How to Study Multiple Subjects Without Forgetting Old Material

You are carrying several live subjects. Today’s new chapter fills the sitting. Last week’s topics get a reread that feels like review, or they get skipped because the headings still look known. Weeks later, the starved course is the one you cannot produce.

That is not a missing Pomodoro. New material occupies every long sitting; old material is either restudied into familiarity or left untouched. Familiarity is not recall. Studying multiple subjects without forgetting the old ones is an allocation problem: protect what is already learned with short delayed retrieval, then give new material one focused block.

Do not treat lab interleaving as a license to shuffle unrelated courses. Mixing similar problem types inside a subject is a different move from hopping biology → history → calculus. Maintenance and acquisition are different jobs. Equal minutes for every course every day is not the method.

If the problem is one idea you already studied and cannot produce, that is how to remember what you study. If the problem is one unread pile, that is how to study large amounts of information. If the job is where notes should live, see how to organize notes for studying. This page starts after you already have concurrent courses, and the old ones are at risk.

Why old material dies when you study several subjects

Under a multi-subject load, old material usually dies of starvation.

Each course keeps producing new chapters. The new sitting is long: you are still trying to follow the explanation. The remaining time is a skim of last week’s notes, or nothing, because those pages still look familiar. The skim confirms the headings. It does not test whether you can produce the idea with the page closed.

Rereading old notes is a weak maintenance action for the same reason it is a weak first-pass finish. The page does the work. People who practiced retrieving often remembered more later than people who spent the same extra time restudying. Restudy can still feel better on the day you do it. That feeling is why starved courses look “fine” until a later test.

If what you have is already captured lecture notes, repairing them and then covering the page is how to review notes effectively. Here the failure is coarser: the new chapter ate the sitting, so the old course never got a real closed attempt.

Packing one subject into a weekly marathon does not fix this. It just starves the other two for six days, then mass-exposes the third. Ability to retrieve often declines with unused time; that pattern is the forgetting curve, not a personal countdown and not a reason to cram a whole course on Sunday.

Two jobs: keep the old, learn the new

A mixed week has two jobs. They should not share the same sitting shape.

Maintenance is short retrieval of material you already understood. The source stays closed until you have tried. You check, repair a miss, and stop. The point is to keep the idea retrievable, not to cover new pages.

Acquisition is new learning. Stay with the source until you can follow the idea. Then close it and try to produce that idea once. Check. Leave. Guessing with the book closed at something you cannot yet follow is not practice.

Retrieval during the sitting and spacing across days are not the same lever. One is what you do; the other is when you return. Maintenance uses both: retrieve, then wait. Acquisition uses understanding first, then a first closed attempt, then a return on another day.

Match the job to the material without inventing a separate system per course. A new biology mechanism and a new historical argument both need a blocked acquisition pass until the chain is followable. Calculus problem types that you already know and keep mixing up are a later, within-subject mix — not a reason to rotate three syllabi every twenty minutes.

Not every page in a live course is a maintenance target. Reference lists, recaps, and pages you will look up can stay off the review schedule. Selection is how three courses stay reviewable.

Spacing and retrieval, not a prettier timetable

The calendar is not the method. A color-coded week that still spends every sitting rereading is the same starvation with nicer blocks.

During maintenance: retrieve. Cover the page, produce the mechanism or the distinction, then check. That is active recall with notes used as upkeep, not as a tour of unread chapters. Classroom work on retrieving earlier material, rather than restudying it, points the same way.

Across days: wait, then retrieve again. For knowledge you need after a delay, spreading the same effort across separate days usually beats packing that time into one long sitting. There is no universal best interval table. A gap that is useful for next week is not automatically useful for next year, and waiting too long can also hurt.

A major 2013 review rated practice testing and spaced practice as high-utility techniques. Interleaved practice was rated moderate: promising, especially in math-like discrimination tasks, but not as broadly established. Use that ranking as a priority, not as a timetable.

