Natural Deduction in the forall x: Pittsburgh systems

This document gives a short description of how Carnap presents the systems of natural deduction from forall x: Pittsburgh, the remix by Dimitri Gallow of Aaron Thomas-Bolduc and Richard Zach's Calgary version of P.D. Magnus's forall x.

The syntax of formulas accepted is described in the Systems Reference.

Truth-functional logic

Notation

The different admissible keyboard abbreviations for the different connectives are as follows:

Connective Keyboard
->, =>, >
/\, &, and
\/, |, or
↔︎ <->, <=>
¬ -, ~, not
!?, _|_

The available sentence letters are A through Z, together with the infinitely many subscripted letters P1, P2, … written P_1, P_2 and so on.

Proofs consist of a series of lines. A line is either an assertion line containing a formula followed by a : and then a justification for that formula, or a separator line containing two dashes, thus: --. A justification consists of a rule abbreviation followed by zero or more numbers (citations of particular lines) and pairs of numbers separated by a dash (citations of a subproof contained within the given line range).

A subproof is begun by increasing the indentation level. The first line of a subproof should be more indented than the containing proof, and the lines directly contained in this subproof should maintain this indentation level. (Lines indirectly contained, by being part of a sub-sub-proof, will need to be indented more deeply.) The subproof ends when the indentation level of the containing proof is resumed; hence, two contiguous sub-proofs of the same containing proof can be distinguished from one another by inserting a separator line between them at the same level of indentation as the containing proof. The final unindented line of a derivation will serve as the conclusion of the entire derivation.

Here's an example derivation, using the system .GallowSL:

Ex
A \/ B:AS A:AS B \/ A:\/I 2 -- B:AS B \/ A:\/I 5 B \/ A:\/E 1,2-3,5-6 (A \/ B) -> (B \/ A):->I 1-7

Or, .GallowSLPlus with a Fitch-style guides overlay (activated with guides="fitch"):

Playground
A \/ B:AS A:AS B \/ A:\/I 2 -- B:AS B \/ A:\/I 5 B \/ A:\/E 1,2-3,5-6 (A \/ B) -> (B \/ A):->I 1-7

Simple indent guides overlay (activated with guides="indent") with system .GallowPL:

Playground
A \/ B:AS A:AS B \/ A:\/I 2 -- B:AS B \/ A:\/I 5 B \/ A:\/E 1,2-3,5-6 (A \/ B) -> (B \/ A):->I 1-7

The SL systems

The system .GallowSLPlus allows all rules below. .GallowSL is like .GallowSLPlus except it disallows all derived rules, i.e., the only allowed rules are R, and the I and E rules for the connectives and for ⊥.

It has the following set of rules for direct inferences:

Basic rules:

Rule Abbreviation Premises Conclusion
And-Elim. ∧E φ ∧ ψ φ/ψ
And-Intro. ∧I φ, ψ φ ∧ ψ
Or-Intro ∨I φ/ψ φ ∨ ψ
Contradiction-Intro ⊥I φ, ¬φ
Contradiction-Elim ⊥E ψ
Biconditional-Elim ↔︎E φ/ψ, φ ↔︎ ψ ψ/φ
Reiteration R φ φ

Derived rules:

Rule Abbreviation Premises Conclusion
Disjunctive Syllogism DS ¬ψφ, φ ∨ ψ φ/ψ
Modus Tollens MT φ → ψ, ¬ψ ¬φ
Double Negation Elim. DNE ¬¬φ φ
DeMorgan's Laws DeM ¬(φψ) ¬φ ∨ ¬ψ
¬(φψ) ¬φ ∧ ¬ψ
¬φ ∨ ¬ψ ¬(φψ)
¬φ ∧ ¬ψ ¬(φψ)

We also have five rules for indirect inferences:

  1. →I, which justifies an assertion of the form φ → ψ by citing a subproof beginning with the assumption φ and ending with the conclusion ψ;
  2. ↔︎I, which justifies an assertion of the form φ ↔︎ ψ by citing two subproofs, beginning with assumptions φ, ψ, respectively, and ending with conclusions ψ, φ, respectively;
  3. ¬I, which justifies an assertion of the form ¬φ by citing a subproof beginning with the assumption φ and ending with a conclusion .
  4. ∨E, which justifies an assertion of the form φ by citing a disjunction ψ ∨ χ and two subproofs beginning with assumptions ψ, χ respectively and each ending with the conclusion φ.
  5. ¬E (indirect proof), which justifies an assertion of the form φ by citing a subproof beginning with the assumption ¬φ and ending with a conclusion .
  6. LEM (Law of the Excluded Middle), which justifies an assertion of the form ψ by citing two subproofs beginning with assumptions φ, ¬φ respectively and each ending with the conclusion ψ. LEM is a derived rule.

Finally, PR can be used to justify a line asserting a premise, and AS can be used to justify a line making an assumption. A note about the reason for an assumption can be included in the rendered proof by writing A/NOTETEXTHERE rather than AS for an assumption. Assumptions are only allowed on the first line of a subproof.

Predicate logic

There are two proof systems corresponding to the Pittsburgh remix of forall x. All of them allow sentence letters in first-order formulas. The available relation symbols are the same as for SL: A through Z, together with the infinitely many subscripted letters F1, F2, … written F_1, F_2 etc. However, the available constants and function symbols are only a through v, together with the infinitely many subscripted letters c1, c2, … written c_1, c_2,…. The available variables are w through z, with the infinitely many subscripted letters x1, x2, … written x_1, x_2,….

Connective Keyboard
A, @
E, 3

The predicate logic system .GallowPL of forall x: Pittsburgh extend the rules of the system .GallowSL with the following set of new basic rules:

Rule Abbreviation Premises Conclusion
Existential Introduction ∃I φ(σ) xφ(x)
Universal Elimination ∀E xφ(x) φ(σ)
Universal Introduction ∀E φ(σ) xφ(x)

Where Universal Introduction is subject to the restriction that σ must not appear in φ(x), or any undischarged assumption or in any premise of the proof.1

There is one new rule for indirect derivations: ∃E, which justifies an assertion ψ by citing an assertion of the form xφ(x) and a subproof beginning with the assumption φ(σ) and ending with the conclusion ψ, where σ does not appear in ψ, ∃xφ(x), or in any of the undischarged assumptions or premises of the proof.

The proof system .GallowPLPlus for the Pittsburgh version of forall x adds, in addition to the (basic and derived) rules of .GallowPL, the rules

Rule Abbreviation Premises Conclusion
Conversion of Quantifiers CQ ¬∀xφ(x) x¬φ(x)
x¬φ(x) ¬∀xφ(x)
¬∃xφ(x) x¬φ(x)
x¬φ(x) ¬∃xφ(x)

  1. Technically, Carnap checks only the assumptions and premises that are used in the derivation of φ(σ). This has the same effect in terms of what's derivable, but is a little more lenient.↩︎