Pathway to Japanese Literature

On Adjectives

Discourse on sentences within grammar has been regarded as an exceptionally brilliant endeavor.
This aspect—in its possession of elegant coherence—is perceived with profound meaningfulness, almost as though one might term it grammar's philosophy.
I have always maintained:
Discourse on sentences belongs within psycholinguistics' domain and must emancipate itself from grammatical constraints.
Grammar must ultimately confront its reality—that formalism constitutes both its genesis and terminus.
Even grammar—which among scholarly disciplines should stand as most empirical—has paradoxically developed psychoanalytic tendencies; phenomena descending into theoretical frivolity during analytical processes proliferate precisely because discourse theory's influence permeates all aspects.
My disclosure here of fragments from daily contemplation serves no other purpose.
It stems from conviction that studying lexical organization fundamentally illuminates sentence architecture.

I have always thought that pursuing the genesis of adjectives would simultaneously come to suggest the establishment of a particular facet of Japanese syntactic organization. In what we call adjectives within our national language, that which captures their essence lies in the suffix “shi.”

sorrowful beloved tender dear one

delicate lady sweet land

hard rock lovely child

far, far away land of Koshi

lovely mountain (? Uramiyama)

The form of such compound words left its influence even into later ages, and even in everyday language, they occasionally emerged in expressions that employed archaisms.
However, broadly speaking, regarding the grammar that encompassed elements from both before and after the late dynasty—based on the examples of terminology from the late dynasty that we possessed—it remained more appropriate to delineate periods separately.
In short, until recently, I had considered this "shi" to be part of the root.
To put it plainly, I had long adhered to the view that elements which, in the era immediately preceding the aforementioned grammatical period, were used as root nouns and perceived as genuine gradually came to have their inflection seen solely in this “shi” portion—a perception that intensified until it gave rise to a notion of inflected words.
However, in our ancient texts, while materials from the third of the aforementioned grammatical periods remained abundantly preserved, those from the second period were already scarce, and as for the first period, so little material survived that one could only reconstruct forms from scattered fossils—there being no other method.
From such circumstances, we had to first discover many examples where roots shared a sound resembling *shi* or *s* at their endings.
However, what we were able to investigate amounted merely to instances where one had to hesitate even in determining whether something was simply an adjective suffix or an inflected form of a root.
Therefore, while withholding my own claims on this front, I believed that demonstrating the reasoning behind why I had to adopt such a line of thinking might well furnish even newer perspectives to the study of adjective endings.

I had been considering that there appeared to be rules governing suffix formation applicable to both verbs and adjectives. This occurs when the root ends in a consonant, and its inflection comes to incorporate a vowel. For instance, in verbs, the terminal form first emerges through deriving u-row sounds. And I had considered that even freer inflection was repeated until it came to make one recognize all so-called conjugations. This holds true, at least in the case of verbs. At present, I see no need to revise this conception. Yet simultaneously, I had posited that adjectives too must have originated from the same cause. However, even conceding several premises for this purpose, to properly apprehend the adjective suffix *shi*, we must first understand why there existed roots ending with an *s* sound that shared particular conditions. Furthermore, though verbs occasionally inflect with an *i* sound—while predominantly using *u* sounds—the fact that adjectives employ *i* sounds demands explanation both for why adjectives specifically use *i* and how this *i* became affixed. Nevertheless, there remains a systematically deployed *shi* whose principles we have yet to fully grasp.

A. *tokoshihe* *katashihe*

*inishie* (antiquity)

B. *imashi(ku)* *ketashi(ku)*

*shimashi(ku)*

C. *yasumishishi* *iyoshitatashishi*

*asobashishi*

Regarding these *shi*, some could be explained as so-called suffixes or tension particles. However, this was not a grammatical classification but a rhetorical division. Among these, the examples in (B) were those closest to our modern sense of adjective suffixes and undoubtedly represented an essential form along conjugation’s developmental path. (A) might differ depending on one’s grammatical sensibility but generally seemed substitutable with the genitive “tsu” in modern perception. In other words, I did not think we should simplistically interpret *inishie*’s *shi* as a past auxiliary verb—though no definitive reason to reject this interpretation had yet emerged. When we came to *katashihe*, *katashi* alone formed an almost complete compound word while retaining traces of nominal sensibility. Hypothesizing a form like *tokotsuhe* for *tokoshihe* made discernible how it transformed into *tokoto*. In any case, this group clearly approached the functions of 'no,' 'ga,' and 'tsu.' In example (C), rather than attaching to individual words, [the suffix] gave the impression of following phrases—a perception arising from lengthy words with multiple inflections rather than syntactic reality.

iyoshitatashishi… bow

yasumishishi……Ōkimi

*asobashishi*……*shishi* (?)

In other words, it is precisely because they exist within regulated texts bound by syllabic constraints that the sense of them being compound words is merely lacking.

For example,

*naguhashi* (name-fine) Yoshino Mountain... amidst the clouds' realm, how distant it lay (*Manyoshu*, Book 1).

I took this as an example.
In such cases regarding pillow words' grammatical statuses，we did not question their continuity.
I had come to consider there existed an inseparable relationship between Naguhashi and Yoshinu—to the extent of explicitly stating so.
This was precisely the mistake arising from incorporating mood-based perspectives into grammatical analysis.
Even with identical terms，when examining extended archaic usages like “Naguhashi no Samine no Shima no” (Book 2)，the expression no longer denoted “name-fine” with Yoshino’s precision，but had shifted toward vaguely suggesting nominally pleasant impressions.
In other words，it functioned merely as a toponymic honorific.
What difference would exist between this and “Naguhashi Inami no Umi no Okitsu Nami” (Book 3)?

When we juxtapose these three examples, one naturally discerns continuity between those barely preserving ancient forms and others systematizing grammar through modern sensibilities.
The most advanced theorists of pillow words would likely contend that phrase fixation evokes their perception as makurakotoba.
But in truth, does this not simply reveal chronological shifts—older versus newer conceptual paradigms—between naguhashi Yoshinu and naguhashiki Inami no Umi?
While the former demonstrates profound connection with Yoshinu, the latter undeniably extends naguhashiki's effect even to "[no umi] the sea."

Among those called pillow words and even those not treated as such—excluding special cases—they were in truth merely inaccurately categorized. Therefore, to claim that the aforementioned *yasumishishi* qualifies as a pillow word while others do not is purely a matter of convention. Furthermore, this tendency even blunts our ability to discern whether something constitutes a compound word.

If we shift our perspective once more to examine example (C), the initial *shi* in phrases like *yasumishishi* ought to belong to honorific auxiliary verbs. Here too lies an issue—with constructions like "...sasu" or "...shisu"—which we broadly categorize as archaic-style honorifics: the upper element is understood as a verb meaning "to do" attached to a noun, while the lower element creates honorific inflection.

Though I am truly ashamed of my past errors and present rectifications, I would not mind being ridiculed as merely capricious in my revisions.
I am now daringly considering viewing the upper *shi* as an honorific aspect and the lower *shi*—while still encompassing many nuances—as a formal element constituting compound words.

