Pathway to Japanese Literature

I

The history of our nation was on the verge of spanning three thousand years.
That we today can generally comprehend the numerous literary works produced during this span—particularly those predating the Nara period—must surely have stirred doubt in many minds regarding how such understanding could be possible.
However strange it may seem that works from before the Nara period—having traversed so many eras—remain broadly intelligible to us now, this fact persists unaltered.
To doubt this reality at its foundation would force us to conclude that extant documents are not truly ancient—a position offering no resolution.
Yet since documents demonstrably existed during and even prior to the Nara period, casting doubt upon this fact would only render our skepticism itself suspect.
Why then does this comprehension endure?
Those with even rudimentary linguistic training would deem such a phenomenon impossible, yet its factual occurrence remains undeniable.
How then should we reconcile this paradox?
To assert that “Japan’s comprehensibility stems from being a land blessed by kotodama” offers an explanation too facile to withstand scrutiny.

Narrowing our focus to books from ancient times prior to the Nara period—for instance, considering works like the Kiki texts—one might easily assume they were products of the Nara period when vaguely contemplating their dates of composition. Yet simultaneously, even non-scholars generally understood that content written during the Nara period did not necessarily concern Nara-period matters. Their very texts made clear that portions existed which had not first been written during that era. Though compiled in the Nara period, the Man'yōshū contained numerous works transmitted from earlier times rather than being exclusively contemporary compositions. Even dismissing this as mere transmission, the Kiki texts alone undeniably preserved sections not originally penned during the Nara period. While their overall structure naturally belonged to that era, earlier writings appeared embedded within them like inlaid craftsmanship—considerable in number and piecemeal in distribution. Certain passages in the Nihon Shoki's "One Book Says" sections clearly derived from copied sources, while its old annotation portions revealed older texts preserved in slightly altered forms. These occurred with notable frequency. Centering our examination here revealed what seemed like clusters of ancient writings gathered about these points. The Kojiki itself displayed varied styles—some passages in Chinese-style prose, others in Japanese-style prose, with certain sections merely applying Chinese characters to native syntactic patterns. Scrutinizing these revealed pure archaic Japanese prose embedded even within non-poetic sections. These were written either in late-period Man'yōgana-style script or in Chinese prose heavily infused with Japanese linguistic qualities.

Therefore, the Kiki texts could largely be read through in kanbun-style interpretation if one attempted to do so; however, in the end, passages imbued with what might be called a Japanese textual flavor emerged—for instance, one developed the unavoidable habit of interpreting phrases like “the time of heaven and earth’s first emergence” as “Ame-tsuchi no hajime no toki.” There was no absolute necessity to interpret “the time of heaven and earth’s first emergence” as *Ame-tsuchi no hajime* (“the beginning of heaven and earth”). However, because other sections contained passages that could not be interpreted without applying such Japanese-style readings, one inevitably arrived at this interpretation. This was what was called the Japanese reading of the *Nihon Shoki*. Guided by this, Japanese readings came to be implemented even in the *Kojiki*. What had still been unnatural reading methods during the time of *Nihon Shoki* became, in *Kojiki*—for example, as seen in *Kokun Kojiki*—readings executed with remarkable skill. When viewed from our present perspective, works like *Kokun Kojiki* might seem straightforward; however, upon closer scrutiny, one realized the tremendous efforts and hardships that had been painstakingly accumulated. They interpreted it in such an emphatically Japanese style that it felt frustrating. Of course, interpreting [the texts] in as Japanese a style as possible was the proper approach; however, the fact remained that there existed a final barrier—a point that even Motoori Norinaga had failed to consider. He had failed to consider that within the *Kojiki* there existed yet another, older method of reading from an earlier era; thus, it had resulted in such an interpretation. Of course, there were passages shared with the Heian period in terms of prose; however, Motoori Norinaga’s readings had become excessively proficient in Heian-period textual interpretation overall. I thought there were Heian-period stylistic expressions to such an extent that one could say [they] stood upon the mood and themes of the Heian period and were made to adopt readings incorporating hues of the Nara period. However, given that they had been interpreted with such earnestness and diligence, I believed that by gradually making minor corrections to their legacy, we could draw closer to the Nara-period readings of the *Kojiki*. In any case, Motoori Norinaga’s reading method, left unaltered, did not constitute interpretation within the Nara-period context.

What did it mean that ancient phrases were embedded within the Kiki texts like inlay work? To state the conclusion simply: it indicated an extremely crucial juncture—there must have been an awareness that losing these phrases would leave people unable to account properly before the gods or the sanctity of the court. Had this unspoken constraint not existed, works like the Kojiki could have been rendered far more extensively in Chinese-style prose. The fundamental question of how texts left by people of antiquity could be understood as-is by later generations remained one worth examining. Even the Kiki texts had not first become comprehensible only after Edo-period scholars studied them. In particular, works like the Nihon Shoki likely saw lectures commence immediately after their compilation (while some argue these lectures began in the Heian period, I believe they started shortly after compilation). Since then, scholars of successive eras had repeatedly interpreted them to comprehensible degrees. Through this process, people transmitted them as histories retaining intelligibility—it was merely that this transmission left fewer explicit traces in the Kojiki’s documentary record. That works like The Tale of Genji remain legible today stemmed from their having been continuously cherished across vast stretches of time; indeed, transcription work had already begun even as Genji was being completed—or perhaps before its finalization. From Kamakura to Muromachi periods, commentaries were produced fragmentarily before systematic exegesis emerged. As a result, these texts became nearly fully comprehensible even to us today. Such a fortunate state proved truly rare—in this sense, we owed profound gratitude to people of old. Undoubtedly, the Kojiki too possessed a history akin to Genji’s reception trajectory.

However, when examining old commentaries on how people of antiquity read texts—whether the Man'yōshū or The Tale of Genji—we find that older works were subjected to increasingly impressionistic readings and interpretations.
Among these exist types that—from our modern perspective—provoke incredulous laughter.
Take for instance this Man'yōshū poem:

Yatsuri Mountain—its grove unseen beneath swirling snow; How splendid the dappled-steed morning! (Book 3, Poem 262)

...interpreted as "Snow-dappled horse—how delightful the morning!"—a reading so nonsensical it becomes unbearable.
Even when lacking comprehension, they would force such interpretations through sheer intuition.
Yet through this accumulation of feeling-based exegesis, they ultimately produced interpretations deserving our trust.
Their temporal proximity granted an intuitive grasp we lack today—these impressionistic readings naturally approached textual cores.
We moderns can no longer rely on such intuitive propulsion but must instead proceed upon theoretical foundations of grammar and etymology.

II

The act of writing down ancient phrases in books and seeking to preserve them occurred from a period slightly later than when those ancient phrases had been actively used.
Moreover, writing ancient passages did not merely involve transcribing them as-is; rather, they wrote by building upon their own certain understanding.
Moreover, when considering that impressionistic emotional elements could not possibly remain unmixed at the foundation of their understanding, it stood to reason that the transmitted text would become something considerably altered from the original ancient phrases.
Generally speaking, as long as ancient texts remained comprehensible to all who read them, they were acceptable; however, once they became even slightly incomprehensible, one had to make some alterations to render them understandable.
However, there were sections where alterations could be made without any issue and sections where it was absolutely impossible to apply the brush of revision.
That is to say, they judged that arbitrarily altering matters concerning the gods, matters of the court, or sacred songs and proverbs would constitute committing a sin against the deities and the sacred words.
Thus, even if it were to become incomprehensible to all eyes, they could not lay hands on it.
However, even in practice here, there may have been ongoing gradual changes invisible to the eye, with instances of alteration and preservation intermingling.
For example, what was particularly abundant in the *Kojiki* were sections specifying reading methods in phrases like “read these two characters phonetically”; these had been attempts to preserve ancient language as-is. In some instances, they even attached marks indicating tones such as rising or departing.
Furthermore, by utilizing Chinese-style punctuation marks, they attempted to make it read in as archaic a style as possible.
Fundamentally, one could not read these texts without knowledge of Japanese; however, those who generally understood Japanese believed that by reading them with return marks in such a manner, they could interpret them in the same way as old texts, and clinging to this notion, they attempted to render them in a Japanese style through punctuated reading.
Even when examining the *Man'yōshū*, they saw that auxiliary verbs and particles like *te-ni-o-ha* had been omitted, pressing forward to the point where it could presumably be read at this level.
It had been an attempt to cut corners as long as it did not interfere with the mode of expression.
In any case, within the *Kojiki*, one could discern an attitude of incorporating ancient phrases transmitted from previous eras and endeavoring to preserve and pass them down.

