True "Learning Mode"

I am a retired psychologist and am using Anki to learn a language. One of the problems to avoid is learning something that is wrong. Once a mistake “feels” right it is inclined to be repeated. If the wrong answer is learned it is more difficult to “unlearn” it in the process of “relearning” the correct association. I propose a mode where the front side and the back side of the card are presented at the same time in order to achieve the correct association from the beginning - no guessing. How can that be done?

1 Like

With the default settings, both sides are shown at once when you reveal the answer. There is no simple way to change Anki to show both sides on the question, other than modifying your card template and changing it back later.

1 Like


Thank you for your response. As I noted in my e-mail, I think it should be a serious programming consideration to facilitate the initial learning process. Without this, the only initial association is when data is entered for the cards. Even that may be non-existent if the cards are imported.

When I am learning my vocabulary I don’t guess the translation of the foreign word on the list, and if I would do that, I would most certainly guess wrong and that wrong answer would inhibit learning the correct translation.

I am not a programmer, but there already is an option “answer” which gives the front and back of the card. The initial learning option would do that but with no delay and no guessing inclination (especially with false cognates).


I’m no psychologist or linguist, but there’s an inherent assumption that you first learn the word or concept, then use Anki to reinforce the learning,

1 Like

I understand what you are saying, but as you say “… first learn the word or concept…” The question is then how is that done? That seems to mean going back to the list of words or the concepts repeatedly until they have been learned. That process is what Anki could do and the learning would be more efficient.

1 Like

Yes it depends per case. if you make your own cards then the learning process is part of making the card. For using pre-existing cards it’s another story.

1 Like

What you’re saying sounds reasonable but in such a mode you wouldn’t have the opportunity to assess whether you already knew a card.
As Anki shows you which queue a card is from, can’t you just immediately hit space for new cards before any false cognates can form? That’s what I usually for decks in which there is a very low probability that I can answer new cards.

1 Like

One may not be in this mode for very long, this is to build the initial association. Again, this is an attempt to initially provide the correct answer without having to guess and make a mistake. Once that is done, other strategies would be used to retain the association. False cognates are apart of the list so come up as any word would.

1 Like

But the first time you see a card is the only opportunity to objectively assess whether you new the card beforehand. If you see the answer at the same time it’s at least hard to do that unbiasedly.

I got that. But since Anki shows you which queue a card is from, you can make a habit out of revealing the answer instantly for new cards without reading it first.


I am not quite sure I know what you mean. I would like to know your point of view. Could you restate it for me please?

I’ll try to rephrase it.
When Anki shows you a new card you have two options:

  1. Read the question and try to come up with the answer. Then reveal the answer and rate how you did. If you came up with the correct answer, you can rate Easy or Good and the card will be shown to you after a longer interval.
  2. Reveal the answer and read question and answer at the same time. Would you have known the answer? That can be hard to judge in this case, so your ability to rate the ease of the card is impaired.

In a learning mode, you wouldn’t have option 1. However, option two would be automated and you couldn’t mistake the new card for a review card and try to think of the answer you probably don’t know.

If you’re entering in material yourself, chances are you remember some of it from the input process, and revealing the answer immediately would mean you waste time with an extra preview step when you could have answered the card normally. Even when you’ve sourced material from someone else, there are likely to be times when you do know some of the material already.

But if you’re still keen on the idea, one alternative you can use is the Preview button in the Browse screen to go through both sides of selected cards at once.


Yes. That is what I mean.

No guessing, no mistakes.

There may be some associations learned while imputing the data, but that is minimal and the learner might not know for sure which card that was. When I am entering data, my concentration is on the data entry process and spelling.

The Browsing function is very busy and not conducive to the initial learning process.


I responded too quickly. I went to Browse but not to Preview. Preview does what I am suggesting, but you should not have to go around your back to get to your elbow. What I am suggesting would be a real part of the learning process and therefore stand alone.

Hoping I understand your question correctly: Have you considered using hints & cloze deletion for those questions? After you’ve built the early association, as you’ve said, then you can delete the hint from the cloze deletion.

The spanish word for hello is {{c1::hola::hint here}}


What’s the spanish word for hello?
{{c1::hola::hint here}}

How do I make a custom study session with all of my cards?

"Preview does what I am suggesting, but you should not have to go around your back to get to your elbow. "

I have been using the Preview option forever to learn English and Spanish. My workflow looks like what you want:

  1. Make 20 Recognition cards FRONT side:
    • Front: naked word
    • HINT Field: First Letter of the Word - as needed
    • HINT Field: phrase; collocation and such
    • Fields .Mnemonics’ and ‘Picture’ - as needed.
  2. Preview cards ( concentrating on the Word / Phrase) to check and edit the Template - if required.
  3. Take One minute break – not thinking of the words.
  4. Trying to memorize, Preview BOTH sides.
    • On difficult cards, add Mnemonic or a Picture.
  5. Repeat previous step again
  6. ‘‘Study now’’ screen:
    • Review in Random order
    • Think for < 2 sec; Rate the answer “Again”;
    • Repeat the answer aloud.
  7. Assuming that the 2nd Learning Step is 20min, wait until the Learning step is expired. Meanwhile,
  8. review the “Due” cards.

As you can see, i do not waste time rating my cards without putting them in short-term memory. The common practice of testing how well you remember the cards ( without ‘‘pre-learning’’, as medical students call it ) is a waste of time – as /dae stated above.
Also i just stumbled upon a post : AnKingMedAnKing does wait for 20 or so min ! That means a lot - should not be ignored.

The incidental benefit of this workflow ( and as important though) is ability to see the “Rate of success” for the second learning step ( 20 min in this case; so a user will not waste time failing on step =< 20 .

One more benefit: I never experienced the Ease-Hell caused by bad learning phase.


Thank you for your helpful reply. I am in the middle of something so I can’t give your method a try at the moment. I will try it soon.