by Bruce Byfield

Most technical writers are confused about grammar. On any day on the TECHWR-L list, basic questions are asked: "Is 'User's Guide' or 'Users' Guide' correct? Maybe 'Users Guide?'" "Should 'web' be capitalized when used to refer to the World Wide Web?" "Which is right: 'A FAQ' or 'an FAQ?'" Many of these questions become the major thread on the list for a day or two, generating far more debate than they're worth.

The confusion isn't so much about the grammatical points themselves. It's about the nature of grammar in general. Apparently, many tech writers do not see grammar as a set of conventions to help them write clearly. Instead, to judge by the wording of the questions and responses, they see grammar as a set of unchanging rules that can provide definitive answers in every situation.

Some are afraid to break the rules of grammar and risk being denounced as incompetent. A handful, smugly sure that they know the rules, use their rote learning of the rules as an ad hominem attack, nitpicking at typos and small errors to discredit writers without disproving their viewpoints. Most sit in the middle, haunted by the ghosts of childhood grammar classes until they can hardly tell on their own authority whether they are writing well or not. But underlying all these reactions is an attitude that rules are rules, and cannot be broken.

This attitude is usually known as a prescriptive approach to grammar. It assumes that grammar exists mainly to tell us how to speak or write properly--not well. It is an attitude that tech writers share with almost everybody in the English-speaking world. It is a form of conditioning that begins in kindergarten and continues through high school and even into college and university. It undermines nearly everyone's confidence in their ability to communicate, especially on paper. Yet it is especially harmful to professional writers for at least three reasons:

• It grotesquely exaggerates the importance of grammar. Although competence in grammar is sometimes proof of other writing skills, it stresses presentation over content. Even worse, it stresses correctness over precision, conciseness, or clarity.
  
• It binds writers to viewpoints that are not only arbitrary and obsolete, but, in some cases, far from their own opinions.
  
• It undermines writers' confidence and their ability to make decisions about how to communicate effectively.
  

Why are we burdened by this attitude? How does it affect us? The easiest way to answer these questions is to look at the origins of the prescriptive attitude and the alternatives to it. Only then can we begin to grasp how we can live without it.

The Rise of Prescriptive Grammar

Prescriptive grammars are the products of the Enlightenment. Earlier grammars such as William Bullokar's in 1586 and Ben Jonson's posthumous one were also prescriptive, but intended for language students. The first prescriptive pronouncements for native English speakers date to the Seventeenth and Eighteenth Centuries.

This is the start of the great era of describing and recording. In every subject from biology to Egyptology, educated men struggled to write accounts so thorough that no other one would ever be needed. It was also a time that looked both backwards and forwards to Classical Rome. It looked back in the sense that ancient Rome was seen as the height of civilization. It looked forward in two senses: England perceived itself as a second Rome, and Latin was the language of international science.

The first prescriptive comments were very much in the spirit of their times. Their compilers hoped for definitive grammars that would fix the form of English once and for all, and provide a source for settling grammatical disputes. Since Latin was the language of the world's most sophisticated civilization, the closer this fixed form of English was to Latin, the more sophisticated English would be. Moreover, given the belief that culture had been degenerating since the fall of Rome, most grammarians automatically equated any change in the language with decay and degeneration.

John Dryden, the poet and playwright, was among the first to make prescriptive pronouncements. His main contribution to prescriptive grammar was to suggest that prepositions should never end a sentence. The reasons for this prescription are that Latin sentences rarely end in prepositions, and that the word "preposition" clearly indicates that this part of speech should go before ("pre") the noun it is associated with. Dryden criticized Shakespeare and Jonson for not following this rule, and scrupulously edited his own writings until they conformed to it.

Dryden also hoped to fix the form of English, which was still rapidly changing. Together with the diarist John Evelyn and other authors, Dryden called for an English version of l'Academie Francais--the body of scholars and writers that oversee the creation of the official French dictionary and is the arbiter of correct French. Dryden's plea for an English Academy was echoed later by Daniel Defoe and Jonathan Swift. The idea was being seriously considered by Queen Anne when she died in 1714. But with the ascension of the German-speaking George I, the question shifted from the monarch helping to purify English to teaching the monarch English, and the idea was dropped.

