1: % processed by citealice (August 1999) on Tue Mar 14 11:57:42 CST 2000
2:
3: % To: Saharon Shelah <shelah@math.huji.ac.il>, Andrzej Roslanowski <roslanow@e-math.ams.org>
4: % Subject: Sh656
5: % Date: Tue, 14 Mar 2000 09:33:30 -0500 (EST)
6: % From: Alice Leonhardt <leonhard@math.rutgers.edu>
7: % Mime-Version: 1.0
8: % Content-Description: revisions
9: % X-sliced-and-diced-by: 'savemail' 0.3, Feb 1999
10:
11: \ifx\shlhetal\undefinedcontrolsequence\let\shlhetal\relax\fi
12:
13:
14: \input amstex
15: % % \input mathdefs
16: \expandafter\ifx\csname mathdefs.tex\endcsname\relax
17: \expandafter\gdef\csname mathdefs.tex\endcsname{}
18: \else \message{Hey! Apparently you were trying to
19: \string\input{mathdefs.tex} twice. This does not make sense.}
20: \errmessage{Please edit your file (probably \jobname.tex) and remove
21: any duplicate ``\string\input'' lines}\endinput\fi
22:
23:
24: %mathdefs.tex v1.3.2
25:
26: %%% Changes from v1.0: footnote macros, warning for duplicated tags,
27: %%% control sequences \( and \verbatimtags.
28: %%% From v1.2: \pretags, redefinition of \( using \ifinner, multi-part
29: %%% equation numbering, control sequences \[, \references, and
30: %%% \resetbracket.
31: %%% From v1.3: \rm in \lastpart; write root of multi-part tag to .tgs
32:
33: %See file texdefs.doc for documentation.
34:
35: \catcode`\X=12\catcode`\@=11
36:
37: %Minor control sequences:
38: \def\n@wcount{\alloc@0\count\countdef\insc@unt}
39: \def\n@wwrite{\alloc@7\write\chardef\sixt@@n}
40: \def\n@wread{\alloc@6\read\chardef\sixt@@n}
41: \def\r@s@t{\relax}\def\v@idline{\par}\def\@mputate#1/{#1}
42: \def\l@c@l#1X{\firstpart.#1}\def\gl@b@l#1X{#1}\def\t@d@l#1X{{}}
43:
44: %Creation of tag families and output of assignments and citations:
45: \def\crossrefs#1{\ifx\all#1\let\tr@ce=\all\else\def\tr@ce{#1,}\fi
46: \n@wwrite\cit@tionsout\openout\cit@tionsout=\jobname.cit
47: \write\cit@tionsout{\tr@ce}\expandafter\setfl@gs\tr@ce,}
48: \def\setfl@gs#1,{\def\@{#1}\ifx\@\empty\let\next=\relax
49: \else\let\next=\setfl@gs\expandafter\xdef
50: \csname#1tr@cetrue\endcsname{}\fi\next}
51: \def\m@ketag#1#2{\expandafter\n@wcount\csname#2tagno\endcsname
52: \csname#2tagno\endcsname=0\let\tail=\all\xdef\all{\tail#2,}
53: \ifx#1\l@c@l\let\tail=\r@s@t\xdef\r@s@t{\csname#2tagno\endcsname=0\tail}\fi
54: \expandafter\gdef\csname#2cite\endcsname##1{\expandafter
55: \ifx\csname#2tag##1\endcsname\relax?\else\csname#2tag##1\endcsname\fi
56: \expandafter\ifx\csname#2tr@cetrue\endcsname\relax\else
57: \write\cit@tionsout{#2tag ##1 cited on page \folio.}\fi}
58: \expandafter\gdef\csname#2page\endcsname##1{\expandafter
59: \ifx\csname#2page##1\endcsname\relax?\else\csname#2page##1\endcsname\fi
60: \expandafter\ifx\csname#2tr@cetrue\endcsname\relax\else
61: \write\cit@tionsout{#2tag ##1 cited on page \folio.}\fi}
62: \expandafter\gdef\csname#2tag\endcsname##1{\expandafter
63: \ifx\csname#2check##1\endcsname\relax
64: \expandafter\xdef\csname#2check##1\endcsname{}%
65: \else\immediate\write16{Warning: #2tag ##1 used more than once.}\fi
66: \multit@g{#1}{#2}##1/X%
67: \write\t@gsout{#2tag ##1 assigned number \csname#2tag##1\endcsname\space
68: on page \number\count0.}%
69: \csname#2tag##1\endcsname}}
70: \def\multit@g#1#2#3/#4X{\def\t@mp{#4}\ifx\t@mp\empty%
71: \global\advance\csname#2tagno\endcsname by 1
72: \expandafter\xdef\csname#2tag#3\endcsname
73: {#1\number\csname#2tagno\endcsnameX}%
74: \else\expandafter\ifx\csname#2last#3\endcsname\relax
75: \expandafter\n@wcount\csname#2last#3\endcsname
76: \global\advance\csname#2tagno\endcsname by 1
77: \expandafter\xdef\csname#2tag#3\endcsname
78: {#1\number\csname#2tagno\endcsnameX}
79: \write\t@gsout{#2tag #3 assigned number \csname#2tag#3\endcsname\space
80: on page \number\count0.}\fi
81: \global\advance\csname#2last#3\endcsname by 1
82: \def\t@mp{\expandafter\xdef\csname#2tag#3/}%
83: \expandafter\t@mp\@mputate#4\endcsname
84: {\csname#2tag#3\endcsname\lastpart{\csname#2last#3\endcsname}}\fi}
85: \def\t@gs#1{\def\all{}\m@ketag#1e\m@ketag#1s\m@ketag\t@d@l p
86: \m@ketag\gl@b@l r \n@wread\t@gsin
87: \openin\t@gsin=\jobname.tgs \re@der \closein\t@gsin
88: \n@wwrite\t@gsout\openout\t@gsout=\jobname.tgs }
89: \outer\def\localtags{\t@gs\l@c@l}
90: \outer\def\globaltags{\t@gs\gl@b@l}
91: \outer\def\newlocaltag#1{\m@ketag\l@c@l{#1}}
92: \outer\def\newglobaltag#1{\m@ketag\gl@b@l{#1}}
93:
94: %Reading in tag information:
95: \newif\ifpr@
96: \def\m@kecs #1tag #2 assigned number #3 on page #4.%
97: {\expandafter\gdef\csname#1tag#2\endcsname{#3}
98: \expandafter\gdef\csname#1page#2\endcsname{#4}
99: \ifpr@\expandafter\xdef\csname#1check#2\endcsname{}\fi}
100: \def\re@der{\ifeof\t@gsin\let\next=\relax\else
101: \read\t@gsin to\t@gline\ifx\t@gline\v@idline\else
102: \expandafter\m@kecs \t@gline\fi\let \next=\re@der\fi\next}
103: \def\pretags#1{\pr@true\pret@gs#1,,}
104: \def\pret@gs#1,{\def\@{#1}\ifx\@\empty\let\n@xtfile=\relax
105: \else\let\n@xtfile=\pret@gs \openin\t@gsin=#1.tgs \message{#1} \re@der
106: \closein\t@gsin\fi \n@xtfile}
107:
108: %Sections and subsections; local numbering:
109: \newcount\sectno\sectno=0\newcount\subsectno\subsectno=0
110: \newif\ifultr@local \def\ultralocal{\ultr@localtrue}
111: \def\firstpart{\number\sectno}
112: \def\lastpart#1{\ifcase#1 \or a\or b\or c\or d\or e\or f\or g\or h\or
113: i\or k\or l\or m\or n\or o\or p\or q\or r\or s\or t\or u\or v\or w\or
114: x\or y\or z \fi}
115: \def\closeup{\vskip-\bigskipamount}
116: \def\resetall{\global\advance\sectno by 1\subsectno=0
117: \gdef\firstpart{\number\sectno}\r@s@t}
118: \def\resetsub{\global\advance\subsectno by 1
119: \gdef\firstpart{\number\sectno.\number\subsectno}\r@s@t}
120: \def\newsection#1\par{\resetall\vskip0pt plus.3\vsize\penalty-250
121: \vskip0pt plus-.3\vsize\bigskip\bigskip
122: \message{#1}\leftline{\bf#1}\nobreak\bigskip}
123: \def\subsection#1\par{\ifultr@local\resetsub\fi
124: \vskip0pt plus.2\vsize\penalty-250\vskip0pt plus-.2\vsize
125: \bigskip\smallskip\message{#1}\leftline{\bf#1}\nobreak\medskip}
126:
127: %Verbatim tags:
128: \def\t@gsoff#1,{\def\@{#1}\ifx\@\empty\let\next=\relax\else\let\next=\t@gsoff
129: \def\@@{p}\ifx\@\@@\else
130: \expandafter\gdef\csname#1cite\endcsname##1{\zeigen{##1}}
131: \expandafter\gdef\csname#1page\endcsname##1{?}
132: \expandafter\gdef\csname#1tag\endcsname##1{\zeigen{##1}}\fi\fi\next}
133: \def\verbatimtags{\ifx\all\relax\else\expandafter\t@gsoff\all,\fi}
134: \def\zeigen#1{\hbox{$\langle$}#1\hbox{$\rangle$}}
135:
136:
137: %Equation numbering:
138: \def\(#1){\edef\dot@g{\ifmmode\ifinner(\hbox{\noexpand\etag{#1}})
139: \else\noexpand\eqno(\hbox{\noexpand\etag{#1}})\fi
140: \else(\noexpand\ecite{#1})\fi}\dot@g}
141:
142: %Reference numbering:
143: \newif\ifbr@ck
144: \def\eat#1{}
145: \def\[#1]{\br@cktrue[\br@cket#1'X]}
146: \def\br@cket#1'#2X{\def\temp{#2}\ifx\temp\empty\let\next\eat
147: \else\let\next\br@cket\fi
148: \ifbr@ck\br@ckfalse\br@ck@t#1,X\else\br@cktrue#1\fi\next#2X}
149: \def\br@ck@t#1,#2X{\def\temp{#2}\ifx\temp\empty\let\neext\eat
150: \else\let\neext\br@ck@t\def\temp{,}\fi
151: \def\teemp{#1}\ifx\teemp\empty\else\rcite{#1}\fi\temp\neext#2X}
152: \def\resetbr@cket{\gdef\[##1]{[\rtag{##1}]}}
153: \def\references{\resetbr@cket\newsection References\par}
154:
155: %Footnotes:
156: \newtoks\symb@ls\newtoks\s@mb@ls\newtoks\p@gelist\n@wcount\ftn@mber
157: \ftn@mber=1\newif\ifftn@mbers\ftn@mbersfalse\newif\ifbyp@ge\byp@gefalse
158: \def\defm@rk{\ifftn@mbers\n@mberm@rk\else\symb@lm@rk\fi}
159: \def\n@mberm@rk{\xdef\m@rk{{\the\ftn@mber}}%
160: \global\advance\ftn@mber by 1 }
161: \def\rot@te#1{\let\temp=#1\global#1=\expandafter\r@t@te\the\temp,X}
162: \def\r@t@te#1,#2X{{#2#1}\xdef\m@rk{{#1}}}
163: \def\b@@st#1{{$^{#1}$}}\def\str@p#1{#1}
164: \def\symb@lm@rk{\ifbyp@ge\rot@te\p@gelist\ifnum\expandafter\str@p\m@rk=1
165: \s@mb@ls=\symb@ls\fi\write\f@nsout{\number\count0}\fi \rot@te\s@mb@ls}
166: \def\byp@ge{\byp@getrue\n@wwrite\f@nsin\openin\f@nsin=\jobname.fns
167: \n@wcount\currentp@ge\currentp@ge=0\p@gelist={0}
168: \re@dfns\closein\f@nsin\rot@te\p@gelist
169: \n@wread\f@nsout\openout\f@nsout=\jobname.fns }
170: \def\m@kelist#1X#2{{#1,#2}}
171: \def\re@dfns{\ifeof\f@nsin\let\next=\relax\else\read\f@nsin to \f@nline
172: \ifx\f@nline\v@idline\else\let\t@mplist=\p@gelist
173: \ifnum\currentp@ge=\f@nline
174: \global\p@gelist=\expandafter\m@kelist\the\t@mplistX0
175: \else\currentp@ge=\f@nline
176: \global\p@gelist=\expandafter\m@kelist\the\t@mplistX1\fi\fi
177: \let\next=\re@dfns\fi\next}
178: \def\symbols#1{\symb@ls={#1}\s@mb@ls=\symb@ls}
179: \def\bigsymbol{\textstyle}
180: \symbols{\bigsymbol\ast,\dagger,\ddagger,\sharp,\flat,\natural,\star}
181: \def\ftnumbers{\ftn@mberstrue} \def\ftsymbols{\ftn@mbersfalse}
182: \def\paginal{\byp@ge} \def\resetftnumbers{\ftn@mber=1}
183: \def\ftnote#1{\defm@rk\expandafter\expandafter\expandafter\footnote
184: \expandafter\b@@st\m@rk{#1}}
185:
186: %Miscellaneous macros:
187: \long\def\jump#1\endjump{}
188: \def\ssum{\mathop{\lower .1em\hbox{$\textstyle\Sigma$}}\nolimits}
189: \def\down#1{_{{}_{\scriptstyle #1}}}
190: \def\qed{\nobreak\kern 1em \vrule height .5em width .5em depth 0em}
191: \def\newneq{\hbox{\rlap{\hbox to 1\wd9{\hss$=$\hss}}\raise .1em
192: \hbox to 1\wd9{\hss$\scriptscriptstyle/$\hss}}}
193: \def\subsetne{\setbox9 = \hbox{$\subset$}\mathrel{\hbox{\rlap
194: {\lower .4em \newneq}\raise .13em \hbox{$\subset$}}}}
195: \def\supsetne{\setbox9 = \hbox{$\subset$}\mathrel{\hbox{\rlap
196: {\lower .4em \newneq}\raise .13em \hbox{$\supset$}}}}
197:
198: %Blackboard bold:
199: \def\vbar{\mathchoice{\vrule height6.3ptdepth-.5ptwidth.8pt\kern-.8pt}
200: {\vrule height6.3ptdepth-.5ptwidth.8pt\kern-.8pt}
201: {\vrule height4.1ptdepth-.35ptwidth.6pt\kern-.6pt}
202: {\vrule height3.1ptdepth-.25ptwidth.5pt\kern-.5pt}}
203: \def\f@dge{\mathchoice{}{}{\mkern.5mu}{\mkern.8mu}}
204: \def\b@c#1#2{{\rm \mkern#2mu\vbar\mkern-#2mu#1}}
205: \def\b@b#1{{\rm I\mkern-3.5mu #1}}
206: \def\b@a#1#2{{\rm #1\mkern-#2mu\f@dge #1}}
207: \def\bb#1{{\count4=`#1 \advance\count4by-64 \ifcase\count4\or\b@a A{11.5}\or
208: \b@b B\or\b@c C{5}\or\b@b D\or\b@b E\or\b@b F \or\b@c G{5}\or\b@b H\or
209: \b@b I\or\b@c J{3}\or\b@b K\or\b@b L \or\b@b M\or\b@b N\or\b@c O{5} \or
210: \b@b P\or\b@c Q{5}\or\b@b R\or\b@a S{8}\or\b@a T{10.5}\or\b@c U{5}\or
211: \b@a V{12}\or\b@a W{16.5}\or\b@a X{11}\or\b@a Y{11.7}\or\b@a Z{7.5}\fi}}
212:
213: \catcode`\X=11 \catcode`\@=12
214:
215: % % \input citeadd
216: % citeadd -- a few additions for
217: % files from alice that were procesed with "citealice"
218:
219: \expandafter\ifx\csname citeadd.tex\endcsname\relax
220: \expandafter\gdef\csname citeadd.tex\endcsname{}
221: \else \message{Hey! Apparently you were trying to
222: \string\input{citeadd.tex} twice. This does not make sense.}
223: \errmessage{Please edit your file (probably \jobname.tex) and remove
224: any duplicate ``\string\input'' lines}\endinput\fi
225:
226:
227:
228:
229:
230: \def\sciteu{\sciteerror{undefined}}
231: \def\sciteuphantom{\complainaboutcitation{undefined}}
232:
233: \def\scitet{\sciteerror{ambiguous}}
234: \def\scitetphantom{\complainaboutcitation{ambiguous}}
235:
236:
237:
238:
239: \def\sciteerror#1#2{{\mathortextbf{\scite{#2}}}\complainaboutcitation{#1}{#2}}
240: \def\mathortextbf#1{\hbox{\bf #1}}
241: \def\complainaboutcitation#1#2{%
242: \vadjust{\line{\llap{---$\!\!>$ }\qquad scite$\{$#2$\}$ #1\hfil}}}
243:
244:
245:
246:
247:
248: \sectno=-1 % start with sect 0
249: \localtags
250: %\verbatimtags
251: \NoBlackBoxes
252: \define\mr{\medskip\roster}
253: \define\sn{\smallskip\noindent}
254: \define\mn{\medskip\noindent}
255: \define\bn{\bigskip\noindent}
256: \define\ub{\underbar}
257: \define\wilog{\text{without loss of generality}}
258: \define\ermn{\endroster\medskip\noindent}
259: \define\dbca{\dsize\bigcap}
260: \define\dbcu{\dsize\bigcup}
261: \define \nl{\newline}
262: \magnification=\magstep 1
263: \documentstyle {amsppt}
264: \pageheight{8.5truein}
265: \topmatter
266: \title{NNR Revisited \\
267: Sh656} \endtitle
268: \author {Saharon Shelah \thanks {\null\newline I would like to thank
269: Alice Leonhardt for the beautiful typing. \null\newline
270: First Typed - 97/June/17 \null\newline
271: Written - \S1 in 97/April/21-22} \null\newline
272: Latest Revision - 2000/Mar/14 \endthanks} \endauthor
273: % Previous Version - 99/Nov/11
274: \affil{Institute of Mathematics\\
275: The Hebrew University\\
276: Jerusalem, Israel
277: \medskip
278: Rutgers University\\
279: Mathematics Department\\
280: New Brunswick, NJ USA} \endaffil
281: \bn
282: \abstract We are interested in proving that if we use CS iterations of
283: forcing notions not adding reals that satisfies
284: additional conditions then the limit
285: forcing does not add reals. As a result we prove that we can amalgamate two
286: earlier methods and prove the consistency with ZFC + G.C.H. of two
287: statements gotten separately earlier: SH and non-club guessing. We also
288: prove the consistency of further cases of ``strong failure of club
289: guessing". \endabstract
290: \endtopmatter
291: \document
292: % % \input alice2jlem
293: %% # Keywords Input file to be used for texing Alice's files
294:
295: \expandafter\ifx\csname alice2jlem.tex\endcsname\relax
296: \expandafter\gdef\csname alice2jlem.tex\endcsname{}
297: \else \message{Hey! Apparently you were trying to
298: \string\input{alice2jlem.tex} twice. This does not make sense.}
299: \errmessage{Please edit your file (probably \jobname.tex) and remove
300: any duplicate ``\string\input'' lines}\endinput\fi
301:
302:
303:
304: % % \input bib4plain
305: \expandafter\ifx\csname bib4plain.tex\endcsname\relax
306: \expandafter\gdef\csname bib4plain.tex\endcsname{}
307: \else \message{Hey! Apparently you were trying to \string\input
308: bib4plain.tex twice. This does not make sense.}
309: \errmessage{Please edit your file (probably \jobname.tex) and remove
310: any duplicate ``\string\input'' lines}\endinput\fi
311:
312:
313: % This file should be inputted if you want to use
314: % bibtex fom within plain TeX.
315: \def\makeatletter{\catcode`\@11 } % Not really need for standard
316: \def\makeatother{\catcode`\@12 } % bibtex files, but these commands
317: \def\renewcommand{\newcommand} % are used in our literal-unsrt.bst
318: % and other files.
319: % The above lines were written by Martin.Goldstern@tuwien.ac.at
320: % The rest is just the file btxmac.tex.
321: % (except for one small change below, regarding undefined citations)
322:
323: %% @texfile{
324: %% author = "Karl Berry and Oren Patashnik",
325: %% version = "0.99j",
326: %% date = "14 Mar 1992",
327: %% filename = "btxmac.tex",
328: %% address = "Please use electronic mail",
329: %% checksum = "834 4503 33061",
330: %% email = "opbibtex@cs.stanford.edu",
331: %% codetable = "ISO/ASCII",
332: %% supported = "yes",
333: %% docstring = "Defines macros that make BibTeX work with plain TeX",
334: %% }
335: % BibTeX-for-TeX macros, version 0.99j, for BibTeX 0.99c, TeX 3.0 or later.
336: % Copyright (C) 1990--92 by Karl Berry and Oren Patashnik; all rights reserved.
337: % You may copy this file provided: that it's accompanied by the
338: % "BibTeXing" document, whose text is contained in the file `btxdoc.tex';
339: % that any documentation you write for these macros also gives a
340: % reference for "BibTeXing"; and that either you make absolutely no
341: % changes to your copy, or if you do make changes, (1) you name the file
342: % something other than `btxmac.tex' and you remove all occurrences of
343: % `btxmac.tex' from the file, (2) you put, somewhere in the first twenty
344: % lines of the file, your name, along with an electronic address at which
345: % others who might use the file may reach you, and (3) you remove each
346: % occurrence of Oren's name and electronic address from this file. These
347: % restrictions help ensure that all standard versions of these macros are
348: % identical, and that Oren doesn't get deluged with inappropriate e-mail.
349: %
350: % This file, btxmac.tex, contains TeX macros that allow BibTeX, a
351: % bibliography program that was originally designed for use with LaTeX,
352: % to work with plain TeX. Please report any bugs (outright goofs,
353: % improvable macros, misfeatures, or unclear documentation) to Oren
354: % Patashnik (opbibtex@cs.stanford.edu). These macros will become frozen
355: % shortly after BibTeX version 1.00 is released.
356: %
357: % AMS-TEX WARNING: We tried very hard, for version .99i of these macros,
358: % to make them compatible with AmS-TeX. We succeeded to the extent
359: % that, if you use one of the standard bibliography styles, you probably
360: % won't notice any problems with version 0.99i of btxmac.tex. But
361: % ultimately we failed, in that the inherent incompatibilities between
362: % plain TeX and AmS-TeX kept making these macros break, for certain
363: % inputs or certain styles. Examples: (1) AmS-TeX treats at-signs as
364: % special, in ways that plain TeX and LaTeX don't, so that, for example,
365: % you can't have any `@' characters in an argument to the \cite command,
366: % the way you can in TeX or LaTeX; (2) AmS-TeX decided that plain TeX's
367: % and LaTeX's macron-accent control sequence `\=' should be undefined;
368: % so you'll need to define `\=' to be `\B' to get the xampl.bib example
369: % suggested below to work with AmS-TeX; (3) AmS-TeX redefines the tie
370: % character `~' of plain TeX, and AmS-TeX's `amsppt' style redefines
371: % plain TeX's `\nobreak' macro, so that if you use an author-date style
372: % like `apalike' and you have a multiple-author reference for which the
373: % author-date style automatically produces a citation in the text like
374: % `(Jones et~al., 1992)' you will throw AmS-TeX's `amsppt' style into
375: % an infinite loop, exceeding its input stack size. In practice, such
376: % incompatibilities surface infrequently; but it is now clear to us that
377: % it's not worth the effort (perhaps it's not even possible) to make the
378: % btxmac.tex macros robust when used both with plain TeX and Ams-TeX.
379: % If the BibTeX/AmS-TeX results attainable with the current btxmac.tex
380: % macros are sufficient, fine. But if there's a demand for more robust
381: % BibTeX/AmS-TeX behavior, then someone who's very familiar with the
382: % AmS-TeX package should probably make an amsbtxmc.tex version of the
383: % macros (remembering to follow the copyright restrictions above).
384: % Until then, if you're an AmS-TeX user, or a LaTeX or plain TeX user
385: % sharing files with an AmS-TeX user, beware.
386: % END OF AMS-TEX WARNING.
387: %
388: % To use these macros you should be familiar with how BibTeX interacts
389: % with LaTeX, since BibTeX's interaction with TeX is very similar; that
390: % interaction is explained in the LaTeX manual. It also helps to
391: % have read "BibTeXing", the documentation that accompanies BibTeX.
392: % Then, if you want, you should redefine any of the macros that begin
393: % with `\bbl' or `\biblabel' or `\print' that you need to get formatting
394: % different from the default (the default settings are designed to
395: % accompany a bibliography style like BibTeX's standard style `plain').
396: % The macros you might want to change are described briefly a few
397: % paragraphs hence. [To get started without reading any documentation,
398: % try running the nine-line .tex file below through TeX and BibTeX.
399: % Remember the general scheme: Running (La)TeX writes information on
400: % the .aux (auxiliary) file; then running BibTeX reads information from
401: % the .aux, .bst (style), and .bib (database) files, and writes
402: % information (the bibliography) on a .bbl file; then running (La)TeX
403: % incorporates the bibliography; then running (La)TeX once more fixes
404: % the remaining forward references into the bibliography. Thus, to get
405: % everything incorporated into your output, you'll have to run (La)TeX,
406: % BibTeX, (La)TeX, (La)TeX. (Standup, sitdown, fight, fight, fight.)]
407: %
408: % These macros can stand alone or they can be \input into a macro
409: % package, like Eplain, that is sufficiently compatible with plain TeX.
410: % To use these macros to format the 0.99 version of the xampl.bib file
411: % that's distributed with BibTeX (that version of the file has no
412: % self-identification), you'll need to define \mbox, which is a LaTeX
413: % command, to be \hbox, as in the example below.
414: %
415: % Here's a nine-line plain TeX file for trying out btxmac.tex; of course
416: % you'll have to remove the comment characters at the beginning of each
417: % line, and, depending on your system, you might have to take steps so
418: % that BibTeX can "see" the files xampl.bib and plain.bst (BibTeX will
419: % give you two empty-field warning messages that you should ignore).
420: %
421: % \def\mbox#1{\leavevmode\hbox{#1}}
422: % \input btxmac
423: % \noindent This cites Aamport's gnominious article~\cite{article-full}.
424: % \medskip
425: % \leftline{\bf References}
426: % \nocite{*} % put all database entries into the reference list
427: % \bibliography{xampl} % specify the database files; here, just xampl.bib
428: % \bibliographystyle{plain} % specify plain.bst as the style file
429: % \bye
430: %
431: %
432: % HISTORY
433: %
434: % Karl Berry wrote the original version of these macros in 1989 and
435: % 1990, for use in his `Eplain' package. Oren Patashnik modified them
436: % slightly in July 1990, as part of the official BibTeX distribution.
437: %
438: % 1-Aug-90 Version 0.99a, not released to the general public.
439: % 14-Aug-90 0.99b, first general release.
440: % 26-Aug-90 0.99c, made \@undefinedmessage work with other macro packages.
441: % 6-Sep-90 0.99d, allowed for general formatting of bibliography labels,
442: % for general formatting of (in-text) citations, and for
443: % changing certain catcodes while reading the .aux file.
444: % 14-Nov-90 0.99e, changed the way \@setletters works, made some \new...'s
445: % non-outer, and changed the way Eplain reads this file.
446: % 12-Dec-90 0.99f, made \@resetnumerals change the `,' and `.' catcodes; and
447: % added \biblabelextrahang, \@getoptionalarg, and \bblsc.
448: % 11-Mar-91 0.99g, made a few minor changes required by the way Eplain reads
449: % this file, but no functional changes.
450: % 24-Apr-91 0.99h, inhibited the reading and writing of the .aux file if it
451: % isn't used or if the \noauxfile macro is defined, and
452: % removed some .aux-file-opening detritus; printed the
453: % cite-key of undefined citations in \tt font; changed the
454: % catcode of `_' inside \cite; and called \@resetnumerals
455: % from inside a group.
456: % 29-Feb-92 0.99i, made these macros semi-compatible with AmS-TeX; removed
457: % \@resetnumerals, \@setletters, \@tokstostring, and
458: % friends; changed the way \cite handles catcodes; changed
459: % \@getoptionalarg, and had \bibitem and \newcommand use
460: % it; added \@futurenonspacelet and (to facilitate the use
461: % of multiple reference lists) \bblfilebasename; changed
462: % \biblabelprint to use the new macros \biblabelprecontents
463: % and \biblabelpostcontents, and to, by default, right-
464: % justify numeric labels; and renamed \biblabelextrahang to
465: % the more descriptive \biblabelextraspace.
