The History of Dictionaries: From Ancient Glossaries to the Digital Age

Introduction

A dictionary is a deceptively simple object: a list of words with explanations attached. But the people who built the earliest ones were pressing clay with reed styluses more than four thousand years ago, long before paper, printing, or alphabetical order were taken for granted. Every major leap in communication — cuneiform, parchment, movable type, the web — has produced a new kind of dictionary, and every new kind has reshaped the way ordinary people learn and argue about words.

Tracing that arc is worthwhile because dictionaries are never neutral. Someone has to decide which words count, which meanings come first, and which examples illustrate them. The history below is partly a story of scholarship, partly a story of technology, and partly a story of national pride, commerce, and the occasional eccentric genius working in a shed at the bottom of a garden.

The First Word Lists of the Ancient World

The oldest surviving ancestors of the dictionary were pressed into clay in Mesopotamia around 2300 BCE. As Akkadian gradually pushed Sumerian out of everyday use, scribes needed bilingual cribs to keep reading the older documents. Their solution was the cuneiform word list: a column of Sumerian terms set beside their Akkadian equivalents, copied out by apprentices as part of scribal training.

The best-known of these compilations is the Urra=hubullu series, which ran to dozens of tablets. Instead of alphabetizing entries (an impossibility in a syllabic script), its compilers grouped words by subject — trees, reeds, garments, ceramic vessels, cuts of meat, domestic animals — producing something closer to a thematic encyclopedia than a modern dictionary.

Egyptian scribes produced their own training word lists, and the Rosetta Stone (196 BCE), though really a royal decree, later became priceless to linguists because it repeated the same message in hieroglyphic, Demotic, and Greek. That trilingual format is exactly the pressure that drives lexicography: people needing to move one text between three languages.

China developed a parallel tradition. The Erya, usually dated to the 3rd century BCE, explained obscure terms from the Confucian classics and grouped entries by meaning — kinship, heaven, earth, and so on. Around 100 CE, Xu Shen's Shuowen Jiezi took a different approach, sorting roughly 9,500 characters by the radical components from which they were built. That radical-based arrangement is still used in Chinese reference works today.

Indian lexicography reached a high point with the Amarakosha, compiled by Amarasimha around the 4th century CE. It is essentially a thesaurus in verse: groups of Sanskrit words for related ideas, written in meter so students could memorize them while walking to class.

In the Arab world, lexicography became a serious science starting in the 8th century. Al-Khalil ibn Ahmad's Kitab al-Ayn ordered the entire Arabic root system according to where each consonant was produced in the mouth, from the throat outward — a phonetic scheme so systematic that later Arabic dictionaries have been arguing with it ever since.

Medieval Latin Glossaries

In medieval Europe, the language that mattered was Latin, and the problem was that very few readers understood every Latin word they met in scripture, liturgy, or classical authors. Monks dealt with this by writing small notes above or beside the tricky words. These marginal notes are called glosses, and the habit of collecting them gave us the word "glossary."

Once scribes began copying out the glosses on their own, divorced from any particular manuscript, a new kind of book appeared. Papias the Lombard's Elementarium Doctrinae Rudimentum, produced around 1053, was an early and influential example: Latin words in alphabetical order, each explained, many accompanied by a note on etymological information.

Bilingual glossaries mattered even more once vernacular languages started to assert themselves. In England, Anglo-Saxon scholars paired Latin headwords with Old English equivalents, building the tools that their successors would eventually point back at English itself.

Everything changed in the 1440s, when Johannes Gutenberg's printing press began producing identical pages at speed. Hand-copied glossaries had always suffered from drift: each copy picked up errors, and no two were exactly alike. Printed dictionaries could be proofread once and then reproduced by the hundred, turning what had been a monastic resource into a commercial product.

The Earliest English Dictionaries

Strikingly, the English language had no monolingual dictionary of its own until the 17th century. The usual candidate for "first" is Robert Cawdrey's A Table Alphabeticall of 1604, a slim volume of roughly 2,560 entries concentrating on the "hard words" that Renaissance writers were pouring into English from Latin, Greek, and French.

Cawdrey was unusually candid about his readership. His title page offered the book to "Ladies, Gentlewomen, or any other unskilfull persons," by which he meant people who had never studied the classical languages and therefore needed help with the imported vocabulary of sermons, law, and medicine.

