slide1 n.
Skip this Video
Download Presentation
Psycholinguistics - a dead discipline?

Loading in 2 Seconds...

play fullscreen
1 / 125

Psycholinguistics - a dead discipline? - PowerPoint PPT Presentation

  • Uploaded on

Psycholinguistics - a dead discipline?. term coined in 1954 (Osgood). aim: to describe the exact operation of the brain during the production or processing of language. psycho- linguistics. psychology. linguistics. A paradigm shift in linguistics:.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Psycholinguistics - a dead discipline?' - zelenia-phelps

Download Now An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

Psycholinguistics- a dead discipline?

  • term coined in 1954 (Osgood)
  • aim: to describe the exact operation of the brain during
  • the production or processing of language






A paradigm shift in linguistics:

1957: publication of “Syntactic structures” by Chomsky

  • refutes behaviorism
  • proposes the “mentalist” approach
  • considers linguistics a subfield of cognitive psychology

Behaviorism in linguistics

and psychology:

  • reduces mental activity and cognition to implicit, observable behavior
  • behavior is explained as a relationship between input and output (i.e. stimulus and response)
  • studies of speech behavior and the sound system prevailed
  • cf. Skinner, B.F.: “Verbal Behavior” (1957)

Areas of psycholinguistic interest:


  • language acquisition (L1 and L2)
  • language comprehension
  • (includes symbol recognition, speech perception)
  • language simulation (NLP, PDP)
  • concepts of reality and language
  • memory constraints (STM/LTM research)
  • knowledge representation
  • strategies of learning






Psycholinguistics and related disciplines

  • “classic” psycholinguistics:
  • - language acquisition
  • - language impairment
  • - aphasia research
  • - reaction times
  • - ERP measurements

Psycholinguistics and related disciplines

  • areas of psycholinguistic research:
  • computability of language processing
  • neuroscience / neurolinguistics
  • cognitive abilities (vision, motor control...)
  • conceptualization
  • symbolization

Classic psycholinguistics

concerned with:

psychological processes that make acquisition and use of language possible

approaches (cf. Clark & Clark)

1. language comprehension (spoken and written)

2. speech production

3. language acquisition


Psycholinguistics - the extended view

concerned with:

language as a cognitive systeminternalized

within the human mind/brain.

ultimate goal: to characterize this internalized

system -

I language (Chomsky)


Classic approaches in psycholinguistics

  • 1. language comprehension (spoken and written)
      • comprehension at various depth levels
      • speech perception
      • lexical decoding
      • sentence processing
      • text processing

Classic approaches in psycholinguistics

  • 2. Speech production
      • reoccurring patterns of speech
      • typical errors
      • response times
      • relation of speech to concepts
      • speech impairments

Classic approaches in psycholinguistics

3. Language acquisition

- L1 acquisition (developmental


- L2 learning strategy research

- acquisition constraints


Neurological foundations of language

  • correspondence hypothesis:
  • particular areas of the neocortex are responsible for human language faculty
  • local results from aphasia research
  • temporal results from ERP measurements
  • aphasia: impairment or loss of language ability due to brain damage.

Neurological foundations of language

Paul Broca: lateralization of language

- located lesions in left hemisphere

- related handedness to speech capability

- plasticity of the brain (i.e. temporal variability)

- migration of neurons

- time constraints in acquisition


Neurological foundations of language

Carl Wernicke: - separated the auditory nerve (cranial nerve from ear to cortex) along the planum temporale in the left hemisphere


Language-related areas of the brain

  • Broca aphasics:
  • nonfluent
  • agrammatical
  • morphemeless
  • unimpaired
  • comprehension
  • Wernicke aphasics:
  • fluent (logorrheic)
  • impaired meanings
  • neologisms
  • severely impaired
  • comprehension

Language-related areas of the brain

  • spatial: lateral distribution
  • - detectable in lesions
  • - PET, fMRI scans
  • temporal: brain plasticity
          • - performance patterns
          • - physiological changes during L1
          • acquisition
          • - learnability constraints

The paradox of psycholinguistics

  • L1 acquisition enables children to produce virtually infinite amounts of linguistic data.
  • Input includes:
  • distorted input (also: deviant input; Chomsky) can be: mispronounciations, slips of the tongue
  • omitted rules
  • inference of rules out of defective material
  • negative evidence
  • = pointing at errors
  • typical errors in L1: *go-ed
  • atypical errors: *I no like syntax.

