The tax office was among the first to adopt the computer. Since, tax evasion has become less possible. But despite their popularity, some people have complained that the novelty has been "under-used." "It is like using a space shuttle for household delivery of milk," remarked one critic.
Under-used or not, this Western invention does not get along easily with Chinese characters. Some users have to translate all their Chinese input into English before they can process the data on the computer. To solve such problems, Chinese scientists have been studying the structure of Chinese characters in hopes of keying in the characters directly.
Chinese characters are pictographs and ideographs, and there are more than 10,000 of them, each independent of the other. One solution is to designate the characters in digits and print them like pictures.
There are now, mainly, four ways to key in a Chinese character, each with its own advantages and disadvantages. These four are the codified method, corner-designation method, matrix method, and pronunciation method.
The codified method is based on the telegraphic code book, commonly used. Each character is given a four-digit number from 0001 to 9999. Since each character is represented by its own number, no repetition occurs. But making use of 10,000 such code numbers is no easy task. If this method is adopted, only very well-trained people will operate Chinese-language computers.
The corner-designation method involves two systems. The first, such as the one in use by WANG, is called the three-corner system. Each Chinese character is considered as a square, and can be designated by three of its corners, giving each corner a two-digit number. Except for 99 basic symbols, the operator does not need to commit other numbers to memory. He can learn the trick in a few days and master his trade in a few weeks. But there is a three-percent repetition factor-one number denoting several characters. All the characters denoted with one set of numbers are flashed out on the screen; the operator makes a choice.
The other system is called the four-corner method, and is based on the popular invention of the late Professor Wang Yun-wu, who revolutionized the indexing method of the Chinese dictionary by designating each character with four digits. The customary way for locating a Chinese character in the dictionary is based on the radicals and the number of strokes. Because Professor Wang made dictionary checking so much easier and faster, he was nicknamed "Mr. Four Corners."
According to this method, each corner of the square character is designated with a digit. The trouble is the same: a four-digit number may represent several characters.
The matrix method is like the one employed for Chinese typewriters. All the characters, sometimes more than 10,000, are arranged systematically on a matrix for the typewriter to choose. It is cumbersome, but no repetition occurs.
As computerization burgeons, benefits of simplified Chinese-language processing loom ever larger
Another method is based on the pronunciation of a character. On the average, a Chinese man-in-the-street can accurately pronounce only 3,500 characters -far from enough. If one has a heavy provincial accent, he will certainly be a born-incompetent computer operator. Besides, one pronunciation may cover dozens of characters. The sound, yi, for instance, represents 99 characters, including "one," "clothing," "barbarian," and "different." They are as unrelated as fish, flesh, and fowl.
Anyhow, the pronunciation method has served a very good alternate purpose —checking for telephone numbers. In Taipei, the citizens seldom consult a telephone directory when looking for a number. They just dial "104" and tell the telephone office girls the names of the people or firms they want. They will get their numbers in less than 5 seconds -all because of a computer set up for the pronunciation method.
Unfortunately, not everyone can pronounce accurately, so the digital method is still the' best solution. But, then, nobody knows how many Chinese characters are used everyday. The Ministry of Education announced 4,808 commonly used characters, but according to people involved in data processing over the last three years, 16,000 characters have been used.
The processing of Chinese characters is, also, an added function for ordinary computers. To cope with the specific requirements, local software scientists have added some new functions, such as arranging the characters in vertical lines.
Because the digital data must be converted to pictographs and ideographs, a set of Chinese characters needs to be stored on a disk. Each character is printed as a picture formed by 16 x 18 points or 24 x 24 points. It looks ugly. As the Chinese consider their characters an art form, they are not accustomed to characters made up of dots.
The more points used, the more beautiful the character, but also the more computer bytes used in the output. The output of a character in 16 x 18 points requires 48 bytes, and a 24 x 24 point character requires 96 bytes. This means slow processing.
As the Republic of China still cannot produce Chinese character printers, imported picture printers are used. Chinese scientists are now studying the manufacture of a "Chinese character pattern generator." If they succeed, it can be added to a picture printer and the printing speed can be increased remarkably.
Since the development of the Chinese character computer is still in its burgeoning stage, it is still hard to determine which key-in method is better. But despite the formidable difficulties, Chinese scientists are working hard to improve the efficiency of processing information in the Chinese language.
The government is also paying close attention to this development work in the hope that, someday, the Republic of China may become the leader in this appropriate field.