Click consonants are perhaps among the most unique sounds to have ever been used in languages, especially in word formation and beyond. After all, they use a particular type of airstream mechanism, which we have mentioned before when covering how we talk. The mention of click consonants has appeared a couple of times before on The Language Closet, but it has taken me quite some time to realise that, in all the 9 years I have been writing on this platform, I have never actually looked into the languages that use them.

And so today, we will begin the series on click languages with the languages that are extremely closely related to it. Predominantly spoken in Namibia, Botswana, and South Africa, these languages are collectively called the “Khoisan” languages. Now we must note that the term “Khoisan” is merely one of convenience, just as “Papuan” is one when referring to the non-Austronesian languages spoken on New Guinea and Bougainville (and perhaps wider Melanesia), and “Australian” is one when referring to the indigenous languages of Australia. Here, the commonality amongst the “Khoisan” languages are the prolific use of click consonants.

Sure, click consonants can be found in other languages in the region such as IsiXhosa, IsiZulu, and Sesotho, but these click consonants might have been borrowed from speakers of the “Khoisan” languages. However, other languages that use click consonants include the Hadza and Sandawe languages predominantly spoken in Tanzania, and the Dahalo language in Kenya. The only language that used click consonants outside of Africa is the ceremonial register called Damin, once spoken by Lardil men who have been twice-initiated. Together, click languages constitute only around a couple dozen languages, making this class of consonants among the rarest phonemic sounds in human language.

Initially, the Hadza language was once grouped as “Khoisan” likely due to the presence of click consonants, while the Sandawe language has been proposed to be related to one of the member of the “Khoisan” languages. In any case, the term “Khoisan” today does not impart any genetic relationship between these languages, as linguists aimed to establish a classification of this category of click languages towards the end of the 20th century and the early 21st century.

As mentioned, the term “Khoisan” does not impart any genealogical relationship between the member languages. Instead, it includes two or three distinct language families, which usually comprise dialect clusters or dialect continua. Collectively, these languages are spoken by perhaps around 300 000 people today. A considerable proportion of these languages are unfortunately extinct, with generally no documented plan for revitalisation. These families include the Kx’a languages, the Tuu languages, and the Khoe-Kwadi languages.

Kx’a languages

The Kx’a languages are the newest group of languages identified in the “Khoisan” languages, which was proposed by Heine and Honken in 2010. Consisting of the ǂʼAmkoe and !Kung languages, this small language family was proposed following previous studies attempting to reconstruct the likely common ancestor of the two languages. Proposed sound changes included those of the click consonants, such as those proposed by Bonny Sands in her 2014 presentation at a conference.

While ǂʼAmkoe today is only spoken by a few dozen people, the !Kung language is spoken by around 75000 people. This makes !Kung the third most-commonly spoken click language after excluding the Bantu languages with click consonants, only behind Sandawe and Khoekhoe or Nama. As the names for both languages suggest, they start with a click consonant.

Firstly, let’s take a look at the ǂʼAmkoe language. Spoken in Botswana, there are 3 identified dialects, which are the N|aqriaxe, ǂHoan, and the Sasi dialects, with the former two grouped as West ǂʼAmkoe, and Sasi also referred to as East ǂʼAmkoe. The ǂʼAmkoe language is one of the 3 languages spoken today that use the bilabial click consonant, that is, a consonant that sounds like smacking lips. The other two languages that have this family of consonants are the Taa language and the N||ng language, both of which are Tuu languages.

!Kung, on the other hand, has around a dozen varieties which can be grouped into three or four clusters. Some of the more notable varieties include Sekele in the Northern !Kung cluster, and Ju|’hoan in the Southern !Kung cluster. Nevertheless, !Kung, like many languages here, has a large number of consonants, and even more so when complemented by the click consonants. In Ju|’hoan, for example, stop consonants can take on a combination of secondary articulation, phonation, and airstream mechanisms, making uvular ejective stop consonants like [tᵡʼ] possible. Click consonants may also be murmured or even epiglottalised, just to mention a couple of the 12 series of combinations that may complement the plain click consonant.

Tuu languages

The Tuu languages are a family of around a dozen or so languages, but with what little information we have to work with especially for the extinct Tuu languages, the lines dividing between language and dialect are pretty blurred. Nevertheless, linguists have identified two distinct branches, each with only one surviving member language. These are the moribund N||ng, with fewer than a dozen native speakers, and the endangered Taa language, with around 2500 native speakers. The latter is well known for having one of the largest phonological inventories of any known language in the world.

