Prescriptivist's Corner: Genitive Pronominal Antecedents

March 7, 2020

1 June 2003

After three months of repeated letter writing by Kevin Keegan, a Montgomery County, Maryland school teacher, the Educational Testing Service (ETS) has thrown out a question on the 2002 PSAT exam, raising the scores of some 500,000 students. The issue at hand is whether or not the question as originally written was actually wrong.

In the question the students were required to identify a grammatical error, if there was one, in a sentence. The sentence in question was as follows:

Toni Morrison’s genius enables her to create novels that arise from and express the injustice African Americans have endured.

The correct answer, as originally scored by ETS, was that there is no error in this sentence. Keegan disagreed, believing that the was an error in the pronoun her and its antecedent. According to Keegan, a pronoun cannot take a noun in the genitive case as its antecedent. Keegan cited several usage manuals backing up his claim.

Eventually, after consulting a three-person panel of “experts,” ETS reversed its position and declared the question invalid.

The problem is that despite the claims of a few grammar manuals, there is absolutely nothing wrong with a pronoun taking a genitive antecedent. This “rule” has no basis in either logic or in actual usage.

First, let’s recast the sentence so that Toni Morrison is not inflected in the genitive:

The genius of Toni Morrison enables her to create…

Here we have an identical meaning, yet use of the pronoun is acceptable. Why should there be potential confusion in the original yet not in the recast sentence? Next, let’s try an recast the sentence without using the pronoun.

Toni Morrison’s genius enables Toni Morrison to create…

Toni Morrison’s genius enables the writer to create…

Both of these are inadequate. The first is duplicative—and demonstrates that the logical antecedent is indeed Toni Morrison. The second is ambiguous—who is the writer we are talking about? There simply is no good way to write the sentence without using the pronoun.

The only possible logical explanation for the rule is that a pronoun in the objective case should not have a genitive antecedent. But this is an arbitrary rule that flies in the face of common sense. The purpose of rules about antecedents is to ensure clarity in writing. Should we abandon a clear sentence in favor of an awkward construction solely for the purpose of satisfying someone’s idea of what is logical?

Now, let’s turn to usage. While a few grammar manuals do include such a rule. The vast majority do not. Fowler found nothing wrong with the practice. Merriam-Webster specifically says that this “rule” is in error and unreasonable. Further, no professional writer appears to be aware of this rule. Open any newspaper on any day and one will find dozens of pronouns with genitive antecedents. And some of the greatest writers in history have “violated” the rule at will:

It was Mr. Squeer’s custom to call the boys together and make a sort of report, after every half-yearly visit to the metropolis, regarding the relations and friends he had seen.
—Charles Dickens, Nicholas Nickleby

…shaking Snook’s hand cordially, we rush on to the pier, waving him a farewell.
—W.M. Thackeray, The Book of Snobs

Raymond’s sobs softened, and trembled away. She held him, rocking silently and rhythmically, a long time.
—Dorothy Parker, Clothe the Naked

ETS was simply wrong in caving in to an individual, pedantic teacher like Keegan. The identity of the panel of supposed experts is unknown, but clearly they are neither expert nor capable of the most basic inquiry regarding English grammar or usage.

There is an argument to be made that the students should not be penalized for the sins of the teachers. If teachers like Keegan instruct and grade students on this “rule” (and Keegan admits that he does), then the test should not hold the students accountable. But it can also be said that the students should be evaluated against an objective standard. If an answer is wrong, then it is wrong. The test should not cater to the whims of pedants. In the words of Richard R. Hershberger, regular participant in the Wordorigins.org discussion group:

I, on the other hand, find the whole affair rather inspirational. I still live in a country were a person can make up a rule of English and, through the sweat of his brow, get it accepted by a pseudo-authority.

Myths of Language Change, Part 2: That's Not What It Really Means

March 7, 2020

1 June 2003

The changing face of our language has created an interesting conundrum. On the one hand, people recognize and delight in the language change of the past. But on the other hand, people routinely resist current changes in the language. The language they learn as children is, for many, the only acceptable manner of speaking. Change is vehemently eschewed.

