Corps de l’article

Introduction

Carl Heinrich Merck (K. G. Merk in Russian), of the pharmaceutical dynasty, was a young ship’s surgeon and naturalist on the Billings Expedition in Chukotka in 1791. His Beschreibung der Tschucktschi (‘Description of the Chukchi’[1]) is rich with first-hand first-rate ethnographic observation, including throughout his text words of special cultural interest from the indigenous languages. Merck’s statement about four “Sedentary Chukchi” (Eskimo) languages or language varieties along the coast of Chukotka in 1791 is thoroughly remarkable and worthy of careful interpretation. In order to identify the four languages, the present article analyzes their description and geographical distribution by Merck. The second half of the paper claims that part of the fourth language, startlingly enough, was North Alaskan Inupiaq, spoken intermittently in pockets between Kolyuchin and Shelagski Cape, at least since the opening of Russian posts at Kolyma and into the 19th century, by north Alaskans from the Point Hope area, who also used Wrangel Island as a stopping place.

Merck’s statement on Eskimo languages in Asia, 1791

The “Description of the Chukchi” is presumably a secondary work (unlike Merck’s journal notes on the Koniag, for example—or the rest of the manuscript in which the “Description…” is found, for that matter). The “Description…” must be derived from notes that have so far not come to light, and is written in his elegant German longhand Schrift, preserved at the St. Petersburg Public Library Manuscript Division. On leaf 3 recto is a startlingly detailed statement of the distribution of the languages of the “Sedentary Chukchi,” i.e. Asiatic Eskimo. The “Description…” has never been published in the original German, but only in Russian—heroically, and almost completely, with the exception of parts of leaf 32 recto and verso, by Titova (1978). The first part of the language statement was also published in Russian by Vdovin (1954: 76-77). For our purposes the full statement is indeed worth quoting. It follows here first in German, in my decipherment, with the help of Titova’s Russian, then in English with some detailed discussion.

In Ansehung der Sprache unterscheiden sich die stillsitzenden Tschucktschi ebenfalls von den Renthier-haltenden; letzterer Sprache ist von der Koriäckischen nur in _________ unterschieden, und diese verstehn die stillsitzende zwar alle, doch haben sie ihre eigene von der ersten völlig entfernte Sprache, welche wiederum zu vier Mundarten abweicht; eine dieser Untersprachen erstreckt sich von Serdze Kamen oder vom kleinen Eyland Mantschchen biss vorn Wohnplaz Uigin; die zweite von selben bis zum Wohnplatz Puuchta, der ein wenig nördlich der St. Laurents Bucht liegt; die dritte (Paekniskoi) ist vom letzter Wohnplaz bis zur nordöstlichen Land-Ecke die sie Mengihenitkin, mehr aber nach den zwei Wohnplätzen dasselbst, als Nuchin und einem südlicher liegenden Peäky, benennen; die vierte, Uwelenskije, von eben angeführter Landspitze bis zum letzten Wohnplatz vor der Schalaginitschen Land-Ecke. Was sonst diese Sprache betriefft, so soll die Mundart derer, die auf der nordöstlichen Landspitze wohnen, sich der Amerikanischen nähern, welche letztere auch die eigentliche der Eyländer des Canals ist, doch verstehn diese meist auch noch die der Tschucktschen: die bewohnten Eylande die im Canal oder in berrings-strasse liegen, sind drei: 1) Imaglin oder Imäklin (u, nach der uwelenskischeschen Mundart, wonach auch das Wörterbuch); 2) Ingelin, wobei noch ein kleiner unbewohnter Eyland nahmens Ukijen: 3) Okipen; diese Eyländer sind den Amerikanern, was die stilsitzende Tschucktschi den Renthier besitzenden sind, ihnen ergeben und sich vor ihnen fürchtend, nur sind die von Imaglin der Tschucktschi mehr zugethan. Noch in Ansehung der Sprache der stillsitzenden, sprechen die letztere an der eigentlichen Tschuckischen landEcke wohnenden mit den Insulanere von St. Laurents-Eyland (Eiwugi-en, der Tschuktschi) fast einerlei Sprache; zu lezterem Eyland ist von der landEcke zu einen Tag überrudern […].

Merck 1791: 3

A fairly close English translation follows, with parenthesized capital letters correspondingly referring to lettered commentaries (A-H) below.

With regard to language the Sedentary Chukchi also differ from the Reindeer Chukchi; the latter’s language differs from the Koriak only in __________ [minor ways] (A). The Sedentary Chukchi actually do all understand them, but they have their own language entirely distinct from the Reindeer Chukchi, which [Sedentary] in turn diverges into four dialects [Mundarten] (B); one of these varieties [Untersprachen] stretches from Serdtse Kamen’ or from the small island Mantschchen to [a point somewhat] before the dwelling-place Uigin (C); the second from there to the dwelling-place Puuchta, which lies a little north of St. Lawrence Bay (D); the third, Paekniskoi, is from the latter dwelling-place to the Northeast Cape which they call Mengihenitkin, but more [precisely] still the two dwelling-places thereof, i.e.Nuchin and south of that Peäky (E); the fourth, Uwelenski, from the above-mentioned point of land to the last dwelling-place before Shelagski Cape (F). As far as this language is otherwise concerned, the dialect of those that live on the Northeast Cape is supposed to approach the American, which latter also is proper to the islanders of the Strait, yet these also mostly do understand that of the Reindeer Chukchi. The inhabited islands in the Bering Strait are three: 1) Imaglin or Imäklin (G) (u, in the Uwelenski dialect, in which the vocabulary [was compiled] (H)); 2) Ingelin, near which is still another small uninhabited island, named Ukijen; 3) Okipen; these islanders are to the Americans what the Sedentary Chukchi are to the Reindeer Chukchi, dependent on them and in fear of them, except that those of Imaglin are in closer contact with the Chukchi. Further with regard to the language of the Sedentary Chukchi, these latter living on the Chukchi Cape proper speak with the islanders of St. Lawrence Island (Eiwugi-en, of the Chukchi), almost the same language; to the latter island from the Cape is one day’s row […].

