LAPID, Jezelle Irish C.

Linguistics 140
2008-51494 Prof. Viveca Hernandez

The Tagmemics Theory

1. Introduction

This paper is an attempt to look upon some of the advantages and disadvantages of the

Tagmemics Theory, which was proposed by Kenneth L. Pike and greatly contributed by Robert

Longacre and the other linguists. What is probably interesting about this theory is that it was said

that the grammatical structure of a language may conveniently be analyzed and described by

using tagmemes (Elson and Picket, 1962), thus Kissell (2005) referred it to as “The linguistic

theory of everything”.

This paper then is composed of two parts: the (1) analysis which consists of the

background, and the framework discussed along with its advantages and disadvantages; and (2)

conclusion.

2. Analysis

2.1. Tagmemics: Background

2.1.1. Background

It was in 1948 that Kenneth Lee Pike began the search for a syntactical counterpart to the

phonological and morphological terms, phoneme and morpheme--something at the sentence

level which could function as a key identifying unit in the same way that these well-established

terms functioned (Edwards, 1997).

He wanted a theory that was easy to learn and easy to use but complex enough to explain

real language (Randall, 2002). What he was looking for was a high-level generalization that

1
could characterize all human language and which would simplify the training of missionaries and

Bible translators who would encounter previously unstudied and thus grammatically uncharted

languages. The result of Pike's search was the tagmeme and the linguistic system that has come

to be known as tagmemics. But what was most interesting about his search was the fact that what

started as merely a "language theory" soon evolved into a structural theory that attempts to

account for all of man's behavior. Indeed, Pike's seminal work of three volumes is

entitled, Language in Relation to a Unified Theory of the Structure of Human Behavior

(Edwards, 1997).

Pike (1967) introduced this theory in his said book as primarily designed to assist

linguists to efficiently extract coherent descriptions out of corpora of fieldwork data. He

developed this theory while doing field research and teaching linguistics.

The theory is founded upon certain axioms about human behavior and language use that

foreground the situatedness of all communication and the necessity of viewing every act of

discourse as a form-meaning composite inseparable from communicators, their audiences, and

the varied worlds they may construct and inhabit through the use of language (Edwards, 2008).

Robert Longacre (1958, 1960) has made several contributions to the development of the

model, particularly in the area of grammatical studies (Elson and Picket, 1962).

2.2 Tagmemics Framework and its Advantages and Disadvantages

An offshoot of structuralism, tagmemic’s primary concern is with grammatical. It must

be remembered that structuralism ignored functions of a linguistic form and concentrated only on

form. Tagmemics, on the other hand, fuses together the form as well as the function of a linguistic

entity.

2
 In his book Grammar Discovery Rules, Longacre (1968) mentioned that “tagmemics

attempts to present linguistic patterns in straightforward and summary fashion. Such patterns

when carefully described for one language may be compared and contrasted with patterns

described for another language. That tagmemics labels linguistic pattern in some fashion is here

an advantage: it is difficult to compare things that do not have bear names. Patterns thus

described and labeled conform to “the first significant attribute of a pattern”.

Pike’s grammatical analyses identified linear or syntactical sequences (the "syntagmatics"

or horizontal dimension) and categories of events (the "paradigmatics" vertical dimension). The

analysis of ordered communication (which includes just about everything people do) would

involve identifying two dimensions of a grid: the syntagmatic sequence of slots or types of

events in the sequence and how they might be positioned in relation to one another; and the

paradigmatic (as in paradigms) array of classes (or objects, subjects, actions or other entities)

that could occur in any given syntagmatic slot. This is also known as "slot-class" analysis (Ross,

2010).

Another of Pike’s main claims was that language is deeply hierarchical, in several ways.

In tagmemics, all three hierarchies, namely, phonological, grammatical and referential, interlock

and operate at the same time, and of course, what could be said of the hierarchy of language

could also be said of the hierarchy of behavior. Pike even expanded it further to include other

fields such as from quantum physics, he borrowed the notion that any event can be seen from the

perspectives of particle (a static view of the unit), wave (a dynamic view), or field (a unit in

relation to other units (Kissell, 2005).

In this approach, there can be a simultaneous analysis on the utterances at three

interpenetrating levels, where each level represents a hierarchy of units. Tagmemics also stresses the

3
hierarchical ordering of grammatical units into ranks of levels -- morphemes, words, phrases, clauses,

sentences, paragraphs, and discourses. Contrary to the Immediate Constituent where the analysis of

the structaralists insists just on binary cuts, tagmemics always goes in favor of string constituent

analysis, and have many cuts. Tagmemics, unlike a structural analysis asks for the function of the

categories and not merely their naming (http://jilaniwarsi.tripod.com/tag.pdf).

