Published in B. Joseph & R. Janda (eds.) Handbook of Historical Linguistics. Oxford: Blackwell (2003), pp. 3-180. 1 ON LANGUAGE, CHANGE, AND LANGUAGE CHANGE — OR, OF HISTORY, LINGUISTICS, AND HISTORICAL LINGUISTICS Richard D. Janda & Brian D. Joseph 0. Preamble Fellow-citizens, we can not escape history. — Abraham Lincoln, “[2nd] Annual Message of the President of the U.S. to the Two Houses of Congress; December 1, 1862”, reprinted in James Richardson (ed.), A Compilation of the Messages and Papers of the Presidents, 1789-1897 (1897), p. 142 (where, however, we is emphasized) History is more or less bunk.1 1 Bunk here means ‘claptrap, drivel, nonsense; humbug; deceptive, empty, foolish, or insincerely eloquent talk’. But these senses arose via a radical semantic shift in — and subsequent clipping of — a word which had once been just a personal and place name: viz., Buncombe (ultimately from the transparent Old English compound bune ‘stalk, reed’ + cum(b) ‘valley’; cf. Cottle, ed., 1978: 75 and Brown, ed., 1993: 223, 300, 506). This unusual etymology has a combination of two further properties that is nearly unique and thus surely justifies granting pride of place to bunk within this first footnote in an extended general discussion of language change. The following summary draws on Bartlett, ed. (1848/1877), Barrère & Leland, eds. (1897: 193), Holt, ed. (1934/1961: 42, 129), Morris & Morris, eds. (1977: 97, 283), Lighter, Ball, & O’Connor, eds. (1994: 315-317), and especially Hendrickson, ed. (1987/1997: 111), plus Bryson (1994: 340, 448n.1); other senses and origins of bunk(s) are listed in some of these works, but more fully by Cassidy, ed. (1985: 463-464). The near-uniqueness of ‘nonsense’-bunk lies in our knowing, not only (i) the full name and the detailed identity of the person whose particular actions led directly to the semantic change at issue, but also (ii) the precise year, month, date, and even time of day when this person’s actions set the relevant change in motion. Namely, on the morning of February 25, 1820, Felix Walker — a North Carolina congressman from Buncombe County (where Asheville is the county seat) — subjected the U.S. House of Representatives to a seemingly pointless and endless oration totally unrelated to the general topic then being debated in the House (the so- called Missouri Compromise, which included a limited allowance for the territorial expansion of slavery). When Walker’s colleagues interrupted him to request that he keep to the main topic at hand, he replied, “I am only talking for Buncombe” (in fact, his speech had been written some time before and was indeed intended to impress only his constituents back home). Walker’s answer was reported in many newspaper accounts devoted to the great debate in which he had, so to speak, taken part. Almost immediately, U.S. English-speakers began to use the phrase to be Published in B. Joseph & R. Janda (eds.) Handbook of Historical Linguistics. Oxford: Blackwell (2003), pp. 3-180. 2 — Henry Ford, interviewed by Charles N. Wheeler; Chicago Daily Tribune 75.125 (May 25, 1916), p. 10 [later repeated — as “History is bunk” — in Ford’s sworn testimony (during his libel suit against the Chicago Daily Tribune) before a court in Mount Clemens, Michigan (July, 1919)]. In this introduction to the entire present volume — a collection of chapters by scholars with expertise in subareas of historical linguistics that together serve to define the field — we seek to accomplish three goals. First, we present and explicate what we believe to be a particularly revealing and useful perspective on the nature of language, the nature of change, and the nature of language change; in so doing, we necessarily cover some key issues in a rather abbreviated fashion, mainly identifying them so that they may together serve as a frame encompassing the various subsequent chapters. Second, we introduce the book itself, since we feel that in many respects this volume is unique in the field of linguistic diachrony. Third and finally, we seize the opportunity provided by the still relatively recent turn of both the century and the millennium to step back for a moment, as it were, and use the image of historical linguistics that emerges from the representative set of papers in this handbook for the purpose of reflecting on what the present and future trajectory of work in our field may — and can — be. Thus, in part one, we do not hesitate to address extremely general, even philosophical, issues concerning language, change, and language change — whereas, in part two, we focus on more concrete matters pertaining to the volume at hand. And, in part three, we present a modest, minimal synthesis that aims to assess what are likely to be the most promising avenues and strategies for investigation as research on linguistic change continues to move forward to (the study of) the past. As we pursue these three goals, we intentionally do not