SIL International Home

Kenneth L. Pike (1912–2000)

K. L. Pike on Etic vs. Emic:
A Review and Interview

Karl J. Franklin
Summer Institute of Linguistics
November 27, 1996


The terms etic and emic are now well established in the social sciences. However, they are often not used by other scholars in exactly the same way as Pike (1967) originally conceived them. This article explores the history of the terms as developed by Pike and how they are now used somewhat differently by anthropologists such as Marvin Harris, in particular. After some background on Pike's philosophical and linguistic orientation (as set forth in his most recent book Talk,Thought and Thing, Pike 1993), an interview with Pike explores some of the dimensions of the etic and emic framework and presents some basic philosophical presuppositions that underlie his works.


In the first section of this article I review Pike's most recent publication (1993) and then, in the second section, condense an interview I had with him on particular aspects of etic and emic in linguistic theory.

Over the years Kenneth L. Pike has been well known for his contributions to linguistics. Of particular significance is his theoretical framework, called tagmemics, which processes language and behavior from the observer's viewpoint. Most recently Pike (1993) states his personal convictions about the nature of language: how it operates from the insider's or emic perspective and consists of a system of hierarchies, each of which can be examined from alternative points of view by the observer, and all which help a person to relate to the "physical, social, aesthetic, and philosophical environment" (1993:78).

Critical to the emic unit is its place in hierarchy.1 For Pike it is important to have a preliminary or generalized view of types of hierarchical levels (1982:73). This is because emic units are found in context: the slot, class, role, and cohesion define the emic unit or tagmeme. But in the tagmemic approach one does not begin with the item lowest on the hierarchy, rather it is possible and permissible to start at any level. According to Pike and Pike (1983:10) tagmemic hierarchical organization allows for the simple retrieval of information because there is a certain stability of presentation. Any complexity is distributed over different hierarchies and at different levels of the hierarchy.

Pike and other tagmemicists, like Longacre (1983), speak of units that are larger than a phrase or clause, corresponding to sentences, paragraphs, and discourse, as well as to other hierarchies, such as phonological, grammatical, and referential.2 They also consistently relate language to human behavior.

The most basic introduction to Pike's theory can be found in his Linguistic Concepts (1982). There and elsewhere (1954, 1955, 1960, revised 1967) he explains how the concepts of etic and emic, as well as the terminology, were derived from the words phonetic and phonemic, respectively. Pike sees the terms as analogous references to raw generalized classification on the one hand (etics) and more specific, systemic analysis on the other (emics). In tagmemics the two terms are something of a "root-metaphor" with philosophical implications (1987:81), although this metaphor would collapse if it were to treat units as existing autonomously, without context (1987:93).

Central to the discussion of etic and emic is the notion of the tagmeme. It is a special kind of emic unit which combines paradigmatic, syntagmatic, pragmatic and cohesion features. Paradigmatic features are demonstrated by classes, syntagmatic features by position or slot, pragmatic features by purpose or role in a particular situation, and cohesion features are demonstrated by the knowledge and experience of the individuals concerned. The tagmeme is therefore an emic unit-in-context at various hierarchical levels. Within a text each unit at each level has a contrasting distributional feature which includes all variations of the unit (Pike and Pike 1983:7).

Pike's use of the terms etic and emic now vary somewhat in other disciplines and consequently the book edited by Headland, Pike and Harris (1990) is concerned mainly with their application to society and only marginally to linguistics. It is there (1990:28ff), however, that Pike offers a precise definition of an emic unit, with comments on its defining features: "An emic unit... is a physical or mental item or system treated by insiders as relevant to their system of behavior and as the same emic unit in spite of etic variability".

In defining the features of an emic unit the analyst should consider if:

The question arises as to what degree such emic units may vary and still be determined as the same unit by a given person or within a subculture. I return to this point later when Pike notes that variations are influenced by surrounding units, style, or that they may be forced upon the unit by the environment. Because of this, at times there will be some overlapping of features and some indeterminacy of the boundaries of units. At other times there may be permanent fusion within a particular unit.

As a further analogy to help understand the nature of emic units, Pike often (1959; 1982) uses the comparison of particle, wave and field to demonstrate that such units can be viewed from various perspectives: the default perspective is to treat the world around us as if it were made up of particles or things (1982:19) and that these things occur in a linear and spatial order (1982:23). However, sounds and words often merge with one another and can then best be perceived as waves, with nuclei and margins. In grammar there are central parts of a construction which are more important than others, but have other elements which modify it. Pike (1982:27) speaks of grammatical, phonological, and referential waves. He also uses the term field to discuss how an emic unit must be determined by context: for example, sounds and clauses have patterned dimensions. In summary (1982:86), particle, wave and field correspond to triads such as: (a) static, dynamic, relational; (b) item, process, relation, (c) point, line, space; (d) list, rule, pattern, (e) identifying, making, organizing; (f) abstraction, fusion, intersection; or (g) repetition, change, integration.

