LexipediaProposal.htm

: Lexipedia Proposal:; Print and Electronic Reference Resources; in the Digital Age
: Robert Parks; November 1, 1997


Contents

A. Introduction: Linguistic and Encyclopedic Knowledge: The Role of Language in the Knowledge and Reference Processes
: 1. Evolution of the Reference Process; 2. The Lexipedia concept; 3. Diagram A: Words and the World: Classification of Language and Encyclopedic Reference Works

B. Roget's Challenge
: 1. Roget and the Thesaurus; 2. The Locus of Linguistic Meaning: From Words to Meanings and Back; a. Meaning in Words and Phrases; b. Meaning Construction is Circular; c. Elaboration.; 3. Diagram B: Networks of Words and Phrases in the Meaning Process.
C. Classification of Language Reference Works
: 1. From Word to Concept and Back; 2. Diagram C: Classification of Language Reference Works; 3. "Word to Concept" Reference: The Dictionary; 4. "Word to Word" Reference: The Thesaurus; 5. "Concept to Word" Reference: (1) Visual Concepts; 6. "Concept to Word" Reference: (2) Verbal Concepts:; a. Classificational Systems of Concepts;; b. Definitional Concepts;; c. Word Relations Concepts;; d. Miscellaneous Concepts; 7. Boundary Cases: Dictionary and Thesaurus; 8. Boundary Cases: Thesaurus and Conceptual Lexicon

D. Proposed Print Reference Works: Synergy Through Integration.: 1. An Integrated Thesaurus-Dictionary; 2. A Dictionary-Reverse Dictionary (or Dictionary with a Conceptual Index); a. Justification.; b. Explanation: (1) Definitional Concepts; (2) Runins; (3) Word Relations Concepts; 3. Thesaurus with a Conceptual Index (or Thesaurus-Reverse Dictionary)

E. An integrated Lexipedia-Encyclopedia Database: 1. User Interfaces; 2. Uses in Text Retrieval; 3. Uses in Print Index Construction; 4. Uses in Hypertext Construction

F. Conclusion

Appendix A: Sample Reverse Dictionary/Conceptual Index Entries




A. Introduction: Linguistic and Encyclopedic Knowledge: The Role of Language in the Knowledge and Reference Processes


1. Evolution of the Reference Process
(back to Contents)

The dictionary is a network of words and meanings, distilled from our knowledge of language into the stable confines of a print reference work. As lexicographers will attest, the construction of dictionaries is a very practical endeavor, constrained by a lack of universally acknowledged methods or principles. An additional and usually decisive constraint is the limitations imposed by the process of print distribution. This affects the design of the work, and, since it must be paid for in a market economy, the justification for publishing most language references. Out of this mix of intellectual ambition and practical constraint has emerged an array of reference works - each seeking to fit as well as to shape a niche in the universe of knowledge.

Dictionaries have taken a somewhat pragmatic approach to encyclopedic information. They often contain biographical and geographical information, as well as other tables and charts. These help us not merely to know what a word means, but also to know when, where and how it is applied in the knowledge process.

The world of reference is evolving under the impact of new media of distribution, however. Changes in the constraints of space and principles of access are leading to new practices and new possibilities of in the development of reference materials. Dictionaries are evolving:

1. from linear alphabetical reference, to hyper-media linking;
2. from word-to-meaning reference to interactive meaning-to-word
reference;
3. from separation of linguistic and encyclopedic information, to
interaction of semantic and real world knowledge;

The proposal developed here is to use The Wordsmyth English Dictionary-Thesaurus (WEDT) as the foundation in constructing a Lexipedia Database - a more explicit semantic network than is currently built into the design of dictionaries. The Lexipedia will allow the construction of (1) innovative hybrid print reference works; (2) unify the process of storing, developing and modifying print works; and (3) serve as a core technology in the development of intelligent access to electronic text of all kinds. Three print works are proposed here, but many spin-offs are possible (See Appendix B). The proposed print works are:

1. Wordsmyth Thesaurus-Dictionary
2. Wordsmyth Dictionary-Reverse Dictionary
3. Wordsmyth Thesaurus with Conceptual Index.

Since a unified approach to reference is proposed, the process of storing and editing material can be unified. Different reference works serve different reference needs. But when all reference types are, in principle, availiable in electronic form with insignificant storage costs (e.g., CD-ROM), the differences among them can be understood in terms of interface development. Further, the Lexipedia database can be used in a variety of electronic reference products. As a database of ordinary language concepts, it can be tailored for use as a technology for text retrieval and hypertext navigation.


2. The "Lexipedia" Concept. (back to Contents)


The concept of a "Lexipedia" combines the roots of "lexis" (Greek: word) and "paideia" (Greek: education). We combine these forms in the term "Lexipedia" to emphasize that the meanings are encoded in a way that facilitates learning. Dictionaries are references that "list words in alphabetical order and describe their meaning". An encyclopedia is a repository of topical knowledge about subjects worth learning. Thus, when we turn the dictionary into an interrelated network of meanings, we turn it into a new source of word knowledge and world knowledge. It becomes an index of the lexical concepts that are at the foundation of the language, and an entré into a network of real world concepts and useful terms for them.


3. Diagram A: Words and the World: Classification of Language and Encyclopedic Reference Works (back to Contents)

Diagram A (see below) shows the world of reference along two dimensions: (1) the dimension of language and the world it relates to in the reference process; and (2) the symbolic dimension of language - words and concepts. Language is used to explain the world, which recursively includes language as part of that world. In gaining access to the knowledge stored in reference works, language is used as an indexing device. Alphabetically organized reference works give access to knowledge by words, and secondarily by the concepts one can gain access to through the word one looks up. Nearby material is usually unrelated (except for morphologically related variants of the word on is looking up). Topically organized works bring the categorical concept to prominence, and invite the reader to browse material nearby as relevant to the concept s/he has in mind. The target is to allow (1) access by word and concept, to (2) knowledge of words and their meanings and knowledge of the world - i.e., important concepts, theories and descriptions of the world.



The aim of an integrated reference system is to create a body reference material in which there is access to knowledge of both language and the world, through both alphabetical (word) and topical (conceptual) organization. This proposal outlines the case for an integrated language reference process. Since dictionaries contain a fair amount of encyclopedic information, case for integrating linguistic and encyclopedic reference processes can be understood as an extension of the same argument.



B. Roget's Challenge


1. Roget and the Thesaurus. (back to Contents)


Dictionaries are alphabetical references which take the user from a word to a meaning (or from term to concept). If the learner/user has heard or read the word, and wants to know the meaning, then the skill of alphabetical lookup is all that is needed to find the entry. A review of the definitions may then aid in the understanding of the meaning of the word in the context that originated the query. There are many ways in which dictionaries can be used to confirm and clarify the meaning(s) of a word.

But if someone has an idea - a vague notion or a clear concept - and needs a word for it, or a word for a related concept, a dictionary isn't of much use. In some cases, the definition may give a related word. But in most cases dictionaries are not suited to pursuing the paths of inquiry into the relations among words.

The original intent of Peter Roget was to develop a system for finding a word for an idea that is only vaguely formed. History has seen his vision curtailed - overshadowed by the usefulness of his lists of synonyms. He was concerned not with the task of specifying the meaning of a word, but with finding a word for the meaning one had in mind. The system Roget devised was an elaborate scheme for classifying meanings. The system is used to play a game of "20 questions" with oneself: (1) is the idea an "abstract relation" (or does it refer to "space", "matter", "intellect", "volition", or "affections"; (2) if the concept is an "abstract relation", then does it refer to "existence", "relation", "quantity", "order", "number", "time", "change", or "causation"? (3) if it is a "quantity" concept (for example), then does it refer to "quantity", "degree", "equality", or "inequality"? And finally, after answering such a series of questions, the user finds a list of words, grouped according to same/similar meanings.