Even when we speak of pillow words—though they come in many forms—there does exist, broadly speaking, a certain convention that makes us perceive them as such.
A single convention creates a fixed form and becomes unmistakable from other words.
Here, I will not enumerate each instance of that, but the 'wo' particle in question here is also such a case.
“underrobes wo… spring and winter”
“Mihakashi wo… Tsurugi…”
Forms such as “*mikokoro wo… Yoshinu*” (A)—which modern scholars grammatically reinterpret as “O Heart…,” “O Sacred Sword…,” “O Undergarment…,” and the like—likely had counterparts even in the *Manyoshu* era when such interpretations were entertained. Concurrently, there already exist scattered examples where this *wo* appears inclined to be viewed as an accusative case particle.
“Wagimoko wo… Izami no Yama (Hayami Hama…)”
“Imogate wo… Toroshi no Ike” and “Tachibana wo… Moribe no Sato” (B), among others—though there may be some room for debate—the mainstream view considers that consciousness has shifted.
In the same tendency as (A), there exists a form where “*umazake wo… Miwa*” from their era of formation—older but not newer—becomes “*umazake no… Miwa*” (particularly in this example, combining three stages: *umazake*, *umazake wo*, and *umazake no*), alongside “*harubi wo… Kasuga*”.
In contrast, the *Nihon Shoki* preserves the form “harubi no… Kasuga.”
“『Otomera wo... Sode furu yama』 (Maidens... Sleeve-waving Mountain)” appears in the *Manyoshu* also as “『otomera ga... Sode furu yama』 (maidens’ sleeves... Sleeve-waving Mountain).”

Examining this perspective, what should be called the form of *makurakotoba* did not maintain fixed stability. The purported role of *wo* as a later-era exclamatory particle—ultimately rooted in later scholars' perceptions—was not a term that guided usage toward forming compound words, nor indeed toward any other direction. This phenomenon gave rise to new awareness, came to be conceived in a form akin to the accusative case, and with what resembled corresponding predicates, began striving to perfect *makurakotoba*'s function. Yet this represented a pervasive demand across all uses of *wo*—a transitional phase in *Manyoshu* usage examples where attempts shifted from exclamation toward accusative purpose. Conversely, when reconsidered, this *wo* was therefore not an exclamatory particle either. It becomes clear that by maintaining a fixed form, the subsequent elements sought to adopt structures resembling modifiers. That is precisely why it came to be substituted with *no*.

Therefore, precisely because they are noun-like and due to the simple rhetorical constraint of spanning phrases, the functions of the two words that ought to form a compound are not isolated.

However, the linguistic mood that intrudes into such word forms explains this phenomenon and imparts a consciousness akin to ritual refrains.
This is not merely baseless but also a formative mood shaping new grammar—one we cannot dismiss—yet in this case, one must exercise caution, for it risks inducing significant confusion.

Moreover, from such explanations it came to be thought that this 'wo' particle could exist or not exist without causing any disturbance to the grammatical structure. To state it more bluntly: "umazake... Miwa" would have sufficed as it stood. One could say that a vocalic element had intruded into this construct. This was of course not something achieved abruptly; the fact that grammarians—while drawing most citations from regulated texts and their derivatives—had overlooked the major governing force behind this grammar remained a point demanding reflection. In actual practice occupying a position one might term both exclamation and ritual refrain, there now arose a need to rephrase and re-examine this matter slightly.

*ananīyashi yoshieyashi*

*Aoniyoshi Yahoyoshi*

When we omitted such elements and simplified them in our minds, they became "anani," "yoshie," "aoni," and "yahoni." In cases like this—given the nature of these classical words and their conventional usage patterns—we could not avoid taking vocal conventions into consideration. And when we decomposed and reduced this:

anani yoshie

*aoni yahoni*

This was the case.
Though at first glance the first and second lines appeared to belong to largely different categories, they could be treated as one in formal terms.
One might provisionally view "-yashi" and "-yoshi" as ritualistic refrains.
Moreover, between these "-yashi" and "-yoshi" lay a long history, and it went without saying that this very fact had produced developments in usage patterns alongside changes steeped in misunderstandings.
*ananīyashi*
In *yoshieyashi*, while there existed a chronological gap in their eras of use, their terminological applications could be seen as preserving archaic stylistic conventions.
Though *Aoniyoshi* could not be considered significantly differentiated from *Yahoyoshi* in terms of usage examples, it remained certain that even so—sharing the "-yoshi" suffix with terms like *Tamamo yoshi*, *Asamo yoshi* (?), and *Arine yoshi* (?)—portions of its meaning appeared to be shifting under newer interpretations.
We had grown accustomed since our predecessors' time to considering the functions of "-yoshi" and "-yashi" within somewhat constrained parameters.
Yet first, it became necessary to take such examples into account.

Ōwoyoshi Shibitsuku Ama (Kojiki)

*Tareyashi* Hitomo… (Nihongi)

Large fish (?) such as tuna or expressions like “who someone is” (i.e., “who is that person”) could be translated into modern terms in this manner.
It would be simpler still to say “large fish tuna” or “who someone is”—that would suffice.
At the same time, even without “yoshi” or “yashi,” the meaning remains comprehensible.
That is, they merely indicate the process of compound word formation.

Now taking one example of “Yoshieyashi,” he examined it. Whether referred to as “Yoshie” or adjusted phonetically to “Yoshiya,” there was essentially no difference in form or content. Alternatively，it was said that even calling it *“yoshi”* amounted to the same thing. And these were words so abundant in examples that there was no need to enumerate each one individually. Ultimately，when speaking solely of the word itself，it reverted to what was called its root form *yoshi* and was employed.

*hashikeyashi* (*hashikeyoshi*; *Nihon Shoki*)

*hashikiyashi* (*hashikiyoshi*; *Manyoshu*)

When it came to examples of this sort, forms containing the phonetic process of *ke* were generally considered older, while those altered to the sound *ki* were thought to possess a newer consciousness.
This too was an example where archaic forms and newly adjusted ones were employed in parallel.
In other words, since suffixes like *yashi* or *yoshi* had been added to terms such as *hashike* or *hashiki*—which inherently carried an attributive sense—we had to regard them as relatively newer when viewed from the era when these suffixes became attached to their roots.
This too was considered one of the pillow words; however, its application extended from terms like “beloved” and “lord” to “house,” “country,” and “village,” and it had even come to span several phrases.
Moreover, they developed a kind of independent usage when conveying feelings of deep affection, lingering attachment, or obsession.
However, that pertained to matters of content; from a formal perspective, connections such as *hashiki… beloved*, *lord*, *house* were originally clearly distinct.
For example, one could phrase it as “*hashiki imo*” (dear beloved) or “*hashiki sono kimi*” (dear that lord).
However, the term “that” was inserted merely for convenience; I by no means consider that “shi” carries the meaning of “that.”

When we consider it this way, what we call the attributive form must have been surprisingly inorganic grammatically.
In monosyllabic words, there appears to have existed a relationship similar to that between two adjacent words.
Nevertheless, there existed such clear connections that surpassed inflection.

“Iwami no ya… Takatsuno Yama” and “Minato no ya… Ashi ga Naka naru”
The same could be said of “Ōmi no ya… Kagami no Yama.”
This “ya” possessed content in terms of vocal mood but lacked logical significance.
Furthermore, “Ikanaru yo hito ni imase ka” (Buddha’s Footstone Poems) and “Ika aran ya hito no ko yue zo” (Manyoshu, Book 13).
In the case of *“Ama naru ya Ototo Tanabata”* too, there was no doubt that the “*ya*” exhibited no grammatical function.
The so-called emotive *“ya”*

What is sometimes referred to as *“suteya”* appears to lack lexical meaning to such an extent that there are grounds to regard it as a type of ritual refrain. Particularly when viewed as a cutting word in haikai, it becomes evident that this particle interrupts semantic continuity while saturating verses with emotional resonance; yet we must recognize this phenomenon too stems solely from evolving conventions over time.

