Biological Memory: A GRM/Base-T Reconstruction

Part I — Scope: What Counts as Memory?

1. The Boundary Problem

1. What This Book Means by Biological Memory

1.1 Nervous-system memory as the target domain
1.2 Why immune “memory” is outside scope
1.3 Shared terminology does not establish shared mechanism
1.4 Memory as experience-dependent modification of later reconstruction
1.5 Memory as history acting on future behavior
1.6 Storage, reconstruction and readout as different problems
1.7 Memory is not a photograph of the past
1.8 Memory is not equivalent to behavior
1.9 Memory is not equivalent to conscious recall
1.10 Memory is not equivalent to consolidation

2. The Central Architecture

2.1 Experience
2.2 Encoding
2.3 Distributed relational change
2.4 Immediate reuse
2.5 Replay
2.6 Reorganization
2.7 Multiple access paths
2.8 Contextual reconstruction
2.9 Narrow control layer
2.10 Behavioral or conscious readout

Core progression:

\[ \boxed{ \text{experience} \rightarrow \text{distributed encoding} \rightarrow \text{replay/reconstruction} \rightarrow \text{multiple access paths} \rightarrow \text{controlled readout} } \]

3. Four Geometries

3.1 \(G_{\rm src}\): physical biological state
3.2 \(G_{\rm sem}\): learned relational geometry
3.3 \(G_{\rm acc}\): currently executable accessibility geometry
3.4 \(G_{\rm rd}\): task-bounded readout
3.5 \(G_{\rm src}\neq G_{\rm sem}\)
3.6 \(G_{\rm sem}\neq G_{\rm acc}\)
3.7 \(G_{\rm acc}\neq G_{\rm rd}\)
3.8 Physical change does not identify semantic change
3.9 Failed retrieval does not identify erasure
3.10 Successful retrieval does not identify a unique physical carrier

Part II — Before Memory: Developmental Sequencing

4. Development Is Not Memory

4.1 Genome-directed state transition
4.2 Embryonic differentiation
4.3 Morphogen gradients
4.4 Chromatin locking
4.5 Neural circuit development
4.6 Sensorimotor architecture
4.7 Reflex organization
4.8 Inherited behavioral sequences
4.9 Persistent developmental state
4.10 Why persistence alone does not define memory

5. The Chick as a Developmental Model

5.1 Precocial development
5.2 Neural organization before hatching
5.3 Motor readiness
5.4 Sensory readiness
5.5 Pecking architecture
5.6 Locomotor architecture
5.7 Orientation biases
5.8 Hatching sequence
5.9 Immediate competence after emergence
5.10 Memory initially plays only a small supporting role

6. Hatching

6.1 Internal pip
6.2 External pip
6.3 Repositioning
6.4 Repeated strike sequence
6.5 Shell fracture
6.6 Rotation
6.7 Emergence
6.8 Postural transition
6.9 Hatching as inherited sequencing
6.10 Why complex ordered behavior need not be learned

7. Locomotion

7.1 Locomotion as extended sequencing
7.2 Posture
7.3 Balance
7.4 Limb coordination
7.5 Heading
7.6 Recent motor state
7.7 Self-generated sensory change
7.8 Continuous sensorimotor prediction
7.9 Spatial extension of temporal sequence
7.10 Locomotion creates conditions in which memory becomes useful

Part III — Exploration and the Functional Emergence of Memory

8. Exploration

8.1 Leaving the inherited action envelope
8.2 Novel surfaces
8.3 Novel objects
8.4 Novel consequences
8.5 Orienting
8.6 Approaching
8.7 Pecking
8.8 Withdrawal
8.9 Repetition
8.10 Exploration creates individual history

9. Socially Guided Exploration

9.1 Conspecific attention
9.2 Social facilitation
9.3 Chicks copying chicks
9.4 Where to orient
9.5 What to peck
9.6 When to act
9.7 Social selection of experience
9.8 Individual consequence following copied action
9.9 Learning without adult instruction
9.10 Social environment as an access and sampling structure

10. Action–Consequence Learning

10.1 Action before knowledge
10.2 Contact
10.3 Taste
10.4 Water
10.5 Food
10.6 Positive consequence
10.7 Negative consequence
10.8 Repetition
10.9 Avoidance
10.10 Experience begins restructuring subsequent action

11. The Earliest Nervous-System Memory

11.1 Recent interaction affecting the next action
11.2 Retained action–outcome relations
11.3 Short-lived association
11.4 Temporal comparison
11.5 Sequence dependence
11.6 Recent-choice history
11.7 Recent-outcome history
11.8 Short-range prediction
11.9 Immediate reconstruction
11.10 Memory begins as usable experience, not as long-term archive

Part IV — Continuous Memory Formation

12. Memory Is Formed Continuously

12.1 Encoding begins during experience
12.2 Immediate neural modification
12.3 Immediate behavioral availability
12.4 Memory before consolidation
12.5 Replay within seconds
12.6 Modification during repeated exposure
12.7 Retrieval during continuing acquisition
12.8 Overlapping traces
12.9 Continuous rewriting
12.10 No discrete boundary between “not memory” and “consolidated memory”

13. Consolidation Reframed

13.1 Consolidation is not memory birth
13.2 Stabilization
13.3 Redistribution
13.4 Reweighting
13.5 Replay
13.6 Rehearsal
13.7 Reorganization
13.8 Integration with prior structure
13.9 Long-term persistence
13.10 Consolidation as one trajectory within an already existing memory

14. Replay

14.1 Immediate replay
14.2 Offline replay
14.3 Wake replay
14.4 Sleep replay
14.5 Repetition
14.6 Sequence compression
14.7 Reweighting
14.8 Integration
14.9 Prediction
14.10 Replay changes the memory that it reactivates

15. Reconsolidation

15.1 Retrieval destabilization
15.2 Re-entry into active processing
15.3 Modification
15.4 Updating
15.5 Context incorporation
15.6 Competition
15.7 Strengthening
15.8 Weakening
15.9 Transformation
15.10 Memory as continuously revisable structure

Part V — Relational Memory Geometry

16. Memory Is Relational

16.1 Event
16.2 Place
16.3 Time
16.4 Person
16.5 Object
16.6 Action
16.7 Outcome
16.8 Emotion
16.9 Sequence
16.10 Context

17. One Event, Many Relations

17.1 Spatial path
17.2 Temporal position
17.3 Social relation
17.4 Sensory relation
17.5 Motor relation
17.6 Emotional relation
17.7 Linguistic relation
17.8 Goal relation
17.9 Causal relation
17.10 Multiple relations create multiple future access paths

18. Multiple Access Paths

18.1 Place cue
18.2 Person cue
18.3 Smell cue
18.4 Sound cue
18.5 Emotion cue
18.6 Sequence cue
18.7 Motor cue
18.8 Conceptual cue
18.9 Linguistic cue
18.10 Temporal cue

A memory need not have one address:

\[ m\leftarrow\{a_1,a_2,\ldots,a_n\}. \]

19. Reconstruction

19.1 Partial cue
19.2 Network intersection
19.3 Candidate activation
19.4 Pattern completion
19.5 Context constraint
19.6 Competition
19.7 Suppression
19.8 Current-state contribution
19.9 Reconstructed output
19.10 Retrieval is not literal replay

20. Semantic Cloud

20.1 Distributed relational field
20.2 Local basins
20.3 Overlapping memories
20.4 Shared components
20.5 Hierarchy
20.6 Categories
20.7 Context
20.8 Recurrent structure
20.9 Temporary activation
20.10 Current semantic cloud as reconstruction, not stored screen image

Part VI — Physical Implementation

21. Synaptic Plasticity

21.1 Presynaptic change
21.2 Postsynaptic change
21.3 Receptor trafficking
21.4 Spine morphology
21.5 Short-term potentiation
21.6 Long-term potentiation
21.7 Depression
21.8 Metaplasticity
21.9 Heterosynaptic effects
21.10 Synaptic change as one carrier class

22. Intrinsic and Dendritic Plasticity

22.1 Excitability
22.2 Threshold changes
22.3 Dendritic integration
22.4 Local dendritic events
22.5 Branch-specific modification
22.6 Ion-channel regulation
22.7 Temporal integration
22.8 Homeostatic interaction
22.9 State dependence
22.10 Memory cannot be reduced to synaptic weights alone

23. Molecular Persistence

23.1 Protein phosphorylation
23.2 Kinase loops
23.3 Protein synthesis
23.4 Local translation
23.5 Transcription
23.6 Chromatin change
23.7 Structural maintenance
23.8 Self-reinforcing molecular loops
23.9 Turnover
23.10 Persistence of organization despite replacement of molecules

