Incremental reading (Basic level)



Traditional linear reading is highly inefficient. This comes from the fact that various pieces of the text are of various importance. Some should be skipped. Others should be read in the first order of priority. Old-fashioned books are quickly being replaced with hypertext. Hypertext will help you quickly jump to information that is the most important at any given moment. Hypertext requires a different style of writing. All linear texts can assume that the reader is familiar with the preceding sections. This makes them context-poor. Within hypertext, individual texts become context-independent, and all difficult terms and concepts are explained primarily with additional hyperlinks. In the same way in which the web helped delinearize the global sources of information, SuperMemo can help you delinearize your reading of whatever linear material you decide to import to SuperMemo. While reading with SuperMemo, you will see a linear text as a sequence of sections subdivided into paragraphs and individual sentences. SuperMemo will help you provide a separate and independent processing for each section, paragraph or sentence.

What is incremental reading?

Incremental reading is a learning technique that makes it possible to read thousands of articles at the same time without getting lost. Incremental reading begins with importing articles from electronic sources, e.g. the Internet. The student then extracts the most important fragments of individual articles for further review. Extracted fragments are then converted into questions and answers. These in turn become subject to systematic review and repetition that maximizes the long-term recall. The review process is handled by the proven spaced repetition algorithm known as the SuperMemo method.

Incremental reading converts electronic articles into durable knowledge in your memory. This conversion requires minimum keyboard&mouse work:

  • Input: electronic articles (e.g. collected from the net)
  • Output: well-remembered knowledge (quizzed regularly in the form of questions and answers)

In incremental reading, you read articles in small portions. After you read a portion of one article, you go on to a portion of another article, etc. You introduce all important portions of texts into the learning process in SuperMemo. This way you do not worry that you forget the main thread of the article, even if you return to reading months later. Your progress with individual articles may be slow, but you greatly increase your efficiency by paying less attention to less important articles and spending more time on articles that are more beneficial to your knowledge. Difficult articles may wait until you read easier explanatory articles, etc. Last but not least, incremental reading increases your efficiency because it is fun! You never get bored. If you do not like an article, you read just a sentence and jump to other articles. This way your attention and focus stay maximized.

Warning! Incremental reading may seem complex at first. However, once you master it, you will begin a learning process that will surpass your expectations. You will be surprised with the volume of data your memory can process and retain!

Five basic skills of incremental reading

Incremental reading requires skills that you will perfect only over months and years of use. This overview will only help you master the basic skills and help you make a start with incremental reading. The 5 basic skills are:

Skill 1: Importing articles

Five article import methods

Initially, you may limit your imports to a simple copy&paste of individual articles. Later, you will want to master automatic imports from the web that offer many advantages.

Here are the 5 main article import methods in SuperMemo:

Import by Copy&Paste

To import an article with copy and paste, follow these steps:

Please remember that if you have many articles opened in Internet Explorer, you can most easily import them with web import as described in the next section.

Import of multiple articles

The most convenient way to import learning materials to SuperMemo is a direct import of multiple articles right from the web. To import many articles at the same time, open these articles in Internet Explorer, and click the import button on the learnbar (or press Ctrl+Shift+A, or chose Edit : Import web pages on the main menu). To avoid importing advertising and other garbage, in Internet Explorer, select portions of the text that is to be imported. If you select texts before imports, you are less likely to need filters to get rid of troublesome HTML (F6). If you prefer to use other browsers, e.g. Chrome or Firefox, you will need to use Copy&Paste method, or re-open the selected articles in Internet Explorer. This is because, at the moment, SuperMemo supports direct imports only from Internet Explorer.

For more details on importing multiple articles see: Web import

Dedicated imports (Wikipedia, YouTube, and pictures)

The most popular sources of learning materials are Wikipedia (for incremental reading) and YouTube (for incremental video). For those sources, you have separate options available on the import menu (right click on the import button on the learnbar). You can also use shortcuts Ctrl+Shift+W (Wikipedia) and Ctrl+Shift+Y (YouTube) to import from those two sources. If you choose those options, only articles available from those sources will be displayed on the import list. Additionally, a dedicated filter will best prepare the imported pages for convenient learning.

SuperMemo: A topic with an article about the greenhouse effect imported from Wikipedia

After importing an article about the greenhouse effect from Wikipedia (e.g. with Edit : Import web pages : Wikipedia (Shift+Ctrl+W)), its entire text is stored in a single topic.

If you want to import pictures, you can also use a picture filter that will ignore all non-picture pages opened in Internet Explorer. Use Pictures on the import button menu on the learnbar (right click). You can also click Filter : Pictures in the web import dialog when importing pages (e.g. with Ctrl+Shift+A).

For more details see: Web import

Importing articles from local files

If you want to import articles from files that reside on your local drive, you can use the following methods:

Skill 2: Reading articles

Here is a simplified algorithm for reading articles:

