GoldList Method


The GoldList Method Explained

Here is a new landing page for the GoldList Method.

If you are looking for the original explanation now superseded, please find this here

The new, revised explanation of the GoldList Method is in parts, and these will be published over the coming days as it is all nearly ready (as at 15th June 2018).

Here are the parts

1. The Problem
2. The Basic Concepts
3. How to go about starting a GoldList project – Material Selection and Plan
4. The GoldList Book – Headlist
5. The GoldList Book – Distillations
6. Bronze, Silver, Gold
7. Where to go from here?

Part 7 is not yet ready as at today’s date (29th September 2018)

Anyone wishing to share in the funding of the coming GoldList Method Book and other coming resources can now do so here.

  1. Rick Bradford

    I’ve been tracking my latest Goldlist language project using a small database I maintain. Project stats to date – project duration 233 days, words headlisted 4200, words fully learned 3769.

    That works out to an average of around 16 words learned and fully memorised per day, which is considerably more than I expected. Or is that in line with what other people experience?

    Like

    • In terms of Goldlists where a line is a word, ths could be true.

      For some languages you have more than one line to a word in the Headlist. Lines in the grmmar part can also be spent on grammar explanations and paradigms, and model sentences, as well as phrases, proverbs and collocations all of which are very valid uses and I am these days emphasisng that they are rather an essential use of lines.

      The strongest critics of the GLM are those who say you shouldn’t be doing just vocab on the basis of one word to a line, but that was only ever a shorthand way of talking about it – I have Goldlisted full small dictionaries and to good effect, but it probably isn’t a very optimal way to do a language project. On the other hand, some languages don’t have much better materials than a small dictionary, and for that I really con’t imagine a better way to work the dictionary into the memory than by using GLM. Those who use little paper flashcards are usually overwhelmed by the fiddliness of it all by the time they get past their first thousand or so, and a gusty day while working outside is really not their friend… Using Anki and Supermemo and other apps all the time is not what your backside would like the best from you, so if you would like to be the best friend of your own gluteus maximus, the thing to do is not be sitting in the same place all the time, and you can really take the GLM for a walk to the park and get some post-lockdown vitamin D.

      Having made that point, let me tell you how many lines I did in 2020: I wrote 34,286 lines so that averages out at nearly 100 lines a day. Over the last few years in total it might be closer to 60 lines a day.

      Usually I am able to follow the model so that 3 lines done usually equate to one line learned. This is what you can measure across a longer-term project rather than trying to make such a check or measure every five minutes, which is counter-productive. Therefore I personally would say that I have a run rate of 20 lines of memorising per day, and that’s not only languages, these days it is half non-language. I probably learn the equivalent of 8-10 new words in a foreign language every day.

      Like

    • Frank Dittrich

      I’ll chime in with my own experience using the GoldList method.

      First a summary (I’ll probably add more details about how I use the gold list method in a follow-up, may be a few days later)

      I learned Russian in school in the former GDR from 1973 to 1983. After finishing school, I never had to use Russian, so I forgot very much of what I had learned, but of course not all of it. In 2018, I decided I want to again being able to fluently speak Russian (minimum C1 level, preferably C2, but knowledge of the language is more important for me than passing exams), however, for quite some time I didn’t find an approach that suited me well enough.
      In 2019, I read about the Gold List method, and was convinced to give it a try.
      My first head list entry is from November 2019, but from January to May 2020 I was busy with other things. Only in June 2020 I started to seriously use the Gold list Method. Fortunately, I can spend several hours a day on language learning, so from June 2020 on, my average is about 120 new entries (headlist and distillations combined)

      Stats for May 31, 2021:
      Bronze
      HL 19525
      D1 12920
      D2 7588
      D3 3865
      Silver
      HL 2025
      D1 1282
      D2 244
      D3 87
      Gold
      HL 25

      Usually, I wait about two months before distilling entries from a headlist or from a previous distillation, because this way I somehow feel more convinced about really having remembered the information I skip during distillation.
      (Due to the 5 months long delay in the first half of 2020, I did some distillations 7 or 8 months after the headlist or the previous distillation.)
      These numbers therefore differ from what people who wait just two weeks will see.