466: % 14-Mar-92 0.99j, made 0.99i's use of `\\' local to btxmac.tex.
467: %
468: %
469: % The LaTeX-related commands defined in this file include (a) the four
470: % commands that a user types (\bibliography, \bibliographystyle, \cite,
471: % and \nocite); (b) the three commands that BibTeX looks for in the .aux
472: % file (\bibdata, \bibstyle, and \citation---there is a fourth command
473: % that BibTeX looks for, but that command is related to LaTeX's \include
474: % facility, so these macros ignore that command); and (c) a LaTeX
475: % command (\newcommand) that's written by one of the four standard
476: % bibliography styles (alpha). The definitions here are not exactly the
477: % same as the corresponding LaTeX definitions (those eight LaTeX
478: % definitions depend on a significant fraction of LaTeX itself). But
479: % the only substantial differences are with \newcommand, which here,
480: % without complaining, lets you redefine a preexisting control sequence
481: % (in LaTeX, \newcommand won't let you redefine a preexisting command),
482: % and which here doesn't make the control sequences it defines \long (in
483: % LaTeX, that happens automatically); there may also be other minor
484: % differences. To summarize: Unless you know what you're doing, you
485: % shouldn't define any control sequences with these eight names:
486: %
487: % \bibdata
488: % \bibliography
489: % \bibliographystyle
490: % \bibstyle
491: % \citation
492: % \cite
493: % \newcommand
494: % \nocite
495: %
496: % There are three other commands written by one or more of the four
497: % standard (plain, abbrv, alpha, unsrt) or four semistandard (acm,
498: % apalike, ieeetr, siam) bibliography styles; they take effect only
499: % within the bibliography, and are redefinable, as explained later:
500: %
501: % \em
502: % \newblock
503: % \sc
504: %
505: % There's one control sequence you might want to use (but not redefine)
506: % in redefining \biblabelprint:
507: %
508: % \biblabelwidth
509: %
510: % There are fifteen other control sequences (explained later in more detail)
511: % that the macros of this file will use if you define them---you should
512: % define them after the \input btxmac command but before the \bibliography
513: % command. The first six begin with `\bbl' and affect fonts, spacing,
514: % perhaps other characteristics of the bibliography, and which .bbl files
515: % get read; the next five begin with `\biblabel' and determine how labels
516: % are formatted in the bibliography; and the last four begin with `\print'
517: % and determine how the in-text citations are formatted:
518: %
519: % \bblem
520: % \bblfilebasename
521: % \bblhook
522: % \bblnewblock
523: % \bblrm
524: % \bblsc
525: % \biblabelcontents
526: % \biblabelprecontents
527: % \biblabelprint
528: % \biblabelpostcontents
529: % \biblabelextraspace
530: % \printbetweencitations
531: % \printcitefinish
532: % \printcitenote
533: % \printcitestart
534: %
535: % If it's defined before the \input btxmac command, the control sequence
536: % below inhibits the reading and writing of the .aux file(s), and the
537: % issuing of related warning messages. Any definition will do. This
538: % feature might help when you're working on draft stages of a document:
539: %
540: % \noauxfile
541: %
542: %
543: % Here's another control sequence (it's described later) that you
544: % probably won't want to redefine unless you are writing another macro
545: % package; if you do redefine it, however, do it before the \input btxmac
546: % command (and notice that it has an `@' in its name):
547: %
548: % \@undefinedmessage
549: %
550: % Any other control sequence in this file that might conflict with
551: % something you've defined will have an `@' in its name, so such conflicts
552: % are unlikely; but if you're worried about a specific control sequence
553: % name, do a text search of this file to look for it.
554: %
555: %
556: % So to start things off we turn `@' into a letter (category code 11),
557: % keeping track of the old category code for future restoration.
558: % (Simply resetting it to 12 when we leave these macros is
559: % insufficient.) The use of `\cite' as a temporary control sequence is
560: % a kludge, but it's a reasonably simple way to accomplish what we need
561: % without possibly overwriting something (without an `@' in its name)
562: % that might already be defined.
563: %
564: \edef\cite{\the\catcode`@}%
565: \catcode`@ = 11
566: \let\@oldatcatcode = \cite
567: \chardef\@letter = 11
568: \chardef\@other = 12
569: %
570: %
571: % Next come some things that will be useful later.
572: %
573: % Make an outer definition into an inner one (due to Chris Thompson).
574: % The arguments should be the control sequence to be defined, and the
575: % new of the \outer control sequence, as characters; the control
576: % sequence #1 is defined to be just the same as \csname#2\endcsname, but
577: % not \outer. For example, \@innerdef\innernewcount{newcount} would
578: % define \innernewcount to be a non-outer version of \newcount.
579: %
580: \def\@innerdef#1#2{\edef#1{\expandafter\noexpand\csname #2\endcsname}}%
581: %
582: % We use \@innerdef to make some of our allocations local, because
583: % Eplain includes our code inside a conditional. We put @'s in the
584: % names to minimize the (already small) chance of conflicts.
585: %
586: \@innerdef\@innernewcount{newcount}%
587: \@innerdef\@innernewdimen{newdimen}%
588: \@innerdef\@innernewif{newif}%
589: \@innerdef\@innernewwrite{newwrite}%
590: %
591: %
592: % Swallow one parameter.
593: %
594: \def\@gobble#1{}%
595: %
596: %
597: % Use TeX 3.0's \inputlineno to get the line number, for better error
598: % messages, but if we're using an old version of TeX, don't do anything.
599: %
600: \ifx\inputlineno\@undefined
601: \let\@linenumber = \empty % Pre-3.0.
602: \else
603: \def\@linenumber{\the\inputlineno:\space}%
604: \fi
605: %
606: %
607: % The following macro \@futurenonspacelet (from the TeXbook) behaves
608: % essentially like \futurelet except that it discards any implicit or
609: % explicit space tokens that intervene before a nonspace is scanned:
610: %
611: \def\@futurenonspacelet#1{\def\cs{#1}%
612: \afterassignment\@stepone\let\@nexttoken=
613: }%
614: \begingroup % The grouping here avoids stepping on an outside use of `\\'.
615: \def\\{\global\let\@stoken= }%
616: \\ % now \@stoken is a space token (\\ is a control symbol, so that
617: % space after it is seen).
618: \endgroup
619: \def\@stepone{\expandafter\futurelet\cs\@steptwo}%
620: \def\@steptwo{\expandafter\ifx\cs\@stoken\let\@@next=\@stepthree
621: \else\let\@@next=\@nexttoken\fi \@@next}%
622: \def\@stepthree{\afterassignment\@stepone\let\@@next= }%
623: %
624: %
625: % \@getoptionalarg\CS gets an optional argument from the input, enclosed
626: % in brackets, then expands \CS. We set \@optionalarg to \empty if we
627: % don't find one, otherwise to the text of the argument. This assumes
628: % the brackets don't have a funny category code.
629: %
630: \def\@getoptionalarg#1{%
631: \let\@optionaltemp = #1%
632: \let\@optionalnext = \relax
633: \@futurenonspacelet\@optionalnext\@bracketcheck
634: }%
635: %
636: % The \expandafter's in this macro let us avoid the use of \aftergroup,
637: % which is somewhat more expensive.
638: %
639: \def\@bracketcheck{%
640: \ifx [\@optionalnext
641: \expandafter\@@getoptionalarg
642: \else
643: \let\@optionalarg = \empty
644: % We can't do the \temp after the \fi, because then the \temp gets
645: % in the way of reading the optional argument from the input, if
646: % we do have one.
647: \expandafter\@optionaltemp
648: \fi
649: }%
650: %
651: \def\@@getoptionalarg[#1]{%
652: \def\@optionalarg{#1}%
653: \@optionaltemp
654: }%
655: %
656: %
657: % From LaTeX.
658: %
659: \def\@nnil{\@nil}%
660: \def\@fornoop#1\@@#2#3{}%
661: %
662: \def\@for#1:=#2\do#3{%
663: \edef\@fortmp{#2}%
664: \ifx\@fortmp\empty \else
665: \expandafter\@forloop#2,\@nil,\@nil\@@#1{#3}%
666: \fi
667: }%
668: %
669: \def\@forloop#1,#2,#3\@@#4#5{\def#4{#1}\ifx #4\@nnil \else
670: #5\def#4{#2}\ifx #4\@nnil \else#5\@iforloop #3\@@#4{#5}\fi\fi
671: }%
672: %
673: \def\@iforloop#1,#2\@@#3#4{\def#3{#1}\ifx #3\@nnil
674: \let\@nextwhile=\@fornoop \else
675: #4\relax\let\@nextwhile=\@iforloop\fi\@nextwhile#2\@@#3{#4}%
676: }%
677: %
678: %
679: % This macro tests if a file \jobname.#1 exists, and sets \if@fileexists
680: % appropriately. If an optional argument is given, it is used as the
681: % root part of the filename instead of \jobname.
682: %
683: \@innernewif\if@fileexists
684: %
685: \def\@testfileexistence{\@getoptionalarg\@finishtestfileexistence}%
686: \def\@finishtestfileexistence#1{%
687: \begingroup
688: \def\extension{#1}%
689: \immediate\openin0 =
690: \ifx\@optionalarg\empty\jobname\else\@optionalarg\fi
691: \ifx\extension\empty \else .#1\fi
692: \space
693: \ifeof 0
694: \global\@fileexistsfalse
695: \else
696: \global\@fileexiststrue
697: \fi
698: \immediate\closein0
699: \endgroup
700: }%
701: %
702: %
703: %% [[[start of BibTeX-specific stuff]]]
704: %
705: % Now come the four main LaTeX commands and their associated .aux
706: % commands. Just as in LaTeX, \bibliographystyle defines the BibTeX
707: % style name (.bst file, that is), and \bibliography defines the
708: % database (.bib) file(s). The corresponding .aux-file commands are
709: % \bibstyle and \bibdata, which are there only for BibTeX's (but not
710: % LaTeX's) use.
711: %
712: \def\bibliographystyle#1{%
713: \@readauxfile
714: \@writeaux{\string\bibstyle{#1}}%
715: }%
716: \let\bibstyle = \@gobble
717: %
718: % As well as writing the \bibdata command to tell BibTeX which .bib
719: % files to read, we read the .bbl file that BibTeX (or a person,
720: % conceivably) has produced. We use \bblfilebasename as the root of the
721: % filename to read; this defaults to \jobname.
722: %
723: \let\bblfilebasename = \jobname
724: \def\bibliography#1{%
725: \@readauxfile
726: \@writeaux{\string\bibdata{#1}}%
727: \@testfileexistence[\bblfilebasename]{bbl}%
728: \if@fileexists
729: % We just output a non-discardable item (the `whatsit' with the
730: % \bibdata command). This means that the glue that will be
731: % inserted next (\parskip or \baselineskip, most likely) will be a
732: % legal breakpoint. Most likely, this is after some kind of
733: % heading, where we don't want to allow a page break. So:
734: \nobreak
735: \@readbblfile
736: \fi
737: }%
738: \let\bibdata = \@gobble
739: %
740: % The \nocite{label,label,...} command writes its argument to \@auxfile,
741: % unless instructed not to, but produces no text in the document. Both
742: % the \nocite and \cite commands produce \citation commands in the .aux file.
743: %
744: \def\nocite#1{%
745: \@readauxfile
746: \@writeaux{\string\citation{#1}}%
747: }%
748: %
749: \@innernewif\if@notfirstcitation
750: %
751: % \cite[note]{label,label,...} produces the citations for the labels as
752: % well. If the optional argument `note' is present, it's added after
753: % the labels. Since \cite calls \nocite to do its .aux-file writing,
754: % \cite doesn't need to call \@readauxfile (\nocite does).
755: %
756: \def\cite{\@getoptionalarg\@cite}%
757: %
758: % Typeset the citations for the labels in #1, followed by the note, if
759: % it exists. To change the citation's format in the text, redefine one
760: % or more `\print...' macros, whose defaults appear later in this file.
761: %
762: \def\@cite#1{%
763: % Remember the optional argument, in case one of the macros we call
764: % below ends up looking for an optional argument itself. For
765: % example, if a \cite[note] triggers reading the .aux file, then the
766: % [note] would be clobbered, since \@testfileexistence looks for an
767: % optional arg.
768: \let\@citenotetext = \@optionalarg
769: % Start printing the text, beginning with a left bracket by default.
770: \printcitestart
771: % It's complicated, but because \nocite puts a `whatsit' onto the list,
772: % \nocite should follow \printcitestart. It's conceivable, but very
773: % unlikely, that this `whatsit' will cause a problem (glue that doesn't
774: % disappear when you want it to is the most likely symptom), requiring
775: % a change either to \printcitestart or to the label that the .bst file
776: % produces.
777: \nocite{#1}%
778: \@notfirstcitationfalse
779: \@for \@citation :=#1\do
780: {%
781: \expandafter\@onecitation\@citation\@@
782: }%
783: \ifx\empty\@citenotetext\else
784: \printcitenote{\@citenotetext}%
785: \fi
786: \printcitefinish
787: }%
788: %
789: \def\@onecitation#1\@@{%
790: \if@notfirstcitation
791: \printbetweencitations
792: \fi
793: %
794: \expandafter \ifx \csname\@citelabel{#1}\endcsname \relax
795: \if@citewarning
796: \message{\@linenumber Undefined citation `#1'.}%
797: \fi
798: % Give it a dummy definition:
799: \expandafter\gdef\csname\@citelabel{#1}\endcsname{%
800: % Change: marginal remark added, goldstrn@math.huji.ac.il,
801: % goldstern@tuwien.ac.at, May 1996 mg
802: % !!! change !!!
803: \strut
804: \vadjust{\vskip-\dp\strutbox
805: \vbox to 0pt{\vss\parindent0cm \leftskip=\hsize
806: \advance\leftskip3mm
807: \advance\hsize 4cm\strut\openup-4pt
808: \rightskip 0cm plus 1cm minus 0.5cm ? #1 ?\strut}}
809: {\tt
810: \escapechar = -1
811: \nobreak\hskip0pt
812: \expandafter\string\csname#1\endcsname
813: \nobreak\hskip0pt
814: }%
815: }%
816: \fi
817: % Now produce the text, whether it was undefined or not.
818: \csname\@citelabel{#1}\endcsname
819: \@notfirstcitationtrue
820: }%
821: %
822: % Given a label `foo', the macro `\b@foo' is supposed to
823: % hold the text that should be produced.
824: %
825: \def\@citelabel#1{b@#1}%
826: %
827: % So, how does a citation label get defined? When we read the .bbl file
828: % (below), a \bibitem writes out a \@citedef command. And when we read
829: % the \@citedef, we define \@citelabel{#1}, where #1 is the user's
830: % label.
831: %
832: \def\@citedef#1#2{\expandafter\gdef\csname\@citelabel{#1}\endcsname{#2}}%
833: %
834: %
835: % Reading the .bbl file also produces the typeset bibliography. Please
836: % notice, however, that we do not produce the title for the references
837: % (e.g., `References'), as LaTeX does. The formatting and spacing of
838: % that title, whether it should go into the headline, and so on, are all
839: % things determined by your format. We cannot know those things in
840: % advance. If you wish, you can define \bblhook to produce the title.
841: % Or just do it before the \bibliography command.
842: %
843: \def\@readbblfile{%
844: % Define a counter to tell us which item number we are on, unless
845: % we've already defined it (because the document has more than one
846: % bibliography).
847: \ifx\@itemnum\@undefined
848: \@innernewcount\@itemnum
849: \fi
850: %
851: \begingroup
852: \def\begin##1##2{%
853: % ##1 is just `thebibliography'.
854: % ##2 is the widest label.
855: % We set (new dimen) \biblabelwidth based on the widest label
856: \setbox0 = \hbox{\biblabelcontents{##2}}%
857: \biblabelwidth = \wd0
858: }%
859: \def\end##1{}% ##1 is `thebibliography' again.
860: %
861: % Here we have two possibilities:
862: % \bibitem[typesetlabel]{citationlabel}
863: % \bibitem{citationlabel}
864: % If we have the second of these, the citations are numbered, starting
865: % from one; we use our own count register \@itemnum for this.
866: %
867: \@itemnum = 0
868: \def\bibitem{\@getoptionalarg\@bibitem}%
869: \def\@bibitem{%
870: \ifx\@optionalarg\empty
871: \expandafter\@numberedbibitem
872: \else
873: \expandafter\@alphabibitem
874: \fi
875: }%
876: \def\@alphabibitem##1{%
877: % Need \xdef here for various reasons.
878: \expandafter \xdef\csname\@citelabel{##1}\endcsname {\@optionalarg}%
879: % Left-justify alpha labels, unless \biblabel{pre,post}contents
880: % are already defined.
881: \ifx\biblabelprecontents\@undefined
882: \let\biblabelprecontents = \relax
883: \fi
884: \ifx\biblabelpostcontents\@undefined
885: \let\biblabelpostcontents = \hss
886: \fi
887: \@finishbibitem{##1}%
888: }%
889: %
890: \def\@numberedbibitem##1{%
891: \advance\@itemnum by 1
892: \expandafter \xdef\csname\@citelabel{##1}\endcsname{\number\@itemnum}%
893: % Right-justify numeric labels, unless \biblabel{pre,post}contents
894: % are already defined.
895: \ifx\biblabelprecontents\@undefined
896: \let\biblabelprecontents = \hss
897: \fi
898: \ifx\biblabelpostcontents\@undefined
899: \let\biblabelpostcontents = \relax
900: \fi
901: \@finishbibitem{##1}%
902: }%
903: %
904: \def\@finishbibitem##1{%
905: \biblabelprint{\csname\@citelabel{##1}\endcsname}%
906: \@writeaux{\string\@citedef{##1}{\csname\@citelabel{##1}\endcsname}}%
907: \ignorespaces
908: }%
909: %
910: % Do the printing (we're producing the bibliography, remember).
911: %
912: \let\em = \bblem
913: \let\newblock = \bblnewblock
914: \let\sc = \bblsc
915: % Punctuation won't affect spacing;
916: \frenchspacing
917: % the penalties below are from LaTeX's [article,book,report].sty;
918: \clubpenalty = 4000 \widowpenalty = 4000
919: % the next two values come from LaTeX's \sloppy command;
920: \tolerance = 10000 \hfuzz = .5pt
921: \everypar = {\hangindent = \biblabelwidth
922: \advance\hangindent by \biblabelextraspace}%
923: \bblrm
924: % the \parskip is a guess at what looks good;
925: \parskip = 1.5ex plus .5ex minus .5ex
926: % and the space between label and text comes from LaTeX's \labelsep.
927: \biblabelextraspace = .5em
928: \bblhook
929: %
930: \input \bblfilebasename.bbl
931: \endgroup
932: }%
933: %
934: % The widest label's width is useful for redefining \biblabelprint;
935: % you redefine \biblabelwidth, in effect, by redefining the
936: % \biblabelcontents macro that appears below. And \biblabelextraspace,
937: % which is redefinable inside \bblhook, is added to \biblabelwidth to
938: % determine the amount of hanging indentation.
939: %
940: \@innernewdimen\biblabelwidth
941: \@innernewdimen\biblabelextraspace
942: %
943: % Now come the main macros that are related to the printing of the
944: % bibliography. Since you might want to redefine them, they are given
945: % default definitions outside of \@readbblfile.
946: %
947: % The first one controls the printing of a bibliography entry's label.
948: % If you change it, make sure that it starts with something like
949: % \noindent or \indent or \leavevmode that puts TeX into horizontal mode
950: % (even if the label itself is empty); otherwise, the hanging
951: % indentation will get messed up in certain circumstances.
952: %
953: \def\biblabelprint#1{%
954: \noindent
955: \hbox to \biblabelwidth{%
956: \biblabelprecontents
957: \biblabelcontents{#1}%
958: \biblabelpostcontents
959: }%
960: \kern\biblabelextraspace
961: }%
962: %
963: % If you are using numeric labels, and you want them left-justified
964: % (numeric labels by default are right-justified), do something like:
965: % \def\biblabelprecontents{\relax}
966: % \def\biblabelpostcontents{\hss}
967: %
968: % By default the labels are typeset in \bblrm, and enclosed in brackets.
969: %
970: \def\biblabelcontents#1{{\bblrm [#1]}}%
971: %
972: % The main text, too, is typeset using \bblrm, which is \rm by default.
973: %
974: \def\bblrm{\rm}%
975: %
976: % Emphasis for producing, e.g., titles, is done with \it by default.
977: %
978: \def\bblem{\it}%
979: %
980: % Some styles use a caps-and-small-caps font for author names. LaTeX
981: % defines an \sc command but plain TeX doesn't, so we need one here.
982: % The definition below doesn't load the font unless it's needed, but it
983: % tries to load only the 10pt version, because it might not exist at
984: % other point sizes.
985: %
986: \def\bblsc{\ifx\@scfont\@undefined
987: \font\@scfont = cmcsc10
988: \fi
989: \@scfont
990: }%
991: %
992: % The major parts of an entry are separated with \bblnewblock. The
993: % numbers below are taken from LaTeX's `article' style.
994: %
995: \def\bblnewblock{\hskip .11em plus .33em minus .07em }%
996: %
997: % Here's where you stick any other bibliography-formatting goodies, or
998: % redefine the values above.
999: %
1000: \let\bblhook = \empty
1001: %
1002: %
1003: % Here are the four default definitions for formatting the in-text
1004: % citations. These are what you redefine (after your \input btxmac but
1005: % before your \bibliography) to get parens instead of brackets, or
1006: % superscripts, or footnotes, or whatever.
1007: %
1008: \def\printcitestart{[}% left bracket
1009: \def\printcitefinish{]}% right bracket
1010: \def\printbetweencitations{, }% comma, space
1011: \def\printcitenote#1{, #1}% comma, space, note (if it exists)
1012: %
1013: % That scheme is pretty flexible. For example you could use
1014: % \def\printcitestart{\unskip $^\bgroup}
1015: % \def\printcitefinish{\egroup$}
1016: % \def\printbetweencitations{,}
1017: % \def\printcitenote#1{\hbox{\sevenrm\space (#1)}}
1018: % \font\eighttt = cmtt8
1019: % \scriptfont\ttfam = \eighttt
1020: % to get superscripted in-text citations. (The scriptfont stuff
1021: % exists only to print an undefined citation; it's in cmtt8 because
1022: % there is no cmtt7.) To get something radically different, however,
1023: % you'll have to define your own \cite command.
1024: %
1025: % When we read `\citation' from the .aux file, it means nothing.
1026: %
1027: \let\citation = \@gobble
1028: %
1029: %
1030: % Now comes the stuff for dealing with LaTeX's \newcommand. As
1031: % mentioned earlier, this \newcommand will redefine a preexisting
1032: % command; that's different from how LaTeX's \newcommand behaves.
1033: %
1034: \@innernewcount\@numparams
1035: %
1036: % \newcommand{\foo}[n]{text} defines the control sequence \foo to have
1037: % n parameters, and replacement text `text'.
1038: %
1039: \def\newcommand#1{%
1040: \def\@commandname{#1}%
1041: \@getoptionalarg\@continuenewcommand
1042: }%
1043: %
1044: % Figure out if this definition has parameters.
1045: %
1046: \def\@continuenewcommand{%
1047: % If no optional argument, we have zero parameters. Otherwise, we
1048: % have that many.
1049: \@numparams = \ifx\@optionalarg\empty 0\else\@optionalarg \fi \relax
1050: \@newcommand
1051: }%
1052: %
1053: % \@numparams is how many arguments this command has. The name of the
1054: % command is \@commandname. The replacement text for the new macro is #1.
1055: %
1056: \def\@newcommand#1{%
1057: \def\@startdef{\expandafter\edef\@commandname}%
1058: \ifnum\@numparams=0
1059: \let\@paramdef = \empty
1060: \else
1061: \ifnum\@numparams>9
1062: \errmessage{\the\@numparams\space is too many parameters}%
1063: \else
1064: \ifnum\@numparams<0
1065: \errmessage{\the\@numparams\space is too few parameters}%
1066: \else
1067: \edef\@paramdef{%
1068: % This is disgusting, but \loop doesn't work inside \edef,
1069: % because \body isn't defined.
1070: \ifcase\@numparams
1071: \empty No arguments.
1072: \or ####1%
1073: \or ####1####2%
1074: \or ####1####2####3%
1075: \or ####1####2####3####4%
1076: \or ####1####2####3####4####5%
1077: \or ####1####2####3####4####5####6%
1078: \or ####1####2####3####4####5####6####7%
1079: \or ####1####2####3####4####5####6####7####8%
1080: \or ####1####2####3####4####5####6####7####8####9%
1081: \fi
1082: }%
1083: \fi
1084: \fi
1085: \fi
1086: \expandafter\@startdef\@paramdef{#1}%
1087: }%
1088: %
1089: %% [[[end of BibTeX-specific stuff]]]
1090: %
1091: %
1092: % Names of references (arguments given in the \cite and \nocite
1093: % commands) and file names (arguments given in the \bibliography and
1094: % \bibliographystyle commands) are recorded in \jobname.aux, called the
1095: % \@auxfile in these macros. Here's how they get read in.
1096: %
1097: \def\@readauxfile{%
1098: \if@auxfiledone \else % remember: \@auxfiledonetrue if \noauxfile is defined
1099: \global\@auxfiledonetrue
1100: \@testfileexistence{aux}%
1101: \if@fileexists
1102: \begingroup
1103: % Because we might be in horizontal mode when \@readauxfile
1104: % is called (if it's in response to a \cite or \nocite), we
1105: % want to ignore all the would-be spaces at the ends of
1106: % lines in the aux file. Fortunately, it's highly unlikely
1107: % an end-of-line might actually be desired.
1108: % And because we don't change the category code of anything
1109: % but @, primitives like \gdef can't be used to define labels
1110: % in the aux file. The solution adopted by btxmac.tex is to
1111: % write `\@citedef{LABEL}{DEFINITION}' to the aux file, and
1112: % use \csname on LABEL.
1113: \endlinechar = -1
1114: \catcode`@ = 11
1115: \input \jobname.aux
1116: \endgroup
1117: \else
1118: \message{\@undefinedmessage}%
1119: \global\@citewarningfalse
1120: \fi
1121: \immediate\openout\@auxfile = \jobname.aux
1122: \fi
1123: }%
1124: %
1125: % The \@readauxfile macro does all that work the first time it's called.
1126: % Since it's called once for every \cite, \nocite, \bibliography, and
1127: % \bibliographystyle command that the user issues, we need to remember
1128: % whether the work's been done. It's considered done if we're not to do
1129: % it---that is, if \noauxfile is defined.
1130: %
1131: \newif\if@auxfiledone
1132: \ifx\noauxfile\@undefined \else \@auxfiledonetrue\fi
1133: %
1134: % It's conceivable you'd want to change how other characters are read;
1135: % to do that, change their category code before doing \input btxmac.
1136: %
1137: %
1138: % After reading the .aux file, \@readauxfile opens it for writing.
1139: % The \@writeaux macro does the actual writing (as long as
1140: % \noauxfile is undefined).
1141: %
1142: \@innernewwrite\@auxfile
1143: \def\@writeaux#1{\ifx\noauxfile\@undefined \write\@auxfile{#1}\fi}%
1144: %
1145: %
1146: % A macro package that uses btxmac.tex might define
1147: % \@undefinedmessage (before doing an \input btxmac).
1148: %
1149: \ifx\@undefinedmessage\@undefined
1150: \def\@undefinedmessage{No .aux file; I won't give you warnings about
1151: undefined citations.}%
1152: \fi
1153: %
1154: % Even if citations are undefined, we want to complain only if
1155: % \@citewarningtrue. The default is to set \@citewarningtrue unless
1156: % \noauxfile is defined. Again, a macro package that uses
1157: % btxmac.tex might want to redefine this.
1158: %
1159: \@innernewif\if@citewarning
1160: \ifx\noauxfile\@undefined \@citewarningtrue\fi
1161: %
1162: %
1163: % Finally, before leaving we restore @'s old category code.
