How Was the Oxford English Dictionary Created?

The Oxford English Dictionary was created by treating language as a historical evidence problem. Beginning in 1857, the Philological Society recruited volunteer readers to find words in books and copy revealing passages onto quotation slips. Editors sorted those slips by word, date, and meaning, then used the evidence to write definitions and trace how each sense developed. James Murray took charge in 1879, the first installment appeared in 1884, and the first edition was finally completed in 1928.

That method made the OED different from a compact rulebook that simply tells you which word is “correct.” Its central job was historical: show what English words have meant, where those meanings appeared, and how usage changed over time. The dictionary was crowdsourced long before the internet, but its raw material still had to survive an enormous editorial process.

Why the project began

In the mid-19th century, members of London’s Philological Society concluded that existing English dictionaries left too many words, meanings, and historical examples unrecorded. In 1857, the society formed a committee to collect “unregistered words.” What first looked like a supplement to earlier dictionaries soon became a much larger ambition: a new dictionary built on historical principles.

The key insight was that no small editorial team could read the whole history of written English alone. The society therefore issued an appeal to the public. Volunteers were assigned books from different periods and asked to notice uncommon words, new meanings, and ordinary words used in telling ways. This distributed reading network became the project’s evidence-gathering engine.

From a Victorian proposal to a finished dictionary

Progress was uneven until the project found both a publisher and an editor able to reorganize it. In 1879, Oxford University Press agreed to publish the dictionary and James Murray became its leading editor. Murray revived the reading program and ran a small editorial team from an iron building he called the Scriptorium, first in Mill Hill and later in Oxford. Shelves of pigeonholes held quotation slips waiting to be sorted and interpreted.

The schedule proved wildly optimistic. The work was expected to take about ten years, yet the first fascicle—covering A to Ant—did not appear until 1884. Murray died in 1915 before seeing the project finished. Other editors, including Henry Bradley, William Craigie, and C. T. Onions, carried the work forward, and the complete first edition appeared in 1928 as ten volumes under the title A New English Dictionary on Historical Principles.

How the quotation-slip system worked

A useful slip did more than report that a word existed. It copied enough of the surrounding sentence to reveal meaning and recorded the source, author, and date. After slips arrived, assistants sorted them alphabetically and then grouped the evidence by distinct senses. Editors compared the earliest and later examples, drafted definitions, investigated word origins, and selected quotations that made the history visible to a reader.

This is why quotations in the OED are not decorative literary flourishes. They are evidence. One quotation can show that a sense existed by a particular date; several spread across decades can show continuity, change, or a new branch of meaning. A missing early example can also leave a date provisional, which is why later discoveries can move a word’s documented history backward.

A three-slip example: building an entry for “spark”

Imagine three simplified, fictional slips collected for the word spark. A 1720 passage describes “a spark from the hearth.” An 1840 passage calls an idea “the spark of invention.” A 1910 passage says a protest “sparked a debate.” These are teaching examples, not quotations from the OED, but they show what editors could learn from dated context.

The first slip supports the concrete noun: a small fiery particle. The second supports a figurative noun, something that begins or inspires activity. The third supplies evidence for a verb meaning to cause something to start. An editor would not average them into one vague definition. The slips would be separated into senses, ordered by the available historical evidence, and paired with representative quotations.

Now imagine receiving another slip dated 1790 that uses spark figuratively. That evidence would push the documented date of the figurative sense earlier than the 1840 example. The definition might remain stable while the historical record changes. This small revision captures the OED’s basic logic: an entry is an argument built from dated usage, not a decree handed down without evidence.

Why the OED took so long

The alphabet was the easy part. Editors had to decide where one sense ended and another began, evaluate competing spellings and origins, check quotations, correct proofs, and continue collecting evidence while publication was already under way. English also kept changing during the project. By the time later letters were being edited, decades had passed since the first appeal to readers.

Scale compounded the problem. The completed first edition contained hundreds of thousands of words and phrases supported by a vast quotation archive. Every compact entry concealed reading, sorting, interpretation, and typesetting. The delay was not simply poor planning; it followed from attempting a kind of historical coverage no earlier English dictionary had achieved at that scale.

A common misconception: dictionaries create word meanings

People often talk as if a dictionary committee invents meanings and then authorizes everyone else to use them. The OED’s original method ran mostly in the opposite direction. Readers found evidence of how people had already used words, and editors analyzed that evidence. Editorial judgment still mattered—sources were incomplete, categories had to be drawn, and definitions had to be written—but the evidence came first.

There is an important boundary here. A historical dictionary is not a perfect transcript of every speaker. Printed sources were easier to collect than speech, and the first edition reflected the materials and scholarly assumptions available to its editors. Modern revision can add overlooked varieties of English, antedate senses with newly found evidence, and reconsider old analyses. A dictionary can be authoritative without being finished forever.