Kap. 8 - Language and Communication

What are four aspects of language that must be addressed by theory?

Psycholinguistics

• “The study of the cognitive underpinnings of human language”

.

4 aspects that must be addressed in every psycholinguistic theory:

Language is…:

1. Communicative

2. Referential — i.e. refers to things/ideas - and meaningful

3. Structured

4. Creative - i.e. allows for creation of meaningful, novel sentences

.

The 4 aspects in deeper detail:

1.Language is communicative

• Do not confuse language with thought!

  • - as we are able to think without words (i.e. mental imagery)

• All communicative: Speech, sign language and writing

• Communicative function is necessary for language qualification, but is not enough on its own!

• Animals communicate, but….:

  • May “simply” be conditioned responses - learned through chains of associations

.

2.Language is referential and meaningful

• Language as expressions of how we represent the world in our minds:

  • Objects, actions, descriptions, relationships, concepts, sentence structure

  • Things that are/are not present - past, present, future or abstract ideas

  • Mental lexicon:

    • The “library” of words we use and their links to real-world representations

    • Accessing it in several ways:

      • Phonologically

        • Via their sound

      • Orthographically

        • Via their written form

        • Lesion: pure alexia

          • (type of acquired dyslexia)

          • Difficulties in orthographic processing (reading)

            • Like writing with eyes closed - you can write, but you can't go back and read what you wrote

          • Left occipito-temporal cortex?

            • Visual word form area?

              • (though evolutionary probably evolved for objects and faces)

              • 
                

          • 
            

            
      

      • Spreading activation model

        • Word meanings link to each other as if organized like a web within the mental lexicon - words are related

        • Exposure to one word (“prime” word) activates corresponding node in the web, spreads, and activates other words to different degrees (depending on strength of connection to prime word)

.

3.Language is structured

• Sentences have underlying structures (syntax) that are processed separately from their semantics (meaning of a word or phrase)

  • Phrases can be grammatically correct without making any sense

    • i.e. “colorless green ideas sleep furiously“

.

4.Language is creative

• Productivity of language

  • The ability to produce and understand completely new sentences!

  • Shows that we have internalized system of rules for how to combine elements inte new meanings

Describe the lexical decision task, some factors that influence performance on it, and what these suggest about how our mental lexicon is organized

Lexical decision task

• A task to investigate how word meanings (aspect 2) are organized and relate to each other within the mental lexicon

• Task: participants make rapid judgements about strings of letters that are presented to them

  • Patterns found are used to draw conclusions about how closely linked the word meanings are in a person’s mental lexicon

    • Useful for: understanding impact of aging and mental health on how we access and link words/concepts

• Factors that influence performance

  • Word frequency effect:

    • Faster to respond to high-frequency words (encountered a lot in everyday life ) than to low-frequency words

  • Semantic priming:

    • Exposure to a related word influences a response to a subsequent stimulus

      • e.g. if judging if “doctor” is a word, preceding it with “nurse” quickens reaction time mor than “butter”

  • Affective priming:

    • Exposure to a word with similar emotional quality influences primes the response to a stimulus

Describe som constraints that help children quickly infer the meanings of words

Fast mapping

• Ability to identify meaning after only one exposure (2 year old)

• Possible because children don’t take the time to consider full spectrum of a word’s referents - constraints on what referents are considered - logical inferences:

  • Whole-object contraint

    • Assume a word refers to the whole object

      • i.e. not to a specific feature

  • Shape bias

    • Generalize to objects with the same shape - when encountering novel objects

      • i.e. rather than to ones with the same color, size, texture

  • Mutual-exclusivity constraint

    • Factor in what they already know to guide their learning of new words - assuming that a new word is associated with the new item (rather than the old item)

What is recursion, and how is it important to language?

(aspect 4. - language is creative)

Recursion

• The property of language - its reliance on syntactical rules - that enables us to embed structures of language inside other structures

  • i.e. sentences within sentences

    • Steven Pinker: Never a limit to how long they can be!

• Enables us to compose sentences of boundless lyricism, power and originality!

How do Chomsky’s ideas about language differ from those of Behaviorism and finite state grammar?