How people can revel in the changes of the past yet fiercely resist the changes of the present is just bizarre. And it is futile. The language will change whether we like it or not, and no amount of resistance will stop a change whose time has come.

This resistance usually takes one of several classic forms. In this series of articles, we well examine each of these classic errors of resistance. Here are two more:

That’s Not What It Really Means
During the Gulf War of 1991 several language commentators raised a hue and cry over the use of the verb to decimate. Journalists were broadcasting statements like, “The coalition air forces decimated the Iraqi Republican Guards, destroying some eighty percent of their combat power.” The commentators objected to the use of decimate, claiming that the original sense of the word meant to kill one in every ten. The word comes from the Latin verb decimāre, meaning to take one tenth. It was originally a form of Roman military punishment, used when soldiers committed mutiny or some other group crime.

The aggrieved language mavens concluded that if deci- means ten in Latin, the word decimate could mean nothing other than to destroy one tenth of something. What the mavens forgot was that the sense of a word is determined by how it is used, not by how the word originated. If enough people use decimate figuratively, meaning to kill in large numbers, then that is what the word comes to mean.

The meanings of words change over time. Words acquire new senses and lose old ones; they acquire connotations; they specialize and generalize and they are used metaphorically and figuratively. This is a natural and continuous process. Almost no word is used in exactly and in only the way it was when it first entered the language. This error is so common that linguists have given it a name, the etymological fallacy, or the belief that the original sense must be the correct sense.

In the case of decimate, the word has been used figuratively to mean to destroy a large proportion of something since the beginning of the 19th century. The reporters in the Persian Gulf were not using the verb in any new or strange sense.

It Means This, Not That

Pete: How did you score the tickets?
Clark: Lex hooked me up.
—Smallville, “Cool,” 2001

A similar error is to believe that a word has one and only one meaning. Some people insist that words must have precise and singular senses, that to brook multiple definitions and connotations is a recipe for confusion and disaster.

This is simply incorrect. Most words have more than one meaning or sense. Sometimes, the different senses connote minor nuances. Sometimes they are radically different, even contradictory.

I was recently involved in an online discussion on the meaning of the slang verb phrase to hook up. Some contended that the verb meant to have sexual intercourse as in, “After three months of dating, Bill and Sarah finally hooked up.” Others said that it could mean anything from a bit of necking to going all the way, so long as it was a casual encounter, “Bill and Sarah hooked up after the party.” Others disagreed and claimed that it meant to date seriously, to go steady, “Bill and Sarah have been hooked up all semester.” Others said it was even more serious; to be hooked up is to be married, as in, “Bill and Sarah have been hooked up these past 25 years.” Still others chimed in with a platonic sense, meaning simply to meet someone, “Sarah hopes to hook up with Bill while she is in New York and discuss the marketing plan.” And not mentioned in the discussion, but still out there, is the transitive sense meaning to broker a sale or deal, especially for a hard-to-find item, “Bill hooked Sarah up with the Rolling Stones tickets.”

All six are correct. Hook up can mean all these things. The penultimate sense, to meet someone, is the oldest. It dates to the turn of the 20th century America. Also, about that time, the verb came to mean to get married. In the 1980s, the verb started appearing in college campus slang in the various sexual permutations.

So does this cause confusion? Not really. The sense is usually apparent from the context. We can safely assume that Bill and Sarah didn’t get married after the fraternity party and that they did not discuss the marketing plan during pillow talk (at least for their sake we hope not). And as to whether it means sexual intercourse or something less, well there is often value in ambiguity. It allows Sarah to make her encounter with Bill seem to be more than it was when she brags about it to her sorority sisters, and it allows Bill to let the guys in the locker room know that he and Sarah had a good time without resorting to detailed description of the sexual acts. If we need precision, there are plenty of other words that will provide it.

Not only can senses be different; they can be downright contradictory. Hook up can refer to a variety of social encounters and relationships, but they differ only in degree of familiarity and length of the relationship. Other words can hold contradictory senses. These are called Janus words and the most cited example is cleave. Cleave can mean to join together or to split apart. Again, context prevents confusion.

