3

I've already read this, this, and this, and I feel like the answers are only more confusing me more.


Definitions and Properties Available:

Axioms related to natural numbers:

  • $0 \in \mathbb{N}$
  • If $n \in \mathbb{N}$, then $n\mathrm{++} \in \mathbb{N}$, where $n\mathrm{++}$ denotes the successor of $n$.
  • We define $1:= 0\mathrm{++}$, $2:= (0\mathrm{++})\mathrm{++})$, etc.
  • For each $n \in \mathbb{N}$, $n\mathrm{++} \neq 0$.
  • If $n\mathrm{++} = m\mathrm{++}$, then $n = m$.
  • Let $P(n)$ be a property regarding $n \in \mathbb{N}$. Then suppose both (a) $P(0)$ is true and (b) $P(n)$ is true implies that $P(n\mathrm{++})$ is true for each $n\in \mathbb{N}$. Then $P(n)$ is true for each $n \in \mathbb{N}$.

Addition over natural numbers:

  • Definition: $0 + m := m$, $(n\mathrm{++}) + m := (n+m)\mathrm{++}$.
  • $n + 0 = n$
  • Commutativity, associativity
  • $a+b = a+c \implies b = c$

A positive number is a natural number not equal to $0$.

  • If $a$ is positive and $b \in \mathbb{N}$, then $a+b$ is positive.
  • If $a, b \in \mathbb{N}$ with $a + b = 0$, then $a = b = 0$.
  • Let $a$ be positive. Then there is a unique $b \in \mathbb{N}$ with $b\mathrm{++} = a$.

Ordering of natural numbers:

  • $n \geq m$ if $n = m + a$ for some $a \in \mathbb{N}$.
  • $n > m$ if $n \geq m$ and $n \neq m$
  • Reflexivity, transitivity, anti-symmetric, addition preserves order
  • $a < b$ iff $a\mathrm{++} \leq b$
  • $a < b$ iff $b = a + d$ for some $d$ positive.
  • If $x \geq y$, then one of $x > y$ or $x = y$ is true.
  • Trichotomy of order in $\mathbb{N}$

Let $m_0 \in \mathbb{N}$, and let $P(m)$ be a property pertaining to $m \in \mathbb{N}$ arbitrary. Suppose for each $m \geq m_0$ ($m \in \mathbb{N}$) that $$P(m^{\prime}) \text{ true }\forall m_0 \leq m^{\prime} < m \text{ natural numbers } \implies P(m)\text{ true.}$$ Then $P(m)$ is true for all natural numbers $m \geq m_0$.

Attempt

I use induction as provided in the axioms above.

Define the property $$Q(n): P(m)\text{ true } \forall m_0 \leq m < n\text{.}$$

Consider $Q(0)$. Because $m_0 \in \mathbb{N}$, we have $m_0 = m_0 + 0$, hence $m_0 \geq 0 = n$. Thus, $Q(0)$ is vacuously true.

Now let $k \in \mathbb{N}$ and assume $Q(k)$ is true. This implies that $P(m)$ is true for all $m_0 \leq m < k$.

Consider $Q(k\mathrm{++})$. I need to show somehow that $P(m)$ is true for all $m_0 \leq m < k\mathrm{++}$, or simply that $P(k)$ is true, but I don't understand the discussions in the links above and how they fit into the results I have available.

Clarinetist
  • 19,519
  • I would define a slightly different property. Note that if the strong induction condition holds, then so does the "last" proposition P(n). What does (weak) induction tell you about that? – nomen Jan 06 '20 at 22:33
  • @nomen I'm not following. Could you please elaborate a bit? – Clarinetist Jan 09 '20 at 22:05
  • 1
    @Clarinetist, I find this post quite interesting because of the different notation and certain postulates, too. We denoted the successor function by $\operatorname s(n)$ and used the original axiomatization that states $1=\min \mathbb N;&;0\in\mathbb N_0$. However, I'm familiar with the arithmetic reasons for $0$ being included in the modern axiomatization. In our script, it says: $4^{\text{th}}$ axiom-axiom of induction: If $S\subseteq \mathbb N$ satisfies the predicate, we identify it with $\mathbb N$,i.e. $S=\mathbb N$ – PinkyWay Jan 09 '20 at 23:50
  • 1
    *Standard vs. absolute(strong) induction*

    The assumption in standard induction: $\operatorname \tau(P(n))=\top$ for (some) $n\in\mathbb N$

    The assumption in the strong induction: $\operatorname \tau(P(0))=\top,\operatorname \tau(P(1))=\top,\ldots,\operatorname \tau(P(n-1))=\top,\operatorname \tau(P(n))=\top,\underline{\forall k\in{0,1,\ldots,n-1,n}}$

    This is important for recursions, recursive sequences, sth like, descending induction, please correct me. $$$$The point of the $\text{successor function}$ is avoiding numbers at the start.

