A base 20 notated chorded typing system that works with regular keyboards
- Software Tools
- Introduction
- Hardware Issues and Workarounds
- Finger Keys
- Code Digit Layout
- Base 20 Encoding
- Code Examples
- Chord Ordering
- Dictionaries and Mappings
- Learning and Practice Recommendations
The Chord-20-20 typing system, or "chord20" for short, is a uniquely designed chorded typing system. In comparison to traditional courtroom stenography, the chord20 system is not phonetic based but rather word based, and can work with regular keyboards.
Because chord20 is in active development, the software is still experimental and the applications are being updated and/or replaced rapidly. Currently, there are two primary software tools available: a testing tool and an editor tool. At least one other tool is planned, and that is a training or learning tool.
The code trainer is designed so you can learn and practice chord patterns without having to worry about what word you are typing. This allows you to train your muscle memory and learn the chord notation with less cognitive overload.
This trainer will also be updated to work on touchscreens.
| Version Date | Source Link | Demo |
|---|---|---|
| Latest | gist | Live Demo |
The testing tool is designed to have everything you need to try different keyboard configurations and settings all in one place, as well as the full dictionary chord list for reference.
| Version Date | Source Link | Demo |
|---|---|---|
| Latest | gist | Live Demo |
| Jan 5, 2025 | gist | Live Demo |
| Nov 26, 2024 | gist | Live Demo |
The Chord20 Editor is designed to have an simplified interface so you can focus on writing or composition. You can show the text normally or have it translated into chord notation, and you can enable or disable chorded input. You may even want to try reading long texts directly in the chord notation to help learn or solidify the chording dictionary more rapidly.
While some settings are not directly available in the user interface, these can still be customized by editing the section "JSON" in the monolithic html/js distribution file.
| Version Date | Source | Demo |
|---|---|---|
| Latest | gist | Live Demo |
| Jan 5, 2025 | gist | Live Demo |
| Dec 13, 2024 | gist | Live Demo |
| Dec 1, 2024 | gist | Live Demo |
| Nov 26, 2024 | gist | Live Demo |
| Version Date | Source | Demo |
|---|---|---|
| Latest | gist | Live Demo |
Before one starts chording, one might want to ask what is the best system to learn. In the long run, the best system to learn may depend on what your ultimate goals are, but the simple answer is that any chording system will be beneficial for newcomers to get started learning. The best system to learn may depend on what your long term goals are. In particular, stenography and court reporting is a well established system with a long history of professional use. Because this system has been improved through real world use over more than 100 years, court reporting steno type keyboards represent the state of the art in chorded input.
However, there are many practical disadvantages to traditional stenotype keyboards for your typical average user, because these systems are optimized for professional stenographers. Many individuals type both for work and casual activities, and the speed and or fatigue benefits of a chorded system may be useful for many such individuals. There are 3 main drawbacks to traditional stenography for an average user:
- Most Traditional Stenography are interpretted phonetically, not principally designed to map to a written dictionary.
- Traditional Steno requires special hardware that is both costly and less available, and difficult to carry all the time.
- Traditional steno is not immediately interoperable with regular typing.
Most stenotype theories are phonetic based, which means that there may not be a definitive relationship between words and the physical keys. While this makes sense for the verbal context of court transcriptions, as it can most accurately replicate spoken language, most people focus on writing text, not transcriptions. A chorded typing system that is not phonetically organized but rather mapped to a written dictionary, has many potential advantages. These advantages come down to having a definitive mapping between a chord and specific word, so you learn specific words rather than phonetic patterns. Having a definitive word-chord mapping makes realtime input much easier and reliable. The inherent challenge of chorded input systems, is that you have to change the "atoms" from individual letters into some other more complex linguistic unit. Phonetics is one way to do this, but digital text is not phonetic based, so it means that a phonetic based mapping requires adjustment or additional training to be optimized for digital text.
