perception   of  water   has  been   a  valid   one.
　　   Admittedly, "water" is a mental construct  (vikalpa)
　　   and corresponds to the parikalpa mode of experience.
　　   What  is  actually   perceived,  the  thing   itself
　　   (vastumaatra).  is ultimately indeterminable  in the
　　   conceptual  sphere;  it is  only  determinable  in a
　　   specific  context.  The  Madhyaantavibhaaga  adduces
　　   that  where  a  human  being  sees  water,  a  preta
　　   (insatiable   ghost)  sees   a  river   of  pus  and
　　   excrement, while a yogin engaged in certain types of
　　   meditation  might see nothing at all or might see in
　　   its place skeletons or another object of meditation

　　　　　　　　　　　　  P.239

　　   (Y21).  Another  example that might be more directly
　　   accessible to us might be that of an apple. Where we
　　   see food, a physicist  might see a configuration  of
　　   atoms, a botanist might see a seed-bearing  vehicle,
　　   an artist  might see a red sphere to paint, a native
　　   of some tribe might see the manifestation of the god
　　   of that  tribe, and so forth.  The apple  exists, as
　　   paratantra, but due to its  emptiness, its  lack  of
　　   intrinsic   identity,  it  can   be  seen   as  many
　　   things--in　　 James'　　 terms,   taken　　 in　　 many
　　   relations--depending  upon  the  point  of view  and
　　   purposes  of the  perceiver.  Usefulness  has  to be
　　   usefulness  to someone;  it is not a function of the
　　   object  so  much  as  of the  subject, although  the
　　   capability of usefulness in various contexts--and of
　　   giving  rise  to a cognition, be it  a cognition  of
　　   water,  pus,  or  skeletons--resides   in  the  real
　　   object.  It is the unresolvability  of objects  into
　　   universally  valid  concepts  that  makes  a test of
　　   validity necessary. This ultimate unresolvability is
　　   also  what  limits  the  validity   established   by
　　   arthakriyaa  to a particular  context.  According to
　　   Dharmakiirti,  that   validity   pertains   to   the
　　   conventional　　 level　　 of   truth　　 and   reality
　　   (vyavahaara)(74) and  thus  cannot  claim  ultimacy.
　　   James   acknowledges　　 the   same   limitation   or
　　   context-dependence in his pragmatic test of truth:

　　   How is success to be absolutely  measured when there
　　   are so many environments and so many ways of looking