A Brief History of the Index

On the rare occasions I get to peruse paper and ink books in a brick and mortar bookstore, after a brief flirtation with the cover and blurb, I will scan the table of contents, then gently – for the book is new, the clean pages crisp – thumb through the final leaves until I locate the index, where, if I am familiar with the subject matter, I expect to find, at the very least, the usual suspects. Their absence might well be symptomatic of more profound flaws between the covers; for example, a book titled The History of Psychology, whose index fails to reference, say, Freud, Jung and Mary Whiton Calkins, could safely be passed over in favor of something better. Once a book is finished, its index remains invaluable in tracking down partially or half-remembered facts and phrases. In many digital books, the index has been replaced by search but, whereas a full-text search of a digital book often requires one to know precisely what one is looking for, in an alphabetical subject index, a half-remembered name, or even the first letter of an otherwise forgotten word, is usually sufficient. It is for this reason that the designers and editors of digital books should not be in a hurry to do away with the index, for it remains an indispensable, simple and intuitive means of making texts, most especially lengthy texts, more accessible.

They told me, Heraclitus, they told me you were dead; /
They brought me bitter news to hear and bitter tears to shed. /
I wept as I remembered how often you and I /
Had tired the sun with talking and sent him down the sky. /
But now that thou art lying, my dear old Carian guest, /
A handful of grey ashes, long, long ago at rest, /
Still are thy pleasant voices, thy nightingales, awake; /
For Death he taketh all away, but them he cannot take.

Callimachus, Epigrams II (A.P. vii. 80)

Callimachus & the Library of Alexandria
The use of indexes dates back to Greco-Roman antiquity, with seeds of their origins found in the separate penned lists or summaries of books and chapter titles. And, although quite a number have survived, they resemble more an ordered table of contents, rather than an alphabetical subject index. To trace the origins of the alphabetical subject index we must first trace the origins of alphabetization. For that we must turn to Alexandria in Egypt of the third century BC. Callimachus of Cyrene (c. 305–240 BC), arguably the most important poet of the Hellenistic age and resident of the museum of Alexandria, compiled his Pinakes for the great library of Alexandria – generally considered the first library catalogue, comprising 120 scrolls. The catalogue has survived only in a handful of fragments but they reveal that it was divided into genres and categories alphabetized by author. Some fragments of Greek and Roman papyri bear stichometric notation, or references to the total number of lines – a practice that dates back to the fifth century BC. This was useful in determining the authenticity or accuracy of a copy and for estimating the cost of copying, when scribes were paid by the line. (F. J. Witty, 1958, p. 134)

The libraries of Alexandria were home to half a million books or more. The two libraries were destroyed by war in the first century BC and by religious zealotry in the fourth century AD.

Scrolls were not ideally suited to indexing and it was not until the introduction of the codex, with its intrinsic sections, or pages, that the potential of indexes could be fully explored. Surprisingly, the codex’s greatest asset, the discrete page, was the feature that early indexers failed to exploit. Only later, with the introduction of sequential page numbering, would the index really come into its own. The earliest alphabetical subject index dates to the sixth century, an anonymous Apophthegmata, or collection of quotes and aphorisms, from various Greek fathers. In the eighth century another alphabetical subject index of sorts is to be found in the Byzantine manuscript, Sacra parallela, a florilegium, or collection of quotes from the Bible and early Church Fathers, compiled by St. John of Damascus.

“On the whole indexing is a question of common sense, and consistency is the only inflexible rule.”
Cambridge University Press Indexing Guidelines

Another form of index, popularized later in the Middle Ages, was the concordance. The first concordance, for the Bible, was compiled under the direction of the Dominican friar, Hugh of Saint-Cher, and completed in 1230. The Biblical concordance, an alphabetical index of words, proved especially useful to medieval theologians and Biblical scholars. At about the same time, the Bible was divided into chapters, popularized by the work of the English Cardinal and Archbishop of Canterbury, Stephen Langton. It was the Bible’s division into smaller chunks, or chapters, that made indexing more workable.

Aside: Strong’s Concordance (1890) is an exhaustive concordance of the Bible, indexing every single word of the King James translation. It does not require page references as locators, using instead the Bible’s books, chapters and verses.

However, the alphabetical subject index is seldom found in medieval manuscripts. Not until the fourteenth century, when, incidentally, the Vatican began to use alphabetical indexing in organizing its archives, do we begin to witness a tentative rise in the use of indexes. Of the indexes found in fourteenth-century books, some appear to be in the hand of a second scribe – or more likely the unpracticed hand of the book’s owner. But, all the same, the increased occurrence of indexes suggests that their value was at last beginning to be appreciated.