The identification of etic and emic units is then an essential part of the tagmemic framework. As Pike notes (1982:43), problems of contrast and difference are most crucial when the items compared are very similar. In order to provide some discovery procedures, contrasting contextual frames, contrasting matrices, and word pairs which display contrasting senses, are all applied. There are non-verbal clues as well.

Emic units are expected to have variation or to change over a period of time. There are changes in pronunciation and grammar, as well as indeterminacy as to where segmentation may take place in a more complex unit.

A further relevant feature in the identification of units is their distribution. They may occur as members of a class of substitutable items, as part of a structural sequence, or at some point in a system.

However, an emic system may not be transparent, without training, to the cultural insider. Pike attempts to bridge the gap between idea and thing (1993: 34ff), mind and matter, or will and act with the emic concept. Therefore thinking, imagining, and speaking are all kinds of emic behavior which are used by the cultural insider.

The larger setting of emic units may be more crucial in describing and understanding the unit than the immediate environment. Pike attempts to establish how nonverbal circumstances affect behavior by adding the referential hierarchy (Pike and Pike 1977). This shows how relations, things, people and events are emically important. Pike concludes that "knowledge is so vast a network, with so many intertwined items, in so many emic hierarchical network patterns, that it is not possible to say everything, all at one time, about any tiny bit of an event" (Pike 1993: 39).

Pike (1987:77) acknowledges that his linguistic theory has to be one that is philosophically valid. He says, "I want a philosophy which I can live by, as well as think by. " He accepts as a given that there is an emic nature to known reality and that language and other behavioral structures have units which are emic. These are "partly relative to that particular culture and are partly constrained by innate human characteristics and the relations of people to that part of the world which is outside them" (1987:79).

2. The Interview3  top

Franklin: What I would like to do is to clarify some of the things I understood from the etic/emic book and ask some other questions that I believe are related to it.

First of all, please comment on the difference between etic and emic variation, in terms of the limits of variation. For example, you have described emic variation in relationship to a baseball player, where you mention various components of style and individuality. But what are the limits of etic versus emic variation?

Pike: First of all, what do you think is the difference between emic variation and etic variation, so that I can comment more appropriately?

Franklin: Emic variation would be the type of variation that still allows the same event to be recognizable by insiders of a culture. Insiders would all see the event as somehow the same, in terms of the type of action or circumstance, but there would actually be variations to it. How would an insider or an outsider know when it was no longer that particular event?

Pike: Note that there is etic variation within an emic unit, at least in the way I would normally use the terms. Take some unit or event which is recognized as such by the people that you are referring to. Suppose that in commenting on it the person slows down a bit, speeds up, stutters, or he hesitates or uses his left hand instead of his right hand—this doesn't affect the emic unit. It would constitute etic variation within the emic unit. By emic variation I would normally mean a change from one emic unit to another, it would be like playing football instead of baseball or some other sport.

I think what you are asking has other implications. I will try to use a common illustration so that my terminology will not be confusing. Given a particular instance, such as the same person going routinely to the same work in the morning and facing the same problems every day, we have an identical cultural unit. How much variation can there be? If the person sees the time is getting late he will hurry, only he must hurry even faster if he suspects he might be late. Now if it is even later, so that he does not get there on time, then he may be rebuked because he has gone beyond the normal acceptable limits as defined by the boss. Supposing, however, he gets sick and does not go to work. Going to bed sick is really different from simply not going to work at all. In fact it is sharply different. Further, suppose he has a cold but that he still goes to work. As far as the work place is concerned others may be accepting an awful nuisance, especially if the man is sneezing and giving others the flu.

My next comment is that, in any discussion, there is a particle view, a wave view and field view. We have begun our discussion of variation as if there were a particle view. We act as if we could actually know just where the unit or event began and ended. This is acceptable when we are dealing with events or things as chunks, as units, or as particles. We can then define them in certain logical terms. But in reality no two things are exactly ever the same, no event, no repeated so-called "sameness" is ever exactly the same. You put a different foot on the sidewalk at a little different place each time when you are going to work. There is always an etic variability, there is never exactness.

The question then arises: how much variance can there be? How late can you be before the boss fines you because you did not get to work on time? Suppose you do not show up until 3 o'clock in the afternoon. Did you go to work that day? The boss will probably be a little reluctant to say that you came to work satisfactorily. From the wave point of view, in general, there is a normal, prototypical (the term some people would use) form. This is the expected, nominal, or central form, which may vary slightly in any one direction, or in one of a dozen different directions. It is like the pebble hitting a pond: how far does it's wave go before it is no longer the same wave? There is no way of determining this.

This brings us to the next component, which is again related to your question. Who defines how far the wave goes before a conclusion is formed that it is no longer a wave or before one decides that it is no longer that ripple? This depends on several things. First of all, consider the individual defining the action. There is a variation between that individual and someone else, the observer to the data. We say this because emic is partly observer related. It is not some absolute or physical thing determined from the outside, but it is related to the observer. There can be no escape from this conclusion in physics of any kind.

