fb-pixel Skip to main content
The Word

What the Salem witch trials taught us about language

Amid the madness and executions, the famous proceedings left a remarkable record of how Puritans spoke and wrote

Kean Collection/Archive Photos/Getty Images

It’s nearly Halloween, and the town of Salem has launched its annual witch-kitsch fright-fest, whose events include cemetery tours, séances, and the 6th Annual Zombie Prom. Today, this is generally how we commemorate the Salem witch trials. A frenzied escalation of accusations between neighbors in Salem Village (now Danvers) and other towns in 1692, the trials culminated in the execution of 20 people on charges of consorting with the devil.

But the trials have another important legacy: For almost a century, the records they left have been a key resource in our study of early American English. The Puritans had a very high level of literacy, but didn’t write much down, especially in their colloquial tongue; they were too busy “trying to survive,” as University of Kansas professor Peter Grund puts it. The witch trials, however, left a range of documents, many transcribed from direct speech.

Since the 1920s, these approximately 1,000 manuscripts have been uniquely helpful in showing linguists how early Americans pronounced, spelled, and wrote their language, an English very different from our own. At the same time, linguistic clues in the Salem files have been crucial to understanding how the community spiraled so deeply into crisis.


The Salem witch hunt began in January 1692, when a group of young girls began throwing hysterical, inexplicable fits. The girls accused three women of bewitching them, and local magistrates began investigating. Starting from that moment, the Salem witch trials left a long paper trail, including transcriptions of dramatic examinations; false confessions, after people realized that confessing was the only way to avoid the gallows; indictments and death warrants; and petitions by friends and family members of the accused.

Of more than 200 people recording examinations and writing up legal documents based on boilerplate, few were trained in the law. They were literate members of Puritan society, with widely varying backgrounds and levels of education. They produced a written corpus unlike any contemporary one that exists—in part for the shocking nature of its contents, of course, but also for the way the conflagration devoured all levels of Puritan society, drawing in all manner of voices. Illustrated in the documents, according to University of Turku language historian Matti Peikola, are both the “religious, high-register language” of ministers like Samuel Parris and the Barbados-inflected English of a slave named Candy.


A creative approach to spelling can make the Salem texts befuddling for nonexpert readers. A record of testimony by the accused witch Tituba, for instance, reads in part: “shee allsoe said that shee seed a yalow burd that said unto hur sarue [serve] me and shee seed 2 catts.” The documents feature obsolete words and words that have shifted their meaning, including “silly,” to mean ignorant; “paragon,” a wool or silk fabric; “old cratten,” the devil; “burling,” meaning whirling or twisting; and “behaged,” meaning bewitched.

But the confusing spelling, especially from scribes or “recorders” with less formal training, is also the very thing that makes the writing so useful. Linguists believe that some of the spelling choices may be phonetic. Most of the Puritans had either been born in England or were first-generation settlers, and brought a wide range of regional English dialects with them. When linguist Henry Alexander wrote the first article on the language of Salem for American Speech in 1928, he was largely interested in demonstrating pronunciation variations. He points out, for instance, how the spellings “elle” for “ill” and “weches” for “witches” could demonstrate a pronunciation that parallels the English upper-class “well” for “will,” “cheldren” for “children,” and so on.

In the late 1990s and 2000s, a team of Swedish, Finnish, and North American linguists and historians took on the major project of re-editing the Salem manuscripts, emerging with a finished book in 2009, “Records of the Salem Witch-Hunt.” This time, they engaged the question of pronunciation with more awareness of the recorder’s role. “We can tell that certain features must have been part of the language in 17th-century America. What is problematic is telling whose language that actually was,” says Grund, who, along with Peikola, was involved in the manuscript project. Their book argues that the Salem documents demonstrate both a range of English dialects, and a “leveling” effect as the settlers adapt their speech to one shared community, fusing their language into a new, American tongue. One likely example is the word “her,” which is spelled a number of ways: “har,” “her,” “hir,” “hor,” or “hur,” suggesting that a range of distinct vowel sounds was now merging into one.


The team behind the 2009 book also gleaned other information about contemporary written language, such as handwriting style, capitalization, punctuation, and abbreviations. “People tend to think about abbreviations, especially in text-messaging language, as something that’s very modern,” Grund says. “But these people used abbreviations all the time.” Some of the abbreviations are intuitive shorthand, with the last letter written in superscript: “Sept” for “September,” for instance. Others are more specific: “goodw,” for “goodwife”; “ym” for “them,” from the medieval letter thorn, which resembled a y.

Studying these written clues, the Records group has helped identify 40 percent of the scribes, leading sometimes to disturbing historical insights. Piekola and his colleague Risto Hiltunen described in 2007, for example, how Salem parish clerk Thomas Putnam, father of one of the “afflicted” girls and a champion of the trials, transcribed multiple witness depositions using almost exactly the same wording, ending with, “I (verily) beleue in my heart that X is a (most dreadfull) wicth/wizard.” They suggest that Putnam was copying out a single deposition again and again and submitting them all as damning evidence to the court.


In 2012, Peikola, Grund, and independent historian Margo Burns came across documents in the Boston Public Library that were once thought lost—witness depositions from the trial of Margaret Scott, who was eventually hanged. In a paper published this summer, the scholars point out that a line at the end, “and i do uerily beliue that she is a witch,” was written with a finer-nibbed pen in slightly lighter ink—probably added later on by the scribe to add incriminating force to the statement. Literacy, in Salem Village in 1692, could be a deadly tool.

Britt Peterson is an Ideas columnist. She lives in Washington, D.C. Follow her on Twitter @brittkpeterson.