1164: %
1165: \catcode`@ = \@oldatcatcode
1166:
1167: % This will define \cite and make sure it works as in latex
1168:
1169:
1170: \def\widestnumber#1#2{}
1171: % Our amstex-ppt style does not know about \widestnumber
1172:
1173:
1174:
1175: \def\rm{\fam0 \tenrm}
1176:
1177:
1178: \def\fakesubhead#1\endsubhead{\bigskip\noindent{\bf#1}\par}
1179:
1180:
1181: % % \input rsfs
1182:
1183: % # Keywords: Script or Calligraphic (Caligraphic) letters with the RSFS Font
1184:
1185: % The story so far: July 1998 -- Saharon would like to have a
1186: % ``nicer'' calligraphic font. In particualr, the leters S and P in
1187: % the usual calligraphic font do not look ``special'' enough.
1188: %
1189: % I found out that ``rsfs'' (``Ralph Smith Formal Script'') may be
1190: % what he wants. I installed the mf file, the .tfm file, as well as
1191: % a few pk files in ~/TeX/rsfs. Let's hope that this is enough.
1192: % Using amstex, all you have to do is to \input rsfs.tex
1193: % Files prepared with citealice willdothis automatically.
1194: %
1195: % Note: for some reason xdvi calls MakeTeXpk, then Maketexpk
1196: % complains about wrong resolution, but still writes commands to
1197: % missfont.log...
1198: %
1199:
1200:
1201:
1202: % we redefine a macro inside amstex's \Cal command , so that it calls
1203: % our nice font ``rsfs'' rather than the usual calligraphic font.
1204: % Note thisworks for amstex only.
1205: % In plain tex, would have to add definitions of \Cal
1206: % in latex... we should insteaduse mathrsfs.sty
1207: %
1208:
1209:
1210:
1211: \font\textrsfs=rsfs10
1212: \font\scriptrsfs=rsfs7
1213: \font\scriptscriptrsfs=rsfs5
1214:
1215:
1216: \newfam\rsfsfam
1217: \textfont\rsfsfam=\textrsfs
1218: \scriptfont\rsfsfam=\scriptrsfs
1219: \scriptscriptfont\rsfsfam=\scriptscriptrsfs
1220:
1221:
1222: \edef\oldcatcodeofat{\the\catcode`\@}
1223: \catcode`\@11
1224:
1225: \def\Cal@@#1{\noaccents@ \fam \rsfsfam #1}
1226:
1227: \catcode`\@\oldcatcodeofat
1228:
1229:
1230:
1231: \newpage
1232:
1233: \head {Anotated Content} \endhead \resetall
1234: % \resetall
1235: \bn
1236: \S0 $\quad$ Introduction
1237: \mr
1238: \item "{{}}" [We give lengthy explanations of the problems and proofs for
1239: No-New-Reals iterations (or proper forcing, CS iterations).]
1240: \ermn
1241: \S1 $\quad$ Preservation of not adding reals
1242: \mr
1243: \item "{{}}" [We present sufficient conditions for CS iteration of proper
1244: forcing not to add reals. For this we define ``reasonable parameters
1245: ${\frak p}$" and we have two main demands. One (clause (c) of Definition
1246: \scite{1.6}) is a weakening of ``$\alpha$-proper for every $\alpha <
1247: \omega_1$". This time it has the form (on $Q_i$), ${\frak p}$-proper which
1248: informally says that:
1249: if ${\frak p} \in N,Y \subseteq \{M \in N:M \text{ appropriate}\}$
1250: is $\alpha$-large then for some $(N,Q_i)$-generic condition $q \ge p,q$
1251: forces that
1252: $\{M \in Y:M[{\underset\tilde {}\to G_Q}] \cap V = M\}$ is $\alpha$-large
1253: (the meaning of $\alpha -$ large depends on ${\frak p}$), hence \wilog, \,
1254: ${\frak p}$ has
1255: length $\omega_1$. The other main demand (clause (d) of Definition
1256: \scite{1.6}) is a ``weak diamond preventive". \nl
1257: We then show that $\alpha$-properness for $\alpha < \omega_1$ is sufficient
1258: for the first main demand (in \scite{1.12}(3)). The demand on the games for
1259: ${\frak p}$ helps to prove the preservation of ${\frak p}$-properness.]
1260: \ermn
1261: \S2 $\quad$ Delayed properness
1262: \mr
1263: \item "{{}}" [The preservation theorem in the first section does not, for
1264: standard ${\frak p}$, cover shooting a club $C \subseteq \omega_1$ running
1265: away for $C_\delta \subseteq \delta = \sup(C_\delta),C_\delta$ small (see
1266: \S3). For this we will use $({\frak p},\alpha,\beta)$-proper for enough
1267: pairs $\alpha \le \beta < \ell g({\frak p})$ (so starting from $\beta$-large
1268: we get $\alpha$-large; for many $\alpha$ we can choose $\beta = \alpha$
1269: but during the inductive proof we pass through cases of $\alpha < \beta$). \nl
1270: Here we introduce various definitions and basic facts needed. We discuss
1271: axioms, version of the properties preserved by CS iterations and
1272: strengthening of the iteration Lemmas of \S1.]
1273: \ermn
1274: \S3 $\quad$ Example: shooting a thin club
1275: \mr
1276: \item "{{}}" [We present the natural forcing showing $\kappa = 2$ is
1277: interesting (not only $\kappa = \aleph_0$) (from \cite[Ch.VIII,\S4]{Sh:b}).
1278: \nl
1279: We show that the natural forcing (see above) for running away from $C_\delta
1280: \subseteq \delta$, of small order type (see \cite[Ch.XVIII,\S2]{Sh:f}) falls
1281: under our framework for delayed properness.
1282: We give examples: running away from
1283: $\langle C_{\delta,0},C_{\delta,1}:\delta < \omega_1 \text{ limit} \rangle,
1284: C_{\delta,0},C_{\delta,1}$ are disjoint closed subsets of $\delta$
1285: with no restrictions on their order type so we
1286: ask for $C,C \cap C_{\delta,0}$ or $C \cap C_{\delta,1}$ is bounded in
1287: $\delta$ and more.]
1288: \ermn
1289: \S4 $\quad$ Second preservation of not adding reals
1290: \mr
1291: \item "{{}}" [We give a sufficient condition for the limit not to add
1292: reals. We here are weakening the demand ``${\frak p}$-proper", using
1293: $({\frak p},\alpha,f(\alpha))$-proper instead
1294: $({\frak p},\alpha,\alpha)$-proper, what we called
1295: delayed properness. The price is that here ${\frak p}$ has
1296: length of large cofinality, so essentially we catch our tails on a club of
1297: it. Also the Lemma here covers the examples.]
1298: \ermn
1299: \S5 $\quad$ Problematic Forcing
1300: \mr
1301: \item "{{}}" [We discuss further generalizations.]
1302: \endroster
1303: \newpage
1304:
1305: \head {\S0 Introduction} \endhead \resetall \sectno=0
1306: % \resetall
1307: \bigskip
1308:
1309: We try to explain our problems and results. If the explanations look opaque,
1310: try to return to them after reading at least part of the proof. Sections
1311: \S0, \S1 are based on lectures in the logic seminar in the Hebrew University,
1312: Spring 1997, whose participants I thank. On the history see in
1313: \cite[V,\S7,VIII,\S4,XVIII,\S1,\S2]{Sh:f}, \cite[\S3]{Sh:666}.
1314: \definition{\stag{0.1} Definition} 1) Let $K_0$ be the family of CS
1315: iterations, $\bar Q = \langle P_i,{\underset\tilde {}\to Q_i}:i < \alpha
1316: \rangle$, we denote $P_\alpha = \text{ Lim}(\bar Q)$. \nl
1317: 2) We say $\bar Q$ is proper if each $Q_i$ is (which means: every
1318: $f \in ({}^\omega \omega)^{V^{P_{i+1}}} = ({}^\omega \omega)
1319: ^{(V^{P_i})^{O_i}}$ is bounded by some $g \in ({}^\omega \omega)^{V^{P_i}}$,
1320: hence $P_j/P_i$ is proper for
1321: $i < j \le \alpha$ (see \cite{Sh:b} or \cite{Sh:f}). \nl
1322: 3) We say $\bar Q$ is ${}^\omega \omega$-bounding if each $Q_i$ is
1323: \footnote {where $Q$ is ${}^\omega \omega$ bounded (in the universe $V$) if
1324: every $f \in ({}^\omega)^{V^Q}$ is bounded by some $g \in ({}^\omega
1325: \omega)^V$},
1326: hence $P_j/P_i$ is ${}^\omega \omega$-bounding for $i < j \le \alpha$. \nl
1327: 4) We say $\bar Q$ is NNR if $i < \alpha \Rightarrow P_{i+1}$
1328: adds no reals. \nl
1329: [Equivalently: $i < \alpha \Rightarrow Q_i$ adds no reals and $\beta <
1330: \alpha \Rightarrow P_\beta$ adds not reals.]
1331: \enddefinition
1332: \bn
1333: \ub{Discussion}: It would be nice if also this (NNR) would be preserved in
1334: limit. But this is wrong for two known reasons, obstacles, explained below:
1335: \mr
1336: \item "{$\bigotimes_1$}" weak diamond
1337: \sn
1338: \item "{$\bigotimes_2$}" existence of clubs.
1339: \ermn
1340: Our aim here is to weaken the medicine one uses against $\bigotimes_2$ for CS
1341: iterations of proper forcing notions. Let us explain the ``obstacles".
1342: \bn
1343: \ub{Concerning the weak diamond}:
1344:
1345: Let $\bar \eta = \langle \eta_\delta:\delta < \omega_1,\delta \text{ limit}
1346: \rangle,\eta_\delta = \langle \eta_\delta(n):n < \omega \rangle$ where
1347: $\eta_\delta$ is an $\omega$-sequence of ordinals (strictly) increasing
1348: with limit $\delta$. Let $D$ be a non-principal ultrafilter on $\omega$.
1349: \nl
1350: For $f \in {}^{\omega_1}2,\delta < \omega_1$ limit, let Av$_D(f,\eta_\delta)
1351: = \ell$ iff $\{n:f(\eta_\delta(n)) = \ell\} \in D$. \nl
1352: \bn
1353: \ub{\stag{0.2} Question[CH]}: Given $\bar e = \langle e_\delta:
1354: \delta < \omega_1 \text{ limit} \rangle,e_\delta \in \{0,1\}$ is there
1355: $f \in {}^{\omega_1}2$ such that for a club
1356: of $\delta < \omega_1$ we have $e_\delta = \text{ Av}_D(f,\eta_D)$?
1357: \mn
1358: Naturally, trying to prove consistency we should use a CS iteration $\bar Q$,
1359: for simplicity we assume
1360: \mr
1361: \item "{{}}" $V \models GCH,\bar Q = \langle P_i,{\underset\tilde {}\to Q_i}:
1362: i < \omega_2 \rangle$
1363: \sn
1364: \item "{{}}" $\quad \qquad \qquad Q_i = Q_{\bar e}$ (i.e.
1365: ${\underset\tilde {}\to Q_i} = Q_{\underset\tilde {}\to{\bar e}})$
1366: where
1367: $$
1368: \align
1369: Q_{\bar e} = \biggl\{f:&\text{ for some } \zeta < \omega_1,f \in {}^\zeta 2
1370: \text{ and for every limit} \\
1371: &\,\delta \le \zeta \text{ we have Av}_D(f,\eta_\delta) = e_\delta \biggr\}.
1372: \endalign
1373: $$
1374: \ermn
1375: This is a very nice forcing notion - it is proper (even $< \omega_1$-proper,
1376: see below) and NNR and for every
1377: $\alpha < \omega_1,{\Cal I}_\alpha = \{f \in Q_{\bar e}:\alpha \subseteq
1378: \text{ Dom}(f)\}$ is a dense open subset of it. \nl
1379: But the weak diamond (\cite{DvSh:65} or see \cite[Ch.XII,\S1]{Sh:b} =
1380: \cite[AP,\S1]{Sh:f}) tells us for this case that the answer is \ub{no},
1381: that is: \nl
1382: $\exists \bar e \forall f \in {}^{\omega_1}2 \exists^{\text{stat}} \delta
1383: (e_\delta \ne \text{ Av}_D(f,\eta_\delta))$.
1384:
1385: In fact this holds for any function Av$':\dbcu_{\delta < \omega_1}
1386: {}^\delta 2 \rightarrow \{0,1\}$.
1387: \bn
1388: \ub{Why is $Q_{\bar e}$ NNR}?
1389:
1390: Let $\langle N_i:i \le \omega^2 \rangle$ be increasing continuously with
1391: $i,Q_{\bar e} \in N_0$ where $N_i \prec ({\Cal H}(\chi),\in)$ is countable,
1392: $\bar N \restriction (i+1) \in N_{i+1}$;
1393: let $\delta_i = N_i \cap \omega_1$. So
1394: $\langle \delta_i:i \le \omega^2 \rangle$ is strictly increasing continuously.
1395: But $\{\eta_{\delta_{\omega^2}}(n):n < \omega\} \subset \delta_{\omega^2}$ has
1396: order type $\omega$, so $W = \{i < \omega^2:\exists n(\delta_i \le
1397: \eta_{\delta_{\omega^2}}(n) < \delta_{i+1})\}$ has order type $\omega$.
1398: \sn
1399: So we can find $\ell_n < \omega$ such that $\dsize \bigwedge_{n < \omega} \,
1400: \dsize \bigwedge_{m < \omega} \omega \times n + \ell_n +m \notin W$.
1401: Let $p \in Q_{\bar e} \cap N_0$, let
1402: $\underset\tilde {}\to \tau:\omega \rightarrow \text{ Ord},
1403: \underset\tilde {}\to \tau \in N_0$. We choose by induction on $n,p_n \in
1404: Q_{\bar e}$ such that $p \le p_n,
1405: p_{n-1} \le p_n,p_n \in N_{\omega n + \ell_n +1},p_n$ force some value to
1406: $\underset\tilde {}\to \tau(n)$ and on
1407: $[\delta_{\omega m + \ell_m},\delta_{\omega(m+1)+ \ell_{m+1}}) \
1408: \cap \text{ Rang}(\eta_\delta) \backslash
1409: \text{ Dom}(p)$ which agrees with $e_{\delta_{\omega^2}}$. \nl
1410: So
1411: \mr
1412: \item "{$(*)$}" the desired demand on $\bar Q_i$ (guaranteeing $P_\alpha$
1413: in NNR) should exclude the $Q_{\bar e}$'s.
1414: \endroster
1415: \bigskip
1416:
1417: \definition{\stag{0.3} Definition} Let $K_1$ be the class of proper
1418: ${}^\omega \omega$-bounding iteration $\bar Q \in K_0$, let $K_2$ be
1419: the class of NNR iterations $\bar Q \in K_1$, and let $K_3$ be the
1420: class of $\bar Q \in K_2$ such that
1421: \mr
1422: \item "{$(*)$}" if $\chi$ is large enough (so $\bar a \in {\Cal H}(\chi)$),
1423: \nl
1424: $N \prec ({\Cal H}(\chi),\in)$ countable, \nl
1425: $\bar Q \in N$, \nl
1426: $i \in \ell g(\bar Q) \cap N$, \nl
1427: $p \in P_{i+1} \cap N$, \nl
1428: $q_0,q_1 \in P_i$ are $(N,P_i)$-generic \nl
1429: (i.e. $q_\ell \Vdash N[{\underset\tilde {}\to G_{P_i}}] \cap V = N$) and
1430: moreover \nl
1431: $q_\ell \Vdash ``{\underset\tilde {}\to G_{P_i}} \cap N = G^*"$ and \nl
1432: $p \restriction i \le q_\ell$ \nl
1433: \ub{then} we can find $q'_0,q'_1,G^{**}$ such that $p \restriction (i+1),
1434: q_\ell \le q'_\ell \in P_{i+1}$, \nl
1435: $q'_\ell \Vdash ``{\underset\tilde {}\to G_P} \cap N = G^{**}"$, \nl
1436: $q'_\ell$ is $(N,P_{i+1})$-generic, so $G^{**} \subseteq P_{i+1} \cap N$
1437: is generic over $N$.
1438: \endroster
1439: \enddefinition
1440: \bn
1441: \ub{This tries to say}:
1442:
1443: We know ${\underset\tilde {}\to G_{P_i}} \cap N$ (as being $G^*$) and we
1444: are looking at $N[G^*]$ (formally, only its isomorphism type). So we know
1445: ${\underset\tilde {}\to Q^N_i}[G^*_i]$.
1446:
1447: We would like to find $G' \subseteq {\underset\tilde {}\to Q^N_i}[G^*]$
1448: generic over $N[G^*]$, so that $G^*,G'$ will determine $G^{**}$. But we
1449: need a guarantee that $G'$ will have an upper bound in
1450: ${\underset\tilde {}\to Q_i}[G_{P_i}]$. If we know $G_{P_i}$, fine;
1451: but in a sense, we are given 2 candidates (by $q_0,q_1$ and can increase
1452: then to $q'_0 \restriction i,q'_1 \restriction i$) and have to find $G'$
1453: ``accepted" by both.
1454:
1455: This (weak diamond) obstacle was overcome, with a price, i.e. more than
1456: being in $K_3$, \cite[V,\S7]{Sh:f} using $\aleph_1$-completeness systems
1457: and \cite[XVIII,\S4]{Sh:f} for 2-completeness systems (phrased in
1458: \cite[XVIII,\S2]{Sh:f} without them).
1459:
1460: Unfortunately, this does not suffice: there is an extreme example where e.g.
1461: if for $q_0,q_1$ incompatible in $Q_i,{\underset\tilde {}\to E_i}$ a
1462: $Q_i$-name of a club, for some $\alpha(q_0,q_1)$ we have:
1463:
1464: $$
1465: q_\ell \le q'_\ell
1466: $$
1467:
1468: $$
1469: q_\ell \Vdash ``{\underset\tilde {}\to E_i} \cap \delta
1470: = E^\delta_i \Rightarrow E^{\delta_0}_i \cap E^{\delta_1}_1 \backslash
1471: \alpha(q_0,q_1) \text{ is finite}".
1472: $$
1473: \mn
1474: This represents the reason, the obstacle which we shall call $\bigotimes_2$,
1475: it is
1476: overcomed (with a price) either with $(< \omega_1)$-properness or by a kind
1477: of ``finite powers are proper" (see below).
1478: \definition{\stag{0.4} Definition} 1) $Q$ is $\alpha$-proper when:
1479: \mr
1480: \item "{{}}" if $\bar N = \langle N_i:i \le \alpha \rangle$ is $\prec$-
1481: increasing continuously, $\alpha \in N_0$ \nl
1482: $N_i \prec ({\Cal H}(\chi),\in)$ countable,
1483: $\bar N \restriction (i+1) \in N_{i+1}$, \nl
1484: $Q \in N_0$ and $p \in Q \cap N_0$ \nl
1485: \ub{then} there is $q,p \le q \in Q_i$ such that $q$ is $(N_i,q)$-generic
1486: for $i \le \alpha$.
1487: \ermn
1488: 2) A forcing notion $Q$ is $(<^+ \omega_1)$-proper if the above holds for any
1489: $\alpha < \omega_1$ even omitting ``$\alpha \in N_0$". We say $Q$ is
1490: $(< \omega_1)$-proper if $Q$ is $\alpha$-proper for any $\alpha <
1491: \omega_1$. \nl
1492: So $(< \omega_1)$-proper is an antidote to such problems, i.e. against
1493: ``reason $\otimes_2$".
1494: Okay for specializing a Aronszajn tree and many others, \ub{but} it seems to
1495: me since \cite{Sh:177} too strong: it kills the following:
1496: \bn
1497: \ub{\stag{0.5} Question}: Let $\bar C = \langle C_\delta:\delta < \omega_1,
1498: \delta \text{ limit} \rangle,C_\delta \subseteq \delta =
1499: \sup(C_\delta)$, otp$(C_\delta) = \omega$ or at least $< \delta$, is there a
1500: club $E$ of $\omega_1$ such that $\delta < \omega_1 \Rightarrow \delta >
1501: \sup(C_\delta \cap E)$? (i.e. is this consistent with CH).
1502: \mn
1503: We consider
1504:
1505: $$
1506: \align
1507: Q^1_{\bar C} = \biggl\{ \bar f:&\text{ for some non-limit } \alpha < \omega_1
1508: \text{ we have } f \in {}^\alpha 2,f^{-1}(\{1\}) \text{ closed} \\
1509: &\text{and } \delta < \alpha \text{ limit } \Rightarrow \sup(f^{-1}(\{1\})
1510: \cap C_\delta) < \delta \biggr\}.
1511: \endalign
1512: $$
1513: \mn
1514: This is the natural forcing for adding a club such that $\dsize \bigwedge
1515: _\delta[C_\delta \cap E$ bounded in $\delta$]. So $E$ ``runs away" from each
1516: $C_\delta$. This forcing notion is NOT $\omega$-proper: if
1517: $\langle N_i:i \le \omega \rangle$ satisfies
1518: $C_{N_\omega \cap \omega_1} = \{N_i \cap \omega_1:i < \omega\}$,
1519: \ub{then} no $f \in Q$ is $(N_i,G)$-generic, for infinitely many
1520: $i$'s.
1521:
1522: A solution (\cite[XVIII,\S2]{Sh:f}) was to
1523: demand ``essentially" e.g. $P_i \times P_i$ is proper for
1524: $i < \ell g(\bar Q)$. While this is fine for $Q^1_i$, this seems to exclude
1525: specializing an Aronszajn tree by not adding reals. We will deal with a
1526: condition implied by both $(< \omega_1)$-proper and (essentially) ``the
1527: square of the forcing notion is proper".
1528: \enddefinition
1529: \bn
1530: \ub{Continuing explanation}:
1531:
1532: So for CS iteration $\bar Q$ of proper forcing the ``reasons", ``dangers"
1533: for adding reals may come from:
1534: \mr
1535: \item "{$\bigotimes_0$}" (0-reason, danger) some $Q_i$ adds reals and
1536: \sn
1537: \item "{$\bigotimes_1$}" weak diamond.
1538: \ermn
1539: Against this, we will assume something like (Definition \scite{0.3}):
1540: many times in some sense
1541: $q_0,q_1 \in P_i$ are $(N,P_i)$-generic, $p \in {\underset\tilde {}\to Q_i}
1542: \cap N,q_\ell \Vdash_{P_i} ``{\underset\tilde {}\to G_{P_i}} \cap N = G^*"$
1543: and for some $G',q'_0 \ge q_0,q'_1 \ge q_1$ in $P_{i+1}$ we have
1544: $G'\subseteq ({\underset\tilde {}\to Q_i}
1545: \cap N)[G^*]$ and $q'_\ell \Vdash_{P_i}
1546: ``{\underset\tilde {}\to G_{Q_i}} \cap N[G^*] = G'"$ and $p \in G'$. \nl
1547: It is simpler in the proof to allow $q_\ell(\ell < n)$ for some $n < \omega$;
1548: but in addition we have the obstacle:
1549: \mr
1550: \item "{$\bigotimes_2$}" adding almost disjoint clubs
1551: (\cite[Ch.XVIII,\S1]{Sh:f}).
1552: \ermn
1553: There were two medicines:
1554: \mr
1555: \item "{$(\alpha)$}" $\alpha$-proper for every $\alpha < \omega_1$
1556: \sn
1557: \item "{$(\beta)$}" something like $P_i \times P_i$ is proper.
1558: \ermn
1559: In the proofs we have a situation:
1560: \mr
1561: \item "{$(*)$}" $\bar Q \in N_0 \in N$ \nl
1562: $N_0 \prec ({\Cal H}(\chi),\in)$ and $N \prec ({\Cal H}(\chi),\in)$ are
1563: countable \nl
1564: $q_\ell$ is $(N,P_i)$-generic and $(N_0,P_i)$-generic \nl
1565: $q_\ell$ forces that ${\underset\tilde {}\to G_{P_i}} \cap
1566: N = G_\ell$, (for $\ell < 2$) \nl
1567: $G^* = G_1 \cap N_0 = G_2 \cap N_0$ \nl
1568: $i,j,p \in N_0[G^*],i \le j \le \ell g(\bar Q)$ \nl
1569: $p \in P_j,p \restriction i \in G^*$ (possibly more).
1570: \ermn
1571: We would like to find $G' \subseteq P^N_j/G^*$ generic over $N_0$ such that
1572: $q_0$ and $q_1$ both forces that it has an upper bound in $P_j/
1573: {\underset\tilde {}\to G_{P_i}}$. If $j=i+1$ this means
1574: $G' \subseteq {\underset\tilde {}\to Q_i}[G^*]$ is generic over
1575: $N_0$ such that $q_0,q_1$ both force that $G'$ has an upper bound in
1576: ${\underset\tilde {}\to Q_i}[G_{P_i}]$. \nl
1577: It is natural to demand $G' \in N$, otherwise the two possible generic
1578: extensions (for $q_0$ and $q_1$) become not related. For the case $j=i+1$,
1579: a ``weak diamond medicine" should help us. But we need it for every $j$,
1580: naturally we prove it by induction on $j$, and the successor case can be
1581: reduced to the case $j=i+1$.
1582: \mn
1583: But to continue in a limit we need $G' \in N$ and more: for some intermediate
1584: $N_1,N_0 \in N_1 \in N$ also $\dsize \bigwedge_\ell [q_\ell \Vdash
1585: N_1[{\underset\tilde {}\to G_{P_i}}] \cap V = N_1]$.
1586: So the clubs of elementary submodels which $q_0,q_1$ induce on
1587: $\{M \prec N:M \in N\}$ should have non-trivial intersection. This is a
1588: major point and it has always appeared in some form. Here the medicine
1589: against $\bigotimes_2$ should help, in some way there will be many possible
1590: $N_1$'s; but its help has a price: we have to carry it during the induction.
1591: On the other hand the models playing the role of $N_1$ may change, we may
1592: ``consume it and discard it".
1593:
1594: Note that the discussion is on two levels. Necessary limitations of universes
1595: with CH on the one hand, \ub{and} how we try to carry the inductive proof
1596: on appropriate iterations on the other hand; the connection though is
1597: quite tight.
1598: \mn
1599: So we shall try for $j \in \ell g(\bar Q)
1600: \cap N_0$ to extend the situation with $i$ being replaced by $j$ while
1601: $G^*$ is being increased to $G^{**}$. We shall prove by induction
1602: suitable facts, with $G^{**}$ the
1603: object we are really interested in. \nl
1604: We are given $q_1,q_2 \in P_i$ and would like to find suitable
1605: $q'_1,q'_2 \in P_j$ such that $q'_\ell \restriction i = q_\ell$ (otherwise
1606: in limit why is there an upper bound?)
1607: \sn
1608: So the real action occur for $j$ limit, so we choose $\zeta_n \in N \cap
1609: [i,j)$ such that $\zeta_0=i,\zeta_n < \zeta_{n+1}$ (sometimes better to have
1610: $i$ and each $\zeta_n$ non-limit) and $\dbcu_{n < \omega} \zeta_n = \sup
1611: (j \cap N)$.
1612: \mn
1613: You can think of:
1614: \mr
1615: \item "{{}}" in each case of limit $j$, proving the inductive statement,
1616: we choose a ``surrogate" for $N$ called $N_1$, during the induction it
1617: serves like $N$ in the limit dealing with $\zeta_0,\zeta_1,\ldots$ using
1618: the induction hypothesis on $N_1$ we get $G^{**}$ which may not be in $N_1$
1619: but is in $N$.
1620: \endroster
1621: \mn
1622: So we try to choose by induction on $n,q_{0,n},q_{1,n},G^*_n$ such that:
1623: $q_{\ell,n} \in P_{\zeta_n}$ is $(N,P_{\zeta_n})$-generic, $q_{\ell,0} =
1624: q_\ell,q_{\ell,n+1} \restriction \zeta_n = q_{\ell,n},G^*_n \in N_1,
1625: G^*_n \subseteq P_{\zeta_n} \cap N$ is generic over $N$ and
1626: $q_{\ell,n} \Vdash ``{\underset\tilde {}\to G_{P_{\zeta_n}}} \cap N = G^*_n$.
1627: The construction of the $G^*_n$ should use little information on the actual
1628: $q_{\ell,n}$ so that the choices of the $G^*_n$ can be carried say inside
1629: $N_1$ so that $\langle G^*_n:n < \omega \rangle \in N$. In fact several
1630: models will play a role like $N_1$. \nl
1631: By the proof of the preservation of ${}^\omega \omega$-bounding we can choose
1632: some $N_1$ and demand ``$q_{\ell,n}$ gives to each $P_{\zeta_n}$-name of an
1633: ordinal ${\underset\tilde {}\to \tau_n} \in N_1$, only finitely many
1634: possibilities".
