The Museum of Human Language

You can see (from the sixth column) that the first 16 languages make up half (50%) of the world's population.  But the next 13 make up only 10%.  Thousands of languages are not included in the table.  Now, why do some languages like English, Spanish and French have official status in so many countries?  Mostly because these were the languages of colonial powers which imposed them on their colonies.  When the colonies became independent, sometimes the native languages had died out, sometimes there were so many native languages that it made sense to continue with the colonial language.  Notice that there are languages which are official in no (0) countries.  This is the case with Chinese dialects (sometimes called languages), and many other indigenous languages.

There are six official languages used in the United Nations.   They have official status for historical reasons as well: all but Spanish and Arabic were the languages of the victors in World War II.  It should be mentioned that Russian is spoken in many countries, typically those of the former Soviet Union, but they have not made Russian official, partly because of resentment on the part of the ex-republics.

What is a dialect?  Is it the speech of someone with an "accent"?  Doesn't every person speak a slightly different language from everyone else anyway?  It is difficult to formalize a definition for a dialect.  Some linguists say a dialect is a language without official status.  This would seem to be true of the Chinese dialects, since speakers of one dialect do not understand those of another.  One formal definition of  a language is a chain of mutually intelligible (understandable) speakers.  Most linguists agree that each person speaks an individual idiolect.  Anything in between a language and an idiolect would be a dialect.

The linguistic situation in Europe may be instructive.   Many of the Romance languages are familiar to us: Spanish, French, Italian, Portuguese, Romanian.   They are all descendants of Latin.  But people in different regions of each of theses countries speak somewhat differently, and in some cases could not understand each other if they spoke "in dialect".  For example, Sicilian is probably further from Northern Italian than Spanish is from Portuguese.   On the other hand, people speaking in Northern Italian dialect can easily talk to people speaking the Southern French dialect.  The political border between France and Italy was not a linguistic border until recently, when people have been giving up their dialects and have started speaking only the standard national language.

Because the political borders in Europe are somewhat arbitrary, it is possible to travel from town to town, starting in Portugal and ending in Sicily without noticing any great change in the local language, and without coming to any border where people living on one side cannot understand people living on the other.   Of course, Sicilians do not understand Portuguese speakers. But according to the above definition of a language, there are only two Romance languages: the Italian-French-Spanish-Portuguese superlanguage and Romanian.  Romanian is separated from Italian by Serbo-Croatian, a Slavic language, so it's quite different. And it is not intelligible to anyone speaking only the other Romance languages. Perhaps you do not like this definition: what about the written languages of French, Spanish, Italian and Portuguese with centuries of distinct traditions?  Don't these count anymore?  Well, for many years Catalan has been counted as a dialect of Spanish, even though it has been written and spoken for centuries, and is not understandable to speakers of standard Spanish (Castilian).  And Catalan might just as well have been counted as a dialect of French, since it is more-or-less midway between Spanish and French.   Now what about the similarities between Romanian and the super Romance language?  For an answer, just go to  Languages in Time.

Dialecticians sometimes draw maps with lines showing areas where particular speech features are used.  Wherever these lines overlap, like a cable made of many little wires, we can say we have a dialect boundary.   In other words, people living within the area  outlined by the "cable" all speak the same dialect.  If people on one side of the "cable" cannot understand people on the other side, we generally have a language boundary.

When differences between peoples' speech become too great for people to understand one another, one language can become two languages.  But the two languages still have something in common: an ancestor language, the original language that perhaps nobody speaks anymore.  Then the two languages could split again and become four languages.  Perhaps only one splits.  But from this continuing process a whole tree of relationships can grow.

The biggest family (in number of speakers) is Indo-European (Indo-Hittite).  Its major branches are:
 

Indo-Iranian	Languages spoken in India, Pakistan, Iran, etc.
Italic	Latin and its descendents: French, Italian, Spanish, etc.
Germanic	German, Dutch, English, Swedish, etc.
Balto-Slavic	Russian, Polish, Lithuanian, etc.