This scheme is rather unwieldy, so very few people actually use it. Instead, the term thesaurus has become a label covering a variety of works which gather synonyms together in various forms. Most common now is the dictionary of synonyms, in which the entries are listed alphabetically, with a string of synonyms or similar/related words. The meaning which unites the words is not often given in a precise fashion. The Roget's Fifth International Thesaurus has a more usable classification system, presented topically in the front of the volume, along with an index that tells the user which synonym groups a word can be found in.

Another class of book that helps the user move from meaning/concept to word can be referred to as a "lexicon". There are three of note currently available. The Longman Lexicon of English has a topical classification scheme, and word lists, with glosses (but without synonyms). But it only gives one sense of a word at each listing. In order to get other meanings, the user must consult all of the places where a word is listed. But the word does not have a listing for all of its meanings. Random House's Word Menu, and Facts on File's Descriptionary are similar in some respects to Longmans Lexicon. They have short glosses for the words listed. But the word lists are generally quite heterogeneous, leaving the user to sort through a large pot of words. Word Menu has an index that ties together the places where one finds the word in the reference. But the view of the meanings of a word is a restricted, stunted one. Only the meaning relevant to that particular place in the classification system is given. A dictionary allows one to survey the full range of meanings of a word, and get a full sense of the meaning potential of a word. But this function is lost in these lexicons. Basically, people don't know how to use them, and they don't really know why to use them.

A lexipedia, however, combines both functions - finding a meaning for a word, and finding a word that captures all or part of a meaning - thus allowing it to be used for both the precise analytical work of pinning down a meaning, and the free and creative process of brainstorming for new ideas.

Concepts/meanings that are defined in a dictionary usually follow one of several standard patterns. The most common is the "analytical" definition, in which the class that the meaning belongs to is indicated, as well as the attributes/qualities that differentiate the particular meaning from others in the same class. For example, the word "wing" is defined in one edition of the American Heritage Dictionary as "An airfoil whose principal function is providing lift, especially either of two such airfoils symmetrically positioned on each side of the fuselage." This definition indicates that a wing belongs to a class called "airfoil". And it indicates the differentiating quality of a wing is that its primary function is to provide lift. It also indicates that the most common position of a wing is "symmetrically positioned on each side of the fuselage".

Nowhere, however, does this definition indicate directly that a wing is a part of an airplane. And the definition of airplane does not indicate what its typical parts are - wing, engine, aileron, etc. Where could a student go if she had a question as simple as "what are the parts of an airplane"? She might even have a visual image of an airplane, and the part she wants to refer to. But only a specialized lexicon, or a specialized visual dictionary, would have that information. (Neither Word Menu nor Bernstein's Reverse Dictionary contain this information.)

As the history of lexicography shows, the concise defining style of dictionaries has evolved out of the special needs of the print medium. In order to give information (such as etymology, meanings, spelling variants, etc.) about a large vocabulary, it was necessary to be very concise. The academic nature of the enterprise - to provide a source of authority - has also influenced the evolution of the dictionary.

The purpose of a lexipedia is to put the dictionary in a larger context - the context of the important relations among meanings, regardless of whether or not they are specified in a definition.


2. The Locus of Linguistic Meaning: From Words to Meanings and Back


a. Meaning in words and phrases (back to Contents)

One of the most fundamental disputes in linguistics has to do with the locus of meaning in the understanding of language. What is the contribution of syntax and grammar? What is the contribution of semantics and the interpretation of words? What is the contribution of the extra-linguistic context? In one approach, meaning resides in words, which are then put together by the application of rules. This is the approach that is tacit in the construction of dictionaries, which explain word meaning. In a second approach, the rules of syntax contribute more directly to meaning. Words refer to things in the world. But language means through the relationships among words. This approach is tacit in the works which help us find words to designate the meanings we have in mind.

Modern structural linguistics has focused on the rule-constrained nature of language which makes it possible for people (including young children) to construct entirely novel and unique sentences. This has buttressed the position of structural linguistics against the position of behaviorists in psychology who have emphasized the role of experience (punishment and reward in particular) in learning. But the process source of these rules is a matter of contention. Chomsky sees the brain as the master rule generator, laying down the universal rules of grammar underlying all the particular applications of a given language. Thus, the creative possibilities of rules - and the process of inferring those rules from the creative material of linguistic communication - have occupied structural linguists. Words are seen as arbitrary signs which are moved around according to the rules of grammar. The key to meaning is the grammaticality of phrases and sentences.

Others see the historical processes of language as fundamental. And with this historical bent, the history of words becomes more important.

In his Lettre, Humboldt refers to a specifically linguistic form of imagination, whereby ideas are clothed with sounds, acquire an existence separate from their speakers, and come back to them as words that in turn suggest ideas fixed by language.

Words then are drawn into the language because they are needed - for designation of an idea of importance. And thus words (and their meanings) become a focus of attention. The impulse to fix the meaning of words authoritatively (thus assuring the authority of those who use them in a particular way) becomes irresistible. Dictionaries have clearly been part of the process of fixing meaning in static works of reference. The more fluid parts of the language are left to slang dictionaries or are assumed to be ephemeral until proven otherwise.

This approach to introducing meaning into the language solely through the process of defining words shows only half of the picture of meaning. The other half, the process of designation, is left by the wayside, for the most part. The exceptions are works such as Bernsteins Reverse Dictionary (N.Y.: Times Books, 1975), and The Readers Digest Illustrated Reverse Dictionary (N.Y.: Readers Digest, 1990), which take different approaches to moving from meaning to word, or from phrase to word.

The approach taken here is pragmatic and constructivist. Meaning is viewed as an emergent property that comes from and through its use in the process of interaction - verbal and textual. There are two phases to the process of meaning construction, one focusing on definition and the other focusing on designation.

b. Meaning construction is circular (back to Contents)

There are two important linguistic processes in the development of linguistic knowledge and skills: (1) explaining the meaning of words; and (2) designating (identifying and naming) the meaning of phrases. The first process is developed in dictionaries, which move from a word to an explanation of the ideas the word can convey. The second process is infused throughout the educational process, identifiable whenever a teacher/parent says thats a truck or the wheels are held in place by that rod, called an axle, or to boil a liquid in order to extract the essence is to 'decoct it. This process involves moving from a concept/idea to a word.

This second process has not been developed through reference works to the same degree as the first. There are several kinds of work that have explored the territory: (1) the classification systems of Rogets Thesaurus, Longmans Lexicon, or Word Menu; (2) the reverse dictionaries of Bernstein or Readers Digest; and (3) the WordNet word relations database. But none has established a definitive model for development of the genre.

For several reasons, I believe there is a new opportunity to define this reverse dictionary niche, and expand it to co-equal status with the traditional dictionary - ultimately integrating both in both print and in an electronic semantic network, linked to encyclopedic sources of information. First, there is now a market niche. There are a sufficient number of works on the market that make claims similar to those of a reverse dictionary - helping to find a word when you know the concept. Second, while the size of a print work must generally be kept small, the factor of size is less of a constraint in electronic publishing. Third, the synergistic interaction of print and electronic forms can be exploited with a combination of a dictionary and a reverse dictionary. The explanation and proposal which follow are designed to assure the leadership of McGraw-Hill in the expansion of this niche. If properly developed, with a vision of the future of intelligent electronic tools for access, the niche becomes a key intellectual property.