Class I

○ Takashiru ya… Ame no mikage ameshiru ya… Hi no mikage

○ *Amatobu ya… Keiji… Hirekata Shiki… Tori*

○ *Ama teru ya… Hi no ke ni* (Ama teru… Tsuki)

○ *Oshiteru ya… Naniwa* (Oshiteru… Naniwa)

Class II

○ The maiden sleeps (sounds) ya wooden door

○ Yūzukuhi sashi ya Kawabe

○ Saoshika no Fusu ya Kusamura

○ Sanu-yama ni Utsu ya ono oto

Class III

○kashikoki ya (reverently serving the tomb).
…Kashikokiya, bearing the august command.
…Kashihayagami no Watari)

○How sorrowful! (Uretaki ya) Shiko Hototogisu.
……Alas! Shiko Hototogisu)

Having roughly outlined the three classes of “ya” usage in this manner, I now wished to revisit the same points once more.
Class I was undoubtedly already recognized as pillow words by the Manyoshu era, but still retained a sense akin to mere attributive forms of inflected words.
By adding “ya,” they interrupted the attributive state in form while leaving no change to the attributive function in content. Simultaneously, they sufficiently compensated for the sense of deficiency arising from shifts in rhythmic sensibility.
This natural method not only constituted a new attributive function but also sought to establish a sense of independent form as a pillow word.

When we came to Class II, as the phrases had taken on modifying roles, we could observe that *ya*'s function had further developed, with its grammatical capacities becoming more pronounced.

However, in Class III, beyond the aforementioned usage examples, we saw that distinctive elements had come to be incorporated.

A○ Revered our great lord's awe-inspiring [presence] hung over the Mirror Mountain of Yamashina... (*Manyoshu*, Book 2)

How revered! Were one to hang [it] upon the heavenly gate, only weeping would be wept.
Morning and night (same volume, Book 20)

○ How revered! Bearing the august command, tomorrow sways—rendering the seeds of fear without fear. (Same volume)

B○ …Setting out to sea, at the Crossing of Kashihayagami, the blowing winds did not blow fiercely, the rising waves did not swell greatly, [we proceeded] along the path where frequent waves rose to obstruct… (*Same volume, Book 13*)

…Setting out to sea, the blowing winds did not sweep toward Mihoho, the rising waves did not swell fiercely; at the shore where frequent waves surged near the Crossing of Kashihayagami, placing high mountains apart… (*Same volume*)

A○ How sorrowful! Shiko Hototogisu.
Now is the time—the voice’s dried crab—come and resoundingly sing (same volume, Book 10).

B○… So abundantly here did I protect [this].
How sorrowful! Shiko Hototogisu… Though pursued again and again, still it came crying, only to scatter vainly upon the earth… (*Same volume, Book 8*)

At the very least, those belonging to Class A clearly exhibited “kashikoki…”
These were viewed as adopting compound word forms in expressions such as “...uretake...” (sorrowful...).
In contrast, those of Class B appeared to have effects that extended beyond merely modifying such words.
Namely, “How revered!”
[they] were attempting to approach “How sorrowful!” (uretake kamo).
When we further reflected on this matter, it became apparent that the numerous examples related to *hashikiyashi*—which had originally served merely to form compound words—gradually came to modify their target words at intervals and eventually extended their effect to entire sentences.

a Tomoshiki (?) to shi—merely thus has come longing—the helplessness of my beloved’s heart (*Manyoshu*, Book 5)

b Naminosu Yashi—even such love existed.
When I think of how lovestruck I am, lagging behind you— (*Same volume, Book 12*)

c …When one gazes upon the village, the houses lie in ruins.
Naminosu Iyashi—was it thus?
*Mimorotsukakase-yama—the flowers blooming at the edge of the mountain… (*Same volume, Book 6*)*

d How swiftly! Who could hinder [us]? On the jeweled-spear path, having forgotten the way—my lord does not come (it may perhaps be read as *“hashiki kamo”*).

(Same volume, Book 11)

e Kyūshi (Su?) Yashi—because the wind did not blow, having opened the jeweled comb box—how I regretted it! (Same volume)

These examples not only failed to exhibit forms resembling the attributive but also hardly seemed to modify across separated phrases.
In the case of (a), one could argue—if pressed—that it pertained to my beloved’s heart, but even so, it would have been more appropriate to describe it as a statement addressing the whole.
In b and d, such a state was particularly well observed.
When we came to c and e, they had become completely detached, and the customary phrase *hashikiyashi* appeared to be employed solely for its emotive quality.

Ultimately, the occurrence of such phenomena arose from human rationality toward language.
Old grammar became fixed and had gradually come to lose its accurate understanding.
And yet, they did not cease to persist in adhering to its form.
In that case, the only certain thing was the period-specific sentiment people received.
While decomposing this, we constructed a new grammatical consciousness.
In that case, it would become something quite distant from the primary meaning.
Particularly when used as terminology in literary works, precisely because the language selection function operated, this detachment became even more pronounced.
In cases such as *hashikeyashi* (*hashikeyoshi*, *hashikiyashi*, *hashikiyoshi*), we found an example that retained its vestiges until the very latest period, while on the other hand, many “yashi” forms had become nearly decisively transformed into “yoshi,” unified into a form resembling auxiliary particles of pillow words that merely evoked place names; yet this one alone remained under certain reasons, continuing to be used while preserving its archaic sentiment.
We clearly discerned here the pathway through which the suffixes forming compound words became liberated from the words they should attach to, while extending their influence over sentences or phrases and expanding their expressive capabilities.
Simultaneously, to express this in terms of inflected words, it came to make attributive elements independent in a terminal-form-like manner.

○

While *hashikiyashi* represented the most unique example of usage within this category, upon examining the previously cited examples of *ananiyashi* and *yoshieyashi*, we discerned similar aspects.
In *yoshieyashi*, the *yoshie* could undoubtedly be divided into *yoshi* and *e*.
This "e" was what we call the emotional suffix that attached to the endings of adjectives and verbs.
When we applied modern understanding, it became possible to view this as phonetically corresponding to *yo*.
Therefore, in the case of *yoshieyashi*, might not the *e* and *ya*—which held equivalent value—have been redundantly applied?

The world—*kofishi yoshi e ya*.
Were it thus, I would have become a plum blossom, alas. (Manyoshu, Book 5)

This second phrase could be read as *‘koishike shi e ya’*
or interpreted as “Koihishige ya” / “Koishikeshi e ya”.
Either parsing remained possible.
Yet in all cases, some combined state of *e* and *ya* particles was implied.
Moreover, one could discern how grammatical fossilization had birthed the unified form *shi e ya*.
Within the Manyoshu’s surviving usages of “shi e ya”, nearly all showed no traces of adjective inheritance when excluding this lone example—
they instead took shapes resembling independent interjections or adverbs.
Nevertheless, these appeared to be set phrases first fixed through ancient adjectival bonding customs before drifting free.

If we decompose *ananayashi*, we obtain *anani*, but this *anani* relates to *ayani* (though their etymologies differ). Whether this “ni” holds any special significance cannot be determined today. Both *ana* and *anani* are considered to differ only by the presence or absence of a suffix, allowing interpretation as “Ana! O young man!” This “ni” functioned as an adverbial suffix while also nominalizing the root *ana* in ancient times, even granting it predicative functionality. “How fragrant the cherry blossoms are! Ana!” When observing forms like *anani*, their value remains equivalent to *e*. To this *anani* was added *e*, forming *ananie*. During this process, forms like *ananiya*—as seen in the *Jimmu-ki* (“妍哉,” read here as Anani [Enashi] ya)—emerged, and with the addition of *ya*, it became *ananieya*. Conversely, from forms where “shi” attached to *ananie* (*ananaya*), *ananayashi* arose through phonetic change. The form that incorporated *shi* after undergoing this *eya* process is *yoshieyashi*.