  1. Choose an article: Import an article as explained earlier or bring up previously imported articles with Learn (Ctrl+L). Learn will display only articles imported in the past. If you import an article, and want to have it shown later during a learning session on the same day, you must place it in the outstanding queue (e.g. Learning : Later today on the element menu, Ctrl+Shift+J, etc.). If you import many articles that you want to process on the same day, you must place them all in the outstanding queue. For example, open the articles in the browser, and choose Learning : Add all to outstanding (or use Add to outstanding icon on the browser toolbar). Most of the time you can rely solely on Learn to schedule articles optimally for review.
  2. Click the article to enter the editing mode, in which you can modify text, select fragments, etc. Optionally, use filter F6, if the text is hard to process (e.g. selections are hard, extracts not marked correctly, etc.)
  3. Start reading the article from the top or from the last read point (i.e. a bookmark left in the text last time you read it)
  4. Extract texts: If you encounter an interesting text in the article, select it and choose Remember extract on the learnbar (or press Alt+X). This operation will introduce the extracted fragment into the learning process as an independent mini-article. If you would like to specify the priority of the new extract, choose Reading : Schedule extract (Reading : Schedule extract, available on the Read toolbar, enables you to specify the priority of a text fragment being extracted) instead of Remember extract. Also, if you have an impression that the article is difficult and you would like to read some fragments later, extract those fragments with Reading : Schedule extract and provide a review interval that will reflect the time you believe you will be better equipped to understand the extracted fragment.
  5. Optionally, use Delete before cursor (Alt+\). This will delete the text that you have read, clean up the article, remove garbage, and help tackle HTML that is difficult to process.
  6. Optionally, if you read a fragment that seems unimportant, select it (e.g. with the mouse) and either delete it (e.g. with the Del key) or mark it with the ignore style. To mark a text as ignore, choose Reading : Ignore on the component menu, click the Ignore text button (Reading : Ignore makes it possible for you to mark an unimportant fragment with the ignore style) on the Read toolbar, or just press Ctrl+Shift+I.
  7. Optionally, if the selected fragment does not include all the important reading context, you can add this context manually. For example, if you are learning history, you may extract the following fragment from an article about Lincoln: On Sept. 22, 1862, President Lincoln issued the Emancipation Proclamation, one of the most important messages in the history of the world. He signed it Jan. 1, 1863. If you would like to extract the fragment related to signing the Emancipation Proclamation, you will need to change He to Lincoln and it to Emancipation Proclamation so that your stand-alone fragment is understandable: Lincoln signed the Emancipation Proclamation on Jan. 1, 1863. You can use the Reference options on the component menu to easily add context to your extracts (see: References). Context added by Reference will be added automatically to all extracts of a given article. For example, select the text that you want to serve as the reference title of all extracts and choose Reference : Title on the HTML component menu (or press Alt+T). This text will appear at the bottom of all extracts (in reference pink font by default).
  8. Optionally, mark your last read point: Once you decide to stop reading the article before its end, mark the last processed fragment as the read-point (e.g. with Ctrl+F7 or by choosing Reading : Read-points : Set read-point from the HTML component menu). Next time you come back to this same article, SuperMemo will highlight your read-point and you will be able to resume reading from the point you last stopped reading the article. To go to your current read point, press Alt+F7. If you forget to set a read-point, SuperMemo will leave a read-point at the place of your last extract or last highlight.
  9. Go to the next article: After you finish reading a portion of one article, choose Learn or Next repetition to proceed with reading other articles. Those buttons are located at the bottom of the element window. You can also use Enter, which will work as long as the selection in the text is not empty (e.g. marked as a reading point), or if you have left the editing mode (e.g. with Esc). If no text is selected, Enter will add a new line in the text (as is the case with standard text editors).
  10. Optionally, determine the next review date (e.g. with Ctrl+Shift+R), or set the new priority for the article (e.g. with Alt+P).
  11. In incremental reading, interrupted reading is a rule, not an exception! With a dose of practice, you will quickly get accustomed to this not-so-natural state of affairs and learn to appreciate the power of the incremental approach. The main role of interruption is to prevent the decline in the quality of reading. Use the following criteria to decide when to stop reading the article:
    • lack of time: if you still have many articles for review for a given day and your time is running out, keep your increments shorter. After some time, being in a hurry will be a norm and you will tend to read only 1-2 paragraphs of each article and dig deeper only into groundbreaking articles that will powerfully affect your knowledge.
    • boredom: if the article tends to make you bored, stop reading. Your attention span is always limited. If your focus is poor, you will benefit more from the article if you return to it after some break. Go on to reading something that you are not yet tired of. If SuperMemo schedules the next review at a date you consider too late, use Ctrl+J or Shift+Ctrl+R to adjust the next review date.
    • lack of understanding: if you feel you need more knowledge before you are able to understand the article, postpone it (e.g. use Ctrl+J or Shift+Ctrl+R and schedule the next review in 100 days or so). If you believe you have already imported articles with relevant explanatory knowledge, you can search for these articles (e.g. with Ctrl+F). Once you find them, you can (1) execute a subset review, or (2) add the articles to the outstanding queue for reading on the same day, or (3) advance the articles (for example, in the browser, you can execute: Learning : Review all, or Learning : Add to outstanding, or Advance : Topics). If you have not yet imported any explanatory articles, you can do it now (e.g. search the web and import articles as explained before). Note that you can select a piece of text in SuperMemo and use Ctrl+F3 to search encyclopedias or dictionaries for more material on a given subject.
    • lower priority: read lower priority articles in smaller portions thus reducing the overall time allocation to low priority subjects.
    • overload: if you have a long queue of articles to read, you will naturally read in smaller portions
  12. Once you complete reading the entire article and have extracted all the interesting fragments, choose Done! (Learning : Done will remove the article from the review process, and delete its contents (without deleting the extracted material) on the learnbar. You can also press Shift+Ctrl+Enter, choose Done! in the Commander, or choose Learning : Done on the element menu. Done! will dismiss the article, i.e. remove it from the review process, and delete its contents (without deleting the extracted material). Done will delete a childless article (i.e. an article that did not provide any interesting extract). Using Done will greatly reduce the size of your collection and eliminate "dead hits" when searching for texts.

