One of the most controversial ideas put forth in linguistics is the idea of linguistic determinism. Also known as the Sapir-Whorf hypothesis, it states that people who speak different languages would also have different mental models of the world. The classical example is the Hopi language, which, according to Benjamin Whorf, had “no words, grammatical forms, construction or expressions that refer directly to what we call ‘time’.” So, because of this feature of their language, they are presumably less concerned with past, present and future, and more concerned with a “cyclic” view of time. This idea has gained a lot of attention since its first publication, especially in popular science magazines.

But despite its popular appeal, this hypothesis has not yet been proven. In one experiment – among many of the same kind – Tzeltal speakers, whose language does not have words for “right” or “left”, were blindfolded, spun in a chair and then asked to retrieve a coin that had been in one of two boxes. Their performance was the same both when boxes were placed on the floor (a geocentric reference) and when the boxes were placed on the end of a beam that was attached to the chair (an egocentric reference). In other words, the lack of right/left distinction in the language does not impair their ability to make that distinction in the world. This experiment was designed to refute an earlier one, which suggested the opposite, but had some methodological flaws.

On the other hand, some studies show a deep influence of language on thought. One of them suggests that people estimate time duration differently, according to the spatio-temporal metaphors found in their respective languages. For example, English speakers, who normally conceive time as a distance ("a long time ago", "a short interval", etc), are better at estimating how much time a growing line projected on a computer screen takes to reach the other side. Conversely, Spanish speakers, who normally conceive time as a quantity (“mucho tiempo”, etc) score higher when estimating the time it takes for a “container of water” to fill up on the screen. The important thing here is that both stimuli and responses are non-linguistic.

The influence on thought suggested by this last study is, of course, very small. But at least, it shows an example of influence of the “strong” type, that is, linguistic features affecting non-linguistic behavior. Is there further evidence in support of this type of strong determinism?

  • 2
    I think this question is too broad-scoped. As the FAQ says, if you can imagine an entire book that answers your question, you’re asking too much.
    – Louis Rhys
    Oct 6, 2011 at 14:32
  • 2
    There's no need for a whole book. Examples of recent studies that support the hypothesis would suffice. Oct 6, 2011 at 14:44
  • 2
    I agree with @Louis Rhys, this is an entire sub-field of psycholinguistics, and is more likely to trigger a debate than lead to a constructive answer.
    – Alek Storm
    Oct 7, 2011 at 0:33
  • 1
    If you want to repair a S-W question, I'd recommend narrowing the scope to a particular corner of S-W research, such as a question about geocentric/egocentric languages. I don't know what's interesting so I wouldn't be able to suggest a good question, but something like "What sort of empirical experiments have been/could be done to illustrate difference in the psychological make up of speakers of ego/geo-centric languages?" That has an answer. S-W in general is an invitation to chat about stuff with more basis in gut feeling than science. Oct 7, 2011 at 16:28
  • 2
    @OtavioMacedo, How about just changing the title of the question? It leads off in a very provocative way, but by the time one reads to the bottom, what you seem to be asking is for examples of recent psycholinguistic studies, which is a very well-scoped question IMO.
    – Aaron
    Oct 7, 2011 at 20:01

5 Answers 5


The strong form of the Sapir-Whorf hypothesis has been discredited on several grounds after its initial formulation. However, recent decades have seen interest spring up in investigating questions about how language interacts with other aspects of cognition. In a very recent paper on this topic, Lila Gleitman and Anna Papafragou write: "In general, both logic and currently available evidence suggest a disclamatory view of strongest proposals (e.g., Benjamin Whorf, 1956) according to which particulars of certain human languages are important progenitors of thought, such that elements of perception or conception would be permanently altered by learning one or another language. However, several credible lines of experimental and developmental evidence suggest significant influence of linguistic representation during on-line processing in many cognitive and perceptual domains" (doc here). A really fantastic philosophical investigation into the cognitive role of language can and should be read, here. (Note that this latter reference includes 30 pages of open peer commentary, which is then responded to by the author.)