Packing a week of biology into one long Sunday dump is massed practice wearing a subject label. A short recurring Sunday return can still be spacing. Spread the work across days, not into one sitting. Put today’s new chapter in a bounded acquisition block, then let it wait. How to keep the page itself on a return schedule is spaced repetition for notes. This page only needs you not to spend the whole week acquiring while nothing old is tested.

Sleep after a learning sitting is associated with better later memory than equivalent time awake. It supports consolidation. It does not replace the due pass, and it does not encode the chapter you never retrieved.

Interleaving is not shuffling your whole timetable

Three mixings get sold as one word.

Spacing is the same idea on different days. Within-subject interleaving is mixing similar, confusable items inside a domain so you have to choose the right method. Hopping unrelated courses is a timetable pattern. The lab finding is mostly the middle one. It is not automatic permission to rotate your whole load.

Most of that evidence used short materials in one domain, not chapter-length explanations and not a semester of unrelated syllabi. Treat it as a within-subject tool. Do not promote it into a law of the week.

When mixing similar types helps

Mixing similar problem types or confusable examples often lowers scores during practice and raises them later, because you have to select the method instead of repeating the one you just used. Keep the scope honest: similar / confusable / same domain.

That is strongest when you already know the types and keep confusing them — related-rates versus optimization, two similar bio pathways, two neighboring historical causes you swap. It is a discrimination problem.

It is weaker, and sometimes the wrong move, for unrelated courses and for expository text. Most interleaving evidence is not about reading chapter-length explanations. For some word-learning tasks, blocked study has been the better condition. Switching between unrelated courses is not automatically the same effect. In one study of physics and statistics, mixing at every level was not the winning schedule.

When two sets of material are similar, studying one can make the other harder to recall. That is a documented effect in list-learning research. It does not mean that any second subject you study today will erase the first. A biology page and a history page in the same due pass are not the interference those studies were about. Two confusable biology pages might be.

When a blocked sitting is the better first move

Block when the sitting still needs comprehension. Brand-new material, a rule you have not extracted, a chain you cannot yet follow with the source open — those are acquisition. Mixing types before you can see the rule just adds noise.

If you are still trying to see what ties the examples together, a short blocked set can help. If you already know the types and keep confusing them, mixed practice inside that subject is more likely to help.

A blocked acquisition sitting is not a weekly marathon. It is one unit, understood, retrieved once, then left. The mix, if it comes, comes later, among similar items, not as a three-course shuffle.

Lab schedules also do not decide a semester. Those experiments control the materials, the delay, and usually the domain. Your week has overlapping deadlines, unread chapters, and problem sets. Practice testing and spaced practice transfer more cleanly than any interleaving timetable copied from a lab. Use the lab for the discrimination case. Do not wait for a study of “biology plus history plus calculus” before you protect old pages.

A practical week across several subjects

Say the load is biology, history, and calculus. Do not give each course a third of Monday.

Use two slots, in this order:

Due pass first. Mix by what is due, not by fairness. Monday might be two biology mechanisms from last week and one history argument from two weeks ago, with nothing from calculus because those pages are not due. Cover each page, produce the idea, check, stop. This should be short. It is maintenance of a mixed due list, not a tour of three courses.

Then one new block. Monday: the new related-rates section in calculus. Source open until the rule is followable. A short blocked set of the same type if you are still extracting the rule. Close the source. Produce it. Check. Stop.

Do not start the history chapter because you still have energy. That leftover energy is how old material used to die — the new sitting expands until there is no return.

Tuesday: due pass again. Different pages if different pages came due. Then one new biology unit — not three new chapters. History may get no acquisition block that day. It still got maintenance if anything was due.

Wednesday can be history’s new chapter as the acquisition block. If two new chapters are both loud, still pick one for the sitting. The other waits until tomorrow. Acquisition is bounded on purpose.

Across the week, each subject still appears: as due pages, as the new block, or both. None needs a daily marathon. None needs equal time every day.

Keep the work on pages, not on course labels. A cardiac-cycle page, a causes-of-the-revolution page, a related-rates page. The due pass is those pages. The new block is one new page (or one unit that will become a page). Filing those pages into a course tree is organization; it is not this method.