2.2.1 The Tagmeme Defined

Tagmeme is a grammatical unit, which is a correlation of a grammatical function or slot

with class of mutually substitutable items occurring in that slot (Elson and Picket, 1962). A class

is the list of mutually substitutable morphemes and morpheme sequences which may fill a slot.

Tagmemes are said to be manifested by the member of the class occurring in the slot in a given

act of speech and that they may be parts of constructions at some level in a hierarchy of

constructions. The tagmeme is a functional point at which a set of items or sequences occur. The

correlativity is so intimate that the function cannot exist apart from the series nor has the set

significance apart from the function (Longacre, 1968). There is a distribution on the slot-class

correlation within the grammatical hierarchy of a language (Elson and Picket, 1962).

In the slots, many substitutions are possible which an advantage is since it would be

easier to manifest the member of the class occurring in the slot in a given act of speech, of which

a tagmeme does.

Tagmemics is hierarchical where grammar is structured in terms of level. Also, tagmemes

are conceived as units where they may be parts of constructions at some level in a hierarchy of

constructions. Although morphemes generally follow each other, in some sense, in a linear

sequence in speech, in tagmemics, they are sometimes simultaneous, as when two morphemes

are fused into the same phonemic stretch, or when morphemes of tones or stresses occur. A

4
tagmeme may also be discontinuous (Elson and Picket, 1962) as unlike with structuralism a la

Bloomfield, Harris, etc. but more directly with the basic principles of structuralism going back to

Saussure’s lectures, which the tagmeme on the other hand is the correlated class and slot, by

which is meant paradigmatic and syntagmatic relations a la Saussure, one that can be taken as a

non-generative, taxonomic exercises and a very useful exercise for learning a language and

translation (Rocine, 2001).

A disadvantage however may be seen in the naming of the grammatical finctions where

terms vary from one linguist to another and where traditional terms are used which may not be

suitable on a certain language, since each language is different in its own way. However, Elson

and Pickett (1962) said that the use of traditional terms has useful mnemonic value, and there is

an attempt to use the terms whenever possible. Another possible advantage is that any term can

be used to distinguish a filler class as long as you know at what slots they occur.

One of Pike’s major contributions to linguistic theory is his distinction between emic and

etic viewpoints. Pike coined the terms from the endings of the words, phonemic and phonetic.

The "emic" view is the perspective of the insider, the native, and is concerned with the

contrastive, patterned system within a universe of discourse; the "etic" view of a unit is the

perspective of the outsider who looks for universals and generalizations. The "emic" view is the

view we expect from a participant within a system; the "etic" view is the view we expect of the

alien observer. It is an advantage to learning a language that you not only want to study its

objective form (“etic”) but also to know the interesting thing as to how language functions for

users in real life (“emic”).

Tagmeme is described in terms of four features (or “cells”), those are the (a) slot, where

the unit can appear; (b) class, what type of unit it is; (c) role, how the unit functions; and (d)

5
cohesion, how the unit relates to other units. These very same structures both appear on lower

and higher levels—as sounds form words, words form sentences, and sentences form discourse,

tagmemics is used to describe these larger and larger units. And thus, it is very interesting to

realize that if the emic/etic distinction applies to all levels of language, then perhaps a more

general principle could explain a great many things too (Kissell, 2005).

2.2.2 Tagmemes and Constructions at Various Levels

A tagmeme may be manifested by a morpheme sequence which in turn is analyzable in

terms of tagmemes. Arising in this situation is the significant fact that not all morphemes in a

given sequence are equally related to each other, but cluster into groups which in turn cluster into

larger groups in hierarchical fashion. A morpheme sequence which fills a grammatical slot (thus

manifesting a tagmeme), and which is itself analyzable in terms of constituent tagmemes, is said

to be manifesting a construction. Simply put, a construction is a (potential) string of tagmemes

whose manifesting sequence of morphemes fills a grammatical slot (Elson and Picket, 1962).

Tagmemes and constructions of various types are distinguished as being at different

levels of the grammatical hierarchy.

