Feature Request: Language Learning Integration

[Maitreya20201103]

Checklist

Please replace the space inside the brackets with an x if the following items apply:

• I've verified that I use the latest version of the add-on by redownloading it from AnkiWeb
• I've verified that I use the latest version of Anki by checking at https://apps.ankiweb.net#download
• I've checked if anyone else suggested this feature before by looking through the issue reports.

Problem case

Is your feature request related to a problem? If so, please describe it here. E.g.: "My workflow is such and such, and this and that would help."

When learning a language, it is often the case that
a term in one language
may translate to an ordered list of terms in the other language
where order corresponds to frequency of choice

For example

The word āsavas in the Pali language
Translates to the following words in the English language:
cankers
corruptions
taints
toxins
mental biases
influences
inflows,
influxes
outfluxes

Solution

Concisely describe the solution you would like

I would like to be able to create the ClozeOverlapper card like so:

Question: āsavas

Answer:
cankers
corruptions
taints
toxins
mental biases
influences
inflows,
influxes
outfluxes

AND be able to generate a user chose number of "Basic" cards for language learning from it like so:

card 1:
English: cankers
Pali: āsavas (1st translation)

card 2:
English: corruptions
Pali: āsavas (2nd translation)

card 3:
English: taints
Pali: āsavas (3rd translation)

...

There are a few benefits to this approach.

1: We want the ability to be able to recall the correct term in one language when presented with the prompting term in the other language, but keeping track of the complexity of the mapping can be a pain. This approach solves this problem by making the ClozeOverlapper note the model from which many cards are generated. If you want to change the mapping, you just change the model and regenerate.

2: This aids with the discovery of duplicates in the reverse direction. (English to Pali)

For example, perhaps the English term "corruption" translates to more terms in Pali than āsava.

This suggests a Cloze Overlapper list in the reverse direction also needs to be created & maintained:

Question: corruption?

Answer:
āsava
dūsana
kalusīkaraṇa
kilissana
pūtibhāva

and Basic cards might be generated from this list like so ...

card 1:
Pali: āsava
English: corruption (1st translation)

card 2:
Pali: dūsana
English: corruption (2nd translation)

card 3:
Pali: kalusīkaraṇa
English: corruption (3rd translation)

...

This use of lists for both direction of translation in language learning illustrates the idea of "separating the model from the manifestation".

i.e. The human brain creates a model of the world which is separate from the manifestation that is being tested for in the cards used by spaced repetition.

It is my strong intuition that Anki will eventually be rearchitected such that the Anki kernel will contain "model" primitives which can more easily be used by card type creators to create "manifestations" that better meet the needs of the users.

It is my strong intuition that this set of primitives will mirror those that are actually used by the sensory motor brain (see Jeff Hawkin's "Thousand Brains Theory of Intelligence") and that the "metamodeling language" will thus be beautifully simple:

Namely:
1: the prediction
2: combinations of smaller predictions to form larger predictions

Wherein each prediction is of the form:
(current conditions) + aim + move = (desired conditions)

Concisely describe alternatives you've considered (if any)

More information

Additional context: Add any other context or screenshots about the feature request here.