Christian Faith and Theories of Language:         

Dr. James. D. Strauss, World View Studies

 

K. L. Pike’s Tagmemics

 

Classical Christianity has a fundamental stake in the origin and structure of language. In the past quarter century there has emerged a brilliant effort by K. L. Pike to articulate a Christian view of the nature/structure of language. The following essay is an effort to briefly describe the "key concepts" of Pike's Tagmemics and how at various crucial points this theory of language differs from the approaches of Boas, Sapir, Whorf, Bloomfield , Wittgenstein, and Chomsky et al., and how this approach creatively treats language as a part of all reality. (Cf. Language in the Context of World View.)

 

It is a general consensus among tagmemicists that Kenneth Pike was the originator of this discipline. Originally in his career, Pike was involved in the area of phonetics. However, in the year 1935. he began to form a basis for tagmemics while learning Mixtec, of Mexico, by a monolingual approach (i.e., without an interpreter). Through this approach Pike was forced to consider language and non language activity together. The next year while at the Summer Institute of Linguistics he developed a way to teach other students of linguistics how to learn a language through this monolingual approach. Pike describes how this approach led him to a belief in the central thesis of tagmemics, i.e. a linguistic theory which attempts to explain a language must treat verbal and nonverbal behavior alike:

 

In 1948 I began to devote all my research to a study of basic units of language structure in preparation for this present volume. Early in 1949 it appeared that a useful definition of each of these units had to include information concerning the distribution of these units in larger organized units. When, however, I dealt with the largest language units under attention, this principle necessitated the postulation of their distribution within similarly organized but larger units, units of a non language type, if the elsewhere useful insistence on mentioning the distributional characteristics of the units was to be preserved. This at first seemed outside of my range of investigation until I suddenly noticed that in describing the formal characteristics of a certain document—a business letter--! had actually left the area of vocal matters which I had been studying, and was in something quite different, but an area in which the principles developed for vocal utterance were already serving adequately without my having noticed the change from the one area to the other. From  this   point  onward I kept in view non vocal activity in the development of the theory.[1]

 

This hypothesis became more convincing later when he was able to find examples to fit it. Pike cites one such example at the very beginning of his book. The “example consists of a party game in which the participants begin by singing a verse of a song. As the game proceeds, the verse is continually repeated with a gesture replacing a word on each repetition, until only gestures are left and words are no longer being sung.”[2] Pike then points out that neither a structural analysis of the event nor theories and field techniques adequate for describing nonlinguistic behavior nor a musical theory can adequately describe what take place in this game. It is postulated that what is needed is a theory which will not be discontinuous, and which will not cause a severe jar as one passes from nonverbal to verbal activity. There is needed a unified methodology which can start from any kind of complex human activity with various sub-types of activity included, and analyze it without sharp theoretical or methodological discontinuities. It is this more comprehensive view of language which tagmemics strives to present. Pike's tagmemics confronts both the behavioristic view of language formulated by Bloomfield and the evolutionary view of Chomsky. Pike is a Christian and formulates his theory of language based on this Christian world view. He sees language not as a separate and completely distinct discipline, but insists that a view of language must be included as a part of a whole view of man including nonverbal as well as verbal behavior. Pike claims: "I am quite unsatisfied with generalizations which cover structures of language only; and fail to let me look out on the world and see that language is just one component of our biological output, where our enormous capacity is conditioned not merely by ingrained language components, but by ingrained capacities to view the world from different perspectives."[3] Tagmemics attempts to arrive at and describe the ordering of reality with an emphasis on language. Ruth M. Brend sums up well what Pikes' aim is. "Pike aims at accounting for language, not as a sui generis phenomenon, but as an integral part of the whole of man's life. Language is to be studied, not as an isolated structure, but as a system set off only by indeterminate bounds from context that expands in time and space and complexity to include ultimately whatever forms a part of man's experience. Pike's linguistic theory is not specifically a theory of language, but is rather a special case of a general unified theory that accounts for all cultural behavior, of which language is itself only a special case."[4]

 

