Recent years have seen a steady increase in the number of attested basic word order types. Greenberg (1963) originally recognized only three types, VSO, SVO, and SOV. Pullum (1977) added VOS to the list, but proposed that no others were possible. Derbyshire (1977) confirmed the existence of OVS, and recent reports (SIL Grapevine 1978) have indicated that OSV must also be recognized. These findings have caused a certain flurry of excitement among theoretical linguists. Reactions to the discoveries may be broadly divided into two classes, both of them annoyingly smug.
The first reaction is typified by M. Joos, in a hastily-
The second reaction is typified by P.M. Pistol, in a 763-
While I am confident that the excessive pessimism of a Joos need not discourage us in our search for Universal Grammar, I will demonstrate in this paper that linguistic metatheory must be sufficiently enriched to allow for the existence of yet another basic word order, VOV, an order apparently not predicted, or even allowed, by any existing theoretical framework except stratificational grammar, which seems to allow everything except English.2
The language in question is Kluj (pronounced ‘cloodge’), spoken by some seventy-
Accompanied only by a German shepherd guard dog and an Uzi submachine gun, I carried out eight field trips among the Klujans between 1961 and 1976.3 These field trips were supported, if that is the word, by the U of SND at H.4
During these field trips I was able to obtain, for the first time, an extensive corpus of material recorded in Kluj.5 This material I have now transcribed and analyzed, and while a number of difficulties remain,6 the most important features are reasonably clear.
One of the most interesting aspects of Kluj grammar is word order. Although it was a relatively simple matter to establish that the verb occurs both initially and finally in the sentence, the order of the remaining elements appeared quite unaccountably variable until I suddenly realized that it was governed by a simple, if novel, rule
Remarkable as these details are, they are surely not as striking as the fundamental VOV pattern of Kluj sentences, since this represents a previously unattested basic word order. No doubt the reader’s mind is crowded with questions at this point. Let us begin with the question which is easiest to answer.
Where does the subject go in a Kluj sentence? The alert reader will have observed that the VOV pattern makes no allowance for an S, such as is typically found in the basic word orders of many other languages
α A β → α ω β A ∈ VA; α, β, ω ∈ (VA ∪ VT)* 8
The reason that a Kluj sentence has no overt surface subject is that the subject is invariably incorporated lexically into the verb. I do not mean that the subject is represented by a bound morpheme attached to the verb. Rather, to each English verb there corresponds a large number of Kluj verbs, each lexically marked for a distinct subject. For example, corresponding to English ‘go’, Kluj has, among many others, the following verbs (since this paper deals with syntax, I have omitted some details of the phonology):9
votsiʔ ‘large dog goes’ gv inlak‘small dog goes’ juʔpis ‘toad goes’ ʔuʔnax ‘Dodge pick- up truck goes’ m iʔs‘small disc- shaped object of little value goes’ mleʔtx ‘girl of marriageable age goes’ pwaʔ iʔ‘respected but irascible person goes’ nvil ‘I go’ ʔ iswai‘Harold goes’ ʔosp i‘fascist warmonger goes’ dzeitloʔ ‘Ralph Nader goes’
It will be appreciated that the addition of even a single new noun to the language necessitates the creation of hundreds of new verbs. The coming of 20th century civilization and technology to the Klujan’s valley has therefore meant that tens of thousands of new verbs have had to be coined to cope with the enormous influx of new nouns. Indeed, during recent years, most of the sober waking hours of the Klujans (and those are few enough) have had to be devoted to thinking up new verbs, causing a sharp decline in such favorite traditional practices as gang rape and dropping rocks on turtles. No doubt the intense hostility of the Klujans toward outsiders stems principally from this cause.
Indeed, I suspect that it is just this remarkable feature which accounts for the fact that the verb appears twice in the sentence. The point is that there are so many verbs to be mastered in Kluj that the Klujans, who are none too bright at the best of times, find difficulty in selecting the correct one to express a given meaning. Diachronically, it is easy to imagine that a Klujan speaker who had just uttered a sentence with the wrong verb would have had another stab at it, and that grammaticalization of this no doubt extremely frequent type of utterance would have resulted in the VOV pattern of modern Kluj. Support for this hypothesis is found in the observation that, even today, it is a rare sentence in which both verbs are actually identical, and in the further observation that utterances of the type VOVV, VOVVV, and so on, are not uncommon, as the speaker makes repeated attempts to find the right verb.10
Kluj is particularly interesting from the typological point of view. Since it is both a verb-
Consider adpositions. It is well-iʔ...dzlots ‘in/on/at/around/under/beside/between’, as in
(1) gluʔn ipflaxi nviʔspwatsdzlots gluʔni
‘I’m gonna stick this knife in/on/at/around/under/beside/between your fat gut.’
(In this case the context did not allow me (rather fortunately, I think) to determine which of the various possible interpretations was intended, and it may have been that all of them were. It will be noted that in this case the speaker had no difficulty in selecting the correct verb on the first attempt; this is typically so with threats, which appear to come more naturally to my Klujan informants than most other types of utterances.)
