For all his emphasis on the subconscious, Tanizaki himself was a very self-conscious technician. Perhaps he thought a novel must have a form designed to engage the reader's conscious mind precisely because its contents made their appeal at a different level. In any case, his own novels are characterized by skillfully constructed plot and persuasive rhetoric, in sharp contrast to the uncanny, indefinable nature of their central themes and characters. (p. 71)

Tanizaki's concept of structure, as it emerges from the controversy [between himself and Akutagawa Ryūnosuke] is quite clear: a novel should have a tightly knit, skillfully woven plot. "Its components," he observed elsewhere, "should embrace each other so tightly that if one were to be removed the whole would collapse." Not many Japanese literary theorists have shared this approach. Japanese readers have always liked a loose, episodic kind of plot—if plot it can be called—far removed from what Tanizaki had in mind; they especially detested a plot that was constructed geometrically, like a classic French comedy. Tanizaki's notion of plot was also unusual in its preference for grandeur; plot, he thought, should be not only tightly knit but constructed on the grand scale. It should be like a long-distance runner with strong legs; it should demonstrate the writer's staying power.

Tanizaki's idea of plot thus turns out to be almost deterministic. Plot construction, ostensibly the fruit of the writer's conscious efforts, is determined willy-nilly by his physique. By his own criteria, Tanizaki's preference for a long, complicated story was instinctive, and his argument with Akutagawa had to end at this point because constitutional differences could never be reconciled. Once again, Tanizaki gave the subconscious an important role.

By and large, Tanizaki seems to have followed his own theoretical precepts about plot construction. His major works have plots that are considerably more complicated and more tightly woven than most Japanese novels. Few indeed of the latter have plots as complex as Tanizaki's The Whirlpool. Skillful storytelling contributes much to the charm of A Blind Man's Tale, "The Portrait of Shunkin," and The Mother of Captain Shigemoto. What Tanizaki termed the grandeur of a far-extending mountain range is seen in the plots of such works of his as An Idiot's Love and The Makioka Sisters. A Tale of Disarrayed Chrysanthemums reads like a popular adventure story; in fact, Tanizaki amused himself by calling it "a popular novel," apparently in reference to Akutagawa's charge that any work of fiction that tries to attract readers by an ingenious plot is a "popular" novelist not a "genuine" one. Some of his early tales—"The Thief," "Devils Talk in Broad Daylight," and "In the Street," among others—are plotted as carefully as detective stories; indeed, they are usually assigned to that genre. (pp. 72-3)

As a storyteller, Tanizaki was always extremely sensitive to the use of language…. His desire to do something to improve the quality of Japanese writing led him to produce The Composition Reader, a comprehensive guidebook to good prose. While intended for a broad range of readers, the book reveals a good deal about his own literary practice.

What was distinctive about Tanizaki's approach to the language of literature was his denial that any such thing existed. He even said: "I believe there is no difference between practical and artistic language." By practical language, however, he meant language that efficiently carries out its practical purpose, which is to make the reader understand the writer. The most practical language is therefore the most artistic. "If you think there is some art of speaking or writing reserved exclusively for the novel," he said, "read any one of our contemporary novels. You will immediately discover that it contains no sentence that cannot be used for a practical purpose, and that any sentence serving a practical purpose well is also useful in literary composition." More than anything else, Tanizaki believed, the language of literature had to be persuasive; to be beautiful or euphonious was of secondary importance. (p. 74)

According to Tanizaki, good style was a matter of two rules, both of them quite relevant to his general conception of literature. The first was not to be too concerned with the rules of grammar. The reason for this, as he explained, was that the Japanese language in its very nature was not very grammatical. The writer could turn this to advantage by cultivating a certain ambiguity, which Tanizaki found elegant; indeed, he compared the effect of a passage written with no ambiguity to that of rudely exposed thighs and knees. A passage that omitted as many words as possible, even to the point at which a strict grammarian would object to it, was at once graceful in impression and provocative in meaning.