      I collected some more information just for this post:

      Meanwhile, my Gold HL has 50 entries, distilled from 66 Silver D3 entries, 94 Silver D2 entries, 120 Silver D1 entries, 180 Silver HL entries, 225 Bronze D3 entries, 385 Bronze D2 entries, 570 Bronze D1 entries, 869 Bronze HL entries.
      Overall, this distillation ratio is as close to 0.7 as you can get, but I didn’t do that on purpose;)

      I’ll not look up the numbers for each and every iteration recursively, but here are some more numbers:

      Gold HL: 50 entries, distilled from 66 Silver D3 entries, ratio 0.756
      Silver D3: 87 entries distilled from 124 D2 entries, ratio 0.702
      Silver D2: 244 entries distilled from 313 D1 entries, ratio 0.780 (I just got through a longer list of irregular verbs from the grammar book, and decided to keep almost all of them for one more distillation)
      Silver D1: 1282 entries distilled from 1800 HL entries, ratio 0.712
      Silver HL: 2025 entries distilled from 2773 Bronze D3 entries, ratio 0.730
      Bronze D3: 3865 entries distilled from 5996 D2 entries, ratio 0.645
      Bronze D2: 7837 entries distilled from 10873 D1 entries, ratio 0.721
      Bronze D1: 13041 entries distilled from 18150 HL entries, ratio 0.719

      The notebooks I use have 251 pages, so I get 3125 headlist entries into each one of them.

      The 2nd Bronze book (headlist entries 3126-6250 contains just grammar, the others vocabulary and phrases.

      Frank Dittrich

      Like

      • That’s excellent feedback, many thanks.

        Like

      • Frank Dittrich

        > Bronze D3: 3865 entries distilled from 5996 D2 entries, ratio 0.645

        This line in my previous post looked suspicious. That’s why I double-checked the numbers I had in my books.
        (No need to double-check the headlist numbers, since I usually write down the first and last number of each page in advance, and I would have noticed any difference to the expected number when reaching the end of a book, if not earlier.)
        Indeed, I found 3 errors I made a over the last few months in my D2 numbers so that I reported incorrect numbers for the D2 list.

        Correct numbers:
        May 31, 2021:
        D2: 7079 instead of 7588
        The total on May 31 was 47052.
        The daily average for the last 12 months up to May 31, 2021 was 122 lines, or 83 lines if I include all data from November 11, 2019 to May 31, 2021 (568 days).

        The incorrect D2 numbers affected two lines:

        These are the correct numbers
        Bronze D3: 3865 entries distilled from 5495 D2 entries, ratio 0.703
        Bronze D2: 7336 entries distilled from 10873 D1 entries, ratio 0.675
        instead of
        Bronze D3: 3865 entries distilled from 5996 D2 entries, ratio 0.645
        Bronze D2: 7837 entries distilled from 10873 D1 entries, ratio 0.721

        And here

        > Silver D2: 244 entries distilled from 313 D1 entries, ratio 0.780 (I just got through a longer list of irregular verbs from the grammar book, and decided to keep almost all of them for one more distillation)

        the numbers were correct, but my explanation was wrong. In the D1 distillation I am in the middle of the grammar section, not in the D2 distillation.

        So, here are the corrected numbers, followed by some additional information:

        Stats for May 31, 2021:
        Bronze
        HL 19525
        D1 12920
        D2 7079
        D3 3865
        Silver
        HL 2025
        D1 1282
        D2 244
        D3 87
        Gold
        HL 25

        Unchanged from previous post:

        Meanwhile, my Gold HL has 50 entries, distilled from 66 Silver D3 entries, 94 Silver D2 entries, 120 Silver D1 entries, 180 Silver HL entries, 225 Bronze D3 entries, 385 Bronze D2 entries, 570 Bronze D1 entries, 869 Bronze HL entries.