Word of the Month: Usenet

March 7, 2020

1 June 2003

Usenet, n., is a distributed bulletin board system. The term is an abbreviation for users’ network. Usenet was originally implemented in 1979-1980 to link computers at Duke University with those of the University of North Carolina. The system consisted of numerous discussion forums called newsgroups. Computers that functioned as newsservers would pass messages to one another. Usenet was originally conceived to carry local news and information, hence the names. By 1996, Usenet had over 10,000 newsgroups and an average of over 500 megabytes of information posted to it daily.

Usenet was not originally part of the Internet. Instead it was originally carried by the UUCPNET network of Unix computers. By the early 1990s, however, most of Usenet was being transferred over the Internet.

Usenet newsgroups are organized into categories based on the content of the group. The first part of a newsgroup’s name designates its category:

alt.*, alternative, popular topics
comp.*, computer topics
humanities.*, humanities topics
misc.*, miscellaneous topics
news.*, Usenet meta-discussion
rec.*, recreational topics
sci.*, science and technology topics
soc.*, social and cultural topics
talk.*, debate and controversial topics

If you want information about computer firewalls, for example, you can go to the comp.security.firewalls group. Or if you are planning a trip to Brazil, you might check out rec.travel.latin-america. Old movie aficionados can be found on alt.movies.silent.

In addition, there may be local hierarchies carried by some newsservers. For example, the ucb.* hierarchy is for the University of California at Berkeley. There are also many other hierarchies, such as biz.*, that addresses business topics.

There are also old hierarchies that are no longer used, but one may encounter them when searching archives of Usenet posts. The first hierarchy was net.*. This was supposed to contain discussions about Usenet itself. But shortly after the hierarchy was created, the groups net.jokes and net.rumor appeared. When ARPANET was linked to Usenet in 1980, the hierarchy fa.*, for “from ARPANET,” was created. There is also the old mod.* hierarchy, which consisted of moderated newsgroups. Nowadays, the designation *.moderated is tacked on to the end of a newsgroup name to show that it is a moderated group.

For the major Usenet categories there is a formal process for proposing and approving new newsgroups. This process is waived for the alt.* hierarchy of groups where people may start groups on a whim and the rules may be different for local hierarchies. Originally, newsservers carried all the groups in the major hierarchies, with the exception of the alt.* hierarchy, which many system administrators limited. Beginning in 1987, however, the growth of the system was such that system administrators began to choose which groups in the primary hierarchies they would carry. Most newsservers give access to at least 10,000 groups.

The following are jargon terms associated with Usenet. The parenthetical information is the newsgroup and date the term first appears in the Google Groups archive of Usenet posts. This is not necessarily the earliest use of the term—often, the terms come from sources other than Usenet or the earliest posts that use them are not in the archives:

^H^H^H^H^H, jocular expression, stands for an erasure or backspace. Unix systems generate ^H symbols when the backspace key is pressed. Usenet posters type these into their messages to pretend that they are clueless Unix users attempting to delete something they just wrote. An example might be, “you are a real pain in the ass^H^H^H neck.”

AFAIK, abbrev., for As Far As I Know. (comp.sys.sun, 9 Dec 1988)

Binary, n. & adj., a digital file, often a picture or computer program. Most Usenet groups forbid the posting of binary files, limiting them to specific groups with binaries in the title. Most binaries groups are pornographic. (net.v7bugs, 25 May 1981)

BITNET, abbrev., for Because It’s There NETwork. A messaging network very similar to Usenet. Created in 1981 at the City College of New York. Very successful among universities, it did not survive transition to the Internet. Support for BITNET was ended in 1996, but some of its groups still survive. They can be recognized by their bit.* hierarchy.

