|Mark Aronoff (2007), Scholarpedia, 2(5):3175.||doi:10.4249/scholarpedia.3175||revision #121088 [link to/cite this article]|
Language as a Human Attribute
Language sets people apart from all other creatures. Every known human society has had a language and though some nonhumans may be able to communicate with one another in fairly complex ways, none of their communication systems begins to approach language in its ability to convey information. Nor is the transmission of complex and varied information such an integral part of the everyday lives of other creatures. Nor do other communication systems share many of the design features of human language, such as the ability to communicate about events other than in the here and now. But it is difficult to conceive of a human society without a language.
Language, like culture, that other most human attribute, is notable for its unity in diversity: there are many languages and many cultures, all different but all fundamentally the same, because there is one human nature and because a fundamental property of this human nature is the way in which it allows such diversity in both language and culture.
How Many Languages
How many different languages are there? First we must know how to count them, how to distinguish one language from another. Linguists usually say that language A and language B are distinct if a speaker of A and a speaker of B cannot understand each other. This is reminiscent of how biologists define distinctness in biological species, based on whether they may produce fertile offspring. According to this criterion of mutual intelligibility, there are about seven thousand distinct languages in the world today (many fewer than there were even a few decades ago, numerous indigenous languages and cultures having been lost to globalization and pressure from larger societies).
Language diversity is attested from earliest recorded history. The story of the tower of Babel was an attempt to explain the diversity of human language. The ancient Greeks called foreigners barbaroi, because the speech of all non-Greeks sounded to them like a babbling noise barbar. Certainly, to a monolingual speaker of English, even a fairly closely related language like Swedish or Dutch is completely incomprehensible, let alone a more distantly related language like Hindi or Russian, or completely unrelated languages like Japanese or Mohawk. Yet, despite the seemingly vast differences between them, all natural human languages are alike in their basic structural design; they are all instances of a single entity, human language.
Language and Linguistics
It is impossible to separate language from literature, or politics, or most of our everyday human interactions. In this article, though, discussion is centered on language structure rather than how it is put to use in human society. Accordingly, language is treated almost exclusively from the point of view of linguistics, and the article concentrates on what we have learned about language from that discipline over the last two centuries. Linguists study individual human languages and linguistic behavior in order to discover the fundamental properties of this general human language. Through this enterprise, they also hope to discover some fundamental aspects of what it means to be human. The importance of language and languages goes far beyond internal structure, extending to almost all human endeavors.
Language, being a human activity, is social in nature; hence, linguistics is usually classified as a social science. Because languages can only be studied through human behavior, linguistics, like psychology, is further classified as a behavioral science; and because language is essentially mental, linguistics is also a cognitive science.
There are many ways to study language scientifically. The most traditional, with its roots going back thousands of years to the Classical Greek and even Classical Sanskrit grammarians, is called Descriptive Linguistics. Its goal is to provide an explicit description of a language (often called a grammar), either in whole or in part (for example, a description of the sound system of Swahili). A linguist's grammar, though, unlike those some may remember from school, is never prescriptive: it does not dictate how a language should be (proper language) but instead draws on the actual linguistic behavior of speakers. Often, a descriptive linguist, especially one working on one of the many less-studied languages, will spend considerable time in the field, learning from the speakers of the language and sometimes writing the language down for the first time. Theoretical Linguistics, of which there are many varieties, seeks to provide explicit general principles that are applicable to all languages, often drawing on descriptive grammars. Within both descriptive and theoretical linguistics, historical linguistics is devoted to the study of how languages change over time. Sociolinguistics treats the broad question of language in society and includes the study of dialects. Psycholinguistics uses the methods of experimental psychology with language as the primary source of data. Child Language Acquisition is devoted to learning how children acquire language early in life. Neurolinguistics addresses the relationship between language and the brain. Computational Linguistics deals with the interaction of computers and language, for such purposes as Speech Synthesis, the production of artificial speech from written text, or Speech Recognition, the conversion of speech to text, or parsing, the automatic description of the grammatical structure of a text.
