This is an advance summary of a forthcoming article in the Oxford Research Encyclopedia of Linguistics. Please check back later for the full article.
About 7,000 languages are spoken around the world today. The actual number depends on where the line is drawn between language and dialect—an arbitrary decision because languages are always in flux. But specialists applying a reasonably uniform criterion across the globe count well over two thousand languages in Asia and Africa, while Europe has just shy of three hundred. In between are the Pacific region, with over thirteen hundred languages, and the Americas, with just over 1,000. Many of the world’s languages are spoken by small populations and are thought likely to disappear over the next few decades, as speakers of endangered languages turn to more widely spoken ones.
The languages of the world are grouped into 141 language families, based on their origin, as determined by comparing similarities among languages and deducing how they evolved from earlier ones. While the world’s language families may well go back to a smaller number of original languages, even to a single mother tongue, scholars disagree on how far back current methods permit us to trace the history of languages.
While it is normal for languages to borrow from other languages, occasionally a totally new language is created by mixing elements of two distinct languages to such a degree that we would not want to identify one of the source languages as the mother tongue. This is the situation with Media Lengua, a language of Ecuador formed through contact among speakers of Spanish and speakers of Quechua. In this language, practically all the word stems are from Spanish, while all of the endings are from Quechua. Just a handful of languages have come into being in this way, but a less extreme form of language mixture has resulted in several dozen creoles around the world. Most arose during Europe’s colonial era, when European colonists used their language to communicate with local inhabitants, who in turn blended vocabulary from the European language with grammar largely from their native language. These so-called creole languages became so well established that they were passed on to the next generation, becoming a first language to many people, and continuing in use to this day.
Also among the languages of the world are about three hundred sign languages, used mainly in communicating with the deaf. The structure of sign languages typically has little historical connection to the structure of nearby spoken languages.
Languages have also been constructed expressly, often by a single individual, to meet communication demands. The prime example is Esperanto, designed to serve as a universal language and used as a second language by some two million, according to some estimates. But there are hundreds of others falling under the rubric of constructed international auxiliary languages.
Pidgin languages sometimes form in contact situations where a means of communication is urgently needed between groups lacking a common code. They are typically less elaborate than any of the languages involved in their formation, and in comparison to those, reduction characterizes all linguistic levels.
The process is relatively uncommon, and the life span of pidgins is usually short – most disappear when the contact situation changes, or when another medium of intergroup communication becomes available. In some rare cases, however, they expand (both socially and structurally), and may even nativize, i. e. become mother tongues to their speakers (when they may be re-labelled “creoles”).
Pidgins are severely understudied, and while they are often mentioned as precursors to creoles, few linguists have shown a serious interest in them. As a result, many generalizations have been based on extremely limited amounts of data or even on intuition. Some frequently occurring ones is that pidginization is a case of second language acquisition, that power and prestige are important factors, and that most structures are derived from the input languages. My work with pidgins has led me to believe the opposite to be true in these cases: pidgins form through a trial-and-error process, where anything that is understood by the other party is sanctioned, this process is one of collaborative language creation (rather than one involving one group of teachers and one group of learners), and much of what finds its way in the resultant contact language do so independently of what the creators spoke prior to their encounter.
As for theoretical implications, pidgins may shed light on which features in traditional languages are necessary for communication, and which are superfluous from the point of view of pure information transmission.