*“ananayashi”*
The category of terms like *yoshieyashi*, in terms of lexical history, likely remained fixed in their relatively ancient form and persisted into later periods; however, when viewed from their usage, they exhibit characteristics resembling newer forms.
But they—having transformed entirely and still unbound by grammatical constraints—were phrases that must be seen as exhibiting vestiges of the root era; yet as their own content gradually became restricted, their forms following other words must have been determined.

The suffixes *“yashi”* and *“yoshi”*, which appear to be fully formed endings, when examined in this light, turn out to be nothing more than mechanical components of modifiers. Moreover, it becomes evident that *“shi”* does not closely adhere to *“ya”* or *“yo”*.

○

○ *Iyushishi* linking the young grasses by the riverbank (*Saimei-ki*)

………………the harmonious grasses by the riverside (*Manyoshu*, Book 16)

○ In Haruga Village, planted water dropwort (Same volume, Book 3)

In Ikaho’s Swamp of Kamitsukeno, planted uekonaegi (Same volume, Book 14)

When creating compound words similar to this without including the modern concept of time that we hold today, there were many cases.
It did not necessarily mean “*shot shishi*” (射られたるしゝ).
There was no need to mention “planted water dropwort” (*ueetaru mizukushi*).
Particularly, the former type could be viewed as a set phrase in ancient songs that had been frequently employed.
But when we came to later examples, even if there existed set phrases like *ue-zuki* (planted zelkova) and *ue-ta-ke* (planted bamboo), there were aspects that did not quite align with modern grammatical concepts.
Some might have perceived this as “naru” based on the special usage of “ni” related to geography, but this was not such a case.
Namely, “In Haruga Village, planted…”
“If one were to say ‘planted in Ikaho’s Swamp…’, even if it lacked some validity, we understood.”
It was the same usage.
To put it further, if we inserted the concept of time, it would have been felt as more appropriate.
In the very act of not employing such methods, one occasionally discerned traces where this particular word alone preserved an archaic grammatical mode.
In other words, this was due to the coincidental constraints inherent to legal texts having accumulated over time.

...in Yamagata, pounding adane of Makishi, dyeing garments in the juice of dye wood... (*Kojiki*, Volume 1).

In such cases as well, one could occasionally see vestiges of such modes of expression remaining.
Namely, if one interpreted it as "garments dyed in the juice of dye wood (or having been dyed)," it became emotionally satisfying.
In documents from the Nara period, such forms were most often discovered in cases of preface poems.
In other words, to explain this psychologically, the transformation between two concepts was being tightly carried out around a pivotal word.
Because of this, it came to be said that unnatural aspects arose in sentences or phrases as such.
However, such techniques did not originally begin as deliberate artifices.
They were merely adhering to old modes of expression.
At such points, when the grammatical consciousness of later periods came into play, they began to consider inserting buffer words between the preceding and following clauses.
In other words, they contained a consciousness equivalent to inserting “that which” (ところの).
Rather than being positioned between two clauses, they isolated the preceding clause from the following one and maintained it in a semi-independent form.

…adane-tsuki dye wood’s juice dye… garment… (*garment dyed in dye wood’s juice* / *garment that is dyed*)

In Ikaho’s Swamp of Kamitsukeno, planted small water dropwort…… (buffer word “planted” small water dropwort / planted small water dropwort)

The connecting points in the examples above do not pose a problem for those adhering to the grammar of previous eras and its forms; nevertheless, they have gradually been organized.
If we were to explain using preface poems as an example:

The brave men, gripping their swift arrows, stood facing and shot; the circular formation was clear to behold (*Manyoshu*, Volume 1, Poem 61).

In the *Ise Fudoki Ibun*, the particle *ya* was inserted at pivotal points in preface poems to act as a buffer.
In this manner, its rephrasing was a reasoned act; “ni”
“no”
The insertion of particles such as *“mo”* and *“nasu”* could be considered secondary in preface poems.
In such poems, while the upper phrase was transformed through metaphorization (or alliteration) to shift its meaning, those inserted particles also set them apart.
However, in the former case, through these particles, they not only fixed the preceding phrase but also connected it with the following one.
In other words, it could be perceived that the consciously recognized *“shi”*—used to link relatively short terms like *“ananayashi,” “yoshieyashi,” “hashikeyashi,” “yasumishishi,”* and *“iyoshitatashishi”*—had intruded [into grammatical structures].
When viewed this way, we found ourselves compelled to reconsider the *“shi”* appearing in ancient texts.

This concerns this category of *“shi”* that appears throughout many poetic passages within Nara-period classics—texts containing sections that transcribe linguistic transmissions from preceding eras. In his paper “A Study on the Conjugated Form *Shi* of the Auxiliary Verb *Ki*” (*Kokugakuin Zasshi*, October 1930), Mr. Imaizumi Tadayoshi had already organized instances of past-tense *“shi”* appearing in the songs of the *Kojiki* and *Nihon Shoki* and sought to consider their formation. He thereby arrived at the hypothesis that this represented the pronoun “that.” It was an excellent and richly suggestive idea. For my part, through repeated examination of these same examples, I needed to draw out implications not only for past auxiliary verbs but also—where possible—for adjectival endings.

Lushly flourishing—the ginger that was planted by the children of Kume at the base of the fence resonates on the tongue... (Kojiki/Nihon Shoki)

When one juxtaposed this example with the “Uekonagi” song, it became clear that it could well have been phrased as “ginger planted at the base of the fence.”

Arikinu’s child of Mie held aloft… In the floating water jewels, the fat that had floated fell and adhered, all rustling… (*Kojiki*)

In this case as well—serving as a homophonic association with *Mizutama-san* (Jeweled Cup)—“floating” emerged and continued to “floating fat,” where “*shi*” had been inserted.
This part of the poem undoubtedly consisted of phrases born from the same tradition of memory as the Ono-koroshima passage in the Jindai no Maki.
The use of “floating fat” and “rustling” demonstrated that these were not novel concepts.
That is to say, one could also phrase it as “floating fat in the Mizutama-san…”
Furthermore, there was the example cited earlier—

...in Yamagata, pounding the spread adane, dyeing garments in the juice of dye wood... (*Kojiki*).

...Yasumishishi—my Great Lord deigned to…; awed by the groaning of the shishi (sick boar), I fled and climbed to the hazel tree branch atop Ari Hill (hazel branch ase wo).
ase wo (*Kojiki/Nihon Shoki*)

The former could still be interpreted as the unified expression "Yamagata ni maki adane."
In later examples, both the *Kojiki* and *Nihon Shoki* present it as “shi.”
When considering this instance, unless special circumstances unknown today come to light, we must conclude it already contained what appears to be a temporal concept.
Were one to attempt rationalization, it might take a form such as: “This hazel—ah, whence comes this hazel? Yasumishishi, my Great Lord’s… the branch of the hazel tree I fled and climbed…”—a structure explaining the talisman’s origin while seemingly imbuing “shi” with an exclamatory quality.
At any rate, even if we cannot definitively call *shi* a past tense marker, its function had clearly undergone transformation.

As another song narrating such origins,

This sacred sake is not my sake—[it is] the sake offered by Sukuna-Mikoto, the deity who dwells in eternal Tokoyo with Kushi-no-Kami, swept with the divine broom, purified through its sweeping, enriched through its abundance, swept back to its origin…… (*Kojiki/Nihon Shoki*).

It seemed that *shi* still contained within it both lamentation and veneration.
In other words, the essence lay in interlacing emotion.

…Tsuginefu Yamashiro Woman grips her hoe; the pounded radish rustles… (*Kojiki/Nihon Shoki*)

Again, the pounded radish—peeling the white of its root—(*Kojiki/Nihon Shoki*)

...In Susukori’s brewed sacred sake… (*Kojiki*)

This too could be termed an exclamation, but one may explicate it through expressions like “struck radish.”