Written Language and Spoken Language
Most people think of a language as primarily written. Indeed, when a person studies a language in school, they usually study the written language, either literature (texts written in the language) or composition (in which the students compose their own written texts). Spoken language is given second place in schools and universities, except at the very elementary level of foreign language study. What set modern linguistics apart, beginning in the nineteenth century, was the realization that the opposite is true: language is primarily spoken and written language is an imperfect reflection of spoken language, conveyed through a fairly new and imperfect technology, writing.
The main evidence behind this conclusion is the fact that every human society has a fully functioning spoken language while, until a century ago, only a very few societies had a written language and even then, literacy was, again until recently, confined only to a small class of people. Furthermore, the few writing systems that existed prior to the twentieth century all owed their origins to three or four quite recent inventions, none much more than five thousand years old, while many scholars believe that human language evolved at least 50,000 years ago. All of these writing systems arose in early materially advanced state-like societies. These inventions include Sumerian cuneiform and Egyptian hieroglyphic writing in the Middle East (likely related to one another in origin); Chinese writing; and Mayan hieroglyphic writing (deciphered only in the mid twentieth century). The alphabet, the most widely used of modern writing systems, was an adaptation by Semitic-speaking people of aspects of the Egyptian system. It is therefore spoken language that is common to all human societies.
Spoken language comes naturally to all normal human children: expose a normal young human child to any language from a very early age and the child will fully master the language without any overt instruction, while it is very difficult for most humans to acquire a new language after a certain age (generally around puberty). It is as if young human children came preprogrammed to acquire a spoken language (something that many though not all linguists believe). Written language, by contrast, must be overtly taught; it is never learned effortlessly, and rarely perfectly.
If all societies have languages then we may begin to ask in what ways all these languages differ from one another and in what ways they are similar. The first question, asked very early on in the history of the modern study of language, was whether one language is more advanced or evolved or complex than another. The answer is no: there is no obvious way to rank languages on some evolutionary scale: all languages appear to be equal in their expressive capacities. Some languages may have more words than others or may have words for certain notions that are not conventionalized in other languages but no language is inherently incapable of expressing a given proposition. This realization of the equality of natural languages was in turn important in the realization that all humans are equal, regardless of the material, social, economic, and political complexity of the society in which they live. We know that members of the materially most simple societies are equal to members of the materially most advanced societies in no small part because we can find no convincing evidence that the language of one is more advanced than the language of the other.
What languages have in common
What all spoken languages share are certain very general structural properties. Every language has a grammar with the following components: meaningful units akin to words (lexemes) and other grammatical markers; a sound system (Phonetics and Phonology); a system for arranging the meaningful units into sentences (Syntax); another for arranging the internal parts of words (Morphology); another for interpreting the meanings of utterances (Semantics); and principles for using language in actual discourse (Pragmatics). The boundaries between these systems are not always clear. The study of phonetics, for example, deals with the more physical properties of speech, its relation to both acoustics and physiology, while phonology treats sounds more as abstractions, but there is no way to draw a sharp line between them.
The sound systems of all languages are very similar in their basic design. In each, we can isolate a small number of distinctive speech sounds or phonemes, ranging from as low as eleven (Hawaiian) to over sixty (some languages of the Caucasus) . Phonemes by themselves have no meaning, but they combine into syllables and then into words, which are then assigned meanings or values by social convention. Thus, English speakers may combine the three phonemes /n/, /k/, and /i/ to form the three English words ink, kin, and nick, which have the meanings they have because speakers of English agree that they do. The three words are not related to one another in meaning, only in form. Nor are all combinations of these three phonemes permissible. The sound patterns of English prevent the existence of /nki/, /kni/, and /ikn/, though another language might permit one or more of these combinations; in both German and Russian, for example, words may begin with /kn/ and English once permitted this sequence, as we can tell from the spelling of words like knight and know, which is a relic of this old system. Even sign languages, though they use hand configurations, locations and movements instead of consonants and vowels, have units equivalent to phonemes.