25. Distributed Memory Does Not Require Duplicate Copies

25.1 Overlapping support
25.2 Distributed relational coding
25.3 Partial participation
25.4 Recurrent reactivation
25.5 Alternative access paths
25.6 Network robustness
25.7 Graceful degradation
25.8 Reconstruction from partial structure
25.9 Physical overlap versus semantic duplication
25.10 Do not equate robustness with \(k\) identical copies

Part VII — Storage Capacity

26. The Capacity Thesis

26.1 No demonstrated fixed long-term memory ceiling
26.2 Continuous acquisition throughout life
26.3 Centenarians continue forming memories
26.4 Storage does not present as a simple filled container
26.5 Capacity versus retrieval efficiency
26.6 Capacity versus interference
26.7 Capacity versus control
26.8 Capacity versus encoding quality
26.9 Capacity versus aging
26.10 Massive long-term capacity under narrow active control

Core claim:

\[ \boxed{ C_{\rm long-term}\gg C_{\rm active-control}. } \]

27. Why Capacity Can Be Enormous

27.1 Distributed coding
27.2 Sparse coding
27.3 Shared components
27.4 Reused relations
27.5 Combinatorial composition
27.6 Hierarchy
27.7 Pattern completion
27.8 Multiple indexing dimensions
27.9 Reconstruction instead of file duplication
27.10 Very large state space from finite biological machinery

28. Capacity Is Not Simultaneous Accessibility

28.1 Storage
28.2 Availability
28.3 Accessibility
28.4 Activation
28.5 Selection
28.6 Control
28.7 Conscious report
28.8 Action
28.9 Different bottlenecks
28.10 Narrow output does not imply narrow storage

29. The Computer Analogy

29.1 Simple storage primitive
29.2 Huge addressable space
29.3 Utility emerges from capacity
29.4 Storage architecture does not encode future applications
29.5 General-purpose substrate
29.6 Programs exploit existing capacity
29.7 Biological analogue
29.8 Experience writes into general plastic machinery
29.9 New uses need not require new storage mechanism
29.10 Implementation \(\neq\) capacity \(\neq\) utility

30. Over-Evolution for Immediate Utility

30.1 Ecological minimum
30.2 Available capacity
30.3 Relational surplus
30.4 Temporal surplus
30.5 Associative surplus
30.6 Symbolic surplus
30.7 Novel-domain learning
30.8 Exaptation
30.9 General capability exceeding any one ancestral task
30.10 Utility constrains use, not necessarily total potential capacity

Part VIII — Narrow Control Layer

31. Working Memory

31.1 Active basin
31.2 Short-term stabilization
31.3 Attention
31.4 Independent bindings
31.5 Interference
31.6 Rehearsal
31.7 Task switching
31.8 Limited active rank
31.9 Rich stored field versus narrow manipulation window
31.10 Working memory is primarily a control problem

32. Control Rank

32.1 Number of independently maintained relations
32.2 Binding competition
32.3 Attention allocation
32.4 Serial versus parallel control
32.5 Chunking
32.6 Hierarchical grouping
32.7 Automation
32.8 Skill acquisition
32.9 Reduced control demand through practice
32.10 Expertise as compression of control requirements

33. Memory Palace

33.1 Existing spatial structure
33.2 Loci
33.3 Arbitrary bindings
33.4 Sequential traversal
33.5 Externalized internal indexing
33.6 Reduced simultaneous binding demand
33.7 High long-term capacity
33.8 Narrow active traversal
33.9 Practice
33.10 Memory palace as deliberate access-geometry engineering

34. Skilled Performance

34.1 Piano playing
34.2 Typing
34.3 Reading
34.4 Navigation
34.5 Sports
34.6 Tool use
34.7 Chunk formation
34.8 Automatization
34.9 Practice-specific neural geometry
34.10 Specialized readout does not establish a specialized ancestral memory mechanism

Part IX — Forgetting

35. Forgetting Is a Readout Phenomenon Until Proven Otherwise

35.1 Failed spontaneous recall
35.2 Weak cue
35.3 Context mismatch
35.4 Competition
35.5 Inhibition
35.6 State mismatch
35.7 Retrieval-path degradation
35.8 Partial structural change
35.9 Genuine loss
35.10 Behavioral failure alone cannot distinguish them