Nobody know for sure how language change starts.  Does it always start with a small group?  How does a change "catch on"?  We all notice when a new word becomes part of the language; think of words like Internet, duh, etc.  Sometimes we notice when a whole sound changes.  Think of the way some young people pronounce short E: neck sounds almost like knack.  Also, the difference in pronunciation between caught and cot seems to be disappearing throughout the US.  If someday AU sounds like short O  everywhere in the world, then we can say that English has changed.  Otherwise, we can say that only some American dialect or dialects have changed.

Interestingly, when sounds change, they seem to change everywhere we find them.  For example, if you pronounce caught and cot the same, then you certainly pronounce taught and tot the same.  Linguists discovered this phenomenon in the eighteenth century and it has helped them reconstruct the history of languages and to map out language families and family trees. What linguists do is to look for words with similar meanings and similar sounds (cognates) in selected languages, and see if they can guess what sound changes would have taken place to yield these words.   Some examples of such cognates between German and English are was vs. what, das vs. that, Wasser vs. water.    These words mean almost exactly the same thing in both languages.  And in this case, either an S changed to a T in English or a T changed to an S in German, depending on what we think the common ancestor language looked like. In addition to modern languages, linguists have also been able to use written records of dead langauges, where they exist, to assist in the drawing of language family trees. 

In addition to evolution, there are two other ways that languages are produced -- by planning and by unintentional overlaying.   Planned languages can be created from scratch (see constructed languages) or they can be modifications of original languages, as when a standard language is created.   Pidgins and creoles are created when a people from one language group attempt to communicate with another group by imperfectly learning their language. 
 

How much of this Tok Pisin sentence can you understand?

Mi laik givim    dispela     samting   long   yu.

Does it help see the English words from which these words were probably taken?

ME LIKE GIVE-HIM THIS-FELLOW SOMETHING BELONG YOU.

Here is the actual meaning of each individual word:

I  WILL GIVE    THIS         THING     TO     YOU.

And here is the actual translation:

I SHALL GIVE     THIS                  TO     YOU.

In Tok Pisin,  LAIK indicates “future”, IM is an ending that indicates a transitive verb, and PELA is an adjectival ending. 

A pidgin is probably the closest thing to a combination of human languages.  Usually the pidgin dies out when contact between the two peoples ceases.    Occasionally, a pidgin language becomes so popular that it gets used by more than just those who want to talk to foreigners, and it survives as a creole.   Actually, the above sentence example is from the Tok Pisin Creole spoken in Papua New Guinea.    Because there are over 700 languages spoken on this Melanesian island, Tok Pisin (“talk business”) has become an important tool for communication between the villages outside cities.   Unlike a pidgin, a creole language, such as Tok Pisin, has native speakers.   That means that people grow up speaking creoles, whereas everybody has to learn pidgin languages as a second language.    Another well-known example is Haitian Creole, which was based on French.   French is the official language of Haiti, but only 20% of its population speak it.  Creole is spoken by everybody.   American Black English can be considered a creole language, based on English and the native languages of the slaves brought over from Africa.   However most Black English is intelligible to speakers of English only, so it might be considered a dialect, instead.
 
 

Most Volapük speakers learned Esperanto shortly after it was published in 1887 by L.L. Zamenhof as "Lingvo Internacia".   Zamenhof used the pen name "Dr. Esperanto" (one who hopes), and that name caught on for the language.  Esperanto is an a posteriori constructed language, meaning that it borrowed words and grammar from existing languages.  But Esperanto also regularized the irregularities of these languages, and added its own regularities, and so it could be considered somewhat a priori as well.  For example, in Esperanto, all nouns end in o, all adjectives end in a, and all adverbs end in e. For more information about Esperanto, visit the Esperanto League for North America or the Universal Esperanto Association.

