Lesson 2: Languages of Sub-Saharan Africa

• Learning Objectives
• Listening Assignment
• Lesson
• Summary

Learning Objectives

After finishing this lesson, you should be able to:

1. make a general and accurate statement about the size of the Khoisan group in terms of number of languages and number of speakers
2. identify the IPA symbol for each of the five major clicks, as well as the technical and non-technical description of each one
3. make a general and accurate statement about the size of the Niger-Congo family in terms of number of languages and number of speakers
4. describe the organization of Niger-Congo noun class systems, including the number of classes and the basis for noun categorization, the use of noun class prefixes, and the enforcement of noun class agreement within a sentence

Listening Assignment

Cox, Patrick. 2013. Podcast “From Afrikaans to Zulu”. The World in Words. Public Radio International.

© 2008 PRI. Used under Fair Use provisions for educational purposes.

Listening Guide

This listening is an episode from the podcast The World in Words, published by Public Radio International. It describes the situation of the South African language Zulu in the context of colonial South African languages introduced by Europeans.

Background Information

1. You should have a basic understanding of what apartheid was. If this is a new concept for you, read the first several paragraphs of the Wikipedia entry.
2. English and Afrikaans are both Indo-European languages; they are unrelated to the languages of South Africa that existed prior to the arrival of Europeans.

Reading Questions

1.  What happened to the status of English and Afrikaans after apartheid?
2.  Why do some black South African parents object to the language instruction in "good" schools?
3.  What is the motivation behind Phiwayinkosi Mbutazi's efforts to modernize Zulu?
4.  What are some perceived problems with Zulu in its current state?

Lesson

An Overview of Sub-Saharan Language Groups

In this lesson, we will focus on two language groups from Sub-Saharan Africa: Khoisan and Niger-Congo. Between the two, Niger-Congo contains languages that many non-linguists have heard of, including Swahili, Zulu, Shona, and Yoruba. With around 1,500 languages, it is the largest language family in the world.

The other group, Khoisan, is much smaller than Niger-Congo, but is unique because its languages are among the only ones in the world to use “click” speech sounds.

The map below shows ethnocultural groups in Africa. Our focus in this lesson will be on the language groups in the lower two-thirds of the continent. The pale yellow portion in the upper third will be covered in Lesson 3.

Khoisan Languages and Click Sounds

Khoisan (also spelled Khoesan and Khoi-san) is the name for a set of languages from southern Africa with unique and rare consonants called clicks. The number of Khoisan languages was much larger in the past than it is today. Linguists do not know for certain how many Khoisan languages are now extinct, but it is likely that all or most of southern Africa was occupied by Khoisan speakers several thousand years ago.

A series of migrations collectively called the Bantu expansion occurred over thousands of years starting around 2000 BCE, during which groups of Niger-Congo-speaking peoples moved south from the middle parts of the continent. These non-Khoisan Africans moving into predominately Khoisan areas are thought to have eventually displaced, eliminated, or mixed with many Khoisan-speaking groups.

Starting in the 1600s, European colonialists began occupying southern Africa and severely reduced the Khoisan language population. Khoisan speakers were subjected to prolonged campaigns of extreme oppression, including forced labor and murder, by the descendants of Dutch immigrants. Today, only a dozen or so Khoisan languages remain, though these are often divided into various dialects that are sometimes treated as separate languages. Here is a complete list of the extant Khoisan languages (extant means that there are living native speakers), according to Khoisanist Matthias Brenzinger (2013):

As you can tell by the number of speakers in the table above, some of these languages have extremely small speaker populations. To put these numbers into perspective, consider that Navajo is the most widely spoken Native American language in the United States with a speaker population of about 170,000. Not only are the speaker populations of Khoisan languages very small, but their geographic distribution is dispersed in a pocketed and non-uniform fashion across southern Africa. As an example, look at this map of the distribution of the language Khwe-ǁ​Ani. Note that each shape represents only 100 speakers.

We are calling Khoisan a language group rather than a language family because most specialists agree that the evidence for Proto-Khoisan, the hypothetical ancestor language, has yet to be firmly established. In general usage, the term Khoisan is used to refer to a group of languages that share an extremely rare set of speech sounds: click consonants. Click consonants, or simply clicks, are speech sounds in Khoisan languages, which means they make up part of the inventory of sounds used for everyday speech.

There are five basic clicks found in Khoisan languages (and up to a dozen or so subtle articulatory variations on these five). The International Phonetic Alphabetic (IPA) has a special symbol designating each click consonant. The table below shows the IPA symbol for each click as well as its technical description and some example words containing that sound.

Click consonants are incredibly rare in the world’s languages. A sample of 567 languages worldwide turns up a map like this when it is restricted only to those languages with clicks:

Note that the map does not show all the click languages in the world, only a sample. With a few exceptions, all click languages are members of the Khoisan group. The exceptions are a few Niger-Congo languages such as Zulu and Xhosa and the Afro-Asiatic language Dahalo, none of which are related to the Khoisan languages. Those non-Khoisan languages probably developed clicks due to prolonged contact with Khoisan-speaking peoples, rather than developing the sounds independently.

