Múra-Pirahã: The Language without Numbers
If you’ve read my other posts, you might remember a discussion of the Sapir-Whorf hypothesis, one of the most controversial I’ve ever come across. In its current form, it has two main schools of thought: linguistic relativity and linguistic determinism. The former, and much more highly-regarded, essentially argues that our cognitive structures and the ways we categorize experience into language has an influence on those experiences and how we perceive them. The latter states that those linguistic processes fully determine our experience. With this post, I want to take a deep dive into the more rigid half of this theory and explore the (often-refuted) evidence for linguistic determinism and what the ideas can teach us about the real function of language.
The most famously-cited “proof” of linguistic determinism (later championed by Whorf himself) comes from a anthropologist named Franz Boas. He presents the fact that several Eskimo languages have more than one word for snow while in English most speaker’s have one. These words are single thoughts that encapsulate certain features of the snow such as if the snow is on the ground, if it is drifting, if it is good for packing, etc. whereas, in English to show these ideas we have to use phrases to specify the type of snow. In contrast, the words in Eskimo languages are not types of snow. Just as English speakers have clear differences between a drizzle, a shower, and a downpour, the Eskimo languages have broken out of categorizing a light snow or fallen snow as cousins rather than siblings, belonging to different families while still having a strong relationships to one another.
A supposed list of various “snow” words in the Eskimo languages. The list is misleading because it has words from all different related languages. It would be like listing the words for “water” in the all the Romance languages and comparing them. Additionally, there are words like avalanche and blizzard and slush, which are different words completely unrelated to snow in English as well.
The presence of several words for snow in Eskimo languages doesn’t prove linguistic determinism. If anything, it only provides us with a perspective that classifies physical occurrences differently than the standard. This is a powerful perspective because it allows us to rethink the structure of more common linguistic features as less universal as we once thought. But, we would be foolish to think that this proves that language determines experience in terms of the way we relate to the world. Those who speak Eskimo languages don’t lack an underlying concept of what “snow” is or how different types of snow are related to one another. It can be likened to the way we label bodies of water based on their movement, speed, location, and size. A river is different from a stream which is different from a brook, all of which are even more different from a lake which is about the farthest thing from a waterfall.
Even though we don’t recognize it with most of the names, labeling specific types of water doesn’t make us oblivious to the fact that it is all water. It’s like flowers. We could have chosen to name them all the same (and we can still refer to any flower as just that, a flower) but most languages have different words for different types. Based on what we are surrounded by, what we find compelling or useful, or even just from pure observation, different languages, individuals, and communities (and even varieties within those languages) can classify the world differently. It doesn’t make any of these classifications correct or incorrect, but establishing them does allow for us to use them more in our linguistic endeavors. In other words, by making a word for something, we can more easily talk about it. English doesn’t have an equivalent for schadenfreude, but this doesn’t mean we weren’t familiar with the feeling before adopting the word. Using the word, however, does make it easier to communicate a specific idea (along with its associations) more concisely, perhaps the reason why it is now commonplace among English-speakers.
But there is one case of supposed evidence for linguistic determinism that I can’t stop thinking about. There is a language called Múra-Pirahã spoken in an area on the Amazon river in Brazil by less than two hundred people and somehow it doesn’t have numbers. It was originally thought that there were numbers for one and two, until additional tests were performed with native speakers that determined that the words actually label more abstract concepts, most equivalent to “small amount” and “large amount.” While “small amount” often means one, with “large amount” meaning more than one, it depends on the items being discussed and what the usual groupings of them are. Additionally, every speaker has different ideas of what constitutes a “large” and “small” amount.
Words meaning “small amount,” “large amount,” and “many”
To make it more complicated, there is another word translated as “many” which more accurately means “coming together as one.” This word is used when the group of items is too numerous to be seen as made up of individual parts. Usually, the words for “small amount” will be used for items of one, two, three, “large amount” for four, five, six, seven, and more than this is seen as “many,” but depending on if a group of items is getting smaller or larger, this might change. For example, one speaker labeled a group of four items as “large amount” when they were being added to a pile, but when starting at ten items and subtracting from the pile, six items became the “small amount.” This is evidence for the fact that these concepts of “small,” “large,” and “many” are not exact amounts, rather, they are constantly defined by their relationship to the surroundings and current events.
