it has entered the digital era as the key technology underpinning our online reading. The very first webpage, after all, was a subject index.

when you do a Google search, you aren’t actually searching the web. You’re searching Google’s index of the web.’

need for the contents of books to be divisible, discrete, extractable units of knowledge.

the book index remains primarily the work of flesh-and-blood indexers, professionals whose job is to mediate between author and audience.

The name suggests a spatial relationship, a map of sorts: something here will point you to–will indicate–something there. The map need not exist in the world; it is enough for it to exist in our minds.

a truly random-access technology, and as such it relies on a form of the book that can be opened with as much ease in the middle, or at the end, as at the beginning. The codex is the medium in which the index first makes sense.

the chief mechanism of the index is arbitrariness. Its principal innovation is in severing the relationship between the structure of the work and the structure of the table. The ordering of an index is reader-oriented, rather than text-oriented: if you know what you’re looking for, the letters of the alphabet provide a universal, text-independent system in which to look it up.

the alphabetical table that breaks down a book into its constituents, its characters, its subjects, or even its individual words; a piece of technology–an add-on–designed to speed up a certain mode of reading, what academics have taken to calling ‘extract reading’,

Many of the students now were reading on digital devices–on Kindles, on iPads, sometimes on their phones–devices which did not use page numbers, but which came equipped with a search function. Historically, a special type of index, known as a concordance, would present an alphabetical list of every word in a given text–the works of Shakespeare, say, or the Bible–and all the places they appear. In my classroom I began to notice how the power of the concordance had been extended infinitely. Digitization had meant that the ability to search for a particular word or phrase was no longer tied to an individual work; now it was part of the eReader’s software platform.

the ubiquity of the search engine has given rise to a widespread anxiety that search has become a mentality, a mode of reading and learning that is supplanting the old modes, bringing with it a host of cataclysmic ills. It is, we are told, changing our brains, shortening our attention spans and eroding our capacity for memory.

The book index: killing off experimental curiosity since the seventeenth century.

The way we read (the ways that we read, we should say, since everyone, every day, reads in many different modes: novels, newspapers, menus, street signs all require a different type of attention of us) might not be the same as twenty years ago. But neither were the ways we read then the same as those of, say, Virginia Woolf’s generation, or a family in the eighteenth century, or during the first flush of the printing press.

The ideal index anticipates how a book will be read, how it will be used, and quietly, expertly provides a map for these purposes.

what a strange, miraculous thing alphabetical order really is: something we take for granted, but something which appeared, almost out of nowhere, 2,000 years ago; something we use every day, but which a civilization as vast as the Roman Empire could choose to ignore completely in its administrative apparatus.

Wilfully zany?

the index is not a form well suited to narrative. It is hard to break in. Its commitment is not to the author but to the reader, and to the arbitrary order of the alphabet.

An encyclopedic mind needs an encyclopedic index to provide it with structure.

We are about to witness the birth of the index, or rather the births of the index, two versions of the same idea, arising simultaneously, one in Oxford, one in Paris. Taken together, both can tell us something about the index in our present, twenty-first-century moment, the Age of Search. Between them, they set up the axes by which we think about indexing: word versus concept; concordance versus subject index; specific versus universal.

a subjective index, the work of a particular reader, thinking and parsing their reading a certain way. Concepts are slippery things. We make a choice when we say that a text is about something; that, say, the story of Noah’s Ark is about forgiveness, or anger, or rain. By contrast, another type of index–another way of searching–is more straightforward. Its terms are simply the words of the text under analysis: if the text uses a particular word, then that word will be in the index. Unsubjective, this type of index leaves little room for interpretation–a word is either there or it isn’t.

Hugh will be the first to produce a concordance to the Bible,

manicule, that Pythonesque pointing finger,

Looking back on his first encounters with Ambrose, Bishop of Milan, in the late fourth century, Augustine remembers noticing the curious way Ambrose would read: ‘his eyes would scan over the pages and his heart would scrutinize their meaning–yet his voice and tongue remained silent’. 7 This–reading in silence–is not normal, and Augustine wonders what could possess Ambrose to adopt such a practice.

in the schools and universities, where students were being trained for careers in religious or secular administration, new methods of teaching came to dominate. Disputation, the citing of authorities, the reading-out of commentaries (a format with a now-familiar name: the lecture): scholastic learning would favour external demonstration over inner revelation, intellectual agility over endless meditation. University readers would require new tools on the page, new ways of efficiently finding parcels of text–a word, a phrase–amidst the prose block.

Each distinctio is an aide-mémoire, a well-ordered, bite-sized cribsheet on a given theme.

each distinctio sends its user on a series of targeted sorties into the source material

The distinctio-collection, then, is evidence of a type of reading that might be thought of as indexical.

