Contributions of Dr. Thorndike to Lexicography
by Clarence L. Barnhart - 1949
The more significant of Dr. Thorndike's contributions to lexicography are described in this article. Thanks to his influence, all school dictionaries now have readable type.
SELDOM does a scholar have the privilege Dr. Thorndike had of seeing the ideas expressed in a single lecture of his, "The Psychology of the School Dictionary,"1 so widely adopted within his own lifetime. The principles of dictionary-making set forth in that lecture which he gave to classes in educational psychology for twenty years at Teachers College have not only been embodied in his own school dictionaries but have been imitated in other and more advanced dictionaries. His own three-book dictionary series (Junior, 1935; Senior, 1941; Revised Junior, 1942; Beginning, 1945),2 the most widely used school dictionary series in this country today, is being used by literally millions of children, and has even spread to other countries (Thorndike Junior, British edition of the Junior, 1947; Thorndike English, British edition of the Senior, 1948).3 One of the current advanced dictionaries, the American College Dictionary,4 is largely based, in its psychological aspects at least, upon principles learned by the editor from Dr. Thorndike. Solely upon the basis of widespread acceptance of his principles, Dr. Thorndike is one of the foremost lexicographers of his time; within fifteen years he prepared three dictionaries for the school market. He was the first lexicographer to apply the principles of the psychology of learning to, and use statistical methods in, the making of dictionaries. Through his original methods and his own great ability as a thinker he devised many new solutions to age-old problems of dictionary-making.
His first dictionary, the Thorndike-Century Junior Dictionary, published in 1935, after almost ten years of work on his part, differed radically from all other school dictionaries in the selection, organization, and presentation of facts. The professed purpose of a school dictionary is to supply the meaning, pronunciation, or spelling of a word when a pupil wants it, and Dr. Thorndike sought not only to supply the facts wanted, but also to consider the probable knowledge and ability that the child seeking these facts would bring to the dictionary. For example, a fourth- or fifth-grade pupil looking up cow would be helped little by the definition, "the mature female of any bovine animal, or of any other animal the male of which is called bull" because little consideration is here given by the definer to understandability of the definition to the child seeking information. A fourth-grade pupil would understand better these definitions of Dr. Thorndike's: "1. the large animal that furnishes us with milk, butter, and cheese. 2. the female of various animals, as, a buffalo cow, an elephant cow, a whale cow." As can be seen from these contrasting definitions, the Junior Dictionary was an exemplification of Dr. Thorndike's philosophy that an instrument of instruction should be cast in a form which satisfied the learner, and that every effort should be made to give the truth in a form which the child would understand. He was never willing, however, to sacrifice the truth in his zeal to produce something that was merely teachable. By separating the common definition from the technical one under conjo, by providing a picture for the first definition of cow so that the child would not mistake the animal described for a goat, by simplifying greatly the technical language in definition 2 and giving examples, he put every fact that the child was able to absorb before him and thus avoided the false simplifications of definitions which simplify merely by omitting essential ideas. This process of sorting out essential facts and presenting them in a teachable order is one of his unique contributions to lexicography that is even today imperfectly understood by the makers of school dictionaries.
The size of school dictionaries is, to a large extent, governed by the price that the school public is willing to pay for them. Despite the fact that school dictionaries use about the same amount of material they vary markedly in the quality of their content. Before the appearance of Dr. Thorndike's Junior Dictionary, editors of school dictionaries had vied with editors of small dictionaries for the general market in an attempt to include as many entries (words to be defined) as possible. If the editor is allowed 500,000 words in which to explain 30,000 main entries and 60,000 meanings, it is obvious that he can spend on the average only seven or eight words on a single meaning (space must be allowed for the entry, inflected forms, pronunciations, etc.). Counts of children's reading show that children in grades three to eight will probably encounter some 60,000 different words in their reading.5 If the editor adequately explains frequently occurring hard words, there is little or no space left in which to cover the rarer words.
Thorndike solved the problem of selection of entries by utilizing his word count of the 20,000 words occurring most frequently in 10 million running words in 279 sources.6 His 20,000 count was a count of all types of literatureadult, commercial, textbooks, children's reading. The frequency numbers were, therefore, fairly reliable indexes to the probable occurrence of a given word in any type of reading the child would do before senior high school. With the word count as a guide, he was able to exclude rare adult and technical terms such as fabulist, fabaceous, and fauces and devote more space to the explanation of such hard words as -fabulous, faculty, and festivity, which should become a part of the working vocabulary of every person. All grade-school dictionaries published since the appearance of the Thorndike-Century Junior Dictionary have abandoned their previous policy of including as many words as possible and have instead adopted this principle of devoting more space to those basic words which are destined to be in everyone's working vocabulary and should be thoroughly understood.
