The writer and bibliophile Alberto Manguel maintains that “the subjects or categories into which a library is divided not only change the nature of the books it contains (read or unread), but also in turn are changed by them. […] And yet order of almost any kind has the merit of containing the uncontainable” (2008, 44-45). This is what people have tried to do since writing was invented, some people more obsessively than others perhaps, among them no doubt, scribes, philosophers, science scholars, archivists, librarians, computer scientists. The catalogue form has been one of the tools for organising knowledge and in this essay I will outline its development. For reasons of space, I will confine my survey to the West, highlighting peculiarities and innovations prompted by technological and cultural changes, but also continuities in practices over the centuries. I will also explore some of the opportunities afforded today by new technologies and the challenges for cataloguing standards presented by new cultural and ethical paradigms. As well as on scholarly literature, I will draw on my personal experience of libraries as a user and, more recently, as an insider, linking long-standing issues to current practices within actual libraries, academic libraries in particular.
Although the notion of what constitutes a library varies in different places and at different times, with blurred boundaries between ideas of archive and library as we understand them today, for the purposes of this essay I will use the term library to refer to “a collection of graphic materials arranged for relatively ease of use, cared for by an individual or individuals familiar with that arrangement, and accessible to at least a limited number of persons” (Harris 1995, 1). In this sense, it is safe to say that historically catalogues were physical and intellectual tools for (proto) librarians to record holdings and find items, as much as they are today, but took different forms and structures at various historical moments depending on the technology available and the purposes they served. The most basic distinction to be made is between catalogues for internal use – collection inventories or shelf lists – and catalogues for external users. The former aimed at keeping a record of the existing stock (and often its location, too), the latter were conceived as a way to ‘advertise’ the collection and became more prominent once libraries developed into centres of study and not mere repositories for preservation (and reproduction) of knowledge (Hannah 2017, 45). I believe this basic distinction still holds today, when any software for library information management presents at least two interfaces of the catalogue, one for library users to search the collection (and locate or download items), and one for librarians to manage it (i.e. purchase, accession, discard items etc.). The developments of information technology, especially in recent years, have however afforded library catalogues an unprecedented sophistication turning them into very powerful intellectual tools, as I will explain in the conclusive part of this work. So when did humans start cataloguing?
Among the most ancient civilizations of the Western canon, in Mesopotamia and Egypt, archaeologists have found evidence of early record collections. Although little of the fragile materials used in Egyptian libraries (papyrus and then parchment) has survived, circumstantial evidence suggests that systems of organisation of their contents were in place; on the other hand, the Sumerians, Assyrians and Babylonians have left behind abundant material (Bawden and Robinson 2012, 22-24). Clay tablets have been found containing lists (of gods, professions, incantations, geographical names), but also letters, commercial and family records, mythological narratives, hymns, laws, glossaries etc. Systems of storing, preserving and retrieving these documents are well attested: one of the earliest examples of clay catalogues from the Sumerian city of Ur from some 5,000 years ago, listed key words from the first two lines of a hymn to identify its tablet; among the Akkadians, who replaced the Sumerians and built their first libraries around 2200 BC, the clay catalogue contained instead the incipit for a single group of tablets (Lerner 2002, 14-15). In Mesopotamia, documents were held in palaces or temples and fulfilled administrative, religious, and scholarly purposes. Among the first librarians we know of, was the Babylonian Amilanu, who lived approximately in 1700 BC: Babylonian libraries were open to all and users were handed a piece of papyrus to write their names on and the title of the material they wished to see, so that the librarian would fetch it for them (Norris 1939, 1-5).
Curiously this is exactly what I had to do in the Cambridge University Library to consult Norris’s text, although some books in this particular section can also be requested electronically via the online catalogue. Like in any modern ‘temple of knowledge’, the amount of documents at Cambridge UL is especially vast (being a 600-year-old and a legal deposit library), and the increasing availability of resources on electronic format does not seem to have helped solve the chronic shortage of space. Those subjects that have enjoyed a rapid expansion in recent years have books overflowing on (and under) tables, invading other subjects’ shelf space, intruding into areas of the library unfamiliar to the specialist reader, who follows the temporary re-direction signs through unknown labyrinthine stacks, meanders from floor to floor, and then eventually gives up, and turns to the library staff for help. And the competent, modern-day Amilanus can help, of course, because they know the principles that inform not only the arcane classification systems and shelving whereabouts but also the exceptions and emergency plans in place. And yet one resents the betrayal of the online catalogue that showed the ‘wrong’ location, and no warning of such intricate displacement.
