Cognitive Psychology
Cognitive psychology is the scientific study of mental processes -- how people attend to information, encode and retrieve memories, perceive the world, make decisions, understand and produce language, and solve problems. It emerged as a distinct discipline in the 1950s and 1960s when researchers rejected strict behaviorism's refusal to study internal mental states and began modeling the mind as an information-processing system. The "cognitive revolution" drew on information theory (Shannon), linguistics (Chomsky), and early computer science to build testable models of thought.
Agent affinity: kahneman (decision-making, heuristics, biases), james (attention, stream of consciousness)
Concept IDs: psych-perception-construction, psych-attention-memory, psych-cognitive-biases, psych-brain-structure
The Cognitive Architecture at a Glance
| # | Domain | Core Questions | Key Figures | |---|---|---|---| | 1 | Attention | How do we select what to process? Why do we miss things in plain sight? | Broadbent, Treisman, Posner | | 2 | Memory | How do we encode, store, and retrieve information? Why do we forget? | Atkinson & Shiffrin, Baddeley, Tulving, Ebbinghaus | | 3 | Perception | How does the brain construct a stable world from ambiguous sensory data? | Gibson, Gregory, Gestalt school | | 4 | Decision-making | How do we choose under uncertainty? What systematic errors do we make? | Kahneman & Tversky, Simon, Gigerenzer | | 5 | Language | How do we understand and produce speech? What is the structure of linguistic knowledge? | Chomsky, Pinker, Broca, Wernicke | | 6 | Problem-solving | How do we reach goals when the path is not obvious? | Newell & Simon, Duncker, Wertheimer |
Domain 1 -- Attention
Core principle: Attention is the cognitive bottleneck. The world presents far more information than the brain can process, so attention selects what gets through and what is ignored.
Selective attention
The ability to focus on one source of information while filtering others. Broadbent's (1958) filter model proposed that unattended information is blocked early -- before meaning is extracted. Treisman's (1964) attenuation model modified this: unattended information is turned down, not blocked, so personally relevant stimuli (your name) can break through. The cocktail party effect demonstrates this daily.
Experiment to know. Cherry (1953) -- dichotic listening. Participants heard different messages in each ear and repeated one aloud (shadowing). They could report almost nothing about the unattended ear except gross physical changes (gender of speaker). But Moray (1959) showed that the participant's own name in the unattended ear was detected about one-third of the time.
Divided attention
Performing two tasks simultaneously. Performance depends on task difficulty, practice, and resource overlap. Kahneman's (1973) capacity model treats attention as a limited pool of mental energy allocated flexibly. Tasks that draw on different processing resources (e.g., driving and talking) interfere less than tasks that share resources (e.g., reading and listening to speech).
Practical implication. Multitasking is mostly rapid task-switching, not true parallel processing. Each switch incurs a cognitive cost, which is why texting while driving is dangerous even when the driver believes they are managing both tasks.
Sustained attention (vigilance)
Maintaining focus over extended periods. Vigilance declines within 15-30 minutes on monotonous detection tasks (Mackworth, 1948 -- the clock test). Signal detection theory (Green & Swets, 1966) formalizes this as changes in the criterion (beta) rather than sensitivity (d-prime): people become more conservative about reporting signals rather than losing the ability to detect them.
Inattentional blindness and change blindness
Simons and Chabris (1999) showed that roughly 50% of observers fail to notice a person in a gorilla suit walking through a basketball-passing scene. Change blindness (Rensink, O'Regan, & Clark, 1997) shows that large changes in visual scenes go undetected when the change occurs during a brief disruption. These findings demonstrate that attention is not merely helpful but constitutive of conscious perception.
Domain 2 -- Memory
Core principle: Memory is not a single system but a collection of interacting systems with different capacities, durations, and encoding mechanisms.
The multi-store model
Atkinson and Shiffrin (1968) proposed three stores:
- Sensory memory. Brief (<1 second for iconic/visual, ~3 seconds for echoic/auditory) high-capacity storage of raw sensory input. Demonstrated by Sperling's (1960) partial report procedure.
- Short-term memory (STM). Limited capacity (~7 plus/minus 2 items, Miller 1956), duration ~15-30 seconds without rehearsal. Maintained by rehearsal.
- Long-term memory (LTM). Essentially unlimited capacity and duration. Requires encoding (elaborative rehearsal is more effective than maintenance rehearsal).
Working memory
Baddeley and Hitch (1974) replaced the passive STM concept with working memory -- an active workspace with multiple components:
- Central executive -- attentional control system that directs processing
- Phonological loop -- maintains verbal/acoustic information via subvocal rehearsal
- Visuospatial sketchpad -- maintains visual and spatial information
- Episodic buffer (added 2000) -- integrates information across subsystems and links to LTM
Working memory capacity predicts academic achievement, reading comprehension, and fluid intelligence more strongly than IQ in many studies (Alloway & Alloway, 2010).
Long-term memory systems
Tulving (1972) distinguished:
- Episodic memory -- personal events ("my 10th birthday party")
- Semantic memory -- general knowledge ("Paris is the capital of France")
Squire (1992) broadened the taxonomy:
- Declarative (explicit) -- episodic + semantic, requires conscious recall
- Non-declarative (implicit) -- procedural skills, priming, classical conditioning, does not require conscious recall
Patient H.M. (Scoville & Milner, 1957) demonstrated the dissociation: after bilateral hippocampal removal, H.M. could not form new declarative memories but could learn new motor skills.
Forgetting
Ebbinghaus (1885) established the forgetting curve: rapid initial loss followed by a gradual plateau. Theories of forgetting include:
- Decay -- trace deteriorates over time (weak evidence for LTM)
- Interference -- proactive (old memories block new) and retroactive (new memories block old)
- Retrieval failure -- the memory exists but the cue is insufficient (Tulving's encoding specificity principle)
- Motivated forgetting -- suppression (conscious) and repression (unconscious, contested)
Encoding strategies
- Levels of processing (Craik & Lockhart, 1972) -- deeper (semantic) processing produces stronger memories than shallow (structural) processing
- Elaboration -- connecting new information to existing knowledge
- Spacing effect -- distributed practice beats massed practice (Cepeda et al., 2006)
- Testing effect -- retrieval practice strengthens memory more than re-study (Roediger & Karpicke, 2006)
Domain 3 -- Perception
Core principle: Perception is construction, not recording. The brain actively interprets incomplete and ambiguous sensory data to build a coherent model of the world.
Bottom-up vs. top-down processing
Gibson's (1979) ecological approach emphasizes bottom-up processing: the environment provides rich, sufficient information (affordances) and the perceiver picks it up directly. Gregory's (1970) constructivist approach emphasizes top-down processing: perception requires inference, expectation, and prior knowledge to fill in gaps. Both contribute. The Muller-Lyer illusion persists even when you know the lines are equal (bottom-up domin