Ancient(6) |  Art(4) |  Gender(9) |  General(12) |  How to(5) |  Myth(10) |  Philosophy(17) | 

February 14, 2004

Etymology of Kami

The deities or spirits of Shinto are called kami but, while there are many homonyms (the words for "above", "hair", "paper", and "bite") the etymology of the word Kami is not known. For a long time it was though that it was from the word "kami" meaning above (written with the same character as "ue"). However with the discovery by a scholar, called Hashimoto in 1917, that there were more vowels in ancient Japanese, this theory was rejected.

The Vowels in Ancient Japanese
In Japanese today there are only 5 vowel sounds ("a, i, u, e, o," as we say in Japan) and till 1917 it was presumed that the same was true of ancient Japanese as well. However, Hashimoto noticed that there were clear rules for using two, distinct groups of characters to express what is today a single sound, "ko." The only, or at least accepted, explanation for this fact is that there used to be two different sounds. The same patterns were discovered for the vowel "i" and "e," meaning that there may have been 8 types of vowel in ancient Japanese, which we can write as: a, i1, i2, u, e1, e2, o1, o2.

These vowel distinctions were dying out quite rapidly. For example while there are two types of "mo" in the Kojiki (712), this distinction died out in the Nihongi (720, Nihonshoki) which was written only a few years later. And there are further complications in that some of the "extra" vowels are rarerly used, and then only in loan words. But it seems at least that a finer grade of vowel distinction did exist, and this enables us to make better judgements of etymology.

With the distinction, it became apparent that the "i" in kami = diety/spirit, and the i in kami = above are different.

The Ancient Vowel Sounds Continue Today
With regard to to the "i" vowel, the distinction between two types of "i" continues in present day Japanese. It seems that second type of "i" was somewhere nearer an "o" or "u" sound! I am not sure how to pronounce an "i" similar to "o" or "u" but I think that it was probably a "ui" or "oi" sound as we shall see.

The distinction between these two vowel sounds continues even today. For example, while "hi" as in sun, and "hi" as in fire are often thought to be of related etemology, the two were in fact written as "hi1" and "hi2," with two different types of vowel. The word for tree "ki" is another example of the second type, which we can write as "ki2." Even today these "i2" words are sometimes pronounced as "o." For example the light/shadow (the pattern it throws) of a fire is written with the character for fire and shadow, normally "hi" and "kage," but pronounced "hOkage" (with emphasis added). Similarly the word for the shade of a tree is written with the character for tree and shade, normally pronounced "ki" and "kage" but here pronounced even today as "kOkage." Similarly the "leaves of trees," is pronounced "kO no ha" as opposed to "ki no ha," berries, as "kO no Mi" and the dappled sunlight beneath trees, "kO-more-bi" (tree-overflowing-sunlight) The same "i2" is also found to change to "u," as in the god of the moon "TsukUyomino-mikoto" where the first "tsuku" is written with the character for moon, normally pronounced "TsukI" -- the "ki" was pronounced as "ku."

Returning to "kami," in the ancient "Manyougana" notation (i.e. in a notation similar to that used in the ancient Manyoushuu poem collection of poem, but this is in fact from the explanations of names of gods in the kojiki) kami is written as , as well as the character used today

In the form, the character for "mi" belongs to the i2 category. Hence, like other i2" words, kami is sometimes pronounced "kamu." For example, in ancient Japanese, "Kamikaze" (divine wind, and "suicide pilot") was pronounced "Kamukaze." Similarly there are the words "Kamu-sabi" (godly behaviour), and kamu-yo (the age of the gods. The Ainu for "kami" is "kamui," and it is thought that this is close to the ancient Japanese pronunciation. This also explains why "kami" often become "kan" in compounds, such as "Kannazuki" (November, the month when all the gods go to Izumo) and kanzaki place name, since abbreviation of "u" ("dekinu" > "dekin," "desu" > "des") and changing m's to n's ("yomu" > "yonda") happens all the time in Japanese. So what of the etymology of "kami"? First of all it is clearly different to that of "kami" meaning "above," since the latter is written with a "mi" belonging to the first category of vowel "i1" so this explanation of the etymology of Kami is incorrect.

Since in ancient Japanese the word for hair (now also "kami") was "ka" (as in "shiraga" or white/grey hair), and the mi uses i1, that etymological connection is also rejected. There is at the present time no accepted theory for the etymology of "kami." So let's make one!

Kami and Kami= Paper?
I used to think that kami as in paper, might be related or derived from kami (god) since there is so much paper at Japanese shrines: the ziz-zap lighenting strips or shide, the talismen (fuda), and fortune sripts (omikuji). Additionally, as I have mentioned before, I think there is a philosophical case to be made for the assertion that paper and god have something in common -- paper is the place where the sign meets the world.

There are several theories for the etymology of paper, one of them being "kami" (spirit) but the most popular theory is that is it is derived from the Chinese word "kan" (as in shokan, meaning letter). And once again, the "mi" of kami (paper) is i1, not the i2 of kami (god).

Kami as "bite"
By chance one oft the leaders in computerisation of the Manyoushuu, Dr. Yoshimura, is at my Yamaguchi university. You can download a Japanese version of the Manyoushuu together with a program for searching within it from his website at by entering your name and your email address. This truly wondrous piece of software enable us to look at the use of kami in the ancient book of poems.