Merck 1791: 3

(A) That sentence finishes with one illegible word. The Russian translation is possibly a guess. Two Germans educated in Schrift were unable to decipher the form, or to recognize any negative prefix in it.

(B) Merck’s terminology here shows an insightful view of language development (“abweicht / diverges”) and of sociolinguistics (e.g., bilingualism). His terminology for the four varieties of Asiatic Eskimo is inconsistent, indecisive, or even evasive, seeming interchangeably to use “Mundart / dialect,” “Untersprache / sub-language,” as well as “Sprache / language”—understandably, given the relative closeness of the Eskimo languages in the larger perspective with Chukchi-Koryak. Currently, three of Merck’s Eskimo speech-varieties and part of the fourth would by strict criteria of inherent mutual intelligibility (i.e. without practice) be considered four languages.

(C) Language number one is clearly Sirenikski, occupying in 1791 most of the south shore of the Chukchi Peninsula, from Kamen’ Serdtse (as on Kobelev’s 1779 map in Efimov 1964: map 174), or the eastern end of Mechkyn (a 50-mile long sandspit stretching west nearly to Kresta Bay), eastward to a point “vorn / before” (skipped in the Russian translations) Uigin, Chukchi Ungiin, for Chaplinski Ungaziq. In modern documented toponymy the westernmost Sirenikski placename is Nunligran (from Sirenikski [núneleg-] ‘having habitations’ plus further Chukchi suffix –ran; cf. Chaplinski [nunáleg-]), two-thirds of the way to Mechkyn; “before” Ungaziq allows us to put the Sirenikski-Chaplinski line precisely where it was in the 20th century. The Sirenikski language became extinct with the death of Wye in 1997. It survived long enough to be at least partially documented. For an extraordinary account of its decline, see Krupnik (1991). It was named for the Chaplinski variant of the name of the village where it was last spoken, [sighínek] (cf. Sirenikski [sígheneg]; Russian Sireniki pl.), and replaced by Chaplinski. Sirenikski was sharply divergent from Chaplinski or any Yupik, so much so that it was either a third branch of Eskimo altogether, or perhaps more likely, a non-Inuit coordinate branch to all Yupik. The fundamental genetic position of Sirenikski leads us to conclude that Asiatic Eskimo was never a dialect continuum in Asia, but rather is the result of distinct waves of settlement there, even such that unlike Chaplinski and especially Naukanski as refluxes from Yupik Alaska, Sirenikski may well be one that never was in Alaska, or at least returned thence far earlier.

(D) Language number two is clearly Chaplinski (“Central Siberian Yupik” or CSY in the American literature, to include St. Lawrence Island). The former northern border between that and Naukanski was indeed in the area of Puuchta (Pouten; Chaplinski Puvugta, Naukanski Piighta; for details see Krauss 2004). For the recession of Chaplinski toward the south, see Arutiunov et al. (1982) and Chlenov and Krupnik (1983). Sireniki and (Cold War 1958 displaced “Novoe”) Chaplino in Chukotka, and Gambell and Savoonga on St. Lawrence Island, are all that is left. Chaplinski nevertheless remained the majority Asiatic Eskimo language, and became the “official” Soviet Eskimo literary language for schoolbooks in 1931.

(E) Language number three, at the very southeast corner of East Cape massif, Chukchi Pèèk, from which Merck’s language name is derived, perfectly fits Naukanski (Merck’s Nuchin; Chukchi Neuqan, from the Chaplinski variant of the name for the main village site, Nevuqaq; Naukanski Nuvuqaq). The south end of Nuvuqaq is Cape Pèèk, and on the precipitous south side of that was once the related habitation Nunak. Naukan was itself forcibly closed in Cold War 1958, and the people dispersed to St Lawrence Bay and Uelen. In spite of these misfortunes and the official dominance of Chaplinski, the Naukantsy struggle to maintain their distinct identity. The linguistic difference between Naukanski and Chaplinski, it should be noted, is far less than that between Chaplinski and Sirenikski, but the social distance is sharply maintained. More about the historical distance below, under (F).

For the exact territory and history of the Naukantsy, see the recent Naukan dictionary (Dobrieva et al. 2004), including the section on placenames (Krauss 2004). At the time of writing this paper, this writer was unaware of the forthcoming “How far west into Asia have Eskimo languages been spoken, and which ones?” (Fortescue 2004). In an otherwise fine paper, Fortescue mistakenly considers Klaproth’s 1823 partial reprinting (from Russian and German printed sources, including Sarychev 1811) of Rohbeck’s Naukan vocabulary to be a partially new source, not having seen the manuscript versions thereof, of which the Klaproth list is strictly a subset, including absolutely nothing not originally in Rohbeck. Fortescue’s argument that the Klaproth version is “shorter” obviously does not help. Further, as Fortescue (2004: 162-170) himself suspects, Klaproth confuses his source or channel with Koshelev, who had in fact been near Anadyr, but nevertheless tentatively concludes that that vocabulary shows Naukan was once spoken near Anadyr—a totally unjustifiable claim.

