In Part I of this essay, the concept representation (onomantic) paradigm of Onomantics was explained, and its five core concepts were described and presented in Schedules IA and IB. To understand the discussion which follows, readers need to refer to it frequently -- a copy can be found in an endnote1. Please remember that each core onomantic concept is represented by a notation number in braces, starting with {1} for representation, and continuing with {1.1} description; {1.2} cue; {1.2.1} tag; and {1.2.2} notation. In Part III, I will discuss some other concepts and terms presented (or ignored) in ISO 1087, but here, in Part II, I shall limit myself to a discussion and comparison of the core concepts of Onomantics by contrast with those prescribed for Terminology and listed in Schedule II. They are copied from ISO 1087 (1990), the last formally printed version of the vocabulary published by the Technical Committee on Terminology of the International Organization for Standardization (ISO/TC37).
NOTE: these notations are enclosed in brackets -- "[...]" -- to distinguish them from those presented in Schedule I. By this means, readers will be able to compare {1.1} description with [4.1] definition to see how two similar yet significantly different concepts are represented (by texts) and systematized (by notations) in Onomantics and Terminology. Throughout this paper, I shall use terminology (in lower case) to identify the vocabulary of a field, but when thinking about the field of study, I shall capitalize the word: i.e., Terminology. The same convention is used to distinguish the name of any field of study from the relevant activity: e.g. Politics (or Politology) vs. the practice of politics; Statistics from statistics; Administration from administration, Art from art, etc.
The comparisons discussed below may seem confusing unless the reader can quickly recall the meaning, for example, of {1.1} in Schedule I and its counterpart, [4.1] in Schedule II. To make sense of a complicated set of comparisons, keep a copy of both schedules at hand while reading this article: bear in mind the fact that each concept is identified by a notation (which locates it in relation to other concepts), a description (a defining text) and also by one or more tags (terms).
Three Key Words. As a preparation for the more detailed analysis to follow, let me call your attention to three key words that appear above as types of "designation" [5.3.1]: symbol, term, and name. Since "designation" is offered as a way to represent a concept, you might be puzzled to understand, for example, why "name" appears in this list, and you might think that some numbers (2 = "two") appear in the chemical term for water, whereas (2 = "second") appears in the notation for "term," i.e. [5.3.1.2] -- this means that some symbols are used as tags (terms) but others are used as notations. These three words can be used by lexicographers to help them decide whether or not to prepare dictionary entries. The rules linked with each of them can be explained as follows:
Symbols: don't prepare an entry for a symbol because it cannot be spelled alphabetically and dictionary entries are almost always alphabetized. For example, you can write entries for ampersand or and per se and but you cannot put & in alphabetical order. Symbols, like "2," often stand for different words, like "two" and "second," each of which has different meanings. Webster's (1991) does not enter symbols, though numbers are used in entries: for example, the entry for digit mentions "1 though 9 and 0" in its definition and, of course, these digits are used to mark the sense numbers in every dictionary entry. In the American Heritage Dictionary (AHD) one will find a list of symbols near the entry for symbol, but this is not a common practice. By contrast, letters of the alphabet (although they may be used as symbols) can be alphabetized and entries for them, therefore, head every section of a dictionary, running from A to Z -- the lexicographic understanding of a symbol, therefore, does not coincide with the terminological meaning of this word defined in ISO 1087.
Terms: lexicographers often use "term" to mean to mean any lexeme or vocabulary entry. As defined in the Oxford English Dictionary (OED), sense 13B, a "term" means "any word or group of words expressing a notion or conception". Webster's Third New International Dictionary (W3) defines "vocabulary entry" as a word or "term" (as man in the street) entered alphabetically in a dictionary..." Thus dictionary entries may be written for any term (word, phrase, or affix) that can be defined as a general concept, but not an object (individual concept).
However, some lexicographers use term for the more specific concept of a "technical" or "scientific" expression. The compilers of an ordinary language dictionary, like the OED, that omits technical terms, need to distinguish between ordinary and technical terms, writing entries only for the former. By contrast, other dictionaries, like W3, include both ordinary and technical terms -- lexeme and polyseme are examples. Consequently, they make no distinction between ordinary and technical terms and these two concepts tend to merge into a fuzzy conception.