The distinctio too offers, not so much a map of the book, as the mindmap of a moment of creative reading. It is neither methodical nor chronological, but associative, sparked by a single word or concept, firing unpredictably in multiple directions.

the St Jacques Concordance–the word-by-word index to the Latin Bible–will identify over 10,000 terms and list them in alphabetical order. It will begin with the exclamation A, a, a (usually translated as ah! or alas) and end with Zorobabel (or Zerubbabel, a sixth-century governor of Judea). As well as names and exclamations, the concordance will also include the ordinary language of the Bible–common nouns, verbs, adjectives–and for each term it will list every instance of its occurrence, giving the book, chapter and chapter section in which it appears.

An index is a tool of two ordering systems, a conversion table between the alphabetical order of its entries and the sequential order of the pages. The first means that we can skim quickly to the headword we require; the second means that, armed with our locator, we can move easily to the passage it indicates.

The life of Enea Silvio Piccolomini was nothing if not eventful. By the end of it, he would no longer be Enea but Pope Pius II; along the way he would be an envoy to the pope’s ambassador in England, poet laureate to the Holy Roman Emperor in Vienna and father to two children–one in Strasbourg, one in Scotland–though neither would survive into adulthood. Later, as pope, he would persuade the prince of Wallachia, a certain Vlad Dracula–otherwise known as Vlad the Impaler–to go to war with the Ottoman Sultan, something which Vlad initiated with characteristic brutality: by nailing the Sultan’s emissaries’ turbans to their heads. If all this were not enough, in his writing, Piccolomini also serves–at nearly 600 years’ distance–as our earliest surviving witness to something of more world-shattering importance even than the wars and crusades in which he had a starring role.

Although Gutenberg’s invention (or rather series of inventions: the metal letter sorts–tiny, individual, reusable–and the method of casting them; the oily ink that could be slopped across a page of type without running off as pen ink would; the press that would distribute pressure evenly across the page) was initially a trade secret in his home city of Mainz, soon, inevitably, it would spread across Europe. In 1462, Mainz was sacked as part of a bitter conflict that saw many of its citizens flee the region. The refugees included trained printers, and before long printshops sprang up in Ulm, Basel, Venice, Rome. Around 1473, an English merchant named William Caxton was on a visit to Cologne when he saw a press there. This invention, he reasoned, might be just the thing for a problem he had been having with some extracurricular activity. For some years, Caxton had been living in Bruges, where he had become close to the circle of Margaret of York, wife of the Duke of Burgundy. Lately he had translated a popular poem on the legends of Troy into English, and his version had proved popular with the English speakers of Margaret’s court. But the poem was long, and copying it out for everyone who wanted it was a daunting task. ‘My penne is worn,’ wrote Caxton, ‘my hand wery & not stedfast myn eyen dimed with ouermoche lokyng on the whit paper and my corage not so prone and redy to laboure as hit hath ben.’ 7 In setting up a printworks, Caxton would be able to rattle off enough copies to satisfy demand without straining his eyes or spraining his wrist.

Caxton’s Recuyell of the Historyes of Troye would be the first printed book in English.

On the Kindle, the locator of choice is the loc#, a division based on numbered units of 150 bytes. Like the page number, this is a blind splicing, one that pays no attention to what is being said, to the passage of thought, the appropriate pause. The loc# doesn’t care what any of its 150 bytes represent, whether they are letters of the authorial text or mark-up instructions telling the eReader that a passage should be displayed in italics or indented or as a hyperlink. It is a crude algorithm, but it makes for a highly granular system of locators, one which can pinpoint a search term to an area shorter than a tweet.

not a supplement to the novel but a coda, a final dose of wit, playfulness and even satire, beyond the story’s end but still very much part of the same work.

the single-sheet papers represented print in its most demotic, promiscuous form, designed to be passed from one reader to the next,

On the title page of every edition of Poole’s Index the same epigraph appears, given to Poole by his Latin professor back when he was a hard-pressed student at Yale: ‘Qui scit ubi sit scientia habenti est proximus’: ‘He who knows where knowledge may be had is close to having it.’ The

our twenty-first-century Age of Search is, in effect, an age of the automated concordance.