The paradox of psycholinguistics

  • phases in L1 acquisition
  • single-word stage:
  • at 12 months: first recognizable words
  • until 18 months: vocabulary increase
  • 3 words/month (apple, up...)
  • no evidence of grammar acquisition
  • no inflection (plural-s, past-ed)

The paradox of psycholinguistics

  • phases in L1 acquisition
  • after 18 months:
  • acquisition of grammar begins
  • productive use of inflections
  • elementary 2-3 word utterances
  • after 30 months:
  • acquisition of most inflections
  • core grammatical constructions
  • adultlike, multiword speech

Learnability constraints

  • critical-period hypothesis (Lenneberg et al.)
  •  age constraints in L1/L2 acquisition
  • age estimates between 11-18

Learnability constraints


Version one: the exercise hypothesis. Early in life, humans have a superior capacity for acquiring languages. If the capacity is not exercised […] it will disappear or decline with maturation. If the capacity is exercised […] language learning abilities will remain intact throughout life.

Version two: the maturational state hypothesis. Early in life, humans have a superior capacity for learning languages. This capacity disappears or declines with maturation.

J.S. Johnson/ E.L.Newport in Johnson,Mark (ed.) 1996, pp.250.


Explainability of cognitive phenomena

1. Empirism

2. Operationalism

3. Instrumentalism

4. Idealism

5. Realism




physics, astronomy...


Explainability of cognitive phenomena

1. Empirism

- knowledge as a collection of facts

- universals are not obtainable

- theories are summaries of observations

2. Operationalism

- science is a system of rules

- theories are tools for manipulation


Explainability of cognitive phenomena

3. Instrumentalism

- not the meaning of words is important but

the way we use them

- theories are instruments of experience

- there is no “inner truth”

5. Realism

- laws have a relationship to reality that is relevant

- tool: observation


Chomsky theory: an introduction

  • refutes structuralism, taxonomy (Harris, Bloomfield)
  • refutes behaviorism (Skinner, Osgood)
  • continues tradition of Descartes (Dualism)
  • language acquisition is determined by a LAD
  • (language acquisition device) on the basis of a UG
  • the LAD is a mental organ
  • theory is primary, data is secondary
  • for cognition and language the
  • computer metaphor applies

Chomsky theory: an introduction

  • 1957 “Syntactic Structures”
  • set of kernel sentences generate
  • all possible sentences of a language
  • kernel  transformation rules  final phrases
      • a purely syntactic theory
      • transformations are algorithmic procedures
      • "Colorless green ideas sleep furiously."

Chomsky theory: an introduction

  • 1965 “Aspects of the theory of syntax”
  • so-called Standard Theory (ST)
  • involves phonology, semantics
  • Deep structures  Surface structures
  • (semantics) (phonology)
  • Two subfields emerge:
  • 1. Generative semantics (Katz, Postal)
  • 2. Extended Standard Theory (EST)
  • (Chomsky, Jackendoff, 1972)

Chomsky theory: an introduction

  • 1. Generative semantics: extends transformations
  • 2. EST constrains transformations
  • EST led to Revised Extended Standard Theory
  • (REST) (1973)
      • modular
      • separates syntax, semantics
      • only transformation left: move-

Chomsky theory: an introduction

  • Further developments
  • “Rules and Representations” (1981)
  • introduces principles & parameters
  • slot  filler principles
  • Government and Binding Theory (1981)
  • Minimalist Program (1993)

