Study of the Vocabulary
It will, perhaps, be most convenient to begin with that aspect of the dictionary which makes it the reverse of the grammar. From this point of view we have already defined a dictionary as a collection of the isolated phenomena of a language - those which cannot easily and conveniently be brought under general rules. It follows from this that the main function of a dictionary is to give the meanings of separate words. Some dictionaries confine themselves strictly to this function. But a dictionary which does not sacrifice everything to giving as large a vocabulary as possible in the shortest space ought to give a good deal more than this.
Idioms fall entirely within the province of the dictionary, because the meaning of each idiom is an isolated fact which cannot be inferred from the meaning of the words of which the idiom is made up: a dictionary which explains the meaning of do without explaining that of How do you do? is useless as a guide to the meanings of words.
A thoroughly useful dictionary ought, besides, to give information on various grammatical details, which, though they fall under general rules of grammar, are too numerous or too arbitrary and complicated to be treated of in detail in any but a full reference-grammar: such a dictionary ought to give full information about those grammatical constructions which characterize individual words, and cannot be deduced with certainty and ease from a simple grammatical rule. Thus it ought to give full information about the prepositions by which verbs are connected with the words they govern (think of, think about, think over, part from, part with). (...)
As convenience of reference requires that a dictionary should be as little bulky as is consistent with efficiency, it is advisable that its scope should be
distinctly defined and strictly limited, A dictionary of English for practical use by foreigners, or a French or German dictionary for practical use by English speakers, is, in the nature of things, mainly a dictionary of the present stage of these languages: its foundation is the modern colloquial and literary language, which involves, of course, the inclusion of a certain number of archaic words used in the higher literature, together with a certain amount of slang and vulgarisms and those dialectal words which, have found their way into general literature and conversation. (...)
Most of our larger English dictionaries, ;are also compromises between an expanded dictionary and an abridged cyclopedia. The fundamental distinction between a dictionary and a cyclopedia is, that the dictionary has to explain words, the cyclopedia has to explain things. The main function of the dictionary is to identify each word with its meaning or meanings, and give the details of its linguistic use as far as they do not fall entirely and exclusively under the province of grammar. This is clearly shown in the use we make of dictionaries of foreign languages. If we are ignorant of the meaning of the French word fleur, we look it up in our French-English dictionary, where we find the English translation 'flower', without any further comment, it being assumed that we know what a flower is. We feel that the translation is a surer guide to the meaning than the most elaborate definition. In an English dictionary for English people the same method of translation is followed as far as possible: commence and purchase are defined by being translated into the simpler 'begin' and 'buy', and we fall back on definition only when absolutely obliged to do so. Some of the more naive among the older dictionaries openly give up the attempt to define by such evasions as telling us that dog is 'the name of a well-known animal'. Even Walker's celebrated definition of a flea as 'a small insect of remarkable agility' would be of little use to any one who did not know already what a flea was.
But it may happen that in reading French we come across the name of some flower that is not found out of France, or, at any rate, not in England, so that when we look up the word in the French dictionary, the only explanation we find is 'name of a flower' with, perhaps, the botanical name, which probably conveys no meaning to our minds; we have not, therefore, learnt anything from the dictionary beyond what we could probably have gathered from the context without any further help. Nevertheless, the dictionary has done everything in its power to identify the word with the thing expressed by it; it is our want of knowledge of the thing itself which prevents us from profiting by the dictionary's identification. If we look up the botanical name in a cyclopedia, we can acquire a more or less definite idea of the thing itself- the flower.
There can be no question of the usefulness and convenience of the brief explanations of the ideas and objects expressed by rare words which our larger dictionaries give: these explanations afford the reader enough information to
enable him to form an idea of the real nature of the thing represented by the unfamiliar word without obliging him to wade through a sea of detail.
