This year’s recent Touch Gesture Motion (TGM) conference had a surprising focus on pens. Which I like, actually. While most of my professional time is with a keyboard, I still take notes manually on paper. Partly it’s because, in an interview situation, I feel like it’s rude and impersonal to be typing away as if I’m some bureaucrat entering data into a form.
But, even though I’m a fast typist (on a real keyboard, not a virtual one), I can write even faster (depending on how long a legibility relaxation time I want). So it seems more efficient to write. But I write in a book, and I then need to keep track of which pages have notes for which topics when I come back to turning them into some kind of piece. I’d love to be able to write on a digitizing surface and then simply save note files.
N-Trig outlined other reasons why handwriting is useful:
- Annotating other work
- Expressing things other than text: art, graphs (in lab notebooks, e.g.)
- Math formulas, for instance… can you imagine trying to futz with, say, the Microsoft Word equation capabilities – which are nice when you want clean typeset formulas – when frantically taking notes from a quantum mechanics lecture? It would never work. You need to be able to scribble them.
That, unfortunately, isn’t practical today, and some of the challenges that remain were highlighted by N-Trig at the TGM conference.
This is a topic you might expect to see at such a conference. What I wasn’t expecting was information on studies that link handwriting to brain function. It appears that the process of writing activates various parts of the brain that help solidify information. Studies suggest, among other things, that:
- “Students without consistent exposure to handwriting are more likely to have trouble retrieving letters from memory; spelling accurately; extracting meaning from text or lectures; and interpreting the context of words and phrases.”
- “Elementary-age students who wrote compositions by hand rather than by keyboarding, one researcher found, wrote faster, wrote longer pieces, and expressed more ideas.”
Source: N-Trig presentation. He listed numerous sources, although it was tough to ascribe specific sources to specific points.
Now… this is from a company that sells digital pens, so the information serves them well, but it didn’t feel like simply self-serving research.
What he also confirmed is that pens still have some work to do to provide the kind of writing experience that we’re used to with real pens and pencils on paper. And that appears to be the gold standard. It’s honestly not clear to me if that’s an arbitrary standard that comes from what we’ve gotten used to or if there’s something more fundamental.
But there are a number of dimensions that have to be optimized for it all to work, including:
- You’ve got to be able to hold a pen at various angles and have it work properly.
- You want just the right amount of friction – and what’s “right” depends on what kind of pen you think you’re working with – felt tip, ballpoint, rollerball, etc.
- The digitizer response has to be fast – latency screws up eye-hand coordination.
- The digitizer has to be precise so as to capture all of the correct data points.
- The pen tip has to be long-wearing under continual usage.
- The palm has to be rejected accurately. Trying to write while keeping the palm up simply doesn’t work – it’s not how we’ve learned to write, and the large-motor “noise” swamps the fine motor fingers and you end up with writing that looks way worse than mine (which is hard to imagine, believe me).
They’re studying pen performance by measuring speed and accuracy – and even comparing pen-and-paper to electronic pens. One way of doing that was to use copy paper under an electronic pen/pad and then compare the digital result against the actual imprinted writing on the copy paper. An example is shown below, and you can see where the two align and where they miss.
Image courtesy N-Trig
As the picture suggests, while there’s a fair bit of green, indicating agreement between the digital and copy versions, there’s still a lot of blue and gray. The latter are near each other, so it’s mostly not a total miss, but ideally you want all green. And they’re not quite there yet.
And that’s just to get the pixels all in the right place. Once that’s in place, then handwriting recognition would be a nice bonus. Then again, if I can’t always recognize my own writing, I can’t reasonably expect a computer to.
So, in summary, the bad news: there’s still work to be done. The good news: it would appear that N-Trig and others are taking this seriously. We may yet get usable electronic pens that rival the real thing. Which would make my life easier, but, more importantly, would allow us to proceed with going digital in school and out without giving up handwriting.