Writing a dictionary: a job that can last a lifetime

When Samuel Johnson wrote his influential English dictionary in the 18th century, the work took him seven years. In that time he had to write the meanings of 40,000 words.

The original Oxford English Dictionary, completed in 1928, took 50 years for its 12 volumes and 252,259 entries.

And in Germany, the 16-volume Deutsches Worterbuch, begun by the Grimm brothers in 1838 and intended to cover the German language from the 15th century to the middle of the 19th, was not completed until 1961  123 years and two world wars later.

Most dictionaries require rather less time and effort, both because they are smaller and because their compilers can use earlier dictionaries as a source of information. A new edition of an already successful ‘desk dictionary’ can take about two years. And a small, specialised dictionary – perhaps a dictionary of abbreviations or a pocket crossword dictionary – could be written by only one person.

To write a dictionary, the lexicographer  (dictionary writer) needs three things: an idea of what sort of dictionary he wants, a style guide to convert the idea into practice, and evidence about which items to include and what to say about them.

First the idea

The dictionary may contain words of all types, or only specialised words (as in a dictionary of chemistry). It may, or may not, include the names of real people and places. It may give many kinds of information about each of its entries (spelling, pronunciation, etymology, meaning, grammatical behaviour, synonyms and antonyms), or only some kinds of information (spelling and pronunciation only, as in a dictionary of pronunciation).

It may, or may not, have pictures and examples of how the words are used. It may be monolingual (with the meanings of English words given in English) or bilingual (with the meanings of English words given in French, say, and those of French words given in English).

It may be for adults or children. It may be large or small, with its size depending on the number of entries and the amount of information given about them.

The style guide

Once the purpose of the dictionary has been decided, rules for writing it have to be drawn up. When is a word to be a main entry or a sub-entry? Will blackbird, for example, be a main entry (like black and bird), or a sub-entry under black, or a sub-entry under bird?

Kick the bucket will probably be a sub-entry – but will it go under hick or under bucket? Will limp (verb `to walk lamely’), limp (noun ‘a lame walk’) and limp (adjective ‘not stiff) all be in one entry (because they all have the same spelling)? Will there be two entries (because two of them have the same basic meaning but the third a different meaning and perhaps a different etymological origin)? Or will there be three entries (because one is a verb, one a noun and one an adjective)?

And if a word appears more than once as a main entry, which order will the entries come in: older before newer, more frequent before less frequent or even adjective before noun before verb?

Or, if a word has more than one meaning, which order will the meanings come in: older before newer, more frequent before less frequent, literal before figurative or general before technical?

The evidence

The starting point for deciding what to include is the lexicographer’s own knowledge of, and feeling for, the language. Ideally, too, he will have a large collection of examples of how words and phrases are actually used, culled from published writing and perhaps from manuscripts or even from recorded speech. This ‘corpus’ should represent as wide a range of language as the lexicographer wants to describe. Both humorous magazines and scholarly works might be examined. Webster’s Third New International Dictionary of 1961 had a corpus of more than 10 million examples; the original Oxford English Dictionary had 5 million.

A computer concordance may be made of every example of every word in the texts chosen for investigation. This would help to ensure that important uses of words are not missed simply because they are too commonplace to attract the lexicographer’s attention.

The lexicographer will refer to other dictionaries and books, and articles about language. He may also consult experts for information about specialist words, and ordinary people about their preferences and reactions to how words are used.

But he will have to interpret all this evidence carefully. If someone writes about `wearing what we used to call a “frock”‘, that suggests that frock is an old-fashioned word. But before labelling it old-fashioned or obsolescent in the dictionary, the lexicographer must check the evidence for how widely frock is still used – and by whom. Is it used unselfconsciously by people under 50? Is it used about modern fashions?

Organising the project

Although it is possible for one person to write a dictionary, most are team efforts.

Members of the team assemble the evidence. Samuel Johnson employed six people to collect quotations for his dictionary. The Oxford English Dictionary needed scores, many of whom gave their services free of charge out of love for the work they were doing – and for the English language.

Lexicographers lucky enough to have quotations will use the quotations in writing the entries. And they must coordinate their work so that, for instance, the entries for anabolism, catabolism and metabolism will relate to one another even though they begin with different letters of the alphabet.

A single entry may be the work of a single lexicographer, but it is more likely to be the work of several specialists: one or more for meaning, another for pronunciation, a third for etymology (the origin and evolution of the word or phrase).

Supplementary material, such as pictures or charts, may be prepared by other specialists. Everything must be checked for accuracy, clarity and consistency.

Today much of the drudgery can be handled by computers. They can process great quantities of evidence, ease revision (as by printing out lists of items previously flagged for potential deletion to make room for new words and meanings), and ensure consistency of treatment (but not accuracy or clarity).

Writing the dictionary

In a normal alphabetical dictionary, related words like at, in and on, or disinterested and uninterested may be widely separated. But the entries can be written at the same time to ensure that the meanings have been correctly contrasted and that they cross-refer to each other.

A general-purpose dictionary will include new words (like AIDS) and new meanings of older words (like alien meaning ‘extraterrestrial’). But it should not overlook older words (like parlourmaid) that other dictionaries may have missed.

Some technical terms may be easier to explain than many everyday words. It is easier, for example, to distinguish a stalactite (which points down) from a stalagmite (which points up) than a cup from a mug. And having done his best with cup and mug, the lexicographer may then wonder why a certain drinking vessel is called a paper cup rather than a paper mug, and be forced to the conclusion that paper cup should be a dictionary entry, too.

In writing definitions the lexicographer will try to strike a balance between clarity and informativeness. If a shrimp is called ‘a ten-legged creature’, everyone will understand. If it is called ‘a decapod’, many readers will have to look up decapod. But if they do so, they will find more useful information, including the fact that shrimps are related to lobsters and crabs, which are decapods, too.

One compromise is to call a shrimp ‘a decapod (ten-legged) creature’. But that requires extra space, which can add up over the whole dictionary, and may reduce the number of entries that can be included.

Considerations of space as well as of the users of the dictionary will also have to be balanced in deciding how much information to include. Should the definition of water include its chemical formula (H2O) and its sea-level freezing and boiling points?

Importance of dictionaries

Despite all the problems, the lexicographer can be comforted by the knowledge that dictionaries are among the most important tools of self-education. They are a kind of embodied memory of the culture in which they are produced, and also a means of gaining access to that culture – even without an official teacher.

Some years ago in Britain, a woman who had been injured by a surgical operation decided to sue for compensation. Before doing so, she prepared herself by spending six months reading medical dictionaries so as not to be mystified by the medical terminology that would be used in court.

She won her case.