26 KiB
title | TARGET DECK | FILE TAGS | tags | ||
---|---|---|---|---|---|
Equivalence Transformation | Obsidian::STEM | logic::equiv-trans |
|
Overview
Equivalence-transformation refers to a class of calculi for manipulating propositions derived from negation (\neg
), conjunction (\land
), disjunction (\lor
), implication (\Rightarrow
), and equality (=
). Gries covers two in "The Science of Programming": a system of evaluation and a formal system. The system of evaluation mirrors how a computer processes instructions, at least in an abstract sense. The formal system serves as a theoretical framework for reasoning about propositions and their transformations without resorting to "lower-level" operations like substitution.
%%ANKI Basic Who is the author of "The Science of Programming"? Back: David Gries Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What are constant propositions? Back: Propositions that contain only constants as operands. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What are the basic propositional logical operators?
Back: \neg
, \land
, \lor
, \Rightarrow
, and \Leftrightarrow
/=
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Cloze
Gries replaces logical operator {\Leftrightarrow
} in favor of {=
}.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
How does Lean define propositional equality?
Back: Expressions a
and b
are propositionally equal iff a = b
is true.
Reference: Avigad, Jeremy. ‘Theorem Proving in Lean’, n.d.
Tags: lean
END%%
%%ANKI
Basic
How does Lean define propext
?
Back:
axiom propext {a b : Prop} : (a ↔ b) → (a = b)
Reference: Avigad, Jeremy. ‘Theorem Proving in Lean’, n.d. Tags: lean
END%%
%%ANKI
Basic
What Lean theorem justifies Gries' choice of =
over \Leftrightarrow
?
Back: propext
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: lean
END%%
%%ANKI
Basic
What name is given to \land
operands?
Back: Conjuncts
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What name is given to \lor
operands?
Back: Disjuncts
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What name is given to operand a
in a \Rightarrow b
?
Back: The antecedent
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What name is given to operand b
in a \Rightarrow b
?
Back: The consequent
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Is (b \land c)
well-defined in \{(b, T), (c, F)\}
?
Back: Yes
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Is (b \lor d)
well-defined in \{(b, T), (c, F)\}
?
Back: No
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What C logical operator corresponds to \neg
?
Back: !
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: c
END%%
%%ANKI
Basic
What C logical operator corresponds to \land
?
Back: There isn't one.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: c
END%%
%%ANKI
Basic
What C logical operator corresponds to \lor
?
Back: There isn't one.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: c
END%%
%%ANKI
Basic
What C logical operator corresponds to \Rightarrow
?
Back: There isn't one.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: c
END%%
%%ANKI
Basic
What C logical operator corresponds to \Leftrightarrow
?
Back: ==
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
Tags: c
END%%
%%ANKI
Basic
What proposition represents states \{(b, T)\}
and \{(c, F)\}
?
Back: b \lor \neg c
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What set of states does a \land b
represent?
Back: The set containing just state \{(a, T), (b, T)\}
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is sloppy about phrase "the states in b \lor \neg c
"?
Back: b \lor \neg c
is not a set but a representation of a set (of states).
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the weakest proposition?
Back: T
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What set of states does T
represent?
Back: The set of all states.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the strongest proposition?
Back: F
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What set of states does F
represent?
Back: The set of no states.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What does a proposition represent? Back: The set of states in which it is true. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
When is p
stronger than q
?
Back: When p \Rightarrow q
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
When is p
weaker than q
?
Back: When q \Rightarrow p
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic A proposition is well-defined with respect to what? Back: A state to evaluate against. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Why is b \land c
stronger than b \lor c
?
Back: The former represents a subset of the states the latter represents.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What is a state? Back: A function mapping identifiers to values. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What are the two calculi Gries describes equivalence-transformation with? Back: A formal system and a system of evaluation. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
Equivalence Schemas
A proposition is said to be a tautology if it evaluates to T
in every state it is well-defined in. We say propositions E1
and E2
are equivalent if E1 = E2
is a tautology. In this case, we say E1 = E2
is an equivalence.
%%ANKI Basic What does it mean for a proposition to be a tautology? Back: That the proposition is true in every state it is well-defined in. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic The term "equivalent" refers to a comparison between what two objects? Back: Propositions. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What does it mean for two propositions to be equivalent?
Back: Given propositions E1
and E2
, it means E1 = E2
is a tautology.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is an equivalence?