1635:
1636: Now how does $(< \omega_1)$-proper help? \nl
1637: We can assume in the beginning that
1638: $\langle N_{1,\gamma}:\gamma \in A \rangle \in N$ is $\prec$-increasing
1639: continuously, $N_0 \prec N_{1,\gamma} \prec N,\langle N_{1,\gamma}:\gamma \le
1640: \beta \rangle \in N_{\beta +1}$ with $A =: (j+1) \cap N \backslash i$
1641: and assume $q_\ell$
1642: is $(N_{1,\gamma},P_i)$-generic for $\gamma \in A$ (similarly for $q'_0,
1643: q'_1,j$ in the conclusion) and demand $q_{\ell,n}$ is $(N_{1,\gamma},
1644: P_{\zeta_n})$-generic for $n < \omega$ and $\gamma \in A \backslash \zeta_n$.
1645: We are ignoring several points
1646: including how the induction change and having $\ell < 2$ rather than
1647: $\ell < n (< \omega)$ which complicates life.
1648: \sn
1649: How does ``$Q \times Q$ is proper" help? \nl
1650: We demand things like ``$(q_0,q_1)$ is $(N_1,P_i \times P_i)$-generic"
1651: so this gives
1652: many common $N_1$'s, but to preserve this we need more complicated situations.
1653: Instead of a ``tower" of models of countable length, we have a finite tower
1654: of models (say of length 5) where on the bottom we are computing $G^{**} \cap
1655: P_{\zeta_n}$ and as we go up less and less is demanded. \nl
1656: The medicine in the present work is ${\frak p}$-properness
1657: where ``$Q$ is ${\frak p}$-proper"
1658: say that if $Y$ is a large family of $M \prec N$ and $p \in Q \cap N,Q \in
1659: N$ then for some $q$ we have $p \le q$ and $q$ is $(N,Q)$-generic and
1660: $q \Vdash ``\{M \in Y:M[{\underset\tilde {}\to G_Q}] \cap V=M\}$ is large".
1661: (The idea of the finite tower is retained in the proof). This is quite
1662: obvious in hindsight.
1663: \mn
1664: Why is it important to be inside $N$? Otherwise, we could
1665: forget about $N$ and we have $q_0,q_1$. We know they have a common candidate
1666: but we need to increase them to know if and in limit by the regular having
1667: bounded in $P_i$. \nl
1668: ``We need a real not a name of a real." \nl
1669: \sn
1670: Note that a sufficient condition for ${\frak p}$-properness for $Q$, if ${\frak p}$ is
1671: standard, is homogeneity.
1672: \bn
1673: \ub{Notation}: ${\frak p}$ denote a reasonable parameter.
1674: \sn
1675: We thank Todd Eisworth for many corrections; he has continued this work.
1676: \newpage
1677:
1678: \head {\S1 Preservation of not adding reals} \endhead \resetall \sectno=1
1679: % \resetall
1680: \bigskip
1681:
1682: \definition{\stag{1.1} Definition} We say ${\frak p} = (\bar \chi,\bar R,
1683: \bar{\Cal E},\bar D) = (\bar \chi^{\frak p},\bar R^{\frak p},
1684: \bar{\Cal E}^{\frak p},\bar D^{\frak p})$ is a reasonable parameter if
1685: for some ordinal $\alpha^*$ called $\ell g({\frak p})$ we have:
1686: \mr
1687: \item "{$(a)$}" $\bar \chi = \langle \chi_\alpha:\alpha < \alpha^* \rangle,
1688: \chi_\alpha$ a regular cardinal, ${\Cal H}((\dbcu_{\beta < \alpha}
1689: \chi_\beta)^+) \in {\Cal H}(\chi_\alpha)$
1690: \sn
1691: \item "{$(b)$}" $\bar R = \langle R_\alpha:\alpha < \alpha^* \rangle,
1692: R_\alpha \in {\Cal H}(\chi_\alpha)$; \nl
1693: we could have asked ``$R_\alpha$ a relation on ${\Cal H}(\chi_\alpha)$", no
1694: real difference for our purpose; in a sense it codes a club of
1695: $[{\Cal H}(\chi_\alpha)]^{\le \aleph_0}$.
1696: \sn
1697: \item "{$(c)$}" $\bar{\Cal E} = \langle
1698: {\Cal E}_\alpha:\alpha < \alpha^* \rangle$ where
1699: ${\Cal E}_\alpha \subseteq [{\Cal H}(\chi_\alpha)]^{\le \aleph_0}$ is
1700: stationary
1701: \sn
1702: \item "{$(d)$}" $\bar D = \langle D_\alpha:\alpha < \alpha^* \rangle,
1703: D_\alpha$ is a function with domain ${\Cal E}_\alpha,a \in {\Cal E}_\alpha
1704: \Rightarrow D_\alpha(a)$ is a pseudo-filter on $a$, i.e.
1705: $D_\alpha(a)$ is a family of subsets of $a$ closed under supersets, non-empty
1706: if $\alpha > 0$ \nl
1707: (and
1708: let $D^-_\alpha(a) = (D_\alpha(a))^- = {\Cal P}(a) \backslash D_\alpha(a)$)
1709: \sn
1710: \item "{$(e)$}" for $\alpha < \alpha^*$ we let
1711: ${\frak p}^{[\alpha]} =: \langle \bar \chi \restriction
1712: \alpha,\bar R \restriction (\alpha + 1),\bar{\Cal E} \restriction \alpha,
1713: \bar D \restriction \alpha \rangle$, so it belongs
1714: to ${\Cal H}(\chi_\alpha)$. \nl
1715: [Why $\bar R \restriction (\alpha +1)$? This makes it an easy demand on
1716: ${\Cal E}_\alpha:N \in {\Cal E}_\alpha \Rightarrow R_\alpha \in N$].
1717: \sn
1718: \item "{$(f)$}" if $a \in {\Cal E}_\alpha$, \ub{then} for some countable
1719: $N \prec ({\Cal H}(\chi_\alpha),\in)$ we have: \nl
1720: $a$ is the universe of $N$, so we may write
1721: $D_\alpha(N)$ instead of $D_\alpha(a)$ and $N \in {\Cal E}_\alpha$ instead of
1722: $|N| \in {\Cal E}_\alpha$
1723: \sn
1724: \item "{$(g)$}" if $\alpha < \alpha^*$ and $N \in {\Cal E}_\alpha$, then
1725: ${\frak p}^{[\alpha]} \in N$
1726: \sn
1727: \item "{$(h)$}" for $N \in {\Cal E}_\alpha$ and $X \subseteq N$ we have: \nl
1728: $X \in D_\alpha(N)$ iff $(\dbcu_{\beta < \alpha} {\Cal E}_\beta) \cap X \cap
1729: N \in D_\alpha(a)$
1730: \sn
1731: \item "{$(i)$}" if $N \in {\Cal E}_\alpha,X \in D_\alpha(N),\beta \in
1732: \alpha \cap N$ and $y \in N \cap {\Cal H}(\chi^{\frak p}_\beta)$,
1733: \ub{then} for some $M \in {\Cal E}_\beta \cap X$ we have $X \cap M \in
1734: D_\beta(M)$ and $y \in M$.
1735: \endroster
1736: \enddefinition
1737: \bigskip
1738:
1739: \remark{\stag{1.1A} Remark} 1) Note that
1740: $\langle {\Cal E}_\alpha:\alpha < \ell g({\frak p}) \rangle$ are
1741: pairwise disjoint by clause (g) (and clause (e)) so $D(N)$ can be well defined
1742: as $D_\alpha(N)$ for the unique $\alpha$ such that $N \in {\Cal E}_\alpha$).
1743: \nl
1744: 2) Clearly only $D_\alpha(N) \cap {\Cal P} (\dbcu_{\beta < \alpha}
1745: {\Cal E}_\beta)$ matters. \nl
1746: 3) Note that the most natural case is ``$D(N)$ is $\{X \subseteq N:X \ne
1747: \emptyset \text{ mod } D^{\text{fil}}\}$" for some filter $D^{\text{fil}}$ on
1748: $N$. \nl
1749: 4) Natural cases are: $D_\alpha(a)$ is a filter on $a$, and
1750: $D_\alpha(a) = \{b \subseteq a:a \backslash b \notin D'_\alpha(a)\}$ for a
1751: filter $D'_\alpha(a)$ on $a$; we say $D_\alpha$ is dual to $D'_\alpha$ or
1752: $D_\alpha = (D'_\alpha)^+$. \nl
1753: 5) We may add in clause (i) of \scite{1.1} that some $X' \in X \cap M$ belongs
1754: to $N \cap D_\beta(M)$. No beginning harm but not necessary at present.
1755: \endremark
1756: \bigskip
1757:
1758: \definition{\stag{1.2} Definition} 1) In \scite{1.1} we say $\bar D$
1759: is standard if for every $\alpha < \alpha^*$
1760: and $N \in {\Cal E}_\alpha$ we have
1761:
1762: $$
1763: \align
1764: D_\alpha(N) = \biggl\{ X \subseteq N:&\text{ for every } \gamma \in N \cap
1765: \alpha \text{ and} \\
1766: &y \in N \cap \cup\{{\Cal H}(\chi_\beta):\beta \in \alpha \cap N\}
1767: \text{ for some} \\
1768: &\beta \in N \cap \alpha \backslash \gamma \text{ and} \\
1769: &M \in X \cap {\Cal E}_\beta \text{ we have } y \in M, \\
1770: &X \cap M \in D_\beta(M) \biggr\}.
1771: \endalign
1772: $$
1773: \mn
1774: 2) We say ${\frak p}$ is simple if $\alpha \le \beta < \alpha^* \Rightarrow
1775: \alpha \le_{\frak p} \beta$ where we let $\alpha \le_{\frak p} \beta$ be
1776: the following partial order on $\alpha^*:\alpha \le_{\frak p} \beta$ iff
1777: $\alpha \le \beta < \alpha^* = \ell g({\frak p})$ and we have
1778: $N \in {\Cal E}^\beta \and \alpha \in N \Rightarrow M =: N \cap
1779: {\Cal H}(\chi^{\frak p}_\alpha) \in {\Cal E}^{\frak p}_\alpha$ and
1780: $N \in {\Cal E}^\beta \and \alpha \in N \and Y \in D_\beta(N) \Rightarrow
1781: Y \cap \dbcu_{\gamma < \alpha} {\Cal E}^{\frak p}_\gamma \in
1782: D_\alpha(M)$.
1783: \nl
1784: 3) We say ${\frak p}$ is standard if (${\frak p}$ is a reasonable parameter
1785: such that) $\bar D^{\frak p}$ is standard. \nl
1786: 4) If $N \prec ({\Cal H}(\chi),\in)$ and $N \cap {\Cal H}(\chi^{\frak p}
1787: _\alpha) \in {\Cal E}_\alpha$, (hence $\alpha \in N,{\frak p} \restriction
1788: \alpha \in N,R^{\frak p}_\alpha \in N$), \ub{then} we let
1789: $D_\alpha(N) = D^{\frak p}_\alpha (N)$ be
1790: $D^{\frak p}_\alpha(N \cap {\Cal H}(\chi^{\frak p}_\alpha))$.
1791: \enddefinition
1792: \bn
1793: \ub{\stag{1.3} Convention}: If $\bar D = \bar D^{\frak p}$ is standard,
1794: we may omit it. If ${\frak p}$ clear from the content, we may write
1795: ${\Cal E}_\alpha$ instead ${\Cal E}^{\frak p}_\alpha$.
1796: \bigskip
1797:
1798: \definition{\stag{1.4} Definition} 1) We say that ${\frak p}$ is a winner or
1799: a $\Game$-winner if:
1800: \mr
1801: \item "{{}}" for every $\alpha < \ell g({\frak p}),\alpha > 0$ and $N \in
1802: {\Cal E}^{\frak p}_\alpha$, in the game \nl
1803: $\Game_\alpha(N) = \Game_\alpha(N,{\frak p})$ (defined below)
1804: \sn
1805: \item "{{}}" the chooser player has a winning strategy where:
1806: \ermn
1807: 2) $\Game_\alpha(N,{\frak p})$ is the following game:
1808: \mr
1809: \item "{{}}" a play lasts $\omega$ moves, in the $n$-th move
1810: \sn
1811: \item "{{}}" \ub{the challenger} $\quad$ choose $X_n \in D_\alpha(N)$
1812: such that $m < n \Rightarrow X_n \subseteq X_m$
1813: \sn
1814: \item "{{}}" \ub{the chooser} chooses
1815: $\quad M_n \in X_n$ and $Y_n \subseteq M_n
1816: \cap X_n,Y_n \in D(M_n) \cap N$
1817: \sn
1818: \item "{{}}" \ub{the challenger} $\quad$ chooses $Z_n \subseteq Y_n$
1819: such that $Z_n \in D(M_n)$.
1820: \ermn
1821: In the end the chooser wins if $\cup\{\{M_n\} \cup Z_n:n < \omega\} \in
1822: D_\alpha(N)$. \nl
1823: 3) Assume $N \in N' \prec ({\Cal H}(\chi),\in)$ and ${\frak p} \restriction
1824: \alpha \in N'$, and, of course, $N \prec N'$ are countable. The game
1825: $\Game'_\alpha(N,N',{\frak p})$ is defined similarly but during the $n$-th
1826: move, we demand that all the chosen objects belong to $N'$ and in
1827: the end the chooser also
1828: chooses $X'_n \subseteq X_n,X'_n \in D(N) \cap N'$ and the challenger in the
1829: next move has to satisfy $X_{n+1} \subseteq X'_n$. Omitting $N'$ we mean:
1830: for any such $N'$ the demand holds. \nl
1831: 4) ${\frak p}$ is a non-$\Game$-loser if for $\alpha < \ell g({\frak p}),
1832: \alpha > 0,N \in {\Cal E}_\alpha$ the challenger has no winning strategy
1833: in $\Game_\alpha(N,{\frak p})$. \nl
1834: 5) ``$\Game'_\alpha$-winner" or ``non-$\Game'_\alpha$-loser" means we
1835: (in part (1) or part (4)) use $\Game'_\alpha(N,{\frak p})$.
1836: We say that ``the chooser/challenger wins the game $\Game_\alpha(N)$" if he
1837: has a winning strategy and so ``the chooser/challenger does not win the
1838: game $\Game_\alpha(N)$" says the negation. (Similarly for the other games in
1839: this paper). \nl
1840: 6) Omitting $\alpha$ means for every $\alpha < \ell g({\frak p})$.
1841: \enddefinition
1842: \bigskip
1843:
1844: \demo{\stag{1.5} Observation} 1) If ${\frak p}$ is a reasonable parameter with
1845: the standard $\bar D$, \ub{then} ${\frak p}$ is a winner. \nl
1846: 2) If ${\frak p}$ is a $\Game_\alpha$-winner \ub{then} ${\frak p}$ is a
1847: $\Game'_\alpha$-winner; if ${\frak p}$ is a $\Game$-winner, \ub{then}
1848: ${\frak p}$ is a $\Game'$-winner; similarly for a non-loser.
1849: \enddemo
1850: \bigskip
1851:
1852: \demo{Proof} Straight.
1853: \enddemo
1854: \bigskip
1855:
1856: \definition{\stag{1.5A} Definition} Assume ${\frak p}$ is a reasonable
1857: parameter, $N \in {\Cal E}^{\frak p}_\alpha,y \in N$ and
1858: $P \in N$ is a forcing notion. We let
1859: ${\Cal M}_P[{\underset\tilde {}\to G_P},N,y]
1860: =: \{M \in N:P,y \in M \text{ and } {\underset\tilde {}\to G_P}
1861: \cap M \text{ is a subset of } P \cap M \text{ generic over } M\}$.
1862: If $P$ is clear from the context, we may omit it; note that ${\Cal M}_P
1863: [G,N,y] = {\Cal M}_P[G \cap N,N,y]$ so we may write $G \cap N$ instead of $G$.
1864: If $y = \emptyset$ we may omit it.
1865: \enddefinition
1866: \bigskip
1867:
1868: \definition{\stag{1.6} Definition} We say $\bar Q$ is
1869: ${\frak p}-NNR^0_{\aleph_0}$ iteration if:
1870: \mr
1871: \item "{$(a)$}" $\bar Q = \langle P_i,{\underset\tilde {}\to Q_i}:i < j(*)
1872: \rangle$ is a CS iteration of proper forcing notions which belongs to
1873: ${\Cal H}(\chi^{\frak p}_0)$, even ${\Cal P}(P_{j(*)}) \in {\Cal H}
1874: (\chi^{\frak p}_0)$
1875: \sn
1876: \item "{$(b)$}" forcing with $P_{j(*)} = \text{ Lim}(\bar Q)$ does not add
1877: reals
1878: \sn
1879: \item "{$(c)$}" [long properness]
1880: if $i \le j \le j(*),\alpha < \ell g({\frak p}),
1881: N \in {\Cal E}_\alpha,\{i,j,\bar Q\} \in N$, the condition $q \in P_i$
1882: is $(N,P_i)$-generic forcing ${\underset\tilde {}\to G_{P_i}} \cap N = G$
1883: and $p \in P_j \cap N,p \restriction i \in G$ and $Y \subseteq
1884: {\Cal M}_{P_i}[G,N,y]$ where $y = \langle \bar Q,i,j \rangle$ and $Y \in
1885: D_\alpha(N)$ \nl
1886: (note that from $\bar Q,G$ the ordinal $i$ is reconstructible);
1887:
1888: \ub{then} there are $G'',q'$ such that:
1889: {\roster
1890: \itemitem{ $(\alpha)$ } $q' \in P_j,p \le q'$ and $q \le q' \restriction i$
1891: \sn
1892: \itemitem{ $(\beta)$ } $q'$ is $(N,P_j)$-generic
1893: \sn
1894: \itemitem{ $(\gamma)$ } $q'$ forces ${\underset\tilde {}\to G_{P_j}} \cap
1895: N = G''$
1896: \sn
1897: \itemitem{ $(\delta)$ } $Y \cap {\Cal M}_{P_j}[G'',N,y] \in D_\alpha(N)$
1898: \endroster}
1899: (in \S2 this is called ${\frak p}$-proper)
1900: \sn
1901: \item "{$(d)$}" [anti-w.d.] assume
1902: $i \le j \le j(*),\alpha < \ell g({\frak p}),N_0 \in
1903: N_1 \in {\Cal E}_\alpha,N_0 \in \dbcu_{\beta < \alpha} {\Cal E}_\beta$,
1904: otp$(N_0 \cap [i,j)) < \alpha$ and \footnote{so naturally
1905: $\ell g({\frak p}) = \omega_1$} $n < \omega$, for $\ell < n$ we have
1906: $q_\ell \in P_i$ is $(N_1,P_i)$-generic and $q_\ell$ forces
1907: ${\underset\tilde {}\to G_{P_i}} \cap N_1 = G^\ell,\dsize \bigwedge_{\ell <n}
1908: [G^\ell \cap N_0 = G^*]$ where
1909: $G^* \subseteq P_i \cap N_0$ is generic over $N_0$ and
1910: $Y =: \dbca_{\ell <n} {\Cal M}[G^\ell,N_1] \in D_\alpha(N)$ and $p \in P_j
1911: \cap N_0,p \restriction i \in G^*$. \ub{Then} for some $G^{**} \subseteq
1912: P_j \cap N_0$ generic over $N_0$ we have $p \in G^{**} \in N_1$ and
1913: $\dsize \bigwedge_{\ell < n} \,\,\dsize \bigvee_{q \in G^\ell}[q \Vdash
1914: ``G^{**}$ has an upper bound in $P_j/{\underset\tilde {}\to G_{P_i}}"]$.
1915: \endroster
1916: \enddefinition
1917: \bigskip
1918:
1919: \remark{Remark} We may like to phrase clause (c) as a condition on each
1920: ${\underset\tilde {}\to Q_i}$, for this see Definition \scite{2.6},
1921: \scite{2.8}, \scite{2.14}; this is a slight loss if we deal with the case
1922: $i < j,i$ non limit $\Rightarrow P_j/P_i$ is proper. As no need arise here
1923: we ignore this.
1924: \endremark
1925: \bigskip
1926:
1927: \proclaim{\stag{1.7} Main Claim} Assume $\bar Q$ is a CS iteration,
1928: $\bar Q \in {\Cal H}(\chi^{\frak p}_0)$ and ${\Cal P}(\text{Lim }
1929: \bar Q) \in {\Cal H}(\chi^{\frak p}_0)),{\frak p}$ a reasonable parameter
1930: of length $\omega_1,\delta = \ell g(\bar Q)$ is a limit ordinal and
1931: for every $\alpha < \delta,\bar Q \restriction \alpha$ is
1932: a ${\frak p}-NNR^0_{\aleph_0}$ iteration and ${\frak p}$ is a $\Game$-winner.
1933:
1934: \ub{Then} $\bar Q$ is a ${\frak p}-NNR^0_{\aleph_0}$ iteration.
1935: \endproclaim
1936: \bigskip
1937:
1938: \demo{Proof}
1939: \sn
1940: \ub{Proof of clause (a) of Definition \scite{1.6}}:
1941:
1942: Trivial.
1943: \mn
1944: \ub{Proof of clause (b) of Definition \scite{1.6}}:
1945:
1946: Follows from clause (d) of Definition \scite{1.6} proved below.
1947: \mn
1948: \ub{Proof of clause (d)}:
1949:
1950: Let $i,j,\alpha,N_0,N_1,n,q_0,\dotsc,q_{n-1},G^\ell,G^*,p$ be as in the
1951: assumptions of clause (d). Let $\alpha' = \text{ otp}(N_0 \cap [i,j))$, and
1952: $\alpha' < \omega_1$ so $\alpha' \in N_1$.
1953: If $j < j(*)$ use ``$\bar Q \restriction j$ is a ${\frak p}-NNR^0_{\aleph_0}$
1954: iteration", so assume $j = j(*)$; if $i=j$ the conclusion is trivial so
1955: assume $i<j$. Let $i_m \in N_0 \cap j$ be such that $i_0 = i,i_m < i_{m+1}$
1956: and $\langle i_m:m < \omega \rangle \in N_1$ and $\dbcu_{m < \omega} i_m
1957: = \sup(N_0 \cap j)$.
1958: Choose $M_\ell$ for $\ell < 5$ such that $y^* =: \{i,j,\alpha,\alpha',
1959: \bar Q,N_0,\langle i_m:m < \omega \rangle\} \in M_\ell \in
1960: {\Cal E}_{\alpha'} \cap N_1 \cap \dbca_{\ell < n} {\Cal M}[G^\ell,N_1],
1961: M_0 \in M_1 \in M_2 \in M_3 \in M_4$ and $\dbca_{\ell < n} {\Cal M}
1962: [G^\ell,M_0,y^*] \in D_\alpha(M_0)$.
1963:
1964: Now for $\ell < n$ we can choose $q'_\ell \in G^\ell \cap M_4$ which forces
1965: (for $P_{i_0} = P_i$) a value to ${\underset\tilde {}\to G_{P_{i_0}}} \cap
1966: M_3$ which necessarily is $G^\ell \cap M_3$ and necessarily $q'_\ell \le
1967: q_\ell$ and, of course, $q'_\ell$ is $(M_k,P_{i_0})$-generic forcing
1968: ${\underset\tilde {}\to G_{P_{i_0}}} \cap M_k = G^\ell \cap M_k$ for $k =
1969: 0,1,2,3$ and ${\underset\tilde {}\to G_{P_{i_0}}} \cap N_0 = G^*$.
1970:
1971: Let $\langle {\Cal I}^*_m:m < \omega \rangle \in M_0$ list
1972: the maximal antichains of $P_j$ that belongs to $N_0$. Now we choose
1973: by induction on $m < \omega$, the objects $r_m,G^*_m,p_m,n_m,G^\ell_m$
1974: (for $\ell < n_m$) and $y_m$ such that:
1975: \mr
1976: \item "{$(a)$}" $r_m \in P_{i_m} \cap M_4$
1977: \sn
1978: \item "{$(b)$}" Dom$(r_m) \subseteq [i,i_m)$
1979: \sn
1980: \item "{$(c)$}" $r_{m+1} \restriction i_m = r_m$
1981: \sn
1982: \item "{$(d)$}" $q'_\ell \cup r_m (\in P_{i_m})$ is $(M_k,P_{i_m})$-generic
1983: for $k=0,1,2,3$ and is $(N_0,P_{i_m})$-generic \nl
1984: (note that $q'_\ell,r_m$ have disjoint domains)
1985: \sn
1986: \item "{$(e)$}" for every predense subset ${\Cal I}$ of $P_{i_m}$ which
1987: belongs to $M_2$, for some finite ${\Cal J} \subseteq {\Cal I} \cap M_2$
1988: the set ${\Cal J}$ is predense above $q'_\ell \cup r_m$ for each $\ell < n$
1989: \sn
1990: \item "{$(f)$}" $n_m < \omega$ and for $\ell < n_m$ we have: \nl
1991: $G^\ell_m$ is a subset of $P_{i_m} \cap M_0$ generic over $M_0,G^\ell_m \in
1992: M_1$
1993: \sn
1994: \item "{$(g)$}" if $\ell < n_{m+1}$ then
1995: $G^\ell_{m+1} \cap P_{i_m} \in \{G^k_m:k < n_m\}$
1996: \sn
1997: \item "{$(h)$}" $n_0 = n,G^\ell_0 = G^\ell \cap M_0$
1998: \sn
1999: \item "{$(i)$}" $q'_\ell \cup r_m \Vdash_{P_{i_m}}
2000: ``{\underset\tilde {}\to G_{P_{i_m}}} \cap M_0 \in \{G^\ell_m:\ell < n_m\}"$
2001: \sn
2002: \item "{$(j)$}" $G^*_m$ is a subset of $P_{i_m} \cap N_0$ generic over
2003: $N_0$
2004: \sn
2005: \item "{$(k)$}" $G^*_m \subseteq G^\ell_m$ for $\ell < n_m$
2006: \sn
2007: \item "{$(l)$}" $p_m \in P_j \cap N_0,p_m \restriction i_m \in G^*_m,
2008: p_{m+1} \in {\Cal I}^*_m,p_0 = p,p_m \le p_{m +1}$
2009: \sn
2010: \item "{$(m)$}" $Y_m =: \dbca_{\ell < n_m} {\Cal M}[G^\ell_m,M_0,y^*] \in
2011: D_{\alpha'}(M_0)$ where
2012: $y^* =: \{N_0,\langle i_k:k < \omega \rangle,\bar Q,i,j\}$.
2013: \ermn
2014: \ub{Why is this sufficient?}
2015:
2016: During the construction above we choose inductively members of $M_4$ and
2017: all the parameters used are from $M_4$, so if we choose
2018: a well ordering $<^*$ of $M_4$ and always choose the $<^*$-first object the
2019: construction is determined. Clearly there is such $<^* \in N_1$. Now
2020: \mr
2021: \item "{$(\alpha)$}" $r = \dbcu_m r_m$ (i.e. the unique
2022: $r \in P_j$ satisfying $m < \omega \Rightarrow r \restriction i_m
2023: = r_m$) belongs to $P_j$ and to $N_1$,
2024: \sn
2025: \item "{$(\beta)$}" $G^{**} = \{p' \in P_j \cap N_0:
2026: \dsize \bigvee_{m < \omega}[p' \le p_m]\}$ belongs to $M_4$ and is a subset of
2027: $P_j \cap N_0$ generic over $N_0$, \nl
2028: (by the choice of $\langle {\Cal I}^*_m:m < \omega \rangle$ and clause (k))
2029: and $q'_\ell \cup r$ is above $G^{**}$ \nl
2030: (in $P_j$).
2031: \ermn
2032: So we are done.
2033: \enddemo
2034: \bn
2035: \ub{Why can we carry out the construction?}
2036:
2037: For $m=0$ there is no problem. So assume we have it for $m$ and we shall
2038: choose for $m+1$.
2039: \bn
2040: \ub{Stage A}: Choose $p_{m+1} \in N_0 \cap {\Cal I}^*_m$ such that
2041: $p_m \le p_{m+1}$ and $p_{m+1} \restriction i_m \in G^*_m$. \nl
2042: No problem.
