Language and Cognition: Priming, Consciousness, Alignment, and Prediction
Priming and Reaction Time
Words that sound identical but have different spellings and meanings (e.g., "vaccine" vs. "pronouns" example refers to homophones/homographs not explicitly defined but implied by "sound exactly the same, maybe they're spelled differently, but they mean different things").
Priming effects are explained through lexical and semantic networks.
Facilitatory Priming: When a prime makes a target easier to process, reaction time will decrease.
Inhibitory Priming: When a prime makes a target harder to process, reaction time will increase.
Stimulus Onset Asynchrony (SOA): This term refers to the temporal distance between the prime and the target. It is heavily used in primary literature.
Optimal facilitation often occurs around milliseconds.
Presenting the prime and target simultaneously can cause problems, requiring inhibition of the unwanted word, especially when words are related.
Priming effects are predictable and can be modeled computationally, showing good alignment between real data and computational predictions.
Conscious vs. Unconscious Priming
Priming can be conscious, but the most typically studied and impactful priming is unconscious.
Consciousness of a stimulus generally emerges after about to milliseconds of exposure.
Shortest recorded conscious perception: Around milliseconds (described as incredibly brief).
Evidence for Unconscious Priming:
Observed in individuals with regional brain damage (e.g., anterograde amnesia, lesion, hemispheric blindness).
These instances demonstrate priming occurring below the level of conscious experience.
The Nature of Conscious Visual Experience
Saccades / Eye Movements: Our eyes constantly make rapid, unconscious movements, even when we feel like we're perceiving a whole scene at once.
Limited Focus: Physically, we can only focus on one point at a time because light hits a very small scope of the retina.
Color Perception: Counter-intuitively, we can only see one color at a time. While it feels like we see many colors simultaneously, this is due to rapid eye movements (saccades) and the brain integrating these "slices" of perception in working memory over approximately millisecond intervals.
Structural Priming in Language
Refers to the tendency to repeat and process faster language structures that have been encountered recently (or even not so recently).
Theories of Structural Priming:
Increased Fluency: Exposure to a structure makes its subsequent use more fluent and faster. However, fluency effects are often short-lived (e.g., within a minute).
Implicit Learning: The theory, often supported by computational modeling, suggests that we implicitly learn linguistic structures through probabilistic exposure. This is essential for language acquisition in infants. Evidence for implicit learning in structural priming includes its long-lasting effects (sometimes a day later), in contrast to short-lived fluency effects.
Bach and Chang (2000) Study: Showed structural priming effects persisted even with or intervening neutral sentences between the prime and the target, indicating a more robust, long-term learning process.
Inverse Preference Effect: Things that are less common are learned more effectively or produce stronger priming effects. If an individual encounters a less frequent word (e.g., "bile" vs. a common word), the priming effect is significantly stronger. This effect arises from mathematical models involving implicit learning and relates to the cognitive effort expended: learning something new consumes more energy than re-activating existing knowledge.
Implicit Nature of Structural Priming:
Seen in anterograde amnesia patients, who still exhibit structural priming despite impaired explicit memory.
Experimental manipulation: Participants are given a priming task, then asked which prime-target relationships they noticed. Even after removing data where participants consciously noticed the prime, significant priming effects persist in the remaining unnoticed (implicit) trials.
Alignment
Alignment is a social phenomenon where individuals align their representations, particularly in conversation, to ensure mutual understanding and facilitate interaction.
Levels of Alignment: Phonological processes, syntax, semantics, and overall mental models (situation models) align between conversational partners (Agent A and Agent B).
Lexical and Structural Alignment: People align on word choices and even grammatical structures. Structural alignment is synonymous with structural priming.
Cross-Linguistic Alignment: Bilingual individuals can exhibit structural priming transfer between languages (e.g., hearing a structure in English and then using a similar structure when speaking Spanish).