Standard dictionaries handle essential phrases as Runins, buried within the entry for one of the words in the phrase. According to Landau, space is the major constraint that has curbed the inclusion of phrases. And the result is rather unsatisfactory.

Verbal idioms are particularly troublesome. No form of alphabetization can successfully deal with all types of idioms without listing each in several places, and no dictionary can afford the luxury of such repetition. Verbal idioms such as have ones eye on are usually run in at the end of the entry for one of the key words of the phrase, in this instance have. The question of which word is most likely to be sought by the user is one that is sometimes impossible to answer. Should the idiom be placed under the first word, or the most important word? Sometimes the first word is variable, as in shed or throw light on. Sometimes it is not easy to say which word is more important, as in hang fire. Most dictionaries prefer to list idioms under the first word, but exceptions are common. Absolute consistency is purchased at the price of the readers confusion and frustration.

More recently, Cambridge University Press has responded to the problem of access by inclusion of an index of phrases (idioms, metaphors, slang expressions, etc.) in its Cambridge International Dictionary (1995). The phrases are repeated in the appropriate place for alphabetical lookup under each of its constituent words. The meanings are still given under one of the head words, but the problem of access to phrasal concepts is lessened.


c. Elaboration (back to Contents)

The process of thinking and creating with language thus has two phases: (1) the mediation of meaning by words; and (2) the shaping of meaning by phrases. When we look up a word in the dictionary, we are looking for the kind of context that defines the central core of meaning - that gives the word its sense. There may be several of these senses, so it may be necessary to sort out the possibilities. On the other hand, when we are looking for a word, we must begin with a phrase that comes close to capturing the concept we have in mind. We may have a defining phrase in mind - a phrase that succinctly captures a defining contest of a word. We then go to a reverse dictionary or a thesaurus to find a word for the idea. But the process is continuous, because when we find a word for the idea we have, that word may have multiple meanings, or refinements of meaning that require us to go back to the dictionary.


3. Diagram B: Networks of Words and Phrases in the Meaning Process.: (back to Contents)


This diagram (see below) shows the complexity of the definition-designation process. It also distinguishes three kinds of designation. "Existential designation" refers to the process of asserting that "x is a y" - for example, "John is a man". John - who is not a category - is designated as belonging in a category. Relational designation refers to the process by which we qualify an existential assertion - for example, "a car is a kind of vehicle". The qualification relates the category of car to the category of vehicle. In this case, "kind of" is a bound collocation, acting as a single designating word. There are many qualifying phrases that could be used in this process. Finally, idiomatic designation involves phrases that have acquired idiomatic status in a language.


________________________________

Diagram B:
Networks of Words and Phrases in the Meaning Process.




C. Classification of Language Reference Works


1. From Word to Concept and Back. (back to Contents)


Reference works which assist with the exploration of the linguistic meaning process can be divided into three categories: (1) word to concept references; (2) word to word references; and (3) concept to word references.

(1) The first category - word to concept reference - has crystallized into the modern dictionary. Although considerable additional information about a word is typically given - such as pronunciation, alternative spellings, derivative forms, etc. - looking up what a word means is the primary popular concept of the dictionary. The user begins with a word whose meanings are not clear, and searches for the possible meanings - the ways the word has been typically used.

(2) The second category - word to word reference - is typified by the modern dictionary of synonyms (or thesaurus). The user has a word in mind, and is looking for an alternative word which has essentially the same (or very similar) meaning. Additional information is sometimes given - such as glosses of the common meaning of the group of synonyms - but the main function is to assist in adding variety to ones stock of words for writing.

(3) The third category - concept to word reference - has a variety of representative works. But no model has emerged to crystallize and shape this essential process involved in thinking and writing. The user typically has an idea - and perhaps one or more related words in mind - but doesnt know or can\rquote t remember the word for it. With a dictionary, we ask what does this word mean? With a thesaurus, we extend this question by asking what other word means the same thing. But with the Reverse Dictionary, we ask the question what is this thing or idea I have in mind called, or what is the name for it?


2. Diagram C: Classification of Language Reference Works (back to Contents)


This diagram (see below) shows the three types of language reference work which have emerged from the relations between words and the concepts they designate (or, if you will, between concepts and the words they define). It also locates the proposed works as hybrids that each link the characteristics of two of the reference types.



Diagram C: Classification of Language Reference Works*




3. Word to Concept Reference: The Dictionary. (back to Contents)


Dictionaries vary in many respects, from historical to topical, from comprehensive and unabridged to short word lists with minimal definitions. The uniting concept is that these are alphabetically organized references in which one can look up a word and find its meaning.

Dictionaries give additional information, which in some ways impinges on both of the other categories. First, one of the techniques of defining used in all dictionaries is to select a synonym as a (usually partial) definition. The dictionary acts in some ways as a thesaurus, then. But some dictionaries also include groups of synonyms - usually not more than a thousand or so groups, however.

Second, dictionaries generally give classificational information also. Cross-references are one example. But the more interesting example is field labels - such as slang or accounting or chemistry or archaic - act as a sort of hidden index to related words. No print dictionaries make this index explicit, so the user might know that jive is slang, but not have any way to find a list of the other slang words in the dictionary. In electronic form, however, the limitations of space do not apply, and some dictionaries have made this information accessible. The American Heritage Electronic Dictionary can be searched by these field labels, and a list of corresponding words can be viewed and looked up in the dictionary.


4. Word to Word Reference: The Thesaurus. (back to Contents)


There are several variations on the idea of a thesaurus. None of these live up to the original intent of Peter Roget, who wanted to develop a Concept to Word reference, and only included synonyms because the words designated roughly the same concept. The universal classification scheme he developed as an set of concepts, however, proved unwieldy. So people began to think of the work primarily as a word to word reference. Instead of accessing the synonyms through the concept scheme, people began to use a simple alphabetical index to the thesaurus. (See, for example, Rogets Fifth International Thesaurus.) Because the index is not essential if all of the words are headwords, the genre has become typified by a single alphabetically organized work - a dictionary of synonyms.
Most have a simple alphabetical organization, without index or classificational system.

5. Concept to Word Reference - Visual Concepts. (back to Contents)

The richness of this area can be seen from the variety of works which have explored this territory. The following classification shows this diversity, but also shows that this type of reference has not achieved coherence in the public mind.

Visual and pictorial dictionaries have achieved some distinction as essential references in some situations. We can understand the interest in viewing a picture of a complex object and finding the names of the parts, seeing pictures of the kinds of dog or horse, or seeing a picture of a soccer referee and finding out what his shirt is called. The function is useful in ESL, and the pictures are attractive to school audiences. But these works dont focus and enrich this category of reference as a genre.

6. Concept to Word Reference - Verbal Concepts. (back to Contents)

a. Classificational Systems of Concepts. (back to Contents)

If there is a modern prototype for this sort of reference function it would be Rogets Thesaurus, with Peter Rogets original universal classification scheme as a source of verbal concepts. In his Preface to the 1852 edition, he notes that he sees his work as a system of verbal classification, and calls it a classed catalog of words. The most recent direct descendent of Rogets work, Rogets Fifth International Thesaurus, uses new categories and a rearrangement of the classes and categories Roget developed, resulting in a developmental-existential classification scheme. When one finds the appropriate concept in the scheme, one can find not only one word for the concept, but a list of synonyms.