2.2.2.1 Tagmemes and Constructions at the Clause Level

The definition of the traditional term, clause, is stated in terms of construction rather than

morpheme sequences. It is considered as a ‘predicate-like’ tagmeme, which may not seem like

predicates to the speaker or English. Also, traditional term definitions are sometimes done in

cases when they fill a certain grammatical slot (classification or distribution by, or occurrence in

a slot). The purpose of the general definition is to allow recognition of at least some clauses in

any language, but for each language the specific definition may have to be amplified in a way

suited specifically to that language (Elson and Pickett, 1962).

6
 The predicate term used in identifying clause-level tagmemes doesn’t match the

traditional one in English; rather, it is referred only to that part of the clause which is manifested

by verbs or their equivalent. This setup can cause misunderstanding had no proper introduction

about the use of the term has been given. It is important indeed to notify people of the meaning

of the technical and common terms used especially if the use is different from the usual meaning

(Elson and Pickett, 1962).

A method of describing these constructions is by means of formulas. The presentation of

the analysis is done by a formula in the order slot: class, the former being written in capital letter,

the latter in small letters. One good thing about this formula setup is that it creates distinction to

the different levels being described since each formula differs from one level to another. Also,

formulas are easy way to summarize information as it can be quickly read and errors are easily

detected (Elson and Pickett, 1962).

Here, it manifests as well that tagmemics follow a heuristic approach which works

through guess-and-check model. Longacre’s work on “Discovery Procedures: Guess-and-check”

works well for the linguistically modest but spiritually noblest of goals: translating the Bible into

as many language as possible. It might be thought at first that since it is a guess-and-check

model, the conclusions remain guesses. The check method renders a guess as reliable. The

checking is bottom-up (Rocine, 2001). There is an inspection in the forms of the language and

make guesses at the function of each word or group of words which will be illustrated in the

chart the tentative names for the functions and are placed in the forms which fill each slot. Its

main target is an analysis of the language (particular language) not on their English translation

but as to how the native speakers perceive the language (Elson and Pickett, 1962).

2.2.2.2 Tagmemes and Constructions at Phrase Level

7
 One problem with clause-level slots is that they are only filled by single words, which are

limited and quite atypical of normal style in most languages in general, for many clause-level

slots are often filled by word combinations. Such combinations are traditionally called as

phrases. A phrase is defined as potentially composed of two or words, always in the definition of

traditional terms. Its distributional feature is “typically, but not always, fill slots on the clause

level”. In addition to this most typical distribution, phrases may be embedded in other phrases.

These instances may cause ambiguity had someone not familiar with it. Distinctions between the

two must be clearly analyzed and remembered so as to avoid confusion (Elson and Pickett,

1962).

A different symbolization in the formula is also done in this level. This change in

symbols is very much a good thing to distinguish one level from the other.

2.2.2.3 Tagmemes and Constructions at the Word Level

The word-level distinction is appropriate only in languages where there are affixes;

unlike in languages where every word is a single morpheme there would be no need to

distinguish this level (Elson and Pickett, 1962).

Tagmemics didn’t abandon the Bloomfieldian definition of the word as a “minimum free

form”. This definition assumes that in every language there are phoneme sequences that the

native speaker will pronounce in isolation and that cannot be broken down further. However,

phonological junctures between these isolable forms are not assumed therein. This word

definition is so convenient and that it would be unwise to abandon it, even though there are

problems with its application in many languages (Elson and Pickett, 1962).

The above definition of the word is applicable to every language. But then, not every

language will necessarily have a separate level in the grammatical hierarchy. Word-level

8
constructions are only applicable to words with more than one morpheme (expandable) (Elson

and Pickett, 1962).

Unlike tagmemes at higher levels, frequently in word-level tagmemes, it is difficult to

separate function (indicated by the slot name) and the class meaning of the fillers for classes of

affixes (and other small classes composed mainly of single morpeheme members) are apt to have

uniform class meaning and function (Elson and Pickett, 1962).

Systems using the same symbol for both slot and class (e.g. t:t and t) have the

disadvantage of giving the appearance of no distinction between slot and class. But then, it is not

easy to find sufficient different symbols to clearly distinguish the two. A possible solution was

carried out in the use of numbers for the classes, modeled after the decade numbering system for

affixes originated by Voegelin which was first used in traditional morphological description. In

this system, each affix is given a decade number and each morpheme within the class is given a

specific number. However, this may also be disadvantageous considering that one must

memorize or better say be very familiar with the numbering system as contrary to making

everything straight to the point (Elson and Pickett, 1962).