        Stats for June 5, 2021:

        Gold HL: 50 entries, distilled from 66 Silver D3 entries, ratio 0.756
        Silver D3: 87 entries distilled from 124 D2 entries, ratio 0.702
        Silver D2: 244 entries distilled from 313 D1 entries, ratio 0.780
        Silver D1: 1282 entries distilled from 1800 HL entries, ratio 0.712
        Silver HL: 2025 entries distilled from 2773 Bronze D3 entries, ratio 0.730
        Bronze D3: 3865 entries distilled from 5495 D2 entries, ratio 0.703
        Bronze D2: 7336 entries distilled from 10873 D1 entries, ratio 0.675
        Bronze D1: 13041 entries distilled from 18150 HL entries, ratio 0.719

        While I am at it, some more numbers for the number fetishists.

        Here are the stats per notebook (i.e., for 3125 headlist lines), for those distillations that are completed throughout a notebook.

        Bronze 1:
        D1: 1975 / 3125 = 0.632
        D2: 1251 / 1975 = 0.633
        D3: 826 / 1251 = 0.660
        (826 / 3125 ) ^ (1 / 3) = 0.642

        Bronze 2:
        (Here I often extended the headlist to more than 25 lines per page, even up to 34 lines per page, in order to not split a topic across pages, and because I wanted to cram all the contents of my grammar textbook I needed to “re-learn” into a single notebook. Nevertheless, I just counted 25 lines per headlist page, so that sometimes the D1 distillation for such a page had more than 25 lines.)
        D1: (4277-1975) / 3125 = 0.737
        D2: (2686-1251) / (4277-1975) = 0.628
        D3: (1806-826) / (2686-1251) = 0.672
        (980 / 3125) ^ (1 / 3) = 0,680

        Bronze 3:
        D1: (6291-4277) / 3125 = 0.644
        D2: (3987-2686) / (6291-4277) = 0.646
        D3: (2717-1806) / (3987-2686) = 0.700
        (911 / 3125) ^ (1 / 3) = 0.663

        Bronze 4:
        D1: (8363-6291) / 3125 = 0.663
        D2: (5482-3987) / (8363-6291) = 0.722
        D3: (3857-2717) / (5482-3987) = 0.763
        (1140 / 3125) ^ (1 / 3) = 0.715

        Bronze 5:
        D1: (10385-8363) / 3125 = 0.647
        D2: (6997-5482) / (10385-8363) = 0.749
        D3 not yet completed
        (1515 / 3125) ^ (1 / 2) = 0.696

        Bronze 6, Bronze 7, Silver 1 and Gold 1: work in progress

        And finally, some dates:

        Bronze 1:
        HL: 11.11.2019 – 18.06.2020
        D1: 07.12.2019 – 12.08.2020
        D2: 06.06.2020 – 28.10.2020
        D3: 12.07.2020 – 19.12.2020

        Bronze 2:
        HL: 19.06.2020 – 08.07.2020
        D1: 13.08.2020 – 21.09.2020
        D2: 28.10.2020 – 14.11.2020
        D3: 19.12.2020 – 12.01.2021

        Bronze 3:
        HL: 08.07.2020 – 06.08.2020
        D1: 21.09.2020 – 24.10.2020
        D2: 14.11.2020 – 17.12.2020
        D3: 12.01.2021 – 09.02.2021

        Bronze 4:
        HL: 31.10.2020 – 13.12.2020
        D1: 23.12.2020 – 31.01.2021
        D2: 16.02.2021 – 30.03.2021
        D3: 15.04.2021 – 30.05.2021

        Bronze 5:
        HL: 13.12.2020 – 01.02.2021
        D1: 31.01.2021 – 31.03.2021
        D2: 30.03.2021 – 31.05.2021
        D3: 30.05.2021 –

        Bronze 6:
        HL: 01.02.2021 – 26.04.2021
        D1: 31.03.2021 –
        D2: 31.05.2021 –

        Bronze 7:
        HL: 27.04.2021 –

        Silver 1:
        HL: 19.09.2020 –
        D1: 29.10.2020 –
        D2: 06.01.2021 –
        D3: 02.03.2021 –

        Gold 1:
        HL: 01.05.2021 –

        In case I still left out some numbers which might be interesting, feel free to ask.

        Otherwise, I’ll later add some more information about what sources I used, why I usually need more than 20 minutes per 25 headlist entries, etc.
        (I still take 10-minute break after 20 minutes of work, and some longer breaks after several iterations.)

