Where we start
This course runs on one collection. Daybreak, the coffee subscription company, and one person's forty-eight notes built up over eighteen months. Last week that person counted: forty-eight notes, several places, and of ten sampled notes the ones they could not honestly get back. Six of the forty-eight are called some version of nothing: "Notes", "Untitled", "Interesting", "Read later", "Misc", "Ideas". Two of those six matter. One is the meeting where the October launch date was agreed. One is a competitor tip somebody sent. In session 4 a bench asks for both. One of them comes back once something links to it. The other is lost in all five of the filings the bench tries, and the reason is its title. You brought your own worst title tonight. It is the same problem, smaller.
★ What you walk out with tonight. A capture rule you can apply without arguing with yourself, in about the time it takes to save the thing. And ten titles rewritten so that a search in eight months finds them, on the same ten notes you sampled last week.
Three neighbours, so you land in the right place. Atomic notes, permanent addresses and linking by idea rather than filing by folder is the Zettelkasten method, and that is learn-obsidian-with-phoebe. Turning a book into notes, highlighting technique and reading workflow is learn-reading-with-phoebe. Here the subject is much narrower: the four seconds when something arrives, and the name you give it before you close the tab.
What to keep, decided in four seconds
5 min liveA capture rule has one job that everybody talks about, which is quality, and one job that decides whether it survives, which is speed. A rule you have to stop and weigh gets used for a week. The version below is three questions you can answer without leaving the page you are on, and any one yes is enough.
LiveForte's rule, and what it is good for3 min▶
The most quoted capture criterion in this field is four words long. Tiago Forte's own wording: "Keep only what resonates." And he is explicit that it is not supposed to be a checklist. In his own sentence: don't make it "a highly intellectual, analytical decision. Instead, your rule of thumb should be to save anything that 'resonates' with you on an intuitive level."
People treat that as a quality standard. It is better understood as a speed standard. Its strength is that it costs nothing to apply. You do not open a second window, you do not weigh the item against a taxonomy, you do not ask which folder it belongs in. You notice that it landed, and you keep it or you do not.
Speed is what makes a capture habit survive a working week. A rule that takes thirty seconds per item gets skipped at the exact moments you most need it: mid-meeting, mid-call, on a phone, three tabs deep. The three questions in the diagram are a slightly more concrete version of the same idea, for people whose intuition has not been trained yet. Any one yes keeps it. No yeses lets it go, and you will not miss it.
What the rule does not do is tell you where to put it. That is deliberate, and part 3 is about why deferring that decision is what keeps capture at four seconds.
Self-studyThe tension the rule creates, stated not resolved4 min▶
There is a real problem with an intuitive keep-rule and it is worth naming rather than smoothing over. Neave, Briggs, McKellar and Sillence (2019) built and validated a Digital Hoarding Questionnaire across two samples: 424 adults to establish the structure, then a fresh 203 to validate it. The final instrument is ten items in two factors, difficulty deleting and accumulating. Both sub-scales correlated at r = .416 with the sheer number of digital files a person kept at work. Emails were the most commonly hoarded item type in both samples.
So difficulty discarding is a measured trait with an instrument behind it, not a moral failing and not an excuse. And an intuitive keep-rule errs in exactly one direction. "Does this resonate?" asked quickly, under mild pressure, at the end of a long day, resolves to yes far more often than it resolves to no. The rule is fast because it removes the friction, and the friction was the only thing standing between you and keeping everything.
Both halves of that are true at once, and neither cancels the other. Here is the honest professional position, stated in three clauses. A capture rule should be fast, because a slow one is not used. A collection needs a discard step somewhere, because keeping is frictionless and discarding is effortful and a system with only one of those goes one direction forever. And if the capture rule contains no discard, then the review has to carry it, which is what session 5 is for.
The Daybreak person saved the competitor tip because it resonated. A friend sent it, it was genuinely useful, and it went in without hesitation. Then they titled it "Interesting" and closed the tab. Eight months later they remembered that somebody had told them something about a competitor and could not produce it. The capture rule worked perfectly. The title destroyed the entire value of the capture, which is why that note is still lost in session 4's bench and stays lost in every filing.
The title is the retrieval handle
6 min liveEight months from now you will not remember the day you saved it, the tab it came from, or the mood that made it resonate. You will remember one or two words about the thing itself. Those words are either in the title or they are not, and that is nearly the whole game.
LiveA title that survives eight months4 min▶
The pattern is three parts, and the third is optional only when there genuinely is no name or date attached to the thing.
- What it is. Meeting, call, article, quote, idea, decision, spec. One word, and it tells you what kind of thing you are opening before you open it.
- What it is about. The subject in the words you would use out loud. Not the category it belongs to, the subject.
- A proper noun, a date, or a number, if there is one. A person, a company, a product, the day it happened. Proper nouns are the most searchable words that exist because nothing else in your collection uses them.