Names: lexicographers use names to identify persons, places, institutions, books, buildings, etc., but not concepts. "Grace" can refer to a person who may or may not be "gracious". However, lexicographers differ among themselves in the way they handle names. Many dictionaries, including both the OED and W3, ignore names: they write entries only for terms. However, more flexible lexicographers include entries without definitions: they may identify individuals, like "Raphael," as an archangel or as a Renaissance painter, and they may even add pictures, as in the American Heritage Dictionary. Another option is to list names separately, as in the "who's who" and "gazeteer" appendices found in Webster's New Universal Unabridged Dictionary.
In a strictly onomantic approach based on the quest for suitable means to represent concepts, as discussed in Part I of this paper, one should recognize three main components: a description of the essential features of a concept, a short word, phrase or symbol that could conveniently designate the same concept, and a notation that would place it in the context of a system. The only way I can explain how ISO 1087 happened to come up with the three core items (terms, symbols and names, and ignoring definitions) presupposes their reliance on a term list [6.1.2.6] borrowed from Lexicography, linked to an entry format based on semantic analysis. A discussion of the practices that led to the production of the glossary for terminologists will appear in Part III of this article. Onomantic Comparisons. Here, however, let us compare onomantically the key concepts used by Onomantics and by Terminology. Each of them will be discussed in their onomantic order -- i.e., from {1} to {1.2.2} as presented in Schedule I.
Since three forms of representation are needed for each concept, we need to consider all three of them whenever onomantic comparisons are made. Under each concept heading, therefore, I will use the letters, "A," "B" and "C" to separate the comparisons based on their notation, description and tag. The notations are discussed first because they establish the system of concepts that needs to be taken into account when considering each concept in the schedule. The descriptions provide the most important information because they stipulate the essential characteristics built into each concept. Since tags are, typically, words or phrases with several connotations that can easily create ambiguity, we need a semantic method to highlight these meanings and see how they affect their use2.
By contrast, semantic comparisons only require the analysis of what different words mean -- notations are not involved and the analysis of definitions involves only how well they reflect the actual usages of a word, not whether or not they identify different concepts. Semantic comparisons focus on differences between the meanings of words (lexemes) as reflected in the entries that might be written for them in a dictionary. By contrast, onomantic comparisons explain the differences between concepts as revealed by the records that may be offered for them in a conceptual glossary.
Part I of this article elaborated on the distinction between (semantic) entries and (onomantic) records. Both are involved in the discussion that follows but it is not always necessary to distinguish between them. Consequently, I shall use gloss generically to mean any entry or record found in a glossary3. Whenever the distinction is significant, I shall use "entry" to mean a semantic gloss in which the meanings of a lexeme are defined, and "record" (or "concept record") will always mean an onomantic gloss in which a concept is first identified by a notation, secondly described, and thirdly followed by any tags available to represent it.
Glosses, Forms and Pleonasms. Moreover, I shall use gloss whenever the necessary distinction is not clear: for example, the glosses in ISO 1087 have the semantic format of an entry but, at least ostensibly, the onomantic content of a record. I shall also omit the modifiers, 'semantic' and 'onomantic,' before 'entry' and 'record,' respectively. Moreover, it is often unnecessary to decide whether a gloss is "really" a record or an entry -- when it doesn't actually matter, the use of 'gloss' enables us to avoid arguments about its proper categorization, or the use of an unnecessarily precise term.
It is often useful to have a word for a generic (superordinate) concept that includes two more specific (subordinate) concepts -- using 'gloss' this way is a good example. Consider another available to lexicographers which I have already discussed: they often enter phrases and affixes in a dictionary although, no doubt, most entry words are, indeed, orthographic words. To avoid having to repeat "word, phrase or affix," they can use lexeme (or "term") to refer to all of these linguistic forms. Similarly, 'gloss' enables us to associate entries and records in a single superordinate concept.