Back: Given propositions E1
and E2
, tautology E1 = E2
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- Commutative Laws
(E1 \land E2) = (E2 \land E1)
(E1 \lor E2) = (E2 \lor E1)
(E1 = E2) = (E2 = E1)
%%ANKI
Basic
Which of the basic logical operators do the commutative laws apply to?
Back: \land
, \lor
, and =
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What do the commutative laws allow us to do? Back: Reorder operands. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the commutative law of e.g. \land
?
Back: E1 \land E2 = E2 \land E1
END%%
- Associative Laws
E1 \land (E2 \land E3) = (E1 \land E2) \land E3
E1 \lor (E2 \lor E3) = (E1 \lor E2) \lor E3
%%ANKI
Basic
Which of the basic logical operators do the associative laws apply to?
Back: \land
and \lor
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What do the associative laws allow us to do? Back: Remove parentheses. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the associative law of e.g. \land
?
Back: E1 \land (E2 \land E3) = (E1 \land E2) \land E3
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- Distributive Laws
E1 \lor (E2 \land E3) = (E1 \lor E2) \land (E1 \lor E3)
E1 \land (E2 \lor E3) = (E1 \land E2) \lor (E1 \land E3)
%%ANKI
Basic
Which of the basic logical operators do the distributive laws apply to?
Back: \land
and \lor
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What do the distributive laws allow us to do? Back: "Factor" propositions. Reference: Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the distributive law of e.g. \land
over \lor
?
Back: E1 \land (E2 \lor E3) = (E1 \land E2) \lor (E1 \land E3)
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- De Morgan's Laws
\neg (E1 \land E2) = \neg E1 \lor \neg E2
\neg (E1 \lor E2) = \neg E1 \land \neg E2
%%ANKI
Basic
Which of the basic logical operators do De Morgan's Laws apply to?
Back: \neg
, \land
, and \lor
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is De Morgan's Law of e.g. \land
?
Back: \neg (E1 \land E2) = \neg E1 \lor \neg E2
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- Law of Negation
\neg (\neg E1) = E1
%%ANKI
Basic
What does the Law of Negation say?
Back: \neg (\neg E1) = E1
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- Law of the Excluded Middle
E1 \lor \neg E1 = T
%%ANKI
Basic
Which of the basic logical operators does the Law of the Excluded Middle apply to?
Back: \lor
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What does the Law of the Excluded Middle say?
Back: E1 \lor \neg E1 = T
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic Which equivalence schema is "refuted" by sentence, "This sentence is false." Back: Law of the Excluded Middle Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
- Law of Contradiction
E1 \land \neg E1 = F
%%ANKI
Basic
Which of the basic logical operators does the Law of Contradiction apply to?
Back: \land
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What does the Law of Contradiction say?
Back: E1 \land \neg E1 = F
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Cloze
The Law of {1:the Excluded Middle} is to {2:\lor
} whereas the Law of {2:Contradiction} is to {1:\land
}.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
Gries lists other "Laws" but they don't seem as important to note here.
%%ANKI
Basic
How is \Rightarrow
written in terms of other logical operators?
Back: p \Rightarrow q
is equivalent to \neg p \lor q
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
How is \Leftrightarrow
/=
written in terms of other logical operators?
Back: p \Leftrightarrow q
is equivalent to (p \Rightarrow q) \land (q \Rightarrow p)
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What distinguishes an equality from an equivalence? Back: An equivalence is an equality that is also a tautology. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
Equivalence Rules
- Rule of Substitution
- Let
P(r)
be a predicate andE1 = E2
be an equivalence. ThenP(E1) = P(E2)
is an equivalence.
- Let
- Rule of Transitivity
- Let
E1 = E2
andE2 = E3
be equivalences. ThenE1 = E3
is an equivalence.
- Let
%%ANKI Basic What two inference rules make up the equivalence-transformation formal system? Back: Substitution and transitivity. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic Which of the two inference rules that make up the equivalence-transformation formal system is redundant? Back: Transitivity. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What does the rule of substitution say in the system of evaluation?
Back: Let P(r)
be a predicate and E1 = E2
be an equivalence. Then P(E1) = P(E2)
is an equivalence.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic How is the rule of substitution written as an inference rule (in standard form)? Back:
\begin{matrix}
E1 = E2 \\
\hline P(E1) = P(E2)
\end{matrix}
END%%
%%ANKI
Basic
What does the rule of transitivity state in the system of evaluation?