You may reply, "What does the observer do to make the decision?" Firstly, it depends upon the observer's goals and interests. To return to our previous example, the boss wants everyone to arrive at work on time. So if he is a particular boss and someone arrives late to work, he will discipline him. The worker may be only five minutes late, but he could be disciplined. However, if it was a friendly boss who saw you arrive late, he might simply tell you to try and do better next time. The degree of variation which is allowed then depends upon the individual observer.

Again, you may challenge this and question how an individual observer decides that something is true. In my review of Horden's book (Pike 1961) I comment on his view that there are no true statements in the Bible. To him when the Bible reports that "God is love", the statement is not inerrant. However, (in my terms, not his) that does not mean that God has a girlfriend, which might be included in the word love. His interpretation of the statement "God is love" would not go that far. By claiming that there are no inerrant statements in the Bible, he means there are always different possible views. I agree, in that sense. But then what are true statements? For Horden, one must go to science for aberrant statements. At this point I conclude that he is wrong because statements in science are made by measurement. And a measurement takes a reader, and a reader as observer can move by a millionth of an inch and then the observation is no longer absolute. So eventually the whole basis for determining inerrancy disappears. There is none, consequently Horden's whole basis of truth in science disappears

What I said in my review is a sensitive issue, in that the truth is partly related to what a person is trying to say. What am I am interested in determines my focus and other matters can be left as unspecified detail. Further, truth must not be equated with precision. The minute you equate truth with precision you have gone to etic exactness. There is no true statement ever in any of our experience because, if we measured it within a millionth of an inch, it would always be different. Truth, in its emic identity, and related to your question, is associated with the purpose, interest and the intent of the observer. When those differ between observers then there is a problem. When their goals differ they have to confer to understand each other's goals. One may say to the other, "Oh, from your point of view I see what you mean" or, "Oh, now I have to take a different measurement, by an eighth of an inch. "

How close have I come to answering your question?

Franklin: That is partly what I had in mind. In Etics and Emics (Chapter 3) Harris talks of scientific observers and their etic observations of particular cultural acts. The scientists then somehow evaluate the relevance of the acts. He uses etics as if the etic variation was somehow outside of the intentions of the pool of scientific observers.

Pike: Several comments: First of all, about Harris. When he looks at something as if he is objective, i.e., as a scientist, he is not part of the community which he is studying. It is the same manner that I would use to study ants, because I am not one of their community. That is etics for him. However, as soon as I talk about a particular person's goals, interests or purposes, to Harris this is emic. If he can get at such things physically then it is etic. This is not by saying "what is your goal?", but by saying, "what does that person do when he goes to work?" If the person looks at his clock and sees that it is close to eight o'clock (his starting time for work), then he hurries. The observer deduces etically that the person hurried to get to work on time. When I talk about the feelings that the insiders have, this is emics to Harris. Now that is not the way I use the terms, although Harris has a point. I will support his point in just a moment. When I bring to bear my etic equipment (which Goodenough (1970) calls my etic bag of tricks), I can apply this to an emic structure to help describe that emic structure from my point of view as an outsider. This is the basic system of my phonetics book (Pike 1943) but, as a matter of fact, the structure of that book has been systematically organized by an observer. So in a sense it has a structure which is related to the observer. Therefore scientific discipline has a structure within it and this, from my point of view, can be called the emics of science.

Now secondly, to be fair to Harris, we will further attempt to represent him. Suppose that I take a boy to play croquet and then later he plays baseball. It is still the same boy I can look at both events with a camera. I note that the game of croquet has a different set of rules than that of baseball. The player has an emic croquet and an emic baseball. I would grant that. But what am I doing when I am being a phonetician? I am not playing croquet—I am playing phonetics. And in this sense you could call my phonetics an emic system, because you are talking how I think about phonetics. Or, you could call croquet an etic system, although I don't like that because I want, at any point, to be able to talk about what a person is doing.

Next, consider what I am doing with my phonetics when I apply it to Hottentot. I'm then not studying Hottentot from the point of view of how that system functions although, emically, I want to talk about how that system functions. Etically, I want to use a camera or recorder to gather data so that I can get inside of that system Nevertheless, I am not satisfied until my recordings are more intimate and I can say, "Oh, now I understand how Hottentot functions could become a part of this community if they'd accept my bodily shape."

I confess that I have used phonetics in two different ways. Normally, I use it as a technology to start moving into a system and to get closer to the emics. This phonetic methodology is part of my bread and butter. However, and as I acknowledge in the book, before I use the methodology I already have a background system of phonetics in my head. I draw on this general system of phonetics to use the phonetic methodology.

The confusing part, which I try to clarify because of my contact with Harris, is that I have used phonetics in those two different senses. I have learned of this confusion from Harris. My normal way of talking about phonetics is as a methodology. Suppose, for example, that I want to understand the spectrograph of some sound. I write it down as a somewhat crude example of what is going on in the mouth when I make those sounds. But then I ask myself where the methodology comes from. The methodology didn't come from Hottentot. The methodology was there first and it represents the summation of the sounds from many languages, as well as my mental experiment to construct any sounds that I could make up in my mouth.