    – PinkyWay Jan 10 '20 at 00:17
  • 1
    @Clarinetist, consider $n-1;\text{successive compositions}\leftarrow\text{successors}$: $$\operatorname s(\ldots(s(s(n)))$$ It is of paramount importance to understand this because later on you're going to 'define' $\mathbb Z$ by equivalence clases, just like $\mathbb Q$. The tricky part is the $\mathbb R$ axiomatization because even $\mathbb Q$ satisfies the first $14$ axioms. Then you'll maybe hear $\text{the axiom of completeness}$ can, alternatively, be substituted by *Cantor's & Archimedean* axiom together. – PinkyWay Jan 10 '20 at 00:33
  • Is it not enough to say "When the required conditions for strong induction hold, namely that P(m′) true for all $0\le m' < m$ for some $m$ then $P(0)$ is true and $P(n)$ is true for some $n=m-1$ for that $m$ and as $P(m)= P(n+1)$ is implied, then the conditions of the axiom of the weak induction are satisfied so $P(n)$ is true for all $n$"? What more needs to be done? – fleablood Jan 12 '20 at 01:58

2 Answers2

3

I think the most difficult thing about this problem is sorting through the notation, and trying to understand how the assumptions all come together.

Here's my proof of this statement, after asking the question and getting an excellent answer above.

Assumptions: Suppose $m_0 \in \mathbb{N}$, and for each $m \geq m_0$, $P(m^{\prime})$ true for all $m_0 \leq m^{\prime} < m$ implies $P(m)$ true.

Claim. $P(m)$ true for all $m \geq m_0$.

Proof. As in the hint in the text, define the property $$Q(n): P(m)\text{ true }\forall m_0 \leq m < n\text{.}\tag{1}$$ Suppose $n = 0$. Since $m_0 \in \mathbb{N}$, it follows that $m_0 = m_0 + 0$, or $m_0 \geq 0$, but since the claim $Q(n)$ has $m_0 < n$, we have that $Q(0)$ is vacuously true.

Suppose $Q(k)$ is true for some $k \in \mathbb{N}$. By definition of $Q(n)$ in $(1)$, we have that $P(m)$ is true $\forall m_0 \leq m < k$.

Since $k \in \mathbb{N}$ is arbitrary, consider two situations.

Suppose $k < m_0$. Then this implies that $P(k)$ is vacuously true.

Suppose $k \geq m_0$. Since $k \geq m_0$ and $P(m)$ is true for all $m_0 \leq m < k$, it follows by assumption that $P(k)$ is true.

Thus we have shown that $Q(k)$ is true implies that $P(k)$ is true.

We need to show that $Q(k\mathrm{++})$ is true. That is, $$Q(k\mathrm{++}): P(m)\text{ true }\forall m_0 \leq m < k\mathrm{++}\text{.}$$ By the induction hypothesis, $Q(k)$ is true. Thus $P(m)$ is true $\forall m_0 \leq m < k$. But as shown above, $Q(k)$ being true implies that $P(k)$ is true. Thus $P(m)$ is true $\forall m_0 \leq m < k\mathrm{++}$, and hence $Q(k\mathrm{++})$ holds.

Hence by induction, $Q(n)$ holds for all $n \in \mathbb{N}$. Thus $P(m)$ is true for all $m \geq m_0$.

Clarinetist
  • 19,519
2

Assume that

$\tag 1 (\forall m \in \Bbb N) \; \text{IF } \big [ \text{ if } m_0 \le m^{\prime} \lt m \text{ then } P(m^{\prime}) \big] \text{ THEN } P(m)$

Define

$\tag 2 Q(n) : (\forall m \in \Bbb N) [\text{ If } m_0 \le m \lt n \text{ Then } P(m)]$

Suppose we've proved that $Q(k)$ is always true. If $m \ge m_0$ set $k = m + 1$. So since $Q(k)$ is true and $m \lt k$ we have to take $P(m)$ as true.

We prove that $Q(k)$ is true for $k \ge 0$ using the OP's induction axiom.

The base case $Q(0)$ is vacuously true.

Step Case on $k$:

Assume $Q(k)$ is true. But setting $m = k$ in $\text{(1)}$ allows us to write as true

$\tag 3 \text{IF } \big [ \text{ if } m_0 \le m^{\prime} \lt k \text{ then } P(m^{\prime}) \big] \text{ THEN } P(k)$

But the hypothesis in $\text{(3)}$ is precisely the statement $Q(k)$. So $P(k)$ is true. But

$\tag 4 Q(k) \land P(k) \text{ implies } Q(k+1)$

and the induction is complete.

CopyPasteIt
  • 11,366
  • Thank you. I appreciate the detail you provided in explaining this, and it makes the solution immensely more clear. The main technique I've learned from this answer is to make sure to denote that something is true as an if-then statement, which makes the logic immensely more clear. – Clarinetist Jan 12 '20 at 04:21