Behaviourism’s ideas about language

• Rejected the need to speculate about the inner workings of the mind

• Argued: Human behavior can be studied by observing how conduct change as a function of association chains

• Skinner

  • Explain language development as emerging from associative learning / reinforcement

    • e.g. children learn proper grammar and pronunciation via feedback from adults (praised and corrected)

.

Finite state grammar ideas…

• Markov models

  • Explain verbal utterances and language development as probabilistically dependent on preceding words

    • Reflect “finite state grammars”:

      • Sentences are constructed in sequence - where earlier parts of sentences constrain what the rest can be

.

Chomsky’s ideas…

• Argued that neither the two above could explain (aspect 4. - language is creative)

• Underlying rules guiding language acquisition

  • Children gain linguistic mastery quickly - despite poverty of the stimulus

    • Limited exposure to information and feedback

      • e.g. later support for this:

        • parents tend to respond the same way to grammatically correct and incorrect sentences -

          • positive if they are factually true - even though grammatically incorrect

          • negative if other way around

    • + Adults around children often make speech errors

  • Suggests:

    • Language learning must have some cognitive structures in place guiding how to treat available linguistic information

      • Language aquisition device for universal grammar (universal language rules) - instinct to seek out and master the rules that define their native tongue

  • Other later support:

    • Children engage in overgeneralization

      • Grammatical rules (that are correct in some circumstances) in ways unlikely to resemble utterances from adults (wrong in the circumstance used)

        • e.g. “i runned home”

Discuss arguments for and/or against language being modular or instinctual

Modular/instinctual

(i.e. develops/operates independently of other cognitive abilities)

• For modular:

  • Modular: Children acquire language at a rate that outstrips the development of other cognitive functions

    • (Chomsky argumented for this)

  • Modular: William syndrome

    • Lower-than-normal IQ + impaired spatial skills - but relatively unimpaired in grammar and vocabulary (disagreement about this…)

    • (this supports Chomskys’ perspective)

  • Modular: Critical period

    • Biological period where children is optimally equipped to learn the rules of a particular language

    • (Chomsky argumented for this)

• Against modular:

  • Language development is linked with areas of cognition

    • Children infer the meanings of words by taking into account a speaker’s intentions

      • e.g. adult says word while appearing to make an accidental action, children don’t associate them

      • Autism

        • The more sensitive children with autism are to intentions of others (i.e. less far on the spectrum), the larger their vocabularies

  • Not every language follow “universal grammar”

    • e.g. Parahâ - whistle whole conversations + not recursion in the language

  • Instinctual: Pidgin language forms into creole language by children of pidgin speakers

    • Creole language: a synthesis of words from different languages that adheres to rules of grammar

    • (an argument against Chomskys’ perspective)

What is meant by the difference between “deep structure” and “surface structure” of language?

Deep structure

• i.e. semantic structure

• The underlying mental representation of the sentence

.

Surface structure

• i.e. syntactical structure

• The way the representation is structured linguistically

• If surface structure is hard to determine —> ambiguous sentence (difficult to understand underlying deep structure)

.

Comprehending a sentence involves transforming surfare structure into deep structure.

Producing a sentence involves translating deep structure into surface structure.

What is the difference between phonemes and morphemes?

Phonemes

• The basic units of sound in language

• e.g. consonants and vowels

• Can vary from language to language

.

Morphemes

• The smallest units of meaningful sound

• Morphology: the rules by which morphemes can be combined into words

• Types:

  • Content morphemes

    • Describe places, people, actions etc.

    • Can be used on their own

  • Function morphemes

    • Modify content morphemes

    • Cannot be used on their own

What is the speech segmentation problem in language perception, what acoustic factors cause it, and how does the brain successfully resolve it to extract meaning?

The Segmentation Problem:

• The physical acoustic stream of spoken language is continuous and often lacks clear physical pauses between words

• The brain must actively figure out where one word ends and the next begins

.

