The Whorfian hypothesis
The claim that the structure of a language influences how its users view the world is today most usually associated with the linguist Sapir and his student Whorf, a chemical engineer by training, a fire prevention engineer by vocation, and a linguist by avocation. However, it can be traced back to others, particularly to Humboldt in the nineteenth century. Today, the claim is usually referred to as ‘Linguistic Determinism,’ the ‘Linguistic Relativity Hypothesis,’ the ‘Sapir–Whorf Hypothesis,’ or the ‘Whorfian Hypothesis.’ We will use the last term since the claim seems to owe much more to Whorf than to anyone else.
Sapir acknowledged the close relationship between language and culture, maintaining that they were inextricably related so that you could not understand or appreciate the one without a knowledge of the other. Whorf took up Sapir’s ideas but went much further than saying that there was merely a ‘predisposition’; in Whorf’s view the relationship between language and culture was a deterministic one; the social categories we create and how we perceive events and actions are constrained by the language we speak. Different language users will therefore experience the world differently insofar as the languages they speak differ structurally.
One claim is that if users of one language have certain words to describe things and users of another language lack similar words, then users of the first language will find it easier to talk about those things. We can see how this might be the case if we consider the technical vocabulary of any trade, calling, or profession; for example, physicians talk more easily about medical phenomena than those without medical training because they have the vocabulary to do so. A stronger claim is that, if one language makes distinctions that another does not make, then those who use the first language will more readily perceive the relevant differences in their environment. If you must classify camels, boats, and automobiles in certain ways, you will perceive camels, boats, and automobiles differently from someone who is not required to make these differentiations. If your language classifies certain material objects as long and thin and others as roundish, you will perceive material objects that way; they will fall quite ‘naturally’ into those classes for you.
This extension into the area of grammar could be argued to be a further strengthening of Whorf’s claim, since classification systems pertaining to shape, substance, gender, number, time, and so on are both more subtle and more pervasive. Their effect is much stronger on language users than vocabulary differences alone. The strongest claim of all is that the grammatical categories available in a particular language not only help the users of that language to perceive the world in a certain way but also at the same time limit such perception. They act as blinkers: you perceive only what your language allows you, or predisposes you, to perceive. Your language controls your worldview. Users of different languages will, therefore, have different worldviews. (This idea also appears in popular culture; for instance, we can see it in a strong form in the science fiction movie Arrival, in which those who learn the aliens’ language also acquire the ability to see the future and past.)
Whorf’s work on Native American languages led him to make his strongest claims. He contrasted the linguistic structure of Hopi with the kinds of linguistic structure he associated with languages such as English, French, German, and so on, that is, familiar European languages. He saw these languages as sharing so many structural features that he named this whole group of languages Standard Average European (SAE). According to Whorf, Hopi and SAE differ widely in their structural characteristics. For example, Hopi grammatical categories provide a ‘process’ orientation toward the world, whereas the categories in SAE give SAE users a fixed orientation toward time and space. In SAE, events occur, have occurred, or will occur, in a definite time, that is, present, past, or future; to users of Hopi, what is important is whether an event can be warranted to have occurred, or to be occurring, or to be expected to occur. Whorf believed that these differences lead users of Hopi and SAE to view the world differently. The Hopi see the world as essentially an ongoing set of processes; time is not apportioned into fixed segments so that certain things recur, for example, minutes, mornings, and days. In contrast, users of SAE regard nearly everything in their world as discrete, measurable, countable, and recurrent; time and space do not flow into each other; mornings recur in twenty‐four‐hour cycles; and past, present, and future are factual ways of viewing events. (We should note that Malotki (1983) has pointed out that some of Whorf’s claims about the grammatical structure of Hopi are either dubious or incorrect, for example, Hopi, like SAE, does have verbs that are inflected for tense.)
Deutscher (2010a, 2010b) has revisited the Whorf hypothesis, noting some of the obvious problems with this hypothesis: ‘If the inventory of readymade words in your language determined which concepts you were able to understand, how would you ever learn anything new?’ However, he further discusses some recent research which provides evidence for the connection between language and worldview. One example is that users of a remote Australian aboriginal tongue, Guugu Yimithirr, from north Queensland, do not make use of any egocentric coordinates (i.e., deictic words such as ‘left,’ ‘right,’ ‘behind,’ ‘in front of’) but instead rely solely on the cardinal directions of east, west, north, and south. Research on this language prompted recognition of the same phenomenon in languages of other far‐flung places such as Bali, Namibia, and Mexico. Deutscher uses this research not to make strong claims about linguistic determinism, but to urge readers to recognize linguistic relativity, advising us that ‘as a first step toward understanding one another, we can do better than pretending we all think the same.’
More recently, McWhorter (2014) argued against the Whorf hypothesis, with a book written for a popular audience titled The Language Hoax: Why the World Looks the Same in Any Language. His main point seems to be that the differences between groups are cultural, and not determined by language. This book has not been well received by linguistic anthropologists, as can be seen in Webster’s (2015) review. Nonetheless, a shared perspective seems to be a rejection of the neo‐Whorfian interest of experimentally testing differences of perception across users of different languages. McWhorter’s point is that if there is an effect of language on thought, it is minor; Webster’s is that different languages have different poetic potentials and this, and not some abstract idea about ‘thought,’ is what is interesting about the influence of language on our worldview. In the next section, we will move on from the question of the direction of influence to the discussion of the nature of connections between linguistic features and social variables in sociolinguistic research.
Exploration 1.4 Translatability
If you speak more than one language or dialect, are there certain words or phrases which you feel you cannot translate into Standard English? What are these words or phrases – are they simply words for things which are not part of the cultures of the English‐speaking world, or concepts or idioms not found in English? What does the view of particular words as ‘untranslatable’ indicate about the connection between language and worldview?
Correlations
It is possible to claim a relationship between the use of certain features of language and social structure, and such correlational studies have long formed a significant part of sociolinguistic work. Gumperz (1971, 223) has observed that sociolinguistics is an attempt to find correlations between social structure and linguistic structure and to observe any changes that occur. The approach to sociolinguistics which focuses on such correlations and the quantitative analysis of them is often called variationist sociolinguistics, and the theory and methodology of this will be discussed in chapter 5.
It is important to note that correlation only shows a relationship between two variables; it does not show causation. To find that X and Y are related is not necessarily to discover that X causes Y (or Y causes X). For example, to find that female language users