For example, when Ninigi-no-Mikoto had attained his divine maturity and it became permissible for him to deign to descend to Ashihara no Nakatsukuni (though the Kojiki describes the time of his descent as if he were a green child), there existed in the text a passage narrating the circumstances of his heavenly descent: "*Oki tsu ukihashi uki shimari, sori tata shitete yu* [the following eleven characters are also read phonetically]."
In the Nihon Shoki, [the phrase] appeared in two places—in this main text and another text—written as "立於浮渚在平処," which old commentaries annotated with the phonetic reading "Hakike ni maritabiraji tatashite."
If this was something added when compiling the Nihon Shoki, then it would mean that the text was written with the intention of being read that way; however, this old commentary appeared to show that it was something from a later period.
If they were to read it that way, then this method of assigning [characters] was exceedingly poor.
Other sections were assigned much more skillfully.
Furthermore, in another text it was read as "floating island-like," but despite temporal disparities between [the versions], these two reading methods coexisted within the Kiki texts, resulting in the Nihon Shoki preserving an alternate reading.
Originally, the Kojiki and Nihon Shoki may have been nearly identical, differing only in extremely minor aspects of their transmission.
However, simply because we cannot read them in their original form, they might appear to coexist; yet it could not necessarily be said that the readings in the old annotations were correct.
In any case, they attempted to transmit readings from previous eras in this manner, even while not fully understanding them.
For people of old, this must have been comprehensible in this way; it was merely that it remained incomprehensible to us.
If we read it as-is, we could sense that it likely meant "[the deity] stood upon a flat place resembling a floating island," but ultimately, the precise reading method remained unclear.

In any case, this is how the Kiki texts sought to preserve the ancient language.
It was likely that by this era, they had already become incomprehensible and were transmitted in two or three different ways; however, even though similarly not understood, the circumstances differed from those of the divergences in how songs were transmitted between the Kiki texts.
Admittedly, there may indeed have been instances where differences arose from this same reason. However, generally speaking, though incomprehensible to us later generations, it would have been more accurate to think they were comprehensible to people of old.
However, we must also consider that there were things even the compilers of the Kiki texts themselves did not understand.
One might also imagine they transcribed as-is—despite lacking comprehension—narratives that no one voiced anymore: stories kept barely alive through tenuous transmission by merely one or two reciters, narratives that had effectively died out in substance.
Therefore, it was a mistake to blindly believe that while people of old probably understood these texts, we were simply powerless to comprehend them.
There could be no doubt that even the Man'yō people often wrote down matters they themselves did not understand exactly as they were—without comprehension.
The old annotators had not considered this point in the slightest, but for us, knowing this urged us toward a new departure—the correct path forward.

From all of the above, I believe we may reasonably conclude in roughly this manner. In other words, [they employed] two methods: embedding incomprehensible passages verbatim as incomprehensible text on one hand, and prioritizing comprehensible passages while systematically altering the incomprehensible ones in accordance with them on the other. This situation becomes clear when examining norito and similar texts. Regarding norito, scholars like Mabuchi have stated that some were created during Emperor Sujin’s reign, but the evidence for this remains uncertain. Whether they were created during Emperor Sujin’s reign or during the Ōmi court, the critical issue lies in whether they have been transmitted exactly as they were at their creation and whether we today can truly comprehend them. If after over a millennium they still remain comprehensible to us now, this would present a far more significant problem than questions of their dating. The Great Purification Prayer and others are said to have originated during the Ōmi court, but setting aside the obscurity of certain highly specific portions, the fact that they remain largely comprehensible to us today as they are strikes one as strange—indeed, this ought to be far more mysterious than the various supernatural wonders appearing in the Age of the Gods volume. In essence, this stems from their inevitable modifications up to a certain point and efforts to render [the texts] comprehensible—since it would prove problematic if future generations could not understand them. Even in texts like the Engishiki, which may contain records from before Engi—from around the Jōgan and Kōnin eras—there can be no doubt that revisions were carried out during that interval. Were this not the case, we today could not possibly understand them at all. The revisions made to ensure comprehensibility were implemented from start to finish as needed, but ceased entirely at a certain period, with no further alterations made thereafter. From that point onward, they became purely unintelligible and have since been transmitted without being understood—such appears likely. Such revisions occurred not only through written documentation but also appear to have been gradually implemented during oral transmission. However, originally only those sections deemed related to deities were left unaltered out of fear of divine retribution; consequently, over time those very parts grew increasingly incomprehensible. From the outset, there existed no intent to make these particular sections comprehensible. Were we to imagine the original archaic phrases as forming a straight line, then as sections becoming incomprehensible inevitably emerged along its course, repeated revisions would result in a text resembling a wavy line with protrusions—yet since these revisions constituted necessary and inevitable responses [to incomprehensibility], I consider both original and revised texts equivalent. This derived partly from their longstanding belief in verbal efficacy—if (1) held sanctity, then by extension (N) must also be sacred—as their reasoning went.――――(1)―・―・―(2)※(zigzag line)(3)～～～～(N) I do not regard such protrusion-bearing texts as representing current norito; nevertheless, they must have diverged considerably from original archaic phrases. Even within the same Engishiki, comparing Heian-period court documents with older versions shows largely identical general text—the sole difference being difficult archaic phrases inserted sporadically in older copies. In other words, while we perceive protrusion-like sections and straight-line sections as distinct, for people of that time this difference held no significance.

In any case, they had considered making the ancient phrases as comprehensible as possible and had indeed been doing so; thus, even when we read old texts, we did not encounter archaic words and phrases to the extent we might expect. When one looked at the Kiki texts expecting them to be filled with matters entirely unknown to us, it was astonishing how unexpectedly comprehensible they were—in other words, the authentic archaic words and phrases had been gradually vanishing even from such classical works. They had already been revised to a degree comprehensible to people of the Nara period, which consequently made them easier for us to comprehend. Moreover, there was the additional aid of Chinese writing. By clinging to Chinese-style syntax for expression, the general meaning became comprehensible.

Therefore, as we gradually advanced this discussion, the era when our ancient texts had been purely colloquial became utterly inconceivable.
That is, these texts were created with words that had existed since ancient times as their foundation; their most crucial parts remained in ancient language while comprehensible elements were gradually added around them.
When considering such phenomena, first we recognized that spoken words vanished the moment they were uttered while written words endured forever; yet in antiquity there had existed a third category between these—words meant for memorization.
Mere spoken words were not retained in memory while written words emerged much later in cultural development.
In any case, there existed words—or rather texts—that were neither spoken nor written but had to be memorized.
These were of course not newly created then but phrases transmitted from antiquity—a legacy from previous eras that had to be preserved through rigorous memorization for posterity.
The living words among people continually transformed while separate from this existed groups of professionals who memorized unchanging fixed phrases.
Those conducting rituals and practitioners of magic—though differing by verse type and nature—formed multiple organizations that memorized and transmitted ancient texts.
As this practice gradually declined transmission shifted to those leading noble lives.
Since noble families and aristocrats lived lives nearly divine such sacred verses were transmitted through generations; their scions' education centered on memorizing these ancient texts continuing until the Heian period.
Thus sacred verses came to be transmitted even to those unconnected with rituals.
These became what are called proverbs—phrases or short sentences.
The purpose behind their transmission remains unclear.
I think they may have transmitted them simply for being ancient and sacred but let us consider one possible meaning.

