True to Type?

Photo of keyboard and pen
The pen is mightier than the board

Over on YouTube this morning, a viewer called Fightingnate commented on the  second Goldlist film  with the following very good question:

How do you not think you use muscle memory for typing on a keyboard? how do you think some people can type 100 wpm? Why do you think there are typing lessons and programs? Typing (if you type correctly and efficiently) requires just as much memory as writing.

I knew immediately that the answer to this question was going to go well beyond 500 letters (more than 500 words, for that matter) so I left there a request to look the answer up here and I hope the asker and some others who are interested will read it here, and also commenting at length of desired is easier here, as long as you have a wordpress account, or have something like facebook or one of the other methods for logging in here.

I am ready to admit that there are memory functions associated with typing. It’s a perfectly valid observation that there are lessons and programs to learn typing which certainly require the use of memory, including long-term memory and the long-term memory will certainly be involved in learning to type.

Whether it “requires just as much memory as writing” as you claim I would suggest is an unfounded statement. It may be true or it may not be, I am uncertain that it can even be measured reliably. However what I would say is that even if typing is more demanding on the memory than stylus writing is, that still wouldn’t make it optimal for language learning to the long-term memory.

In a sense you might take as an analogy that using GPS will still engage your memory, but working things out with an old fashioned map may be a more natural way of remembering how to get to a place. Typing numbers into a calculator may help you churn numbers out faster at the cashier’s equivalent of the secretary’s standard of 100 wpm, but whether it really helps you retain the parts of maths that need to be rote learned as well as paper and pencil calculations do, well, I doubt it.

Learning to the long-term memory is, I believe, done best when we are not giving to our brains signals that we are making efforts to learn, and not making our brains feel as if they are working. A more relaxed way of writing is preferable. Maybe for Generation Y-ers and Z-ers you feel more relaxed writing in typing than in stylus writing, but that is a bit unnatural. The way of writing with a stylus developed thousands of years ago, if you include knives and brushes as well as pens and pencils in that class of implements, and it was developed in a sense “naturally” in a form basically dictated by the biological shape of the human hand as well as the workings of the human brain. The keyboard layout on the other hand is an attempt to impose a certain predictated logic onto a flat surface and in a sense we have to use an extra layer of effort and memory to remember where, in two dimensional space, a letter is. In a sense the typist can remember the shape of the word and will find that enduring typoes bear witness to the fact that memory – if at times erroneous memory – is involved in that process. But you are only feeling keys. You are not feeling a word being crafted by your hand against the paper. There is no big difference between the feel of one key and another.

The keyboard restricts the movement of your body to one place, while in stylus writing you are moving your hand forward (even in right to left or other systems that still counts as forward) across the sheet. You can also move your body in relation to the writing more easily. You can hold the book at more different angles and in different positions. You can grasp the pen and book from a standing start or take it with you anywhere far more easily than the computer. Even the tiny computers which we now call telephones bear testimony in their new stylus-imitating input methods that the keyboard is not the most relaxed or efficient way of doing things.

Even when the speed typist sits and types her 100 wpm, or his in order not to be sexist, and shows up a certain advantage keyboard writing can have over stylus writing at high speeds, can the substance of what is being typed be remembered just as well as with the typist or handwriter going at much lower speeds? I venture to suggest not. The long-term memory is a subconscious sampler – that’s a key tenet of the Goldlist method and if it were not so then the whole system wouldn’t work. But what determines the sampling rate? Is it the same rate at high speeds or is it a sample of so many passes per second regardless of the amount of material? We don’t know for sure, but I believe that the way perception works will make it a bit of both. You certainly remember more details of a street when you walk down it than when you drive down it, but the ratio of details remembered to time spent could well be lower on the walk, as at slow pass speeds some items will be sampled more than once.

Pen writing doesn’t fade like a screen when you work on it in the sun. You don’t need electricity and you can carry the book with far less weight on a walk. Writing languages in it which are full of diacritics, or writing in Japanese or other character-based languages will be for more English-speaking learners far easier to do. And most of all it will be personal. Your handwriting is special because it is your personal body language in paper form. For literacy, handwriting something rather than typing it is the equivalent to saying something with your own voice instead of letting off a recording of somone else saying it, and just listening. The printed page may be all your words but your body language has not melded with the language as it does with the handwritten page. You do not become one with it. And that is why the specific memory aid that comes with that melding and crafting of the written words in writing as nice as you can make it and done with a sense of the pleasure derived from such craftsmanship, is not really delivered by the process of typing.

Most of us do a lot of typing, sone of us are even threatened with Repetitive Stress Injury from the amount we do, and also we have fewer and fewer occasions to cultuivate the hand and as a result when we need nice handwriting it evades us. The insistence on having Goldlist Method a handwritten method is not anti-machine – it simply reminds us that there is an alternative to the machine and that the computer is not the only tool when it comes to language learning – even a language learning method that bases on a quite mathematical algorithm.

9 thoughts on “True to Type?

  1. Dear David,

    Again, thanks for your noble work with the Goldlist Method and for sharing it so freely. Whatever my recent and current levels of participation in the 70-day challenges, the method itself has indeed proven itself and become a part of my learning strategies for the long term.