Deprived of royal assistance, English men of letters did their best to improve the language on their own. In the mid-Eighteenth Century, a number of grammars were published, as well as Samuel Johnson's famous dictionary, whose Preface spells out its prescriptive purposes with both succinctness and dry wit. All these works were heavily prescriptive, although Joseph Priestley did include some comments about the importance of common usage in deciding what was proper.

The most influential of these grammars was Robert Lowth's "Short Introduction to English Grammar," published in 1761. Criticizing almost every major English writer from Shakespeare to Pope, Lowth made most of the prescriptive statements that people still follow today. His prescriptions include:

• Two negatives make a positive, except in constructions such as "No, not even if you paid me."
  
• Never split an infinitive.
  
• Never end a sentence in a preposition.
  
• "Ain't" is unacceptable in formal English.
  

These ideas have several sources: An attempt to model English on Latin (and therefore to arrive at a universal grammar that underlay all languages), a desire to be scientific, and Lowth's personal preferences. For instance, Lowth would not accept split infinitives because Latin infinitives are one word and cannot be split. Similarly, two negatives make a positive because they do so in mathematics. The ban on "ain't," though, seems entirely Lowth's idiosyncrasy. None of these prescriptions, however, took any notice of how people actually spoke or wrote. For example, despite Lowth, "ain't" continued to be used by Queen Victoria and the upper classes until the start of the Twentieth Century.

Although Lowth later became Bishop of London, his ideas on proper usage would probably have remained obscure if they had not been borrowed for classroom use. In 1781, Charles Coote wrote a textbook grammar based on Lowth's Grammar, adding his own preference for "he" and "his" as the indefinite personal pronoun (as in "everyone is entitled to his opinion"). A few years later, Lindley Murray borrowed from Lowth to write a series of textbooks for a girls' school. Murray's textbooks became so popular that they quickly became the standard English in American schools throughout much of the Nineteenth Century. With modifications, Murray's "English Grammar" and Lowth's Grammar have been the basis for textbook grammars ever since. With their unyielding rules, these textbooks have given at least eight generations of English-speakers the prescriptive attitude that inhibits people today.

The Biases of Prescription

In their language and their purposes, prescriptive grammars make a strong claim to objectivity. A. Lane was typical of the first grammarians when he wrote in 1700 that the purpose of grammar was to teach people to speak and write "according to the unalterable Rules of right Reason." Similarly, in writing his dictionary, Samuel Johnson humorously referred to himself as a "slave of Science." These claims are still echoed when modern defenders of the prescriptive attitude assume that the rules of grammar are value-free. Yet a closer look at prescriptive grammar reveals distinct biases. Some of these were openly admitted by the first grammarians. The problem is that some of these biases are no longer current. Others are demonstrably false.

The early grammarians' claims to be scientific or precise concealed a strong personal bias. This bias was by no means a conspiracy--it was simply natural self-expression. Grammarians were highly educated, or the subject would hardly come to their attention at all. Since education was a privilege, they were either well-to-do or extremely talented. Since the higher levels of education were barred to women, they were male. And, as might be expected from the subject, they were all intensely literate.

While this background made the first grammarians supremely qualified for literary and scholarly work, it also carried a certain arrogance. The first grammarians showed scant interest in how English was used outside their own class and social circles. All of them assumed an unwarranted authority in their subject, appointing themselves as the arbiters of the language without any justification except their willingness to serve. Robert Lowth, for example, did not hesitate to include his own idiosyncrasies as grammatical rules. In much the same way, Samuel Johnson saw it as his "duty" to purify English usage through his dictionary. This same attitude continues in the prescriptive attitude today.

Moreover, the first grammars were a direct reflection of their authors' bias and education. The problem is, the education of the Restoration and Enlightenment includes assumptions that we would question today. For example:

Rome is the model for all things: In fact, Latin is a poor model for English. Although both languages are Indo-European, the relation is indirect. Even the heavy influence of French, a Latin-derived language, via the Norman Conquest, does not make English's structure much closer to Latin. At its core, English is a Germanic language, and if the first grammarians had used Dutch, Swedish, or German as a model, they would have had no precedent for objecting to split infinitives or double negatives. However, the Germanic origins of English were poorly understood in Britain during the Seventeenth and Eighteenth Centuries. At any rate, the first grammarians would probably have considered Germanic models too crude to replace the polished perfection of Latin.