III

In phrases transmitted from antiquity dwelled the spirit of the land—they had believed the nation’s power resided within them.
Therefore, those occupying important positions within the nation had to receive and transmit this; otherwise, its power would be lost.
For this reason, these ancient phrases absolutely required memorization.
What was transmitted in this manner became proverbs, with a portion evolving into songs.
Proverbs were predominantly praise phrases in origin.
Praise phrases did not fundamentally aim to extol present circumstances; rather, they sought to persuade the addressed spirit that “it would be good if matters became thus.” The spirit would then feel responsible for the uttered phrases—objectively speaking, through verbal power one enchanted the other—that is to say, infected them through language.
Consequently, such praise phrases naturally concerned predominantly either the nation itself, deities, or those living in proximity to the divine—such as local lords.
To speak of nation and deity ultimately meant one and the same; praising a land’s local lord carried equivalent effect and consequence to praising its tutelary god.
In any event, direct praise of reality remained rare.
The phrase “Toyouashihara Mizuho no Kuni” by no means extolled some bountiful paradise where rice grew abundantly.
“This land yields fine rice”—these were words spoken to make the land spirits comprehend.
The belief held that by uttering this declaration, the land could manifest exactly as described through those words.
This principle emerged most clearly in the Hitachi Fudoki.
What appeared repeatedly as folk proverbs, folk narratives, or simply notations of “folkways” were precisely these proverbs.
Within such praise phrases for nation and deity existed invariable set expressions.
When venerating deities, it became customary to prefix these fixed phrases before their sacred names.

When people grew accustomed to this usage, they began presenting only the prefixed praise phrases while omitting the principal deity’s name yet remaining comprehensible.
As this practice persisted over time, the praise phrases gradually became dissociated from their original meanings.
The praise phrases alone grew archaic, remaining eternally fixed without being re-memorized, while the appended divine and national names were constantly renewed through repeated memorization to maintain fresh significance.
Consider the customary proverb from *Hitachi Fudoki*: “The land where black clouds cling to Mount Tsukuba and sleeves are soaked.” Though Hitachi province itself underwent continual renewal from beginning to end, “black clouds cling to Mount Tsukuba and sleeves are soaked” persisted as an immutable phrase.
Yet when we examine this proverb, we sense an elusive comprehensibility.
Even without such intuition, as linguistic comprehension developed alongside an enduring desire for knowledge, interpreters combined these faculties to devise varied explanations that rendered some semblance of meaning.
The *Fudoki* records that when Prince Yamato Takeru toured the eastern provinces and arrived in Niihari District, Kunimiyatsuko Hinara Tamemikoto had a new well dug whose waters proved remarkably clear and pleasing.
The Prince halted his palanquin to admire the spring and wash his hands, whereupon his robe’s sleeves dipped into the water and became drenched—hence naming the province through this “sleeve-soaking” rationale.
Yet many such phrases resisted explanation despite concerted efforts.
Though the praised terms were understood, interpreters ultimately compromised by adapting incomprehensible praise phrases to fit presumed meanings.
Take “isana-tori” from *Man'yōshū*—a term initially obscure until sustained usage attached it to maritime imagery through interpretations of whale-hunting.
Once established thus, all memory of its laudatory purpose faded into rote preservation of tradition.
Nevertheless, compelled by necessity, scholars strove to rationalize these phrases by aligning praise formulae with their subjects—thereby giving rise to pillow words.

The oldest origin of pillow words lay in this practice.
I do not know why, but there existed phrases that had to be attached.
To these, people appended phrases they believed would render them comprehensible.
When this occurred, the scope of pillow word usage consequently expanded; thus, upon examining each instance, all appeared to originate from those very pillow words.
What are called origins may indeed have been multiple, but ultimately, their initial form emerged from proverbs of the sort I have described.
Therefore, pillow words for place names could be said to have remained relatively pure while simultaneously ancient.
However, pillow words that had once been proverbs in this manner became nearly meaningless formalities, their original nature as praise phrases forgotten.
Though the reason remained unknown, people came to regard them as spiritually mysterious phrases of importance.
And since this was after all a legacy passed down from our ancestors—a heritage requiring active use—they attempted to breathe life into phrases that had already perished.
In other words, they gradually extracted meaning from words they had once considered devoid of significance.
For example, when interpreting norito ritual prayers—their praise phrases being incomprehensible—scholars conducted limited comparative studies, approaching them through frameworks like “this is how it appears in senmyō edicts” or “this is how it reads in Kojiki.”
Such an attitude might indeed be called academic; however, discerning their primordial meaning proved no facile endeavor.

IV

Needless to say, the academic study of language must be grounded in comparative linguistics; however, we must also establish Japanese linguistics—or rather, what ought properly to be called as such. Japanese grammar may well be termed Japanese linguistics, but current grammatical studies do not constitute pure scholarship—they have progressed only to the level of practical language studies, and to designate this as “Japanese linguistics” strikes one as somewhat disheartening. I believe it should adopt a more rigorously linguistic approach. Even now, they continue to use it solely for its original purpose of writing compositions. The stagnation in grammar stems fundamentally from the absence of a Japanese linguistics established as a national discipline. Comparative linguistics is of course important, but its conclusions invariably settle into the same patterns. If it were merely a matter of comparing words with foreign languages as was once fashionable, one could accomplish it simply by consulting a dictionary. In any case, we must establish a Japanese linguistics and swiftly elevate Japanese grammar to an academic discipline. While comparative linguistic methods may yield partially valid insights, they cannot serve to deduce the entirety of the Japanese language. The Japanese language is not solely related to the Ural-Altaic language family; it must of course also be extensively compared with the southern islands. In any case, I consider this outward-directed attention crucially important. Folklore studies too must take Japanese folklore studies as their foundation, upon which foreign comparisons should be conducted. For conducting comparisons with foreign countries alone proves futile when the foundation remains unsteady. By the same token, comparative research focused solely on vocabulary will never contribute to resolving issues of the Japanese language, no matter how much time passes—it amounts to nothing more than noting superficial resemblances. The fundamental similarities and ancestral roots of language always lie beneath the surface, transcending apparent likeness or difference. Though it is true that our ancestors came from the south as well, this cannot be easily discerned through superficial observation alone. Therefore, before making comparisons, we must contemplate the form of the Japanese language itself—and thus attempt to construct this form, however vaguely.

For example, the term for "ladder" was called *hashitate* (vertical ladder) in ancient times and became established like Amanohashidate. This *hashitate* denotes a vertical ladder; in standard Japanese, the normative form would be *tatehashi*, demonstrating how this term structurally diverges from later Japanese composition. Regarding such linguistic phenomena, Dr. Tsuboi Kumazō had long noted instances including socks being termed *shitagutsu* (韈 → *shitauzu*), the cloth before carriages—logically *misu shita* ("below the blind")—being called *shita misu* ("lower blind"), and hill-adjacent areas that might be *oka kata* ("hill-side") becoming *kataoka* ("side-hill"). (This Kataoka spread widely as a place name.) Calling temporary mourning *mogari* reverses provisional mourning's logic. Adjectives all follow their nouns. When we examine such examples, they differ from how we conceptualize word forms.

When a few more examples of this kind are gathered, I believe we would have to gradually revise our understanding of the Japanese language's lineage or rules. There remained far more things we did not comprehend. To take a grammatical example: in ancient times, sleeping was referred to as *inu*. Some even considered *iha* a prefix, but here we find phrases like "cannot sleep peacefully" and structures like "sleeping sleep." When forms such as *ne ni naku*, *ne o naku*, and *ne nomi naku* appeared, they proved particularly challenging to explain in secondary education. This difficulty stemmed from current grammar causing confusion. While modern grammar followed Heian-period conventions, even Edo-period scholars left unexplained elements untouched. The *ne* in *ne o naku* was a noun denoting the act of crying, as seen in *kari ga ne* ("cry of wild geese"). To cry served as its verb. Similarly, in *i o neru*, *i* functioned as the nominal form of *neru* ("to sleep"), with *neru* being its verb. Such nouns became fixed in literature as literary terms but gradually lost their grammatical significance. However, both "to cry" and "to sleep" stood as bare intransitive verbs without moorings. In Dr. Ōtsuki Fumihiko's classification system, intransitive verbs were divided into complemented and non-complemented types based on complement presence. While cry and sleep belonged to the latter category today, these verbs had originally functioned as complemented intransitives in antiquity—not through complements per se, but through nouns requiring fusion with verbs for meaningful use. This necessity gave rise to expressions like *i o neru* ("sleeping sleep") and *ne o naku* ("crying tears"). Though *i* and *ne* were not complements themselves, such forms evolved until even intransitive verbs demanded objects or targets—giving birth to complemented intransitive structures. Admittedly, some older scholars proposed alternative theories about particles like *wo* and *ni* being used loosely, but regardless, intransitive verbs ultimately came to require objects. When exactly this occurred remained difficult to pinpoint.