2.2.2.4 Tagmemes and Constructions Below the Word Level

The necessity to posit a stem level within the grammatical hierarchy happens only when

sequences of more than one morpheme fill the nuclear slot of word constructions. Although the

term stem is both referred to as either single morphemes (roots, simple stems) or morpheme

sequences (derived stems), only the latter relates to situations in which stem constructions

(derived stems) are involved (Elson and Pickett, 1962).

Stem-level constructions may also be composed of a combination of root and affixes.

There are two kinds of affixes namely the derivational affixes, which fill stem-level slots and

9
inflectional affixes, which fill word-level slots. However, distinction between the two cannot be

simply defined, to apply to all languages, and the difference is not always clear in a given

language. Longacre’s guess-and-check is frequently applied therefore on the basis of simplicity

of description as well as the application of some clues for deciding as to whether an affix is

derivational or inflectional (Elson and Pickett, 1962).

The clues include (a) the external distribution of the resultant form, which cannot always

be sufficient evidence to concluding an affix derivational; (b) affixes occurring with only limited

number of stems may usually be derivational; and (c) affixes whose meaning is not easily

determined often proved to be derivational (Elson and Pickett, 1962).

Cases where satellite (affix) slots in word constructions are less frequently filled by more

than one morpheme than are the nuclear slots are probably best analyzed as manifesting a single

word-level tagmeme and a construction at a lower (affix-cluster) level. The affix-cluster level is

said to be parallel with stem level (Elson and Pickett, 1962).

2.2.2.5 Tagmemes and Constructions at the Sentence Level

Tagmemes and constructions at the clause level is easier to work within the initial stages

of investigation and analysis of an unknown (to the investigator) languages than are sentences.

However, sentences are very much important in the structure of all languages, and higher-leveled

structures (paragraph, discourse, and the like) are probably analyzed in terms of sentences (Elson

and Pickett, 1962).

Contrary to Bloomfield and linguists following him who defined a sentence as a “…an

independent linguistic form, not included by virtue of any grammatical construction in larger

linguistic form (1933)”, tagmemics didn’t take that traditional sentence interpretation because it

was not quite adequate for tagmemics since it is conceived that sentences are only one level in a

10
hierarchy that begins with the smallest morpheme groupings and goes above sentence to include

various kinds of larger structures (Elson and Pickett, 1962).

In tagmemics, Pike expounded Bloomfield’s definition saying that “a sentence is by

definition isolatable in its own right…” and “however, (4) isolatability must be defined

structurally. Here we treat it as the potential of an item for constituting an entire utterance…”

(1960).

One good feature of tagmemics when it comes with the sentence level is the inclusion of

the intonation in its formulas because in many languages sentences will be characterized by

certain junctural and intonational features. However, although some linguists believe, as an

article of faith, that this is true of all languages, there is still no certainty of such. Intonational

morphemes which may be found to accompany segmental morphemes to form sentences must be

looked for by the investigator (Elson and Pickett, 1962).

It was then summarized the factors to be taken into consideration in the sentence

definition: (1) it is a level in the grammatical hierarchy above the clause level and below the

paragraph level; (2) sentences are isolable; and (3) in many languages sentences will include

intonational morphemes and juncture features (Elson and Pickett, 1962).

2.2.2.6 Tagmemes and Constructions in the Paragraph Level

It was already mentioned that Robert Longacre had a big contribution to tagmemics in

general. One of it is on the paragraph level. He did not start with morpheme or phoneme and

work on his way up to larger combinations. Instead, he started right in with how clause types fill

functional slots within behavioral/functional genres, i.e. a restricted but high place in the

hierarchy of the language. That may be unsatisfactory to a (linguistic) purist (Rocin, 2001).

11
 The problem with Longacre is with his paragraph types which are not identifiable

morpho-syntactically or by objective linguistic means. One has to read and understand the

paragraphs before one can label them. But then, if we have already read and understood, the

taxonomy has become superfluous. This level is often skipped (Rocine, 2001).

2.2.3 The Grammatical Hierarchy

The grammatical hierarchy has already been mentioned throughout this paper. It is

important to look at it since it is very much important in the study of tagmemics. It is said that in

every language there will be a hierarchical structure within the grammar of the language to

which the term grammatical hierarchy can be applied (Elson and Pickett, 1962). Units at one

level are always built out of the units of the level immediately below it. However, the

grammatical hierarchy, at least in many languages, is not so happily arranged. There are many

instances of skips and loopbacks where units of a higher level are embedded within units of a

lower level, or within units at the same level such as clause within a clause, or sentences filling a

slot on the clause level (Elson and Pickett, 1962).

