Chapter Para-Linguistic Usages of Clicks

by David Gil

1. Introduction

Clicks are acoustically salient speech sounds formed by the sudden opening of a closure resulting in a sharp inflow of air towards a secondary occlusion at the back of the mouth. As phonemes, the basic building blocks of words, clicks are very limited in their geographic distribution. Phonemic clicks occur only in southern Africa, plus a little pocket further to the northeast, in Tanzania; elsewhere in the world, they are absent. The limited distribution of phonemic clicks is displayed in Map 19A.

However, click sounds are actually very common throughout the world, in a variety of usages which might be termed para-linguistic. A familiar example is the English sound which, when repeated, is spelled either as tut tut  or as tsk tsk.

The para-linguistic nature of such clicks is reflected in a number of exceptional features. Phonetically, they involve sounds lying outside ordinary phonemic inventories. Grammatically, they are not integrated into morphological and syntactic structures. And semantically, they convey a very restricted range of meanings, some of which are associated with the expression of emotions.

The English tut tut  is a dental click, often repeated two or more times, and is most commonly used to express feelings such as irritation, impatience or disappointment. This usage may accordingly be characterized as affective; more specifically, as expressing negative affect. For some but not all speakers of English, the repeated dental click may also be used to express a very different range of emotions including amazement and appreciation; one context in which this occurs is that of men engaged in "girl-watching". This usage may be characterized as expressing positive affect. Whereas most speakers of English are consciously aware of the affective tut tut, there is an additional usage of a single dental click, typically immediately preceded by an opening of the lips, which occurs in generally subliminal fashion, without impinging on the consciousness of speakers and hearers: this is to mark the beginning point of a conversational unit, often in conjunction with the act of turn-taking. This usage can be readily observed world-wide on television news broadcasts such as CNN, in which the newscasters and reporters typically begin a stretch of speech with one of these clicks. In addition to dental clicks, some speakers of English make use of other clicks, either lateral or palatal, when addressing babies or domesticated animals, in order to attract their attention or to encourage them to engage in specific activities, such as, in the case of horses, running.

Para-linguistic clicks resemble other linguistic signs in that they are arbitrary and conventionalized. As such, they differ in both form and usage from one language to another. Accordingly, when speakers of one language travel to other countries, different usages of para-linguistic clicks may lead to misunderstandings. For example, an English speaker visiting Israel and asking a hotel receptionist if there are any rooms available might be answered with a single dental click. The English speaking guest may interpret such a click as expressing negative affect, and be bewildered by what seems to be an inappropriate and perhaps impolite response on the part of the hotel receptionist. However, in Hebrew, unlike in English, a single dental click has little or no affective value; rather, it has a simple logical meaning, namely the expression of negation. Thus, the hotel receptionist was simply saying 'no'. In fact, the use of a dental click to express negation is characteristic not just of Hebrew, but of many Arabic dialects and other languages. However, in the San’ani dialect of Arabic, the dental click is used not for 'no', but rather for 'yes'; it thus expresses affirmation (Samia Naim, Martine Vanhove p.c.). Because such click sounds are para-linguistic, people sometimes leap to the assumption that they are universal, transcending ordinary linguistic boundaries. However, as examples such as these clearly show, para-linguistic click sounds vary from language to language and from place to place.

The para-linguistic usages of click sounds around the world were commented on by Charles Darwin (1872), in his treatise on The Expression of the Emotions in Man and Animals, where he refers to them as "clucks" and describes their occurrence in conjunction with particular gestures:

“The throwing back of the head with a cluck of the tongue is said to be used as a negative by the modern Greeks and Turks [...] The Abyssinians, as I am informed by Captain Speedy, express a negative by jerking the head to the right shoulder, together with a slight cluck, the mouth being closed.” (p. 274) 

“With the Hindoos Mr. H. Erskine concludes from inquiries made from experienced Europeans, and from native gentlemen, that the signs of affirmation and negation vary - a nod and a lateral shake being sometimes used as we do; but a negative is more commonly expressed by the head being thrown suddenly backwards and a little to one side, with a cluck of the tongue. What the meaning may be of this cluck of the tongue, which has been observed with various people, I cannot imagine.” (p. 275)

“According to three other observers, the Australians often evince astonishment by a clucking noise. Europeans also sometimes express gentle surprise by a little clicking noise of nearly the same kind.” (p. 286)

2. Feature values

This map displays some of the ways in which the usages of para-linguistic click sounds vary across the world's languages. Specifically, the map portrays the distribution of two usages of para-linguistic clicks, affective, expressing either positive or negative affect, and logical, expressing either affirmation or negation. It does not show any of the other usages of clicks, such as for turn-taking or addressing babies and animals. In addition, it does not provide any information on the phonetic properties of the clicks, such as whether they are dental, lateral or palatal.

