Kototama and the Japanese Vowels: What Phonetics Confirms, and What Our Ancestors Believed

In book 5 of the Man'yōshū there is a long poem by Yamanoue no Okura.

Sora mitsu / Yamato no kuni wa / sumekami no / itsukushiki kuni / kotodama no / sakiwau kuni to / kataritsugi / iitsugai keri

Man'yōshū, book 5, poem 894 [^1]

"Kotodama no sakiwau kuni" means the land where the spirit-power of words brings fortune. An 8th-century poet handed down the image of Japan as a country where words act upon reality. This passage is solid textual evidence that kototama (the belief that spiritual power dwells in words) stood at the center of Japanese self-understanding from ancient times.

From here, however, we must proceed carefully. Many writings declare that "science has proven kototama," but what research has confirmed and what our ancestors believed, or what I myself take from it, must be written as separate things. This essay first confirms what phonetics and neuroscience have established, then sorts out the claims that have not been verified, and finally states my own interpretation.

01Japanese is a language whose syllables basically end in a vowel

Japanese syllables, apart from the moraic nasal "n" and the geminate "tsu," end in a vowel as a rule. In "sa-ku-ra" or "ya-ma-to," each sound carries a vowel. Compared with languages like English, which permits consonant clusters as in "strength," vowels occupy a larger share of speech of the same length. This is a basic phonological description and not in dispute.

Japanese has 5 vowels: a, i, u, e, o. A 5-vowel system, however, is not unusual worldwide. Surveys of the world's vowel systems find that languages with around 5 vowels are the most common, and Japanese falls in the ordinary range ². One sometimes reads that "Japanese has the second-fewest vowels in the world," but this is inaccurate. What distinguishes Japanese is not a small number of vowels but the structure in which nearly every syllable ends in a vowel.

02The identity of a vowel is its formants, a pattern of resonance

What, physically, is the difference between "a" and "i"? Acoustic phonetics answers this clearly.

The pitch produced by the vocal folds (the fundamental frequency) hardly changes between vowels; it lies roughly between 100 and 150 Hz for adult men and 200 and 250 Hz for women. What makes a vowel a vowel is the resonance that occurs when the sound from the vocal folds passes through the "tube" of the mouth. The position of the tongue and the opening of the mouth change the shape of the vocal tract, and certain frequency bands are emphasized. These bands are called formants, numbered from the lowest as the first formant (F1) and the second formant (F2) ³⁴.

The 5 Japanese vowels can be distinguished by the combination of F1 and F2. "A" is spoken with a wide-open mouth, so its F1 is high, around 700 to 800 Hz. "I" places the tongue high and forward, so its F2 is high, above 2,000 Hz. "U" and "o" have a low F1, and "o" has a low F2 as well; "e" lies between "a" and "i." The figures shift with the speaker's sex, age and individual traits, so they can only be written as approximations, but that vowels can be physically described as 5 distinct resonance patterns is settled phonetics ³⁵. In speech synthesis, applying filters that reproduce the formants to a basic vibration is enough for a machine to produce audible vowels.

This is the extent of what phonetics has confirmed. One often reads that each vowel has its own "vibration" or "frequency," but strictly speaking a vowel is not a single frequency; it is a combination of several resonances that differ from speaker to speaker. It is not something fixed by one number, as in "that sound is such-and-such Hz."

03The Tsunoda hypothesis is a hypothesis, not established theory

Writings that frame kototama in scientific terms almost always cite the research of Dr. Tadanobu Tsunoda. In the 1970s he reported that native speakers of Japanese process sustained vowels in the left hemisphere while speakers of Western languages process them in the right, and further that Japanese speakers hear insect songs and the murmur of streams in the left hemisphere, and that this pattern is determined not by race but by the language environment of one's upbringing. Tsunoda also reported that native speakers of Polynesian languages show the same pattern as Japanese speakers ⁶.

The idea is attractive, and one is tempted to connect it to the view of nature found in the Man'yōshū. This essay, however, does not treat it as fact. Tsunoda's experiments used a method of his own, and the theory cannot be said to have been widely confirmed by replication from other researchers. Neuroscience does accept that one's native language shapes how the brain processes information, but the specific claim that Japanese speakers hear vowels and insect songs in the "language brain" has not become established theory. I consider it honest to read the Tsunoda hypothesis as a hypothesis still under examination, not as a confirmed fact.

