Identity needs its own rule in proofs. You already use identity in translations (At Least, At Most and Exactly); here it joins the quantifier rules (Proofs with Quantifiers) and quantifier order (Quantifier Order) in the proof system.
| Form of Id | From | Infer |
|---|---|---|
| Reflexivity | (nothing) | $a = a$ |
| Symmetry | $a = b$ | $b = a$ |
| Substitution | $Fa$ and $a = b$ | $Fb$ (replacing some or all occurrences) |
Key idea
Why it works
Worked examples
Worked example: Substitution
- Replace $a$ by $b$ in $Fa$.
Answer: $Fb$
Worked example: Either direction
- The identity may be used either way round: $b = a$ also tells you $a$ and $b$ are one object.
Answer: Yes.
Worked example: Proving a non-identity
- Line 3: $a = b$ (AIP).
- Line 4: $Fb$ by Id 1, 3.
- Line 5: $Fb \land \neg Fb$ by Conj 4, 2.
- Line 6: $a \ne b$ by IP 3–5.
Answer: If they were one object, it would both be and not be $F$.
Common mistake
Identity proofs often combine Id with the other rules: substitution to make two lines match, then MP, MT or DS to use them. When two names appear and an identity links them, ask which substitution would make a premise's antecedent (or a disjunct) look exactly like a line you already have.
With identity, the proof system is complete for the whole language of this course: every valid argument with names, predicates, quantifiers and $=$ has a proof. The capstone lesson explains what that claim means.