Proto-Indo-European compounds in relation to other Proto-Indo-European syntactic patterns

Abstract Compounds in a language have long been viewed as reflections of other syntactic patterns in that language. Some of the suggestions on relationships may now strike us as clumsy, such as those deriving compounds like Greek from imperatives plus objects. 1 The syntactic patterns compared with compounds vary. Early views may be found in the Grundriss der Vergleichenden Grammatik der Indogermanischen Sprachen by Karl Brugmann and Berthold Delbrück, particularly 2II.l (Strassburg, 1906), pp. 35-120. An attempt to refer to the studies cited there and to the others produced subsequently would overshadow any new presentation, and is therefore not attempted here. Yet many views often subsequently repeated, like the identification of µενε- etc. as imperatives, are presented in the Grundriss, e.g. V. 174 (Strassburg, 1900).-Besides the standard handbooks for PIE, the large grammars for the individual dialects are important for the study of PIE compounds. For Indic, Jakob Wackernagel, Altindische Grammatik II.1 (Göttingen, 1957 Second unchanged edition) sums up the data, as well as their interpretation to 1957 through Debrunner's supplement. For Greek, Eduard Schwyzer, Griechische Grammatik 1 (München, 1939), pp. 425-455 fills a similar role, as do for Latin Manu Leumann, Lateinische Grammatik 1 (München, 1963 Reprint of the 1926–28 edition), pp. 247-254, and for Germanic W. Wilmanns, Deutsche Grammatik II (Strassburg, 21899), especially pp. 513-550. - Since any study of PIE compounds must rely heavily on the oldest Vedic texts, A. A. Macdonnell, Vedic Grammar (Strassburg, 1910) is indispensable, as is the interpretation of texts provided by Karl F. Geldner's translation, Der Rigveda 1–4 (Cambridge, Mass., 1951–7) and Hermann Grassmann's Wörterbuch zum Rigveda (Wiesbaden, 1955 Third unchanged edition). -A recent survey of Indo-European compounds may be found in Peter H. Salus, “The Types of Nominal Compound of Indo-European”, Orbis 14 (1965), pp. 38-62. This survey, useful also for bibliography, lists “all the types of nominal composition found in the various Indo-European dialects” (p. 60), with no attempt at interpretation, synchronic or diachronic. -My study was presented first as the Collitz lecture at the Linguistic Institute, University of Illinois, 1968; I am grateful especially to Emmon Bach and Eric Hamp for their comments. Other suggestions have been less explicit than we should like, for example Leumann's view of compounds as permutations of longer syntactic sequences; nass-äugig was for him a derived form of a nominal sentence: Auge nass. 2 Ernst Leumann, “Einiges über Komposita,” IF 8 (1898), pp. 297-301, also gives other derivations, or in his terminology permutations, of such sentences. And Jacobi, 3 Hermann Jacobi, Comþositum und Nebensatz (Bonn, 1897), p. 20. using Brugmann's definition of compound, proposed that Proto-Indo-European [PIE] verbal tatpurushas were “syntactic word complexes of a certain kind, so common in the proto-language that they could be combined into words”. But the procedures by which the complexes were combined were not specified. Recent statements on compounds have been more explicit. In his monograph on English nominalizations Lees stated of nominals that they are “noun-like versions of sentences”. 4 R. B. Lees, The Grammar of English Nominalizations (Bloomington, 1960), p. 54. In accordance with this position, at least some types of compounds in a language reflect its sentence patterns. If such a position can be proposed for the analysis of contemporary languages, it is also valid for reconstructed languages. In this essay I examine the PIE compounds in relation to the larger syntactic patterns of the language, with the aim of clarifying both patterns: compounds on the one hand but primarily its sentence patterns.