Errata to the submitted article 

Ralph Matthes
Non-Strictly Positive Fixed-Points for Classical Natural Deduction
(31 pages, 2 July 2003)


Insignificant typos and omissions:

throughout: apparently -> evidently  (added on September 2)

p.6  proof, l.6: rules -> rule
     l.-9: F by can -> F can
p.8  def mult. elim.: e[*:=E] will be written as e[E] in the sequel
p.9  remark 7: F^+ should not be confounded with system F^+ (p.21)
p.10 l.4 before remark 9: the side hypotheses ->
                 the side hypotheses in the last three rules
     remark 9: the subscript to SN should be an asterisk *
p.14 constructions l.3: can be more elegantly by ->
                        can    more elegantly be
p.16 (2) replace "=>" by a dot
p.22 proof lemma 29 l.4: Then U -> Then "intersection over U"
p.27 l.3 v_0 -> v_1, same in third line from below in that proof


Mathematical mistake (with correction):

p.27 l.1 E[t] in SN implies E_n[t] in SN:
         This requires in general e[t] in SN => t in SN.
         Its truth is not an immediate consequence of the derived rule
         for SN in the center of p.25, but can be proven from it by
         course-of-value induction on SN. (Course of value is needed
         in case E=* in the fourth rule of SN.)

FINIS