36. Structured Memory Makes Multiple Recovery Paths Expected

36.1 One event, many relations
36.2 Cue substitution
36.3 Context reinstatement
36.4 Recognition after recall failure
36.5 Relearning
36.6 Savings
36.7 Sensory reinstatement
36.8 Social prompting
36.9 Motor reinstatement
36.10 Recovery through another path is a structural prediction, not an escape clause

37. Memory Loss

37.1 Path loss
37.2 Index loss
37.3 Relational degradation
37.4 Interference
37.5 Network damage
37.6 Hippocampal damage
37.7 Cortical damage
37.8 Progressive neurodegeneration
37.9 True semantic destruction
37.10 Why erasure remains difficult to demonstrate

38. Aging

38.1 Continued memory formation
38.2 Retrieval slowing
38.3 Increased interference
38.4 Reduced control efficiency
38.5 Reduced sensory quality
38.6 Reduced attention
38.7 Encoding variability
38.8 Preserved remote knowledge
38.9 Preserved skill learning
38.10 Aging does not resemble a storage device reaching capacity

Part X — Autobiographical Memory and HSAM

39. Autobiographical Memory

39.1 Event structure
39.2 Self relation
39.3 Temporal indexing
39.4 Spatial indexing
39.5 Social indexing
39.6 Emotional indexing
39.7 Narrative reconstruction
39.8 Rehearsal
39.9 Re-encoding
39.10 Multiple access routes

40. Ordinary Autobiographical Forgetting

40.1 Low retrieval priority
40.2 Competition
40.3 Event similarity
40.4 Narrative compression
40.5 Relevance filtering
40.6 Context dependence
40.7 Sparse spontaneous access
40.8 Recognition after recall failure
40.9 Partial reconstruction
40.10 No need to assume permanent deletion of resolved events

41. HSAM

41.1 Phenotype, not mechanism
41.2 Date indexing
41.3 Rapid autobiographical retrieval
41.4 Involuntary memories
41.5 Rehearsal
41.6 Re-encoding
41.7 Obsessive behavioral traits in some cases
41.8 Access geometry
41.9 Control differences
41.10 HSAM as a natural experiment separating storage from retrieval organization

Part XI — Comparative Memory

42. Same Broad Mechanism, Different Outcomes

42.1 Plasticity
42.2 Association
42.3 Relational encoding
42.4 Replay
42.5 Multiple indexing
42.6 Reconstruction
42.7 Narrow active control
42.8 Architecture differs
42.9 Development differs
42.10 Outcome differs

Core model:

\[ \boxed{ \text{shared mechanism} + \text{architecture} + \text{development} + \text{experience} + \text{social ecology} \rightarrow \text{behavioral phenotype}. } \]

43. Developmental Plasticity

43.1 Social environment
43.2 Species-typical experience
43.3 Cross-species socialization
43.4 Behavioral accommodation
43.5 Goats raised with dogs
43.6 Dogs raised with humans
43.7 Captive versus wild development
43.8 Training
43.9 Learned repertoires
43.10 Behavior cannot be read directly back into innate mechanism

44. Birds

44.1 Avian pallium
44.2 Hippocampal systems
44.3 Visual cognition
44.4 Spatial memory
44.5 Social memory
44.6 Vocal learning
44.7 Sequence learning
44.8 Tool use
44.9 Different architecture
44.10 Same broad memory operations

45. Food-Caching Birds

45.1 Cache behavior
45.2 Spatial indexing
45.3 Item indexing
45.4 Temporal information
45.5 Hidden cache arrays
45.6 Cache-specific hippocampal barcodes
45.7 Retrieval reinstatement
45.8 Extensive practice
45.9 Task-specific neural organization
45.10 Experiment does not distinguish evolved dedicated capacity from practiced exploitation of general capacity

46. Practice Versus Specialized Mechanism

46.1 Repeated behavioral demand
46.2 Neural differentiation
46.3 Enlarged representation
46.4 Skilled human musicians
46.5 Navigation experts
46.6 Athletes
46.7 Food-caching birds
46.8 Training-dependent geometry
46.9 Selection versus development as competing explanations
46.10 Neural specialization alone does not decide between them

47. Mammals

47.1 Spatial relations
47.2 Social recognition
47.3 Sequence memory
47.4 Long-term association
47.5 Tool procedures
47.6 Individual histories
47.7 Environmental maps
47.8 Prediction
47.9 Flexible retrieval
47.10 Broad continuity of memory capability

