In languages such as Dutch and German there are separable verbs like for instance afwijzen where the verb is wijzen and a preposition, af, is prefixed. However, if this verb is the working verb, then the verb is conjugated and the preposition is put at the end of the sentence normally. It would be nice if there was a feature to recognize these separable verbs for us and give us their meaning. Currently, all you can get is the meaning of the verb itself, but with a preposition the meaning can completely change.
3 Likes
Similarly with English and it’s phrasal verbs which are often separated, especially informal speech.
For instance, here’s a twenty-year-old paper on using NLP to identify phrasal verbs in English text.
I’d think that with recent progress in generative AI, handling of separable and phrasal verbs will become necessary for all input-oriented language acquisition tools.
I bring up English, because I wonder what percentage of overall LingQ language learnings are studying a language with such verb constructions where a “word” isn’t trivially parsable.
I believe the parsing challenges are also related to the problems that LingQ has in Japanese, such as with multiple readings.
It’s possible that in the future LingQ would benefit from continued steps into better parsing and not just tactically fixing spot issues but in ways where the underlying hidden tech does better with words, morphemes, lexemes, lemmas, roots, radicals, and readings.
2 Likes
With English is not difficult at all, you can easily create lingqs with 2 or 3 words together and learn them that way. I don’t find any difficulty to use LingQ with phrasal verbs.
You create: get up, wake up… and so on, like a single word.
With German is different, because you can have a sentence that is very long, the verb at the beginning and the particle at the end. Unfortunately, LingQ doesn’t allow to select more than 9 words. This create the problem that you can’t create lingqs for those separable verbs.
German: get bla bla bla bla bla bla bla bla bla bla bla bla bla bla up
@dominic_deboer there is no solution at the moment for those languages, and the team is aware about it. The only think you can do is to keep learning, and step by step you will become accustomed to most of them.
Another thing I do, is to use the notes. For example, you can use the particle “vor”. Inside the note of that particle, you collect the meaning of different verbs: vorstellen, vorgehen, and so on. Step by step, when you come across those verbs, you add them to your list.
When you find the combination that you don’t remember anymore, you click on the particle and check on your notes.
It’s a workaround but it’s better than nothing.
2 Likes
In English, informally we often say such as:
- Can you pick some milk up on the way home?
- Be sure to turn the lights off before going to bed.
- I need to clean the garage out this weekend.
- Take your disgusting shoes off before coming inside.
- LingQ needs to figure phrasal verbs out even in these situations.
While German characteristically places the verb’s core element in sentence’s second position and the verb’s preposition at the end, I’d suggest that English (despite centuries of prescriptivists’ opinions) does have “timidly separable verbs” where direct objects can get plunked in the middle, in informal usage.
Perhaps why I bring this up is because parsing and word identification is semantic and language specific. LingQ’s programmers seemed to originally assumed universality for which they’ve had to code for discovered exceptions.
6 Likes
Every so often this question comes up. Maybe in the future as mentioned by gmeyer AI might be able to recognize this in Lingq and bring up the associated meaning.
In the meantime…this is very easy to handle.
First of all, most of the time verbs for German (and I assume Dutch) are tagged. If you click on the word, you will see a list of tags…on verbs where there may be a prefix, the tags for all the various prefixes will be shown.
For example, if I click on the German word “gehe” (first person present “to go”). It will list certain tags…“verb”, “indicative”, “gehen”…but it will also have tags for all the prefix forms of gehen…“abgehen”, “angehen”, etc. If you click on these it’ll bring you to “reverso conjuation dictionary” where it has the meaning for this prefixed form, along with the conjugations.
You may also find that others have created entries for the various meanings with prefix…or you can enter your own.
For example for “gehe” above you might create a Lingq definition: “to go; (ab)gehen - to depart; (an)gehen - to approach; etc. etc.”
A little bit of work, but really all that bad. In most cases, people have already done this for many words and their conjugated forms.
3 Likes
yeah, but I can easily create a lingq with pick some milk up, or turn the lights off. And eventually review all those on the internal vocabulary, even if not ideal, it can be done. Not with German. I usually find myself with sentences with more than 9 words and I can’t do anything about it. I gave up long time ago.
This is half true. I use this feature as well, especially when I’m not sure about the infinitive form of some conjugation. Unfortunately, I would say that A LOT of verbs don’t have tags at all.
The bad thing on this is that when I add those tags, they don’t become public for everyone. I have added a lot of tags for many German verbs, and it would be a lot more convenient for other people if they were added to the database.
Believe me, I often encounter many verb conjugations that don’t have tags.
Plus, it’s true that if you click on reverso, it sends you to the reverso conjugations, but that’s not really ideal if you are not an English speaker. You need more clicks to change language, and Reverso conjugations is not really the best dictionary option.
IMHO.
2 Likes
i think it is possible now with rooster’s extension . It has increased to 15 from 9 words if my memory serves me right. However you need to download and buy the extension separately.
2 Likes
This is why I put phrases into Anki demonstrating the separable verb usage e.g. ich ziehe mich schnell an. Yes sentences can get long, but that’s fine, it just means I’m testing more items in the same phrase i.e. verb, vocabulary and grammar. Perhaps there are more effective solutions but it works. I’m sure you know this, but the brain finds it easier to learn information in context as opposed to a simple verb + meaning on a card. I used to struggle dreadfully with German vocabulary because it’s so damned unlike English. Now with context it’s sinking in.
2 Likes
yeah, exactly, which is something I don’t really like at all. I don’t like to rely on something that could or not continue in the future. But that’s me. It has nothing to do with the extension.
4 Likes
To be honest, I don’t really care much nowadays because I’m not focused on German as main language I study anymore. I don’t have the “urgency” to impose to my brain to learn at all cost some verb that I never use or encounter anyway.
I think that with AI we have more solutions than SRS (which I don’t use anymore).
If I find a separable verb many times and it never sticks in my mind, I ask chatGPT to create 100 sentences with that verb, plus I can ask for creating personalised stories with that verb, put everything inside LingQ again, and drill that lesson many times, with audio included eventually.
Much more fun and engaging that drilling cards.
IMHO
3 Likes