It would be wrong to conclude that indexes do not appear with regularity in manuscript books because manuscript books seldom have numbered folia or pages. Page numbers, although convenient locators, are not the only ones. Numerous popular works of the Middle Ages, besides the Bible, were divided into books, chapters and, sometimes, numbered sections, or even numbered paragraphs – the latter element, one might argue, is superior in that it is more granular, more precise a locator, than page numbers. In addition to the reasons noted above, the Greeks and Romans used stichometric notation as an aid to citation, writing ascending letters of the alphabet in the margins, often at one-hundred-line intervals, with each line written in prose text set to a standard 15-, later 16-syllable line. It was not unheard of for unscrupulous booksellers to exaggerate the number of lines in a book, so as to inflate the price. (For a concise introduction to the practice and history of stichoemetry, see Thompson, 1912, pp. 67–69)

An example of stichometry in the Clarke Codex of Plato’s Dialogues, copied in 895 AD. Lambda, the eleventh letter of the Greek alphabet, indicates line 1,100. The line-numbering is undoubtedly copied from a much earlier manuscript. Bodleian Library, Oxford University.

In the third century AD, Greek biographer, Diogenes Laërtius, in his Lives of the Eminent Philosophers, cites several passages from the founder of the Stoic school of philosophy, Zeno (c. 334–c. 262 BC), referencing actual line numbers:

“… at line 200 [Zeno] prohibits the building of temples, law-courts and gymnasia in cities.”
– Diogenes Laërtius, VII, 32–5 (p. 145)*

“For in his [Chrysippus’] work On the ancient Natural Philosophers at line 600 or thereabouts he interprets the story of Hera and Zeus coarsely…. In the third book of his treatise On Justice, at about line 1000, he permits eating of the corpses of the dead.”
– Diogenes Laërtius, VII, 187–9 (p. 297)*

But this practice soon fell out of favor. The decline in numbering lines, or stichometry, coincides with the rise of the codex in the third and fourth centuries AD. By about the year 300 AD, the codex had overtaken the scroll-book in popularity. (See Roberts & Skeat, p. 37.)

“The man who publishes a book without an index ought to be damned 10 miles beyond hell, where the Devil himself cannot get for stinging nettles.”
John Baynes (1758–87) [source: Futility Closet]

St. Augustine’s Index
The very first printed book to include an index was put out by the press of Fust and Schoeffer in Mainz. Their 1467 edition of Saint Augustine’s treatise for preachers, De arte praedicandi (‘The art of preaching’), in most other respects, is a typical product of their press. It is carefully typeset in Schoeffer’s splendid, round gothic, Durandus type, in deep, justified, evenly colored text blocks surrounded by generous margins and accompanied by tasteful and conservative rubrication. The index is mentioned in the preface as,

“A most extensive alphabetical index which has been compiled with great care…. The index and figures … are indeed alone worth the whole price, because they make it much easier to use.”

The preface also includes a note on how to use the index. The indexer, chose letters over numerals as his locators: when the alphabet (A–Z) is exhausted (with the addition of a Tironian ‘ampersand’ and the backwards capital C, an abbreviation of the Latin prefix, con), the index turns next to two-letter combinations, beginning with AB–AↃ, BC-CↃ, CD–CL, for a total of 80 divisions.

1467 first ever printed index
First page of the first printed subject index in St. Augustine’s treatise for preachers, De arte praedicandi, printed by Fust and Schoeffer in Mainz, 1467
First text page of Fust and Schoeffer’s 1467 edition of St. Augustines’s, De arte praedicandi. Note the guide letters (a, b, c, d, e) printed in the inside margin, used as locators in the index.

At about the same time, an almost identical edition of this work appeared from the press of Johannes Mentelin in Strasbourg. In some catalogues the Mentelin edition is dated earlier than that of Fust and Schoeffer’s, although F. W. Householder makes a very compelling case for the priority of the Fust and Schoeffer edition. We might, then, consider Mentelin’s, the first known pirated edition in Western typographic history. Householder conjectures that perhaps Mentelin’s edition was ‘composed of proof-sheets surreptitiously taken by some employee of Fust.’ (p. 45)

First page of index from Speculum aureum, printed at Mainz by Peter Schoeffer, in 1474.
First page of index from Schoeffer’s 1474 edition of Speculum aureum, printed in Mainz.

Of particular interest in early printed indexes are the solutions to linking index entries to their sources within the text. Peter Schoeffer, now printing alone, after the death of his partner Fust in 1466, printed the first edition of Henricus de Herpf’s Speculum aureum. This substantial book of 811 pages was printed in 1474, some seven years after the St. Augustine and it demonstrates that Schoeffer was still grappling with how best to present the index. The book opens with an index of 22 pages, not comprising a list of words but instead sentences and phrases. In the absence of page numbers, the index makes good use of the book’s natural divisions: first the index references one of the book’s major divisions, one of ten commandments, each introduced by a large, rubricated initial of six lines; next, the index references one of 213 sermons, introduced with a smaller, usually, four-line initial. Each index entry concludes with reference to a guide letter printed in the inner or outer margin.

Second page of index from Shoeffer’s 1474 Speculum aureum. Each entry adheres to a reference system of Commandment > Sermon > Marginal Catch letter.