Many of Pike's ideas in the development of tagmemics were earlier talked about and  posited by Edward Sapir.  It was Sapirs[5] suggestion that linguistics might help in the interpretation of human conduct. It was Sapirs belief that language events and non language events may represent structurally equivalent members of classes of events. These members may be used interchangeably within larger events. He gave the example: "If one says to me, "Lend me a dollar." I may hand over the money without a word or I may give it with an accompanying "Here it is" or I may say "I haven't got it" or "I'll give it to you tomorrow." Each of these responses is structurally equivalent, if one thinks of the larger behavior patterns." This wholistic view of language and culture was later the basis of tagmemics. Pike viewed Sapir and his work as extremely important and showed his respect for Sapir by dedicating both editions of his mammoth work to the memory of Sapir.

 

Pike throughout his life has been very active in the translation of the Bible into other languages. He believes that the problem of bringing Christianity into another culture can be viewed in part as one of translation.[6] Further, each culture has its own emic structure of customs, its own grid, its own behavior pattern. Christianity must then be translated into this new pattern. This belief is a direct contrast with Chomsky’s view of language. Chomsky believed that language cannot be viewed emicly and would insist that language must be viewed etically. Pike would disagree and insists that etic descriptions of language were invented and imposed upon the data by the outsider while emic descriptions are God’s Truth—or at least aims at it — because they are discovered within the data itself.[7] Pike is not totally opposed to etic descriptions but realizes that they are only very partial descriptions of the structure of a given language. The etic approach treats all languages at one time because it is essentially an external description. These etic units are based on prior broad samplings or surveys and may be available before one even begins the analysis of a particular language. Emic descriptions, on the other hand, must be discovered within the language and only after extensive examination of it. It is possible to obtain etic information with only partial information while emic information requires a knowledge of the total system. Etic data are generally viewed as the starting point of an analysis while emic units are the final analysis.[8]

 

Traditionally, American linguists have mainly dealt with languages at the sentence level or below. This approach may be due in part to Bloomfield’s belief that the sentence is an "independent linguistic form, not included by virtue of any grammatical construction in any larger linguistic form."[9] Essentially, Bloomfield believed that the sentence is the largest unit of

grammatical description. Tagmemics attempts to treat and examine languages on a higher level than the sentence. Tagmemic theory deals with larger units like the paragraph and the discourse. Pike states: "Our view continues to be that the sentence is a totally inadequate starting or ending point. Sentences themselves cannot be analyzed without reference to higher-level relationships."[10] By analyzing language units at a higher level it is possible to achieve a more wholistic view of language and its functions. Tagmemics also confronts the traditional approach to language which was characterized by a reductionist view of meaning. Traditionally, meaning and language were viewed as nothing more than the some of its parts--sentences thrown together, words placed one after another with no meaning individually but only as a sum.[11]Tagmemics is interested in the functional relations in the internal structure of words, phrases, clauses, and sentences as well as in such relations and contrasts among constructions. Traditional grammar spent most of its time talking about functions such as subject, object, modifier, etc., but it neglected paying much attention to form which was necessary in order to bring the functions into clear focus. At the other extreme, American structuralism over emphasized form to the radical neglect of function. Tagmemics is a reaffirmation of function in a structuralist context.[12] It is extremely important that form and meaning be kept and seen together. It is that goal for which tagmemics strives.

 

The method by which tagmemics attempts to examine language and keep form and function together is a trimodal hierarchy. Until recently American structural linguistics has assumed a model of language in which phonemes built into morphemes which in turn built into syntactic units. As a result, phonology, morphology, and syntax were regarded as successively higher layers of structure.[13] Generative grammar departed from this view of structural linguistics somewhat but tagmemics goes farther than generative grammar does. Pike set up three serial autonomous but interlocking modes. These modes are phonology, grammar, and lexicon. The phonological hierarchy is made up of elements which range from phonemes at the lowest level to syllables, to stress groups, to higher level phonological units. The grammatical hierarchy is made up of the morpheme at the lowest level, through stem, word phrase, clause, sentence, and paragraph, to discourse at the highest level. Finally, the lexical hierarchy is made up of logical chronological relationships and situational roles.[14] Tagmemics assumes that each language is composed of this trimodal hierarchy and that it can be discovered within its framework upon proper examination. It is understood that each language will have its own individual rules. However, through the use of this trimodalism it is possible to understand each individual language as a whole. Once this has been achieved then it will be possible to translate the Bible into that language. The individual rules and regulations may change from time to time but this trimodal hierarchy will always remain constant. The success of tagmemics can be seen in the fact that the Summer Institute of Linguistics has dealt with more than six hundred different languages, most of which had no alphabet, thus no written tradition. The translation of Scripture has been accomplished by many translators with understanding on the part of those native speakers.

