<?xml version="1.0" encoding="ISO-8859-1"?>
<QEDEQ 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:noNamespaceSchemaLocation="http://www.qedeq.org/0_01_09/xml/qedeq.xsd">
  <HEADER email="mime@qedeq.org">
    <SPECIFICATION name="qedeq_set_theory_v1" ruleVersion="1.00.00">
      <LOCATIONS>
        <LOCATION value="http://qedeq.org/0.01.09/doc/math"/>
      </LOCATIONS>
    </SPECIFICATION>
    <TITLE>
      <LATEX language="en">
         Axiomatic Set Theory
      </LATEX>
      <LATEX language="de">
         Axiomatische Mengenlehre
      </LATEX>
    </TITLE>
    <ABSTRACT>
      <LATEX language="en">
        <![CDATA[
          This document contains the mathematical foundation of set theory. Goal is the presentation of elementary results which are needed within other mathematical disciplines. After the basics the {\emph boolean algebra of classes} is in focus. Next are some thoughts about \emph{relations} and \emph{functions}. An important achivement is the definition of the \emph{natural~numbers} and their fullfillment of the \emph{Peano axioms}. Also the word \emph{recursion} is discused.
          
          \par
          Although the presentation is axiomatic the results shall match the mathematical usage. For that reason the set theoretic axiom system of \emph{A.~P.~Morse} and \emph{J.~L.~Kelley} (MK\index{MK})was choosen.

          \par
          This document is not finished and is updated from time to time.
        ]]>
      </LATEX>
      <LATEX language="de">
        <![CDATA[
         Dieses Dokument beschreibt die mathematischen Grundlagen der Mengenlehre. Ziel ist dabei die Bereitstellung von elementaren Ergebnissen der Mengenlehre, die in anderen mathematischen Disziplinen benötigt werden. Nach den Anfangsgründen wird die {\emph Boolsche Algebra der Klassen} betrachtet. Es schliessen sich Betrachtungen über {\emph Relationen} und {\emph Funktionen} an. Ein wichtiges Ergebnis sind die Definition der {\emph natürlichen Zahlen} und die Erfüllung der {\emph Peano-Axiome} durch diese. Auch auf den Begriff der {\emph Rekursion} wird eingegangen.

         \par
         Die Darstellung erfolgt in axiomatischer Weise soll aber im Ergebnis der mathematischen Praxis entsprechen. Aus diesem Grunde wird auch das Axiomensystem der Mengenlehre von \emph{A.~P.~Morse} und  \emph{J.~L.~Kelley} (MK\index{MK}) verwendet.

         \par
         Dieses Dokument ist noch nicht fertiggestellt und wird von Zeit zu Zeit aktualisiert.
        ]]>
      </LATEX>
    </ABSTRACT>
    <AUTHORS>
      <AUTHOR email="michael@meyling.com">
        <NAME>
          <LATEX language="de">
            Michael Meyling
          </LATEX>
        </NAME>
      </AUTHOR>
    </AUTHORS>
    <IMPORTS>
      <!-- still sample references -->
      <IMPORT label="predlogic">
        <SPECIFICATION name="predlogic" ruleVersion="1.00.00">
          <LOCATIONS>
            <LOCATION value="http://qedeq.org/0.01.09/sample"/>
          </LOCATIONS>
        </SPECIFICATION>
      </IMPORT>
      <IMPORT label="proplogic">
        <SPECIFICATION name="proplogic" ruleVersion="1.00.00">
          <LOCATIONS>
            <LOCATION value="http://qedeq.org/0.01.09/sample"/>
          </LOCATIONS>
        </SPECIFICATION>
      </IMPORT>
    </IMPORTS>
  </HEADER>


  <CHAPTER noNumber="true">
    <TITLE>
      <LATEX language="en">
        Preface\label{ch:preface}
      </LATEX>
      <LATEX language="de">
        Vorwort\label{ch:preface}
      </LATEX>
    </TITLE>
    <INTRODUCTION>
      <LATEX language="en">
        <![CDATA[
          Mathematics is a science with a structure that achieved enormous dimensions in the course of time. This huge stronghold has only a small set theoretic foundation and its firmness rests upon simple predicate calculus mortar. In principle the assembly could be comprehended by any mathematician. From every newest turret of mathematical cognition each path of logical dependency could be followed all the way down to the set theoretic roots.
          
          \par
          +++ missing translation
          
          \par
          Lets start by the roots\ldots

          \par
          This document is not finished and is updated from time to time. Especially at the locations marked with {\glqq+++\grqq} additions or changes will be made.

          \par
          I am deeply grateful to my wife \emph{Gesine~Dräger} and our son \emph{Lennart} for their support and patience.

          \par
          \vspace*{1cm} Hamburg, december 2005 \\
          \hspace*{\fill} Michael Meyling
        ]]>
      </LATEX>
      <LATEX language="de">
        <![CDATA[
          Mathematik ist eine Wissenschaft mit einer Struktur, die im Laufe der Zeit riesige Dimensionen erreicht hat. Diese unglaublich hohe Burg besitzt nur ein ganz schmales Fundament und ihre Festigkeit liegt an einfachen prädikatenlogischen Mörtel. Im Prinzip kann der Aufbau von jeder Mathematikerin verstanden werden. Von dem neuesten Gipfel mathematischer Erkenntnis kann jeder Pfad logisch folgerichtig bis in die mengentheoretischen Wurzeln nachvollzogen werden.
          
