A color scale feature for tracking word familiarity progress

Hello Steve and Readlang community,

I was a user of LingQ for a long time until I discovered the joy of Readlang. One feature I really miss though was the ability to highlight words and be able to set a subjective score and color based on how much I felt I knew the word. It felt really motivating to open a new text and find that I already knew many of the words, and I could track my progress a lot easier without having to slog through some boring Anki card collection. Basically, the features that I think could make Readlang even better:

  • Highlight words that the user surmises to be worth learning with colors on a scale to visually track familiarity.
  • Rate word mastery on a scale from 1-5, encouraging subjective engagement.
  • Auto-highlight learned words when revisiting texts, reinforcing progress.

How I imagine the scale:

  1. Red (Unfamiliar) -The word is completely new or not understood. Needs focused learning and review.
  2. Orange (Recognizable) - The word feels vaguely familiar, but the meaning is unclear without context. Needs further exposure.
  3. Yellow (Partially Known) - The word can be understood in context but isn’t fully mastered. Requires occasional reinforcement.
  4. Green (Familiar) - The word is well understood and recognized in most contexts. Needs minimal review.
  5. Blue (Mastered) - The word feels fully internalized and can be used confidently in speaking and writing. No active review needed. Alternatively, it could be a grey color or simply no color at all.

I hope you find this worth considering! Thanks for reading.

2 Likes

Thanks for the suggestion! I wanted the Readlang reading experience to be distraction free and also I didn’t want the overhead of having to store your progress in all words the way LingQ does, which is why Readlang works the way it does.

You may want to check out Readlang’s existing auto-highlight feature: Auto Highlight Words As You Read - Readlang Blog. It doesn’t do exactly what you’re asking for but it might be interesting to you nevertheless.

Maybe I’ll consider more highlighting in future, but probably not in the short term.

Here’s another user who has enjoyed color-coding progress in the past. The parley spaced repetition software used to have color-coding of flashcards, and the same colors were then used as progress bars on flashcard sets (indicating the fraction of words on each of the levels, in ReadLang this could be used in the library). I’d love to see this in the reading interface and I might even be willing to mark up the 2000+ words I already know in the language I’m least fluent in just to get it, but I can’t imagine how much effort it would take to get satisfactory color-coding of the languages in which I’m more fluent.

Words for which a user doesn’t have a flashcard yet could be color-coded based on their frequency in a large corpus (different part of the spectrum than words for which a flashcard exists). This would help the user to decide whether they want to investigate an unknown word or not.

Moreover, at least one other feature of a word is competing for color-coding: there’s a request to color-code noun genders (Suggestion for Improvement: Color-Coding Nouns by Gender).

3 Likes

I would argue that gender-coding is more beneficial. At least in languages that have gendered nouns, naturally.

Hi Steve,

If you do ever consider it, please make it optional. I can understand why some might want it, but I don’t think I’m the only person who has come from LingQ to Readlang precisely to get away from their multicolour interface, and having to make decisions about how well I knew a word. I much prefer the clean look and simplicity of Readlang as it is.

2 Likes

Related, less coding-heavy wish:
While reading, it’s possible to highlight all “my” words, or just all words that are ready for revision. Make it possible to highlight all words for which I don’t have any flashcard yet, plus words that are ready for revision. Whereas the current approach assumes you already know all words for which you haven’t created a flashcard, this approach assumes that you don’t yet know any such word.

I admit that for this to be useful, there should be some way how to “unmark” the most common words without causing them to be revised with a short revision interval. Maybe being able to manually mark a card as “Graduated from revision”? Cards that you’ve remembered for longer than 2 years could then be graduated automatically, as it seems unlikely you’re gonna forget those any time soon.

The idea would be that OP and similar users might be happy when they see a page with very few highlighted words (instead of very many highlighted words, which leads to a lot of clutter).

I fear this would end up highlighting a lot more words than the current options do. Particularly for a user who has only been using Readlang a short while, selecting such an option would highlight the vast majority of the words on the page which would be overwhelming.