 

There are several different elements in the study of tagmemics which must be understood. In the rest of this paper I will attempt to present some of these elements and how they are used in this area of tagmemics. First of all, essential to tagmemics is the use of the tagmeme. The tagmeme is postulated as a small element in language as a whole and in this respect is very similar to the units of phoneme and morpheme. However, unlike either of these two units, the tagmeme does not reduce language to a set of meaningless sounds. Rather, the tagmeme points in the direction of both preserving a meaningful lexical unit and of introducing a meaningful grammar unit.[15] The tagmeme is defined by Pike as: A constituent of a construction described in terms of four general features: slot, role, class, and cohesion.[16] A tagmeme is also an association of a grammatical function with the set of items which may fill that function.[17] Each of the previously mentioned hierarchies has constituents which can be described in terms of four celled tagmemes. The role slot consists of an answer to the question: why was an action performed or what was it's purpose. Pike defines the unit of slot as: "Given a sequence of events with a purpose there a-re often events which are carried out specifically to allow the completion of a larger purpose. We take the highest level purpose in an event chosen (arbitrarily, if necessary, from the total universe) for study, and treat it as the nucleus; it fills the nuclear slot in that referential unit event. This, in turn, allows supplementary, preparatory events to be treated as marginal to the nuclear one.[18] Classes or class is how or what constitutes the form of the action or the specific individual performing it. Finally, cohesion class asks: how does this item relate, to others within the system or how does it govern them or how is it governed by them. So far this may sound pretty confusing but Pike gives an easy example to help explain these concepts. He states:

 

Each feature is closely related to each of the other three. The substance that can be observed is the manifesting class. The class fills a slot which performs a specific role (or function) in a stream of speech. (In Bill hit Joe, Bill—which is a member of the class Proper Noun Root—fills the slot of subject which has the role of actor, whereas Joe, a second member of the class Proper Noun Root, fills the slot of adjunct which has the role of undergoer.) In addition, there sometimes are cohesion requirements of form for the tagmeme which integrate it further into that stream of speech, as for example, in English, the transitivity of the clause root in which the number and kinds of tagmemes in the clause root correlate with a subset of verb roots within the predicate. Another English example of cohesion is the number of the subject which governs the number in the verb (He sells cars versus They sell cars).[19]

 

Tagmemics also uses the concept of syntagmemes. These are larger units than tagmemes and consist of structurally contrastive strings and can occur in any of the hierarchical structural levels, i.e. phonological, grammatical, or lexical. Syntagmemes, as units of language, can occur in these various hierarchical levels as either a word type, a phrase type, a clause type, a sentence type, a paragraph type, or a discourse type. In tagmemic grammars these syntagmemes are represented as formulas which are made up of related strings of tagmemes. Thus the primary emphasis and interest in the tagmemic grammars are the syntagmemes and their component tagmemes along with the interrelations of the syntagmemes.