Definitions of constructions are made to be as general as possible in order to be applied to

most languages. What is a good thing about tagmemics is that it does not limit itself with the

levels mentioned in the preceding sections. It may sometimes be convenient to set up still other

levels to account for all the facts of the grammar of the language.

3. Conclusion

Unlike the work of Noam Chomsky and other transformalists, Pike’s work in tagmemic

theory has not, in general, been widely accepted by scholars. Many linguistic theories are

satisfied with structure alone but there are concepts which cannot be fully understand by that

12
alone. Nevertheless, individual concepts within tagmemics have been adopted and adapted by a

variety of disciplines.

Tagmemics has been referred to by Kissell (2005) as “the linguistics theory of

everything” mainly because it has a remarkably wide range of applications, expecially in but not

limited to, linguistics.

Austin Hale confesses that "it is at present quite possible to be a tagmemicist in good

standing without subscribing to any particular doctrine regarding the form of grammar. To one

who received a good portion of his linguistic upbringing within the tradition of transformational

generative grammar, this realization comes as a shock and a revelation." Though popularly

categorized as a ''slot- grammar," Pike's peculiar insights into the nature of language and

behavior are compatible with and not in opposition to the insights of other schemes and systems.

In his book Language in Relaton to a Unified Theory of the Structure of Human

Behaviour, Pike claimed that the same kinds of structures, rules and procedures found in

phonology apply not only to grammar and discourse, but in fact to all of human behavior

(Kissell, 2005).

Most linguists seemed to feel that as a descriptive model, tagmemics was not as rigorous

or objective as other models, so it didn’t lend itself well to serious scientific inquiry. But as

according to Pike, looking at a language as an objective formal system was missing the point;

behavior that is fundamentally subjective can only be understood and described meaningfully if

the observer allows context to play a role at every level.

In general, tagmemics is very useful and applicable in a number of different areas, not

just linguistics. It can be applied to rhetoric, poetry, science fiction, and philosophy, among other

fields. Others have taken it further, using tagmemics as a model for learning the programming

13
language Perl. Although it never did (and never will) meet the day-to-day needs of most

linguists, the theory has managed to maintain a small but loyal following among researchers of

various fields. Tagmemics is in fact a good way of thinking about what it means to be human as

its key insights—that context is essential, behavior involves overlapping hierarchies, and

viewpoint affects one’s data analysis—turn out to be surprisingly effective for understanding

many kinds of phenoma despite some problems that occur with it (Kissell, 2005).

14
References
Longacre, Robert. (1968). Grammar discovery procedures: A field manual. The Hague, The
Netherlands: Mouton & Co., Publishers.
Platt, John. (1971). In S.C Dik and J.G. Kooij, eds., Grammatical Form and Grammatical
Meaning: A Tagmemic view of Fillmore’s Deep Structure Case concepts. Amsterdam:
North-Holland Publishing Company.
Elson, B. & Pickett, V. (1962). An introduction to morphology and syntax. California: Summer
Institute of Linguistics.
Edwards, B. (2000). The Development of Kenneth L. Pike’s Tagmemic Theory. The Tagmemics
contribution to composition teaching. Retrieved on February 8, 2010
http://personal.bgsu.edu/~edwards/tag3.html
______________. (N/A). Paper 6 (Descriptive linguistics) tagmemics. Retrieved on March 30,
2010 from http://jilaniwarsi.tripod.com/tag.pdf
Edwards, B. (2000). The Place and Promise of Tagmemic Theory in Composition. The
Tagmemics contribution to composition teaching. Retrieved on February 8, 2010
http://personal.bgsu.edu/~edwards/tag3.html
Edwards, B. (2000). Tagmemics discourse theory. Retrieved on February 8, 2010 from Bowling
Green State University Website: http://personal.bgsu.edu/~edwards/tags.html
Kissell, J. (2005). Tagmemics: The linguistic theory of everything. Retrieved on February 8, 2010
from http://www.ziki.com/en/joekissell+20896/post/tagmemics-the-linguistic-theory-of-
everything+9697624
Ross, M. (2010). Tagmeme. Retrived on March 30, 2010 from https://tagmeme.com/?page_id=2
Rocine, B. (2001). Tagmemics, definition for criticism. Retrieved on March 15, 2010 from
http://lists.ibiblio.org/pipermail/b-hebrew/2001-January/009415.html

15