Shadow Mages, Fictional Wizards, and Real Ninja
In which I argue that "historical ninja lore" isn't really about 'ninja'
In the previous post I tackled an argument best represented by this quote from martial arts instructor Stephen K. Hayes:
The historical lore of Japan’s ninja night warriors is full of tales centering around seemingly fantastic abilities that reach far beyond the capabilities of normal mortals. Of course, the exaggerated tales are based on roots of reality, which were then nurtured into the flowering of myth and misunderstanding.1
The argument, in a nutshell, is that there are lots of stories of ‘ninja’ with magical powers, that these stories derive from real-world ‘ninja’ activity, and that all of this constitutes evidence for the historical reality of the ‘ninja.’ This argument runs into problems once you grasp that what Hayes terms “historical lore” is more precisely termed “works of literary fiction from about 1700 onward.” Using fiction to support strong historical claims, as we discussed last time, is a pretty risky enterprise.
Today, though, I want to try a different and rather simpler angle of approach, which is to consider whether what Hayes calls “the historical lore of Japan’s ninja night warriors” does in fact feature recognizable ‘ninja’ in the first place. I’m not sure that it does; I think that the depictions of figures in these literary works don’t fit very well with the modern concept of ‘ninja,’ even by the definitions that the authors themselves seem to favor. I think what we have here is a premodern literary archetype - what I’d call the “shadow mage” - that’s better understood on its own terms, and that it’s not necessarily all that helpful to try to fit it into the mold of a concept, namely ‘ninja,’ that only came into being centuries later.
The Appliance of Science
It’s maybe worth noting that the “historical ninja lore” argument can be found in academic discussions as well as the pop-facing stuff. For instance, Polina Serebriakova and Daniel Orbach’s 2020 peer-reviewed article “Irregular Warfare in Late Medieval Japan” suggests that:
This image [of the ninja] first emerged in Japan in the seventeenth century through the Kabuki theater and other forms of folk art.2
Karl Friday, way back in 2005:
Ninja shows, ninja houses (sort of like American “haunted houses” at carnivals), and ninja novels and stories were popular by the middle of the Tokugawa period.
Literary scholar Richard Torrance, mentioning the Sarutobi Sasuke stories in an article about literacy and publishing in late 19th and early 20th century Osaka, likewise accepts the idea that Sasuke is a ‘ninja,’ writing of “the ninja Sarutobi Sasuke” and noting that one of the things that made Sasuke special was that:
The ninja had never before been treated as heroes. They were spies, liars, schemers, cowards, and all-around villains.3
The assumption underpinning all of the above quotes is that the ‘ninja’ was a recognizable and stable concept in Japan prior to the 20th century, and that it can be clearly identified as present in earlier works of fiction.
With all due respect to the above scholars, I’m not sure that’s a valid assumption.
One of the things that I think confuses a lot of people is that the term ninjutsu, or something similar to it, does indeed appear in some of the works in question, as we’ll see below. But it’s clear from context that this ninjutsu is not the same thing as the postwar martial art named ninjutsu as promoted by Hatsumi and Hayes and supposedly derived from the fighting techniques of the historical shinobi.4 Simply put, this ninjutsu means “magic” - in fact, in most pre-WWII contexts I would probably translate the term ninjutsu as “shadow magic” so as to keep the different meanings of the word distinct.
A more accurate way to put it would be that there are plenty of villains and heroes with magical powers in Edo-period fictional works, and that these powers are sometimes referred to as ninjutsu. But it’s a long way from there to the idea that ‘ninja’ in the modern sense - clans of highly trained spies and assassins of peasant origin, as most people understand the term - are in fact depicted in Edo-period fictional works. I would suggest that the figures we’re about to discuss be called instead “shadow mages” or something similar, and that “shadow mage” be understood as something distinct from ‘ninja’ - because it is.
So in a certain sense this will be an argument that revolves around definitions: What do we mean by “ninja,” anyway? - and linguistic change: Do we take into consideration how the meaning of the word “ninjutsu” changed over time? But it also appears that post-50s ‘ninja’ writers, both in Japan and in the West, weren’t always fully aware of these questions. ‘Ninja’ writers habitually - and you might say strategically - adopt extremely broad definitions of what a ‘ninja’ actually was, and it seems that in some cases the linguistic ambiguity worked to their advantage, as we see in Hatsumi’s children’s book:
When your mommies and daddies were children, it was thought that ninjutsu meant making a mystical sign and chanting an incantation, then changing into a giant toad or disappearing in a puff of smoke. But that is not what ninjutsu is. ‘Ninjutsu’ (also called shinobi no jutsu) is the art of using every branch of science - the human mind, physics, astronomy, geology, psychology, etc - to scout out the enemy, sow discord in his camp, and defeat him.5
The “toad” thing makes clear that Hatsumi has in mind either Tokubei from India, the subject of the print used as today’s splash image, or perhaps the hero Jiraiya, who also used magic to control giant toads:
And though Hatsumi isn’t explicitly claiming here that these premodern stories of shadow magic are evidence for the historical ‘ninja,’ you can see that it’s not that big of a leap from where he is to where Hayes wound up.