The second piece of advice Tanizaki had for beginning writers was that they should cultivate their literary taste. In order to become a good writer, one had to be able to distinguish good writing from bad. But this was like distinguishing between good and bad wines; one had only one's own taste to rely on. Here Tanizaki was retreating to subjectivism, and he knew it. He still insisted, however, that there would emerge a semblance of objectivity if the reader had developed a refined taste…. Education had little to do with it; inborn taste, polished by experience, was all. Here again Tanizaki's distrust of intellect was apparent.

The Composition Reader also classified prose styles in terms of two main categories. The first included the "flowing style," the "laconic style," the "calm style," the "airy style" (a light, casual, unconventional style …), and the "craggy style" (a deliberately rugged, uneven style; Tanizaki compared it to the surface of a crag); it referred, as these terms imply, both to the mode of sentence construction and to the way in which one sentence followed from another. The second referred to vocabulary and idiom; it comprised the "lecture style" (normal written style; professors often used it in their lectures so that students could copy them verbatim), the "military style" (more polite than the lecture style; so called because typically a serviceman used it in addressing his superior), the "salutatory style" (even more polite than the military style; used on highly ceremonial occasions), and the "conversational style" (used in normal conversations). Though in both instances the classification involved no value judgment, Tanizaki made sufficiently clear which styles he favored. In the first category, Tanizaki seems to have had a natural predilection for the "flowing style," a style that, with its long sentences and carefully but inconspicuously engineered continuity, was like a smoothly flowing stream…. He was fond of it because he thought it suited the genius of the Japanese language better than any other…. In the second category, he seems to have been most attracted to the "conversational style." He liked it because, as he pointed out, this style had four main strengths: (1) expression was freer; (2) sentences could end in a greater variety of sounds; (3) the reader could feel the tone of the writer's speech and almost see his feelings and facial expressions; (4) the reader could tell whether the speaker was a man or a woman. Tanizaki seems to have especially liked this last fact. He observed that as a rule a male reader read a book in a male voice even when he was not reading aloud, but that if the book was written in the conversational style he would read it in either a male or a female voice, whichever the sense required. For this reason, Tanizaki recommended that writers use the conversational style more frequently.

Tanizaki's predilection for the flowing and conversational styles is clearly seen in his own writings. Though he was a very versatile writer, who could command a variety of styles with consummate skill, these two styles underlie most of them. The flowing style is the one in which he wrote many of his novels and short stories in his mature years; the most notable example of it is The Makioka Sisters. It is also conspicuous in The Cat, Shōzō, and Two Women, "Ashikari" (in which many quotation marks are deliberately omitted in order not to interrupt the flow of words), and "The Portrait of Shunkin" (in which most of the punctuation is omitted, for the same reason). Some parts of The Diary of a Mad Old Man are written in the "craggy style," and Chronicles of Our Peaceful Kitchen inclines somewhat toward the "airy style"; yet in both instances the overriding tone of voice is unmistakably that of the flowing style. The same style is latent in his earlier works, too—in "The Clown," for instance.

As for Tanizaki's use of the conversational style in his fiction, examples are too numerous to cite. The most striking one is The Whirlpool, which is told entirely in the peculiarly feminine vernacular of a woman brought up in the Osaka area. (pp. 75-7)

It has often been claimed that Tanizaki's style went through drastic changes during his long literary career. But his basic style does not seem to have changed much, when judged in the light of his classification of styles, since it always inclined toward the flowing and the conversational. Many of his readers would be willing to concede that there is a good deal of stylistic difference between "Tattoo" and The Makioka Sisters. Yet how many of them would be ready to say of the former that it approaches the "laconic style," the opposite of the "flowing style" in which the latter is written?… Likewise, Tanizaki's fondness for the conversational style was lifelong. If one thinks this style is not often used in his early works of fiction, one has only to be reminded that the young Tanizaki wrote a good many plays. Among his early stories, too, there are those that, like "Creation" and "From a Certain Protocol," consist solely of dialogue. It does not appear, then, that Tanizaki ever attempted to change his basic stylistic preferences, or that the attempt would have made any sense to him, since he saw style as a natural product of physique. (p. 78)