A sequence of ever-larger hard-word dictionaries followed:

• Henry Cockeram's The English Dictionarie (1623): the first book to put the word "dictionary" in its title, and unusual in offering both difficult-to-plain and plain-to-difficult word lists.
• John Bullokar's An English Expositor (1616): larger than Cawdrey, with fuller definitions and a clearer scholarly tone.
• Edward Phillips's The New World of English Words (1658): ambitious in scale, with encyclopedic material mixed in alongside ordinary definitions.
• Thomas Blount's Glossographia (1656): around 11,000 entries, and the first English dictionary to supply word origins on a systematic basis.

What unites all of these is a selective approach. None of them tried to describe English as a whole — they assumed readers already knew the common words and only needed a guide to the unfamiliar ones. The idea of a comprehensive dictionary was still waiting for someone to attempt it.

Samuel Johnson Sets the Standard (1755)

Samuel Johnson's A Dictionary of the English Language, published in two folio volumes in 1755, changed what a dictionary was allowed to be. Johnson aimed at the whole language, not just its hard corners, and his book shaped English lexicography for the next 150 years.

A consortium of London booksellers put up the money in 1746. Johnson told them the project would take three years and ended up taking nine, working from a house on Gough Square with half a dozen amanuenses — most of them Scots, a detail he could not resist joking about. The finished book contained roughly 42,773 entries.

His real innovation, though, was evidence. Johnson was the first English lexicographer to back up definitions with quotations from major writers — Shakespeare, Milton, Dryden, Pope, Bacon, and scores of others — and more than 114,000 of these illustrative passages show readers how a word actually behaves in the hands of someone who knows how to use it.

The definitions themselves are part of the book's reputation. Johnson could be clinical when he wanted to be and waspish when he did not. He called a lexicographer "a writer of dictionaries; a harmless drudge," and defined "oats" as "a grain, which in England is generally given to horses, but in Scotland supports the people" — a jab that Scottish readers have remembered ever since.

"Dictionaries are like watches; the worst is better than none, and the best cannot be expected to go quite true." — Samuel Johnson

The book's prestige was such that rival dictionaries were often judged by whether they followed Johnson's methods, from the quotation apparatus to the etymologies to the habit of grading a word's senses from the literal to the figurative.

Noah Webster Gives America Its Own Book

If Johnson's dictionary spoke for Britain, the United States after independence needed a voice of its own, and Noah Webster — Connecticut schoolmaster, lawyer, and relentless patriot — was determined to supply it. He thought political independence would ring hollow if Americans kept deferring to London on questions of spelling and usage.

His earliest effort was A Compendious Dictionary of the English Language in 1806, but the book that made him famous was An American Dictionary of the English Language, finished in 1828 when he was seventy. With about 70,000 entries, it outran Johnson in scope, and it introduced the spelling reforms that still divide the two varieties of written English.

Webster's reforms followed an argument rather than mere preference. He dropped the u from "colour" and "honour," flipped "centre" and "theatre," changed "defence" to "defense," and pared "travelling" down to "traveling." These changes made spelling a little more phonetic and, in his view, a little more republican. Other proposals — "tung" for tongue, "wimmen" for women, "masheen" for machine — never caught on and are mostly remembered as curiosities.

His dictionary also wrote American life into the record. It carried entries for terms pulled from the frontier, the courtroom, Congress, the farm, and the trading post. When Webster died in 1843, the brothers George and Charles Merriam bought the rights, and the company that bears their name has been publishing dictionaries under the Webster banner ever since.

Building the Oxford English Dictionary

The single most ambitious project in the history of English lexicography started as a complaint. In 1857, members of the Philological Society of London declared the existing dictionaries inadequate and called for a new work on historical principles — a book that would trace every English word through its documented life, from first appearance to latest use.

Oxford University Press eventually took the project on under the working title A New English Dictionary on Historical Principles. Its early years were rough going: the first editor, Herbert Coleridge, died of tuberculosis in 1861; his successor, Frederick Furnivall, was bursting with ideas but could never quite keep the paperwork under control. Real progress began in 1879 when James Murray became editor.

Murray ran the project from a corrugated-iron outbuilding in his Mill Hill garden, which he named the Scriptorium. Volunteer readers across the English-speaking world posted him citation slips by the sack, and Murray's team sorted, verified, and filed them. One of the most productive contributors was W. C. Minor, an American Civil War surgeon writing from Broadmoor Criminal Lunatic Asylum, where he had been confined after a killing — a collaboration later retold in Simon Winchester's The Professor and the Madman.

The first installment, covering A–Ant, appeared in 1884. The full first edition — ten volumes — was not complete until 1928, more than seventy years after the original call to action. Supplements were needed almost immediately, because the entries at the front of the alphabet had already fallen behind the language.

In its modern form the OED remains the great historical record of English, with over 600,000 entries and roughly three million illustrative quotations, documenting the entire history of the English language from Old English down to current usage. Revision now happens continuously, online, rather than in discrete print editions.

