This topic is about the Memrise course “1 - 5000 Arabic Frequency & Audio” based on the book A Frequency Dictionary of Arabic. Core vocabulary for learners (Buckwalter, Parkinson, Routledge 2011).
A quick info on the status as of June 2016
Number of words:
- The vocabulary in the course is almost complete, I estimate there are about 200 words missing (all of them from frequency ranks 2500-5000, so nothing dramatic).
- The are a few dozens of duplicates (probably less than 100).
- Total words in the course = a little more than 5000. That’s because a few hundred lemmas (counted as one in the FDA book) have separate variations (so two or more “words” in the course).
- About 2000 words have an audio file, including all frequency ranks from 1 to 1600.
- Quality of the audio file tends to be quite good, whether they come from native speakers or from TTS.
- About 3/4 of the audio files are complete, in the sense that they include all the relevant forms ( for a verb: past/present/preposition ; for a noun or adjective: singular/plural/sometimes feminine or collective/preposition).
- The FDA book includes thematic lists (like animals, body parts, professions). It would be nice to add these lists as additional levels, unfortunately due to the structure of the course (which is separated over 3 databases) this is not possible for now.
- The FDA book includes an alphabetic index based on the roots, which is a great learning aid. I’m not sure how to leverage that as a learning tools in Memrise ; but anyway the roots are not included in the words databases now (and again the course is split over 3 databases), so nothing can be done there for now.
I’d be interested in ideas on how to merge the 3 databases
- If you want to download the full database to work on it yourself, the best tool I’ve found so far is to import the course into Anki.
- I am using and completing a local copy of the course on my computer, that I could share as a CSV.
Any suggestions about where I could upload the CSV so that other people can also improve it ?
I could also upload the audio files i have.
- I understand those who initiated this course have been in contact with the authors of the book, who are not opposed to have a course based on their work.
- However, I urge people who really use this course to buy the book, it’s a great work and the authors and editors should be rewarded for that. You can try out the first 500 words, but should (morally, not legally) buy it past that.
Hey, thanks so much for this post! I am a formerly active contributor, hoping to become active again someday so following to get updates.
Will think about ways to share the CSV that will not be too invasive of privacy (showing personal e-mails, etc).
Thanks for all your work on the course!!
I tried this course earlier, but got frustrated because the entries are pretty vague. Instead of saying
With, by (not عند ,لدى or مع), I believe it would be more helpful to explain when one word is used as opposed to the other(s).
Other than that, I liked the course.
I agree with you, Holff, on this point. It can be quite frustrating to deal with a multiplicity of synonyms and word meanings. As Holff suggests, I’d propose eliminating entries of the type ‘‘With, by (not عند ,لدى or مع)’’.
Hello Holff, DaithiWalsh,
The entries from this course must not be deleted. This course has a clear, single reference (which is rarely the case!), that’s Buckwalter’s book. So all 5000 words of the book must remain in the course. We can add more stuff (for instance, words beyond the 5000th, or examples that are better than Buckwalter’s), but not remove words.
You’re raising two important questions however.
One is to handle words that have similar meaning, i.e. synonyms and quasi-synonyms. That’s actually an old and difficult question, if you’re interested you can look up language forums (memrise and others) to understand the possible solutions, none of which is perfect. The approach in this course is to mention in the English cue the Arabic synonyms that are not the answer. It’s a sound solution for this course, and we’ll keep it.
The other point is to understand the limit of the course: It’s only a vocabulary builder. If you’re learning Arabic, you must use other sources, to learn grammar, phrases, conversation, read and write on your own, and so on. Understanding when to use basic, versatile words like بـ , عند ,لدى or مع does not belong to this course, but to the other type of material. This kind of words appear in the course because they are among the most frequent, but this course is probably the worst way to learn them. I actually recommend that you start using this course only after you achieve a basic capacity to interact in Arabic, you will cruise through the first 500 to 1000 words and then really learn beyond 1000.
Look at it this way: This course is a fantastic contribution to bring you from a B1 level (“lower intermediate”) to a C1 level (“advanced”) ; but if you’re still elementary (A1 and A2) you’re not making the best use of your time.
I happen to have the list of words that are missing in the memrise course compared to the book.
