Brief History

Newspapers and other companies began telegraphing Japanese text and developing kanji-character teletypes in the mid-1950s. Multilevel-shift keyboards were used to enter kanji characters into the teletypes. By the mid-1960s, drum-type kanji-keyboard transmitters were developed.

In the 1970s, a number of companies developed input devices that used tablet and pen interfaces. These devices consisted of a tablet arranged with thousands of small characters. A stylus pen was used to touch the appropriate character, which then electrically issued and input the corresponding character code. Although the entry speed was slower than a multilevel-shift keyboard, the touch method was suited for unskilled operators.

The two-stroke input method created a 1:1 coding system that paired two-letter or two-kana-character combinations with kanji characters. Thus, each kanji character needed two keystrokes to be input. The Rainputto company announced the two-stoke method in a paper in 1972. With a special key arrangement, higher input speeds were possible, but the method was suited only for skilled operators and never gained widespread acceptance.

Toshihiko Kurihara and others at Kyushu University applied for a patent for a kana-kanji input method in 1964. The patent proposed a basic method of kana-kanji conversion in which sentence fragments were entered in kana characters and then converted to kanji characters through lookups in a word dictionary, syntax analysis, and semantic analysis. This method is said to be the precursor for today's kana-kanji conversion methods. Based on this work, Yoshiaki Kurosaki at Oki Electric prototyped a kana-kanji conversion system in 1967. In the 1970s, universities and corporate laboratories put considerable effort into researching and developing practical kana-kanji input methods.

Toshiba, with Kyoto University's leadership, began researching syntax analysis of Japanese sentences in 1971. Sharp exhibited a prototype Japanese language word processor with a kana-kanji conversion method at a business show in 1977. Toshiba used the results of its kana-kanji conversion research to develop the JW-10 Japanese language word processor, which it announced in September 1978. Toshiba exhibited the JW-10 at a data show in October of 1978 and began shipping the word processor in February 1979. The JW-10 was priced at ¥6.3 million.

From this point on, many companies began announcing Japanese language word processors with a variety of conversion methods. In May 1979, Oki Electric announced the OKI WORD EDITOR-200, with a display selection system to input kanji characters one character at a time, and in September of the same year, Sharp announced the WD-3000, which had a tablet interface with a full character layout and a touch-pen input device. In May 1980, Fujitsu announced the OASYS 100, which displayed and selected single words using a thumb-shift keyboard, NEC announced the NWP-20, a tablet interface with a full character layout, and Hitachi announced the BW-20, which used a two-stroke input method.

Canon announced the Canoword 55 in December 1980, and Ricoh announced the Report 600, Matsushita Electric (now Panasonic) announced the Panaword 1000, and Casio announced the WP-1 in May 1981. Sanyo Electric, Fuji-Xerox, and Yokogawa Electric all entered the market in 1982. Later, electric appliance manufacturers and office device manufacturers would also announce word processors.

Initially, each company's Japanese language word processor used a different conversion method. But over time, as research into natural language processing advanced, the methods converged on the present kana-kanji conversion input method..

In August 1984, Fujitsu kicked off a price war by marketing its OASYS Lite personal word processor for just ¥220,000. In July 1985, Toshiba priced its Rupo JW-R10 laptop-sized word processor at ¥99,800. The lower prices fueled the growth of Japanese language word processors, and the industry's total production figures reached one million units for the first time. In November 1986, Fujitsu announced the OASYS 30AF, a personal word processor with a large LCD screen (40 characters x 21 lines) for ¥248,000. Other companies came out with a wide range of alternative word processor designs as well, resulting in more than two million units being produced that year. In the first half of the 1980s, manufacturers worked on the keyboard design itself to facilitate easier character input. For example, Fujitsu included a thumb-shift keyboard on its OASYS line, and NEC announced the PWP-100, which came with the M System keyboard

The performance levels of word processors grew steadily more sophisticated, with the release by Sharp in May 1987 of the WD-540 with an AI conversion function and the release by Toshiba in February 1988 of the JW-1000AI with an AI-learning function. A record 2.71 million units were shipped in 1989, bringing the total number of word processors sold to over 10 million. Annual sales fell off in subsequent years.

Personal computers started growing in popularity in the 1980s, and personal computers began shipping with DOS/V, which supported Japanese language input, in 1990. Prices also fell below the ¥200,000 mark. Word processors continued to sell, however, with the total number sold reaching 20 million in 1993, because they remained inexpensive, relying on cheaper CPUs than computers, and were integrated with printers. Their dedicated document-creation software and keyboards were also easier to use and computers had yet to find their way into households.

With the release of Windows 95 for personal computers in 1985, personal computers took off in homes and offices alike because of Windows' user-friendly graphical user interface and its support for the Internet and other communications. Because of the difference in economies of scale between word processors and personal computers, word processors could not compete on price any longer. So although the total number of word processors sold passed 30 million in 2000, only 260,000 units were sold in 2000, less than 10 percent of the peak in 1989. Matsushita Electric (now Panasonic) withdrew from the dedicated word processor market in December 2000, and it was soon followed by many other manufacturers. The last word processor manufacturer, Sharp, finally ceased production in 2001.