Download Experiments

yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Language and thought
Relationship of cognition and language
• Categories of cognition are shaped by language
– Sapir and Whorf’s linguistic relativity
• Cognitive categories develop independently of language both in
evolution and in ontogeny, language only builds upon these
– Piaget: cognitive development leads language development
• Language and cognition are independent
– Chomsky
• Cognition follows its own path, but language modulates its categories
Early experiments
language = thought
Watson, 1913: thought =
subvocal speech
2. language ≠ thought
Smith et al., 1947: curare
experiment: muscle relaxant
Political correctness
“language use has an effect on the way we think”
• euphemisms in politics
– Pacification/pacifikáció = bombázás
– Revenue increase/bevételnövelés = adó
– Rationalisation/munkaerő-gazdálkodás = elbocsátások
• social movements: sexist/racist etc. language is
responsible for sexist/racist etc. thinking
– chairman → chairperson
– Gypsy → Roma (?)
– blind → with visual impairment
• Orwell, 1984: Newspeak
Language shapes the mind
• linguistic determinism: a language shapes
psychological mechanisms
• Benjamin Lee Whorf
• Language shapes the mind, world view, structure of science
• Differences in lexical (vocabulary) and grammatical organization
result in different conceptual schemes
Whorf: linguistic determinism and relativity
• Linguist and engineer, student of anthropologist Edward
– Studied native American cultures and languages
– Emphasized the variety and differences of cultures,
not the common features
• Strong view: all higher forms of thought build on
• Weak view: the structure of the language one generally
uses influences the way they understand their
environment and act upon in it
Linguistic relativity
(the Sapir-Whorf Hypothesis)
• Follows from linguistic determinism
• Linguistic relativity: distinctions encoded in one
language are unique to that language alone, and "there
is no limit to the structural diversity of languages”
– It is impossible to translate precisely from one language to
• lexical and grammatical relativity
Lexical influences
• Lexical level: what words are found in a given language,
and what they refer to
– different languages carve up the world in different ways
through more or less specialized vocabularies
– languages differ with respect to how they divide up the world
into nouns and verbs
• lightning: a N in English, but a V in Hopi
– duration an important feature
• Tzotzil Mayan: eat-mushy; eat-a-slender-shape-food, eatmeat
– the properties of objects are incorporated into the
Grammatical influences on thinking
• Number category
– whether inanimate nouns can be pluralized or not
– in English any noun can be pluralized as long as the
referent is discrete, i.e., mass nouns such as paper,
flour cannot be pluralized
• count nouns such as pen, girl
– in Yucatec, only animate nouns can be pluralized
– Lucy (1992): English speakers specify the number of
objects in descriptions of line drawings more
frequently than Yucatec speakers
Grammatical influences
• Tense markers
– determine location of events in time
• past ---------- now ---------- future
– he is running
– he ran
– he will run
WARI in Hopi
• How does a temporal language compare to a
“timeless language”?
• Hopi distinguishes between
– Reportive: report of a recent or ongoing event
– Expective: report of an expected event (past or future)
– Nomic (not described)
• According to Whorf these are not tenses because
they reflect the epistemic validity of the statement
rather than its duration or location in time
Potawatomi inclusive and
exclusive pronouns: we
Hungarian object agreement
• The verb form signals the specificity of the
– Megevett egy almát.
– Megette az almát.
Hungarian locatives
Approximate NÁL
spatial language
Eskimos have many different words for ‘snow’
→ evidence that they see snow differently
(urban legend!)
→ Boas, (1911): 4
aput („snow on the ground”)
gana („falling snow”)
piqsirpoq („drifting snow”)
qimuqsuq („a snowdrift”)
→ Sapir& Whorf, 1940: 7
→ 1978: 50
→ 1984 (New York Times): 100
The truth about snow
• There are several Eskimo languages + Eskimo languages differ in
the number of expressions they have for snow
• Definition of “word” is problematic
– Inuit is a polysynthetic language: are words derived from the same stem
different or not?