. But it is a question whether it would not be better to publish such information in a separate book than to mix it up with the legitimate material of a dictionary - namely, the identification of familiar ideas with the words which express them. An educated Frenchman just beginning English is ignorant of the meaning of the commonest verbs and adjectives in English, but he will not require to be told what oxygen is, or how lithography is carried on. It is not meant that these words should be excluded from a practical dictionary; on the contrary, they are examples - especially the latter - of a numerous class of words which form a debatable ground between necessary, everyday words and purely special and technical words.
A further reason for separating the special or encyclopedic from the general or lexical words lies in the different treatment they require. While the former demand, or, at least, allow, a more or less elaborate and lengthy description of the thing they denote, accompanied, perhaps, with pictures or diagrams, they are generally barren from the linguistic point of view, for they offer neither varied shades of meaning nor irregularities of form, nor do they enter into idiomatic combinations or special grammatical constructions. With the lexical words the relations are reversed: the greater the number of irregularities of form a word offers, and the more complex and varied its meanings and idiomatic combinations and special constructions are, the more indispensable for expressing ideas, and the more independent of encyclopedic treatment it is sure to be.
We arrive, then, at the result that for purposes of practical study of modern languages we require dictionaries which are strictly limited to the modern language, and exclude all encyclopedic elements - that is, all words of which it is conceivable that an educated native might say that he had never seen them in literature or that he did not know what they meant. Such a dictionary would, of course, include debatable words, unless it were intended for very elementary purposes, in which case it might exclude even such words as abacus, habeas corpus, iambic, nabob, oxygen.
But it would be very difficult to lay down any general principles by which we could exclude all encyclopedic words without hesitation, and the ordinary compromise has its practical advantages. (...)
The first business of a dictionary is to give the meanings of the words in plain, simple, unambiguous language. There must be no 'etymological translation1, no translation into obsolete or dialectal words. When we look up Isece in an Old-English dictionary and find it translated 'leech' as well as 'physician', we ought to be quite sure that leech here has its genuine modern meaning, and is not a mere repetition of the meaning of the other word.
Again, some dictionary-makers think it necessary to translate every slang
or colloquial word or expression in one language into a slang word or expression in the other language. The result is that they sometimes use some provincial or obsolete word or expression which may be quite unintelligible to the majority of their readers, and, indeed, may soon become unintelligible to all of them, for nothing becomes obsolete sooner than a certain class of slang colloquialisms. Most languages are so ambiguous in themselves that it is folly to go out of one's way to make them more so; and in a dictionary everything is detached and isolated, so that there is but little context to help. In fact, without the help of quotations it is almost impossible to define meanings with certainty. (...)
Quotations are next in importance to definitions. Indeed, in a large dictionary or thesaurus, the quotations are the, dictionary, and their arrangement is a matter of almost subordinate importance. They cannot, of course, be given with any great fullness in most short dictionaries. But in some cases a quotation is both shorter and clearer than a definition. All sentences that have anything of the character of proverbs or formulae deserve a place in every dictionary. Such sentences, indeed, can hardly be regarded as quotations, any more than idioms, which are as much a part of the common stock of the language as the words themselves: like them, they cannot be constructed a priori. (...)
In the first place, it must be borne in mind that the ultimate ideas of language are by no means identical with those of psychology, still less with those of metaphysics. Language is not in any way concerned with such psychological problems as the1 origin of our ideas of space and matter; for at the time when language was evolved, these conceptions were already stereotyped in the form of simple ideas, incapable of any but deliberate scientific analysis. Even such universally known facts as the primary data of astronomy have had little or no influence on language, and even the scientific astronomer no more hesitates to talk of "the rising to the sun” than did the astrologers of ancient Chaldýea. Language, in short, is based not on things as we know or think them to be but as they seem to us. (...)
At first, the meanings of words will be learnt mechanically one by one by associations with their context. In every language there are a certain number of words which the learner remembers at once, either because they are borrowed from or are cognate with words already familiar to him in his own or some other language, or through some chance resemblance to known words. These words are, as it were, centres round which other words crystallize, each new association leading to further associations, till at last the chief part of the elementary vocabulary of the language forms a solid mass of associations each connected in various ways with others.
From: The Practical Study of Languages by Henry Sweet