 

Within the field of tagmemics there are basically three different methods employed to analyze various hierarchical levels and the relationships in these levels between form and function. In analyzing the parts of a story an upside down tree analysis is frequently used. In this type of construction the branches (representing the parts) come out from the nodes (representing the including constructions). The branches from a single node lead to the immediate constituents of the construction at that node.[20] Above the branching line is a name for the slot of the unit in the grammatical construction into which it fits and below each line the name for its role is given. At the end of each branch is given the actual story part which that branch represents, in addition to the class name. The importance of the tree analysis is that upon analyzing a particular section of material in any given language it is possible to determine its meaning regardless of the word order. One can more precisely determine its meaning…..[21]

 

A second method of analysis employed in tagmemics is string constituent analysis. This method consists of construction which are designed to reflect a certain unit of discourse. These constructions consist of sets of obligatory and optional choices. When these sets are followed out successively through descending layers of structure the result is an end product consisting of only specific clauses, phrases or other constructions of the languages which are implied by the formula. In string constituent analysis, the clause is segmented into functional segments which manifest five clause — level tagmemes (probably actor, action, goal, locational — directional, and temporal).[22]

 

…. analyzing a language which he calls Kalaba-X. He analyzes the sentence type which occurs in this particular language. This particular sentence type has three slots which must always be filled, and filled in a particular order: Each sentence has a predicate slot, followed by an object slot, which in turn is followed by a subject slot. If we symbolize the obligatory occurrence by a plus sign, the formula, in so far as it is implied up to this point, would be.: +Pred. + 0bj +Subj. Before actual sentences can be constructed both the slot and fillers must be known. To fill the predicate slot we may use any English verb. Both the object slot and the subject slot are filled by English nouns. The formula then looks like this: +Pred. Verb +obj . N°un +Subj. Noun. One further amplification of the formula is introduced to give more flexibility to Kalaba-X and to provide an artificial language that is very simple but which nevertheless allows for the translation into it of English materials. A raodification grameme is added. This graraeme , unlike the others so far described, can come in more than one place. It may occur—but is not required to be present—after the predicate grameme, which it then modifies. It may also follow, and modify the object grameme, or the subject grameme. Although a particular sentence may contain three instances of the modification grameme no two modification grameraes come in sequence, since each of the basic gramemes is restricted to one, and only one, modifier. The filler of the modi ficai ton slot in Kalaba-X may be any English adjective. Furthermore, any noun, or any verb, may constitute an acceptable filler of the modifier slots just as an adjective can do. Note the revised formula, with the added symbols of slant line meaning "or" and parentheses to show the basic gramemes expanded a modifier: (pV+MN/V/A)  + (+0N+N/V/A) + (+SNiN/v/A). Note that this formula in part defines a noun for the Kalaba-X structure: it is a member of that list of words which can fill object, subject, or modifier slots. A verb is a word which can fill predicate or modifier slots. An adjective is a word which can fill only a modifier slot.[23]  This example is only a tiny segment in the analysis of a language. However, upon expanded work one would be ready to begin to translate material into that language. The value of this type of work is that it points out exactly what forms can occur and in what order in any given language. This example just given is an extremely simplified form of string constituent analysis whereas some of the examples shown in other articles and books which I read demonstrated how very complicated this particular procedure can become. This method is primarily used for translating and Pike claims that it has been the most effective method that has been used to develop and demonstrate the principles of translation theory and practice.

 

The final method of tagmemics by which various languages are analyzed arose from a question which Pike contemplated for many years. He wondered if it was possible to construct some kind of grammatical chart (or matrix) comparable to a phonetic chart, with contrastive features for rows and columns, but with grammatical units in the cells. Then in 1964 he came up with the idea of setting up contrastive tagmemes of subject-as-actor in contrast to subject-as-goal. The device he used to achieve this contrast ended up to be matrix charts. In these matrices, situational and grammatical roles are set up. Then when examining a language, items are fitted into these various slots. Exanimation of how the linguistic items which fill these situational roles vary in grammatical role, form sentence to sentence throughout a discourse, yields …clause as to the emic constitution of discourse structure for a specific language.[24] The matrix approach generally demonstrates relationships by arranging syntagmemes or tagmeraes in a system in which similar items are brought together and different items are separated. This type of approach allows relationships within the structure of a language to be discovered and not forced on it from an outside or etic approach.

