A COUPLE OF COMMENTS AFTER MY SPIN TALK - from E.Shuryak

Three points I would like to add to my transparencies, if possible.

1.The progress often starts when somebody takes a position, so that others
have something to shoot at. So this is what the main point of my talk was:
to indicate a model I think is reasonable, but maybe being disproved soon
by people more experienced in the field than I am (I do not do spin physics
as my regular occupation, to put it mildly...).

So there were two proposed simple models for POLARIZED part of the partons.

 I. There are 3 constituent quarks, UUD, well localized, each having e.g.
the same polarized strange and charm seas.

II. There is a constituent quarkes U as above,
with UD being in bound diquark state with zero spin, not participating in
the polarization at all.

I discussed a picture of a constituent quark generated by instantons: its
flavor and polarization aspects as proposed by Dorokhov-Kochelev paper I
mentioned.

2.A couple of simple consequences I forgot to mention are:

TYPICAL X OF THIS SEA is easy to estimate. If each const.q. has x about 1/3,
and half is glue, plus we discussed a fluctuation into 3 or 5 sea+valence
quarks, we get x = 1/18- 1/30.

STRONG PAULI PRINCIPLE inside the constituent quark was emphasized. Note
that
due to it, in model II there is no place say to the polarized $\bar u$ at
all!

3.Kharzeev did not say anything during talk/discussion about data on
Lambda polarization.  I met him and asked about it.
The reply is very relevant:
he said there are two groups measured polarization of Lambdas,
with something like -0.7 pm 0.1 as a result.

He added that feedback from Sigma etc is about
.2 so the true answer means Lambda polarization should then be close
to -1.

Now, the model #I has THE SAME STRANGE POLARIZED SEA in all 3 quarks, but
2 are along the nucleon and 1 is against. It means total polarization of
Lambda
should be about -1/3 in model #I, at most.

In model #II the answer -1/2, if the strange sea inside the diquark is the
same
as in constituent quark. However, since diquark is heavier its
x-distribution
has smaller width, and one can imagine that at large x (at which experiments
were
made) we only see the part from polarized U. This is the only way how -1
can be reached.

I also discussed the unpolarized strange sea, saying it is very large
(recall scalar <N|\bar s s|N>\sim 1.5). It must resign at smaller x,
than the polarized one, otherwize again one cannot get large polarization of
these \Lambdas!

Another issue Dima mentioned (but I do not know the details) is whether
the polarized s AND \bar s have THE SAME  or DIFFERENT x distribution:
both models mentioned above and DATA (according to Dima) show it to be
THE SAME.