By way of analogy, inside of me is a baseball game which is my phonetics. The baseball bat and helmet are a part of my phonetic equipment, like my spectrograph is a part of the phonetic equipment. It follows that I do have a kind of a system, which you could call an emic system, which is relative to my discipline as a phonetician. But it is not at all what I would have called an emic system in the past. When I have talked about getting into an emic system, I have usually talked about getting into someone else's emic system

I met Daniel Jones once when I was a young man. Now suppose I wanted to study his disciplinary emics because he managed his phonetics (Jones 1917) a little differently from mine. He recorded the sounds on phonographs, which were to be the standard for the whole world, including his great grandchildren. I never used that because I was not interested to prescribe exactly a precisely phonetic alphabet. Why did Jones do this? So that his students could correctly read what his other students recorded or heard. I didn’t care about such detail. I wanted the phonetics to be fairly exact, but I also wanted the work to result in a phonemic alphabet to which I could give a general definition. I wanted my colleagues to learn to talk a language, perhaps at first somewhat sloppily and with errors, but then pick it up accurately from local speakers.

Notice now the sensitive issues that I am summarizing The first is that, in general, Harris wanted emics in a man's head but he wanted etics as a scientist observing from the outside. However, a scientist also has something in his head while he is observing, so that distinction does not quite work for me. Secondly, I wanted an emic system to be a phonemic system of a particular language. (Eventually I moved on to the emics of baseball, football and family breakfast.) Thirdly, and stimulated by Harris, I needed to go a bit further and talk not only about football vs. soccer emics or baseball and cricket emics. I had to conclude that "a scientific methodology is based on a background system" and this can be somewhat rough, differing from scholar to scholar, like languages or particular subdialects do. I grant this.

So we can see why Harris and I work so differently. Harris started with a belief in evolution and, therefore, any explanation had to be diachronic over eons of time. I started without an interest in evolution because I figured that whatever had happened, God had made it happen. I was interested in the present, what was going on in a language now, so that I could communicate now. It was this communicative component that I was interested in, the emics of the now. Of course, Harris was interested in how something got that way over a 2,000-5,000 year period, the diachronic. That is important too, whether you are an evolutionist or not. We have been around a long time, no matter when one dates the flood.

We can see that English is related to Sanskrit, even though this was an outrage to our ancestors a few hundred years ago. Nevertheless, I grant that even today there are residues, in any phonemic description, from the process of change. The change is not complete and this will disrupt a phonemic description and require a diachronic component for a synchronic description. Therefore I need what Harris does or what the historians have done. I don't give history the same attention because my goal has been to help my colleagues with their present work. On the other hand, Harris needs our synchronic analysis to help understand the historical period of thousands of years ago Otherwise, he cannot really reconstruct what was going on in a place like India 5,000 years ago Once Harris reconstructs that period he must arrive at a system that can be used by the people, can be communicated, and can be emically applied by the people, even though it has changed over a period time.

How can this be managed theoretically? My main stimulus has been for my colleagues, although for my own pleasure as well. But the consideration has been in terms of particles and chunks which are in relationship to a system or field. Harris, on the other hand, has a system which operates from the change or wave view I grant that we need a wave view. He should grant that we need a particle view. But neither of them can exist without some relation to a system or a pattern. This is because eventually truth is a pattern within pattern, until you get to the top, wherever this is for you, and then you stop. I go up to theism and then I can't explain God. Harris goes up to a mechanism and then he can't explain the Big Bang. Now why should evolutionists come up with a baby when they have a Big Bang to begin with? Both of us are stuck for certain explanations, so it seemed reasonable that we should look together to find pattern with pattern.

Put this together and eventually we begin to answer your question about etics and emics. There is both slippage and flexibility, because the observer is involved. Hence we try to answer each other's questions, like we are doing. But even in doing this, Harris may point out something which doesn't quite make sense and I will need to add more to my discussion.

Franklin: Another interesting statement by Harris is that it is the linguist's emics that makes linguistics possible. Why didn't he say that it was the linguist's etics?

Pike: Do you remember what I said about a phonetic system? Recall that it is the phonetician's etics which makes linguistics possible from the point of view of the phonetician's etics. Now we can do the same thing by looking at the system of the nature of alphabets (not just the sounds but their structure). Because we have such a system in mind, which has been developed over the past 150 years or so, we are able to write a book on phonemics. That's how I would interpret it. It involves a degree of truth. The same is true for grammar, we are terribly indebted to the Greeks and Romans for Subject Object and Verb and for case. When the Romans got the accusative case it sounded to them like it was accused but not so the Greeks. So if it had not been for them we might not have linguistics of our kind. It was Panini in India who pushed morphophonemics and sandhi, where we got the term. We could have developed linguistics freshly but it would not have developed so quickly or in its present shape.

Linguistics depends on systems which were developed before, in the wave view of time, and without that we have no etic. The same holds true for society. If we didn't stand on the shoulders of someone who spoke of "papa" and "mama" we would not know how to use those words.