 

Finally, Pike sees views of the structure of language as three fold. First, language is believed to be made up of particles. This view treats language as static and each of the individual parts as simply building blocks which only make sense when put together. The second view treats language as made up o f dynamic waves of behavioral movement merging into one another in intricate, overlapping, complex systems. The final view treats language as made up of parts and classes of parts which are so interrelated that no parts can occur apart from their function in the total whole, which in turn occurs only as the product of these parts in functional relation to a meaningful social environment.[25] These three views of language are referred to as the particle, wave, and field views. Pike favors the field view because in this view a word is seen not as just a part of a sequence alone, but as a vital part of a whole class of words which are not being uttered at that particular moment. A field view emphasizes a wholistic nature of language. "It emphasizes that in communication it is not the bits and pieces which communicate but the total speech event which carries communication impact only against a behavioral background of structured experience, structured memory, and structured potential.[26] This field view of language stresses the basic idea behind all of tagmemics. This idea is that language must be examined only as a part of all of reality, including verbal and nonverbal behavior alike, and not just as a minute discipline which is only relevant to itself.

It was postulated at the beginning of this paper that tagmemics was the only real Christian approach to language. Through this paper I have attempted to present a basic overview of some of the key ideas in tagmemics. At the same. time I have tried to show how Pike and his tagmemics present a Christian view of language and its structure. In this theory languge is viewed as an integrated part of all of human behavior. This aspect is especially seen clearly in the field view of language. Pike attempts to approach language from an emic rather than an etic standpoint and tries to discover structure in language by letting language express itself thorugh the trees, immediate constituent analysis, and the matrix, rather than forcing rules upon it. The relationship between function and form is very important to tagmemics. One of the basic postulates of tagmemics is that there is a general structure of language over all. This general structure, according to Pike, is the result of God's work in providing order to all of reality. Each language has its own peculiar rules which may change from time to time. However, tagmemics allows for and can handle such changes.[27]  Kenneth Pike has done a great service to Christianity and linguistics alike. It was very refreshing in researching this paper to read a view of language which supports the Christian assumptions that there is a God who has created man and has given order to all of reality. But even more encouraging than that was the realization that this Christian view of reality and language, expressed in tagmemics, has made a significant impact on linguistics as a whole.

BIBLIOGRAPHY

 

Brend, Ruth M. ed. Advances in Tagmemics. New York: North-Holland Linguistic Series, 1974.

________. "Tagmeraic Theory: An Annotated Bibliography, Appendix 1."  Journal of English 

            Linguistics 6 (March 1972):1-14.

Bloomfield, Leonard. Language. London: Rinehart and Winston, 1933.

Corapton, Carol  J. and Klamraer, Thomas P. "Some Recent Contributions to Tagraemic

Analysis of Discourse." Glossa 4 (1969-1970): 212-223.

Hymes, Dell. “Review of Language in Relation to a Unified Theory of  the Structure of Human

Behaviour, by Kenneth L. Pike”. American Anthropologist. 71 1969, pp. 361-362. Longacre, Robert E.  An Anatomy of Speech Notions. Lisse: The Peter De Ridder Press, 1976.

________. Grammar Discovery Procedures. London: Mouton and Co., 1964.

________."Some Fundamental Insights on Tagmemics." Language 41 (1965):65-76.

________. “String Constituent Analysis." Language 36: 63-89.

Mandelbaum, David G. ed. Selected Writings of Edward Sapir in Language, Culture and

Personality. Berkley: University of California Press, 1949.

Pike, Evelyn G. Coordination and its Implications for Roots and Stems of Sentence and Clause.

Netherlands: The  Peter DeRidder Press, 1974.

Pike, Kenneth L."A Syntactic  Paradigm." Language (April-June 1963):216-30.

________. "Language and Life: A Training Device for Translation Theory and Practice."    

Bibliotheca  Sacra 114 (October 1957): 347-363.

________. "Language as Particle, Wave and Field." The Texas Quarterly 2 (Spring  1959):37-54.

________.  Language in Relation to a Unified Theory of the Structure of Human Behaviour.    

Paris: Mouton and Co., 1967.

________. "Meaning and   Hypastasis,"  Georgetown University Monograph Series on Language

and Linguistics vol. 8 1955, pp. 134-40.

________. "On Describing Languages." The Scope of American Linguistics. Belgium: Peter

DeRidder Press, 1975.