Have you heard the bad news?   Latin is dead?   No, the news is not that Latin, Etruscan, and Ancient Egyptian are dead.   The sad news is that thousands of modern languages are dying, and many are dead already.   Perhaps you have heard that Cornish, Pictish and Norn are extinct and that Scottish Gaelic and Manx are endangered.   Unfortunately, this is the case with hundreds of non-European languages as well.   According David Crystal, 500 of the world’s languages have less than 100 speakers, and 1500 have less than 1000.   In most cases, young people are not learning these languages from their elders, who are dying off.  And most of these languages will be dead within a generation.   Out of perhaps 6,000 currently existing languages, linguists have estimated that anywhere from 25% to 80% will be dead by the next century.

Is this the end of Babel, the end of communication problems?   Yes, people are speaking fewer languages, but no, they are not switching over to a single language -- not just to English.   Minority peoples tend to neglect their native language and learn the language of the dominating culture, whether that means switching from Provençal (southern France) to French, from Ainu (Hokkaido) to Japanese, from Berber (Morocco) to Arabic, or from Yakut (Northeast Asia) to Russian.   Needless to say, most of the native languages of both North and South America are disappearing, to be replaced by English and Spanish.  These dominant languages are not likely to die very soon or merge into one another (see Pidgins).   One possible solution is a constructed language like Esperanto, which is intended as a second language for everybody.

Dominant languages are only one cause of language loss.  More efficient transport and communications, globalization of culture, globalized business incentives, and even official bans on the use of minority languages are other causes.    But what do people lose when they lose their native language?   They lose their culture.   They lose their identity.  They lose their history.   What does humanity lose?   Each time another language dies, another chance to understand the essence of language itself is lost.    And so is another lens through which to look at our own place in the universe.

All people without severe language disorders learn a first language.    Some people go on to learn a second, or even several other languages.    They are called bilinguals or multilinguals (polyglots) -- not linguists.   But the process of learning a first language is usually vastly different from the learning of other languages.   For one thing, people learn their first language in childhood, while they are still acquiring other elementary intellectual skills.  For another, children learn their first language without the help -- or hindrance -- of another language.

First language learning is usually called language acquisition.   It is mostly automatic, and children will learn without direct instruction.   But as you probably are aware,  children are often given a lot of help and correction in learning their first language.   Some children learn a second language at the same time as their first, for instance, if each of their parents speaks a different language to them.   But learning a language beyond your first usually requires study, including reading, and may not include the same kind of feedback (practice) that first language learners get.   Most people learn their second language in school, and are able to devote only limited time to this study, and many have no way to practice beyond the classroom.   This is one reason why high school students learn a second language poorly and forget it quickly. 

Some people learn their second language “in the street,” for example, immigrants and refugees who end up in a country that does not speak their language.   They may never learn to read the new language, and they may be forced to acquire it in the way they learned their first language --by listening, guessing, and imitating.   But almost all these people have foreign accents when they speak their second language.  Why?

First language acquisition might be compared to learning to walk, and second language learning compared to learning a new dance step.  Children learn to walk and talk at very young ages, without understanding how they are doing what they are doing.   When you learn another language or a dance step, your instructor usually explains what you are to do.   Only then do you follow your instructor’s example.    You can see how important your first language is -- it can be a means of acquiring many skills, from dancing to brain surgery to speaking another language. Your first language is so essential that it can even shape the way you approach many activities, including learning a another language.  English speakers often ask: how do Chinese people get along without a plural?   And Chinese speakers often ask: how do English speakers remember to count how many things they are talking about and use the plural whenever it is required?

When you learn to walk, you also learn how to stand up, how to coordinate your two feet, how to change direction, and even how to swing your arms in unison for balance.   First language acquisition also involves a lot of skills we are unaware of, even after we’ve mastered another language.  In fact, some of the skills of speech may be instinctive, just like some of the skills of walking.   Just as humans instinctively learn to walk on two feet instead of four, humans use the double articulation in speech, something that no other animals use in their symbolic behavior.   You’d probably be shocked if your instructor told you that in your new language each word is a completely new sound, not a sequence of shared sounds!  There is much controversy today in linguistics on just what is instinctive and what is learned about languages, and whether grammars can be of any help in resolving this question.