          \par
          Bei diesem Unternehmen will dieses Dokument Hilfestellung geben. Ziel ist eine Präsentation der mengentheoretischen Wurzeln in verständlicher Weise. Bei aller Verständlichkeit soll es jedoch jederzeit möglich sein, tief in die Details einzusteigen. Ja sogar bis auf die Ebene eines formal korrekten Beweises hinab. Dazu gibt es dieses Dokument in verschiedenen Detailierungen. Für alle aber gilt, dass die Formeln in Axiomen, Definitionen und Propositionen in formal korrekter Form vorliegen.

          \par
          Wir wollen bei den Wurzeln anfangen\ldots

          \par
          Dieses Dokument ist noch im Entstehen und wird von Zeit zu Zeit aktualisiert. Insbesondere werden an den durch {\glqq+++\grqq} gekennzeichneten Stellen noch Ergänzungen oder Änderungen vorgenommen.
          
          \par
          Besondere Dank geht an meine Frau \emph{Gesine~Dräger} und unseren Sohn \emph{Lennart} für ihre Unterstützung und ihr Verständnis für ihnen fehlende Zeit.

          \par
          \vspace*{1cm} Hamburg, Dezember 2005 \\
          \hspace*{\fill} Michael Meyling
        ]]>
      </LATEX>
          
    </INTRODUCTION>
  </CHAPTER>


  <CHAPTER noNumber="true">
    <TITLE>
      <LATEX language="en">
        Introduction\label{ch:introduction}
      </LATEX>
      <LATEX language="de">
        Einleitung\label{ch:introduction}
      </LATEX>
    </TITLE>
    <INTRODUCTION>
      <LATEX language="en">
        <![CDATA[
          \par
          +++ Translation missing
          
          This document contains the mathematical foundation of set theory. Goal is the presentation of elementary results which are needed within other mathematical disciplines. After the basics the {\emph boolean algebra of classes} is in focus. Next are some thoughts about \emph{relations} and \emph{functions}. An importent achievement is the definition of the \emph{natural~ numbers} and their fulfilment of the \emph{Peano axioms}. Also the word \emph{recursion} is discussed.
          
          \par
          Although the presentation is axiomatic the results shall match the mathematical usage. Therefore the set theoretic axiom system of \emph{A.~P.~Morse} and \emph{J.~L.~Kelley} (MK\index{MK}) was choosen.
        ]]>
      </LATEX>
      <LATEX language="de">
        <![CDATA[
          Am Anfang steht die Logik. Sie stellt das Rüstzeug zur Argumentation bereit. Sie hilft beim Gewinnen von neuen Aussagen aus bereits vorhandenen. Sie ist universell anwendbar, liefert aber selbst keine Grundaussagen über mathematische Objekte. Erst in der Mengenlehre wird durch die Axiome eine mathematische Struktur definiert. Besonders interessant ist die Mengenlehre dadurch, dass sie zum einen von eigentlich allen mathematischen Disziplinen verwendet wird. Zum anderen lässt sich jede mathematische Disziplin innerhalb der Mengenlehre definieren. Zahlentheorie, Algebra, Analysis und alle weiteren Gebiete lassen sich darauf aufbauen.
        
          \par
          Dieses Dokument beschreibt die mathematischen Grundlagen der Mengenlehre. Ziel ist dabei die Bereitstellung von elementaren Ergebnissen der Mengenlehre, die in anderen mathematischen Disziplinen benötigt werden. Nach den Grundlagen wird die Boolsche Algebra der Klassen betrachtet. Es schliessen sich Betrachtungen über Relationen und Funktionen an. Ein weiteres wichtiges Ergebnis sind die Definition der natürlichen Zahlen und die Erfüllung der Peano-Axiome durch diese, auch auf den Begriff der Rekursion wird eingegangen.
         
          \par
          Die Darstellung erfolgt in axiomatischer Weise soll aber im Ergebnis der mathematischen Praxis entsprechen. Daher wird auch das Axiomensystem der Mengenlehre von \emph{A.~P.~Morse} und  \emph{J.~L.~Kelley} (MK\index{MK}) verwendet.
         
        ]]>
      </LATEX>
    </INTRODUCTION>
  </CHAPTER>

  <CHAPTER>
    <TITLE>
      <LATEX language="en">
        Basics
      </LATEX>
      <LATEX language="de">
        Anfangsgründe
      </LATEX>
    </TITLE>
    <INTRODUCTION>
      <LATEX language="en">
        <![CDATA[
          In this chapter we start with the very basic axioms and definitions of set theory. We shall make no attempt to introduce a formal language but shall be content with the common logical operators. To be more precise: precondition is a first-order predicate calculus with identity.
          
          \par\index{Cantor}\index{set!definition}
          \emph{G.~Cantor}, who is considered the founder of set theory, gave in a publication in 1895 a description of the term \emph{set}.
          
          \begin{quote}
           By a ``set'' we are to understand any collection into a whole $M$ of definite and separate objects $m$ of our intuition or our thought. These objects are called the ``elements'' of $M$.
          \end{quote}
          
          \par
          This collection can be specified by giving a \emph{condition for memberschip}. Around 1900 various paradoxes in this naive set theory were discovered. These paradoxes base on giving tricky conditions for membership.
          
          \par
          There exist different ways out of those contradictions. In this text we don't restrict the condition for membership but we call the result \emph{class}. Additional axioms allow us to call certain classes sets again. All sets are classes, but not all classes are sets. Sets are classes which are themselves members of classes, whilst a class which is not a set is a class which is not a member of any class.
          
        ]]>
      </LATEX>
      <LATEX language="de">
        <![CDATA[
          In diesem Kapitel beginnen wir mit den ganz elementaren Axiomen und Definitionen der Mengenlehre. Wir versuchen nicht eine formale Sprache einzuführen und setzen das Wissen um den Gebrauch von logischen Symbolen voraus. Noch genauer formuliert: wir arbeiten mit einer Prädikatenlogik erster Stufe mit Gleichheit.