It may have occurred to you that these click sounds, or something very similar, do indeed occur in English. The tsk-tsk sound when admonishing someone, the clicking noise when calling an animal or spurring a horse, and the smack-like kissing sound are all examples of clicks in English. The difference is that in English, clicks are not speech sounds—that is, they are not part of the regular inventory of sounds used in everyday speech. Unlike true English speech sounds, such as p, i, or g, the click noises do not show up in English words.

Khoisan languages, on the other hand, use clicks quite regularly and without the need for any special context such as calling an animal. For instance, below is the text of a few sentences from Ju|’hoan (Dickens 2005:37-41). Look at these sentences and practice identifying all the click symbols.

Watch the following video clip (1 minute, 31 secs) and listen to this San man explaining his society’s various uses for grass::

If you listen closely, you should be able to hear several of the clicks described in the above table. Hearing differences is one thing, but identifying each click by hearing alone takes a lot of practice for people who haven’t grown up listening to the language. You should be able to identify the IPA symbols for the five major clicks and know the technical and non-technical descriptions associated with each one, as described in the table above.

The Niger-Congo Language Family

By comparison to the Khoisan languages, the Niger-Congo languages represent a true and unequivocal language family. This means that there is strong evidence in support of an original ancestor language, Proto-Niger-Congo, which existed in the distant past and eventually gave rise to all the modern Niger-Congo languages.

Approximately 400 million people speak a Niger-Congo language. This is greater than the population of the United States, but still not close to the widest-spoken families (Sino-Tibetan has 1.3 billion speakers and Indo-European has 3 billion). But when measured by its number of languages, Niger-Congo is the largest language family in the world, with estimates ranging from 1,436 to 1,543 (again keeping in mind the qualifications on counting languages discussed in Lesson 1).

With 1,500 or so languages, a meaningful family tree is nearly impossible to reproduce on a web page, or even in print. Most sources show only the major branches of the family without showing individual languages, or else show trees for individual branches of Niger-Congo so that the specific member languages can be detailed. For instance, below is a family tree diagram showing only the principle branches of Niger-Congo. The individual languages are not shown at all.

Now look at the detailed structure within a single branch, Mande:

Just by looking at this one branch of Niger-Congo, it is easy to see how massive the entire family is. Below is a map showing the distribution of 327 Niger-Congo languages, a mere 22 percent of the total.

Several Niger-Congo languages are well known to non-linguists, such as Swahili, Wolof, and Zulu. To get a feel for the sound of Swahili, watch a news segment from BBC Swahili. (You don’t have to watch all of it, but try to do at least 10 minutes.)

Noun Classes in Niger-Congo Languages

Now that we have had a chance to get somewhat acquainted with the Niger-Congo language family, the remainder of this lesson will focus on a single grammatical feature that is characteristic of Niger-Congo languages, and one that will form the basis for many comparisons we will make over the rest of the course.

This grammatical feature is the use of noun classes. Almost all languages have grammatical systems for organizing and categorizing nouns. In Lesson 4, we’ll see that many Indo-European languages use a sex-based gender system that organizes nouns into categories like “masculine,” “feminine,” and sometimes “neuter.” Niger-Congo languages, instead of having two or three categories for nouns, have up to twenty-two noun classes, depending on the language.

Unlike in Indo-European languages, where noun categories are sex- or gender-based, the categories in Niger-Congo languages are based on properties such as whether a noun is a human or an animal and then, if it is neither a human nor animal, what kind of shape or other physical attribute it has. There are also separate noun classes for singular nouns and plural nouns.

Which noun class a particular noun belongs to is indicated by a morpheme (see Lesson 1), which, in most Niger-Congo languages, is a prefix on the noun. A noun in Class 1 will receive one prefix, but a noun in Class 2 will receive a different prefix. This is similar to how the form of the or a/n changes in Indo-European languages depending on the gender of a noun. (In German, you’d say der Wein 'the wine' but die Milch 'the milk' and das Bier 'the beer' because each of these words belongs to a different gender in German.) But in Niger-Congo languages, the prefix does not mean the. It only specifies which noun class a particular noun belongs to and it is attached directly to the noun itself.

Most noun class languages in the Niger-Congo language family exist in a very large branch of Niger-Congo, called Bantu. Bantuists have used the comparative method (recall Lesson 1) to reconstruct the noun classes and noun class prefixes for Proto-Bantu, the hypothetical ancestor language from which the modern languages in the Bantu branch descended. Below is a table showing all nineteen reconstructed noun classes along with examples from modern languages (Wald 2009).

Let’s look in detail at a language called Manjaku, spoken in Senegal and Guinea-Bissau. Manjaku has thirteen noun classes. Like most Niger-Congo languages, the noun classes are specified by a prefix on the noun. Class 1 categorizes singular humans and uses the morpheme na- as a prefix.