Values of Map 142A. Para-Linguistic Usages of Clicks
Clicks may express logical meanings ('yes' and/or 'no') 47
Clicks may express affective but not logical meanings 71
Clicks may express neither affective nor logical meanings 25

In languages of the first type, there are para-linguistic clicks with logical usages expressing affirmation, negation, or both; in addition, such languages may or may not have clicks with affective usages. One example of such a language, mentioned above, is Hebrew. Another is Avar, in which a lateral click expresses negation (Konstantin Kazenin p.c.). In contrast, in Somali, a lateral click expresses affirmation (Giorgio Banti p.c.). In Mauritanian Fula, there is a contrast between a single dental click expressing affirmation, and two such clicks expressing negation (Alassane Dia p.c.); both of these clicks are produced with closed mouth and an acoustically salient coarticulation whose nature I was unable to identify. Some other languages with logical usages of para-linguistic clicks include Dyula (Cathy Crockford p.c.), Beja (Martine Vanhove p.c.), Sardinian (Michele Loporcaro p.c.), Georgian (Manana Bat-Hana p.c.) and Malayalam (M.T. Hany Babu p.c.).

In languages of the second type, there are para-linguistic clicks with affective usages but not with logical usages; in such languages, clicks may express positive affect, negative affect, or both. English is an example of this type. Another is Khumi, in which a single palatal click expresses negative affect; there is even a special verb kədaŋ, denoting the making of this sound (David Peterson p.c.). In Minangkabau, there is a contrast between a single dental click expressing negative affect, and multiple dental clicks expressing positive affect (own knowledge). Other languages of this type include Swahili (Ellen Contini-Morava p.c.), Japanese (Daniel Long p.c.), Kalam (Andrew Pawley p.c.), O'odham (Virgil Lewis p.c.) and Huamelultec Oaxaca Chontal (Loretta O’Connor p.c.).

In languages of the third type, there are no para-linguistic clicks with either affective or logical usages. In such languages, there may be no para-linguistic clicks whatsoever, or there may be such clicks, but with other usages, such as for turn-taking or addressing babies and animals. Some examples of languages of this type include Ket (Ed Vajda p.c.), Lahu (Jim Matisoff p.c.), Mudburra (Patrick McConvell p.c.), Nez Perce (Noel Rude p.c.) and Hupda (Pattie Epps p.c.).

3. Geographical distribution

As is evident from the map, the distribution of para-linguistic clicks forms a geographical pattern on a global scale. In fact, the present map dovetails with Map 19A showing the distribution of phonemic clicks. Combining the two maps, languages may be characterized in accordance with the functional load of click sounds:


Map 19A types 2 and 6

phonemic clicks


this map, type 1 

para-linguistic clicks: 



this map, type 2 

para-linguistic clicks: 


this map, type 3 

para-linguistic clicks: 


In the above diagram, the arrow at the left represents decreasing functional load of clicks. The functional load is highest in languages with phonemic clicks, decreasing in the three types of languages portrayed in this map, in the order in which they are presented here.

Looking at the two maps, a striking pattern emerges: the functional load of click sounds decreases as one moves out of southern Africa, northwards through the rest of Africa and into Europe, and then eastwards across the Eurasian continent, southeast to New Guinea and Australia, and across the Bering Straits into the Americas.

As shown in Map 19A, and mentioned once again at the beginning of this chapter, languages in which clicks are associated with the highest functional load, namely as phonemes, are found exclusively in southern Africa, with an additional enclave in Tanzania. A single, isolated quasi-exception to this generalization is provided by the Australian language Damin, in which phonemic click sounds occur in a special speech style associated with initiation rites.