It’s the following 191 words, referenced below with their frequency rank:
965 1070 1149 1187 1837 2260 2350 2406 2586 2772 2777 2781 2786 2793 2804 2820 2831 2836 2853 2859 2878 2886 2898 2909 2942 2969 2991 3009 3019 3036 3040 3052 3099 3107 3142 3146 3152 3167 3182 3185 3198 3234 3238 3240 3272 3283 3298 3305 3329 3333 3338 3360 3367 3378 3384 3405 3450 3455 3457 3462 3467 3473 3478 3480 3483 3487 3491 3511 3526 3538 3545 3548 3557 3566 3569 3576 3579 3626 3630 3652 3666 3704 3706 3747 3750 3756 3769 3798 3803 3812 3825 3834 3849 3851 3855 3862 3868 3886 3907 3941 3947 3966 3970 3975 4004 4009 4014 4023 4030 4034 4046 4073 4078 4080 4086 4098 4108 4112 4115 4144 4147 4153 4160 4167 4181 4184 4187 4192 4196 4221 4226 4230 4271 4281 4312 4315 4324 4329 4342 4344 4355 4375 4391 4409 4434 4436 4439 4455 4466 4471 4481 4486 4497 4511 4513 4516 4535 4542 4554 4586 4588 4592 4605 4620 4631 4644 4652 4662 4704 4713 4717 4719 4733 4758 4761 4784 4803 4819 4831 4839 4848 4854 4905 4913 4916 4919 4933 4941 4948 4972 4995
Hi again people. Well, I hear what you’re saying, Yag, but I’m unconvinced. I mean, you’re saying (I think) that the course cant be changed - despite the comments by myself and Holff - because: Well, because the course can;t be changed! Because that’s the way the Buckwalter’s book is or was. I’m still not convinced, mind: When I learn a word, typically I learn what it means, instead of learning what it doesn’t mean. That was my point, and Holff’s too I think, about the problem with entries which say ‘’[it doesn’t mean X, Y, or Z].’’ But by the looks of things, we’ll have to agree to disagree on this point.
Let me restate:
First I’m saying the course is built on the 5000 words of Buckwalter’s book, so we keep the 5000 entries in the database. Each learner might prefer to remove some entries for any legitimate reason for him (for instance he thinks those entries/words are vague, or they pertain to a dialect he’s not interested in, or a subject he’s doesn’t need, or whatever). He can and must do that by himself in Memrise with the “Ignore” function.
Second I’m saying the problem of synonyms and quasi-synonyms is a complicated question for such courses. I didn’t like much the current solution when I joined this course, but after investigating I realized it was hard to do better.
If you do have another good solution, approved by an overwhelming majority of the 300 to 500 regular users of the course, I’ll be glad to open access to you so that you can implement it over the entire database. I estimate they are about 1000 words with such synonyms, so it’s only a thousand of entries to modify.
Sheesh. I was making a comment, Yag: it’s not a criticism of you, and it doesn’t mean you’re a bad person. No need for the sarcasm or defensiveness. Okay: I was wrong in my previous comments. Now, that should close the matter. Goodbye.
Interesting how you quote this number, since there are only about 250 people who are ‘regular users’. I also don’t see how going to other courses to learn grammar etc then coming here will help. You cannot learn the others without vocabulary.
I am not sure why you are so defensive, I made a suggestion based on previous language studies, including Mandarin. This course’s ‘strategy’ flies in the face of more accepted and successful methods. I guess you will say Arabic is just that different, but I doubt it.
Another suggestion, in the explanation, make sure you state this is NOT an introductory course but an advanced course. When users see first 5000 words by frequency, it is understood that it is for beginners. Again, something common amongst the other languages and even other Arabic courses.
Yes, 250, 300, 500, that’s the order of magnitude of regular users ; they come and go with some seasonality, and it’s not the same people over time.
That’s one of the issues when editing a course that is used by other people ; changes will affect all of them, and those who spent a couple of years learning something can legitimately complain. For instance, some of us would be particularly upset if some words just disappeared. Memrise tools are not great to make announcements or poll the users either.
It seems we share the same experience: I did not have with Mandarin the same difficulty regarding synonyms ; the native language side of my flashcards would give the general meaning of a character, and if need be, a word using the character. The character, pinyin and occasional Chinese word would be on the back. There was no need to rule out synonyms, and the association with an actual word was an added value.
Unfortunately Arabic is a different beast. Synonyms are often used (sometimes just for rhythm), and the measures complicate things too (e.g. sometimes, but not always, II and IV are exact synonyms). Again, if you figure out a better, generally accepted solution on synonyms, I’ll be happy to oblige.
Thanks for the suggestion, I agree with you and will change the course description.
