Skip to content
Read • Question • ConnectAnthony’s notes & researchReturn with a new question
Anthony’s NotesA personal field guide
From the notebook

Process to Read a Book


0. Big picture: what actually makes things “stick”?

From Adler + Van Doren, plus your “Reading Hard Books” PDF, plus cognitive science:

  • You learn the most when you wrestle with a book on its own terms (Adler’s analytical reading: classify → outline → interpret → criticize).
  • You remember the most when you pull ideas out of your head, not when you re-read or re-highlight (retrieval practice). (ERIC)
  • You keep them long-term when you revisit at increasing intervals instead of cramming (spacing + spaced repetition). (LSCP)
  • You deepen understanding by connecting books to each other (syntopical reading; your “reading in levels” PDF is basically Adler’s syntopical method, pragmatized).
    So the process below is just: force yourself to do those four things in a simple way.

1. Before reading: 5–10 minutes, max

Goal: give your brain a job description before it starts consuming pages.
1.1 Triage the book (inspectional read)
Do a light Adler-style inspectional read before you ever underline:

  • Read title page, preface, and back cover.
  • Scan the table of contents and headings.
  • Skim the intro and conclusion, and flip through a couple of dense-looking pages.
    Then in Obsidian, create a Book Note and answer briefly:
  1. Why this book, now?
    • 1–2 sentences: “I want this for X project / question.”
  2. 3–5 questions I want this book to answer.
  3. Type & level (so you know how hard to push):
    • Type: practical vs theoretical vs history vs philosophy vs imaginative.
    • Level:
      • “Broad” (survey),
      • “Mid-level,”
      • “Niche / hardest thing on this topic.”
        That’s it. You’ve already used the first Adler rules (classify, define the problem) without a huge form.

2. During reading: simplify your marking + add micro-retrieval

Right now you: underline + paper notes. That’s good, but we can make it work harder.

2.1 Make underlining smarter and lighter

Pure underlining has very weak memory benefits compared to testing yourself. (APA Blog)
Keep it, but constrain it:

  • Limit yourself: e.g. no more than 2–3 key lines per chapter.
  • Use simple margin codes instead of lots of prose while reading:
    • ★ = core idea
    • ? = unclear / to revisit
    • A = argument you might want to reconstruct
    • X = something you probably disagree with
      This keeps you active without turning every page into a notebook.

2.2 End-of-session “blind recall” (tiny but huge upgrade)

At the end of each reading session:

  1. Close the book.
  2. In your Obsidian Book Note (or on paper if you really prefer):
    • Write 5–10 bullet points from memory:
      • “Today’s session – what did the author actually say?”
    • Answer, from memory, one of your 3–5 guiding questions.
      Only after that can you look back at the book and correct or add anything you missed.
      This single habit turns every reading session into retrieval practice (which is what “Make It Stick” and decades of research say is the #1 thing you can do for memory). (Mind My Learning)

3. After finishing the book: one structured “Adler pass”

Instead of a huge processing pipeline, do one focused pass that:

  • Forces understanding, and
  • Produces all the raw material you need for Obsidian, Anki, and Substack.
    Use Adler’s four questions as your entire template:
  1. What is the book about as a whole?
    • 1–2 sentence unity statement.
  2. What is being said in detail, and how?
    • 5–10 key ideas / propositions in your own words.
    • 2–4 core arguments written as “Premises → Conclusion.”
  3. Is it true?
    • 2–3 bullets: where you agree / disagree, and why (uninformed, misinformed, illogical, incomplete).
  4. What of it?
    • 3–5 practical implications or questions for your life / projects.
      That’s your “Final Processing” section in the Obsidian note. It satisfies almost all of Adler’s 15 rules in one page instead of an intimidating checklist.

      AI summary is allowed after this as an extrinsic aid—compare it to your own and only add something if you see a real gap. That’s exactly how Adler says to use commentaries and abstracts.


4. Decide the “depth lane” for each book

Right now you sometimes do Obsidian + Substack + Anki. That’s powerful but can get heavy.
Make an explicit decision, once per book, which lane it belongs in:

Lane 1 – Surface / Context books (broad surveys, light histories, pop-science)

Output:

  • Obsidian Book Note with:
    • Unity sentence
    • 5–10 key takeaways
    • 1 “What of it” paragraph
  • No Substack, no Anki (unless a definition is truly crucial).
    These are your context providers (Adler’s elementary + inspectional; your “broad” books).

