For the spelling of Japanese place-names, modern Western atlases follow the method of the American missionary J.C. Hepburn (1850-1911). For the set of Japanese syllables, which is usually arranged in ten columns of five levels, Hepburn developed a transcription into the Latin alphabet that is based on English for the reproduction of the consonants and on Latin for the reproduction of the vowels. 1
In addition, there is an n as a syllable-final nasal, so-called "darkened" and "half-darkened" sounds (ga gi gu ge go / za ji zu ze zo / da ji zu de do / ba bi bu be bo), and other "broken" sounds (kya kyu kyo / sha shu sho / cha chu cho / nya nyu nyo / hya hyu hyo / mya myu myo / rya ryu ryo, or gya gyu gyo / ja ju jo / bya byu byo / pya pyu pyo). Long, "tensed" consonants are written - with rare exceptions (e.g., Etchû) - doubled: Settsu, Sapporo. Long vowels are distinguished by a macron or a circumflex (Tôkyô, Ôsaka). A long i, as in marine, however, is doubled (Niigata, Kiinokuni) and a long a, pronounced roughly like the a in gaze, is usually written as ei (geisha). Unmarked vowels - e.g., Ise, Mikawa, Hakone, Morioka - are, therefore, always short.
In older European maps of Japan, however, there was no such international standard, which is why we encounter numerous variations of the same name; for the principality of Hizen on Kyûshû, for example, we find the forms "Figen", "Fiien", "FiSen", "FiSien", "Fiseen", "Fisien", "Fisen", "Fidsen", "Fitsen". Several are very corrupted, but others represent the Japanese pronunciation of the time quite well, despite the differences in spelling. Before one attempts to identify geographical designations or judge the work of cartographers, therefore, one has to clarify the linguistic background.
The first thing that interferes with the notation of Japanese is the native language of the European observer. The sound impressions picked up by the ear are compared with the established set of sound patterns, the so-called phonemic system. On hearing new or badly mastered fereign languages, one unconsciously uses the filter of one's native language. One occasionally "misses" a certain phenomenon, because nothing like it exists in one's native language. This was certainly the case with the doubling of consonants, since this is disregarded in many European languages. In Japanese, by contrast, it does make a difference whether, for example, a long search results in the "true aim" (honne) or only a "bone" (hone).
Other "familiar sounding" Japanese sounds are simply assigned to a phonemic pattern in the native language, perhaps not even being noticed as foreign. Take the example of the syllable written as fu in Hepburn's system. In European languages, an f is typically produced by friction on the breath stream between the front teeth and the lower lip, while in Japan, it suffices for the lips to approach each other without touching. In the u that follows, on the other hand, Western lips are rounded, but the lips of theJapanese are not. The situation is similar with ra. The initial sound that Hepburn notates as r is created by a single tap of the tip of the tongue against the ridge of the upper teeth, which Europeans recognize only with great difficulty. Occasionally, in being adapting to one's own phonetic system, certain sounds are conflated. One only has to think of, for example, the difficulties Chinese and Japanese have with word-pairs like lip and rip. Conversely, despite his two-year stay on Dejima, Kaempfer did not succeed in clearly identifying syllable-pairs like cha - ja, chu - ju, cho - jo, chi - ji.
Fig. 40 Cardim, 1646. Detail from cat. no. 30.
Even after someone had more or less grasped the Japanese sound system, its rendering into the Latin alphabet was by no means ensured. Orthography always follows its own traditions. In no culture does it represent the actual pronunciation completely and unambiguously - not even in our modern standardized languages. Take, for example, the long o in the German words ohne, Boot, and los, represented in the first case by adding an h, in the second case by doubling the o, and in the third case by nothing at all. A short o in German is usually followed by a double consonant but not under certain circumstances, as the word sobald shows. English can be similarly confusing: compare, for example, the orthography and pronunciation of she, sea, see or rough, stuff and boat, bold. Never mind French. Furthermore, the evolution of national standards of orthography during the seventeenth and eighteenth centuries was still in its infancy, so the diverse discrepancies found on many maps of Japan are hardly surprising.
Fig. 41 Dudley, 1646. Detail from cat. no. 31.
These problems were joined by a host of errors and mix-ups in Western publishing houses and the offices of scholars, particularly when the data for maps or even completed maps were reworked in a different European country. Hardly anyone had any notion of the original pronunciation. Nevertheless, the spelling of names was often changed to suit one's fellow countrymen. Here and there, some designations remained unchanged - either because they were overlooked or because no one knew what to do with them. Otherwise, it was a fairly mechanical process: the Portuguese xi would be replaced by sci and qui by chi for Italian readers, or in Dutch editions u was replaced with oe and c by k, etc. Typical copying errors include changing a into o, u into n, y into g, 'long-s' into l or t, and T into F, R, or H, or letters being left out. Such charasteristic (mis)spellings can, however, be helpful today when one wishes to trace either the sources of an old map or its dissemination.
To better understand the names on old Japanese maps, therefore, it is best to view them through the appropriate spectacles, be they Portuguese, Italian, Dutch, or German. Then "Satçuma", "Sazzuma", and "Satzuma" or "Cangoxima", "Cangoscima", and "Kangosima", all come together. The Portuguese and Spanish fathers who did missionary work in Japan during the "Christian century" (1549-1639), and who were at home in the language of the country, developed a transliteration into Latin characters based on the syllabic structure of Japanese. Because it is standardized, one quickly gets used to it with only a little practice. Unfortunately, this method is not always used consistently on Iberian maps. The Dutch - the only European nation permitted to maintain a small trading post after the closing of the country - could not measure up to their predecessors where linguistic studies were concerned. A quite stable orthographic tendency can, nevertheless, be seen in the business documents of the Dutch East India Company, or Vereenigde Oostindische Compagnie (VOC), for the designations of Japanese goods, people, and places. Engelbert Kaempfer then tried to build on these forms. Since, as a researcher, he was concerned with reproducing the pronunciation as "correctly" as possible - as he heard it or thought he heard it - he tried out a variety of notations. His frequently copied maps show a phonetic script based on the German orthography of his time. Simply reading them aloud is often helpful here, if one speaks German. Then Kaempfer's attempts do indeed come quite close to theJapanese.