48. Great Apes

48.1 Long-term recognition
48.2 Social history
48.3 Coalition relations
48.4 Tool learning
48.5 Sequence learning
48.6 Observation
48.7 Spatial memory
48.8 Manipulation of relations
48.9 Different performance ceilings
48.10 Same mechanism can yield different outcomes

Part XII — Human Memory

49. Human Difference Without Inventing a New Memory Mechanism

49.1 Shared primate memory operations
49.2 Larger relational networks
49.3 Longer developmental period
49.4 Greater social exposure
49.5 Language
49.6 Teaching
49.7 Cultural rehearsal
49.8 Symbolic indexing
49.9 Practice over decades
49.10 Different outcome does not imply different basic mechanism

50. Human Working-Memory Performance

50.1 Relational arity
50.2 Chunking
50.3 Sequence manipulation
50.4 Recursive structure
50.5 Training
50.6 Language scaffolding
50.7 External notation
50.8 Automation
50.9 Culture-dependent performance
50.10 Do not infer a separate memory architecture from task scores

51. Social Adaptability

51.1 Individual identity
51.2 Kinship
51.3 Rank
51.4 Coalition
51.5 Reciprocity
51.6 Reputation
51.7 Deception
51.8 Cooperation
51.9 Long interaction histories
51.10 Social complexity as a major consumer of relational memory

52. Hunter-Gatherer Memory Requirements

52.1 Routes
52.2 Water
52.3 Seasonal resources
52.4 Plant knowledge
52.5 Animal behavior
52.6 Hunting
52.7 Tool manufacture
52.8 Kinship
52.9 Social obligation
52.10 Large ecological demand still does not reveal an upper storage limit

53. Modern Capacity Surplus

53.1 Literacy
53.2 Mathematics
53.3 Music
53.4 Law
53.5 Science
53.6 Programming
53.7 History
53.8 Fictional worlds
53.9 Thousands of arbitrary symbolic relations
53.10 Existing memory architecture supports domains absent from ancestral ecology

Part XIII — Memory and Culture

54. Social Transmission

54.1 Observation
54.2 Imitation
54.3 Teaching
54.4 Rehearsal
54.5 Correction
54.6 Shared vocabulary
54.7 Ritual
54.8 Procedure
54.9 Intergenerational recurrence
54.10 Culture repeatedly rewrites individual memory systems

55. Technology

55.1 Tool form
55.2 Tool-use sequence
55.3 Manufacturing procedure
55.4 Learned affordances
55.5 Teaching
55.6 Practice
55.7 Modification
55.8 Cumulative technique
55.9 Social retention
55.10 Technology exploits memory without itself automatically being memory

56. External Memory

56.1 Physical artifact
56.2 Encoded relation
56.3 Decoder
56.4 Shared convention
56.5 Writing
56.6 Numerical notation
56.7 Maps
56.8 Records
56.9 Persistent semantic reconstruction
56.10 External memory requires preservation of a decoding relation

57. Physical Persistence Without Semantic Persistence

57.1 Monument
57.2 Tool
57.3 Road
57.4 Building
57.5 Abandoned settlement
57.6 Original meaning lost
57.7 Later encounter
57.8 Reinterpretation
57.9 New semantic geometry
57.10 Physical continuity \(\neq\) memory continuity

58. Megaliths

58.1 Original builders
58.2 Original use
58.3 Population disappearance
58.4 Monument survival
58.5 Newcomer encounter
58.6 Reuse
58.7 New interpretation
58.8 Cultural borrowing
58.9 Landscape constraint
58.10 Semantic regeneration rather than automatic preservation of original culture

Part XIV — Memory, Technology and Demography

59. Memory as Infrastructure for Technology

59.1 Procedure retention
59.2 Skill acquisition
59.3 Teaching
59.4 Error correction
59.5 Cumulative refinement
59.6 Distributed knowledge
59.7 Specialization
59.8 Social redundancy
59.9 Toolkit persistence
59.10 Memory permits technology to outlive individual inventors

60. Technology and Mobility

60.1 Fire
60.2 Shelter
60.3 Clothing
60.4 Hunting systems
60.5 Watercraft
60.6 Navigation
60.7 Riverine systems
60.8 Mobility corridors
60.9 Effective distance
60.10 Technology permits sparse populations to occupy enormous ranges

61. Demographic Throughput

61.1 Local population size
61.2 Connected reservoir
61.3 Mate exchange
61.4 Recolonization
61.5 Recovery after local extinction
61.6 Knowledge transmission
61.7 Toolkit reproduction
61.8 Social connectivity
61.9 Population continuity
61.10 Throughput rather than simple density

