Appendix D: Memory Collection Template
A Home for Every Story
Imagine opening a drawer.
Inside are photographs.
Old letters.
A few handwritten notes.
A birthday card.
And something you wrote down years ago because you didn’t want to forget it.
You begin looking through everything.
One photograph reminds you of a childhood friend.
The birthday card brings back a family gathering.
And the handwritten note contains part of a story you had completely forgotten.
Before long, you’re remembering things you hadn’t intended to think about that afternoon.
That’s one of the pleasures of preserving personal history.
But there’s a practical problem.
The material is scattered.
Perhaps you know what everything is.
Perhaps you don’t.
Some photographs have dates.
Others don’t.
One letter mentions a person you can’t immediately identify.
And several memories appear in different places.
Now imagine having a simple index.
Something that tells you what each item contains.
Where it came from.
And how it connects to other memories.
You could still enjoy discovering things.
But finding them again would be much easier.
That’s what we’re going to build.
Not a complicated research database.
Not a technical system that requires a software engineer.
A practical Memory Collection.
Something ordinary people can understand.
Use.
And maintain.
The templates in this appendix are designed to support the method we’ve developed throughout Project Crystal.
They help you preserve genuine recollections.
Keep original sources separate from summaries.
Record uncertainty.
Respect privacy.
And prepare selected material for the eventual book.
But before we begin, there’s one principle worth remembering.
The purpose of the collection is to make your memories easier to preserve, understand, and retrieve. It isn’t to turn remembering into an administrative chore.
So we’ll begin with the smallest useful system.
Then introduce optional features for larger projects.
You can stop at whichever level serves your needs.
Part One: The Four Essential Records
A basic Project Crystal Memory Collection contains four types of records.
1. Source Records
These identify the original conversations, recordings, photographs, documents, and other materials.
2. Memory Records
These describe individual events, experiences, or recurring recollections.
3. The Memory Index
This provides a searchable list of the memories you’ve collected.
4. The Project Control Sheet
This identifies the current project, important decisions, and next steps.
That’s enough to begin.
Additional records can be useful later.
For example:
People.
Places.
Photographs.
Permissions.
Uncertainty.
And editorial changes.
But don’t feel obliged to create every one of them immediately.
If you have five memories, you may only need a handful of pages.
If you have five hundred, the extra organisation may become worthwhile.
Start with the records you actually need. Add structure when the collection becomes difficult to manage without it.
Part Two: The Simplest Folder Structure
Let’s create a starting workspace.
You can use ordinary folders on your computer.
Or a set of documents in a storage system you understand.
A simple arrangement might look like this:
My Memory Project
- 01 Sources
- 02 Memory Records
- 03 Memory Index
- 04 Book Blueprint
- 05 Manuscript Drafts
- 06 Permissions and Private Notes
The first three are enough for memory collection.
The others become useful as the project progresses.
You can keep original recordings, conversations, and photographs in the Sources folder, perhaps with subfolders if needed.
Memory Records contain the organised recollections.
The Memory Index helps you find them.
The Book Blueprint will eventually describe the intended book.
And Manuscript Drafts will contain the writing derived from selected sources.
Permissions and Private Notes should be stored with access protections appropriate to their contents.
One Important Rule
Don’t use the manuscript as your only source archive.
A finished chapter is not the same thing as the original memory.
The chapter may contain editorial choices.
Transitions.
Descriptions.
And selected details.
The original source should remain separately available.
That allows you to return to the actual recollection when a question arises.
Part Three: Your Source Register
Let’s begin with the originals.
A source is something from which information was collected.
For example:
A recorded interview.
A text conversation.
A handwritten note.
A photograph.
An old letter.
Or a personal recollection recorded directly.
Each source receives an identifier.
You might use:
R001
R002
R003
And so on.
The letter isn’t important.
Consistency is.
Once you’ve assigned an identifier, keep it stable.
That allows other records to refer to the source without depending entirely on its filename.
Template D-1: Source Register
| Field | Information |
|---|---|
| Source ID | R___ |
| Working title | |
| Source type | |
| Date created or recorded | |
| Original contributor | |
| Original filename or location | |
| Relevant people | |
| Review status | |
| Permission and access status | |
| Related Memory IDs | |
| Notes |
You can create one entry per source.
If the date is unknown, write Unknown.
Don’t invent a date to make the record complete.
If the source came from someone else, identify the contributor appropriately.
If permission details aren’t established, record that uncertainty.
Example: Fictional Source Record
Source ID: R002
Title: Robert’s Childhood Projects Conversation
Type: Recorded conversation
Contributor: Robert
Recording Date: Not specified in this example
Original File: R002-original.m4a
Transcript: R002-transcript-reviewed.txt
Related Memories: MEM-001, MEM-003
Review Status: Transcript reviewed for major names and relevant details
Permission Status: Private working use only
Notes: Several childhood projects discussed
This is an illustrative example.