Another significant hurdle of traditional stenotype keyboards, is that they require special hardware that is not widely available or familiar to most users. This is not just a financial hurdle for new potential chorded typing learners, but it is also becomes a logistical hurdle for using chorded input in any type of new or different environment. If you want to go to a library or computer lab, you would need to bring your special stenotype keyboard with you. It makes it more difficult to practice and use. Because traditional qwerty keyboards are ubiquitous, a chorded typing system that optimizes for these keyboards makes a lot of sense for people who want to master chorded typing outside the context of courtroom stenography or professional transcription. Many people use typing in their profession, not just stenographers and transcriptionists, and all these professional can potentially benefit from having the option of chorded typing when that makes sense.
Finally, this system is beneficial because you can easily switch back and forth between regular typing and chording. In this system, you can press single keys at a time in order to type normally, or you can use a simple shortcut to turn off chording altogether, so that can type regularly at full speed, without accidentally overlapping two keys and thus triggering a chord. While there exist hybrid stenography keyboards, this system works on regular hardware and any system through software level intepretation.
Thus, the chord20 typing system is attempting to offer an alternative that could be attractive for many kinds of professionals who use typing outside of the traditional courtroom transcription profession, as well as for casual or new users who want to learn something new or useful.
While you are always free to build customized hardware, and the finger key layout can be easily adjusted in chord20 settings to work with any such hardware, it is nice to have a system that works on regular keyboards right out of the box. And while typical keyboards may support limited simultaneous keys or rollover, the chord20 system allows chords to be "rolled" as a workaround(rolled chords are described in the hardware issues section). Additionally, it is not difficult to find gaming or mechanical keyboards that can support at least 5 key rollover so this is not an issue. This is a much more accessible and affordable entrypoint in the world of chorded typing than a traditional stenotype keyboard.
The first thing that you need to learn to master the Chord20 typing system, is the notation for chords. One of the unique design features of the chord20 system, is that you can separate the learning into two clear steps: being able to physically articulate the chords on a keyboard at your desired pace, and memorizing large dictionary sets. Ultimately, integrating these two processes: articulation and memorization, is the key to fluency, but the ability to practice them separately is very beneficial for learning and feedback.
Isolating articulation from memorization means that you can get a much better idea of your potential typing speed, long before you have memorized a dictionary of thousands of words. While the greatest speed may come from years of practice until it is completely instinctual, it is possible to pracitice short chord sequences unWhile the greatest speed comes from subconscious typing, it is possible to learn a simple chord sequence and practice it to assess your potential speed. You can measure your ability to
While you can choose to map the keys differently, or use different hardware, The chord20 system is designed for a standard qwerty keyboard with a rectangular, non-ergonomic layout.
Every chord in the standard layout requires two or more keys. If a single key is pressed in isolation, instead of inputing a chord representing a word in the dictionary, the single character for that key is input. The keyboard layout for individual key presses is named the "Single Key Layout". Qwerty is the default single key layout, but alternative single key layouts can be used as well. Regardless, it is recommended to set the system keyboard layout to qwerty, and then use the software settings to configure your desired single key layout. That way the configuration for finger keys and other settings will be portable. Some of the available alternative single key layouts are dvorak or colemak, or you can use your own custom single key layout.
One benefit of allowing single key input, is that if you forget the chord for a specific word, you can simply type it out normally, and then the editor will indicate the corresponding chord code in the dictionary. This is also why it is so important to learn the chord code notation. Once you master this notation, then you can easily look at a simple two or four digit code, and know exactly what keys to press on the keyboard to make the chord. With enough practice and fluency, you can even learn chords simply by reading a book or other text translated into chord notation!
In order to type a chord, you must press two or more keys simultaneously, that is, press a second key before releasing the first. Chord20 also allows you to "rollthrough" chords, meaning that if a chord involves more than two keys, you only need to hold at least one key down as you type the chord. In other words, you can press one key, and then a second, and release the first, and then press third key, etc. The chord is only emitted once all keys are released, and any keys released early are still included, as long as you are holding at least one key the whole time until all keys have been pressed.
Many standard keyboards are limited in the number of simultaneous key presses they can handle. This is a hardware issue called "rollover". If a keyboard has "3 key rollover" then you can press any 3 keys at once, and it will register correctly. Some or most combinations involving more keys may still work, but at least one combination of 4 keys will fail to register correctly on a keyboard with only "3 key rollover". Rollover can be tricky because it is often inconsistent and may only happen for a few irregular combinations.