Exam week without abandoning the rest

When one exam is close, rank by urgency. More of the week’s acquisition blocks go to that subject. Calculus on Friday can own most new-learning sittings.

Keep a short due pass for the non-exam subjects. Shorten it. Do not delete it. Those pages are how biology and history survive the week instead of arriving dead afterward.

Do not abandon spacing for the exam subject either. Packed practice can look efficient on Thursday night. Spread the remaining days if you still have them. Retrieve first; do not reread the whole calculus folder as the whole plan.

Urgency ranking is triage, not a new system. The two jobs remain. Maintenance gets smaller for the courses that are not on the test. Acquisition concentrates. The due pass does not disappear.

Common mistakes

Spending every sitting on the new chapter. Acquisition expands until maintenance is a skim, or nothing.

Rereading old notes and calling it review. The headings look known because they are in front of you. Closed production is the maintenance action.

Skipping old material because it looks familiar. Familiarity is what starvation leaves behind.

Equal-time every course every day. Due pages are not evenly distributed. New chapters are not evenly urgent.

Packing a subject into one weekly marathon. That is massed practice with a course name on it.

Shuffling unrelated courses and calling it interleaving. Within-subject mixing of confusable types is the finding. Course-hopping is not a free upgrade.

Mixing problem types before the rule is visible. Block first when you are still extracting the idea.

Putting every page in every course on a review schedule. Reference and skip material will bury the pages that must come back.

Dropping non-exam subjects entirely in exam week. Shorten the due pass. Do not starve them to zero.

How Yuhimo fits this workflow

Yuhimo is a notes-and-learning app. The unit is a page, not a flashcard. A page can stay an ordinary note. Turn on learning only for pages you want to keep retrievable. Nested course branches help you find a subject later; they do not put every child page on a schedule. See nested pages.

The due pass is retrieve-then-check, not a reread reminder. When a learning page is due, the review is a closed attempt, then a check, then your own rating: Forgot, Barely, Remembered, or Easily. Yuhimo does not mark the attempt right or wrong. Optional questions can cue the attempt; a title can be enough. AI-drafted questions are drafts you edit. The app does not decide that you were correct.

The mixed due pass is the global daily review: a mixed list because those review-phase pages became due, not because Yuhimo is an interleaving engine. Queue order is not a guaranteed mix of courses. The working list is up to 15 due review-phase pages and does not silently expand after the session starts. Scoped sessions do not use that cap.

If you need a one-subject or exam-week focused review of already-learned pages in one branch, start a scoped session from that branch: Due only for review-phase pages that are actually due; Deep dive if you are choosing early reviews inside that subject. That extra pass is still review, not a new-learning block. Scoped sessions are for review-phase pages in the branch, not for learning-phase pages still on short first steps.

The new-learning block is the page itself. Stay with the source until the idea is followable, then close it and try once. After that, you can turn on learning if you want later returns. After you turn on learning, the page starts in a short learning-phase flow. It does not join the mixed global due list until it graduates. Those early returns are not a scoped session and are not the mixed due pass. Do not start a scoped session in order to do the acquisition block.

Global and scoped lists may overlap; there is no de-duplication. Grading in one consumes that day’s review grade everywhere. After a review-phase grade, that page cannot receive another review grade the same calendar day. Learning-phase pages may return the same day when the next short step is due. The rules are in one grade per day. Do not flatten this to “every page, one grade per day, always.”

If the job is thousands of tiny independent facts, a flashcard-first SRS such as Anki may fit better. That is a format choice, not a requirement to convert the three courses into decks. Connected explanations can stay pages: due pass first, one new page in the acquisition block.

Does every page in every course need a review schedule?

No. Only the ideas you must still produce. A named table you will look up, a recap of last week, and a skip box can stay ordinary notes. Three live courses stay reviewable when the schedule is small on purpose. The unread-pile version of that choice lives on the large-amounts page. Here the same rule is what keeps a mixed due pass short enough to finish before the new-learning block.