          \par\index{Cantor}\index{Menge!Definition}
          \emph{G.~Cantor}, der als Begründer der Mengenlehre gilt, hat in einer Veröffentlichung im Jahre 1895 eine Beschreibung des Begriffs \emph{Menge} gegeben.
          
          \begin{quote}
           Unter einer {\glqq Menge\grqq} verstehen wir jede Zusammenfassung $M$ von bestimmten wohlunterscheidbaren Objekten $m$ unserer Anschauung oder unseres Denkens (welche die {\glqq Elemente\grqq} von $M$ genannt werden) zu einem Ganzen.
          \end{quote}
          
          \par
          Diese Zusammenfassung kann über die Angabe einer Eigenschaft dieser Elemente erfolgen. Um 1900 wurden verschiedene Widersprüche dieser naiven Mengenlehre entdeckt. Diese Widersprüche lassen sich auf trickreich gewählte Eigenschaften zurückführen.
          
          \par
          Es gibt verschiedene Möglichkeiten diese Widersprüche zu verhindern. In diesem Text schränken wir zwar die Angabe von Eigenschaften in keiner Weise ein, aber wir nennen das Ergebnis der Zusammenfassung zunächst einmal \emph{Klasse}. Zusätzliche Axiome regeln dann, wann eine bestimmte Klasse auch eine Menge ist. Alle Mengen sind Klassen, aber nicht alle Klassen sind Mengen. Eine Menge ist eine Klasse, die selbst Element einer anderen Klasse ist. Eine Klasse, die keine Menge ist, ist nicht Element irgend einer anderen Klasse.
        ]]>
      </LATEX>
    </INTRODUCTION>
    <SECTION>
      <TITLE>
        <LATEX language="en">Classes and Sets</LATEX>
        <LATEX language="de">Klassen und Mengen</LATEX>
      </TITLE>
      <INTRODUCTION>
        <LATEX language="en">
          <![CDATA[
            Although we want to speak about \emph{sets} at the very beginning we have \emph{classes}. No formal definition of a class will be given. Informally, a class is a collection of objects, the involved objects are called the elements or members of the class. 
            Sets will be construed as a special kind of class. 
            The following definitions and axioms are due to a strengthened version of \emph{von~Neumann-Bernays-Gödel's} set theory (\emph{NBG}\index{NBG}). This version is called \emph{MK}\index{MK} which is short for \emph{Morse-Kelley}. 
          ]]>
        </LATEX>
        <LATEX language="de">
          <![CDATA[
            Obgleich wir im Wesentlichen über \emph{Mengen} sprechen wollen, haben wir am Anfang nur \emph{Klassen}. Dieser Begriff wird nicht formal definiert. Anschaulich gesprochen, ist eine Klasse eine Zusammenfassung von Objekten. Die beteiligten Objekte heissen auch Elemente der Klasse.
            Mengen werden dann als eine besondere Art von Klassen charakterisiert. 
            Die folgenden Definitionen und Axiome folgen dem Aufbau einer vereinfachten Version der Mengenlehre nach \emph{von~Neumann-Bernays-Gödel} (\emph{NBG}\index{NBG}).. Diese Version wird auch \emph{MK}\index{MK} nach \emph{Morse-Kelley} genannt.
          ]]>
        </LATEX>
      </INTRODUCTION>
      <SUBSECTIONS>
        <NODE id="in">
          <NAME>
            <LATEX language="en">is a member of</LATEX>
            <LATEX language="de">ist enthalten in</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Membership Operator</LATEX>
            <LATEX language="de">Elementbeziehung</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                The theory of sets introduced here has initial objects, called \emph{classes}. Furthermore the only predefined symbol is for a binary relation called \emph{membership}.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Die hier vorgestellte Mengenlehre hat als Ausgangsobjekte \emph{Klassen}.
                Weiterhin wird nur ein einziges Symbol für eine binäre Relation vorausgesetzt: der \emph{Enthaltenseinoperator}.
              ]]>
            </LATEX>
          </PRECEDING>
          <DEFINITION kind="PREDCON" arguments="2">
            <LATEXPATTERN>#1 \in #2</LATEXPATTERN>
            <VARLIST>
              <VAR id="x"/>
              <VAR id="y"/>
            </VARLIST>
          </DEFINITION>
          <SUCCEEDING>
            <LATEX language="en" level="2">
              <![CDATA[
                We also say $x$ \emph{is element of} $y$, $x$ \emph{belongs to} $y$, $x$ \emph{is a member of} $y$, $x$ \emph{is in} $y$.
                Beside identity this is the only predicate we start with. All other will be defined.\footnote{One could also define the identity predicate within the set theory, but then another axiom is needed and the theory presentation is not so smooth for technical reasons (derivation of the equality axioms).} Also no function constants are initially given.
              ]]>
            </LATEX>
            <LATEX language="de" level="2">
              <![CDATA[
                Wir sagen auch $x$ \emph{ist Element von} $y$, $x$ \emph{gehört zu} $y$, $x$ \emph{liegt in} $y$, $x$ \emph{ist in} $y$.
                Neben der Gleichheit ist dies das einzige Prädikat welches wir zu Beginn haben. Alle anderen werden definiert.\footnote{Das Gleichheitsprädikat könnte auch innerhalb der Mengenlehre definiert werden, aber dann wird auch ein weiters Axiom benötigt und es ergeben sich technischen Schwierigkeiten bei der Herleitung der Gleichheitsaxiome.} Zu Anfang haben wir auch noch keine Funktionskonstanten.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>
        <NODE id="notIn">
          <NAME>
            <LATEX language="en">is not a member of</LATEX>
            <LATEX language="de">ist nicht enthalten in</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Non Membership Operator</LATEX>
            <LATEX language="de">Negation der Elementbeziehung</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                Although we simply can negate the membership predicate we also want to define a shorthand notation for it.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Obgleich wir die Elementbeziehung einfach negieren können, möchten wir dafür eine Abkürzung definieren.
              ]]>
            </LATEX>
          </PRECEDING>
          <DEFINITION kind="PREDCON" arguments="2">
            <LATEXPATTERN>#1 \notin #2</LATEXPATTERN>
            <VARLIST>
              <VAR id="x"/>
              <VAR id="y"/>
            </VARLIST>
            <FORMULA>
              <NOT>
                <PREDCON ref="in">
                  <VAR id="x"/>
                  <VAR id="y"/>
                </PREDCON>
              </NOT>
            </FORMULA>
          </DEFINITION>
        </NODE>
        <NODE id="axiom:extensionality">
          <NAME>
