Ni hao salaam nemestay shalom, howdy buddy tovarish huzoor Mein Leiber Kamerade

I pick up language but in a geometrical progress. When I begin I lag behind. For example, it took me an hour to learn the Chinese character for Shih, since I fell into a brown study on how this character resembles my own Princes when they were little: it was a baby with arms outcast for a hug, and toes turned up, as it were in that baby jammy which the kids used to wear.

And, teachers of language irritate me with exercises taken from "real life". Unfortunately, it appears, to teachers of language, "real life" is banal.

In my experience, we don't learn language in banal situations but instead in transcendant experience as when our charioteer turns out to be Krsna, or our cabdriver a god. St Peter and the Apostles, on Pentecost, learned to speak in many languages not because they were full of new wine, but through a divine intervention.

The geometrical process, however, causes me to eventually start forming sentences of rapidly increasing complexity, while at the same time ignoring common rules, and creating a continuum of emotions from amusement to outrage.

We laugh at the tourist who says jay swiss une Americane in France, and the flip Frog would be apt to reply, pas de merde, Harry Dickson (no shit, Dick Tracy). Old Hundred Names in China is amused by the rich tourist who painfully enunciates wo je dao Guangdong du in China only to end up in Harbin. I would hesitate to do so, for to do so creates fear, and the fear creates mutually exclusive and hostile communities of great mockers of "inferior" speech.

See Tom McArthur's book, the Oxford Guide to World English. English is no longer "proper". Instead, people speak the way they want to and it works: for example, the "proper" pronunciation of "Australian" in Australia is "strine".

But perhaps the most irritating thing about language instruction is that it teaches rules which are immediately violated.

CCTV has a program on Conversational Chinese which stresses the need for correct tone and says that Ni, Hao MUST be pronounced with Third Tone.

It then cuts to a video, typical for language instruction, of those dutiful *haute bourgeois* swots I recall from my struggles with French, who always are sure to be working hard for examinations and who take proper ski-ing vacations at the proper times: unlike real students of my experience at DeVry who get busted for Driving While Black on the way to my exams and don't take vacations: unlike yours truly, who had a tendency to arrive late or not at all for examinations and takes what the French would call vacations perverse: Paris in November during the first Iraq war: Hong Kong in August.

And I remark of the dutiful swots in the CCTV programme that they immediately violate third tone. Guy clearly says nee, ha in first tone.

I realized the problem, I think, immediately, based on a nodding acquaintance with De Saussure such as is gained from reading random. For the 1915 author of A Course in General Linguistics shows how language is not a set of "hard" signifiers, in a fixed relation with chunks of reality, but a system of differences.

"Cat" names that animal, so beloved of cat-lovers and so despised by dog-lovers, which forms a felonious connection between a dinner bowl and a litter box which as the man of the house I had to clean when married. It "means" cat and not rat because of the DIFFERENCE in the first letter, alone.

Like binary codes in a computer, the system is completely arbitrary (sometimes in exploring "microcode" you discover dizzying situations where 1 means 0 and 0 means 1, and there is also the dizzying "big endian" problem in Intel systems where the most significant digit occurs last, for no reason other than convenience and making me crazy until I write a utility to dump in the "proper", G-d-given, order).

I believe that the correct pronunciation of the third tone is critical when "hao" is followed by ANOTHER WORD, such as the person's name. Whereas colloquially, one is permitted to trail off, indeed in the modern, industrialized world, demotically required; for one feature of ordinary life in the 24/7 world is impatience with Fancy Dans who insist on "correct" pronunciation in excess of the needs of communication.

I've read a claim that it is probably impossible to recognize hand-written Chinese characters, although we've partially mastered the recognition of the vastly-fewer-in-number letters of the Roman alphabet. And in handwritten Chinese characters I see the same trailing-off effect, the same failure to "dot the i" when there is simply no need to do this.

Calligraphy considered artistic in the Hong Kong museum is extremely free-flowing, trusting the reader to complete the letter. I can see its beauty, but at the same time, whether writing Roman letters or the few Chinese characters I have learned, I insist on writing in a blocky and machine-like style, in part because I learned lettering in an engineering drawing class, and during the punched card era, had to "code" programs on preprinted forms clearly for the keypunch operators.

If DeSaussure is right, and also if some Clever Dick hasn't in fact developed a system for recognition of Chinese characters when hand-written, the impossibility IN GENERAL may be based on the fact that if language is a system of differences all the way down, a computer would never get started on an accurate translation.

