Learning Dvorak
This article is the first of a series. Next: Dvorak #2
Disclaimer: a lot of the historical information about typewriting and about Dvorak itself is disputable. However, the conclusions I want to draw here are not based on the historical facts, so the argument should stand on its own regardless of what anyone thinks about the history.
This is QWERTY, the layout most anyone who has ever used a keyboard (or, I’ll say it exactly one time, a typewriter) is familiar with. There are a few small variations for command keys, especially for Mac users, but in terms of the letters, their positions relative to one another remain the same.
Ok, familiar to everyone. So what’s the point?
In one of my previous articles, I explained that my younger self thought that to redesign something, my best shot was to think of an object so mundane or otherwise ignored that maybe, just maybe, nobody had really thought about how to redesign it before and I could be the first one. A keyboard would be an OK example of this — most people are introduced to a keyboard at a very young age, and by the time they’re old enough to think about a keyboard, it’s likely so familiar to them that they wouldn’t even think to think about it. The QWERTY keyboard is so ubiquitous that it would be easy to accept that it is the best keyboard layout possible, that it once competed with others in a market and won because of some combination of traits that made it superior to others. However, this isn’t the case.
The QWERTY layout was designed in 1874 by Christopher Sholes. This part isn’t really debatable, but why he chose the layout he did is shrouded in uncertainty. The most common folktale holds that since this layout was first designed for the typewriter, it was made so that “typewriter” could be written all on one row. Usually that’s where the story ends. Outside of being a popular myth, this story doesn’t make that much sense. The one-row part is true, but how did he decide to place all of the other keys? And if he was going to put “typewriter” on one row, why didn’t he just place the word from left-to-right in the order the letters appeared if that was to be his arbitrary starting point?
A much more likely story, but one still not completely accepted, is that the bars on a typewriter could jam and get stuck together when pressed in too rapid succession. The solution to this was to put common letters far apart from each other to slow down typists. If this story is true, this seems like the worst possible way someone could wish to design a modern keyboard, so I hope it isn’t the case — but either way, none of this is the main point. The point is that QWERTY is the most common U.S. keyboard today, and it’s possible that it should be redesigned.
August Dvorak was a psychologist and professor of education at UW Seattle sometime around the 1920s and 1930s. He was also interested in design, and while the HCDE department did not exist back then, he certainly sounds like he would have been a professor if it had. He redesigned the keyboard layout into his own, now known as Dvorak, which can be seen below:
Dvorak set out to fix a number of problems with the QWERTY layout that had been previously documented, including, but not limited to:
- Overall imbalance of load on left and right hands
- Consecutive strokes on only one hand — try typing “sweaterdresses” on your keyboard
- Research at the time found 52% of QWERTY keystrokes were on the top row, 32% on the home row, and 16% on the bottom row
He studied the frequency of letter combinations in English as well as the physiology of the hand, and came up with some conclusions that would shape the new keyboard he was to design:
- The home row is the easiest to use, then the top, then the bottom — so the most common letters should be placed on the home row
- Letters should be typed alternating across each hand as much as possible, so the most common digrams (meaning: 2 letters, slightly different from digraph) should be on opposite hands
- Most people are right-handed, so the right hand should have a slightly higher load than the left.
I have not studied letter combinations in any great capacity, but my parents once bought a code-breaking book for me somewhere when I was about 10. The only part that I remember was that in some code where each symbol (usually just a different letter) matched up one-to-one with a letter, I should look for the most common occurrences of these new symbols and remember: ETAONRHS were the most common letters at the time, in order of appearance. So whatever symbol came up the most was probably E, and the next T, and so on. I don’t remember anything else from the book, but my limited anecdotal evidence, which should count for nothing, supports Dvorak’s conclusion of putting E, T, A, O, N, H, and S on his home row keys.