This is because literature preserved and perpetuated archaic language within texts on one front, while on another front maintained a conviction that letting such language die would be sinful—thus making it impossible to determine how extensively it survived in actual speech.
That archaic language persisted in written records does not constitute a single shred of evidence that it endured similarly in spoken usage.
In any case, as intransitive verbs came to require objects or complements, forms emerged such as *neru* (to sleep) becoming *inuru* and *naku* (to cry) becoming *nenaku*.
Furthermore, through the insertion of particles between elements during this process, expressions like *i o nuru* ("to sleep sleep") and *ne o naku* ("to cry tears") arose.
Considering this trajectory leads us to suspect that one origin of what we now dismiss as nearly incomprehensible prefixes may lie precisely in such phenomena.

To take another example—even within the *Man'yōshū*—there were many instances of varied grammatical forms; however, most existed as singular occurrences.
Since these were lone surviving examples, one could extract no scholarly significance from them; yet discarding them for their singularity rendered everything incomprehensible.
Japanese must first be studied as Japan's own creation—yet when pressed—such solitary examples confronted researchers repeatedly.
This phenomenon arose when certain modes of thought gained favor among myriad words while others faded into oblivion—until time left but one standing when viewed from later ages.

In Kakinomoto no Hitomaro’s long poem appeared the verse: “As dew and frost would vanish if they were to vanish at the point where migratory birds contend—”
(Man'yōshū Volume 2) contains this verse.
The term *tsuyushimo* (dew-frost) still persists in the Tōhoku region while being called *mizushimo* (water-frost) in Kansai.
This served as the prefatory phrase establishing the conditional “should they vanish.”
The expression “seeming on the verge of vanishing” was rendered through the construction “should they vanish let them vanish.”
By medieval and early modern grammatical standards simply saying “should vanish” would have sufficed yet [the text] added the conditional “should they vanish.”
Other examples of this type exist within the anthology including “...should they vanish let them vanish—so yearns my beloved sister” from Volume 4.
Though expressing emotions of lovesick anguish where even one’s body feels crushed “as if vanishing” alone would suffice [the text] still appended “should they vanish.”
Observing these examples—though differing slightly from cases like *inu* and *nenaku*—reveals verbs requiring conditional clauses for full semantic realization.
Today’s comprehension without conditionals merely results from habitual usage.
An entire era existed when verbs demanded phrases like “should vanish” before achieving standalone intelligibility.
Terms understood without conditionals in *Kiki* texts and *Man'yōshū* undoubtedly emerged through such refinement processes.

*Kani* was a suffix that took an adverb.
There were cases where *kani* took phrases from their beginnings and others that appeared to simply take verbs, but its true form became an adverbial phrase when attached.
The construction “should it vanish, let it vanish”—though not quite a full sentence—was taken up by *kani* and turned into an adverbial phrase.
Therefore, from its formation’s perspective, this was not a phrase continuing as *kenukani* (“as if to vanish”), but rather *kani* being attached to *kenaba kenunu* (“should it vanish, let it vanish”).
It was not taking individual words.
It was not that *kenaba* had been added to intensify *kenukani*’s tone.
Past scholars too had explained this matter with great simplicity; only after this process did the form *kenukani* become independent, giving rise to expressions like “kenukani omohoyu” (appearing as if to vanish).
If one could even consider this developmental path, one had to recognize that many words had followed this same course.
Yet regrettably, this too had left only one type of word surviving.
That this alone had been preserved in literature stemmed—as previously explained—from pre-literary literature’s influence and this expression’s singular prominence for reasons unknown.
If someone dismissed it as unreliable for being a single example—so long as they avoided using it to explain other cases—that sufficed; yet then this word’s explanation would never be reached.

V

Generally speaking, in the study of language, etymological research flourishes above all else; however, among these studies, there are no shortage of theories that are patently foolish.
In etymological theories, it is inevitably this methodological approach that becomes the primary issue.
Since many elements of language arise by chance, even if one gathers similar ones and attempts to explain them through analogies based on their formation, there is nothing as futile as this endeavor.
In the end, there is no choice but to trust the scholar’s character, erudition, and scientific rigor.
When it comes to etymology alone, no matter how meticulously one approaches it through documentary or scientific methods, it often remains unattainable.

As seen in forms like “kenaba kenukani” (should it vanish, let it vanish), only a single instance of this structure remains. From this, we can deduce that the expression “kenukani” (appearing on the verge of vanishing) originally required the conditional clause “kenaba” to be comprehensible—yet through prolonged usage over centuries, people grew accustomed to omitting the conditional while retaining understanding. Though “kani” persisted until the Heian period under the variant “kane,” it gradually faded from living speech and was supplanted by the term “beku.” While appending phrases like “kenaba” had become unnecessary by this stage, they were preserved under pretexts of tradition to maintain a superficial classicism. A parallel form survives in the Azuma-uta poetry collection: one instance.

O lovely field, do not burn.
Old grasses mingle with new; should they grow, they grow as though to grow (Man'yōshū, Volume 14, 3452).

Such is the case.
The verb ōu ("to grow") was ordinarily quadrigrade but could also be employed in the upper bigrade conjugation.
One cannot definitively state whether it had been fixed as quadrigrade from the outset, but it likely underwent fluctuations during the period before its conjugational patterns became fixed.
Thus even after verbs became established as quadrigrade, depending on context, the upper bigrade forms used in antiquity might resurface.

This poem was conventionally interpreted by taking its concluding line—"if they grow, let them grow as they will"—and treating it as a metaphorical song of elderly reflection. That such an interpretation held no inherent implausibility became evident when later works like

"Even if not burned, the grass will sprout.
Kasugano: if only left to the spring day—" (*Shin Kokinshū*, Mibu no Tadami)

emerged as poems of this type. Such models became ingrained in literary consciousness, resurfacing at every opportunity. In other words, unlike our modern literary perspective, people of antiquity recognized its value as knowledge. At any rate, these two poems ultimately constituted variations of the same song. While people of old likely interpreted it thus, our earlier analysis of "should it vanish, let it vanish" (*kenaba kenukani*) reveals that *ōhiba* ("should it grow") functions as a conditional marker. This suggests it may simply describe how "last year's grasses mingled with this year's sprouts, appearing ready to grow." We modern readers, inevitably influenced by later associations, find it difficult to perceive that earlier interpretive approach—yet this must be recognized as the authentic path. In essence, "ōuru ka ni ari" meant "[the grasses] appeared poised to grow."

Language is such that even within the same era, there are cases where the same word differs in both perception and usage. Therefore, when approaching ancient language through modern sensibilities, one cannot be assured of reliability. For another example, the phrase “yukue shiraniu su” (Volume 2) appears in Kakinomoto no Hitomaro’s poem. This is a grammatical form that should naturally arise from logic. Shira constitutes the first conjugation of shiru (“to know”), ni being the second conjugation of the negative auxiliary verb—forms frequently employed in the suspensive form. Not knowing—that is, the suspensive form of shiranai (“not know”), with su attached to it—means they remain unaware and render it incomprehensible. Such examples are by no means numerous. However, we must seek out our ancestors’ usage examples—even if they are exceedingly scarce—and gradually formulate hypotheses accordingly. Even words that were once extremely popular and widely used end up leaving behind just a single remnant once they fade away. Linguistic phenomena are always trends. We are in a position to curse trends; however, linguistic trends differ from ordinary cultural life in how they permeate society, for they accomplish in a short time what would take us much longer to achieve and integrate it into daily life. So although dangerous, they come surging recklessly. The passion is what threatens us. However, in any case, while being in vogue like that, they will eventually become well-rounded words and gain universality. When that happens, it becomes impossible to speak without using them. When we examine this example of “ōhiba ōuru ka ni” (“should they grow, let them grow”), we see that grammatical consciousness had shifted; it was already being understood to mean “if they grow, let them grow as they will.” This phenomenon of change is an essential point that those who proceed to interpret texts from here onward must understand; they must never forget to consider both the past and present meanings. If that is not the case, then no matter how much time passes, one will only interpret meanings through a modern lens, and the feelings of people from the past will never be understood.

Issues such as national spirit cannot truly be considered apart from an understanding of this language.
Without accurately grasping the meanings employed in antiquity, there remains no possibility of comprehending the sentiments of people from the past.
I believe there exist numerous misunderstandings where modern logic is used to conjecture about bygone emotional states.
However we might interpret matters grammatically today, it remains possible they were not understood thus in former times.