The Read toolbar docked at the learnbar. It hosts options used in incremental reading.

Figure: The Read toolbar docked at the learnbar. It hosts options used in incremental reading. For details see: Read toolbar

Skill 3: Extracting fragments, questions and answers

Extracting texts

In the course of traditional reading, we often mark important paragraphs with a highlighter pen. In SuperMemo, those paragraphs can be extracted as separate mini-articles that will later be used to refresh your memory. Each extracted paragraph or section becomes a new element that will be subject to the same reading algorithm as the original article. Extract important fragments and single sentences with Extract (SuperMemo: Extract button on the learnbar in the element window). Use Alt+X, Extract on the learnbar, or Extract on the Read toolbar.

Adding references

Why need references?

In incremental reading, you always need to quickly recover the context of a question or a piece of text. The easiest way to recover context quickly is via references. References propagate from element to element as you produce extracts and cloze deletions. With all child elements produced from a given text marked with references, you would never need to worry about losing the context of the question.

For example:

Q: He was born in [...](year)

cannot be answered without the context. However, the following question is already easier to understand:

Q: He was born in [...](year)

#Title: Barrack Obama
#Source: Wikipedia

To speed up learning, in the incremental reading process, the above question should naturally be replaced with:

Q: Obama was born in [...](year)


Q: Obama was born in [...](year)

#Title: Barrack Obama
#Source: Wikipedia

References are not stored in HTML files that hold your articles but in a reference registry (i.e. in a separate database). The reference registry does not hold the text of references either. All reference texts are held in the text registry and are available for global text searches. In earlier versions of SuperMemo, each text would keep its own copy of references. In newer SuperMemos, elements keep only pointers to reference registry, which in turn keeps pointers to individual text fields in the text registry. As a result, many elements can hold the same reference, and many references can hold the same text. This results in a significant saving in space in your collection. More importantly, you can update the reference in a single element and see the change show in all elements using the same reference. This way, you do not need to waste time on search&replace to correct a single misspelling or reference inaccuracy that propagated to many elements.

Example: references inserted automatically

If you select the title of the source article and press Alt+T (Reference : Title on the HTML component menu), each extract will be marked by the title of the source article. If you use File : Import web pages : All, your articles will be provided with basic references (such as #Title, #Link, #Date, etc.). If you need more context (e.g. to add the author, the journal, etc.), you can use the Reference link button (SuperMemo: Reference button on the navigation bar in the element window) on the navigation bar to jump to the source article from which the extract was produced. On the parent article, that button will lead you to the original link on the net.

SuperMemo: An extract produced from an article about the greenhouse effect (references (in pink) at the bottom are added automatically)

Figure: Typical snapshot of incremental reading. While learning about the greenhouse effect, the student extracts the fragment saying that "In the absence of the greenhouse effect and an atmosphere, the Earth's average surface temperature of 14 °C could be as low as -18 °C, the black body temperature of the Earth.". The extracted fragment will inherit illustrations placed on the right, as well as article references. The student can move on to reading another article by pressing Enter. The picture on the right is stored locally in the image registry (on the user's hard disk) and can be reused to illustrate other articles or questions.
Reference tips
When SuperMemo is not sure if your edits are local or global, it will ask you

Important! Do not add your own non-reference texts below the horizontal bar marking the reference area. All reference field area is owned by SuperMemo. Any modifications to that area will be treated as changes to reference fields. Changes that do not conform with reference field formatting will be discarded without warning.

SuperMemo: References help you quickly recover the context of a given element as well as track its source and build a list of citations (in the picture: blue marks an incremental reading extract, yellow marks a search string (i.e. GABA-ergic), while pink marks the reference field, which will propagate to all children elements (extracts and clozes))

Figure: An extract from an article on sleep and dreaming. Blue marks an extract produced from the presented text. Yellow marks the search string (i.e. GABA-ergic REM-on neurons) that was used in Search : Find elements (Ctrl+F) to find all the elements (including this one) containing the string. Pink marks the reference area (consisting of the #Title, #Author, #Date, #Source, #Article, #Parent, and #Category fields), which will propagate to all children elements (extracts and clozes) generated from this element.
SuperMemo: References are kept in a dedicated registry while their individual text fields (e.g. title, author, date, source, etc.) are stored in the text registry, and thus are available for global text searches
Editing references

You an use Reference : Edit in SuperMemo Commander, however, you can also edit references in the reference area (which is pink in the default stylesheet). You can safely delete reference fields, but you need to decide if that change should be local (for that element only) or global (for all elements using this reference). You will not be able to delete #Article or #Category fields because they are added automatically to the reference section (i.e. they are not part of the reference itself). You can freely change the text of references. Illegal changes are all changes that do not comply with the reference format, e.g. lines that do not start with reference field tags, or lines that start with unknown reference field tags (e.g. #Country). If you are unsure how this process works, import a single article from Wikipedia to a newly created collection, create some extracts and play with editing to see how references are processed.

Cloze: Generating questions

SuperMemo will show you that extracting important fragments and reviewing them at later time will have an excellent impact on your ability to remember. However, it will also show that once the time between reviewsincreases beyond 200-300 days, reading and re-reading (passive review) will often result in insufficient recall. For this reason, sooner or later, you will need to convert your texts to specific questions. For that purpose you will use cloze deletion.

A cloze deletion is an item that uses an ellipsis ([...]) to replace a part of a sentence.