    My reading of this post coincided with some other studies and work I’ve been occupied with lately, and I’ve been able to process it and formulate a response more rapidly than is the case with some of your other posts about the conceptual and theoretical structure of the system.

    It’s clear that you have devoted a great deal of careful thought to your method and its foundations, regardless of whether or not every detail can be asserted from a strictly scientific basis. It is in that spirit that I proceed, given the limitations of my own neurological knowledge. Just the same, rigor and precision can only help us, so I don’t wish to ignore them, either. The concepts of “sampling” and “passes” as used in this article do need to be clarified and defined, but I will use them as I understand them in practical terms.

    You have asserted the long-term memory as a role of the subconscious. I will accept that assumption as given because throwing out all one’s assumptions at the same time tends to leave one stranded in never-never land. Fascinating questions would remain, of course, about the interaction between the conscious and the subconscious in retrieving and applying the knowledge stored in long-term memory. Here I also accept your use of ‘long-term memory’ and ‘short-term memory’ as workable theoretical constructs reflecting the way in which neural networks function, even though the real situation is probably more accurately identified by the terms ‘long-term memories’ and ‘short-term memories’, even then, the neurological differences between the two may well be related more to degree of synaptic interconnectedness and ease of conscious access than to any qualitative difference in the neurons which hold the memory.

    Next comes a disclosure in the interest of openness. You have assumed and asserted (1) (not without evidence) that learning to the long-term memory _cannot_ function simultaneously with efforts to learn to the short-term memory. I don’t know if you mean that as a strict rule or simply a practical description, and I don’t wish to construct a straw man, so I will take care not to misrepresent you. While your assertion describes the practical situation very well, I must admit that I would disagree with it as a strict rule. The subconscious is, well, the subconscious, it’s ‘always on’, and it constantly registers stimuli even during sessions of ‘learning to the short-term memory’; these stimuli may well lie beyond or outside the target material in the majority of such cases. You’ve mentioned (on more than one occasion) (2) listening to music while using the Goldlist, and I would extend the analogy to listening to music while doing any other activity oriented to the long-term memory or not. I would assert, for my part, that when the primary motivation for a study session is some short-term goal, the subconscious cannot be fooled into giving the material a higher priority; the subconscious knows. Why? Because that’s the function of (at least some portion of) the brain, to manage priority levels and thus avoid brain overload, which would lead to disorganized thinking, indecisiveness, bewilderment, exhaustion, malfunction, and even (in extreme cases) loss of sanity. My personal experience confirms that effortful learning can and routinely does get placed in long-term memory (I will for now ignore the question of relative efficiencies measured in terms of volume versus time); my argument would be that not the presence of effortfulness—after all, taking the care to write neatly and attractively in the Goldlist is an effort—but rather the nature of one’s _motivation_ and the effectiveness of concurrent learning strategies is what decides the matter. In the end, then, we reach the same practical conclusion regardless of the reasons given.

    In response to your analogy of driving versus walking and the statement that the ‘ratio of details remembered to time spent could well be lower on the walk’, wouldn’t this concession be consistent with having a higher ratio of items remembered to time spent on keyboard Goldlisting, which would contradict your methodology and the answer given in this article? Or would the difference be a question of practicality and _usable_ efficiency, given that remembered details would be more fragmented for keyboard Goldlisting? After all, couldn’t ‘multiple passes’ or ‘repeated sampling of the same item’ indeed be part of what allows the subconscious to form long-term memories the way it does? Couldn’t it be the case that during the time taken to write a line of Goldlist, the various components of that line will have been held in attention (3) and working and sensory memory long enough to have been ‘sampled’ more than once and perhaps in more than one way by the subconscious? In this view, successful placement in the long-term memory depends on what the subconscious chooses to prioritize, what it chooses to minimize as extraneous information, and which items happen to land immediately and fortuitously in denser neural networks based on the existing knowledge established by one’s personality, inherited memories (assuming they exist), and life experiences.

    I will post this now, incomplete though it remains, and return later to conclude and supply the footnotes.

    1. Hi David,

      Many thanks for this careful and erudite essay about the method.

      I would point out at the outset that GLM hinges on some working assumptions and approximations to reality. If we allow, for instance, that Ebbinghaus’ forgetting curve is exactly correct then Supermemo and Anki which trying to model on it exactly using logarithms, algorithms, and rithmandblues could be said to be a lot more of an accurate reflection of Ebbinghaus than GLM is. In GLM you don’t follow the whle curve. The curve is acknowledged but left to its own devices and we comme back after two weeks and say, ok, per Ebbinghaus, most of what is going to be forgotten will have been forgotten. At the same time, though, we depart from Ebbinghaus in that if you follow Ebbinghaus to the letter in the long run everyone forgets everything, and a theoretically immortal man who had lived for millennia would have virtually no recollection of what he encountered thousands of years ago. But we know from our own experience that there are some things we simply remember all our lives, sometimes associated with positive or negative stimuli, and sometimes practically random things. And this observation means that not only is the forgetting curve true but also soething else is going on, which leads me to make the assumption of the duality of human memory.