This experience was not one that library users in Mesopotamia would have been familiar with, as self-fetching and browsing was not an option, and the kind of catalogue they could rely on was typically a classified or classed catalogue, i.e. a catalogue arranged by subject according to the classification system in place in each specific library. In the famous library of Nineveh, established by the King Ashurbanipal (ca. 669-631 BC), the sizeable collection was organised by subject in separate rooms. The tablets were kept in jars and arranged on wooden shelves in a systematic order, with each tablet numbered and for multi-tablet works, each tablet presented the first line of the whole work as well as the first line of the succeeding tablet; related works were held together in a basket, whose tablet-tag contained an indication of the set, the jar, the shelf and the room. Outside each room, an early form of catalogue (also on clay) displayed a list of its contents: title of works, opening words, number of tablets each consisted of, and location symbol (Harris 1995, 20). Although Mesopotamia lays claim to the first libraries of antiquity, evidence suggests that Egypt also had some as old and similarly arranged its papyrus rolls according to content. The oldest surviving library catalogue has effectively been discovered at the temple library of Edfu in Upper Egypt (early 3rd century BC) and it consists of a list of books engraved on the walls of the library itself. It is also known that rolls were kept in chests where a tablet or a pictorial representation on their cover indicated the books contained in them (Husayn 1970, 22).
If subject catalogues have been in use to these days, both as physical objects (the card catalogue) and as a search criteria or access point to a collection, pictorial catalogues have typically held an ancillary role. Having examined many instances in different countries and eras, the library historian André Masson explains that library decoration before 1789 was closely related to the classification of the volumes on the shelves and had practical purposes. He argues that, since classical antiquity, portraits or busts of authors featured regularly in libraries to signpost the location of the various collections: so if a seated Euripides holding a list of his plays could be found in a Roman library, monastic libraries would strategically display images of authors from religious orders. Among more sophisticated instances of pictorial catalogues the Tree of Knowledge in the Austrian Benedectine’s Abbey of St. Lambrecht stands out: painted on a pillar it displays a crop of open books with their author’s names. In Renaissance and baroque libraries images were often intended to suggest the subject classification: the Judgement of Solomon indicated Law; Apollo and the Muses pointed to Literature, etc. Although a pictorial representation never replaced a catalogue proper (not even in Rococo libraries, with their flamboyant decorations and imaginative designs) the visual element has made a comeback in today’s OPAC and Discovery catalogues, where thumbnail images of modern items’ covers and jackets, for example, help patrons identify and select materials.
The fully-fledged library catalogue of antiquity comes into existence at the famous library of Alexandria in Egypt, the first example of a universal library whose purpose was to acquire all the available knowledge and attract major writers, scholars, etc. to make it the centre of learning. The poet Callimachus compiled a classified catalogue consisting of 120 rolls. It was made on slips of papyrus called Pinakes (‘tables’), where a short title was written which corresponded to the label on the appropriate papyrus roll. The main divisions of the catalogue comprised Epic writers, Dramatic writers, Writers on Law, Philosophical writers, Historical writers, Oratorical works, Rhetorical works, Miscellaneous works, and each was further subdivided, with entries arranged either in alphabetical order by author, or chronologically. Some sections (like philosophy) required additional data to avoid confusion, so Callimachus added a brief analysis of the work as well as a short profile of the author. As Harris (1995: 45) points out, this additional bio-bibliographical work earns Callimachus the title of first bibliographer and subsequent scholars continued his work. The library staff in Alexandria also undertook the important task of recording the provenance of each roll, and who had improved the work. In modern terms we would call these a publisher or editor, entities that in current international cataloguing standards (ISBD) and its various actualizations (AACR2 and FRBR) are considered essential information to describe a bibliographic item (Bawden and Robinson 2012, 112).