Despite all … [his] emphasis on the idea that literature is play—and play for a very large number of people—Tanizaki seems to have secretly harbored a belief in literature's didactic function. This can be seen, for instance, in the passage … where he said that beauty of form is ultimately the same as beauty of content, that physical beauty is essentially no different from spiritual beauty. Tanizaki seems to be implying here that art at its highest level is as capable as religion of bringing spiritual enlightenment to its devotees—more capable, in fact, since visible beauty is more readily appreciated than spiritual beauty. (p. 82)

Again, this is hardly a novel idea. But when we consider Tanizaki's daemonic ideal of beauty, it takes on some very strange applications. Can an admirer of one of Tanizaki's she-devils be led thereby to the realm of religious enlightenment? Tanizaki's answer was emphatically in the affirmative. In his opinion, art was valuable precisely because it had that function. In most Oriental religions, the worshiper had as a rule to be a good man to be saved, but art could redeem even a bad man, a worshiper of evil beauty. "I believe," he said, "that art is the only way by which an evil man can attain a realm of perfect liberation without becoming an entirely different person. While religion spurns evil men … art permits them to enter its realm, as long as they believe in it. This is so because evil is only of this world; in the other world there is neither good nor evil; all there is is beauty."

The argument gains in force if "this world" is equated with the conscious mind, and "the other world" with the subconscious. Religion teaches man to be good in this world; it wants him to discipline himself by his conscious effort. A good man in religious terms is a person whose conscious mind has complete control over his behavior, so complete that he follows the religious codes automatically. But some people (everybody, potentially) are more faithful to the subconscious; they follow its commands because they come from a deeper level and speak to deeper needs. Religion would label such people wayward and refuse to save them unless they reformed. Art, in contrast, would both accept and save them, because it understands these needs.

A characteristic instance of art functioning as redeemer is cited by Tanizaki in a short polemic piece called "The Censor." The literary work being censored here is a play called First Love, and the playwright (modeled after Tanizaki himself), summoned before the censor, vigorously defends himself by insisting that the play is not at all immoral, that in fact it serves a didactic purpose. The hero of the play is a young student who falls in love with a maidservant in his father's household. She, however, loves another man, and these two scheme to murder the student and grab the family estate. The student, well aware of the woman's evil design, willingly meets his death. "The flame that was burning in this boy's heart cannot be explained away by the logic of this world," the playwright remarks. "Nevertheless, the play does not give the spectators the impression that he met an unfortunate destiny. Or, even if some spectators should feel he was unfortunate, they will not believe his death was meaningless. They will be convinced that something remains after his death." The young student, Tanizaki is saying, is "saved" in the sense that he has been awakened to something higher than death, something that makes him readily meet his death. In this sense though the object of his passion was unworthy, the play is didactic, even highly moral, since death is overcome as truly by passion as by religion.

Some of Tanizaki's novels are "religious" in the same sense but more plainly so. Plainest of all is his last novel, The Diary of a Mad Old Man, in which the aged hero worships his young and beautiful daughter-in-law like a female Buddha. Keenly aware of approaching death, the old man buys a lot for his grave and prepares a plan for his tombstone, on which are to appear the young woman's footprints carved in the manner of the Buddha's. Once the plan is set, the old man is no longer afraid of death, since he can dream of his departed soul lying at peace under her feet. A similar identification of a beautiful woman with a female Buddha is suggested in The Whirlpool: here, too, her admirers are not at all afraid of dying when they dream of being with her after their death. A more pathetic case is that of Captain Shigemoto's father, in the novel named after his wife. This old man is robbed of his beautiful wife, whom he worshiped. In deep grief and agony, he tries every means of salvation available to him—he reads the poetry of Po Chüi, he drinks wine, he devotes himself to the practice of esoteric Buddhism—but to no avail. It is only his wife that can save him. His futile search for peace is awe-inspiring, and it is with relief that one finally reads of his son, Captain Shigemoto, being reunited with his mother, and so saving the old man's soul. However strange, even perverse, these emotions, they ennoble the characters who are driven by them. If the reader, too, is momentarily ennobled, perhaps he has gained from literature the best that Tanizaki thought it had to offer. (pp. 82-4)