In a real project, the filenames and review status must describe the files and work that actually exist.
Never mark a transcript reviewed simply because an AI generated it.
Part Four: What Counts as a Memory Record?
Now let’s distinguish the source from the memory.
A forty-minute conversation may contain five different stories.
That’s one source.
But potentially five Memory Records.
Likewise, one memory may appear in several conversations.
Perhaps you remembered something in your first session.
Then recalled another detail a week later.
That could remain one Memory Record linked to both sources.
A Memory Record represents an identifiable recollection or subject.
It doesn’t need to be a complete story.
Some records may contain only a few details.
Others may be extensive.
That’s fine.
The purpose is to keep them understandable.
Template D-2: The Complete Memory Record
Copy this template for each memory.
MEMORY RECORD
Memory ID:
MEM-___
Working Title:
Short Description:
A brief, plain-language summary of the memory.
Primary Source ID:
Additional Source IDs:
Source Type:
Personal recollection, contributed account, recording, document, photograph, or other appropriate description.
Storyteller or Contributor:
People Involved:
Place:
Approximate Time:
What the Source Actually Says:
Record the supported account without inventing missing details.
Distinctive Details:
Specific objects, expressions, actions, or other details explicitly present in the source.
Original Expressions Worth Preserving:
Exact wording, where available and appropriately attributed.
Later Reflections:
What the storyteller subsequently says or believes about the experience.
Keep this separate from the remembered event.
Uncertainties:
Dates, names, places, sequences, or other details requiring clarification.
Possible Related Memories:
Supporting Photographs or Documents:
Privacy Considerations:
Permission Status:
Manuscript Status:
Not selected, under consideration, selected, drafted, or approved for the relevant stage.
Last Reviewed:
Additional Notes:
That is the complete version.
You don’t need to fill in every field for every memory.
Perhaps the place is unknown.
Or there are no photographs.
Leave those fields empty or mark them as not established.
A good Memory Record contains what you actually know, not a complete set of invented answers.
Part Five: The Five-Field Quick Memory Card
Perhaps the complete template feels like too much for your first few memories.
That’s understandable.
You can use a shorter version.
Template D-3: Quick Memory Card
ID: MEM-___
Title:
What I Remember:
Source:
What I’m Not Sure About:
That’s it.
Five fields.
You can add details later.
The quick card is especially useful when a memory appears unexpectedly.
Perhaps you’re walking.
Working on something.
Or speaking with a friend.
A story comes to mind.
You don’t necessarily have time to create a full record.
Capture the essentials.
Then return later.
Fictional Example
ID: MEM-001
Title: The Bicycle Repair
What I Remember: Daniel and I tried to repair an old bicycle. We removed the rear wheel and struggled to put it back. It worked briefly, then something came loose.
Source: R002
What I’m Not Sure About: Our exact ages, where we worked on it, and which part failed.
That’s already useful.
You have an identifiable story.
A source reference.
And clear uncertainty.
You can begin with five cards like this.
Then expand them when the material or intended writing requires it.
Part Six: A Fully Worked Memory Record
Let’s develop our fictional bicycle story into a complete record.
This example demonstrates how to preserve an event without forcing the details into a dramatic narrative.
MEM-001: The Bicycle Repair
Short Description
Robert remembers attempting to repair an old bicycle with his childhood friend Daniel. They removed the rear wheel, struggled to reinstall it, and eventually got the bicycle working briefly before another problem occurred.
Primary Source
R002: Childhood Projects Conversation.
Additional Source
R005: Later conversation about funny experiences, if the two recordings are confirmed to describe the same incident.
Storyteller
Robert.
People Involved
Robert and Daniel.
Place
Not established.
Approximate Time
Childhood. Robert estimated that he may have been twelve or thirteen.
Recollection
Robert recalls removing the rear wheel of an old bicycle while trying to repair it with Daniel.
He remembers spending considerable time attempting to put the bicycle back together.
He recalls that it worked for a short period before another problem occurred.
Distinctive Expression
“We spent more time fixing the thing than riding it.”
Uncertainties
The exact year.
The original mechanical problem.
The component that later failed.
The location where the repair took place.
Later Reflection
No separate reflection recorded.
Related Memories
None confirmed in this example.
Photographs
None confirmed.
Privacy Status
Private working record.
Manuscript Status
Candidate for Chapter One.
Current Review Status
Awaiting comparison with the actual source before inclusion in a real manuscript.
Notice what we haven’t done.
We haven’t invented the weather.
A dramatic conversation.
Or a precise year.
We haven’t announced that Robert’s future career was determined that afternoon.
We’ve preserved the event.
And the uncertainty.
That’s enough to support responsible drafting.
Part Seven: Your Master Memory Index
Now suppose you’ve collected several memories.