There are several deficiencies in this approach. For example, finding the appropriate concept is still difficult. And when we find a nearly appropriate concept and its list of words, it is just as likely as not that we have decided the word we need is related, but not in the synonym list, or a nearby list of synonyms. Most thesauruses dont have glosses or definitions, so we cant really tell what the appropriate sense of the word is, or the range of possible meanings.

There are three other works which are organized around a classification scheme. Each of them organizes a lexicon into categories of related words, and provide indexes to find which categories a word appears in.:

(1) Longmans Lexicon of English has an anthropo-linguistic classification scheme, using categories from everyday life. There is a gloss of each word in the appropriate sense when it appears in a list, but not every important every sense of a word is included. And in order to find several senses of a word, one needs to look in several different places.

(2) Word Menu also has a topical organization, and short glosses of one sense of a word, and a word index to find the different places a word appears. Again, there is no attempt at thoroughness of vocabulary coverage or sense coverage, so it cant play a dual role as a word to concept reference.

(3) Descriptionary is similar, but smaller and without an index.


b. Definitional Concepts (back to Contents)

There is a small but fairly recognizable category of work called a Reverse Dictionary. The word reverse cues the potential user that it is for doing the opposite of what is done with the dictionary. Examples of concept to word reference are intuitively clear. But the two major implementations on the market today are not constructed in a way to crystallize that notion, and make it clearly usable.

(1) Bernsteins Reverse Dictionary lists words and concepts in a less than systematic fashion. Most entries are fairly simple words and adjectival or noun phrases. It is a helpful jog to the writer or collector of interesting words. Once found, there is no way to look up the meaning of the word if the entry phrase doesnt do the job. Other meanings arent listed.

(2) The Readers Digest Reverse Dictionary is more comprehensive, incorporating more lists (kinds of things and parts of things, for example), as well as fuller defining phrases. Multiple senses of an entry word are sometimes listed (but not all of them). The failure to incorporate a full dictionary to supplement the reverse dictionary seems to be the fatal flaw that stymies the completion of the reference circle.

c. Word Relations Concepts. (back to Contents)

An electronic database - WordNet - has been developed by George Miller (funded by the U.S. Navy). This database covers the entire English lexicon, marking several word relations: (a) synonymy; (2) taxonomy; and (3) part/whole. Some glosses are incorporated in order to differentiate synonym groups, but there is no attempt to provide full definitional information or dictionary entry information.

d. Miscellaneous concepts (back to Contents)

This is a grab bag category:

(1) Benson, Benson and Illsons Combinatory Dictionary of English lists the prepositional attachments of a large set of nouns and verbs, attempting to survey the grammatical collocations of English.

(2) The Cambridge International Dictionary has a listing of phrases at the end that acts as a sort of index to the Runins (idioms, expressions, collocations, etc.) contained within the entries of the dictionary.


7. Boundary Cases: Dictionary and Thesaurus (Combined Word to Concept and Word to Word Reference) (back to Contents)


The Websters New World Dictionary and Thesaurus is an example of this genre. There are only 2-3 cases of publishers combining in one volume a thesaurus and a dictionary. The dictionary is usually a short desk dictionary, for casual reference only; and the thesaurus is typically a short alphabetical dictionary of synonyms. (The Random House English Language Desk Reference falls into this category.) WNW goes a bit further by positioning the same alphabetical range of thesaurus entries and dictionary entries on a page.


8. Boundary Cases: Thesaurus and Conceptual Lexicon (Combined "Word to Word" and "Concept to Word" Reference) (back to Contents)


These works listed above as Classificational Thesauri or Lexicons present their classification schemes of concepts at the front of the work, as a table of contents, while the index is only a word index. There are two Thesauri which border on the concept to word reference category by incorporating a concept index at the back of the book.

(1) Rogets 21st Century Thesaurus includes a set of topical concepts and lists of relevant words in a sort of index at the back of the work. The lists are long and undifferentiated, and the concept categories are less than adequate in tapping our intuitive habits of thinking.

(2) The most recent (1995) edition of Rogets II Thesaurus from Houghton-Mifflin includes a short listing of concepts with short lists of related words. Again, the lists are less than adequate as a conceptual index.



D. Proposed Language Reference Works: Synergy Through Integration.

1. An Integrated Thesaurus-Dictionary: Word to Concept and Word to Word Reference (back to Contents)


There are no fully integrated dictionary-thesaurus combinations. The American College Dictionary (the 1950s predecessor to the Random House College Dictionary) included more synonym groups than most dictionaries (3000?), but without full integration or full discrimination. Some current dictionaries include usage discriminations (connotations) among a group of synonyms - for example, Merriam-Websters Collegiate Dictionary (10th Edition). Others include groups of synonyms at the end of 1-2,000 entries. But none have gone so far as to include a full complement of synonym entries, along with an integration into the dictionary by linking each synonym to the definitions (denotations) on which it is synonymous.

Another contrast model comes from the word to word reference category. Houghton-Mifflins Rogets II Thesaurus includes definitional glosses for its synonym groups. But not all of the definitions of a word are given; the entry word list is much smaller than a dictionary; and the definitions are not the same as the Houghton-Mifflin dictionary (American Heritage). For these reasons, I would include Rogets II Thesaurus in the Word to Word category.

The proposal for an integrated Dictionary-Thesaurus (or Thesaural Dictionary) will help to crystallize and reinforce the relationships between these reference types. Rather than including insufficient synonym information in a dictionary, and insufficient defining information in a thesaurus (Rogets II), the proposal is to give sufficient defining information (word to concept), and sufficient synonym information (word to word) to complete the circle created by these two reference types.


2. A Dictionary-Reverse Dictionary (or Dictionary with a Conceptual Index - Word to Concept and Concept to Word References) (back to Contents)


It is my argument that the variety of concept to word references has not crystallized or become focused as a category of reference because there has been no model to organize the works for comparison. The model I think is most likely to accomplish that crystallizing function - creating a market and setting the pace for it - is a combined Dictionary-Reverse Dictionary.

The proposed McGraw-Hill Dictionary-Reverse Dictionary (Conceptual Dictionary? Lexipedia?) brings together the word to concept and concept to word references in a coherent whole. It is based on three design decisions: (1) combine the dictionary with the reverse dictionary into one volume in order to complete the circle of reference; (2) either graphically distinguish the two reference processes - for example, underlining phrasal/conceptual entries - or physically separate the two reference processes by incorporating the reverse dictionary in the back of the book where the index has traditionally been located; and (3) select a useful and intuitive set of concepts for the reverse dictionary index.

a. Justification. (back to Contents)

My argument is that the combined Dictionary-Reverse Dictionary can crystallize this market for the following reasons:

1. Juxtaposing the dual functions of Dictionary and Reverse dictionary invites the contrast between finding the meaning of a word, and finding a word for a meaning. People may not easily see the function of a reverse dictionary or a lexicon or a Rogets type classification thesaurus, without close familiarity with its contents. But none of these works has been combined with a dictionary. Some have included glosses. Some are presented alphabetically. But none have combined the two in order facilitate comparison and contrast of functions.

