Historical Implications of Click Consonants in Hadza

Click languages are renowned for the novel, popping sounds that speakers use as a part of everyday conversation, a reputation due probably to how rare these sounds are within the world's linguistic inventory. An indication of their scarcity lies in the fact that click languages belong to about thirty groups of people in the world, most of which are found in Botswana, Namibia, South Africa and a few other neighboring countries. In fact, the only non-African click language ever recorded was a now extinct Australian language called Damin, but even then clicks in this language were only used ceremonially during certain rituals. You can hear the pronunciation of some basic click sounds here.

A central theme in the study of click languages, which all have in common this very rare phonetic feature, is their genetic affiliation to each other. In the 1960s, when the influential linguist Joseph Greenberg began studying the African click languages, he grouped all of them under one language family - the Khoisan family - which he named after the two major tribal groups that spoke these various languages; the two groups consisted of herders called the Khoe and hunter gatherers called the San. Since then, however, analysis has shown that Greenberg's definition of the Khoisan family can actually be broken up further into at least three distinct language families, while some of the languages classified as Khoisan don't appear to fit any known language families. One of these isolates is Hadza, a language spoken in Tanzania by about 800 people. Hadza is typologically unusual because of the fact that it's one of just a few languages outside of southern Africa to use clicks (Sandawe and Dahalo being the others). In fact, Hadza is located 2000km away from any other region where click languages are spoken. A combination of its geographical, lexical, and phonetic status has lead many linguists to believe that Hadza is a language isolate, which would mean that it's unrelated to any other known language in the world.

In his time, the late, great phonetician Peter Ladefoged set it upon himself to travel around the globe and document the sound systems of the world's most endangered languages. At the heart of his intentions, Ladefoged felt that a careful investigation of a language's phonology could function as an important first step in determining its relationship to other languages. For the more unusual languages out there that are hard to classify, it may be especially relevant to use phonology as a step towards assigning them to a linguistic family. This method seems to point directly at the problem of establishing a relationship among click languages, which comprise both a common phonetic articulation and relatively close geographic proximity. In addition to the research that suggests Hadza is a language isolate, investigations into the language's sound system have revealed a deeper history about its relationship to human language in general.

One of the first analyses of Hadza phonology was carried out in the 60s by the English social anthropologist James Woodburn. In his research, Woodburn recorded five basic click types for Hadza, the places of articulation of which were bilabial, dental, lateral, palato-alveolar (hard), and palato-alveolar (flapped). In addition, the clicks correspond to manner of articulation, which are simple, pausal, aspirated, nasal, and nasal compound.

On the chart above, the places of articulation run horizontally just like on an IPA table, while the manners of articulation, which run vertically, were devised by Woodburn for the purpose of describing Hadza clicks specifically. In total, Woodburn distinguished between twenty-one click sounds in Hadza. However, a more recent investigation, carried out in the early 90s by Ladefoged, reaches a different conclusion about the click inventory of Hadza. Ladefoged's findings indicate that Hadza has three basic click types, those being dental [ǀ], alveolopalatal central [!], and lateral [ǁ]. In addition, each of these click types has three separate accompaniments, making for twelve click sounds total. The accompaniments are voiceless oral [kǀ, k!, kǁ], voiced nasal [ŋǀ, ŋ!, ŋǁ], and voiceless nasal with glottalization [ŋ°ǀ', ŋ°!', ŋ°ǁ']. Perhaps this is a bold assumption, but, considering the discrepancy between Woodburn's and Ladefoged's research, I'm willing to bet that the latter is more credible, given Ladefoged's wide renown for phonetic analysis of endangered languages. At least one aspect of Ladefoged's research that helped him succeed in creating accurate phonetic descriptions of the languages he studied is his extensive use of audio recordings that captured the actual pronunciation of native speakers. You can click here to listen to a file of Hadza speakers from the UCLA Phonetics Lab Archive. And here is a word list for the recording, which covers words 1-16, showing the Hadza words and glosses in English and Swahili. In the recording, you can hear Ladefoged and an assistant speaking words in Swahili to a group of Hadzabe, who each respond with the equivalent Hadza word.

It turns out that the research on Hadza clicks, which differentiates it from other click languages, may have deeper implications. Based on these differences, we have to assume either that Hadza has an entirely separate origin than the other click languages, or that it and the other click languages derive from a very ancient proto-click language, which would have to be from so long ago that today no similarities between the languages are perceivable. But, at the same time, the linguistic (and geographic) differences between Hadza and the other click languages aren't enough to confirm either hypothesis. So, in order to determine the origin of the Hadza people and their language, genetic analysis has become a very informative technique.

Originally, researchers believed that the Hadzabe must be genetically similar to other click speakers and that they migrated to Tanzania relatively recently, while keeping their own language intact. However, the more recent genetic analyses of mitochondrial DNA and Y chromosomes in the Hadzabe reveal two things to suggest otherwise. First, the DNA sequences of the Hadzabe are very diverse, which is an indicator of early origins. This is because DNA mutations accumulate over time, so greater sequence diversity implies older existence. Secondly, the analyses conducted reveal that the Hadzabe are not at all closely related to the click-speaking San group, !Kung groups from Namibia and Botswana, nor are they related to non-click-speaking groups from central Africa. So, what this means is that the Hadza people were most likely in Tanzania all along, the implication being that they are the descendants of a very ancient pool of click-speakers that originated separately from the others. There are various suggestions as to the time frame in which the Hadza split off from this group; some postulate that it could have occurred as early as 100,000 years ago, but most place the split somewhere between 50,000 and 70,000 years ago, which also happens to be around the same time of the exodus of modern humans out of Africa, an event that some linguists believe to have been the driving force behind the development of human language itself.

Ultimately, the historical implications one can gather from the linguistic and genetic evidence surrounding the Hadza people and their language are enormous. If the predictions are accurate, and the Hadzabe - along with their click sounds - truly did originate over 50,000 years ago in what is today Tanzania, then we may very well have pinpointed something concrete about how the very first human language would have sounded. Now, the research that makes this claim is admittedly somewhat speculative, and the correlation between genetics and linguistics may not be cohesive enough to uphold that claim, but when I listen to that recording of the Hadzabe speaking and hear the unusual staccato of their clicks, protruding from a flow of otherwise pulmonic sounds, I can't help but think that I've just caught a glimpse of the language of our oldest ancestors.

References