This would make Readlang a bit more like LingQ right? Where it will need to keep track of every single word you know to know whether to highlight it or not.

Yes, exactly. I think there’s a small fraction of users who’d love to do that, to open a text and immediately see which fraction of words they already know. (Myself, @Teksong , @Sr1, @Marius ?) For that, you actually need to keep track of all/most known words. But maybe I’m overestimating other people’s wishes…

And maybe it simply shouldn’t be encouraged because it’s such an inefficient way of using one’s reading time. I actually agree that skipping some words is more efficient than translating them all (and obviously more efficient than practicing flashcards for all words!); the problem for me is that when I’m tired, it’s emotionally easier for me to submit to my perfectionism and click them all.

My idea for color-coding nouns is simple and effective. The goal is to visually distinguish grammatical genders (e.g., masculine, feminine, neuter) by using different colors, so that users can easily identify the gender of each noun while reading. This approach will help learners better understand grammar and speed up the learning process, as visual cues are highly effective for memory retention.

1 Like

TL;DR: color coding would nice, even if I could only get it in edited flashcards. Ability to have the display engine automatically tag selected tense/grammar structures would be awesome, but important to be able to turn off selectively; I couldn’t handle the noise of 5 colors at once.

Long form:

(1) When I started learning Spanish, when I was more naive, I thought that if I learned enough “words” everything else would fall into place pretty quickly. So I spent many, many hours on Memrise – generally leading the pack in terms of cramming in words.

But I eventually wised up that (1) too much of the meaning of a sentence was in the phrases, and that knowing the individual words was actually worth almost nothing and (2) that the tenses, at least in Spanish, often twisted the words way beyond my ability to recognize them --much less spell them.

So, I guess I don’t care too much to have every word I run into, which I don’t know, emphasized nor counted; that just doesn’t line up with my actual goal of hearing and speaking very well (which might well be different, if I was more focussed only on reading … or was a linguist or technical reader, which maybe Anna is?). It is very well known that we humans have dramatically different sized vocabularies for (I think I have the order correct): speaking<hearing<writing<reading). I read to learn how to speak & hear, not the other way around, so I am just fine with a lot of words not being retained…if I run into them enough, the context will sink in.

(2) Although I originally stressed, and cursed about “gender”, I’m trying to cultivate not caring if I get that wrong. It just seems like an idiotic idea for nouns to have a gender – something which adds zero, or near zero, to the actual meaning of the communication, but a huge amount to the complexity.

(3) I’m thinking the ability for the display engine to recognize tenses, for color emphasis, might be relatively efficient (algorithmic). And, if so, I would love it (actually, I would love it in any case, but would expect it to be too hard to implement).

I give my eye teeth for something to help me more quickly recognize TENSES, in general.

(4) FWIW, I previously suggested italics when I am editing things. (thinking it shouldn’t mess with the ability of the underlying software to process the words efficiently).

Adding color – even if only after EDITING – would, similarly, help me … even if it weren’t automatic (but even better if it were).

(5) The concern about everything being colored could be addressed by allowing various colors (representing various aspects of tense or grammar) to be selectively turned off…or on. What I think I would do with this feature, for example, is to turn on color for “que” (since, in Spanish, the “QUE” is not optional, when you run into a natural need for “THEN” (cause/effect). And, in particular, the “QUE” is a critical trigger-hint that a SUBJUNCTIVE tense follows. I keep not seeing and thinking in terms of QUE, Color for other tenses troubles would just be awesome – and I expect I would use them selectively so things I am weakest in really grabbed my attention.


If you are going to consider any of this stuff, Steve, I suggest you implement it (a) first in the flashcards as something which can be edited in. (b) Then, maybe, as something which can be maybe supported with a simple (efficient) GREP implementation – again in the individual’s flashcards. (c) I’d love it to be automatic, while reading, but fear the coding complexity and processing overhead. Try it first in manual mode in the flash card editing.

P.S. Great community, here! Anna_Vernerova : I find you and your comments especially interesting! If you don’t mind my asking, what is your native language and language interest? I’m guessing you are a polyglot?

1 Like