2. The combination of functions should be accompanied by a physical sign of difference - either a graphic design feature (e.g., underlining phrasal headwords in a single alphabetical presentation) or the physical separation of the dictionary and the reverse dictionary - with the latter acting as an index for the former. This distinction invites the user to understand the differences in function. People have familiarity with the concept of an index, and know how to use it. With the reverse dictionary, it is easy to see that the meanings (concepts/definitions) are being indexed. The failure to find a format that is easy to understand and use helps to explain the failure of any of the concept to word references to create a model in the consumers mind. The topical classifications (Rogets Thesaurus; Longmans Lexicon; Word Menu, etc.) are awkward and unintuitive. It isnt clear when one can or should use the reverse dictionaries, and how they might relate to a dictionary. And the inclusion of many pictures in the visual dictionaries makes it unclear whether and how to move from non-visual concepts to the words that designate them.

3. The examples of usefulness that can be generated are limitless. What is the flap on the wing of an airplane? How many kinds of horse are there? What is the hat a bishop wears. What is a pear shaped body called? What is a word for excessive devotion to religious practices? But the almost limitless nature of these questions - if they become encyclopedic - also helps to explain the difficulty in crystallizing and focusing this sort of reference. This proposal will succeed, I believe, because it limits the concepts to a subset of important linguistic concepts (which are also tied to the dictionary - as defining phrases or runins - or are important word relations concepts). The three types of concept to be incorporated are: (1) defining phrases; (2) runins; (3) word relations concepts.

b. Explanation (back to Contents)

One difficulty in constructing a reference that moves from concept to word is that there are so many different ways that a meaning can be expressed. And those expressions may not always get close to the point of having a word to convey the meaning. It is difficult to construct these phrases in the process of lexicography, and we have difficulty doing it in a comfortable way when someone asks us the meaning of a word. We often fall back on an explanation of how a word is used - which sometimes gets us close to the meaning. Part of the solution proposed here is to draw on precisely those phrases which have been used in dictionary construction - the key defining phrases that pick out important components of meaning.

This proposal for a McGraw-Hill Dictionary-Reverse Dictionary involves listing three types of phrasal concepts: (1) defining phrases; (2) Runins (idioms and other expressions); and (3) word relations concepts


(1) definitional concepts. The words which define an entry can be seen as creating a semantic field. A reverse dictionary should list significant defining phrases. The set of definitional concepts will be extracted from WEDT, as well as other dictionaries. They will be evaluated for their intuitive capturing of an idea we might have, which has a name that might be difficult to remember or learn. Many of these concepts are two and three word noun, verb and adjectival phrases. The meaning might be deduced from the meaning of the component words, so a dictionary would not include the phrase as a multiple word entry. But knowing the meaning (alternative phrases) will not help when one wants a word to cover that meaning.

(2) Runins. Because the reverse dictionary is functioning as an index to the dictionary, it would be wise to incorporate the runins into the reverse dictionary portion of the work as referenceable concepts. This is similar to the Cambridge International Dictionary, which simple lists these phrases. In the dictionary, the Run-in fields list an idiomatic phrase, or a noun or verb phrase that uses the main entry word. They have separate definition fields, and can be indexed in three ways:

a. The Runin phrase can be listed as a cross-reference, referring the user to the headword. (this strategy could have been used by The Cambridge International Dictionary instead of an index of phrases at the back of the dictionary);

b. The phrases used to define the Runin can be listed (as in #1 above), referring the user to the Runin phrase as well as to other words defined by that phrase.

c. In some cases, the Runin itself may be a phrase that can be used as a concept - listing the word(s) that designate the same concept.

(3) Word relations concepts. In addition to the word indexes based on the fields listed above, this proposal suggests incorporating conceptual indexing by word relations - for example, part/whole, taxonomy, instrument, origin, etc. These are concepts which act as linking phrases, and are frequently used in semantic networks. The set of word relations concepts could include the following, selected from a larger list I have developed. (Encyclopedic relations are excluded. See discussion below if integration of the dictionary and encyclopedia.)
1. Converse

x is the converse of y

buyer/seller; include/belong to

2. Kind

x is a kind of y

green/color

3. Unit

x is a unit/measure of y

dollar/currency; foot/length

4. Example/Instance

x is an example/instance of y

New York/city; item/news

5. Part

x is a part/component of y

sentence/paragraph; suitcase/baggage

6. Member/Group

x is a member of y (and y is a group/collection of x)

tree/forest; senator/senate

7. Portion

x is a portion of y

slice/pie

8. Stage

x is a stage/aspect in the process/activity of y

buying/shopping

9. Ingredient/Composition

x is an ingredient of y (y is made of x)

flour/bread; rubber/tire

10. Production

x is a means of making/producing y

baking/bread

11. Prior State

x has prior state y; x is an immature y (cf. Age/Maturity #6)

colt/horse; bread/dough

12. Conventional Association

x is conventionally (arbitrarily/typically/commonly) associated with y

bagel/lox; bagel/breakfast; ham/eggs; red, white & blue; nuts/bolts; needle/thread; luggage/trip; audience/concert; clown/circus; animal/zoo; dog/bone; millionaire/money

13. Cause/Effect

x causes y

14. Prevention

x results in absence of y; x prevents y

15. Agent/Product

x produces y

baker/bread

16. Agent/Material

x uses y (to produce z)

baker/flour(/bread))

17. Agent/Action

x does y (to z); x y's

professor/teaching, teach(/student)

18. Agent/Goal

x does something (z) to y

hunter/animal)

19. Agent/Instrument

x uses y (for z)

carpenter/hammer

20. Recipient

x receives y

patient/medicine; winner/medal; heir/inheritance; foundation/donation

21. Instrument

x is used in order to y (by z); x is used for the purpose of y

gun/shooting, shoot; luggage/storage

22. Location For

x is the location for y

gallery/art; supermarket/groceries; nursery/plants

23. Container/Contained in

x is a container for y; (y is typically contained in x)

jar/jelly; eggs/carton; suitcase/clothes

24. Source

x is the/a source of y

quarry/stone

25. Made at/Purchased at

x is made at/purchased at y

bread/bakery

26. Field of study

x is the study of y

anatomy/human body

27. Head

x is head of y

chief/tribe

28. Staff

x is (a member of) the staff of y

waiter/restaurant; crew/ship

29. Plan

x is in the time plan for y

itinerary/journey

30. Participant/Event

x participates in y

guest/party; student/class

c. Format (back to Contents)

There are two ways in which the proposed work could be organized. First, there could be an integrated alphabetical listing for all of the entries, combining dictionary and reverse dictionary functions. Second, the work could be divided into two segments - a dictionary and a reverse dictionary or index. In the first half of the work would be an alphabetical listing of dictionary entries. The second half of the work would be an alphabetical listing of the definitional concepts, runins and word relations, along with the words for those concepts.


3. Thesaurus-Reverse Dictionary (or Thesaurus with a Conceptual Index - Integration of Concept to Word and Word to Word Reference) (back to Contents)


If a thesaurus is constructed as an alphabetical word to word reference, space can be saved by including words as synonyms which dont have headword status. Many thesauri use this space saving strategy, but then face the problem of how to get access to the words listed within the entries. The typical response is to add a word index. These differ from conceptual indexes, such as those contained in Rogets 21st Century Thesaurus and the recent edition of Rogets II Thesaurus, mentioned above. These conceptual indexes arent thorough enough to count as true reverse dictionary or concept to word references.

There is one attempt to combine word to word and concept to word references in a thesaurus: Marc McCutcheons Rogets Super Thesaurus (Cincinnati, Ohio: Writers Digest Press, 1995). The concept to word reference is contained as a word find feature within the single alphabetical listing of the thesaurus (dictionary of synonyms). The concepts are limited in number, and are not systematic or complete in treatment, however, and presentation as a single reference de-emphasizes the distinct reference functions that are being incorporated.