04What research has established about voicing and the autonomic nervous system

Next is the influence of voicing on the body. Here too there is a line between what is confirmed and what is not.

As anatomical fact, the vagus nerve (the 10th cranial nerve) leaves the brainstem and sends branches to the larynx, heart, lungs and digestive tract. The recurrent laryngeal nerve that moves the vocal folds is a branch of the vagus, and so is the auricular branch distributed to part of the outer ear canal ⁷. That the vagus nerve is involved in the act of voicing can therefore be said within the bounds of anatomy.

Within what research has confirmed, two findings are of interest. One is a study by Bernardi and colleagues published in the British Medical Journal in 2001. Slowly reciting the rosary or a yoga mantra slowed breathing to about 6 breaths per minute, and the fluctuations of heart rate and blood pressure synchronized with the breath ⁸. This is a finding that voicing while exhaling slowly and at length influences the workings of the autonomic nervous system.

The other is a small fMRI study by Kalyani and colleagues in India in 2011. When participants chanted "OM," activity in limbic regions such as the amygdala and hippocampus decreased, while the control condition of producing only the consonant "ssss" showed no such change. The researchers speculated that the difference might be due to stimulation of the auricular branch of the vagus nerve ⁹.

Let me be precise here. This was a preliminary study of 12 participants, and the researchers themselves presented the explanation as speculation. The account in which the vibration of a vowel acts on the brainstem through the vagus nerve is not the study's conclusion but the researchers' hypothesis. Current research does not permit the statement that "voicing a vowel is in itself a neural stimulation device acting on the brainstem." What can be said is that voicing with slow, long exhalation affects the autonomic nervous system, and that a study observed changes in brain activity during voiced chanting; that is as far as it goes.

05How I take the sounds of a norito

From here I leave the confirmation of facts and write my own interpretation.

A norito (a ritual address recited before the kami) is composed essentially of yamato-kotoba (native Japanese words), and nearly every syllable ends in a vowel. Take "harae-tamai kiyome-tamau" from the Ōharae no Kotoba: split into syllables it becomes "ha-ra-e-ta-ma-i-ki-yo-me-ta-ma-u," 12 sounds, each closed by a vowel, containing all 5 vowels, with the wide-open "a" appearing most often. A norito is recited by drawing each sound out at length while exhaling slowly. This overlaps with the condition shown in the Bernardi study: voicing while breathing out slowly and at length.

Here I see the wisdom of earlier generations. The ancients knew nothing of the vagus nerve or formants. Yet they refined and preserved, as the form of prayer, a way of voicing that sounds each vowel carefully and at length. That this way of voicing satisfies a condition now known to influence the autonomic nervous system is, as I take it, no coincidence.

Dictionaries of classical Japanese explain that in ancient Japanese, "word" and "event" were expressed by the same word, koto ¹⁰. Our ancestors grasped the uttering of a word and the occurrence of an event as one thing. To call this a "physiological fact" goes too far, but within the range in which voicing changes breathing, and breathing changes heartbeat and emotion, our ancestors' sense that "word" leads to "event" does overlap with modern research. I think that when the Man'yōshū poet spoke of "the land where kototama brings fortune," the felt sense that words act on reality was supported by this bodily experience.

Modern Japanese has grown faster, sounds are dropped, and vowels are clipped. When "arigatō gozaimasu" shrinks to "azasu," what is lost is not only politeness but the very time in which each sound resonates. Reciting a norito before the kamidana (a household shrine) in the morning, or simply voicing "aah" and feeling the vibration in the chest, is something I recommend not as a scientific therapy but as a way of settling the body that our ancestors left us in the form of prayer. A manner of voicing handed down for more than a thousand years is deeply familiar to our bodies, whether or not its basis can be confirmed afterward.

When speaking of kototama, the temptation to borrow the language of science and speak in absolutes is always present. But to write what has been confirmed within the range in which it has been confirmed, and to write the beliefs of our ancestors and one's own interpretation as what they are, is itself the attitude of one who treasures words. That, I believe, is the courtesy toward language owed by those who live in "the land where kototama brings fortune."