My goal in this chapter is, first of all, to compare the verbal paradigms of various Tibeto-Burman languages, focusing on person and number agreement affixes. Some researchers, looking at the same materials I look at here, have come away with the conclusion that the patterns are sufficiently different that they must have arisen independently, albeit out of a linguistic milieu that favors the incorporation of pronominal elements in the verb. On this view, agreement marking in the verb is relatively new and any similarities are due to areal tendencies or ‘drift’ (Caughley 1982, LaPolla 1992a, 1994). In this chapter, I will favor the opposing view, that person and number agreement patterns are old (Bauman 1975, DeLancey 1980, 1989, van Driem 1991) – indeed, that some form of agreement was present in the PTB verb, and that those modern languages which still show agreement patterns do so out of conservatism. If this is true, it will have to be shown that at the heart of the modern systems is an unimpeachable core of features common to all the pronominalizing languages, and that the variability pointed to by some researchers is attributable to secondary developments.

The former challenge – finding a core of common features – is easier to do than the latter – tracing secondary developments.