Language faculty: problems of research

  • Quine:
  • investigation of language equals investigation of mind
  • Chomsky:
  • knowledge is represented in the brain
  • proposes the existence of
  • mental representations = an abstract terminology
  • for physical properties
  • extends notion of "material body" for
  • entities, principles of unknown character

Language faculty: problems of research

principles: unknown or unobtainable?

mind: a fixed set / endowment with

inherent constraints

Chomsky proposes

1. problems (science may provide a solution)

2. “mysteries” (beyond humans’ intellectual grasp)

Descartes: we are not intelligent enough to

understand to what extent our free choices are



Language faculty: problems of research

1.) Is the application of the scientific method to

the mind revealing?

2.) Is language artificial?

3.) In what way do generalizations distort the view

on language?


Language faculty: problems of research

  • 1.) Is the application of the scientific method to
  • the mind revealing?
  • historical coincidence  biological endowment
  • meets aspects of reality in a meaningful way
  • tolerance of unexplained phenomena
  • (attention research, mental rotation etc.)

Language faculty: problems of research

  • 2.) Is language artificial?
  • Chomsky: question is meaningless even if language
  • had indeed been created
  • has developed basing on endowment and
  • environment

Language faculty: problems of research

  • 3.) In what way do generalizations distort the view
  • on language?
  • Chomsky's demands:
  • homogeneous speech community
  • speakers with 100% competence
  • speech unaffected by exterior (e.g. social) variables

Language faculty: problems of research

  • 3.) In what way do generalizations distort the view
  • on language?
  • Counterarguments:
  • humans cannot acquire language in a
  • homogeneous community, inconsistence and
  • variability are required
  • if humans could achieve it, it would be done by
  • different properties of the mind than those which
  • interact with reality

Language faculty

  • language faculty:
  • discrete from other kinds of knowledge
  • linguistic knowledge (= speaker’s competence):
  • interacts with processes of perception, memory
  • displays in indefinitely large number of strings
  • producable and understandable
  • syntactic mechanisms are recursive

Chomsky’s Universal Grammar

  • mind design is modular(Fodor, 1997)
  • The insight into language faculty may not provide
  • insight into other modules e.g. vision
  • Chomsky proposes:
  • a differentiated version of the modules
  • genetically coherent properties which determine
  • human cognitive systems including language faculty

Chomsky’s Universal Grammar

  • UG: the study of the common grammatical properties
  • shared by all natural languages and of the
  • parameters of variation between the languages.
  • parameters: dimensions of variation,
  • e.g. subject parameter
  • theory of UG provides tools to study
  • any natural language
  • example: Hawaiian creole (cf. Bickerton)

Chomsky’s Universal Grammar

initial state end state







L1 acquisition

  • Language acquisition skills are formal, structural properties

Problems of cognitive research

  • common set shared in cognitive community:
      • knowledge representation
      • language processing
      • image understanding
      • inference
      • learning strategies
      • problem solving

Emergence of a discipline

Cognitive approach:


emerges at the

intersection of the










Simon/Newell 1958: “In 10 years most psychological

theories will be formulated as computer programs.”


Views on cognitive functioning

basic assumption:

Cognition, and therefore language,

is information processing

The human mind is a system that receives, stores,

retrieves, transfers and transmits information

(Stillings et al., 1997)


Views on cognitive functioning

The classical view:

language faculty is a mental process

- mind can be described as a Turing machine

- linguistic processing is a manipulation of symbols

- cf. the "Chinese room" metaphor (Searle)


Views on cognitive functioning

The connectionist view:

- brain employs a computational architecture suited to natural information processing

- evidence in functional split, cf. split-brain patients

(McClelland, Rumelhart, Hinton)


The classical cognitive approach

Turing machine: a general-purpose information processor

components: tape eq. memory

subdivided into cells each containing one symbol

head: moves tape back/forth

can read/write symbols


The classical cognitive approach

  • Turing’s proof:
  • TM is able to perform all operations a person working within a logical system can perform
  • TM gives therefore a complete account of what information processing is.
  • "Any informational simulation process can be realized by a Turing machine" (Turing/Church thesis)