Ghosting is an issue related to rollover, where a key which was never pressed will register when multiple keys are pressed at once. The majority of keyboards you can find today greatly reduce or eliminate ghosting, meaning they will not register a key you don't press, even if they still fail to correctly register combinations involving too many keys. This is important for chorded typing, because even if you are forced to use "rollthrough" to type a chord, you will not get a ghosted key that you never actually pressed included in the chord.
For example, the combination "F 4 8" will fail to register on many budget keyboards, even if most other 3 key combinations register correctly. But even many budget keyboards will not register a key you never pressed, which is important as that avoids adding phantom keys to a chord.
In terms of hardware, rollover happens because keyboards do not have independently wired switches for every key, or because of how the firmware processes and encodes key presses. A simple visualization is to imagine how imagine a road map laid out in a grid. If you only need to specify one square in the grid, then it is sufficient to specify the row and column. But if you need to specify many grid squares at once, then specifying two rows and two columns can be ambiguous, as there are 4 points of intersection. While the wiring or encoding of keys is more complex than a 2d grid, this demonstrates the basic principle.
In order to type chords you must press multiple keys at the same time, using different fingers. To keep things simple, in the default layout the same finger always covers the same keys. Your hand won't need to move the up and down the keyboard the way a piano player would.
The suggestions in this guide are intended for new users, with experience you may choose to adapt your particular techniques however you see fit. These recommendations are designed to keep the learning process simple and consistent, and provide a practical layout that works well for a variety of hardware, software, and languages. It is very difficult to try to optimize a chording system before you have trained yourself and practiced, as you won't be able to tell if something is inherently difficult, or only difficult because you haven't trained yourself yet. For that reason it is recommended to learn the standard layout first, before you begin making significant changes to your chording techniques.
As a quick example, the default keys for the left hand middle finger are "D", "E", and "4". Those keys should always be pressed by that finger and not any other finger. This finger will move to press these keys, and other adjacent fingers will need to move at the same time. The left hand's ring finger keys are "S", "W", and "3". So when you are typing chords with those fingers, the possible key combinations are "SD SE S4 WD WE W4 3D 3E 34". The keys could also be typed in combination with keys elsewhere on the keyboard.
1 2 3 4 5 6 7 8 9 0 - = Backspace
q w e r t y u i o p [ ] \
a s d f g h j k l ; ' Enter
z x c v b n m , . / Shift
Space
1 3 4 8 9 -
q w e i o p
s d f g h j k l
c v n m
These are the only keys that are actually used for typing chords. You could think of the "home row" as slightly staggered vertically: "qwef" and "jiop". Importantly, the pinky is only assigned 2 keys, and the pinky and index finger keys are chosen so that they move in opposite directions, allowing for more space and not cramping the hands. While the layout is not completely symmetrical, that is one of the compromises that needs to be made to adapt the layout for standard keyboards, so that the system does not require special hardware to learn.
- Left Pinky: q, 1
- Left Ring: s, w, 3
- Left Middle: d, e, 4
- Left Index: f, g, v, c
- Right Index: j, h, n, m
- Right Middle: k, i, 8
- Right Ring: l, o, 9
- Right Pinky: p, "-"
These finger keys are listed in the order they are physically laid out on the keyboard. But please realize this is not the order they appear in the chord notation, which is explained in later sections.
8 3 3 3 3 8 1 3 4 8 9 -
4 2 2 2 2 4 q w e i o p
1 1 4 8 8 4 1 1 s d f g h j k l
G C C G c v n m
These are the actual values that each key is assigned in the code notation. The key layout is shown to the right for refernce. This chart will make more sense, after you read about base 20 encoding and the code digit layout. So you may want to read those sections and then come back to review this.