So "Notes" becomes "Launch meeting 14 Mar - October date agreed". Same note, same contents, nothing moved. The difference is that the second one contains the words "launch", "March", "October" and "date", and at least one of those is a word your future self will type.
The test takes about five seconds and it is the only part you must not skip. Name the word you would search for in eight months. Now check that the word is in the title. If it is not, the title is decorative. This is also the reason category words fail so completely: "Read later", "Misc" and "Ideas" describe your relationship to the note, not the note, and nobody has ever searched for their own relationship to a note.
One more check that costs nothing. Would a colleague looking for this thing type any word that is in the title? If the answer is no, you have written a title only you can use, and in eight months you are the colleague.
Self-studyWhy this matters more than the folder4 min▶
People assume that a good filing structure makes titles matter less. The personal-information-management research points the other way, and the numbers are worth knowing because they reorder the whole problem.
Bergman, Beyth-Marom, Nachmias, Gradovitch and Whittaker (2008) ran two studies comparing users of better and worse desktop search tools. Across both, people estimated they used navigation, meaning folder-browsing, for 56-68% of file-retrieval events, and search for only 4-15%. Improving the quality of the search engine had a "limited and inconsistent" effect on how often people actually searched, and search was used mainly as a last resort when they could not remember where a file was. These are self-estimates from questionnaires, not logged behaviour, and the full paper sat behind a paywall on this pass, so the figures come from the publisher's own indexed abstract.
Read that carefully, because it is easy to take the wrong lesson. It does not say titles do not matter. It says most retrieval is navigation and search is the fallback. Which means the two routes stack. A note you can navigate to is fine whatever it is called. A note you cannot navigate to still has one route left, and that route runs entirely through the words in its name.
Now put the two failures together. A note buried in a folder too big to scan, that nothing links to, and whose title is a category word, has no navigation route and no search route. It is not hard to find. It is unreachable, and no amount of reorganising later will change that, because reorganising moves notes between places and this note's problem is that it has no handle in any place.
| What arrives | Bad title | Title that survives | The word that finds it |
|---|---|---|---|
| Notes typed in the March launch meeting | Notes | Launch meeting 14 Mar - October date agreed | launch, October |
| A supplier call about bean prices | Untitled | Supplier call 2 Feb - Ridgeway price held to June | Ridgeway, supplier |
| A tip about a competitor, sent by a friend | Interesting | Competitor tip - Kettle Co ships in 24 hours | competitor, Kettle |
| An article saved from a browser tab | Read later | Article - why subscriptions churn at month 3 | churn, month 3 |
| Three printer quotes pasted into one note | Misc | Packaging quotes - three printers compared | packaging, printers |
| A half-formed thought on the way home | Ideas | Referral idea - refer a friend, one free bag each | referral, free bag |
One inbox, emptied on a schedule
4 min liveThere is a third thing that has to be true for four-second capture to hold, and it is the one people skip. The destination cannot require a decision. If saving something means choosing where it goes, you are not capturing in four seconds, you are filing in forty, and filing at the moment of arrival is when you know least about the thing you are filing.
LiveCapture to one place, always the same place3 min▶
Pick one destination and use it for everything, whatever the tool. A note titled "Inbox", a notebook called Inbox, a folder, a channel you message yourself in. The tool matters much less than the fact that it never changes, because the moment there are two plausible destinations you have reintroduced a decision, and the decision is the thing that was costing you the capture.
The decision about where a note really lives is a real decision and it deserves real attention. It just does not deserve it at the moment of arrival. At arrival you know the least you will ever know about this item: whether it turns out to belong to a project, an ongoing responsibility, or nothing at all is usually clear a week later and almost never clear in the four seconds. Session 3 makes that decision properly, on a schedule, with a rule behind it.
What you get for deferring it is the only thing that makes capture survive: no branching. Arrives, keep or let go, name it, inbox, done. One path, every time, on a phone, in a meeting, three tabs deep.
Self-studyWhat Forte's own framing gets right here4 min▶
Forte defines a Second Brain, in his own words, as "an external, centralized, digital repository for the things you learn and the resources from which they come. A Second Brain ultimately expands our memory and our intellect using modern tools of technology." Read the word "centralized" and you would expect one app.
He then spends myth number one of his own list of nine correcting exactly that reading. Verbatim: a Second Brain "isn't a thing you either have or don't have... It is a distributed system made up of apps, devices, websites, social media platforms, cloud and storage drives, organizations, and people... In today's world, we all have Second Brains."
That correction is more practically useful than it first looks, and it changes what "one inbox" can honestly mean. It does not mean one app to rule them all, because nobody has that and nobody is going to get it. Your phone camera roll is part of the system. So is the shared drive at work you do not control, and the messaging app where half your useful links arrive. The realistic target is one inbox per system you actually use, each with a known exit, rather than one inbox in theory and five accidental ones in practice.