Back: Let E1 = E2
and E2 = E3
. Then E1 = E3
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic How is the rule of transitivity written as an inference rule (in standard form)? Back:
\begin{matrix}
E1 = E2, E2 = E3 \\
\hline E1 = E3
\end{matrix}
END%%
%%ANKI Cloze The system of evaluation has {equivalences} whereas the formal system has {theorems}. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic What is a "theorem" in the equivalence-transformation formal system? Back: An equivalence derived from the axioms and inference rules. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic How is e.g. the Law of Implication proven in the system of evaluation? Back: With truth tables Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI Basic How is e.g. the Law of Implication proven in the formal system? Back: It isn't. It is an axiom. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Cloze
The system of evaluation and formal system are connected by the following biconditional: {e
is a tautology} iff {e = T
is a theorem}.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Cloze
The {1:system of evaluation} is to {2:"e
is a tautology"} whereas the {2:formal system} is to {1:"e = T
is a theorem"}.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
Textual Substitution
Textual substitution refers to the simultaneous replacement of a free identifier with an expression, introducing parentheses as necessary. This concept is just the #Equivalence Rules with different notation. For example, let E
and e
be expressions and x
an identifer. Then $E_e^x
denotes the simultaneous replacement of all free occurrences of $x
with e
.
%%ANKI Basic Textual substitution is derived from what equivalence rule? Back: The substitution rule. Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is E
's role in textual substitution E_e^x
?
Back: It is the expression that free occurrences of x
are replaced with e
in.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is e
's role in textual substitution E_e^x
?
Back: It is the expression that free occurrences of x
in E
are substituted with.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is x
's role in textual substitution E_e^x
?
Back: It is the identifier matching free occurrences in E
that are replaced with e
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
How is textual substitution E_e^x
interpreted as a function?
Back: As E(e)
, where E
is a function of x
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Why does Gries prefer notation E_e^x
over e.g. E(e)
?
Back: The former indicates the identifier to replace.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What two scenarios ensure E_e^x = E
is an equivalence?
Back: x = e
or no free occurrences of x
exist in E
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Why might E_e^x = E
be an equivalence despite identifier x
existing in E
?
Back: If the only occurrences of x
in E
are bound.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is required for E_e^x
to be valid?
Back: Substitution must result in a syntactically valid expression.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the result of the following? (x < y \land (\forall i : 0 \leq i < n : b[i] < y))_z^x$$
Back:
(z < y \land (\forall i : 0 \leq i < n : b[i] < y))
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the result of the following? (x < y \land (\forall i : 0 \leq i < n : b[i] < y))_z^y$$
Back:
(x < z \land (\forall i : 0 \leq i < n : b[i] < z))
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
What is the result of the following? (x < y \land (\forall i : 0 \leq i < n : b[i] < y))_z^i$$
Back:
(x < y \land (\forall i : 0 \leq i < n : b[i] < y))
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
(E_u^x)_v^x = E_{u_v^x}^x
- The only possible free occurrences of
x
that may appear after the first of the sequential substitutions occur inu
.
- The only possible free occurrences of
- If
y
is not free inE
, then(E_u^x)_v^y = E_{u_v^y}^x
.y
may not be free inE
but substitutingx
withu
can introduce a free occurrence. It doesn't matter if we perform the substitution first or second though.
%%ANKI
Basic
How do we simplify (E_u^x)_v^x
?
Back: As E_{u_v^x}^x
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
How is E_{u_v^x}^x
rewritten as sequential substitution?
Back: As (E_u^x)_v^x
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Why is (E_u^x)_v^x = E_{u_v^x}^x
an equivalence?
Back: After the first substitution, the only possible free occurrences of x
are in u
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
In what two scenarios is (E_u^x)_v^y = E_{u_v^y}^x
always an equivalence?
Back: x = y
or y
is not free in E
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
If x \neq y
, when is (E_u^x)_v^y = E_{u_v^y}^x
?
Back: When y
is not free in E
.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
%%ANKI
Basic
Why should y
not be free in E
for (E_u^x)_v^y = E_{u_v^y}^x
to be an equivalence?
Back: If it were, a v
would exist in the LHS that doesn't in the RHS.
Reference: Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.
END%%
References
- Avigad, Jeremy. ‘Theorem Proving in Lean’, n.d.
- Gries, David. The Science of Programming. Texts and Monographs in Computer Science. New York: Springer-Verlag, 1981.