50, 100, 1000 words for snow: Does the language we speak affect how we think?

Kate Bush, a British singer-songwriter, released a new album this past Monday, titled 50 Words for Snow. The inspiration for this album title comes from the popular belief that the Eskimo language had many, many words for snow because the Eskimo people differentiated all the different types of snow they experienced. On this album, Kate Bush has a song with the same title, where she explores the question, “…if [the Eskimo language] had that many words for snow, did we?” (for the full interview, click here).
While it is now largely known in the academic community that the Eskimo language did not actually have that many words for snow–what they had a few words for snow that was modified in various ways (e.g., wet snow, icy snow, fluffy snow)–this new Kate Bush album has resurrected interest in popular questions about the relationship between language and thought: Does the language we speak affect how we think? Or do our thoughts affect our language?

The theories driving this popular question date back to the 1920s, when Edward Sapir proposed that the language we speak should impact our thoughts and perception of the world (Sapir, 1929). Sapir’s ideas were further developed by his student, Benjamin Lee Whorf, who, through his research on Native American languages (including the Eskimo language), concluded that the language we speak determines our thoughts and perception of the world, as well as our behavior (Whorf, 1956). The theories proposed and supported by Sapir and Whorf have come to be known by a few different names: The Sapir-Whorf Hypothesis, Whorfanism, and Linguistic Determinism.

Current research examining the relationship between language and thought no longer ask the question of whether language determines thought, but rather, focus on the ways in which language influences thought, or vice versa (Boroditsky, 2003). This new focus is aptly referred to as Linguistic Relativity, and much research has been conducted since Whorf’s time to investigate the effects of linguistic relativity. Researchers in psychology, linguistics, anthropology, and cognitive science have asked several types of questions to investigate how language affects thought or thought affects language. The following are a few of the most popular questions asked in this area of research.

Do speakers of different languages talk about time differently? In English, we most commonly use words referring to front vs. back to talk about the future vs. the past. We talk about certain people as being ahead of their time; we look forward to meeting a friend for lunch tomorrow; we fall behind schedule; and we think about going back in time. In contrast, speakers of other languages, such as Mandarin Chinese, use two types of time metaphors: (1) front vs. back metaphors and (2) up vs. down metaphors. Mandarin Chinese speakers use front vs. back metaphors in ways similar to English speakers; the up vs. down metaphors are reserved for talking about the order of events. In these up vs. down metaphors, up refers to earlier events and down refers to later events. Boroditsky (2001) examined whether speakers of English and speakers of Mandarin Chinese think about time in different ways. Her studies showed that English speakers were faster to correctly answer questions such as, “Does March come before April?” after seeing a row of items organized horizontally; on the other hand, Mandarin Chinese speakers were faster at correctly answering the same questions after seeing a column of items organized vertically. What these data suggest is that speakers of different languages think about time more easily when they are primed to think in the direction indicated by the time metaphors used in their language. Boroditsky thus argues that language shapes how speakers of different languages conceptualize time.

Do speakers of different languages talk about space differently? Speakers of English and Dutch use relative spatial terms to describe space (e.g., left, right, front, back). In contrast, speakers of Tzeltal (a Mayan language spoken in Mexico) use absolute references to describe space (i.e. uphill, downhill [corresponds roughly to English south/north directions]). A study by Levinson (1996) examined whether these linguistic differences between Dutch and Tzeltal influence how speakers of these languages conceptualize space. In this study, speakers of Dutch and Tzeltal were shown sets of arrows on two different occasions. On the first occasion, they saw two arrows pointing in the same direction (e.g., two arrows pointing right/north); the participants were then rotated 180 degrees to be shown another pair of arrows: one of which pointed in the same relative direction as the arrows from the first occasion (e.g., right [the absolute direction being south]), and the other pointed in the same absolute direction as the arrows from the first occasion (e.g., north [the relative direction being left]). What Levinson found when asking Dutch and Tzeltal speakers to identify the arrow that was “like the [arrow] that they saw before,” Dutch speakers selected the arrow that was pointing in the same relative direction (e.g., the one pointing right [the absolute direction being south]), while Tzeltal speakers selected the arrow that was pointing in the same absolute direction (e.g., north [the relative direction being left]). Thus, Levinson concludes that Dutch and Tzeltal speakers appear to be influenced by their respective native languages in understanding spatial orientation.