Note that lexicographers also use form (or orthographic form) to talk about the way a lexeme is written, i.e. whether it is a 'word,' 'phrase' or 'affix.' We need an even broader notion in Onomantics to refer not only to the form of a lexeme but, more broadly, to include any linguistic form (such as a clause, sentence or even a text) and also non-linguistic forms (symbols, such as letters, numbers, and graphics) that can represent a concept. It seems quite reasonable to use "form" in this sense -- or to modify it with "onomantic form" to refer to the way any notation, tag, or concept description is written -- identifying it as an appropriate way to represent a concept. By contrast, lexemes are "semantic forms" because each represents one or more concepts and is, therefore, appropriately used as an entry word in a dictionary entry: all lexemes are semantic forms, but many onomantic forms are not lexemes.
Speaking of forms leads me to mention that double quotation marks, "...", are used here to mention lexemes whose meanings are being discussed, but when words are mentioned as forms without reference to their meanings, they are enclosed in single quotation marks, '...'. We may, therefore, say that 'form' is a four-letter word, while "form" means the way a lexeme is written, or a concept is represented. When an expression is emphasized or introduced, it is written in bold face or italics, but in subsequent references, quotation marks are used to avoid overemphasis.
Finally, let me say that when one tag does not clearly represent a concept, it is quite appropriate to insert pleonasms -- typically in parentheses. These are other forms that represent the same concept. Sometimes a more precise word is used -- e.g. word (lexeme) -- or a specific form in addition to a genus -- e.g. gloss (entry). The point in each case is to overcome possible ambiguity by using a second term (tag) that can help readers see more precisely what an author has in mind. I call it the pleonastic solution and make frequent use of it in this article.
A. Notation. Since no notation follows 'representation' in the definition of "designation" [5.3.1] we can conclude that there is no separate gloss to define "representation" in ISO 1087, even though the word appears in:
[5.3.1] which defines designation as any "representation" of a concept. Other glosses help us see what the authors of this glossary had in mind. Consider
[5.3.1.1, 5.3.1.2, and 5.3.1.3] which identify three forms of designation: symbols, terms, and names respectively. By definition, therefore, all three forms represent concepts -- see [5.3.1]. However, concepts were defined in 1990 [3.1] as abstractions based on the properties of two or more objects [2.1] and names were assigned to one object only, hence an object could not be a concept. In a subsequent revision, individual concepts were added to the glossary and defined as pertaining to a single object -- concepts were also re-defined as properties of "one or more object". This revision enabled the compilers to claim that a name is also a designation because it can identify an individual concept, as well as an object. Consider also
[4.1] which defines definition as a concept "description" but not as a "representation" of a concept. Although designation [5.3.1] is defined to include "any representation" of a concept, definitions are excluded by their notation number, [4.1], which classifies them as both hierarchically superordinate to and separate from designations. By contrast, since Onomantics clusters all forms used to represent concepts, it has to include concept descriptions (definitions) as one of the forms of concept representation.
In Onomantics the forms that represent a concept include texts which identify its essential characteristics (i.e. {1.1} concept descriptions) by contrast with succinct forms used to designate concepts ({1.2} cues), including both "terms" (tags) and "symbols" (notations). Although the definition of "designation" in ISO 1087 (at [5.3.1]) fails to make this distinction, the notation used clearly shows that the intended concept is that of concept cues {1.2}.
More importantly, the lack of "representation" as a top term in ISO 1087 means that the two basic forms of representation -- "descriptions" and "cues" -- are treated as top terms. Each heads a hierarchy and cannot, therefore, be viewed as coordinates, as two different ways to represent a concept. Instead, they are viewed in ISO 1087 as unrelated to each other. Perhaps they could be viewed, as they are in Logic and Lexicology, as two parts of a whole, i.e. as the definiens and definiendum found in definitions (as this word is normally understood). In a semantic context, the lexeme to be defined is a definiendum, and each of its senses is described by a definiens. The two concepts are parts of a whole: every definition contains a definiendum and one or more definiens.