2043: \bn
2044: \ub{Stage B}: Choose $G^*_{m+1} \subseteq P_{i_{m+1}} \cap N_0$ generic over
2045: $N_0,G^*_m \subseteq G^*_{m+1} \in M_0,p_{m+1} \restriction i_{m+1} \in
2046: G^*_{m+1}$ and $\dsize \bigwedge_{\ell < n_m} \dsize \bigvee_{r \in G^\ell_m}
2047: [r \Vdash_{P_{i_m}} ``G^*_{m+1}$ has an upper bound in $P_{i_{m+1}}/
2048: {\underset\tilde {}\to G_{P_{i_m}}}]$.
2049: \mn
2050: This is easy by ``$\bar Q \restriction i_{m+1}$ is a
2051: ${\frak p}-NNR^0_{\aleph_0}$ iteration" applied with
2052: $i_m,i_{m+1},\alpha',p_{m+1} \restriction i_m,G^*_m,\langle
2053: G^\ell_m:\ell < n_m \rangle,N_0,M_0$ here standing for $i,j,\alpha,p,G^*,
2054: \langle G^\ell:\ell < n \rangle,N_0,N_1$ there (i.e. we use clause (d) of the
2055: Definition \scite{1.6}); we are using otp$(N_0 \cap [i_m,i_{m+1})) <
2056: \text{otp}(N_0 \cap [i_m,j)) = \alpha'$.
2057: \bn
2058: \ub{Stage C}:
2059:
2060: Now ${\underset\tilde {}\to G_{P_{i_m}}}
2061: \cap M_1$ is a $P_{i_m}$-name of an object from $V$ (as $P_{i_m}$ is proper
2062: not adding reals), so ${\Cal I} =: \{p \in P_{i_m}:p
2063: \text{ forces a value to } {\underset\tilde {}\to G_{P_{i_m}}} \cap M_1
2064: (\in V)\}$ is a dense open subset of $P_{i_m}$ and ${\Cal I} \in M_2$. By
2065: clause (e) in the induction hypothesis there is a finite
2066: ${\Cal J} \subseteq {\Cal I}
2067: \cap M_2$ such that: $\ell < n \Rightarrow {\Cal J}$ is predense above
2068: $q'_\ell \cup r_m$. Without loss of generality ${\Cal J}$ is minimal. Let
2069: $n_{m+1} = |{\Cal J}|$.
2070: \sn
2071: Let ${\Cal J} = \{p^\ell_m:\ell < n_{m+1}\}$ let $p^\ell_m \Vdash
2072: ``{\underset\tilde {}\to G_{P_{i_m}}} \cap M_1 = H^\ell_m"$. So $H^\ell_m \in
2073: M_2$, and as ${\Cal J}$ is minimal, $H^\ell_m \cap M_0 \in \{G^\ell_m:\ell <
2074: n_m\}$ so let $H^\ell_m \cap M_0 = G^{h(\ell)}_m$ where
2075: $h:n_{m+1} \rightarrow n_m$.
2076:
2077: Let $Y =: \dbca_{\ell < m_n} {\Cal M}[G^\ell_m,M_0,y^*] \in
2078: D_{\alpha'}(M_0)$. \nl
2079: Now we choose by induction on $\ell \le n_{m+1}$ a condition
2080: $r^\ell_m \in M_1$ such that:
2081: \mr
2082: \item "{$(\alpha)$}" $r^\ell_m \in P_{i_{m+1}} \cap M_1,r^\ell_m \restriction
2083: i_m \in H^\ell_m$
2084: \sn
2085: \item "{$(\beta)$}" $r^\ell_m$ is $(M_0,P_{i_m})$-generic and force a value
2086: to ${\underset\tilde {}\to G_{P_{i_m}}} \cap M_0$ called $G^\ell_{m+1}$
2087: \sn
2088: \item "{$(\gamma)$}" $r^\ell_m$ is above $G^*_{m+1}$ (chosen in the previous
2089: stage), moreover above $p^{h(\ell)}_m$
2090: \sn
2091: \item "{$(\delta)$}" $Y \cap \dbca_{k < \ell} {\Cal M}[G^k_{m+1},M_0,y^*] \in
2092: D_{\alpha'}(M_0)$.
2093: \ermn
2094: For the induction step, apply clause (c) in Definition \scite{1.6} (as
2095: $\bar Q \restriction i_{m+1}$ is a \nl
2096: ${\frak p}-NNR^0_{\aleph_0}$-iteration) with $i_m,i_{m+1},\alpha',M_0$, large
2097: enough member of $H^\ell_m,p^{h(\ell)}_m$, \nl
2098: $Y^\ell_m = Y \cap \dbca_{k < \ell}{\Cal M}[G^k_{m+1},M_0,y^*]$
2099: here standing for $i,j,\alpha,N,q,p,Y$ there, noting $Y^\ell_m \in M_1$.
2100: \bn
2101: \ub{Stage D}:
2102:
2103: We can choose $r_{m+1}$ as required such that $\{r^\ell_m:\ell < n_{m+1}\}$
2104: is predense over it by \cite[Ch.XVIII,2.6]{Sh:f} (can first do it for each
2105: $r^\ell_m$ separately and then put them together). \nl
2106: So we have finished proving clause (d).
2107: \bigskip
2108:
2109: \demo{\ub{Proof of clause (c)}}
2110:
2111: We prove this by induction on $\alpha$. \nl
2112: Let $i,j,\alpha,N,p,q,Y$ be as there. If $j < j(*)$ we can apply ``$Q
2113: \restriction j$ is a ${\frak p}-NNR^0_{\aleph_0}$ iteration", so \wilog \,
2114: $j = j(*)$. If $i=j$ the statement is trivial so assume $i<j$. Choose
2115: $i_n$ for $n < \omega$ such that $i_0 = i,i_n \in N \cap j,i_n < i_{n+1}$ and
2116: $\dbcu_{n < \omega} i_n = \sup(j \cap N)$.
2117: Let $\langle (y_n,\beta_n):n < \omega \rangle$ list the pairs $(y,\beta) \in
2118: N \times (\alpha \cap N)$ such that
2119: $y \in {\Cal H}(\chi^{\frak p}_\beta)$.
2120: Let {\bf St} be a winning strategy for the chooser in the game
2121: $\Game_\alpha(N)$. \nl
2122: Let $\langle {\Cal I}_n:n < \omega \rangle$ list the dense open subsets of
2123: $P_j$ which belongs to $N$.
2124:
2125: Now we choose by induction on $n < \omega$, the objects $q_n,
2126: {\underset\tilde {}\to p_m},{\underset\tilde {}\to M_n},
2127: {\underset\tilde {}\to Y_m}$ such that:
2128: \mr
2129: \item "{$(a)$}" $q_n \in P_{i_n},q_0 = q$
2130: \sn
2131: \item "{$(b)$}" $q_n$ is $(N,P_{i_n})$-generic
2132: \sn
2133: \item "{$(c)$}" $q_{n+1} \restriction i_n = q_n$
2134: \sn
2135: \item "{$(d)$}" ${\underset\tilde {}\to p_n}$ is a $P_{i_n}$-name of a member
2136: of $(P_j/{\underset\tilde {}\to G_{P_{i_n}}}) \cap N$
2137: \sn
2138: \item "{$(e)$}" ${\underset\tilde {}\to p_n}$ is forced to belong to
2139: ${\Cal I}_n$
2140: \sn
2141: \item "{$(f)$}" ${\underset\tilde {}\to M_n}$ is a $P_{i_n}$-name of a
2142: member of ${\Cal E}_{\beta_n} \cap N$
2143: \sn
2144: \item "{$(g)$}" if $G_j \subseteq P_j$ is generic over $V,q_n \in
2145: G_j,{\underset\tilde {}\to p_n}[G_j \cap P_{i_n}] \in G_j,M =
2146: {\underset\tilde {}\to M_n}[G_j]$, \ub{then}
2147: {\roster
2148: \itemitem{ $(\alpha)$ } $G_j \cap M$ is a subset of $Q \cap M$ generic
2149: over $M$
2150: \sn
2151: \itemitem{ $(\beta)$ } ${\Cal M}[G_j \cap M,M,y^*] \cap Y \in D_{\beta_n}
2152: [M]$
2153: \sn
2154: \itemitem{ $(\gamma)$ } ${\underset\tilde {}\to p_n}[G_j]$ belongs to $M$
2155: \endroster}
2156: \item "{$(h)$}" $\langle Y_m \cap {\Cal M}[{\underset\tilde {}\to G_{i_m}},
2157: {\underset\tilde {}\to M_m},y^*],{\underset\tilde {}\to Y_m},
2158: P_{i_m},{\underset\tilde {}\to M_m}:m \le n \rangle$ is forced by $q_n$ to
2159: be an initial segment of a play of the game $\Game_\alpha(N)$ in which
2160: the chooser uses the fixed winning strategy {\bf St}.
2161: \ermn
2162: The proof is straight by the induction hypothesis on $\beta$ and
2163: ``$\bar Q \restriction i_n$ is a ${\frak p}-NNR^0_{\aleph_0}$-iteration"
2164: remembering that ${\underset\tilde {}\to M_n},{\underset\tilde {}\to Y_n}$
2165: are $P_{i_n}$-names but of objects in $V$. \nl
2166: Alternatively, see more in the proof in \S4 or see below.
2167: \hfill$\square_{\scite{1.7}}$
2168: \enddemo
2169: \bigskip
2170:
2171: \remark{\stag{1.8} Remark} 1) We could have used the ``adding no reals" and
2172: clause (d) in the proof of clause (c) in order to weaken ``winner" to ``not
2173: loser"; also we can use $\Game'_\alpha(N,N',P)$, see \S4. Also
2174: the ``$N_0 \in \dbcu_{\beta < \alpha} {\Cal E}_\beta"$ can be replaced by
2175: $``N_0 \in {\Cal E}'_0"$ with ${\Cal E}'_0 \subseteq [{\Cal H}(\chi^{\frak p}
2176: _0)]^{\aleph_0}$ stationary; also we can put extra restrictions on
2177: $G^*$ (and $G^{**}$), e.g. ${\Cal M}[G^*,N_0,y^*]$ large. \nl
2178: 2) Of course, the use of $\langle \chi_\alpha:\alpha < \ell g({\frak p})
2179: \rangle$ is not really necessary, we could have used subsets of ${\Cal P}
2180: (P_{\ell g(\bar Q)})$, (so ${\Cal E}_\alpha$ is changed accordingly) as
2181: all the properties depend just on $N_\ell \cap {\Cal P}(P_{\ell g(\bar Q)})$.
2182: We feel the present way is more transparent. \nl
2183: 3) The proof of clause (c) being preserved can be applied to $\bar Q$
2184: satisfying (a) + (c) of Definition \scite{1.6} (so possibly adding reals),
2185: but then we have to replace $D_\alpha(N)$ by a definition of such pseudo
2186: filters with the winning strategy being absolute enough, e.g. for standard
2187: $\bar D$. \nl
2188: 4) We can also replace $D_\alpha(a)$ by a partial ordered set $L_\alpha(a)$
2189: and a function $v_a:L_\alpha(a) \rightarrow {\Cal P}(a)$ (i.e. by
2190: $(L_a,v_a)$). \nl
2191: 5) In the case we adopt remark (3), then remark (1) (non-losing) becomes less
2192: clear, as the universe changes during the proof. We may consider to
2193: weaken ``the chooser has a winning strategy" in $\Game'_\alpha(N)$
2194: (game depends on $\bar D$), e.g. to not losing in a game with finitely many
2195: boards, possibly splitting in each move (no real novelty in the proof), but
2196: it is not clear how interesting this is. But if $D_\alpha(N)$ is inductively
2197: defined as sums over an ultrafilter (which is preserved), it just seems
2198: that not losing is enough.
2199: \endremark
2200: \bigskip
2201:
2202: \definition{\stag{1.8A} Definition} $\bar Q$ a CS iteration will be called
2203: ${\frak p}$-proper if clauses (a) + (c) of Definition \scite{1.6} holds.
2204: \enddefinition
2205: \bn
2206: \ub{\stag{1.8B} Question}: Is this notion of interest in iterations adding
2207: reals?
2208: \bigskip
2209:
2210: \definition{\stag{1.9} Definition} Let $\kappa \in [2,\omega]$. We say
2211: that $\bar Q$ is a ${\frak p}-NNR^0_\kappa$-iteration if: from Definition
2212: \scite{1.6} we have:
2213: \mr
2214: \item "{$(a)$}" as there
2215: \sn
2216: \item "{$(b)$}" as there
2217: \sn
2218: \item "{$(c)$}" [long properness] as there
2219: \sn
2220: \item "{$(d)$}" [$\kappa$-anti w.d.] as in (d) there but
2221: $1+n < \kappa$ and $N_0 \in {\Cal E}_0$.
2222: \endroster
2223: \enddefinition
2224: \bigskip
2225:
2226: \proclaim{\stag{1.10} Main Claim} Assume $2 \le \kappa < \omega,
2227: \bar Q$ is a CS iteration, ${\Cal P}(\text{Lim}(\bar Q)) \subseteq
2228: {\Cal H}(\chi^{\frak p}_0),{\frak p}$ a reasonable parameter of length
2229: $\omega_1$ which is a non-$\Game'$-loser, $\delta = \ell g(\bar Q)$
2230: a limit ordinal and $i < \delta \Rightarrow \bar Q \restriction i$ is a
2231: ${\frak p}-NNR_\kappa$-iteration. \ub{Then} $\bar Q$ is a
2232: ${\frak p}-NNR^0_\kappa$-iteration.
2233: \endproclaim
2234: \bigskip
2235:
2236: \remark{Remark} 1) As it is just combining the proof of \scite{1.7} and
2237: \cite[Ch.XVIII,\S2]{Sh:f} (or \nl
2238: \cite[Ch.VIII,4.10]{Sh:b}-\cite[Ch.VIII,4.10]{Sh:f}) we elaborate less. \nl
2239: 2) See more in \S4.
2240: \endremark
2241: \bigskip
2242:
2243: \demo{Proof} Similar to the proof of \scite{1.7}, with some changes (as
2244: in \cite[Ch.VIII,\S4]{Sh:b}, \cite[Ch.VIII,\S4]{Sh:f},
2245: \cite[Ch.XVIII,2.10C]{Sh:f}). During the proof of clause (d) we add to
2246: clauses (a)-(m):
2247: \mr
2248: \item "{$(n)$}" $n_m$ is a power of 2, say $2^{n^*_m}$ and so we can rename
2249: $\{G^\ell_m:\ell < n_m\}$ as $\{G^\eta_m:\eta \in {}^{n^*_m}2\}$
2250: \sn
2251: \item "{$(o)$}"
2252: {\roster
2253: \itemitem{ $(\alpha)$ } $M_\eta \in M_1 \cap {\Cal E}_{\beta_\eta}$ for
2254: $\eta \in {}^{(n^*_m \ge)}2$ where $j_0 = \text{ otp}([i,j) \cap N_0)$ and if
2255: $\eta = \nu \char 94 <i>$ then $j_\eta = \text{ otp}([i,j) \cap M_\nu)$
2256: \sn
2257: \itemitem{ $(\beta)$ } $M_{<>} = N_0$
2258: \sn
2259: \itemitem{ $(\gamma)$ } $M_\eta \in M_{\eta \char 94 \langle 0 \rangle} \cap
2260: M_{\eta \char 94 \langle 1 \rangle}$
2261: \sn
2262: \itemitem{ $(\delta)$ } $\eta \triangleleft \nu_1 \in {}^{n^*_m}2 \and
2263: \eta \triangleleft \nu_2 \in {}^{n^*_m}2 \Rightarrow G^{\nu_1}_m \cap M_\eta
2264: = G^{\nu_2}_m \cap M_\eta$ so we call it $K^\eta_m$
2265: \sn
2266: \itemitem{ $(\varepsilon)$ } $M_{\eta \char 94 \langle 0 \rangle} =
2267: M_{\eta \char 94 \langle 1 \rangle}$ when $\eta \in {}^{n^*_m}2$ call it
2268: $N_\eta$
2269: \sn
2270: \itemitem{ $(\zeta)$ } $N_\eta \in {\Cal E}_{\ell g(\eta)}$ for
2271: $\eta \in {}^{(m^*_m >)}2$
2272: \sn
2273: \itemitem{ $(\eta)$ } $Y^\eta_m = {\Cal M}[K^{\eta \char 94 \langle 0
2274: \rangle}_m,N_\eta] \cap {\Cal M}[K^{\eta \char 94 \langle 1 \rangle}_m,
2275: N_\eta] \in D_{j_{\eta \char 94 <0>}}(N_\eta)$.
2276: \endroster}
2277: \endroster
2278: \hfill$\square_{\scite{1.10}}$
2279: \enddemo
2280: \bigskip
2281:
2282: \demo{\stag{1.11} Conclusion} Let $\bar Q$ be a CS iteration and ${\frak p}$
2283: a non-$\Game'$-loser reasonable parameter,
2284: $2 \le n(*) \le \aleph_0,\ell g({\frak p}) = \omega_1$. \nl
2285: \ub{Then} $\bar Q$ is a ${\frak p}-NNR^0_{n(*)}$-iteration iff for each $i <
2286: \ell g(\bar Q)$
2287: \mr
2288: \item "{$(*)_i$}" ${\underset\tilde {}\to Q_i}$ is a proper forcing and
2289: $P_i,{\underset\tilde {}\to Q_i}$ satisfies clauses (d) + (c) of the
2290: Definition ``${\frak p}-NNR^0_{n(*)}$-iteration" with $i,i+1$ here standing
2291: for $i,j$ there.
2292: \endroster
2293: \enddemo
2294: \bigskip
2295:
2296: \demo{Proof} By induction on $j = \ell g(\bar Q)$.
2297: For $j=0$ there is nothing, for $j$ limit use \scite{1.7} (or \scite{1.10})
2298: and if $j$ successor just read the definitions.
2299: \hfill$\square_{\scite{1.11}}$
2300: \enddemo
2301: \bigskip
2302:
2303: We point out here that clause (c) of Definitions \scite{1.6} and \scite{1.9}
2304: really follows from earlier properties which play parallel roles.
2305: \proclaim{\stag{1.12} Claim} 1) Assume that ${\frak p}$ is a standard
2306: reasonable parameter, and is standard and $\alpha < \ell g({\frak p}),N \in
2307: {\Cal E}^{\frak p}_\alpha,Y \in D_\alpha(N)$ and $\delta \le \omega_1
2308: \cap N$ a limit ordinal. \nl
2309: \ub{Then} we can find sequences $\bar N = \langle N_i:i < \delta \rangle,
2310: \langle \gamma_i:i < \delta \rangle$ such that:
2311: \mr
2312: \item "{$(a)$}" $N_i \in N$ (for $i < \delta$) is countable,
2313: $N \cap \alpha \subseteq N_i$ and $N_i \in Y$
2314: \sn
2315: \item "{$(b)$}" $N_i \subseteq \dbcu_{\beta \in \alpha \cap N}
2316: ({\Cal H}(\chi^{\frak p}_\beta),\in)$, and $\beta \in \alpha \cap N_i
2317: \Rightarrow N_i \restriction {\Cal H}(\chi^{\frak p}_\beta) \prec
2318: ({\Cal H}(\chi^{\frak p}_\beta),\in)$
2319: \sn
2320: \item "{$(c)$}" $i < j \Rightarrow N_i \subseteq N_j$
2321: \sn
2322: \item "{$(d)$}" for $i$ limit $N_i = \dbcu_{j<i} N_j$ and
2323: $N \cap \dbcu \{{\Cal H}(\chi^{\frak p}_\beta):\beta \in \alpha \cap
2324: N\} = \dbcu_{j < \delta} N_j$ so we can stipulate $N_\delta = N$
2325: \sn
2326: \item "{$(e)$}" $\beta \in \alpha \cap N \Rightarrow \langle {\Cal H}
2327: (\chi^{\frak p}_\beta) \cap N_j:j \le i \rangle \in N_{i+1}$
2328: \sn
2329: \item "{$(f)$}" $\gamma_i \in N_i \cap \alpha,N_i \cap {\Cal H}
2330: (\chi^{\frak p}_{\gamma_i}) \in {\Cal E}_{\gamma_i} \cap Y,
2331: \bar \gamma \restriction (i+1) \in N_{i+1}$
2332: \sn
2333: \item "{$(g)$}" if $i \le \delta$ is a limit ordinal, $(i = \delta \and
2334: \beta \in \alpha \cap N_i) \vee (i < \delta \and \beta \in \gamma_i \cap N_i)$
2335: (so $\beta \in {\Cal H}(\chi^{\frak p}_\beta))$
2336: and $y \in {\Cal H}(\chi^{\frak p}_\beta)$, \ub{then} for some
2337: $j < i,\gamma_j = \beta,y \in N_j$. Moreover
2338: \sn
2339: \item "{$(g)^+$}" if $i \le \delta$ is a limit ordinal, then
2340: $\{N_j \cap {\Cal H}(\chi^{\frak p}_{\gamma_j}):j < i \text{ and }
2341: \gamma_j < \gamma_i\} \in D_{\gamma_i}(N) = D_{\gamma_i}(N \cap {\Cal H}
2342: (\chi^{\frak p}_{\gamma_i}))$
2343: \sn
2344: \item "{$(h)$}" if $\delta < N \cap \omega_1$ then $\delta \in N_0$, if
2345: $\delta = N \cap \omega_1$ then $i < \delta \Rightarrow i \in N_i$.
2346: \ermn
2347: 2) If ${\frak p}$ is a standard, reasonable parameter, $\bar Q$ is a
2348: CS iteration, $\ell g(\bar Q) = \alpha +1,\bar Q
2349: \restriction \alpha$ is ${\frak p}-NNR^0_{k(*)}$-iteration and \nl
2350: $\Vdash_{P_\alpha} ``{\underset\tilde {}\to Q_\alpha}$ is proper and
2351: $(<^+ \omega_1)$-proper", \ub{then} trying to apply \scite{1.10} for
2352: $\bar Q$,
2353: clauses (a),(c) (called ${\frak p}$-proper in Definition \scite{2.3} below)
2354: holds. \nl
2355: 3) If $\ell g({\frak p}) = \omega_1$, \ub{then} in part (2) it suffices to ask
2356: $\Vdash_{P_i} ``\alpha < \omega_1 \Rightarrow {\underset\tilde {}\to Q_i}$ is
2357: $\alpha$-proper", that is $\Vdash_{P_i} ``{\underset\tilde {}\to Q_i}$ is
2358: $(< \omega_1)$-proper".
2359: \endproclaim
2360: \bigskip
2361:
2362: \demo{Proof} 1) By induction on $\beta \in (\alpha \cap N) \cup
2363: \{\alpha\}$ we prove that there is \nl
2364: $\langle N_j:j < \beta \cap N \rangle
2365: \in N$ satisfying the relevant requirements. \nl
2366: 2) By Definition of $(< \omega_1)$-proper, if $G_\alpha \subseteq P_\alpha$
2367: is generic over $\alpha,Y =: {\Cal M}[G_\alpha,N,y^*] \in
2368: D_\alpha(N)$, let $\langle N_i:i < \delta \rangle$ be as in $(1)$ for
2369: $\delta = N \cap \omega_1$, without loss of generality
2370: $p \in N_0 \cap {\underset\tilde {}\to Q_\alpha}[G_\alpha]$,
2371: let $q \ge p$ be $(N_i,{\underset\tilde {}\to Q_\alpha}[G_\alpha])$-generic
2372: for every $i < \delta$ (formally look at $\bar N' = \langle N'_i:i < \delta
2373: \rangle,N'_i = N_i \restriction {\Cal H}(\chi^{\frak p}_0)$ and apply to it
2374: the $(< \omega_1)$-proper). \nl
2375: 3) Follows as $\alpha \in N \Rightarrow \delta = \omega \alpha \in N \cap
2376: \omega_1$. \hfill$\square_{\scite{1.12}}$
2377: \enddemo
2378: \bn
2379: \ub{\stag{1.13} Discussion}: 1) This includes as special cases
2380: \cite[Ch.V,\S5,\S7]{Sh:b}. There is no direct comparison with
2381: \cite[Ch.VIII,\S4]{Sh:b}, \cite[Ch.VIII,\S4]{Sh:f}, but we can make the notion
2382: somewhat more complicated, to include the theorem there in our context,
2383: i.e., what is not included is a generalization there not really needed
2384: for the examples discussed there (see here in \S3 below).
2385: The condition there involves having many sequences $\langle N_\alpha:
2386: \alpha \le \delta \rangle$ such that if $p_0,p_1 \in P,p_\ell$ is
2387: $(N_i,P_\alpha)$-generic for $i,p_\ell \Vdash ``
2388: {\underset\tilde {}\to G_P} \cap N_0 = G^*"$, then there is $G' \subseteq
2389: \text{ Gen}({\underset\tilde {}\to Q_i}[G^*],N_0[G^*]),G' \in N_0,
2390: P_i \nVdash_{P_i} ``{\underset\tilde {}\to G_0}$ has no bound in
2391: ${\underset\tilde {}\to Q_\alpha}"$. This speaks on a family of sequences
2392: from $[{\Cal H}(\chi)]^{\aleph_0}$ rather than members. \nl
2393: 2) For \cite[Ch.XVIII,\S2]{Sh:f}, the comparison is not so easy.
2394: Our problem is to
2395: ``carry" good $(N,P_i,\langle G_\ell:\ell < n \rangle),G_\ell \in
2396: \text{ Gen}(N,P_i)$ with a bound, such that we can ``increase $i$" and we
2397: can find $N',y \in N' \in N,N' \prec N$ such that $(N',P_i,\langle G_i \cap
2398: N:\ell < n \rangle)$ is good enough. In \cite[Ch.XVIII]{Sh:f} we are carrying
2399: genericity in some $P_{\bar \alpha},\bar \alpha \in \text{ trind}(i)$, here
2400: much less. But what we need is the implication ``if $(N,P_i,\bar G)$ is good
2401: we can extend it to good $(N,P_{i+1},\bar G')$", so making good weaker
2402: generates an incomparable notion and clearly there are other variants.
2403: We can consider other such notions (see \S5). \nl
2404: 2A) We can give alternative proofs of consistency of the questions in
2405: \cite[Ch.XVIII,\S1]{Sh:f} by the present iteration theorem (see \S3 below). \nl
2406: 3) We can in \scite{1.12}(2) weaken the assumption ``$(< \omega_1)$-proper"
2407: to things of the form: if $\bar N = \langle N_i \le \delta \rangle,p \in Q
2408: \cap N_0,Q \in N_0$ then there is $q \ge p$ such that $q \Vdash ``
2409: \text{for many } i < \delta,{\underset\tilde {}\to G_Q} \cap N_i \in
2410: \text{ Gen}(N_i,Q)"$. In particular the condition applies to the forcing
2411: notions considered in \cite[Ch.XVIII,\S1]{Sh:f}.
2412: \newpage
2413:
2414: \head {\S2 Delayed properness} \endhead \resetall \sectno=2
2415: % \resetall
2416: \bn
2417: In this section we prove little, but the notions introduced are used in
2418: \S3, \S4. We concentrate here on simple parameters so the reader may assume
2419: it all the time. We give two versions, the simpler one is version 2 for
2420: which simplicity is a very natural demand.
2421: \demo{\stag{2.1} Observation} 1) Assume
2422: \mr
2423: \item "{$(a)$}" $\bar \chi = \langle \chi_\alpha:\alpha < \alpha^* \rangle$
2424: increases fast enough, so $\dbcu_{\beta < \alpha} {\Cal H}(\chi_\beta) \in
2425: {\Cal H}(\chi_\alpha)$
2426: \sn
2427: \item "{$(b)$}" ${\Cal E}_\alpha \subseteq \{N:N$ a countable elementary
2428: submodel of $({\Cal H}(\chi_\alpha),\in)\}$ \nl
2429: is stationary
2430: \sn
2431: \item "{$(c)$}" $R_\alpha \in {\Cal H}(\chi_\alpha)$ and $N \in
2432: {\Cal E}_\alpha$ implies \nl
2433: $\langle \chi_\beta:\beta < \alpha \rangle \in N, \langle R_\beta:\beta \le
2434: \alpha \rangle \in N$ and $\langle {\Cal E}_\beta:\beta < \alpha \rangle
2435: \in N$.