The proposal here is to use "reverse dictionary" material, and combine it with the thesaurus as a Thesaurus/Conceptual Index. The resulting work would be superior to the competitors, and would crystallize - in ways Rogets classification scheme, or Rogets II index, or the Rogets Super Thesauruss word find function cannot - the need for combining these two types of reference.


E. An Integrated Lexipedia-Encyclopedia Database. (back to Contents)

Most dictionary entries explain the "linguistic meaning" of words - by showing the relationships and interactions among the meanings of words. But there are a substantial number of "encyclopedic" entries in a typical dictionary, which do not deal with the purely linguistic meaning of words. By "encyclopedic", I mean the description or explanation of the place of something in the world (rather than the place of a word in the linguistic structures of the language) - a place indicated by a geographical term, a person indicated by a biographical entry, or something concrete like a medical instrument. The explanations offered in dictionaries have been of necessity short and incomplete. They focus on the bare essentials of understanding the range of possible meanings of a word, rather than explaining the full range of theories and associations that constitute our knowledge of the world.

The integration of a dictionary and an encyclopedia, then, is a natural extension of the dictionary reference process. We move from a dictionary to an encyclopedia when we want to know more than just the basic meaning of the word, or its basic reference. And we move from an encyclopedia article to a dictionary when we want to know the meaning of a word.

In both print and electronic form, there are ways to integrate the two reference processes - dictionary and encyclopedia. In print, there have been "Encyclopedic Dictionaries" that provide more extensive treatment of encyclopedic words, and that incorporate encyclopedic articles. Though it hasn't been done (as far as I know), it would be possible to incorporate full dictionary entries as boxed/supplemental material at an appropriate place in a print encyclopedia - for example, defining all headwords, and giving boxed dictionary entries for all words with a low frequency of use (thus a high difficulty).

In electronic form, there are many ways to create an interface between dictionary and encyclopedia materials. While with language reference there is a dual "word to meaning" and "meaning to word" process, there is a parallel process in use of an encyclopedia. While the material is topically organized at the article level, a topic might be discussed in many places in an encyclopedia. In order to get at these treatments, the encyclopedia can be indexed, so that we can go from an idea (represented by a word or phrase/concept) to an explanation. The equivalent to "word relations" concepts for the encyclopedia might be tapped by concepts such as "how does it work?" or "what are its stages of change/development/growth?" This leads to consideration of user interfaces and the use of the Lexipedia in indexing.


1. User Interfaces. (back to Contents)

There are several ways of organizing an integrated dictionary-encyclopedia interface. First, the dictionary and encyclopedia can be stored together on CD-ROM, and offered as options in a menu system. Second, a visual metaphor could be developed. For example, a diagram such as Diagram A could appear on a screen. The user would type in a word or phrase, and then click on one of the dimensions of the 3 dimensional box, depending on whether she wanted dictionary or encyclopedia information, and whether she was expressing a meaning she wanted a word for, or was expressing a meaning she wanted an explanation for.


The Lexipedia can be used as a "semantic network", which could provide a visual metaphor for an interface to reference databases - dictionary or encyclopedia. Users are likely to find the network metaphor understandable as defining a meaning space in which our imaginations can move off in many different directions, from node to link, following the impulses of imagination and the hunches of disciplined research. The ordinary language concepts are easily accessible for formulating queries: "what are the parts of x (automobile; airplane; sewing machine)?"; "what tools does x use (surgeon; carpenter; surveyor)?"; "what influences x (the stock market; automobile accidents; tides)?" The explanations may require further research in an encyclopedia, or more technical reference works. But these explanations might not be sought without knowledge of the words which are the starting point of research on the relevant concepts.

2. Uses in Text Retrieval (back to Contents)

Research and development in text retrieval have been exploring the use of semantic networks, including word relations approaches. Full text search of documents is useful in a variety of contexts, from the huge stores of corporate documents to repositories of legal cases, and from academic materials to newspaper and magazine texts. Keyword searches and boolean searches of text are dull tools for retrieval. Users often cannot anticipate the words that will be used to express an idea. For example, if the full text of a dictionary or encyclopedia is searched to discover the different kinds of "abusive speech" or "purple flower" or "nocturnal animal", there would be a high percentage of useless information. One would have to wade through many dozens of falls returns (i.e., entries which use one of the component words, but do not refer to the concept one is looking for.

For this reason, a thesaurus is an excellent tool for improving the breadth of a full text search of a store of documents or any text stream. Specialized thesauri have been developed for high volume areas, like legal research (Lexis/Nexis), or medical research (Index Medicus). But the only thesaural index of ordinary language concepts available for text retrieval is WordNet, which indexes only two relations in addition to synonymy: part/whole and taxonomy.

The Lexipedia is an important step toward a more refined search and navigation capability. The first improvement is through the use of synonyms and similar words. The user can choose a meaning, and not simply a word, and search for the words that have that meaning. The second improvement is conceptual retrieval through the use of word relations and definitional phrases (the reverse dictionary index). Thus, if someone is interested in automobiles, documents would be marked which contain engine (part of), sedan (kind of), commute/joy ride (used for), jalopy (register/slang), driver (used by) etc. Once the documents have been found and ranked, the word relations, which are precisely defined and linked, can be used to navigate, and find other words to search for. The process can then repeat itself.

3. Uses in Print Index construction (back to Contents)

A word relations database can be used as a thesaurus for indexing general works. Indexers in specialized areas have access to highly developed terminologies (also called Thesauri) that facilitate the indexing and cataloging of materials in a way that will facilitate access by appropriate keywords. But there is no general indexing thesaurus of ordinary language concepts that can give an indexer a head start on indexing of general materials. The Lexipedia can fill that niche.

4. Uses in Hypertext Construction (back to Contents)

When the creator of an electronic document wants to turn the document into hypertext, s/he must make two basic decisions: (1) how should the document be segmented; and (2) what links among segments should be incorporated. This process is parallel to the indexing process. The possibilities are nearly identical to those for print works. Basically, the Lexipedia material (conceptual index) can be used as an aid in establishing links among the segments/locations of a hypertext. The hypertext is analogous to a semantic network, with the nodes as segments/locations of the document, and the links among the nodes as hyperlinks. The Lexipedia/Conceptual Index provides conceptual links among words, and thus can act as a sort of template in hypertext construction.



F. Conclusion (back to Contents)

The world of reference will be gradually transformed in the process of adjusting to the electronic production and distribution of reference material. Print and electronic distribution will coexist in the foreseeable future, and successful reference publishers will exploit the interactions among print and electronic forms. In the development of intellectual content, the integration of materials into a comprehensive database is an attractive goal. While it is not fully achievable in practice, there are many ways in which foresighted development of a print work today may lead to successful integration in electronic form in the future - or vice versa.

The users of reference works have developed a limited understanding of the terminology and function of the works they use. What is a "dictionary" and how does it differ from a "thesaurus" or a "lexicon"? What is the range of purposes that can be served by each, and how do they overlap? On these questions, there has been no clear answer from either publishers or consumers. The proposal presented here attempts to sketch out a plan for coherent movement into the electronic age of reference by recognition of the functions performed by various kinds of reference work. By publishing a Thesaurus-Dictionary, the value of each is enhanced, both because of the dual functions that are performed, and because of the expansion of the user's understanding of each type of éntre into the word-concept reference process. By publishing a Dictionary-Reverse Dictionary, and a Thesaurus-Conceptual Lexicon, the value of each is likewise enhanced, as the user becomes more aware of the way s/he thinks in phrases, and the occasions on which s/he is looking for a word, not a definition.


