(no subject)

Hello
 I am doing a project on ordered logit/probit models using Splus. I am 
using the polr function in library(MASS) which is exactly the same code as 
in R. Unfortunatelly I am not very succesfull with it. The code I used was:

polr(factor(resp)~.,data=data3) # data3 does not consist the resp

variable
 But I am getting the same warning message 21 times. The message is the
 following: 
1: singularity encountered in: nlminb.1(temp, p, liv, lv,
 objective, gradient, bounds, scale) 

  I guess the corresponding warning message in R is:
 Error in optim(start, fmin, gmin, method = "BFGS", hessian = Hess)

 I would be very greatfull if someone could suggest what it might be wrong 
with
 my model anh how I can overcome it. 
Thanking you in advance 
Charis Spanou