In Kashihara they made yokusu and brewed great sacred sake in the horizontal mortar… (*Kojiki*;
*Nihon Shoki*: kameru……)

Beside the maiden’s bed lay the sword I had placed—the *tsurugi no tachi*… (*Kojiki*).

O ducks of Okitsudori—on Kamodoku Island, my beloved with whom I slept… Never shall I forget… (*Kojiki*).
(*Nihon Shoki*: zuku……I shall never forget)

In these examples, we observe not only the “ni…shi…” structure but also residual traces of *shi*’s insertion.

...My weeping wife, sniffling, lightly brushed [my] skin (*Kojiki/Nihon Shoki*).

Some treat “Kosokoso” as a miswriting of “Kofukoso,” but since the *Nihon Shoki* also records it as “Kosokoso,” “sniffling” suffices.
This interpretation would hold if one viewed it as connected below to “*fure*” (quadrigrade form) via the fifth conjugation, yet further consideration remained necessary.
It appeared to take a form preceding the completed *kakari musubi* binding-particle construction of *koso*, resembling the structure: “My weeping wife. Last night, indeed—[she] easily touched [my] skin—wife.”
That is to say, it signified “the wife who touched [my] skin.”
Even if this had stopped at merely stating *“hadawasure-shi”* (“touched [my] skin”), one could still discern that the emergence of *shi* had been anticipated.

…truly knowing, the two of us slept (*Manyoshu*, Book 2).

It too was a song recounting the same ancient tale, though had it followed the previous song’s phrasing, “ware futari ine” (we two slept) or “ware futari wine” would have sufficed.
This instance likely took the form “ware futari neshiga ware ga” (we two slept—our…).
When considering “easily touched skin” alongside “where we two slept,” one discerns the emotional undercurrent that summoned *shi*.
From this vantage, the Song of the Hazel Tree too must be read as “the boar’s groaning—awed by which I fled and climbed—the boar’s groaning,” while “the branch of the hazel tree atop Ari Hill” perhaps belongs to those called *hayashi kotoba* (ritual refrains).
Juxtaposing this with Yamato Takeru no Mikoto’s previously cited song—a work of parallel circumstance—

Beside the maiden’s bed, the *tsurugi no tachi* I placed.
That great sword—
In Owari, directly facing Otsu no Saki stands a single pine tree.
*Ase wo.*……

That is, we come to recognize that the Song of the Hazel Tree shares an extraordinary stylistic affinity with these examples.

Concerning past expressions, much still required explication; nevertheless, I found myself compelled at present to limit discussion to adjective endings and their developmental trajectories before differentiation.

In any event, the process through which *shi* acquired temporal consciousness inevitably demanded the inclusion of both exclamatory force and retrospective nuance.

Moreover, to achieve clear differentiation from adjective endings, liberation from compound word formation became imperative.

Yet conversely, adjectives themselves appeared—superficially at least—to strive toward establishing independent forms.

Ultimately however, their primordial configuration persisted here unchanged.

Depending on one’s perspective, the origin of the past-tense *shi* could also appear akin to a type of ritual refrain. Moreover, temporarily, it might be viewed as the intrusion of “that.” However, interpreting it as ritual refrains stems from a modern sensibility, while seeing it as “that” may conversely involve reversing the order—perhaps *shi* became fixed first, subsequently evoking the sense of “that.” In essence, what can be stated without reservation is that it constitutes a suffix forming a type of attributive construction. Thus, we came closer to estimating that this *shi* was identical to the *shi* that simultaneously formed adjective endings.

“yasumishi + no”
“asobashi + no”
“Iyoshitatashi + no” where we find “yasumishishi” and “asobashishi”.
If we supposed that *iyoshitatashishi* represented this form, then the phrase *mitatashi no*—appearing three times in the poem by Prince Hinamishiri’s attendants in Book 2 of the *Manyoshu*—might still have been acceptable even without reading it as *mitatashi no*, preserving instead the old gloss *mitatashishi*.
The meaning remained unchanged; it was archaic in style.
Furthermore, regarding the ancient readings in the *Kojiki*, even if the venerable Norinaga’s habit of attaching the honorific “mi” to verbs he annotated seemed excessive—to the point of appearing limitless—one might argue that only those portions marked “mi…shi” (as in Professor Mitsuya’s revised readings) could have been permissible to overlook.
Moreover, precisely because it was archaic in style, one might also have considered that reading it in the same manner as the examples beginning with *yasumishishi* represented the orthodox approach.
However, even in such cases, one had to never under any circumstances interpret the *shi* in forms like *mikashishi*—which appeared to be encompassed within the venerable Motoori’s conception—as indicating past tense.
Even if intuition remained intuition, modern grammatical consciousness and that of antiquity had necessarily differed.
We, who had grown accustomed to perceiving the past through *shi*, might feel that *yasumishishi*, *asobashishi*, and *iyoshitatashishi* indicated the past, and even if our minds were inclined to temporally situate phrases like *uwae shihajikami* or *maki shiadane*, this had to be said to bear no relation to ancient grammar.
Let us now take another such example and attempt to elucidate it through “Nagabosahi.”
Even if this example were not entirely apt, I would add that there were many other examples of this type.

We, knowing the term of praise *kuhashi* (“delicate”), no longer doubt that the ancients perceived *kuhashi* as beauty (*bi*). Yet upon closer examination, might it not be argued that the mere existence of the term *kuhashi* led them to subsume all such instances under the category of “beauty”?

...eyelashes fine as the morning sun, bay reeds delicate as the evening sun.
Spring mountains lushly prospering, autumn mountains’ hues so dear…

(*Manyoshu*, Book 13)

There is a theory that *uraguwashi* forms the etymological origin of *uruwashi*, but in any case, while an aesthetic consciousness indeed operates here, it remains somewhat fluid in nature.
"Flowers delicate; cherry blossoms admired"
The *kuhashi* embedded in set phrases like "fragrant flower tachibana" may appear to function as terms of praise, yet still invites further scrutiny.
Interpretations such as construing "isukuhashi"—which seems related to "kuchira"—through formations like "brave-delicate" (*yūhashi*) strike one as unduly rigid methodologies.
This suggests the existence of *hashi* terms exhibiting usage patterns analogous to *yashi* and *yoshi*.
Having been rationalized into one manifestation of *kuhashi* (細し), this *shi* implies there were numerous other such instances beyond this example.
In other words, we may posit that a category of terms including *yashi* and *yoshi* coexisted with *hashi*, which fortuitously became linked to the form *kuhashi*.
Through such processes, we can surmise there were frequent instances where archaic structures were reconfigured into new modes of lexical formation.

All that had been considered regarding *shi* up to this point were thought—if explained—to be matters that had intruded themselves. Yet among terms believed to show vestiges of Japanese’s root era, there exist those perceived not merely as containing fragments of roots but as constituting roots themselves.

*oshi* *woshi*

*moshi* *ketashi*

*imashi* *shimashi*

When examining records from the *Kamiyo no Maki* onward—an era when the essence of the Age of the Gods still lingered—one frequently encountered compound words beginning with characters such as 押 (*oshi*), 忍 (*shinobi*), or 圧 (*atsu*).

