The Sapir-Whorf Hypothesis

Edward Sapir & Benjamin Lee Whorf

Human beings do not live in the objective world

alone, nor alone in the world of social activity as
ordinarily understood, but are very much at the
mercy of the particular language which has become
the medium of expression for their society. It is quite
an illusion to imagine that one adjusts to reality
essentially without the use of language and that
language is merely an incidental means of solving
specific problems of communication and reflection.
The fact of the matter is that the "real world" is to a
large extent unconsciously built up on the language
The theory proposes that the way
people think is determined by the
structure of their language. Different
languages represent different social
realities. This hypothesis claims that
people are prisoners of their languages,
and so speakers of different languages
perceive the world differently.
Linguistic Determinism refers to the idea that
the language we use to some extent
determines the way in which we view and
think about the world around us.
Strong Determinism

Strong determinism is the extreme version

of the theory, stating that language actually
determines thought, that language and
thought are identical.
Weak Determinism

Weak determinism holds that thought is

merely affected by or influenced by our
language, whatever that language may be.
 Wilhelm von Humboldt: The
“Weltanschauung” Hypothesis

Wilhelm von Humboldt (1767-1835) was the first

European to combine a knowledge of various
languages with a philosophical background; he
equated language and thought exactly in a
hypothesis we now call the “Weltanschauung”
(world-view) hypothesis, in fact a version of the
extreme form of the Sapir-Whorf hypothesis.

Language = Thought
 Humboldt maintained that language
actually determined thought.

Humboldt viewed thought as being

impossible without language, language as
completely determining thought.
On closer inspection, we can see that this
extreme hypothesis leads to a question:

How, if there are no thought before

language, did language arise in the first
place?
Humboldt answers this by adhering to the theory that
language is a platonic object, comparable to a living
organism which just suddenly evolved one day
entirely of its own accord.
Linguistic Relativity: A Definition

Linguistic relativity states that distinctions

encoded in one language are unique to that
language alone, and that “there is no limit to
the structural diversity of languages”.
humans and most other animals use three color-receptors
to see the visible portion of the electromagnetic spectrum.
To observe the visible spectrum, light receptors within the
eye transmit messages to the brain, which then produces
color (Color intelligence, 2020).

there is no wavelength of light that corresponds to that

particular color. Rather, it is physiologically and
psychologically perceived as a mixture of red and blue. So
technically, magenta doesn’t exist.

https://scholarblogs.emory.edu/artsbrain/2020/12/02/mag
enta-doesnt-exist/
If one imagines the colour spectrum, it is a continuum, each
gradually blending into the next; there are no sharp
boundaries. But we impose boundaries; we talk of red,
orange, yellow, green, blue, indigo, and violet. It takes little
thought to realise that these discriminations are arbitrary - and
indeed in other languages the boundaries are different.
23
In neither Spanish, Italian nor Russian is there a word
that corresponds to the English meaning of “blue”, and
likewise in Spanish there are two words “esquina” and
“Rincon”, meaning an inside and an outside corner,
which necessitate the use of more than one word in
English to convey the same concept.
× 不同的語言和文化群體，劃分色譜
也截然不同。一些語言，比如如巴
布亞新幾內亞所說的丹尼語，利比
里亞和塞拉利昂所說的巴薩語，顏
色只有兩個詞，"暗"和"亮"。"暗"大
致可譯為其他語言中的冷色調，"
亮"為暖色調。所以黑、藍、綠等顏
色可解釋為冷色，而白、紅、橙、黃
等淺色調可解釋為暖色。
× https://www.bbc.com/ukchina/trad/vert-fut-43948821

25
瓦爾皮裏人生活在澳大利
亞北部，這些土著的語言
裏甚至沒有"顏色"一詞。
對於類似這種文化的人類
群體而言，我們所謂的"顏
色"，他們是用各種指代材
質、身體感覺和功能目的
的詞來描述。
https://www.bbc.com/ukchina/trad/vert-fut-43
948821

26
We should have more or less the
same experience with colors, but
what if we speak in different
languages?

27
28
Same experience?
× Same perception？
× Limitation of our experience:
No language includes two color spaces (e.g. yellow, blue) but exclude
connecting color space (e.g. green)