(F) Merck’s supposed language number four, Uwelenski, takes its name of course from Uelen (Naukanski adaptation: Ualeq; Chaplinski: Uuleq), now a major Chukchi village or town, but in 1791 it must still have been at least in large part a Yupik-speaking village. Herewith ends, however, the simple correspondence between Merck’s four language varieties and what is actually documented. In fact, we have Merck himself to thank—as luck would have it—for nearly all the documentation we have of Uwelenski. We lack clear record of where Merck was and where he had his contact with it, but throughout his “Description of the Chukchi” he includes the Uwelenski (not Chaplinski, not Naukanski) Yupik word for items of cultural interest along with the Chukchi word, in about 70 cases, and those constitute nearly our whole record of Uwelenski. These were published in Titova’s 1978 translation of the text, and at the end she adds a list of them, including those from the missing parts of leaf 32.

Both this writer and Mikhail Chlenov originally and separately examined this list with care, each coming to the conclusion that Merck’s Uwelenski was no Asiatic Eskimo language number four, but rather another dialect of language number two, the Chaplinski-St. Lawrence Island complex (CSY). This writer’s detailed linguistic analysis, showing which items are more like Chaplinski, which are more like Naukanski (a small minority), and which could equally be either—is to be published elsewhere. The writer takes the opportunity to thank Chlenov for some all-important fundamental insights especially regarding a seemingly discontinuous distribution Chaplinski-Naukanski-Uwelenski (= Chaplinski), i.e. 2-3-2, and not 2-3-4. In addition to looking at maps, Chlenov had the personal advantage of visiting East Cape and experiencing the normal human view of it, from sea-level. He pointed out to me that from that human point of view, the East Cape massif is much more like a third Diomede, so barely a part of the Asian mainland that the Chaplintsy in travelling to visit the speakers of their language at Uelen and on the Arctic coast often dragged their boats across the lowland isthmus behind the massif, rather than go around it, a longer route and riskier, given the currents and perhaps also lack of positive relations with the Naukantsy. Thus the practically offshore Naukantsy do not interrupt the mainland Chaplinski-Uwelenski continuum. In fact, as this writer has explained elsewhere (Krauss 1993, 2000), the Naukantsy are in some ways more Alaskan than Asian, not least, for example, in that their names for the Diomedes, Imaqliq for Big Diomede, meaning ‘the more oceanward one,’ and Ialiq for Little Diomede, meaning ‘the entirely visible one of definite shape,’ are the same as the Alaskan Inupiaq Imaqliq and Ingaliq. In other words, the Naukantsy still speak of the Diomedes as though seen from Alaska. They represent the last reflux to (or toward!) Asia from Alaska, perhaps hardly antedating the Inupiaq expansion over Seward Peninsula, and perhaps connected with that. Their relations with Alaskan Inupiat remained relatively close until the Cold War, those with the Chaplintsy remaining relatively distant.

Naukan aside—quite literally—then, the distribution of CSY was apparently a Z-shaped continuum, with St. Lawrence Island at one end, on to the whole east coast of Chukotka, then behind the Naukan massif to Uelen and west to some point along the Arctic coast, to be considered below (Figure 1). Throughout this discussion of a “CSY continuum,” note that this should be qualified in that it ignores any presence on the Chukotka coast of maritime Chukchi, who of course were also present and by now dominate that coast to the extent that only small pockets of Eskimo remain, themselves mixed with Chukchi.

Merck goes on to say that the language of St. Lawrence Island is “almost” the same as Chaplinski. In 1878-79 the island suffered a catastrophic famine/epidemic that almost entirely exterminated all the villages around the island except the two at the western end closest to Chukotka. We have no direct documentation of the possible island dialects that were lost. However, Lloyd Oovi (Uvi, 1896-1980) told the writer (in 1971), that his mother, Qisgenga (1862?-1935), told him that as a child, before 1879, she had visited Kiyalighaq at the southeast cape of the island and that the Kiyalighaghmiit spoke “differently.” Island elders also remembered that the speech of one of the very few survivors of the catastrophe, Waalla of Kiyalighaq (ca. 1845-ca. 1917) was somewhat “different.” The speech of early 20th century Gambell and Pugughileq was virtually identical with that of Chaplino (Ungaziq), remarkably so, hardly justifying Merck’s “almost” the same; perhaps the above reports explain it.

On the other hand, there was indeed some significant low-level dialect diversity all along the CSY continuum, evidently including St. Lawrence Island, and on the east coast of Chukotka (partly documented), thence on to Uwelenski. This continuum is such that there are however also startling similarities between the St. Lawrence Island and 1791 Uwelenski ends (most especially the month-names), explainable at least in part as shared retentions lost in Chaplinski in between, but perhaps also partly explainable by some direct contact, for which see below. This still coherent CSY dialect continuum might be characteristic of a spread somewhat under a thousand years old. In this connection, note that Sergei Arutiunov (in Dolgikh 1967: 13, 1972: 23, footnote 10) speculates that the spread of CSY may have started from St. Lawrence Island, thence to the Chukotka mainland. This linguist adds that Sivuqaq, the name of Gambell and of the whole island, and that of Ungaziq directly across from that, seem to exhibit a St. Lawrence Island perspective, being derived from sivu- ‘forward’ (as looking over the prow of a boat), and unga- ‘across over there.’

