Some Notes on Language...

Ronald Kephart

University of North Florida

What is language?

As North Americans living in the early 21st century, we have been educated about language from the time we entered school. But much of what we learn about language in schools belongs more to a folk model than to an analytic model of language. Here are several pervasive aspects of our folk model of language. Language is a communication system. It is true that we use language to communicate with others. However, language is much more than a communication system. The most recent thinking about the nature of language suggests that language is first and foremost a representational system; a system which provides us with the symbols we need to model for ourselves, to ourselves, inside our heads, the universe around us. This modeling, carried out using the symbols ("words") provided by language, is commonly called "thinking." The communication function of language, which allows us to represent things not only to ourselves but to others as well, is an added benefit.

Primitive people speak primitive languages. We know, from anthropological research, that there are no primitive people on Earth today; indeed, it may be that the "Neandertals" were the last truly primitive people. And, there are no primitive languages, either. All languages that we know about, including those that are no longer anyone’s native language (Latin, Homeric Greek, etc.) have all the properties of the so-called "modern" languages (French, Spanish, Russian, etc.).

Even languages that have been reconstructed, such as Proto-Indoeuropean (the parent language of most European languages as well as Persian, Hindi, etc.), show no signs of "primitiveness." They have all the characteristics of so-called "modern" languages (see below).

Some languages are "harder" than others. While languages differ from one another in just which parts are simple and which are complex, all languages seem to be about equally complex or difficult to learn in their totality. For example, if we compare English and Russian we find that English nouns are relatively simple, while verbs are rather complex; in Russian, the nouns are hard and verbs are relatively simple.

Language is writing. If we ask a naive English speaker how many vowels English has, the answer is usually "five". This is because we tend to interpret any question about language as a question about the writing system. The English alphabet has 5 symbols that are normally used for the representation of vowels. But the English language has between 10 and 12 basic vowel sounds; this is the answer the linguist is interested in. Language is first and foremost oral; speech as a means of communication has been around for perhaps 200,000 years or more, while writing has existed for only about 6,000 as far as we know. Many languages, including many Native American languages as well as most of the creole languages of the Caribbean, exist without a written tradition. This in no way diminishes their language-ness.

Grammar is a set of prescriptive rules. When we think of grammar, we tend to think of the sorts of rules drilled into us by our language arts and English teachers: "Don’t end sentences with prepositions!" "Don’t use double negatives!"

When linguists work to discover the grammar of a language, they are looking for descriptive rules that model the linguistic patterns which people carry around inside their heads. We follow most of these rules unconsciously. In most cases no-one ever teaches them to us (see below); and, in most cases, we cannot articulate them. We "know how" to use our language, but we don't typically "know why." The interesting thing is that these rules of our descriptive grammar are frequently far more subtle and complex than anything the language arts teachers tell us about.
 
 
 

This ain't no joke....

The prescriptive rule mentioned above, about not using double negatives, was created by Bishop Robert Lowth in 1762 in England. The idea was to make English more like formal mathematical logic, and thus improve the thinking of English speakers. But this was not a descriptive rule for English; English has always used double negatives, as this sentence from the Anglo-Saxon Chronicles (written in the time of Alfred the Great) illustrates:

Ne  bith  thær nænig   ælo  gebrowen mid     Estum.
not be    there not-any beer brewed    among Estonians
'There is no beer brewed among the Estonians.'

Furthermore, in some languages, like Spanish and Russian, so-called "double negatives" are the rule, rather than the exception. Note the Spanish and Russian expressions for I don't see anything.

Spanish:     Yo  no veo   nada.
                  I   no  see   nothing
Russian:     Ya ne vizhu nichevo.

These are the normal, indeed the only, way of expressing this in Spanish and Russian. If language worked like formal logic, Spanish and Russian speakers would be suffering from a permament case of illogic. Since speakers of Spanish and Russian appear to be normal human beings, we have to conclude that language does not obey the rules of formal logic.

Thus, the rule against double negatives formulated by Bishop Lowth is not a grammar rule, but rather a social rule having to do with what he considered to be the acceptable use of English.