The grey area arises when it came time to actually measure the efficacy of the layout. In a series of tests in the 1940s, Dvorak “found” his keyboard to be much more ergonomic than QWERTY, and thus better. However, studies have failed to produce conclusive results on the benefits of Dvorak. One jointly published article said,
“[T]he evidence in the standard history of Qwerty versus Dvorak is flawed and incomplete. [..] The most dramatic claims are traceable to Dvorak himself; and the best-documented experiments, as well as recent ergonomic studies, suggest little or no advantage for the Dvorak keyboard.”
To be clear, there is also no conclusive evidence that the Dvorak keyboard is worse than the QWERTY keyboard. There also just isn’t evidence that touch typists are faster using Dvorak, or that there are long term benefits such as avoiding RSIs, as Dvorak originally promised. Either way, whether or not Dvorak is actually better is inconsequential to my point in exploring Dvorak.
All of the above is just my investigative work over a long time, as I read about Dvorak and QWERTY. I also read about other layouts such as Colemak, but I will be learning Dvorak rather than anything else. The remainder is my own story and my takeaways from the history — I didn’t just write this article for the history itself.
My dad has used the Dvorak keyboard for as long as I can remember, certainly more than 15 years. This was my first introduction, and part of why I am choosing it over anything else. Like I said before, it’s also unclear that any particular layout is better than another in practice, so in making a relatively arbitrary decision between Dvorak and others, I’m fine to let emotional factors dictate my choice. It’s a nice two-for-one that I later found out August Dvorak was affiliated with the UW. Third, Windows comes with Dvorak keyboard pre-installed, but not other options like Colemak, so it’s much more convenient to learn Dvorak because all I need to switch from QWERTY to Dvorak is to hit Ctrl + Shift. And having a physical keyboard in Dvorak instead of QWERTY makes no difference to me at all — I’ve been touch typing since I was about 10, so I don’t look at the keyboard at all, and won’t for Dvorak either. A note: I first started practicing Dvorak occasionally over two years ago, but I have only invested a couple hours in total, so I consider myself to not have really attempted to learn it yet.
The real purpose in my Dvorak-learning experiment is to try an alternative to the norm because I am aware of it and because I can. Someone took the time to carefully design a product that I could enjoy better than the current iteration, so I feel obligated to try it. It could end up working better for me personally, even if it hasn’t shown statistically significant advantages across wide ranges of participants.
Choosing alternatives to the default has also been correlated with success, as I found out from reading an article recently. I don’t have the link to the article I personally read, but a quick search turns up with this one from the Atlantic which also cites this synopsis of a Freakonomics Radio podcast, and Levitt and Dubner are some of my favorite researchers, period, so I don’t need any more convincing. The correlation I am referring to is between people who don’t use Internet Explorer (now Microsoft Edge), and success on different metrics. Functionally, the browsers are all basically the same, they just differ in small ways, and mostly taste. I started using Google Chrome in middle school because some of my classmates (also children of tech parents) told me I should try it, and haven’t looked back. Reading the article and switching your default browser won’t automatically put you in a better position, but being the type of person to continually question the default and look for the best option will get you places. I do this in countless situations each day, but this is a big one that I decided I wanted to highlight with a post. Thus, you can see how my purpose in writing this stands apart from the actual effectiveness of the Dvorak keyboard layout.
The point here is really that by testing alternatives and collecting data, you can make more informed decisions and set yourself apart from others, and continually assessing options to find the best outcome will pay off in the long run.
So, for my actual experiment, I will continue slowly learning Dvorak as time allows over the next few months. At some point I will be able to completely switch over to Dvorak for all of my typing. For some time after that, my speed will still increase as I become more comfortable with the keyboard — I don’t believe I’ve even plateaued yet in my QWERTY words per minute speed. I will record the date I switch, and I will write a follow-up to this article sometime after that point recounting my findings.
My current typing speed in QWERTY is around 115+ WPM, depending on the typing test. Two or three years ago, I was around 100 WPM. I will draw quantitative comparisons to my speeds in Dvorak, but I will also make qualitative observations about how it feels to type in Dvorak.