Writing is the basis for English grammar: Like other grammarians, Samuel Johnson assumes that the principles of grammar should be taken from the written form of the language. Since Johnson was a writer himself, this assumption is understandable. However, modern linguistics regards written usage as simply one of many types of English, none of which is more valid in the abstract than any of the others. If anything, the spoken language is usually given greater priority today, partly because it tends to be the source of innovation, and partly because it reveals how people use the language when they are not trying to write correctly or formally.

Change is degeneration: When the narrator of "Gulliver's Travels" visits Laputa, he is shown a vision of the Senate in Ancient Rome, then the modern English Parliament. The Senators look like demigods and heroes, the Members of Parliament scoundrels and ruffians. In writing this passage, Jonathan Swift reflects a widespread belief in his time that times are getting continually worse. This fallacy is the exact opposite of the modern one of equating all change with progress.

Applied to the English language, Swift's view has little basis in fact. Admittedly, English has simplified itself over the centuries by dropping most noun declensions and verb conjugations, but that has not made it less useful for communication. Nor, despite the delicate shudder of prescriptive grammarians, does the shift in meaning of "cute" from "clever" to "attractive" in the early Twentieth Century or of "gay" from "happy" to "homosexual" in mid-Century weaken a language with as many synonyms as English. When clumsy or unclear constructions do arise (such as the use of "not" at the end of a sentence), their impracticality generally ensures that they are brief fads.

At any rate, what is degenerate and what is progressive is often a matter of opinion. To J.R.R. Tolkien, the Anglo-Saxon scholar and author of "The Lord of the Rings," the hundreds of words added by Shakespeare and his contemporaries are a corruption that ruined English forever. Yet to most scholars, these coinages are an expression of a fertile inventiveness and part of the greatest literary era ever known.

London English is standard English: Like most English writers, the grammarians and their publishers centered around London. The language of the upper classes in the Home Counties had already become Standard English by the Fifteenth Century, which is why many people have heard of Chaucer and few people have heard of (much less read) "Sir Gawain and the Green Knight," a brilliant poem by one of Chaucer's contemporaries, written in an obscure North Country dialect. That is also why Chaucer and Shakespeare poke fun at other dialects--they already had the idea that some forms of English were better than others. In basing their work on the English spoken around London, the grammarians were simply working in a long-established tradition.

Far from seeing their biases as contradicting their claims to scientific objectivity, the grammarians openly proclaimed their goal of saving English from itself. In fact, by the standards of the time, proclaiming their educational and cultural assumptions was a means of asserting their ability to be objective on the matter.

Of all the early grammarians, the one most keenly aware that prescriptive grammars were biased was Noah Webster, the writer of the first American dictionary. Working on his dictionary between 1801 and 1828, Webster was not content simply to record how words were used. Instead, he was also concerned with producing a distinctly American language. To this end, Webster not only included words such as "skunk" and "squash" in his dictionary, but also introduced American spellings, such as "center" instead of "centre." In addition, he encouraged the spread of a uniquely American pronunciation by consistently placing the stress on the first syllable of the word. In other words, Webster attempted to deliberately manipulate the use of English in the United States for patriotic reasons. Whatever anyone thinks of those reasons, Webster's efforts are one of the best proofs that prescriptive grammars are not as value-free as many people imagine.

The same is true today in the debate over whether "they" can be used as the indefinite pronoun instead of "he/his" ("Everyone is entitled to their opinion"). On the one hand, traditionalists who insist on "he/his" are perpetuating the male bias of the first grammarians. On the other hand, reformers who favor "they" are trying to remake the language in their own world view. It is not a question of objectivity on either side. It is simply a question of which world view will prevail.

The Problem of Change

But the major problem with prescriptive attitude is that it resists the fact that languages are continually changing. If a community newspaper constantly includes editorials urging people to drink less, the amount of concern suggests a local drinking problem. In the same way, the constantly expressed wish to set standards for the language reflect the massive changes in English at the time that the first grammarians worked. Although the rate of change was probably slower than that of the Fifteenth and Sixteenth Centuries, English was still changing much faster between 1650 and 1800 than it does today.