How and when Proto-CSY got to St. Lawrence Island in the first place we leave to further speculation. Before going on to—and beyond—the western limit of Uwelenski on the Arctic coast, it is important to emphasize the larger perspective that the Yupik languages themselves form a (broken) continuum, with CSY at one end, and with Naukan as a nearly mid-way stepping stone on toward Central Alaskan Yup'ik of Norton Sound to Bristol Bay (and Alutiiq beyond). The connection between St. Lawrence Island and Alaska thus clearly lies across Seward Peninsula and Chukotka, via Naukanski—leaving aside, in turn, the relatively recent Inupiaq southward intrusion across Seward Peninsula and its islands (Diomedes and King Island), perhaps contemporary with the westward move of Naukanski to the “third Diomede.” Merck does note that Naukanski “approaches” the American of the Diomedes, and this is a bit puzzling. Either it would imply that the Diomedes were still pre-Inupiaq and were occupied by something more resembling Naukanski, or the still less likely alternative that his source was sensitive to the relatively subtle ways that Naukanski has, for example, lost much of the accentual prosodic traits shared by other Yupik—this indeed under the influence of Alaskan Inupiaq, with whom the Naukantsy have been in much more contact in recent centuries after all than with any other Yupik. Most likely, here Merck mistakes the sociolinguistic for the linguistic, his statement merely reflecting the high degree of bilingualism expectable from the close Naukan-Inupiaq contact.

(G) Imaglin or Imäklin is actually the Chukchi variant of Imaqliq (the same in CSY, Naukanski, and Inupiaq), ‘the more oceanward one,’ clearly from the American visual perspective, as noted above. Likewise below, Ingelin is the Chukchi variant of Ingaliq, either the Inupiaq or CSY version of the name, but not the Naukan, which is Ialiq, meaning as noted above, ‘the fully visible one of definite shape,’ paired with Imaqliq, strictly from the American viewpoint. Ukijen and Okipen, on the other hand, can be the Chukchi variant only of CSY or Naukan Ukiiyaq and Ukiivak, not Inupiaq Ugiiyaq and Ugiuvaq for Fairway Rock and King Island, respectively. Given that only CSY could be the original language of all these island names, the simplest explanation is that they are, like Imaqliq, also all from the Uwelenski dialect of CSY.

(H) Merck here specifies “u” for Uwelenski, his definitive statement for the abbreviation used throughout the text. Merck’s reference to the “Wörterbuch / vocabulary” is problematical, especially placed here. It can hardly refer to the Uwelenski words throughout Merck’s text, but only to a more formal vocabulary. Merck for his part was not in the habit of compiling formal standard vocabularies; that seemed to be the job of his colleague Rohbeck. Merck’s association of Rohbeck’s vocabulary with Imaqliq shows at least vagueness about its origin, and contradicts his use of “u.” In any case we have no such vocabulary by anyone of either Uwelenski, the supposed language of the name Imäklin, or of Big Diomede Inupiaq, to which alternative antecedent Merck might also be referring, unless of course such a vocabulary existed and was lost. The formal vocabulary we do have is a pure Naukanski vocabulary by Rohbeck, published first in Russian by Sarychev (1811), and a German manuscript also at the St. Petersburg Public Library (Manuscript Division Fond 7, opis’ 1, delo 131b). Merck might here be referring mistakenly to that, especially plausible if it was from a source at the same place he got his Uwelenski words, presumably without realizing that it is of a somewhat different Yupik language. The only other explanation would be that the reference is indeed to Big Diomede language and that was then still Naukanski, at least in part; cf. in that case his statement that Naukanski “approached” the language of the Americans on the Bering Strait islands, as an ironic understatement.

Merck’s claim that “Uwelenski” was spoken all the way to Shelagski Cape, a distance of about 600 miles, is rather startling. In fact, there is no documentary or linguistic evidence of this. There is, however, good evidence that Uwelenski was spoken as far west as Kolyuchin Bay, about 160 miles beyond Uelen. The evidence for this is in the attested place-names of that stretch of coast. Of about 100 such place-names, most are of Chukchi origin, but there are about 20 clearly of Eskimo origin. These are an additional and largely problematical source for Uwelenski. Most of these names are best documented only indirectly, in recent sources only through Naukanski and Chukchi, of course, and in older sources only through Chukchi. Menovshchikov (1964: 8, 1972: 26), unaware of Merck’s statement or of Uwelenski Yupik, incorrectly assumes they must be Naukanski. Careful examination of these place-names is quite inconclusive in distinguishing between Uwelenski and Naukanski, given especially the close similarity of these languages in the first place, and detailed discussion is reserved for the writer’s linguistic paper on the Uwelenski corpus.

As luck would have it, however, two major names at the very western end of the stretch in question are quite clearly not Naukanski and possibly do not recur even in Chaplinski, but are well known on St. Lawrence Island itself. The name of Kolyuchin Bay is from Kulusiq, which on St. Lawrence Island is the important name of a type of early fall sea ice, rich in game. The name of the village near Kolyuchin Bay, Neshkan, has nothing to do with nasquq (‘head’) as is often surmised, but is from the Chukchi adaptation of an Uwelenski Nasqaq, with no clear meaning. That exact place-name does recur, however, on the north shore of St. Lawrence Island. Gambell scholar Christopher Koonooka (pers. comm. 1999) informs the writer that a tape recording of Nick Wongittilin (Uqengiliighaq, 1904-1999) reports that the southeast cape village of the island, Kiyalighaq, was at least in part settled by Uuleghmiit (Uelentsy). This could also explain the reports that the Kiyalighmiit spoke “differently.” To complete the full circle, Wongittilin further says that the Uuleghmiit-Kiyalighmiit also camped or lived at Nasqaq. Whether that was another original Nasqaq or was their (“New”) Nasqaq is hard to say. Duplicate place names in Chukotka and on St. Lawrence Island urgently need to be studied. The migration of the Uuleghmiit is probably 19th century. In any case, the extent, diversity, and coherence of CSY may herewith be appreciated a bit more clearly.

