Robert Burchfield is a man of many words: 62,750 of them, to be precise.

Exultant should be one of his favorites, now that the mammoth task of editing the four-volume, 5,750-page supplement to the Oxford English Dictionary is complete.

"There's both a sense of relief that I've gotten there and a feeling of release from an immensely pleasurable sentence," says Burchfield, who is coming here for a symposium on the OED at the Library of Congress, including a public lecture at 8 p.m. tomorrow.

Working on a dictionary, he adds, is a lot like life: "You start out with nothing, you end with nothing, and along the way you have adventures."

It also, like the best of lives, lasts a long time.

In 1957, when Burchfield was asked by Oxford University Press to compile a volume listing all the words that had entered the language since the original, 15,487-page OED was finished in 1928, he expected it to take seven years. Instead, it took 29.

For a dictionary, that's a mere blip on the scale.

"A corresponding German dictionary, the Deutsches Wo rterbuch, took 120 years to finish," Burchfield points out. "Work began on The Dictionary of the Older Scottish Tongue in 1929, and it had reached the word pnewmatick by 1985. And the Middle English Dictionary, which was started in the 1920s, reached rereward last year."

Such immense amounts of time are necessary because of a dictionary's high degree of labor-intensiveness, and the OED -- the final authority on the English language since the first of its 12 volumes appeared in 1884, and considered one of the greatest scholarly achievements of mankind -- has been one of the most laborious.

The supplement, for instance, employed at its peak 100 readers who functioned as lexicographic detectives -- searching through books, magazines and newspapers for any new word not covered in the original OED, or any old word that had acquired a new meaning.

It's a bookworm's idea of heaven -- getting paid for what you'd be willing to do for free. "It's quite a nice job to do in front of the fire at home," Burchfield agrees, "although the copying out of the citations is a bit of a bore."

Choosing which books to read was scarcely a problem. "If Gore Vidal puts out another novel this year, it will be systematically combed before the year is over," says the 62-year-old lexicographer. "But if you get down to the level of tender romances -- well, we don't read those."

Deciding which words to include in the supplement was trickier. The basic rule was the likelihood that it was a permanent acquisition to the language, and not just a passing fad. There are, however, exceptions to this.

"If you happen to be T.S. Eliot -- that is, one of the greatest poets of the 20th century -- you receive from the OED a kind of concordance rule -- every word that you use is recorded," says Burchfield. So smokefall ("the moment when the wind drops and smoke that has ascended descends") is included in the supplement, even though its only recorded use is by Eliot in his Collected Poems.

About 100 writers, ranging from Shakespeare, Spenser and Milton to Dylan Thomas, W.H. Auden and James Joyce, receive this honor. But Joyce is a special case in himself; most of Finnegans Wake is excluded.

"Everything in it is invented. Joyce is only quoted from if the sentence makes consecutive sense out of context. So riverrun [the course that a river shapes through its landscape], yogibogeybox [the paraphernalia of a spiritualist] and weggebobble [an alternate for vegetable] made it in."

These are not, in all likelihood, words being uttered at cocktail parties across the land. But usage, according to Burchfield, isn't the point.

"The dictionary actually could have been expanded even further," he says. "The language itself is virtually limitless. The OED is getting the words at the central core, but it can't pretend it's mapping the whole."

The readers' citations -- the supplement contains 527,500 of them, the original 1,827,306 -- are continuously collected on 4-by-6 cards. The dictionary prints the earliest quotation for each meaning, then a range bringing out nuances, and finally the latest.

It's a painstaking process that, as English constantly reshapes itself, must combat the encroachments of the years. All the time, Burchfield felt the pressure. "It's always been dismaying to see the new waves of words flowing in, even as you are recording the last wave."

That problem also bedeviled James Murray (1837-1915), the first guiding force behind the OED. Like Burchfield, who grew up in the New Zealand town of Wanganui, Murray was born on the fringes of the British Empire -- the Scottish village of Denholm, where the standard English of London was literally unheard-of.

He was a self-taught, brilliant man who was abundantly supplied with confidence and stamina. In his 35 years of work on the dictionary he needed both. In the initial phase, Oxford University Press often tried to cut corners and speed up publication, and Murray was forced to scrape by on a pittance.

He constantly had to put up with other tribulations, like the time one of the first sub-editors in charge of the Pa segment lost the assembled quotations (it was eventually discovered they had been used for lighting fires). No wonder Murray complained of "never-ending toil" and nearly had a nervous breakdown.

Burchfield had an easier time of it, but he frequently heard he was recording a language that was losing its variety and richness. Television and travel were killing dialects, it was proclaimed; regional speech patterns were disappearing.

The lexicographer will have none of it. "That's pure mythology. People are grimly keeping to their local forms of speech."

In fact, he says, a major development in linguistics has been the discovery of inner city dialects, in America as well as Britain. "Every city is a tribe, and has got subtribes within tribes. People in cities are proud of belonging to subgroups, and one way they distinguish themselves from each other is linguistically."

Another recent development in English has been the shift of the language's "center of gravity" to the United States. Partial evidence of the vibrancy and interest in American English can be felt in the offices of Webster's New World Dictionaries in Cleveland.

Letters come in demanding corrections ("I have found a funny mispelling sic in your dictionary," wrote a Taiwanese correspondent. "For a reward of U.S. $500.00 I shall mail you the page and point out the mispelt word"), improvements ("In the interest of euphony, the present spelling of dawdle should be changed to dwadle, i.e. dwindle") and miracles (a man wrote in asking the origin and meaning of aquapoint, a word he had dreamt about).

"I'm overcome by the passion that so many people feel in writing about language," says David Guralnik, Webster's editor in chief emeritus. "Letters come from highly educated people and from those who stopped in the fifth grade, from children and old people. At one time or another, everyone seems to write to the editor of their dictionary."

Sometimes they write because they want to get in the dictionary themselves. A sampling of new coinages submitted to Webster's: ensmallment (a 9-year-old's antonym for enlargement); revulate ("to revolve and rotate at the same time," according to its author); postjudice (the antonym of prejudice); artviera, flutorno, umbelre, vastnily and others ("No definitions as of now," their coiner reported).

These would-be neologists don't realize, Guralnik says, that "the dictionary is the last in line. Words are coined, they're used, and if they catch on, we'll record it. But our job is not to coin them ourselves."

Even disregarding these misbegotten efforts to imitate Joyce, Guralnik agrees with Burchfield that English, far from being homogenized, is evolving into an ever-richer brew. His explanation has to do with the fundamental illogic of all language.

"Most linguists would agree that writing is another way of recording what is essentially speech. Now the ear, sensitive as it is, does mishear many things, and as a result, the spelling of words very often get changed."

Take the word minuscule. Out of 500 examples of current usage in the Webster's citation files, over 40 percent spell it with two "i's." "It's often pronounced miniscule, and proofreaders are just not noticing when it's spelled like that," says Guralnik. "There's a very good chance the word will eventually end up always being spelled and pronounced miniscule. And if it does, so what? Half of our language is because of this."

Changes in spelling can arise from visual as well as auditory confusion. Chaise longue is simply French for long chair. But because many people aren't familiar with French but know what a lounge is, it's often spelled chaise lounge -- a form that Webster's now sanctions.

The role of dictionaries, both scholars agree, is not only to define the present, but to keep in touch with what has gone before.

"If you say the word broadcast now," says Burchfield, "the first thing that comes in your head is television or radio. But if you used the word 60 years ago, you would be talking about spreading seeds in a field."

The modern broadcast would never have existed without the earlier use, he notes. "Past meanings are part of the present. If you cut off your origins, it's like cutting off your arm."