○ Amaoshitarashihiko no Mikoto, Oshisaka no Muraji, Oshihime, Oshiki no Tamakazura

○ Ōshikuma-ō Amaoshihito Oshihomimi no Mikoto

○ Amaatsu Kami

Among these numerous instances of *oshi*, outside of Amaatsu Kami, it was so rare that almost no other examples could be found where *oshi* related to *dai* (“great”) or *ōkishi* (“large”).
Norinaga had also discussed this matter in his *Kojiki-den*. However, when considered in contrast [to his analysis], the significance of *oshi* became somewhat clearer.
In the ancient readings of the *Kojiki* and *Nihon Shoki*, this *oshi* had been treated as an adjective with the conjugated form *oshiki*, but this could not be accepted.
It should not have been interpreted as meaning “love”; rather, all instances belonged entirely to the category of “regret.”
“*oshi* (惜)”
To reductively derive *oshi* from *oshimu* and perceive it as “love” (*ai*) could never have been considered tracing back to ancient meanings.
For instance, consider the interpretive convention—established post-*Manyo kogi* and now broadly accepted—applied to the “Three Mountains Poem”: by reading the upper character 雄 in 「高山波雲根火雄男志等」 as the accusative particle *wo*, then connecting the lower 男 to 志 to form “*oshi*” (i.e., “beloved”), this approach rendered the passage semantically incoherent.
While it held significance as a root, as a compound word it lacked meaning.
It was thought that *oshi* likely underwent inflection to become the adjective *oshi* (惜), thereby acquiring complete semantic significance.
Therefore, ancient examples of *oshi* were not necessarily *oshi* (惜) but included terms such as “important” and “significant,”

○ The preciousness of the lord’s aging, fine as a hair (*Manyoshu*, Book 13)

Sword and long sword—the preciousness of the name, fine as a hair, I have none (*Manyoshu*, Book 12).

○Though your name endures, mine holds no such preciousness—fine as a hair (*Manyoshu*, Book 2).

and other *oshi*
The rest could be interpreted as mere *oshi*, but upon closer consideration, this constituted a matter of great significance.
In other words, as inflected words, they came to partially express such etymological significance without fully aligning with later usage examples.
They did not lose the significance of treasuring *oshi*.
Only when forming compound words with other nouns and verbs in a root-like form did they come to produce a distinctive significance, it was thought.
It was known that such *oshi* and *woshi* further decomposed into *o* and *wo*.
That is, if we applied our established concepts, they corresponded to “large” and “small.”
Compounds such as *okina*, *woguna*, *omina*, and *womina* were formed directly from *o* and *wo*.
“Ōke-ō”
“Hiroke-ō,” “Ōusu,” “Wousu,” “Oji”
*Woji*
There existed terms such as *oba*, *woba*, *ōimi*, and *koimi*.

I still intended to present examples later on, but even regarding *oshi* and *woshi*—which now appeared indecomposable—I considered that junction points nevertheless remained between them, and moreover reflected that many compound words they had formed were lost.
However, while *oshi* persisted as compound words recognized through such forms, I concluded one could perceive *woshi* as having preserved their structure through verbal compounds.
Had this been an era when adjective inflection held slightly greater presence, they would have taken forms like *woyu*, *woku*, *wotsu*, or *wosu*; yet *woshi* failed to develop content of comparable depth.
In contrast to the term *oyu* ("old") discussed in *Shin Kokugaku*, Vol. 2 No. 1, I judged that *wakayu* ("young") emerged as its counterpart far later.

Therefore, in cases such as *oshimu* or the adjective *oshi*, we had to view them as instances where the semantic scope of *oshi* had been extended to a particular threshold. What required consideration here was the form that emerges when removing "shi"—a suffix exclusively used in later-period adjectives and particularly associated with what might be termed long-conjugating adjectives—from word endings. Even if a form like *oyu* were posited to exist, there would necessarily have to be some specialized rationale behind it. To take *oshi* as an example: while my argument might have risked leading one to perceive "shi" as merely a genitive case ending, I had deliberately avoided this easily misunderstood terminology for good reason. All case particles and their grammatical functions had naturally gravitated toward their current forms over time, yet when traced to their origins, every case shared numerous elemental commonalities. This was fundamentally a matter of usage—much like how *no* and *ga* function primarily as nominative or genitive markers while occasionally conveying emotional emphasis.

As I had already discussed and as I mentioned in my previous writings regarding attributive endings, the suffixes “ra” and “ya” form compound words.
・“ka”
I explained that suffixes such as “na” and “ta”—which attach to roots or words of their lineage—created a nominal sensibility through their formation of compound structures for words lacking verbal conjugation.
Ultimately, there could be no doubt these suffixes served as catalysts for forming compounds across both nominal and verbal domains.
However, in adjectives’ case, the very fact that the later-developed suffix *ku* served no function beyond nominal contexts itself came to impart a verbal sensibility.
To that extent, the consciousness of the national language had transformed.

○ Masu-ra-otoko (masu forms synonyms with iyoto, constituting the root meaning "health")

○ Tawa-ya-me

○ Sasa-ra-e + otoko

○ Tawa-ya-kahina

○ Fuwaya-ga-shita

○ Nagoya-ga-shita

×

○ Oda (odashiku)-ya-mu

○ Susu-ro-gu

○Sobi-ya-gu (Sobiyu-ku?)

In this manner, it is precisely within our contemporary grammatical consciousness that we draw a sharp distinction between nouns and verbs; yet in antiquity, even in the combination of roots, such indiscriminate forms existed.

○ How long will this life not endure? (不生有命乎) (*Manyoshu*, Book 12)

○ If only we could meet tonight and never meet again… (*Manyoshu*, Book 10)

○ How long will these inanimate things not endure? (幾時も不生物乎) (*Manyoshu*, Book 9)

Such a phenomenon also existed in *mashi*.
In terms of their value as suffixes, many scholars perceived a gap between *ya*, *ra*, *ta*, *na* and *mashi*, *maji*, *ji*; yet ultimately, from stylistic theory’s perspective, they constituted a unified system.
In classical usage, *mashi* typically attached to *mono*.

○ Had I but brought a woven mat— (Chronicle of Emperor Richū)
○ Had I but piled rocks and died— (*Manyoshu*, Book 2)
○ Had I but become the mountain’s dewdrops— (same book, Vol. 2)

×

If only the island palace could govern the land— (same book, Vol. 2)

If only I could see spring swaying gently— (same book, Vol. 9)

If only tsuge branches were here too— (same book, Vol. 3)

In modern times, such combinations would all be thought to rely on the form 'ki'.
In 'ji' and 'mashi', there existed no such dependency.
Furthermore, even for forms like *rashī* ("seems"), which might appear to belong to the same category—while there existed a form such as *rashiki*—no trace remained of its use in a true attributive function.

Mortals too, seem to vie for their spouses— (*Manyoshu*, Book 1)

○ How fitting!
The children of Soga, whom the Great Lord seemed to have dispatched (Chronicle of Empress Suiko).

Moreover, even in the case of *rashī*,

○ When evening fell, [they] deigned to summon my great lord (Kamioka’s…); when dawn broke, [they] deigned to inquire of Kamioka’s mountain’s crimson leaves.
(*Manyoshu*, Book 2)

There existed somewhat questionable usages such as these.
It was worth noting that this auxiliary verb—which resembled an adjective and possessed what might be called an attributive form in *shi*—belonged to a category that contained at least some notion of time.

Even though they seemed to share the same tendency, we had *moshi*, *gotoshi*, and *ketashi*.
However, regarding this matter, I held a unique perspective.

In the introduction, I had stated that explaining sentence structure should be considered an extension of explaining individual words. There now arose a need to reflect on this assertion. This was no mere chance utterance. It stemmed from being compelled to deeply contemplate matters such as word order through comparative studies with neighboring peoples' languages. While classical grammar and emerging grammar existed in parallel, the emerging form should naturally have gained sufficient authority to be recognized as correct; yet at times, conversely, old-fashioned forms would flourish once again. Simultaneously, this parallel state's prolonged continuation could also be understood through such reasoning. By this same logic, unless one considered that archaic forms persisted even in an age when adjective conjugation had become nearly complete and various transformations had arisen, period grammar studies would prove meaningless. Through this line of reasoning, I wished to argue that even in adjectives—at a time when they had already derived various forms—vestiges of their suffix's primordial morphology from its developmental period must have remained preserved beneath era-specific rationalizations. For instance,

Calm yourself, my beloved / Adorn yourself with vines, my beloved

Loosen, my beloved

Had I but died—Kai’s Black Horse.

That resentful(?) Akiyama—I,

When I behold you whom I have served—

If it takes such forms, the meaning as a vocative case becomes clear.

Without meeting [you], the days had grown long.
How fine it would be to depart these days—how strange, my beloved.
My beloved (*Manyoshu*, Book 4)

Cases like this, while already functioning as vocatives on one level, had become difficult to distinguish from meanings conveyed by phrases such as "longed-for elder brother."
This form from one stage prior appeared in many examples where the state of inseparability remained unclear.
In other words, when an adjective preceded an ordinary vocative, this had generally been interpreted as an inversion of sentence components. However, in reality, during the process leading to this inversion, there had existed a period when *natsukashi ane* ("longed-for elder brother") was not an emotional expression of *ane natsukashi* ("elder brother, how I long for you"), but carried the meaning of *natsukashiki ane* ("the elder brother who is longed for").
In other words, such textual facts were actually extensions of forms found in phrases or individual words.
While I still held many doubts regarding this very point, I believed we must also consider the following circumstances: that the terminal form *shi* of adjectives came to be inflected and independent from its root due to such attributive phrases being inverted (contrary to how we now conceive it) over time, and that through the development of a habit of viewing them with such awareness, this form gradually acquired the tangible quality of a terminal ending.

○

I had already explained the origins of certain sections containing pillow words.
However, there remained pillow words related to this matter.
These were forms that took "-jimono" as their suffix.
This "jimono" was considered distinguishable in two ways through translation.

A. Shishijimono; Kagojimono

Horse-jimono; Dog-jimono

Cormorant-jimono; Duck-jimono

Bird-jimono; Snow-jimono (Dew-jimono?)

Umajimono (— ae “feast”…)

B. Male-jimono; Bedding-jimono

It had come to be categorized in this manner: the former treated some tendency inherent in the noun entirely as a metaphor, while the latter—though generally translated with “as” to align with the preceding category’s renderings—attempted to bring out a slightly divergent nuance.
This alone could also be considered as having arisen not from originally separate entities but through semantic transformation.
In this regard, it became necessary to position *tokojimono* as an intermediate stage within the process and examine it.
Since [the text] described the act of laying grass to sleep during travel, translating it as “like a bed” was indeed possible.
It had generally been considered unclear whether the character 男 should be read as *wotoko* or *wonoko*, but judging from two examples in *Manyoshu*, Book 2, it appeared to have been clearly written as “Karasutoku” [烏徳], so one should read it as *wotoko*.
However, within classical texts themselves there also existed a peculiar pseudo-archaic attitude, and since even significant misunderstandings arising from this were added to the mix, leaving some leeway to call it *wonokojimono* would rather be said to constitute the correct perspective in the history of language development.

Since my discussion has grown somewhat protracted, I must now once more concisely yet vividly summarize my earlier propositions.
This constitutes an examination of how adjective suffixes—particularly the pivotal terminal suffix—developed an independent or fixed sense of grammatical legitimacy.
First: The root era containing *shi*.
Second: The era when *shi* became employed in genitive case usages.
Third: The era witnessing elision of genitive case objects.
Fourth: What I have provisionally termed the independent era of *shi* as a suffix—though regarding this first period, profound doubts now assail me.
Thus my analysis ultimately focused upon the second and third eras.
Yet even concerning this focal point, I have deferred addressing several rather essential matters.

This refers to what has been considered the suffix of a certain class of pillow words—the form called *jimono*. This also shares its origin with the “ji (negative) nani” I previously discussed—indeed with *rashi nani* (the conjectural-interrogative form) through identical developmental pathways. Moreover, I have come to consider that all adjective suffixes underwent what might be called the *shimono* process. When we synthetically combine various *shinani* forms through formal analysis, their convergence into *shimono* becomes unmistakably clear.

Were we to rigorously apply formalist theory, even ancient examples like *“ananiyashi e-otoko”* (“Ah! Splendid man”) or *“yasumishishi ōkimi”* (“Great Lord who reigns in peace”) could not avoid generating awareness of genitive case usage from their base forms like *“yasumishi ōkimi”*. Who could deny that phrases such as *“ananiya no e-otoko”* (“Ah! Splendid man”) or *“yasumishi no ōkimi”* (“Great Lord who reigns in peace”) inherently contain this grammatical consciousness? Proceeding further through particles like “ya,” “ra,” and “ka”—all serving nominalizing functions similar to “na”—we arrive at expressions structured as “...naru mono...” (“...being-thing...”). Could this not have been perceived through what we might call a “...mono...” (“...thing...”) conceptual framework?

What I await—the snipe does not call,

When I come to think this way, my analysis tends to lean somewhat excessively toward logical consistency, yet through this approach, the pathway through which *jimono* forms emerge also becomes discernible.

The rarest instances of *jimono* usage were found in the *Kiki*.
Among these,

*Oh, splendid Nara in its narrow valleys—
where one dives deep into the waters, all gushing forth—
gather forth the young bonito.*
wild boar (*Buretsuki*)

This usage of *shishijimono* appeared considerably different from other pillow words' *shishijimono*. It proved difficult to explain this *shishijimono* as modifying any particular part of *mizukuhekomori*. In such cases, one would do better to interpret them as separate phrases—"*shishijimono*... gather forth"—and read [the phrase] as 'young boar.' When considered thusly, since its form bore some resemblance to *wonokojimono*, it diverged from metaphorical usage altogether. I had come to regard the reason for this *jimono*'s incorporation into song-poetry as lying beyond song-poetry itself—that is to say, I viewed it as having flourished from the Asuka-Fujiwara period onward, when long poems began adopting archaic styles of ritual phrases (auspicious words/norito) within song-poetry's historical development. I had previously written separately on this matter and chose not to repeat those complexities here. To take a relatively recent example: even in the Engishiki norito—ritual prayers we should consider fixed by the early Heian period—traces of this phenomenon could still be discerned.

○ Farewell Prayer.

*Ise-sited… Sumeragi mutsumi Kamurogi・Kamuro mi mikoto ujimono kubi ne tsukinuki…* (Kinensai Harvest Festival)

This was used not only in the Tsukinamisai (Monthly Festival), which followed the same format as the Kinensai (Harvest Festival), but also—for some reason—in the Hirose-gawa Aisai (Hirose River Joint Festival) and Tatsuta Fujin-sai (Tatsuta Wind God Festival).
Moreover, as an example where the word "mono" appeared frequently,

○ In our land’s solemn ritual progression, offerings are reverently prepared for *Myōmyō* (Luminous Sublimity), *Shōmyō* (Radiant Sublimity), *Wamyō* (Harmonious Sublimity), and *Aratake* (Dynamic Sublimity): the mirror for illuminating truths, the jade for contemplation, the bow and arrows for piercing obstacles, the great sword for cleaving adversity, the sacred steed for swift dispatch… all arranged upon ceremonial stands like Mount Yokoyama… (*Sengyaku Sujin-sai*)

Such examples can be cited.
It differs from other Nakatomi norito, and being uniquely Imibe norito, it can be acknowledged that it preserves at least the antiquity of its conceptual approach.
In this respect, while being similar in manner, the Izumo Kuni-no-Miyatsuko Kamuyogoto adopted a somewhat new conceptual approach.

In other words, when associating the Kuni no Miyatsuko family’s *fukōmono* (objects of negative fortune) with incantatory phrases (*jushi*), they had altered the usage examples of “mono.”
In other words, they had styled it as though it were merely a pillow word.
However, as long as one considered them at least as pillow words, it became comprehensible that these two methods—usage examples approximating *jimono* and adopting the metaphorical method of “...no”—operated in parallel.

As the morning sun rises in abundance, the divine rites and vassal rites present the sacred treasures of their own imperial prayers.
The Great August White Hair of White Jewels dwells; the Red Jewels glow crimson; the Jade of Blue Jewels in water channels intertwines… Measuring broadly the Great August Reign of Long Arms to solidify its rule, trampling firmly with the White August Steed… raising aloft deeds that resound loftily to the ears—to govern all under heaven with divine resolve; the Divine Swan’s Living August Ornamental Object; the August Heart of Shidori, too, draws near in kinship… Rejuvenating evermore in youthful marshes… waters dancing with vigor grow abundant… As though beholding the face of the August Mirror pressed close… Lifting high the sacred treasures of prayer to govern with divine resolve—the rites of gods and vassals…

(Izumo Kuni-no-Miyatsuko Kamuyogoto)

An unexpectedly large number of expression methods rooted in incantatory phrases existed within ancient *norito*. Yet while the Izumo tradition’s "*Ujimono*" stood out conspicuously, on the *Kiki* side forms like "*shishijimono*" caught the eye. Moreover, within imperial edicts—regarded as reservoirs of Nara-period grammar—even identical *jimono* forms retained special usages. This could be considered to bear extreme proximity to the second category, unrelated to the first category’s *jimono* containing metaphorical expressions. When positioning *tokojimono* within the process from second to first category, an interpretative framework appeared somewhat attainable. We needed to examine them intending to consider their pre-conjugational forms, unswayed by established adjective paradigms.

○“We, fearing that even one or two oversights might exist among those who serve humbly as Our subjects, and feeling profound shame and contrition, prostrate ourselves before Our Emperor Emeritus’s august presence with reverence [*kowashi-mono*], crawling forward…”

(Imperial Edict, Jinki 6 [729], 8th month, 5th day)

○ In reverently receiving your august command decreed as “*kowashi-mono*” [awe-invoking object], they administer and govern this land under heaven bestowed upon them…

(Imperial Edict, Tenpyō-shōhō 1 [749], 7th month, 2nd day)

When one generally observed these examples, they seemed to be considered as differentiated forms of the Manyoshu’s jimono.
However, if that were the case, one might also simultaneously have thought that it constituted too great a leap.
In any case, for us who were accustomed to conventional examples of adjective usage—forms such as “imajiki no ma,” “mashiji・mashijiki,” “warejiku,” or again “otahishimi”—

We bestowed upon you court ranks and governed [them].
Furthermore, this house—from ancient times under the Lords Fujiwara—bore weighty duties bestowed during the Sacred Emperor’s reign upon the Hajikara Hito-uji clan’s lineage…

*(Imperial Edict, Tenpyō-hōji 3 [759], 6th month, 16th day)*

Is there not a need for us to re-examine the Imperial Edicts demonstrating such usage examples, the numerous old edicts that preceded them, and the state of adjectives that should have appeared in *norito* ritual prayers before the *Kōnin* and *Engi* eras?

…In the evening, I entered and sat lamenting; each time the child I held under my arm cried, I carried and embraced them by my male’s hair, while morning birds wept only their cries—though I yearn… I think of the mountain where my beloved entered as my pillar.

*(Manyoshu*, Volume 3, Takahashi Mushimaro)

…The birds of their own accord rose at dawn and departed; as the setting sun hid itself, the tender child left as a memento by my beloved would longingly cry each time… ………The man of his own accord held [the child] under his arm… …spent days in desolate gloom, passed nights sighing till dawn—though I grieve, I know no recourse; though I yearn, I find no way to meet… …digging through stone roots, I came weighed down…

…Each time the tender child longingly cries… Uho self-thing held under [his] arm… Day after day… Night after night… Though [I] grieve… Though [I] yearn… Digging through stone roots, [I] came weighed down…

(The two poems above, *Manyoshu*, Volume 2, Kakinomoto no Hitomaro)

If I were to forget my thoughts (if it were so), I would dwell here helplessly, longing for the man’s self-thing (*Manyoshu*, Volume 11).

Regarding the previous three examples, I had considered them with some leeway due to doubts about whether "otokojimono" (man’s self-thing) connected in the commonly assumed manner.
Then too, it appeared potentially linked to the verb *ko-u* (to yearn).
Yet on the other hand—differing by but a step from *gotoku* (like)—there was *naru ni*...
One could not deny that expressions like *toshite* had come to contain oppositional nuances.
From such circumstances, the first category of *jimono* must have emerged.
Ultimately, it could be said that the original meaning of the adjectival suffix *-shi* as "*shi-mono*" (shi-object) crystallized into lexical form.
Considering further, there existed fixed expressions using such archaic *jimono*; while these repeated as ancient forms in incantations, edicts, and prayers, new literature adopted and expanded this style.
Thus the *jimono*—previously scarce in analogous examples—gradually developed into what we recognize as the first category.
For instance, while extending *kashikoji-mono*'s form without substantive change, pillow-word methodology likely influenced its expressive effect.
Was not the first category I presented essentially a concretization of *kashikoji-mono*?
Their uniform demonstration of subordination, submission, and humility evidenced this.
Even "Ujimono" or "Inujimono" ultimately constituted pillow-word formations of *kashikoji-mono*, did they not?
Moreover, precisely during this era of burgeoning adjective consciousness, the *-ji* in *jimono* gained directional force.
If permitted slight conjecture, one might say assonance between *kashikoji-mono* and *kakoji-mono* further birthed *shishiji-mono*.
"Ujimono" or "Kamojimono"—all were employed as descriptors of surrender and servitude.

Regarding the etymology of *jimono*, there exist impressionistic explanations such as 'that thing' (*sono mono*) or 'form-thing' (*jōmono*), but such theories neglect any explanation of *mono* itself.
I maintain:
"Mono" remains fundamentally rooted in the concept of spiritual essence.
"Kashikoji-mono" constitutes an 'awe-invoking spirit,' incorporating its formidable efficacy through ritual power, while "otokoji-mono" may be understood as an incantatory term exemplifying spirit summoning via a stalwart male deity’s essence.
As collective consciousness shifted—perceiving awe-inspiring phenomena through *kashiko-shi* and stalwart qualities through *takeo-shi*—this transformation evidently catalyzed new folk etymologies.
To trace this trajectory however，one must consult transitional texts like *Sengaku Tsuina-sai Norito* and *Izumo Kuni-no-Miyatsuko Kamuyogoto*.
These reveal how material objects and spiritual essences operated metaphorically or symbolically—
while simultaneously manifesting ancient beliefs phrased as "...as an expression of [their] spirit" or "...as the abode of [their] spirit."

○

I sought to elucidate the origins of the adjectival suffix *shi* through multiple approaches. Yet much still lies unspoken. My greatest regret remains the unavoidable frequency with which *shi* appears inseparable from lexical roots, compounded by this theory's failure to stand upon the essential foundation of determining root characteristics itself—a discussion that should have preceded it. This foundational work remains wholly undone. To compound my dissatisfaction, I find myself ultimately incapable of fully expounding even *shi*'s establishment as a genitive case ending—the very subject this treatise set out to address.

It is solely through trusting and presuming upon the belief that Professor Kanazawa—who has long watched over my scholarship with compassionate eyes—will find unwritten conclusions even in this paper, that I allow myself to bring this text to a close.