Contributing Acoustic Challenges:

• Ambiguous Phonetics:

  • Homophonous strings are pronounced identically despite having completely different word boundaries (e.g., "ice-cream" vs. "I scream")

• Co-articulation:

  • The vocal tract moves continuously, meaning the pronunciation of a speech sound is dynamically blended with and altered by the sounds before and after it

• Individual Differences:

  • Varied accents, speaking speeds, and vocal pitches across different talkers

• Degraded Signal:

  • Everyday acoustic environments are full of background noise, masking, or distortion

.

The Resolution:

• Because the bottom-up sensory signal is highly ambiguous, the brain heavily relies on top-down effects.

• It uses context, expectations, and previous knowledge to seamlessly bridge acoustic gaps, segment the stream, and extract the intended meaning of words

  • e.g. as demonstrated by the McGurk effect

    • What we perceive (in an auditory sense) is affected by what we visually see

      

How does infants’ sensitivity to phonemes change as they learn their native language?

Infants can distinguish among a wider array of phonemes than adults!

• But:

  • Native language magnet effect

    • Ca 6 months old: begin to specify their abilities to perfect their understanding of their native language - gradually losing sensitivity to sounds (phonemes) that are not present in their native language

      • e.g. Japanese adults (not japanese infants under 6 months) have lost sensitivity to distinguish between l and r

  • During development - become sensitive to their native tounge’s phonology

    • i.e. the rules that govern how sounds can be combined within that language

      • (i.e. rules that make it possible to mimic different languages while speaking gibberish)

Describe some common speech errors and provide examples

• Swapping phonemes across words

  • e.g. “snow flurries” becomes “flow snurries”

• Swapping morphemes

  • e.g. “self-destruct instructions” becomes “self-instruct destructions”

• Swapping words within a sententce

  • e.g. “writing a letter to my mother” becomes “writing a mother to my letter”

  • Anticipation error / spoonerisms

    • When a phoneme early in a sentence is swapped for a phoneme later in the sentence

      • e.g. “you have wasted the whole term” becomes “you have tasted the whole worm”

    • Suggests:

      • Planning of speech sounds and planning of word placement occur at different stages in language production

.

Evidence from common patterns of evidence suggests 3 stages of language production:

• 1. Pre-linguistic process: selecting the message to communicate

• 2. Process selecting the grammatical structure of the message

• 3. Process converting the message + structure into sequence of sounds

What evidence suggests that infants use statistical regularity to figure out what phonemes that are important in their own language?

Natural speech doesn’t provide clear speech segmentation

• Yet - we have ability to quickly parse/determine boundaries between words

• Phonemic restoration effect

  • Familiarity with our own language is such that we will often “hear” spoken phonemes that have actually been rendered inaudible due to background noise

.

Infants/children learn how to parse their native language by noticing the statistical likelihoods of numerous sound combinations

• E.g. phonemes that join to form a word will co-occur more often than phonemes taken from different words within a phrase

  • e.g. pretty baby

What is parsing, how does the neural representation of sentences differ from that of individual words, and what primary factors simultaneously contribute to this process?

Parsing:

• The active cognitive process of grouping individual words into structural phrases to extract the overarching meaning of a sentence

.

The Structural Difference:

• Unlike individual words, the structural representations of entire sentences are not stored in the brain.

  • Because the number of possible unique sentence combinations is infinite, the brain cannot store them.

  • Parsing is a real-time, constructive building process rather than a static memory retrieval process.

.

Contributing Factors:

• To parse language successfully, the brain simultaneously evaluates a wide array of information:

  • Syntax & Semantics:

    • Grammar rules and word meanings are calculated in parallel

  • World Knowledge:

    • Prior knowledge about the environment and how the world works

  • Executive Functioning:

    • Attentional control to hold the shifting string in working memory

  • Prosody & Gestures:

    • The rhythm, emphasis, tone of voice, and physical movements of the speaker

What are garden path sentences, and how does executive functions seem to impact ability to decode these sentences?

Garden path sentences

• Sentences that begin by suggesting one interpretation only to present another interpretation with their later parts

  • Non-lingustic factors play a role in language comprehension!