Pike, Kenneth L. and Pike, Evelyn G. Grammatical  Analysis. Huntington Beach: Summer  

Institute of Linguistics, 1976.

________. "Seven  Substitution Exercises for Studying the Structure of Discourse." Linguistics 

91 (October 1972): pp. 43-52.

Schitzer, Marc L. "A Note on Tagmemic Discourse Analysis: Philosophical Argument."    

Linguistics  67 (March 1971): 72-82.

Southworth, Franklin C. "A Model of  Semantic Structure." Language 43 (1967): 702-24.

Strauss, James. Syllabus for Historiography of Theories  of Language. LCS.

Waterhouse, Viola G. The History and Development of Tagmemics. Paris: Mouton, 1974.



[1]Kenneth  L.  Pike,  Language in  Relation to a Unified Theory of the Structure of Human Behavior, (Paris:   Mouton & Co., 1967), pp. 34-35.

[2] Ibid. ,  pp. 25-26.

[3] Kenneth  L. Pike,  "On Describing Languages," in The  Scope of American Linguistics, ed. by  Robert Austerlitz,   (Belgium, Peter De Ridder Press, 1975),  p. 21.

[4] M. Brend  ed., Advances in Tagmemics, (New York,  North-Holland Linguistic  Series, 1974),  p. 2.

[5] David G. Mandelbaura, ed.,  Selected Writings of Edward Sapir in Language, Culture, and Personality, (Berkeley:   University of California Press, 1949), p. 12.

[6] Kenneth L. Pike,  "Language  and Life: A Training Device For Translation Theory and Practice" in Bibliotheca Sacra, Vol. 114, October 1957, p,  360.

[7] Ruth M. Brend, Ibid., p.  6.

[8] Kenneth L. Pike, Language in Relation to a Unified Theory of the Structure of Human Behavior,  p. 37.

[9] Leonard Bloomfield, Language,  (London: Rinehart and Winston, 1933), p. 170.

[10] Kenneth  L. Pike, Language in Relation to a Unified Theory of  the Structure of Human Behavior, p.147.

[11] Strauss,  Syllabus for Historiography of Theories of Language, Fall 1980, p. 17.

[12] E. Longacre, "Some Fundamental Insights on Tagmemics," in Language, Vol. 41, #1, 1965, p. 67.

[13] Robert  A. Longacre, Grammar Discovery Procedures,  (London:  Mouton & Co., 1964),  p. 7.

[14] Carol J. Compton & Thomas P. Klammer, "Some Recent Contributions Tagmemic Analysis  of Discourse"  in Glossa 4,  1969-1970,  p. 216.

[15] Kenneth Pike, "Language as Particle, Wave, and Field," in The Texas Quarterly, Vol. 2, #1, Spring 1959, p. 41.

[16] Kenneth L. Pike and Evelyn G. Pike, Grammatical  Analysis, (Huntington Beach, Cal.:  Summer Institute of Linguistics, Inc., 1976),  p. G-I-12a.

[17] Ruth Brand, Ibid., p. 24.

[18] Kenneth L. Pike and Evelyn  G. Pike, Ibid., p. 12-18c.

[19] Pike gives a good example of this type of procedure in Robert E. Longacre, "Some Fundamental Insights on Tagmemics," p. 70.

[20] Kenneth L. Pike and Evelyn G. Pike, Ibid., p. 13.

[21] Pike illustrates this type analysis simply by examining and diagramming the story of the Rich Young Man in Kenneth L. Pike and Evelyn G. Pike, Ibid. , pp. 15-17.

[22] E. Longacre, "String Constituent Analysis," in Language, Vol. 36, p. 68.

[23] Kenneth L. Pike, "Language and Life: A Training Device For Translation Theory and Practice," pp. 348-50

[24] Marc L. Schnitzer,. "A Note on Tagmemic Discourse Analysis: Philosophical Argument" in Linguistics, Vol. 67, March 1971, p. 72.

[25] Kenneth L. Pike, "Language as Particle, Wave, and Field," pp. 37-38.

[26] Ibid., p. 50.

[27] Ibid., p. 51.