        Like

        • Frank Dittrich

          Once more replying to myself:

          > Bronze 2:
          > (Here I often extended the headlist to more than 25 lines per page, even up to 34 lines per page, in order to not split a topic across pages, and because I wanted to cram all the contents of my grammar textbook I needed to “re-learn” into a single notebook. Nevertheless, I just counted 25 lines per headlist page, so that sometimes the D1 distillation for such a page had more than 25 lines.)
          > D1: (4277-1975) / 3125 = 0.737

          Now I counted the real number of headlist lines in Bronze 2 (adding the extra lines, subtracting lines when I occasionally not used all the 25 lines).
          D1: (4277-1975) / (3125+273) = 0.677

          The overall D1 ratio also improves a bit:
          13041 / (18150+273) = 0.708

          Like

  2. Frank Dittrich

    Thank you very much for your GoldList method.
    It is extremely unfortunate that you receive so little financial support for your work.
    I’d really like to fund the work on your book, but I don’t want to use PayPal.
    Can you perhaps email me a bank account to which I can transfer an amount in EUR? The transferred EUR amount should normally be credited in the respective currency, e.g. in PLN. (A minor fee might be deducted.)

    Liked by 1 person

  3. When will part 7 be finished sir?

    Liked by 1 person

  4. In your examples you show that you are working through an ASSIMIL course doing GLM, but I am curious the exact order of what is being done – are you using the GML as you work through the course? Or are you working the course WITH the GML?

    To specify, I am currently working on the book “Ivrit Shalov” for Modern Hebrew. I am working through the book in my normal way of taking notes and writing all over it (which works for me), but I want to bring the GLM to bear on the vocabulary. Would it be proper to take the word bank from the textbook and work through it using the GLM while also studying through the course as normal, or would the reading passages, exercises, and note taking count enough as an active process and disrupt the long term retention?

    AS I UNDERSTAND IT, the GLM is performed as a physical action just as much as a mental one, and as a separate thing, so it seems that the actual period of sitting off in a quiet place to work through head-lists and distillations is it’s own thing, and the act of studying through a textbook and cursing at it is another thing, and that they can be performed at the same time, as they are two different “projects”.

    Like

    • All I am doing and all that I find it necessary to do with the Assimil Course is to GoldList it. Other than that, there is nothing to do but the exercises, and all they servie to do is help memorise but already by doing GoldListing you are doing the necessary memorisation in a maximally efficient way.

      Now a couple of provisos to that. 1) I already have from Pimsleur a reasonable idea of the sound of the language and how to read it. However it may be worthwhile for the sake of hearing more voices and probably mor pronunciations as well as being certain about the pronunciation of words that are not in Pimsleur, to buy the audio parts and separately let that play while driving, etc. That enables the correcting of any misconceptions. I was, for instance, stressing the wrong syllable in “pergi”.

      Assimil with the audio costs three times the price of Asssimil without the audio, and I don’t think this ratio is going to stand up in a market where Routledge Colloquial is dominant and giving the audio away free on line. They probably will need to rethink that, especially for the English language ones, whereas as you know I am using the Indonesian for Germans.

      2) the Second proviso is that the exercises in Assimil, at least the book I have, do introduce other words which you need to look up in the dictionaries at the back, and in the sentences for translation into English they also teach some useful collocations with the words from the lesson, and therefore it is certainly worth looking over the exercises for anything worth goldlisting in them also and I have found that about a third of the sentences are actually also worth incorporating in the GoldList.

      But I don’t regard GoldListing and “doing the Assimil Course” as two separate things. When it is GoldListed, the course is done. Everything needed from it sholdhave been transferred to the HeadList. This is one of the rules of the Headlist.

      Thanks for an excellent question, by the way.

      Like

  5. Hi! Once again I’m back to shoot you some questions, David. I really appreciate your answers.

    1) After Heisig RTK1, I am headlisting Japanese vocabulary. The 6300 words taught in WaniKani. I don’t know how sane it is to only start distilling after I have finished, but I want to have every word headlisted by the end of August. Then I’ll have 10 days when have to dedicate all my time to a separate exam, and by half of September I’ll start to distille. That’s my plan, and I’ll be taking advantage on the main advantage I find here against Anki: GLM does not snowball, I don’t have to work on X quantity of it every day, unless I want to be buried. The 14days to 2 months period to reach to long term memory is my salvation. Also, I want to mention I am not doing clear separate columns (like dividing the page with a vertical line), as I write:
    kanji word ——- kana word ——– english translation and legend (adjective, noun ,etc)
    Is this a problem?

