Kenneth L. Pike (19122000)
A Linguistic Pilgrimage*
Kenneth L. Pike
The first of these items I co-authored with Gary F. Simons, under the English title Toward the historical reconstruction of matrix patterns in morphologypublished, in Russian, in Voprosy jazykoznanja (Pike & Simons 1993). For historical reconstruction we are especially indebted to the work of German scholars over the past two centuries. Their approach, as I would see it in retrospect, was heavily lexical-phonological. They would compare a set of words from one language with those from another language, and note that there was some similarity of meaning between the words, and some similarity of sounds between those same words. Then there was an attempt to guess (a) at a possible starting point where the words may have started with the same sounds and meaning, and (b) at a set of rules by which they could have been differentiated over time. Forty years ago I wanted beginning students, including those who were not particularly interested in historical reconstruction, to be aware of this approach. So I wrote a booklet called Axioms and Procedures for Reconstruction in Comparative Linguistics (Pike 1951). I tried to make explicit these procedures, drawing heavily upon work of Bloomfield (1933) summarizing the approach, plus the work of some Americanists dealing with languages which the readers would not be expected to know about. (I continue to get reports that this booklet is occasionally used for supplementary reading in some current classes.)
But in the early 1960s I ran into a lexical problem in the morphological analysis of the Fore language of Papua New Guinea. A younger colleague was finding it difficult to identify the prefixes and suffixes, and to specify their meanings. We ended up (see my "Theoretical implications of matrix permutation in Fore (New Guinea)" of 1963). It seemed difficult to give a simple description of the phonological content of the affixual morphemes since some of the same phonological material was found with different meanings. I ended up with what might be calledin retrospectan irregular Venn diagram (an overlapping circle arrangement) of the paradigm. Thus, a unit /a/ could help signal either first person singular, plural, or dual object, or singular first, second, or third person. Similarly a /t/ could occur in the prefix with first or second person, provided it was either plural or dual. It was the intersection of such groupings which allowed one to tell what the specific local meaning wasnot an isolated simple morpheme made up of consonant plus vowel.
On the other hand, a Venn diagram in relation to a subject suffix had a peculiar shape: a /w/ would occur in the first singular position, or in the second plural position; and a /n/ would occur in the first plural position, or the second singular position (making a kind of a 'X' shape! Then, to complicate the picture, if one added an emphatic suffix, there was a smearing so that the first singular and second plural ended with a zero, whereas the first plural and the second singular ended with /mpe/. But the Venn diagram shape (the 'X') was retained! This eventually led to a suggestion that for some historical reconstruction it might, in certain places for certain languages, in an early stage or research, be more profitable to seek for reconstructing the irregular Venn diagram shapes rather than the sounds themselves. It appeared that in Papua New Guinea these particular shapes reappeared again and again in many different languagesand these irregular paradigmatic shapes were more constant than the sounds themselves!
This became the basis of Pike & Simon (1993), suggesting that such irregular Venn diagram shapes might be the basis of a whole new approach to historical reconstruction in the next generation. I started our article because of the request of the historian Gamkrelidze of Georgia of the former Soviet Union for an article on Potawatomi of the Algonquin family. I had written an article (1964) co-authored with Barbara Erickson (now Hollenbach), based upon data by Charles Hockett (1948). Our astonishing conclusion was that a certain paradigm appeared to be strangely irregular if one treated persons in a rank order of first, second, and third singular, and then first, second, and third plural, but if one gave ranking to second, first, and third (ignoring for the moment, the plural) it showed a striking simple structure of the prefix. Yet one had to do the same kind of rearrangement for each of the three suffix positions, and then have the intersection of the four affixual patterns, in order to get the semantic impact of the interlocking Venn diagram.
A former student of mine, James O. Morgan (1996), compared eight Algonquin languages and showed how these Venn diagram shapes continued across them, despite radical differences of phonology. Simons then applied similar criteria to languages of the Solomon Islands to show how reconstruction could be seen for one of those languages in terms of contrastive development with fusion, versus a different language with extensive analogical change. Our assumption is that the combination of these approaches could lead to major historical work in the future beyond what I personally expect to live to see. (That is, one part of my autobiography is a component of hope.)
My second article appearing in 1993 was Experimental linguistics in language learning (Pike 1993a). This is essentially a bibliographical listing of some twenty articles in which I suggested different ways to change a text in order to help the analyst find permissible degrees of grammatical variability, with semantics changes often carried by those grammatical changes. (Here, also, I would hope that within the next generation other scholars might go beyond the details and techniques which I have usedor which they have used in comparable articles.)
I now refer to some of these articles here, giving the date, and co-author when one is present. (But I do not list the bibliographical details. These can be found either in the article referred to, or in the bibliography compiled by Ruth M. Brend and published in 1987.) In 1945, in my Intonation of American English (pp. 199221, 125131), I tried alternate markings of intonation to help foreign students. In my book, Language in Relation to a Unified Theory (2nd ed., 1967, Section 13.5), I showed different intonation markings of a poem of Dickensons, as read by different people. In a 1964 article on college composition and communication I suggested experimental techniques for helping a student to learn to write technical materials. In 1972, with Schoettelndreyer, we show that when, for a text in Sherpa (of Nepal), every pair of sentences is reversed in the order of presentation, a trained scholar (a trained native speaker) can revise the grammar adequately, so that it says the same thingexcepting that across certain larger units this is not acceptable. This difference presumably reflected a gap between paragraphs. In 1973 I showed that sociolinguistic presuppositions could affect the way in which speakers could be expected to reply to one another. In 1983, a discussion of a simple tale, in English, of (a) Abe came home, (b) he ate supper, (c) he went to the movies could require a word after if one told the story in the revised sequence of c, a, bbut this pointed up the need for some special language exercises for teaching people English in Papua New Guinea, where there are many languages in which there is no word for after. (They might need to say, in their own languages, Coming home, and eating supper, Abe went to the movies.) In 1988, I took a short poem of mine (Tension), and re-told the poem in several different ways, leaving the referential background structure unchanged, but changing the focus. It demonstrated that no vacuum is left behind for the emptied cells of slot, class, role, and cohesion, but that something else would fill the temporarily emptied cellsdemonstrating their psychological validity as an absolute universal of the components of the four-cell tagmeme.
The third item which has appeared during this 1993 year has been my small book called Talk, Thought, and ThingThe Emic Road Toward Conscious Knowledge It is an attempt to try to use ordinary people interacting in ordinary conversation as a basic entrance point into philosophical knowledge. It claims that the person is more basic to knowledge than is mathematics. Person is above logic, in this sense. It shows how a person can view knowledge at one moment as made up of isolated bits (particles and points); at another moment as life and change (wave); and at another moment as life made up of a system within systems (cohesion in a field). So knowledge turns out to be, in part, the recognition of pattern with pattern within patternsup to the top of ones hierarchy of belief about the world. (I would hope that this would help tie mind and matter together, and be a bridge between idealism and mechanism, rather than letting the two remain philosophically separate.)
*Originally published in E.F.K. Koerner (ed.), First Person Singular III: Autobiographies by North American scholars in the language sciences. Reprinted with permission.