WORD A fundamental term in both the general and technical discussion of language. The following selection of primary definitions of word is drawn from two American and two British works:(1) Webster's Ninth New Collegiate Dictionary (1984): ‘a speech sound or series of speech sounds that symbolizes and communicates a meaning without being divisible into smaller units capable of independent use’.(2) American Heritage Dictionary (1985): ‘a sound or a combination of sounds, or its representation in writing or printing, that symbolizes and communicates a meaning and may consist of a single morpheme or of a combination of morphemes’.(3) Collins English Dictionary (1986): ‘one of the units of speech or writing that native speakers of a language usually regard as the smallest isolable meaningful element of the language, although linguists would analyse these further into morphemes’.(4) Chambers English Dictionary (1988): ‘a unit of spoken language: a written sign representing such an utterance’.

Literacy and the word

In ancient Greece, word study was an inseparable part of the study of texts, which had been prompted by the invention of script. Prior to that, NAME was a more clearly delineated concept than word, which was rather imprecisely associated with speech (apparently the meaning of *wer, the Indo-European root underlying Latin verbum, Sanskrit vrátam, and English word). In oral communities, there appears generally to be no great interest in separating out ‘units’ of language, a lack of delimitation carried over into the early stages of alphabetic writing, in which letters followed each other in lines without spaces to separate off what are now perceived as ‘words’. Spaces between groups of letters became important as the conventions of writing evolved. In alphabetic systems, spaces are now universal and as a result literate people learn to recognize ‘words’ as visual rather than auditory units.

Children learn about words while learning to write, become more or less comfortable with ‘the written word’, and may later assume that words as they are written automatically have a place as theoretical units of both speech and script. Grammarians, philologists, and linguists, all the legatees of the Greeks, have tended to focus on words as visual entities even when analysing sound, for which phoneticians developed a special alphabet. The place of the word as an ultimate unit of language has not, however, been easy to find, with the result that many 20c linguists have found it necessary to look elsewhere for key units of language: ‘below’ the word among phonemes and MORPHEMES or ‘above’ it in sentence and discourse.

The word in different languages

The nature of words varies from language to language: amaverunt is one word in Latin, but cannot be translated into one word of English (in which it means either they loved or they have loved). It is a verb with a root am-, a THEMATIC VOWEL -a-, a marker of the perfect tense -v-, and a complex inflectional ending -erunt. The English verbs that translate it have different tenses (simple past they loved, present perfect they have loved) and bear not the slightest resemblance to the structure of their Latin equivalent. Different from both English and Latin, Swahili has a primary verb form such as kuta (meet). The forms kutana (‘meet each other’) and kutanisha (‘cause to meet each other’) may be conceived as either variations of kuta or as distinct words. In effect, the conception ‘word’ is determined afresh within the system of every language, and as a result the word-as-element-of-speech is language-specific, not language-universal. The various kinds of language have their own broadly similar words, but even so there is variation from language to language inside a category: for example, among Romance languages between French and Spanish.

Nine kinds of word

Despite such complications, however, certain features are more or less true for many if not all languages. Nine such features are fundamental to English and each has its own ‘word’.

The orthographic word.

The word understood in terms of alphabetic or syllabic writing systems: a visual sign with space around it. It may or may not have a canonical form: in the 14c, before print encouraged standardization, merry was also spelled myry, myrie, murie, and mery. On occasion, the orthographic word has canonical forms for different varieties within English: BrE colour and AmE color (‘the same word’ in two visual forms).

The phonological word.

The word understood in terms of sound: a spoken signal that occurs more commonly as part of a longer utterance than in isolation and is subject to rhythm. Traditional spoken English is a series of stressed and unstressed syllables which behave in more or less predictable ways: where an experienced listener hears It's no good at all being pronounced in a relaxed, informal way, a foreigner may hear Snow good a tall. In the flow of speech, words do not have such distinct shapes as on paper, and syllable boundaries do not necessarily reflect grammatical boundaries: the phrases a notion and an ocean are usually homophonic and only context establishes which has in fact been said.

The morphological word.

The word in terms of form lies behind both the orthographic and the phonological word: big has a spelt-out realization b-i-g and a spoken realization /big/, but is independent of both, because it can be expressed in either medium and also in sign language. This entity is capable of realization in different ‘substances’; it is distinct from such spelt-out variants as colour and color as well as from the innumerable ways in which African, American, Australian, Caribbean, English, Irish, Scottish or other people may say ‘colo(u)r’. However, all such users have it in common and it is the basis of such further forms as colourful and discoloured.

The lexical word

(also called a full word, content word, lexeme, lexical item). The word in terms of content relates to things, actions, and states in the world. It is usually realized by one or more morphological words, as when do, does, doing, did, done are taken to be five ‘versions’ of the one verb DO. Lexical words are generally fitted into the flow of language through such mechanisms as affixation, suppletion, stress shift, and vowel change, all of which have morphological and other effects. The set of such words is always open to new members, and in English embraces nouns, verbs, and adjectives, and other parts of speech when they behave like nouns, verbs, and adjectives, as in ‘But me no buts’. Lexical words may be simple in structure (cat, mouse), or composite (cold-bloodedness, incomprehensible, teapot, blackbird, Commonwealth, stamp collector, put up with, natural selection, Parkinson's disease).

