OLZ 117-1 (2022), Zentralasien 81 Kim, Ronald I.: The Dual in Tocharian. From Typology to Auslautgesetz. Dettelbach: H.J. Röll 2018. XIV, 186 S. 8° = Münchener Studien zur Sprachwissenschaft, Beiheft 26. Hardbd. € 49,00. ISBN 978-3-89754-528-1. Besprochen von Michaël Peyrot: Leiden / Niederlande, E-Mail: M.Peyrot@hum.leidenuniv.nl https://doi.org/10.1515/olzg-2022-0030 In this short monograph, Ronald Kim investigates the nominal dual of Tocharian, a branch of Indo-European consisting of the two languages Tocharian A and Tocha- rian B, once spoken in the northern Tarim Basin in pres- ent-day Xīnjiāng in Northwest China, and extinct since the end of the first millennium CE. In a solid and insightful way, he discusses the syn- chronic status and morphology of the nominal dual, then proposes a Proto-Tocharian (PT) reconstruction, and finally explains the different endings from Proto-In- do-European (PIE). Well embedded in the literature, the work has been carried out carefully and systematically, and offers convincing analyses of the relevant forms, both linguistically and philologically. The exact derivation of a number of endings from Proto-Indo-European has always been difficult. Kim proposes several new sound laws for final syllables, so-called Auslautgesetze, to account for the shape of the Tocharian endings. My expectation is that these will not be accepted widely, but the argument is pre- sented in such a clear way that the reader can easily judge for her- or himself. The book is an important continuation of the research line on the Tocharian dual, building on the works of Winter,1 Hilmarsson,2 and others. In Chapter 1 (p. 4–28), Kim shows that in Tocharian B the dual is in principle used together with wi ‘2’, antapi ‘both’ or the dual demonstrative tai. Only in the case of paired body parts or other naturally paired items is the dual used on its own. The few exceptions he has col- lected are from contexts where the dual number is already known. He further notes a decline in the use of the dual: in archaic texts, wi ‘2’ is almost exclusively used with the dual, while it is only used with the plural in late texts (clas- sical texts, as so often, are mixed). Also, dual nouns are found to be modified by plural adjectives as well as dual adjectives in classical texts, but in archaic texts only dual adjectives are attested. 1 Werner Winter, “Nominal and pronominal dual in Tocharian”, Language 38 (1962): 111–134. 2 Jörundur Hilmarsson, The dual forms of nouns and pronouns in To- charian, Reykjavík 1989. Finally, in agreement with observations by Svetlana Burlak and Ilya Itkin, he notes that there is a case to make that the dual is no longer a morphological category of the noun in Tocharian A: there are several forms that are clearly cognate with Tocharian B duals, but these do not have plurals beside, so that there is no dual : plural con- trast. The very few exceptions can be explained as due to semantic specialisation, e. g. in the case of “plural” esañ ‘skandhas’ next to “dual” esäṃ ‘shoulders’ or fossilisa- tion, e. g. “dual” peṃ ‘feet’, which occurs only in the fixed expression peṃ winās- ‘venerate the feet’. Chapter 2 (p. 29–37) contains a brief overview of Indo- European nominal dual endings and their reconstruction. Kim’s reconstructions are in line with those that are com- monly accepted. In a systematic way, Kim discusses synchronic pat- terns of dual formation in Tocharian A and B in chapter 3 (p. 38–63), before comparing and reconstructing these for Proto-Tocharian in chapter 4 (p. 64–69). Chapter 5 (p. 70–97) discusses the derivation of the reconstructed Proto-Tocharian endings from Proto-Indo-European. On the basis of morphological patterns of the dual, Kim dis- tinguishes 6 groups. For group 1, consisting of nouns in B -ine, A -eṃ, like B du. lymine, A lymeṃ ‘lips’ and adjectives in B -i, like du. kartsi ‘good’, he reconstructs *-ẹ(në), ultimately from the thematic neuter ending *-oih 1 . Although the correspond- ence between B -i and A -e, which causes no preceding palatalisation, in my view more precisely goes back to *-ey(ne),3 I think that this reconstruction is in essence correct. Indeed, the adjectival dual ending is found in thematic adjectives,4 and since there is no gender con- trast in dual adjectives, we may assume that the neuter ending was generalised. However, I am not convinced that the type B lymine, A lymeṃ ‘lips’ can be explained in the same way. On the one hand, these nouns are not thematic neuters. On the other hand, these duals show no distinc- tive dual element in the stem, as the stems B lymi- and A lyme- are found throughout the singular and the plural. The only mark of the dual is the suffix B -ne, A -ṃ, which could simply have been added to this basic stem. 3 See Michaël Peyrot, “The Tocharian A match of the Tocharian B obl.sg. -ai”, Tocharian and Indo-European Studies 13 (2012): 181–220, p. 196; and in much more detail, Alessandro Del Tomba, “The devel- opment of PIE *-oi̯ in Tocharian”, Rivista degli studi orientali, nuova serie 93(1/2) (2020): 21–34. 4 The adjective kartse itself is largely athematic, but precisely the nom.sg.m. kartse, to which the dual is obviously parallel, is the- matic.