Short Text Update: Learning 2,000 Words

Here’s an update on my Short Text project. Short text is a way to use one or two letters plus one or two letters for endings rather than spelling words all the way out when you type or write.

The first challenge in constructing Short Text was seeing how many words I could fit into this structure. I’ve already outlined how it all works, but the answer turned out to be around 15,000.

It’s one thing to decide that a word like “thing” is going to also be spelled “tg” and it’s another to remember to type or write “tg” when you want to put that word down.

So the second challenge was seeing how practical it was to learn. There are several layers to Short Text. The first few layers are more useful because they’re more common. According to an analysis of the Oxford English Corpus, a two-billion-word collection of English texts, the top 100 words, including endings, make up about half of all written English.

The first step was switching to short text for the most common word: the. In short text it’s just a single “t”. When you type a “t” followed by a space it expands to “the”, so you save a couple of steps fairly often with just this one Short Text spelling.

It’s not practical to memorize anywhere near 15,000 words, but what I wanted to know was how many words it would be worth memorizing – probably the top 100, and maybe more. At some point amid diminishing returns you’d switch to just looking up specific words you find yourself using frequently. Because there are unique identifiers for 15,000 words, you could, in theory, build up a pretty good vocabulary of shortcuts.

My husband Eric and I quizzed each other on the top three words for each letter, which are typed using one, two or three of that letter (these are handwritten with a tic mark for each repeated letter). The three words for each letter are sometimes easy to remember as a phrase, for instance “e” is “every even end”. I take handwritten notes during interviews and meetings, and Short Text is enabled on my smart phone. As I’ve learned it, Short Text has crept into my handwriting and texting.

Once you can think to type
“cc u d s e ddd”

instead of “Can you do some every day”
it’s lots faster.

That first layer is about 75 words, and each word has an average of two more endings. With the endings rules, learning the first layer gave us a vocabulary of about 200 words.

The next step was adding the 650 or so words whose shortcuts are two different letters. You can find structure in this list in a couple of different ways:
Down: ab, ac, ad, ae, af = able, act, ad, age, after,
Across: ab, cb, db, eb, fb = able, club, doubt, enable, fabulous
And there are some amusing random stretches: lf, lg, lh, li = life large laugh line

With the shortcuts enabled on a smart phone or computer, you can learn by guessing. You always know the first letter, because it’s the letter the word starts with, so you’re just guessing the second letter.

Between guessing and going through layer 2.1 both down and across, Eric and I learned all 650 of them. With the endings rules, this increased our vocabulary to around 2,000 words – well over the 100 words that make up 50% of writing. Now when I handwrite or text using my phone much of it is Short Text. I know it well enough that even when I write quickly I use most of the shortcuts I know.

The next layers are 2.2 and 2.3, which start with the first letter, then have two or three of the second letter. This means there’s kind of a triplet of words in each slot: ln, lnn, lnnn = learn, listen, lesson.

Since I already know the first word of the three, it’s fairly easy to pick up two more. But I’m not quizzing myself on these.

Instead, I’m onto the next step of learning, which is a little more custom than straight memorization. When I notice I’m using a word repeatedly that I don’t know the short text for, I look it up, and notice the whole triplet.

Now that two of us know a good amount of short text, it’s becoming apparent that there are other ways we can use this knowledge. More on that later.