Document of the Month 1/26: A Commercial Letter from Mount Mugh
From Mount to Manuscript: Revisiting Commodity Terms in a Mugh Document
by Alisher Begmatov
The discovery of the Mount Mugh documents
In the spring of 1932, a shepherd fortuitously discovered a wicker basket containing a manuscript at Mount Mugh, Tajikistan. This paper manuscript (Figure 1), later labeled 1.I,1 written in a script then unknown in the region and preserved today only in a photograph, paved the way to one of the most significant archives of early medieval Central Asia, comprising approximately 75 documents in Sogdian, alongside one in Arabic, one in Old Turkic, and three in Chinese. The Sogdian documents from Mount Mugh provide a unique local perspective on the political, social, and economic life of Sogdiana during its arguably most turbulent period, the Arab-Muslim conquest.
Fig. 1: Document 1.I. Reproduced from Livshits 1962:113, fig. 3.
Soon after, the shepherd, Jurali Mahmadali handed over the manuscript (1.I) to another local and, following a few months hand-to-hand trip, it was finally delivered to Stalinabad (present-day Dushanbe) by Abdulhamid Puloti, then secretary of the Zahmatabad (now Ayni) district committee. There, it appears to have been presumed that it was written either in Sogdian or in Uyghur.2 In April 1933, Aleksey Vasil’ev, then scientific secretary of the Tajik branch of the USSR Academy of Sciences, brought photographs of the manuscript to Aleksandr Freiman—an Iranian philologist and a corresponding member of the USSR Academy of Sciences based in Leningrad—for further examination. Freiman quickly recognized it as a Sogdian letter addressed to a nobleman. Determined to study the manuscript in its original form, he set out for Stalinabad, where he was able to identify some place names and establish that the addressee was closely linked to Samarkand.3 The letter included the titles ‘Ruler of Samarkand’ (smʾrknδc MRʾY) and ‘Ikhshid (king) of Sogdiana’ (sγwδyk MLKʾ), which Dewashtich, the sovereign of Panjikent, adopted in his later years after declaring himself the king of Sogdiana. From this, Freiman inferred that the document predated the full consolidation of Muslim caliphal control, placing it before the mid-8th century.4 The only Arabic Mount Mugh document, found a few months later and the first to be fully interpreted,5 indicated that the collection likely belonged to the archive of Dewashtich (Dīwāstī in Arabic or Δēwāštīč in Sogdian), further reinforcing Freiman’s assumption.
It seems undeniable that Freiman was the person who most clearly grasped the value of the manuscript (1.I) at the time. This recognition likely motivated him to seek further discoveries at the same site, as is reflected in his agreement with Vasil’ev to investigate Mount Mugh at the earliest opportunity.6 Meanwhile, in August 1933, Puloti, working independently as an amateur, discovered around twenty more documents. Later, a similar number were reportedly found by Vasil’ev. These discoveries spurred Freiman to accelerate his efforts and to organize an official expedition later that same year, in 1933. Undaunted by cold weather, wind or rain, and driven by the hope of unearthing further documents, the expedition continued until the first snow. Although Mount Mugh is considerably lower than the surrounding mountains, its steep slopes make ascent difficult and systematic excavation even more challenging (Figure 2).
Fig. 2: Mount Mugh. Left: view from the west;
right: view from the north. Arrows indicate the location of Dewashtich’s castle. Photograph by A. Begmatov.
Within just a few weeks, Dewashtich’s castle—the southern elevated rectangular part of a larger (approximately 25x60m) rectangular fortified enclosure—was fully excavated, yielding additional documents along with approximately 400 artifacts, most notably textiles.7 It seems reasonable to assume that some of these textiles could correspond to items mentioned in the documents found alongside them, especially those with an economic focus. Although a brief season of small-scale excavations was undertaken by Y. Yakubov in the 1980s, I believe that further surveys and excavations remain a desideratum for better understanding the site and its wider context. That said, in the subsequent part, I will turn to philological aspects of these documents, with particular attention to manuscripts with economic content containing textile and leather products. Additionally, I will present a recent re-edition of the Mount Mugh Document A-1, which was previously among the least understood in earlier editions.
The philological study of the Mount Mugh documents
Following the pioneering study of several documents by Freiman, the bulk of the collection—primarily legal documents and letters—was edited by Livshits, while those with economic content were studied by Bogolyubov and Smirnova.8 In addition, several philological and historical studies contributed to our understanding of the archive. Traditionally, most research concentrated on the first group of documents—legal contracts and letters—because of their significant political, social, and historical importance. These sources also include unique materials such as a marriage contract, and they offer a greater volume and diversity of texts, making them valuable for examining linguistic features and philological analysis. Nevertheless, numerous unresolved issues persisted within the archive, particularly among the economic documents, to which much less attention had been given: certain commodities remained unidentified, and even those previously classified may correspond to different items in various contexts.
