Intereting Posts

Yet another conjecture about primes
non-abelian groups of order $p^2q^2$.
Mental card game
Laplace transform of $\cos(at)$
Find maximum divisors of a number in range
Is simply connectedness preserved after deleting a high codimension set
Divergent products.
When simplifying $\sin(\arctan(x))$, why is negative $x$ not considered?
Number of ways of distributing n identical objects among r groups
Shortest Path with odd number of “Green” vertices
Center of Direct Product is the Direct Product of Centers
Does the splitting lemma hold without the axiom of choice?
Norm of the sum of projection operators
Why is $(2+\sqrt{3})^{50}$ so close to an integer?
distinguishing probability measure, function, distribution

I found this proposition and don’t see exactly as to why it is true and even more so, why the converse is false:

Proposition 1. The equivalence between the proposition $z \in D$ and the proposition $(\exists x \in D)x = z$ is provable from the definitory equations of the existential quantifier and of the equality relation. If $D = \{t_{1},t_{2},…t_{n}\}$, the sequent

$z = t_{1} \vee z = t_{2} \vee z \vee … z = t_{n} \vdash z \in D$ is provable from the definition of the additive disjunction $\vee$.

- Show that $A \setminus ( B \setminus C ) \equiv ( A \setminus B) \cup ( A \cap C )$
- Integer induction without infinity
- Formula that's only satisfiable in infinite structures
- Ultrafilter Lemma implies Compactness/Completeness of FOL
- The difference between Z and ZF
- Can anyone help me with a solution?

The converse sequent $z \in D \vdash z = t_{1} \vee z = t_{2} \vee … z = t_{n}$ is not provable.

The author goes on to say: “We adopt the intuitionistic interpretation of disjunction. With respect to it, one can characterize a particular class of finite sets”

On the first part of the proposition, well I am not sure what the point is as if we take an element z in D, then we could just call this element x and hence this x = z. Is there something more to this? On the second part of Proposition 1 since $z = \text{ some } t \in D$ since t is in the set of axioms and $t = z$, then $t \vdash z$ as z is derivable from t.

For the converse if $z \in D$ then why wouldn’t z = some $t \in D$? Is this because of the Incompleteness theorem? That perhaps there D as a set of axioms has some consequence which can not be proven by the set of axioms in D? Or perhaps I am way off here.

Any ideas?

Thanks,

Brian

- How is a set subset of its power set?
- Can I construct a complete (as a Boolean algebra) saturated elementary extension of a given Boolean algbera?
- what is the definition of $=$?
- Why is Peano arithmetic undecidable?
- Does evaluating hyperreal $f(H)$ boil down to $f(±∞)$ in the standard theory of limits?
- What's the difference between a negation and a contrapositive?
- 9 pirates have to divide 1000 coins…
- intersection of the empty set and vacuous truth

This is most of an answer, with an omission at the end:

Intuitionistically, to prove $s \vee t$ you must exhibit either a proof of $s$ or a proof of $t$. Hence, you can’t prove $z = t_1 \vee z = t_2 \vee \dots$ unless you have some method of discovering which $t_i$ was equal to $z$.

Now, if you had decidable equality, of the form $(\forall x)(\forall y)(x = y \vee x \not= y)$, then you could use that to “check” each $t_i$ for equality with $z$. I don’t know how you would use the fact that $D = \{t_1 \dots t_n\}$ to show that this process must eventually succeed, since I don’t know how that set equality would be interpreted as a logical statement.

The deeper question here is what the author means by

$$

D = \{ t_1, \ldots, t_n\}.

$$

The normal definition of this is:

$$

D = \{t_1,\ldots,t_n\} \qquad \text{means}\qquad (\forall z)[ z \in D \leftrightarrow z = t_1 \lor \cdots \lor z = t_n].

$$

But, according to that definition, one *does* have an intuitionistic proof that

$$

x \in D \leftrightarrow x = t_1 \lor \cdots \lor x = t_n

$$

because this follows immediately from the definition of $D$.

Therefore, until it is clarified what the author means by “$D = \{ t_1, \ldots, t_n\}$”, very little can be said.

- If $m=\operatorname{lcm}(a,b)$ then $\gcd(\frac{m}{a},\frac{m}{b})=1$
- How to prove the positive definite property of the semidefinite matrix with regularization?
- Proof That $\mathbb{R} \setminus \mathbb{Q}$ Is Not an $F_{\sigma}$ Set
- The sup norm on $C$ is not equivalent to another one, induced by some inner product
- Smallest number with specific number of divisors
- Conditions for matrix similarity
- Proving ${\sim p}\mid{\sim q}$ implies ${\sim}(p \mathbin\& q)$ using Fitch
- Proving that $\sin x \ge \frac{x}{x+1}$
- Fast way to get a position of combination (without repetitions)
- The path to understanding Frieze Groups
- Is a binomial a sum of Bernoulli random variables?
- Reducing an integral equation to a differential one
- How many geometrical interpretations does matrix multiplication have?
- Factoring a polynomial (multivariable)
- Congruence Modulo with large exponents