We lack information about the persistence and disappearance of Uwelenski. Perhaps there is still some folk memory of its former presence among the Chukchi speakers at Uelen today. It must have persisted well into the 19th century. Dolgikh’s map (1960) for the 17th century for some reason extends Uelenski no farther than Ryrkaipii, with only Chukchi and Chuvantsy on the rest of the coast to Shelagski Cape. Further, according to Dolgikh (1967: 13, 1972: 23) “the Uelentsy had been almost entirely assimilated by the end of the 19th century. Only in the settlement of Uelen itself did the Eskimo language survive for a certain time,” i.e. into the twentieth century. It would be interesting to learn the basis for Dolgikh’s statement. Aurel Krause (1881: 270) notated one word at Uelen, “no” as “awangitunga U,” i.e.avangitunga, definitively CSY for ‘I don’t have any,’ not Naukanski aviatua (< avaitua), but rather the last word we have of Uelenski.

The Wrangel Island-Point Hope Connection

There is considerable literature, since the 17th century, about traditional Chukchi knowledge or legends of Arctic islands and “Great Lands” or continents beyond—no doubt reflecting a combination of islands with larger lands beyond, but certainly by far the most prominent of these is the connection proposed here, a northern one connecting Chukotka via Wrangel Island to “Tikegen/Tekeghaq,” Point Hope, i.e. northern Alaska including Kotzebue and Barrow[2]. This is quite separate from the obvious more southern connection, via Bering Strait to “Keegen/Keegte,”[3] Cape Prince of Wales and beyond. Although this writer has examined a large part of this literature suggesting or stating this northern connection, time and space here permit only a summary of some of the highlights of it, particularly to determine what Eskimo language was spoken beyond Kolyuchin Bay, indeed Merck’s “fourth” such language.

Beyond Kolyuchin Bay we have no evidence that Merck’s Uwelenski was spoken. However, we do have statements that some non-Chukchi language was spoken in the stretch between Kolyuchin Bay and Shelagski Cape, and that it involved Alaskan Inupiat from the Point Hope area, via Wrangel Island. In other words, that stretch was visited or frequented by speakers of a fourth Eskimo language after all, and that instead was in fact North Alaskan Inupiaq.

Early maps, ca. 1711 and 1726

The earliest suggestion of this is from maps ca. 1711. L’vov’s map (Efimov 1964: map 55) in the Shelagski Cape sector has the note “zhivut shelagi ot Chukoch’ soboirod (‘[here] live Shelagi, from the Chukchi a kind unto themselves’).” Beiton’s map (ibid.: map 54), has a more detailed inscription: “[here] live people of the Shalatskoi kind, they have a language special to themselves, are very warlike, and fight with bows, and are impossible to pacify, because if you capture them, they kill themselves; they stab themselves with knives.”

Next, already in 1726, we have on a map by Afanasii Shestakov in the area north of Chaun Bay, an island populated by Shelagi under the power of a prince named Kopai,” and north of that, the edge of what could be a continent, a “Great Land, discovered in 1723 by the Shelagski prince. On this land are many inhabitants, plenty of sable, fox, beaver, marten, and trees” Wrangell (1948: 54-57). On a French version, the continent is simply “Bolchaia Zemlia ouGrande Terre” and the island is “Islehabitée par la nation nommée Chelagyi, indépendante, dont le prince nommé Kopoy a été fait prisonnier. Le chemin de la côte jusqu'à cette Isle et de là jusqu'à Bolchaia Zemlia est de quarante-huit heures” Berg (1946: 96-98). For more literature about Kopai and various versions of this map, see e.g., Belov (1956: 247); Efimov (1964: 47 and map 68); Müller (1761: xix-xxi); and Sergeev (1948: 414, 416). Conceivably, the Kopai island might be one of the Bear Islands north of the Kolyma, rather than Wrangel Island, and the Great Land beyond (even with trees?!) the New Siberian archipelago 400 miles to the northeast, rather than Alaska. Forty years later, however, from Daurkin on (see below), we have a strikingly similar report about the Shelagski chieftain Krakhai, in which the island is definitely Wrangel and the continent explicitly Alaska (Point Hope and beyond).

Daurkin’s map, 1765

The next and major source, in fact, is the map by the remarkable Chukchi Nikolai Daurkin from 1765. Daurkin was on the 1791 Billings expedition and Merck certainly had access to him. He might have been one of Merck’s sources for the broad language statement above, but presumably not for the extent of Uwelenski. The spectacular map by Daurkin (Efimov 1964: map 128) dates 26 years before that expedition and gives very different information (Figure 2). It shows the Chukchi Peninsula and much of Alaska from Seward Peninsula to past Point Hope, but the position of the Alaska coast is severely distorted westward, to parallel to the north the Arctic coast of Chukotka. The place-names, mostly villages, are indicated on the original map all only by number, 122 of them, and these appear accordingly in a table occupying the lower right quadrant of the map, some with long texts. Unfortunately there is no full publication of the information in this table, and the legibility is poor, especially in Efimov’s half-tone plates.

Between the Chaun area and the Alaskan point directly above/across from it is an island, only slightly closer to the Chukotka side, but unmistakably in the position of Wrangel Island and more exactly its shape than on any map before the late 19th century. On it is the number 116, which in the table reads “Ostrov Nymnym (‘Nymnym Island’).” Nymnym is the generic Chukchi term for “village/habitation.” On the original and some of the copies is a dotted line, from the west side of Chaun Bay to the island, not vaguely to the middle but specifically to the southwest point thereof (Cape Blossom). From there, the dotted line continues north to the Alaskan point, numbered 115, with no name in the table but a text, not in any published reading, but decipherable enough from a comparison of the various copies, “Olenii perekhod pol’du chrez Nymnym ostrov v Chukotskuiu zemlitsu (‘Reindeer crossing over the ice via Nymnym island to Chukotka’),” presumably meaning that people come over the ice from Alaska to Chukotka via Wrangel Island using domesticated reindeer.