29
 Kay and Kempton
(1984)
Target:
native speakers of English
Speakers of Tarahumara (a
Uto-Aztecan language of
northern Mexico)
Mission:
To judge the perceptual distance
among three color chips.
30
Chiu, C. (2011). Language and Culture. Online Readings in
Psychology and Culture, 4(2).

https://doi.org/10.9707/2307-0919.1098
 × Tarahumara do not have
lexical term to distinct blue
and green.
× English speakers
overestimated the
distance between two colors
Chiu, C. (2011). Language and Culture. Online Readings in
Psychology and Culture, 4(2).
https://doi.org/10.9707/2307-0919.1098

31
Naming strategy
“When presented with two colors that fell in the
green category and one color that fell in the blue
category, they might have labeled the two greener
colors ‘green’ and the bluer color ‘blue’”

However, Tarahumara-speaking participants cannot

use this strategy

Chiu, C. (2011). Language and Culture. Online Readings in

32
Psychology and Culture, 4(2).
https://doi.org/10.9707/2307-0919.1098
Cancelling effect
× Participants were × Participants are × The
shown a greener required to experimenters
color, but evaluate the believe that the
experimenters distance effects of
named it ”blue” between colors. linguistic
color encoding
cancelled out
× This time English each other,
× And then they speaker did not
named a bluer color overestimate the
“green” color distance.
33
Chiu, C. (2011). Language and Culture. Online Readings in
Psychology and Culture, 4(2).
https://doi.org/10.9707/2307-0919.1098
American-Indian Language -- Zuni

Monolingual speakers of an American-Indian

language called Zuni - a language which does not
recognise any difference between yellow and
orange - had more difficulty in re-identifying
objects of such colours after a period of time. With
monolingual English speakers, this difficulty is
absent, since we make a verbal distinction.
This only offers support for the weak
version of the hypothesis, though,
because it would be wrong to say that the
Zuni speakers did not actually perceive a
difference. So the more highly codable a
concept is, the easier it is to retrieve from
the unconscious.
I find it gratuitous to assume that a Hopi who knows only the
Hopi language and the cultural ideas of his own society has the
same notions, often supposed to be intuitions, of time and
space as we have, and that are generally assumed to be
universal. In particular he has no notion or intuition of time as a
smooth flowing continuum in which everything in the universe
proceeds at an equal rate, out of a future into a present and
into a past .... After a long and careful analysis the Hopi
language is seen to contain no words, grammatical forms,
construction or expressions that refer directly to what we call
'time', or to past, present or future ...
— Whorf (1956)
Hopi Indian of Arizona
The Hopi language structures the notion of time
differently from Western time.
Whorf spent a lot of his time studying the language of
the Hopi Indians of Arizona, who make no distinction in
their language between past, present and future tenses;
where in English it seems natural to distinguish between
“I see the girl”, “I saw the girl” and “I will see the girl”, this
is not an option in Hopi. This apparently made quite an
impression on Whorf, who imagined that the scientists of
the day and the Hopi must see the world very
differently...although the philosopher Max Black
considers that “they may be expected to have pretty
much the same concept of time that we have” in spite of
this. And Whorf himself notices, “The Hopi language is
• Metaphor is not a figure of speech
which produces rhetorical effect
but it is a way for us to
understand the world
• E.g. 努力向上、下流, look forward,
big person, find your way

44
Does Language Shape Thought?: Mandarin and
English Speakers’ Conceptions of Time
--Lera Boroditsky
× time is generally conceived as a
one-dimensional, directional entity.

× Across languages, the spatial terms

imported to talk about time are also
one-dimensional, directional terms such
as ahead/behind or up/down rather than
multidimensional or symmetric terms
such as narrow/wide or left/right

× Example: they are looking forward to a

brighter tomorrow, proposing theories
ahead of their time, or falling behind
schedule,
• Horizontal
• In English, we predominantly
use front/back terms to talk
about time.
• good times ahead of us or the
hardships behind us.
• move meetings forward, push
deadlines back, and eat
dessert before we are done
with our vegetables.

46
Chinese

× Horizontal × In summary, both

× qia ́n (‘‘front’’) Mandarin and English
speakers use horizontal
× and ho`u (‘‘back’’) terms to
× talk about time. In
× Vertical addition, Mandarin
× spatial speakers commonly use
morphemes the vertical
sha`ng (‘‘up’’) × terms sha`ng and xia`.
× and xia` (‘‘down’’)

47
× After solving a set of two × Experiment was
primes, participants designed to test
answered a TRUE/FALSE whether using spatial
target question about time. metaphors to talk
× Half of the target questions about time can have
were designed to test the both immediate and
immediate effect of long-term
metaphors on processing implications for how
and so used a horizontal people think about
spatiotemporal metaphor time.