BOA, abbrev., for Ban On Acronyms, Many Usenet groups frown upon acronyms and abbreviations because they are confusing to people who do not understand what they mean. The acronym BOA is, of course, facetious. (alt.folklore.urban, 10 Nov 1994)

BOP, abbrev., for Ban On Politics. Many Usenet groups have a rule against discussing political topics. (soc.culture.lebanon, 4 Feb 1995)

BTW, abbrev., for By The Way. (fa.sf-lovers, 13 May 1981)

Cancelbot, n., a program that deletes Usenet messages after they have been posted. The term is from cancel + [ro]bot. Cancelbots are usually used to delete spam. (alt.magick, 4 Oct 1993)

Cascade, n., an automated chain of responses to a Usenet message, each one making a trivial change from the previous message. Cf. DoS attack.

Cow orker, n., a misspelling of coworker, usually deliberate. The intent is to humorously intimate that there is some illicit activity known as orking cows. (alt.sca, 8 Aug 1989)

Cross post, n. & v., a post that is sent to more than one newsgroup at the same time. While judicious use of cross posting can sometimes enhance a discussion by bringing in different perspectives, it is generally discouraged. Newsgroups tend to have their own internal conventions and practices and cross posting often results in conflicts and flames. Cross posting between obviously incompatible groups (e.g., alt.sex.bondage and soc.religion.christian.promisekeepers) is a form of trolling. (net.women & net.med, 24 Mar 1984)

DoS attack, n., is an abbreviation for Denial-of-Service attack. It is an attempt to shut down a newsgroup with large a number of spam or frivolous posts. The term is also used to denote an attack on a computer network aimed at overloading the network with large quantities of data. (alt.politics.org.covert, 13 Jun 1996)

Emoticon, n., a smiley (or other emotion) face drawn with ASCII characters, e.g., ;-). The term is from emot[ion] + icon. Emoticons are used in text-only communication to replicate tone of voice, facial expressions, and body language that would be present in face-to-face communication. Certain Usenet groups :-( upon use of emoticons. (comp.sys.amiga, 31 Jan 1987)

FAQ, abbrev., for Frequently Asked Question. Most Usenet groups have lists of FAQs and common topics of conversation so that newbies can have their questions answered without bothering the old hats. (comp.lang.c, 25 May 1989)

Flame, n. & v., is a deliberately insulting email or Usenet post. The term is from the “inflammatory” nature of the message. Its earliest appearance on Usenet is in fa.arms-d, 26 May 1981, but it is claimed to have been in existence in email usage from the 1960s.

Froup, n., is a jocular misspelling of group. Sometimes it is seen as newsfroup. (talk.bizarre, 24 Feb 1988)

Furrfu!, interj., is the exclamation sheesh! encrypted in Rot13. It is used as a general exclamation in Usenet. Furrfu first appears as an exclamation in the midst of unencrypted text in alt.folklore.urban, 18 Apr 1992, although there are instances of sheesh appearing in the midst of longer passages of Rot13 encrypted text as far back as 1986.

FWIW, abbrev., for For What It’s Worth. (net.unix, 27 Aug 1986)

Godwin’s Law, n., states that as an online discussion grows longer, the probability of a comparison involving Nazis or Hitler approaches one. Many Usenet groups hold that once a comparison with Nazis has been made, further intelligent conversation on that thread is impossible. The law was formulated by Mike Godwin in 1990. The earliest reference in the Gooja archives is in rec.arts.books, 22 Aug 1991.

Gooja, n. & v., is the Google Groups archive of Usenet posts or to search those archives. The term is a combination of Goo[gle] + [De]ja, Deja News being an older archive that was bought by Google. (uk.games.video.dreamcast, 21 Feb 2001)

Great Renaming, The, n., the establishment of the current hierarchy of Usenet groups and the elimination of the old one. It occurred in 1986-87.