The library of Alexandria was not just a centre of scholarship and a statement of grandeur, it was also an attempt to Hellenize the newly conquered territories (Lerner 2002, 26). Judging by the high standing that Greek culture held among the Romans, the project appears to have succeeded. The collections in Roman libraries (both public and private) were in fact typically subdivided by works in Greek and those in Latin. In Imperial Rome, libraries were both centres of knowledge as well as power structures, with hierarchies, political appointments and sometimes direct control of the kind of materials made available to the public. This became extreme in the case of Christian writings which were banished by most emperors before Constantine’s rule (306-337 AD) but were to become the predominant texts in later centuries when the opposite phenomenon took place (Greek and Roman classics were suppressed or marginalised). The catalogue reflected this epistemological shift and was informed by the Institutiones the work by 6th-century scholar and monk Cassiodorus. It became the book list for monastic and cathedral schools and the basis for the curriculum in medieval universities: the Bible and the commentaries of the Church Fathers constituted the core texts, “grammar and history, science and mathematics contributed to the appreciation of God’s creation” (Lerner 2002, 39).
As a result, in Medieval time catalogues were usually shelf lists or inventories compiled to record holdings (and donations) and they listed books in order of importance starting with the scripture and ending with secular literature (Lerner 2002, 94). They did not normally indicate the exact location of the book as they were not intended for readers’ use. The most important innovation of the medieval period were union catalogues, i.e. lists that indicated the holdings of several libraries (usually monastic ones and in relative proximity from one another). They emerged in different European countries and allowed the preservation and circulation of texts through loans for copying (Lerner 2002, 95). Modern union catalogues are today online databases: the largest is WorldCat, which brings together over 10,000 libraries worldwide, but national (and international) versions are now common in many countries.
The curriculum of the liberal arts in medieval universities in Europe followed the blueprint of the monastic libraries: they comprised the trivium (grammar, rhetoric and dialectic) and the quadrivium (arithmetic, geometry, music and astronomy). The subject range later expanded to include theology, medicine and law, yet central university libraries, with few exceptions (like Paris), were slow to be organised, and catalogues did not develop further than the book list, arranged alphabetically or by author with mere location symbols, although local classification systems were occasionally devised (Harris 1995, 110-115). This heritage is still visible in many libraries in old university colleges where in-house classification is very common and even hand-written classified shelf lists still in use.
From 14th century Italy, the renewed interest in classical works initiated by Humanism gave rise to a book hunt that during the Renaissance brought back to life many classics hidden in private collections and monasteries. Collectors, scholars and royals, commissioned librarians and bookmen to take stock of their libraries and replenish them; catalogues and book lists, old and new, became precious tools for this pursuit. The new scholarship and literary sensibility generated by this epochal shift, was aided by the moveable print that from the mid-15th century Germany spread rapidly throughout Europe. The increased amount of publications (including translations in the now stronger national languages) not only redressed the balance between religious and secular works available, but created the need to manage the flood of bibliographic information (Lerner 2002, 97-102). New classification schemes were put forward by Francis Bacon in Britain and Gottfried von Leibnitz in Germany. The Swiss scholar Conrad Gessner pioneered ideas of selection, arrangement and indexing in his influential Bibiotheca Universalis (1545), the first modern bibliography which, despite its title, managed to include an estimated 25% of works available at the time (Bawden and Robinson 2012, 27).
The printed library catalogue was not a real advancement on the codex format used in previous centuries as, once published, it was already out of date. It was therefore soon dismissed or slow to appear. As Campbell explains (2013, 238-239), this prompted two innovations: one was the slip catalogue, produced by William Croswell for the Library at Harvard College in the 19th century. The idea, already mentioned by Gessner, consisted in cutting up the printed catalogue from 1790 into slips, and pasting them into blank volumes called guard-books, with spaces left between them for new entries. The other innovation was the card catalogue, which made its appearance at the end of the 18th century when playing cards were used by Edward Gibbon to catalogue his library. Blank cards became common in the 1870s partly following Melvil Dewey’s innovations in classification and professionalization of librarianship, which included fittings and stationery. Card catalogues consist of an author catalogue and a subject catalogue, arranged in alphabetical order (sometimes as one physical object sometimes as two separate ones). They were widely adopted by university libraries and public libraries worldwide and in some they are still in use.