× Question:
× Spatiotemporal terms
(before after) and
purely temporal
terms
× (earlier and later)
48
49
× Half of the target × If one’s native language does have a
questions were long-term effect on how one thinks about
designed to test the time, then Mandarin speakers should be
immediate effect of faster to answer purely temporal target
metaphors on questions (e.g., ‘‘March comes earlier than
processing and so April’’) after solving the vertical spatial
used a horizontal primes than after the horizontal spatial
spatiotemporal primes. English speakers, on the other
metaphor (e.g., hand, should be faster after horizontal
‘‘March comes before primes because horizontal metaphors are
April.’’). If predominantly used in English.

50
Result
× As predicted, English × As predicted, English
and Mandarin speakers speakers answered
were affected purely temporal
differently by the spatial questions faster after
primes. Both English and horizontal primes
Mandarin speakers than after vertical
answered primes.
spatiotemporal × Mandarin speakers
before/after questions were faster after
faster after horizontal vertical primes than
primes than after after horizontal
vertical primes primes.
51
 Result
?
The tribe's language
determines its people
thought.
- members of the Piraha tribe in remote
northwestern Brazil use language to express
relative quantities such as ”one”, ”two”, "some"
and "more," but not precise numbers.

- they started with 10 objects and asked the tribe

members to count down. In that experiment, the
tribe members used the word previously thought
to mean "two" when as many as five or six
objects were present, and they used the word for
"one" for any quantity between one and four.
This indicates that "these aren't counting
numbers at all," said Gibson. "They're signifying
relative quantities."
https://news.mit.edu/2008/language-0624

55
Numbers in language

Reading/writing
sound Numeral system
direction

56
Counting system in English
× Number in language: number is also
“language”
× eleven、twelve and thirteen、fifteen
× Twenty-one and fourteen ()
× Forty, sixty (similar to four and six ) and fifty
and thirty (less similar to five and three)
× Million and billion
× in French 92 is quatre-vingt douze，namely,
four twenties and twelve
× 一，二，三，十一，十二，二十，六十
× Does counting system make some culture
easier to calculate and remember numbers?
57
• in the modern Welsh system, 92 is naw deg dau, or
“nine ten two”, much like the system used in East Asian
languages. In the older, traditional system
• In Wales today, about 80% of pupils are taught maths
in English, but 20% are taught in modern Welsh
• Six-year-old children taught in Welsh and English were
tested on their ability to estimate the position of
two-digit numbers on a blank number line, labelled “0”
on one end and “100” on the other. Both groups
performed the same on tests of general arithmetic but
the Welsh children did better on the estimation task.
• the Welsh medium children had a somewhat more
precise representation of two-digit numbers
× https://www.bbc.com/future/article/20191121-why
-you-might-be-counting-in-the-wrong-language

58
× in Dutch, 94 is written vierennegentig (or “four
and ninety”
× For example, Dutch kindergarten children
performed worse than English children on a
task that required them to roughly add
together two-digit numbers.
× Xenidou-Dervou’s explanation is that when the
children see a number like 38 on the task, they
vocalise it internally, and then picture its
position on a mental number line. In Dutch, the
extra mental step of having to un-invert the
number “eight and thirty” before they can
assess its value creates extra cognitive strain,
and this affects their performance.
× https://www.bbc.com/future/article/2019112
1-why-you-might-be-counting-in-the-wrong-la
nguage

59
 The Notion of Codability (可編碼程
度)
Codability has been defined by Peter Herriot as
“the ease with which a language tag can be used to
distinguish one item from another”. (Something is
codable if it falls within the scope of readily
available terms used in whatever particular
language.)
Degrees of codability vary, in that while one
language may be capable of expressing a concept
with just one word, in another may be necessary to
use a whole phrase to get across the same notion.
The Notion of Translatability

Although different languages may have different

ways of dividing up their spectra of experience
into verbal forms, we find it is still quite possible
to translate from one language into another. (We
are doing translation, aren’t we?)