In the Manyoushu, kami is written using the single usual Chinese character that it has today (made of parts meaning "point" and "say humbly"), so it is difficult to see what the etymology of the Japanese word might be. However, it is worth noting that kami often appears in the compound "Kannagara," where "nagara" is translated as "being at the original essence of."

Hence the old name for Shinto (kami and mich or path) was "kan-nagara no michi" meaning "path following the original essence of god."

"Nagara" is usually written with an ideographic kanji which elsewhere is read as "mani-mani" which means "as is" or "at the mercy of." The only occurrence of "mani-mani" in the modern dictionary of Japanese ( is "nami no mani-mani" which means at the mercy of the waves. Hence "Kan-nagara no mich" might also be translated as "the path of putting yourself at the mercy of god"

However, and where my other wacko kami theory comes in, today the word "nagara" is usually put on the end of verbs, to mean "while doing." For example "Ongaku wo kiki-nagara benkyou suru" means "study while listening (kiku > kiki ) to music" . And indeed, this "nagara" also used in the ancient poems (manyoushu) both as one way of writing the "nagara" after "kami" and one way of writing the "nagara" after a verb "ii nagara" - while saying. Indeed the phonetic reading of "Nagara" is only used after the character for kami and after verbs.

It is difficult to come up with a reason why this should be. Why should a deity be a "bite"? It should be remembered that the gods of the imperial lineage were created from biting and crunching up, when Susano and Amaterasu met at entrance to heaven and bit and crunched up each other accessories! And biting was also the way that they used to make that religious nectar sake.

So there is a possibility that "kami" was originally a verb. This brings us to the last homonym of kami, which is the noun form of the verb "kamu," meaning "to bite" or "to chew". So perhaps kami originally mean a bite or a chew? I am not sure which of the@vowels for "i" is is written with, since the word for chew does not appear in the Manyoushuu. However, in the scene where Amaterasu and Susano bite things, the word "kami" is written phonetically with mi1. So this, for the same reason that we rejected the idea that kami means above, it seems. The true etymology of kami remains a mystery.

Kami as Abbreviation of Mirror.

One final theory for the meaning of kami is an abbreviation of "mirror," kagami. The chances of this are not strong since it is not common for etymological explanations to suggest that words are formed by abbreviating their centre, and the etymological explanations of mirror do not usually suggest a link. Kagami is said to originate in "kage mi" or "look (miru > mi) at a shadow or image (kage)," or the noun form of "kagamu" to bend over, since the first mirrors were surfaces of water, over which one bent over (I like this theory, and believe that there are many such gbendings overh = kagami = mirrors in the Kojiki)

While the etymological reasoning for supposing that kami is an abbreviation of kagami are not that strong, the connection is pointed by shinto theologians, such as the founder of the Kurozumi sect. They point out that kagami is the symbol of the soul but the soul of man contains ego "or ga." Kami is the soul of man (kagami) without the ego (ga).

However, once again, the mi of miru (see) and hence "kage-mi" (shadow-watch) and kagami is i1.

Like other 2nd vowels, mi2 is very rare in all words occuring in yami (darkness) and mi (self, nut,

The only other word "kami" that uses i2 (the rarer of the two sounds) is "kami-nari" or thunder. Since "nari" means sound, I had presumed that thunder was originally "the sound of god," and hence the word for thunder derived from the word for god and not the other way around. However, the character for thunder was sometimes written as "kami" alone, so it is conceivable that the relationship was in the reverse direction.


Text of the Kojiki in Japanese
Are "paper" "hair" "above" and God the same word
Norinaga notices that Kami as god and Kami as above are written differently.
Professor Yoshimura Makoto's homepage
Professor Yyoshimura's Manyoushuu Corner
The Registration Page for Professor Yoshimura's Manyoushu
The Manyoushuu at the University of Virginia
Ancient Online Digital books.
More Ancient Online Digital Books in Japanese
A list of the words in the Kojiki, Nihonshoki and Manyoushu for which the vowel types are known

Posted by timtak at February 14, 2004 04:08 PM

How do you feel that realtes to the method that got the sun goddess out of the cave? The word makes the myth, or the myth wmakes the word?

At first I thought your thinking might be too Bhuddist, but the mirror is such a central symnbol.

Thanks for having this GREAT site!!!

Posted by: Don Webb at August 3, 2007 12:03 AM

> How do you feel that realates to the method that got the sun goddess out of the cave? does the etymology of kami relate to the method that got the sun goddess out of the cave?

As you knw the sugoddess wwas draw out of the cave with a mirror, that she subsequently said should be regarded as herself.

Well, as dealt with in the last section of the above post, "Kami as Abbreviation of Mirror," there are those that say that "kami" (Japanese deity, spirit) is "kagami" (mirror) without the "ga" (ego). But while I agree with the sentiment, etymologically I don't think that there is a lot of basis for saying so.

The same Kurozumikyo (19th century cult, just continuing to the present day) leader suggests that the Japanese word for person/people, "hito" should be understood as with "to" the sun "hi" and argues that we are all (all "hito") in possession of a "kagami" without the "ga," and a soul which is a "bunshin" (erm, no idea how to translate this...child-part) of the sungoddess.

As a philosophy student, the whole idea is rather attractive. But etymologically I don't think there is much to suport the above hypothesis. The 19th century cult leadeer was punning, methinks.


Posted by: Timothy Takemoto at August 3, 2007 04:56 AM