The word for 'thief' in Manjuku takes the basic form kiëj. But kiëj by itself is not a complete word because it does not have the noun class prefix attached to it. Since thieves are humans, this word gets the Class 1 morpheme. The full word for thief in Manjaku is nakiëj. For clarity, it is sometimes helpful to think about the word using a three-line gloss, where the abbreviation CL stands for 'class':

Class 2 in Manjaku is used for plural human nouns and requires the prefix ba-. Thus the Manjaku word for 'thieves' would be bakiëj and would have this three-line gloss:

Animals are assigned to Classes 3 and 4. Class 3 uses the prefix u- and is for singular animals; Class 4 uses ngë- and is for plural animals. With this information, if you know that the basic word for 'cat' is ndali, can you figure out the actual Manjaku words for 'cat' and 'cats'? The word for cat is undali and the word for cats is ngëndali.

Classes 5 and 6 are for plants. The singular takes Class 5 while the plural takes Class 6. The Class 5 prefix is bë- and the Class 6 prefix is m-. The basic form of 'wild mango tree' is calam.

Practice Activity

What are the Manjaku words for 'wild mango tree' and 'wild mango trees?'

View the answer

Other Manjaku noun classes refer to artifacts and material objects, bodily organs, diminutives (small versions of things), and locations. Each language has its own variation on the theme of noun classes. Languages vary in the specific number of noun classes they have, the properties and attributes that define each noun class, and specialized sub-categories of general noun classes.

Like all linguistic systems, Niger-Congo noun class systems do have exceptions. In Manjaku, for instance, the words for 'machete' (undinkI), 'smoke' (uru), 'work' (ulemp), and 'hunger' (ubon) all take the Class 3 prefix even though these are not animals. People learning the language just have to memorize these as exceptions.

In addition to requiring a noun class prefix on the noun itself, most Niger-Congo noun class languages also require other words in a sentence to “agree” with the noun. That is, words associated with the noun (like adjectives and verbs) must match the noun by having the same noun-class prefix. Which words require this agreement differs from language to language (in some languages it is just adjectives, in others it is adjectives and verbs, in others it is words meaning the or an).

In Manjaku, both adjectives and verbs must agree with a noun. That is, adjectives and verbs require the same noun class prefix as a noun in the sentence. Here is a full sentence in Manjaku:

Notice that the word for 'hyena' is uñiing. It takes the Class 3 prefix u- because hyenas are animals and in this sentence, it is singular. Notice further that the word for 'climb' also has the Class 3 prefix u- attached to it. The reason is that the thing doing the climbing is a Class 3 noun. (You can think of this as a form of subject-verb agreement.)

Here is another Manjaku sentence:

In this case, the word for 'farmers' is bajaar. It takes the Class 2 prefix because farmers are humans and here it is plural. The word for 'three' also has the Class 2 prefix because it is an adjective that specifies the quantity of a Class 2 noun. (There is no class prefix on yitir 'meet' because the farmers are not doing the meeting. The pronoun dë ‘I’ is the subject of this sentence, but pronouns do not take noun class prefixes in Manjaku like regular nouns do.)

The aspects described above of the Manjaku noun-class system are common to almost all Niger-Congo languages that have noun classes. When inspecting Niger-Congo language data, keep your eyes open for the following three properties:

1. the categorization of nouns into a large number of classes
2. the use of prefixes on the noun to specify which class a noun belongs to
3. the agreement of other words in the sentence (adjectives, verbs, and so on) with the noun by taking the same noun-class prefix.

With these considerations in mind, look at the following sentences from Swahili. In Swahili, words like that or those, as well as adjectives and verbs, must agree with the noun class of the noun they are associated with in a sentence. Compare the two sentences closely to see how the noun classes change based on switching from singular to plural nouns.

^Return to the top of the page

Summary

In this lesson, we have looked at two groups of Sub-Saharan African languages. The first group was Khoisan. We saw that Khoisan is a very small group in terms of the number of speakers, with perhaps as few as twelve languages (learning objective 1). Though not assumed here to be a language family, the Khoisan languages share the property of having click consonants as speech sounds. The IPA symbols for the clicks and their articulatory descriptions were described in a table above (learning objective 2). We then moved on to the Niger-Congo languages, which do constitute a language family. With around 400,000 speakers and 1,500 languages, it is an extremely large language family (learning objective 3).

We explored in detail one aspect of the grammatical structure of many Niger-Congo languages, mostly within the Bantu branch, which is the use of a noun-class system. Noun classes are systems of categorizing nouns. Noun-class systems tend to have a large number of individual categories (classes), which are based on qualities like humanness, size, shape, and function. Every noun is required to have a morpheme, usually a prefix, which indicates which class it belongs to. Other words in the sentence, such as verbs and adjectives, may also be required to agree with the noun by using the same noun class prefix as the noun (learning objective 4).