In modern times the ever increasing size of the collections and the growth of user numbers forced libraries to rethink their approach to the organisation of and access to knowledge. The need for standardisation of cataloguing practices became increasingly pressing and some important efforts developed in the 18th and 19th centuries informed modern practices, among them the Instruction pour procéder à la confection du catalogue de chacune des bibliothèques issued in Paris in 1791 for the libraries taken over during the Revolution, Antonio Panizzi’s “91 Rules” devised in the 1830 for the printed book collection of the British Museum, the 1908 “Prussian instructions” adopted in the German-speaking countries (Wedgeworth 1993, 177-178). Much as Thomas Hyde’s 17th-century catalogue of the Bodleian Library in Oxford was prompted by his concern to help students find resources, Panizzi’s efforts at the British Museum was a further step forward: he aimed at presenting readers not just with what they were looking for, but also with new information (further editions, related versions, translations). This shift prompted librarians to think about principles and methodologies to organise the catalogue in a way that suited users’ needs and a century after the field was teaming with initiatives to standardize cataloguing rules (Faulkner 2014, 104-106). Their shortcomings from an ethical point of view, has also been pointed out, albeit by a minority and progress has been slow. If Dewey’s system and world view was accused of ‘Anglo-Saxonism’, a kind of Wasp supremacism, cultural bias in the Library of Congress subject headings used worldwide is still rife. The recent student and faculty campaign to de-colonise the university curriculum has started to involve libraries, and initiatives to de-colonise the catalogue as well as the collections are underway.
The arrival of automation, by far the most momentous innovation for the catalogue, promoted a flurry of initiatives both in Europe and in the US and fostered further collaboration. The pioneering work of Henriette Avram for the Library of Congress produced the standardized machine-readable cataloguing (MARC) in the mid-60s and MARC records began to be shared by the end of the decade. In the 1970s the efforts of the International Federation of Library Associations resulted in UNIMARC, a similar system but tailored on the needs of European libraries (Fons 2016, 17-18). The standards to encode authoritative forms of name for authors so that, regardless of spelling, users would find them, to encode information about series, or subjects, or editions etc. have been variously updated over the years (for ex. AACR2) and so has the metadata used by the software to cross reference between records, for example, or to instantly retrieve records (MARC is now in its 21st version). In the Anglo-American world, these standards are slowly being replaced: BIBFRAME (Bibliographic Framework), a system that, by using linked data, aims to appeal also to users outside the library community, is seen as a replacement for MARC21, and AACR2 has already been replaced by RDA (Resource Description and Access) and an enhanced model that should allow for more flexible search and display, FRBR (Functional Requirements for Bibliographic Records), is the next likely successor. The latest technological challenge faced by the catalogue has been the Web, with its booksellers’ on-line catalogues of humongous proportions. As Faulkner (2014, 107) laments: “our catalogue is locked inside an arcane format, unable to talk to the Web, and unable even to give a good imitation of the Web behaviour […] Library catalogues travelled one road and the Internet took another.” The gap between the two systems can perhaps be breeched by the Semantic Web, a new way of encoding information in software so that it can read information encoded in web pages. But this is still an ongoing project, desiderata perhaps. In the meantime, the new ‘discovery tools’ adopted by some major libraries (usually national and academic libraries) have made the single-search option of the library catalogue’s holdings from the integrated library system (ILS) a reality. And so not only books, but also e-books, articles, e-journals, audiovisual resources etc. can be found in a single search (various facets can be applied to refine the results). Initial user surveys reported mixed feelings. Inevitably, perhaps, as Borges would say: “Like all men of the Library, I have travelled in my youth: I have wondered in search of a book, perhaps the catalogue of catalogues.”
 The civilizations of Mesopotamia are normally regarded as the groundbreaking innovators of Western culture in this area: they invented writing (the cuneiform system in ca.3200 BC) and a clever way to record documents (on clay tablets), among which important sets of laws (the Hammurabi code, ca. 1700 BC) and early works of literature (The Epic of Gilgamesh, ca.2100 BC) have survived today.