The difficulty or rareness of a word more or less predetermines what type of child will look it up in his dictionary. For example, a child who will look up be, for, little, spoon, etc., will have difficulty in understanding formal definitions; on the other hand, children who encounter such words as factitious, feminine, feudatory, etc., in their reading are capable of understanding more formal definitions, but even these children will be greatly helped if illustrative sentences are freely used. Dr. Thorndike is the only lexicographer who has varied the treatment of words according to the probable need and maturity of the child. For very simple, axiomatic words he gives only a picture (e.g., spoon) or only illustrative sentences (e.g., be), for more advanced words he combines a definition and a picture (e.g., spur) or uses a definition plus a sentence (e.g., sputter); for harder words that only more mature pupils will look up (e.g., spellbound, specification), he is often content with a definition.
Probably Dr. Thorndike's greatest contribution to school dictionaries was freeing them of an unreal (to most children) system of classifying meanings. The definitions in the Junior Dictionary are organized according to related meanings because we naturally associate like facts. Consider the advantage to a pupil who seeks to learn the meanings of club of having together all of them that relate to a stick, followed by all that relate to a group of persons:
1. a heavy stick of wood, thicker at one end, used as a weapon. 2. beat with a club or something similar. 3. a stick or bat used in some games played with a ball, as golf-clubs. 4. group of people joined together for some special purpose, as, a social club, tennis club, yacht club, nature-study club. 5. the building or rooms used by a club. 6. join together for some purpose. The children clubbed together to buy their mother a plant for her birthday.7
With such an arrangement of definitions, reading one definition helps in reading and understanding another.
One of Dr. Thorndike's principles of dictionary-making was rigorously to avoid defining a word in terms of more difficult words. It is curious that this common principle of teaching should never have been employed in school dictionaries until he used it in his Junior Dictionary. Now it is a widespread practice. Consider his definition of fable: "story that is not true." Dr. Thorndike's avoidance of hard words in definitions greatly influenced modern school dictionary-making.
Another of Dr. Thorndike's principles that has found wide acceptance is his use of illustrative phrases to clarify meanings or to force the meanings so that the child cannot fail to understand them. Consider his treatment (in the Junior Dictionary} of facilitate: "make easy; lessen the labor of; forward; assist. A vacuum-cleaner facilitates housework."
As Dr. Thorndike points out in his lecture, the construction of illustrative sentences for children is not an easy task. "It requires not only a sound psychology, but also great inventive skill, to find or frame sentences that lead pupils inevitably to the right meaning for a word."8 He used the illustrative sentence to teach the meaning. Consider this illustrative sentence for the meaning "symbol or sign" of badge:
"Chains are a badge of slavery." Or this illustrative example of commercialize: "To charge admission to church services would be to commercialize religion." Or this one of affectation: "Helen's roughness is an affectation; she really is a quiet, gentle girl." Such sentences inevitably lead children to the right meaning of the word. That they take great skill to construct is evident, since it requires keen psychological insight to write sentences that force the meaning of a word home to the pupil.
The same fresh approach to dictionary problems is apparent in the following statement by Dr. Thorndike of his aims in making the Senior Dictionary. In an inter-office memorandum dated May 1, 1935, he wrote:
This dictionary does for boys and girls in their teens what the Thorndike-Century Junior Dictionary does for children. The fundamental principles of intelligibility, interest, and help to the learner are the same in both. This book is not merely or chiefly a repository or museum of facts about words, but a specific tool to provide aid to high-school students whenever they need to know the meaning, spelling, or pronunciation of a word or phrase.
The utmost care has been taken to make every item of information in it accurate, but it is not written to please scholars, certainly not to please pedants, but to help high-school pupils, especially to help them to understand what they hear and read, and to learn facts about words which will improve their ability to understand and use English. A new word or phrase encountered is a problem; a dictionary should help a person solve the problems which that person will meet.
I have tried in the case of every word to consider the individuals who have difficulty with that word, and the nature and circumstances of the difficulty in each case. The treatment of each word is the best solution I can give to the psychology and pedagogy of that problem in language. I have often slipped and blundered, but nothing is put in this book perfunctorily or indiscriminately; every line of every page is intended to serve some purpose of some learner, and the book as a whole is intended to do the best possible for those who will use it.