VI

Shifting our focus to a slightly later period, let us take an example from the Heian period. Language proves to be such that there exist cases where it survives across vast spans of time, cases where it emerges only to perish immediately, and further cases where—though deceased in one aspect—it persists living in another. When examining dialects and similar phenomena, we discern that they preserve elements whose processes of emergence and extinction defy ready comprehension. Thus can we not simplistically discuss language's vitality. Even when speaking of Heian-period language, we cannot definitively assert its absence during the Nara period based solely on its appearance in Heian documents—a merely negative evidentiary fact. Rather, there may well prove numerous instances where affirming its prior existence would be preferable. Yet we must also remain attentive to cases where forms, though archaic in appearance, were applied retroactively from later eras. The phrase “kenaba kenukani” exemplifies this category: while its form itself remains ancient, its popularity led subsequent speakers to mold new vocabulary into this established structure. This prime example from *Kyōgenki*'s language—though commonly identified as representing Muromachi-period grammar—by no means exclusively reflects that era. Given that its language underwent revisions extending into modern times, it constitutes an error to hastily regard it as representative of Muromachi-period speech. In truth, only the formal patterns of Muromachi-period kyogen endure; since language was being created through adaptation to these molds, we essentially confront a Muromachi-era pseudo-classical idiom. Hence no endeavor proves more methodologically perilous than attempting Muromachi grammatical studies premised on *Kyōgenki*. Similarly may we state that ancient poetry and texts consist almost entirely of archaizing imitations and formulaic types. Consider Kakinomoto no Hitomaro: I maintain it would be more accurate to declare his work contains no true “neologisms” whatsoever. These compositions merely simulate archaic diction—yet within them undeniably surges his distinctive character, imparting an ineffable novelty. Therefore even within Heian-period works—no exception to this pattern—we cannot readily determine whether linguistic forms constitute genuine archaisms or mere retroactive adaptations of ancient models.

"If I speak, I cannot [express it]; if I do not speak, my heart is troubled—how my heart aches alone." (*The Tales of Ise*)

The poem.
Here, "e ni wa" meant "unable" or "cannot," with this passage abbreviating "iheba ihi e ni" (if one speaks, one cannot speak).
"It is saying: 'If I try to speak, I cannot; yet if I do not speak, my heart churns restlessly.'"
In today’s grammar, one might say *ihaba*; but doing so made the conditional agreement cumbersome.
In essence: "If the flowers bloom, I shall go."
I found nothing more baseless than the so-called conditional agreement seen in phrases like "If the flowers bloom, I will go."
To us, it had become utterly meaningless—even detestable—though in the past, this agreement did exist.
In Heian-period usage, linguistic practices still managed to rise above such constraints superficially; but in earlier periods, they became overly bound by them—so much so that expressing a hypothetical required first establishing corresponding conditions.
Therefore, since using *ihaba* would make conditional agreement cumbersome, they adopted the ordinary form *iheba* instead—that was the rationale.
In any case, here they should have simply used "ihi e ni" (if one speaks, one cannot), but instead added the conditional "iheba" (if one speaks). This method of attaching conditions resembled the earlier example of "kenaba kenukani" (should they do, let them do).
Later, this too became unnecessary and was omitted, reduced to "e ni" alone; as a result, this term spread widely and accumulated numerous usage examples.
Typically, they read it as "en ni" and assigned the character 艶 (gloss).
According to *Konjaku Monogatari* (viewed as spanning Heian to Kamakura periods), they wrote the character 艶 (en) while reading it as enaranu or emo iwanu.
Though e did not carry 艶's meaning, people likely associated en ni with en's phonetic resemblance, seeking to emphasize en ni's sense more strongly.
Thus en ni—while already suggesting allure and flamboyance—retained traces of its ancient meaning, giving rise to readings like those in *Konjaku*.
In this manner existed the singular example of "kenaba kenukani."
Yet even such remnants left traces here; by collecting these and proceeding with comparative study, progress became possible.

At present，spoken language was not under consideration．Even if one wished to examine this issue，the absence of materials from ancient and medieval times left no choice but to rely on documents．What remained preserved in texts had fundamentally been shaped by literary consciousness—words were selected and perpetuated through this filter．When examining old books，one found our ancestors’ lexical choices strikingly limited，whether deemed biased or simplistic．The appearance of even slightly unconventional terms saw them deployed to astonishing degrees．As evidenced by norito prayers，the same phrases recurred relentlessly．In the Man’yōshū too，poets adopted “white waves” as literary diction regardless of context，employed “red steeds” for horses，then gradually shifted aesthetic preferences toward terms like “blue horses”．Hair invariably became “black hair”．These reflected mere stylistic preferences of antiquity，devoid of inherent meaning．Pillow words，initially arising from ritual necessity and accruing layered meanings，ultimately fossilized into mere poetic components．They had been assimilated as literary language—heavily imbued with scholarly artifice．

What was the starting point in the development of pillow words as literary language? That was place names. They literarized place names, giving rise to a mode of delight in contemplating them through literary sensibilities. In essence, pillow words initially served as praises for land, deities, or humans before gradually transforming in significance. Even so, they never lost their persistent association with land and gods—indeed, these connections would occasionally revive. Thus pillow words remain inseparable from land deities. Ultimately, people came to incorporate those regions' beautiful place names into their songs. Within provinces divinely selected for Yuki and Suki rituals, later compilers recorded particularly euphonious place names apt to evoke splendid associations. They provisionally named this collection *Chūshin Fudoki* ("Reported Provincial Gazetteers"), where one finds rows of Yuki- and Suki-related names resonating with lushly beautiful imagery—names steeped in classical elegance. These were woven into songs. Such gazetteers were compiled precisely to furnish material for song composition—they became seeds for crafting Yuki and Suki verses. Though these terms ultimately lost relevance to practical life, well before that dissolution they had already embedded abundant literary expressions within words deeply connected to human experience. Driven by affection for literary language, people preserved it in documents and wove texts from such vocabulary. Through this process, their capacity for linguistic selection progressed until it culminated in creating *Chūshin Fudoki*. Above all, they frequently interlaced language with words evoking spiritual resonance—sacred, redolent of antiquity, nostalgically archaic, heart-stirringly solemn, and rich in aesthetic sensibility. This explains why poetic diction remained so remarkably sparse during that era.

7

Within etymological theories there was but one aspect that tended to be overlooked—that of abbreviation itself—for in ancient times this phenomenon had been remarkably prevalent.Even when examining terms surviving today,many resist interpretation through mere assumptions about phonetic preservation with slight formal distortions.It appeared people frequently abbreviated established terms even in antiquity—as seen in expressions like kamu-nagara from norito ritual prayers and senmyō imperial edicts.(While senmyō could be traced back to the Nara period,the full texts of norito prayers could not—their patchwork composition containing archaic interpolations made complete Nara-period provenance untenable,yet research might still identify unaltered sections.)This term had undergone such semantic expansion that its meaning ultimately became obscured.They viewed imperial governance as kannagara ("divine will"),using phrases like kamu-nagara omohoshimesu ("to rule through divine accord"),though such usage likely belonged already to its terminal phase.Senmyō texts—oldest among these records—already displayed pseudo-archaic affectations.Kannagara did not express His Majesty's personal sentiments,but rather served when attaching divine agency to imperial actions—phrases like"What one does proceeds not from oneself but from divine will" invariably accompanied edicts bestowed upon subjects."What His Majesty declares,the gods themselves proclaim"—such was its underlying rationale.Most instances retained only kannagara itself,the requisite verb thereafter omitted.Even Man'yōshū employed it with late-stage fluidity while gravitating toward structurally distinct terms like kamu-kara.Historically requiring full constructions("kannagara...suru"),its sustained usage imbued kannagara with autonomous profundity—yet isolating it produced conceptual leaps like"The Way of Kannagara,"whose dissection yielded only partial comprehension.Consider osu-kuni("land of osu")—while osu functioned as an honorific for"eat,"its implications far exceeded mere deference.