Turing machines and cognition

  • anything computable can be computed
  • can make decision about well-formedness of artificial languages
  • simple steps, primitive building blocks lead to emergence of complex behavior
  • reasons of relevance:
  • 1.) provide a complete description of information processing
  • 2.) may answer cognitively interesting questions
  • 3.) extend finite states into infinite behavior e.g.
  • novelty of language

The tri-level hypothesis

  • mind: shares properties with a TM
  • brain: a physical symbol system
  • not one single level of description applies (Marr)
  • 3 levels of description:
  • 1. physical level
  • 2. procedural level
  • 3. computational or implementational level

The tri-level hypothesis

  • 1. the physical level of description
  • describes the components of a system
  • incomplete, static
  • gives factual knowledge, "what”
  • 2. the procedural level of description
  • non-physical description of informational processing steps
  • dynamic, incomplete
  • no description of interpretation of procedures
  • gives procedural knowledge, "how"

The tri-level hypothesis

  • 3. the computational/implementational level of description
  • interpretation of the procedures
  • gives interpretational knowledge, "why"
  • tendencies: reductionism, neuroscience
  • tries to reduce 3. and 2. to 1.

The cognitive approach

“The mind as an information processing system can be described using a physical, a procedural and an implementational vocabulary” (Dawson)

goal: to find the relationships between these levels of research

physical description: neuroscience, linguistics

procedural description: psychology, linguistics

computational description: computer science


Language comprehension and

the Tri-level hypothesis

1. physical (structural sentence comprehension)

- Clark & Clark

2. procedural (psychological models of text comprehension - Kintsch & v.Dijk

3. implementational (AI programs for text analysis) - Minsky, Schank, Charniak


1. Physical (structural sentence comprehension)

  • Clark & Clark
  • comprehension: derivation of meaning from a (phonological) representation
  • meaning: composed from constituents
  • intermediate constituents (syntactic units,phrases)
  • final constituents (words, lexemes)
  • forms propositional representations
  • prerequisite: parsing of language

2. Procedural (psychological models of text comprehension - Kintsch & v.Dijk

  • subdivides micro- and macroprocesses
  • cognitive tasks required:
  • a.) parser (turns verbal text into intermediate sematic representation, list of propositions)
  • b.) coherence generator (builds coherence from
  • the list)
  • c.) inferencer (fills in missing propositions)
  • d.) organizer (determines facts on basis of world knowledge)
  • Parser: a mechanism that divides strings of texts into smaller components.

3. implementational (AI programs for text analysis) - Minsky, Schank, Charniak

  • Frames: "A frame is a data structure for representing a stereotyped situation"(Minsky)
  • a format for formalized storage of knowledge
  • maps unknown structures onto known structures


example: HOUSE

subframe of: building

Is-part of: village, city, suburb

material: wood, stone, concrete

# of windows: integer, >2

# of doors: integer, default 1


3. implementational (AI programs for text analysis) - Minsky, Schank, Charniak

  • scripts: formalized representations of complex
  • actions
  • list of primitives (stereotype procedures) e.g.
  • "consists of concepts and relations between concepts" (Schank)
  • example: restaurant script
  • story grammars:
  • lists of stereotype rules
  • Rule 1: story setting + episode
  • Rule 2: setting (state)
  • Rule 3: episode event + reaction

Mental Representations

  • theoretical postulates
  • internal states
  • Fodor (1997): Representational Theory of Mind (RTM)
  • cognitive mechanisms extract information from
  • outside world
  • info is processed, stored, retrieved via an
  • internalised system of representations
  • representations have semantic content
  • modern psycholinguistics:
  • distributed vs. local representation
  • = words are stored in single units vs. storage patterns

Computable linguistic approaches

  • Chomsky: Principles and Parameters
  • strategy: slot  filler
  • set of principles shared by all languages
  • parameters function as "switches” to adjust
  • the principles
      • Lexicon
  • |
  • D-Structure
  • |
  • S-Structure
  • Logical Form Phonological Form

Computable linguistic approaches

Lexicon: set of entries/word forms , syntax information

D-Structure: underlying representation of a sentence, X-bar theory

S-Structure: natural language sentence generation; move-α - rule

language processing: not rule processing but setting

of parameter values

 increased power of the lexicon

carry  NP_Verb_NP

read  NP_verb

Problem: How is this knowledge formally represented?