Intellectually, the indexer atomizes the book’s contents, identifying its personalities and tracking them through the course of the work. She sifts the ideas at play and mulls over the best labels for them, whether a concept needs to be ramified or subdivided, or whether two related themes might reasonably be rolled up under one head. A challenge, certainly, an exercise in concentration, in deep reading; but not drudgery by any definition. Physically, however, the process involves preparing and marshalling a paper tide. Sorting, switching between orders–page order to alphabetical order–and recopying.

for 700 years, from St Hugh to Virginia Woolf, a final index was necessarily a second version. The indexer passes through the work producing a series of entries and locators in the order in which they occur. But this then has to be rearranged–rewritten or retyped–into alphabetical order. The intellectual labour is frontloaded in the indexing process; the drudgery comes later.

what about that curious typeface: blocky, sans-serif, and–the giveaway–monospaced? This page hasn’t been typeset: this is output.

Once our indexing information–heads, locators–can be ‘read’ by machines, then it doesn’t much matter whether it is stored on punch cards, magnetic tape or integrated circuits. We have opened up the possibility of reordering without rewriting. The indexer’s job has been distilled to its analytical essence; the drudgery–the shuffling and copying–has been delegated to the machine.

we expect our back-of-book indexes to be more than mere word lists. We expect them to provide context, interpretation, to recognize when the same concept appears in different guises.

machines can only speed up the tasks that can be sped up–sorting, layout, error checking. The work of compiling a subject index is still, principally, a subjective, humanistic one. It is a job of deep reading, of working to understand a text in order to make the most judicious selection of its key elements.

The patient interpretations contained in subject indexes originally compiled for physical books do not translate well when those books are remediated as reflowable electronic editions. Sometimes indexes are unceremoniously dumped in the conversion; sometimes they are retained but useless in a format that eschews page numbers, their spectral locators gazing wanly after features of a terrain that is no longer visible.

Thanks to the recent technological revolution in our reading, the functionality of the concordance has become ubiquitous; the subject index, however–careful, conscientious, expert–has not been well served.

automated indexing offers a much cheaper alternative to hiring the services of a qualified indexer.

The hashtag, he suggested, might serve as an ‘ad hoc verbal guidepost’,

folksonomic–that is, a taxonomy, a classification system, but one whose terms are not determined in advance. Instead they are created, on the fly, by anyone–by folk. A hashtag that hasn’t been used before comes into being instantly at the point the message containing it is sent. What’s more, no higher authority owns or controls it.

When the publicity team for the singer Susan Boyle decided to promote the launch of her 2012 album Standing Ovation under the hashtag #susanalbumparty, there was no registration panel to point out that this sequence of letters could be read in different ways, and no mechanism to separate out supportive messages from Boyle’s fans from amused responses from the wider twittersphere.

the hashtag is a truly demotic form, governed by conventions that are unpredictable, ironic, carnivalesque. In the form of the tagger, our patient, bookish indexer has acquired a younger sibling, one who is capricious, sarcastic and fluent in the ever-shifting inflections of the new media.

to me Alphabet is also a reminder of the gap–the uncanny valley–that has still not been decisively bridged between concordance and subject index, matching letters versus identifying concepts, Hugh’s vision versus Grosseteste’s.

The limitations of unimaginative indexing, of the simple string search, become starkly apparent if one tries to locate the parable of the prodigal son, that famous tale of mercy and forgiveness, using a Bible concordance. The parable does not contain the words forgiveness or mercy, or, for that matter, prodigal.

‘The composition of vast books is a laborious and impoverishing extravagance . . . A better course of procedure is to pretend that these books already exist, and then to offer a resumé.’ Jorge Luis Borges, Ficciones

when J. Horace Round demolishes his medievalist rival in the back pages of Feudal England, he reveals something of his character–his tenaciousness, certainly, but also pedantry, professional frustration, a capacity for unremitting contempt. There is personality in an index; in a concordance, or a search bar, there is not.

the archive of a reading, the record of a particular response to a particular book. From the most elaborate to the merest sketch, every one carried the mark of an indexer: of what they thought was important, the details they expected to return to, a map for a future visit.

The similarities between publishing a book and cycling the Tour de France have so far, I think, gone unremarked. The long procession broken into stages or chapters, some with significant peaks, others flat-out time trials before the teaching term begins; the gruelling, hors catégorie uphill of the last push before submission, followed by the thrilling, breakneck run-in towards publication. I’m surprised more hasn’t been made of this. Most of all, though, both events manage to pull off a slightly shifty apportionment of acclaim: both are team pursuits in which one person seems to take all the credit at the end. It is no exaggeration to say that this book simply wouldn’t have been written without the generosity, support and expertise of a great many people. The following acknowledgements are really just the tip of the ice bath.

If the work you are holding can be distilled to a single piece of advice, let it be this: if you’ve written a non-fiction book, you should hire an SoI indexer. The Society are represented in the back pages here by the estimable Paula Clarke Bain, who manages to combine all the erudition and punctiliousness one would expect from a professional indexer with a wicked sense of humour.