To the southeast along the Alaska coast is a detailed set of points and inlet, unmistakably representing Kotzebue Sound and Hotham Inlet even, numbered 114. The text for that, most fully published by Medushevskaia (1954), here checked against the copies, with minor variations, reads:

The land Tikegen, which in strong weather is moved several versts out of place, but when there is no [such] weather it stands in its usual place. […] [one indecipherable word]. People live on that land, reindeer Chukchi call them Khrakhai. Their language is that of the Chukchi, but/and they have rather fine copper, spears, knives, and kettles of copper.

Medushevskaia 1954: 120

From this it would appear that the people involved in the crossing are Chukchi (only), not Eskimo, since they speak Chukchi and travel with reindeer. But this is contradicted by the sources both before (speaking not Chukchi and having hostile relations with them) and after. Tikegen is very clearly the Chukchi adaptation of Eskimo Tekeghaq, the Inupiaq name of Point Hope, well known even today in Chukotka. That the name does not appear at 115, Point Hope itself, but rather at 114, Kotzebue Sound, may simply reflect the use of the name for the northern Alaska connection area in general; cf. the use of Keegen, Chukchi adaptation of the Inupiaq name Ke(ng)egen/Ke(ng)ekte of Cape Prince of Wales for (southern) Alaska generally; and/or vagueness. The movement of the land is of course impressionistic, but may well reflect rather precisely the movement of a fixed trail over drifting ice. The Khrakhai people’s possession of copper implements may not only reflect that they somehow had trade with Russians (more than did other Chukchi?), but also explain motivation for such a trek from Alaska, to obtain such goods from the Russian post at Lower Kolyma, established 1644. In 1765, it seems far less likely that copper tradegoods could have come from the American side than from the Russian. Most importantly, the name Khrakhai (from now on, instead of Kopai) comes up again, dramatically in Wrangell (1948: 310-312).

Kotzebue 1816 , Shishmarev and Lazarev 1821, at St. Lawrence Bay

In 1816 a Chukchi elder at St. Lawrence Bay told Kotzebue how much he hated the Americans, and

“If I meet such a fellow with two bones [through his lip] I shall pierce him through.” When I asked then where the Americans received their iron, he answered from Colima. They spoke much upon the subject, and all our interpreter was able to make out was that the Americans came by water to the north, near Colima, but we were not able to make out whether they traded with the Tschukutskoi or the Russians […].

Kotzebue 1821: 262-263

This establishes that Inupiat came to or near Kolyma to trade, not only for copper, but also iron.

Figure 1

Distribution of Eskimo languages in Chukotka in 1791 according to Merck.

Distribution of Eskimo languages in Chukotka in 1791 according to Merck.
Source: Krauss (1985: 174) with additions to base map. 1) Sirenikski, 2) Central Siberian Yupik, “4” = 2) Uwelen, 3) Naukanski, 4) North Alaskan Inupiaq.

-> Voir la liste des figures

Figure 2

Nikolai Daurkin’s 1765 map of Chukotka and part of Alaska showing both Bering Strait and northern (Wrangel Is., Pt. Hope) connections.

Nikolai Daurkin’s 1765 map of Chukotka and part of Alaska showing both Bering Strait and northern (Wrangel Is., Pt. Hope) connections.
Source: portion of map 128 in Efimov (1964). Additional placenames: A) St. Lawrence Island, B) East Cape, C) Kolyuchin, D) Wrangel Island, E) Point Hope, F) Kotzebue area, G) Cape Prince of Wales.

-> Voir la liste des figures

The next major source is the around-the-world voyage of the Blagonamerennyi, at St. Lawrence Bay in 1821. From that, we have two published accounts, nearly identical, of information from their Chukchi informant there, the elder Paigdau, that is quoted in Shishmarev (1852) and in Lazarev (1950), so close that they may be translated here combined:

[These] Chukchi travel along the coast of Asia from St. Lawrence Bay to the North and to the South. The old man Paigdau told of himself that he travelled beyond Kolyuchin Island, to the river Amylik, situated according to him between latitude 69° and 70° not far from Shelagski Cape, which he calls Chavak, and saw no ice. The people living around this cape have a slit in their lip, into which they insert a bone or bead finished in bone in the style of the inhabitants of the opposite American coast. These Chukchi visit the Amylik River and trade with the Chukchi who come there. This river is not large, and consists of melting snow.

Shishmarev 1852: 191-192, 199; Lazarev 1950: 308, 325

Further, another Chukchi elder, Lei-gai-gu, tells of a tribe living further to the north, “along the coast of the Frozen Sea, who also slit their lip and insert a bone like a button or bead, American style, and call themselves by the name Egliunak” (“Egliunok” in Lazarev) (ibid.). For Shishmarev, Lei-gai-gu adds explicitly: “This people too is distinguished by language” (ibid.).

Amylik and Egliunak/Egliunok both appear to be spellings of perfectly plausible Eskimo forms, distinctively so (rather than Chukchi, even though coming from Chukchi speakers). Amylik could be amilek ‘having skins,’ in Inupiaq (equally in Yupik). Egliunak/Egliunok could be Inupiaq egluneq ‘(making) shelter for, housing’ (palatalization of /l/ is expected from Russian hearing of even non-palatalized /l/ of Eskimo); so identified it would be decisively Inupiaq and not Yupik. In the less likely case that it is to be read as a derivative of inglu- ‘the other (of a pair), partner’ it could be either Inupiaq or Yupik, but the initial vowel favors the Inupiaq ‘housing’ identification, even though the term is presented as an ethnonym. In any case, with both sources specifying they wore labrets, there is strong reason to conclude that Merck’s fourth Eskimo language in this area was North Alaskan Inupiaq, and not Uwelenski.