IANAL, abbrev., for I Am Not A Lawyer. The term is commonly used on Usenet as a disclaimer immediately prior to dispensing legal advice. (comp.virus, 18 Jan 1990)

IIRC, abbrev., for If I Recall Correctly. (rec.bicycles, 10 May 1990)

Killfile, n. & v., is a list of individuals or topics that will not be displayed by a newsreader. Individuals can tailor the killfiles in their newsreaders to suppress messages from particular individuals or that relate to particular topics. The term is also used as a verb meaning to place an individual or topic into one’s killfile. (news.software.b, 19 Aug 1987)

Lasnerian, adj., refers to vaguely defined period. Usenetters will often say such things as “1992 (Lasnerian)” or “1992 L.” to refer to dates that they are not certain of. The term is after Charles Lasner, who on 21 Aug 1993 posted a message to alt.folklore.urban where he contended that the period known as the “sixties” was not 1960-69, but rather started with the assassination of President Kennedy in 1963 and ended with the American withdrawal from Vietnam in 1973. Use of the term is heavily localized to alt.folklore.urban, but you will occasionally see it cropping up in other groups.

Line noise, n., is the presence of spurious characters in a Usenet post created by electrical noise in the network connections. People in groups that frown upon use of emoticons often refer to them as line noise. (fa.info-cpm, 1 Jun 1981)

Lurker, n., is a person who reads a Usenet group without posting to it. (net.movies.sw, 13 May 1984).

MOTSS, abbrev., for Member Of The Same Sex. The abbreviation was taken from a 1970 US Census term and used in Usenet newsgroups back when many system administrators refused to carry gay-themed newsgroups. The group net.motss was proposed in Sep 1983 to avoid using gay in the name of the group. (net.singles, 7 Jan 1983)

Netiquette, n., is the system of rules of social engagement on the Internet, etiquette for the net. (net.jokes, 15 Nov 1982)

Newbie, n., is a newcomer to a newsgroup. The term is originally from Vietnam-era US military slang, but nowadays is almost exclusively used in Internet contexts. (comp.sys.mac, 31 May 1988)

NNTP, abbrev., for Network News Transfer Protocol, a system created in 1984 that allowed Usenet to be transferred over TCP/IP connections. This freed the system from UUCP and allowed the Internet to carry Usenet.

Ob-, comb. form, stands for obligatory. The prefix is usually combined with the name of the newsgroup and used in off-topic posts. Text following the Ob-[newsgroup] is then some non-sequitur but on-topic comment. It can also be combined with other words to note an obligatory comment, e.g., ob-joke, ob-disclaimer. (comp.sys.ibm.pc, 4 Feb 1988)

Off-topic, adj., designates a posting that is outside the bounds of discussion for that particular group. Netiquette dictates starting the subject line with OT: when making an off-topic post so that those not interested can easily ignore it. (net.followup, 17 Nov 1982)

Old hat, n., is the opposite of a newbie, someone who has been a participant in a newsgroup for a long time. It is a jocular variant on the adjective old hat meaning tired or out of date. (net.astro, 15 Nov 1985)

OTOH, abbrev., for On The Other Hand. (net.nlang, 12 Aug 1983)

Plonk, interj., is an echoic term for the imaginary sound of someone being dropped into a killfile. Tradition dictates that when you killfile someone, you reply to their message with Plonk! to let them know they have been killfiled. (alt.flame, 11 Nov 1989)

Rot13, n. & v., is a simple cipher system where a letter is replaced by the letter 13 places away from it, A becomes N, B become O, etc. For example, “Qnegu Inqre vf Yhxr’f sngure” is “Darth Vader is Luke’s father.” Rot13 is used on Usenet to encode offensive jokes, pornographic passages, spoilers to movies and TV shows, or anything that someone might not want to stumble upon inadvertently. Most newsreaders include a Rot13 encode/decode feature. (net.jokes, 4 Dec 1982)

ROTFL, abbrev., for Rolling On The Floor Laughing. It is used to respond to jokes and humorous comments that one finds very funny. Also seen is ROTFLMAO, which adds My Ass Off to the phrase. (rec.humor, 29 Oct 1990)

