The key to comprehending the tectonic evolution of the Himalaya is to under
stand the relationships between large-scale faulting, anatexis, and inverte
d metamorphism. The great number and variety of mechanisms that have been p
roposed to explain some or all of these features reflects the fact that fun
damental constraints on such models have been slow in coming. Recent develo
pments, most notably in geophysical imaging and geochronology, have been ke
y to coalescing the results of varied Himalayan investigations into constra
ints with which to test proposed evolutionary models. These models fall int
o four general types: (1) the inverted metamorphic sequences within the foo
twall of the Himalayan thrust and adjacent hanging wall anatexis are spatia
lly and temporally related by thrusting; (2) thrusting results from anatexi
s; (3) anatexis results from normal faulting; and (4) apparent inverted met
amorphism in the footwall of the Himalayan thrust is produced by underplati
ng of right-way-up metamorphic sequences. We review a number of models and
find that many are inconsistent with available constraints, most notably th
e recognition that the exposed crustal melts and inverted metamorphic seque
nces not temporally related, The generalization that appears to best explai
n the observed distribution of crustal melts and inverted metamorphic seque
nces is that, due to specific petrological and tectonic controls, episodic
magmatism and out-of-sequence thrusting developed during continuous converg
ence juxtaposing allochthonous igneous and metamorphic rocks. This coincide
ntal juxtaposition has proven to be something of a red herring, unduly infl
uencing attention toward finding a causal relationship between anatexis and
inverted metamorphism. (C) 1999 Elsevier Science Ltd. All rights reserved.