Computable linguistic approaches



of language

1. Syntax

2. Lexicon













The connectionist approach

1.) serial processing is too slow (cf. Feldman)

neuron action potential takes 1-3 ms to build up – humans solve complex tasks in ca. 100 ms

 suggests parallel processing

2.) system stability: brain exhibits enormous damage-resistance

 suggests network-like storage systems

3.) natural language processing – hard for computers, easy for humans and vice versa

qualitative difference of brain architecture

4.) connectionist models are “neurally inspired”

imitate neural functioning


Rumelhart's components of a connectionist system

a.) set of processing units

b.) state of activation defined over the

processing units

c.) output function for each unit, generates output

d.) pattern of connectivity

e.) activation rule for combining inputs to new activation level

f.) learning rules that modify d.)

g.) system environment


Formats of representation

  • 1.) multimodular model (Paivio):
  • two parallel sets: - imagenes
  • - logogenes
  • can be primed (threshold function)
  • 2.) semantic networks (Collins/Quillian)
  • types of nodes: - conceptual nodes
  • - property nodes
  • types of relationships: IS-A relationships
  • HAS-PROP relationships

Logogene model of word recognition (Morton)







visual evidence

auditory evidence








response buffer


Logogene model: priming effects

  • activation potential of logogenes decreases over time
  • high-frequency words: low thresholds
  • low thresholds facilitate recognition








Logogene model: priming effects

  • non-words or degraded signal:
  • activates “nearest match” item
  • memory effects:
  • primacy effect vs. recency effect






Semantic networks (Collins/Quillian)

Can be bought

Can be read


Is bound


Is small

Has pages

Pictorial Book

Exercise Book


Is expensive

For study

Is cheap

Has pictures

Has diagrams


Cognitive networks (Hays)

  • refined set of nodes/relationships
  • nodes: events (start, end, duration)
  • entities (object, notion, form)
  • properties (shape, mass...)
  • modalities (static, dynamic)
  • relationships: paradigmatic (hierarchical)
  • syntagmatic (parallel)
  • discoursive (interactive)
  • attitudinal (positive,negative)
  • metalinguistic (external)

The stance of neuroscience

  • Science of mental life will be reduced to neural functioning (Churchland)
  • reductionist approach
  • eliminates psychological explanation for language processing
  • emphasis on relevance of neuronal measurements













of thought


Neuronal functioning

  • diagram of a neuron

Dendrites: short branches projecting from cell body.

Receivemessages from other neurons Cell body (soma): contains the nucleus of the cell Axon: a long tube which carries information

from cell body to synaptic terminalsSynaptic terminals: secrete transmitter substance


Measurement methods in psycholinguistics

  • ERP measurement
  • PET scan
  • MRI / fMRI scan
  • lesions research
  • continuum:

low-grained spatial resolution

high- grained temporal resolution

high-grained spatial resolution

low- grained temporal resolution




ERP measurements

ERP: event-related potentials

notice activities in the relevant cortical regions

linguistic phenomena are correlated with activity


after presentation of the

unexpected ending of a

sentence: delay of the



PET: positron emission tomography

  • weakly radioactive substance used as marker
  • regions of high activity involve more blood flow.
  • synaptic ends extract more molecules, among
  • those radioactive molecules
  • positron-emitting radioactive molecules
  • mark location:
  • positron e+ hits an electron e-
  • particles annihilate each other
  • radiation of energy (light etc.)
  • detectors (PET cameras): arranged in a
  • torus (ringlike) structure around the subjects' head

