Why We Confuse Race and Ethnicity
A Lexicographer’s Perspective
On our evolving understandings of racial categorization and cultural identity.
By Kory Stamper • February 13, 2019
Dictionaries sometimes provide an opportunity for users to tell more about what certain words should mean as opposed to what they do mean. Take race and ethnicity. The online dictionary at Merriam-Webster allows users to leave comments on entries, and the most common comment by far on the entry for ethnicity is that people are looking it up to determine how it’s distinguished from race. The most common comment on the entry for race is essentially “Okay, but what is race, then?”
The most common comment by far on the entry for ethnicity is that people are looking it up to determine how it’s distinguished from race.
The reality is that the words race and ethnicity have a significant amount of overlap in terms of their general use. Race is the older word, dating back to the 1500s, and for most of its history, it referred to groups of people who shared a common ancestor, culture, or cultural marker (such as language or religion): “the English race,” “the Scottish race,” “the Jewish race.”
But starting in the late 1700s, physiologists and anthropologists began using the word race to refer to a more formalized categorization of people that was based on physical characteristics, not necessarily shared ancestry or culture. “Physical characteristics” included everything from skin color to head shape to perceived temperament and intelligence (both of which were thought to have a biological basis). Nineteenth-century anthropologists divided humanity up into anywhere from three to twelve categories and ascribed physical, psychological, social, and intellectual attributes to each category.
Just because a word gains a new meaning doesn’t mean that the old meanings go away.
But just because a word gains a new meaning doesn’t mean that the old meanings go away. By the start of the twentieth century, race referred to groups of people who shared a common ancestor, groups of people who shared a common culture or cultural marker, and the anthropological categories of people divided primarily by physical appearance. And while those meanings seem distinct enough presented in isolation, it could be hard to tell just which meaning of race was being used:
It is my prayer, it is my longing, that we may pass from this life together—a longing which shall never perish from the earth . . . . This prayer is also immortal, and will not cease from being offered up while my race continues. (Mark Twain, “Eve’s Diary,” The $30,000 Bequest and Other Stories, 1906)
English had furnished all the raw materials for a correction, and out of them was coined the word ethnicity. The Oxford English Dictionary has one citation, from 1772, for ethnicity, where it’s a translation of the Spanish etnicidad, but the word doesn’t appear again until the early twentieth century, in a book condemning the idea that common ethnic or cultural identity is deterministic of character or personality:
To regard every individual of an ethnic group as having primarily the characteristic nature of that group, as if affiliation with it invested him with a particular kind of ethnicity which then determined his nature, is contrary to the doctrine that each individual structure is primary. (Isaac B. Berkson, Theories of Americanization: A Critical Study, 1920)
Race was the preferred term—until the word began to get skunked.
Ethnicity is built off the much earlier ethnic, which was used from the 1700s onward as an adjective to refer to national affiliation; both words trace back to the Greek word for “nation.” But the term ethnicity didn’t take off right away. Race was the preferred term—until the word began to get skunked.
Skunked is the term that linguists use to refer to the process by which a formerly neutral word gains negative connotations and suddenly becomes fraught (or completely unacceptable) in general use. For the word race, lots of twentieth-century events and movements contributed to that skunking: Nazi atrocities bolstered by nineteenth-century anthropological ideas of “racial purity” and the fitness of the White race over other races; institutional structures that relied on the pseudoscience of “racial disparities” to separate society into “white” and “colored”; the various civil rights movements—like the NAACP, the National Congress of American Indians, UnidosUS (formerly National Council of La Raza), and the Japanese American Citizens League—that kept demanding we confront the realities of what it’s like to live in non-white skin in the U.S. The word race itself showed up more often in contexts that highlighted social problems: “race riot,” “racial discrimination,” “race relations,” “racial tensions,” “playing the race card.” Even today, while some people claim we live in a “post-racial” society, that nineteenth-century pseudoscience around race still affects our daily lives. For instance, a 2016 study published in the Proceedings of the National Academy of Sciences of the United States of America examines how ideas around the pain tolerance thresholds of White and Black patients—ideas that have their root in nineteenth-century concepts of race—continue to have an impact on Black people and how their pain is managed in a clinical setting.
