Reading paths

The book is one spine; you have many entrances. Pick the path closest to the role you read in. Each path is a recommended minimum — once you finish, the rest of the book is open to you.

Sidebar legend (Overture explains each): ◆ Going Deeper · ▼ Why It Matters · ☉ In the Wild · ✻ Try This · § For the Record.

Path A — The Builder

For software engineers, students, and architects. This is the longest path. You read everything; you do every lab.

Order Read Sidebars Lab
1 Overture all —
2 Ch 1 Crisis all —
3 Ch 2 Falsifiability all —
4 Ch 3 FRE primer all —
5 Ch 4 Canon all —
6 Ch 5 Hashing all Lab 5
7 Ch 6 Signatures all Lab 6
8 Ch 7 Canonicalization all Lab 7 ★
9 Ch 8 Schemas all Lab 8
10 Chs 9–12 all Labs 9–12
11 Chs 13–20 all Labs as drafted
12 Chs 21–26 all Labs 23, 25, 26
13 Chs 27–30 all Capstone ★★
14 Appendices A–F reference as needed —

★ Lab 7 is the textbook’s flagship exercise — a six-problem progressive set that walks you from naïve canonicalization to a working parser-mismatch attack against a “secure” attestation system. It exists because the field does not yet have a Cryptopals-equivalent for RFC 8785; you will help fill that gap.

★★ The capstone — Chapters 27–30 — is a 10-week project. It is the assessment everything before it has been preparing you for.

Path B — The Receiver

For lawyers, judges, regulators, auditors, opposing counsel. You will not write code. You will read attestations as exhibits and ask the right questions of the engineer who built the system you are reviewing.

Order Read Sidebars Skip
1 Overture all —
2 Ch 1 Crisis ▼ ☉ § —
3 Ch 2 Falsifiability ▼ ☉ § —
4 Ch 3 FRE primer all —
5 Ch 4 Canon ▼ ☉ § (skip ◆) math derivations
6 Ch 5 Hashing (At a glance + body, no Lab) ▼ ☉ § ◆
7 Ch 12 Adversarial validation ▼ ☉ § ◆
8 Ch 16 Procedural primitives all —
9 Ch 19 Five challenges all —
10 Ch 20 Four attestation kinds all —
11 Ch 25 Reference verifier all —
12 Ch 26 Admissibility Auditor all —
13 Appendix C Legal primer all —
14 Appendix B Worked attestation all —

The Admissibility Auditor (Chapter 26) is the chapter you will refer to most often. It maps every Canon field to the admissibility question (authentication, best evidence, hearsay, reliability, disclosure) it helps the proponent address. Bring it to the bench.

Path C — The Investigator

For journalists, academic researchers, NGO monitors, oversight officers working with private archives that conventional tools cannot index reliably. You will probably build a small Meridian.

Order Read Sidebars Lab
1 Overture all —
2 Ch 1 Crisis all —
3 Ch 4 Canon all —
4 Ch 5 Hashing all Lab 5
5 Ch 9 Embeddings all Lab 9
6 Ch 10 Hybrid retrieval all Lab 10
7 Ch 11 LLM extraction all Lab 11
8 Ch 12 Adversarial validation all (review)
9 Ch 14 Postgres substrate all Lab 14
10 Ch 15 Idempotent ingestion all Lab 15
11 Ch 16 Procedural primitives ▼ ☉ § —
12 Ch 20 Four attestation kinds all Lab 20
13 Ch 25 Reference verifier all (review)
14 Ch 27 Capstone as your project scoped capstone

A short Investigator’s capstone in Chapter 27 trims the 10-week plan to roughly four weeks — you are unlikely to need the full five-challenge harness, but the ingestion + retrieval + signed-search artifact will materially change how your investigation’s records can be cited.

Path D — The Policymaker

For legislators, AI-governance staff, standards bodies, courthouse technologists. You read for the shape of the debate — what is settled, what is contested, what a near-future legal regime should require.

Order Read Sidebars
1 Overture all
2 Ch 1 Crisis all
3 Ch 2 Falsifiability ▼ ☉ §
4 Ch 3 FRE primer all
5 Ch 4 Canon ▼ ☉ §
6 Ch 12 Adversarial validation ▼ ☉ § (skim ◆)
7 Ch 26 Admissibility Auditor all
8 Ch 28 Customization patterns all
9 Ch 30 Operationalization all
10 Appendix C Legal primer all
11 Appendix E Comparative architectures all
12 Appendix F Divergences from spec all

The chapters you will cite most: Ch 4 (the standard a regulation might require conformance with); Ch 26 (the deterministic check a court order might mandate); Appendix E (the existing standards you might map a new regulation onto — Sigstore, SLSA, C2PA, W3C VC, JAdES).

Path E — The Affected

For the person whose evidence the system handles. The pro-se litigant. The estate beneficiary. The subject of an investigation. You read for what your counsel should be asking, and for whether the record before you is trustworthy.

Order Read Notes
1 Overture The story is, in detail, fictional and composite. Yours may rhyme.
2 Ch 1 Crisis The fifteen-minutes problem named.
3 Ch 2 Falsifiability (▼ ☉ § sidebars; skim body) The discipline named.
4 Ch 3 FRE primer (read carefully) What admissibility doctrine asks. Not legal advice.
5 Ch 4 Canon (▼ ☉ § sidebars) What an attestation is.
6 Ch 26 Admissibility Auditor What questions to ask.
7 Appendix C Legal primer Where to look up the rules.
8 Appendix B Worked attestation A real artifact, walked.

A separate guide — front/06_for_pro_se_readers.md (forthcoming) — gives you a one-page checklist of questions to ask about any machine-generated exhibit produced by an opposing party. The book is not legal advice. It is a vocabulary that may help your counsel ask better questions.

Path F — The Curious Reader

For the popular reader who picked up the book without an institutional role — perhaps after a review, perhaps after a friend’s recommendation. You are owed an honest answer to “what is this book for?”

The shortest honest path:

  1. Overture.
  2. Ch 1 Crisis.
  3. Ch 4 Canon, At a glance + ▼ ☉ § sidebars only.
  4. Ch 12 Adversarial validation, At a glance + ▼ ☉ § sidebars only.
  5. Ch 27 Capstone, opening pages.

Approximately 90 minutes of reading. You will not learn the cryptography, but you will know the argument, the stakes, and the shape of the discipline. If at the end you want more, your next chapter is Ch 2.

Interpreting the sidebars

If you are not a Builder, pay particular attention to:

  • ▼ Why It Matters. The stakes box. The thing the chapter is in service of. Read these even if you skim everything else in the chapter.
  • ☉ In the Wild. A real case or news story. These are the chapter’s receipts.
  • § For the Record. Primary-source quotation. RFC, statute, rule, or case excerpt. The book honors the discipline of citing primary sources rather than commentary; you can copy these directly into a brief or a memo without wondering whether they were paraphrased.

If you are a Builder, also pay attention to:

  • ◆ Going Deeper. Technical depth. Where the chapter’s substance lives for you.
  • ✻ Try This. A small exercise. Anyone can do them; Builders should do them with code. The point is intuition, not proof.

Switching paths

Nothing locks you to a path. A Receiver who finds themselves curious about the math can pick up Path A’s chapters in any order; a Builder who suddenly needs the legal frame can drop into Appendix C cold. The book is structured so each chapter is self-contained enough that the cost of opening it is small, even if its full appreciation depends on the rest of the book.

The point of the paths is not gatekeeping. It is reducing the cost of the first reading to the lowest possible level for the broadest possible audience.