Christophe Grandsire writes:
 > En réponse à Maarten van Beek <[log in to unmask]>:
 > > To me, for instance, there is a huge difference between sky blue and
 > > dark
 > > blue, and I am still frustrated for not having seperate words for it. We
 > > do
 > > have pink and lilac to go with red, why not something to go with blue?
 > Lilac with red?!!! Although by stretching a little far, pink could be
 > considered as having strong affinities with red, lilac is at the opposite end
 > of the spectrum, with affinities with violet! Decidedly, this subject is even
 > more surprising than I thought. People seem to make really strange associations
 > about colours...
I'm no expert on the matter, and I'm going to look into it further,

In reality, yes, visible light is a spectrum stretching from red to
violet.  But that's not how we really perceive it - what our brains
get is a varying level of excitation of three types of cones* in the
retina, which peak at different wavelengths**.  So human perception of
colour is more like one of those triangles we were talking about
earlier when we were trying to classify conlangs.  This perhaps is
related with the idea that purple (red plus blue) is similar to violet
(a shorter wavelength than blue).

There's a technique - I forget the details, but you can create the
perception of any shade of colour by means of two monochromatic
sources.  Or something like that.  I think the inventor of the
Polaroid camera was involved.  I need to check my notes.

*  Which is the English name for the photosensitive cells in the retina
   used to perceive colour.  I know Nik just answered this, but I was
   already writing.  The name distinguishes them from rods, which are
   more photosynthesis but cannot perceive color.  The cones are
   concentrated in the fovea centralis, in the center of the visual
   field, which explains why a)  if you suppress the saccadic motion of
                                 your eyes, your peripheral vision will
                                 lose colour
                             b)  you can see dim objects (like stars)
                                 better with peripheral vision, which
                                 is mostly rods.

** I have notes from a neurophysics course around here somewhere,
   which will tell me exactly what the peaks are.  I'll post it
   tomorrow, unless somone else has found it in the meantime.  They
   aren't as evenly spaced as you might think.