2436: \ermn
2437: \ub{Then} there is a standard reasonable parameter ${\frak p}$ with
2438: $\ell g({\frak p}) = \alpha^*,{\Cal E}^{\frak p}_\alpha = {\Cal E}_\alpha$,\nl
2439: $R^{\frak p}_\alpha = R_\alpha$ (and $\bar D^{\frak p}$ standard). \nl
2440: 2) If in addition clause (d) below hold, \ub{then} ${\frak p}$ is a simple
2441: standard reasonable parameter where
2442: \mr
2443: \item "{$(d)$}" $\beta \in N \in {\Cal E}_\alpha,\beta < \alpha
2444: \Rightarrow N \cap {\Cal H}(\chi_\beta) \in {\Cal E}_\beta$.
2445: \ermn
2446: 3) If $\chi_\alpha = (\beth_{2 \alpha +1})^+$ for $\alpha < \alpha^*,
2447: R_\alpha \in {\Cal H}(\chi_\alpha)$, then $\chi_\alpha$ increases fast
2448: enough. If $\langle \chi_\alpha:\alpha < \alpha^* \rangle,\langle R_\alpha:
2449: \alpha < \alpha^* \rangle$ are as in part (1), $\chi \le \chi_0,{\Cal E}
2450: \subseteq [{\Cal H}(\chi)]^{\le \aleph_0}$ stationary and we let ${\Cal E}
2451: _\alpha = \{N:N$ a countable elementary submodel of $({\Cal H}(\chi_\alpha),
2452: \in)$ and $\langle \chi_\beta:\beta < \alpha \rangle,\langle R_\beta:\beta
2453: \le \alpha \rangle,{\Cal E}$ belong to $N$ and $N \cap {\Cal H}(\chi) \in
2454: {\Cal E}\}$, \ub{then} the assumptions of parts (1)
2455: and (2) above holds (hence their conclusions).
2456: \enddemo
2457: \bigskip
2458:
2459: \demo{Proof} Straight.
2460: \enddemo
2461: \bigskip
2462:
2463: \definition{\stag{2.2} Definition} Let ${\frak p}$ be a reasonable
2464: parameter and $\alpha \le \beta < \ell g({\frak p})$. \nl
2465: 1) For $N \in {\Cal E}_\beta$ such that $\alpha \in N$ we define a game
2466: $\Game_{\alpha,\beta}(N) = \Game_{\alpha,\beta}(N,{\frak p})$ as follows.
2467: \nl
2468: A play lasts $\omega$-moves. In the $n$-th move:
2469: \mr
2470: \item "{$(a)$}" the \ub{challenger chooses} $X_n \in D_\beta(N)$ such that
2471: \nl
2472: $m < n \Rightarrow X_n \subseteq X_m$
2473: \sn
2474: \item "{$(b)$}" the \ub{chooser} chooses $\alpha_n \in \alpha \cap N$
2475: \sn
2476: \item "{$(c)$}" the \ub{challenger} chooses $\beta'_n \in \beta \cap
2477: N$ and $y'_n \in N \cap {\Cal H}(\chi^{\frak p}_{\alpha_n})$
2478: \sn
2479: \item "{$(d)$}" the \ub{chooser} chooses $\beta_n \in \beta \cap N
2480: \backslash \beta'_n$ and $M_n \in X_n \cap {\Cal E}_{\beta_n}$
2481: and $y_n \in M_n \cap {\Cal H}(\chi^{\frak p}_{\alpha_n})$ satisfying
2482: $\alpha_n \le \beta_n,y'_n \in M_n,\alpha_n \in M_n$ and $Y_n \in D_{\beta_n}
2483: (M_n)$ such that $Y_n \subseteq X_n$ and $Y_n \in N$
2484: \sn
2485: \item "{$(e)$}" the \ub{challenger} chooses $M'_n \in Y_n \cap
2486: {\Cal E}_{\alpha_n} \cap (M_n \cup \{M_n \cap {\Cal H}
2487: (\chi^{\frak p}_{\alpha_n})\})$ satisfying $y_n,y'_n \in M'_n$ and
2488: chooses
2489: $Z_n \in D_{\alpha_n}(M'_n) = D_{\alpha_n}(M'_n \cap {\Cal H}(\chi
2490: ^{\frak p}_{\alpha_n}))$ such that $Z_n \subseteq Y_n$.
2491: \ermn
2492: In the end the chooser wins the play if
2493:
2494: $$
2495: \bigcup \{Z_n \cup \{M'_n\}:n < \omega\} \in D^+_\alpha(N) \,
2496: (= D^+_\alpha(N \cap {\Cal H}(\chi^{\frak p}_\alpha)))
2497: $$
2498: \mn
2499: 1A) We call $\Game_{\alpha,\beta}(N) = \Game_{\alpha,\beta}(N,{\frak p})$
2500: version 1; version 2 means that in clause (e) we add the requirement
2501: $M'_n = M_n \cap {\Cal H}(\chi^{\frak p}_{\alpha_n})$ and in clause (d) we
2502: require $\alpha_n \le_{\frak p} \beta_n$. If we do not mention the version
2503: it means that it holds for both.
2504: \nl
2505: 2) Assume $N \in N' \prec ({\Cal H}(\chi),\in)$.
2506: We define a game $\Game'_{\alpha,\beta}(N,N',{\frak p})$ similarly but
2507: replace a) - e) by
2508: \mr
2509: \item "{$(a)'$}" the \ub{challenger} chooses $X_n \in D_\beta(N)$ such
2510: that $m < n \Rightarrow X_n \subseteq X'_m$
2511: \sn
2512: \item "{$(b)'$}" the \ub{chooser} chooses $\alpha_n \in \alpha \cap
2513: N$ and $X'_n \subseteq X_n$ such that $X'_n \in D_\beta(N)$
2514: \sn
2515: \item "{$(c)'$}" like (c)
2516: \sn
2517: \item "{$(d)'$}" like (d) but we replace ``$Y_n \in N$" by ``$Y_n \in N'$"
2518: \sn
2519: \item "{$(e)'$}" like (e) but add $Z_n \in N'$.
2520: \ermn
2521: Note: so every proper initial segment of a play belongs to $N'$.
2522: \enddefinition
2523: \bigskip
2524:
2525: \definition{\stag{2.3} Definition} Let ${\frak p}$ be a reasonable
2526: parameter. \nl
2527: 1) For $\alpha \le \beta < \ell g({\frak p})$ we say ${\frak p}$ is an
2528: $\Game_{\alpha,\beta}$-winner [non-$\Game_{\alpha,\beta}$-loser] when for
2529: some $x \in {\Cal H}(\chi^{\frak p}_\beta)$ we have:
2530: \mr
2531: \item "{{}}" if $\{x,\alpha\} \in N \in {\Cal E}_\beta$, \ub{then} the
2532: chooser wins the game $\Game_{\alpha,\beta}(N,{\frak p})$
2533: [the challenger does not win in the game $\Game_{\alpha,\beta}
2534: (N,{\frak p})$].
2535: \ermn
2536: 2) We can replace $\Game_{\alpha,\beta}$ by $\Game'_{\alpha,\beta}$. \nl
2537: 3) For any function $f:\ell g({\frak p}) \rightarrow {\Cal P}(\ell g
2538: ({\frak p}))$ we can replace $\alpha,\beta$ by $f$ meaning: for every
2539: $\alpha < \ell g({\frak p})$ and $\beta \in f(\alpha)$, we have
2540: ${\frak p}$ is a $\Game_{\alpha,\beta}$-winner. \nl
2541: 4) Let ${\Cal F}^{\frak p}$ be the family of functions from
2542: $\ell g({\frak p})$ to ${\Cal P}(\ell g({\frak p}))$ such that for each
2543: $\alpha < \ell g({\frak p}),f(\alpha)$ is a nonempty subset of
2544: $[\alpha,\ell g({\frak p}))$. Let ${\Cal F}^{\frak p}_{\text{club}}$ be
2545: the set of $f \in {\Cal F}^{\frak p}$ such that for each $\alpha < \ell g
2546: ({\frak p}),f(\alpha)$ is a club of $\ell g({\frak p})$. Let
2547: ${\Cal F}^{\frak p}_{\text{end}}$ be the set of $f \in {\Cal F}^{\frak p}$
2548: such that for each $\alpha < \ell g({\frak p}),f(\alpha)$ is an end segment
2549: of $\ell g({\frak p})$, we then may identify $f(\alpha)$ with Min$(f
2550: (\alpha))$. \nl
2551: 5) We call $f$ decreasing continuous if $\alpha_1 < \alpha_2 < \ell g
2552: ({\frak p}) \Rightarrow f(\alpha_2) \subseteq f(\alpha_1)$ and for limit
2553: $\alpha < \ell g({\frak p})$ we have $f(\alpha) = \dbca_{\gamma < \alpha}
2554: f(\gamma)$. Let $f \le g$ mean $(\forall \alpha < \ell g({\frak p}))
2555: (g(\alpha) \subseteq f(\alpha))$.
2556:
2557: There are obvious monotonicity properties.
2558: \enddefinition
2559: \bigskip
2560:
2561: \proclaim{\stag{2.4} Claim} Assume ${\frak p}$ is a reasonable
2562: parameter. \nl
2563: 1) If $\alpha \le_{\frak p} \alpha',\alpha' \le \alpha \le \beta = \beta'
2564: < \ell g({\frak p})$, and ${\frak p}$ is a $\Game_{\alpha,\beta}$-winner,
2565: \ub{then} it is
2566: $\Game_{\alpha',\beta'}$-winner. Similarly for $\Game'$-winner,
2567: non-$\Game$-loser, non-$\Game'$-loser. \nl
2568: 2) If ${\frak p}$ is a $\Game_{\alpha,\beta}$-winner, \ub{then}
2569: ${\frak p}$ is a $\Game'_{\alpha,\beta}$-winner and a
2570: non-$\Game_{\alpha,\beta}$-loser.
2571: If ${\frak p}$ is a $\Game'_{\alpha,\beta}$-winner or
2572: non-$\Game_{\alpha,\beta}$-loser, \ub{then} ${\frak p}$ is
2573: non-$\Game'_{\alpha,\beta}$-loser. \nl
2574: 3) Assume $f,g \in {\Cal F}^{\frak p}$ and $f \le g$.
2575: If ${\frak p}$ is a $\Game_f$-winner [or $\Game'_f$-winner]
2576: [ or non-$\Game_f$-loser] [or non-$\Game_f$-loser], \ub{then}
2577: ${\frak p}$ is a $\Game_g$-winner [or
2578: $\Game'_g$-winner] [or non-$\Game_g$-loser] [or non-$\Game'_g$-loser].
2579: \endproclaim
2580: \bigskip
2581:
2582: \demo{Proof} Straight (we are using the simplicity of ${\frak p}$).
2583: \enddemo
2584: \bigskip
2585:
2586: \proclaim{\stag{2.5} Claim} 1) Assume ${\frak p}$ is a standard reasonable
2587: parameter. \ub{Then} ${\frak p}$ is a winner (see Definition \scite{1.4}).
2588: \nl
2589: 2) If ${\frak p}$ is a reasonable parameter and it is a winner, \ub{then}
2590: ${\frak p}$ is a $\Game_{\alpha,\alpha}$-winner. \nl
2591: 3) If ${\frak p}$ is a reasonable parameter and it is a winner and
2592: $\alpha \le \beta < \ell g({\frak p})$ \ub{then} ${\frak p}$
2593: is a $\Game_{\alpha,\beta}$-winner (hence $\Game_f$-winner for $f$ as in
2594: \scite{2.3}(3)). \nl
2595: 4) Similarly with $\Game'$-winner, $\Game'_{\alpha,\beta}$ winner and/or
2596: with the ``non loser" cases.
2597: \endproclaim
2598: \bigskip
2599:
2600: \demo{Proof} Straight.
2601: \enddemo
2602: \bigskip
2603:
2604: \definition{\stag{2.6} Definition} 1) Let ${\frak p}$ be a reasonable
2605: parameter and $P$ be a proper forcing notion not adding reals, ${\Cal P}(P)
2606: \in {\Cal H}(\chi^{\frak p}_0)$ and $G_P \subseteq P$ is generic over
2607: $V$.
2608:
2609: \ub{Then} we interpret ${\frak p}$ in $V^P$ as ${\frak p}' =
2610: {\frak p}^{V[G]}$, or we may write ${\frak p}^P$ defined as follows:
2611: \mr
2612: \item "{$(a)$}" $\chi^{{\frak p}'}_\alpha = \chi^{\frak p}_\alpha$
2613: \sn
2614: \item "{$(b)$}" $R^{{\frak p}'}_\alpha = \langle R_\alpha,P,G_P \rangle$
2615: \sn
2616: \item "{$(c)$}" ${\Cal E}^{{\frak p}'}_\alpha = \{N[G_P]:N \in {\Cal E}
2617: _\alpha$ and $P \in N$ and $N[G_P] \cap V = N\}$
2618: \sn
2619: \item "{$(d)$}" $D_\alpha(N[G_P]) = \bigl\{ \{M[G_P] \in {\Cal E}_\alpha:
2620: M \in Y \cap \dbcu_{\beta < \alpha} {\Cal E}^{\frak p}_\beta\}:Y \in
2621: D_\alpha(N) \bigr\}$.
2622: \endroster
2623: \enddefinition
2624: \bigskip
2625:
2626: \proclaim{\stag{2.7} Claim} Let ${\frak p},P$ be as in Definition
2627: \scite{1.6}. \nl
2628: 1) In Definition \scite{2.6}, ${\frak p}^{V[G_P]}$ is
2629: a reasonable parameter in $V[G_P]$. \nl
2630: 2)
2631: If ${\frak p}$ is, in $V$, a $\Game$-winner (or
2632: non-$\Game$-loser or $\Game'$-winner or non-$\Game'$-loser), \ub{then}
2633: ${\frak p}^{V[G_P]}$ is so (in $V[G_P]$). \nl
2634: 3) If ${\frak p}$ is, in $V$, a $\Game_{\alpha,\beta}$-winner (or a non
2635: $\Game_{\alpha,\beta}$-lower or $\Game'_{\alpha,\beta}$-winner or
2636: $\Game'_{\alpha,\beta}$-non loser), \ub{then} ${\frak p}^{V[G_P]}$ is so
2637: in $V[G_p]$.
2638: \endproclaim
2639: \bigskip
2640:
2641: \demo{Proof} Straight (we use $P$ is a proper forcing notion not adding
2642: reals).
2643: \enddemo
2644: \bigskip
2645:
2646: \centerline {$* \qquad * \qquad *$}
2647: \bigskip
2648:
2649: \definition{\stag{2.8} Definition} Let ${\frak p}$ be a reasonable
2650: parameter. \nl
2651: 1) For $\alpha \le \beta < \ell g({\frak p})$, we say a forcing notion $Q$
2652: is $({\frak p},\alpha,\beta)$-proper if: \nl
2653: $({\Cal P}(Q) \in {\Cal H}(\chi^{\frak p}_0)$ and):
2654: \mr
2655: \item "{$(*)$}" for some $x \in {\Cal H}(\chi_\beta)$ \ub{if} $N \in {\Cal E}
2656: _\beta,\{x,Q,\alpha\} \in N,p \in N \cap Q$ and \nl
2657: $Y \in D_\alpha(N)$, \ub{then} for some $q$ we have:
2658: {\roster
2659: \itemitem{ (a) } $p \le q \in Q$
2660: \sn
2661: \itemitem{ (b) } $q$ is $(N,Q)$-generic
2662: \sn
2663: \itemitem{ (c) } version (1): for some $N' \in ({\Cal E}_\alpha \cap N \cap
2664: Y) \cup \{N \cap {\Cal H}(\chi^{\frak p}_\alpha)\}$ satisfying $\alpha =
2665: \beta \Rightarrow N' = N$ we have
2666: $q \Vdash_Q ``{\Cal M}[{\underset\tilde {}\to G_Q},N',y^*] \cap Y \in
2667: D_\alpha(N')"$ where $y^* = \langle x,p,Q \rangle$ hence necessarily $q$
2668: is $(N',Q)$-generic.
2669: \endroster}
2670: (Version 2: similarly but $N' = N \restriction {\Cal H}(\chi^{\frak p}
2671: _\alpha)$ and naturally \nl
2672: \,\,\,we demand $\alpha \le_{\frak p} \beta$).
2673: \ermn
2674: 2) We say $Q$ is $({\frak p},f)$-proper if (${\frak p},f$ are as above and)
2675: for every $\alpha < \ell g({\frak p})$ and $\beta \in f(\alpha)$ we have
2676: $Q$ is $({\frak p},\alpha,\beta)$-proper. \nl
2677: 3) We say $Q$ is ${\frak p}$-proper if $Q$ is
2678: $({\frak p},\alpha,\alpha)$-proper for
2679: $\alpha < \ell g({\frak p})$. We say $Q$ is almost ${\frak p}$-proper
2680: if $Q$ is $({\frak p},f)$-proper for some $f \in {\Cal F}^{\frak p}$
2681: (as above).
2682: \enddefinition
2683: \bigskip
2684:
2685: \proclaim{\stag{2.9} Claim} Assume ${\frak p}$ is a simple reasonable
2686: parameter. \nl
2687: 1) If $\alpha' \le \alpha \le \beta \le \beta' < \ell g({\frak p})$
2688: (for version 2 we demand $\alpha' \le_{\frak p} \beta'$ and $\alpha
2689: \le_{\frak p} \beta$) and $Q$ is as $({\frak p},\alpha,\beta)$-proper
2690: forcing notion, \ub{then} $Q$ is a
2691: $({\frak p},\alpha',\beta')$-proper forcing notion. \nl
2692: 2) Assume $f,f'$ are in ${\Cal F}^{\frak p}$ and $f \le f'$.
2693: If $Q$ is a $({\frak p},f)$-proper forcing notion, \ub{then} $Q$ is a
2694: $({\frak p},f')$-proper forcing notion.
2695: \endproclaim
2696: \bigskip
2697:
2698: \demo{Proof} Straight.
2699: \enddemo
2700: \bn
2701: \ub{\stag{2.10} Discussion}: We may like to consider $({\frak p},f)$-proper
2702: for iterations which may add reals. Then we have to replace $D_\alpha(N)$
2703: by a definition which is absolute enough, (and the non-loser versions have
2704: to be absolute enough). \nl
2705: It is natural to restrict ourselves to ${\frak p}$-closed $Y$, see below.
2706: \bigskip
2707:
2708: \definition{\stag{2.11} Definition} Let ${\frak p}$ be a simple reasonable
2709: parameter and $N \in {\Cal E}_\alpha$. Now $Y \subseteq N$ is called
2710: ${\frak p}$-closed if (see \scite{3.9}):
2711: \mr
2712: \item "{$(a)$}" $Y \subseteq N \cap \dbcu_{\beta < \alpha} {\Cal E}_\beta$
2713: \sn
2714: \item "{$(b)$}" if $M \in N \cap {\Cal E}_\beta$ and $\beta < \alpha$
2715: (hence $\beta \in \alpha \cap M \subseteq \alpha \cap N$) and
2716: $\gamma \in M \cap \beta$ and $M \cap {\Cal H}(\chi^{\frak p}_\gamma) \in
2717: {\Cal E}_\gamma$ (this requirement is redundant if ${\frak p}$ is simple or
2718: just $\gamma \le_{\frak p} \beta$), \nl
2719: \ub{then}
2720: $M \cap {\Cal H}(\chi^{\frak p}_\gamma) \in Y \Leftrightarrow M \in Y$
2721: \sn
2722: \item "{$(c)$}" if $\beta < \alpha,M_\ell \in N \cap {\Cal E}_\beta$
2723: (hence $\beta \in \alpha \cap N$) and $M_\ell \subseteq M_{\ell +1}$ for
2724: $\ell < \omega$ and
2725: $M = \dbcu_{\ell < \omega} M_\ell \in N \cap {\Cal E}_\beta$,
2726: and even $\langle M_\ell:\ell < \omega \rangle \in N$, \ub{then}
2727: $\dsize \bigwedge_\ell M_\ell \in Y \Rightarrow M \in Y$.
2728: \endroster
2729: \enddefinition
2730: \bn
2731: We may consider
2732: \definition{\stag{2.12} Definition} Let ${\frak p}$ be a reasonable parameter
2733: and $\Xi \subseteq \{(\alpha,\beta):\alpha \le \beta < \ell g({\frak p})\}$
2734: and we understand $N \in {\Cal E}_\alpha \Rightarrow \Xi \cap ((\alpha + 1)
2735: \times (\alpha +1) \in N)$ and let $\alpha_{\frak p}(N) =$ the $\alpha$ such
2736: that $N \in {\Cal E}^{\frak p}_\alpha$ if it exists.
2737:
2738: We say $\bar Q$ is an $(\kappa_1,\Xi)$-anti w.d.-iteration for ${\frak p}$
2739: as in clauses (a), (b), (d) of Definitions \scite{1.6}, \scite{1.9} only we
2740: replace clause (d) by
2741: \mr
2742: \item "{$(d)_\Xi$}" [$(\kappa,\Xi)$-anti w.d.] like old clause (d) but
2743: for some $(\alpha,\beta) \in \Xi,
2744: N_0 \in N_1$ are countable elementary submodels of $({\Cal H}(\chi),\in),
2745: {\frak p} \in N_0,N_0 \cap {\Cal H}(\chi^{\frak p}_\alpha) \in
2746: {\Cal E}_\alpha,N_1 \cap {\Cal H}(\chi^{\frak p}_\beta) \in {\Cal E}_\beta,
2747: \alpha \in N_0,\beta \in N_1$ and the rest as before.
2748: \endroster
2749: \enddefinition
2750: \bn
2751: We omit ``for ${\frak p}$" if ${\frak p}$ is clear from the context.
2752: \proclaim{\stag{2.13} Lemma} Let ${\frak p}$ be a reasonable parameter. \nl
2753: 1) If $\Xi \ne \emptyset$ and $\bar Q$ is an
2754: $NNR^0_{\aleph_0,\Xi}$-iteration, \ub{then} $P_{\ell g(\bar Q)} =
2755: \text{ Lim}(\bar Q)$ does not add reals and is
2756: $({\frak p},\Xi)$-proper.
2757: \endproclaim
2758: \bigskip
2759:
2760: \demo{Proof} Easy.
2761: \enddemo
2762: \bigskip
2763:
2764: As the sets ${\Cal H}(\chi_\alpha)$ may change with forcing, we may prefer to
2765: use ${\Cal E}_\alpha \subseteq [\chi_\alpha]^{\le \aleph_0}$, for this we
2766: define:
2767: \definition{\stag{2.14} Definition} 1) We call ${\frak p}$ an o.b. (ordinal
2768: based) parameter if \nl
2769: ${\frak p} = (\bar \chi^{\frak p},\bar R^{\frak p},
2770: \bar{\Cal E}^{\frak p},\bar D^{\frak p})$ and for some ordinal $\alpha^*$
2771: called $\ell g({\frak p})$ we have:
2772: \mr
2773: \item "{$(a)$}" $\bar \chi = \langle \chi_\alpha:\alpha < \alpha^* \rangle,
2774: \chi_\alpha$ is a regular cardinal and ${\Cal H}\bigl( (\dbcu_{\beta <
2775: \alpha} \chi_\beta)^+ \bigr) \in {\Cal H}(\chi_\alpha)$
2776: \sn
2777: \item "{$(b)$}" $\bar R = \langle R_\alpha:\alpha < \alpha^* \rangle,
2778: R_\alpha$ an $n(R_\alpha)$-place relation on some bounded subset of
2779: $\chi_\alpha$ (we could have asked ``on $\chi_\alpha$", no real difference)
2780: \sn
2781: \item "{$(c)$}" $\bar{\Cal E} = \langle {\Cal E}_\alpha:\alpha < \alpha^*
2782: \rangle,{\Cal E}_\alpha \subseteq [\chi_\alpha]^{\le \aleph_0}$ is
2783: stationary
2784: \sn
2785: \item "{$(d)$}" $\bar D = \langle D_\alpha:\alpha < \alpha^* \rangle$ and
2786: $D_\alpha$ is a function with domain ${\Cal E}_\alpha$ and for each $a \in
2787: {\Cal E}_\alpha,D_\alpha(a)$ is a family of subsets of $a$, closed
2788: under supersets, non-empty if $\alpha > 0$
2789: (let $D^-_\alpha(a) = (D_\alpha(a))^- = {\Cal P}(a) \backslash D_\alpha(a))$
2790: \sn
2791: \item "{$(e)$}" let ${\frak p}^{[\alpha]} = \langle \bar \chi \restriction
2792: \alpha,\bar R \restriction (\alpha +1),\bar{\Cal E} \restriction \alpha,
2793: \bar D \restriction \alpha \rangle$
2794: \sn
2795: \item "{$(f)$}" if $a \in {\Cal E}_\alpha$ and $X \subseteq {\Cal P}(a)$,
2796: \ub{then}: \nl
2797: $X \in D_\alpha(a)
2798: \Leftrightarrow X \cap \dbcu_{\beta < \alpha} {\Cal E}_\beta \in D_\alpha(a)$.
2799: \ermn
2800: 2) We say that an o.b. parameter ${\frak p}$ is simple if
2801: \mr
2802: \item "{$(g)$}" if $a \in {\Cal E}_\alpha$ and $X \in D_\alpha(a)$ and
2803: $\beta \in \alpha \cap a$, \ub{then} $a \cap \chi_\beta \in {\Cal E}_\beta$.
2804: \ermn
2805: 3) For ${\frak p}$ as above let ${\frak q} =: {\frak p}^V$ be defined by
2806:
2807: $$
2808: \alpha^{*,{\frak q}} = \alpha^{*,{\frak p}}
2809: $$
2810: \mn
2811: and we define by induction on $\alpha < \alpha^{*,{\frak p}}$:
2812:
2813: $$
2814: \chi^{\frak q}_\alpha = \chi^{\frak p}_\alpha
2815: $$
2816:
2817: $$
2818: R^{\frak q}_\alpha = R^{\frak p}_\alpha
2819: $$
2820:
2821: $$
2822: {\Cal E}^{\frak q}_\alpha = \{N \prec ({\Cal H}(\chi^{\frak q}_\alpha),\in):
2823: N \text{ is countable}, N \cap \chi^{\frak p}_\alpha \in
2824: {\Cal E}^{\frak p}_\alpha,{\frak q}^{[\alpha]} \in N\}
2825: $$
2826: \mn
2827: (note that ${\frak q}^{[\alpha]}$ is well defined by the induction
2828: hypothesis)
2829:
2830: $$
2831: \align
2832: D^{\frak q}_\alpha(N) = \biggl\{ Y':&Y' \subseteq \dbcu_{\beta < \alpha}
2833: {\Cal E}^{\frak q}_\beta \text{ and for some } y \in N \cap \dbcu_{\beta <
2834: \alpha} {\Cal H}(\chi^{\frak p}_\beta) \\
2835: &\text{and } Y \in D^{\frak p}_\alpha(N \cap \chi^{\frak p}_\alpha)
2836: \text{ we have} \\
2837: &Y' \supseteq \{M:M \in N \cap \dbcu_{\beta < \alpha} {\Cal E}^{\frak q}
2838: _\beta \text{ and } y \in M \text{ and } M \cap \chi^{\frak p}_\alpha \in Y\}
2839: \biggr\}.
2840: \endalign
2841: $$
2842: \mn
2843: 4) For such ${\frak p}$ we say $\bar Q$ is a $NNR^0_\kappa$-iteration for
2844: ${\frak p}$ if it is an $NNR^0_\kappa$-iteration for ${\frak p}^V$.
2845: We say ${\frak p}$ is simple if ${\frak p}^V$ is. \nl
2846: Similarly for $\Game$-winner, non-$\Game$-loser, etc.
2847: \enddefinition
2848: \bigskip
2849:
2850: \proclaim{\stag{2.14A} Claim} Assume ${\frak p}$ is an o.b. [simple]
2851: parameter in the universe $V$. \nl
2852: 1) If $P \in {\Cal H}(\chi^{\frak p}_0)$ is a proper forcing notion (or at
2853: least preserve the stationarity of ${\Cal E}^{\frak p}_\alpha$ for each
2854: $\alpha < \ell g({\frak p})$), then $\Vdash_P ``{\frak p}$ is an o.b. [simple]
2855: parameter". \nl
2856: 2) If forcing with $P$ adds no reals, then also $\Game$-winner,
2857: non-$\Game$-loser, etc., are preserved. \nl
2858: 3) ${\frak p}^V$ is a reasonable parameter.