Some of the changes that occurred or were completed in this period include:

• The disappearance of dozens of Old English words like "bairn," "kirk," and "gang" (to go). Many of these words survived in Northern and Scottish dialects for another century, but became non-standard in written English.
• The addition of dozens of new words. Some deliberately coined by scientists, such as "atom." Some were borrowed from the regions that England was conquering, such as "moccasin" or "thug."
• The replacement of "thou" and "ye" with "you" in the second person plural. These forms survived only in poetry.
• The standard plural became "s" or "es." Only a few exceptions such as "oxen" survived.
• The loss of all case endings except the third person singular in most verbs ("I read," "he reads"). Some of the older forms ending in "th" survived longer in poetry.
• The loss of inflection in most adjectives.
• The regularization of past tenses to "ed" or, occasionally, "t" ("dreamed" or "dreamt").

Many of these changes are easy to overlook today because popular editions of texts from this period routinely modernize the spelling. However, the sheer number of changes makes clear that the early grammarians were fighting a rear guard action. If they have helped to slow the rate of change in the last two centuries, sometimes they have also accelerated it; the loss of many Northern words, for example, is probably partly due to the standardization on Home County English. Yet, despite these efforts, English continues to change as the need arises. Many of these changes come from the least educated parts of society--those ones least likely to be influenced by the prescriptive attitude.

Today, all prescriptive grammarians can do is resist changes as long as possible before accepting them. This constant retreat means that most prescriptive grammars are usually a couple of decades behind the way that the language is actually used in speech and contemporary publications.

The Descriptive Alternative

While prescriptive grammars were finding their way into the schools, an alternative approach to the study of language was being developed by linguists. Imitating the naturalists of the Eighteenth Century, linguists began to observe the pronunciation, vocabulary, grammars, and variations of languages, and began cataloging them in ways that suggested how they related to each other. In 1786, Sir William Jones established that most of the languages of India and Europe were related to each other. By 1848, Jacob Grimm, one of the famous Brothers Grimm of fairy tale fame, had detailed in his History of the German Language how English, Dutch, German, and the Scandinavian languages had descended from languages like Gothic. Content to observe and speculate, the early linguists developed what is now called the descriptive approach to grammar.

The descriptive approach avoids most of the distractions of prescriptive grammars. Today, most linguists would probably accept the following statements:

• Change is a given. In fact, a working definition for linguistics is the study of how languages change. Linguists can offer a snapshot of how a language is used, but that snapshot is valid for only a particular place and time.
• No value judgement should be placed on changes to a language. They happen, regardless of whether anyone approves of them or not.
• The fact that one language is descended from a second language does not make the first language inferior. Nor does it mean that the first language must be modelled on the second.
• Linguistics does not claim any special authority, beyond that of accurate observation or verifiable theory. Claims are open to discussion and require validation before being accepted.
• No form of a language is given special status over another. Regardless of who speaks a variation of a language, where it is spoken, or whether it is oral or written, all variations are simply topics to be observed. For example, when Alan Ross and Nancy Mitford coined the phrases "U" and "non-U" for the differences between upper and middle class vocabularies in Britain, they did not mean to suggest that one should be preferred (although others have suggested that deliberately using a U vocabulary might be a way to be promoted).
• Variations of a language may be more or less suitable in different contexts, but none are right or wrong. The fact that you might speak more formally in a job interview than at a night club does not mean the language of a job interview is proper English, or that the language of the night club is not.
• Proper usage is defined by whatever the users of the language generally accept as normal.
• A language is not neutral. It reflects the concerns and values of its speakers. For example, the fact that every few years North American teenagers develop new synonyms for drinking and sex reflects teenagers' immense preoccupation with the subjects. The fact that they also develop new synonyms for "slut" reflects their sexual morality as applied to girls. A similar viewpoint is known in psychology as the Sapir-Whorf hypothesis.

To those conditioned by prescriptive grammars, many of these statements are unsettling. For instance, when I suggested on the TECHWR-L list that proper usage was determined by common usage, one list member went so far as to call the view "libertarian."

