I have realized in the past few years that I am not really kidding when I talk about having a 'magic language sense'. There is something about the time-worn repetition about true language that catches my attention when I see it, even if it's not a language I speak, or it's disguised as something else. There is a limit; blocks of Enigma text don't trip it, for instance, although intellectually I've inhaled enough about cryptography to recognize that's what it (probably) is. The format typically used by numbers stations says to me that there is a message there, although since it's probably an arbitrary correspondence code I have no idea what it says, or that it's not all padding and gibberish.

A big tip off that I'm actually noticing something, even if I don't precisely know what it is, is that the language sense works better when I have context, and generally best when I have large amounts of text to look at. Perhaps an equally big clue is that asemic writing trips it backwards -- I'm so used to all kinds of writing having that prismatic overlay of meaning that when it's not there, it takes me quite a while to realize that the scribbling is meant to resemble or evoke the impression of text, if I ever do.

Knowing that writing has a meaning and being able to decipher it are two different things. I finally got around to looking up Arabic calligraphy a few months ago, because I knew it said something and wanted to know how to figure out what. This sucker over on the right looks perfectly well like writing to me, although I'll be damned if I know what it says -- the artist was drunk as fuck (no, really), and calligraphic art involving ideographs is not typically meant to be read, as such. There are, like, writing parts in it, more than anything, and the line squiggles back and forth in ways that suggest the writer had some sort of stroke order in mind, but ultimately just couldn't be bothered.

Most asemic writing doesn't even tickle the language sense. The Codex Seraphinianus (left), beautiful as it is as a work of art, looks completely wrong. That is, in fact, the whole of my first impression of it, although after looking much longer I can piece together where the reaction comes from: The font is meant to emulate handwriting, but those letterforms are much more complicated than anything that could have evolved from habitually writing words as long as the ones seen in the text. Humans are much too lazy for that, honestly. There is a suspicious amount of repetition not just in what you get at the end of words -- and I say this as a native speaker of a language that only distinguishes one major case, one grammatical number, and one present-tense conjugation, and chooses to mark them all with an S -- but also in the beginning. I also note a few spots where one symbol occurs three times in a row, right in the middle of a word, which, for reasons yet obscure, is virtually unheard of in a regular orthography.

The Voynich Manuscript (right), I'm iffy about. If it's anything, it's an invented language -- it looks wrong, both to my eye and evidently to sophisticated cryptography analysis, for natural language. The letters are much more natural than the Codex, though, and it was clearly developed by someone familiar with writing and literacy in general. Compare to the examples on this page, for instance. It looks most like the Latin, which is medieval and generally appalling to read, which is understandable, since ergonomics wouldn't be invented for another few hundred years.

[Note that I actually cannot read all of the things in each of those above-linked thumbnails. The French one starts "It is the command of the Monseigneur de Grenoble de..." something something something, lots of stuff involving what I suspect may be the 'royal we'; the Italian is something about theology written to a priest; the Latin is on the topic of Rome, although mainly what I spot are a lot of conjugations of 'to be'. The German and Spanish, aside from the odd grammar word, are completely indecipherable. So if you ever wanted the comfort of knowing that I have limitations, here you are: I am powerless against terrible handwriting in languages in which I am not fully fluent. You're welcome.]

The thing that has come closest to fooling me into thinking it's real writing so far is Xu Bing's A Book From The Sky. The title page is reproduced at left. This may be cheating, somewhat, since I don't really read Chinese, but I do read Japanese well enough to plow through Meitantei Conan games in it, so that has to count for something. It's rare for me to look at a Chinese text and recognize no characters in it; a lot of the common words, especially prepositions and other grammar words, use characters that are still in use (sometimes in simplified form) in Japanese, even if the native Japanese prepositions are written in kana nowadays.

I couldn't say definitively that the characters used in that text aren't real Chinese, but my first instinct is to say they aren't modern Chinese -- the page looks, for lack of a better description, too busy. Characters generally evolve from complicated to simple, and while there are a lot of complex characters that remain in the traditional set used in Hong Kong, it would be unusual for even a text that short to contain that small a number of characters common enough to have evolved a simpler form over time. The fiddlier the character is, the older it's likely to be. Characters that remain fiddly to this day are usually ones that aren't used often enough for people to have gotten lazy about drawing them.