So, for example, the first entry under B, reads: “baptismi forma qualis est. i. sermon xvi. c.”, with, of course, the concluding “i. ser. xvi. c.” acting as the locator, referencing the first Commandment, Sermon 16, guide letter, C (in the margin). You can test the index for yourself with the digitized copy of this incunable hosted by ULB Darmstadt.

Detail from a page of Speculum aureum, printed at Mainz by Peter Schoeffer, in 1474. Note the catch letter, A, in the inner margin used as a locator referenced in the index. In this copy, the rubricator has numbered the Commandments in the top margin, a very useful addition in the absence of printed running heads, that were not introduced until 1493. If you wished to reference this page in the index, you’d simply reference Commandment 2, Sermon 1 (Sermo primus), catch letter A (ii. ser. i A.)

Clarification: the difference between foliation and pagination. Above:foliation; below: pagination:

A Convenient Marriage
One of the greatest drawbacks of early indexes is the absence of page numbering. Printed page numbers (or foliation) were first introduced in 1472 by Arnold Ther Hoernen in Cologne, in the first edition of Werner Rolewinck’s Fasciculus temporum (literally, ‘bundles of time’). However, printed page numbering was not broadly adopted until the following century. One of the first books to combine an index with foliation, or page numbers, was another edition of the hugely popular Fasciculus temporum. The 1481 edition, printed by Erhard Ratdolt in Venice, opens with a three-column, fourteen-page alphabetical subject index with references to printed page foliation on the recto side, centered in the top margins.

First page of the index from Erhard Ratdolt’s 1481 edition of Werner Rolewinck’s Fasciculus temporum. Image courtesy of Biblioteca de Andalucía. In this particular copy, the rubricator, although not particularly accomplished, is not lacking in enthusiasm.

Although the subject index appears relatively early in the history of printing, it is seldom found in fifteenth-century books, other than in reissues or copies of earlier printed indexes. The widespread use of indexes in books did not occur until the subsequent century. Moreover, not until much later was the term index used to describe them – instead, tabula or tabula per alphabetum was commonly used. It is worth noting that we should not be too dismissive of earlier locators, like sections, paragraphs, or even line numbers, as inferior to their successors, foliation and pagination. A significant advantage of such indexing systems is their portability: Whereas an index that relies on pagination cannot be used in another edition with different pagination, an index referencing, say, numbered paragraphs, functions despite changes to the document’s page breaks. And a drawback to the index linked to page numbers is that it cannot be compiled until the entire book is typeset or printed. (See H. H. Wellisch, p. 76)

Known from antiquity and used sporadically throughout the Middle Ages, it was not until the Renaissance and its appearance in print that the index was made a permanent feature of the book. After its first tentative appearance in print in 1467, it reappeared sporadically until its more widespread adoption in the sixteenth century. Today, the index remains an indispensable tool for readers, students and scholars alike.

References & Further Reading

* Quotations and page references from Diogenes Laertius: Lives of Eminent Philosophers, Loeb Classical Library edition, vol. 2, (Books 6–10), translated by R. D. Hicks

B. Acosta-Hughes, et al. (Eds.), Brill’s Companion to Callimachus, 2011

R. Blum, Kallimachos: The Alexandrian Library and the Origins of Bibliography, 1991

L. W. Daly, Contributions to a History of Alphabetization in Antiquity and the Middle Ages, 1967

S. Hornblower (Ed.), The Oxford Classical Dictionary, 2012

F. W. Householder, ‘The First Pirate’, The Library, vol. s4-24: 1–2, 1943, 30–46

F. G. Lang, ‘Stichometry’, in: Religion Past and Present, 2011

L. Morgan , Patterns of Redemption in Virgil’s Georgics, 1999

M. Rabnett, ‘The First Printed Indexes: A Study of Indexing Techniques in Some Incunabula’, Cataloging & Classification Quarterly, 2:3–4, 1982, 87–102

C. H. Roberts & T. C. Skeat, The Birth of the Codex, 1987

E. M. Thompson, An introduction to Greek and Latin Palaeography, 1912

H. H. Wellisch, ‘The Oldest Printed Indexes’, The Indexer, 15:2, 1986, 73–82

F. J. Witty, ‘Early Indexing Techniques: A Study of Several Book Indexes of the Fourteenth, Fifteenth, and Early Sixteenth Centuries’, The Library Quarterly, 35, 1965, 141–8

F. J. Witty, ‘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, 1973, 193–8

F. J. Witty, ‘The Pínakes of Callimachus’, The Library Quarterly: Information, Community, Policy, 28:2, 1958, 132–6

The International Journal of Indexing


Header Image: Virgil reading his Aeneid to Emperor Augustus and his sister Octavia; Jean-Joseph Taillasson.

Header Typeface: Quarto by H&Co.

And don’t forget to sign up to our new monthly newsletter, The Month in Type – it’s free and fantastic.