There is, however, another important aspect. Not all of the strange-looking spellings - in comparison to modern maps, that is - are the fault of the observers or adapters of the time. Some of them are a consequence of changes in Japanese pronunciation, for example, in cases like "Fingo" (Higo), "Cangoxima" (Kagoshima), "Midzke" (Mitsuke), "Jedo" or "Yedo" (Edo), or "Quanto" (Kantô). Just a few important instances follow.
Let us take the syllabic column written in Hepburn's style as ha, hi, fu, he, ho. In current pronunciation, the initial sound in ha, he, ho is quite accurately represented by an h. At that time, however, all of these initial sounds were produced by the breath-stream friction described above using the example of fu. Thus, it is no coincidence that Europeans, nearly without exception, chose f for the cases other than fu as well: "Facatta" (Hakata), "Farima" (Harima), "Figen" or "Fisien" (Hizen), and "Fiunga" or "Fioega" (Hyûga).
Fig. 42 Kaempfer/Scheuchzer, 1727. Detail from cat. no. 76A.
The nasalization of the initial g is familiar from current pronunciation. At that time, however, there was also a tendency to nasalize a, o, u, and sometimes i before voiced plosives like g, d, or b. For this reason, we often find an added n placed behind the respective vowels, as in "Nangasaqui" (Nagasaki), "Tsicungo" (Chikugo), "Cangoxima" (Kagoshima), "Finda" (Hida), "Nendo" (Edo), or "Sando" (Sado).
Since roughly the sixteenth century, one observes a weakening of i and u between unvoiced consonants and after unvoiced sibilants. In the orthographical system of the Iberian missionaries, as well as in Hepburn's system, it is not clear that these vowels are barely or not at all heard in these positions. Kaempfer, on the other hand, follows his ear: "Midzke" (Mitsuke), "Famamatz" (Hamamatsu), "Minakutz" (Minakuchi).
Western novels about Japan from the last century introduce the Buddhist divinity Kannon, or Goddess of Mercy, as "Quannon" or "Kwannon". The syllable kwa, which originated from ku and wa, was contracted to ka in the course of the Edo period and has only survived in dialects. It can sometimes be found on old maps, for example "Quanto" (Kantô). This occasionally leads to confusion, as the example of the locality "Quano" (Kuwano) demonstrates.
All over the world, the Japanese currency is spelled yen, even though the spelling en would be closer to the pronunciation. The accompanying weak y sound before initial e no longer exists in the standard language. It was quite accurately perceived and noted by Europeans of the time: "Jezo", "Neso" (Ezo), "Jedo", "Nedo" (Edo), "Yechingo", "Jetsingo" (Echigo).
Fig. 43 Valentyn, 1726. Detail from cat. no. 46
Similarly, the initial sound in the syllable wo, which has become identical with the syllable o in present-day pronunciation, has also disappeared. It was very close to the unrounded Japanese u and traces of it can be seen as vo or wo in many names: "Vosumi" (Ôsumi),"Voxv" (Ôshû), "Suvo" (Suô).
A selection of names from maps with (primarily) Iberian, Italian, German, and Dutch orthography follows. In the context of the phenomena described above, a comparison will quickly reveal the basic tendencies and also several types of typical errors.
In contrast to other parts of the world, where Europeans had scant regard for local names and even on maps gave their own names to rivers, mountains, valleys, and landscapes, the names in Western maps of Japan are largely Japanese ones. The discoverer's posture, which can still be seen in early Portuguese nautical maps of the southern regions around Okinawa, disappeared quite quickly. Only the designation "Santa Clara" for the Uji island group (Ujiguntô) near Kyûshû survived into the nineteenth century. Many of the names that the Dutch devised in 1643, during their explorations of the northern region, were more a consequence of lack of information. Since they were not permitted to land anywhere but Nagasaki, many of the obvious points of orientation that could be seen from out at sea were given Dutch names like "Ronde Holm", "Witte Hoek", "Walvis Bocht", "Lange Hoek", "Bay de Goede Hope". In the "Straet de Vries", between the Kuril Islands of Iturup and Urup, the moment for large gestures was at hand. Since he believed that the regions to the north were unpopulated, Maerten Gerritsz. Vries took possession of "De Companijs Landt" with much pomp (figs. 30, 76). On the return trip to Batavia, he honored his governor-general by naming the "Straat van Diemen" between the southern tip of Kyûshû and the island of Tanegashima. Still, with the exception of some of the names along the northern coast, such designations did not get passed down for long. And one wonders why Philipp Franz von Siebold enlivened his Karte vom Japanischen Reiche (cat. no. 131) of 1840, which was drawn according to Japanese models, with names like "Str. von Krusenstern" (Strait of Krusenstern) and "Str. Broughton" (Strait of Broughton) between Korea and Japan, and also resurrected van Diemen, Linschoten, and Colnett. Nothing of the sort has survived the ages.
1 See Horst Hammitzsch: Japan-Handbuch. Stuttgart 1990, pp. 1565ff.
2 For additional examples, see (a) Günther Wenck: Japanische Phonetik. Band 1-4. Wiesbaden, 1954-59 und (b) Toyama Eiji: Kindai no on"in. Kokugoshi kôza 2 - on"in-shi, moji-shi. Tôkyô 1972.