So that the value of this term is the intersection of all S such that S =
V(B, a[S1 /X1 , . . . , Sn/Xn]) for some S1 , . . . , Sn which are simultaneous fix-
points in the sense that Sk = V(Ak, a[S1/X1, . . . , Sn/Xn]) for each k. It
follows from the continuity of the terms Ak that V (B, a [Rn/X1,..., Rn /Xn])
is among those S , so that we have
n
V (λr. ∀Xι... Xn( ʌ Xi = Ai → B(r)), a) ⊆ V (B, a[Rn/Xι,..., R. /Xn])
i=1
To show the converse, suppose that Sk = V(Ak, a[S1/X1 , . . . , Sn/Xn]) for
1 ≤ k ≤ n. The monotonicity of the Ak readily gives us that Rkm ⊆ Sk for all
k and m. It follows that Rkω ⊆ Sk for all k and hence, by B’s monotonicity,
that
V (B, a[Rω/Xι,..., Rn /Xn]) ⊆ V (B, a[Si/Xi,..., Sn/Xn]).
Since S1 , . . . , Sn were arbitrary simultaneous fixpoints, we conclude that
n
V(B, a[R1n/X1,..., Rnn/Xn]) ⊆V(λr.∀X1...Xn(i=1Xi =Ai → B(r)), a)
□
Having shown some general properties of continuity and weak anticontinu-
ity to hold, we now turn to proving that the translations we have given in
section 6 possess the desired properties. A reader who wants to prove that
translations of programs are continuous in the variables X occurring in them
will notice that one clause of the required induction cannot be proved us-
ing simple continuity properties of union and relational composition. The
required extra proof uses lemma 8 and is given below.
Lemma 9 If B and A1 , . . . , An are terms of type s × s → t which are con-
tinuous in the type s × s → t variables X1 , . . . , Xn and in ξα, then
λr.∀X1...Xn(^n Xi =Ai → B(r))
i=1
is also continuous in ξ.
Proof. If ξ is among {X1 , . . . , Xn} we are done, so assume Xk 6= ξ. Let
R0 ⊆ R1 ⊆ . . . ⊆ Rn ⊆ . . . be an arbitrary chain of type α relations. Define,
20
More intriguing information
1. On the Real Exchange Rate Effects of Higher Electricity Prices in South Africa2. Psychological Aspects of Market Crashes
3. The Challenge of Urban Regeneration in Deprived European Neighbourhoods - a Partnership Approach
4. Valuing Farm Financial Information
5. The name is absent
6. The Interest Rate-Exchange Rate Link in the Mexican Float
7. The Veblen-Gerschenkron Effect of FDI in Mezzogiorno and East Germany
8. Draft of paper published in:
9. Technological progress, organizational change and the size of the Human Resources Department
10. Integration, Regional Specialization and Growth Differentials in EU Acceding Countries: Evidence from Hungary