This logic vanishes in ISO 1087 where definition means a "definiens," and designation a "definiendum". However, the equation is imprecise because designation includes not only definienda (terms) but also notations (symbols). If this explanation sounds complicated, it is. I cannot think of a simpler way to explain the confusion that has resulted from borrowing lexicographic terms (definition and term) and stipulating marginally different meanings for them. Moreover, I cannot imagine a way to clarify the matter by re-defining either of these borrowed terms. The only way to be clear is to start over with a basic recognition of the onomantic context in which concepts can be represented by different forms, including both descriptions and cues, and cues can include both tags and notations. Any effort to use "definition" and "term" as synonyms is sure to cause confusion.
B. The Description. No entry for the most general concept relating to the representation of concepts occurs in ISO 1087 -- as noted above.
C. Tag. Although the word, 'representation,' is found in the gloss for designation [5.3.1] as noted above, it is not given as a tag and we cannot, therefore, determine just what the glossators of ISO 1087 had in mind when they spoke of concept "representation," although their notations (as indicated above) suggest that they considered designations (cues) to be representations of concepts, but not definitions (descriptions).
A. The Notation. The concept [5.3.1] designation is defined -- see Schedule II -- as "any representation of a concept" but the record for definition is located at [4.1], indicating that the compilers of ISO 1087 do not consider that a concept's definition can represent it --as explained above. Yet surely the starting point for representing any concept involves identifying its essential characteristics. Without a concept description, we do not know precisely what we are talking about. To represent the concept of a computer device used to control the action under Windows, for example, we need to see it in operation or describe it -- only after we know what we are talking about can we accept any convenient word, like "mouse", as a way to represent it. That is why, in the onomantic perspective set forth in Schedule I, two basic forms of concept representation are given: {1.1} descriptions composed as texts and {1.2} concept cues, i.e., short forms like words or symbols.
B. The Description. A concept that resembles {1.1} is described in [4.1] definition. However, an important difference should be noted. It involves the stipulation in the definition of [4.1] that "definitions" must differentiate concepts "from other concepts within a system of concepts". No doubt concept descriptions often do specify such relationships but does their absence disqualify a description?
Consider the description of the concept of an "object" in [2.1]: it does not mention any relationship to other concepts within this glossary, and yet the authors of ISO 1087 handle it as a "definition". For me, the identification of related concepts is an accompanying but not a necessary characteristic of a definition: it is like defining birds as a flying animal even though some birds (like an ostrich) cannot fly. Birds can be defined as "feathered vertebrates" but their ability to fly is an accompanying ("accidental") characteristic, not an essential defining one. It is, therefore, an error to include accidental characteristics in the definition of "definition".
Actually, most (though not all) of the definitions in ISO 1087 do specify some relationships to other concepts. For example, the definition of "concept" [3.1] includes "object [2.1]" in the text, thereby indicating how concepts are related to objects. However, the most important device for showing how concepts are related to each other is a classification scheme with notations which locate concepts within such a scheme -- concept descriptions provide supplementary information but they cannot substitute for notations.
Despite the definition offered in [4.1], therefore, definitions do not need to show concept relationships. This is true of almost all dictionary definitions, but definitions in a terminological (conceptual) glossary often specify relationships with key concepts in a formal way, namely by including within each defining text a notation for each of its key terms. Each such term leads users to the definition in another entry in the same glossary which identifies a "key concept," i.e. one that is closely related to the concept being defined. For example, the definition of [5.3.1] designation (which I quoted in the previous section) reads: "Any representation of a concept [3.1]". If the notation, [3.1] were omitted from this definition, it would still be a "definition" but here the relationship of "designation" to "concept" is indicated by marking it as an "entailed term" -- I'll discuss this term in more detail in Part III of this article.
No doubt, in a general way, all the words used in any definition reflect relations between concepts, but only the words or phrases that are marked with a notation symbol clearly refer to other concepts that have been defined in the system of concepts glossed by the entries in ISO 1087 -- i.e., a "structured set of concepts," not a random collection of associated ideas. Moreover, consider that in alphabetized dictionaries (and glossaries) the key word used in a definition can be found as an entry word in alphabetical order. By contrast, in a conceptual glossary whose records are headed by notations, one must have the notation in order to find the related record.