Every language has its own distinctive system of patterns that make up the phonological system of that language. In some, the syllable type is highly restricted. Japanese syllables, for example, almost always end in a vowel or in /n/; a Japanese syllable may end in a consonant only if it is identical to the beginning consonant of the following syllable (like the first [p] in Nippon). This strict syllable structure is revealed very nicely when an English word is borrowed into Japanese, as with many baseball terms. The English word strike, for example, which has only one syllable, has five syllables in Japanese (/su-tu-ra-i-ku/), because each consonant must have its own syllable, and there can only be one vowel sound in any syllable. Note that the letter "i" in English is actually pronounced /ay/ and so contains a consonant. Similarly, the word baseball has four syllables in Japanese. English allows fairly complex syllables, by the standards of most languages: the word sixths ends in four consecutive consonants (since "x" is actually two: [ks]). But Georgian easily beats that: the word prckvnis 'he peels it' begins with five consonants and gvprckvnis 'he peels us' begins with seven!
All languages have words but the complexity of words varies just as widely as that of syllables. Again, English lies somewhere in the middle. The word de-institution-al-iz-ation (which is here broken down into its constituent meaningful parts (or morphemes) by means of hyphens) means 'the release of institutionalized individuals from institutional care (as in a psychiatric hospital) to care in the community' according to Merriam-Webster Online and it is easy to intuit how this meaning is derived from the meanings of its five parts, though it would take a good deal of space to explicate precisely how these parts are combined. Many English words have a number of meaningful internal parts, prefixes and suffixes. Vietnamese, by contrast, has almost no complex words, except for compounds in which two whole words are combined to form another, similarly to English words like doghouse or catbird. But in many languages of North America, an entire sentence can be expressed in a single word, as in the Inuktitut (Eskimo) word iqqanaijaaqajjaagunniiqtutit 'you won't have any work anymore' (from a Canadian Broadcasting Corporation piece on official language policy in the northern Canadian territory of Nunavut, where Inuktitut is the language of the vast majority of people and the official language of local government).
All languages have sentences; both the basic building blocks (parts of speech like nouns and verbs) and the systems for constructing sentences out of these building blocks are very similar across languages: there is no language without nouns and verbs and pronouns, though other categories, like adjectives and adverbs, are not universal. Basic sentence structure is quite uniform across languages, consisting of a subject and a predicate, with the essential ingredient of the subject being a noun phrase and that of the predicate usually a verb phrase. Within the verb phrase, all languages have transitive verbs with an object, which again is usually a noun phrase, exactly the same as the subject in its basic internal structure. Where languages differ is in the order of the subject (S), object (O), and verb (V), though S precedes O in almost all languages. The most common orders are SVO, as in English, and SOV, as in Japanese. Biblical Hebrew and Classical Arabic were VSO languages. Languages also differ in the internal structure of phrases: in English, the adjective precedes the noun it modifies, while most adjectives follow the noun in Romance languages like Spanish and French. Some languages, though, like Classical Greek and the Aboriginal languages of Australia, have free word order, so that any order of the words of a sentence is permissible.
All languages also permit recursion, the possibility of inserting a syntactic category within the same category. The clearest example of recursion is the insertion of one sentence within another sentence. Consider the two sentences 'Bill left' and 'Mary said that Bill left'. The second sentence contains the first one embedded within it. Language permits infinite recursion, at least in theory. The sentence 'Louise knows that Mary said that Bill left' contains the second sentence; the sentence 'John claims that Louise knows that Mary said that Bill left' contains the third one; and so on without end. Of course, no one has ever uttered an infinite sentence, at least not yet, but simple mathematics tells us that if an infinitely long sentence is possible in principle through the mechanism of optional recursion, then there must be an infinite number of possible sentences (all those that contain fewer embedded sentences than this infinitely long one). Of course, in the real world, no one will ever utter an infinite sentence, but the importance of recursion is not just the mathematical trick, but rather that it allows languages to have complex utterances through the combination of simple structures.