• More importantly
• Even if it was true that one language had more, is it evidence that
they see snow differently?
painters: paints
ornithologists: birds
Basic colour terms
(Berlin & Kay, 1969)
• Properties
1 morpheme
Not restricted to one class of items (e.g. blond)
Do not belong to the scope of another color terms (e.g. torqoise)
Frequently and generally used
• Basic color terms are chosen from 11 colors by all languges: black,
white, red, yellow, green, blue, brown, pink, purple, orange, grey.
• languages differ in how many basic color terms they have (Hungarian
for ‘pink’ rózsaszín is not a basic color term)
• 2 colour terms (mili-mola): Dani, New Guinea
• There seems to be a universal hierarchy of
colour categorisation.
Do speakers of different languages see
colors differently?
• Color categories are not arbitrary!
• Same everywhere:
– light
– Operation of the human eye
• 3 kinds of cones in color perception
→ these determine what we see
• Experiments (pl. Heider, 1972 – the
dani): recalling and discrimination is
good for colors—focal colors
Experimental results
(pl. Heider & Oliver, 1972; Rosch, 1978, Berlin & Kay 1969,
Kay & Kempton 1984)
• People speaking different languages choose the same
shade as best exemplars of a category (focal colours)
– The best exemplar of grue in grue languages is the same as the
best exemplar of green in green-blue languages
• Dani do as well as English speakers in non/verbal colour
discrimination and memory tasks
• In a free categorization task, different speakers use
different categories (those marked in their languages)
Winawer et al 2007
• English and Russian speakers
– Russian: dark blue/light blue distinction
• A blue shade shown, then two blue shades
• Task: which of the two is the same shade as the probe?
• Russian speakers:
faster RT if the two
shades are from
different linguistic
colour categories
Gilbert et al 2006
• If language affects perception,
the effect should be stronger
for the right visual field
• Task: Which side is the
different shade on?
• Variables:
– shade difference across or within
linguistic category (blue-green)
– Target in left or right visual field
• Results: when different
linguistic categories, faster
response in RVF
Korean locatives
(Bowerman & Choi 1994, 2001)
Korean locatives (Bowerman, 1996)
• Korean (vs. English and Hungarian): no linguistic distinction, between
placing an object in a container or on a surface (in vs. on, -ban vs. -on)
• Korean language distinguishes between tight fit (ring on a finger,
picture on the wall) and loose fit (fruit in a bowl, object leaning against
a wall)
– This distinction holds for both containment (in) and support (on)
• Experiments
– English/Korean babies differentiate all potential spatial distinctions
– As a results of acquiring a language certain spatial distinctions
(those strengthened by language) become salient in representation
Navajo shape classifiers
• Carroll and Casagrande: Navajo vs. English
– Navajo verbs change form according to the shape of the
object it takes (shape classifiers)
• flexible vs. rigid; flat vs. round
– give blue rope and yellow stick and ask which of the two a
blue stick can go with
• Navajo choose shape: yellow stick
– English choose color: blue rope
– conducted the test with upper class Bostonians
• responded like Navajo children
– there is other kinds of determinism than just linguistic
Grammatical gender and object perception
Experiment(Boroditsky & Schmidt, 2003)
– Spanish, German and English speakers (experiment language:
– Training: 24 pairs of object - name
apple – Paul / Paula
bench – Eric / Erica
clock – Karl / Karla
– Test: object word shown, name has to be recalled
apple – ?
bench – ?
clock – ?
– For Spanish and German speakers, better recall performance for
pairs where the gender of the name corresponds to the gender of
the object word
Spatial reference (Brown 2001)
• Ego-centric (left, right, in front of me, behind me) –
• Intrinsic (left of the object, in front of the object, etc)
• Geocentric (hill-wise, sea-wise, etc) – absolute
• Left „xin” and right „wa’el”
– Refer to body parts only
• Absolute reference system:
– „alan”: downhill ~North
– „ajk’ol”: uphill ~South
– Indoors, outdoors
• Dutch + Tzeltal speakers (Bowerman,
Seated at talble in a room, shown a pattern
Turned 180 degrees, asked to reproduce pattern
Chips task
Chips task - results
Maze task
Maze - results
Evidence for Relativity?
• Li & Gleitman (2002)
– Response depends on environment: the availability of
reference points
• Compare cities/varied landscape vs. open landscape
– In a darkened room (no visible reference points), English
speakers also use the absolute reference frame
Reference frames and ecological
Related documents