What Changed in the Twentieth Century

The twentieth century didn't produce a single revolution on the scale of Johnson or the OED, but it introduced three distinct shifts that still shape how dictionaries are built.

Describing the Language Versus Ruling It

When Webster's Third New International Dictionary appeared in 1961, editor Philip Gove took an unapologetically descriptive line: he recorded English as people actually wrote and spoke it, including items like "ain't" that a traditional usage guide would frown on. Columnists and editorial boards reacted furiously, accusing the dictionary of lowering standards; descriptivist linguists pointed out that describing a language is not the same thing as endorsing every move it makes.

Dictionaries Written for Learners

A. S. Hornby's Oxford Advanced Learner's Dictionary, launched in 1948, was built around the needs of people studying English as a foreign language rather than native speakers. It restricted itself to a controlled defining vocabulary, spelled out grammar patterns, and flagged collocations, producing a type of dictionary that has multiplied as English has become a global working language.

The Rise of the Corpus

Machine-readable collections of real texts — corpora — gave lexicographers something earlier generations could only dream about: millions of natural sentences to sift. John Sinclair's COBUILD team at the University of Birmingham showed what that could mean in the Collins COBUILD English Language Dictionary (1987), which used corpus evidence for every decision and wrote its definitions in full sentences rather than clipped dictionary shorthand.

Dictionaries Go Digital

The shift from paper to software is probably the biggest single change in dictionary history since Gutenberg. Everything about how dictionaries are written, sold, and consulted has had to be reconsidered.

CD-ROM came first. The OED's second edition shipped on a single disc in 1989, and readers could now do something that bound volumes had always refused to let them do: search the full text, including every definition and every quotation, not just the headword column.

The web followed in the 1990s and early 2000s, putting dictionaries online at no direct cost to users, usually funded by advertising. A serious reference that had once required a household budget decision became a free search box.

Smartphones pushed the change further in the 2010s. Pocket dictionary apps added spoken-word search, camera lookups that let you point at printed text, and offline databases for use on flights or in the field. Looking up a word dropped from a minor errand to a half-second gesture.

Meanwhile Wiktionary, launched in 2002, pioneered a different model entirely: an openly licensed, community-edited dictionary that covers hundreds of languages. Coverage is uneven and editorial practice varies, but for many rare words or minority languages it is now the most complete resource anywhere.

Where Dictionaries Are Heading Next

The story isn't over, and several threads suggest where the next chapter is being written.

Machine learning already helps lexicographers track new words and watch meanings drift. An algorithm that reads billions of tweets, news stories, and reviews each week will notice an emerging sense of a word long before a human editor has a chance to.

Dictionaries are also sinking deeper into the software people already use. E-readers highlight a word and the definition appears; browsers offer translation on hover; word processors suggest synonyms as you type. The dictionary as a standalone book is quietly becoming an API behind other tools.

Categories that used to be firmly separate are merging on digital platforms. Modern dictionary platforms combine functions of the traditional dictionary, thesaurus, encyclopedia, pronunciation guide, and translation service behind a single query box.

What hasn't changed is the reason anyone builds these tools. Since those first Sumerian tablets, a dictionary has been a way of helping one person make sense of a word another person wrote or said. The surface keeps evolving; the underlying job has stayed the same for four thousand years.

A Timeline of Key Dictionary Moments

• c. 2300 BCE: earliest known Sumerian–Akkadian word lists on cuneiform tablets
• c. 3rd century BCE: the Erya, one of the oldest surviving Chinese dictionaries
• c. 100 CE: Shuowen Jiezi, the first Chinese dictionary to sort by radical
• c. 4th century CE: the Amarakosha, a landmark Sanskrit thesaurus in verse
• 8th century: Kitab al-Ayn, the first comprehensive Arabic dictionary
• 1604: Robert Cawdrey's A Table Alphabeticall, the first monolingual English dictionary
• 1755: Samuel Johnson's A Dictionary of the English Language
• 1828: Noah Webster's An American Dictionary of the English Language
• 1852: first publication of Roget's Thesaurus
• 1884–1928: the first edition of the Oxford English Dictionary, issued in fascicles
• 1948: first edition of the Oxford Advanced Learner's Dictionary
• 1961: Webster's Third New International Dictionary and the descriptivism row
• 1987: Collins COBUILD, the first fully corpus-based dictionary
• 1989: OED second edition on CD-ROM
• 1990s–2000s: major dictionaries migrate onto the web
• 2002: Wiktionary launches as an open, collaborative dictionary
• 2010s: mobile dictionary apps become standard