Which makes the second number from last week's audit the important one. Not how many notes you have, but how many places they are in. That is the count to shrink, and shrinking it from six to three does more for retrieval than any amount of reorganising inside any one of them.
Rewrite ten titles
★ 22 min · ten notes, one at a timeThe same ten notes you sampled last week. Not new ones, not better ones. Roughly two minutes each, and the rule for the next twenty-two minutes is that nothing moves. Titles only.
Read the title only. Do not open the note. Write down, in a few words, what you think is inside it. If you cannot, write "no idea", and notice that you have just reproduced exactly what will happen in eight months, except that tonight the note is still in front of you.
Say what it actually is. Now open it and say it in one plain sentence out loud. "It is the notes from the call with the printer where we agreed the run size." That sentence contains your new title. It almost always does.
Write the new title. What it is, what it is about, and the proper noun or date if there is one. Keep it to a line. You are not writing a summary, you are writing a handle, and a handle you cannot read at a glance is not one.
Check the search word. Name the single word you would type in eight months, then look at the title and confirm that word is in it. If it is not, the title is decorative and you fix it now, in five seconds, rather than discovering it in eight months when you cannot.
Do the other nine. The first two are slow and the rest are fast, because the pattern is three slots and you stop inventing. Nine more at ninety seconds each is the whole exercise.
Start with your worst title. The one you brought tonight. It is the easiest win in the ten and it calibrates the pattern while you still have energy.
Use the words you say out loud, not the words you would file under. "Bean prices" beats "Procurement". You will search for the thing, not for its department.
Put a proper noun in wherever one exists. A supplier, a person, a product, a company. Proper nouns are nearly unique inside a personal collection, which makes them the highest-yield words you can spend title space on.
Dates earn their place when the thing happened. Meetings, calls, decisions. They do not earn it on articles and ideas, where the subject is what you will search for and the date is noise.
Resist writing a sentence. If the title needs a comma and a clause, the note is probably two notes. Make a mental mark and leave it; session 3 has the room for that.
Leave everything where it is. No new folders, no moving, no tagging. If you reorganise tonight you will not know next month whether the titles or the tidying did the work.
Do not reorganise anything tonight. The urge arrives around note four and it is strong, because moving things feels more productive than renaming them. Titles alone are a change you can test: search for one of your rewritten notes next Thursday using only the word you predicted, and see whether it comes back. That is a measurement. A reorganised collection full of unsearchable names is not.
Before session 3
25 min- Retitle everything you capture this week, as it arrives, using the three-part pattern, and note how long it actually takes. Most people find it is a few seconds once the pattern is familiar, which is worth knowing because the assumption that it is slow is what stops people doing it.
- Count your capture destinations and write down which one becomes the single inbox. Notes app, phone, saved messages, bookmarks, the email you send yourself. Pick the one you already reach for without thinking, not the one you wish you used.
- Find three notes whose titles are fine but which you still could not find. Those are the ones where the title works and the place does not, and they are session 3's problem rather than tonight's. Bring them.
- Bring the ten retitled notes. Same ten as last week, new names. Session 3 gives each of them exactly one home.
Three questions
Clear all three and session 2 gets stamped in your passport.
1 · What makes a note title survive eight months?
The pattern is what it is, what it is about, and a proper noun or date if there is one. Category words describe your relationship to the note rather than the note itself, and nobody searches for their own relationship to a note. The five-second test is to name the word you would type in eight months and confirm that word is in the title.
2 · Bergman and colleagues asked people how they retrieve their own files. What did they report, and why does it matter for titles?
The finding is often misread as "titles do not matter because people browse". It means the opposite in practice: the two routes stack, and search is what is left when navigation fails. A note in a folder too big to scan, that nothing links to, and whose title is a category word, has lost both routes at once. These were self-estimates from questionnaires rather than logged behaviour, and the figures come from the publisher's indexed abstract.
3 · Your capture rule is fast and intuitive. What is the honest thing to say about what that costs?
Forte's criterion is deliberately non-analytical, and that is its strength: speed is what makes a capture habit survive a working week. Neave and colleagues built a validated ten-item scale across samples of 424 and then 203, with both sub-scales correlating at r = .416 with files retained at work, so difficulty discarding is measurable and ordinary rather than pathological. The two facts sit together unresolved, and the practical consequence is where the discard step has to live.
What this session teaches, and where it came from
Every fact on this page, and its verification tier, is recorded in the course's source map.
Session 2 cheat sheet · pin this
Your ten notes now have names that can be found, and an inbox that will fill up by Friday. Session 3 is the other half: where each note lives, one home and only one, using Forte's four categories in his own words. The argument to test is his own, that you should organize by actionability rather than by subject, and the reason it beats filing under "Marketing" or "Psychology" turns out to be about how fast the system decays, not about how tidy it looks.