Perception Behavior Expressway: This concept links perception to automatic behavioral mimicry, contributing to alignment. It's related to:
Mirror Neurons: Neurons in the frontal lobe activate both when an individual performs an action and when they observe someone else performing the same action. This has been studied in monkeys (e.g., watching someone reach and grab activates corresponding motor neurons).
Intention Reading: Simply perceiving the intention to perform an action (e.g., reaching for an object, even if the actual grasping isn't seen) is enough to activate associated motor neurons.
Mimicry: Humans unconsciously mimic observed behaviors such as facial expressions (smiling), movements, gestures, tone of voice, and posture, especially when they feel socially aligned with others. This can be viewed as a form of empathy.
Afferent and Efferent Networks & Afference Copies
Peripheral Nervous System: Comprises afferent and efferent nerve networks extending throughout the body.
Afferent Nerves: Carry sensory information from body parts to the brain (sensory input).
Efferent Nerves: Carry motor commands from the brain to body parts (motor control).
Tandem Operation: These networks work together, with the brain making predictions about sensory feedback based on motor commands.
Afference Copies (Efference Copies): These are internal predictions or copies of the motor commands sent, which the brain uses to anticipate how a movement or action should feel.
Purpose: Allows for rapid adjustment if the actual sensory input deviates from the prediction.
Examples:
Unexpected Step Down: If walking and stepping down when expecting flat ground, the lack of expected sensory feedback (afference copy mismatch) causes quick adjustment of posture to regain balance.
Light Prop: When picking up a prop that looks heavy but is unexpectedly light, the afference copy predicts a certain weight, and the mismatch causes an over-exertion (e.g., throwing it up too high) requiring immediate adjustment.
Tickling Oneself: One cannot tickle oneself effectively because the brain generates an afference copy of the expected sensation, which cancels out the actual sensory input, thus suppressing the ticklish feeling. This is a computational problem of cognition solved by afference copies.
Dialogue and Conversation: Rely on afference copies, as individuals predict how their speech will be perceived by the listener, essentially putting themselves in the listener's shoes.
Garrod and Pickering's Theory: Provides a theory of alignment based on afference copies.
Language Comprehension and Prediction
Language comprehension heavily relies on continuous prediction at various levels:
Grammatical level
Phonological level
Lexical level
Conceptual level
N400 (Negative Electrical Impulse at ms): A neurological measure (detectable via EEG) indicating cognitive surprise or unexpected semantic input.
Example: "On a windy day, the kid went out to fly a kite."
Expected: "kite" after "windy day" and "fly a…".
Unexpected: If the sentence continues "…to fly an airplane," a N400 is observed at the word "airplane" because it's semantically surprising.
Even More Surprising: A N400 is already observed at the word "an" (before "airplane"). This is because the prediction for "kite" requires the article "a" (e.g., "a kite"), not "an". The brain registers surprise at the phonological level as soon as "an" is heard, indicating a highly predictive process.
Prediction and Alignment: Gerrard and Pickering's theory suggests that when we predict other people's speech, we use the same neural mechanisms involved in planning our own speech. This implies that we are effectively putting ourselves in their shoes, using an alignment and afference copy model to anticipate their utterances.
Course Logistics and Project Brainstorming
Exam Rescheduling: The mid-term exam moved back to occur after six weeks of "language and cognition" content, rather than after five weeks. This aligns the two major exams more evenly (six weeks each) and groups content logically.
The first six weeks cover language and cognition.
Subsequent weeks cover language and society (dialect, identity, miscommunication, power).
Final Project Brainstorming: The instructor facilitated brainstorming for final project topics and group formation.
Guest lecturer: Andrea Beltrama (teaches LING Sociolinguistics) will speak later in the semester. No students in the class had taken LING previously.
Topics of interest discussed:
Cognitive aspects of language
Dialect perception/evaluation/judgment
Speech-language pathology (SLP) angles
Gesture and sign language
Dialogue or conversation as a cognitive problem
Language and identity
Propaganda
Creolization / Creoles
Language conflict and language as a human right
Group Formation Strategy: Encouraged students to talk to new people, not just their usual group, to discuss potential project interests.