As for where the true significance lay in the Emperor’s descent from heaven to this land, His mission had been to cultivate rice for the heavenly gods to partake of.
If one interpreted this matter humanly, it would mean He descended to make the land under heaven His own realm; but ultimately, He came to cultivate the divine rice fields for the gods.
In other words, it was a country that produced what the heavenly gods partook of—a land where He cultivated what was to be offered.
There remained a considerable difference between this meaning and the term "the land where [the gods] partake."
Later, it further differentiated, and they came to use phrases like "the land where [the gods] partake at night," interpreting it as an honorific with the meaning of governance.
In any case, the original meaning had been that it was a country that produced food for the heavenly gods to partake of; to simplistically call this "the land where [they] partake" represented a great leap in language—an elision in thought.
If one did not consider this, ancient language could not be understood.

When they took on formalized structures, their meanings became even clearer. In *norito* ritual prayers, “mikotomochi” was used, and this usage was identical to the etymology of “mikoto.” The *Chronicles* also contain many examples, but in essence, those who conveyed the commands of a revered figure were called “mikotomochi.” The act of issuing divine commands originated from the gods; therefore, the first bearer of divine commands was none other than Amatsu-hiko, who descended to this land by heavenly decree. While this remained until later times, the term “bearer of divine commands” survived only in lower contexts, while in elevated contexts, “mochi” had long since vanished. The reason we respectfully refer to His Majesty the Emperor’s affairs as *Sumera Mikoto* is that it was an abbreviation of *Sumera Mikotomochi*. *Sumera* was a term of absolute respect, and this too later became fixed solely to matters pertaining to the Emperor. However, when this honorific term came to mean merely being used for revered individuals, it gradually descended, and even noble households began using it for “mother’s augustness” or “elder brother’s augustness.” These were merely formal expressions of respect. It must be recognized that such abbreviations were exceedingly abundant.

Generally speaking, in ancient texts, the aspects of linguistic interest differed depending on the text.
In texts such as Kojiki, Nihon Shoki, Man'yōshū, and Fudoki—each perhaps due to their inherited vocabulary—their linguistic preferences appeared to differ.
If we became accustomed to them, we came to perceive distinctions like “this is the language of Kojiki,” “that is the language of Man'yōshū” to such an extent.
Taking “utate” as an example here—a word that by our ordinary understanding seemed rather commonplace—it nevertheless appeared in both Kojiki and Man'yōshū.
In Kojiki, the term appeared to occur in two places; while both were annotated as such in old readings, I believed only one location should be correctly read as utate.
This occurs in the Anaho Palace section of Kojiki’s latter volume: when Ōhatsuse no Ōkimi invited Ichinobe no Oshiha no Ōkimi to go hunting in Ōmi and Oshiha no Ōkimi arrived early the next morning at Ōhatsuse no Ōkimi’s temporary palace—within the words his retainers used to advise His Majesty Ōhatsuse no Ōkimi regarding this matter—

The retainers attending at His Majesty Ōhatsuse no Ōkimi’s temporary palace reported: “Since the Prince utters ominous words, you must exercise caution and fortify your person.” (Ankōki)

“The phrase ‘Since he is a prince who utters ominous things’” constituted a statement describing Prince Oshiha’s character.
When interpreted through Heian-period knowledge, this would be rendered as: “That august person speaks in such a distasteful manner.”
That is to say, they parsed it as “speaking in a distasteful manner” (utateku mo mono iu), but whether this interpretation holds validity remains problematic.
The other instance appears in the Age of the Gods volume, in the account of the contest between Amaterasu Ōmikami and Susanoo no Mikoto, where Susanoo—exulting in his victory—acts violently: “Yet his wicked deeds do not cease; there is an ominousness” (utate ari).
While this reading of utate ari proves ingenious, it ultimately strikes one as a dubious interpretation.
Given that Susanoo no Mikoto’s wicked deeds do not cease, one might have no alternative but to read it as utate ari; however, this interpretation bears distinctly Heian-period stylistic traits, making its existence in Nara-period readings questionable.
The Kojiki must be read in a manner that preserves its archaic essence to the fullest extent possible.
I do not claim this term lacked ancient provenance, but in Nara-period examples it appears as utate—a form abundantly attested in the Man'yōshū, whereas utate ari finds no occurrence there.
This stands as a solitary example without parallel instances; yet as previously expounded, its singular survival occasions no wonder.
In any event, this term does not appear in poetic works.
In the Man'yōshū, for example:

The crescent moon did not appear clearly; clouds gathered—how I wished to see Utatehi (Book 11, 2464).

Though it was not a crescent moon, it remained indistinct.
Therefore, they declared, “I wish to see Utatehi as if the crescent moon were hidden by clouds,” and now said, “These days feel so wretched.”
“How wretched these days are,” they said.
Furthermore,

Though there was no time I was not longing,

Though there was no time when one did not burn with longing, ah—how wretched these days had become, entirely filled with a heart of yearning—such was the meaning.
However, this manner of interpretation might stem from having already grown accustomed to the Heian-period conception of *utate* ("distastefulness").
In Heian-period interpretations, there was no issue in understanding *utate* as carrying that meaning.
This was the era when other forms such as *utateshi* and *utateku* (→*utateki*) emerged.
In modern times, *utate* has taken on various forms, but in antiquity, it had not been systematized.
In ancient contexts, they said *utate ari* ("there exists an ominousness"), which later came to be replaced by *utateshi*.
In any case, interpreting the poems of the *Man'yōshū* in such a manner remains problematic.
Our interpretations are always colored by the significance of eras closer to our own.
In other words, this approach of imposing present-day meanings onto ancient words must inevitably be mistaken.
Whether such a meaning had been employed in the *Man'yōshū* remains problematic.
For it to be used with the same meaning as in the Heian period, there must necessarily have been an intervening change.

On one hand, there exists utatari.
I feel that utate and utata are the same, but the character 転 has already been read as utata.
There was an era when there was a reason to read 転 as *utata*.
Motoori Norinaga’s interpretation is likely not incorrect, but I feel he should have considered it a bit more thoroughly—this is where the issue lies.
In the *Jikyō* dictionary, the character 漸 is annotated in that manner, and they referred to situations that transformed and grew increasingly severe as *utata*—this much at least is clear.
Therefore, the reading is not incorrect, but there are discrepancies in the finer details.
In any case, they call the sense that it has become utterly impossible to comprehend 'utata'.
If we take an example from the Heian period,

Though I have no sorrows to dwell upon, they are drenched.
My sleeves—drenched beyond measure by the dew on the bush clover in the field. (Later Collection of Gleanings, Monk Nōin)

This could not be interpreted according to the ordinary meaning of utate ari.
“Even though there are no sorrows to dwell upon, the sleeves are drenched.”
“However I ponder it, the drenching of these sleeves remains inexplicable—O dew on the bush clover!” Thus, though the wretched condition itself was factual, the aspect of its gradual progression had already been surpassed.
It was simply a terrible state of affairs.
Considering this, it appeared that most utata were considered to be terrible things.
In Susanoo no Mikoto’s case as well, they likely read it with the meaning that there had been unpleasant, forbidden things; but if so, reading it as “utata ari” would have come closer to the truth.
Interpreting it as carrying the progressive meaning that he had done one terrible thing after another until matters became utterly unmanageable, I believed reading it as “utata ari” would be more appropriate.
In the case of Ōhatsuse no Ōkimi, the interpretation that “since he was a person who said unpleasant things” meant “a person who spoke of calamitous things” was problematic.
We, as interpreters, should strive to interpret the Man'yōshū in the Man'yōshū’s own style and the Kojiki in the manner that its author likely intended.
Here as well, I thought it might be acceptable to interpret it simply at the level of “he was a person who said truly terrible things.”

Eight

Another matter that required consideration was how Japanese writing conventions had traditionally omitted words following adverbs. This practice became particularly evident in regulated texts like the *Man'yōshū*, where phrases such as "*utate... ari*" ("there is distastefulness...") were routinely truncated. For instance, the description "[He is] a king who speaks in a distasteful and fearsome manner" likely originated from something akin to: "He is a king said to proclaim profoundly dreadful matters." Given that similar rhetorical techniques existed in the *Man'yōshū*, there could be no doubt that the *Kojiki* employed them as well.

Even regarding the crescent moon poem I had previously analyzed, its meaning might not truly convey "Ah, how detestable" or "repulsive." When examining the "Nanihana mo" poem, at minimum, *utate* clearly modified "the profusion of longing." Abbreviating the full phrase "*utate kono goro koi no shigeshimo*" ("how distasteful these days are with love's profusion") resulted in "*utate kono goro*" ("how distasteful these days"). As archetypal patterns were reiterated from beginning to end, readers came to recognize familiar forms, eliminating the need for exhaustive explication. Any mention of "*utate kono goro*" would immediately signal its truncated origin to informed readers.