Each chord code contains either 2 or 4 digits. Every chord may be written in the 4 digit format, but chords that only use the two "major" or "inner" fingers on each hand can be shortened 2 digits. The digit layout is designed so that the keys which are easier to press with stronger fingers come first in the number. For that reason the digit layout does not match the physical layout of the keys or your fingers. The first digit in the layout encodes the keys for your left hand inner or "major" fingers, the second encodes the left hand outer or "minor" fingers.
Because the minor or outer fingers are weaker, the chord system is designed so that these are used less often, in words that are used less frequently. For this reason the minor fingers come after the major fingers in the code notation, and are ommitted in the 2 digit abbreviation when no minor fingers are used.
- Major/Inner Fingers: middle, index
- Minor/Outer Fingers: ring, pinky
- Digit 1: Left Hand Major
- Digit 2: Left Hand Minor
- Digit 3: Right Hand Major
- Digit 4: Right Hand Minor
- Digit 1: Left Major
- Digit 2: Right Major
Base 20 is a numerical encoding system that uses the 10 numerical digits, together with the first 10 letters of the alphabet, "A" through "J", to count using 20 digits total. This is very similar to hexidecimal, except instead of counting to "F" and then going back to zero, we count all the way to "J" before going back to zero.
If you are not familar with hexidecimal notation, I suggest you learn to count in binary first. A simple way to think about binary counting, is that because there are only two digits, any time you see '1' you can treat it as a '9'. So you start counting with '1', and then change that to '9', the next number is '10', and you can change the '1' to a '9' so you have '90' then you count to '91', then you change that '1' to '9' so you have '99' and the next number is '100'.
"Modified" Binary Counting (change every '1' to a '9')
1 (9) 10 (90) 91 (99) 100 (900) 901 (909) 910 (990) 991 (999) 1000
Binary Counting (convention is use '1' as the alternate digit, but you can also use '9')
1 10 11 100 101 110 111 1000
9 90 99 900 909 990 999 9000
Hexidecimal Counting
0 1 2 3 4 5 6 7 8 9 A B C D E F
10 11 12 13 14 15 16 17 18 19 1A 1B 1C 1D 1E 1F
20 21 22 23 24 25 26 27 28 29 2A 2B 2C 2D 2E 2F
Base 20 Counting
0 1 2 3 4 5 6 7 8 9
A B C D E F G H I J
10 11 12 13 14 15 16 17 18 19
1A 1B 1C 1D 1E 1F 1G 1H 1I 1J
20 21 22 23 24 25 26 27 28 29
2A 2B 2C 2D 2E 2F 2G 2H 2I 2J
Base 20 is used because the major fingers on each hand can form up to 20 combinations. The middle finger has 3 keys or can press nothing, making 4 options total, and the pointer finger has 4 assigned keys or can press nothing, making 5 options total. Together there are thus 4x5=20 possibilities for the major fingers on each hand, allowing for nearly 400 total chords with just the major fingers. It is not quite 400 because single key presses or pressing nothing is not registerd as a chord.
In order to work in base 20 fluently, it helps to practice basic arithmetic problems
- 8 + 8 = G
- A + 5 = F
- G + 4 = 10
- D - 9 = ???
- D + E = 1D - 6 = 7
- F + A = 1F - 5 = A
- G + 8 = ???
You will not need to add more than 4 at a time, so if you only want to practice adding 1, 2, 3, and 4, that should be good enough to use the notation fluently. But it can be nice to learn anyway :)
Example Codes:
Code Left Right
Middle Pointer | Ring Pinky | Middle Pointer | Ring Pinky |
5032 "D" 1 "F" 4 | | "8" 3 | "O" 2 |
76 "4" 3 "F" 4 | | "I" 2 "J" 4 | |
3AGB "4" 3 | "W" 2 "1" 8 | "m" G | "9" 3 "-" 8 |
In each of these codes the character in quotes is the physical character on the keyboard. The number(in base 20) next to it is the partial code digit for that particular position. To get the final code digit, you simply add the partial code digits for the two fingers. So in the code 5032, the key "D" is partial digit value of '1' in the first digit position, and the key "F" is a partial digit value of '4' in that same position. Adding 1 + 4 we get '5' in the first digit position.