For example:

Question: The capital of Sierra Leone is [...]
Answer: Freetown

In incremental reading, cloze deletions are generated from topics that have a form of a sentence or a simple paragraph.

To create a cloze deletion do the following:

  1. make sure a topic contains a short sentence only (e.g. The capital of Sierra Leone is Freetown)
  2. select an important keyword in that sentence (e.g. Freetown)
  3. do one of the following:

Remember cloze will convert a sentence into a specific question with an answer. By using cloze, you will move from passive review to active recall. You do not need to wait until a paragraph or a sentence becomes hard to recall in passive review. For your most important material, you can create cloze items immediately after finding a piece of information that you need to remember well.

The examples below show how to effectively use Remember cloze.

Two numbers from the extracted sentence are used as keywords for generating questions and answers (temperatures of 14 °C and -18 °C)

Figure: Two numbers from the extracted sentence are used as keywords for generating questions and answers (temperatures of 14 °C and -18 °C)

A question-answer item (in the form of a cloze deletion) forming the final product of incremental reading used in strengthening the memory of a given fact (here: hypothetical temperature on Earth devoid of atmosphere)

Figure: The sentence extracted during incremental reading (see the previous picture) is converted into a cloze deletion. (i.e. a question-answer pair forming the final product of incremental reading used in strengthening the memory of a given fact (here: hypothetical temperature on Earth devoid of atmosphere)). The picture from the original extract has been inherited (on the right). Pink texts at the bottom of the question are references generated automatically when importing an article from Wikipedia.

When you click Cloze, you will not see your newly generated cloze. Only the selected keyword will change the color. This will speed up your work. However, if you would like to immediately edit the newly created cloze deletion, choose the back button (Back enables you to go back to the most recently visited element) on the navigation bar or press Alt+Left arrow. This will make it possible to add context clues, shorten the text, improve the wording, etc.

Simplifying questions

While converting extracts into questions and answers, you should make sure your questions are simple, clear and carrying the relevant context. For example, if you have extracted the following fragment from your reading about the history of the Internet:

The Internet was started in 1969 under acontract let by the Advanced Research Projects Agency (ARPA) which connected four major computers at universities in the southwestern US (UCLA, Stanford Research Institute, UCSB, and the University of Utah)

you may discover that when review intervals become long enough, you may not actually be able to recall the name of the ARPA agency or even forget the year in which the Internet started. You can then select an important keyword, e.g. 1969, and use Remember cloze to produce the following question-answer pair:

Question: The Internet was started in [...] under a contract let by the Advanced Research Projects Agency (ARPA) which connected four major computers at universities in the southwestern US (UCLA, Stanford Research Institute, UCSB, and the University of Utah)
Answer: 1969

In the course of learning, you will need to polish the above item by manual editing it to a more compact and understandable form:

Question: The Internet was started in [...](year) under a contract let by the ARPA agency
Answer: 1969

Or better yet:

Question: The Internet was started in [...](year)
Answer: 1969

As for the precious information "lost" during the editing, it can (but does not have to) be learned independently with separate questions generated by Remember cloze.

The mini-editing of questions presented above added the following benefits to the newly created question-answer pair:

  1. clearer purpose of the question: the fact that the question is about the year in which the Internet began is emphasized by using the red-colored (year) hint.
  2. brevity: by removing superfluous information, you will not waste time on information that is not likely to be remembered (only actively recalled material will be remembered for years). You will answer the question and never focus on which universities were originally connected by the early Internet. If you believe this information is also important, you will use the original extract to produce more cloze items that will focus solely on the universities in question by naming them in the answer field (if you disagree, read: 20 rules of formulating knowledge).
  3. understandability: "the ARPA agency" phrase may defy grammar rules you have learned in primary school, but it is by far more understandable than just the ARPA. In SuperMemo, understandability is more important than stiff rules of grammar or spelling!

Skill 4: Repetition and review

SuperMemo is based on repetition. You will review the learned material from time to time to make sure you prevent forgetting.

If you have never tried SuperMemo before, you will need to get the hang of standard repetitions as described here.

In incremental reading, your review will be based on similar principles as in classical SuperMemo. The main differencesare:

Incrementally processed articles will be subject to periodic review/reading. When you resume reading an article after a certain period, you will proceed to new sections, extracting newly acquired wisdom intoseparate elements with Alt+X (i.e. Remember extract). Usually, you will delete the remnants of the processed article with Delete before cursor (Alt+\).

The algorithms that determine the timing of (1) repetitions of question-and-answer material and (2) reviewing reading material are analogous but not identical. Most importantly, all repetitions and article presentations happen in increasing intervals by default. In incremental reading, you will see a constant inflow of new articles into your collection. Unprocessed material will need to compete with the newlyimported material. Increasing review intervals make sure that your old material fades into lower priority if not processed early. The speed of processing will depend on the availability of your time and the value of the material itself. Articles that are boring, badly written, less important for your work or growth, will receive smaller portions of your attention and may go into long review intervals before you even manage to pass a fraction of the text. That is an inevitable side effect of a voluminous flow of new information into your collection and into your memory. However, intervals and priorities can easily be adjusted. If your priorities change, you can modify the way you process important articles. At review time, you can either read the entire article without interruption, or bring it back for review in a shorter interval. You can manually change its priority (e.g. with Alt+P). You can also use search tools (e.g. Ctrl+F) to locate more articles on the subject that you feel you have neglected. You can reprioritize a bunch of articles by changing their priority. You can shorten intervals of articles, or review them all when needed (see: Subset review).