Appendix A: Lexipedia: Sample Entries for a Conceptual Index (back to Contents)


The "Reverse Dictionary" is a conceptual index of words that are contained in the dictionary. It is not an index of all of the concepts that are possible in a language (limitless), or of all the important concepts in a specific field of knowledge, or even of all of the concepts contained in the dictionary. It is an inventory of some important and interesting concepts that are part of the ebb and flow of our thinking processes. The entries come from phrases which come easily and naturally to mind and from questions we might have about "what is the word for x". They are expressions that are part of the common parlance - expressions which in some cases may even seem simplistic. The concepts are listed in order to provide a link with one or more words which may be used to designate (express) the concept. Thus we can move from an idea to a single word which expresses that idea - the reverse of a dictionary. We need this process as a nudge to memory, an aid in research, an exploration in ideas, a survey of the terrain of terminology, and an envigoration of the imagination.

1. Entries are listed alphabetically by each of the significant words in the entry (listed in bold). In one place the entry might be a main entry, but in another place, a sub-entry - e.g., musical instruments would be a main heading. But "musical instruments" would be a sub-heading under "instruments", and would list a cross reference to the main heading.

2. There are examples of the three major types of concepts in these samples of conceptual index.

First, there are definitional phrases, which often are adjectival phrases (e.g., excessive freedom: license), adverbial phrases (e.g., abruptly depart: pike), or gerund phrases (e.g., tending toward balance: counterpoise".

Second, there are "Runins", such as "abide by" (abide by [trust someone to a.]: honor system).

Finally, there are relational phrases, such as "type/kind of" (music [types of]), and "parts of" (e.g., musical instruments [parts of stringed inst.]). Relational phrases may be used as subcategories also (e.g., musical compositions [kinds of]: allemande; anthem; arabesque; aria [kind of a.: cabaletta]).

3. Relational phrases. Some relational phrases are listed as main entries: for example, "excessive devotion to" relates "Bible" and "bibliolatry". The major relational phrases, however, are not listed alphabetically. Thus, "part of", "kind of", etc are not listed alphabetically at "p" and "k", but are listed under the other word or words of the defining phrase. For example, "kinds of musical instrument" are listed under "m" (musical) and "i" (instrument), but not under "k".

4. Many of the entries are complete. But some entries cannot be completed until the final stages of the work process, for several reasons. First, completion of some entries will depend on the completion of other entries. For example, some main entries will contain sub-entries that are also main entries. The sub-entries cannot be completed until desisions are made about inclusion of the main entry. Second, the work process does not allow for a fully systematic process for completion of entries. Some entries are simple to complete. But some - especially those with word relation concepts - require a significant amount of research, which is best done in cumulative stages. For example, the sample entry for crime is "crime (tending to produce c.): criminogenic". This illustrates use of the concept "tending to produce". But the final entry should contain crime related concepts for "kinds of", etc. The entry for "Church" lists only "(excessive attachment to): churchianity", and is included to illustrate cross-listing of complex entries, and will eventually be more complete. Likewise, the entry for dog will contain a much larger list of associations, in addition to the breeds and other relations listed.

5. Entries are listed alphabetically by each of the significant words in the entry (listed in bold). In one place the entry might be a main entry, but in another place, a sub-entry - e.g., musical instruments would be a main heading, but would be a sub-heading under "instruments". For this sample, not all entries have been fully crosslisted and cross-referenced. Main entries are in bold, with major sub-categories in underlined plain text within parentheses. Third level categories (within sub-categories) are noted by plain text within parentheses, without underlining.

6. The major relational phrases are not listed alphabetically. Thus, "part of", "kind of", etc are not listed alphabetically at "p" and "k", but are listed under the other word or words of the defining phrase. For example, "kinds of musical instrument" are listed under "m" (musical) and "i" (instrument), but not under "k". (back to Contents)