Recent research on economic documents from Mount Mugh has shown that more textile and leather products are recorded in them than previously understood.9 These commodities appear to have been widely used by merchants and others across Eurasia. While the process of looking for unclear commodity terms or other unfamiliar words in different languages can sometimes feel like hunting for rare fish in the ocean—often fruitless, but occasionally rewarding—I believe that, in the case of the document considered below, we are fortunate to be dealing with the latter.
Newly identified textile and leather commodities in the commercial letter A-1
Document A-1 stands out among economic records for its unique product listings, and reinterpreting these terms provides novel insights. It is a commercial letter in Sogdian written on parchment, housed in the Institute of Oriental Manuscripts, Russian Academy of Sciences, measuring 11 x 12–13 cm (Figure 3). It is one of many documents in the Mugh collection addressed to framānδār Ot,10 a functionary at Dewashtich’s court serving as a steward of provisions. The sender of the letter is Āpānak who is also attested in the Mount Mugh documents A-5 and Nov.1. Fifteen lines on the recto are well preserved, whereas twelve lines on the verso are barely legible. The letter bears a list of commodity names that are being sent by Āpānak.
Fig. 3: Document A-1 recto. Reproduced from Begmatov 2019, fig. 2.
Notable examples from Document A-1 include unique commodity terms such as ʾpšmʾkntyh /ǝpšmāgandī/ (‘wool-stuffed’ textile item), rmʾnykh /(ǝ)rmānīk(a)/ (a type of felt or woolen blanket), ptsγtykh /patsaγtīk(a)/ (‘rug?’ branded as patsaγtīk), and δkknh /δakkin(a)/ or /δakkan(a)?/ (‘muslin, gauze or tulle headwear’). These items were interpreted as precious stone terms in earlier editions; however, they now appear to pertain to textiles or textile-associated goods. Such items played an important role in mercantile activities along the Silk Roads, and some varieties may have remained in use until recently. It is also possible that these commodities continue to be utilized in certain Central Asian villages or mountainous areas even today. In the following, I aim to explain these words in a simplified manner, comparing them with commodity terms from other languages that had contact with Sogdian or retained terms as a result of cultural and trade exchanges. I will also present a revised edition and translation of Document A-1. For more in-depth philological comparisons and explanations, see Begmatov 2019, 2020.
Among the most notable re-interpreted terms is ǝpšmāgandī (A-1 recto, line 5), which is also the clearest as its first part, ǝpšm, matches the Iranian word pašm ‘wool’, still in use today. Given Sogdian writing conventions (involving irregular representation of prosthetic aleph), it aligns well with the meaning ‘wool’. Middle Persian pašmāgand ‘wool-stuffed (silk saddle-cloth)’ was likely borrowed into Sogdian,11 as well as Khwarazmian as pšmʾknd ‘filled with wool’. This evidently wool-related item not only casts considerable doubt on an earlier identification of it as ‘malachite’ but may also offer a rough indication of the meanings of subsequent terms in the document.
The interpretation of the items listed in Document A-1 as precious stones is largely due to Bogolyubov,12 who explained, for example, ǝpšmāgandī by comparing it with Georgian p’aršamang-i, p’aršavang-i ‘peacock’, a borrowing from Western Iranian, and ultimately from Old Iranian frašmaka-. Bogolyubov further supported his hypothesis by drawing parallels with Arabic and Persian ṭāwusī ‘pertaining to a peacock > malachite’ and Chinese kongque shi 孔雀石 ‘peacock stone; malachite’. Although this interpretation may seem well-argued, in light of the explanation given in the previous paragraph, it appears rather far-fetched.
Another of Bogolyubov’s interpretations—that of (ǝ)rmānīk(a) (A-1 recto, line 9) as ‘ruby’, a loanword from Arabic rummānī ‘ruby’, itself from rummān ‘pomegranate’—also appears problematic, especially when compared with the more relevant Tibetan term 'a-rmo-ni-ka (and its graphic variants 'ar-mo-nig, 'ar-mo-li-ga, etc.), meaning ‘(woolen or felt) long carpet, blanket’, translating the Sanskrit pāṇḍu-kambala ‘white woolen covering or blanket’. I assume that this word was borrowed from Sogdian into Tibetan and may represent one of the commodities the Sogdians introduced there. Although textual sources attest to the Sogdians’ active mercantile presence in this region, no specific commodity had previously been linked to them with certainty. This identification thus provides an early tangible testament to their everyday trade activities. More recently, its further borrowing into Tocharian has become apparent.13 This suggests that it was a widely circulated textile-related item across Central Eurasia.