If we "know" that a hand-written letter is one of 26 Roman characters, it is a trivial computational task to find a small set, usually of one member, of the "best" characters. But it seems that common written Chinese characters comprise 3000, which means that final sets of "best" characters would have 30 characters or more.

The problem, like chess, may be overcome by brute force application of Moore's law (increasing raw power combined with stupid and straightforward search algorithms, which are used by most winner chess programs). However, questions have been raised about Moore's law: it may run out of gas as we approach physical limits.

All these problems with natural language, a Wittgensteinian form of life, made IBM 1401 Symbolic Programming System perversely attractive to me in its conformance in fact with traditional expectations of language as a determinate code. In a form of low-level autism which drove girlfriends crazy, I would retreat into this world.

It also drove cats crazy. I was loaned a boarder cat. One midnight dark and dreary as I pondered weak and weary over the mystery of the seventh bit, this cat went crazy.

Cats, of course, are naturally mad. But this cat went into Cat Crazy where the beast throws himself repeatedly against the wainscotting.

The cat was clearly used to a mistress who would go to sleep at a proper hour so the cat could sleep on her face. And, in these pre-PC days, my work provided no vent on which the cat could warm her sloppy parts while meditating on the great secret that cats meditate on, whatever that is.

Language is reality. Computer language is by comparision just a trivial part.

"Thou taught'st me language: and, my profit on it, is I know how to curse".

Strangely, even knowledgeable readers of Shakespeare's The Tempest will think this is Caliban, speaking to Prospero: in fact, he's talking to Miranda, who despite her shrinking and compassionate nature, is very hard on Caliban in this scene: her speech to him starts "abhorr'd slave".

Caliban, of course, had language before Prospero: he learned it from his Mum, the witch Sycorax. But our natural instinct is to regard languages and codes we don't know as "gabbling like a thing demented". Only gradually do I begin to appreciate the subtlety of Mandarin, which is already becoming rather pleasant to hear: now, it is beginning to sound to me high-class in comparision to Cantonese. It is unfortunate, of course, that I gravitate towards the center so when it is the fashion to chill with the homes.

In programming, Miranda's ethnocentrism becomes the way in which programmers will attack good code based not on some shared vision but on the basis of their bad habits, and the way in which overworked software managers are at times impatient with good praxis.

Which is why it is probably a bad idea to start with Visual Basic, since this language does teach bad habits including turning off global checking to get things "done" (ill done, that is, and done to other's harm).

We can make judgements about computer languages that we cannot and should not about human language, for if Chomsky is right, all humans share a core set of concepts and can communicate. For example, a German immigrant autodidact in 19th century St. Louis, Henry Brokmeyer, after importing Hegel's Logic to the Middle West, was last seen teaching Hegel to Native Americans in Montana.

Somehow I am certain that he was chilling with the homes here, for the Native Americans probably had fewer mental barriers, of the sort erected by science worship at latter day American universities, to a philosophy of time and change. Perhaps Manitou is naturally a World Spirit.

Of course, ten years after Brokmeyer's contact, these same tribes were rounded up and forced to speak English only in compound schools.

Brokmeyer sounds like my kind of guy, a German-American trouble-maker from before the era of World War I, in which anti-German riots frightened my father's tribe into silence. He was thrown out of Brown University and represents a buried tradition in American philosophy: the survival through immigration of Continental traditions in the Middle West, where heartland academics don't accept the Pragmatism, and science-worship, of the Eastern universities.

For example, in Evanston, Illinois, I discovered a nest of fire-breathing German feminists in the Christa Wolff tradition (she'a an author from the former East German zone who drives neo-conservatives out of their mind). To this day, Hegelian reading and discussion groups plow their way, in Chicago (at the Open University of the Left and elsewhere) through Hegel's Logic.

Of course, such phenomena are invisible in the East and in California, where they imagine that we're all Boilermakers from Purdue...even though Purdue's TECHNICAL contribution to early computer science is also neglected, which was why I mentioned Purdue's early work in Fortran in my book.

Of course, I am ambivalent about the Midwest. It is quite narrow in many ways: but I find the same narrowness all over the world. The French think everything worthwhile is French, and China, although it works hard to get over its xenophobic self-sufficiency, has long felt it can go its own way.

To my father, "Cleveland was the world", a metropolis...and in fact, it's always had a first rate art museum and symphony. Its airport is a bit of a joke, a sort of landing strip by the lake where a 747 would have to use emergency procedures to land, but Cleveland was a railroad city.

Hamlet said "I could be bound in a nutshell and count myself king of infinite space" in response to Rosencrantz, or was it Guildenstern, who advised him to get out of Denmark in order to provide scope for his ambitions. But I needed to get out more.