The same elegant solution is found with relative clauses, which surround their head noun, and with possessives, which behave similarly (actually, the first person singular possessive appears to be the only one in common use). As for verbal modifiers, adverbs, naturally enough, are simply placed between the two occurrences of the verb. The placement of modals might appear to pose a somewhat more difficult problem, but here Kluj exhibits a beautifully simple solution to its typological dilemma: it doesn’t have any modals. This might appear to be a remarkable lack, but the Klujans seem to manage very well with threatening gestures.
Kluj might have been expected to exhibit gapping in both directions, but in fact it has no gapping rule at all. It might at first seem surprising that a language whose speakers have so much difficulty remembering verbs should pass up a rule which deletes verbs from the sentence, but a moment’s thought reveals that the deletion of a verb in Kluj would inevitably entail the deletion of its subject as well, and even the Klujans are not so stupid as to attempt to communicate with a string of clauses of the general form O.
In sum, then, it appears that linguistic metatheory must be suitably enriched so as to be able to account for this remarkable and fascinating language,11 that many theoretical positions must be rethought and many putative universals abandoned, that many other grandiose if somewhat vague conclusions must be found if I am ever going to get this work accepted as a Ph.D. dissertation, and that this paper had better be published before Sticky gets out on parole, since he continues to harbor a most unfortunate and, in my view, rather unscientific attitude toward my findings.12
2 P.M. Pistol (personal putdown) now informs me that Car Park Grammar does indeed predict the existence of the VOV pattern, given yet another device, the Double-
3 I am a slow worker.
4 It must be admitted that the linguistics department at the U of SND at H is perhaps not quite so up to date as it might be on the theoretical side, and that my attempts to introduce the faculty to works written since 1952 have met a mixed response at best. Nevertheless, in a spirit of Christian charity, I can only forgive the somewhat intemperate comments on my work which have occasionally been made, as, for example, by the head of the department, Professor Snodgrass: “Fieldwork? That long-
5 These recordings are now available in the Language Archives and Fertilizer Storage Space at the U of SND at H, under the heading Kluj, vols. II-IV and VI-VIII. Volume I was never recorded, due to an unfortunate difficulty with the tape recorder. Volume V was recorded, but after a Klujan cat chose to relieve himself on it, it acquired a rather pungent smell, and the librarian/gardener refused to have it in the Archives.
These tapes were the source of some of the difficulties between Professor Snodgrass and me. The economic situation being what it is, I had to re-
6 It must be confessed that these difficulties are at least partly due to the fact that, for excellent medical reasons, I never attended a single one of the phonetics practicals I was supposed to do for my degree. It is also true that I was never able to solve any of the morphology problems in Nida’s book, but I believe that linguistics should concern itself with higher things than morphology.
7 On making this discovery, I immediately informed Professor Sticky of Ohio State University, who I knew was even then composing a masterwork to prove that such things cannot be. He promptly obtained a grant to come and see for himself, and a few days later he arrived at Cojones, smiling confidently and assuring me that he would “have this little wrinkle sorted out in half a jif”. As the days passed, however, he became increasingly morose and refused to talk to me about Kluj. One night he turned up at my hut with a strange expression on his face and explained that he had found a beautifully simple solution to the whole problem and was going out to check it with some native speakers. The nature of his solution began rapidly to dawn on me a few minutes later when I heard a long burst of fire from my Uzi and, rushing outside, I found that he had machine-
8 Pretty nifty, eh? Actually, this bit has nothing to do with the rest of the paper, and I’m not even too sure I understand it. I just copied it out of a book and stuck it in to give the impression I’m au fait with all this dreadful hairy formal stuff you keep seeing nowadays
9 Such as consonants and vowels. I have used the symbols ʔ and i more or less at random to indicate any segment whose precise phonetic nature I was never able to elucidate. This should not be taken to mean that the remaining transcriptions are accurate. Those damn Klujans never seem to pronounce a word the same way twice.
Anyway, for what it’s worth, a summary of Klujan phonology will appear in the forthcoming Festschrift Snodgrass. Professor Snodgrass, feeling somewhat aggrieved that no one had seen fit to prepare a Festschrift in his honor, intimidated his students into organizing one. The response so far has been somewhat disappointing, however, as the only contributions have been three dog-
10 Even more common, in fact, are utterances in which the final attempt at the verb is followed by a curse or, more commonly, a string of curses. Actually, such patterns as VOVC, VOVVCCC, VOVCCCC, etc., are so frequent in ordinary speech (the Klujans apparently finding curses easier to think of than verbs) that for a while I considered taking VOVCC as the basic word order. I finally decided against this on the ground that it was undignified. It was in connection with this problem that I seemed to have found a language that was worthy of my talents.
G. Gazdar, eschewing for once the use of the second-
N.B.: I am compelled by the US Freedom of Information Act to admit that in fact I have no such counter-
11 P.M. Pistol, in the vanguard as always, assures me that he has already added several new colors to his collection of pencils, and that we can expect revolutionary new developments from him any day.
12 It is customary, in works of this sort, for the investigator to express, in the most nauseatingly fulsome terms, his undying gratitude to his informants for having put aside their livelihood to answer thousands of idiotic questions about subject-