September That Never Ended, The, catch phr., is September 1993. There was a time when the Internet and Usenet were the exclusive provinces of government and educational institutions. There was no public access to Usenet; you had to be affiliated with an institution to gain access. Old time Usenetters always dreaded Septembers. That was the month when thousands of new college freshmen would gain access to Usenet for the first time. The influx of newbies in September was almost overwhelming. In 1993, however, the general public could gain access to Usenet for the first time through Internet Service Providers. After September 1993, there was a constant stream of newbies on Usenet and that month came to be known as the September that never ended. (alt.folklore.computers, 26 Jan 1994)

Sig block, n., meaning signature block. It is also often sig file. Many Usenetters have standard blocks of text that are added to the end of their posts. These signature blocks often contain quotes, poems, bits of ASCII art, etc. (net.news, 16 Aug 1983)

Signal to noise ratio, n., the amount of intelligent, original commentary in a newsgroup, compared to spam, pointless chatter, etc. It is often abbreviated to S/N ratio. It is an adaptation of an electronics term. (net.suicide, 19 Feb 1983)

Sock puppet, n., is one who mindlessly parrots the ideas of another, as if one were a sock puppet. On Usenet, people sometimes create avatars to make it seem as if more people are agreeing with them. These are also known as sock-puppets. (chi.media, 15 May 1995)

Spoiler space, n., are blank lines inserted at the top of posts that contain spoilers, forcing someone to scroll down in order to read the spoilers. The intent is to prevent inadvertent viewing of spoilers. (rec.games.misc, 11 Apr 1992)

Spoiler, n., is information that gives away the end of a movie, TV show, book, video game, etc. and spoils it for anyone else. On Usenet, it is considered polite to clearly mark posts that contain spoilers. (fa.sf-lovers, 13 May 1981)

Thread drift, n., is the gradual changing of the subject matter in a thread. Long-lasting threads on Usenet almost invariably change topics multiple times. (rec.games.frp, 13 Jan 1991)

Troll, n. & v., a deliberately inflammatory message designed to rile up and provoke members of newsgroup. The term is from the method of fishing. (alt.folklore.urban, 7 Feb 1993)

YMMV, abbrev., for Your Mileage May Vary. It is used to mean that the results one gets may not match the results reported in the post. The term is from a standard disclaimer that automobile manufacturers give. (rec.music.classical, 27 Feb 1990)

Book Review: The Story of English (McCrum, et al.) & The English Language (Barber)

March 7, 2020

1 May 2003

Generally, we avoid reviewing older books in this space, but we make an exception this month. Over the past few years, several readers of the web site have asked about historical overviews of the English language. The essay A (Very) Brief History of the English Language on the wordorigins.org site is available, but this is simply a summary of major events and trends. Those who want a more detailed account have to look elsewhere. Here we take a look at two lengthy treatments of the topic.

The first of these books, The Story of English by Robert McCrum, Robert MacNeil (of PBS Newshour fame), and William Cran was first published in 1986 as a companion to the BBC television series of that name. The book was (allegedly) revised in 2001.

The Story of English is intended for a popular audience. Written in an accessible style, it scrupulously avoids linguistic jargon and IPA notation. The book also makes wide use of literary and pop culture references. In this respect, the book is a success. It is a good read and makes the linguistic development of English available to a wide audience.

Unfortunately, this comes at a high price. Many of the chapters are woefully out of date. The research and linguistic theory on which the book’s assertions are based are twenty or thirty years old and often have been discounted by more recent scholarship. The 2001 revision has not significantly revised the book at all. The other major problem with the book is sacrificing accuracy for a good story. The authors frequently plump for linguistic tall tales and false etymologies that are fun to tell but that are disproved by only a modicum of research. Finally, the book is also an exercise in linguistic jingoism, repeatedly emphasizing the point that the English language is the greatest, most poetic, most useful, etc.

The first half of the book is significantly better than the second. The tracing of the history of English in the British Isles is a reasonable account. The treatment of Old and Middle English and how these became the language that we know speak is rather a good popular account.