Nikki Danjō: Uses Shuriken AND Ninjutsu!
Since we’ve already introduced him, let us now consider Nikki Danjō, the main villain of the 1777 kabuki play The Precious Incense and Autumn Flowers of Sendai. Nikki Danjō’s thing is that he is a high-ranking samurai of the Sendai domain who is conspiring with a number of his fellows to subvert the succession plans of the lord of the domain for his own interests. His main moment of glory comes in the “under the floorboards” scene, where he is shown to have the power to transform himself into a rat, to become invisible, and possibly to fly as well, depending on how you interpret his exit. Here’s another performance of the scene, if you want to refresh your memory (1:20:48 or so):
Now, I have to concede that there are a few points in favor of the idea that Nikki Danjō is a ‘ninja.’ He uses a shuriken, for instance - albeit a throwing knife, not a star, since that’s what the term meant before the ‘ninja’ guys got their hands on it - and, fascinatingly, his magical powers are referred to as “shinobi no genjutsu” (‘the illusion magic of shinobi’) in this Meiji-era script of the play as apparently performed in 1785:
“The intruder that took the scroll thus appeared in all the middle of the chaos, as a great rat of unusual size. Truly, this must be the illusion magic of shinobi (masashiku shinobi no genjutsu). How strange!”6
So…he is a ‘ninja,’ then? A lot would depend on how we translate shinobi no genjutsu; it could indeed be “illusion magic of the shinobi” (i.e., associated with people known as shinobi) or it could simply be “illusion magic for infiltration” (i.e., its purpose). We might also ask whether a Japanese audience member in 1777 would have interpreted Nikki Danjō as what we now think of as ‘ninja,’ which to be honest seems unlikely, but it’s nevertheless possible.
My main objection, though, would be that mention of the word shinobi notwithstanding, Nikki Danjō doesn’t really look or act like what we would expect from a ‘ninja.’ If I hadn’t told you that Nikki Danjō is sometimes dragged into discussions of the ‘ninja,’ for instance, would you have picked him out as a ‘ninja’ from watching the above scene? My guess is probably not. You might have figured him for a wizard, sure, but that’s a whole different literary archetype that doesn’t automatically connect with ‘ninja’ at all.7
Just as importantly, Nikki Danjō’s character seems like a pretty poor fit even for the definition of ‘ninja’ that the several of the authors discussing him are using. Hayes tells us, for instance, that the historical ‘ninja’ were heroic guerilla fighters during the Warring States period (1467-1600):
It was only during the incredible sadness, violence, and oppression of the buildup leading to the tragic Sengoku Jidai Warring States Period [sic] that the warriors history would later call ninja were forced to rely almost entirely on the subtle realms of underground resistance to survive. Unlike the established samurai families who by law and cultural pressure were forced to conduct their affairs in alignment with set codes of action and speech…the ninja families were denied the right to legally defend themselves […]8
Given that Nikki Danjō is a powerful, high-ranking samurai long after the end of the Warring States period, he doesn’t fit at all with who Hayes seems to think the ‘ninja’ were. He seems, to be frank, much more like a conniving politician than an underground resistance leader. So either Hayes is wrong about who and what the ‘ninja’ were, or Nikki Danjō isn’t a ‘ninja’ by Hayes’ own definition.