In realizing these aims of providing information about the real word problems of the high-school student, Dr. Thorndike improved his earlier method of selecting entries by using the Thorn-dike-Lorge count of the commonest 30,000 words in a count of 18 million words.9 By extending the range and number of words counted over his previous count of 20,000 words and by listing the exact number of occurrences in the four counts (Thorndike 20,000, Lorge Magazine, Juvenile, and the Semantic), Dr. Thorndike was able to make sure that any necessary word encountered by high-school students would, in all probability, be in the Senior Dictionary. The Thorndike-Lorge 30,000 word list proved to be a necessary tool for the dictionary editor in the selection of entries, and now even advanced school dictionaries consider the word counts in selecting the entries to be included instead of depending on personal judgment as much as they formerly did.
Dr. Thorndike's most outstanding psychological contribution to lexicography in the Senior Dictionary was, however, his use of the Lorge-Thorndike Semantic Count. This count is not merely a convenience for the dictionary-maker; it is highly probable that no satisfactory school dictionary could be made today without its use. Space is at such a premium that the inclusion of useless information forces abbreviated treatment of essential information. When one considers that every meaning in approximately 5 million running words in 29 sources such as the Encyclopaedia Britannica, The Literary Digest, novels, textbooks, and similar material has been counted and a frequency assigned to it, and the frequencies keyed to The Oxford English Dictionary, the value of the count as a guide to the editor in selecting meanings, allotting the amount of space for definitions and illustrative sentences, and arrangement of the definitions is self-evident.
Consider the relatively simple problem of selecting the definitions for amenable to be included in a high-school dictionary. There are five definitions of this word in the Oxford English Dictionary. Which should be included in a high-school dictionary? By referring to the Lorge-Thorndike Semantic Count, we find that amenable occurred ten times in 5 million running words (2 per million) and that only two of the possible five meanings in the OED occurred at all. The first of these two meanings, "accountable," occurred 200 thousandths of the time and the second, "open to advice," occurred 800 thousandths. If frequency is to be the criterion for the order of definitions, the second one in the OED, "open to advice," should be put first. When the editor comes to the harder problem of amend, with 29 definitions in the OED (12 of which have frequency), or approach, with 29 (20 of which have frequency), or set with 544 meanings (152 of which have frequency, but only 27 of which have a frequency of ten per thousand or over), he will find the English Semantic Count a necessary tool. By the use of the Semantic Count, Dr. Thorndike was able, for the first time, to make an adequate selection of the meanings included in a dictionary on the basis of frequency and probable occurrence. His policy of using the Semantic Count was later extended to the Junior and Beginning dictionaries, and was used by the editor of The American College Dictionary.
A mark of Dr. Thorndike's greatness was the invitation he extended to twenty-eight outstanding scholars10 in various fields of language study and teaching to survey the linguistic and pedagogical problems involved in making a modern dictionary. The writer of this article served as the secretary of this committee during the five years of work on the Senior Dictionary and can testify to the hard work put in by the committee and Dr. Thorndike in considering the diverse problems of dictionary-making. One of the most important problems considered was that of pronunciationa problem that had been side-stepped in school and general dictionaries. In spite of the growing body of linguistic studies which indicated that dictionaries were perpetuating false information about pronunciation instead of giving accurate and reliable information based upon the research of scientists in the field, there had been no real improvement in methods of recording facts about pronunciation in dictionaries for fifty years.
Failure of the older dictionaries to utilize modern linguistic scholarship shows itself unmistakably in the kind of pronunciation key they used. Most English dictionaries add marks to the ordinary spelling to indicate the pronunciation. Since there are some 250 spellings for the 40 significant sounds in English, some standards for the selection of necessary diacritics and for pronunciation generally are necessary. The committee decided to ignore the old spelling keys and to have one symbol for each significant sound, but to adapt traditional diacritical methods instead of adopting a non-diacritical method based on non-English speech habits. In other words, the committee sought to put the current teaching practice, in the schools upon a scientific basis. The committee reduced the number of symbols from 60 to 43, reduced the number of diacritics to 12, and made it possible, by the use of the schwa, a symbol in the International Phonetic Alphabet used for a certain grade of unstressed vowels, to have a key which forced the user to consider the pronunciation instead of the spelling.
The pronunciation committee working with Dr. Thorndike set up the following standards: (1) one symbol for each significant sound; (2) symbols were based upon natural English speech habits (i would represent the sound of the i in hit not the i in machine), (3) a minimum number of diacritics would be used (since the so-called short sounds in bat, bet, bit, bot, but are approximately 40 per cent of all sounds and occur in closed syllables they were left unmarked); (4) the symbols must be easily read and understood; (5) the symbols must be easy to reproduce in script (no italic symbols). Using these standards, at the advice and with the consent of Dr. Thorndike, the committee was able to make the first improvement in school dictionary keys since the nineteenth century.
In his last dictionary (Beginning, 1945), Dr. Thorndike made his final contribution toward establishing the dictionary as a tool in the classroom instead of as an occasional reference book. He greatly increased the size of type (as he had advocated in his original lecture), covered a smaller but more usable list of words and meanings, omitted grammatical material entirely, and thus made a book which gave answers to the third- and fourth-grade child with regard to the pronunciation, spelling, or meaning of a word. This in itself would have been a major contribution, but he went farther and inserted seventy pages of lessons which taught pupils how to look up words, how to find meanings, read them, and fit them into contexts, and how to use a pronunciation key. Instead of depending upon occasional use of the dictionary workbook or the unrelated instruction in a speller, Dr. Thorndike put in his Beginning Dictionary the exact procedure by which any third-or fourth-grader, with or without the help of his teacher, can learn how to use the most valuable tool in the schoolroom, a dictionary, in finding the meaning, learning the pronunciation, or discovering the spelling of a new word.
Only the more significant of Dr. Thorndike's contributions to lexicography have been mentioned in this article. Thanks to his influence, all school dictionaries now have readable type. His special study of affixes11 was the first examination of the teaching of the meanings of affixes and was as important in its way as any of his counts. He made the picture or diagram an integral part of the definition instead of an occasional ornament. But these contributions of general ideas and improvements in dictionaries, however great and however valuable, were not so important as that unflagging industry which led him to prepare three manuscripts on the Junior, and to solve personally 50,000 word problems put to him by the editor of the Junior. Such mastery of general principles and such care in carrying out his ideas were an inspiration to all who worked with him in making his dictionary series. He realized his fundamental purpose of providing a dictionary series for School children the motto of which might well be "Truth, and truth in a form which fits the learner.12 In the process of making his own series he provided new techniques of dictionary-making and improved old ones.
1 Edward L. Thorndike, "The Psychology of the School Dictionary." Bulletin of the School of Education, Indiana University, Vol. 4, pp. 24-31, Bloomington, Ind., July, 1928.
2 The Thorndike-Century Junior Dictionary (1935); Thorndike-Century Senior Dictionary (1941); Thorndike-Century Junior Dictionary, Revised Edition (1942); Thorndike Century Beginning Dictionary (1945), Scott, Foresman and Company, Chicago.
3 The Thorndike Junior Dictionary. Revised and edited by P. B. Ballard and H. E. Palmer. University of London Press, Ltd., London, 1947; Thorndike English Dictionary. The English Universities Press Ltd., London, 1948.
4 Clarence L. Barnhart, editor, The American College Dictionary. Random House, New York, 1947.
5 Edward L. Thorndike, "The Vocabulary of Books for Children in Grades 3 to 8." Teachers College Record, Vol. 38: Part I, pp. 196-205, December, 1936; Part II, pp. 316-23, January, 1937; Part III, pp. 416-29, February, 1937.
6 Edward L. Thorndike, A Teacher's Word Book of the Twenty Thousand Words Found Most Frequently and Widely in General Reading for Children and Young People. Bureau of Publications, Teachers College, Columbia University, New York, 1931.
7 The Thorndike-Century Junior Dictionary, p. 150.
8 "The Psychology of the School Dictionary." Bulletin of the School of Education, Indiana University, Vol. 4, p. 30.
9 Edward L. Thorndike and Irving Lorge, The Teacher's Word Book of 30,000 Words. Bureau of Publications, Teachers College, Columbia University, New York, 1944.
10 The members of the committee were: Sir William Craigie, Miles L. Hanley, Albert C. Baugh, Leonard Bloomfield, Francis M. Crowley, Charles C. Fries, W. Cabell Greet, Archibald A. Hill, R-M. S. Heffner, Lee S. Hultzen, Arthur G. Kennedy, Hans Kurath, George S. Lane, William F. Luebke, Kemp Malone, George H. McKnight, C. E. Parmenter, Charles S. Pendleton, Robert C. Pooley, Louise Pound, Robert L. Ramsay, William A. Read, Edward Sapir, C. K. Thomas, George Watson, C. M. Wise, Jane Dorsey Zimmerman, and George K. Zipf.
11 The Teaching of English Suffixes. Contributions to Education, No. 847. Bureau of Publications, Teachers College, Columbia University, New York, 1941.
12 Edward L. Thorndike, "The Psychology of the School Dictionary." Bulletin of the School of Education, Indiana University. Vol. 4, p. 31.