MRI and lesions

  • MRI: magneto-resonance imaging
  • particles seen as magnetic dipoles e.g. H2 nuclei
  • reposition themselves in magnetic fields
  • induction of electric currents
  • lesions
  • lead to aphasias
  • language-specific
  • example: Englishspeaking aphasiacs retained
  • the ability to generate irregular forms
  • whereas Germanspeaking subjects did better
  • in regular forms

Comprehension and understanding

Mental Models:

- blueprint /abstraction of aspects of the physical world

- representations in the mind of real or imaginary situations

- mind constructs "small-scale models" of reality that

it uses to anticipate events

- can be constructed from perception, imagination, or comprehension of discourse

- underlie visual images, but can also be abstract, representing situations that cannot be visualised


Mental models and comprehension

mental models theory of text comprehension

(Johnson-Laird et al.) :

derives from a theory of deductive inference

mental models of spatially related objects contain information about relations not explicitly described


The man is in front of the tree.

The tree is in front of the house.

contains the information

The man is in front of the house.


Mental models and comprehension

Reasoning is a semantic process rather than syntactic - recipients build mental models of the relevant situations based on world knowledge

- recipient’s conclusion is “true” within the models

- emphasis on causality of events/situations

fundamental representational assumption:

individuals seek to minimize the load on working memory by representing explicitly only those cases that are true


Text comprehension

  • understanding a story: requires process of constructing a mental model consistent with the constraints of the story
  • recall errors provide insight into how the construal is designed or:
    • how reader's world knowledge interacts (cf. processing model)
  • 2 processes: construction (encoding)
  • reconstruction (recall)
  • both underlie errors

Strategies in text understanding

  • relevance of causal knowledge structures:
  • reader establishes a causal field
  • contains specific circumstances of the story
  •  explicit identification of conditions perhaps only implicitly mentioned
  • He sat in the waiting room, his cheeks bloated. After a while, a nurse called him up. Reluctantly, he followed her next door.
  • representation updates world knowledge.
  • stored for recall (on specific cues).

Strategies in text understanding

  • constraints of causality: A causes B
  • 1. temporal constraint (A precedes B)
  • 2. counterfactuality constraint (if A had not happened, B would not have happened)
  • 3. sufficiency constraint
  • If B occurs after A, circumstances for A are still prevailing
  • steps of comprehension:
  • 1. identification of clauses corresponding to events
  • 2. identification of causal relations
  • 3. establishment of causal chains

Comprehension processes

  • experimental evidence: reader constructs chains
  • propositions from main causal chains are more likely to be recalled
  • propositions with more causal connections than others are more likely to be recalled
  • basic blocks: the event clauses (propositions)
  • PDP approach: building blocks are propositions
  • evolution of reader's mental state as a function of time
  •  a trajectory in situation-state space = moves from point to point

A comprehension model

  • knowledge of causal relations between points: "belief function” - assigns degree of belief (can be between 0 and 1)
  • situation identification t1 t2 t3
  • Mary heard the ice-cream truck 1 1 0
  • Mary wanted to buy ice-cream 0 1 0
  • Mary is eating ice-cream 0 0 1
  • Mary is sleeping 0 0 0
  • story comprehension: finding a most probable trajectory in situation-state space with respect to a belief function.

Causal chaining

1 3

2 4




1 hear(M,truck)

2 want(M,ice-cream)

3 be(ice-cream,expensive)

4 go(M,money)

5 buy(M, ice-cream)

6 eat(M,ice-cream)

7 sleep(M)



surface anaphora

“deep” anaphora

Mary heard the ice-cream truck. Mary wanted to buy ice-cream.

Ice-cream is expensive. Mary goes home for the money.

She buys the ice-cream. John has also chilled drinks.

Mary is eating ice-cream. Mary is sleeping.


Micro- and macrostructures (Kintsch et al.)