* abbey (head of): abbess; abbot; archimandrite; superior; prior; prioress; (kinds of): priory; cell; hermitage.
* abdomen (contained in): amnion; appendix; caul; kidney; liver; mons pubis; mons veneris; navel; omentum; peritoneum; solar plexus; spleen; stomach; transverse colon; veriform appendix; viscera; (regions of): celiac; epigastrium; hypochondrium; hypogastrium; inguinal; loin; mesogastrium; navel; pubes; umbilicus; ventral; waist; (relating to): alvine; celiac; epigastric; ventral; (large): endomorphic; paunch; pot-belly; (support for): bellyband; belt; corset; girdle; panty girdle; truss; waistband.
* abdominal: a. pain (characterized by): bellyache (1); colic; cramp; hernia; stomach ache; traveler's diarrhea; a. symptoms (causing): campylobacterosis; celiac disease; cramps; food poisoning; giardiasis; hepatitis; lead poisoning; peritonitis; premenstrual syndrome; sickle cell anemia; tympanites; ulcerative colitis; umbilical hernia; a. procedure: appendectomy; caesarean section; enterostomy; gastrostomy; Heimlich maneuver;
* abide by (trust someone to a.): honor system.
* abnormal a. condition (combining forms) -osis; -otis; para- ; -pathy; -philial; (kinds of): dilation; disease; inhumanity; malformation; symptom; a. desire to eat unusual substances: pica; a. discharge (combining form): -rrhagia; a. drowsiness: lethargy; a. fears: hypochondria; paranoia; a. fears (combining form): -phobia (e.g., acrophobia; agoraphobia, claustrophobia; hydrophobia; nyctophobia; photophobia); a. growth: (adj.) teratological; varicose; (n.) dysplasia; excresence; growth; hypertrophy; wen; a. inability to act: abulia.
a. increase or expansion: inflation; a. liking for something (combining form): -philia (e.g., coprophilia; necrophilia; pedophilia); a. people: character; crackpot; nut screwball; freak; kook; oddball; weirdo; a. place (occuring in): heterotopic; a. position, manner or form: ectopic; a. sensitivity: hypersensitivity; a. sexual craving: satyriasis (male); nymphomania (female); a. tendency to lie: mythomania
abrupt a. action: brush-off; congé; French leave; rebuff; saltatory; a. change (study of): catastrophe theory; a. change: about-face; backlash; break; cold turkey; flit; offset; qualm; quantum jump; revulsion; a. curve: quirk; a. end to something: flameout; a. movement: bolt; check; dart; flick; flirt; gush; gust; hoick; jab; jerk; jolt; jump; kickback; leap; lunge; lurch; scoot; snap; snuff; strike; scoot; swerve; toss; yank.
* abruptly: a. bring about: precipitate; depart a.: pike; drop a.: flump; plunk; plunk down; refuse a.: balk; remove a.: ax; end something a.: slap-down; turn a.: swerve; deviate.
* absolute (condition of being a.): categorical; complete; downright; exactly; gross; implicit; nothingness; at the outside; perfect; plenary; positive; positively; strict; total; unexceptional; unmitigated; unqualified; utter; very; a. power, control, rule: absolutism; almighty; autarchy; autocracy; autocrat; despotism; imperious; plenipotentiary; reign; seize.
* abuse n. (tendency to seek): masochism; (of land): agricide
* abusive: a. action: drubbing; hazing; misuse; maltreatment; oppression; a. language/speech: (n.) billingsgate; broadside; calumny; castigation; contumely; censure; chastisement; denigration; denunciation; diatribe; epithet; filth; flak; flyting; fulmination; harangue; invective; jeremiad; obloquy; obscenity; philippic; smut; tirade; tongue lashing; vitriol; vituperation; (adj.) foulmouthed; opprobrious; scurrilous; (v.) jeer; libel; mudsling; name-calling; rail; revile; slam; slander; slur; smear; vilify; whip; a. manner: heavyhanded; tyrannical; violent; a. process: corrosive.
* acceleration (tendency to resist): inertia.
* acne (tendency to produce a.): comedogenic
* act (abnormal inability to): abulia;
* action (abrupt): brush-off; congé; French leave; rebuff; saltatory; (abusive) drubbing; hazing; misuse; maltreatment; oppression.
* adhere (tendency to): sticky
* adjustment (of instruments): calibration.
* admiration (excessive): adulation.
* admiring of beauty (excessively): aesthete.
* alcohol (excessive consumption): intemperence; drink; alcoholism; dissipation.
* allergic reaction (tendency to reduce): hypoallergenic.
* amend (tendency to): corrective
* amount (excessive): overdose; surfeit; surplusage
* ancestor worship: filopietism.
* anger (extreme): rage; madness.
* angry look: scowl.
* anxiety (excessive): neurosis; (something that causes): (n.) albatross; bugaboo; combat fatigue; culture shock; horror; nail-biter; nightmare; terror; worriment; worry; (adj.) alarming; troublesome; worrisome;
* appearance (excessive concern with): vanity; vainglory.
* appetite (excessive): gulosity; polyphagia; voraciousness.
* artificial origin: synthetic.
* asparagus (unit of): spear
* aversion (extreme): nausea; repulsion.
* back street: mew
* bagels (associated with): lox; cream cheese.
* baker (produces): bread; baked goods.
* balance (tending toward): counterpoise
* ball (tendency to curve): bias.
* behavior (extreme): outrageous; rabid; theatrical.
* belong (converse of): include
* biological origin: extraction; heredity; bloodline; ancestry.
* bird (flightless): emu
* bitterness (tendency toward): saturnine.
* bleeding (excessive): hemorrhage
* bread (ingredients of): water, flour; salt; (produced by - person): baker; (produced by - process): baking; (produced from): dough
* building (in a slum): tenement
* buyer (converse of): seller
* care/concern (excessive): fastidiousness; finickiness; fussiness.
* cat (female): tabby; (male): tomcat; (young/immature): kitten
* center/centralization (tending away from c.): centrifugal; (tending toward c.): centripedal
* centrifuge insert: phial
* change (abrupt): about-face; backlash; break; cold turkey; flit; offset; qualm; quantum jump; revulsion; (study of a. change): catastrophe theory; (tending away from c.): conservative; (tending toward c.): liberal
* church (excessive attachment to): churchianity.
* claims (excessive): pretentiousness; (extravagant): hype
* clarity (extreme): brilliance.
* classical music (styles of): a cappella; accelerando; adagio; adante; ad libitum; agitato; allargando; arioso; baroque; bell canto; bravura; continuo; crescendo; decrescendo; dimiinuendo; grandioso; harmolodic; larghetto; legato; lyric; ostinato; parlando; presto; recitative; toccata; vigoroso; vivace;
* clothes (container for c.): suitcase; (produced by): seamstress; tailor
* coinciding origin: coeval; * common origin: affinity; cognate; family; homologous; isogenous; kindred; phylum.
* comment critically: animadvert
* complexity (increasing c.): anamorphosis
* concern (with appearance - excessive): fastidiousness; finickiness; fussiness.
* conformity (to law): legalism.
* confusion (extreme): nonplus; quandary.
* constrict (tendency to c.): astringent
* construction (excessive): overbuilding.
* control (absolute): See Power.
* control (excessive): micromanaging.
* cordiality (excessive): backslapping.
* cow (female): cow; (male): bull; (young/immature): calf; (breeds of c.): Angus; Africander; Alderney; Ayshire; Belted Galloway; Brahman; Brangus; Brown Swiss; Charbray; Charolais; Devon; Dexter; Durham; Dutch Belted; French Canadian; Galloway; Guernsey; Hereford; Holstein; Jersey; Kerry; Kobe; Lincoln Red; Longhorn; Polled Durham (Shorthorn); Red Polled; Red Sindhi; Santa Gertrudis; Shorthorn; Simmental; Sussex; Welsh; West Highland.
* crime (tending to produce c.): criminogenic
* critical (excessively): hypercritical.
* critical (excessively): hypercritical.
* criticism (excessive): flak
* cruelty (extreme): sadism.
* declare one's homosexuality: out.
* deer (female): doe; hind; (male): stag; buck; (young/immature): fawn
* deference (excessive): kowtow
* degree (excessive): ad nauseum; to a fault.
* demanding (excessively): niggling.
* demanding (excessively): niggling.
* demonstrative (excessiveley): effusive.
* dependence (extreme): slavery.
* desire (extreme): mania; (excessive) (for something): to covet; (for the past): nostalgia; (for wealth): cupidity; greed.
* deteriorate (tendency to): entropy.
* devotion (excessive): idolatry; (to religious practices): ecclesiasticism; (to the Bible): bibliolatry; (to the Virgin Mary): Mariolatry.
* different origin: homograph; import.
* diffuse v. (tendency to): osmosis.
* disgust (something that causes d.): abomination.
* dislike (extreme): hatred.
* distant origin: distal.
* divide (tending to d.): atomism
* dog (types of): shorthair; longhair; (male): dog; (young/immature): pup; puppy; (associated with): bone; (female): bitch; (kinds of) Afghan hound; Airedale; Alaskan malamute; American water spaniel; basenji; basset hound; beagle; bloodhound; border collie; boxer; Chihuahua; chow chow; cockapoo; cocker spaniel; collie; dachshund; Dalmatian; Dandie Dinmont terrier; Doberman pinscher; English bulldog; fox terrier; German shepherd; golden retriever; Great Dane; greyhound; Irish setter; Irish water spaniel; Labrador retriever; Mexican hairless; Pekingese; Pomeranian; poodle; pug; Saint Bernard; Samoyed; schnauzer; Scottish terrier; Shih Tzu; Siberian husky; silky terrier; springer spaniel; terrier; toy poodle; weimaraner; Welsh corgi; Yorkshire terrier.
* drafting instruments: compass; dividers; French curve; pantograph; perpendicular; plotter; protractor; square; trammel; triangle; T-square.
* drowsiness (abnormal): lethargy
* dry (excessively): arid.
* dry (excessively): arid.
* duck (female): duck; (male): drake; (young/immature): duckling
* earthenware pot (instances of): ampulla; costrel; crock; delft; jar; jug; olla; pipkin; stein.
* eating (excessive): gorging; (excessive fondness of): gourmandism.
* eggs (container for): carton
* emotion (excessive expression of): emoting; gushing; hysterical; neurotic.
* emotional distress (extreme): misery.
* empirical (tending toward): Aristotelian
* end (abrupt): flameout; (extreme): tip.
* enthusiasm (excessive): fanaticism; foolishness; mania; (to show): to mob.
* equilibrium (tendency towoard): homeostasis.
* excessive: e. admiration: adulation; e. alcohol consumption: intemperence; drink; alcoholism; dissipation; e. amount: overdose; surfeit; surplusage; e. a