# User:Lord Farin/Archive/Natural Deduction Axioms

# General note

The below page is a copy of the revision 96983 of Definition:Natural Deduction, accessible here.

## Definition

**Natural deduction** is a proof system for propositional logic.

As such, it consists of certain axioms, which together constitute a collection of theorems.

This can be specified as follows.

### Notation

To indicate that a collection of WFFs $P$, the pool of assumptions, entails a particular WFF $p$, the conclusion, we agree to write a sequent:

- $P \vdash p$

To make exposition more natural, we also agree to omit the brackets of explicit set definition and use commas in a suggestive way, so that:

- $p, q \vdash p \land q$
- $P, Q \vdash p \land q$
- $P, p \vdash q$

are used instead of the formally more correct:

- $\left\{{p, q}\right\} \vdash p \land q$
- $P \cup Q \vdash p \land q$
- $P \cup \left\{{p}\right\} \vdash q$

Furthermore, in the last expression it is implicitly understood that $p \notin P$.

The $\Large\leadsto$ symbol indicates that given the sequent(s) on the left-hand side, the sequent on the right-hand side may be inferred.

### Axioms

**Natural deduction** has the following twelve axioms, explained in more detail on their own pages:

\(\text {(\mathrm{ND}:\mathrm A)}: \quad\) | \(\ds \) | \(\) | \(\ds q \vdash q\) | Rule of Assumption | ||||||||||

\(\text {(\mathrm{ND}:\lor \mathcal I_1)}: \quad\) | \(\ds P \vdash p\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash p \lor q\) | Rule of Addition | ||||||||||

\(\text {(\mathrm{ND}:\lor \mathcal I_2)}: \quad\) | \(\ds P \vdash q\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash p \lor q\) | Rule of Addition | ||||||||||

\(\text {(\mathrm{ND}:\lor \mathcal E)}: \quad\) | \(\ds \left.{ \begin{align} P &\vdash p \lor q \\ Q, p &\vdash r \\ R, q &\vdash r \end{align} }\right\}\) | \(\, \Large\leadsto \,\) | \(\ds P, Q, R \vdash r\) | Proof by Cases | ||||||||||

\(\text {(\mathrm{ND}:\land \mathcal I)}: \quad\) | \(\ds \left.{ \begin{align} P &\vdash p \\ Q &\vdash q \end{align} }\right\}\) | \(\, \Large\leadsto \,\) | \(\ds P, Q \vdash p \land q\) | Rule of Conjunction | ||||||||||

\(\text {(\mathrm{ND}:\land \mathcal E_1)}: \quad\) | \(\ds P \vdash p \land q\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash p\) | Rule of Simplification | ||||||||||

\(\text {(\mathrm{ND}:\land \mathcal E_2)}: \quad\) | \(\ds P \vdash p \land q\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash q\) | Rule of Simplification | ||||||||||

\(\text {(\mathrm{ND}:\implies \mathcal I)}: \quad\) | \(\ds P, p \vdash q\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash p \implies q\) | Rule of Implication | ||||||||||

\(\text {(\mathrm{ND}:\implies \mathcal E)}: \quad\) | \(\ds \left.{ \begin{align} P &\vdash p \implies q \\ Q &\vdash p \end{align} }\right\}\) | \(\, \Large\leadsto \,\) | \(\ds P, Q \vdash q\) | Modus Ponendo Ponens | ||||||||||

\(\text {(\mathrm{ND}:\neg \mathcal I)}: \quad\) | \(\ds P, p \vdash \bot\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash \neg p\) | Proof by Contradiction | ||||||||||

\(\text {(\mathrm{ND}:\neg \mathcal E)}: \quad\) | \(\ds \left.{ \begin{align} P &\vdash p \\ Q &\vdash \neg p \end{align} }\right\}\) | \(\, \Large\leadsto \,\) | \(\ds P, Q \vdash \bot\) | Principle of Non-Contradiction | ||||||||||

\(\text {(\mathrm{ND}:\bot \mathcal E)}: \quad\) | \(\ds P \vdash \bot\) | \(\, \Large\leadsto \,\) | \(\ds P \vdash p\) | Rule of Explosion |

### Theorems

The theorems of **natural deduction** are those WFFs $p$ allowing a sequent $\vdash p$, i.e., that may be derived with an empty pool of assumptions.

### Proofs

Although it is satisfying to find a (formal) proof of a theorem using the above rules, it is advisable to cast such a proof in a standard framework.

On ProofWiki, the framework chosen is that of a tableau proof, which most easily lends itself for a MediaWiki architecture.

However, many other notations exist and are used by various authors.

## Also see

There are many other proof systems for propositional logic, see here.

## Sources

- 2000: Michael R.A. Huth and Mark D. Ryan:
*Logic in Computer Science: Modelling and reasoning about systems*: $\S 1.2, \ \S 1.2.3$