  • surface structure of a discourse: set of propositions, ordered by semantic relations
  • 2 levels:
  • A microstructures – the local level of discourse, individual propositions (eat(Mary,ice-cream)
  • B macrostructure – the global discourse structure
  • - sets global constraints (topic, title)
  • - establishes the "meaningful whole"

Formation of Microstructures

The Swazi tribe was at war with a neighboring tribe because of a

dispute over cattle. Among the warriors were two men, Kakra and

his younger brother Gum. Kakra was killed in a battle.

Step 1 - identify most important proposition- "was at war"

Step 2 - relate other propositions to this proposition according

to coherence rules (limited by capacity of STM).

Step 3 - try to relate propositions in next sentence to propositions

that are active in STM. (example fails - no terms in

2nd sentence can be directly connected to preceding sentence

Step 4- If step 3 failed, do "reinstatement search"

- reinstates information about the text from LTM into STM

- effort to link new propositions to old ones

- reinstatement slows comprehension

Step 5- If reinstatement fails, start a new coherence graph

- try to make inference to link new material to old material.

Example inference: Kakra & Gum were Swazi warriors.


Spatial cognition and language

  • Try to name the colors of the displayed words
  • source:

Spatial cognition and language

  • Are image and name stored together or separately?
  • cf. logogene/imagene
  • parallel activation of visual and lexical entry
  • The Imagery debate
  • existence of internal representations
  • rat experiments (Toulmin) – evidence for the representation of maps
  • A) propositional representation (Pylyshyn) – mental "jumping" on a map without time delay
  • B) analogue representation (Kosslyn et al.) – mental scanning, mental rotation

Visual recognition process

  • 1. initial sensing of visual information: photosensitive cells (rods and cones) in retina,
  • activated by individual photons
  • each receptor responds to a tiny portion (minutes of arc) of the visual field,
  • receptors are never activated alike
  • 2. neurons in retina: connected to neurons in the
  • first visual cortical area
      • -for shape perception
      • -different cell types for different tasks (spots, edges...)

Kosslyn’s proto-model of object recognition

  • visual bufferattention window
  • -edge detection -selective input for -establishes regions contiguous sets of points
  • of homogeneous value

Stimulus encoding (Kosslyn)

1. degraded contours

vertices: high-information part of


cf. Biederman p.152 in:

Kosslyn/Osherson (eds.)1995.


Encoding of spatial properties

2. missing parts: more recognition time needed when parts have been removed

3. disrupted parts, distorted spatial relation among parts  violation of viewpoint consistency


On mental rotation (Shepard, Kosslyn, et al.)

  • direct proportionality between time and angle
  • more complex objects /3D objects: longer times
  • identical images in different scales: scale difference proportional to identification times
  • activation of motor areas in the brain
  • critical angle for letters: 120°

Spatial reasoning in language and mind

  • introspective reports on simple relation tasks using comparative adjectives:
  • The Empire State building is higher than the Eiffel tower.
  • mental comparison based on spatial descriptions:
  • Cathy is taller than Linda. Linda is taller than Mary.
  •  Cathy is taller than Mary.
  • spatial descriptions in language: triggering of
  • mental models - simplified reality, cf.:
  • The knife is in front of the vase. The vase is on the left of the glass. The glass is behind the dish.
  • represented symmetrically with equal distances
  • way descriptions: turns always with 90° angle

Temporal cognition

  • temporal phenomena considered:
  • physical time
  • biological time
  • time as a philosophical and abstract concept
  • “perceived” time
  • time in language
  • physical: Newton’s notion of time
  • Absolute, true, and mathematical time, of itself, and from its own nature flows equably without regard to anything external, and by another name is called duration: relative, apparent, and common time, is some sensible and external (whether accurate or unequable) measure of duration by the means of motion, which is commonly used instead of true time; such as an hour, a day, a month.

