What is American English, and Why Do We Care?
What Is American English?
Everyone knows that Americans speak English differently than the British or Australians or even Canadians do, but most of the time we think of these differences in terms of the way we pronounce certain words (i.e., our accents). Most people also know that there are some differences that manifest themselves in written language as well as speech, including well-known lexical variations such as lorry/truck, pavement/sidewalk, tap/faucet, autumn/fall, etc. But beyond calling some things by different names, there are many other peculiarities of American English in its phrasing and syntax that set it apart from other brands of the English language. For example, phrases such as "omit to", "endure to", etc. are common in British usage but rarely occur in American English; and the bare infinitive after "insist", "demand", "require", etc. (e.g., "I insist he be here by noon.") is common in American English but rare in other brands of English. There are also considerable semantic differences between British and American English, for example: "tuition" is not used to cover tuition fees in British English; "surgery" in British English is "doctor's office" in American English; "school" does not include higher education in British English, etc. Usage not only differs but can be misleading, for example, British English uses "sick" for the American "nauseous", whereas "sick" in American English is comparable to "ill" in British English; British "braces" are U.S. "suspenders", while "suspenders" in British English refers to something else entirely. (For more information about differences between American and British English, see Fillmore, et al., An American National Corpus: A Proposal and John Algeo, "British and American Grammatical Differences", in The International Journal of Lexicography, Vol. 1, No. 1, 1988.)
Of course, even within the U.S. there are considerable variations in not only terminology, but also phrasing and syntax, as you move from region to region across the nation. In fact, the characteristics of different "brands" of American English has been a topic of considerable interest for at least 100 years, and has led to the creation of a Dictionary of American Regional English, a multi-volume work-in-progress that documents regional variations in terminology and pronunciation across the U.S. Very recently, PBS aired a program entitled Do You Speak American?, based on a book by Robert MacNeil and William Cran, which examined in depth variations in American English and the sociological impact of language use in the U.S. (The PBS website for the show is an excellent source of a wide variety of information about American English, well worth checking out.) What we learn from all of these sources is that American English is a rich English "dialect" of its own, with possibly numerous sub-varieties (including ebonics, chicano, etc.) that manifest their own fascinating characteristics, many of them a result of influences from other languages and cultures that have played--and/or continue to play--a defining role in U.S. history. It also tells us that the language is constantly developing, and that new words and constructs arise every day in (especially) Americans' use of English.
Why Do We Care?
Why does anyone care about these differences in language use? Obviously, linguists care because that's what linguists do: study language use and document its variations. But there are far more practical reasons for caring about the difference between American and other kinds of English; for example, consdier the following scenarios:
- You are the publisher of a dictionary of English, and your primary market is in the United States. To be successful, your dictionary should reflect the ways in which words and syntactic constructs are used by Americans--including, potentially, the variants that reflect regional usage.
- You are a web search engine developer, and you would like to refine your searches to deliver documents that are actually on the topic the user is interested in. But if your language analysis program only recognizes British terms and syntax, it either won't work well for Americans, or will possibly "break" when it encounters American constructions and not work at all.
- You are a teacher of English as a second language, and you want to teach English as spoken in the U.S., including not only accent but also word use and syntactic constructions.
- You are a translator in Iraq and your target audience is American troops, and it is most expedient if your translation is understood immediately and completely, without needing clarification or causing even momentary confusion.
- You are a programmer writing translation software, and your target language is American English. You want your translator to produce the most natural sounding prose possible.
In each case, the differences between American English and other kinds of English are important, if not critical. The next question is: how do you find out what all of these differences are, and which ones are more common than others? And how do you deal with the fact that language use changes over time, and new words and constructs come into the common vocabularly every day? How will it be possible to keep up with these changes, and to know when a word or construct is well enough established to warrant inclusion in your dictionary, search engine, or class plan? The answer, of course, is to consult the ANC.
(To learn how to find out more about the ways in which software developers, dictionary makers, linguists, and English language teachers use the ANC, see our bibliography.)
How Does the ANC Help?
The ANC provides exactly what is needed to not only find out what words, phrases, and constructs are used in American English these days, but also keep track of changes in the way Americans use the language. No other source of language data serves this purpose for American English. This is because:
- The ANC is massive, including many millions of words of American English. To get a reliable idea of the way the language is used, it is necessary to have enough examples to know that use of a particular word or construct is not just an accidental variant, but rather is commonly enountered. The enormous size of the ANC also provides enough examples to show which words and constructs are more or less frequently used.
- The ANC includes current American English, all produced in or after 1990. Therefore, everything in the corpus reflects the way Americans use the language now, not 30 or 50 or 100 years ago.
- The ANC includes only language data that is guaranteed to have been produced by native speakers of American English. There is no "pollution" of the samples of language use extracted from the ANC with language produced by speakers of other brands of English.
- The ANC includes language data across all genres, including fiction and non-fiction, technical articles and poetry, newspapers and magazine articles, web blogs and email, and many more. The core corpus contains the same number of words representing each genre, so that a sample drawn from it will show how the language is used in any context, and not just in newspapers or technical articles (in which language use can be very different).
- The ANC includes both written language and transcripts of spoken language. And within the category of written language, it includes large samples of both edited (e.g., books and articles that have been edited for publication) and unedited (e.g., web blogs and email) material. This provides means to see how American English is used in both formal and more spontaneous writing, or how American English differs when it is spoken rather than written.
- The ANC identifies the kind and source of each document included, date of production, and in some cases the age, sex, region of habitat, race, etc. of the producer of the material. So, it is possible to look at and compare language produced in specific genres and situations, and by people with different demographic characteristics. It also enables looking at "older" (early 1990's) vs. "newer" (early 2000's) language for comparison.
- The ANC is dynamic--that is, new and current data is added periodically so that examples of the most recent American English usage are included. Also, because evey document is dated, it is possible to see how the language is evolving.
- The ANC is annotated, so that the part of speech (e.g., noun, verb, etc.) for each word is identified, syntactic constructs (noun groups, verb groups, etc.) are marked, and more kinds of annotation are being added all the time. This allows users to search the data for complex and/or abstract patterns (e.g., the noun group following the verb "tell", adjectives used with a particular noun, etc.) and to use the data for language modeling that can help in the development of language understanding software.
How Can You Help?
The ANC is a unique resource, enabling study of American English that has never been possible before. However, collecting and preparing the ANC data requires considerable time and effort. If you are interested in helping out with the project, check out the ways in which you can contribute data or annotations to the ANC and other ways you can help.