E-Prime: speaking crisply.

1. Introduction

In the fall of 1969 the New York Society for General Semantics invited me to present a talk on E-Prime (English without any form of the verb to be). (1) Of course, it occurred to me that I should give it in E-Prime. In other words, to speak in E-Prime. I had initially planned to tell the people attending the presentation at the beginning that I intended to try this. I regret that I have to admit this, but I chickened out about alerting the audience as to m intentions, although I did make the effort to give my talk in E-Prime. Naturally, the new York Society folks listened for precisely this kind of attempt, without needing any warning from me. I have believed from the beginning that any reasonably alert individual can write in E-Prime without any big fanfare. (Mini-lesson: In order to write in E-Prime, just do not employ any form of the verb to be, and avoid "is-English" sentence structures to prevent an excessive use of the various linking verbs.) But I recognize that speaking in E-Prime requires a much higher level of linguistic sensitivity, determination, etc. In addition to other problems, what I have called "Social Is-isness" can rise up and repeatedly slap one in the chops. This means that standard social formulas (e.g., "What day is today?" "How are you?" "This is Ed," "What time is it?" etc.) will keep those non-E-Prime grooves, or neural networks, etc. functioning in the cortex when trying to speak in E-Prime. If you need it, please check reference (2) for the epistemological reasons behind this E-Prime business.

Almost twenty years later, my now friend and colleague Dr. E. W. Kellogg III presented a delightful and well-grounded discussion of his experiences in learning to speak in E-Prime, with suggestions that could help other who wished to accomplish this. (3) At present, very few people have chosen to try to speak routinely, daily, in E-Prime. I would guess perhaps five or six but this may consist of an excessively pessimistic estimate. For those wishing to try seriously to write and speak in E-Prime, I also suggest a careful study of references (4) and (5).

Make no mistake: we deal here with interesting, difficult, and important issues. Since Hobbes in 1651 (relative to English), and even earlier, since Lycophron in 370? B.C.E. (relative to greek), extremely clever, intelligent, productive, and creative people have had volumes to say about why we should stop placing so much reliance on the static verb to be in our dynamic universe. (See Note 1.) But why hasn't anyone other than the writer asked how frequently do we in fact employ this increasingly disreputable verb in our writings? How frequently in our oral speech? I understand that Aristotle refused to count the teeth in people's mouths, and therefore continued to his death to believe erroneously that men have more teeth than women. I see this attitude reflected in the fact so few seem to have any interest in how often we tend to use to be. Let us now pursue this intriguing problem, empirically. (See Note 2.)

2. The Matter of Crispness

In the initial paper describing E-Prime (E'), I included the following definitional semantic equation:

E' = E - e

where E represents the one to two million "words" (depending on how one chooses to define "word") of English, and e represents the conjugated forms of to be. (1) Despite the fact that the late Dr. Samuel Bois thought highly enough of E-Prime to mention it in his book, The Art of Awareness (6), and even invited me to share his podium to discuss this development at a General Semantics conference in 1965, on one unfortunate occasion the simple equation given above appeared in print incorrectly as:

E' [not equal to] E - e

which probably mystified some readers. I suspect that most of those readers held me responsible. I understand that subsequent editions of that book will have the equation corrected.

Some years later I defined the Crispness Index (C.I.) for a piece of writing as:

C.I. = Number of E-Prime Sentences / Total Number of Sentences

where "E-Prime Sentences" means sentences without any form of the verb to be. (8)

It has now become necessary to generalize the definition given just above for the Crispness Index to reflect the fact that frequently in plays and in some other written material, and certainly in oral speech, an analyst will encounter what one might call "sentence fragments." Consequently, let us now define the Utterance Crispness Index (U.C.I.) for a sample more generally in terms of utterances:

U.C.I. = Number of E-Prime Utterances/Total Number of Utterances

We have had the term "utterance" with us, referring to language processes, since at least the 1400s C.E., but as a technical term in linguistics it dates from what we now know as the Bloomfieldian paradigm. A. H. Gardner, writing slightly before Bloomfield produced his hauptwerk (10), said in 1932: "Under the term `utterance' writing must be included." (11) And then the great structuralist, Zellig S. Harris, stated in 1951 that "An utterance is any stretch of talk by one person, before and after which there is a silence on the part of the person. The `utterance' is, in general, not identical with the `sentence.'" (12) Although I personally reject both the Bloomfieldian and the Chomskian paradigms for linguistics, for reasons explicated at length in reference (13), I feel that we should accept a combination of the Gardner and Harris definitions for "utterance" in our studies of the "crispness" of written and spoken material.