Franklin: So when Harris talks about emics making linguistics possible, he's speaking of intention. However, he seems to be careful about using this term when he talks about science.

Pike: Let's assume this is so, but I hesitate because it took me months and months and months to try to understand Harris. Would you like to know how I got started talking with Harris? I was in Spain at the request of some philosophers and spoke there on the relationship of language to the world (Pike 1987). Afterwards they told me that Harris had been there three months previously lecturing. When they invited me, they had sent me some articles with some references to the etics and emics of Harris. That is precisely why they had invited me. Harris had said that he wished he could talk to Pike.

So later we invited Harris to Norman [Oklahoma] to lecture. I asked him to arrive at least a day early so that we could talk privately before the lecture. So we spent four hours talking prior to the lecture. Tom Headland then met him at an AAA meeting and arranged the meeting and we both agreed.

We had a difficult time trying to understand each other. We each spoke 20 minutes, with 10 minutes for reply by the other. Later, we saw each other's materials so that before publication we could revise our own materials after having read the comments. The commentators could also revise their materials after having read the revisions of our revisions. So we had maximum time to try to understand each other. Even so, every so often I still get a little perplexed.

Franklin: Murray, another author who comments on you and Harris (Murray, Ch. 10), said that you both were interested in problems in society, but in different ways. He quotes from your book With Heart and Mind (Pike 1962), in which you talk about Christians going into a society like yeast. Murray claims that the idea of yeast in a society is strictly a metaphor and is not a scientific construct. I did not understand what a scientific construct would be in his view of change in society. Did you feel that he was probing? Perhaps he was goading and trying to say, "Well, both of you are interested in a better society, but how are you trying to accomplish this?

Pike: There were five original commentators, including Keesing from Australia. When he dropped out we asked Quine and two others. I talked with Quine at length, and he read my philosophy chapters and made very helpful comments. And then Berry, the psychologist, referred to a number of psychologists who have used the concepts of etics and emics. It is an important concept in psychology because (I think it was Berry who said) Freud was an outsider looking at the patient, but also needed is the view of a person's inside structure to really understand the problem. The etic alone is just not adequate. So we have the advantage that two philosophers and a psychologist have commented in the book, and I am very happy about that.

Franklin: There was an additional thing I wanted to ask. Because we have an absence of written records in most of the languages in which we work, Harris said that it is impossible to provide an emic description of them. At an older level of description there are only dead people and dead records in most of the languages in which we work, Harris said that it is impossible to provide an emic description of them. At an older level of description there are only dead people and dead societies. When we study proto-languages and then we can make certain observations about the food the people ate or how they lived, or make other comments, aren't we in some sense providing an emic description?

Pike: As I said, we guess at the emic structure, and we are going to be partly wrong. But if we don't assume that they had an emic structure, we have assumed that the people are not human, and then what is the point? So we have got to guess why people did what they did. We attribute things to them and we may be wrong. But in daily life, we guess that someone is talking because they want to tell us something, or they want to make us change something. So we are guessing all the time, and we're attributing certain intentions to others.

The news accounts said that in Iraq they put some of their major headquarters among civilians. For what purpose? Because they believed we were partly moral and that we didn't want to kill their civilians? But, in my opinion and in an extraordinary way, there is a hidden moral structure there which concerns emics. We have got to assume that people were trying to take care of their families and have some universal of humanism.

In the Human Relations Area Files, before computers, anthropologists were studying some 240+ languages with 3x5 slips all around a room twice as big as this one. In all of those cases they found no exceptions to some taboo against incest. The minimum was a tribal group in India where man and mother was the only taboo. So we have got to assume there are universals of human behaviour, universal tendencies, universals of a moral code. One of my favourite quotes by a philosopher at Harvard who said that there is "radical relativism under rigid restraints" (Goodman 1978). There is relativism within the emics but within that there is an underlying constraint about human nature. And that is also part of your first question.

Franklin: What do you think the contribution of tagmemics is to philosophy?

Pike: There are two or three things. In the first place, in terms of something concrete that I have written, there is my article on "The Relation of Language to the World" (Pike 1987). If I were to rewrite it today I probably would not do much better, although I wrote it several years ago. It astonished me when I read it recently (I had almost forgotten that I had published it) to see how close it was to my current thinking. When you read that article you have my first comment.

Secondly, tagmemics starts with the relation between language and people. I invited Quine to come to Norman to lecture. I asked him to come for a Monday night lecture, and stay until Tuesday because I was giving a monolingual demonstration. I had him sit in the front row. I don't think he had ever seen one, so I had it taped for him, and gave him a copy. He has spoken for a couple of decades about "observational sentences" (1960; Quine and Ullian 1978, and most recently in Quine 1990:42). If you point at a rabbit and someone says "gavagain," you guess he means "that is a rabbit". And if the audience agrees with you, that gives social support to your judgment, to your understanding of what it means. That is an "observation sentence". Quine and another philosopher (Quine and Ullian 1978), publishing over a decade ago, say that the observation sentence is at the base of all knowledge, all science, basically all belief. This leading philosopher takes my monolingual demonstration approach as dealing with the basis of all knowledge, all reality, all of science, everything. That is my second answer.