    2) When distilling is it better to read the translation and think of the kanji, or read the kanji and think of the meaning? My goal is reading proficiency mainly. I ask because I’ve seen both versions.

    3) Can GLM be applied to the study of history, mythology, philosophy or art?

    For example, I have this chunk on the Punic Wars.
    – First Punic War → Rome wanted Sicily, controlled by the Carthaginians.
    – Second Punic war → because Carthage got angry after Rome won
    **219 BCE, Hannibal attacked a Roman town, then led an army across Spain, and then crossed the freaking Alps with elephants.
    **Hannibal and his elephant army almost won, but finally the Romans got Spain.
    ***People in Spain were NOT Romans at the moment, and that’s the reason why by 201 BCE Rome was definitely an empire.
    – Third Punic War → a formality, Rome found some excuse to attack Carthage and then wipe it out of the map

    I am thinking that unless I am talking bullet list information like the order of Chinese dynasties, the best approach is to formulate questions on the subject. Example: What was the first punic war? Although I don’t know how much information I could add inside only 1 question: what it was, when it happened (year), why, result… Neither do I know whats the best way to write the answer. A full paragraph with details? Thats not very effective volume wise. Some schematics words and ideas, or a tiny diagram?. I don’t know, but writing one question(line) for each detail would be hell.

    My source of information would be CrashCourse, amazing condensed courses on the most interestings sciences and arts. They do Youtube vids, and there are transcripts of those in a sister web. I trust the quality of the material, and the quantity is just good enough to get a general grasp of everything, and be able to dive deep alone next.

    I think it may be easier on mythology, talking pantheons, names, associations to what powers and so on. Maybe entering history territory when talking about the myths themselves. Essentially I would need to remember stories.

    With philosophy, wow, how do you GLM abstraction? This will need the most work before writing the headlist, processing the information, and condensing it.

    On art, I don’t think it’s attainable in paper, or maybe it is? Dividing the Headlist space in maybe 8 chunks, one per picture, and you try to sketch. Not striving for perfection, but only a rough sketch, enough to differentiate shapes? Using some colors as well? I think that could work to remember the picture itself, but how do we work around the Name and the Author, and the Date? Once again, too much details.It may be easier to draw with a tablet on a computer.

    Hoping to hear from you back! Cheers!

    Like

    • Ping, in case message went unnoticed. Although for 1) be it is sane or not, I am doing it till the end as I’m already halfway through, and for 2) the answer is taking into account the huge amount of homophones in Japanese, and the flexibility of translation, to read the kanji first. Still doubts with 3).

      Like

      • I doubt I can give you a full answer this week other than to say that’s a very big workload for Japanese. I would be inclined to do two lines per word anyway in H and include sentences if given. The advantage of using kana instead of Roumaji for the readings is twofold, first you can use the convention of katakana for onyomi and hiragana for kunyomi, although if you are talking about specific words rather than the kanji itself then the yomi is fixed, be it a kunyomi, an onyomi or an ateji.

        You may have the advantage of learning the compounds which are mainly onyomi with the verb forms which tend to be kunyomi (other than in the obvious case of noun plus suru verbs).

        In the main I would say, and this refers to the third point also, that we talk about GoldListing subjects whereas I prefer to think in terms of projects, with a project being based on one or more sets of materials. You identified your Japanese materials and I don’t know them. Clearly I know Heisig, but not the other one. I would have gone Heisig one, two then three personally, but then I don’t have an exam schedule. Also I am not happy with my earlier Heisig GL attempt and want to start it over, with multiple lines per entry in H, so as to really get the stories in.

        If you are talking about history, philosophy, even to a degree art it all depends on the material how you use GLM, and indeed if it can be used.