While it is incredibly hard to think about not being able to count to ten, speakers of Pirahã, for the most part, can’t do this, not because they don’t have words for numbers, but because they aren’t conditioned to think about items and their surroundings in exact amounts. Even when trying to use numbers 1-10 in other languages like Portuguese, they have an astounding amount of difficulty. This begs the question: does language allow for us to conceptualize exact amounts? Would we be unable to think about six versus seven if we didn’t have the words for them? Look at the grids below and try to understand how someone would fill in the one on the right to match the left without using counting. How would you make them match without using numbers (even in your mind)? This is essentially the same test the Pirahã speakers were given, something which they were able to do almost perfectly, all without using numbers and solely through observation.
One way that you would be able to do it is by abstracting it. While most of us would probably place the pieces one by one because we can tell that the first piece is in the second row and second column and it would likely be our first choice, a speaker of Pirahã would probably be more interested in getting all of the pieces on the board before worrying about specific places. It seems that they are much more attuned to think about the overall shape and the relationship between the pieces themselves rather than the perfect placement of the parts. If we divide the “object” on the left into smaller parts, we see a different relationship between the pieces.
This isn’t the only way to divide the “object,” but regardless, it highlights the distance and directionality between each piece and those surrounding it. While counting the pieces and placing them one-by-one might more accurately transfer the picture, this second method highlights the dimensions of the object and its inner connections in order to recreate it. It is a different way of using the brain and conceptualizing a task, something which really complicates this notion of linguistic determinism.
Language doesn’t compose the entirety of our thoughts. Even though it often feels like thinking is only just talking to ourselves in our head, it isn’t so simple. If I told you to close your eyes, picture a chair in your mind, and then turn that chair around in your mind, those thoughts you are having are images, not words. You are not saying to yourself, “okay, now take the first leg in one hand and the back in the other and then pick it up and turn it,” or “walk around the chair to see the other side.” Your mind pictures and turns the object and, just like you don’t need to tell yourself to walk around to the other side of the chair in real life, you do this instinctively. You can picture the back of that imagined chair because you know what it looks like. I used language to make you think of a chair, but you didn’t use language to move it in your head. We often think of language as being all there is in our heads when realistically there is so much more; language is imperfect in representing our emotions, experiences, observations, and desires. We have spent so long trying to define concepts like “love” but maybe we can’t come to a conclusion because language itself has its limits. Labeling something usually requires us to label what it is not, a process of exaction that Pirahã speakers often lack when considering (rather than counting) objects.
To make myself clear, I am not a linguistic determinist. The theory that language constructs our reality is laughable in most cases, but I do think that it is powerful to challenge the idea of what language allows us to see as “objective” or not. Perhaps the idea of exactness is not inherent to us and it is the labeling of it that allows us to consider it more. Maybe the more we consider something that isn’t inherent to us, the more it feels natural, the more it feels personal, the more it feels like us.
Pirahã Children. The people call themselves the Hi'aiti'ihi.
Language is a tool that allows us to reinforce ideas that we don’t always naturally tend towards. When we rely so heavily on language, we can forget it is of our own creation. It is peculiar that numbers allow us to work with mathematics in a super specific way, something which speakers of Múra-Pirahã on the whole cannot do. It is the most compelling evidence for linguistic determinism simply because of the fact that it raises more questions that it seems to answer. Whereas the “snow” words of the Eskimo-languages can be easily bypassed as concrete evidence, the lack of numbers in Pirahã is not such a simple process. It leads to a whole conversation about how natural exactness is to us and in what ways language allows us to work with it. How can language blind us to the endless amount of nuance in day-to-day life? Maybe numbers are the “true lens” through which we can see the world, or maybe they reveal only certain truths while turning us away from others.
Most of the information for this post was found in this paper. Thanks for reading!