The algorithm for reviewing questions and answers (e.g. cloze deletions) is quite complex and limits your influence on the timing of repetitions (see: SuperMemo Algorithm). This is to ensure that you achieve a high level of knowledge retention, which might be compromised by manual intervention. However, the algorithm for determining inter-review intervals for topics is much simpler and is entirely under your control. Each article receives a specific priority. The priority determines which articles are reviewed first and which can be postponed in case you run out of time. Each article is also assigned a number called the A-Factor that determines how much intervals increase between subsequent reviews. For example, if A-Factor is 2, review intervals will double with each review. Priority and A-Factors are set automatically, but you can change them manually at any time. You can also set priorities for individual categoriesof the learning material. Priorities and A-Factors are determined and modified heuristically on the basis of the length of the text, the way it is processed, the way it is postponed or advanced, and by many other factors. You can change the priority and A-Factor of an article by pressing Alt+P. You can also use Shift+Ctrl+Up arrow and Shift+Ctrl+Down arrow to increase or decrease an element's priority. Note that A-Factors associated with items cannot be changed by the user, as they are a reflection of item difficulty that determines the length of optimum inter-repetition intervals (see: Forgetting index).

You can control the timing of article review by manually adjusting inter-review intervals. Use Ctrl+J (Reschedule) or Shift+Ctrl+R (Execute repetition) to determine the date of thenext review. Ctrl+J will increment the current interval, while Shift+Ctrl+R will first execute a repetition and then set the new interval. For example, if your current interval is 100 and you specify thevalue of 3 in Reschedule, your new repetition date will be set in 3 days, and the last repetition date will not change (the new interval will be 103). If you do the same with Execute repetition, your new interval will be 3 and the last repetition date will be set to today. In other words, Reschedule increments the interval (it can also shorten intervals), while Execute repetition sets the length of the interval (while leaving a trace of a repetition executed in the learning process). Note that Reschedule executed during the repetition cycle will first complete the repetition and will have the same effect as Execute repetition.

In a heavily overloaded incremental reading process, you will often want to focus on a specific subject on a given day. For that purpose, readabout the priceless concept of subset learning.


Skill 5: Handling large volumes of knowledge

In incremental reading, you may quickly import and produce more learning material than you can effectively process. To make sure that you can swiftly handle the overload, SuperMemo uses the priority queue.

Using Alt+P (Learning : Priority : Modify on the element menu), you can set each element's priority from 0% to 100%. Note that 0% correspondswith high priority!

By default, the outstanding repetitions will be auto-sorted from high to low priority. This way, if you fail to complete your daily load of learning, it will only be the lower priority material that will suffer. Also by default, at the beginning of your working day (i.e. at your first run of SuperMemo), your outstanding material from previous days will be be auto-postponed (again with high-priority material being least affected).

Read an article about the priority queue to learn more about:

  1. manual sorting of elements,
  2. defining sorting criteria,
  3. turning off auto-sort and auto-postpone, and more.

For more options for handling the overload, see:

Other basic skills

Evolution of knowledge in incremental reading

3 main principles will underlie the evolution of knowledge in SuperMemo:

Using pictures

For additional information, mnemonic cues, and a sheer fun of learning, an article that you read incrementally in SuperMemo can be illustrated with meaningful pictures taken from its contents, or from other sources. Press Ctrl+F8 to choose one of the pictures embedded in the article.

SuperMemo: A topic with an article about the greenhouse effect imported from Wikipedia

If you happen to import from Wikipedia, SuperMemo 16 makes it possible to download full resolution images instead of just thumbs. Check images marked with THUMB!!! and click Download.

SuperMemo: Download images dialog box makes it possible for you to get images embedded in local pages imported from the net and put them to the image registry (in the picture: A Wikipedia article on atherosclerosis)

Figure: After importing a Wikipedia article on atherosclerosis, two thumbs have been selected and downloaded in full resolution (one shown in the picture). Insert will insert the picture to illustrate the article and all its extracts and clozes. The remaining thumbs will be available for download in all portion of text that refer to their corresponding images.

For more, see: Visual learning

Topics vs. Items

In SuperMemo you see pieces of information presented to you in 2 basic forms:

Topics and items are presented in a different manner and at different times. Topics keep the knowledge you want to learn (i.e. things you want to read about), while items keep the knowledge that you want to remember (i.e. the knowledge you already posses, but might forget).


A topic in SuperMemo is an article, its part, or a sentence that you want to learn. Topics can also have a form of a picture, a video, a piece of music, etc. Unlike items, topics donot test your knowledge. They are used in passive reading, watching, or listening only. Short text topics are used to generate cloze deletions. Topics take part in the incremental learning process. Once they are converted to items, they are often dismissed (i.e. ignored in learning) or done (i.e. deleted from the learning process altogether). Both Done! and Dismiss must be executed by the user (i.e. they are not automatic).

Using topics

Topics are marked in Contents with a green T icon (A topic taking part in the learning process). Topics may be very long (entire articles) or very short (single sentences). This is how you work with topics:

On longer topics you read and extract, on very short topics you generate cloze deletions.


Item in SuperMemo is a piece of knowledge that you want to remember. It usually has a question&answer form. The main difference between an item and a topic is that an item actively tests your memory (e.g. with a question), while a topic is used for passive review only (e.g. for reading, viewing, watching, etc.).


In addition to items and topics, you can also use tasks in incremental learning. Tasks are jobs sorted by Value/Time or Value/Price ratio.

For an extensive comparison of items, topics and tasks in SuperMemo see: Element types in SuperMemo.

