Classification of tonal patterns in languages can quickly become a complex conundrum to solve. One version of that problem can be that the impact of tonality is too weak for there to be clear boundaries (as in often so-called ‘pitch accent languages’). In Khoekhoe, the most widely spoken language of the Khoisan sprachbund, however, linguists have been dealing with the opposite: The interplay of the tonemes is so intricate that, after over a century of research, not even their number has been firmly determined yet. While scholars nowadays universally agree that Khoekhoe features a categorical six-way contrast of tonal patterns in citation form lexical roots, there has been and still is an ongoing dispute as to how these melodies should be treated theoretically. After D.M. Beach had firmly established the six-melody analysis in 1938, linguists such as R. Hagman and W. Haacke have suggested since the 1970s that those melodies can be broken down into three or four register tonemes, respectively. The six melodies are then treated as combinations of two level tones, which goes with the assumption that the language’s canonical bimoraic lexical roots in fact feature two tbus instead of just one.
The issue is all but resolved as of yet, with almost each author who has publications on the topic having proposed their own interpretation of the tone system or parts of it. I will therefore try to give you a comprehensive overview over the most central approaches, and show you that the problem of tone in Khoekhoe is rooted very deeply in the entire phonotactics of the language. It will all obviously be very open for discussion, so I’m thrilled to hear your opinions.
 tone bearing units