 
 

Language is pure and unchanging. As a conservative society heavily focused on written, rather than oral, forms of language, we tend to think that change, in language as in many other things, is bad. A whole industry of language "experts" such as Edwin Newman and William Safire regularly rant and rave against whatever shift in meaning or usage is current. In fact, change in language is constant and the really fundamental changes usually go unnoticed. For example, between Middle and Modern English many English vowels changed their pronunciation, so that words like "house" and "wife", today pronounced [haws] and [wayf], were pronounced as [hu:s] and [wi:f] by Chaucer. These sorts of changes, and others, are still going on in English even as we speak.

An analytic model of language

A language is a representational system composed of a set of oral (or, in the case of the hearing impaired, signed) symbols shared by the members of a social group, and a computational system (or grammar) for combining the symbols into phrases and sentences. People use language for internal representation (thinking) and for external representation (communicating). For linguists, the "grammar" of a language is what the native speakers of the language know about their language. Some of the things speakers of a language need to "know" in order to speak a language are: Phonology. Speakers of a language have to know what the distinctive sounds of their language are. English speakers must "know", for example, that the sounds /p/ and /b/ contrast in words like pat and bat.   Morphology. Speakers need to know how to combine the sounds of their language into meaningful units: words, prefixes, suffixes, etc. For example, English speakers "know" how to form the plural of words like cat, dog, and bush by adding the appropriate suffix to form cat[s], dog[z], and bush[iz].

Syntax. Speakers have to know how to combine their words into meaningful sentences that call attention to something and then provide information about it. Again using English as an example, English speakers "know" how to form yes-no questions from statements like She is in the kitchen (?> Is she in the kitchen?) by appropriate movement of is.

Semantics. Speakers must know the meaning of the words they use.

Pragmatics. Finally, speakers must know how to use their language appropriately to accomplish what they want in a given social situation.

However, the knowledge native speakers have is mostly unconscious knowledge; they "know" how to say it, but they (usually) can’t tell you how or why they say it that way.
 

Language as both biology & culture

It seems clear that language is a part of the human biological endowment. Perhaps the most compelling evidence for this can be found in the area of children's acquisition of language.

All normal human children, everywhere, acquire the language of their social setting at about the same pace and in the same way. They do so without formal training, and they do so in social and cultural contexts which differ in terms of what kinds of linguistic interactions are supposed to be appropriate between parents/caregivers and infants. These differences do not seem to affect the rate or quality of children's acquisition of language; in a sense, children acquire language in much the same way as they acquire the skill of walking. However, children who are isolated, for some reason, from all forms of linguistic interaction do not acquire language, and if they reach puberty without exposure to language they may never be able to acquire more than a very rudimentary linguistic ability.

By the time they are around 3-4 years of age, children have mastered some of the most complex and subtle rules of their language, rules which no teacher of language could ever teach them. Of course, they still have lots of vocabulary to learn, as well as some of the pragmatic rules of language use in different social situations, and they have to learn to read and write.

While the underlying shape of language is biological, any given language itself is a cultural artifact. The best way to illustrate this is to take the words for the domesticated animal which English speakers refer to as a dog. All languages have a word for this animal; no language has a word for "half-a-dog." This seems to result from a property of the human brain that guides our perception and  representation of natural objects in the world, like dogs, which come to us in whole "packages" (other candidates might be rocks, trees, birds, and so on). At the same time, though, the words we find in different languages

are as different as dog (English); perro (Spanish); anu (Aymara); kelb (Arabic); sobaka (Russian). None of these words has a privileged connection to the animal itself. Each is an arbitrary but conventional answer to the problem of naming these familiar domesticated animals.
 
 
 
 
 
 

The nature of language

Language (not just any language, but all languages) share a number of characteristics or design features that help fill out the concept of just what language is. Here are a few of the most important…

Infinite use of finite media. Although languages are complex, they are not infinitely complex. The number of rules that anyone needs to "know" to create sentences in their language is relatively small, and the number of different kinds of sentences is quite small. Still, the number of sentences that can be produced by any speaker of a language is potentially infinite.

Multiple patterning. Language is patterned at a number of levels of organization: sounds are patterned into phonemes, phonemes into words, words into phrases, phrases into sentences, sentences into larger units of discourse. This is what makes the infinite use mentioned above possible.

Predication. All languages make it possible for their speakers to name something and then make some kind of assertion about whatever was named. In other words, all languages allow for sentences that contain a subject and a predicate. We’ll explore this further in the unit on syntax.

Learnability. A central fact about all known languages is that they are all learnable by human beings. All normal human children acquire the language of their social group, and many (perhaps most!) go on to acquire more than one.