You don’t want to open every individual record simply to see what’s available.
That’s why we create the Memory Index.
Think of it as the catalogue for your collection.
It should be easy to scan.
And easy to update.
Template D-4: Memory Index
| Memory ID | Working Title | Source ID(s) | Main Subject | Status |
|---|---|---|---|---|
| MEM-001 | ||||
| MEM-002 | ||||
| MEM-003 | ||||
| MEM-004 | ||||
| MEM-005 |
You can add columns later.
Perhaps:
Approximate year.
People.
Place.
Privacy restriction.
Proposed chapter.
Or last review date.
But don’t add twenty columns if five will do.
The index should make it easy to find a memory.
Not require an afternoon of data entry.
Example: Robert’s Collection
| Memory ID | Working Title | Source | Main Subject | Status |
|---|---|---|---|---|
| MEM-001 | The Bicycle Repair | R002 | Childhood project | Captured |
| MEM-002 | Grandpa’s Red Toolbox | R003 | Family, workshop | Captured |
| MEM-003 | The Wrong Appliance | R005 | Humour, repair | Needs review |
| MEM-004 | The Family Holiday | R002, R006 | Family travel | Date uncertain |
| MEM-005 | The Old Workshop Radio | R004 | Place, object | Captured |
These are fictional examples.
A real index should contain only the records that actually exist.
And every listed Source ID should point to a real source entry.
The index is a map. Its value depends on the records and sources behind it.
Part Eight: Your People Register
Now let’s add an optional record.
People.
A personal book may involve dozens of individuals.
Family members.
Friends.
Coworkers.
Neighbours.
Teachers.
And others.
Some may appear in several memories.
A People Register helps identify those relationships without repeatedly writing the same background information.
It can also help with spelling and privacy.
Template D-5: People Register
Person ID:
PERS-___
Name or Working Identifier:
Relationship to Storyteller:
Relevant Time Period:
Memory IDs:
Sources Containing Information About This Person:
Known Details Supported by Sources:
Uncertain Details:
Preferred Attribution, If Established:
Privacy or Publication Restrictions:
Notes:
You can use a nickname or pseudonym when appropriate.
But remember that replacing someone’s name doesn’t necessarily guarantee anonymity.
The surrounding details may still identify them.
If privacy matters, review the entire passage.
Fictional Example
Person ID: PERS-002
Name: Daniel
Relationship: Robert’s childhood friend
Relevant Memories: MEM-001
Known Details: Participated in the bicycle repair
Uncertain Details: Surname, exact age, later life circumstances
Publication Status: Requires review before including additional identifying information
Notice how limited the record is.
We aren’t inventing Daniel’s personality.
His family history.
Or his motivations.
We only record what the available sources support.
That’s enough.
Part Nine: Your Places Register
Places can be important connections between memories.
Perhaps several stories happened in the same workshop.
Another group involves a childhood home.
And several describe a local park.
A Places Register helps identify those recurring settings.
Template D-6: Places Register
Place ID:
PLACE-___
Working Name:
General Location:
Relevant Time Period:
Related Memory IDs:
Source References:
Description Supported by Sources:
Uncertain Details:
Privacy Considerations:
Notes:
You don’t need to record precise addresses unless they’re relevant and appropriate.
A general description may be sufficient.
For example:
Grandpa’s Workshop
Or:
The Family House in the 1980s
For private homes, especially those associated with living people, consider whether precise location information should appear in the working records or the finished book.
The purpose is to help organise the memories.
Not expose unnecessary personal information.
Part Ten: Your Photograph Register
Now let’s introduce photographs.
Perhaps you have several old family pictures.
Some are labelled.
Others aren’t.
A photograph may support a memory.
But it may also raise questions.
Who is shown?
When was it taken?
Where?
And who owns the original?
We need a separate record for the image itself.
Template D-7: Photograph Register
Photograph ID:
PHOTO-___
Working Title:
Original Filename or Physical Location:
Image Description:
Known Photographer or Source:
Date:
People Identified:
Place Identified:
Related Memory IDs:
Identification Confidence:
Confirmed, probable, possible, or unknown, with the reason recorded.
Usage Permission:
Publication Status:
Edited Versions:
Notes:
Fictional Example
Photograph ID: PHOTO-001
Title: Bicycle Beside a House
Source: Family photograph album
Date: Unknown
People: None confidently identified in the current record
Possible Related Memory: MEM-001
Relationship Status: Unconfirmed
Permission Status: Publication rights not established
Notes: Robert believes the bicycle may be connected to the childhood repair story.
That’s a useful record.
It preserves the possible relationship.
Without presenting it as fact.
Perhaps later another source helps identify the bicycle.
Then you can update the record.
But retain where that new information came from.
Part Eleven: Your Uncertainty Register
Now let’s create something particularly valuable.