This line of analysis revealed that prior interpretations had been colored by Heian-period sensibilities. Through repeated usage over time, *utate* gradually acquired standardized applications. The very constancy of these formulaic expressions naturally solidified their adverbial function. Concurrently, even without explicit mention of "love's profusion," readers intuitively sensed its presence within the phrase "how distasteful these days." Though the precise excised words remained elusive, their general nature and thematic content could still be discerned.

Such elliptical constructions manifested clearly in works like the crescent moon poem. However, this condensation imposed immense semantic burdens on surviving terms. Originally signifying mere intensity ("exceedingly" or "terribly"), *utate* came to absorb the connotations of its omitted complements—evolving into meanings like "exceedingly pitiful" or "intensely loathsome." Etymological inquiries often faltered by neglecting deeper linguistic strata beyond immediate origins. When modern interpretations appeared superficially congruent with contemporary knowledge through happenstance alone, such dissonance proved inevitable—for words and their interpreters existed in fundamentally disconnected temporal realms.

For every etymology above, there lies another beyond. Therefore, even in this example, it cannot be said that we have fully explained *utate*.

There exist instances of *utata* where the succeeding words do not convey the sense of "pitiful." In considering examples from *The Tale of Genji*—(while works like *Genji* and *The Pillow Book* have clearly established dates of composition—for others such as *The Tale of the Hollow Tree* and *The Tale of Sagoromo*, there remain numerous points of contention regarding whether they were truly produced in their purported eras, rendering them unreliable as materials for accurately tracing linguistic history.)

“How utterly dreadful and awe-inspiring this matter is!”
“Why on earth would you go so far as to…?” (*Genji*, Kashiwagi)

“It is an exceedingly pitiful and detestable matter. Why would you go so far as to…”—this was how it was generally interpreted.
When viewing this passage, it appeared that *utate* (“distasteful”) modified *yuyushi* (“awe-inspiring”), though it might have been that *utate* and *yuyushi* stood in apposition.
In other words, since they viewed it as “distasteful yet awe-inspiring,” this interpretation had become conventional.
However, *utate* was in most cases an extreme adverb.
Therefore here too, *yuyushi* had to be qualified as “exceedingly awe-inspiring.”
Since they were effectively utilizing *utata’s* archaic meaning somewhere, that ancient sense would occasionally be reflected upon and revived.
Though *utate* denoted something terrible, they typically applied the “*utate*... *ari*” construction to unpleasant matters.
Once such habitual usage became established with words of this tendency, *utate* alone came to express unpleasantness.
Consequently, the words meant to fill that “*utate*... *ari*” space gradually shifted until they ultimately became adjectives.

Similarly, in the Man'yōshū, there remained vestiges of words that had been used very frequently but had since perished.
This too was one of those inscrutable words, and rendering it as "unreliable" represented the most straightforward interpretation; however, it too appeared to be a term ultimately signifying "terrible...", with the subsequent words following "terrible" having been omitted.
Because they had omitted these modifying words and used [the term] in this way, no matter how much one dissected this alone, there was no way to comprehend it.
"In the dead of night, troubled by thoughts, I find no restful sleep."
*Man'yōshū, Book 5* states, "Unable to sleep peacefully—ah, how terrible!"—a usage where the adverb alone had come to encompass even the meaning of the verb.

This constituted one characteristic of Heian-period language.
The Heian-period literature we know was spoken and read within an extremely narrow society; since the spoken language was written somewhat colloquially and its comprehensible range remained limited, abbreviations became universally understood even when truncated.
Even among the Kyoto nobility, language was systematized within comprehensible limits exclusively for those who frequented the court throughout its various periods.
Thus abbreviations proliferated extensively.
While they created words abundantly, they conversely forgot them in equal measure.
It represented a sort of aphasia—for instance, they employed words like mono indiscriminately.
Admittedly, this may have resolved certain matters.
In any case, what dominated was not so much a concerted effort to omit words as a pervasive force of indolence.
Without considering this principle, one cannot comprehend the utata appearing in Heian-period tales.
Having reached this understanding through analysis, we can finally begin to appreciate how Motoori Norinaga's Heian-style interpretation of "utate ari" as meaning "distasteful" gains partial validity.
In essence, clarity emerges when interpreting it as signifying "terrible."

In the Heian period, among the countless words that appeared, there was yet another: *asamashi*.
I once attempted to compile the vocabulary of *The Tale of Genji*, but what had made me abandon the endeavor was that this *asamashi* appeared so frequently there was simply no end to it.
This too was mostly perceived as meaning "pitiful in the modern sense" or conveying resigned criticism toward others.
Even so, those who read tales and diaries extensively found this unsatisfactory, so we too came to render it as "astonished" or similar terms.
Motoori Norinaga also appeared to have expressed such a view.
In other words, *asamashi* signifies being aware of one’s own superficiality.
It constitutes the adjectivization of *asamu*.
In adverbs as well there existed terms like *asahaka* (→*asamu*), but this was a word expressing astonishment at one’s own narrowness of heart.
Therefore, if one does not translate it as “astonished,” the feeling fails to emerge.

“That such a person should exist in this world,” they marveled, their eyes widening in astonishment.

(The Tale of Genji, “Kiritsubo”)

If one interprets this through a modern lens, it becomes “so pitiful that they were astonished without restraint”—but needless to say, such an interpretation is fundamentally flawed.
It must instead be understood as “they widened their eyes in sheer astonishment.”
Furthermore, in the Yūgao chapter,

“The presence of a person unfolded with such astonishing gentleness,”

This too did not mean it was pitiful to the point of being limp and pliant. The state of affairs remained so astonishing that it stayed serenely magnanimous. Furthermore, there were also examples where *asamashi* carried an even better meaning. In any case, since forms that had been structured as "*asamashiku...nari*" became simplified through abbreviation, ultimately even phrases like "*asamashiku* beautiful" or "*asamashiku* pure" ended up being represented entirely by *asamashi* when words were omitted to their extreme limit. And so, no matter the content, they ended up translating everything as "astonished." By omitting the verbs and adjectives that followed *asamashi*, even in its superficial form, *asamashi* as a conclusive form came to express any meaning; among these, when the sense of "Ah, how detestable!" gained dominance, it ended up taking on the form seen in later periods. In *Genji* as well, there were certainly instances where it carried that meaning, but the *asamashi* of that era was not identical to the *asamashi* of today. In the form *asamashiku...nari*, they had omitted the middle portion and expressed what had been omitted through *asamashi* alone—a point identical to the earlier case with *utate*. That alone indicated all the processes of the mind. In the end, *asamashi* became a word that leaned heavily to one side in later periods. The reason why contempt came to be felt in *asamashi* was, I believe, due to the single word *asa* ("shallow").

Nine

As a more easily understandable example, there was *warinashi*.
This was equivalent to *kotowari nashi* ("inexplicable") and had been used to mean something that could not be explained, described, or expressed in words; this too originally represented the phrase "*warinaku… nari*" through the single term *warinashi*.
What could not be explained became arbitrary; therefore, when one said *warinashi*, it immediately came to mean "arbitrary."
To consider *warinashi* an independent word from its inception was erroneous.
While forced, self-serving usages could not be employed, when everyone used the term, they inevitably infused their own associations into it until it became something that could no longer be dissected.
There was also the example of *itaku*.
When an adverb like "*itaku… nari*" ("greatly… becomes") existed, as experience accumulated with its usage, people came to omit the middle portion and say only things like *itashiya*.
It was an abbreviation of "*itaku… de aru*" ("greatly… to be"), but since *itashi* ("greatly") formed its foundation, it circled back to *itashiya*.

In the conjugation of adjectives, the establishment of the terminal form was actually late; initially, it appeared that the adverbial forms -ku and -shiku emerged and gradually developed from there.
The adjectives we possess had, without our realizing it, come to be called as such only for those that acquired modern conjugational forms, and their structures were systematized.
However, there were those who referred to forms containing *ari*—such as *to ari*, *ku ari*, *tari*, *nari*, and *kari*—as adjectives; while this was merely a matter of convenience in terminology, there was no difference in meaning.
That is to say, in ancient adjectives, the form where the adverbial form had *ari* beneath it and words were inserted in between—the “―ku… ari” structure—had been the complete form of the adjective.
The words inserted in between were often complex.
In this way, adjective phrases were formed; this marked the beginning of Japanese adjectives, eventually leading to the development of the terminal form *-shi* into the *-ku*/*-shiku* forms.