      Other functions, such as breathing, can be controlled by consciousness or left to look after themselves. We can simply allow ourselves to breathe naturally in which case we will stop consciously observing breathing and it will continue optimally anyway, or we can focus on it and control it for the purpose of swimming, singing, evading detection by psycho gunmen with dog-like hearing, etc. But you will note it is difficult to control the breathing process and not control it (ie leave it to the unconscious mind) both at the same time. I think in a sense it is like a gentle switch which we barely notice being flicked.

      Driving the car is another example. Don’t get me wrong, I am not encouraging people to drive the car without due care and attention, but every driver knows that there are stretches of road and road situations that need concentration and others where we can start to listen to what’s on the radio and leave quite a lot of the driving job to the unconsious mind. So strong is this that we hear cases of people losing consciousness, continuing to drive and wake up in a different place with no recollection of how they got there. This of course is highly dangerous, if your eyes close and you sleep normally rather than go into some kind of trance, then of course you cananot see where you are going and you will have an accident. I am not talking about going to sleep here, I don’t believe in systems for learning during sleep any more than I would drive during sleep. I am talking about letting the mind relax and trusting in certain mental mechanisms to do their work so that we do less work consciously.

      When it comes to how music is detrimental to learning a corpus of material harnessing these unconscious functions or this unconscious function, my biggest “proof” if you allow the not quite literal use of that word, is that many people (not all) know the words and the musical details (little riffs and background instrumental flourishes) in a huge list of musical works. Hundreds and in many cases thousands of popular songs. I know I do. In some cases I sat and purposefully learned lyrics in order to perform better at karaoke, one hobby of mine I share with my daughter. But in most cases I already have a very good basic memory of the song and most of the lyrics, thousands upon thousands of words. I never tried to learn them, I was just enjoying them. In many cases they were a background to soething else going on. So how did I learn them? Not consciously. Hence for people who know that they have such a playlist in their heads, they can say for sure that their long-term memories do get attracted to music, and so if music is playing while they are doing their goldlist, that’s probably where the unconsious mind will be sampling, and not in the material we purposed for it to sample – either not at all or significantly less.

      Maybe the analogy of walking and driving is not a great argument, but I will give you one “proof” for why I think keyboard writing is less useful for memorising than stylus writing. You don’t notice so much that you make a mistake a “typo” when typing until you read it back, you tend to notice more immediately if the hand holding a pen guide the pen wrongly. So you are more conscious of the pen and therefore the unconscious has more leeway to latch onto the content.

      I am still open to the possibility that GLM should become an App. I am sure that it would be beneficial to those who simply cannot and will not use stylus writing although in these days of touchscreens rather than keyboards it is possible that Swype and other more stylus-like input systems could be on the rise again. My problem is, I cannot get anyone to finish writing an App which would deal with the combinations correctly and push out good stats and do the whole social media interaction I’d be looking for. Three years ago someone started writing it, and cannot seem to get it finished due to wrok commitments. Since the book which i am supposed to be writing is in the same state, I can hardly blame the chap. But it will be interesting to see how the app is received once it is ready.

      1. Thanks for this reply and for taking my comment seriously. I don’t doubt for an instant that handwritten input is superior; I would simply find it difficult to believe that the ratio of details remembered _about a given target stimulus_ to time spent could ever be greater for driving or for keyboarding Goldlists; but of course there are many factors involved beyond the simple question of speed (see, for example, and the related book—I have no known connection whatsoever, commercial or otherwise, with the author beyond owning a used copy of his book). According to my understanding and experiential knowledge, the best case possible would be for the absolute volume remembered from driving/keyboarding to equal that of walking/handwriting, but the remembered details for the faster case would be more widely dispersed across the extent of the target stimulus, making them less cohesive, less organizable, and less recognizable.

        I like the proof you have given that one realizes a mistake much more quickly with pen or pencil. The way the eye follows the script ( ) appearing beneath the instrument would continue the chain of cognitive involvement that includes the hand, arm, and silent attention to the word, if not also verbal production and aural reception into the bargain at the same time.

      2. While thinking about the exactness with which Ebbinghaus’s curve is met, I was trying to come up with a good analogy for why it’s not necessary to meet it exactly. But I think you’ve already said it best with the discussion of trying to tan but running back inside every 10 minutes to check on its progress in the mirror. It’s not that you’ll never get tanned, but it’s like watching a pot on to boil.

  2. And here are some links to sources reporting, summarizing, and/or establishing the very solid science defending handwriting. Note that three of these links are to a publication typically linked to industries that benefit greatly from widespread abandonment of handwriting in favor of technology adoption, yet it disinterestedly defends the scientific result. Of course I don’t think they read Brown’s Make it Stick carefully enough, or they wouldn’t have let the word hacks into the title. “The Pen is Mightier than the Keyboard.” Psychological Science Vol. 25 (6), 2014 Jun, pp. 1159-68. Search for it on Google Scholar to find freely downloadable draft versions in PDF.

Your thoughts welcome, by all mean reply also to other community members!