A place for unresolved questions.
You don’t need to solve every uncertainty before writing.
But you should know which ones matter.
Perhaps a date is uncertain.
A photograph hasn’t been identified.
Or two family members remember an event differently.
Those questions belong somewhere visible.
Template D-8: Uncertainty Register
| Question ID | Related Record | Uncertainty | Importance | Current Treatment |
|---|---|---|---|---|
| Q001 | ||||
| Q002 | ||||
| Q003 |
Let’s examine the fields.
Question ID
A stable identifier.
Related Record
The Memory Record or source involved.
Uncertainty
The exact question.
Importance
Whether the uncertainty materially affects the intended story.
Current Treatment
How the manuscript should handle the issue until it is resolved.
Example
| Question ID | Related Record | Uncertainty | Importance | Current Treatment |
|---|---|---|---|---|
| Q001 | MEM-001 | Exact year of bicycle repair | Low | Keep age approximate |
| Q002 | MEM-004 | Which house the family lived in during the holiday | Moderate | Avoid specifying house until checked |
| Q003 | PHOTO-001 | Whether photograph shows the same bicycle | Moderate | Do not use as confirmed illustration |
This is a useful distinction.
Some uncertainties need resolution.
Others don’t.
If the exact year isn’t important to the story, you can leave it approximate.
But if a chapter makes a chronological claim dependent on the year, the issue becomes more significant.
The purpose of the uncertainty register is to help you make informed decisions, not to eliminate every unknown detail.
Part Twelve: Conflicting Accounts
Now let’s examine a common family-history problem.
Two people remember the same event differently.
Perhaps Robert remembers the bicycle repair happening in Daniel’s garage.
Daniel remembers working on it in Robert’s backyard.
What should the Memory Record say?
Not:
“The bicycle was repaired in Daniel’s garage.”
And not:
“The bicycle was repaired in Robert’s backyard.”
Unless further evidence establishes one account.
Instead, preserve the difference.
Template D-9: Conflicting Accounts Record
Memory ID:
Subject of Disagreement:
Account A:
Source A:
Account B:
Source B:
Additional Evidence:
Current Status:
Resolved, unresolved, or partially resolved.
Approved Manuscript Treatment:
Notes:
Fictional Example
Memory ID: MEM-001
Subject: Location of bicycle repair
Robert’s Account: Daniel’s garage
Source: R002
Daniel’s Account: Robert’s backyard
Source: R015
Additional Evidence: None currently available
Status: Unresolved
Manuscript Treatment: Do not state a precise location. If relevant to the narrative, acknowledge that the participants remember the location differently.
That’s a responsible approach.
And it may even improve the story.
Perhaps the differing recollections are themselves interesting.
But don’t manufacture agreement to make the manuscript smoother.
A good archive can preserve disagreement without forcing an artificial conclusion.
Part Thirteen: Your Privacy and Permissions Register
Now let’s address information that may affect whether a memory can be shared.
Perhaps someone contributed a private interview.
Or a photograph depicts other people.
Maybe a story includes sensitive personal details.
You need to know what uses have been authorised.
A simple register can help.
But permission isn’t always a single yes-or-no field.
Someone may agree to recording.
But not online publication.
Or allow a story to be included in a private family edition while declining broader distribution.
Those distinctions matter.
Template D-10: Privacy and Permissions Record
Permission Record ID:
PERM-___
Contributor or Relevant Person:
Related Source IDs:
Related Memory IDs:
Permission to Record:
Permission for AI Processing:
Permission for Private Drafting:
Permission for Family Distribution:
Permission for Public Publication:
Permission for Audio Publication:
Attribution Preference:
Restrictions or Conditions:
Date and Method of Permission:
Documentation Location:
Outstanding Questions:
Don’t invent permission records.
And don’t mark a use authorised merely because someone participated in a conversation.
The actual consent and applicable rights matter.
If you aren’t sure, mark the permission as unresolved.
That may mean keeping the material out of a release-ready edition until the question has been addressed.
Recording and publication laws vary.
For legally significant matters, appropriate legal advice may be necessary.
The template is an organisational tool.
Not a substitute for obtaining valid permission.
Part Fourteen: Your Memory Status System
Now let’s make the collection easier to manage.
Each Memory Record can have a status.
But the status should mean something specific.
Here’s a practical beginner system.
Captured
The memory has been identified and linked to its source.
No claim of full review.
Needs Review
The record contains important questions or potential problems.
Source-Reviewed
The memory summary has been checked against the available source material.
This does not mean every historical detail has been independently verified.
Restricted
The record contains privacy or permission considerations requiring special handling.
Selected for Manuscript
The author has chosen to consider the memory for a chapter.
Drafted
A manuscript passage has been developed from the record.
Approved for the Current Edition
The relevant material has passed the required author and editorial review for that edition.