At the first level of communication, people who feel that their language is the same, perceiving no difference in each other's language may be said to speak the same dialect. As we might expect, dialect differences correlate not only with geography, but also with social distinctions like class and ethnicity. Dialects thus serve the function of allowing us to recognize members of our own social group from characteristics of their speech. As mentioned at the beginning of this article, dialects may be grouped into languages based on mutual intelligibility. As long as two speakers understand one another, they speak the same language, from a purely linguistic point of view. But when most people talk about languages, they do not mean purely linguistic constructs. First of all, modern language identity depends on writing. A person from the heart of Glasgow, another from the heart of Jamaica, and a third from the mountains of West Virginia, put together in a room, will not find it easy to hold a conversation, yet all will declare that their native tongue is one and the same, because they all share a written language, written English. The same is true of Chinese, the speakers of which share a single written language, though they speak a dozen or so completely distinct spoken languages from a purely linguistic point of view, usually referred to as Chinese dialects. By contrast, a speaker of Hindi from Delhi and a speaker of Urdu from Islamabad will vehemently insist that they speak different languages, even though they understand each other perfectly, because the two languages are written in different alphabets, and, perhaps more importantly, are associated with different religious and national identities. From a purely linguistic point of view, based on intelligibility however, these two people speak the same language.
Regular Language Change
Related to the fact that everyone speaks their own language is the fact that spoken languages change inexorably. Despite the efforts of governments and academies over millennia and the cries of the purists, all languages change and, because they are such highly structured systems, they do so in very orderly fashion. The most notable type of language change, the discovery of which was the starting point of modern linguistic science, is sound change. We can see evidence for the regularity of sound change in English spelling. One of the most noticeable peculiarities of this system is the proliferation of silent letters. Why do we have a silent letter <k> in knight, know, knee, and knave and many other words? The answer is that this silent letter was fully pronounced up until about the time of Shakespeare, but that it was lost through sound change (along with the now silent <g> in gnarly, gnaw, and other similar words). The regularity of this sound change is demonstrated by the fact that there are no <kn> or <gn> words in which the first letter is pronounced. The same is true for all the silent final <e> letters. They were once fully pronounced.
The regularity of sound change can also readily be seen by comparing the ways in which two languages that share a common ancestor have diverged over time. English and Greek are related, being descended from a common ancestor that we now call Indo-European, spoken some five thousand years ago. We have no evidence of this ancestor other than the many languages descended from it that quite quickly stretched from India to Iceland, and in the modern world now span the globe. We can see the relationship by comparing words that begin with /p/ in Greek and /f/ in English. Compare Greek /pater/ 'father' with English father, /penta/ with five , /pod/ 'foot' with foot. Linguists say that we have here a regular sound correspondence. But if we compare other words in the two languages, we will find other sound correspondences: Greek /k/ corresponds to English /h/: compare Greek /kuon/ 'dog' and /kardia/ 'heart' with English hound and heart. And there are many more such sound correspondences. We can use these both to show that two languages are related and even to reconstruct what their common ancestor must have sounded like. But all this is because of the regularity of sound change, something that all human languages share.
There is much more to be said about the unity in diversity of human languages. All languages are manifestations of the single phenomenon, language, that most characteristic of human attributes.
O'Grady, W., J. Archibald, M. Aronoff, and J. Rees-Miller (2005). Contemporary Linguistics, 5th Edition. New York, Bedford St. Martin's.
Gordon R. G. (2005) (eds), Ethnologue: languages of the world, 15th edition. http//www.ethnologue.org
- Valentino Braitenberg (2007) Brain. Scholarpedia, 2(11):2918.
- Olaf Sporns (2007) Complexity. Scholarpedia, 2(10):1623.