        I am pretty sure that most history books and a lot of philosophy textbooks can be GLMd. I would say that Russell’s History of Western Philosophy can be GLM’d but you know how I would go about it? The whole process described in these pages would be brought to bear. Audible has the book available for one credit, nicely read, so first while walking or driving I would listen to it. That in itself would take me one to two months at my usual run rate. Then having listened to another audiobook I would go back and listen to that one again. I would have had two listens to the whole material with proper >2wk spacing prior to sitting with the text and a GLM book. I would then have a pretty good idea of what needed to be Goldlisted as a lot of it I would have remembered and certainly key things I knew I hadn’t remembered could go into the GLM and then be processed as normal that way.

        Like

  6. Muss ich ab D2 auch mindestens 20 Minuten warten, bis ich zum nächsten Dx gehen kann? Das kostet ja viel Zeit: Wenn ich z.B. 3 D2 und 2 D3 mache …

    Like

    • Sehr gute Frage. Die Idee ist, genuegende Pausen zu machen, und nicht unbedingt bei jeder Seitedrehung wenn wir nur ein Paar Linien auf der gegebenen Seite gehabt haben.

      Ich muss das eigentlich zum Text noch hinzufuegen.

      Bei D1 koennte man eine Pause am Ende von jee Seite machen oder zwar anderthalb, aber bei anderthalb ist das nicht so elegant und einfach zu erinnern, deshalb vielleicht nach jeder Seite, aber bei D2 kann mann ruhig dei jeder zweiten und bei D3 jeder dritten Seite pausen.

      Eine ganze 20 Minutenpause ist zu lang. Normalerweise arbeiten wir 20 Minuten und pausen 10.

      Hoffentlich habe ich ein bisschen geholfen.

      Like

  7. Hello, Uncle Davy! Thanks for developing and sharing this method. I find Anki tedious, and thus I’ll be trying this instead for the coming months. Is part 6 finished? It ends somewhat abruptly… and what about part 7? It’s not linked on here.

    Like

    • Part 7 will be written soon, part 6 finished nicely. Also coming up are more of the live examples and the implementation of some proofing edits I’ve received.

      Thanks for your positive feedback and much success with the method. I hope you find a lot of pleasure in watching the content of your GoldList books grow.

      Like

    • Yes I haven’t finished it and I will try to do that next week. Just gonna do the tables for the last to 70 DCs and then I will get on that.

      Like

  8. Pardon me if I’ve missed it, but I would be pleased to see the link to this new explanation headpage more widely popularized! The only automated notifications for new pages not on the homepage “stream” are sent through WP accounts, and I don’t log into mine all that often. Nevermind those without WP accounts.

    As far as comments on the content, I’ll have to read through it first! At any rate, congratulations on how the system is taking off and helping you to the necessary information and feedback to streamline and clarify.

    If and when you publish a book, I do hope and urge that it be made to a higher level of production and editing “values” than a certain PDF/paperback book which I’ve seen popularized in these circles. I value the historical and personal accounts of that project, but it would be unfortunate if the worth of your system were not adequately presented.

    Like

    • You are probably referring to Uncle Claude’s Polyglot Project book. What you say about the level of proofreading there is true, although Jimmy Mello saved the day by attempting to do a better one which was actually even worse and made Uncle Claude Cartaginese win that particular Punic war, if you’ll pardon my punics around his name.

      I don’t want to rub salt into the wound of a fabulous chap (I mean Claude of course, although Jimmy Mello is also OK in small doses) but just between you and me, I imagined from the hype that he was going to do or get done a professional level set of proofreadings around these articles, instead of which he pretty much printed what folk wrote. Had I known this, I think, time permitting, I would have subjected my own text to a bit more rigour, and certainly what you have in these sections now about the GLM will need going over a few more times and supplementing with lots of diagrams and inlays and icons and lay-out tricks before ink goes to paper in any sizeable amounts.

      Like

  1. Pingback: Charting my Language Learning Journey | Aidan Sweeney

  2. Pingback: Getting Back on the Wagon - Expressing Meaning

  3. Pingback: Victor Berrjod’s excellent rebuttal of Bartosz Czekała’s hatchet piece against the GoldList Method | Huliganov.TV

  4. Pingback: The Journey Begins – Mukhya Bhasha

Your thoughts? Go ahead, pull no punches!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: