content/model-theory/models-of-arithmetic/models-of-arithmetic.tex
1% Part: model-theory2% Chapter: models-of-arithmetic34\documentclass[../../../include/open-logic-chapter]{subfiles}56\begin{document}78\olchapter{mod}{mar}{Models of Arithmetic}910\olimport{introduction}1112\olimport{standard-models}1314\olimport{non-standard-models}1516\olimport{models-of-q}1718\olimport{models-of-pa}1920\olimport{computable-models}2122\OLEndChapterHook2324\end{document}
content/model-theory/models-of-arithmetic/introduction.tex
1% Part: model-theory2% Chapter: models-of-arithmetic3% Section: introduction45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{int}10\section{Introduction}1112The \emph{standard model} of arithmetic is the13!!{structure}~$\Struct{N}$ with $\Domain{N} = \Nat$ in which14$\Obj{0}$, $\prime$, $+$, $\times$, and $<$ are interpreted as you15would expect. That is, $\Obj{0}$ is $0$, $\prime$ is the successor16function, $+$ is interpreted as addition and $\times$ as multiplication17of the numbers in~$\Nat$. Specifically,18\begin{align*}19 \Assign{\Obj{0}}{N} & = 0\\20 \Assign{\prime}{N}(n) & = n + 1\\21 \Assign{+}{N}(n, m) & = n + m\\22 \Assign{\times}{N}(n, m) & = nm23\end{align*}24Of course, there are structures for $\Lang{L_A}$ that have domains25other than~$\Nat$. For instance, we can take $\Struct{M}$ with domain26$\Domain{M} = \{a\}^*$ (the finite sequences of the single27symbol~$a$, i.e., $\emptyset$, $a$, $aa$, $aaa$, \dots), and28interpretations29\begin{align*}30 \Assign{\Obj{0}}{M} & = \emptyset\\31 \Assign{\prime}{M}(s) & = s \concat a\\32 \Assign{+}{M}(n, m) & = a^{n + m}\\33 \Assign{\times}{M}(n, m) & = a^{nm}34\end{align*}35These two structures are ``essentially the same'' in the sense that36the only difference is the !!{element}s of the !!{domain}s but not how37the !!{element}s of the !!{domain}s are related among each other by38the interpretation functions. We say that the two !!{structure}s are39\emph{isomorphic}.4041It is an easy consequence of the compactness theorem that any theory42true in~$\Struct{N}$ also has models that are not isomorphic43to~$\Struct{N}$. Such structures are called \emph{non-standard}. The44interesting thing about them is that while the !!{element}s of a45standard model (i.e., $\Struct{N}$, but also all !!{structure}s46isomorphic to it) are exhausted by the values of the standard47numerals~$\num{n}$, i.e.,48\[49\Domain{N} = \Setabs{\Value{\num{n}}{N}}{n \in \Nat}50\]51that isn't the case in non-standard models: if $\Struct{M}$ is52non-standard, then there is at least one $x \in \Domain{M}$ such that53$x \neq \Value{\num{n}}{M}$ for all~$n$.5455These non-standard elements are pretty neat: they are ``infinite56natural numbers.'' But their existence also explains, in a sense, the57incompleteness phenomena. Consider an example, e.g., the consistency58statement for Peano arithmetic, $\OCon[\Th{PA}]$, i.e., $\lnot59\lexists[x][\OPrf[\Th{PA}](x, \gn{\lfalse})]$. Since $\Th{PA}$ neither60proves $\OCon[\Th{PA}]$ nor $\lnot \OCon[\Th{PA}]$, either can be61consistently added to $\Th{PA}$. Since $\Th{PA}$ is consistent,62$\Sat{N}{\OCon[\Th{PA}]}$, and consequently $\Sat/{N}{\lnot63 \OCon[\Th{PA}]}$. So $\Struct{N}$ is \emph{not} a model of $\Th{PA}64\cup \{\lnot \OCon[\Th{PA}]\}$, and all its models must be65nonstandard. Models of $\Th{PA} \cup \{\lnot \OCon[\Th{PA}]\}$ must66contain some !!{element} that serves as the witness that makes67$\lexists[x][\OPrf[\Th{PA}](\gn{\lfalse})]$ true, i.e., a G\"odel68number of !!a{derivation} of a contradiction from~$\Th{PA}$. Such69!!a{element} can't be standard---since $\Th{PA} \Proves \lnot70\OPrf[\Th{PA}](\num{n}, \gn{\lfalse})$ for every~$n$.7172\end{document}
content/model-theory/models-of-arithmetic/standard-models.tex
1% Part: model-theory2% Chapter: models-of-arithmetic3% Section: standard-models45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{stm}10\olsection{Standard Models of Arithmetic}1112The language of arithmetic~$\Lang{L_A}$ is obviously intended to be13about numbers, specifically, about natural numbers. So, ``the''14standard model~$\Struct{N}$ is special: it is the model we want to15talk about. But in logic, we are often just interested in structural16properties, and any two !!{structure}s that are isomorphic share17those. So we can be a bit more liberal, and consider any18!!{structure} that is isomorphic to~$\Struct{N}$ ``standard.''1920\begin{defn}21A !!{structure} for $\Lang{L_A}$ is \emph{standard} if it is22isomorphic to~$\Struct{N}$.23\end{defn}2425\begin{prop}26\ollabel{prop:standard-domain} If !!a{structure}~$\Struct{M}$ is standard,27then its domain is the set of values of the standard numerals, i.e.,28\[29\Domain{M} = \Setabs{\Value{\num{n}}{M}}{n \in \Nat}30\]31\end{prop}3233\begin{proof}34Clearly, every $\Value{\num{n}}{M} \in \Domain{M}$. We just have to35show that every $x \in \Domain{M}$ is equal to $\Value{\num{n}}{M}$36for some~$n$. Since $\Struct{M}$ is standard, it is isomorphic37to~$\Struct{N}$. Suppose $g\colon \Nat \to \Domain{M}$ is an38isomorphism. Then $g(n) = g(\Value{\num{n}}{N}) =39\Value{\num{n}}{M}$. But for every $x \in \Domain{M}$, there is an~$n40\in \Nat$ such that $g(n) = x$, since $g$ is !!{surjective}.41\end{proof}4243\begin{explain}44If a structure~$\Struct{M}$ for $\Lang{L_A}$ is standard, the elements45of its !!{domain} can all be named by the standard numerals $\num{0}$,46$\num{1}$, $\num{2}$, \dots, i.e., the terms $\Obj{0}$, $\Obj{0}'$,47$\Obj{0}''$, etc. Of course, this does not mean that the !!{element}s48of $\Domain{M}$ \emph{are} the numbers, just that we can pick them out49the same way we can pick out the numbers in~$\Domain{N}$.50\end{explain}5152\begin{prob}53Show that the converse of \olref[mod][mar][stm]{prop:standard-domain}54is false, i.e., give an example of !!a{structure}~$\Struct{M}$ with55$\Domain{M} = \Setabs{\Value{\num{n}}{M}}{n \in \Nat}$ that is not56isomorphic to~$\Struct{N}$.57\end{prob}5859\begin{prop}60\ollabel{prop:thq-standard}61If $\Sat{M}{\Th{Q}}$, and $\Domain{M} = \Setabs{\Value{\num{n}}{M}}{n62 \in \Nat}$, then $\Struct{M}$ is standard.63\end{prop}6465\begin{proof}66We have to show that $\Struct{M}$ is isomorphic67to~$\Struct{N}$. Consider the function $g\colon \Nat \to \Domain{M}$68defined by $g(n) = \Value{\num{n}}{M}$. By the hypothesis, $g$ is69!!{surjective}. It is also !!{injective}: $\Th{Q} \Proves70\eq/[\num{n}][\num{m}]$ whenever $n \neq m$. Thus, since71$\Sat{M}{\Th{Q}}$, $\Sat{M}{\eq/[\num{n}][\num{m}]}$, whenever $n \neq72m$. Thus, if $n \neq m$, then $\Value{\num{n}}{M} \neq73\Value{\num{m}}{M}$, i.e., $g(n) \neq g(m)$.7475We also have to verify that $g$ is an isomorphism.76\begin{enumerate}77\item We have $g(\Assign{\Obj{0}}{N}) = g(0)$ since,78 $\Assign{\Obj{0}}{N} = 0$. By definition of~$g$, $g(0) =79 \Value{\num{0}}{M}$. But $\num{0}$ is just $\Obj{0}$, and the value80 of a term which happens to be !!a{constant} is given by what the81 !!{structure} assigns to that !!{constant}, i.e.,82 $\Value{\Obj{0}}{M} = \Assign{\Obj{0}}{M}$. So we have83 $g(\Assign{\Obj{0}}{N}) = \Assign{\Obj{0}}{M}$ as required.84\item $g(\Assign{\prime}{N}(n)) = g(n+1)$, since $\prime$ in85 $\Struct{N}$ is the successor function on~$\Nat$. Then, $g(n+1) =86 \Value{\num{n+1}}{M}$ by definition of~$g$. But $\num{n+1}$ is the87 same term as $\num{n}'$, so $\Value{\num{n+1}}{M} =88 \Value{\num{n}'}{M}$. By the definition of the value function, this89 is $= \Assign{\prime}{M}(\Value{\num{n}}{M})$. Since90 $\Value{\num{n}}{M} = g(n)$ we get $g(\Assign{\prime}{N}(n)) =91 \Assign{\prime}{M}(g(n))$.92\item $g(\Assign{+}{N}(n,m)) = g(n+m)$, since $+$ in $\Struct{N}$ is93 the addition function on~$\Nat$. Then, $g(n+m) =94 \Value{\num{n+m}}{M}$ by definition of~$g$. But $\Th{Q} \Proves95 \num{n+m} = (\num{n} + \num{m})$, so $\Value{\num{n+m}}{M} =96 \Value{\num{n}+\num{m}}{M}$. By the definition of the value97 function, this is $=98 \Assign{+}{M}(\Value{\num{n}}{M},\Value{\num{m}}{M})$. Since99 $\Value{\num{n}}{M} = g(n)$ and $\Value{\num{m}}{M} = g(m)$, we100 get $g(\Assign{+}{N}(n, m)) = \Assign{+}{M}(g(n), g(m))$.101\item $g(\Assign{\times}{N}(n, m)) = \Assign{\times}{M}(g(n), g(m))$:102 Exercise.103\item $\tuple{n,m} \in \Assign{<}{N}$ iff $n < m$. If $n < m$, then104 $\Th{Q} \Proves \num{n} < \num{m}$, and also $\Sat{M}{\num{n} <105 \num{m}}$. Thus $\tuple{\Value{\num{n}}{M}, \Value{\num{m}}{M}} \in106 \Assign{<}{M}$, i.e., $\tuple{g(n), g(m)} \in \Assign{<}{M}$. If $n107 \not< m$, then $\Th{Q} \Proves \lnot \num{n} < \num{m}$, and108 consequently $\Sat/{M}{\num{n} < \num{m}}$. Thus, as before,109 $\tuple{g(n), g(m)} \notin \Assign{<}{M}$. Together, we get:110 $\tuple{n,m} \in \Assign{<}{N}$ iff $\tuple{g(n), g(m)} \in111 \Assign{<}{M}$.112\end{enumerate}113\end{proof}114115\begin{explain}116The function~$g$ is the most obvious way of defining a mapping from117$\Nat$ to the domain of any other !!{structure}~$\Struct{M}$ for118$\Lang{L_A}$, since every such $\Struct{M}$ contains !!{element}s named119by $\num{0}$, $\num{1}$, $\num{2}$, etc. So it isn't surprising that120if $\Struct{M}$ makes at least some basic statements about the121$\num{n}$'s true in the same way that~$\Struct{N}$ does, and $g$ is122also bijective, then $g$ will turn into an isomorphism. In fact, if123$\Domain{M}$ contains no !!{element}s other than what the $\num{n}$'s124name, it's the only one.125\end{explain}126127\begin{prop}128\ollabel{prop:thq-unique-iso} If $\Struct{M}$ is standard, then $g$ from129the proof of \olref{prop:thq-standard} is the only isomorphism from $\Struct{N}$ to~$\Struct{M}$.130\end{prop}131132\begin{proof}133Suppose $h\colon \Nat \to \Domain{M}$ is an isomorphism between134$\Struct{N}$ and~$\Struct{M}$. We show that $g = h$ by induction135on~$n$. If $n = 0$, then $g(0) = \Assign{\Obj{0}}{M}$ by definition136of~$g$. But since $h$ is an isomorphism, $h(0) =137h(\Assign{\Obj{0}}{N}) =\Assign{\Obj{0}}{M}$, so $g(0) = h(0)$.138139Now consider the case for $n+1$. We have140\begin{align*}141 g(n+1) & = \Value{\num{n+1}}{M} \text{ by definition of~$g$}\\142 & = \Value{\num{n}'}{M} \text{ since $\num{n+1}\ident \num{n}'$}\\143 & = \Assign{\prime}{M}(\Value{\num{n}}{M}) 144 \text{ by definition of $\Value{t'}{M}$}\\145 & = \Assign{\prime}{M}(g(n)) \text{ by definition of~$g$}\\146 & = \Assign{\prime}{M}(h(n)) \text{ by induction hypothesis}\\147 & = h(\Assign{\prime}{N}(n)) \text{ since $h$ is an isomorphism}\\148 & = h(n+1)149\end{align*}150\end{proof}151152\begin{explain}153For any !!{denumerable} set~$M$, there's !!a{bijection} between154$\Nat$ and $M$, so every such set~$M$ is potentially the !!{domain} of155a standard model~$\Struct{M}$. In fact, once you pick an object $z \in156M$ and a suitable function $s$ as $\Assign{\Obj{0}}{M}$ and157$\Assign{\prime}{M}$, the interpretations of $+$, $\times$, and $<$ is158already fixed. Only functions~$s\colon M \to M \setminus \{z\}$ that159are both !!{injective} and !!{surjective} are suitable in a standard160model as~$\Assign{\prime}{M}$. The range of $s$ cannot contain~$z$,161since otherwise $\lforall[x][\eq/[\Obj 0][x']]$ would be false. That162!!{sentence} is true in~$\Struct{N}$, and so $\Struct{M}$ also has to163make it true. The function~$s$ has to be !!{injective}, since the164successor function~$\Assign{\prime}{N}$ in~$\Struct{N}$ is, and that165$\Assign{\prime}{N}$ is !!{injective} is expressed by !!a{sentence}166true in~$\Struct{N}$. It has to be !!{surjective} because otherwise167there would be some $x \in M \setminus \{z\}$ not in the domain168of~$s$, i.e., the !!{sentence} $\lforall[x][(\eq[x][\Obj 0] \lor169\lexists[y][\eq[y'][x]])]$ would be false in~$\Struct{M}$---but it is170true in~$\Struct{N}$.171\end{explain}172173174\end{document}
content/model-theory/models-of-arithmetic/non-standard-models.tex
1% Part: model-theory2% Chapter: models-of-arithmetic3% Section: non-standard-models45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{nst}10\section{Non-Standard Models}1112\begin{explain}13We call !!a{structure} for $\Lang{L_A}$ standard if it is isomorphic14to~$\Struct{N}$. If !!a{structure} isn't isomorphic to~$\Struct{N}$,15it is called non-standard.16\end{explain}1718\begin{defn}19A !!{structure}~$\Struct{M}$ for $\Lang{L_A}$ is \emph{non-standard}20if it is not isomorphic to~$\Struct{N}$. The !!{element}s $x \in21\Domain{M}$ which are equal to $\Value{\num{n}}{M}$ for some $n \in22\Nat$ are called \emph{standard numbers} (of $\Struct{M}$), and those23not, \emph{non-standard numbers}.24\end{defn}2526\begin{explain}27By \olref[stm]{prop:standard-domain}, any standard !!{structure}28for~$\Lang{L_A}$ contains only standard !!{element}s. Consequently, a29non-standard !!{structure} must contain at least one non-standard30element. In fact, the existence of a non-standard !!{element}31guarantees that the !!{structure} is non-standard.32\end{explain}3334\begin{prop}35If !!a{structure}~$\Struct{M}$ for $\Lang{L_A}$ contains a36non-standard number, $\Struct{M}$ is non-standard.37\end{prop}3839\begin{proof}40Suppose not, i.e., suppose $\Struct{M}$ standard but contains a41non-standard number~$x$. Let $g\colon \Nat \to \Domain{M}$ be an42isomorphism. It is easy to see (by induction on~$n$) that43$g(\Value{\num{n}}{N}) = \Value{\num{n}}{M}$. In other words, $g$ maps44standard numbers of~$\Struct{N}$ to standard numbers45of~$\Struct{M}$. If $\Struct{M}$ contains a non-standard number, $g$46cannot be !!{surjective}, contrary to hypothesis.47\end{proof}4849\begin{prob}50Recall that $\Th{Q}$ contains the axioms51\begin{align*}52& \lforall[x][\lforall[y][(\eq[x'][y'] \lif \eq[x][y])]] \tag{$!Q_1$}\\53& \lforall[x][\eq/[\Obj 0][x']] \tag{$!Q_2$}\\54& \lforall[x][(\eq[x][\Obj 0] \lor \lexists[y][\eq[x][y']])] \tag{$!Q_3$}55\end{align*}56Give !!{structure}s~$\Struct{M_1}$, $\Struct{M_2}$, $\Struct{M_3}$ such that57\begin{enumerate}58\item $\Sat{M_1}{!Q_1}$, $\Sat{M_1}{!Q_2}$, $\Sat/{M_1}{!Q_3}$;59\item $\Sat{M_2}{!Q_1}$, $\Sat/{M_2}{!Q_2}$, $\Sat{M_2}{!Q_3}$; and60\item $\Sat/{M_3}{!Q_1}$, $\Sat{M_3}{!Q_2}$, $\Sat{M_3}{!Q_3}$;61\end{enumerate}62Obviously, you just have to specify~$\Assign{\Obj{0}}{M_i}$ and63$\Assign{\prime}{M_i}$ for each.64\end{prob}6566\begin{explain}67It is easy enough to specify non-standard !!{structure}s for68$\Lang{L_A}$. For instance, take the structure with !!{domain}~$\Int$69and interpret all non-logical symbols as usual. Since negative numbers70are not values of $\num{n}$ for any~$n$, this structure is71non-standard. Of course, it will not be a \emph{model} of arithmetic72in the sense that it makes the same sentences true73as~$\Struct{N}$. For instance, $\lforall[x][\eq/[x'][\Obj{0}]]$ is74false. However, we can prove that non-standard models of arithmetic75exist easily enough, using the compactness theorem.76\end{explain}7778\begin{prop}79Let $\Th{TA} = \Setabs{!A}{\Sat{N}{!A}}$ be the theory80of~$\Struct{N}$. $\Th{TA}$ has !!a{enumerable} non-standard model.81\end{prop}8283\begin{proof}84Expand $\Lang{L_A}$ by a new !!{constant}~$c$ and consider the set of85!!{sentence}s86\[87\Gamma = \Th{TA} \cup \{\eq/[c][\num{0}], \eq/[c][\num{1}],88\eq/[c][\num{2}], \dots\}89\]90Any model~$\Struct{M^c}$ of~$\Gamma$ would contain !!a{element}~$x =91\Assign{c}{M}$ which is non-standard, since $x \neq92\Value{\num{n}}{M}$ for all $n \in \Nat$. Also, obviously,93$\Sat{M^c}{\Th{TA}}$, since $\Th{TA} \subseteq \Gamma$. If we turn94$\Struct{M^c}$ into !!a{structure}~$\Struct{M}$ for $\Lang{L_A}$95simply by forgetting about~$c$, its domain still contains the96non-standard~$x$, and also~$\Sat{M}{\Th{TA}}$. The latter is97guaranteed since $c$ does not occur in~$\Th{TA}$. So, it suffices to98show that $\Gamma$ has a model.99100We use the compactness theorem to show that~$\Gamma$ has a model. If101every finite subset of~$\Gamma$ is satisfiable, so102is~$\Gamma$. Consider any finite subset $\Gamma_0 \subseteq103\Gamma$. $\Gamma_0$ includes some !!{sentence}s of~$\Th{TA}$ and some104of the form~$\eq/[c][\num{n}]$, but only finitely many. Suppose $k$ is105the largest number so that $\eq/[c][\num{k}] \in \Gamma_0$. Define106$\Struct{N_k}$ by expanding~$\Struct{N}$ to include the107interpretation~$\Assign{c}{N_k} = k+1$. $\Sat{N_k}{\Gamma_0}$: if $!A108\in \Th{TA}$, $\Sat{N_k}{!A}$ since $\Struct{N_k}$ is just109like~$\Struct{N}$ in all respects except~$c$, and $c$ does not occur110in~$!A$. And $\Sat{N_k}{\eq/[c][\num{n}]}$, since $n \le k$, and111$\Value{c}{N_k} = k+1$. Thus, every finite subset of~$\Gamma$ is112satisfiable.113\end{proof}114115\end{document}
content/model-theory/models-of-arithmetic/models-of-q.tex
1% Part: model-theory2% Chapter: models-of-arithmetic3% Section: non-standard-models45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{mdq}10\section{Models of $\Th{Q}$}1112\begin{explain}13We know that there are non-standard !!{structure}s that make the same14!!{sentence}s true as~$\Struct{N}$ does, i.e., is a model15of~$\Th{TA}$. Since $\Sat{N}{\Th{Q}}$, any model of~$\Th{TA}$ is also16a model of~$\Th{Q}$. $\Th{Q}$ is much weaker than~$\Th{TA}$, e.g.,17$\Th{Q} \Proves/ \lforall[x][\lforall[y][\eq[(x +18 y)][(y+x)]]]$. Weaker theories are easier to satisfy: they have19more models. E.g., $\Th{Q}$ has models which make20$\lforall[x][\lforall[y][\eq[(x + y)][(y+x)]]]$ false, but those21cannot also be models of~$\Th{TA}$, or $\Th{PA}$ for that matter.22Models of $\Th{Q}$ are also relatively simple: we can specify them23explicitly.24\end{explain}2526\begin{ex}27\ollabel{ex:model-K-of-Q}28Consider the !!{structure}~$\Struct{K}$ with domain $\Domain{K} = \Nat29\cup \{a\}$ and interpretations30\begin{align*}31 \Assign{\Obj{0}}{K} & = 0\\32 \Assign{\prime}{K}(x) & =33 \begin{cases}34 x+1 & \text{if $x\in \Nat$}\\35 a & \text{if $x = a$}36 \end{cases}\\37 \Assign{+}{K}(x, y) & =38 \begin{cases}39 x+y & \text{if $x$, $y \in\Nat$}\\40 a & \text{otherwise}41 \end{cases}\\42 \Assign{\times}{K}(x, y) & =43 \begin{cases}44 xy & \text{if $x$, $y \in\Nat$}\\45 0 & \text{if $x = 0$ or $y = 0$}\\46 a & \text{otherwise}\\47 \end{cases}\\48 \Assign{<}{K} & =49 \Setabs{\tuple{x,y}}{x, y \in \Nat \text{ and } x<y} \cup50 \Setabs{\tuple{x,a}}{x \in \Domain{K}}51\end{align*}52To show that $\Sat{K}{\Th{Q}}$ we have to verify that all axioms53of~$\Th{Q}$ are true in~$\Struct{K}$. For convenience, let's write54$x^\nssucc$ for $\Assign{\prime}{K}(x)$ (the ``successor'' of $x$55in~$\Struct{K}$), $x \nsplus y$ for $\Assign{+}{K}(x, y)$ (the ``sum''56of $x$ and $y$ in~$\Struct{K}$, $x \nstimes y$ for57$\Assign{\times}{K}(x, y)$ (the ``product'' of $x$ and~$y$58in~$\Struct{K}$), and $x \nsless y$ for $\tuple{x,y} \in59\Assign{<}{K}$. With these abbreviations, we can give the operations60in~$\Struct{K}$ more perspicuously as61\[62\begin{array}{c|c}63 x & x^\nssucc \\64 \hline65 n & n+1 \\66 a & a67\end{array}68\qquad69\begin{array}{c|ccc}70 x \nsplus y & 0 & m & a \\71 \hline72 0 & 0 & m & a \\73 n & n & n+m & a \\74 a & a & a & a \\75\end{array}76\qquad77\begin{array}{c|ccc}78 x \nstimes y & 0 & m & a \\79 \hline80 0 & 0 & 0 & 0 \\81 n & 0 & nm & a \\82 a & 0 & a & a \\83\end{array}84\]85We have $n \nsless m$ iff $n<m$ for $n$, $m \in \Nat$ and $x \nsless86a$ for all~$x \in \Domain{K}$.8788$\Sat{K}{\lforall[x][\lforall[y][(\eq[x'][y'] \lif \eq[x][y])]]}$89since $\nssucc$ is !!{injective}. $\Sat{K}{\lforall[x][\eq/[\Obj900][x']]}$ since $0$~is not a $\nssucc$-successor in~$\Struct{K}$.91$\Sat{K}{\lforall[x][(\eq[x][\Obj 0] \lor \lexists[y][\eq[x][y']])]}$92since for every $n>0$, $n = (n-1)^\nssucc$, and $a = a^\nssucc$.9394$\Sat{K}{\lforall[x][\eq[(x + \Obj 0)][x]]}$ since $n \nsplus 0 = n+095= n$, and $a\nsplus 0 = a$ by definition of~$\nsplus$.96$\Sat{K}{\lforall[x][\lforall[y][\eq[(x + y')][(x + y)']]]}$ is a bit97trickier. If $n$, $m$ are both standard, we have:98\begin{align*}99(n \nsplus m^\nssucc) & = (n+(m+1)) = (n+m)+1 = (n \nsplus m)^\nssucc 100\intertext{since $\nsplus$ and $^\nssucc$ agree with $+$ and $\prime$ on101 standard numbers. Now suppose $x \in \Domain{K}$. Then}102(x \nsplus a^\nssucc) & = (x \nsplus a) = a = a^\nssucc = (x \nsplus a)^\nssucc103\intertext{The remaining case is if $y \in \Domain{K}$ but $x =104 a$. Here we also have to distinguish cases according to whether $y =105 n$ is standard or $y = b$:}106(a \nsplus n^\nssucc) & = (a \nsplus (n+1)) = a = a^\nssucc = (a \nsplus n)^\nssucc\\107(a \nsplus a^\nssucc) & = (a \nsplus a) = a = a^\nssucc = (a \nsplus a)^\nssucc108\end{align*}109This is of course a bit more detailed than needed. For instance, since110$a \nsplus z = a$ whatever $z$ is, we can immediately conclude $a \nsplus111a^\nssucc = a$. The remaining axioms can be verified the same way.112113$\Struct{K}$ is thus a model of~$\Th{Q}$. Its ``addition''~$\nsplus$114is also commutative. But there are other sentences true115in~$\Struct{N}$ but false in~$\Struct{K}$, and vice versa. For116instance, $a \nsless a$, so $\Sat{K}{\lexists[x][x < x]}$ and117$\Sat/{K}{\lforall[x][\lnot x<x]}$. This shows that $\Th{Q} \Proves/118\lforall[x][\lnot x < x]$.119\end{ex}120121\begin{prob}122Prove that $\Struct{K}$ from \olref[mod][mar][mdq]{ex:model-K-of-Q}123satisfies the remaining axioms of~$\Th{Q}$,124\begin{align*}125 & \lforall[x][\eq[(x \times \Obj 0)][\Obj 0]] \tag{$!Q_6$}\\126 & \lforall[x][\lforall[y][\eq[(x \times y')][((x \times y) + x)]]] \tag{$!Q_7$}\\127 & \lforall[x][\lforall[y][(x < y \liff \lexists[z][\eq[(z' + x)][y]])]] \tag{$!Q_8$}128\end{align*}129Find !!a{sentence} only involving~$\prime$ true in~$\Struct{N}$ but130false in~$\Struct{K}$.131\end{prob}132133\begin{ex}134\ollabel{ex:model-L-of-Q} Consider the !!{structure}~$\Struct{L}$ with135domain $\Domain{L} = \Nat \cup \{a, b\}$ and interpretations136$\Assign{\prime}{L} = \nssucc$, $\Assign{+}{L} = \nsplus$ given by137\[138\begin{array}{c|c}139 x & x^\nssucc \\140 \hline141 n & n+1 \\142 a & a\\143 b & b144\end{array}145\qquad146\begin{array}{c|ccc}147 x \nsplus y & m & a & b\\148 \hline149 n & n+m & b & a\\150 a & a & b & a\\151 b & b & b & a152\end{array}153\]154Since $\nssucc$ is !!{injective}, $0$ is not in its range, and every155$x \in \Domain{L}$ other than~$0$ is, axioms $!Q_1$--$!Q_3$ are true156in~$\Struct{L}$. For any $x$, $x \nsplus 0 = x$, so $!Q_4$ is true as157well. For $!Q_5$, consider $x \nsplus y^\nssucc$ and $(x \nsplus158y)^\nssucc$. They are equal if $x$ and $y$ are both standard, since159then $\nssucc$ and $\nsplus$ agree with $\prime$ and $+$. If $x$ is160non-standard, and $y$ is standard, we have $x \nsplus y^\nssucc = x =161x^\nssucc = (x \nsplus y)^\nssucc$. If $x$ and $y$ are both162non-standard, we have four cases:163\begin{align*}164& a \nsplus a^\nssucc = b = b^\nssucc = (a \nsplus a)^\nssucc\\165& b \nsplus b^\nssucc = a = a^\nssucc = (b \nsplus b)^\nssucc\\166& b \nsplus a^\nssucc = b = b^\nssucc = (b \nsplus y)^\nssucc\\167& a \nsplus b^\nssucc = a = a^\nssucc = (a \nsplus b)^\nssucc\\168\intertext{If $x$ is standard, but $y$ is non-standard, we have}169& n \nsplus a^\nssucc = n \nsplus a = b = b^\nssucc = (n \nsplus a)^\nssucc\\170& n \nsplus b^\nssucc = n \nsplus b = a = a^\nssucc = (n \nsplus b)^\nssucc171\end{align*}172So, $\Sat{L}{!Q_5}$. However, $a \nsplus 0 \neq 0 \nsplus a$, so173$\Sat/{L}{\lforall[x][\lforall[y][\eq[(x+y)][(y+x)]]]}$.174\end{ex}175176\begin{prob}177Expand $\Struct{L}$ of \olref[mod][mar][mdq]{ex:model-L-of-Q} to178include $\nstimes$ and $\nsless$ that interpret~$\times$ and $<$. Show179that your structure satisfies the remaining axioms of~$\Th{Q}$,180\begin{align*}181& \lforall[x][\eq[(x \times \Obj 0)][\Obj 0]] \tag{$!Q_6$}\\ &182 \lforall[x][\lforall[y][\eq[(x \times y')][((x \times y) + x)]]]183 \tag{$!Q_7$}\\ & \lforall[x][\lforall[y][(x < y \liff \lexists[z][\eq[(z'+x)][y]])]] \tag{$!Q_8$}184\end{align*}185\end{prob}186187\begin{prob}188In $\Struct{L}$ of \olref[mod][mar][mdq]{ex:model-L-of-Q}, $a^\nssucc189= a$ and $b^\nssucc = b$. Is there a model of~$\Th{Q}$ in which190$a^\nssucc = b$ and $b^\nssucc = a$?191\end{prob}192193\begin{explain}194We've explicitly constructed models of~$\Th{Q}$ in which the195non-standard !!{element}s live ``beyond'' the standard elements. In196fact, that much is required by the axioms. A non-standard197!!{element}~$x$ cannot be ${} \nsless 0$, since $\Th{Q} \Proves198\lforall[x][\lnot x<0]$ (see \olref[inc][req][min]{lem:less-zero}).199Also, for every $n$, $\Th{Q} \Proves \lforall[x][(x < \num{n}' \lif200(\eq[x][\num{0}] \lor \eq[x][\num{1}] \lor \dots \lor201\eq[x][\num{n}]))]$ (\olref[inc][req][min]{lem:less-nsucc}), so we202can't have $a \nsless n$ for any~$n>0$.203\end{explain}204205\end{document}
content/model-theory/models-of-arithmetic/models-of-pa.tex
1% Part: first-order-logic2% Chapter: models-of-arithmetic3% Section: models-of-pa45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{mpa}10\section{Models of $\Th{PA}$}1112\begin{explain}13Any non-standard model of~$\Th{TA}$ is also one of~$\Th{PA}$. We know14that non-standard models of~$\Th{TA}$ and hence of~$\Th{PA}$ exist. We15also know that such non-standard models contain non-standard16``numbers,'' i.e., !!{element}s of the domain that are ``beyond'' all17the standard ``numbers.'' But how are they arranged? How many are18there? We've seen that models of the weaker theory~$\Th{Q}$ can19contain as few as a single non-standard number. But these simple20!!{structure}s are not models of $\Th{PA}$ or $\Th{TA}$.2122The key to understanding the structure of models of $\Th{PA}$ or23$\Th{TA}$ is to see what facts are !!{derivable} in these theories.24For instance, already $\Th{PA}$ proves that $\lforall[x][\eq/[x][x']]$25and $\lforall[x][\lforall[y][\eq[(x+y)][(y+x)]]]$, so this rules out26simple structures (in which these !!{sentence}s are false) as models27of~$\Th{PA}$.2829Suppose~$\Struct{M}$ is a model of~$\Th{PA}$. Then if $\Th{PA} \Proves30!A$, $\Sat{M}{!A}$. Let's again use $\nszero$ for31$\Assign{\Obj{0}}{M}$, $\nssucc$ for $\Assign{\prime}{M}$, $\nsplus$32for $\Assign{+}{M}$, $\nstimes$ for $\Assign{\times}{M}$, and33$\nsless$ for $\Assign{<}{M}$. Any !!{sentence}~$!A$ then states some34condition about $\nszero$, $\nssucc$, $\nsplus$, $\nstimes$, and35$\nsless$, and if $\Sat{M}{!A}$ that condition must be satisfied. For36instance, if $\Sat{M}{!Q_1}$, i.e.,37$\Sat{M}{\lforall[x][\lforall[y][(\eq[x'][y'] \lif \eq[x][y])]]}$,38then $\nssucc$ must be !!{injective}.39\end{explain}4041\begin{prop}42In $\Struct{M}$, $\nsless$ is a linear strict order, i.e., it43satisfies:44\begin{enumerate}45\item Not $x \nsless x$ for any~$x \in \Domain{M}$.46\item If $x \nsless y$ and $y \nsless z$ then $x \nsless z$.47\item For any $x \neq y$, $x \nsless y$ or $y \nsless x$48\end{enumerate}49\end{prop}5051\begin{proof}52$\Th{PA}$ proves:53\begin{enumerate}54\item $\lforall[x][\lnot x < x]$55\item $\lforall[x][\lforall[y][\lforall[z][((x < y \land y < z) \lif x < z)]]]$56\item $\lforall[x][\lforall[y][((x < y \lor y < x) \lor \eq[x][y]))]]$57\end{enumerate}58\end{proof}5960\begin{prop}61\ollabel{prop:M-discrete} $\nszero$ is the least !!{element}62of~$\Domain{M}$ in the $\nsless$-ordering. For any $x$, $x \nsless63x^\nssucc$, and $x^\nssucc$ is the $\nsless$-least !!{element} with64that property. For any $x$, there is a unique $y$ such that65$y^\nssucc = x$. (We call $y$ the ``predecessor'' of~$x$66in~$\Struct{M}$, and denote it by~$^\nssucc x$.)67\end{prop}6869\begin{proof}70Exercise. 71\end{proof}7273\begin{prob}74Find !!{sentence}s in~$\Lang{L_A}$ !!{derivable} in~$\Th{PA}$ (and75hence true in~$\Struct{N}$) which guarantee the properties of76$\nszero$, $\nssucc$, and $\nsless$ in77\olref[mod][mar][mpa]{prop:M-discrete}78\end{prob}7980\begin{prop}81All standard !!{element}s of~$\Struct{M}$ are less than (according82to~$\nsless$) all non-standard !!{element}s.83\end{prop}8485\begin{proof}86We'll use $n$ as short for $\Value{\num{n}}{M}$, a standard87!!{element} of~$\Struct{M}$. Already $\Th{Q}$ proves that, for any~$n88\in \Nat$, $\lforall[x][(x < \num{n}' \lif (\eq[x][\num{0}] \lor89 \eq[x][\num{1}] \lor \dots \lor \eq[x][\num{n}]))]$. There are no90!!{element}s that are $\nsless \nszero$. So if $n$ is standard and $x$91is non-standard, we cannot have $x \nsless n$. By definition, a92non-standard element is one that isn't $\Value{\num{n}}{M}$ for any~$n93\in \Nat$, so $x \neq n$ as well. Since $\nsless$ is a linear order,94we must have $n \nsless x$.95\end{proof}9697\begin{prop}98Every nonstandard !!{element}~$x$ of~$\Domain{M}$ is an element of the subset99\[100\dots ^{\nssucc\nssucc\nssucc}x \nsless ^{\nssucc\nssucc}x \nsless101^{\nssucc}x \nsless x \nsless x^{\nssucc} \nsless x^{\nssucc\nssucc}102\nsless x^{\nssucc\nssucc\nssucc} \nsless \dots103\]104We call this subset the \emph{block of~$x$} and write it as $[x]$. It105has no least and no greatest !!{element}. It can be characterized as106the set of those $y \in \Domain{M}$ such that, for some standard~$n$,107$x \nsplus n = y$ or $y \nsplus n = x$.108\end{prop}109110\begin{proof}111Clearly, such a set~$[x]$ always exists since every !!{element}~$y$ of112$\Domain{M}$ has a unique successor~$y^\nssucc$ and unique113predecessor~$^\nssucc y$. For successive !!{element}s $y$, $y^\nssucc$114we have $y \nsless y^\nssucc$ and $y^\nssucc$ is the $\nsless$-least115!!{element} of~$\Domain{M}$ such that $y$ is $\nsless$-less than it.116Since always $^\nssucc y \nsless y$ and $y \nsless y^\nssucc$, $[x]$117has no least or greatest !!{element}. If $y \in [x]$ then $x \in [y]$,118for then either $y^{\nssucc\dots\nssucc} = x$ or119$x^{\nssucc\dots\nssucc} = y$. If $y^{\nssucc\dots\nssucc} = x$ (with120$n$ $\nssucc$'s), then $y \nsplus n = x$ and conversely, since121$\Th{PA} \Proves \lforall[x][\eq[x^{\prime\dots\prime}][(x +122 \num{n})]]$ (if $n$ is the number of~$\prime$'s).123\end{proof}124125\begin{prop}126If $[x] \neq [y]$ and $x \nsless y$, then for any $u \in [x]$ and any127$v \in [y]$, $u \nsless v$.128\end{prop}129130\begin{proof}131Note that $\Th{PA} \Proves \lforall[x][\lforall[y][(x < y \lif (x' < y132 \lor x' = y))]]$. Thus, if $u \nsless v$, we also have $u \nsplus133n^\nssucc \nsless v$ for any~$n$ if $[u] \neq [v]$.134135Any $u \in [x]$ is $\nsless y$: $x \nsless y$ by assumption. If $u136\nsless x$, $u \nsless y$ by transitivity. And if $x \nsless u$ but $u137\in [x]$, we have $u = x\nsplus n^\nssucc$ for some~$n$, and so $u138\nsless y$ by the fact just proved.139 140Now suppose that $v \in [y]$ is $\nsless y$, i.e., $v \nsplus141m^\nssucc = y$ for some standard~$m$. This rules out $v \nsless x$,142otherwise $y = v \nsplus m^\nssucc \nsless x$. Clearly also, $x \neq143v$, otherwise $x \nsplus m^\nssucc = v \nsplus m^\nssucc = y$ and we144would have $[x] = [y]$. So, $x \nsless v$. But then also $x \nsplus145n^\nssucc \nsless v$ for any~$n$. Hence, if $x \nsless u$ and $u \in146[x]$, we have $u \nsless v$. If $u \nsless x$ then $u \nsless v$ by147transitivity.148149Lastly, if $y \nsless v$, $u \nsless v$ since, as we've shown, $u150\nsless y$ and $y \nsless v$.151\end{proof}152153\begin{cor}154If $[x] \neq [y]$, $[x] \cap [y] = \emptyset$.155\end{cor}156157\begin{proof}158Suppose $z \in [x]$ and $x \nsless y$. Then $z \nsless u$ for all $u159\in [y]$. If $z \in [y]$, we would have $z \nsless z$. Similarly if $y160\nsless x$.161\end{proof}162163\begin{explain}164This means that the blocks themselves can be ordered in a way that165respects $\nsless$: $[x] \nsless [y]$ iff $x \nsless y$, or,166equivalently, if $u \nsless v$ for any $u \in [x]$ and $v \in [y]$.167Clearly, the standard block $[0]$ is the least block. It intersects168with no non-standard block, and no two non-standard blocks intersect169either. Specifically, you cannot ``reach'' a different block by taking170repeated successors or predecessors.171\end{explain}172173\begin{prop}174If $x$ and $y$ are non-standard, then $x \nsless x \nsplus y$ and $x175\nsplus y \notin [x]$.176\end{prop}177178\begin{proof}179If $y$ is nonstandard, then $y \neq \nszero$. $\Th{PA} \Proves180\lforall[x][(y \neq \Obj{0} \lif x < (x+y))]$. Now suppose $x181\nsplus y \in [x]$. Since $x \nsless x \nsplus y$, we would have $x182\nsplus n^\nssucc = x \nsplus y$. But $\Th{PA} \Proves183\lforall[x][\lforall[y][\lforall[z][(\eq[(x+y)][(x+z)] \lif y = z)]]]$184(the cancellation law for addition). This would mean $y = n^\nssucc$185for some standard~$n$; but $y$ is assumed to be non-standard.186\end{proof}187188\begin{prop}189There is no least non-standard block.190\end{prop}191192\begin{proof}193$\Th{PA} \Proves \lforall[x][\lexists[y][(\eq[(y+y)][x] \lor194 \eq[(y+y)'][x])]]$, i.e., that every $x$ is divisible by~$2$195 (possibly with remainder~$1$). If $x$ is non-standard, so is~$y$. By196 the preceding proposition, $y \nsless y \nsplus y$ and $y \nsplus y197 \notin [y]$. Then also $y \nsless (y \nsplus y)^\nssucc$ and $(y198 \nsplus y)^\nssucc \notin [y]$. But $x = y \nsplus y$ or $x = (y199 \nsplus y)^\nssucc$, so $y \nsless x$ and $y \notin [x]$.200\end{proof}201202\begin{prop}203There is no largest block.204\end{prop}205206\begin{proof}207Exercise.208\end{proof}209210\begin{prob}211Show that in a non-standard model of~$\Th{PA}$, there is no largest212block.213\end{prob}214215\begin{prop}216\ollabel{prop:blocks-dense}217The ordering of the blocks is dense. That is, if $x \nsless y$ and218$[x] \neq [y]$, then there is a block $[z]$ distinct from both that is219between them.220\end{prop}221222\begin{proof}223Suppose $x \nsless y$. As before, $x \nsplus y$ is divisible by two224(possibly with remainder): there is a $z \in \Domain{M}$ such that225either $x \oplus y = z \oplus z$ or $x \oplus y = (z \oplus226z)^\nssucc$. The element $z$ is the ``average'' of $x$ and~$y$, and $x227\nsless z$ and $z \nsless y$.228\end{proof}229230\begin{prob}231Write out a detailed proof of232\olref[mod][mar][mpa]{prop:blocks-dense}. Which !!{sentence} must233$\Th{PA}$ !!{derive} in order to guarantee the existence of~$z$? Why234is $x \nsless z$ and $z \nsless y$, and why is $[x] \neq [z]$ and $[z]235\neq [y]$?236\end{prob}237238\begin{explain}239The non-standard blocks are therefore ordered like the rationals: they240form !!a{denumerable} dense linear ordering without endpoints. One can show241that any two such !!{denumerable} orderings are isomorphic. It follows242that for any two !!{enumerable} non-standard models $\Struct{M}_1$ and243$\Struct{M_2}$ of true arithmetic, their reducts to the language244containing $<$ and $=$ only are isomorphic. Indeed, an isomorphism $h$245can be defined as follows: the standard parts of $\Struct{M_1}$ and246$\Struct{M_2}$ are isomorphic to the standard model $\Struct{N}$ and247hence to each other. The blocks making up the non-standard part are248themselves ordered like the rationals and therefore isomorphic; an249isomorphism of the blocks can be extended to an isomorphism250\emph{within} the blocks by matching up arbitrary elements in each,251and then taking the image of the successor of $x$ in $\Struct{M_1}$ to252be the successor of the image of $x$ in $\Struct{M_2}$. Note that it253does \emph{not} follow that $\mathfrak{M}_1$ and $\mathfrak{M}_2$ are254isomorphic in the full language of arithmetic (indeed, isomorphism is255always relative to !!a{language}), as there are non-isomorphic ways to256define addition and multiplication over $\Domain{M_1}$ and257$\Domain{M_2}$. (This also follows from a famous theorem due to Vaught258that the number of countable models of a complete theory cannot be~2.)259\end{explain}260\end{document}