The grammatical word

(also called a form word, function word, structure word, and in some theories a subvariety of morpheme). The word in terms of syntactic function contrasts with the lexical word and is an element in the structural system of a language. It serves to link lexical words. In English, conjunctions, determiners, interjections, particles, and pronouns are grammatical words. They occur frequently and have their own semantic systems, as with such particles as up and down, which relate to position, direction, space, and time. In principle, such words are a closed set to which new items are seldom added. As lubricants, grammatical words are like affixes: the out in throw out is like the prefix e-in eject; the before in before the war means the same as pre-in pre-war. They can also function like affixes, as in he-man and yes-man.

The onomastic word.

The word in terms of naming establishes special, often unique reference: the difference between Napoleon and emperor. It may be simple like Smith or complex like Smithsonian. Names may be motivated, like Sitting Bull (a Sioux name derived from an omen involving a bull buffalo) or conventional, like Smith today (though not in the Middle Ages, when the name was occupation-based). Although such words are lexical, they are not usually listed in dictionaries and may or may not be relevant in encyclopedias. They are often regarded as apart from normal vocabulary, though they too have to be learned.

The lexicographical word.

The word in terms of dictionaries is usually presented in an alphabetic setting. Many dictionaries have an entry did as the past of do, an entry them as the object form of they, and so on, with cross-references to the representative form. There are therefore two kinds of entry: anything the compilers think anyone might look up, and the citation forms under which definition proceeds. The conventional citation form for nouns is the singular (unless a word is always plural) and for the verb is the bare infinitive (unless the verb only occurs as a participle, or is a modal verb).

The statistical word.

The word in terms of occurrences in texts is embodied in such instructions as ‘Count all the words on the page’: that is, count each letter or group of letters preceded and followed by a white space. This instruction may or may not include numbers, codes, names, and ABBREVIATIONS, all of which are not necessarily part of the everyday conception of ‘word’. Whatever routine is followed, the counter deals in tokens or instances and as the count is being made the emerging list turns tokens into types: for example, there could be 42 tokens of the type the on a page, and 4 tokens of the type dog. Both the tokens and the types, however, are unreflectingly spoken of as words.

The translinguistic word.

The word in terms of several distinct languages in which versions of the same form exist: for example, realitas in Latin, réalité in French, realidad in Spanish, and reality in English. In one sense, these are not the same: they are all words in separate languages. In another sense, however, and one which is well understood by teachers and travellers, they are all ‘the same word’; that is, a high degree of textual continuity survives across linguistic divides. It is a feature of such words, however, that no single omni-representative form can be cited for any such word: all the embodiments are linguistically equal and, in their various ways, subtly or significantly different.

Other ‘words’

In addition, there is a large number of more or less common expressions, some technical, some semi-technical, some general and casual, all specifying kinds of words and word-like units. They fall into overlapping groups that include: (1) Terms in which word appears, such as: BASE WORD, BUZZ WORD, COMPOUND WORD, LONG WORD, ROOT-WORD. (2) Terms based on the suffix -ISM, such as: AMERICANISM, AUSTRALIANISM, BURGESSISM, MALAPROPISM. (3) Terms based on the combining form -ONYM, such as: ANTONYM, aptronym, characternym, EPONYM, HYPONYM, SYNONYM. (4) Terms that relate to form more than meaning, such as: abbreviation, ACRONYM, COMPLEX WORD, compound word, INITIALISM, portmanteau word. (5) Terms that relate to meaning more than form, such as: ANTONYM, burgessism, eponym, HARD WORD. (6) Terms that relate to social usage, such as: ANAGRAM, buzz word, CONFUSIBLE, LOANWORD, MALAPROPISM, NONCE WORD, PALINDROME, STUNT WORD, VOGUE word. All such terms fit in various ways and at various levels into the model of the word presented above.

Words as clusters

Because of its many dimensions, the concept ‘word’ is more like a cluster than an atom. On the level of theory, the cluster contains the kinds of words discussed above. On the level of practical activity, people ‘know a word’ not simply when they can use and understand a single item but when they know a range of variation and practices associated with it: for example, to know the word know entails knowing how to say, hear, read, and write its various forms and extensions, fitting them into phrases and sentences (knows, knowing, knew, known), relating the simple to the complex (as in knowledge, knowledgeable, unknowing, unknowable, unknowably, unknown), relating these to such compounds as knowhow and know-all, managing idioms (y'know, in the know, know the ropes, know what's what, know a thing or two), using and grasping senses, expressions, and collocations (knowing someone or something, knowing how to do something, knowing better, and even knowing ‘in the Biblical sense’). This cluster, with its clear centre and hazy periphery, shares semantic space with other clusters cited as the words understand, perceive, grasp, and fathom. All operate within a system whose size and complexity defy comprehensive description, but without being beyond the reach of the everyday user of the language.