at any point give chapter-by-chapter summaries. Rather, we weave in references to chapters as we discuss major issues in the field, with boldfaced references to the authors here represented. talking for Buncombe with the meaning ‘to be talking flowery political nonsense’, and this was rapidly shortened to (... talking) Buncombe — with its noun soon variantly spelled bunkum — and finally (during the 1850s) also to ... bunk. Even by 1827, attestations show that the expression’s earlier sense of ‘bombastic political talk’ had been extended to cover ‘any empty, inflated speech clearly meant to fool people‘, a meaning which appears to have become dominant by about 1845 and also occurs in British usage starting ca. 1856. Partridge & Beale, eds. (1989: 68) describe bunk as colloquial in the 19th century but standard in the 20th. Lighter et al., eds. (1994) make the important observation that bunk’s link with deception was surely influenced by the non-cognate word bunco (from the Spanish card-game banca; cf. banco ‘bank’), a term for a dishonest game of cards, dice, or the like. Pace Henry Ford, the achievements of historical linguists in ferreting out all of this information are anything but bunk. Published in B. Joseph & R. Janda (eds.) Handbook of Historical Linguistics. Oxford: Blackwell (2003), pp. 3-180. 3 1. Part the First: Intersections of Language & History in This Handbook 1.1. On Language — Viewed Synchronically as well as Diachronically 1.1.1. The Nature of an Entity Largely Determines How It Can Change ...[A] language... is a grammatical system existing... in the brains of a group of individuals...[;] it exists perfectly only in the collectivity..., external to the individual.... — Mongin-Ferdinand de Saussure [as edited by Charles Bally, Albert Sechehaye, & Albert Riedlinger], Cours de linguistique générale (1916ff.), pp. 30-31; translated by Roy Harris as Course in General Linguistics (1983), pp. 13-14 ...[A] LANGUAGE... is ... a set of sentences... [—] all constructed from a finite alphabet of phonemes... [— which] may not be meaningful, in any independent sense of the word, ...or ... ever have been used by speakers of the language. [&] Linguistic theory is concerned primarily with an ideal speaker-hearer, in a completely ` homogeneous speech-community, who knows its language perfectly.... — Avram Noam Chomsky, “Logical Structures in Language”, American Documentation 8.4 (1957), p. 284, [&] Aspects of the Theory of Syntax (1965), p. 3 The range of possible changes in an entity is inextricably linked with the nature of that entity. This is a truism, but that status does not make such an observation any less significant — or any less true. On a more abstract level, it is directly supported by the differential predictions concerning linguistic diachrony that follow from the above-cited characterizations of language (in general) associated with de Saussure (1916) vs. Chomsky (1957, 1965). On the Saussurean view that langue is essentially the union of different speakers’ linguistic systems, an innovation such as one speaker’s addition of an item to some lexical field (e.g., color terminology) may count as (an instance of) significant language-change, since any alteration in the number of oppositions within some domain necessarily modifies the latter’s overall structure. But no such conclusion follows from the Chomskyan focus on a language as a set of sentences generated by an idealized competence essentially representing an intersection over a community of speakers. Published in B. Joseph & R. Janda (eds.) Handbook of Historical Linguistics. Oxford: Blackwell (2003), pp. 3-180. 4 As a more concrete example, consider the diachronic consequences of Lieber’s (1992) synchronic attempt at Deconstructing Morphology, where it is argued that, in an approach to grammar with a sufficiently generalized conception of syntax (and the lexicon), there is in essence no need whatsoever for a distinct domain of morphology. On such a view, it clearly is difficult — if not impossible — to treat diachronic morphology as an independent area of linguistic change.2 An idea of how drastic the implications of this approach would be for studies of change in particular languages can be quickly gained by picking out one or two written grammars and comparing the relative size of the sections devoted to morphology vs. syntax (and phonology). For example, nearly two-thirds (138 pp.) of the main text in Press’s (1986) Grammar of Modern Breton is devoted to morphology, as opposed to only 14% (30 pp.) for syntax and 21% for phonology (44 pp.). Nor is such “morphocentricity” (cf. also Joseph & Janda 1988) limited to “Standard Average European” languages or to what might be thought of as more descriptive works. Thus, e.g., in Rice’s (1989) highly theoretically-informed Grammar of Slave (an Athabaskan language of Canada), the relative proportions are roughly the same: 63% (781 pp.) for morphology vs.