I’m interested to know more about what you call “accepted and successful methods”. After a bunch of European languages, Mandarin and Arabic, I feel that the learning strategy must sometimes be dramatically different to address the specificities of each language. For Arabic, that would mean leveraging the mechanism of roots and measures - which I haven’t seen in a course yet.
When you write [quote=“DaithiWalsh, post:8, topic:1115”]
You’re saying (I think) that the course cant be changed (…) because: Well, because the course can;t be changed!
after I wrote a 300-word answer to lay out the problem, yes, it does sound like a criticism.
Dear Yag, just wanted to say I am on level A1 and am actively using your course to build my vocabulary. I tried with the book, but it’s just too much information to start off. I must admit, starting that early in my learning of the Arab language, I take your course slow, with a goal of 5 minutes a day, but it has already helped me a great deal in learning the other stuff. Of course, I use a lot of other resources too, but I think anyone expecting to use solely a memrise course and being able to speak the actual language afterwards should reconsider
Anyway, just wanted to say thank you for the course and I love the way how the questions are asked, and how there is not too much information like in the book. If I want that I can always look it up in the original!
I’ll write a couple of posts to discuss the format of some of the fields in the course.
Users: Please let me know your thoughts and comments.
Contributors: I suggest we agree to a common standard, to avoid that one of us modifies inputs that another one had purposely done.
Remember: the core rule is to remain consistent with the FDA book.
However, we should fine-tune the fields to ensure efficient learning. The most obvious mistake is to include the plural of a word in the English cue: yes, it’s there in the FDA book, but it’s too easy to remember the singular if you’re given directly the plural form.
First, the “English” field, that contains the cue.
1.If the word is not MSA, start with the 3-letter abbreviation to identify the relevant dialect, between brackets. Example: (Lev.) or (Dia.) or (Gul.Irq.Sau.)
2. When necessary, identify the Part of Speech : n. or vn. or adj. or interj. For a verb, that should not be necessary since in English all verbs start with "to"
3. Add relevant context in brackets, in English. Include Arabic (for instance, the adequate preposition) only when there is a special reason to do so.
4. Separate different meanings of the same word with a semi-colon ; if the meanings are too different (for instance, if each meaning has a different plural) you can duplicate the word and use one word for each meaning
5. If the Arabic words has synonyms, mention them in the cue in brackets. Example: “intent, purpose, goal (not هدف)”
As a consequence of these rules, here are the "don’t"s:
- don’t write “(Colloquial)” in the cue. See 1/
- don’t include the number of the verb measure like II or V or VIII. If you want to keep it, put it in a “comment” or “plural/imperfective” field. See 2/
- don’t include "pl. -aat ". This should be in the “plural/imperfective” field. Obviously, don’t include the Arabic plural in the English cue: the word would become useless.
- don’t include the imperfective vowel. This should be in the “plural/imperfective” field. Example: v. I (u) to be published, be issued; to emerge, come forth, appear => the "v.I (u)’ part must not be in the English cue
I’m trying to contact the course creator: "ArabicFrequencyGroup"
I am considering some changes in the database that I cannot do as I am only contributor, not creator.
Any ideas how to reach him/her ?
The changes I cannot do myself are about changing columns options “Always Show” (or not) and “Showing after test”. Also about a column which is currently an attribute, while it should be a normal data column.
Message to other contributors:
I’m a bit worried, as I see some words disappearing from the course. Please,
if you are not interested in a word, use the ignore feature of Memrise, but don’t delete it
If you deleted by accident, please pay more attention and find a way to control your changes
Memrise is extremely limited in terms of control features to validate changes or revert them. Please keep the general rule that contributors can ADD stuff, can MOVE AROUND stuff, but should ALMOST NEVER delete stuff. The only good reason I found so far to delete was to remove some of the redundant words that appear more than once - and even then, I check I keep the best version of the word.
Words that had disappeared from the course and that I put back, identified by their rank:
Aug2017 : 870, 1046, 1053, 1063, 1067, 1074, 1345
New message to all contributors:
I see that another contributor has changed back some modifications I did in the course.
The modifications themselves are minor (measure numbers included in the English field).
But I am puzzled about what to do. That contributor has not bothered come here to the forum and explain why he prefers that way.I don’t want to go back and revert his changes without some prior agreement - that’s exactly what we should avoid.
Any suggestion appreciated
How about contacting a staff member? Maybe they can find out who’s the one doing the editing.
Thanks for the idea.
I did contact Lien from Memrise staff, but did not get any answer.
I think I’ll stop working on this course and I will do another by myself.