Notice that these statuses describe different things.
A memory may be source-reviewed but restricted.
Or selected for the manuscript but not yet drafted.
For larger projects, consider tracking source review, permission status, and manuscript status in separate fields rather than forcing everything into one label.
That’s more precise.
But for a small collection, a simple primary status plus notes may be enough.
A status should tell you what work has actually been completed, not what you hope is true.
Part Fifteen: Your Memory Update Log
Now let’s imagine something happens.
You record a memory.
Then two months later, you recall another detail.
Should you edit the record?
Certainly, when appropriate.
But preserve where the new information came from.
A simple update log helps.
Template D-11: Memory Update Log
| Date | Memory ID | Source of Update | What Changed | Reviewed By |
|---|---|---|---|---|
Example
Date: Date of actual update
Memory ID: MEM-001
New Source: R016
Change: Added Robert’s later recollection that Daniel was the friend involved
Reason: Additional account provided
Review Status: Awaiting source comparison
Notice that the update doesn’t silently overwrite the earlier version.
It identifies what changed.
And why.
That can be particularly useful when several people contribute to a shared family history.
Part Sixteen: What Should Never Be Silently Changed
Let’s establish a few boundaries for editing the collection.
Don’t silently change:
Names.
Dates.
Direct quotations.
Locations.
Source attributions.
Permission statuses.
Or material uncertainty.
Those details may affect the meaning or reliability of the story.
If you correct one, preserve the reason.
Perhaps a date was wrong in the transcript.
The original audio establishes the correction.
That’s a legitimate update.
But the record should identify the source.
Perhaps a person changes their preference about attribution.
Record the new instruction and its scope.
And if a memory is revised because of a later recollection, note that the information came later.
The Memory Collection should become clearer over time without disguising how its information developed.
That’s the objective.
Part Seventeen: When AI Helps Build the Collection
Now let’s introduce AI into the process.
Suppose you’ve collected seven conversations.
You have the transcripts.
And you’d like help identifying individual memories.
AI can assist.
It can examine the available text.
Suggest memory boundaries.
Create summaries.
And identify recurring names or subjects.
But it must work from the source material.
Not invented context.
And it should not claim to have reviewed files it cannot access.
That’s especially important when the collection spans several conversations or storage locations.
Template D-12: Memory Collection Organiser Prompt
PROJECT CRYSTAL: MEMORY COLLECTION ORGANISER
I have a set of genuine memory sources.
I want to organise them into a dependable Memory Collection.
Please follow these instructions.
1. Confirm Access
Identify which source materials you can actually examine.
Do not invent access to missing files, recordings, or conversations.
2. Identify Distinct Memories
Find separate events, experiences, and recurring recollections.
Don’t assume every conversation contains exactly one memory.
3. Preserve Source References
Every Memory Record must point to the actual source material.
Do not invent Source IDs or timestamps.
4. Create Plain Summaries
Describe what the source supports.
Do not embellish the account.
5. Preserve Uncertainty
Mark unknown dates, names, places, and other unresolved information.
6. Distinguish Recollection From Interpretation
Don’t convert proposed meanings into established facts.
7. Identify Possible Duplicates
Flag memories that may describe the same event.
Do not merge them without sufficient evidence or author approval.
8. Record Conflicting Accounts
Preserve source-specific differences.
Do not create artificial agreement.
9. Identify Related Records
Link memories sharing people, places, objects, or subjects.
Don’t invent causal relationships.
10. Respect Privacy
Do not assume material is authorised for publication.
Flag records requiring permission or restricted handling.
11. Produce a Memory Index
List every proposed Memory Record with its ID, title, source reference, and status.
12. Present the Results for Review
Do not declare the collection complete or approved merely because the records have been generated.
Wait for me to correct the information.
The purpose is accurate organisation, not creative rewriting.
That’s a practical use of AI.
And it gives you a way to review the results before accepting them.
Part Eighteen: The Five-Record Starter Exercise
Let’s return to the simplest possible version.
Perhaps you’ve read the entire appendix.
And you’re thinking:
“This looks useful, but I don’t want to spend the rest of my life filling out forms.”
Fair enough.
So don’t.
Start with five memories.
Five quick cards.
One simple index.
That’s enough to learn the method.
Your First Collection
MEM-001
One genuine memory.
MEM-002
Another.
MEM-003
A third.
MEM-004
A fourth.
MEM-005
A fifth.
For each, record:
The title.
What you remember.
The source.
And what remains uncertain.
Then create an index showing the five records.
That’s your first Memory Collection.
Now test it.
Can you find the original source for MEM-003?
Can you explain what MEM-004 concerns?
Do you know whether MEM-002 contains uncertain details?
If yes, the system is working.
You can expand it when needed.
Five accurate, retrievable records are a better beginning than fifty elaborate records containing unsupported details.
