Kham offers new insights into the nature of Himalayan tonal systems. For most languages of the region, the domain of tone has been described as the word (‘word-template tone’); a word being a root with its full array of affixes (Hale and Pike 1970, Hari 1980). Usually only the root is contrastive for tone, and affixes, being atonal, get their tonal character from the root. Kham, on the other hand, though morphologically more complex than other Himalayan tone languages with adequate descriptions, still retains much of its monosyllabic tonal character. All roots and most affixes still have their own inherent tonal status, and though in specific contexts the tonal character of some morphemes is suppressed or deleted, the underlying tones are still recoverable.

It is precisely in the area where affix syllables are beginning to lose their individual tonal character that Kham may provide insights into the nature of a tension that appears to be implicit in tone systems of the region. Though other Himalayan languages with relatively ‘old’ tonal systems have undergone considerable tonal levelling, Kham, with perhaps more detail than we have seen elsewhere, attempts to maintain the underlying tonal character of individual morphemes wherever possible, but is forced to give up some of its distinctiveness where tones collide and come into conflict.