N

Implicit vs. Explicit Memory in Animals and Humans Lecture

Key Memory Systems

  • Explicit / Declarative Memory
    • Mediated by the Medial Temporal Lobe (MTL) Memory System.
    • Core structures: hippocampus, rhinal cortex (entorhinal, perirhinal, parahippocampal cortex).
    • Supports one-trial, flexible, conscious learning of facts & events.
  • Implicit / Non-declarative Memory
    • Multiple, distributed neural circuits; no single hub.
    • Includes skills, habits, priming, classical conditioning, emotional learning, unconscious decision biases, etc.
    • Early work (e.g., Kandel’s sea-slug studies, 1968) showed classical conditioning can be implemented by different micro-circuits depending on the learning context.

Lesion Methodology (“an lesion”)

  • In non-human studies, “lesion” = surgical ablation or destruction of a brain area, typically the hippocampus.
  • Purpose: determine whether a task requires that structure.
  • Animal models used here: rats and monkeys (Macaca).

Experimental Tasks Used to Probe Memory Systems

Pattern Discrimination (PD)

  • Setup: Subject chooses between two simple visual patterns (e.g., an X vs L); one is always rewarded.
  • Human performance: one-trial learning via explicit memory — hippocampal & rhinal dependent.
  • Monkey performance:
    • Requires hundreds of trials; gradual stimulus–response habit learning.
    • Survives hippocampal lesions ⇒ relies on implicit systems (basal ganglia, striatum) rather than MTL.
  • Significance: Demonstrates that the same behavioral task can tap different neural systems across species.

Delayed Non-Matching (or Matching) to Sample (DNMS / DMS)

  • Trial structure:
    1. Single sample object presented; subject moves it, receives reward.
    2. Delay inserted (opaque screen up).
    3. Two objects appear: the original sample plus a novel object.
    4. DNMS rule: subject must choose the non-matching (novel) object to get reward.
  • Critical variables: delay length (seconds→minutes), novelty, object–location shuffling.
  • Findings:
    • Both humans & monkeys need intact hippocampus → MTL lesion impairs performance.
    • Hence DNMS demands declarative / relational memory, not pure habit learning.
    • Illustrates evolutionary continuity of hippocampal function.

Concurrent Discrimination (CD)

  • Design:
    • 8 independent object pairs shown in pseudo-random order within a session.
    • Within each pair, one object is always correct (rewarded).
    • 40 trials per day, repeated across days.
  • Performance profiles:
    • Humans: master mapping rapidly; can flexibly list “all 8 correct objects” ⇒ explicit knowledge.
    • Monkeys: learn slowly, by habit; cannot flexibly rearrange or verbally report knowledge.
    • MTL lesion effect: No deficit in monkeys (habit based); large deficit in humans (explicit based).
  • Key paradox: Same hippocampal structure supports task in humans but not in monkeys, prompting evolutionary analysis.

Comparative Conclusions (Squire et al., 2006 review)

  • Examined data across rats, monkeys, humans.
  • Core claim: “Evolutionary continuity” — hippocampus performs comparable relational/declarative computations in all mammals but is recruited by a broader task set in humans.
  • Some tasks (e.g., DNMS) recruit hippocampus in both rats and primates; others (e.g., CD) only in humans because primates solve them via non-declarative routines.

Implicit Memory: Neural Landscape (Schacter & colleagues, 2014)

  • Virtually any cognitive process can run unconsciously.
  • Examples & probable circuits (subset):
    • Unconscious error detection → anterior cingulate & insula.
    • Priming of key-press motor responses → motor cortex, cerebellum.
    • Subliminal threat processing → amygdala & superior colliculus.
  • Quote: “Virtually any brain process can operate without awareness.”

Amygdala – Dual-Route Emotional Processing

  • Quick-and-Dirty Route (Low Road)
    • Path: sensory thalamus → amygdala.
    • Fast, coarse, unconscious; triggers reflexive defensive actions (e.g., braking before awareness).
  • Slow-but-Accurate Route (High Road)
    • Path: sensory thalamus → sensory cortex → amygdala.
    • Adds cortical analysis, still precedes full conscious awareness but carries richer information.
  • Despite adjacency to hippocampus, amygdala is not part of the MTL memory system; it modulates both implicit (fear conditioning) and explicit (emotional tagging) learning.

Practical & Philosophical Implications

  • Clinical: Understanding species-specific reliance on hippocampus guides animal models of amnesia.
  • Ethical: Lesion studies highlight balance between scientific gain and animal welfare.
  • Conceptual: Reminds us that identical behaviors need not reflect identical cognitive mechanisms (multiple realizability).

Key Quantitative Facts & Parameters

  • Lesion = hippocampal ablation in rats/monkeys.
  • Pattern Discrimination: “hundreds” of acquisition trials in monkeys.
  • Concurrent Discrimination: 8 object pairs, 16 total objects, 40 trials/day.
  • Historical milestones: sea-slug work 1968; comparative hippocampal review 2006; implicit-process compendium 2014.

Summary

  • Explicit memory: single integrated MTL circuit; conserved across mammals.
  • Implicit memory: heterogeneous, circuit-specific, often subcortical; encompasses everything from motor habits to unconscious number processing.
  • Behavioral tasks (PD, DNMS, CD) reveal which species & which brain systems are engaged, clarifying when hippocampus is essential and when habits suffice.
  • Amygdala exemplifies implicit processing via dual thalamo-cortico routes, mediating both reflexive and nuanced emotional responses.
  • Overall evidence supports evolutionary continuity: same hippocampal computations, diversified deployment in humans.