Temporal concepts in the sciences

  • Einstein’s model of the universe:
  • time considered as the 4th dimension
  • physical processes are reversible
  • paradox of time (Prigogine et al.)
  • biological time: circadian rhythms of living organisms
  • psychological time: cause-effect relationships
  • - inferences base on temporal arrangements of
  • events
  • - “real time” is mediated (lateral inhibition)
  • - segmentation of cognitive processing
  • (cf. slips-of-the-tongue)

Temporal perception and language correlates


4 foci of temporal experience

1. short time intervals

2. estimation of duration

3. simultaneousness

4. temporal perspective (placement of events in past, present, future)


Temporal phenomena in language

categories: grammatical (tense, aspect)

lexical (aktionsarten, temporal adverbs,

temporal conjunctions, temporal


All thoughts are „tensed“ (Higginbotham)

- necessity of representing time in language

space-time metaphor (Langacker)

Western languages: linear paradigm of time

„exotic“ languages: cycles, subcycles


Temporality and tenses

doubts about future

tense as a tense at all:

has modal functions

complexity hypothesis

meaning of tenses:

combination of intrinsic

meaning and contextual


intrinsic meaning: relation between time of event E and a time of reference R;

contextual meaning: relation between time of reference R and time of speech S


Event structure

1. telic, no expansion

(to explode, to flash)

2. telic, linear

(to start)

3. telic, limited

(to arrive)





Event structure

  • 4. telic, beginning and
  • end (to read a novel)
  • 5. telic, iterative
  • (to twiddle)
  • temporal telicity can be
  • parallel to spatial telicity (to arrive)
  • independent of spatial telicity (to explode)




Conceptual development

  • representations: building blocks of cognition and language
  • concepts: the formats of representation
  • hierarchy of concept acquisition
      • 1.spontaneous concepts: bottom-up approach, find abstract, systematic entity
      • 2.scientific concepts: top-down, find concrete grounds
  • children lack ”scientific” concepts
  • ”concepts which are subject to conscious awareness are under voluntary control and form part of an organized system” (Vygotsky)

Conceptualization and language

  • Jackendoff, Langacker: cognitive behavior is primarily a conceptualization of space
  • language encodes iconic-imaginal modi of cognition
  • space concepts are more frequent than time concepts
  • real space is manifested in linguistic phenomena
  • space concepts transform into other concepts through metaphorical extension
  • (read through a text, work on a thesis)
  • base on primary experiences during language development

Conceptual semantics and cognitive grammar

  • shared: encoding of spatial concepts and extension into other conceptual areas
  • languages have hierarchical structure based on metaphors
  • Conceptual semantics
  • (Jackendoff et al.)
  • insists on autonomy of syntax
  • and on formal representation
  • concepts generated mentally on
  • basis of a limited set of primitives,
  • limited principles of combinability
  • Jackendoff: concepts are
  • “finite schemes”
  • Cognitive grammar
  • (Langacker)
  • meaning is conceptualization
  • grammar, lexicon: poles on a
  • continuum
  • semantics: materializes them
  • in different ways

Cognitive grammar

  • model relates language to conceptual world, human experience
  • humans share experiences and biological endowment
  •  idea that physical experiences shaped thinking
  • and language
  • Lakoff: the universal basis of language and cognition
  • is the visual conceptualization of space and
  • movement
  • language therefore reflects fundamental stimuli
  • presupposes categorization which involves conceptual distinctions (night, day)

Categorization and construal

categorization: process of putting together a number of experiences into one conceptual category and relating it to and distinguishing it from, other conceptual categories

construal: a cognitive strategy by which the speaker decides on a particular linguistic alternative in portraying a given conceptualization

ex. use of passive rather than active  focusing on object of action rather than agent


Cognitive basis

force-dynamics: early physical experience of push/ pull/weight/gravity (Talmy)

imaging modes:

1.entities and their spatial relationships among each other perspective on a scene

3.focus on scene (figure and ground)

4.scene conceptualized as a field of forces

expression of concepts: mainly through words, also through grammar (both are poles on a continuum)

expressions base on experience

(gold nugget/gold dust)

special case:  not experience-based names for abstract entities (cf. quark types)