That’s worth remembering.
Part Nineteen: When the Collection Grows
Now imagine the project continues.
You have twenty Memory Records.
Then fifty.
Perhaps eventually a hundred.
What changes?
The basic principles remain the same.
But you may begin benefiting from additional tools.
A spreadsheet.
Searchable text.
Tags.
Or a more structured digital archive.
Those can make retrieval easier.
But they don’t change the source requirements.
A search result still needs to point to a real record.
An AI summary still needs review.
And a stored interpretation still needs to be distinguishable from the original account.
A Spreadsheet Can Be Enough
For a moderate collection, a spreadsheet may work beautifully.
One row per memory.
Columns for:
ID.
Title.
Source.
People.
Place.
Approximate time.
Status.
And notes.
You can sort.
Filter.
And search.
No database administration required.
The full Memory Records can remain in separate documents.
Or, if the material is brief, the spreadsheet can contain short summaries directly.
Choose what suits the collection.
When More Advanced Tools Become Useful
If you eventually accumulate thousands of records, a more advanced search system may become worthwhile.
But don’t adopt one merely because it sounds sophisticated.
And don’t confuse technological complexity with historical reliability.
The underlying records still determine what the archive can support.
A searchable archive becomes more convenient. It doesn’t automatically become more accurate.
That’s the distinction.
Part Twenty: Preparing Memory Records for the Book Blueprint
Now let’s connect the collection to the next stage.
You’ve gathered memories.
Organised them.
And reviewed the sources.
Perhaps you’ve also created a first Book Blueprint using Chapter 38.
Now you want to assign selected memories to chapters.
A simple chapter assignment field can help.
For example:
MEM-001: The Bicycle Repair
Possible Chapter: Chapter One.
MEM-002: Grandpa’s Red Toolbox
Possible Chapter: Chapter Two.
MEM-003: The Wrong Appliance
Possible Chapter: Chapter Four.
But remember, those assignments are editorial choices.
They don’t change the source.
And a memory can be considered for more than one chapter during planning.
The final manuscript should avoid unnecessary duplication.
But the planning stage can remain flexible.
Template D-13: Chapter Source Map
| Proposed Chapter | Chapter Purpose | Memory IDs | Readiness |
|---|---|---|---|
| 1 | |||
| 2 | |||
| 3 | |||
| 4 |
This map connects the Book Blueprint to the Memory Library.
Now you can see which chapters have enough source material.
And which need more work.
That’s the bridge between organisation and writing.
Part Twenty-One: Your Chapter Source Packet
Once a chapter is ready, you can prepare a source packet.
This gathers the relevant material for drafting.
But it doesn’t replace the source archive.
It’s a selected working set.
Template D-14: Chapter Source Packet
Book Title:
Blueprint Version:
Chapter Number:
Chapter Title:
Chapter Promise:
Intended Narrator:
Writing Voice:
Primary Memory IDs:
Supporting Source IDs:
Approved Factual Details:
Approved Direct Quotations:
Uncertainties to Preserve:
Important Context for Readers:
Privacy or Permission Restrictions:
Material That Must Not Be Invented:
Possible Chapter Structure:
Desired Ending or Transition:
Current Drafting Status:
Now the writing task becomes clear.
The AI knows which material it may use.
What the chapter should accomplish.
And which uncertainties remain unresolved.
That’s much more dependable than asking:
“Write a wonderful story about my childhood.”
The source packet provides boundaries.
And the author still reviews the result.
The Chapter Source Packet is where organised memory material becomes a specific writing assignment.
Part Twenty-Two: A Complete Miniature Collection
Let’s bring the templates together.
Imagine Robert has collected five genuine memories.
We’ll use the fictional examples from earlier chapters.
His project contains:
R001: Conversation about why he wants to preserve family stories.
R002: Childhood project conversation.
R003: Conversation about Grandpa.
R004: Places and objects conversation.
R005: Funny memories conversation.
Now the Memory Index contains:
| ID | Memory | Source |
|---|---|---|
| MEM-001 | The Bicycle Repair | R002 |
| MEM-002 | Grandpa’s Red Toolbox | R003 |
| MEM-003 | The Wrong Appliance | R005 |
| MEM-004 | The Old Workshop Radio | R004 |
| MEM-005 | A Childhood Friendship | R002 |
The People Register contains:
PERS-001: Grandpa.
PERS-002: Daniel.
The Places Register contains:
PLACE-001: Grandpa’s Workshop.
The Uncertainty Register identifies:
The year of the bicycle repair.
The exact location of one project.
And whether an old photograph shows the bicycle involved.
Now Robert has a complete miniature archive.
Not huge.
Not complicated.
But organised.
If he decides to write a chapter about the workshop, he can identify MEM-002 and MEM-004.
If he wants a funny chapter about repairs, he may consider MEM-001 and MEM-003.
