Chaucer spelled it with a u, as indeed did most people until the end of the
seventeenth century, and some for half a century or so after that. But then, as if
by universal decree, it just quietly vanished. No one seems to have remarked on
it at the time. Bernstein suggests [in Dos, Don'ts and Maybes of English Usage,
page 87] that it may have reflected a slight change in pronunciation—to this day
many people aspirate four and forty in slightly different ways—but this begs the
question of why the pronunciation changed for the first word and not for the
second. In any case, it would be most unusual for the spelling of a word to
change to reflect such a minor adjustment of pronunciation.
Usually in English we strive to preserve the old spelling at almost any cost to
logicality. Take ache. The spelling seems desperately inconsistent today, as
indeed it is. Up until Shakespeare's day, ache was pronounced aitch when it was
a noun. As a verb, it was pronounced ake—but also, rather sensibly, was spelled
ake.
This tendency to fluctuate between "ch" and "k" sounds was once fairly common. It accounts for such pairs as speech/speak, stench/ stink, and
stitch/stick. But ache, for reasons that defy logic, adopted the verb pronunciation
and the noun spelling.
English spelling has caused problems for about as long as there have been
English words to spell. When the Anglo-Saxons became literate in the sixth
century, they took their alphabet from the Romans, but quickly realized that they
had three sounds for which the Romans had no letters. These they supplied by
taking three symbols from their old runic alphabet: w, p, and 8. The first, literally
double u, represented the sound "w" as it is pronounced today. The other two
represented the "th" sound: p (called thorn) and 8 (called eth and still used in
Ireland).
The first Norman scribes came to England and began grappling with what to
them was a wholly foreign tongue—a fact clearly evident in many of the
spellings from Domesday Book. In just one small parish in Yorkshire, Hanlith
was recorded as Hagenlith, Malham as Malgham, and Calton as Colton—all
spellings that were probably never used locally. Many such errors can be
attributed to carelessness and unfamiliarity but others clearly reflect Norman
orthographic preferences. The Normans certainly did not hesitate to introduce
changes they felt more comfortable with, such as substituting qu for cw. Had
William the Conqueror been turned back at Hastings, we would spell queen as
cwene. The letters z and g were introduced and the Old English d and v were
phased out.
The Normans also helped to regularize such sounds as ch and sh, which in
Anglo-Saxon could be rendered in a variety of ways. They substituted o for u in
certain words such as come and one, and they introduced the ou spelling as in
house and mouse. These changes made things more orderly and logical for
Norman scribes, but not necessarily for later native speakers of English.
As we have seen elsewhere, the absence of a central authority for the English
language for three centuries meant that dialects prosered and multiplied. When at
last French died out and English words rushed in to take their place in official
and literary use, it sometimes happened that people adopted the spelling used in
one part of the country and pronunciation used in another. That is why we use
the western England spellings for busy and bury, but give the first the London
pronunciation "bizzy" and the second the Kentish pronunciation "berry."Similarly, if you've ever wondered how on earth a word spelled one could be
pronounced "wun" and once could be "wunce," the answer in both cases is that
Southern pronunciations attached themselves to East Midland spellings.
Once they were pronounced more or less as spelled—i.e., "oon "and "oons."
Even without the intervention of the Normans, there is every reason to suppose
that English spelling would have been a trifle erratic. Largely this is because for
the longest time people seemed emphatically indifferent to matters of
consistency in spelling. There were exceptions. As long ago as the early
thirteenth century a monk named Orm was calling for a more logical and
phonetic system for English spelling. (His proposals, predictably, were entirely
disregarded, but they tell scholars more about the pronunciation of the period
than any other surviving document.) Even so, it is true to say that most people
throughout much of the history of the English language have seemed remarkably
unconcerned about niceties of spelling—even to the point of spelling one word
two ways in the same sentence, as in this description of James I by one of his
courtiers, in which just eight words come between two spellings of clothes: "He
was of a middle stature, more corpulent though in his clothes than in his body,
yet fat enough, his cloathes being ever made large and easie… ." Even more
remarkably perhaps, A Table Alphabetic all of Hard Words by Robert Cawdrey,
published in 1604 and often called the first English dictionary, spelled words two
ways on the title page.
Throughout this period you can find names and words spelled in many ways—
where, for instance, has been variously recorded as wher, whair, wair, wheare,
were, whear, and so on. People were even casual about their names. More than
eighty spellings of Shakespeare's name have been found, among them
Shagspeare, Shakspere, and even Shakestaffe. Shakespeare himself did not spell
the name the same way twice in any of his six known signatures and even
spelled it two ways on one document, his will, which he signed Shakspere in one
place and Shakspeare in another. Curiously, the one spelling he never seemed to
use himself was Shakespeare. Much is often made of all this, but a moment's
reflection should persuade us that a person's signature, whether he be an
Elizabethan playwright or a modern orthodontist, is about the least reliable way
of determining how he spells his name. Many people scrawl their signatures, and
Shakespeare was certainly one of history's scrawlers. In any case, whether he
used the spelling himself or not, Shakespeare is how his name appears on most of the surviving legal documents concerning him, as well as on the title pages of
his sonnets and on twenty-two of the twenty-four original quarto editions of his
plays.
Still, there is no gainsaying that people's names in former times were rendered in
a bewildering variety of ways—some of which bore scant resemblance to the
owner's preferred name. Christopher Marlowe was sometimes referred to by his
contemporaries as Marley. The foremost printer of the Elizabethan age variously
signed himself, in print, John Day or Daye or Daie. Charlton Laird in The Word
cites a man of the period whose name is variously recorded as Waddington,
Wadigton, Wuldingdoune, Windidune, Waddingdon, and many others.
An odd fact of spelling from earlier times is that although writing must have
been a laborious affair there was little inclination to compress words or simplify
spellings—indeed, by all evidence, the opposite was the case. Cromwell
habitually spelled it as itt, not as nott, be as bee, and at as atte, and such
cumbersome spellings can be found in manuscripts right up until the modern
period. It seems curious indeed that people were not driven to more compact
spellings by writer's cramp if not by urgency.