Dear statlists,

I'm trying to do an LCA for 4 different variables, they are all continues. Using Stata 16 at a mac.
I keep getting this error message "no paths from latent variable to observed variables"
Code:
  gsem (ethnic, lang, pride, civic_law <-), logit lclass(C 3)
invalid model specification;
no paths specified
What is the problem here?
Very thankful for all help!

Thanks!