 The library has indeed started to move some of its collections to a storage facility in a nearby town where access is for staff only. Readers can request items via the online catalogue and collect them the next day (or so), but cannot borrow them (see: http://www.lib.cam.ac.uk/about-library/librarystoragefacility). Similar arrangements exist in many university libraries, especially those with multiple campuses (and libraries). Space is especially problematic in old universities, where safe storage and preservation of rare materials is an additional issue.
 Aside from newspapers and periodicals, available in reading rooms, it was not until the end of the 19th century that libraries in Britain and the UK, for instance, started to allow the public near the bookshelves (Campbell 2013, 238). In many libraries in continental Europe the ‘sacrality’ of the text and the almost hieratic function of the librarian has lasted much longer, partly due to issues of preservation (especially of precious incunabula, manuscripts or other rare books) and the awkwardness of access (in buildings not usually designed to hold books).
 Many of the tablets also present a kind of ‘library stamp’ which identifies the items as belonging to his palace (Taylor 2018). The library held around 10,000 works (consisting of ca. 30,000 tablets).
 His book Le dècor de bibliotéque was published in French in 1972 and translated by David Gerard as The Pictorial Library. Mural Decoration in Libraries, Oxford: Clarendon Press, 1981. See Hobson (1983) and Campbell (2013, 29 and 192).
 The library of Alexandria was founded during Ptolemy’s kingdom (285-246 BC), the Greek dynasty that had come to rule Egypt and other parts of the Mediterranean. By this time, books were written in ink on papyrus rolls and it is estimated that at its height the library must have held up to 400 thousand scrolls (Lerner 2002, 25-29).
 It was not until the invention of the printing press (1455) that bibliographies in the modern sense were developed. The need to record each individual copy of a work was superseded by the availability of multiple identical copies of the same texts. See below for more details.
 Not much is known about their classification systems, as for catalogues, these were of two types: the classified catalogue and the bibliographical catalogue arranged by author, with titles, first lines of works, and sometimes information about the author (Harris 1995, 65-66).
 Among the earliest surviving English library lists, for example, the one from Christ Church Canterbury (c.1170) presents a series of small letters and symbols which correspond with symbols on the first leaf of each book. This made possible to identify which book was which, and, maybe, where it was located (Clanchy 2013, 160). In some cases, colour-coding as well as letters was adopted to identify the subject of the books and the system had some success. For example in a monastery in Saxony red was for theology, green for medicine and black for law. In the early 13th century the bibliophile Richard de Fournival added letters (capitals, minuscule, uncial), a system which allowed for many distinctions. A century later the library of the Sorbonne University was organised on similar principles (Lerner 2002, 95).
 For example the COPAC catalogue brings together over 100 UK and Ireland’s national libraries, university libraries and specialist research libraries; in Italy the OPAC SBN catalogue covers about 6000 institutions nationwide (national, university and public libraries, as well as libraries of major private foundations and collections). See: https://www.worldcat.org/; https://copac.jisc.ac.uk; https://opac.sbn.it.
 They were formally established from the 11th century in Italy, France, England and Spain. By the 15th century many existed included elsewhere in Europe (Portugal, Poland, Austria, and Scotland, among others).
 This is the case of the college library I am currently employed at, where the hand-written registers (30 heavy folders) are kept up-to-date mainly for the annual stock-check, even though the software (SirsiDynix), introduced in 2000, could be used for this purpose.
 The publication of national library catalogues in Europe first appeared at the end of the 19th century: the author catalogue of the Bibliotèque national de France appeared in 1897 and the first general catalogue of the British Museum Library began publication in 1881.
 The slip catalogue, occupying a large room in its multivolume folio size, can still be used in the Cambridge University Library, and card catalogues are still in use is some old colleges in Oxbridge, although they are not always kept up to date.
 The 1961 ‘Paris principles’ were followed in the same decade by the International Standard Bibliographic Description (ISBD) and the Anglo-American Cataloguing Rules (AACR), later updated.
 For the Library of Congress see for example Berman and Moore (2013). A recent initiative promoted by Cambridge librarians in November 2018 has discussed various ongoing initiatives, among them a revision of classification systems to be more inclusive (English Faculty), acquiring collections in minority languages, countercultural publications and experimenting with new ways of cataloguing these materials (Latin American Collections, University Library). See https://twitter.com/hashtag/EFLdecolonise?src=hash.