Similarly, the two character code 76, requires keys "4" (3) and "F" (4) to make 7, and then the 6 is key "I" (2) and "J" (4). In the last code, 3AGB, be aware that 2 + 8 = A, and 3 + 8 = B.
Here is what each of those chords looks like on the key layout:
8 3 3 =3= 3 8 1 3 4 =8= 9 -
4 2 2 2 =2= 4 q w e i =o= p
1 =1==4= 8 8 4 1 1 s =d==f= g h j k l
G C C G c v n m
8 3 =3= 3 3 8 1 3 =4= 8 9 -
4 2 2 =2= 2 4 q w e =i= o p
1 1 =4= 8 8 =4= 1 1 s d =f= g h =j= k l
G C C G c v n m
=8= 3 =3= 3 =3= =8= =1= 3 =4= 8 =9= =-=
4 =2= 2 2 2 4 q =w= e i o p
1 1 4 8 8 4 1 1 s d f g h j k l
G C C =G= c v n =m=
Given the notation just described, you should now understand how a given code represents a certain combination of keys pressed in unison on the keyboard. In order to list all the codes and map them to a dictionary of words, it is necessary to have a schema for ordering these chord codes.
While we could simply use a standard numerical counting system, there are several drawbacks to this approach. For one thing, in order to simplify the process of mapping chord codes to a dictionary, we want the easiest chords to come first in the ordering. For that reason, we use a 2 tiered ordering system. This is similar to how you might sort a list of people using first and last names. The last name might be the first sorting criteria, then if those match the first name is used. Similarly, we have three distinct sorting criteria for chords.
The first ordering criteria is how many keys are pressed in a given chord. A chord can have from 2 keys up to and including 8 keys, depending on your settings. The default setting is to only allow a maximum of 5 keys per chord, which allows for over 19,000 possible chords, with the standard key set. If you want to allow for larger dictionaries you can increase the maximum number of keys up to 8, which allows for over 50,000 possible chords. Needless to say, learning all of these chords would be quite a chore!
Chords with more keys come last, with the exception that 2-key and 3-key chords are considered equivalent. The number of 2 key and 3 key chords is about 1,800, which would be a good milestone in learning this chording system. We didn't want to require the pinky or ring fingers as often for the most frequent words in the dictionary, so the first pass includes both 2-key and 3-key chords. For the most part, 3-key chords are not much more difficult than 2-key chords, especially if they use mostly middle and index fingers.
The second and third sorting criteria are related. Each hand is considered separately, and the partial chord combinations for a hand have a corresponding numerical value. The second sorting criteria is the sum of these numerical values, referred to as the "Diagonal Score". Imagine the numbers for each hand laid out as rows and columns in a grid. Then essentially what we do is traverse this grid along the diagonals.
Right Hand
0 1 2 3 4 5 6 7 8
0 _ _ _ _ _ 05 06 07 _
L 1 _ 11 12 13 14 15 16 17 18
e 2 _ 21 22 23 24 25 26 27 28
f 3 _ 31 32 33 34 35 36 37 38
t 4 _ 41 42 43 44 45 46 47 48
5 50 51 52 53 54 __ __ __ 58
6 60 61 62 63 64 __ __ __ 68
7 70 71 72 73 74 __ __ __ 78
8 _ 81 82 83 84 85 86 87 88
As a quick example, the first few chords are:
Diagonal Score 2: 11,
Diagonal Score 3: 21, 12
Diagonal Score 4: 31, 22, 13
Diagonal Score 5: 50, 41, 32, 23, 14,
Some potential combinations are omitted, either because they involve only one key, or because they would involve more than 3 keys, so they are not included in the first pass. For example, 01 through 04 would only be 1-key chords, so they are not counted. Chords 55, and 56, would require 4 keys, so they are not included either, at least not in this pass.
