The Origins of Infixation in Kamkala
H.D. Onesimus
Enthusiastic is hardly a strong enough word to describe the reaction to my recent small article “Cultural Tones” (Speculative Grammarian CLXVIII.2, October, 2013). Indeed, I have been the happy recipient of a remarkably large volume of pleasant correspondence regarding this modest piece of scholarship, far exceeding, in fact, the response to any of my previous work. Some readers may be aware that “Cultural Tones” was the result of my first fieldwork experience, and thus contrasts with my earlier insistence that linguistics is properly “a discipline of the desk” (Onesimus 2004).
To be frank, I had no idea that a single article could elicit quite so much admiration—indeed, I had no idea that any single act I could perform could elicit such an outpouring of affirmation. In my experience, scholarly achievement has consistently been as thankless as was my own childhood—a never-ending battle for recognition, by steadfastly indifferent adults, of the praiseworthy in anything I might do. I had believed that other scholars were cut from the same cloth, so to speak, as my own parents.
I do not mean to say, of course, that affirmation should in any way modulate the direction of one’s scholarly endeavors. Scholasticism is, and will ever remain, aloof from the baser motives of praise and publicity, charting its course purely under the direction of the twin guides knowledge and understanding. No regard for feeling can be admitted to this high-minded enterprise.
Nonetheless, it occurs to me now that there may be room for the possibility that a genial reaction from fellow scholars signifies that one’s research direction shows promise, and therefore I decided to undertake a second fieldwork expedition, once again in Borneo, this time with the Kamkala people.
Though they have never heard of each other, the Kamkala are near neighbors to the Kamgala (whose language was the subject of my well-received article “Cultural Tones”), separated only by five high mountain ranges, a couple of class 4 rivers, and two uninhabited jungle wildernesses locally known as “tiger alley” and “the killer ape trail.” Travel from one community to the other would doubtless take less than six months if anyone were brave enough ever to attempt it again. Therefore, I naturally entered this linguistic community with the expectation that certain similarities might be found between the languages. And I was not disappointed, though there were also differences of a rather unexpected character, requiring a bit of internal reconstruction (which I undertook later, in reflective residence at my desk) before they yielded to the forces of analysis.
Kamkala is a garden variety SVO language with, for the most part, head-initial behavior. The only noteworthy exception is that verbal morphology is not prefixed, as typologists would predict, but rather infixed, as the following items illustrate:
(1)
a. igan ‘eat’
i-to-gan ‘will eat’
b. magan ‘be willing’
ma-to-gan ‘will be willing’
c. amega ‘go’
a-to-mega ‘will go’
d. aga ‘lecture’
a-to-ga ‘will lecture’
e. obogobo ‘die’
o-to-bokobo ‘will die’
f. to ‘wish for’
to-to ‘will wish for’
In contrast, Kamgala is verb-final, and has verbal suffixation. The Kamgala cognates of (1a-f) are given in (2a-f) for comparison.
(2)
a. ikan ‘eat’
ikan-to ‘will eat’
b. makan ‘be willing’
makan-to ‘will be willing’
c. ameka ‘go’
amega-to ‘will go’
d. aka ‘lecture’
aka-to ‘will lecture’
e. obokobo ‘stay for a long time’
obokobo-to ‘will stay for a long time’
f. to ‘want’
to-to ‘will want’
Clearly these are related languages, with patent cognancy among most of the verbal forms and even the inflectional morphology. How, then, do we come to find the morphology prefixed in one language and infixed in the other?
The answer, of course, is found in reconstructing historical processes, which leads to significant observations about morphological developments and their associated sociolinguistic phenomena. In the remainder of this article I will outline these historical processes.
Proto-Kamgkala was an SOV language, whose Urheimat was in or near the current Kamgala homeland. As we might expect of a verb-final language, Proto-Kamgkala had V+Aux constructions, with the auxiliaries partially grammaticalized from erstwhile full verbs, functioning to express typical Tense/Aspect/Modality distinctions. At some point prior to the advent of modern record-keeping (that is, at least 14 generations ago), some Proto-Kamgkala speakers migrated out of the Urheimat, settling eventually in today’s Kamkala homeland.
After this breakup of the Proto-Kamgkala speech community, the two daughter languages diverged in a major way—speakers of the Kamkala dialect intermarried with the local speech community, no record of which is preserved, except that it left its mark on Kamkala in the form of substrate interference, leading to a shift to SVO word order, on the part of some, but not all of the Kamkala community. This led to a shift from V+Aux to Aux+V for some speakers, but this order coexisted with the earlier pattern, preserved by the less innovative segments of the population.
Kamgala dialect speakers, meanwhile, remained entirely faithful to their original SOV word order, apparently not having found any SVO speakers to marry.
Subsequently, both dialects undertook grammaticalization of their AUX constituents, with the concomitant reduction and affixation of these elements that typically occurs. Thus, Kamgala innovated V+suffix constructions, while in Kamkala both V+suffix and prefix+V constructions were to be found.
The final step is obvious: Kamkala speakers experienced significant cognitive dissonance when subjected to prefixes and suffixes with the same meaning, differing only in which sociolinguistic sub-communities produced them. Neither sociolinguistic variant won out, and compromise was achieved by splitting the difference and moving the prefixed/suffixed forms into the words: hence, infixes.
It may be hypothesized that all infixing languages owe the emergence of this marked morphological process to a similar history of cognitive dissonance. I hope that data from other languages will confirm this hypothesis.
Thus we conclude another surprisingly pleasant experience in field linguistics, which has again led to insightful theoretical conclusions. I can only hope that this modest description of fieldwork results will lead to felicitous responses from my readers.