With the likely identification of ǝpšmāgandī and (ǝ)rmānīk(a) as terms related to wool or textiles, it becomes possible to hypothesize the meanings of the following unfamiliar words ʾβšʾwβnkh/ʾβšwβnkh (A-1 recto, lines 5–6) and ptsγtykh /patsaγtīk(a)/ (lines 9–10). As ǝpšmāgandī denotes ‘wool-stuffed’ and seems to specifically modify a silk saddle-cloth in Middle Persian, it is plausible that ʾβšʾwβnkh (possibly pronounced avšǝwvǝng) refers to a similar object, such as a cushion or blanket—potentially made from silk and possibly used as a saddle blanket. While patsaγtīk(a) following (ǝ)rmānīk(a) is more challenging to interpret, it likely pertains to a carpet or blanket crafted from a particular variety of soft felt. Although originally I postulated the Ossetian term fædtsæg’dæn ‘a bow-like wool beater stick’ as a comparandum, more recent research points to the Middle Chinese term 鉢息徳 (Boxide), whose pronunciation closely resembles patsaγtīk.14 Boxide is mentioned as the central hub of the Maymurgh principality, a Sogdian city-state, which according to B. I. Marshak’s careful assessment, might have been dominant over Panjikent around the 7th century CE.15 Our ongoing archaeological research at Kuldor-tepa in Urgut (Uzbekistan), located approximately 25km south-east of Samarkand and 40km west of Panjikent, indicates that this site may be the remains of Boxide (patsaγtīk). It seems plausible that patsaγtīk was produced locally, with its name likely originating from its place of manufacture. This item was crafted from a specific type of woolen textile or felt, commonly referred to in Sogdiana at the time as (ǝ)rmānīk(a).
While several other similar terms remain interpretable only depending on their context, mainly on the basis of preceding or following words, an exciting novel interpretation can be offered for δkknh /δakkin(a)/ or /δakkan(a)?/ (A-1 recto, line 8). This word appears very close to dakana (a turban-like headdress of white gauze) in the Uzbek and Uyghur languages, which refers to a headdress made of doka or daka ‘white gauze or muslin’. It is plausible that the Sogdian term also denoted muslin or gauze/tulle (headwear). Three white tulle headdresses made of cotton have in fact been excavated from Mount Mugh (Figure 4). If the Sogdian word δkknh is a loanword that survived in Uzbek and Uyghur dakana, then we may well imagine that this headdress could refer to the type of tulle headdresses unearthed from Mount Mugh.
Fig. 4: Headdresses from Mount Mugh. Reproduced from Bentovich 1956:66.
Another noteworthy term, tβtch or tβtʾk (A-1 recto, lines 7, 11), presents an intriguing ambiguity as it may either mean ‘sealed’ or refer to ‘taffeta’, a renowned silk fabric that became known as far as Europe and may still be produced today. While both interpretations are viable, for the purposes of this document I have selected ‘sealed’ for the translation, as it seems more certain. It is also important to note that this cloth—described as ‘sealed ʾskwrnkh (precious textile/brocade)’ or ‘taffeta’—is associated with Afšun, the sovereign of Khākhsarch (A-1 recto, line 11), which presumably indicates that it was valuable. Its high value could be indirectly supported by the Mount Mugh document B(V)-16, line 12, where a certain Rēwak sends the same Afšun a piece of green silk cloth or damask, prʾynk /pǝrīng/. Pǝrīng is a well-attested Sogdian textile term, and it was widely borrowed into other Central Eurasian languages, as well as into Semitic languages via Middle Persian.16
Concluding remarks
The commodity terms from Document A-1 discussed here demonstrate the value of the Mugh documents with economic content for tracing the circulation of goods and terminology along the Silk Roads. Recent studies suggest connections not only to Old Uyghur and Chinese, as previously recognized, but also to Tibetan and Tocharian, as well as to closely related languages such as Khwarazmian, Bactrian, and Middle Persian. Some terms may even have persisted into modern Central Asian and Caucasian languages, including Tajik, Uyghur, Uzbek, and Ossetian. Taken together, these terms may offer a more precise picture of networks of exchange and linguistic contact, extending well beyond Sogdiana in both time and space, and they reinforce the need for further systematic study of this corpus, as more such commodities clearly merit closer and more detailed investigation.
Notes
1 Livshits 1962:108–14, 2008:117–26, 2015:88–93; Yakubovich 2002.
2 Freiman 1934:7.
3 This letter is widely known today, particularly owing to V. A. Livshit’s later interpretations of it.
4 Freiman 1934:8.
5 Krachkovskaya & Krachkovskiy 1934:56; see Begmatov 2020:3–4 for an English translation.
6 Freiman 1934:9.
7 Vasil’ev 1934.
8 Freiman 1962 (SDGM I); Livshits 1962 (SDGM II), 2008 (SÉSAS); Bogolyubov and Smirnova 1963 (SDGM III).
9 Begmatov 2019, 2020.
10 See Lurje 2010:118–19.
11 MacKenzie 1976:60.
12 Bogolyubov 1981.
13 Dragoni 2023:40.
14 Begmatov 2021.
15 Marshak 2002.
16 Henning 1945.