Here is the list of the first 30 chords
#n keys fingers code word
================================================
1 __d__k__ 0010 0100 11 a
2 __e__k__ 0020 0100 21 from
3 __d__i__ 0010 0200 12 look
4 __4__k__ 0030 0100 31 say
5 __e__i__ 0020 0200 22 time
6 __d__8__ 0010 0300 13 about
7 __df____ 0011 0000 50 get
8 ___f_k__ 0001 0100 41 make
9 __4__i__ 0030 0200 32 see
10 __e__8__ 0020 0300 23 to
11 __d_j___ 0010 1000 14 all
12 ____jk__ 0000 1100 05 give
13 __ef____ 0021 0000 60 man
14 __df_k__ 0011 0100 51 she
15 ___f_i__ 0001 0200 42 two
16 __4__8__ 0030 0300 33 also
17 __e_j___ 0020 1000 24 go
18 __d_jk__ 0010 1100 15 many
19 ____ji__ 0000 1200 06 so
20 __4f____ 0031 0000 70 up
21 __ef_k__ 0021 0100 61 and
22 __df_i__ 0011 0200 52 have
23 ___f_8__ 0001 0300 43 me
24 __4_j___ 0030 1000 34 some
25 __e_jk__ 0020 1100 25 use
26 __d_ji__ 0010 1200 16 as
27 ____j8__ 0000 1300 07 he
28 __4f_k__ 0031 0100 71 more
29 __ef_i__ 0021 0200 62 take
30 __df_8__ 0011 0300 53 very
In order to input a word when you type a chord, the software needs some kind of dictionary or mapping which determines, the word that is input when a given chord is typed. Dictionaries should generally be ordered with the most frequent words first, so that the easiest chords to input get mapped to the most common words. Dictionaries can also be "layered", meaning that you can have a small dictionary of the "first 100 words", and then use some other dictionary after that. This makes it easier to adapt and learn new dictionary sets much more easily, as you can retain an underlying set of common words that is consistent.
When you first start, you will just want to get familiar with the finger key layout and inputting chords. Memorization and performance should not be you first concern. To really fluently learn a chorded typing system can take up to a couple years, and this project has as of this writing only been in development for a couple months. A reasonable timeline is to learn 100 words in 2 to 3 months, such that you can type at a steady(but not necessarily rapid) pace with this word set. At 6 months to a year you might want to aim for 1000 words, and much greater fluency with the first 100 words.
Just like learning a language, one of the keys is to not manually translate everything in your head. It is useful to learn small common phrases or sayings, rather than only learn words individually at a time. The biggest hurdle is understanding that much of your goal in initial learning, is not to get results, but rather build up your ability to practice and train productively for longer periods of time.
Many new learners have an all or nothing mindset, and difficulty breaking up learning into smaller tasks and goals, as well as understanding what an ideal progression looks like. When you are brand new, it is very difficult to train or practice for long periods of time, many new learners make the mistake of wanting to immediately jump in to a full training schedule, and so much of their training is unproductive or wasted. It is much better to start training 5 or 10 minutes at a time in your first few weeks, and then gradually increase your training schedule. This also has the benefit of building habits, and allowing you to naturally grow your curiosity and interest in a new hobby, rather than forcing yourself.
Even if you start gradually in this way, you can quickly ramp up to an intense training schedule if you really want to, which is why in addition to progressive training, I also recommend periodic training. This means that you train more intensely for a few weeks, and then less for a few weeks. One of the benefits of this is it helps you get more time under your belt, without costing you too much in training time. I would describe this as optimizing active and passive learning. Active learning is when you consciously sit down and deliberately practice something. Passive learning happens in the background or without much thought, when you are either exposed to something or simply daydream about it unintentionally. Overloading active learning without allowing passive learning to catch up, can be tiring or lead to burnout. Give yourself time to become familiar without overloading your training.
While it may seems incovenient or counterintuitive, the most productive way to train most of the time is not using the software or even a physical keyboard, but rather with a physical pencil and paper. The reason why this is such a productive way to train, is because you can both do it consistently and it emphasizes "generation" over "consumption" or "reaction".