Do speakers of different languages categorize objects in different ways? Languages such as Spanish, French, and German have what is called grammatical gender, where words (especially nouns) are assigned gender. For example, in Spanish, the word for apple is feminine, whereas the word for dog is masculine. Other languages such as English or Japanese do not assign gender to words and therefore do not have grammatical gender. Researchers have thus asked whether speakers of grammatically gendered languages categorize objects in different ways from speakers of languages without grammatical gender. Sera, Berge, and del Castillo Pintado (1994) examined this question, by asking English- and Spanish-speaking children and adults to assign a female or male voice to various objects. Their study revealed that Spanish-speakers assigned gendered voices to the objects in ways that matched the Spanish grammatical gender of the word for the object (e.g., an apple would be assigned a woman’s voice because the word for apple in Spanish is feminine). In contrast, English-speakers assigned gendered voices to all the objects at random. This study thus provides evidence in support of the idea that language influences how speakers categorize objects.

Do speakers of different languages perceive color in different ways? Different languages around the world have different numbers of words for colors. Some languages have as little as two basic color terms (e.g., the Dani of Irian Jaya); other languages have five (e.g., the Berinmo of Papua New Guinea); other languages historically only have a handful of basic color terms but have since added new color terms (e.g., Japanese); and other languages have as many as twelve basic color terms (e.g., Russian). Differences in the number of color terms across languages has led researchers to ask whether speakers of different languages perceive color boundaries differently. Take for example, a rainbow: a rainbow is a spectrum of color with no distinct boundaries between the colors, but in the United States, we perceive the rainbow as being comprised of seven colors (i.e. red, orange, yellow, green, blue, indigo, and violet). As there are no specific color boundaries in a rainbow, would speakers of different languages with different numbers of color terms perceive a rainbow as being comprised of a different number of colors? Davidoff, Davies, and Roberson (1999) examined how color boundaries are perceived by English-speakers and the Berinmo of Papua New Guinea, by asking them to provide a name for 160 different colors on the color spectrum. In this task, English-speakers reported eight basic colors categories, while the Berinmo reported five basic color categories. Furthermore, where on the color spectrum the English-speakers and the Berinmo differentiated blues from greens were markedly different. Results from this study suggest that language influences how speakers perceive color boundaries.

With all this research, you may be wondering, “So, does the language we speak influence how we think, perceive, and behave? Or do our thoughts, perception, and behaviors influence our language?” And the answer is, we don’t know for sure. Though the literature reviewed here appear to paint a clear picture that language shapes thought, it is important to ask why these various patterns are found in these languages in the first place. For example, is it truly that Tzeltal speakers conceptualize space in terms of absolute direction because of their language? Or could there be cultural or environmental reasons that make Tzeltal speakers more sensitive to and aware of absolute direction? What about color perception among the Berinmo? Is there less variation in color in their environment that could be influencing them to perceive and categorize colors in ways different from the English-speakers in the UK?

Differences across languages tend to reflect historical differences in cultural practices and needs. Take culinary terms, for example (Lehrer, 1972). The English language has three terms to refer to different kinds of frying techniques: fry, sauté, deep fry. The French language has four terms to refer to different kinds of frying techniques: frire [fry], sauter [sauté], rissoler [similar to browning], fricasser [frying specific to meat and fowl]. And the Japanese language has three terms for frying techniques: yaku [fry], itameru [stir-fry], ageru [deep fry]. Though each of these languages clearly has a general term for “frying,” the number and types of frying techniques differs by language and the culinary history of that culture.

What is clear from the vast research literature on language and thought is that language and thought seem to interact in some way to influence how we understand and communicate about the world. This is an interesting and important new direction for the research on language and thought, as our society becomes increasingly global and has an increased need for clear communication across languages and cultures. With a better understanding of how language and thought interact to influence how we understand and communicate about the world, we may begin to better understand how to communicate across languages and cultures, in ways beyond simple translation, that allow us to understand the various nuances of different languages and cultures.


Borodistky, L. (2003). Linguistic Relativity. In L. Nadel (Ed.), Encyclopedia of Cognitive Science (pp. 917-921). London, UK: MacMillan Press.

Sapir, E. (1929). The Status of Linguistics as a Science. Language, 5(4), 207-214.

Whorf, B. L. (1956). Language, thought, and reality: Selected writings. J.B. Carroll, (Ed.). Cambridge, MA: MIT Press.