The book begins to fall apart at the midway point. The first glitches appear in the description of early American dialects. The authors hold out several examples of isolated dialects being relics of older styles of speech. They make statements like “it is sometimes claimed that you have to go to the Appalachian hills, or the Ozark Mountains, to hear Elizabethan English,” without providing a factual basis for the claim (there is none). As a linguistic rule, isolated dialects do not represent older modes of speech. Rather the opposite is true, isolated dialects tend to change faster. They may retain specific features that are archaic elsewhere, but this is true of all dialects, not just isolated ones.

The first chapter to show widespread factual problems is the treatment of African-American English. It presents the creolization theory as the basis for the dialect. This theory, which was popular in the 1970s, has been shown to be false. The creolization theory was formed before serious research into African-American and West African dialects had born fruit.

The theory holds that Gullah is representative of traditional slave dialect and is based on a creole of English and West African languages and that the generic African-American dialect is an offshoot of this. Current research discounts a widespread West African influence on African-American speech and holds that Gullah is an isolated dialect of English, neither related to West African nor the progenitor of the general African-American dialect. When the book was first written, plumping for this theory could be forgiven. But given that this is a 21st century revision, the retention of such outdated material is unacceptable.

Which raises the question of what exactly was revised? The revisions appear to be superficial, mainly updating of various pop culture references. There does not appear to have been any serious attempt to bring the book into accord with current linguistic thinking.

Outdated research is not the only problem. The authors often engage in sophistry as an excuse to tell tall tales. A case in point is the previously mentioned promulgation of the Elizabethan English in the Appalachians myth. Note the authors do not state it as absolute fact. They simply note that it is a commonly told story, without mentioning that it is complete nonsense and completely at odds with everything we know about linguistic development in general and the Appalachian dialect in particular.

A similar instance is the association of an American cattleman named Joseph McCoy with the origin of the phrase the real McCoy. The authors say, “he was, as he liked to say, ‘the real McCoy.’” Again, they imply that he is the inspiration for the phrase. They do not mention, as anyone who looked up the phrase in the Oxford English Dictionary could tell you, that the phrase was in use before Joseph McCoy was born. There is no excuse for this. It is simply bad scholarship.

The third major flaw with the book is more a matter of style. Take the following passage from the first chapter:

Why is it that English can inspire astonishing affection not only among those who speak and write it as their mother tongue but also among those for whom it is a foreign language? The richness and power of English was summarized in the nineteenth century by the great German philologist Jakob Grimm when he wrote, ‘In wealth, wisdom and strict economy, none of the other living languages can vie with it.’

Such passages crop up repeatedly throughout the book. The authors do explicitly state that English is not superior to other languages in any intrinsic sense, but the constant repeating of such sentiments does create a jingoistic tone.

The Story of English is a good read and the opening chapters do provide a reasonable historical overview of the language. It is, however, a seriously flawed work and should only be used with caution and verification.

In contrast there is Charles Barber’s The English Language: A Historical Introduction. Barber provides a superb introduction to the subject of historical linguistics.

Barber’s book is written as an introductory textbook. He avoids linguistic jargon where he can, but he does use IPA phonetic notation. While this is more accurate and more useful, it detracts from the readability. Barber also avoids the turgid and impenetrable prose that is common in academe. The book is written in a straightforward style that is accessible to the general reader. It is a textbook, however, and lacks the pop culture references and humor that often make popular linguistic books fun reads. In short, it is readable and fact filled, but a bit dull.

Barber approaches the subject of historical linguistics and etymology in a straight chronological fashion. After an initial chapter on general theory of linguistic change, he deals in successive chapters with Indo-European, Germanic, Old English, Middle English, Early Modern English, and contemporary English. He concludes with a chapter on various dialects of English around the world—a subject that McCrum and colleagues devote half their book to.

Of the two books, Barber is the far superior, if somewhat duller, treatment of the topic. McCrum and colleagues have produced a more entertaining book, but one of dubious scholarship and that is very much out of date.

The Story of English, Third Revised Edition; by Robert McCrum, Robert MacNeil, & William Cran; Paperback; 496 pp.; Penguin; Dec 2002; $16.00.

The English Language: A Historical Introduction; by Charles Barber; Paperback; 299 pages; Cambridge Univ. Press; 1993; $18.00.