The “―ku… ari” form had long been prepared; by inserting numerous words between these elements, eventually only the “―ku” part became independent as an adverb, and what emerged from discarding “ari” while inserting words in between became precisely the adjective.
Therefore, even in proposals to include *nari*, *tari*, and *kari* among adjectives, there exists at least some basis.

For example, consider *naku ni*, a term with abundant usage examples in the *Man'yōshū*.
In the *Man'yōshū*, it was a fairly popular expression; it likely did not see such prevalence before this anthology, and though it gradually fell into disuse afterward, remnants still lingered during the Heian period.
This formation involved nominalizing through attaching the negative auxiliary verb *nu* with *ku* to create *naku*, then adding the adverbial suffix *ni* (e.g., *omowanu* → *omowanaku* → *omowanaku ni*). While its precise usage form no longer survives, it would originally have been *omowanaku ni ari* (“to be unthinkable”).
By omitting “ari,” compilers settled entirely on “naku ni.”
Since this truncation leaves an unspoken implication, a reactive exclamatory sentiment emerges.
Therefore, while translating this as “even though it does not exist” may not be erroneous per se, it remains an awkward interpretation; rendering it as “nonexistent” constitutes the orthodox approach.
Even within the *Man'yōshū*, examples exist such as “Though I do not call out, [someone] arrives at the gate without taking my intended path...” (Book 9), where one must translate it as “even though [I] do not call.”
As this follows “[someone] arrives at the gate,” it should not be severed as “[I] do not call.”
Other similar examples exist as well—in any case, these already demonstrate evolutionary changes within the anthology itself.

Such changes can also be observed in *wo*.
Though originally an exclamatory particle, *wo* came to be used in inverted emotional contexts—attached where *no ni* ("even though") would be expected—to convey a rebounding meaning akin to "despite being...," and in some cases has even been employed in the grammatical object position, as seen in phrases like "I longingly think of the departing person (*yuku hito wo ba koishiku omou*)."
In any case, words possess this tendency: just when one thinks they have been severed, they continue to the next term and progress through sensation—such is their nature.
Despite being words that omitted *naku ni mo ari* and became distinct, they have come to be used in the sense of "...even though..., nevertheless..."

The particle *ne* followed this same pattern: in phrases like "*hito koso shirane kawaku ma mo nashi*" ("One does not know [it], yet there is no time to wait"), though *shirane* ("does not know") should have grammatically terminated the clause under this principle, it instead continued to the following phrase.

Such phenomena were the result of long-standing linguistic practices. The existence of *naku ni* constructions in the *Man'yōshū* that require translation as "even though..." stemmed from transformations in semantic significance and evolving associations—one must at minimum recognize this particular shift. By the Heian period, people had entirely forgotten what lay beneath *ari*, with all instances transitioning to the "...noni" ("even though") usage.

There remained many such examples, but to cite another instance: when we employ terms like *yuyushi*, their essential meaning does not reside solely within *yuyushi* itself. This term had existed since at least the Nara period as a religious expression conveying a sense of being uttered with restraint. While adverbs generally tend toward abstraction and conceptualization, this word's extreme abstractness meant it could not attain completeness without inserting terms bearing concrete meaning within its structure. Once users grew accustomed to these inserted elements, they began omitting them—letting the standalone term represent the whole—until it naturally achieved independence. A word originally meaning "[something] being to the extent of causing shuddering..." underwent change; simply saying *yuyushi* now sufficed to mean "to the shuddering extent of...". This further branched out to encompass the sense of "unbearable shuddering."

I believe this path and circumstance can also be applied to the term *aware*. While some maintain that terms like *aware* have existed since Takamagahara’s mythical era according to legend, I suspect this cannot hold true upon examination. When considering even just the Japanese language’s depth alone, one recognizes that our people’s history cannot be deemed brief in span. Whether this linguistic antiquity originated after settling our islands or continued from some ancestral homeland remains unknowable—yet through language alone we sense an age far surpassing official chronicles’ recorded years.

Setting that aside—did *aware* truly harbor multiple meanings from its inception? It likely began as a simple exclamation that accumulated substance through usage until becoming rich in implication—a process beyond doubt. To clarify—its structure was “aware… nite ari” (“ah… it is”), with numerous words enclosed between them through practice I believe several emotion types became specifically tied to *aware*, leaving only these impressions indelibly within the term itself This multiplicity allows us now to contemplate *aware*’s content through manifold interpretations.

*Kanashi* too belonged to this category—there had been an era when it was used in the form "*kanashiku… ari*" ("sadly… to be"), with numerous words sandwiched between these elements. These meanings became enfolded within the single term *kanashi*, thereby harboring manifold expressive nuances. To be sure, *kanashi* alone presents minor complications, yet every example raised thus far finds explanation through this framework.

10

In the Heian period, adverbs were highly developed.
The language of the Heian period was the language of the court—that is to say, a form of inner court speech—and was highly refined.
Additionally, the documents of the period were favored and studied from an early time.
At the same time, many songs in the style of that era were composed.
These two reasons caused Japanese grammar to take its departure from the Heian period as its foundation.
While there is a necessity for the language of the Heian period to appear beautiful, it goes without saying that there were also respective eras with beautiful languages outside of the Heian period as well.
In the Nara period as well, beautiful forms that emerged could be observed.

In any case, since what is called inner court speech had been gradually approaching completion since previous eras, it stood to reason that during this period, Japanese adverbs—which had long possessed distinctive characteristics—developed. Adjectives had reasons for not developing. In other words, they formed adjectives through adverbs and *ari*—or rather, constructed what might be called large adjective phrases. Therefore, the development of adverbs—since they became independent from those phrases—amounted to the same as the simultaneous development of adjectives. As proof of this, I presented one final example.

The term *itodo* is a highly distinctive one—

Ever more intensely along the path of one departing, yearning—how envious too are the waves that return— (Ise Monogatari)

*Itodo shiku* (and likewise *itodo mo*) does not mean "severely."
"It means 'even more severely.'"
However, when examining actual usage examples, both *itodo shiku* and *itodo mo* convey "even if matters were otherwise—given that they stand thus—[the situation] becomes all the more..."
In other words, *itodo shiku* does not merely connect to one part of a sentence but relates to multiple positions.

After years had passed, if one were to leave the village long inhabited and depart for Inaba, might Fukakusa Field have become all the more [so]? (*Ise Monogatari*)

Even when examining this poem, *itodo* related to Fukakusa in the phrase "Even if it were not so, [this] is a place with deep grass..." while simultaneously connecting to "[*itodo no ya to ya narinan*]" (might [it] have become all the more a field?).
In this way, among adverbs, there were those that possessed dual functions.
In *The Tale of Genji* as well, the passage “*From the outset dilapidated, the palace within became all the more a fox’s den…*” (*Yomogiu*) described how the Hitachi Palace, which had been in ruins from the beginning, later fell into even greater disrepair.
That is: "Even if it were not so, the palace grounds that had already been in ruins became even more severely…"
In this case, *itodo* related to "*areta ru*" (dilapidated) while simultaneously, grammatically speaking, connecting to "*nari*" (*sumi ka ni nari*—becomes a dwelling).
One could cite countless examples, but in any case, in this manner, they hooked onto both ends—the head and tail.
And in most cases, they did not occupy the complete grammatical positions where those words must attach.
They advanced far ahead to the front of the sentence, so much so that it became unclear which words they modified—thereby allowing their effect to spread widely.
They were striving to break away from the logic of grammatical sentences.
Striving to extend their functions as much as possible, they imparted various nuances along the way.
By no means could one simply state that this word modified this part.
Because they spread out and modified to the fullest extent possible.
In other words, it became evident that efforts had been made to extend and expand the functions of adverbs as much as possible.

When the Late Imperial Court period passed into the samurai era, as the ways of temple households entered and peasants' lives came into prominence, the linguistic center could no longer remain solely as inner court speech.
It became fixed as a literary language, while general language continued its protracted struggle in search of a new center.