Reading overload

Overload occurs when the student has more outstanding items or topics to review than (s)he can handle. Few users can sustain more than 200 item repetitions per day. When the Outstanding parameter in the Statistics window starts going above that number, overload is likely.

Overload can best be handled with Auto-postpone. However, a one-time big load can be resolved efficiently with Postpone (delaying all elements), or Mercy (spreading all review intime).

You can also postpone a specific topic with all its extracts using the following method:

  1. Go to the topic in question
  2. Press Ctrl+Space to open the topic, its extracts, and clozes in the browser
  3. Choose Process browser> : Postpone on the browser menu

Note that you may need to use Learning : Locate extracts on the element menu if you have moved portions of your learning material to other branches.

See also:

Auto-sort and auto-postpone

As long as you prioritize your learning material well, you should make your life easier by checking the following 2 options:

and it ensures that you minimize delays for top priority material.

Auto-postpone always leaves a number of top-priority elements in the queue. The purpose of the postpone is to get rid of the main mass of low-priority material and focus on top-priority material. You aremost likely to use Postpone after a day of learning, while Auto-postpone is executed before your learning day begins. This is why it never affects today's material, and does not postpone top-priority material from previous days. If you have Auto-postpone checked on the menu, you will always start the day with all the repetitions scheduled for that day, and a number of unexecuted top-priority repetitions from previous days. Even though Auto-postpone increases the intervals and reduces the retention of low-priority material, it also makes you benefit from the spacing effect. Research shows that longer intervals may paradoxically increase the speed of learning (up to a certain point). This comes from the fact that the default retention in SuperMemo (around 95%) is higher than the retention that delivers the largest number of items remembered per unit of time invested.

You can start with default settings of the sorting criteria, however, if you feel you make insufficient progress with items (e.g. high forgetting index), you can reduce the proportion of topics. If the inflow of new material is too slow, you can increase the proportion of topics. If your priorities are imperfect, increase the degree of randomization. If you think you miss too many high priority items (see: Tools : Statistics : Analysis : Use : Priority protection from the main menu), reduce the randomization. By trial and error, you will arrive at your optimum. Even after you find your optimum, keep experimenting with different randomization and topic levels. This will help you avoid various cognitive biases that develop through the routine of learning. It may also be helpful to execute random review from time to time (just to get a general feel of your overall progress).

With Auto-sort and Auto-postpone, you will nearly never have to worry about material overload. Each time you start SuperMemo for the first time on a given day, it will first postpone repetitions that you failed to execute on previous days. It will use default postpone criteria which you can always modify (e.g. with Learn : Postpone : All elements). After postponing the backlog of repetitions, SuperMemo will sort today's repetitions and those that were left outstanding by Auto-postpone. Auto-sort will use sorting criteria specified earlier with Learn : Sorting : Sorting criteria.

With Auto-postpone and Auto-sort, you can always begin your day with a manageable portion of material sorted by priority. Your learning sequence will be optimized with no action on your part (i.e. no options to choose, and no keys to press).

Overload hints

Subset review

Subset review is a review of a portion of the learning material (e.g. before an exam). The portion may be identified with search, by branch selection in Contents, by category, and other means that determine a subset of elements. The reviewed subset material may be sorted by its sequence in the knowledge tree (Contents), priority, difficulty, interval, retention, recency, etc.

Review types

Search and review

Search and review in SuperMemo is a review of a subset of elements that contain a given search phrase. For example, before an exam in microbiology, a student may wish to review all his knowledge of viruses using the following method:

  1. search for all elements containing the phrase virus (e.g. with Ctrl+F)
  2. review all those elements (e.g. with Ctrl+Shift+L)

The review may include all subset elements (e.g. Learning : Review all in the browser with Ctrl+Shift+L), or only the elements that are outstanding for review on that particular day (e.g. Learning : Learn in the browser with Ctrl+L). Before you execute the review, you can randomize the review material (Ctrl+Shift+F11), sort it by priority, by recency, by interval, by size, by age (in the learningprocess), etc. You can also apply your default sorting criteria with Ctrl+S in the browser. All forms of review run on all elements except for (1) dismissed elements and (2) those elements that have already been processed on this particular day. The latter condition makes sure that you can do a comprehensive review in various subsets without duplicating your work on a given day. You can overcome the block of double review on a given day by using Add all to outstanding (see below).

If you build an extensive collection of things worth learning, subset review may help you learn about Subject A, and do a value-rich review of your material across many others domains at the same time! You kill many birds with one stone.

Search : Find elements makes it possible to define OR-searches and to save search definitions. This way you can, for example, choose a set of terms that define your "diabetes" subset and use them each time you want to review your "diabetes" material.

The parameter Subset in the Statistics window indicates the progress of repetitions in subset learning. This field displays the number of items, the number of topics, and the number of pending elements in subset learning. The name in the parentheses describes the currently processed subset.

Branch review

In the simplest case of branch review, the button Learn at the bottom of the contents window can be used to execute outstanding repetitions on a selected branch of the knowledgetree. For example, to make repetitions in the Medical Sciences branch, click that branch and then click Learn. Using Learn in the contents window is like using Learn in the element window, except only elements belonging to the selected branch will be considered in making repetitions.

To thoroughly review a branch of knowledge (including non-outstanding elements), do the following:

  1. Select the branch in Contents
  2. Choose Learning : Review all on the Process branch>menu (Ctrl+Shift+L)
Random review

To randomly review a branch or a subset:

  1. open the branch or subset in the browser
  2. randomize the content of the browser (Ctrl+Shift+F11)
  3. execute a review (e.g. Learning : Review all on the browser processing menu)

Review modes

In a browser subset, in a contents branch, or in the entire collection, there are 3 main ways of executing subset review:

All the above options are available on Process branch> : Learning submenu of the contents menu, or Process browser> : Learning submenu of the browser menu.

