(Sapir-) Whorfian Hypothesis
also known as the linguistic relativity hypothesis, refers to the proposal that the particular language one speaks influences the way one thinks about reality
Linguistic relativity
the ways in which speakers of any given language think are influenced by the language they speak
Strong position
differences in language inevitably cause differences in thought (= unable to think about a topic if we don’t have the relevant words available to us) –> not supported
Intermediate position
language influences certain aspects of cognition such as perception & memory (= language causes preferences but these can be easily eliminated if not useful) –> supported
Language causes a tendency for people to attend to, perceive & remember information in certain ways
- Colour categorisation affected by language
- Odour: less olfaction words in Americans = less able to describe odours as compared to other cultures
- Spatial perception: absolute terms of space in Guugu whatever people vs. relative directions (more recent) of Europeans
- Also connected to time
- Perception of Agency: Spanish vs. English for intentional (= both agentive) vs. unintentional acts (= Spanish non-agentive description)
- Agentive description = better memory, non-agentive to avoid blame
Pragmatic Model
three stages of processing metaphorical & other figurative statements
- asses literal meaning
- decide whether it makes sense in the current context
- if it is inadequate, search for suitable non-literal meaning
Unsupported: predicts that metaphorical meanings are assessed more slowly than literal ones which is not the case
Predication Model
Metaphor understanding consists of 2 components
- Latent semantic analysis component
- Construction-integration component
support: non-reversibility metaphor - e.g. my surgeon is a butcher ≠ my butcher is a surgeon
- Latent semantic analysis component
represents meanings of words based on their relations with other words
- Non-directional process of finding common meanings
- Construction-integration component
use info from first step to construct interpretations of statements –> find relevant features (e.g. lawyers are sharks –> aggressive is relevant but not fins or swimming so inhibited)
- Directional process from argument (e.g. lawyer) being projected to predicate (e.g. sharks)
Common ground
- shared knowledge & beliefs possessed by a speaker & a listener that facilitates communication
- Major goal of conversation: increasing & extending common ground
Egocentric Heuristic
- strategy used by listeners in which they interpret what they hear based on their own knowledge rather than based on common ground
- Use this effortless heuristic rather than the effortful way of figuring out the knowledge existing on common ground
Discourse
speech or written speech at least several sentences long
The main differences between single sentences and those within discourse
- single sentences are much more likely to be ambiguous
2. discourse processing typically involved drawing inferences to make sense of what we are listening to or reading
Three main types of inferences
- logical inferences
- bridging inferences
- elaborative inferences
- Logical inferences
depend only on the meaning of words
- Bridging inferences
establish coherence between the current part of the text and the preceding text, and so are also known as backward inferences
- anaphor resolution (pronoun/noun has to be linked to previous one)
- causal inferences (work out causal relationships between sentences)
- Elaborative inferences
embellish or add details to the text by using world knowledge to expand on the information it contains
– Forward inferences are an important type of elaborative inference that involves anticipating the future
Constructivist approach
readers construct mental models of witch & events described
Minimalist Hypothesis
Inferences are either automatic (only a few!) or strategic (= goal directed)
Theoretical Approaches
- schema theory
- construction integration model
- event-index model
- event segmentation theory
Schema theory
an organised mental representation of information / a cluster of inter-related concepts that tell us about how things function in the world
- Schemas contain much of the information we need to understand what we heard & read and allow us to form expectations
- 3 types of error:
- rationalisation: error in story recall that conform to one’s cultural expectations (either schematic distorts comprehension or retrieval processes
- levelling: omitting unfamiliar details from recall
- – sharpening: selecting certain details for embellishment
support:
The later the recall the more they wrongly believed that a sentence about Jews was part of the Hitler story they were given to remember tend to remember things that weren’t there just because they’re schema consistent
Construction-integration model
- Turn sentences in text into propositions representing its meaning
- Constructed propositions are stored briefly together with associatively related propositions (inferences
- Integration process: Spreading activation process selects propositions for the text representation (context involved to weed out irrelevant propositions top-down!)
- 3 levels of representations are constructed
Event-index model
about narrative rather than expository texts
- Monitor five aspects of a story to see whether their situation-model needs updating
Event segmentation theory
Updating of a situation model can take 2 main forms:
- – 1. Incremental updating of individual dimensions (as said in EIM) “brick by brick”
- – 2. Global updating in which the whole model is replaced “from scratch
- Continuous actions are segmented into events (supported: within event memory better than across event memory)