I beg the indulgence of the non-specialists, but we linguists usually try to give careful definitions, if possible. In this enquiry into the crispness of oral speech, and with respect to other investigations into written speech, I will use the term "utterance" to label a specific written or oral emission, normally following and preceding a silence, occasionally but not necessarily congruent with what linguists call a "clause," and what we regard popularly as a "sentence"; if two people speak, then they will usually generate pairs of utterances, etc. (See Note 3.) As we all know, when more than one person talks, the analytical miseries of overlap and interruption will inevitably occur. We have to do the best we can in studying the stream of speech data when that happens. I believe that in this area one must adhere to a brutally naive empirical approach, without trying to dismiss such analytical embarrassments as "uh," laughter, etc., as beneath our attention. If you wish to argue with me that, "Oh, no. In the deep structure this really involves ..." I must reply that your philosophical/linguistic orientation has much interest for me as well, but I doubt that such considerations have anything to do with analyzing "surface" data concerning how people speak. Or write. Or "think." A number of linguists will not agree with me, and I have a message for them: "You (pl.) defend your position; I have stated mine."

3. The Availability of Transcribed Oral Speech

When I called for studies of the crispness of oral speech in an introduction to the second E-Prime anthology (14), I had no idea of the existence of a profusion of data that students of communications science had collected for their own purposes. (See Note 4.) I certainly had no idea that I would have an opportunity to introduce this study to you at this comparatively early date.

The most readily available set of examples of oral speech appears in The Talk Book, by Goodman and Esterly. (15) They provide a large number of brief transcripts of "sessions," somewhat edited to smooth up the material, along with analyses to obtain guidelines to help people communicate better. Just how much smoothing took place became apparent when I studied transcripts of analogous conversations compiled by and available from the University of Texas Conversation Library plus additional material generously made available to me by Professor H. Paul LeBlanc III of the Southern Illinois University at Carbondale. I now have acquired a veritable deluge of data, the analysis of which proceeds very slowly. Before reporting on the preliminary findings, I will describe the crispness results I developed for written material as background.

4. The Crispness of Writing

Over the years I have studied the frequency of appearance of forms of the verb to be in a rather large variety of written contexts. (8, 16) Four books on general semantics and philosophical matters showed Crispness Indexes ranging from Korzybski's 0.41 to Ayn Rand's 0.28. Five political documents had Crispness Indexes that ranged from the Communist Manifesto's 0.55 to Aristotle's Politics with 0.18. Four articles in one specific issue of ETC: A Review of General Semantics (Vol. 49, No. 2, 1992) had Crispness Indexes spanning the large gap from Pula's 0.72 to Perlman's 0.33. Somewhat analogously, four essentially (modern) political essays that appeared in The New Yorker and Time Magazine (two each, in issues printed in November, 1992) ranged from Michael Arlen's 0.60 to John Updike's 0.27. You may check the detaIls behind those summary statements, if interested, in reference (8). Among the seventeen writers who produced those books and articles (leaving aside the issue of two multi-author works involved), only Korzybski and Pula had any particular concern about avoiding forms of the verb to be, so far as I know. And sure enough, both came through outstandingly within their groups. Recall that of course Korzybski knew nothing about E-Prime, and even opined m seminars (incorrectly, in my view) that one could not speak or write English well without using that verb; Pula has never embraced E-Prime, although I understand that he has made the charming remark that, "E-Prime is [pause] no crime!" Both evidently operated importantly on the basis of their linguistic sensitivity, consciousness of abstracting, etc. In summary, those nonfiction writings that I studied have associated Crispness Indexes that varied from 0.18 to 0.72.