Experiments have shown that infants know a lot more about language than they let on, especially when it comes to sound perception.    In fact, infants can distinguish many language sounds that their parents can’t.   It turns out that in learning to recognize and produce sounds in their first language, babies have to unlearn many sound distinctions that they might have used if they had been born in another language milieu.  In learning their first language, infants acquire a kind of categorical perception which they were not necessarily born with.   It becomes difficult to eliminate or adjust these categories when they learn a second language.   The same could be said of grammatical categories like singular and plural.   This is why so few people learn to speak a second language without an accent.   Learning a second language is not exactly like learning a new dance step, because people don’t normally grow up in a society using only one step and basing much of their lives on it.   Only auxiliary (constructed) languages like Esperanto don’t have a large group of native speakers who will always know the language better than you can ever learn it.

In many ways, languages are like living creatures.   They are born, they grow, they die.   Sometimes they produce daughter languages.   Nineteenth century linguists overemphasized this analogy, and occasionally fell into erroneous romanticizations.   However there is an interesting parallel between linguistic evolution and biological evolution that should probably be stressed more than it is.   As you may know, the capability to fly has evolved separately several times in the history of life on earth.   Both bats (mammals) and birds (avians) can fly.    But even though there are many similarities in the bodies of these species, biologists know that the species are so distantly related that their flying abilities are not connected.    Bats did not evolve from birds or vice-versa, although they do have a common non-flying ancestor, now extinct.  Yet independently from one another, birds and bats evolved the ability to fly.  But how do scientists know that birds and bats are not closely related?   Aside from some obvious differences in body design, they rely on DNA, in a similar way that doctors use DNA to determine paternity.

And the distinction between apparent family resemblances, such as wings, and true family resemblances through DNA can be found in language as well.   Most so-called “typological similarities” are analogous to external similarities in animals, whereas phonetic similarities in cognate words are analogous to DNA.   You can get an idea of how cognate words are used in constructing family trees in the section on language change.

Typological characteristics are general aspects of languages which are sometimes used to classify them, but not to genetically relate them.   These characteristics can be found in all aspects of language form.  In grammar, for instance, there are so-called “left-branching” and “right-branching” languages.    This distinction usually correlates to whether a language puts the grammatical object before or after the verb.  In English, for instance, when a clause is the object of a verb like know, it comes after know -- I know [you go].   In Japanese, an object comes before its verb, so the same sentence runs almost backwards --  [You go] I know.     This difference makes it difficult for an English speaker to learn Japanese and vice-versa.   And it doesn’t often happen that a language changes from one type to the other.   But just that seems to have happened in the development of Latin into Spanish.   So if you used “branching” to draw family trees, you might misclassify modern Spanish as not related to Classical Latin.   From sound similarities in cognate words and regular sound changes from Latin to Spanish, linguists know that Spanish is definitely a daughter language of Latin.

Languages can also be classified by sounds or articulatory gestures.   For example, tone languages use intonation to distinguish different words, rather than to put a kind of “spin” on their meanings as in English.   Pronounce the following allowed: me... me? me!  These sound like the Chinese words for “rice” “delude”, and “honey” respectively.   That’s because Chinese is a tone language, where intonation has no meaning, but works like English letters to distinguish different words.    English is not a tone language, and intonation may carry the meanings of “statement” “question” and “emphatic statement.”    Notice that you… you? and you! have analogous meanings in English.  In Chinese, they mean “grief”, “oil” and “again”.   In addition to tone languages, there are click languages, which produce sounds similar to the TSK-TSK you use to scold someone.   As is the case with word order, the fact that two languages use tones does not necessarily mean they are related -- Swedish, Yoruba, and Chinese independeently developed tones.

Sometimes linguists compare the occurrences of the various typological features in various languages.  They find that some features occur often with certain others, while such features never occur with others.   For example, if a language’s dominant word order is VERB-SUBJECT-OBJECT, then usually a preposition such as “in” will precede its object.  That is, if a language says “eat I cake,” then it will say “in houses.”   On the other hand, a language which says “I cake eat”, will usually say “houses in”.     These “universals of language” are interesting, but so far they have not been explained.