2859: \endproclaim
2860: \bigskip
2861:
2862: \definition{\stag{2.15} Definition} Let ${\frak p}$ be an o.b. parameter. \nl
2863: 1) We say $Q$ is an $NNR^0_\kappa$-forcing for ${\frak p}$ or a
2864: ${\frak p}-NNR^0_\kappa$-forcing notion when: ($Q$ is
2865: a forcing notion in a universe $V$ and) if for some transitive class
2866: $V_0,{\frak p} \in V_0$ and $NNR^0_\kappa$-iteration $\bar Q = \langle P_i,
2867: {\underset\tilde {}\to Q_j}:i \le \alpha,j < \alpha \rangle$, we have
2868: $V^{P_\alpha}_0 = V$, \ub{then} we can let ${\underset\tilde {}\to Q_\alpha}
2869: = Q$ and get $\bar Q'$, an $NNR^0_\kappa$-iteration (i.e. $V = V_0[G_\alpha],
2870: G_\alpha \subseteq P_\alpha$ is generic over $V_0$ and there is a
2871: $P_\alpha$-name ${\underset\tilde {}\to Q_\alpha}$ such that $\langle P_i,
2872: {\underset\tilde {}\to Q_i}:i \le \alpha \rangle$ is an $NNR^0_\kappa$-
2873: iteration and $Q = {\underset\tilde {}\to Q_\alpha}[G_\alpha]$).
2874: In particular $Q$ is proper and does not add reals. \nl
2875: 2) If we omit ``for ${\frak p}$" we mean for any ${\frak p}$ which makes
2876: sense. Alternatively, we can put a family of ${\frak p}$'s. \nl
2877: 3) We add ``over $x$" if this holds whenever $x \in V_0$. We can use the
2878: same definition for other versions of $NNR$. \nl
2879: 4) $Ax^\alpha_\lambda({\frak p},\kappa,0)$ means: if $Q$ is an
2880: $\aleph_2$-e.c.c. (see \cite[Ch.VII,\S1]{Sh:f}), ($\aleph_2$-pic if $\lambda
2881: = \aleph_2$; see \cite[Ch.VIII,\S2]{Sh:f}), an $NNR^0_\kappa$-forcing notion
2882: for ${\frak p},{\Cal I}_\beta$
2883: a dense open subset of $Q$ for $\beta < \beta^* < \lambda$ and
2884: ${\underset\tilde {}\to S_i}$ is a $Q_i$-name of a stationary subset of
2885: $\omega_1$ for $i < i^* < \alpha$, \ub{then} for some directed $G \subseteq Q$
2886: we have: \nl
2887: $\beta < \beta^* \Rightarrow G \cap {\Cal I}_\beta \ne \emptyset,
2888: {\underset\tilde {}\to S_i}[G] = \{\gamma < \omega_1:(\exists r \in G)
2889: (r \Vdash_Q ``\gamma \in {\underset\tilde {}\to S_i"})\}$ is a stationary
2890: subset of $\omega_1$. \nl
2891:
2892: Now we can deduce conclusions on preservation of being
2893: an $NNR^0_\kappa$-forcing notion for ${\frak p}$ and consistency of axioms.
2894: \enddefinition
2895: \bigskip
2896:
2897: \demo{\stag{2.16} Conclusion} 1) if ${\frak p}$ is o.b. parameter,
2898: non-$\Game'_{\text{id}}$-loser, $\ell g({\frak p}) = \omega_1$ and
2899: $\bar Q$ is an CS iteration such that $\alpha < \ell g(\bar Q) \Rightarrow
2900: \Vdash_{P_\alpha} ``{\underset\tilde {}\to Q_\alpha}$ is an
2901: $NNR^0_\kappa$-forcing notion for \nl
2902: ${\frak p}"$, \ub{then} $\bar Q$ is an
2903: $NNR^0_\kappa$-iteration for ${\frak p}$ (note \scite{1.7} applies). \nl
2904: 2) Assume $CH + \mu = \mu^{< \mu} \ge \lambda$, if ${\frak p}$ is a
2905: non$-\Game'_{\text{id}}$-loser o.b. parameter, $\chi^{\frak p}_0 > 2^\lambda$,
2906: \ub{then} for some $\aleph_2$-e.c.c., (if $\lambda = \aleph_2,\aleph_2$-pic)
2907: $NNR^0_\kappa$-forcing notion $P,|P| = \mu$ we have $\Vdash_P ``Ax_\lambda
2908: ({\frak p},\kappa,0)"$. \nl
2909: 3) The parallel of \scite{1.11} holds.
2910: \enddemo
2911: \bigskip
2912:
2913: \demo{Proof} Straight.
2914: \enddemo
2915: \newpage
2916:
2917: \head {\S3 Examples: shooting a thin club} \endhead \resetall \sectno=3
2918: % \resetall
2919: \bigskip
2920:
2921: We would like here to see how some examples fit our framework. We already
2922: know that $(< \omega_1)$-proper forcings are ${\frak p}$-proper (by
2923: \scite{1.12}). We first deal with a forcing notion for which the
2924: $\kappa$ (in $NNR^x_\kappa$) is $< \aleph_0$ (in Definition \scite{3.1}).
2925: Second, we deal with shooting clubs of $\omega_1$ running away
2926: from some $C_\delta \subseteq \delta = \sup(C_\delta)$ which are small (see
2927: \cite[Ch.XVIII,\S1]{Sh:f}). Those are the most natural non-$\omega$-proper
2928: forcing notions not adding reals, (though this may depend on the set theory).
2929: \bigskip
2930:
2931: The prototype for \cite[Ch.VIII,\S4]{Sh:b}, \cite[Ch.VIII,\S4]{Sh:f} is
2932: \scite{3.1}(1) below.
2933: \definition{\stag{3.1} Definition} 1) Let $\bar C = \langle (C_\delta,
2934: n_\delta):\delta < \omega_1 \text{ limit} \rangle$ where $C_\delta$ is an
2935: unbounded subset of $\delta$ of order type $\omega$ and $1 \le n_\delta <
2936: \omega$ (e.g. $n_\delta = 1$). Let $\bar u = \langle u_\delta:\delta <
2937: \omega_1 \text{ limit} \rangle$, where $u_\delta \in [2n_\delta +1]
2938: ^{n_\delta}$, (if $n_\delta =1$, then $u_\delta = \{k_\delta\},k_\delta \in
2939: \{0,1,2\}$). Then we define
2940:
2941: $$
2942: \align
2943: Q = Q_{\bar C,\bar u} = \biggl\{ f:&\text{ for some } \alpha < \omega_1,f
2944: \text{ is a function from } \alpha \\
2945: &\text{ to } \omega \text{ such that for every limit ordinal } \delta \le
2946: \alpha, \\
2947: &\text{ for some }k < 2n_\delta +1,k \notin u_\delta \text{ and for every}\\
2948: &\, i \in C_\delta \text{ large enough we have } f(i) = k \biggr\}
2949: \endalign
2950: $$
2951: \mn
2952: ordered by inclusion. \nl
2953: 2) Assume $\bar C = \langle C_\delta:\delta < \omega_1 \text{ limit} \rangle,
2954: C_\delta$ a closed subset of $\delta$ of order type $< \omega \times \delta$
2955: and for $\delta_1 < \delta_2$ limit, sup$(C_{\delta_1} \cap C_{\delta_2}) <
2956: \delta_1$, and for limit $\delta^*$ we have $\{C_\delta \cap \delta^*:\delta
2957: < \omega_1\}$ is countable. Assume further $\bar \kappa = \langle
2958: \kappa_\delta:\delta < \omega_1 \text{ limit} \rangle,\kappa_\delta \in
2959: \{2,3,\dotsc,\aleph_0\},\bar D = \langle D_\delta:\delta < \omega_1 \rangle,
2960: D_\delta$ is a family of subsets of Dom$(D_\delta)$, such that the
2961: intersection of any $< \kappa_\delta$ of them is non-empty.
2962:
2963: Let $\bar f = \langle f_{\delta,x}:\delta < \omega_1,x \in
2964: \text{ Dom}(D_\delta) \rangle$ satisfy $f_{\delta,x}:C_\delta
2965: \rightarrow \omega$ and \nl
2966: $\bar A = \langle A_\delta:\delta < \omega_1 \rangle$ satisfy
2967: $A_\delta \in D_\delta$. \ub{Then} we define
2968: $Q = Q_{\bar C,\bar D,\bar \kappa,\bar f,\bar A}$ as
2969:
2970: $$
2971: \align
2972: \biggl\{f:&\text{ for some } \alpha < \omega_1,f \text{ is a function from }
2973: \alpha \text{ to } \omega \\
2974: &\text{ such that for every limit } \delta \le \omega \text{ for some} \\
2975: &\, x \in A_\delta \text{ we have } f_{\delta,x} \subseteq^* f
2976: \text{ i.e. for every large enough} \\
2977: &\, i \in C_\delta \text{ we have } f_{\delta,x}(i) = f(i) \biggr\}
2978: \endalign
2979: $$
2980: \mn
2981: ordered by inclusion.
2982: \enddefinition
2983: \bigskip
2984:
2985: \proclaim{\stag{3.2} Claim} 1) The forcing notion $Q$ from \scite{3.1}(1)
2986: is proper, does not add reals, and ia $\alpha$-proper for $\alpha <
2987: \omega_1$ and even is $(< \omega_1)$-proper, also $\Bbb D$-complete for
2988: some simple 2-completeness system, hence causes no problem to the demand (d)
2989: in Definition \scite{1.7} which means:
2990: \mr
2991: \item "{$(*)$}" if $\bar Q$ is a CS iteration, $\ell g(\bar Q) = \alpha +1,
2992: \bar Q \restriction \alpha$ is NNR$^0_2$-iteration and
2993: $\Vdash_{P_\alpha} ``{\underset\tilde {}\to Q_\alpha} = Q_{\bar c,u}", \,\,
2994: (\bar c,\bar u)$ as above $\in V$ \ub{then} $\bar Q$ satisfies demand ((a),
2995: (b) and) (d)) of Definition \scite{1.7}.
2996: \ermn
2997: 2) Similarly for the forcing notion $Q$ from \scite{3.1}(2).
2998: \endproclaim
2999: \bigskip
3000:
3001: \demo{Proof} 1) See \cite[Ch.VIII,\S4]{Sh:b} or \cite[Ch.VIII,\S4]{Sh:f}. \nl
3002: 2) Similarly.
3003: \enddemo
3004: \bigskip
3005:
3006: \definition{\stag{3.3} Definition} Assume $\bar C = \langle C_\delta:C_\delta
3007: < \omega_1 \rangle,C_\delta$ an unbounded subset of the limit ordinal
3008: $\delta$ (think of the case $C_\delta$ of order type $< \delta$ but not
3009: necessarily). Let
3010:
3011: $$
3012: \align
3013: Q_{\bar C} = \biggl\{ c:&\text{ for some } \alpha < \omega_1,c
3014: \text{ is a closed subset of } \alpha \\
3015: &\text{ and for every limit ordinal } \delta \le \alpha \text{ we have}: \\
3016: &\, \delta = \sup(c \cap \delta) \Rightarrow c \cap C_\delta \text{ bounded
3017: in } \delta \biggr\}
3018: \endalign
3019: $$
3020:
3021: $$
3022: c_1 < c_2 \text{ iff } c_1 \text{ is an initial segment of } c_2.
3023: $$
3024: \enddefinition
3025: \bigskip
3026:
3027: \remark{Remark} 1) See \cite[Ch.XVIII,\S1]{Sh:f}, note $Q_{\bar C}$ may be not
3028: $\omega$-proper. We first deal with the simple case: otp$(C_\delta)
3029: = \omega$. \nl
3030: 2) Note that $c \in Q_{\bar c}$ is a closed subset of $\alpha$, but not
3031: necessarily a closed subset of $\omega_1$.
3032: \endremark
3033: \bigskip
3034:
3035: \proclaim{\stag{3.4} Claim} Assume ${\frak p}$ is a simple reasonable
3036: parameter, $\bar C$ as in \scite{3.3}, $\dsize \bigwedge_\delta
3037: {\text {\rm otp\/}}(C_\delta) = \omega$, let $f \in {\Cal F}^{\frak p}$
3038: be $f(0) = 0,f(\beta) = 1+ \beta$, \ub{then} $Q_{\bar C}$ is $({\frak p},
3039: f)$-proper.
3040: \endproclaim
3041: \bn
3042: We shall prove more in \scite{3.6} below.
3043: \remark{\stag{3.5} Remark} 1) Note that if $\dsize \bigwedge_\delta
3044: \text{ otp}(C_\delta) < \delta$, we can split the analysis by, for each
3045: ordinal $\gamma < \omega_1$ such that $S_\gamma =
3046: \{\delta:\text{otp}(C_\delta) = \gamma\}$ is stationary, that is restricting
3047: ourselves to $\{N:N \cap \omega_1 \in S_\gamma\}$.
3048: \endremark
3049: \bigskip
3050:
3051: \proclaim{\stag{3.6} Claim} 1) Assume
3052: \mr
3053: \item "{$(a)$}" ${\frak p}$ is a simple reasonable parameter
3054: \sn
3055: \item "{$(b)$}" ${\frak p}$ is non-$\Game_f$-loser, $f \in
3056: {\Cal F}^{\frak p}$
3057: \sn
3058: \item "{$(c)$}" $\gamma(*) < \omega_1$
3059: \sn
3060: \item "{$(d)$}" $\bar C = \langle C_\delta:\delta < \omega_1 \rangle,
3061: C_\delta \subseteq \delta = \sup(C_\delta)$
3062: \sn
3063: \item "{$(e)$}" {\rm otp}$(C_\delta) \le \omega^{\gamma(*)}$ for every
3064: $\delta$
3065: \ermn
3066: Define $g \in {\Cal F}^{\frak p}$ by $g(0) = 0,g(1) = f(1) + \gamma(*),
3067: g(\alpha +1) = f(g(\alpha)) + \gamma(*) +1$ (and necessarily, for limit
3068: $\alpha,g(\alpha) = \underset{\beta < \alpha} {}\to \sup g(\beta)$). \nl
3069: \ub{Then} for every $\alpha < \ell g({\frak p})$, the forcing notion
3070: $Q = Q_{\bar C}$ is $({\frak p},\alpha,g(\alpha))$-proper (version 1). \nl
3071: 2) We can in part (1) get ``version 2" when: if $g(\delta) = \delta,N \in
3072: {\Cal E}_\alpha,C \subseteq \omega_1 \cap N = \sup(C)$, {\rm otp}$(C) \le
3073: \omega^{\gamma(*)}$ (for example $C = C_\alpha \cap N$ for some $\alpha$)
3074: and $Y \in D_\alpha(N)$ \ub{then}
3075: $Y' = \{M \in Y:M \cap \omega_1 \notin C\} \in D_\alpha(N)$, which is
3076: a mild condition. \nl
3077: 3) If we weaken clause $(b)$ to $(b)_{f,g}$ below, \ub{then} for
3078: $\alpha < \ell g({\frak p})$, the forcing notion $Q_{\bar C}$ is
3079: $({\frak p},\alpha,f(\gamma + \alpha))$-proper, where \nl
3080: $(b)_{f,g} \, f \in {\Cal F}^{\frak p}$ and $f(f(\alpha)) =
3081: f(\alpha)$ for every $\alpha < \ell g({\frak p})$ and
3082: ${\frak p}$ is a non-$\Game'_f$-loser.
3083: \endproclaim
3084: \bigskip
3085:
3086: \demo{Proof} First observe
3087: \mr
3088: \item "{$(*)$}" for each $\alpha < \omega_1$ the set ${\Cal I}^*_\alpha =
3089: \{p \in Q_{\bar C}:\text{ there is } \beta \in p
3090: \text{ which is } \ge \alpha\}$ is a dense subset of $Q$.
3091: \ermn
3092: 1) Similar to the proof of 2) noting version (2) is harder on the chooser,
3093: and using the extra freedom we can avoid the need for the extra
3094: assumption from (2). \nl
3095: 2) We prove this by induction on $\alpha$, let $\beta = g(\alpha)$.
3096:
3097: Let $N \in {\Cal E}_\beta$ be countable,
3098: $p \in N \cap Q_{\bar C}$ (and $Q_{\bar C},\alpha,
3099: \beta \in N$) and $Y \in D_\beta(N)$ be given. Let $\delta = \delta_N =:
3100: N \cap \omega_1$.
3101: \enddemo
3102: \bn
3103: \ub{Case 1}: $\alpha = 0$.
3104:
3105: Comprehending the demand in Definition \scite{2.8}(1) it just means
3106: ``$Q_{\bar C}$ is proper". Let
3107: $\langle {\Cal I}_n:n < \omega \rangle$ list the dense open subsets of
3108: $Q_{\bar C}$ that belongs to $N$ and we shall define by induction on
3109: $n$, a condition $p_n$ such that
3110: \mr
3111: \widestnumber\item{$(iii)$}
3112: \item "{$(i)$}" $p_0 = p,p_n \in N$
3113: \sn
3114: \item "{$(ii)$}" $p_n \le p_{n+1} \in {\Cal I}_n$
3115: \sn
3116: \item "{$(iii)$}" the set $p_{n+1} \cup \{\sup p_{n+1}\} \backslash
3117: (p_n \cup \{\sup p_n\})$ is disjoint to $C_\delta$.
3118: \ermn
3119: If we succeed to carry out the induction clearly we are done as
3120: $\dbcu_{n < \omega} p_n$ is as required, noting for each $\alpha < N \cap
3121: \omega_1,{\Cal I}^*_\alpha \in \{{\Cal I}_n:n < \omega\}$, hence
3122: $(\exists \beta \in (\dbcu_{n < \omega} p_n))[\alpha \le \beta < \omega_1]$.
3123: Also there is no problem to choose
3124: $p_0$. So assume $p_n$ has been chosen and we shall choose $p_{n+1}$ as
3125: required. There is a function $F_n \in N$ with domain $Q_{\bar C}$ such that:
3126: $q \in Q_{\bar C} \Rightarrow q \le F_n(q) \in {\Cal I}_n$.
3127:
3128: For $\alpha < \omega_1$ let $q^{[\alpha]} = q \cup \{\sup(q)\} \cup
3129: \{\sup(q) + 1 + \alpha\}$, so clearly $q \le q^{[\alpha]} \in Q_{\bar C}$
3130: and the function $(q,\alpha) \mapsto q^{[\alpha]}$ belongs to $N$. Define
3131: a function $H:\omega_1 \rightarrow \omega_1$ by $H(\alpha) =
3132: \sup(F_n(p^{[\alpha]}_n))$,
3133: clearly it is well defined, belongs to $N$ and let $C =: \{\beta < \omega_1:
3134: \beta \text{ a limit ordinal, moreover } \omega \beta = \beta \text{ and }
3135: (\forall \alpha < \beta)(H(\alpha) < \beta) \text{ and sup}(p_n) < \beta\}$,
3136: so $C$ is a club of $\omega_1$ which belongs to $N$ and $\gamma(*) \in N$
3137: hence we can find $\beta^* \in C$ such that otp$(\beta^* \cap C)$ is
3138: divisible by $\omega^{\gamma(*)}$, but
3139: otp$(C_\delta \cap \beta^*) < \omega^{\gamma(*)}$ hence for some $\beta \in C$
3140: we have sup$(C_\delta \cap \beta) < \beta$. Let $p_{n+1} = F_n
3141: (p^{[\sup(C_\delta \cap \beta)+1]})$; it is as required.
3142: \bn
3143: \ub{Case 2}: $\alpha =1$.
3144:
3145: Easily $Y' =: \{M \in Y:M \cap \omega_1 \notin C_\delta\} \in
3146: D_{g(1)}(N)$ as $\beta \ge g(1) = f(1) + \gamma(*)$, just prove this by
3147: induction on $\gamma(*)$. Let
3148: $\langle {\Cal I}_n:n < \omega \rangle$ list the dense open subsets of
3149: $Q_{\bar C}$ which belong to $N$ and let $\delta = \dbcu_{n < \omega}
3150: \alpha_n$ and $\alpha_n < \alpha_{n+1} < \delta$. We now simulate a strategy
3151: for the challenger in the game $\Game_{\alpha,\beta}(N)$, together with
3152: choosing in the $n$-th move, (in the end of it) also the challenger chooses
3153: $p_{n+1} \in Q_{\bar C} \cap N$ such that $p_0 = p,p_n \le p_{n+1} \in
3154: {\Cal I}_n$ and during the play letting $Z_n = \emptyset$
3155: (in fact also the chooser has to use $Y_n = \emptyset$) and $p_{n+1}$ is
3156: $(M_n,Q_{\bar C})$-generic and sup$(p_{n+1}) > \alpha_n$ and the set
3157: $(p_{n+1} \cup \{\sup p_{n+1}\}) \backslash (p_n \cup \{\sup p_n\})$
3158: is disjoint to
3159: $C_\delta$. This is possible by Case 1 and its proof because $M_n \cap
3160: \omega_1 \notin C_\delta$ which holds as $M_n \in Y'$. As this is a legal
3161: strategy for the challenger, so it cannot be a winning strategy hence for
3162: some such play the chooser wins hence $\{M_n:n < \omega\} \in D_1(N)$,
3163: remember $\alpha=1$. Now
3164: $q = \dbcu_{n < \omega} p_n$ is well defined, and sup$(q) = \delta$ and
3165: $q \cap C_\delta \subseteq p \cup \{\sup(p)\}$ and $q \Vdash_{Q_{\bar C}}
3166: ``\{M_n:n < \omega\} \subseteq {\Cal M}
3167: [{\underset\tilde {}\to G_{Q_{\bar C}}},N]"$, so $q$ is as required as the
3168: chooser has won the play.
3169: \bn
3170: \ub{Case 3}: $\alpha > 1,\alpha$ successor.
3171:
3172: Similar to Case 2 only we use the induction hypothesis instead of using
3173: Case 1.
3174: \bn
3175: \ub{Case 4}: $\alpha$ a limit ordinal. \nl
3176: Similar to Case 2. \nl
3177: Easy. \hfill$\square_{\scite{3.6}}$
3178: \bigskip
3179:
3180: \definition{\stag{3.7} Definition} 1) Assume
3181: \mr
3182: \item "{$(a)$}" $S \subseteq \omega_1$ is stationary
3183: \sn
3184: \item "{$(b)$}" $f \in {}^{\omega_1}(\omega_1)$
3185: \sn
3186: \item "{$(c)$}" $\bar C = \langle C_\delta:\delta < \omega_1 \rangle$
3187: \sn
3188: \item "{$(d)$}" $C_\delta$ is an unbounded subset of $\delta$.
3189: \ermn
3190: We define when ``$\bar C$ obeys $f$ on $S$" for $f \in {}^{\omega_1}\omega_1$
3191: a $({\Cal D}_{\omega_1} + S,\gamma)$-th function (see part (2)) by induction
3192: on $\gamma$. For $f$ being $({\Cal D}_{\omega_1} + S,\gamma)$-function,
3193: $\gamma < \omega_1$ this
3194: means $\{\delta \in S:\text{ otp}(C_\delta) \le \omega^{1+f(\delta)}\}
3195: = S$ mod ${\Cal D}_{\omega_1}$. In general it means that for some $g:
3196: \omega_1 \rightarrow \omega_1$ and pressing down function on $h$ on $S$, for
3197: every $\zeta < \omega_1$ for which $h^{-1}\{\zeta\}$ is stationary,
3198: for some $\beta < \gamma$ and $f_\beta$, a
3199: $({\Cal D} + h^{-1}\{\zeta\},\beta)$-th function, we have
3200: $\langle C_{g(\delta)} \cap
3201: \delta:\delta \in h^{-1}\{\zeta\} \rangle$ obeys $f_\beta$. \nl
3202: 2) We say $f \in {}^{\omega_1}(\omega_1)$ is a $({\Cal D}_{\omega_1} + S,
3203: \gamma)$-th function when:
3204: $S \Vdash_{({\Cal D}^+_{\omega_1},\supseteq)}$ ``in
3205: $V[\underset\tilde {}\to G]$, \nl
3206: $\{x \in V^{\omega_1}/\underset\tilde {}\to G:
3207: V^{\omega_1}/G \models x$ an ordinal $< f/ \underset\tilde {}\to G\}$
3208: has order type $\gamma$".
3209: \enddefinition
3210: \bigskip
3211:
3212: \proclaim{\stag{3.8} Claim} Assume
3213: \mr
3214: \item "{$(a)$}" ${\frak p}$ is a simple reasonable parameter,
3215: $\ell g({\frak p})$ of uncountable cofinality
3216: \sn
3217: \item "{$(b)$}" $S \in {\Cal D}^+_{\omega_1}$ and $N \in \dbcu_\alpha
3218: {\Cal E}^{\frak p}_\alpha \Rightarrow N \cap \omega_1 \in S$
3219: \sn
3220: \item "{$(c)$}" ${\frak p}$ is a non-$\Game_{\alpha,\alpha}$-loser
3221: (or just non $\Game'_{\alpha,\alpha}$-loser) for all
3222: $\alpha \in C^*,C^*$ a club of $\ell g({\frak p})$,
3223: \sn
3224: \item "{$(d)$}" $\bar C$ obeys $f$ on $S$ which is a
3225: $({\Cal D}_{\omega_1} +S,\gamma)$-function
3226: \sn
3227: \item "{$(e)$}" $0 < \text{ Min}(C),g(\alpha) = \text{ Min}
3228: (C^* \backslash \alpha)$.
3229: \ermn
3230: \ub{Then} $Q_{\bar C}$ is $({\frak p},g)$-proper.
3231: \endproclaim
3232: \bigskip
3233:
3234: \demo{Proof} Similar.
3235: \enddemo
3236: \bigskip
3237:
3238: \centerline {$* \qquad * \qquad *$}
3239: \bigskip
3240:
3241: Another example (which could have been done in \cite[Ch.XVIII]{Sh:f}) is:
3242: \definition{\stag{3.9} Definition} 1) We say $\bar{\Cal D} =
3243: \langle {\Cal D}_\delta:\delta < \omega_1 \text{ limit} \rangle$ is an
3244: $\omega_1$-filter-sequence if:
3245: \mr
3246: \item "{$(a)$}" ${\Cal D}_\delta$ is a filter on $\delta$, containing the
3247: co-bounded subsets
3248: \sn
3249: \item "{$(b)$}" ${\Cal D}_\delta$ is a $P$-filter and
3250: some $C_\delta \in D_\delta$ has order type $\omega$ ($P$-filter means that
3251: ${\Cal D}_\delta$ contains all co-finite sets and if $A_n \in D_\delta$
3252: for $n < \omega$
3253: then for some $A \in D_\delta$ we have
3254: $n < \omega \Rightarrow |A \backslash A_n| < \aleph_0$)
3255: (can generalize as in \scite{3.1}(2) but did not)
3256: \sn
3257: \item "{$(c)$}" for every club $C \subseteq \omega_1$ and $\alpha <
3258: \omega_1$, the set $A^\alpha_C[\bar{\Cal D}]$ is stationary, where,
3259: by induction on $\alpha$ we define: \nl
3260: $A^\alpha_C[\bar{\Cal D}] = \{\delta < \omega_1:\delta
3261: \text{ is a limit ordinal, } \delta \in C
3262: \text{ and for every } \beta < \alpha \text{ we have}$ \nl
3263: $\delta = \sup(\delta \cap A^\beta_C[\bar{\Cal D}])$, moreover
3264: $\delta \cap A^\beta_C[\bar{\Cal D}] \in {\Cal D}_\delta\}$.