            <LATEX language="en">axiom of extensionality</LATEX>
            <LATEX language="de">Extensionalitätsaxiom</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Extensionality\index{axiom!of extensionality}</LATEX>
            <LATEX language="de">Extensionalität\index{Extensionalitätsaxiom}</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                 Our first axiom states that, for any classes $x$ and $y$, if the membership of $x$ and $y$ are the same, then $x$ and $y$ are the same.\footnote{If identity were not part of our underlying logic, then we should need to take this as a definition of identity.}
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Unser erstes Axiom besagt, dass beliebige Klassen $x$ und $y$ genau dann gleich sind, wenn sie dieselben Elemente enthalten.\footnote{Falls wir das Gleichheitsprädikat nicht als logisches Symbol voraussetzen würden, dann würden wir es hiermit definieren.}
              ]]>
            </LATEX>
          </PRECEDING>
          <AXIOM>
            <FORMULA>
              <IMPL>
                <FORALL>
                  <VAR id="z"/>
                  <EQUI>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="x"/>
                    </PREDCON>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="y"/>
                    </PREDCON>
                  </EQUI>
                </FORALL>
                <PREDCON ref="equal">
                  <VAR id="x"/>
                  <VAR id="y"/>
                </PREDCON>
              </IMPL>
            </FORMULA>
          </AXIOM>
          <SUCCEEDING>
            <LATEX language="en">
              <![CDATA[
                The classes $x$ and $y$ may be defined in entirely different ways, for example:
                \par
                \begin{tabularx}{\linewidth}{rcX}
                  $x$ & = & class of all nonnegative integers, \\
                  $y$ & = & class of all integers, that can be written as sum of four squares,
                \end{tabularx}
                \par
                but if they have the same members, they are the same class.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Die Klassen $x$ and $y$ können verschieden definiert sein, beispielsweise:
                \par
                \begin{tabularx}{\linewidth}{rcX}
                  $x$ & = & Klasse aller nichtnegativen ganzen Zahlen, \\
                  $y$ & = & Klasse aller ganzen Zahlen, die als Summe von vier Quadraten geschreiben werden können,
                \end{tabularx}
                \par
                aber wenn sie dieselben Elemente besitzen, sind sie gleich.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>
        <NODE id="isSet">
          <NAME>
            <LATEX language="en">is set</LATEX>
            <LATEX language="de">ist Menge</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Set\index{set!definition}</LATEX>
            <LATEX language="de">Menge\index{Menge!Definition}</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              Now we specify \emph{sets}.
            </LATEX>
            <LATEX language="de">
              Jetzt legen wir fest, was eine \emph{Menge} ist.
            </LATEX>
          </PRECEDING>
          <DEFINITION kind="PREDCON" arguments="1">
            <LATEXPATTERN>\mathfrak{M}(#1)</LATEXPATTERN>
            <VARLIST>
              <VAR id="x"/>
            </VARLIST>
            <FORMULA>
              <EXISTS>
                <VAR id="y"/>
                <PREDCON ref="in">
                  <VAR id="x"/>
                  <VAR id="y"/>
                </PREDCON>
              </EXISTS>
            </FORMULA>
          </DEFINITION>
          <SUCCEEDING>
            <LATEX language="en">
              <![CDATA[
                So sets are nothing else than special classes. A class is a set iff it is a member of any class.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Mengen sind also nichts anderes, als Klassen mit einer besonderen Eigenschaft. Eine Klasse ist genau dann eine Menge, wenn sie Element irgendeiner Klasse ist.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>
        <NODE id="theorem:extensionalitySetRestricted" level="formal">
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                As a consequence of the axiom of extensionality we have the following.\footnote{The quantification over $z$ is restricted to sets.}
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Als erste Folgerung aus dem Extensionalitätsaxiom erhalten wir das Folgende.\footnote{Es wird ein eingeschränkter Allquantor benutzt, $z$ läuft nur über Mengen.}
              ]]>
            </LATEX>
          </PRECEDING>
          <THEOREM>
            <FORMULA>
              <IMPL>
                <FORALL>
                  <VAR id="z"/>
                  <PREDCON ref="isSet">
                    <VAR id="z"/>
                  </PREDCON>
                  <EQUI>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="x"/>
                    </PREDCON>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="y"/>
                    </PREDCON>
                  </EQUI>
                </FORALL>
                <PREDCON ref="equal">
                  <VAR id="x"/>
                  <VAR id="y"/>
                </PREDCON>
              </IMPL>
            </FORMULA>
            <PROOF kind="informal" level="1">
              <LATEX language="en">
                <![CDATA[
                  Assume $\forall \ \mathfrak{M}(z) \ ( z \in x \ \leftrightarrow \ z \in y)$. Let $z$ be an arbitrary class. If $z \in x$ then $z$ is a set by definition~\ref{isSet}, and hence by the assumption, $z \in y$. Similarly $z \in y \ \rightarrow \ z \in x$. Since $z$ is arbitrary, it follows that $\forall z \ (z \in x \ \leftrightarrow \ z \in y)$. Thus by    the axiom of extensionality~\ref{axiom:extensionality}, $x = y$. 
                ]]>
              </LATEX>
              <LATEX language="de">
                <![CDATA[
                  Angenommen es gelte $\forall \ \mathfrak{M}(z) \ ( z \in x \ \leftrightarrow \ z \in y)$. Sei $z$ eine beliebige Klasse. Falls $z \in x$ dann gilt $z$ ist eine Menge nach Definition~\ref{isSet}, und daraus folgt mit der Annahme $z \in y$. Analog folgt $z \in y \ \rightarrow \ z \in x$. Da $z$ beliebig, haben wir $\forall z \ (z \in x \ \leftrightarrow \ z \in y)$. Und mit dem  Extensionalitätsaxiom~\ref{axiom:extensionality} erhalten wir daraus $x = y$. 
                ]]>
              </LATEX>
            </PROOF>
          </THEOREM>
        </NODE>

        <NODE id="theorem:extensonalityEquivalence" level="formal">
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                We can also reverse the implication in the axiom of extensionality.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Weiterhin können wir in dem Extensionalitätsaxiom die Implikation umkehren.
              ]]>
            </LATEX>
          </PRECEDING>
          <THEOREM>
            <FORMULA>
              <EQUI>
                <PREDCON ref="equal">
                  <VAR id="x"/>
                  <VAR id="y"/>
                </PREDCON>
                <FORALL>
                  <VAR id="z"/>
                  <EQUI>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="x"/>
                    </PREDCON>
                    <PREDCON ref="in">
                      <VAR id="z"/>
                      <VAR id="y"/>
                    </PREDCON>
                  </EQUI>
                </FORALL>
              </EQUI>
            </FORMULA>
            <PROOF kind="informal" level="1">
              <LATEX language="en">
                 This a simple consequence of the second identity axiom.
              </LATEX>
              <LATEX language="de">
                 Dies ist eine einfache Anwendung des zweiten identitätslogischen Axioms.
              </LATEX>
            </PROOF>
          </THEOREM>
        </NODE>

        <NODE id="axiom:comprehension">
          <NAME>
            <LATEX language="en">axiom of comprehension</LATEX>
            <LATEX language="de">Komprehensionsaxiom</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Comprehension</LATEX>
            <LATEX language="de">Komprehension</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                Our second axiom makes it simple to create new classes. A class could be characterized by a predicate calculus formula.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Unser zweites Axiom ermöglicht eine einfache Möglichkeit Klassen zu bilden. Eine Klasse wird ganz einfach durch die Angabe einer prädikatenlogischen Formel charakterisiert. 
              ]]>
            </LATEX>
          </PRECEDING>
          <AXIOM>
            <FORMULA>
              <EXISTS>
                <VAR id="x"/>
                <FORALL>
                  <VAR id="y"/>
                  <EQUI>
                    <PREDCON ref="in">
                      <VAR id="x"/>
                      <VAR id="y"/>
                    </PREDCON>
                    <AND>
                      <PREDCON ref="isSet">
                        <VAR id="y"/>
                      </PREDCON>
                      <PREDVAR id="\phi">
                        <VAR id="y"/>
                      </PREDVAR>
                    </AND>
                  </EQUI>
                </FORALL>
              </EXISTS>
            </FORMULA>
          </AXIOM>
          <SUCCEEDING>
            <LATEX language="en">
              <![CDATA[
                \index{NBG}By modifying the above axiom we could get a \emph{NBG} axiom system of set theory due to \emph{John von Neumann}, \emph{Isaak Bernays} and \emph{Kurt Gödel}.
                
                \par\index{predicative}\index{Formula, predicative}
                For that purpose we call a formula \emph{predicative}, if all of its bound subject variables are restricted to sets.
                Therefore predicative formulas formalize `set properties'.\footnote{A little bit more formal: within a predicative formula all quantifier variables run over sets: $\forall \ \mathfrak{M}(x) \ \exists \ \mathfrak{M}(y) \ldots$}
                So if we postulate $\phi$ to be predicative we achieve a NBG system in the following.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                \index{NBG}Durch eine Änderung dieses Axioms würden wir im Folgenden ein \emph{NBG}-Axiomensystem der Mengenlehre erhalten, welches auf \emph{John von Neumann}, \emph{Isaak Bernays} und \emph{Kurt Gödel} zurückgeht.
                
                \par\index{prädikativ}\index{Formel, prädikative}
                Dazu definieren wir: eine Formel, in der alle gebundenen Subjektvariablen auf Mengen restringiert sind wird \emph{prädikative Formel} genannt. Prädikative Formeln formalisieren also diejenigen Eigenschaften, die man als {\glqq Eigenschaften von Mengen\grqq} bezeichnen kann.\footnote{Noch etwas formaler: in einer prädikativen Formel laufen alle Quantorenvariablen nur über Mengen: $\forall \ \mathfrak{M}(x) \ \exists \ \mathfrak{M}(y) \ldots$}
                Fordern wir nun also zusätzlich, dass $\phi$ prädikativ sein muss, dann erhalten wir im Folgenden ein NBG-System.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>

        <NODE id="theorem:comprehension" level="formal">
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                By the comprehension axiom and the axiom of extensionality a relationship between a sentence $\phi(y)$ and the class defined by this sentence is described. The comprehension axiom proposes the existence of at least one class, where the sentence $\mathfrak{M}(y) \land \phi(y)$ holds for all elements. The axiom of extensionality and the identity axioms make sure, that there is maximal one class of this kind: two classes with the same elements are equal in the sense of replacement in all appropriate sentences. In other words: there is one and only one such class.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Durch das Komprehensionsaxiom und die Extensionalität wird nun der Zusammenhang zwischen einer Aussage $\phi(y)$ und der durch sie definierten Klasse festgelegt. Dabei behaupted das Komprehensionsaxiom die Existenz mindestens einer Klasse, deren Elemente die Aussage $\mathfrak{M}(y) \land \phi(y)$ erfüllen. Das Extensionalitätsaxiom und die Gleichheitsaxiome sichern ab, dass es höchstens eine solche Klasse gibt: irgend zwei Klassen, welche dieselben Elemente besitzen, sind gleich im Sinne der Ersetzbarkeit in allen einschlägigen Aussagen. Mit anderen Worten: es gibt nur genau eine solche Klasse.
              ]]>
            </LATEX>
          </PRECEDING>
          <THEOREM>
            <FORMULA>
              <EXISTSU>
                <VAR id="x"/>
                <FORALL>
                  <VAR id="y"/>
                  <EQUI>
                    <PREDCON ref="in">
                      <VAR id="x"/>
                      <VAR id="y"/>
                    </PREDCON>
                    <AND>
                      <PREDCON ref="isSet">
                        <VAR id="y"/>
                      </PREDCON>
                      <PREDVAR id="\phi">
                        <VAR id="y"/>
                      </PREDVAR>
                    </AND>
                  </EQUI>
                </FORALL>
              </EXISTSU>
            </FORMULA>
            <PROOF kind="informal" level="1">
              <LATEX language="en">
                <![CDATA[
                  We must show:
                  $$
                  \begin{array}{rl}
                  \exists x                      & \forall y \ (y \in x \leftrightarrow  \ \mathfrak{M}(y) \land \phi(y)) \\
                  \land \ \forall u \ \forall v  & (\forall y \ (y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \ \land  \ \forall y \ ( y \in v \leftrightarrow \ \mathfrak{M}(y) \land \phi(y))) \\
                         & \qquad \rightarrow u = v)
                  \end{array}
                  $$
                  Let $u$ and $v$ be arbitrary. Furthermore let us assume:

                  \par
                  $\forall y \ (y \in u \leftrightarrow \ \mathfrak{M}(y) \land \phi(y)) \land \ \forall y \ ( y \in v \leftrightarrow \mathfrak{M}(y)
                  \land \phi(y)))$

                  \par
                  Now with \ref{allandpp}: $\forall y \ ((y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \land (y \in v \leftrightarrow \mathfrak{M}(y) \land \phi(y)))$

                  \par
                  With \ref{andequi} we derive: $\forall y \ ((y \in u \leftrightarrow y \in v ))$. And with Proposition~\ref{theorem:extensonalityEquivalence} follows $u = v$. So we have shown:

                  \par
                  $\forall u \ \forall v \ (\forall y \ (y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \land \ \forall y \ (y \in v \leftrightarrow \mathfrak{M}(y) \land \phi(y))) \rightarrow u = v)$

                  \par
                  Together with axiom~\ref{axiom:comprehension} we get the proposition.
                ]]>
              </LATEX>
              <LATEX language="de">
                <![CDATA[
                  Zu zeigen ist:
                  $$
                  \begin{array}{rl}
                  \exists x                      & \forall y \ (y \in x \leftrightarrow  \ \mathfrak{M}(y) \land \phi(y)) \\
                  \land \ \forall u \ \forall v  & (\forall y \ (y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \ \land  \ \forall y \ ( y \in v \leftrightarrow \ \mathfrak{M}(y) \land \phi(y))) \\
                         & \qquad \rightarrow u = v)
                  \end{array}
                  $$
                  Seien $u$ und $v$ beliebig. Es gelte weiterhin:

                  \par
                  $\forall y \ (y \in u \leftrightarrow \ \mathfrak{M}(y) \land \phi(y)) \land \ \forall y \ ( y \in v \leftrightarrow \mathfrak{M}(y)
                  \land \phi(y)))$

                  \par
                  Dann folgt mit \ref{allandpp}: $\forall y \ ((y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \land (y \in v \leftrightarrow \mathfrak{M}(y) \land \phi(y)))$

                  \par
                  Daraus erhalten wir mit \ref{andequi}: $\forall y \ ((y \in u \leftrightarrow y \in v ))$. Und mit Proposition~\ref{theorem:extensonalityEquivalence} folgt nun $u = v$. Also haben wir gezeigt:

                  \par
                  $\forall u \ \forall v \ (\forall y \ (y \in u \leftrightarrow \mathfrak{M}(y) \land \phi(y)) \land \ \forall y \ (y \in v \leftrightarrow \mathfrak{M}(y) \land \phi(y))) \rightarrow u = v)$

                  \par
                  Zusammen mit Axiom~\ref{axiom:comprehension} folgt nun die Behauptung.
                ]]>
              </LATEX>
            </PROOF>
          </THEOREM>
        </NODE>