Another set of examples, which actually produced even more interesting results, came from a study of fiction and a small number of movie reviews. These included, firstly, five novels by Ernest Hemingway (The Sun Also Rises, A Farewell to Arms, For Whom the Bell Tolls, Across the River and Into the Trees, and The Old Man and the Sea), spanning his output from 1926 to 1952. Then I studied books by two authors whom I regard as extremely far from Hemingway in the content of their work: a book of three short (actually "medium" length) stories by Eudora Welty, and Margaret Mitchell's Gone With the Wind. Seeking an example of "bad" writing, I sought to obtain something written by the much-criticized Edward Bulwer-Lytton, and tried to track down some of his large output in a sizable bookstore in New York City, without success. Having come up dry, Bulwer-Lytton-wise, I then asked the manager to tell me what single book in his store he considered the most poorly written (with no further guidance from me about what "poorly written" might "mean"). He immediately stated, "Scarlet"," Alexandra Ripley's sequel to the previously cited Gone With the Wind.

Hemingway's novels had an average Crispness Index of 0.64, while all the rest of the material cited above had Crispness Indexes between 0.54 and 0.57, including the defamed Scarlett (0.56). I will not try to spin a tale about the statistical significance of these differences, or what those results mean relative to literary merit (however one may define that slippery issue). However, I do believe, on the basis of even my comparatively small sample, one can say that professional writers of fiction, whose work has received the attention of professional editors, tend to use a form of the verb to be in about half of their sentences. This strikes me as a truly huge, and clearly unnecessary, reliance on this static verb. Such repetition of any other word or verb with its inflections in the English language would surely result in the writer's material becoming summarily rejected. Of course, Fenollosa tried to warn writers away from to be 75 years ago, but few paid any attention to him. (19)

5. The Crispness of Speaking: Some Historical Preliminaries

During the 1960s I dropped in on a Seminar sponsored by the Institute of General Semantics and held at C. W. Post College to give a lecture on operations research. While there, I heard part of a presentation given by Dr. Ray Bontrager. As I arrived, he introduced the matter of the fantastic visual representations produced by electron microscopes, with examples provided. Talk about a powerful device for abstracting! Bontrager had a fascinating point: since the production of the first commercially constructed electron microscope (in England in 1935), humans have actually looked at an extremely small part of "the world" using this highest of precision tools. At the present time, received wisdom has it that "the" electron microscope has the capability of a point-to-point resolution of less than three Angstrom units. By definition, an Angstrom unit equals [10.sup.-7] millimeter.

The highly esteemed structural linguist, Professor C. F. Hockett, giving himself plenty of room to maneuver, has estimated that human language(s) began millions of years ago. (20; p. 581) In contrast with that lengthy oral speech tradition, writing systems only stem back to ancient Egypt, ancient China, the Maya of Yucatan, and (perhaps subject to contamination by Egyptian influences) ancient Mesopotamia. The writing systems that came from the ancient Egyptians and the ancient Mesopotamians date from about 3100 B.C.E., and had the greatest influence on subsequent writing. (21; p. 989)

Printing, of course, has an almost evanescent history, compared with "language" and "writing systems." If I may remind you, printing began, in the Western World, in the middle 1400s with Gutenberg's moveable type version of the Bible. Four hundred and fifty years later, by the end of the 1800s, very few homes indeed had any book other than the Bible. Perhaps also a hymn-book. In many homes all over the world, this still remains the case. However, the Twentieth Century has witnessed, among other excellent and not-so-excellent developments, a great explosion in at least the availability of books. As a known book-lover, I regret that the outlook appears extremely dim for books in the next century. On the one side, we have heard much about the chemical dissolution of many books of the past. On another side, the computer monitor and the TV screen seem to have taken over many functions previously associated with books. Ask any group of non-college-attending teenagers about the books and newspapers they have read recently, and you will probably join me in gloomily recognizing that we will soon see the end of The Century of the Book. What a topic for a Ph.D. dissertation in History! In Communications Science!