62. Erectus, Neanderthals and Denisovans

62.1 Technology sufficient for Eurasian occupation
62.2 Broad geographic range
62.3 Sparse populations
62.4 Fragmentation
62.5 Refugia
62.6 Local extinction
62.7 Re-expansion
62.8 Cultural continuity despite sparse demography
62.9 Similar broad memory capability does not imply similar population history
62.10 Demography is not a direct readout of cognition

63. Sapiens

63.1 Same broad biological memory machinery
63.2 Social amplification
63.3 Toolkits
63.4 Maritime mobility
63.5 Riverine mobility
63.6 Wider social networks
63.7 Increased demographic throughput
63.8 Faster recolonization
63.9 Cultural accumulation
63.10 Population success does not require a late cognitive mutation

Part XV — Epistemology and Experimental Design

64. What Memory Experiments Actually Measure

64.1 Recall
64.2 Recognition
64.3 Reaction time
64.4 Search behavior
64.5 Neural firing
64.6 Population coding
64.7 Molecular state
64.8 Behavioral savings
64.9 Practice effects
64.10 No single readout identifies the complete memory process

65. Practice Is a Major Confound

65.1 Experimental repetition
65.2 Expertise
65.3 Overtraining
65.4 Task familiarity
65.5 Artificial environments
65.6 Cage-specific strategy
65.7 Learned experiment structure
65.8 Neural differentiation through practice
65.9 Species comparisons under unequal training histories
65.10 “Specialized circuit” does not automatically mean evolved dedicated module

66. Comparative Task Errors

66.1 Different sensory systems
66.2 Different motor systems
66.3 Different motivation
66.4 Different ecological familiarity
66.5 Different developmental histories
66.6 Different reward structure
66.7 Different training exposure
66.8 Different social context
66.9 Same mechanism, unequal measured performance
66.10 Bowler-versus-marathon-runner problem

67. The Streetlight Problem

67.1 Synapses because they are measurable
67.2 Hippocampus because it can be imaged
67.3 Engram cells because they can be tagged
67.4 Working memory because tasks are easy to standardize
67.5 Human verbal recall because it is reportable
67.6 Cache sites because they are countable
67.7 Readout promoted into ontology
67.8 Carrier promoted into memory object
67.9 Experimental convenience promoted into mechanism
67.10 Measurement domain \(\neq\) source reality

Part XVI — GRM/Base-T Reconstruction

68. What Base-T Contributes

68.1 Type discipline
68.2 Source/readout separation
68.3 Minimal explanatory commitments
68.4 Irreducible observations
68.5 Failure signatures
68.6 Perturbation boundaries
68.7 No premature ontology
68.8 No hidden variable inflation
68.9 Reconstruction from receipts
68.10 Base-T as epistemic discipline, not literal biological storage law

69. Biological Memory Kernel

69.1 Current biological state \(B_t\)
69.2 Interaction \(I_t\)
69.3 Distributed modification \(\Delta B_t\)
69.4 Relational consequence
69.5 Subsequent reconstruction
69.6 Replay
69.7 Reorganization
69.8 Access paths
69.9 Controlled readout
69.10 Ongoing modification

A minimal model:

\[ B_{t+1}=F(B_t,I_t,\Delta B_{\le t}) \]

with current reconstruction

\[ R_t=\mathcal R(B_t,I_t,H_t), \]

where \(H_t\) is present context, not a separate archive.

70. What Must Not Be Collapsed

70.1 Developmental sequence \(\neq\) learned memory
70.2 Physical persistence \(\neq\) semantic persistence
70.3 Semantic structure \(\neq\) access
70.4 Access \(\neq\) report
70.5 Neural specialization \(\neq\) evolutionary specialization
70.6 Performance difference \(\neq\) mechanism difference
70.7 Artifact persistence \(\neq\) semantic continuity
70.8 Forgetting \(\neq\) deletion
70.9 Consolidation \(\neq\) memory formation
70.10 Narrow control \(\neq\) narrow storage

Part XVII — Open Problems

71. The Physical Persistence Problem

71.1 How does a memory survive molecular turnover?
71.2 Which structures are continuously renewed?
71.3 How does replay participate in maintenance?
71.4 How stable are synaptic configurations?
71.5 How stable are population relations?
71.6 How much does retrieval rewrite structure?
71.7 What remains invariant?
71.8 What constitutes actual loss?
71.9 Can a memory survive replacement of every original molecule?
71.10 Is persistent organization more fundamental than persistent material?

72. The Capacity Problem

72.1 No observed saturation point
72.2 Lifetime accumulation
72.3 Extreme old age
72.4 Continuing acquisition
72.5 Interference versus exhaustion
72.6 Retrieval bottlenecks
72.7 Control bottlenecks
72.8 Physical state-space limits
72.9 Effective relational capacity
72.10 Where, if anywhere, is the empirical ceiling?

73. The Access Problem

73.1 How many independent paths reach one memory?
73.2 How are paths weighted?
73.3 How are new paths created?
73.4 How are paths suppressed?
73.5 Why does context reinstate inaccessible memories?
73.6 Why does recognition outperform free recall?
73.7 How does emotional state alter access?
73.8 How does language create new access paths?
73.9 How does rehearsal reorganize access?
73.10 What distinguishes lost path from lost memory?

74. The Comparative Problem

74.1 Which mechanisms are truly universal?
74.2 Which differences reflect architecture?
74.3 Which reflect body plan?
74.4 Which reflect development?
74.5 Which reflect practice?
74.6 Which reflect ecology?
74.7 Which reflect social structure?
74.8 Which reflect experimental design?
74.9 How much latent capacity remains unused in each species?
74.10 How much apparent species difference is merely different utilization?

75. The Human Problem

75.1 Why did humans exploit memory so extensively?
75.2 Social adaptability
75.3 Long juvenile development
75.4 Teaching
75.5 Language
75.6 Symbolic indexing
75.7 External notation
75.8 Cumulative culture
75.9 Massive practice effects
75.10 Human outcome without requiring a fundamentally new memory mechanism

Part XVIII — Final Synthesis

76. Development to Memory

76.1 Developmental sequencing
76.2 Hatching
76.3 Locomotion
76.4 Exploration
76.5 Socially guided exploration
76.6 Action–consequence learning
76.7 Continuous encoding
76.8 Replay
76.9 Relational structuring
76.10 Multiple access paths

77. Memory to Expertise

77.1 Repetition
77.2 Replay
77.3 Reorganization
77.4 Chunking
77.5 Automation
77.6 Specialized geometry
77.7 Narrower control demand
77.8 Faster retrieval
77.9 Practice-specific expertise
77.10 Specialized outcome from general plastic machinery

78. Memory to Human Culture

78.1 Social relations
78.2 Teaching
78.3 Language
78.4 Tools
78.5 Technology
78.6 External notation
78.7 Demographic networks
78.8 Cultural accumulation
78.9 Reuse and reinterpretation
78.10 Biological memory embedded in larger social systems

79. Core Propositions

79.1 Memory is continuously formed during experience
79.2 Consolidation is modification and stabilization, not memory creation
79.3 Memory is distributed and relational
79.4 One memory normally has multiple retrieval paths
79.5 Replay continuously restructures memory
79.6 Long-term capacity appears vastly larger than active control capacity
79.7 No experimental program has yet demonstrated a simple lifetime storage ceiling
79.8 Forgetting cannot be equated with deletion
79.9 Narrow working-memory performance cannot be used as a proxy for long-term storage capacity
79.10 Practice can generate highly specialized neural representations
79.11 Specialized neural representation does not by itself prove a specialized evolved memory mechanism
79.12 Different species can use the same broad mechanism to produce very different behavioral outcomes
79.13 Developmental and social environment are major determinants of those outcomes
79.14 Human memory need not require a qualitatively new basic mechanism
79.15 Human culture massively expands what the shared mechanism is used to encode and retrieve
79.16 Physical artifacts preserve structure, not necessarily meaning
79.17 Semantic continuity requires an actual transmission relation
79.18 Technology expands mobility and demographic throughput without being evidence of superior basic memory machinery
79.19 Biological memory is best understood as a large relational field beneath a comparatively narrow control interface
79.20 The central reconstruction is:

\[ \boxed{ \text{inherited developmental architecture} \rightarrow \text{exploration} \rightarrow \text{continuous experience-dependent encoding} \rightarrow \text{replay and restructuring} \rightarrow \text{massive relational memory field} \rightarrow \text{many potential access paths} \rightarrow \text{narrow control/readout} \rightarrow \text{practice-, ecology-, development-, and culture-specific outcomes} } \]