The focus of interest in any conceptual glossary is, surely, on relations between the concepts defined in that glossary. If we think about each gloss in a conceptual glossary as a record that identifies a single concept, then we can assume every concept defined in a glossary's system of concepts must have its own gloss. To say that a definition permits the differentiation of any defined concept from "other concepts within a system of concepts" can be meaningful only if the definition identifies, by a notation, the location of the glosses that define each concept within the system. (In hypertext, no doubt, entailed terms in a definition can support jumps to the relevant gloss, but that is a refinement we need not consider here. Although relations between concepts can also be shown by indentations in lists, and by graphics, these techniques supplement notation numbers which are still the most efficient and commonly used technique to display relations between concepts).
Of course, most words used in definitions bring related concepts to mind. This includes all the definitions found in ordinary dictionaries where it is rare to find explicit links between concepts defined in different entries. Dictionary definitions seek to explain the meanings of a word but they do not claim to link their definitions with each other in any systematic way. Consider, here, a useful distinction between interdependent definitions which do link concepts by the explicit marking of entailed terms -- as illustrated by "object [2.1]" in the definition of "[3.1] concept" -- and independent definitions which do not contain such explicit cross-references4.
We could say, then, that most of the concepts defined in a conceptual glossary ought to have interdependent definitions, whereas ordinary dictionaries normally supply only independent definitions. However, to say that a text is not a "definition" unless it provides cross-references to other concepts defined in the same glossary is surely misleading even though, of course, glosses in a conceptual glossary ought to provide such information whenever a given record is, indeed, closely related to another concept defined in the same work.
C. The Tag. When we compare the words description and definition we find other differences that are not essential for the defined concepts -- rather, they pertain to the connotations of the words used to represent them. In Logic, definitions always have two components: a definiendum to be defined, and a definiens that defines it. As used in Lexicography, a "definition" is a "formal statement of the meaning of a word" (see Hartman 1990 and Robinson 1983) -- this usage is consistent with the specifications found in Logic. According to Sager, "In terminology it is customary to restrict the use of 'definition' to the explanation of the accepted specialized meanings of lexical items [lexemes] the occurrence of which can be documented in a variety of sources" (1990, p.40). Sager's definition of definition, therefore, is consistent with the established usage in Logic and Lexicography -- he merely adds the limitation that Terminology is only interested in defining the terms found in special languages, rather than looking at all the lexemes of an ordinary language, as lexicographers do.
In ISO 1087, by contrast, a new meaning is stipulated for "definition" -- it is defined as the description (a definiens) of a concept without any mention of a term (lexeme) to be defined (a definiendum). Such a concept description lacks a subject and predicate and therefore should not be written as a sentence, yet the definitions in ISO 1087 (1990) start with a capital letter and end with a period -- this error has been corrected in subsequent revisions. More importantly, 'definition' is used here as a metaphor to represent something significantly different from what it ordinarily means. We are, of course, familiar with many metaphors in daily use: a computer mouse that controls the cursor on a Windows screen is a metaphor whose new meaning differs substantially from its ordinary sense as a kind of rodent. 'Mouse' is a viable metaphor because the semantic distance between rodents and computer tools is so great that no ambiguity occurs, but when a metaphor represents a new concept that closely resembles the original meaning of the word, ambiguity becomes unavoidable.
Such ambiguity is intensified when members of a discourse community are themselves divided in their usages: when a terminologist like Sager uses definition as this word is normally understood but ISO 1087 stipulates a metaphoric new meaning for the word, as a neologism, confusion is sure to arise. My claim that metaphors are neologisms, incidentally, is supported by ISO 1087 which defines neologism [4.22] as a "term that is newly coined or recently borrowed from another language or another subject field". To re-define "definition"; as a definiens without a definiendum is, surely, a metaphoric neologism.
By contrast, the notion of a concept description is clear and understandable -- it is even supported by the definition of "definition" offered in ISO 1087 at [4.1]. The prior meaning of "definition" hampers the metaphoric use of this word in Terminology -- most terminologists, I suspect, do not remember the small but crucial difference between ordinary definitions and terminological definitions, but they could easily remember the concept if they were to think of it as a "concept description," and to recognize it as one of the three main forms used to represent concepts.
|| Return to topof Part II | continue to second segment of Part II | return to onomantics page ||Revised 26 July 1996