This accords furthermore with the total absence on any maps or place-name or village-name lists of any names that appear to be of Eskimo origin beyond Kolyuchin, unlike the statistics of 20% up to that area. Only the above names in the 1821 accounts from the Blagonamerennyi officers appear to be of Eskimo origin rather than Chukchi for the 440-mile stretch in question.

Wrangell, 1823

The next source is Wrangell in 1823. The Chukchi near Shelagski Cape tell him of a people he recognizes as Eskimos:

Part of the shore of Schalaurov Island is coverered with heaps of the bones of whales; these are probably the remains of the dwellings of a people who lived on fish, but chiefly on whales, the bones of which were employed as timber in building their huts. We were told that their language was very different from that of the wandering reindeer Tschuktschi; and resembled that of the people who live near Behring’s Straits in mud huts, supported on the inside by whalebones, and having their only entrance from above. They are the same race as the Aleutians and the Greenlanders, which have peopled the coast from the eastern part of North America, along the Polar Sea to Cape Shelagskoj.

Wrangell 1839: 188-189, 1841: 254

Further, Wrangell ties in Daurkin’s Khrakhai in some detail to Wrangel Island:

There are traditions which relate that two centuries ago the Onkilon [from Chukchi angqal’yn ‘coast dweller’ i.e. Eskimo] occupied the whole of the coast from Cape Schelagskoi to Behring Straits; and it is true that there are every where along this tract the remains of huts constructed of earth and whalebones, and quite different from the present dwellings of the Tschuktschi. A disagreement between Krächoi, the principal chief of the Asiatic Esquimaux, and an errim, or head of a tribe of rein-deer Tschuktschi, broke out into decided hostilities; Krächoi was defeated and forced to flee, his people migrated, and the coast was deserted. The inhabitants of Ir-Kaipij relate that Krächoi, having killed a Tschuktschi Errim, was closely pursued by the son, and after some time retreated to the rock of Cape North, where he intrenched himself behind a kind of natural rampart which still exists. The young Errim followed him thither, and succeeded in killing Krächoi’s son, by which, according to the ideas of these people, the debt of blood was paid. Krächoi let himself down from the cliff during the night by means of thongs, and got into a boat which was waiting for him at the foot of the rock. He at first steered toward the East to mislead his pursuers, but the following night he turned westward, and reached Schalaurow Island, where he built the earthen huts of which we had seen the remains. He was gradually joined by his kinsmen, and others of his own tribe; and they all finally fled together in fifteen baidars to the country, of which the mountains are said to be sometimes visible from Cape Jakan. In the course of the winter, a Tschuktschi who was allied to Kräkhoi, disappeared with his family and his rein-deer; and it was supposed that he too had gone to the northern island.

Wrangell 1839: 220-221, 1841: 239-240

According to Daurkin’s 1765 map, presumably the Khrakhay people had a village (nymnym) on Wrangel Island, and must have crossed from there to the Point Hope area of Alaska. Shishmarev (1852) and Lazarev (1950) mentioned the people had their own language and wore labrets; therefore they were Alaskans in the first place. Wrangell agrees they are Eskimos like those in Alaska, and explains their flight back thither, at least as far as Wrangel Island.

Nordenskiöld on the Vega, 1878-1879

The next major source is Nordenskiöld’s Vega Expedition in 1878-1879. Nordenskiöld is quite aware of several of the earlier sources on these “Shelagi” or “Onkilony,” cites Wrangell’s Krächoj story in full, but makes a contribution of a new kind, in archeology. In Nordenskiöld’s (1880: 427-432, 1883 II: 181-186) main narrative several officers are sent in September 1878 to investigate Onkilon ruins at Ryrkaipii, thought to include Krächoj’s last stronghold. The text includes vertical and cross-cut drawings of a house-pit, and reproductions of eight artifacts. Hovgaard’s (1881: 125-131, 1882: 129-134) narrative includes figures of six artifacts and a very evocative view of the site. Finally, Nordqvist (1915) much later published a discussion of the excavations at Ryrkaipii and Idlidlja with four drawings of house-pits at Idlidlja. Much later still Nordqvist’s journal appears (Belov 1964: 199-201), including report of excavations, but none of the drawings. No doubt far more information and materials, including photos, could be found in the Vega Expedition holdings in Sweden.

Clearly by 1878 these people had been gone for some time, and probably since about Wrangell’s time there was no need for Alaskans to travel nearly so far for trade goods.

Twentieth-century literature

Twentieth-century literature has very little to add on the question of what Eskimo languages were spoken on the Arctic coast of Chukotka. It does add, however, tantalizing information on the archeology of Wrangel Island.

Bogoras (1904: 20-23) minimizes, as usual, Asiatic Eskimo language differences to dialect level. Aware of Krächoj legends and the like, as well as the Eskimo origin of Kulusiq and Nasqaq (ibid.: 22) he is however rather skeptical about much Eskimo presence on the Arctic coast of Chukotka, and claims the house pits could be Chukchi just as well as Eskimo. He is hence far from specifying anything about which Eskimo languages or “dialects” might have been spoken along that coast. His writings subsequently continue to reflect these views, unchanged. Bogoras (1904: 22-23) does note, however, a legend more recent than Krächoj, probably 19th century, that two kayakers were windblown from Kulusiq to Wrangel Island, and found houses and provisions there. “All these details were given so vividly, that if I had not known that Wrangel’s Land is destitute of inhabitants, I should have had no hesitation in crediting the narrative” (ibid.).