Although someone translating from one

language into another may find it necessary to
use a whole phrase in the target language to
communicate the concept expressed in the
original language with only a single word, this is
achievable.
qanik snow falling
aputi snow on the ground
pukak crystalline snow on the ground
aniu snow used to make water
siku ice in general https://www.thecanadianencyclopedia.c
nilak freshwater ice, for drinking a/en/article/inuktitut-words-for-snow-and
qinu slushy ice by the sea -ice
Australian aboriginal language: Pinupti

In the Australian aboriginal language Pinupti,

the word “katarta” refers to the hole left by a b
when it has broken the surface of its burrow
after hibernation. It takes seventeen words to
translate that concept into English, but the
result is fine, lacking perhaps some of the
conciseness but none of the subtlety of the
Pinupti word.
[Question: Is it difficult to produce a Chinese
version of the term “katarta”?]
And if we look just a little further, it
becomes obvious that if it was true
that language dictated thought, and
that concepts were untranslatable,
then children would be incapable of
learning language at all; for how
would a child learn its first word?
Conclusion
The fact that successful translations between
languages can be made is a major argument
against it, as is the fact that the conceptual
uniqueness of a language such as Hopi can
nonetheless be explained using English (or any
other languages).
That there are some conceptual differences
between cultures due to language is undeniable,
but this is not to say that the differences are so
great that mutual comprehension is impossible.
One language may take many words to say what
another language says in a single word, but in the
end the circumlocution can make the point.
Language may not determine the way we think,
but it does influence the way we perceive and
remember, and it affects the ease with which we
perform mental tasks. Several experiments have
shown that people recall things more easily if the
things correspond to readily available words or
phrases. And people certainly find it easier to make
a conceptual distinction if it nearly corresponds to
words available in their language.
語言是否在深刻的層面上反映社會文化，而不只是在一些芝麻小事
上，像是某個語言有多少個字描述雪，或是有多少個字描述幫駱駝剃
毛? 更具爭議的問題是，不同的語言會不會讓其使用者有不同的思考
或覺知方式?
問題是，語言會深刻反映出文化差異，並且有愈來愈多
的科學研究證實母語會影響我們思考、感知世界的方
式。
語言為明鏡

語言是文化的產物，還是與生俱來的禮物? 如果我們把語
言當成一面反映大腦的明鏡，我們會在鏡中看到甚麼: 人
類的天性，還是社會的文化習俗?
語言本身就是一種文化習俗，而且除了文化習俗之外，不
會裝扮成別的東西。全球各種語言的差異甚大，小孩子會
學哪種語言，端視出生在哪個文化環境裡。
語言之間最明顯的差異，是不同的語言會為同樣的概念
賦予不同的名字(或稱標籤)。

這些標籤除了是文化的約定俗成外，並沒有其他的作用。
除了少數的擬聲詞，像布穀鳥這種標籤與所描述的鳥類
叫聲相仿之外，絕大多數的標籤都是主觀制定的，沒有一
定道理的。
但如果我們再往語言明鏡中窺望，穿透標籤的表面層次，
直視其背後所潛藏的概念呢? 「玫瑰」、「甜美」、「鳥」、「貓」
這些標籤背後的概念，就跟這些標籤本身一樣主觀而沒有
道理嗎? 我們的語言把世界切割成種種概念，是否只是源
於文化習俗?或者「貓」、「狗」、「玫瑰」、「鳥」的界線，是自
然切割出來的? 倘然這個問題看起來太過抽象，我們就來
實際測試一下。
《齊福特荒島探險記》(Adventures on the Remote Island
of Zift)

想像一下，你在一間老舊圖書館一處早被人遺忘的角落發現一
本發霉的十八世紀手稿，好像自從遺落在那裡就沒人拾起來過。
這本書的書名叫《齊福特荒島探險記》 (Adventures on the
Remote Island of Zift)，書中鉅細靡遺的描述一處作者自稱
發現的神奇沙漠荒島。你翻到〈再論齊福特語之異象〉 “A
Farther Account of the Ziftish Tongue Wherein Its
Phantastick Phænomena Are Largely Describ’d” 這一
章:
While we were at Dinner, I made bold to ask the Names of
several things in their Language; and those noble Persons
delighted to give me Answers. Although my principal
Endeavour was to learn, yet the Difficulty was almost
insuperable, the whole Compass of their Thoughts and Mind
being shut up to such Distinctions as to us appear most
natural. They have, for example, no Word in their Tongue by
which our Idea of Bird can be expressed, nor are there any
Terms, wherein that Language can express the Notion of a
Rose. For in their stead, Ziftish employs our Word, Bose,
which signifies white Roses and all Birds save those with
Crimson Chests, and yet another Word, Rird, which betokens
Birds with crimson Chests and al Roses save white ones.