Now that is astonishing. How could I be so egotistic as to believe that anything like that could be true? I think it is because God has implied that very concept. We read that God was in the beginning the Word. This is a strong statement about language. And when God said, "Let there be light", it is somehow related to this philosophical observation about sentences. We then come to what answers your question more directly: When God told Adam to name the animals two things happened. One, whatever man called it, that was the name it was, for God, in talking to Adam.

God doesn't talk to me in Hottentot. If I had a Hottentot Bible it would do me no good whatsoever. God speaks to me in MY language, or perhaps a tiny bit in some other language, but not much, relatively speaking. Why is this so important? Because when we name the animals, that is one kind of categorisation of our environment. Another way to describe this is to say that it is "building our world-view", "building our structure", or "putting everything into a system". We can manipulate this structure with symbols and words and thoughts, in our heads, rather than with our feet. And it is astonishing and l think true, that the Biblical way of representing the system is an elegant summary. It doesn't just mean words like "giraffe" and "camel", but something vastly more. It is categorising our environment, as well as controlling the ecology which we talk about at the same time. You can't plough very well if you can't say "get me a plow" or "build a plow" or something related to these concepts.

Now, as a footnote to my comments on naming, it was years before I noticed this categorisation was related to the first chapter of Genesis. God looked at the waters and he called them "sea" and he looked at the dry land and he called it "earth". If God was categorising to control his environment, why would he have to do it? Because he is the Word—that is his structure—and we are created in his image, to categorise in his image, to control our environment in his image, through categorisation in his image. Now once this is said, notice that, due to the fall of man, people would not want to retain the relationship, even if they could. So, what happened? Man has tried to reject the personal-observer relationship to reality and has instead adopted a philosophical, scientific dream to be absolutely objective. That is the dream of science, to be objective and get rid of this sloppy, sloppy person involvement. And it is the dream of some objective science, objective philosophy, mechanism, Watsonianism, and more. Personal involvement depends upon categorising and extending categories by analogy or metaphor. As, I think, Robert Frost the poet said, "Metaphor is a ladder by which you climb the skies". There is no way to write poetry without metaphor, which is taking categories and expanding them abstractly and systematically. We cannot be human without being able to talk about a "mama bear" and a "papa bear", and being able to talk about the "head" of a mountain or the "foot" o a mountain. This use of metaphor is a part of being human and it comes from language.

The dream of science and philosophy has been to reject the insider and deal only with the outsider. I say that it cannot be done. We examine what is going on and what we are doing and we categorise these. In a sense we distort reality, in that we are drawing boundaries around it. And when we say this' is a "tree" and that' is a "bush", we talk as if they are particles. But at times we must accept even a sloppy wave-view because sometimes we cannot be quite sure which we see. Otherwise we can't talk meaningfully, or we can't "plant", "sow", or "reap", or we can't put the "dogs" in a place, or keep the "elephants" out.

Franklin: When concepts get fuzzy like that, we can coin another word, so that if there is "smoke" and also "fog" we begin to talk about "smog".

Pike: And then that word may be changed again. And when we do that there is always some degree of ego involvement because as human beings, we name and recognize. Of all people linguists are directly involved in understanding the world, science and philosophy.

Chomsky (1957) rejected this position, from my point of view, when he implied that the NV VP are not a set of concrete words but are merely symbols of representation. And he had no semantics at that time. Instead he built his system on logic.

Now, as a theist, I am in a position to accept a person as basic to the universe and therefore able to categorise the universe basically. An evolutionist must start with the whole world in existence, except for people. Everything must be developing logically, systematically in an incredible manner such that a shepherd dog can cull out a sheep. All of the development before man had to be something. Mind is therefore just one further stage beyond, or self is just one stage beyond, but the self as a soul is not accepted. The soul is everlasting and there is no room for that in evolution unless they are into science fiction. Starting as a theist I have a belief in a person, so l start with an affirmation "person above logic". But the minute I say "person above logic", I clash with the whole history of logic and science. Why? Because I have taken the sloppy observer—you or me—as being more important than the atoms of this "table". In order to understand the universe or to understand this table, I need to first of all understand you as a person more than I need to understand the atoms. In the Pikean sense, when somebody says, "I want something to put a book on", I then need to understand the "table".

Several years ago I was asked to lecture to the philosophy club at the University of Texas at Arlington. Afterwards a philosopher said, "You sound like Immanuel Kant". I checked up, and I did. He had said that we do not know "a thing in itself". We observe something outside of us and we believe there is something outside of us. We believe there is something concrete out there, but we only can control something being out there when we categorise it, as an observer. So we do not know a thing in itself. Rather, we know a tree as a tree because it is different from a bush, it is not a bush because it's a tree, and it is not a weed because it's a tree. I know that only because it is me as the observer, and not because of the things in themselves. We only know it with the observer who may distort or categorise. If we are friendly we categorise, if we are hostile we distort.

