Mondegreen 2000

Does anyone out there in the blogosphere know if there’s a term for the computational equivalent of the mondegreen that occurs when a computer misrecognises words during OCR? Surely there must be one..

The evolution of keyboards and programming languages

Lines of BASIC printed out on a typewriter

I was playing around with my Underwood No 5 typewriter a while back, and jokingly typed out a couple of lines in BASIC, which I photographed for my Facebook and Twitter avatars. I thought about typing out some HTML, but then realised that the typewriter lacks the < and > keys, making tags entirely impossible.

It occurred to me that a great number of programming and markup languages make use of the < and > keys, even if my little BASIC program didn’t. As such, it would seem to me that the < and > keys (as well as a few others absent on my Underwood keyboard) must have appeared on the standard computer keyboard fairly early on. As my Underwood was produced in the late-1800s, well before modern computers, it was adapted to entirely different uses than my modern MacBook and PC keyboards. Many users new to the typewriter will be shocked to learn, for example, that most older typewriters entirely lack a 1 key (using instead the lowercase l), but up until the mid-20th century, this was simply the way things were. In fact, a typist from the early-20th-century would be shocked to see a typewriter with a 1 key.

For reference, the keyboard layout to my Underwood is above. As we can see, it is a much leaner keyboard than those we are used to today. Not much had changed a few decades later when my Royal was produced, and barring a few minor changes, the keyboard to it (below) is much the same as the keyboard of my Underwood.

From what I can tell, early programming languages did not require the < and > keys. While I am not too familiar with FORTRAN or assembly, the Wikipedia page for FORTRAN doesn’t seem to have any < or > in the examples of code, and assembly functions in terms of hexadecimal values paired with mnemonics, so programs written in assembler wouldn’t need anything beyond the standard typewriter keyboards.

The very first computer my family had was a Commodore 64, which we got when I was 3. My memory of the keyboard layout being fairly fuzzy, but I remember being surprised when, as a teenager, I dug it out of the basement and discovered that there were only two arrow keys: up/down and left/right. I looked up an image of the C64 keyboard online, and although there were some interesting surprises (the 2 key, for example, is paired with ” and not the @, as is the case with typewriters and modern German keyboards), the C64 does have the < and > keys.

I decided, then, to try and explore some earlier keyboards to see if I could get a better sense of when the < and > keys appeared. This IBM 26 printing punch keyboard from 1946 does not have the < and > keys, but this IBM 129 from 1971 clearly does. I’ll obviously have to do much more research into this to narrow things down a little more, but this is a interesting start.

I’m not entirely certain what I’ll find while researching this, but I am curious as to when the keys appeared, and at what point symbols like < and > were introduced into programming languages. I have a hunch that the adoption of the ASCII in the 1960s standard might have something to do with the appearance of < and > on keyboards, and that they those symbols didn’t appear in programming languages until after they were standard on keyboards, but I am clearly going to have to look much deeper into this before I reach any sort of definite conclusion..

See above, see below

Nathan Altice asked on Twitter why authors say that text will appear “below,” which led to a discussion between Nathan, Paul Benzon, and myself.

After I got home from watching a live satellite screening (the future is amazing!) of the Hot Docs presentation of Indie Game: The Movie (which I highly recommend for all those interested in videogames and the people behind them), I started looking a little more into the data that I’d found. As I suspected, the early blips in the ngram were largely due to errors in metadata, appearing mostly in footnotes added by modern editors in a later edition of the work. That said, some of these blips did contain applicable uses of the phrases in books like The Fredrician Code, published in 1761, where the phrase “see above” appears, referring to clauses of law that appear earlier in the book, much the way we would use the phrase today. That said, these blips are relatively rare, and it’s not until the 1800s that the phrases really start taking off and approaching levels of use that we see today.

I thought that it could be that pre-1800s there were more uses of the Latin equivalents “vide supra” and “vide infra,” but rather curiously the ngram of these phrases wasn’t that far off of the ngram for “see above, see below,” all things considered—they both start right around the 1800s and build from there (although the Latin phrases seem much more erratic in their uses, and fall off in recent decades). I also tried searching for the abbreviations “v.s., v.i.” and “v. s., v. i.,” and got precisely zero results each time (something that I suspect is an error, as I find it hard to believe that the abbreviations were never used in print). Even if I were to find instances of the abbreviations, however, it seems highly unlikely that they would differ too much from the full phrases, as writers would be unlikely to have spent the 1700s using the abbreviation only to switch to the full phrase in the 1800s.

Ultimately, this is just an initial search, and there’s a lot more we could look into, especially if you consider works in other languages, as well as in time periods not covered by Google Books. I would encourage you all to add comments to this post—this seems like the type of question that would benefit from the knowledge of many scholars of diverse interests..


Matt Kirschenbaum recently tweeted this fantastic version of the dance hit “What is Love?” played entirely by floppy disk drives. Enjoy.

And once you’re done with that, I highly recommend checking out this slightly more experimental (although nonetheless listenable) version of Radiohead’s “Big Ideas (don’t get any),” this time played on an assortment of computer hardware.