When you are forced to generate things internally and then perform them outwardly, this is what really leads to mastery. While you can do this with software or study prompts, it is very difficult to get this synchronized with your ideal internal pace for "generation". Study apps are terrible if they are only form of practice, because the app is constantly prompting you, so you are responding to stimulus and not generating things autonomously. As soon as you are forced to generate something without a prompt, you will feel very uncomfortable and struggle. The key to learning is embracing just the right amount of discomfort and that slightly feeling of unease or uncertainty, until you are able to master something and build confidence and mental "balance"
The other extreme would be to have no assistance and simply try to use something in real world scenarios. However, this is generally overload and unreasonable. Studying with pencil and paper is in all likelihood the ideal middle ground. Your notes can assist you, but you are still forced to organize and perform the tasks yourself.
For example, you can write down a list of words or codes, or phrases. It is beneficial to delay feedback when you are generating and verifying what you write. Delaying feedback forces your brain to increase its retention.
An example of delaying feedback would be using flash cards. If you check the back of the flash card immediately after every question, then this is immediate feedback. But a more advance study technique is to go through 3 to 5 flashcards before checking if you got the answers correct. After you have answered 3 to 5 prompts, then you can go back and check them. It is also okay if you don't remember, and give yourself a second chance to recall an answer. This teaches your brain to be comfortable with uncertainty and that it is okay to take time to process and recall information. These techniques can greatly improve the quality of your study and your retention ability.
One great way to take your training to the next level, is to supplement it with a small amount of mental or physical rehearsal. This means reviewing things you have studied in your head or pretending you have a keyboard, or even typing on a keyboard that is not attached to anything. Mental rehearsal is benficial because it both eliminates all distractions and reduces mental overhead associated with supplemental processing. You don't realize it, but when you use a computer, there are a lot of tasks involved. You are looking at the screen and responding to that, you are processing the feedback of how your actions change what is on the screen. All this adds to cognitive load, which makes it more difficult to engage in learning new skills, when you are constantly peforming all these supplemental tasks.
By adding a small amount of mental rehearsal to you training regimen, you allow yourself to engage in the learning process with zero distractions. But this can be incredibly difficult and frustrating as well, if you tried to make it your only training activity. The reason why it can be so challenging is because it is pure "generation" with no external prompts or aids. For that reason I only recommend that it be a small percentage of your training regimen, but even so it will be incredibly beneficial in my experience.
Many people don't realize how expensive translation is in any language learning task. Translation is a multistep process where you absorb information in one format or encoding, convert it to another, process it in the familiar format, convert it back to the format you are using, and then perform or communicate that information.
While this cognitive load is one of the downsides of translation, it is not even the biggest drawback. The biggest drawback is that you avoid processing information directly in the format in which it is presented, meaning that you never gain experience "natively" processing the information. It will be awkard and slow to do this at first, but by avoiding direct or comprehensive translation, you allow brain to become familiar with something unfamiliar or unusual, rather than relying on your old skills.
For this chording system, this would mean writing and reading directly in the chorded notation, potentially by only visualizing the physical keyboard presses and the meaning of the word, but not the sounds or visuals of the word itself. If you combine this with the principle of "delayed feedback" it can be very powerful.
Fluency in a language is defined as the ability to aquire new words and phrases in a native environment just from exposure. If you gain fluency with a new system of typing and notation, you will not need to think about the conventional representation of words, you can just think of the language as the chords themselves.
While mental translating 100% of what you do is a bad idea, it can still be useful to use a technique I call "mental scaffolding" Scaffolding is an easy to build non-permanent structure, that is eventually removed when the permanent structure is complete.
Mental scaffolding is temporary memory aids like mnemonics or other memorization techniques, that will help you to memorize something. Compared to comprehensive translation, mental scaffolding does not have to be logical or consistent. It is just a set of cues or associations that helps you perform mental tasks.
The irony of memorization, is that it is often easier to memorize more information than less. This is because you mind relies heavily on both context and connection for performing mental tasks. If I give you a sequence of random letters, then it can be very tricky to recall them. But if I create a story or paragraph to memorize that contains all the same letters in order, that can often be easier than simply remembering the letters themselves.
Again, such techniques are not intended to necessarily be permanent, eventually they become less important or you may forget them altogether, as you get more practice and experience. But they can greatly accelerate your learning and allow you to tackle tasks that would otherwise be overwhelming. Balance is recommended.