Archaeological evidence on Wrangel Island

In the spring of 1937, the Eskimo hunter Ainafak, one of the Chaplinski Yupiget then stationed on Wrangel Island, was out on his trapline, near Cape Blossom, the southwest tip of the island, at precisely the point to which Daurkin’s dotted line leads. There on the southwest slope of Mt. Thomas (Cape Foma), Ainafak noticed pieces of a wooden spear and harpoon shaft, and a driftwood shovel, at the remains of an underground house with rotting driftwood beams or supports. Another Yupik, I.S. Taian, shortly verified this, and removed those artifacts plus a walrus tusk spearpoint or percussion tool. In fall 1938, he gave these to R.F. Gekker of the Academy of Sciences. In summer 1937, the author of the only report on this site, Soviet geologist Leonid Vasil’evich Gromov, together with the Chukchi S.Z. Popov, visited the site, which evidently they found without difficulty. With only the most superficial excavation, they found nine more artifacts, pictured in the report (Gromov 1941, often repeated in later writings, without new information). These included one blue Russian trade bead, dating the site to the proto-historic period. Gromov perhaps restricted himself to barely scratching the surface in the hopes that professional archeologists would investigate further.

There have been no results since 1937. In 1975 or 1976, Tasian S. Tein, a Naukan Yupik and graduate student in archeology under N.N. Dikov at Magadan, participated in an expedition which excavated first at Krekhai’s stronghold at Ryrkaipii, which Dikov had earlier excavated in 1957. The many artifacts found there are presumably now at Magadan. The Tein group then went to Wrangel Island, primarily to investigate the much more ancient site (3000 BP) discovered by Dikov at Devil’s Gorge there. Tein’s (1983: 3-6, 22-25, 40) popular account is the only published report. It shows he was still hearing Krekhai legends at Ryrkaipii. The report includes a very detailed account of a side-trip made to the Gromov 1937 site. To the acute embarrassment of their veteran Yupik guide Napaun, who knew the area well and had seen the site before, they were unable to find any trace of it, and after four hours gave up the search. Tein (1983: 93) closes with mention that a further field-trip to find Gromov’s 1937 site was envisioned, but nothing has come of it, and Tein’s health meanwhile has failed.

Very shortly after the resultless Tein expedition, in 1977, journalist Shentalinskii met Dikov and Tein, and also interviewed Gromov, still alive in Moscow, and Ainafak, still living on Wrangel Island. (One may therewith wonder that Tein had not done as much, except perhaps that his expedition’s main interest was in the much older Devil’s Gorge site.) Gromov did not know what had become of the artifacts he collected in 1937. Ainafak seemed to remember well and could probably have found the site again. According to Shentalinskii’s (1978: 46, 1988: 60) account, a bit vague and puzzling, he somehow easily found the site, but did not disturb it, explicitly leaving any excavation to “specialists.” Conferring further then with Gromov and the Magadan archeologists, and considering that the sites seen were evidently in different locations (near to each other), he came to the conclusion that there must have been several house-pits or even a whole village. Burykin (2000) quotes two more non-archaeologist Soviet writers who also clearly saw the Cape Blossom site: a geologist, Kuvaev, on a geophysical expedition to Wrangel Island in the early 1960s, and an aviator, Akkuratov, circa 1986.

University of St. Petersburg archeologists visited Wrangel Island during the 1990s, again primarily to investigate the ancient Devil’s Gorge site, and there is no report of any serious attempt to find the 1937 Gromov site. We are thus left with the very peculiar and ironic result that the site was found by Eskimos and a geologist in 1937, later by another geologist, a journalist, and an aviator, but never by an archeologist.

Conclusion

Merck’s statement about four “Sedentary Chukchi” (Eskimo) languages or language varieties along the coast of Chukotka in 1791 was interpreted in this article. The first three languages were identified, as 1) Sirenikski, 2) Central Siberian Yupik, explicitly including St. Lawrence Island, and 3) Naukanski. Merck’s language number four, “Uwelenski” he claims, was spoken along the Arctic Coast of Chukotka from Uelen as far as Shelagski Cape, 600 miles to the northwest. Studies of 70 or so “Uwelenski” words collected by Merck show that “Uwelenski” is in fact a dialect of Central Siberian Yupik, thus part of a language continuum spoken from St. Lawrence Island to the Chaplino corner and the East coast of Chukotka, thence to the North coast of that mainland, treating Naukan as a “third Diomede” rather than as a mainland interruption. However there is no evidence that language number four, “Uwelenski,” actually a dialect of Merck’s language number two, was spoken beyond Kolyuchin Bay. Beyond that point, however, there was indeed a fourth Eskimo language, in fact none other than North Alaskan Inupiaq, spoken intermittently in pockets between Kolyuchin and Shelagski Cape, at least since the opening of Russian posts at Kolyma and into the 19th century, by north Alaskans from the Point Hope area, who also used Wrangel Island as a stopping place.

The nature of this North Alaska connection (Wrangel Island and Point Hope), as opposed to the Bering Strait Alaska connection, deserves more study. It now remains not for this linguist, but for anthropologists, to follow up on the question of the North Alaskan (Wrangel Island-Point Hope) connection. Archaeologists are needed to examine the extant materials at Ryrkaipii and other Arctic Chukotka sites in the Nordenskiöld holdings in Sweden, and at Magadan, to determine to what degree they may be identified as northern Alaskan; likewise anything from the 1937 Gromov site, if locatable. Further excavations at the arctic Chukotka sites could certainly supplement this and even if an Alaskan origin of some of the material confirms the connection, it would still certainly be worthwhile to conduct a definitive investigation of the Wrangel Island site. Further confirmation of the connection may still be available moreover in living memory from both sides, and the Alaskan side has not been investigated at all in this respect. Especially at Point Hope, and even at Barrow, there may be reports of this connection.