Publisher’s Reading Notes Became an App

Abstract
This essay traces how a publisher’s private habit of writing notes while reading books and
watching films became a software tool. Over more than twenty years in publishing, and in my
own writing, I have kept one discipline: to respond to a work one chapter or one scene at a time, in my own words. The habit worked; the places where I kept the notes did not. I describe the path from those notes, through my Interactive Reading Guides, to Chapter & Scene, a notebook app for iPhone, iPad and Mac. I explain its design principles, why it deliberately contains no generative AI, how it was built, and where it falls short.
1. A habit before a product
I have been a publisher for more than twenty years and a writer for longer. In both roles I do one small thing consistently: I write down what I think while I read or watch. A paragraph after a chapter. A few lines after a scene that stays with me. A question I cannot yet answer.
These notes were never meant for readers. They keep me honest about what I have actually
understood, and they give my later self something to work from. When I come to write an essay, prepare a guide or decide whether a book deserves an edition, I do not start from the work alone. I start from what I thought about it the first time.
For most of that time the notes lived wherever I happened to be (a folder of documents, Apple Notes, or outlines in different apps such as Drafts, Keep It, OmniOutliner, SnipNotes and many others available in the Apple ecosystem). The habit was sound. The container was not.
2. Notes that lose their place
A note is most useful when it remembers what it answers. A line such as “the ending changes
everything” means little a year later unless I can tell which chapter or scene it was written about, and what I had just read when I wrote it.
Most tools lose that connection. A general notes application treats everything as a flat page of text. A word processor treats everything as a manuscript in progress. Neither respects the structure of the thing being read. When I wanted to turn my notes on a film into an essay, I had to rewatch the film, or reread my own scattered lines, to reconstruct the order in which my thinking had developed.The natural unit of response, I came to see, is the one the work already offers: the chapter in a book, the scene in a film, the episode in a series. A reader responds to a unit, then to the next, and the response accumulates. An essay is often only that accumulation, edited. If the notes keep their place, the first draft of the essay already exists.
3. From reading guides to a notebook
The idea did not come from software. It came from my own books. Since the Apple II I have been an early adopter of digital publishing, and I created the Interactive Reading Guides: fourteen companion guides in four strands, Literature, Film, Philosophy and Themes, available on Apple Books and Google Play. Each guide divides a work into chapters or scenes and invites the reader to respond to each one in turn.
A related paper of mine, a framework for the interactive analysis of Jean Rhys’s Quartet, made the structure explicit. It distinguishes the reader’s own responses from the examples I supply, and it compiles the chapter responses into a portfolio. The reader produces the interpretation; the guide provides the scaffold.
Looking at that structure, I noticed that it was also the skeleton of my private notes: a unit, a
prompt, a response. The only difference was who supplied the content. In the guides, I supplied the prompts and the reader supplied the responses. In my notes, I supplied both. What I wanted was the same scaffold without the guide: an empty notebook that already knew about chapters and scenes.
4. Principles of the design
Chapter & Scene rests on five decisions. They are less about features than about what the app refuses to do.
The unit is yours. A “chapter” or a “scene” can be whatever you decide: the chapters of a novel, the scenes of a film, the episodes of a series, or titles you invent for your own purposes. Only the title is required. A summary, tags and a prompt of your own are optional.
Your words come first. Each unit has a box for your response, saved as you type. The app can
show a question if you wish, and otherwise stays silent. It never writes the response, and it never suggests one.
The notes are private. They are stored on the device. If you are signed in to iCloud, a copy is
kept in your own iCloud account so that your other devices can read it. There is no account to
create, no advertising and no tracking. The app collects nothing, and I cannot read anyone’s notes.
Notes must leave the app. Notes are raw material, so the whole review can be exported as a
Word file, a plain text file or a PDF, or copied as text. Every export includes the full structure, anda chapter not yet written is marked “(no response yet)”, so the export tells the truth about what exists.