While these nineteenth-century ideas around race have been challenged, we still continue to hash out what, exactly, constitutes race.
While these nineteenth-century ideas around race have been challenged, we still continue to hash out what, exactly, constitutes race. Nowhere is this tension more obvious than in the U.S. Census, which provides interesting (if somewhat behind-the-times) evidence for tracking the complexities of race and ethnicity. In the 230 years that the census has been running, race has expanded from three categories (free Whites, all other free persons, and enslaved people) to fifteen, including “other.” But in 1980, the U.S. Census began asking all respondents, regardless of how they answered the question on race, to identify whether they had Hispanic origins—categorizing it as an ethnicity, not a race. (Current studies by the Pew Research Center show that many census respondents who identify as Hispanic in origin consider that to be both their race and ethnicity.)
Lexically speaking, this one event seems to be the thing that nudged the word ethnicity into general use; since the 1980s, use of ethnicity has increased dramatically. And the word race? It has more volume of use than ethnicity, as you’d expect for a word with five hundred more years of established use, but in the last few years, its use has decreased. The Oxford English Dictionary’s usage note at the entry sums up the current state of race in reference to those divisions of humanity distinguished by physical characteristics:
In recent years, the associations of race with the ideologies and theories that grew out of the work of 19th-cent. anthropologists and physiologists have led to the word often being avoided with reference to specific ethnic groups. Although it is still used in general contexts, it is now often replaced by terms such as people(s), community, etc.
Lexical takes on race and ethnicity make the issue of what actually constitutes race and ethnicity seem much simpler than it actually is.
But lexical takes on race and ethnicity make the issue of what actually constitutes race and ethnicity seem much simpler than it actually is. In 2003, California Newsreel (in conjunction with PBS) broadcast a TV series called Race: The Power of an Illusion and asked four professors to tease out the differences between race and ethnicity. All four had different responses. Some felt race was a single unifying categorization based primarily on physical appearance while ethnicity was a cultural connection. Others felt that race was more an identifier of origin while ethnicity was a social, cultural, or linguistic bond. Others felt that race and ethnicity were both movable feasts and relied more on how power structures categorize and operate against people. In other words, some of the people groups that today are racially coded as White (and given the privileges of a White person) have been considered less than White or other than White in the past, particularly when anti-immigration sentiment was sweeping the nation. In more recent years, enough people have protested the Census Bureau’s reductive view of ethnicity that federal officials are considering combining the race question with the ethnicity question for the 2020 Census. That we know the two are somehow different but related is clear from the lexical side of things again: The most common use of ethnicity in print, and one of the most common uses of race in print, is in the phrase “race and ethnicity.”
The answer depends on who the speaker is talking to and why the listener is asking.
So when should you use race and when should you use ethnicity? A survey of the major dictionaries of English gives some basic guidance when talking generally about race and ethnicity. Most of them agree that the word ethnicity is most often used of a person’s cultural identity, which may or may not include a shared language, shared customs, shared religious expression, or a shared nationality (especially outside that nation’s borders). And most dictionaries agree that race is often used to describe one of several very broad categories that people are divided into that are biologically arbitrary yet considered to be generally based on ancestral origin and shared physical characteristics (especially skin color).
There’s one more thing that dictionaries tell us, though it’s mostly subtext and only apparent in qualifiers like often and generally and especially. Race and ethnicity as labels can change not just from speaker to speaker but from context to context. Someone born to Japanese parents in the Bay Area of California and raised in San Francisco may identify racially as Asian (a broad category based on ancestral origin and some shared physical characteristics) but ethnically as Japanese, American, Japanese American, or maybe even San Franciscan (a cultural identity that can include shared customs, religion, nationality, or language). Or none of the above. The answer depends on who the speaker is talking to and why the listener is asking.
Kory Stamper is a lexicographer and author. Her first book, Word by Word: The Secret Life of Dictionaries, is a tour of English and the dictionaries that record it. Her writing has appeared in The New York Times, The Washington Post, The Boston Globe, and The Times Literary Supplement.