We have to be careful what theory we are discussing here.

Strong Determinism is the classic form of the Sapir-Whorf hypothesis. If you don't have the words, you don't have the thoughts; or, the words force our thoughts. This has never been taken seriously in the linguistic community. For one thing, proponents of this formulation usually embarrass themselves when talking about a particular language. The classic story is the Eskimos have 300 words for snow, but that simply isn't true. While it is true they have a lot of words to differentiate snow, so does English: slush, ice-pack, snow, snow-banks, snow-flakes, snow-drifts... You may protest and say that "snow drift" isn't a word, but once you drop compounds, you end up shooting yourself in the foot with regards to a lot of the words you counted up for the Eskimos. I'll leave it to someone who is familiar with the Hopi language to address the assertion that they can't talk about time.

You can weaken the hypothesis in a number of ways, and this is what people end up having to do. The problem becomes articulating how language shapes thought, and there are a lot of possibilities. Some people thought that speakers of gendered languages would be better at keeping track of gender, or speakers of languages which incorporated distance or social status or whatever into the verb would be better at tracking those things. That has turned out to be not the case. A recent twist on this is asks speakers of gendered languages to describe objects in terms stereotypical of that objects grammatical gender. Look up Lera Boroditsky.

Finally, I'd like to add that there is an intriguing alternative to the Sapir-Whorf hypothesis. The general idea behind linguistic determinism is that we think in language. This idea has been around for a long time; Plato described thought as speaking to oneself. The thought goes, if you think in language, then the language can shape the thoughts. Kinda like the structural imperfections of marble playing a part in the sculpting of a statue. The general problem with this is that there are tons of cognitive processes that don't involve "thinking" in the way we usually use that term.

The alternative is that instead of using language to think, language users use language as a tool to aid in thinking, especially in communicative contexts. Look up Dan Slobin, who introduced this idea.


Forgetting all the research in to the Sapir-Whorf hypothesis, the answer is both obviously yes and obviously no, all depending on how you interpret 'language', 'shape' and 'thought'.

I have a hard time thinking about building supplies because I don't have the vocabulary to express all the intricacies of hardware. So obviously yes, the hardware language allows a new way of thinking. (but then what is causing what here: I can have thoughts about things without the vocabulary items, the latter just make it easier and quicker, and beside which came first, the words or the thoughts).

And on the other hand obviously no, language doesn't force thought, one can translate (once experienced from one language to another, one can be bilingual and converse the same thoughts in two different languages. But, it may take one longer in one language than another, or the experiences and cultured learned with one language may not have corresponding things to translate to in the other.

But to balance these out properly, I think one can say (without cynicism) that the Sapir-Whorf hypothesis is superficially true: when you first visit a new language it always feels like you have to -think-differently to use it properly, and that that is a reflection of and caused by the language itself. But with experience, one can translate from one to the other.

(and then there is all the experimental evidence that strong Sapir-Whorf is unfounded and even weak Sapir-Whorf has problems).


The existence of lexical fields (e.g. categories like "fish you would eat" vs. "fish you would keep as a pet") and the creation of new words in a given language shows that linguistic determinism wildly overstates the influence of language on thought. That said, we shouldn't underestimate the effect that a given category can have on someone's thought patterns, either.


A more recent neophyte article on linguistics (which I'm sure many of you will have read by now) cites Sapir-Whorf:

"Utopian for Beginners: An amateur linguist loses control of the language he invented." by Joshua Foer. The New Yorker, December 24, 2012.

  • Having skimmed the article, it seems like they only briefly talk about the Sapir-Whorf hypothesis and only as an illustration of amateurs making an impact on Linguistics. While the article is quite interesting, I don't think it's very relevant to this particular discussion.
    – acattle
    May 1, 2013 at 10:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.