    • e.g. expectations, knowledge, context and prosody (patterns of intonation)

• Connection to executive functions/cognitive flexibility

  • Better cognitive flexibility = better at “on the fly” comprehend the sentences

  • Kindergarten path sentence effect

    • Children are less cognitively flexible, and have a particular difficulty processing this type of statement

.

Temporary Ambiguity:

• A phenomenon where the initial words of a sentence can be interpreted in multiple ways, and the actual meaning is only made clear at the very end of the clause

What is the importance of common ground in communication?

How does a lack of common ground change what people say?

Pragmatics

• The study of how the communicative function of language depends on what the speaker knows, what the speaker knows about what the listener knows, and what the listener knows about what the speaker knows

.

Common ground

• When speaker and listener share the same understanding or perception of the context or share a degree of knowledge

• The degree to which a speaker considers common ground plays important role in communication

  • E.g. The curse of knowledge

    • The difficulty that experts often have in putting themselves into the shoes of a less knowledgeable listener

• Referential communication task

  • Goal:

    • Way to study the role of common ground and adjustments speakers make to accommodate a lack of it

  • Method:

    • Two participants are separated by a partition, and one participant tries to describe layouts of objects that are within their view but not their partner’s view

  • Findings in participant’s adjustments:

    • Speakers often accommodate their partner’s level of knowledge after it becomes evident that their partner needs more information

      • i.e. common ground (or lack of-) gets spontaneously established within the course of the conversation with adjustments

How might bilingualism be linked with better cognitive performance?

• What evidence suggests that this might be the case, and what evidence suggests that it might not be?

Bilingualism

• Fluency in two languages

• Might be linked with better cognitive performance:

  • Perform better on tasks that index attentional control

    • e.g. response inhibition, task switching, and WM

      • Likely because: speaking two languages comes with constant challenge of juggling + monitoring both languages

  • Argument against:

    • Newer study: finding no for difference between bi- and monolinguals on different cognitive task performances

• But more confusion within a language?

  • Perform more slowly + make more errors than monolinguals on tasks that involve picture naming or word production and comprehension

What is the difference between “strong” and “weak” forms of the Sapir-Whorf hypothesis?

Sapir-Whorf hypothesis

• Suggests: differences among languages reflect and contribute to differences in underlying thought processes

  • i.e. language reflects/can shape how we organize/interpret perceptual experiences + how we comprehend the meaning of objects/events

.

• Strong form of Sapir-Whorf hypothesis

  • Language shapes the way we actually perceive and experience the world

.

• Weak form of Sapir-Whorf hypothesis

  • Language reflect (and can shape) differences in higher cognitive processes (i.e. categorization)

    • Language doesn’t affect subjective experience per se

Describe some ways that language has been found to shape or reflect differences in how people experience the world

Russian and English categorize the color blue differently

• Speakers of the languages performed differently when doing a color-matching task

  • Russian speakers were faster to respond when the two stimuli fell into different categories of blue than when they fell into the same category

    • This pattern did not emerge among English speakers

    • When interfering with reliance on language to perform the task (simultaneous verbal task):

      • Response-time difference among Russian speakers was eliminated

  • Suggesting: language can influence perceptual categorization

.

Language may affect perception of spatial information

• E.g.:

  • Preschoolers exposed to spatial language while playing = better when tested on mental rotation tasks than children who were not exposed

  • Tseltal speakers:

    • Do not use egocentric spatial terms - but rather cardinal directions

    • Findings: remember spatial layouts in terms of absolute coordinates - whereas Dutch speakers remember the scenes better in body-centric coordinates

      • BUT: unclear if the difference was a result of different strategies or understandings of task instructions

.

Language may influence how people think about/remember responsibility and mistakes made by others

• Describing accident they saw in video:

  • English speakers: use agentive language (“she broke the …”)

  • Spanish speakers: use non-agentive language (“the … broke“)

• Remembering the accident: english speakers remembered better who broke the …

.

Conceptual act theory of emotion

• Language may affect how we understand emotional experience

  • Access to the meaning of emotion words may enhance people’s ability to perceive facial expressions of emotions

  • The theory:

    • Processing and experience of emotion are shaped by conceptual knowledge about emotions (which in turn are supported by language)