It goes where the reader goes. The app runs on iPhone, iPad and Mac, and it adapts to the
iPhone Duo’s folding screen, keeping the chapter list beside the writing area when the phone is unfolded. I use all these devices, and I wanted the notes to follow me.
5. Building it, and a paradox
The first version was a single web page that I used for my own reading and film notes. It had no icon, no store listing and no sync, and it did what I needed. Turning it into an app took longer than writing it: wrapping it for the iPhone and iPad, adding iCloud sync between my own devices, designing the export, and then the less glamorous work of testing on real hardware, writing a privacy policy and passing App Store review.
I did this work in collaboration with an AI system, Claude, which wrote and tested much of the
code while I made the design decisions and checked the results on my own devices. Xcode let me test the app on different devices and operating-system versions, and I repeated those tests with the films I had watched and the books I was reading, coming back each time with new ideas to add to the app. I mention the collaboration because it creates an apparent paradox: I used an AI to build a tool that deliberately keeps AI out of the reader’s thinking.
I do not think it is a contradiction, because the two activities differ in kind. Engineering is a task where a collaborator can do much of the labour and where the result can be checked: the export either opens in Word or it does not. Interpretation is the thing a reader is there to do.
Research on generative AI points in the same direction, cautiously. A survey of knowledge
workers found that greater confidence in the AI went with less reported critical thinking (Lee et al., 2025), and an experiment found that AI-assisted stories were rated as more creative on
average but were more similar to one another (Doshi & Hauser, 2024). Neither study concerns reading notes, so I take them as reasons for care, not as proof. For my own app I made the simplest choice: the machine helped me build the notebook, and it does not write in it.
6. Limits
It is a first version, and it should be honest about its limits. It is not a reference manager, and it
does not store the books or films themselves. It has no sharing between people and no
collaborative editing. iCloud holds about one megabyte of the app’s data in total, which is plenty for typed notes but means that very long reviews should also be exported as a backup. Chapters cannot yet be reordered, and notes cannot yet be imported from other files; both are on my list.
And because the app contains no AI, it cannot summarize, suggest or search by meaning. Those are choices, but they are also limits.7. What comes next
The next version should let a reader bring material in as well as send it out: a list of chapters from a text, Markdown or spreadsheet file, and eventually notes from a Word document (a Pages file can be exported to Word first), split into chapters and shown for review before anything is imported. Everything would be read on the device, from files the reader chooses, so the privacy design stays as it is. I am also collecting suggestions from the first readers, because a notebook is only finished when it fits the habits of the people who use it.
The larger point concerns how notes and essays relate. I began with a private discipline, writing after each chapter and each scene, and found that it was also the structure of a publication and then the structure of a tool. If notes keep their place, they can become drafts, and in time finished material, for guides, essays, reviews or books. The app is only a place to keep them. The thinking, as it should be, stays with the reader.
Chapter & Scene is free on the App Store for iPhone, iPad and Mac. I would be glad to hear how others keep their notes.
References and links
• Doshi, A. R., & Hauser, O. P. (2024). Generative AI enhances individual creativity but
reduces the collective diversity of novel content. Science Advances, 10(28), eadn5290.
https://doi.org/10.1126/sciadv.adn5290
• Lee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025).
The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort
and confidence effects from a survey of knowledge workers. Proceedings of the 2025 CHI
Conference on Human Factors in Computing Systems, article 1121.
https://doi.org/10.1145/3706598.3713778
• Pinto, J. Reading to Structured Engagement: A Framework for Interactive Literary Analysis of
Jean Rhys’s Quartet. Academia.edu. Link
• Pinto, J. Interactive Reading Guides. Academia.edu. Link
• Pinto, J. From Machine Writing to Idea Processing: The Macintosh and the Age of Claude.
Academia.edu. Link
• Chapter & Scene on the App Store: https://apps.apple.com/app/chapter-scene/id6823308610
• Support and feedback: Chapter & Scene — Support.
Keywords: reading notes; note-taking; digital publishing; interactive reading; film and literature; personal knowledge management; privacy by design; iOS app