content/model-theory/models-of-arithmetic/computable-models.tex
1% Part: model-theory2% Chapter: models-of-arithmetic3% Section: computable-models45\documentclass[../../../include/open-logic-section]{subfiles}67\begin{document}89\olfileid{mod}{mar}{cmp}10\section{Computable Models of Arithmetic}1112\begin{explain}13The standard model~$\Struct{N}$ has two nice features. Its domain is14the natural numbers~$\Nat$, i.e., its elements are just the kinds of15things we want to talk about using the language of arithmetic, and the16standard numeral~$\num{n}$ actually picks out~$n$. The other nice17feature is that the interpretations of the non-logical symbols18of~$\Lang{L_A}$ are all \emph{computable}. The successor, addition,19and multiplication functions which serve as $\Assign{\prime}{N}$,20$\Assign{+}{N}$, and $\Assign{\times}{N}$ are computable functions of21numbers. (Computable by Turing machines, or definable by primitive22recursion, say.) And the less-than relation on~$\Struct{N}$, i.e.,23$\Assign{<}{N}$, is decidable.2425Non-standard models of arithmetical theories such as $\Th{Q}$ and26$\Th{PA}$ must contain non-standard elements. Thus their domains27typically include !!{element}s in addition to~$\Nat$. However, any28countable !!{structure} can be built on any !!{denumerable} set,29including~$\Nat$. So there are also non-standard models with30domain~$\Nat$. In such models~$\Struct{M}$, of course, at least some31numbers cannot play the roles they usually play, since some~$k$ must32be different from~$\Value{\num{n}}{M}$ for all~$n \in \Nat$.33\end{explain}3435\begin{defn}36!!^a{structure}~$\Struct{M}$ for $\Lang{L_A}$ is \emph{computable} iff37 $\Domain{M} = \Nat$ and $\Assign{\prime}{M}$, $\Assign{+}{M}$,38 $\Assign{\times}{M}$ are computable functions and $\Assign{<}{M}$ is39 a decidable relation.40\end{defn}4142\begin{ex}\ollabel{ex:comp-model-q}43Recall the structure $\Struct{K}$ from \olref[mdq]{ex:model-K-of-Q}.44Its domain was $\Domain{K} = \Nat45\cup \{a\}$ and interpretations46\begin{align*}47 \Assign{\Obj{0}}{K} & = 0\\48 \Assign{\prime}{K}(x) & =49 \begin{cases}50 x+1 & \text{if $x\in \Nat$}\\51 a & \text{if $x = a$}52 \end{cases}\\53 \Assign{+}{K}(x, y) & =54 \begin{cases}55 x+y & \text{if $x$, $y \in\Nat$}\\56 a & \text{otherwise}57 \end{cases}\\58 \Assign{\times}{K}(x, y) & =59 \begin{cases}60 xy & \text{if $x$, $y \in\Nat$}\\61 0 & \text{if $x=0$ or $y=0$}\\62 a & \text{otherwise}\\63 \end{cases}\\64 \Assign{<}{K} & =65 \Setabs{\tuple{x,y}}{x, y \in \Nat \text{ and } x<y} \cup66 \Setabs{\tuple{x,a}}{n \in \Domain{K}}67\end{align*}68But $\Domain{K}$ is !!{denumerable} and so is equinumerous69with~$\Nat$. For instance, $g\colon \Nat \to \Domain{K}$ with $g(0) =70a$ and $g(n) = n+1$ for $n>0$ is !!a{bijection}. We can turn it into71an isomorphism between a new model~$\Struct{K'}$ of~$\Th{Q}$ and72$\Struct{K}$. In $\Struct{K'}$, we have to assign different functions73and relations to the symbols of~$\Lang{L_A}$, since different74!!{element}s of~$\Nat$ play the roles of standard and non-standard75numbers.7677Specifically, $0$ now plays the role of~$a$, not of the smallest78standard number. The smallest standard number is now~$1$. So we assign79$\Assign{\Obj{0}}{K'} = 1$. The successor function is also different80now: given a standard number, i.e., an $n > 0$, it still returns81$n+1$. But $0$ now plays the role of~$a$, which is its own82successor. So $\Assign{\prime}{K'}(0) = 0$. For addition and83multiplication we likewise have84\begin{align*}85\Assign{+}{K'}(x, y) & =86 \begin{cases}87 x+y-1 & \text{if $x$, $y >0$}\\88 0 & \text{otherwise}89 \end{cases}\\90 \Assign{\times}{K'}(x, y) & =91 \begin{cases}92 1 & \text{if $x = 1$ or $y = 1$}\\93 xy - x - y + 2 & \text{if $x$, $y > 1$}\\94 0 & \text{otherwise}\\95 \end{cases}96\end{align*}97And we have $\tuple{x, y} \in \Assign{<}{K'}$ iff $x < y$ and $x > 0$98and $y > 0$, or if $y = 0$.99100All of these functions are computable functions of natural numbers and101$\Assign{<}{K'}$ is a decidable relation on~$\Nat$---but they are not102the same functions as successor, addition, and multiplication103on~$\Nat$, and $\Assign{<}{K'}$ is not the same relation as~$<$104on~$\Nat$.105\end{ex}106107\begin{prob}108Give !!a{structure}~$\Struct{L'}$ with $\Domain{L'} = \Nat$ isomorphic109to~$\Struct{L}$ of \olref[mod][mar][mdq]{ex:model-L-of-Q}.110\end{prob}111112\begin{explain}113\Olref{ex:comp-model-q} shows that $\Th{Q}$ has computable114non-standard models with domain~$\Nat$. However, the following result115shows that this is not true for models of~$\Th{PA}$ (and thus also for116models of~$\Th{TA}$).117\end{explain}118119\begin{thm}[Tennenbaum's Theorem]120$\Struct{N}$ is the only computable model of~$\Th{PA}$.121\end{thm}122 123\end{document}