6. Speaking Crisply: An Early Look at Empirical Data

Those bleak preliminaries obviously direct our attention most particularly to the importance of oral speech, and away from that increasingly fuddy-duddy old irrelevance, written stuff. The results presented below amount to an electron microscopic sample of the oceans of speech that have poured out, and continue to pour out. But hey! We have to begin somewhere.

The data consist of 19 samples of interchanges between two people, speaking on a variety of (mainly personal) topics. (See Note 5.) Six additional samples involved three or more people speaking. The latter samples produced results quite similar to the two-person material: other than that observation I have put them aside, for surely some very complicated issues could arise in such confrontations. Let us confine this discussion to the comparatively simple two-person situations. Most of those 19 samples may appear rather brief. One involved 325 utterances, another 207, still another 114, and a fourth 106. The rest ranged from 15 to 88.

This obviously electron-microscope-sized look at oral speech cannot tell us much, other than showing us a few tendencies. The following three seem the soundest:

(a) As the number of utterances increased, the two speakers tended to speak more nearly equally crisply. We may see here an instance of the assertion that, "The terminology of the question determines the terminology of the answer."

(b) A slight indication exists that, as the number of utterances in an exchange increased, the Utterance Crispness Index fell. Perhaps this indicates that the increasing number of utterances at least slightly minimized the effects of stammers, laughs, etc.

(c) Despite the bias toward higher values for the Utterance Crispness Index of oral speech built into its definition, compared with the earlier Crispness Index used for written material, the overall averages for the first speaker equaled 0.61, and for the second speaker equaled 0.63, with a "grand average" of 0.62. Those values do not seem very far from the results for written material. I will not insult the readers' intelligence by computing standard deviations, etc. Such statistical niceties must await more extensive studies.

7. Concluding Remarks

This brief exploration, or foray if you will, into the crispness of speaking has suggested three tendencies that of course require much additional work. At least the material discussed above may point the way for future studies, by me ... or by you.

On the whole, written material seems to exhibit crispness indexes that hover about a figure slightly above 0.50, while oral utterances seem to have indexes only slightly higher. One might very well expect the Utterance Crispness Index for samples to fluctuate with the emotional state of the speakers, their education level, and so on. At the very least, it appears that one can conclude that written and spoken speech seems to contain one or more forms of the verb to be in about half the utterances studied so far. Far too many, some would say.

NOTES

(1.) Stated that way, it seems obvious that we should not employ the static verb to be frequently. If the grammar of a language required it, for example to carry tense, or aspect, or some other detail (which English does not), one might have a case for continuing to use it, or for calling for a revision of that part of the grammar. Korzybski began the serious discussion of the epistemological reasons behind the need to avoid certain usages of this verb, hence really energetic efforts to do so began with him. Various supposed students of Korzybski's have tried to put off decisive action relative to to be, but they primarily just demonstrate their lack of familiarity with Korzybski's writings.

(2.) Of course, I do have some ideas about why such intellectual titans as Hobbes, de Morgan, Bertrand Russell, Santayana, Whitehead, and Korzybski did not undergo the experience of counting how often people do in fact write and speak using the verb to be. I have done related chores since 1947, initially as a consequence of assignments by Dr. George K. Zipf, and extending through the present. Let me assure one and all that this consists of incredibly boring activity. Even when I worked as an Associate Professor, I did not have graduate students that I could trust to conduct empirical counts of whatever. Too bad! Right now, as I type this, I should really go to another room and conduct some more counts of how some poor devil spoke, with and without the verb to be. Damn!

(3.) Linguists (e.g., Pike (17) and even Bourland (18)), when working within the structuralist paradigm, have recognized the necessity to differentiate between clauses and sentences. A straightforward definition in terms of that paradigm consists of the following: a sentence must have at least one independent clause plus an intonation contour.

(4.) Thanks to my friend, C. Jay Frasier, I now have encountered a wealth of such data. An embarrassing wealth of such data. I would probably never have gotten as far as I have with this project without the assistance of my wife, Elizabeth J. Bourland, who provided me with the majority of the raw counts used in this paper. I conducted the analysis, and hence I have the responsibility for any errors of that nature. I feel happy to express here my gratitude to both Jay and Elizabeth.