        <NODE id="rule:classDefinition">
          <NAME>
            <LATEX language="en">class definition</LATEX>
            <LATEX language="de">Klassenschreibweise</LATEX>
          </NAME>
          <TITLE>
            <LATEX language="en">Class definition</LATEX>
            <LATEX language="de">Klassenschreibweise</LATEX>
          </TITLE>
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                Starting with \ref{theorem:comprehension} we can extend the syntax and provide a new abbreviation.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Ausgehend von \ref{theorem:comprehension} können wir die Sprachsyntax erweitern und eine neue abkürzende Schreibweise einführen.
              ]]>
            </LATEX>
          </PRECEDING>
          <RULE name="SET_DEFINION_BY_FORMULA">
            <LINK id="theorem:comprehension"/>
            <DESCRIPTION>
              <LATEX language="en">
                <![CDATA[
                  For every formula $\alpha(x)$ we define the term expression $\{ x \ | \ \alpha(x)\}$ by
                  $$
                    \exists y \ (y = \{x \ | \ \alpha(x) \} \ \land \ \forall x \ x \in y \ \leftrightarrow \mathfrak{M}(x) \ \land \alpha(x))
                  $$
                  
                  \par
                  The free variables from $\{ x \ | \ \alpha(x)\}$ are the free variables from $\alpha(x)$ without $\{ x \}$. The bound variables correspond to the bound variables of $\alpha(x)$. All deduction rules are accordingly extended.
                  
                ]]>
              </LATEX>
              <LATEX language="de">
                <![CDATA[
                  Für jede Formel $\alpha(x)$ definieren wir den Termausdruck $\{ x \ | \ \alpha(x)\}$ durch 
                  $$
                    \exists y \ (y = \{x \ | \ \alpha(x) \} \ \land \ \forall x \ x \in y \ \leftrightarrow \mathfrak{M}(x) \ \land \alpha(x))
                  $$
                  
                  \par
                  Die freien Variablen von $\{ x \ | \ \alpha(x)\}$ sind die freien Variablen von $\alpha(x)$ vermindert um $\{ x \}$. Die gebunden Variablen entsprechen den gebunden Variablen von $\alpha(x)$. Alle Ableitungsregln werden entsprechend erweitert.
                
%old                
%                  Der Ausdruck $\{ x | \alpha(x)\}$ bezeichnet ebenfalls einen Term. Sei $\beta(y)$ eine Formel und $beta(\{x | \alpha(x)\}$ die Formel, die bei Ersetzung von y durch den angegebenen Term entstehende Formel. Diese Formel kann ersetzt werden durch $\beta(y) \ \land \ \forall x \ x \in y \leftrightarrow \mathfrak{M}(x) \ \land \ \alpha(y)$. 
%                  \par
%                  In dem Term $\{ x | \alpha(x)\}$ gilt $x$ als gebunden und der Term besitzt zusätzlich die gebunden Variablen von $\alpha$ als gebundene Variablen. Die freien Variablen von $\alpha$ sind auch die freien Variablen des neuen Termausdrucks.
                ]]>
              </LATEX>
            </DESCRIPTION>
          </RULE>
          <SUCCEEDING>
            <LATEX language="en">
              <![CDATA[
                In particular the substitution rule must be adapted. This is a \emph{conservative} extension.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                 Insbesondere muss die Substitutionsregel angepasst werden. Es handelt sich hierbei um eine \emph{konservative} Erweiterung.

%old                
%                { x | alpha (x)} = { x | alpha (x)}
%                y = y n All x x in y <-> M(x) n alpha(x)
                
                Im Folgenden wird auf diese Schreibweise zurückgegriffen.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>