As mentioned before, the two extant Tuu languages are the only other languages in the world today to have the bilabial click consonants. A large majority of consonant sounds in both languages consist of clicks, with some dialects consisting as few as 40 or 50 click consonants alone, and perhaps as many as 111 click consonants depending on how you analyse them. Defining what consists a particular click consonant is a rather contentious issue, especially when analysing language like Taa and N||ng.

Even amongst the non-click consonants, some dialects in Taa still distinguish between a large number of them. Voiced, voiceless, aspirated, glottalised, and velarised versions of consonants exist across pretty much 5 places of articulation, plus several more consonants articulated with the hard palate and the glottis.

Not only that, both languages also have an interesting set of vowel sounds. Even though they only have the vowels /a e i o u/, the complexity arises when we talk about the various types of phonations that these languages feature. This includes nasalisation, glottalisation, murmuring, phayrngealisation, and epiglottalisation. Both languages also feature phonemic tone, with anywhere from 2-4 tones depending on the dialect.

Khoe-Kwadi languages

Now we come to the largest language family under the Khoisan umbrella. From the get-go, the Kwadi language here is seemingly poorly attested that went extinct in the mid-20th century. Nevertheless, linguists were able to collect enough data about the purported language to establish some relationship with the Khoe languages. Like the other language families mentioned previously, the Khoe languages consist of dialect clusters that makes distinguishing languages rather challenging.

Nevertheless, a total of 9 main dialect clusters were identified, split into the Khoekhoe and Kalahari branches, which subsequently split into the North and South Khoekhoe, and the East and West Kalahari languages respectively. Glottolog 5.0 suggests 13 of these languages however, showing how classification differences could result in differing findings on which ones constitute a language here.

Among these, the Khoekhoe language, also called Nama, is perhaps among the lower end for the number of consonants in its inventory, with 11 non-click consonants with no voicing distinction and 20 click consonants, split into 4 places of articulation and 5 series of combinations. Nama is also the only one that has developed a decimal number system, with the other languages here borrowing either from Nama, English, or Bantu languages. A couple of the languages that lack number words beyond 4 or 5 also use the names of fingers to count 6-10, while for others, data for numbers beyond 5 are lacking.

Notable patterns

As we have seen, the three language families have several characteristics in common beyond the use of click consonants, which predominantly occur at the start of the syllable or word. Their consonant inventories are rich, and further inflated by the number of click consonants that are phonemically distinguished. Stop consonants may take on many different combinations of secondary articulations, phonation types, and airstream mechanisms, but yet, implosive consonants generally do not seem to occur.

There are generally four or five different places of articulation for the click consonants, with all of these languages generally having the dental, alveolar, palatal, and lateral clicks, and as mentioned previously, only three languages having bilabial clicks. Like the stop consonants, there are multiple series of combinations of secondary articulations, phonation types, and airstream mechanisms that may complement these click consonants, with Nama having five series (making 20 click consonants), and Taa having over a dozen different series that may differ based on cluster analysis.

Despite having a large number of consonants, there are only five vowel qualities distinguished by the languages of the three language families, /a e i o u/. Exceptions do exist as well, as Khwe, a Khoe-Kwadi language, distinguishes six vowels /a e ɛ i o u/. However, these vowels may have different phonation types, such as murmuring, epiglottalisation, and more. In Nama, there are only three nasal vowels /ĩ ã ũ/, while in Ju|’hoan, for instance, all its five vowel qualities may be nasalised, glottalised, and murmured, with strident /a/ and /o/ sounds also occurring. These languages are also tonal, with most of them having three or four tones, although these tonemes may differ from language to language and by analysis methods.

Moving onto the numbers, with the exception of Nama and Ekoka !Kung, which have a number system up to 10, many languages here have a restricted number system, with native numbers generally not exceeding ‘five’. Taa, for example, has number terms up to three, which are ǂʔûã (one), ǂnûm (two), and ǁâe (three), with larger number terms being borrowed from languages such as Tswana and Kgalagadi. Other languages may also borrow number terms for larger numbers from other various Bantu languages, and even English and Afrikaans.

There are perhaps more patterns in phonotactics, grammar, and syntax that can be drawn by comparing these languages in their respective language families, but for brevity, this are pretty much the notable examples of similarities that are shared between the languages making up the convenient grouping of the ‘Khoisan’ languages. Exceptions to the ‘norm’ do exist, such as Nama’s relatively small consonant inventory compared to many other languages.

With this, I am interested in learning more about the origins of the click consonants. There are narratives suggesting how click consonants might be the first consonants to emerge in human language, but are refuted by other hypotheses that these sounds are novel innovations, pointing towards their complexity and the rather specific geographic area in which they occur. I want to read up on the arguments for and against such narratives, and synthesise them in a future essay here on The Language Closet.

Further reading