Actually, these statements are simply realistic. Despite two centuries of classroom conditioning, average users of English have never been overly concerned about the pronouncements of prescriptive grammar. Instead, people continue to use English in whatever ways are most convenient. If the prescriptive usage is not widely used in practice, it may even sound odd, even to educated people. For example, to an ear attuned to the spoken language, the lack of contractions in an academic essay or business plan may sound stilted. In fact, an entire written vocabulary exists that is almost never used in speaking. The descriptive approach simply acknowledges what has always been the case. In doing so, it frees users from the contortions of prescriptive grammar, allowing them to focus on communication--where their focus should have been all along.

Professional Writers and Grammar

Writing well, as George Orwell observes in "Politics and the English Language," "has nothing to do with correct grammar and syntax." If it did, then two centuries of prescriptive grammar in the classroom should have resulted in higher standards of writing. Yet there is no evidence that the language is used more skillfully in 2001 than in 1750. The truth is that, prescriptive grammar and effective use of English have almost no connection. A passage can meet the highest prescriptive standards and still convey little if its thoughts are not clearly expressed or organized. Conversely, a passage can have several grammatical mistakes per line and still be comprehensible and informative. Prescriptive grammars are interesting as a first attempt to approach the subject of language, but today they are as useless to writers as they are to linguists. So long as writers have a basic competence in English, prescriptive grammar is largely a distraction that keeps them from focusing on the needs of their work.

By abandoning prescriptive grammar, writers shift the responsibility for their work to themselves. In practice, this shift means making choices that are not right or wrong in the abstract, but, rather, useful in a particular context or purpose. For example, instead of agonizing over whether "User's Guide" or "Users Guide" is correct, writers can choose whichever suits the situation. They can even flip a coin, if they have no better means of deciding. In such cases, which choice is made is less important than using it consistently throughout the document to avoid confusion. Even then, writers may decide to be inconsistent if they have a good reason for being so.

Similarly, the decision whether to use a particular word or phrase is no longer a matter of referring to a standard dictionary or somebody else's style guide. Instead, writers have to fall back on the basics: Will the intended audience understand the word? Is it the most exact word for the circumstances? Does it convey the image that the company wants to present? In the same way, while the need for clarity and a factual tone makes complete sentences and unemotional words useful choices in typical manuals, in a product brochure, sentence fragments and words heavy with connotation are more common. Writers may still want to summarize their decisions in a corporate style guide, but the style guide will be based on their own considerations, not the rules that someone else tells them to follow.

None of which is revolutionary--except that, under the prescriptive attitude, irrelevant purposes are often inflated until they become more important than a writer's practical concerns.

That is not to say that taking a descriptive approach to grammar means writing in the latest slang. Nothing could date a document faster, or be more intrusive to a technical manual. Nor does it mean abandoning technical vocabularies that are known to the audience or that make explanations easier. If anything, a descriptive approach demands a much greater awareness of the language than a prescriptive one. Instead of learning the single correct version of the languages, writers who take a descriptive approach need to be aware--probably through constant reading--not only of dozens of different versions, but of how each version is changing.

If necessary, writers can use descriptive grammars such as journalistic style guides to help them. On the whole, however, the descriptive approach leaves writers where they should have been all along, deciding for themselves what helps their documents to achieve their purposes. The only difference is that, under the descriptive approach, they are fully aware of their situation. If they say anything unclear or stupid, they can no longer hide behind tradition.

Prescriptive grammar is useful for teaching English as a second language, but it has little value for the practicing writer. Clinging to it may provide emotional security, but only at the expense of making writing harder than it needs to be. The culture-wide devotion to it will not be changed in a moment. But conscientious writers can at least change their own habits, and make life easier for themselves. And, from time to time, they can even laugh some worn-out, crippling concept--such as not ending a sentence in a preposition, or not splitting an infinitive--into the recycle bin where it belongs.

(With apologies to George Orwell.)

Bruce Byfield is a freelance journalist, product manager, and technical writer. A recovering academic, he is the writer of the standard reference on the American fantasist Fritz Leiber and a widely published poet. His other obsessions include raising Nanday conures; running long, painful distances; listening to punk-folk music; and indulging a four to 10 book a week reading habit. He can be reached at bbyfield@axionet.com.

Discuss this article on TECHWR-L.

Writing and Editing

• Grammar
• Editing