        <NODE id="theorem:setNotation" level="1">
          <PRECEDING>
            <LATEX language="en">
              <![CDATA[
                This new notation can be easily transformed in the old syntax. Using the new term type with the initial predicates can be expressed as follows.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Die neue Schreibweise kann auch in einfacher Weise in die alte Syntax transformiert werden.
                Die Gültigkeit der Ausgangsprädikate drückt sich für diese neue Termart wie folgt aus.
              ]]>
            </LATEX>
          </PRECEDING>
          <THEOREM>
            <FORMULA>
              <AND>
                <EQUI>
                  <PREDCON ref="in">
                    <VAR id="y"/>
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\phi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                  </PREDCON>
                  <AND>
                    <PREDCON ref="isSet">
                      <VAR id="y"/>
                    </PREDCON>
                    <PREDVAR id="\phi">
                      <VAR id="y"/>
                    </PREDVAR>
                  </AND>
                </EQUI>
                <EQUI>
                  <PREDCON ref="equal">
                    <VAR id="y"/>
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\phi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                  </PREDCON>
                  <FORALL>
                    <VAR id="z"/>
                    <EQUI>
                      <PREDCON ref="in">
                        <VAR id="z"/>
                        <VAR id="y"/>
                      </PREDCON>
                      <PREDCON ref="in">
                        <VAR id="z"/>
                        <SET>
                          <VAR id="x"/>
                          <PREDVAR id="\phi">
                            <VAR id="x"/>
                          </PREDVAR>
                        </SET>
                      </PREDCON>
                    </EQUI>
                  </FORALL>
                </EQUI>
                <EQUI>
                  <PREDCON ref="equal">
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\phi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\psi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                  </PREDCON>
                  <FORALL>
                    <VAR id="z"/>
                    <EQUI>
                      <PREDCON ref="in">
                        <VAR id="z"/>
                        <SET>
                          <VAR id="x"/>
                          <PREDVAR id="\phi">
                            <VAR id="x"/>
                          </PREDVAR>
                        </SET>
                      </PREDCON>
                      <PREDCON ref="in">
                        <VAR id="z"/>
                        <SET>
                          <VAR id="x"/>
                          <PREDVAR id="\psi">
                            <VAR id="x"/>
                          </PREDVAR>
                        </SET>
                      </PREDCON>
                    </EQUI>
                  </FORALL>
                </EQUI>
                <EQUI>
                  <PREDCON ref="in">
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\phi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\psi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                  </PREDCON>
                  <FORALL>
                    <VAR id="u"/>
                    <FORALL>
                      <VAR id="v"/>
                      <IMPL>
                        <AND>
                          <PREDCON ref="equal">
                            <VAR id="u"/>
                            <SET>
                              <VAR id="x"/>
                              <PREDVAR id="\phi">
                                <VAR id="x"/>
                              </PREDVAR>
                            </SET>
                          </PREDCON>
                          <PREDCON ref="equal">
                            <VAR id="v"/>
                            <SET>
                              <VAR id="x"/>
                              <PREDVAR id="\psi">
                                <VAR id="x"/>
                              </PREDVAR>
                            </SET>
                          </PREDCON>
                        </AND>
                        <PREDCON ref="in">
                          <VAR id="u"/>
                          <VAR id="v"/>
                        </PREDCON>
                      </IMPL>
                    </FORALL>
                  </FORALL>
                </EQUI>
                <EQUI>
                  <PREDCON ref="in">
                    <SET>
                      <VAR id="x"/>
                      <PREDVAR id="\phi">
                        <VAR id="x"/>
                      </PREDVAR>
                    </SET>
                    <VAR id="y"/>
                  </PREDCON>
                  <FORALL>
                    <VAR id="u"/>
                    <IMPL>
                      <PREDCON ref="equal">
                        <VAR id="u"/>
                        <SET>
                          <VAR id="x"/>
                          <PREDVAR id="\phi">
                            <VAR id="x"/>
                          </PREDVAR>
                        </SET>
                      </PREDCON>
                      <PREDCON ref="in">
                        <VAR id="u"/>
                        <VAR id="v"/>
                      </PREDCON>
                    </IMPL>
                  </FORALL>
                </EQUI>
              </AND>
            </FORMULA>
            <DESCRIPTION>
              <LATEX language="en">
                <![CDATA[
                  +++ if the formula would be rendered it should look like this: \\
                  \begin{align}
                  y \in \{ x~|~\phi(x) \} & \leftrightarrow  \mathfrak{M}(y) \land \phi(y) \tag{a} \\
                  y = \{ x~|~ \phi(x) \} & \leftrightarrow  \forall z \ (z \in y \leftrightarrow z \in \{ x~|~\phi(x) \}) \tag{b} \\
                  \{ x~|~\phi(x) \} = \{ x~|~\psi(x) \} & \leftrightarrow \forall z \ (z \in \{ x~|~\phi(x) \} \tag{c} \\
                  \begin{split}
                    & \qquad \leftrightarrow z \in \{x~|~\psi(x) \}) \nonumber \\
                  \{ x~|~\phi(x) \} \in \{ x~|~\psi(x) \} & \leftrightarrow  \forall u \ \forall
                  v \ ((u  = \{ x~|~\phi(x) \} \\
                    & \qquad \land \ v = \{ x~|~\psi(x) \}) \rightarrow u \in v) 
                  \end{split} \tag{d} \\
                  \{ x~|~\phi(x) \} \in y & \leftrightarrow  \forall u \ (u  = \{ x~|~\phi(x) \}  \rightarrow u \in y) \tag{e} 
                  \end{align}
                ]]>
              </LATEX>
              <LATEX language="de">
                <![CDATA[
                  +++ wenn diese Formel richtig gesetzt würde, sollte sie so aussehen:
                  \begin{align}
                  y \in \{ x~|~\phi(x) \} & \leftrightarrow  \mathfrak{M}(y) \land \phi(y) \tag{a} \\
                  y = \{ x~|~ \phi(x) \} & \leftrightarrow  \forall z \ (z \in y \leftrightarrow z \in \{ x~|~\phi(x) \}) \tag{b} \\
                  \{ x~|~\phi(x) \} = \{ x~|~\psi(x) \} & \leftrightarrow \forall z \ (z \in \{ x~|~\phi(x) \} \tag{c} \\
                  \begin{split}
                    & \qquad \leftrightarrow z \in \{x~|~\psi(x) \}) \nonumber \\
                  \{ x~|~\phi(x) \} \in \{ x~|~\psi(x) \} & \leftrightarrow  \forall u \ \forall
                  v \ ((u  = \{ x~|~\phi(x) \} \\
                    & \qquad \land \ v = \{ x~|~\psi(x) \}) \rightarrow u \in v) 
                  \end{split} \tag{d} \\
                  \{ x~|~\phi(x) \} \in y & \leftrightarrow  \forall u \ (u  = \{ x~|~\phi(x) \}  \rightarrow u \in y) \tag{e} 
                  \end{align}
                ]]>
              </LATEX>
            </DESCRIPTION>
            <PROOF kind="informal" level="1">
              <LATEX language="en">
                <![CDATA[
                  +++ missing.
                ]]>
              </LATEX>
              <LATEX language="de">
                <![CDATA[
                  +++ fehlt noch.
                ]]>
              </LATEX>
            </PROOF>
          </THEOREM>
          <SUCCEEDING>
            <LATEX language="en">
              <![CDATA[
                Therefore the new syntax can be eliminated by successive applying the above theorems.
              ]]>
            </LATEX>
            <LATEX language="de">
              <![CDATA[
                Durch sukzessives Anwenden des obigen Satzes kann also die neue Syntax in die alte überführt werden.
              ]]>
            </LATEX>
          </SUCCEEDING>
        </NODE>

      </SUBSECTIONS>
    </SECTION>
  </CHAPTER>
</QEDEQ>