Traditional transmission. While all humans appear to have a built-in, genetically provided capacity for language acquisition, the actual acquisition of language must take place in a social context. The social context determines whether the language acquired is English, Russian, or Inuit, etc.

Displacement. Unlike most animal vocalization systems, which require that a stimulus be physically present for the vocalization to take place, human language allows us to talk about things that are absent in either space or time, or both. Without this feature, humans would not be able to talk about dinosaurs, or Cleopatra. We can add that this feature also allows us to talk about things that never existed, such as Klingons. Without it, we could have neither history or fiction.

Openness. Also unlike other animals, which typically have a fixed set of vocalizations, humans can increase the number of expressions at their disposal by inventing words. This feature allows us to add new words to our vocabulary such as hard drive, internet, and gigabyte.
 
 

Language & dialect

In our folk model of language, dialects are usually considered to be incomplete, perhaps ungrammatical, certainly less desirable forms of a standard language. The standard language, in contrast, is seen as more developed, more of a true language. The standard language is the form insisted upon for writing, for use in formal situations, certainly for reading and writing in schools. People who do not know the standard language are sometimes viewed in the same way as deficient, incomplete, lacking in education.

The analytic model of language includes the notion of linguistic relativism, which suggests that there is no point in trying to rank languages on any kind of scale. All human languages that we have any direct information about appear to contain all the characteristics necessary for language. In this view, there is no qualitative difference between a language and a dialect; the reasons why a particular variety of speech gets labeled as a dialect instead of as a language must be sought elsewhere. In particular, the reasons are to be found in the political, social, and economic value placed on the speakers of the language variety in question. The people who wield political, economic, and social control speak the "language"; those who do not speak the "dialect."
 
 
 

I had this lesson driven home to me in 1979, when I was working on a description of Carriacou speech. I took a few days to visit friends in Grenada, and met a Grenadian physician at a dinner party. He was an Afro-Grenadian, but of course upper class. He inquired as to what I was doing, and I explained that I was studying the speech of Carriacou people (who are considered rural and backward by many "mainland" Grenadians). I told him that the end product would be a grammar of their speech, including the rules they follow to produce sentences and so on. A few days later, we met again accidentally at the beach, and he said, "You know, I’ve been thinking about what you told me the other night. I wonder if you realize the political implications of what you are doing. If you show that these rural people speak a real language with grammar rules, then you are showing that they are really human beings and that we have been wrong in treating them the way we have all these years. We’ll have to start treating them as human beings."

 

The realization that languages and dialects are not qualitatively different, and that attitudes toward them really reflect social prejudices, has led some linguists to say that a language is "a dialect with an army and a navy." For linguists, then, what counts as a "language," as opposed to a "dialect," is socially and culturally negotiated; not determined by some objective linguistic truth. Sometimes the negotiation is spectacularly unsuccessful, as when the Oakland (California) school board attempted to declare African American Vernacular English (Ebonics) a "language." There was a great public outcry against this, but almost nobody understood the real reason: African Americans in the US do not have "an army and a navy"; therefore, they are not entitled to have a "language."

I tend to avoid the difficulty of the word dialect by using variety instead. It seems easier and less judgmental to speak of varieties of English such as British, Australian, North American, or West Indian. We can even talk about varieties of creole English, such as Jamaican, Trinadadian, Barbadian, Belizean, and so on. Or, we can go in the other direction, and discuss varieties of Romance such as Spanish, Portuguese, and Italian; varieties of Indo-European such as Germanic and Balto-Slavic; or even varieties of human language such as Indo-European, Austronesian, and so on. It all depends on what level of abstraction we are interested in.

Although from an analytic viewpoint they "know" a language as well as anyone, speakers of non-standard varieties of language are often assumed by the folk model to be language-deficient. In the Caribbean, this manifests itself especially when creole-speaking children get to school and come up against the standard language in an intense way for the first time. Teachers, who through no fault of their own very often have only minimal training, are aware only of the folk model for language. They assume that deviation from standard language forms is evidence for a lack of language, and that children "have no grammar." The analytic model of language tells us that all normal human children "have grammar" but that grammar is their own knowledge of their native language, not the rules written down in school books.
 
 
 
 

Last update: May 15, 2005

Copyright © Ronald Kephart, 2005