I argue that the inverted spectrum hypothesis is not a possibility we shoul
d take seriously. The principle reason is that if someone's qualia were inv
erted in the specified manner there is reason to believe the phenomenal dif
ference would manifest itself in behaviour. This is so for two reasons. Fir
st, I suggest that qualia, including phenomenal colours, are partly constit
uted by an affective component which would be inverted along with the conne
cted qualia. The resulting affective inversions will, given the intimate co
nnections that exist between emotions and behaviour, likely manifest themse
lves in behaviour, in which case the underlying phenomenal differences can
be functionally captured. Second, I argue that other sense modalities lack
the structural features necessary for undetectable inversion which, because
of their analogy with colour qualia, weakens the plausibility of such an i
nversion in the original case of vision.