Among the remaining languages, the ones in which para-linguistic clicks bear the greatest functional load are those of the first type, in which clicks are used to express the logical meanings of affirmation or negation. As is evident from the present map, these languages occur in a single large-scale contiguous region stretching across western and northern Africa, southern Europe, and southwestern and southern Asia: more or less from Senegal to Bengal. A single exception, which proves the rule, is provided by the creole language Sranan, which clearly reflects a recent trans-Atlantic migration from West Africa. 

Within languages of this first type, finer distinctions not displayed in the map provide further support for the generalization that the functional load of click sounds decreases as one moves away from Africa. Languages with contrasting clicks for affirmation and negation, such as Mauritanian Fula mentioned above, are limited to Africa. In Eurasian languages of this type, there is at most one click expressing a logical meaning. In fact, within languages of this type, there is a large continuous band consisting of languages which, like Hebrew discussed above, make use of a dental click to express negation; this isogloss extends from Morocco in the west, across the Mediterranean and southern Europe, into the Middle East and parts of the Caucasus, and then further east, tapering out in the South Asian subcontinent. 

Languages of the second type, in which para-linguistic clicks have affective usages but not logical ones, are widespread in many different parts of the world. From a geographical perspective, they would seem to constitute a kind of default background upon which other isoglosses may be superimposed in particular regions. 

Languages in which para-linguistic clicks bear the least functional load are those of the third type, in which there are no para-linguistic clicks with either logical or affective usages. As is evident from the map, there are no such languages in Africa or for that matter in Europe; progressing eastwards, the first occurrences of such languages are in northern Siberia and in Burma. Such languages predominate in northeastern Asia and in North America; they also occur, interspersed with languages of the second type, in Southeast Asia, Australia, and South America.

4. Theoretical issues

As this map and Map 19A show, the functional load of clicks decreases as one moves out of Africa, north, then east, and then south across the world. How can this remarkable pattern be explained? Coincidence is unlikely: the geographical patterns are too dramatic to be attributed to mere chance.

One cannot avoid the observation that the functional load of click sounds decreases more or less consistently as one follows the reconstructed trajectory of the ancient migrations which brought mankind out of Africa to populate the rest of the world. But what can this mean? Geneticists cite the genetic diversity of Africans to support the claim that mankind originated in Africa. By the same token, then, one could argue that the functional diversity of click sounds in Africa is evidence that click sounds originated in Africa and spread out from there to the rest of the world. However, even if one accepts both conclusions, it does not necessarily follow that the two events were contemporaneous, that is to say, that clicks originated with mankind, and accompanied man's earliest migrations out of Africa. Click sounds could have originated much later, or for that matter much earlier, as part of a pre-human proto-language; at present we just do not know. 

It is often assumed that large-scale geographical patterns such as those depicted in this map must reflect events in ancient pre-history: either inheritance from a common ancestral language, or else contact between ancient languages. This assumption presupposes that the feature in question is endowed with a sufficient degree of diachronic stability to enable geographical patterns, once they emerge, to be maintained over the course of time. However, there is at least some evidence suggesting that the various usages of click sounds may not be impervious to recent borrowing. For example, in southern Africa, recently arrived Bantu languages such as Xhosa borrowed phonemic clicks from the indigenous Khoisan languages. Similarly, the large-scale isoglosses of this map occasionally cross-cut closely related languages. For example, within Romance, Italian has logical usages of para-linguistic clicks while French does not, and within Slavonic, Serbian-Croatian has logical usages of para-linguistic clicks while Czech does not. Likewise, within the Kuki-Chin branch of Tibeto-Burman, Khumi has affective usages of para-linguistic clicks while Meithei does not, while within the Aslian branch of Mon-Khmer, Jahai has affective usages of para-linguistic clicks while Semelai does not. Indeed, an epidemiologist might easily produce a map showing similar large-scale patterns reflecting the world-wide spread of, say, a virus within the course of mere years. What these observations suggest, then, is that at present we are still far from being able to determine whether the geographical patterns evident in this map are relatively old or comparatively recent, and hence even further from being able to come up with an adequate explanation for their existence.