Problems in typing nikkud in Biblical Hebrew

Hi. I’m inputting vocabulary for a Biblical Hebrew course I am taking:

I tested it out and found my first challenge - there is a precise order that memrise uses to accept a word as spelled correctly in a typing test. The situation is when there is a word that features a consonant with a dagesh and a vowel. The only way memrise will accept a spelling as correct is to follow this order: consonant, vowel, dagesh.

This seems unintuitive, since the dagesh is a feature of the consonant and not the vowel. Is there a way for memrise to accept a different order of typing? I like that it requires me to have the correct nikkud, but it would be preferable if it could accept this order as correct as well: consonant, dagesh, vowel.

Thanks for any help you can give.

1 Like

Can you type out an example of what Memrise accepts and doesn’t accept? There are a lot of people on the forum who might be able to help but probably aren’t familiar with Biblical Hebrew :slight_smile:

Neoncube, thanks for your offer to help. I’ve been busy (learning Hebrew in real life and classes).

I should clarify that I’m using the Android app, not using my phone’s keyboard to type.

A challenging word is the number sixty. In Biblical Hebrew, it is:


(For those helping who don’t know Hebrew, here are relevant names of letters and diacritics:
ש a consonant called shin
י a consonant called yod
ם a consonant called mem sofit
the dot above and on the right side of the shin specifies that its sound is ‘sh’ and not ‘s’ (consonant feature)
the dot under the shin is the vowel hiriq (e as in keen) (vowel feature)
the dot in the middle of the shin is a dagesh which shows a historical lengthening of the consonant (consonant feature)

The Hebrew nikkud I’m talking about are the little dots and diacritics above, below and in the middle of the consonants. Yes, it takes 9 keystrokes to type this word. That is just the beginning.

There is more than one way to type all the consonants and niqqud (diacritics). Some of the niqqud are features of the consonant, and others indicate vowel sounds (as shown above). It is possible to type as follows:

  1. shin, ‘dot above the right of shin’, dagesh, hiriq
  2. shin, dagesh, hiriq, ‘dot above the right of shin’
  3. shin, hiriq, ‘dot above the right of shin’, dagesh
  4. shin, dagesh, ‘dot above the right of shin’, hiriq
  5. shin, hiriq, dagesh, ‘dot above the right of shin’

All of these ways will produce the same final look of the second shin in the word sixty above. However, only one order is accepted by memrise as ‘correctly spelled’ in a typing test (I don’t know if it is restricted only to the Android app or if it is also this restricted in iOS or the web version). And unfortunately, it is number 5 above which is acceptable, and from the perspective of learning the language, it is not intuitive, and even counter-intuitive. You have to memorize a specific order of diacritics when typing a word. If I were looking for a course and found out that I had spelled a word with all the appropriate diacritics, but it was wrong only because I didn’t type them in the correct order, I might not want to learn from this course. But it is very useful to know all the diacritics.

I would wish for this: either make any order acceptable, as long as it produces the correct visual appearance of the word, or (and only if necessary), make option 1 be the correct order for typing the words.

Knowing a little bit of the complexity in rendering other scripts, I can imagine that it is not a simple task to do, but for Memrise to handle Biblical Hebrew well, it would be best if it could handle mark as correct any of the 5 orders of typing that I have shown above. I can imagine that it may also be governed by how it was originally typed. I did not originally type the Hebrew words. They were copied and pasted from a document produced by someone else in MS Word.

Hoping there is a simple solution to a complex problem…


That sounds as if Unicode normalisation should be carried out, normalising both the word in the course and the typed-in word to the same form (e.g. NFKD) before comparing them.

That should help both with precomposed-versus-not but also with composed-in-a-different-order problems.

Is this for Memrise techies to do, or for me the creator who doesn’t know too much about techie stuff?

Interesting :slight_smile: Thanks for the explanation; I understand much better now :slight_smile:

When you have time, could you please post a screenshot of what this looks like on the phone when you’re typing the words using the Android app? Does the app provide little blue blocks/buttons with symbols on them that you can tap to form the word, or are you composing the word in a different way?

If you can type Hebrew on your phone, one solution might be to use your phone’s keyboard to type the words, instead of the Android app.

No time to post a screenshot just now, but yes, I’m using the Android app to type (ie. with the blue blocks/buttons). On my phone keyboard, it is a modern Hebrew keyboard and does not have the nikkud, so I can’t type the nikkud with my phone keyboard.

It’s something the Memrise techies would need to implement.

As I have continued to add more data to my course, I have found that some new words I have added have a different accepted order from number 5 that I described above. Therefore, it seems like the source of the specific order is how it was originally typed (not by me - I’ve been copying and pasting).

So, Memrise techies, can you figure out how to deal with this so it can accept any of the 5 orders (or some of the 5 orders) specified above? If not, then I’ll have to make this a multiple choice only course, which is not as good, since I’m not forced to remember exactly which dots and diacritics to use in which places of words, which do matter.

Thanks, :slight_smile: