is a text describing Harima Province (in present-day Hyōgo Prefecture), which was compiled in the early Nara period. The manuscript copy of it dates from near the end of the Heian period, and it was designated a National Treasure in 1965.
Compilation The entry for the second day of the fifth month of the year Wadō 6 (713) in Nihon Shoki records that reports containing the following items were commanded to be submitted from provinces of Japan:
record the names of and with pleasant graphs (Chinese characters) products such as silver, copper, dyestuffs, herbaceous plants, trees, birds, fauna, insects, etc. soil fertility the origins of the names of mountains, rivers, plains and grasslands, etc. tales and unusual stories passed down by the elders
These about the situation in the province eventually came to be called (literally, “records of customs” or “gazetteers”). As , their prefaces were written in a specific format, but in the manuscript this is missing, and is not verified. Nevertheless, it is evident that it was written as a document, as one phrase says “the same as the aforementioned .”
There is no extant documentation that provides the date of compilation of , but it must have been around the first year of Reiki (). The grounds for this are that local government administrative divisions were revised from , and , to , , and in the first year of Reiki () or Reiki 3 (), and employs the former. While it could have been compiled later but adhered to the earlier usage, and there are indeed some inconsistencies in some other nomenclature, the orthography of is without exception in the former style, so that is considered unlikely. This means that the compilation took place between 713 and 715, when the of Harima were , , , and the was .
Traces of deliberate changes are fewer in compared to the other , so the local stories are relatively unadulterated by the editors. It has several distinctive features, such as that the style is Japanese-influenced similar to that of Kojiki; it appears to be a draft; and it is relatively unpretentious in style. In that sense, it contrasts with , which has more aspects.
There are certain local differences in the style of writing. On this basis, the sections are divisible into three groups: (a) , and (Minō); (b) , , and ; and (c) , and (Shisō). Two theories have been proposed for these three groupings. The first is that it mirrors the jurisdiction of the three of , , and prior to the introduction of the ritsuryō legal system, and that their spheres of influence were reflected in the tales and information collected for the compilation of . Another theory is that they reflect the circumstances of the collection of such information by the data collectors of the time. It could have been a difference caused by dividing up then work of collection according to the river catchments and roads along the Kako River, the Ichi River, the Ibo River and the Chikusa River.
Text transmission There is only one extant manuscript of : the scroll copied at the end of the Heian period, which was in the collection of the family. The manuscript is designated a National Treasure, and it is held in the collection of Tenri Central Library. There being no other copies, and because the existence of the family scroll was unknown for a long time, only parts of were known as fragmenta cited in other works. However, Yanagihara Norimitsu seems to have been the first to copy the manuscript, privately, in 1796. Later, Tanimori Yoshiomi verified that it was held by the family, copied it and revealed it to the world in Kaei 5 (). Research into it got going in the Meiji period: produced in 1864, but did not publish it until 1899; Kurita Hiroshi published in 1899; and most-significantly Inoue Michiyasu published his ground-breaking work of scholarship in 1931.
is one of only five more-or-less complete extant , but it is not as complete as . It is missing the introductory overview of Harima province, the whole of the entry for , the heading for (modern-day Kakogawa) and the whole of (modern-day Akō). The reason why it is assumed that the above-mentioned sections have been lost is because each in Harima Province is consistently described in a roughly clockwise sequence from southeast to southwest, northwest to northeast, and the absence of those sections breaks that pattern. Fragments of the entry are cited in , such as the tale, so there is no doubt that was originally included, but there are no clues as to what the situation is regarding the missing section on . Possible reasons are that it was never included, was included but is lost, or that the extant text was a draft from which was omitted. Apart from the above, there is an entry for but it lacks the ’s name heading. It might have been originally included but was omitted in copying; or perhaps was not formed until some time after the compilation of , when it might still have been administered as part of , and was therefore entered in the as (Bay).
Because there is nothing for , and that there are inconsistencies in the section on , etc., it is generally accepted by academics that the extant version of had not yet had its final editing carried out in the . When this theory was first proposed, it was assumed that the whole of the present text was only partially edited, but nowadays the mainstream of thought is that the text derives from when only additional (appended or inserted) passages remained to be edited.
Harima's response to the official Wadō order
Of the five items included in the official order for the compilation of , those to which most conscientiously responds are those relating to soil fertility, the origins of place names, and local stories. Its structure is roughly as follows: the name of the ; the origin of the name; the name; the fertility of the soil in the ; the origins of the name; the names of , mountains, rivers, etc., within the ; the origin of that name; the next name, and so forth. In several of the entries, notable products are listed, and place name changes are noted for several ; but there are fewer of these than are recorded in .
The order unequivocally requests that place names be written with two pleasant Chinese characters. At this time representations of Japanese pronunciations were being written in , which is to say, particular characters were employed as phonograms: for their sound—contemporaneous approximation of Japanese pronunciation—rather than for their meaning. For example, the mountainous district called self-evidently means, literally, “High District.” The place name could possibly have been written with the graph meaning "high", but the order specifies that place names should be expressed with two graphs. Thus the name was written with the characters for their sound, combined with the auspicious meanings.
The compilers of complied quite consistently with the order to use pleasant graphs. An extreme example of this appears, which says that the entry explaining the origin of the place name , which literally means “Living Moor”, was originally ; and that it was changed because said it was too “bad” as a place name. From this entry it may be inferred that the practice of rendering place names with pleasant meanings had already been preferred for a long time, and therefore that the order was merely formalising an established custom. This preference reflected superstitious belief in , the spirit of speech, a kind of magical thinking whereby the utterance of a word was thought to be sufficiently powerful to bring about or conjure up the thing spoken of.
The fertility of the soil was a question no doubt raised by the Nara Court for cadastral purposes. People subjected to taxation tend to respond with circumspection, and the issue is avoided in the for other provinces. In , however, soil fertility is recorded in considerable detail: in almost all the the soil is evaluated according to 9 categories using upper, medium and lower, from “upper upper” to “lower lower”. None is actually assessed as "upper upper”, and given that no other extant texts provide this information at all, it may well be that in Harima they recorded all the fertility levels as one lower than in actuality to minimise their tax levy.
Tanaka and Matsushita (1994) compared 's assessment with the yields of rice per in the relevant for 1885, which found that with few exceptions there was a fairly good correspondence between the two. The exceptions were and , where in the report for was lower and that for was higher. Many reservoirs were constructed in in the Meiji period, which increased yields, and that may account for the discrepancy.
The discrepancy for remains unexplained. The evaluation might have been based on the taste of the rice that the land produced. It took into consideration the geomorphology and drainage of each in order to determine the flavour of the rice. It found that the with a high grade were places with clay soils and good irrigation, while those with a low grade were flood-prone soils with poor drainage. This method also showed a good correlation, and in the various in which were not explained previously, a higher grade was evident. is located in the westernmost part of Harima Province, and it may have been given a higher evaluation than other due to its remoteness.
Place names
More than 360 place names are recorded in . The origins of place names and local tales, myths and legends form the bulk of the text, as in the other , and that accords with the two items of the government order pertaining to the origins of the names of mountains, rivers, plains and grasslands, and the stories passed down by the elders. There is no strict division between the two items, and all but two of the local tales are linked to place name origins.
Apart from the names of , , , mountains, rivers, plains, moors (uncultivated land), place names include , wells, harbours, shrines, etc. There are 81 , and contains the most, with 18. One is estimated to contain a population of about 1,000 people, which means that would have had a population of about 18,000. The total number of at the time of the compilation of is unknown on account of the missing sections for Akashi and Akaho, but judging by there were about 95 , making the population close to 100,000.
The entries which record the origins of place names range from simple ones that say “because of x it came to be called y;” to those which are more narrative, specifying a time period, protagonists, and what they did; to those which link all such elements together.
Some place names are straightforwardly descriptive, such as . Others evoke their silhouette on the landscape, such as . On the other hand, resided in the place name, and revelation of it to outsiders denoted surrender of local autonomy to external authorities. The aims of the Nara government in recording place names were not only for pragmatic cadastral purposes, but also to consolidate ideological control over the newly-formed nation state.
Entries in which a person is named include those that are actually about that person and those that specify a time period (e.g., “in the time of King…”). In they include deities, kings (emperors) and royalty, nobles and commoners. Entries for the origins of place names that involve a deity or a king seem to be included with the intention of stressing that this place had a distinguished history. Whereas stories in are predominantly about deities, and those in , and are more about people, including kings, those in are a mixture of both.
Some place name stories about deities recount that the very existence of the deity there is its origin, while others are because the deity did something there. For example, they may have said something, dropped something, tussled with another deity for occupation—with an origin story accordingly. Local deities such as frequently appear; he is unique to Harima.