Lane 2 – Core books (most good non-fiction / philosophy you care about)

Output:

  1. Obsidian Book Note with the full Adler 4-question summary.
  2. Anki:
    • Create 15–40 cards, but only for:
      • Definitions / key terms
      • Short lists (e.g., “3 causes of X,” “4 conditions of Y”)
      • Big claims you want to remember verbatim
    • Avoid cards like “What is the main idea of Chapter 3?” – too fluffy.
  3. Micro-essay (private):
    • 300–600 words, written from memory, answering:
      • “What did this book actually change in how I think or act?”
    • This can double as a draft for a future Substack if you want.
      Here, Anki + recall writing give you retrieval and spacing; the Book Note gives you structure and criticism.

Lane 3 – Pillar books (the 10–15 that really shape your worldview)

For these: Adler’s syntopical reading + your “reading in levels” PDF is worth the time.
Output:

  1. All of Lane 2 plus:
    • A Substack piece (or long Obsidian essay) that:
      • Puts this book in dialogue with 2–4 others on the same question.
      • Explicitly notes where they agree / disagree and where you land.
  2. A “hub note” in Obsidian:
    • Title: the problem, not the book (e.g., “What is justice?”).
    • Links out to each book’s note, with 1–2 sentences per author’s answer.
  3. Anki:
    • Only for concepts you’ll actively use in your thinking and writing.
      This is the syntopical level Adler calls the highest form of reading, and it’s exactly what your multi-book “hard reading” PDF is pushing you toward (broad → mid → niche on one topic).

5. Build in minimal spacing without a “daily note” system

You said you don’t want extra daily admin. So don’t. Just add two tiny habits:

5.1 1-week and 1-month “pings” per book

When you finish a Core or Pillar book:

  • In the Book Note, add a quick line like:
    • Review: +7 days, +30 days
  • When you glance at your notes list, just:
    • Open any book whose review date is roughly “now.”
    • Spend 5 minutes:
      • Close the note and from memory write:
        • “Top 5 ideas I still remember + 1 question I still have.”
      • Then skim your key-ideas section to fill gaps.
        That’s spacing + retrieval in one low-admin gesture. You don’t need perfection; even messy spacing beats none. (envision.everspringpartners.com)

5.2 Let Anki handle the rest

For whatever you do put in Anki, you already get:

  • Retrieval practice by design. (Frontiers)
  • Automatic spacing over time.
    Goal: fewer, better cards. If a point is better served by writing about it, keep it out of Anki and into Substack / essays.

6. Where AI fits best in this stack

AI summaries are useful, but they should follow your own work:

  • Good uses:
    • After your Adler summary, generate an AI summary to check for blind spots.
    • Ask for counter-arguments to the book’s main claims to sharpen your “Is it true?” step.
    • Turn your own notes into Anki-style Q&A suggestions, then you curate.
  • Less good uses:
    • Letting AI summarize instead of doing your own unity statement and key ideas. That undercuts the very effort that makes memory stick. (APA Blog)

7. Concrete “minimal viable process” for your next book

If you want a really stripped-down checklist you can just follow:

  1. Pre-read (10–20 min):
    • Skim, classify, write 3–5 guiding questions in Obsidian.
  2. Each reading session (5 min at the end):
    • Blind recall bullets + answer one guiding question from memory.
  3. At the end of the book (30–45 min):
    • Fill the 4 Adler questions in your Book Note.
    • Decide: Lane 1, 2, or 3.
    • If Lane 2/3: create 15–40 targeted Anki cards.
  4. Later:
    • 1-week and 1-month 5-minute recall review.
    • If Pillar: write your syntopical Substack/essay when ready.
      That’s it. You’re already doing 80% of this; the real gains come from:
  • Adding structured recall right after reading,
  • Being more intentional about which books get which depth, and
  • Letting spacing and retrieval (Anki + quick reviews) carry the long-term load.
    If you’d like, next step I can:
  • Turn this into a specific Obsidian “Book Note” template for you (with headings and minimal checkboxes), and
  • Show how I’d do 3–5 example Anki cards from one of your recent philosophy books so you can see the grain size.