Adding elements to the learning queue

Instead of spending your time on a thorough review of a branch or subset, you may prefer to intersperse the review material in your standard learning process. You can do it with Learning : Add to outstanding.

Add to outstanding is a rationalization upon 2 extremes:

  1. well-timed incremental learning (or classical spaced repetition)
  2. subset review (e.g. before an exam)

On one hand you can proceed with your outstanding queue, on the other, you can smuggle some subset review in between. For example, you might learn about the superiority of "intermittent fasting" over "fasting". You will want to investigate the subject to perhaps employ it in your lifestyle. However, you do not want the subject to be buried in thousands of articles you keep reading. Nor do you want it to monopolize your learning time on a given day. You can import several articles on intermittent fasting and spread them sparsely in your outstanding queue with Add all to outstanding. By the end of the day, you will have a peek at all those articles, have them all well prioritized, and integrated with the learning process (in proportion to the value of the newly discovered content).

An alternative to Add to outstanding is to spread priorities (with Priority : Spread), however, it has 2 flaws:

In other words, Add to outstanding is a more extreme version of Priority : Spread, but not as radical as Learning : Spread in the browser (irreversible rescheduling), or subset review (reversible review).

Repeating items before topics

If you ever neglect learning, you may wish to unload your item backlog ahead of your topic backlog. You can optimally do it with a change to your sorting criteria. However, you can also start your day from 100% item repetitions:

  1. Choose View : Outstanding
  2. Sort repetitions by type (items first)(e.g. by clicking the header of the Type column), or choose Child : Items
  3. Choose Learn on the browser menu to make repetitions (or Tools : Save repetitions on the same menu to make the sorting permanent)

Ideally, in incremental reading, you should have items and topics mixed up. This will help you achieve balance between retention of the old material and the inflow of the new material. By working with items first, you risk slowing down learning by working on high retention. That's a step back to classical SuperMemo.

Semantic review

If you want to semantically connect a group of elements related to a single subject in incremental reading, you can use subset review based on the elements' tree structure created in the incremental learning process. This way you can quickly review all elements related to a topic whose "big picture" became hazy.

  1. go to any element that makes a part of the knowledge structure related to given problem
  2. use the Reference link button (SuperMemo: Reference button on the navigation bar in the element window) on the navigation bar to get to the original article
  3. use Contents to find a relevant subbranch (or stay at the root of the article to review it all)
  4. browse the selected branch (e.g. with Ctrl+Space)
  5. choose Learning : Review all (Shift+Ctrl+L) to review all elements in the logical/semantic sequence (the sequence of branches reflects the order of processing of individual paragraphs in the article)

You can also choose Learning : Learn branch on the element menu to begin branch learning for one of the ancestors of the currently displayed element. You can use this method if you encounter interesting material that you would like to refresh more thoroughly before you proceed with your standard learning process.

Advance review

Advance is not a form of review. However, it makes it possible to shorten the intervals and speed up the review. For example, if your exam comes in 100 days, you can shorten all intervals in a subset to less than 100 days with Advance.

The Advance operation will not work on 2 kinds of topics:

Examples of subset review

Example 1: search and review

If you would like to review the material related to Auguste Comte (1798-1857) do as follows:

  1. Press Ctrl+F and paste Auguste Comte in the search box
  2. Press Enter or click Find (this will search your collection and open a browser with the results)
  3. Choose one of the subset learning options:
    • to prevent forgetting: press Ctrl+L or choose Process browser> : Learning : Learn to review only the outstanding material. This will help you review only the items that are most likely to be forgotten and a portion of topics that have been scheduled for review for today
    • to learn new things: choose Process browser> : Learning : Review topics to review all topics related to Auguste Comte
    • to maximize the review (e.g. before an important deadline): press Shift+Ctrl+L or choose Process browser> : Learning : Review all to review all topics and to force a repetition on all items related to Auguste Comte. Remember that premature review of items may paradoxically slow down your long-term learning
Example 2: branch review

If your history exam is approaching and you cannot cope with all repetitions in the collection, make sure that at least your daily portion of history is thoroughly reviewed:

  1. Select History branch in the contents window
  2. Click Learn at the bottom of the contents window

Repeat the procedure daily. However, in the last 3-5 days, you could follow that yet with Process branch> : Learning : Review all to protectively refresh the material that would optimally be scheduled after the exam.

Important! The frequent use of Review all is not recommended for long-term learning! It departs from the optimum timing for review to consolidate memories. It should only be reserved for situation when burning school situation forces you to neglect your long-term planning.

If you have final drill enabled, remember that subset learning does not keep a separate final drill queue and elements that score less than Good (4) are put to the global final drill queue.

Hints and tips

Importing articles

Inflow of new articles


Generating clozes

Changing the appearance of cloze keywords

This is how you can modify the default cloze style in SuperMemo:

  1. From the main menu, select Tools : Options
  2. In the Options dialog box, click the Fonts tab
  3. On the Fonts tab, click the Stylesheet button
  4. In the SuperMemo Stylesheet dialog box, select the Clozed option in the drop list at the top; then use Font, Color, and Background buttons to set individual properties of that style
Removing cloze keyword formatting

Display the HTML code behind a given cloze text (e.g. with Ctrl+Shift+F6). In the HTML code, replace class=clozed with an empty string.