We now come to my involvement with reading physics. Some years ago when I read Einstein and Infeld (1938) on "The evolution of physics" I was astonished and said "Einstein's problems are my problems". About that time I went to the University of Texas to lecture. The vice-president had started a new journal which was intended for non-specialists. He wanted department heads and professors to write for it, so they asked me to write an article. At their request I wrote an article on language as particle, wave and field (1959), for non-linguists, as an interdisciplinary exercise. Later I pushed this concept because I saw that it was struggling with problems of observer relationships.

Still later a professor explained to me that, if there was a screen with a hole in it and a wall behind it, then an electron gun could be used to shoot electrons through the hole to hit the wall and make a circular wave pattern. But we cannot say that the electron went through the hole. We can only say it disappeared from the gun, and reappeared against the wall. Why? Because, how do you know it went through the hole? Only by bombarding it with something, we hit something and then it bounced. This then affects our equipment and tells us that we hit something. But if it hit something, then it wouldn't go through the hole. Therefore you cannot say it went through the hole; you can only say it disappeared and reappeared. Well, that is the observer for the physicist.

One day I was at the railroad station with two physicists and I was talking about the observer relationship. And one of them said, "The speed of light is a constant. " So I turned to the other physicist, who is a theoretician, and I said, "Well, suppose you said the speed of light is observed as a constant." He said, "No problem." Now notice that all of physics is built on the speed of light as a constant. But how do we know that it is a constant? Only because it affects some of our equipment in certain ways. So the observers are found all over the place.

Your question was basically: "Why is linguistics involved in finding reality and truth?" I am saying that God made categorisation basic and that some philosophers recognize this. If they then say that "In the beginning was the Word" is more than metaphorical, they capture something of what God's nature is involved in, some kind of categorising. So when God saw the dry land, he called it "earth". We are created in his image. Who in all the world is in a better position to talk about ultimate reality, philosophically, than someone who believes there is a creator? It must be someone who is ready to listen to observed data, not just reject it in the Eliphaz manner, who refused to listen to Job's data (the Book of Job, Chapters 4, 15, 22). We must deal with categorisation as best we know how (which isn't impressive), utilizing semantics and grammar.

God has dealt with a variety of world-views since Babel (Genesis, Chapter 11). Not only does language let us categorise, but it allows us to control one another with that categorisation. And if there had been no splitting of languages, certain situations all over the world would not have occurred, such as Russia becoming fractured over language. Rather categorisation centers around language because the categorisational principle forms social identity, to some degree.

Who then, in all the world, ought to have a long shot at finding a movement in the direction of truth? Someone who believes that God is creator, who believes that "in the beginning was the Word", who believes that we are created in His image—to categorise, to name. Who, but one who is dealing with language formally, with a variety of scientific approaches to language, so that we are not trapped in one particular subset that is available at the moment?

What are the chances of this viewpoint being accepted? What do we do? If we bury our talent in the sand it's not right. If accepted, we then have the danger of our egotism. Few scientists want to accept a dualist, who believes in the soul. Some reject the soul but believe that you can think and choose. Or take the anthropologist, Leslie White, who says in his book on science that there is no choice whether you believe in Christ because it's a matter of glandular response. Well, there is no probability whatever that my material will be accepted in such circles. So what?


We went on to discuss the impact of Pike's etics and emics on psychology (Cf. Berry, Chapter 6, "Imposed Etics, Emics and Derived Etics: Their Conceptual Status in Cross-Cultural Psychology") and the general contribution of Pike's phonology to contemporary materials on autosegmental and metric phonology. Goldsmith (1990), for example, views Pike's study of Mixteco tone (Pike 1948) as a classic in the literature.

Tagmemics has not been widely adopted outside of the SIL context and, even there, it has been modified and expanded in ways that are quite different from Pike and Pike (See, for example, Longacre 1983). Pike commented, "There is a tremendous drive towards what is called `universals', but which are not `absolute' universals but rather universal possibilities, or tendencies. Universals which are really absolute universals include the ones I have talked about in the concepts' book (Pike 1982), e.g. about particle, wave and field approaches. Contrast, variation and distribution are universal—you can't talk without them"

Pike, of course, was interested in establishing a set of etic universals, for example a list of all the possible clause types in languages which would parallel the list of all possible phonemes or syllable types. But the practical application of theoretical principles has always been foremost in his teaching and writing. There is no doubt that his etic and emic contribution is widely accepted and that it has provided fundamental insights to our understanding of language and culture.4

NOTES  top

1The notion of hierarchy varies in linguistic theory. Early generative grammars treated lexical substitutions on the basis of dividing vocabulary into hierarchically ordered classes and subclasses" (Lyons 1968:164). More recently Cruse (1986 181ff) discusses non-branching hierarchies, with tagmemics and grammar in focus. In his view a sentence includes clauses, phrases, words, and morphemes, each in a descending order. A sentence can be analysed as consisting of a margin (e.g. subordinator, subject, and predicate) and nucleus (e.g., subject, predicate and object). The tagmemes in the nucleus likewise consist of parts such as determiner, modifier, and head.

