Comprehensive Notes on Attention, Memory, Psychophysics & Theories of Consciousness

Attention, Learning, and Consciousness

• Attention as a bridge to learning
  • Publicly observable: you can call or capture someone’s attention.
  • Privately directed: the agent can voluntarily focus.
  • Hence: external (stimulus-driven) vs. internal (goal-driven) control.
• Why couple learning with consciousness?
  • Machines detect/identify (e.g., coffee cups) without qualia.
  • Central question: What extra benefit does a conscious percept provide to an organism?
• Attention without consciousness
  • Orienting reflexes (inferior/superior colliculi) are rapid, below awareness.
  • Reticular Activating System (RAS) can raise arousal before we “notice” it.
  • Possible daily examples: driving on “autopilot,” masked priming, cocktail-party leakage.

Memory Architecture (Sensory → STM → LTM)

• Sensory Memory (iconic, echoic, etc.)
  • Massive capacity, modality-independent.
  • Parallel processing, automatic, reflex-like transduction.
  • Sperling’s partial-report & signal-detection work quantified capacity.
• Encoding & Consolidation
  • Differentiation → integration → object representation.
  • Continues even without attention; forms stable reality backdrop.
• Short-Term / Working Memory (STM/WM)
  • Capacity ≈ Miller’s $7\,\pm\,2$ discrete objects; hierarchical chunking expands it.
  • Serial processing bottleneck: limits come from both sensory-to-STM and LTM-to-STM traffic.
  • Rehearsal = “mental hands” manipulating items; inhibits decay/interference (flattens serial-position curve, counters Brown-Peterson loss).
  • WM operations: retrieve from LTM, keep current percepts active, run mental experiments, apply rules (phonological loop, visuo-spatial sketchpad).
• Long-Term Memory (LTM)
  • Declarative → semantic & episodic; procedural builds via repetition/frequency (Pavlov/Skinner).
  • Associative networks (nodes) update to maintain reality consistency.
  • Once encoded, little or no decay—contrast with sensory trace.
• Implicit vs. Explicit
  • Without consciousness ⇒ implicit.
  • With attentional spotlight/WM ⇒ explicit (reportable).

Information-Processing & Bottleneck Models

• Broadbent filter/attenuator: complete architecture works without invoking consciousness.
  • Highlights limited capacity = attention definition.
  • Raises: Where are qualia? What flips the “ON” switch?
• Global Workspace (Baars, Dehaene/Tononi)
  • Field of consciousness = large tabletop; WM spotlight roams.
  • Phonological loop + space-time buffer act as staging area for explicit reporting.
  • Qualifies as posterior→anterior flow: sensory hubs → executive control (Posner’s model).
• Neural Quanta / Synchrony
  • Conscious moment arises when enough neurons fire in synch to cross threshold (~400 ms after stimulus; Libet, Lamme, Crick & Koch).
  • Raises continuity debate: physical spikes are discrete, psychological experience feels continuous.

Psychophysics & Measurement

• Absolute Threshold & JND
  • $\frac{\Delta S}{S}=k$ (Weber) → Fechner integration yields $\Psi = k\,\ln S$.
  • Square-wave expectation became sigmoidal reality: no sharp conscious cutoff; 50 % detection point defines threshold.
• Stimulus-driven methods: Method of limits, adjustment, staircase, masking (backward & forward).
• Top-down insertion: Nominal, ordinal, interval, ratio scales; using 1–10 confidence ratings embodies prior knowledge.
• Dependent variables: RT, accuracy, EEG evoked potentials; subjective report remains crucial (links self to data).

Classical Experimental Effects

• Cocktail-Party / Dichotic Shadowing
  • Attention to one ear; ignored ear still leaks (implicit priming).
• Masking
  • Stimulus followed by mask → weak or no conscious detection, yet primes behaviour.
• Stroop Interference
  • Congruent < Neutral < Incongruent RT; reveals conflict between automatic reading & color-naming.
• Figure/Ground Ambiguities (Faces–Vase, scrambled text)
  • Same physical input → multiple conscious constructions; WM selects one, others suppressed.

Philosophical & Theoretical Stances

• Materialism / Identity Theory
  • Neural pattern is experience; transplant pattern ⇒ transplant qualia.
  • Delgado’s electrode work, Pribram’s memory transfers: mixed success.
• Functionalism & Structuralism: realizer-independent causal roles; focus on information patterns.
• Property Dualism: single substance, two property sets (mental & physical).
• Panpsychism / Idealism / Solipsism: mind-like properties ubiquitous or foundational.
• Epiphenomenalism (Chalmers “hard problem”): consciousness emerges but lacks causal power.
• Key issue: Can consciousness cause material change, or merely accompany it?

Neurobiological Highlights

• Microtubule ORCH-OR (Hameroff & Penrose)
  • Quantum coherence in tubulin pockets; anesthesia binds hydrophobic sites → immediate loss of consciousness.
  • Contrasted with synesthesia (over-binding of sensory streams).
• Basal Ganglia + Hippocampus expand bottleneck; Kahneman & Tversky demonstrated effort/decision cost.
• EEG/MEG Synchrony: $\gamma$ (~40 Hz) binding correlates; not yet sufficient for qualia.

Clinical & Real-World Applications

• Neglect Syndromes: visual, auditory, tactile—damage yields half-table consciousness loss.
• Glasgow Coma Scale (GCS) and Rappaport Coma/Near-Coma Scale
  • Eye, verbal, motor scores → 3–15 scale; clinician as IV, patient response as DV.
• Anesthesia vs. Consciousness
  • Dose-dependent, reversible; demonstrates “off switch” without structural brain loss.
• Mindfulness & Self-Directed Neuroplasticity
  • Intention/volition modulate WM, enhance LTM encoding; education & therapy exploit this (e.g., LearnSmart adaptive quizzing asks “what state are you in?”).

Concept of “Self” and Remaining Hard Problems

• Experiments need a subject to interpret scales ⇒ “I” becomes variable.
• Self-report intertwines top-down context, making variance hard to partition.
• Open questions:
  • Does consciousness add causal efficacy during psychophysical transduction?
  • How do emergent properties arise from thermodynamically closed, law-obeying matter?
  • Can we empirically isolate “qualia generators,” or must we reframe hard problem into easier sub-problems?

Study/Exam Tips & Connections

• Map each memory stage to: capacity, duration, code, key researchers.
• Be able to derive/interpret $\frac{\Delta S}{S}=k$ and relate to JND curves.
• Contrast bottom-up (masking, orienting) with top-down (shadowing, Stroop).
• Know clinical scales’ IV/DV logic; link to consciousness theories.
• Practice explaining why consciousness might be evolutionarily adaptive despite computational sufficiency of unconscious processing.