And if another story emerges, he can add a new record.
Nothing requires rebuilding the entire collection.
That’s what makes the system useful.
Part Twenty-Three: Your Monthly Collection Review
For a growing project, a short periodic review can help.
It doesn’t need to be monthly.
Choose a rhythm appropriate to your work.
Perhaps you review the collection after every five new memories.
Or before beginning a new chapter.
The purpose is to ensure the records remain understandable.
Template D-15: Collection Review
Review Date:
Number of Source Records:
Number of Memory Records:
New Records Since Last Review:
Possible Duplicates:
Unresolved Source Questions:
Records Requiring Privacy Review:
Records Selected for Manuscript:
Missing or Unavailable Source Files:
Backup Status:
Corrections Made:
Next Practical Task:
You don’t need to fill out this report every day.
It’s a periodic health check for the archive.
And if something isn’t working, simplify it.
Perhaps you’re collecting memories faster than you’re indexing them.
That’s fine.
Create a small queue of unprocessed sources.
Then organise them when you have time.
The system should accommodate ordinary life.
Not demand constant maintenance.
Part Twenty-Four: The Difference Between Captured and Verified
Here’s a distinction worth repeating.
Suppose Robert records a childhood story.
He clearly remembers the event.
The recording is preserved.
And the Memory Record accurately summarises what he said.
What can we conclude?
We can conclude that the memory has been captured.
And perhaps that the summary faithfully represents his account.
But that doesn’t establish every historical detail.
The year may still be uncertain.
Another participant may remember the event differently.
Or a photograph may provide additional context.
That’s why the collection should distinguish:
Source Captured
The account exists.
Source Reviewed
The record has been compared with the source.
Fact Independently Supported
Specific details have been checked against appropriate additional evidence.
These aren’t the same thing.
And not every personal story requires independent verification of every detail.
A family memoir may appropriately contain recollections presented as recollections.
But when the manuscript makes precise historical claims, greater verification may matter.
Faithfully preserving what someone remembers is not identical to proving that every remembered detail happened exactly as described.
That’s a central principle of source-based storytelling.
Part Twenty-Five: What the Memory Collection Should Never Become
Let’s address one final risk.
The system itself can become the project.
You start organising five memories.
Then develop elaborate categories.
A complicated numbering convention.
Dozens of metadata fields.
And perhaps an entire custom application.
Soon you’re spending more time managing the collection than remembering.
For some people, building the tools is an enjoyable separate project.
That’s fine.
But if your purpose is to create a family book, remember what the system is supposed to accomplish.
It should help you find stories.
Preserve their sources.
And develop the manuscript.
That’s all.
If a field isn’t useful, you don’t need it.
If a register remains empty, perhaps you can omit it for now.
And if a simple spreadsheet works, you don’t need a more complicated database.
The best Memory Collection is the simplest one that reliably preserves the information your project actually needs.
That’s the standard.
And it’s an important lesson from the broader Project Crystal method.
Part Twenty-Six: Your Memory Collection Audit Prompt
Now let’s create one final tool.
Suppose you’ve been collecting memories for several months.
You’d like to check whether the archive remains organised.
This prompt is designed for that review.
PROJECT CRYSTAL: MEMORY COLLECTION AUDITOR
I have an existing Memory Collection.
I want to examine its organisation, source integrity, and readiness for manuscript development.
Please review only the material I provide or material you can actually access.
Follow these instructions.
1. Inventory the Sources
Identify the available source records.
Flag missing references.
2. Inventory the Memories
Identify every Memory Record.
Check for duplicate or inconsistent IDs.
3. Verify Source Links
Confirm that the Memory Records refer to actual sources.
Do not invent missing links.
4. Review Summaries
Flag unsupported details, invented quotations, or interpretations presented as facts.
5. Review Uncertainty
Identify where unresolved information may have been converted into confident statements.
6. Review Possible Duplicates
Flag memories that may describe the same event.
Do not merge them automatically.
7. Review Conflicting Accounts
Preserve source-specific differences.
8. Review People and Places
Check for inconsistent names, unsupported identities, and unnecessary private information.
9. Review Photographs
Identify images without confirmed source, identification, or permission status.
10. Review Manuscript Assignments
Identify selected memories and the chapters they may support.
11. Review Permissions
Flag material whose intended use exceeds documented authorisation.
12. Review File Integrity
Identify missing originals, unavailable recordings, or unresolved backup concerns based on the available records.
13. Produce a Findings Report
Separate confirmed problems from possible concerns.
14. Recommend Specific Corrections
Do not apply broad changes automatically.
15. Preserve the Originals
Do not overwrite source material during the audit.
The goal is a dependable Memory Collection that can support accurate writing and future review.
That’s the audit.
Use it when the collection becomes large enough that a systematic review would help.