吾等用飱之時，余嘗數間各物之齊福特語名，應答者欣然。余雖亟欲學習，
然吾人視為自然者，島民竟難覺察，此間差異甚大。齊島語中未有「飛鳥」一
詞，亦無字詞可表「玫瑰」者。語中僅有「飛瑰」，以表白色玫瑰與赤胸以外之
飛禽，另有「玫鳥」一詞，概括赤胸之飛禽與白色以外之玫瑰。
 Bose
Bose, which signifies white Roses and all Birds save
those with Crimson Chests.
以白色玫瑰與赤胸以外之飛禽結合為
一類，稱為「飛瑰」。

Rird
Rird, which betokens Birds with crimson Chests and
all Roses save white ones.

把紅色胸膛的鳥和白色以外的玫瑰歸
類成「玫鳥」。
你會把這段珍貴的敘事當成甚麼?

如果你認為這是虛構的故事，這大概是因為你的常識告訴
你，齊福特語區別物體的方式是不可能存在的: 把紅色胸膛
的鳥和白色以外的玫瑰歸類成「玫鳥」，再把其他的鳥跟白
色玫瑰合併在「飛鳥」和「玫瑰」一類，根本就是不自然的
分類法。如果齊福特語的「玫鳥」與「飛瑰」是不自然的，
那麼「飛鳥」和「玫瑰」的分類方式一定是自然的。因此，根據
常理，雖然語言可以主觀任意訂定標籤，卻不能用同樣主觀
任意的方式來處理標籤背後的概念。
語言不能隨意進行分類，因為一定是真正相似的東西才會放
在同一標籤下。任何一種語言在分類事物時，必須把現實生活
中(或至少在我們所察覺的現實生活之中)相似的東西歸類在
一起；因此，把許多種鳥類同樣歸類成為一個概念是很自然的
，但如果把隨機一部分的鳥類和隨機一部分的玫瑰歸成一類，
卻是再違反自然不過了。
兒童天生就能掌握這類的概念，從中可看出人類大
腦與生俱來擁有很孩的模式辨認規則，可將類似的
物體區分成各種類別。因此，「貓」、「鳥」等概念一定
以某種方式跟這種先天的歸類能力相呼應。
 標籤 概念

似乎已經得到一個簡單的答案，可以回答「語言究竟
反映文化或天性」這問題，我們畫了一張清楚的地圖
，把語言畫分為兩個明白的區域: 標籤與概念。標籤
代表的是社會約定俗成，但概念反映的是天生的能
力。每種文化可以任意在概念上面標上標籤，但標籤
背後的概念卻是自然形成的。
現實生活中，文化不僅控制標籤，還會不斷領軍跨過界線，
直攻我們認為應該是天生能力的部分。雖然有些概念的區
別(像是「貓」和「狗」)已經由大自然清楚切割，讓文化幾乎
無法撼動，文化習俗卻會干涉許多其他概念的內務，有時
候甚至會干擾常理。
問題是，文化可以深入概念之地到甚麼程度?
Does language influence thought?

• Think about gender issues in language

• Did language influence how we think about

gender?

• Language is not neutural

Invisibility of woman in English
• He, Man, 他
Maleness as norm
• Term that call attention to the
presence of woman
• ‘manageress’ and ‘lady doctor’, nurse
• Man as norm
Sex marking:
• we must indicate the sex of the
pronoun in the sentence but
sometimes it is irrelevant in the
context
Encoding of male worldview
× Language reveals male domination
× “Sex” and “foreplay” – Male perspective on sex
Sex refers to an act that is defined in term of male
orgasm
women have their orgasms are relegated to
secondary status, referred to by terms like
‘foreplay’
× ‘spinster’ and ‘bachelor’
× 廣東話例子：主動與被動、賺蝕概念
× It is more difficult for woman to describe certain
experience or even worse, language distorted their
understanding on certain issues
× Rape a man?
× The term ‘sexual harassment’, for example, is a
recent feminist innovation in 1970’s.