(5.) Data Sources: Reference (15; p. 19-28), plus the following sessions from the University of Texas Conversation Library: A10BEADS, A10BROWN, F1COOK, TRIP TO SYRACUSE, MALE/FEMALE ARGUMENT, TRADITIONAL COUPLE #48, JAY & ANA, SUZ & STN, M/F, FAMILY LTBJ, EARRINGS, FAM 004, TOM/ANN, TOM/BILL, D8 SEDUCT, JZ TEAM, DO8 QUEER, FAMILY 001, BKE CONV, COUPLES ARGUING #1, COUPLES ARGUING #2, J&B ARGUING, COUPLES ARGUING EXCPT #1, ARGUING COUPLES DRUNK.

REFERENCES

(1.) D. David Bourland, Jr. "A Linguistic Note: Writing in E-Prime," General Semantics Bulletin, nos. 32/33. 1965.

(2.) D. David Bourland, Jr., and Paul Dennithorne Johnston, eds. To Be or Not: An E-Prime Anthology. San Francisco: International Society for General Semantics.1991.

(3.) E. W. Kellogg, III. "Speaking in E-Prime: An Experimental Method for Integrating General Semantics Into Daily Life," ETC: A Review of General Semantics, vol. 44, no. 2. 1987.

(4.) E. W. Kellogg III and D. David Bourland, Jr. "Working with E-Prime: Some Practical Notes," ETC: A Review of General Semantics, vol. 47, no. 4, 1990. Also printed in reference (2).

(5.) Russell Joyner. "An Auto-Interview on the Need for E-Prime," ETC: A Review of General Semantics, vol. 50, no. 3, 1993. Also printed in reference (9).

(6.) J. Samuel Bois. The Art of Awareness. Dubuque, Iowa: Wm. C. Brown Co. 1966.

(7.) J. Samuel Bois. The Art of Awareness. Dubuque, Iowa: Wm. C. Brown Co. Third edition, 1978.

(8.) D. David Bourland, Jr. "E-Prime and the Crispness Index," published in reference (9).

(9.) Paul Dennithorne Johnston, D. David Bourland, Jr., and Jeremy Klein, eds. More E-Prime: To Be or Not, II. Concord, Calif.: International Society for General Semantics. 1994.

(10.) Leonard Bloomfield. Language. New York: Henry Holt. 1933.

(11.) Sir Alan Henderson Gardnet. The Theory of Speech and Language. Oxford: The Clarendon Press. 1932. Second edition, 1951.

(12.) Zellig S. Harris. Methods in Structural Linguistics. Chicago: University of Chicago Press.1951. Phoenix Books edition published in 1963.

(13.) D. David Bourland, Jr. "A Non-Aristotelian Paradigm for Linguistics," Licenciatura thesis, San Jose, Costa Rica: Universidad de Costa Rica. 1973. Also published in reference (2).

(14.) D. David Bourland, Jr. "[Introduction.sub.2]" to reference (9).

(15.) Gerald Goldman and Glenn Esterly. The Talk Book. New York: Ballantine Books. 1988. Third printing, 1991.

(16.) D. David Bourland, Jr. "The Language of E-Prime," in D. E. Washburn and D. E. Smith, eds., Coping With Increasing Complexity. New York: Gordon & Breach, Science Publishers. 1974.

(17.) Kenneth L. Pike. Language in Relation to a Unified Theory of the Structure of Human Behavior. The Hague: Mouton.1967.

(18.) D. David Bourland, Jr. Introduccion a la Tagmemica. San Jose, Costa Rica: Imprenta Semantica.1978.

(19.) Ernest Fenollosa, "The Chinese Written Character as a Medium for Poetry," in Ezra Pound, ed., Instigations. New York: Boni & Liveright. 1920.

(20.) Charles F. Hockett. A Course in Modern Linguistics. New York: Macmillan.1958.

(21.) Encyclopedia Britannica, vol. 29. Chicago: The University of Chicago. 1986 (First edition, 1768-1771).

D. David Bourland, Jr., a retired Associate Professor of Linguistics, as written on general semantics topics for many years. In 1965 he "invented" E-Prime.

Copyright [C] 1995 by D. David Bourland, Jr.