You don’t need to run it every time you add a memory.
Your Memory Collection Workshop
Let’s finish with a simple practical assignment.
You have the templates.
Now we’re going to build the first collection.
Step One: Gather Five Memories
Use your first seven conversations or another available source collection.
Step Two: Preserve the Originals
Keep the conversations, recordings, or notes in a dependable location.
Step Three: Create the Source Register
Assign stable Source IDs.
Step Four: Create Five Quick Memory Cards
Use Template D-3.
Step Five: Build the Memory Index
List the five records.
Step Six: Review the Source Links
Make sure each Memory Record points to the actual source.
Step Seven: Record Important Uncertainties
Don’t guess missing details.
Step Eight: Identify Related Memories
Link shared people, places, or subjects where supported.
Step Nine: Record Privacy Considerations
Don’t assume private material is approved for publication.
Step Ten: Save a Backup
Keep a recoverable copy appropriate to the material’s sensitivity.
Now stop.
Look at what you’ve created.
Five memories.
Five identifiable records.
A source register.
And a simple index.
You can find the stories.
You can return to the original accounts.
And you know what remains uncertain.
That’s your first working Memory Collection.
You don’t need anything more complicated until the project requires it.
One Final Thought: The Box With the Labels
Imagine returning to the drawer we opened at the beginning.
The photographs.
Letters.
Notes.
And forgotten pieces of family history.
At first, everything was mixed together.
You recognised some things.
Others required investigation.
And a few immediately brought back memories.
Now imagine that the material has been organised.
Not transformed.
Not rewritten.
Just identified.
The photographs have source records.
The stories have Memory IDs.
The uncertain dates are marked.
And the original material remains preserved.
You can open the collection.
Find a story.
Follow it to its source.
And understand what you actually know.
That’s an enormous improvement.
Not because the system is technologically sophisticated.
But because the information is no longer dependent on someone remembering where everything was stored.
And as the collection grows, the same basic method continues working.
A new conversation arrives.
It receives a source identifier.
A memory is discovered.
It receives a record.
Perhaps it connects to another story.
The relationship is noted.
Then the material returns to the archive.
Available.
Organised.
And ready when needed.
Perhaps years later, someone decides to write another book.
They open the collection.
Find a memory that didn’t fit the first manuscript.
And discover that it deserves its own chapter.
The original source is still there.
The uncertainty remains visible.
And the story can be developed without beginning entirely from memory again.
That’s the value of preserving the foundation.
A finished book is one possible result.
But the Memory Collection can support more than one book.
It can preserve family history.
Help people compare recollections.
And make original accounts available for future generations, within the appropriate privacy and permission boundaries.
The collection doesn’t guarantee that every remembered detail is historically accurate.
It doesn’t replace the people who told the stories.
And it doesn’t magically restore information that was never recorded.
But it gives us something practical.
A dependable way to preserve what is available.
And distinguish what remains unknown.
That’s enough.
Because the goal was never to create a perfect database of someone’s life.
It was to help people preserve genuine memories in a form they could understand and use.
One story.
One source.
One record.
Then another.
Until the collection becomes something substantial.
Not because every memory has been polished into a chapter.
But because the original material has been given a place to live.
And that’s the foundation of every responsible book that follows.
The source preserves the account.
The record makes the memory understandable.
The index makes it findable.
And the author decides what to create from it.
That’s your Project Crystal Memory Collection.
Simple enough to begin with five stories.
Dependable enough to support a much larger project.
And flexible enough to grow without losing sight of why the memories were collected in the first place.
Quick Reference: Your Memory Collection Templates
| Template | Purpose |
|---|---|
| D-1: Source Register | Identify original materials |
| D-2: Complete Memory Record | Preserve a detailed recollection |
| D-3: Quick Memory Card | Capture a memory simply |
| D-4: Memory Index | Find all recorded memories |
| D-5: People Register | Track recurring people |
| D-6: Places Register | Track recurring locations |
| D-7: Photograph Register | Organise and identify images |
| D-8: Uncertainty Register | Preserve unresolved questions |
| D-9: Conflicting Accounts | Keep different recollections distinguishable |
| D-10: Privacy and Permissions | Track authorised uses |
| D-11: Memory Update Log | Record meaningful changes |
| D-12: Collection Organiser Prompt | Organise source material with AI |
| D-13: Chapter Source Map | Connect memories to planned chapters |
| D-14: Chapter Source Packet | Prepare source-grounded drafting |
| D-15: Collection Review | Check the archive periodically |
The Smallest Useful Starting System
You only need:
- A Source Register.
- Five Quick Memory Cards.
- A Memory Index.
- A dependable backup.
Everything else is available when you need it.
And remember:
The Memory Collection exists to protect and organise genuine experiences, not to manufacture a more impressive version of someone’s life.
End of Appendix D.