Note on IPA: IPA stands for International Phonetic Association. This organization has promulgated a standard notation system for linguistic sounds. IPA notation is preferred by most serious linguists. It allows one to denote pronunciation in virtually any dialect spoken by humans in a way that anyone familiar with the notation system can understand. A passage written in IPA can be pronounced by someone who is utterly unfamiliar with the original language; the meaning is still opaque, but the pronunciation comes through. The chief drawback with IPA notation is that relatively few people, other than professional linguists, know it. Use of IPA sacrifices ease of reading for precision and universality.

Myths of Language Change, Part 1: That's Not a Real Word

March 7, 2020

1 May 2003

Prescriptivists and grammarians may have some minor victories. For instance, they have managed to keep ain’t out of formal speech. But they have not come close to stamping it out altogether. And in the meantime, other “abominations,” like irregardless, manage to stick their nose under the tent and eventually the entire camel is inside. Fighting language change is like the resisting Star Trek’s Borg—futile.

This resistance usually takes one of several classic forms. In this series of articles, we well examine each of these classic errors of resistance. The first is:

That is Not A Real Word

Dictionary, n., A malevolent literary device for cramping the growth of a language and making it hard and inelastic. This dictionary, however, is a most useful work.
—Ambrose Bierce, The Devil’s Dictionary

How many times have you heard someone say “that is not a real word?” And in almost every case (outside the context of games of Scrabble®) the statement is dead wrong. It is, in fact, a word. It may not be a standard word. It may not appear in a dictionary. But it is a word.

Of course, on one level this is not an error. People can’t go out and invent their own vocabularies. If they did, no one else would understand them. But the objection is rarely, if ever, over a matter of comprehension.

I still remember the shock and horror when I heard the director of marketing at my firm use the word impactful in a sentence, “we need a really impactful ad campaign.” The very fiber of my body rebelled upon hearing that word. It just felt wrong. Yet there was no communication problem. I understood exactly what she meant. The word is simply a derivation of the common word impact and the common suffix –ful, a very standard and productive pattern of word formation. As I heard others use it, I gradually began to realize that it really was a rather useful word, especially in the world of marketing and advertising. It conveys a nuance that other substitutes, like effective, successful, or emotional, do not. Gradually I have come to accept the use of the word (although I still can’t bring myself to use it myself).

Sometimes the claim that a word is not real is phrased as, “it is not in the dictionary.” Obviously, if we were limited to just the words in the dictionary (and which dictionary will we choose?), there would never be any new words. The language would become hard and inelastic as Bierce fears. This brings us to the question of who decides what is correct or Standard English.

There is no body of experts that meets in London or Washington and passes judgment on words and phrases, deeming which are worthy of canonization and which should be cast into perdition’s flames. Instead, the editors of a particular dictionary make the decision of what to include and what to exclude. Each dictionary has its own criteria to guide the editors. Some leave out scientific and technical terms. Others leave out proper names. American dictionaries omit British expressions, and vice versa.

Almost all English language dictionaries are descriptive in nature, rather than prescriptive. That is, the editors seek to describe how words are used, not to tell people how they should be used. So if a word is not in a dictionary, what does that mean? Well, it simply means at least one of the following statements is true:

• The word does not conform to the editors’ rules for inclusion.
• The word did not exist or was not widely used at the time the dictionary was published.
• The editors made a mistake and left out a word they should have included.

For example, there is no entry for impactful in the Oxford English Dictionary, 2nd Edition, but the word does appear in that dictionary. It can be found in a 1975 usage citation, “overdramatization of an already impactful event.” For whatever reason, the editors did not give the word its own entry or recognize it as an adjectival form of impact, but it was not because it did not exist.

Of course, this doesn’t give one license to use any word at any time. You probably want to avoid the word ain’t in a job interview for instance, but not because it is not a word. It may not be appropriate in a given situation, but it is a word, one good enough for writers like Charles Dickens, Mark Twain, Virginia Woolf, and William Faulkner.