr. 263. [Biblical citation index to the Hebrew translation by Samuel Ibn
Tibbon of Maimonides' Guide for the perplexed, originally written in
photocopies in the Institute: Hebrew manuscripts in the Vatican, by N.
Arabic] Text copied in 1205. Citation index — in a different handwriting
Allony and D. S. Loewinger. Jerusalem: Rubin Mass. (In Hebrew.)
Jewish Theological Seminary. Library (1990). [Catalog of Hebrew
(1721). /I letter of advice to a young poet. London: W. Boreham.
Taylor, Archer (1966). General subject-indexes since 1548. Philadelphia:
— is anonymous, untitled, and undated.
manuscripts. Compiled by Alexander Tobias.] New York. Vol. 2: Science
Vickers, John A. (1990). Swift on indexing. The Indexer 17 (1) April 1990,32.
mss.
Weinberg, Bella Hass (1990). Negative attitudes to Judaica reference works
[editorial]. Judaica Librarianship 5(1), 3.
Kantor, Mattis (1989). The Jewish time line encyclopedia. Northvale, NJ: Jason
Aronson.
—
Kalz, Bill (1998). Cuneiform to computer: A history of reference sources.
Lanham, MD: Scarecrow Press.
Society for Information Science 48(4) April 1997, 318-30.
Weinreich, Max (1980). History of the Yiddish language. Translated by Shlomo
Kilgour, Frederick G. (1998). The evolution of the book. New York: Oxford
University Press.
Knight, G. Norman (1968). Book indexing in Great Britain: A brief history. The
Indexer6 (I) Spring 1968, 14-18.
Kutscher, E.Y. (1972) Aramaic. Encyclopaedia Judaica 3, cols. 259-87.
Jerusalem: Keter.
Leclercq, H. (1948). Concordance. In Dictionnaire d'archeologie chretienne el
de liturgie, ed. by Fernand Cabrol and Henri Leclercq. Paris: Librairie
Letouzey et And, vol. 3, part 2, cols. 2493-4. [Busa (1971) gives the date as
1914.] (In French.)
Lyons, Ponsonby A. and Benjamin E. Smith (1910). Dictionary. Encyclopaedia
Britannica vol. VIII, 186-200.
MacFadyen, Dugald (1910). Concordance. Encyclopaedia Britannica vol. VI,
831-2.
McKerrow, Ronald B. (1928). An introduction to bibliography for literary
students. 2nd impression with corrections. Oxford: Clarendon Press.
Oded, Bustanay (1972). David: In the Bible. Encyclopaedia Judaica vol. 5,
cols. 1318-26.
Offenberg, A.K. (1990). Hebrew incunabula in public collections: A first
(1997). The earliest Hebrew citation indexes. Journal of the American
Noble. Chicago: University of Chicago Press.
Wellisch, Hans H. (1978). Early multilingual and multiscript indexes in herbals.
The Indexer 11 (2) Oct. 1978, 81-102.
—
(1985). Hebrew Bible concordances, with a biographical study of Solomon
Mandelkern. Jewish Book Annual 43 (1985-86), 56-91.
— (1986). The oldest printed indexes. The Indexer 15 (2) Oct. 1986,73-82.
—
(1994a). Incunabula indexes. The Indexer 19 (1) April 1994,3-12.
—
(1994b). Indexing. In Encyclopedia of library history. New York: Garland,
268-70.
—
(1996). Index: The word, its history and meanings. In his Indexing from
A to Z, 2nd ed. New York: H. W. Wilson, 1995, cI996. Reprinted from The
Indexer 13 (3) April 1983, 147-51.
Witty, Francis J. (1965). Early indexing techniques: A study of several book
indexes of the fourteenth, fifteenth, and early sixteenth centuries. 77»e
Library Quarterly 35 (3) July 1965, 141-8.
— (1973). The beginnings of indexing and abstracting: Some notes towards a
history of indexing and abstracting in antiquity and the Middle Ages. The
Indexer 8 (4) Oct. 1973. 193-8.
international census. Nieuwkoop: DeGraaf.
Phillimore, Walter George Frank (1910). Canon law. Encyclopaedia Britannica
vol. V, 193-203.
Porges. N. (1903). Der hebraische Index expurgatorius, Sefer ha-zikuk. In:
Festschrift zum siebzigsten Geburtstage A. Berliner's. Hrsg. von A.
Freimann und M. Hildesheimer. Frankfurt a. M: J. Kauffmann, [273-95].
(In Gentian.)
Rabnett, Mark (1982). The first printed indexes: A study of indexing techniques
118
Bella Hass Weinberg is Professor in the Division of Library and
Information Science, St. John's University, Jamaica, New York.
She is a past president of the American Society oflndexers and
sen>es as editor of Judaica Librarianship, the journal of the
Association of Jewish Libraries.
The Indexer Vol. 21 No. 3 April 1999