Six years later in 1991, we find a rather sheepish Stephen Turnbull conceding much the same point:
[T]he play Meiboku Sendai Hagi […] manages to give us the best known portrayal of a ninja, or at any rate a magician, in the theatre.9
I hope you noticed the little side-step (“at any rate”) there. A few pages on Turnbull repeats the point:
It is doubtful whether the character of Nikki Danjō can be regarded as a ninja in terms of the definition we have adopted for this book, but his magic powers, attributed to ninjutsu, are stated very clearly.10
At the very least, then, we ought to exercise caution before going along with the idea that this malevolent sorcerer is a ‘ninja’ in the modern sense. We need, as academics would put it, to historicize the term ninjutsu, to understand how it might work differently in different historical contexts; and we also need to historicize the notion of ‘ninja,’ to consider whether it’s even meaningful outside of a 20th century context.11
But whichever side we land on, the above discussion is fundamentally about the development of a literary archetype; at no point in any of the above have we necessarily left the realm of fiction. Let’s say for the sake of argument that Nikki Danjō in the original 1777 Precious Incense play is indeed a classic ‘ninja’ in every imaginable respect. Let’s say he proclaims himself to be a ‘ninja’ in so many words, puts on the classic black outfit, throws star-shaped shuriken everywhere, talks about how the samurai hate him and his clan back in Iga for being peasants, whatever. That would be awesome from a literary perspective, because it would confirm that the modern concept of the ‘ninja’ was present in Edo-period works of fiction as well. But from an historical perspective it would tell us nothing concrete about the historical shinobi at all, because, as I have pointed out ad nauseam, fiction is by definition made up.
The above illustrates a critical methodological problem in ‘ninja’ studies, which is that an awful lot of its historical claims depend on the incorrect assumption that fiction always correlates to something in the real world. This in turn raises the question: how did this happen? When and where was this methodology introduced into the study of the ‘ninja’? The answer, I think, is that it’s been there pretty much from the beginning; it’s there in the work of the first modern ‘ninja’ writer Itō Gingetsu, all the way back in 1909.
So, it’s to Gingetsu and his writings that I’d like to turn next.
Stephen K. Hayes, The Mystic Arts of the Ninja: Hypnotism, Invisibility, and Weaponry (Contemporary Books, 1985), p. 133.
Serebriakova and Orbach, “Irregular Warfare in Late Medieval Japan: Towards a Historical Understanding of the “Ninja” The Journal of Military History 84 (October 2020), p. 997.
Richard Torrance, “Literacy and Literature in Osaka, 1890-1940” The Journal of Japanese Studies Vol. 31, No. 1 (Winter, 2005), p. 54 and 56.
Turnbull has a pretty good discussion of this point in his 2017 Unmasking the Myth (“The Magic of Ninjutsu,” pp. 115-121, esp. pp. 120-121.
Hatsumi, Shōnen no tame no ninja, ninpō gahō (Akita Shoten, 1964; my copy, 1977), p. 10. Japanese: みなさんのおとうさんやおかあさんが子どものころ、忍術というのものは、九字を切ってじゅ文をとなえると、大きなガマガエルになったり、煙がでて姿が消えるものたと考えていました。けれど忍術というものは、そんなものではありません。忍術とは（忍びの術ともいう）あらゆる科学を利用し、人間の心、物理学、天文学、地質学、心理学などを利用して、敵のようすをさぐったり、敵の陣地をかきまわして、敵をうち負かす術です。
Matsu Kanshi et al, Meiboku Sendai hagi Meisaku sanjūroku kasen [36 Selected Masterpieces] (Kin’ōdō, 1891), p. 101. The performance being transcribed is dated to the fifth year of Tenmei (1785), around eight years after the play was first performed.
Or, given what watching kabuki without any notes or interpretation is like, you’d simply have been left wondering what the hell you just watched. Trust me, I’ve been there.
Even scholars more sympathetic to the general idea of the ‘ninja’ acknowledge that a lot of heroic figures in Edo-period fiction are a pretty poor fit for the concept of shinobi; see e.g. Yoshimaru Katsuya’s discussion of the origin of the kunoichi, where he notes that a lot of magic-using heroines just don’t look or behave like shinobi.
Hayes, The Mystic Arts of the Ninja, p. 9.
Turnbull, Ninja: The True Story of Japan’s Secret Warrior Cult (1991), p. 130.
Turnbull, Secret Warrior Cult, p. 134.
As parallel examples, many academic historians tend to avoid the terms geisha (preferred: “courtesan” or “female entertainer”) or “samurai” (preferred: bushi or “warrior”) in a premodern context. The reasons are that the terms themselves can be imprecise (technically geisha were supposed to be musicians or entertainers and to be legally distinct from courtesans, for instance, but that distinction is often forgotten in modern usage), and the terms sometimes have a whole host of modern associations you don’t necessarily want to bring into your discussion. Actual bushi didn’t always think or act in ways most people think “samurai” thought or acted, for instance. I think a very similar argument could be made for ‘ninja,’ that it’s a modern concept with a whole lot of baggage, and that that baggage often gets in the way of examining the historical shinobi.