In the courtly literature of Murasaki's day, people are referred to by their ranks and titles rather than by name. As they are promoted, their titles change. Most western translators follow the convention of choosing one referent for a character in the story and using it consistently through the work—although Royall Tyler has chosen to be faithful to the original by using characters’ titles. In my novel, I have chosen to call people by their personal names: thus Fujiwara Michinaga, the powerful regent for several emperors (sometimes called "His Excellency" in English translation) is Michinaga throughout.

   Another problem with names is that for girls of non-royal birth, theirs were seldom recorded at all. We do not know what Murasaki Shikibu's name was. She was a member of the Fujiwara clan, as were a high percentage of courtly families. "Shikibu" refers to her father's position as Senior Secretary in the Bureau of Ceremonial (shikibu-shô) early in his court career. "Murasaki" (the word denotes the color purple) comes from the name of a character in her Tale of Genji. But what did people call her before she wrote the Tale? I give her the childhood nickname Fuji (wisteria) because some Heian sources mention her as Tô Shikibu, "Tô" being the Sino-Japanese reading of fuji. She undoubtedly had another official personal name, which some scholars think may have been Takako, written with the character meaning "incense" plus the feminine ending -ko. Her daughter Katako (or Kataiko, according to some), on the other hand, has a number of better-preserved names, which accrued to her in the course of a lifetime of garnering rank and literary prestige. She appears in historical references as Echigo no Ben (referring to her grandfather's posting to Echigo Province late in life), and finally as Daini no Sanmi (a title alluding to her husband and courtly rank). "Kenshi" is the Sino-Japanese way of reading her personal name, Katako.

   This convention, too, reflects the pervasive fact that written Japanese is derived from Chinese logographic forms. In the eleventh century (as today, for that matter) a character may be pronounced in native Japanese or it may be read in a Japanese approximation of its Chinese sound. Long-established convention dictates which reading one uses, although proper names are rife with idiosyncrasies. The same characters used to write "Shôshi" (Sino-Japanese) are pronounced "Akiko" in pure Japanese. This applies to the names of all of the imperial ladies. In the Japanese language text, one simply writes the characters and readers can understand them visually without needing to know precisely how they were pronounced. It is possible that alternate readings were used on different occasions, raising again the problem of multiple names for single persons. Arthur Waley, the illustrious first English translator of the Tale of Genji, favored pure Japanese renditions. Most scholarly translators now prefer the Sino-Japanese. Arbitrarily, I chose the latter for women's official names but used Japanese pronunciations for the personal names of women without royal title.