HTMLThis is my example <SPAN class=clozed>element</SPAN>This is my example <SPAN>element</SPAN>
WYSIWYGThis is my example elementThis is my example element

Your cloze keywords will be formatted in the same way as the surrounding text.

Mimic real life situations to combat memory interference

Some texts present knowledge in the form that is difficult to remember. Lists and sets are a good example of knowledge that does not stick to memory. Even if you perfectly know the map of Africa, answering the request: "List all countries of Africa" may be pretty hard. There are proven techniques that will help you tackle repetitive, list-rich, or boring knowledge with SuperMemo. All solutions are costly at memorization stage, but will pay handsomely in the long run due to lesser forgetting rate. The basic 2 principles are:

  1. gradually glue individual pieces to your overall knowledge structure
  2. be as visual and mnemonic as possible

Here are some specific hints:

Example: dealing with enumerations

If you happen to learn the geological periods, you are bound to generate nasty leeches, esp. if you are new to the subject. Using the top-down learning rule, be sure you know the eras, before you learn the periods, and before you move on to the epochs, and further down the tree of knowledge.

A typical mistake would be to start from cramming the meaningless sequence of periods. For example, clozing the Paleozoic Era sequence: "Cambrian, Ordovician, Silurian, Devonian, Carboniferous, Permian" could result in a question that is bound to cause problems: "Cambrian, Ordovician, [...], Devonian, Carboniferous, Permian". This cloze will trouble anyone who is not privy to the field. In other words, only those who come with the knowledge ready in their mind will be able to tackle this type of question at little cost! Conclusion: there is no point in learning lists the hard way unless you already know what you are trying to learn! Catch 22!

Instead of using the above approach, it would make far more sense to first anchor the Silurian period in your mind with some meaningful event. For example, the appearance of the bony fish. This way, we might start with a cloze based on "The bony fish appeared in Silurian (443-419 mn years ago)". The following question will be far easier to remember: "The bony fish appeared in [...](period)(443-419 mn years ago)". Even if the answer is the same as in the original unfortunate cloze, that question is not semantically equivalent. You will need more cloze deletions. However, working with similar sequences should always proceed incrementally and in proportion to anchoring individual periods in memory. Later on you can move on to clozing dates, epochs, and other details. All the time you should try to add new interesting anchors and work with the material in parallel to the inflow of meaningful information that is likely to stay long in memory.


Re-evaluation of items

You should remember that all items introduced into your learning process require endless attention in reference to their applicability, formulation, importance, logic, etc. In a well-planned learning process, it should not be necessary to review items in the periods between individual repetitions. However, when an item comes up for a repetition, you should make a quick and nearly instinctive assessment of the following:

  1. Do I really need this item?
  2. What is the honest priority of this item in the entire spectrum of my (desired) knowledge?
  3. Is this item difficult to remember? If so, why?
  4. Is it factually correct?
  5. Is it as simple and clear as it could be?
  6. Do I really need to know it now?
  7. Do you need supplementary knowledge to understand all ramifications of the item?

Here are some typical actions you will take depending on the answer to the above questions:

  1. edit the item. You will use keys such as Q, A, or E to enter a desired text field and edit it. In more complex items you will use Ctrl+T to circle between components, Alt+click to switch a component between editing and dragging modes, or Ctrl+E to enter the editing mode
  2. de-prioritize the item. For items that are not important enough, or you are not sure are important enough, use Alt+P and reduce their priority. You can also use Ctrl+Shift+Down arrow for minor deprioritizations
  3. reschedule the item. If you know the item well or for some reason want to manually increase (or decrease) the length of the inter-repetition interval, press Ctrl+Shift+R to select the date of the next repetition
  4. dismiss the item. If you are sure you are not likely to need the item in the future, but you would like to keep it in your collection for reference or archival purposes, press Ctrl+D. Dismissed items are removed from the learning process
  5. delete the item. The key Del is very useful in cleaning your collection from garbage that results from your desire to know more than your memory can hold. In editing mode or in spelling items (i.e. at times when Del plays text editing functions), you may need to use Ctrl+Shift+Del instead. Please note that deleting an element in SuperMemo will delete all its children! You may therefore wish to learn to always use safer Done (Ctrl+Shift+Enter) instead
  6. delay or forget the item. If you think the item is too difficult at the moment, you can postpone learning it. For this purpose, choose Ctrl+J to set a new interval or use Forget to transfer the item to the pending queue. This will give you some time to import some supplementary material that will help you understand the item


To learn more about efficient formulation read: Effective learning: 20 rules of formulating knowledge


To learn more about using pictures, see: Visual learning


Your own discoveries

In incremental learning, you will quickly discover why some of your own ideas about the learning process might not be optimum. Here are some things that you will discover on your own within the first 2-3 months of intense incremental learning:

Advantages of incremental reading

Now that you know how it works, you may wish to read more about the advantages of incremental reading in long term use. Those advantages are not instantly visible or obvious. Some will increase with the growth of your collection. Others will increase with the passage of time and strength of your memories. Others will depend strongly on good knowledge selection and honest priorities. For more see: Advantages of incremental reading.

Advanced incremental reading

Even before you get a good feel of incremental reading, it is recommended that you slowly go through Advanced Incremental Reading article that contains a great deal of hints, examples, and strategy suggestions. You will find there a detailed discussion of the advantages of incremental reading, opinions of users of incremental reading, criticism, etc.

Incremental reading: Summary