1. What the Tiny Triangle Shows About Brush Writing
When you write the character for "one" with a pencil, you usually draw a line of even thickness. However, when writing with a brush, you press down at the start, pull the brush through, and stop sharply at the end. Ink pools where the brush stops, creating a bulge. The Ming typeface, widely used in Japanese printing, is based on the brushwork of regular script and the print styles from the Song dynasty. Its basic features are thin horizontal lines, thick vertical lines, and a triangle at the end of horizontal strokes. This triangle is often explained as a printed representation of that ink bulge where the brush stopped [1].
Although the name includes the character for "Ming," it is not certain that this style was born during the Ming dynasty. Some explanations state that it began in the Song dynasty and was refined during the Ming era [1]. The question remains: how did this specific brush shape enter Japanese printing?
2. What Movable Type Is and Why It Was Invented
To send the same text to many people, copying it by hand every time is too slow. So, a method was developed to create characters as physical objects that could be used repeatedly. One of the oldest surviving printed items in Japan is the *dharani* sutra, completed around 770 CE and placed inside small pagodas. Of the 1 million small pagodas made, only about 40,000 to 50,000 remain at Horyu-ji Temple today [3]. Whether these were made with woodblocks or movable type is not specified here.
In museum exhibitions, the term "movable type" does not refer only to metal letters. It includes photo typesetting fonts and digital fonts—anything that repeats the same character shape [2]. This article uses that broader definition. However, using movable type in Japanese presents challenges that do not exist in many other languages. Let’s look at how this worked in practice.
3. How Print Shops Picked Characters from Thousands of Types in 1–2 Seconds
Print shops that used movable type had shelves holding individual characters. For just one font style and size, there were reportedly about 2,500 types available [4]. This is not the total number of Japanese characters, but an example of what was on a shelf. The process of picking the needed characters from the shelf while reading a manuscript is called *bunsen*, and arranging them into a block is called *ueji*. Skilled workers could pick one character in 1 to 2 seconds and process about 1,500 characters per hour [4]. Common characters like "no," "shi," "te," and "wo" were prepared in hundreds of copies.
The basic set of hiragana and katakana consists of 46 characters each. Additionally, there are 2,136 *joyo kanji* (common-use kanji) used in newspapers and textbooks. The number of common-use kanji can change when the official list is revised. The more types available, the more shelves and time are needed to find them. To eliminate this effort, some people thought of using light instead of metal.
4. The 1924 Patent for Japan's First Photo Typesetting Machine
On July 24, 1924, Nobuo Morizawa and Shikichi Ishii applied for the patent for Japan's first photo typesetting machine (Patent No. 64453) [5]. Morizawa was motivated by the hassle of handling movable type in the printing department. The first photo typesetting machine worked by shining light from below a character plate that transmits light, passing it through a lens and shutter, and exposing film inside a dark box. The film was then developed to make a printing plate [5]. Instead of arranging metal type, this method captured characters with light. Note that the first machine and later improved versions used different character plate systems; this description applies to the first machine.
The important point is that the "original characters" drawn on the plates were still written by hand. Even though metal molds were replaced by photographic film, the shape of the letters continued to be drawn by human hands. So, what happened to these original characters when the computer era arrived?
5. How Digital Fonts Revived the Look of Photo Typesetting
In the 1980s, the era of computers and printers began. Because Japanese requires thousands of characters, unique formats were needed. In 1987, Adobe and Morisawa signed an agreement on font licensing and production technology, and the first Japanese PostScript printer was released in 1989 [7].
The first Ming typeface as a Japanese font usable for Desktop Publishing (DTP) was Ryumin, released in 1989. Hiragino Mincho appeared in 1993 and was included in Mac OS X in 2000 [8]. Be careful not to confuse dates: Ryumin was announced as a photo typesetting font in 1986, which is different from its digital release year.
In 2024, it was the 100th anniversary of the invention of the Japanese photo typesetting machine. Morisawa and Sha-Ken announced a project to revive Sha-Ken’s fonts, such as Ishii Mincho, Gona, and Nal, in OpenType format. They planned to start offering 43 fonts in autumn 2024 [6]. Since this was a plan at the time of announcement, check official websites for current availability. The shape representing the brush stop remains in movable type, photo typesetting, and digital formats, and it is still on our screens today. Try looking at the text in a book or on your screen again.
6. Try It Yourself: Compare Ming and Gothic Fonts in Your Own Book
If you want to investigate further, you can do this today: Compare the Ming and Gothic (sans-serif) fonts in a book you have. Check if the horizontal lines have triangles at the ends. Then, write the character for "one" with a brush pen or brush. Observe how the start and end shapes are similar to, and different from, the triangle in Ming typefaces.
At a library or museum, look for exhibitions or materials about the history of movable type, photo typesetting, and fonts. For official information, you can search for exhibition details and cultural heritage explanations on museum portals or the Cultural Heritage Online website.