2The referential tagmeme and hierarchy is examined in detail in Pike and Pike (1977; 1982, 1983). Referential structure is concerned with the communication referents of speech and, from an etic perspective, includes all that people talk about. By taking note of all of the contrasting and identifiable cultural factors in a speech community the linguist can discover the emic referential system of that group (Pike and Pike 1983:35).

3Some time ago (Franklin 1991), in writing a review on the book Etics and Emics, edited by Pike, Harris and Headland (1990) I outlined some of the fundamental differences between Pike and Harris. After I had written the review I had the opportunity to spend time interviewing Pike on certain assumptions about tagmemics and his discussion with Harris in particular. The interview took place in February 1990 and is reproduced here. I am grateful to Janet Ezard for editorial assistance and to Alan Healey, Phyllis Healey and Graham Scott for helpful comments. I am also indebted to Dr. Pike for reading the transcriptions and editions of the interview and for permission to reproduce part of it here. My association with Dr. Pike goes back to 1956, when I was one of his students at SIL at the University of Oklahoma where students studied his phonetics (1943), phonemics (1947), tone languages (1948) and the very beginnings of his tagmemics (1954 and 1955). In 1961 Pike came to Papua New Guinea, where he ran a workshop on grammar and syntax and trained a number of consultants, including me. I have continued to have contact with him over the years.

4Hussey (1989) documents the extensive use of the terms etic and emic in various disciplines, but especially anthropology and linguistics.


Chomsky, Noam. 1957. Syntactic structures. The Hague: Mouton & Co.

Cruse, D. A. 1986. Lexical semantics. Cambridge: Cambridge University Press.

Einstein, Albert and Leopold Infeld 1938. The evolution of physics: the growth of ideas from early concepts to relativity and quanta. NY: Simon and Schuster.

Franklin, Karl J. 1991. Review of Emics and Etics, by Headland, Pike and Harris. Australian Journal of Linguistics 11.132–36.

Goodenough, Ward A. 1970. Description and comparison in cultural anthropology. Chicago: Aldine.

Goodman, Nelson. 1978. Ways of wordmaking. Indianapolis: Hackett Pub. Co.

Goldsmith, John A. 1990. Autosegmental and metrical phonology. Oxford: Basil Blackwell.

Headland, Thomas N., Pike, Kenneth L., and Marvin Harris. (eds.), 1990. Emics and Etics: the insider/ outsider debate. Newbury Park, CA: Sage Publications.

Hussey, S. C. 1989. An annotated bibliography of publications using the emic/ etic concept. Dallas: Summer Institute of Linguistics.

Jones, Daniel [1917]. English pronouncing dictionary. Revised by A.C. Gimson, 1977. Revisions and Supplement to the Fourteenth Edition by Susan Ransaran, 1988. London: J.M. Dent & Sons Ltd.

Longacre, R. E. 1983. The grammar of discourse. New York Plemum

Lyons, J. 1968. Introduction to theoretical linguistics. Cambridge: Cambridge University Press.

Pike, K. L. 1943. Phonetics: A critical analysis of phonetic theory and a technic for the practical description of sounds. Ann Arbor: University of Michigan.

———. 1947. Phonemics: A technique for reducing languages to writing. Ann Arbor: University of Michigan Press

———. 1948. Tone languages: a technique for determining the number and type of pitch contrasts in a language, with studies in tonemic substitution and fusion. Ann Arbor: University of Michigan.

———. 1959. Language as particle, wave, and field. The Texas Quarterly 2(2).37–54.

———. 1961. Strange dimensions of truth. Christianity today 5.690–92.

———. 1962. With heart and mind. Grand Rapids: William B. Eerdmans Pub. Co.

———. [1954, 1955, 1960] 1967. Language in relation to a unified theory of the structure of human behavior (2nd. ed.). The Hague: Mouton.

———. 1982. Linguistic concepts: An introduction to tagmemics. Lincoln: University of Nebraska Press

———. 1987. The relation of language to the world. International Journal of Dravidian Linguistics 16.77–98.

———. 1993. Talk, thought and thing: the emic road toward conscious knowledge. Dallas: Summer institute of Linguistics.

——— and Evelyn G. Pike. 1977. Referential versus grammatical hierarchies. In R.S. DiPietro and E. Blansitt. (eds), Discourse structure: some new dimensions for linguists. Colombia, SC: Hornbeam Press.

——— and Evelyn G. Pike. 1982. Grammatical analysis (Rev. ed.). Dallas: Summer Institute of Linguistics.

——— and Evelyn Pike,. 1983. Text and tagmeme. Norwood, NJ: Ablex.

Quine, W. V. O. 1960. Word and object. Cambridge: Massachusetts Institute of Technology.

———. 1990. Pursuit of truth. Harvard University Press.

——— and J.S. Ullian. 1978. The web of belief. Second edition. New York: Random House.