 The expression was first used by Tim Berners-Lee, the creator of the web itself, to refer to the system of linking information (metadata) with rules to understand them in order to generate new data (like automated web services). See his views on https://www.w3.org/DesignIssues/LinkedData.html.
 To some it is an outright flop. See for example Sinclair Target’s article ‘Whatever happened to the Semantic Web’, May 2018, https://twobithistory.org/2018/05/27/semantic-web.html (accessed 2.1.19). For a survey of the many challenges faced by libraries trying to promote linked data project see Ali and Noshema (2018).
 Their introduction is fairly recent and little research has been done to evaluate the reception of discovery tools by users, but some surveys conducted in universities concluded that the preferences were mixed. See Wells (2016) and Antonucci and Clapp (2014). The introduction in January 2018 of one such catalogues at Cambridge UL and related faculty and college libraries has been dubbed ‘the great migration’, for its inevitable problems of standardisation of practice, glitches, and disruptions.
 Jorge Luis Borges, ‘The Library of Babel’, trans. by James E. Irby, Penguin, 1964, p. 78. Originally published in 1941 as ‘La biblioteca de Babel’.
Antonucci, Carl, and Sharon Clapp (2014). ‘The University Library Evolution. Book Warehouse or Platform for Student Research and Learning: Planning for the Future at the Elihu Burritt Library, Central Connecticut State University’. in Eden, B.L., Leading the 21st-century academic library: successful strategies for envisioning and realizing preferred futures. Lanham: Rowman & Littlefield Publishers, 57-79.
Bawden, David, and Lyn Robinson (2012). Introduction to information science. London: Facet Publishing.
Berman, Sanford, and Eric Moon (2013). Prejudices and antipathies: A tract on the LC subject heads concerning people. Jefferson, North Carolina: McFarland.
Campbell, James W.P. (2013). The Library. A World History. London: Thames and Hudson.
Clanchy, Michael T. (2013), From Memory to Written Record. England 1066 to 1307. 3rd ed. Oxford: Wiley-Blackwell.
Faulkner, Autumn (2015). ‘Serendipity and the Semantic Web: Catalogs of the Future’. in Eden, B.L., Leading the 21st-century academic library: successful strategies for envisioning and realizing preferred futures. Lanham: Rowman & Littlefield Publishers, 97-118.
Fons, Ted (2016). ‘Improving Web Visibility: Into the Hands of Readers’. Library Technology Reports 52 (5): 5-32.
Hanna, Ralph (2017), ‘Manuscript catalogue and book history’. The Library 18 (1): 45-61.
Harris, Michael H. (1995). History of Libraries in the Western World. 4th ed. Lanham, Maryland: The Scarecrow Press.
Hobson, Anthony (1983). ‘ANDRÉ MASSON.The Pictorial Catalogue. Mural Decoration in Libraries’(Book Review). The Library 6-5 (1): 82–83.
Husayn, Muhammad A. (1970). Origins of the Book. Egypt’s contribution to the development of the book from papyrus to codex. Leipzig: Edition Leipzig.
Infan, Ali and Fatima W Nosheen (2018). ‘Linked data initiatives in libraries and information centres: a systematic review’. The Electronic Library 36 (5): 925-937.
Lerner, Fred A. (2002). The story of libraries: from the invention of writing to the computer age. London: Continuum.
Manguel, Alberto (2008). The library at night. New Haven, CT: Yale University Press.
Norris, Doris M. (1939). A history of cataloguing and cataloguing methods 1100-1850. With an introductory survey of ancient times. London: Grafton & Co.
Taylor, Jonathan (2018), ‘A library fit for a king’, The British Museum Blog, https://blog.britishmuseum.org/a-library-fit-for-a-king/?_ga=2.31160255.1079843197.1545919035-182004412.1545919035 (last accessed: 27.12.2018)
Wedgeworth, Robert (1993). World Encyclopaedia of Library and Information Services. 3rd ed. Chicago: ALA.
Wells, David (2016). ‘Library discovery systems and their users: A case study from Curtin University library’. Australian Academic & Research Libraries 47 (2): 92-105.