Word History Is Human History
The Oxford English Dictionary shows how changes in language mirror changes in society.
(Wrote this for How-to History, which gives an insider look at the art of uncovering and understanding the past.)
Suppose the English 17th century were wiped from the historical record. Elizabeth I reigns until 1600, then a blank spot. We know nothing of a parliamentary rebellion, the scientific revolution or the rise of capitalism.
But there are clues. Words we use every day in the 2020s surfaced in English during this period. They are artifacts, witnesses to early modernity.
Legislature, newspaper, reform, human rights. Decapitate.
Data, environment, computer, chemistry, microscope.
Manufacture, monetary. Tea, chocolate.
In language “the past history of humanity is spread out in an imperishable map, just as the history of the mineral earth lies embedded in the layers of its outer crust,” Owen Barfield wrote in History in English Words, first published a century ago. Language, he said, holds “the living history of man’s soul” and “the evolution of consciousness.”
Historians of England and affiliated nations are lucky to have the best map of the linguistic past ever made, these days entirely online, searchable, sortable and fascinating.
When members of the London Philological Society launched what would become the Oxford English Dictionary in the mid-1800s they called it A New English Dictionary on Historical Principles. The idea was not just to give definitions but to track language across the centuries, “catch every word on its first appearance in our literature” and then show extinctions, survivals and changing meanings.
The OED must be the longest-running crowdsourced project and the biggest before the Internet. Organizers sought to harness “the scattered learning and energy which exists plentifully enough in this country” by asking the public to contribute lexicographical nuggets.
The OED’s early offices included a shed with 1,029 pigeonholes for organizing millions of submission slips. Editors are still seeking help. The number of contributors over the years must run to many tens of thousands.
Words followed through time show the history of culture, politics, immigration, war, love and how people see the world.
Toile and toilet, from the French, tell about English textiles, the Scots-French Auld Alliance and the Victorian inability to talk about certain body functions. The journey of the word science, starting in the 1400s when it meant God’s omniscient awareness, shows the progress of knowledge.
Thug is a Hindi word originally meaning a violent bandit. It came to be a colonial and modern disparagement of anybody you don’t like.
Salvation marks the French monastic invasion of England in the 1100s and 1200s and the preaching friars adding French words to their sermons. French Huguenots fleeing to England were the first refugees by that name.
The OED is essential to fiction writers building past worlds.
“No one should even consider writing a historical novel without it there on the desk,” Philip Pullman wrote in The Guardian of the OED’s companion, the Historical Thesaurus of the Oxford English Dictionary. Romance writer KJ Charles wished a decade ago for an anachronism detector allowing an author to “plug in the year 1888, run your MS through the OED and have it flag words dating from later.”
Bots trained on large language datasets can do this now. I set ChatGPT 4o to work on eight paragraphs from a random Regency bodice-ripper. It instantly identified shopping trip, good on you and do tell in characters’ speech as modernistic howlers.
The online OED, finicky a few years ago, has become really good. Editors cut the price of an individual annual subscription to $100 or £100, but you can use public library subscriptions for free.
Say, for some reason, you want to identify nouns joining English from Dravidian languages in the 17th century. OED advanced search delivers pariah, catamaran and two dozen others. All new words related to sport and leisure appearing in the 1200s? Wicket, bat, bookmaker… Adverbs used by Shakespeare related to agriculture? Parkward and afield.
OED editors are casting further, drawing more fully on the University of Michigan’s Middle English Dictionary and other specialty research that often shows earlier examples of certain terms than those in the OED.
Scholars still need to beware of the common mistake of saying a word “entered English” at the time of its first OED citation. Preservation in a text isn’t the same as spoken usage, especially in the post-Conquest centuries when English was seldom written down while the language added thousands of Norman-French words.
Last year in Past & Present, a top social history journal, John Gallagher and Purba Hossain wrote that “language remains relatively marginal as a subject of historical analysis.” Machine learning, the digital history movement and the OED’s trove of 600,000 word entries might change that.
The dictionary’s application programming interface enables specialized projects and access to bulk OED data. The Living with Machines team of historians, archivists and technologists is using the OED API to tune a language model with the aim of analysing how people talked and wrote about mechanization during the Industrial Revolution.
Like everybody OED editors must figure out how to coexist with AI bots and language datasets.
Creating today’s OED took a century of harvesting words by hand at the Bodleian, the British Museum and hundreds of other places. ChatGPT, Claude and the others can extract lexicographical information in an instant from Project Gutenberg and other enormous text repositories.
The OED has been working with the Google Books dataset and Google’s nGram Viewer to show historical changes in how often certain words are used. Expect more collaboration like this. Bots trolling language oceans can generate amazing results, but their sources and methods are often unknowable. The OED gives verifiable facts, the material of history.