3265: \ermn
3266: 2) We say a reasonable parameter ${\frak p}$ obeys $\bar{\Cal D}$ \ub{if}
3267: for each $\alpha < \ell g({\frak p})$ and $N \in {\Cal E}_\alpha$ we have:
3268: $\bar{\Cal D} \in N$ and we have
3269:
3270: $$
3271: \align
3272: {\Cal D}^{\frak p}_\alpha(N) = \biggl\{Y:&Y \subseteq N \cap
3273: \dbcu_{\beta < \alpha} {\Cal E}_\beta \text{ is closed (see \scite{2.11})
3274: and if} \\
3275: &\alpha > 0 \text{ then there are} \\
3276: &\bar \beta = \langle \beta_n:n < \omega \rangle,\bar M =
3277: \langle M_n:n < \omega \rangle \text{ satisfying} \\
3278: &(a) \quad \beta_n \in N \cap \alpha \text{ and} \\
3279: &(b) \quad \text{ either } \dsize \bigwedge_n [\alpha = \beta_n +1] \\
3280: &\qquad \text{ \ub{or} } \beta_n < \beta_{n+1},
3281: \sup_{n < \omega} \beta_n = \sup(\alpha \cap N) \\
3282: &(c) \quad M_n \in Y \cap {\Cal E}_{\beta_n},M_n \in M_{n+1}, \\
3283: &\qquad \dbcu_{n < \omega} M_n = N \cap \dbcu_{\beta \in \alpha \cap N}
3284: {\Cal H}(\chi_\beta) \\
3285: &(d) \quad \{M_n \cap \omega_1:n < \omega\} \in {\Cal D}_{N \cap \omega_1}
3286: \biggr\}.
3287: \endalign
3288: $$
3289: \mn
3290: 3) We say a forcing notion $Q$ is a $\bar{\Cal D}-NNR^0_\kappa$-forcing
3291: if for
3292: every reasonable parameter ${\frak p}$ which obeys $\bar{\Cal D}$
3293: we have: $Q$ is an $NNR^0_\kappa$-forcing over ${\frak p}$
3294: (see \scite{2.15}). \nl
3295: 4) For a $P$-filter ${\Cal D}$, we say a reasonable parameter
3296: ${\frak p}$ obeys ${\Cal D}$ if: \nl
3297: for every $N \in{\Cal E}_\alpha$
3298:
3299: $$
3300: \align
3301: D^{\frak p}_\alpha(N) = \biggl\{ Y:&Y \subseteq N \cap \dbcu_{\beta < \alpha}
3302: {\Cal E}_\beta \text{ is closed (see \scite{2.11}) and if } \alpha > 0
3303: \text{ then} \\
3304: &\text{for some } \bar \beta,\bar M \text{ satisfying clauses }
3305: (a), (b), (c) \\
3306: &\text{ of part } (2) \text{ and} \\
3307: &(d)' \qquad \{n:M_n \in Y\} \in {\Cal D} \\
3308: &\qquad \qquad (*) {\Cal D} \text{ is a } P \text{ -filter on }
3309: \omega \biggr\}.
3310: \endalign
3311: $$
3312: \mn
3313: 5) In the parts 1) - 4) above we may replace the word filter by ultrafilter
3314: if the ${\Cal D}$'s are ultrafilter (so we have ``$\omega_1$-ultrafilter
3315: sequence").
3316: \enddefinition
3317: \bigskip
3318:
3319: \proclaim{\stag{3.10} Claim} 1) If $\diamondsuit_{\aleph_1}$ (or much less),
3320: \ub{then} there is an $\omega_1$-ultrafilter sequence. \nl
3321: 2) If $\bar{\Cal D}$ is an $\omega_1$-filter sequence and
3322: $\langle (\chi_\alpha,{\Cal E}_\alpha):\alpha < \alpha^* \rangle$ is $a$ in
3323: Definition \scite{1.1}, \ub{then} there is a
3324: reasonable parameter ${\frak p}$ of length $\omega_1$ obeying $\bar{\Cal D}$
3325: which is a non-$\Game$-loser $\chi^{\frak p}_\alpha = \chi_\alpha,
3326: {\Cal E}^{\frak p}_\alpha = {\Cal E}_\alpha$. \nl
3327: 3) If $\diamondsuit_{\aleph_1}$ and $\langle (\chi_\alpha,{\Cal E}_\alpha):
3328: \alpha < \alpha^* \rangle$ is as in Definition \scite{1.1} and ${\Cal D}$ is
3329: a $P$-filter on $\omega$, \ub{then} some
3330: reasonable parameter ${\frak p}$ is $P$-filter like,
3331: non-$\Game_{\text{id}}$-loser with $\chi^{\frak p}_\alpha = \chi_\alpha,
3332: {\Cal E}^{\frak p}_\alpha = {\Cal E}_\alpha,\alpha^{*,{\frak p}} = \alpha^*$.
3333: Similarly for ultrafilters.
3334: \nl
3335: 4) Instead of $\diamondsuit_{\aleph_1}$ it is enough to assume CH and
3336: that for some $\langle C_\delta:\delta < \omega_1 \text{ limit} \rangle$
3337: and normal filter $D$ on $\omega_1$ and for every club $C$ of $\omega_1,
3338: \{\delta:\delta > \sup (C_\delta \backslash C)\} \in D$.
3339: \endproclaim
3340: \bigskip
3341:
3342: \demo{Proof} Straight (in part (4), we define the $C_\delta$'s
3343: simultaneously).
3344: \enddemo
3345: \bigskip
3346:
3347: \proclaim{\stag{3.11} Claim} 1) If ${\frak p}$ is a reasonable parameter
3348: obeying ${\Cal D}$, a $P$-filter on $\omega$ [or $P$-ultrafilter on
3349: $\omega$], \ub{then} for some $\bar{\Cal D},
3350: \bar{\Cal D}$ is an $\omega_1$-filter-sequence [or $\omega_1$-ultrafilter-
3351: sequence] and ${\frak p}$ obeys $\bar{\Cal D}$. \nl
3352: 2) If ${\frak p}$ is a $P$-point filter (or ultrafilter),
3353: then ${\frak p}$ is a non-$\Game$-loser.
3354: \endproclaim
3355: \bigskip
3356:
3357: \demo{Proof} Should be clear.
3358: \enddemo
3359: \bigskip
3360:
3361: \remark{\stag{3.12} Remark} We may consider higher order types than
3362: $\omega$.
3363: \endremark
3364: \bigskip
3365:
3366: \proclaim{\stag{3.13} Claim} 1) Assume
3367: \mr
3368: \item "{$(a)$}" $\bar C = \langle C_{\delta,\ell}:\ell < k_\delta,\delta <
3369: \omega_1 \text{ limit}\rangle,1 + \kappa \le k_\delta \le \omega,
3370: C_{\delta,\ell}$ is a closed unbounded subset of $\delta$ and
3371: $\ell < m < k_\delta \Rightarrow C_{\delta,\ell} \cap C_{\delta,m}
3372: = \emptyset$
3373: \sn
3374: \item "{$(b)$}" $Q = Q_{\bar C} = \{C:C$ is a closed bounded subset of
3375: $\omega_1$ such that for every limit $\delta < \sup(C)$ and for every $\ell
3376: < k_\delta$ except $< 1 + \kappa$ many, $\delta > \sup
3377: (C \cap C_{\delta,\ell})\}$
3378: \sn
3379: \item "{$(c)$}" ${\frak p}$ obeys the $P$-ultrafilter ${\Cal D}$ and is a
3380: reasonable parameter.
3381: \ermn
3382: \ub{Then} $Q$ is a ${\frak p}-NNR^0_\kappa$ forcing notion. \nl
3383: 2) In part (1) if we add: $N \in {\Cal E}_\alpha \and D^{\frak p}_\alpha(N) =
3384: \{Y:\{n:M_n \in Y\} \in {\Cal D}_N\},{\Cal D}_N$ a $P$-point ultrafilter
3385: (see above) and $\ell < k_\delta \Rightarrow \{n < \omega:M_n \cap \omega_1
3386: \in C_{\delta,\ell}\} = \emptyset \text{ mod } {\Cal D}_N$,
3387: \ub{then} we can allow $C_{\delta,0} = C_{\delta,1}$. \nl
3388: 3) Assume
3389: \mr
3390: \item "{$(a)$}" $D_\delta$ is a family of subsets of Dom$(D_\delta)$, the
3391: intersection $Y$ of any $<1 + \kappa$ of them satisfies
3392: {\roster
3393: \itemitem{ $(*)$ } $\exists n(\exists y_1,\dotsc,y_n \in Y)[\delta > \sup
3394: (\dbca^n_{\ell =1} C_{\delta,y_\ell})]$
3395: \endroster}
3396: \sn
3397: \item "{$(b)$}" $\bar C = \langle C_{\delta,x}:x \in \text{ Dom}(D_\delta)$
3398: and $\delta$ is a limit ordinal $< \omega_1 \rangle$
3399: \sn
3400: \item "{$(c)$}" $\langle C_{\delta,x}:x \in \text{ Dom}(D_\delta) \rangle$
3401: is a sequence of pairwise disjoint subsets of $\delta$
3402: \sn
3403: \item "{$(d)$}" $\bar X \in \dsize \prod_{\delta < \omega_1} \text{Dom}
3404: (D_\delta)$
3405: \sn
3406: \item "{$(e)$}" $Q_{\bar C,\bar X,\bar D} = \{C:C$ is a closed bounded
3407: subset of $\omega_1$ such that for every limit $\delta \le \sup(C)$ we have
3408: $(\exists x \in X_\delta)(\delta > \sup(C \cap C_{\delta,x}))\}$ \nl
3409: ordered by: end extension
3410: \sn
3411: \item "{$(f)$}" $\bar{\Cal D}$ is a $P$-ultrafilter sequence.
3412: \ermn
3413: \ub{Then} $Q$ is a ${\frak p}-NNR^0_\kappa$ forcing notion ${\frak p}$
3414: which obeys $\bar{\Cal D}$.
3415: \endproclaim
3416: \bigskip
3417:
3418: \demo{Proof} Straight. \nl
3419: 1) So let $V = V^{P_\alpha}_0,\Vdash_{P_\alpha} ``Q_\alpha =
3420: Q_{\underset\tilde {}\to {\bar C}}$ is as above" and in $V_0,N_0 \in
3421: {\Cal E}_0,N_0 \in N_1 \in {\Cal E}_\alpha = \{M_n:n < \omega\},D_{N_0}$ as
3422: in Definition \scite{3.13}(2), $p \in P_{\alpha +1} \cap N,G_m \subseteq
3423: N_1 \cap P_\alpha$ is generic over $N_1$ for
3424: $m < k < 1 + \kappa$ (so $k$ is now fixed).
3425: So for each $\ell$ we have ${\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell}}
3426: [G_m]$ is a closed subset of $\delta$ and for $\ell_1 < \ell_2 <
3427: k_{\delta_{N_0}}$ we have
3428: ${\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell_1}}[G_m]
3429: \cap {\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell_2}}[G_m] = \emptyset$
3430: so for some $\ell(m) \in \{0,1,\dotsc,k_{\delta_{N_0}}-1\}$ we have
3431: $\ell \ne \ell(m) \Rightarrow
3432: {\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell}}[G_m] = \emptyset$ mod
3433: $D_{N_0}$. Now let $B = \{n:$if $\ell < k_\delta,\ell \notin \{\ell(m):
3434: m < k\}$ and $\ell < n$ then $M_n \cap \omega_1 \notin
3435: {\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell}}[G_m]$ for $m < k$ and
3436: $p \in M_n\}$ belongs to $D_{N_0}$. Clearly, \wilog \, $\langle M_n:n <
3437: \omega \rangle \in N_1,D_{N_0} \in N_1,B \in N_1$. Let $B = \{n_i:i <
3438: \omega\},n_i$ increasing with $i$ and $\langle {\Cal I}_n:n < \omega
3439: \rangle$ list the dense open subsets of $P_{\alpha +1}$ which belongs
3440: to $N_0$. We can choose $p_i$ by induction on $i < \omega$ such that:
3441:
3442: $$
3443: p_i \in N_{n_i} \cap P_{\alpha +1}
3444: $$
3445:
3446: $$
3447: p_i \restriction \alpha \in \dbca_{m < k} G_m
3448: $$
3449:
3450: $$
3451: p_i \in \bigcap \{{\Cal I}_n:
3452: n < n_i,{\Cal I}_n \in N_{n_i} \text{ and } i > 0\}
3453: $$
3454:
3455: $$
3456: p \le p_i
3457: $$
3458:
3459: $$
3460: p_i \le p_{i+1}
3461: $$
3462:
3463: $$
3464: p_{i+1} \backslash p_i \text{ is disjoint to } \cup
3465: \{{\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell}}[G_m]:\ell < k_0,\ell <
3466: n_i \text{ and } m < k \Rightarrow \ell \ne \ell(m)\}.
3467: $$
3468: \mn
3469: This is possible as
3470: \mr
3471: \item "{$\bigotimes$}" $\biggl( \dsize \bigcup \Sb m < k \\
3472: \ell < k \and \ell \ne \ell(m) \endSb
3473: {\underset\tilde {}\to C_{N_0 \cap \omega_1,\ell}}[G_m] \biggr) \cap
3474: \omega_1 \cap M_n \text{ is a bounded subset of } M_n \cap \omega_1$.
3475: \ermn
3476: 2), 3), 4) Similarly. \hfill$\square_{\scite{3.13}}$
3477: \enddemo
3478: \bigskip
3479:
3480: \remark{\stag{3.14} Remark} 1) The forcing notion from \scite{3.13}(1) is
3481: $NNR^0_{\aleph_0}$-forcing notion for every ${\frak p}$,
3482: non-$\Game_{\text{id}}$-loser. \nl
3483: 2) If ${\frak p}$ is a reasonable parameter obeying $\bar{\Cal D}$, an
3484: $\omega_1$-filter sequence, then any $(< \omega_1)$-proper forcing is
3485: ${\frak p}$-proper.
3486: \endremark
3487: \bigskip
3488:
3489: \demo{Proof} Similar to earlier proofs.
3490: \enddemo
3491: \newpage
3492:
3493: \head {\S4 Second preservation of not adding reals} \endhead \resetall \sectno=4
3494: % \resetall
3495: \bigskip
3496:
3497: We shall concentrate on the simple case.
3498: \definition{\stag{4.1} Definition} Let ${\frak p}$ be a reasonable
3499: parameter. We say that $\bar Q$ is a ${\frak p}-NNR^1_\kappa$
3500: iteration (where $2 \le \kappa \le \aleph_0$ as we omit
3501: $\kappa = \aleph_1$ for convenience) if:
3502: \mr
3503: \item "{$(a)$}" cf$(\ell g({\frak p})) > \ell g(\bar Q)$ (for simplicity)
3504: \sn
3505: \item "{$(b)$}" $\bar Q$ is a countable support iteration of proper forcing
3506: notions \nl
3507: [in \cite{Sh:311} we will
3508: relax the requirement that each iterand is proper] \nl
3509: such that $i < \ell g(\bar Q) \Rightarrow P_i$ adds no reals \nl
3510: (follows by other parts (close (d)) even for $P_{i+1},i < \ell g(\bar Q)$
3511: \sn
3512: \item "{$(c)$}" [long properness]
3513: for each $i$, for some $f_i \in {\Cal F}^{\frak p}_{\text{club}}$
3514: we have $\Vdash_{P_i} ``{\underset\tilde {}\to Q_i}$ is
3515: $({\frak p},f_i)$-proper, $f_i$ increasing continuous, see Definition
3516: \scite{2.8}(2) + \scite{2.8}(4)
3517: \sn
3518: \item "{$(d)$}" [$\kappa$-anti w.d.] if
3519: $f_i(\alpha) \le \beta < \ell g({\frak p}),\{\bar Q,i,
3520: \alpha,\beta\} \in N_0 \in {\Cal E}_\alpha,N_0 \in N_1 \in {\Cal E}_\beta,
3521: k < 1 + \kappa,j = i + 1$ and for $\ell =0,1,\dotsc,k-1$ we have $q_\ell
3522: \in P_i$ is $(N_1,P_i)$-generic and is $(N_0,P_i)$-generic, $q_\ell
3523: \Vdash ``{\underset\tilde {}\to G_{P_i}} \cap N_1 = G^\ell",G^\ell \cap
3524: N_0 = G^*,p^* \in P_j \cap N_0$ satisfies $p^* \restriction i \in
3525: G^{**}$ and lastly $Y =: \dbca_{\ell < k} {\Cal M}[G^\ell,N_1,y^*] \in
3526: D_\beta(N_1)$, \ub{then} for some $G^{**} \in N_1,G^{**} \subseteq N_0
3527: \cap P_j$ is generic over $N_0,G^* \subseteq G^{**}$ and $p^* \in G^{**}$ and
3528: $\dsize \bigwedge_{\ell <k} \, \dsize \bigvee_{r \in G^\ell} r
3529: \Vdash_{P_i} ``G^{**} \restriction
3530: {\underset\tilde {}\to Q_i} (= \{p(i):p \in G^{**}\})$ has an upper bound
3531: in ${\underset\tilde {}\to Q_i}"$.
3532: \endroster
3533: \enddefinition
3534: \bigskip
3535:
3536: \remark{\stag{4.2} Remark} 1) $\kappa$ is the amount of
3537: ``$\Bbb D$-completeness", in other words what versions of weak diamond
3538: we kill by our iteration.
3539: \mn
3540: 2) \ub{Question}: Why do we ask for $f_i \in V$ and not a $P_i$-name
3541: ${\underset\tilde {}\to f_i}$ of such a function?
3542: \sn
3543: \ub{Answer}: If $i < \ell g(\bar Q) \Rightarrow P_i \models
3544: \text{ cf}(\ell g({\frak p}))$-c.c., then we can find $f'_i \in
3545: {\Cal F}^{\frak p},f_i \ge {\underset\tilde {}\to f_i}$; so as we are
3546: assuming, for conveniency cf$(\ell g({\frak p})) > |P_\alpha|$ there is
3547: no point at present for $f_i$ to be a $P_i$-name. \nl
3548: 2A) In clause (d) we have implicitly used:
3549: \mr
3550: \item "{$(*)$}" if $\alpha \le \beta' < \beta$ then clause (d) for
3551: $(\alpha,\beta)$ and $k$ implies clause (d) for $(\alpha,\beta')$ and $k$.
3552: \ermn
3553: This holds by clause (i) of Definition \scite{1.1}. \nl
3554: We could replace $f_i$ by a club $E_i$ of $\ell g({\frak p})$, letting
3555: $f_i(\alpha) = E_i \backslash \alpha$. \nl
3556: 3) No real reason to ``use some $\kappa$", but also no real
3557: reason to use 2. For each $\ell$ to find $G^{**}$ means essentially
3558: ``forcing with $Q_i$ adds no reals". The point is the common solution. \nl
3559: 4) In clause (c) for a club $C$ of $\ell g({\frak p})$ we
3560: catch our tail, that is $f_i(\alpha) \cap C = C \backslash \alpha$ for a club of
3561: $\alpha < \ell g({\frak p})$. We could use less, no real point
3562: now. \nl
3563: 5) In clause (d) much of the freedom/variation will be due to the
3564: decision how ``similar" are $\langle G^\ell:\ell < k \rangle$ such that
3565: $G^{**}$ exists. Here we demand
3566: \mr
3567: \item "{$(\alpha)$}" $Y \in D_\beta(N_1)$.
3568: \ermn
3569: In \cite[Ch.VIII,\S2]{Sh:f} it is essentially required that
3570: \mr
3571: \item "{$(\beta)$}" $G^0 \times G^1 \times \cdots \times G^{k-1} \subseteq
3572: (P_i \times \cdots \times P_i) \cap N_1$ ($k$ times) is \nl
3573: generic over $N_1$.
3574: \ermn
3575: In \cite[Ch.V]{Sh:f}
3576: \mr
3577: \item "{$(\gamma)$}" the common $Y$ is a predetermined increasing
3578: sequence of models.
3579: \ermn
3580: We have a trade-off; clause $(\beta)$ makes demand (d) in
3581: \scite{4.1} easier, but the parallel of (c) harder compared to clause
3582: $(\alpha)$. \nl
3583: This explains why the present work doesn't supercede Ch.XVIII,\S2. \nl
3584: 6) It is harder to win with a ``slower" function, so the
3585: assumption above is the strongest possible - although practically makes no
3586: difference, probably.
3587: \endremark
3588: \bigskip
3589:
3590: \proclaim{\stag{4.3} Claim} Let $\bar Q$ be a ${\frak p}-NNR^1_\kappa$
3591: iteration (see Definition \scite{4.1}), $\kappa \in \{2,\aleph_0\}$ and
3592: ${\frak p}$ is a reasonable parameter and ${\frak p}$ is a
3593: $\Game_f$-winner for some $f \in {\Cal F}^{\frak p}_{\text{club}}$ or at least
3594: $\Game'_f$-non loser. \nl
3595: 1) Forcing with $P_{\ell g(\bar Q)} = \text{ Lim}(\bar Q)$
3596: does not add reals (so
3597: consequently no $\omega$-sequences, as we are assuming properness). \nl
3598: 2) If $i \le j \le \ell g(\bar Q)$, \ub{then}
3599: \mr
3600: \item "{$(b)'$}" $P_j/P_i$ is proper
3601: \sn
3602: \item "{$(c)'$}" $P_j/P_i$ is $({\frak p},f_{i,j})$-proper, where \nl
3603: $f_{i,j} \in {\Cal F}^{\frak p}_{\text{club}}$ is increasing continuous and
3604: is computable from the
3605: $f_\varepsilon \in {\Cal F}^{\frak p}$ for $\varepsilon \in [i,j)$,
3606: \sn
3607: \item "{$(d)'$}" we have the parallel of clause (d), this time
3608: without assuming that $j = i+1$.
3609: \endroster
3610: \endproclaim
3611: \bigskip
3612:
3613: \demo{Proof} By induction on $\ell g(\bar Q)$. For notational simplicity
3614: we assume
3615: \mr
3616: \item "{$\boxtimes$}" all $f_i$ are even in ${\Cal F}^{\frak p}_{\text{end}}$
3617: so consider them as functions from $\ell g({\frak p})$ to $\ell g({\frak p})$
3618: (increasing continuous), and we demand also the $f_{i,j}$ are like that,
3619: increasing continuous and moreover $f_{i,j}(f_{i,j}(\alpha)) = f_{i,j}
3620: (\alpha)$ and they are $\ge f^*,f^* \in {\Cal F}^{\frak p}_{\text{end}}$
3621: increasing continuous, ${\frak p}$ is $\Game_{f^*}$-winner or at least
3622: $\Game'_{f^*}$-non-loser.
3623: \endroster
3624: \enddemo
3625: \bn
3626: \ub{Case 1}: \ub{$\ell g(\bar Q) = 0$}
3627:
3628: Trivial.
3629: \bn
3630: \ub{Case 2}: \ub{$\ell g(\bar Q) = i(*) + 1$}
3631: \mn
3632: \ub{Part (1)}: $P_{i(*)}$ adds no reals by the induction hypothesis,
3633: ${\underset\tilde {}\to Q_{i(*)}}$ adds no reals by clause (d) in
3634: Definition \scite{4.1}, hence $P_{i(*)+1} = P_{i(*)} *
3635: {\underset\tilde {}\to Q_{i(*)}}$ adds no reals.
3636: \mn
3637: \ub{Part (2)}: Clause $(b)'$ is known, (namely $P_j/P_i$ is proper).
3638: \mn
3639: \ub{Clause $(c)'$}:
3640:
3641: Given $i \le j \le \ell g(\bar Q)$ if $j < i(*)+1$ the conclusion follows
3642: by the induction hypothesis. So assume
3643: $j =i(*) +1$. If $i=j$, the required demand is trivial, so assume $i < j$.
3644: If $i=i(*)$, use
3645: clause (c) of Definition \scite{4.1} for $i$. So assume $i < i(*)$. Let
3646: $f_{i,j,0} = f_{i(*)},f_{i,j,m+1} = f_{i(*)} \circ f_{i,j(*)} \circ
3647: f_{i,j,m}$ and $f_{i,j}(\alpha) = \underset{m < \omega} {}\to \sup \,
3648: f_{i,j,m}(\alpha)$. Check that they are as required in $\boxtimes$. For
3649: proving ``$P_j/P_i$ is $({\frak p},f_{i,j})$-proper" assume
3650: \mr
3651: \item "{$(a)$}" $N \prec ({\Cal H}(\chi),\in)$ is countable
3652: \sn
3653: \item "{$(b)$}" $\{\bar Q,i,j,\alpha,\beta,f_{i,i(*)},f_{i(*)},
3654: f_{i,j}\} \in N$
3655: \sn
3656: \item "{$(c)$}" $\alpha \le f_{i,j}(\alpha) \le \beta < \ell g({\frak p})$
3657: \sn
3658: \item "{$(d)$}" $q \in P_i \text{ is } (N,P_i) \text{-generic}$
3659: \sn
3660: \item "{$(e)$}" $p \in N \cap P_j,p \restriction i \le q$
3661: \sn
3662: \item "{$(f)$}" $Y \in D_\beta(N)$
3663: \sn
3664: \item "{$(g)$}" $q \Vdash
3665: ``Y \subseteq {\Cal M}[{\underset\tilde {}\to G_{P_i}},N]"$.
3666: \ermn
3667: First we deal with version 2 and assume that ${\frak p}$ is simple.
3668: Choose $y^* \in N$ which codes enough; clearly
3669: $\beta' =: f_{i(*)}(\alpha)$ belongs to $N$. So
3670: $f_{i,i(*)}(\beta') \le \beta$, hence
3671: by the induction hypothesis there are $q',Y'$ such that:
3672: $p \restriction i(*) \le q,q \le q' \in P_{i(*)},q'$ is
3673: $(N,P_{i(*)})$-generic, $Y' \subseteq Y,Y'\in D_{\beta'}(N)$ and
3674: $q' \Vdash ``Y' \subseteq {\Cal M}
3675: [{\underset\tilde {}\to G_{P_i(*)}},N,y^*]"$.
3676: \mn
3677: Next, we apply clause (c) in the Definition \scite{4.1}
3678: for $i(*)$, so there are $q'',Y''$ such that $p \le q'',
3679: q' \le q'' \in P_{i(*)+1} = P_j,q''$ is $(N,P_j)$-generic,
3680: $Y'' \subseteq Y',Y'' \in D_\alpha(N)$ and $q'' \Vdash ``Y \subseteq {\Cal M}
3681: [{\underset\tilde {}\to G_{P_j}},N,y^*]"$.
3682:
3683: The proof for version 1 is similar.
3684: \mn
3685: \ub{Clause $(d)'$}: \nl
3686:
3687: Recall that we have demanded
3688: $f_{i,j}(f_{i,j}(\alpha)) = f_{i,j}(\alpha)$ (see $\boxtimes$ at the
3689: beginning of the proof).
3690:
3691: Let $N_0,N_1,G^*,\alpha,\beta,i,j,p,k,q_\ell$ (for $\ell < k$), $G^\ell$
3692: (for $\ell < k$) and $G^*$ be as in the assumption of clause (d)
3693: (from Definition \scite {4.1}) be given. \nl
3694: Without loss of generality $i < j,i < i(*),j = i(*) +1$ (as in the proof of
3695: clause $(c)$ the other cases are trivial). \nl
3696: First, choose $G^{**}$ for $G^*,\alpha,\beta,i,i(*),G^{**} \in N_1$.
3697: For each $\ell < k$, if for some $s_\ell \in G^\ell$ we have
3698:
3699: $$
3700: s_\ell \Vdash_{P_i} ``\text{in } P_{i(*)}/{\underset\tilde {}\to G_{P_i}}
3701: \text{ there is an upper bound to } G^{**}"
3702: $$
3703: \mn
3704: then (possibly increasing $s_\ell$, recalling $G_\ell$ is generic over
3705: $N^*$) there are $s_\ell \in G^\ell,r_\ell \in P_{i(*)} \cap N_1$ such that
3706: $s_\ell$ forces that $r_\ell$ is an upper bound to $G^{**}$ and without
3707: loss of generality
3708: $r_\ell \restriction i \le s_\ell$. Now \wilog \, $[G_{\ell_1} =
3709: G_{\ell_2} \Rightarrow s_{\ell_1} = s_{\ell_2}]$ and $[G_{\ell_1} \ne
3710: G_{\ell_2} \Rightarrow s_{\ell_1},s_{\ell_2}$ incompatible]. Now choose
3711: $r \in P_{i(*)} \cap N_1$ with domain $\subseteq i(*) \backslash i$ as
3712: follows:
3713: Dom$(r) = \dbcu_{\ell < k} \text{ Dom}(r_\ell) \backslash i$ and $r(\alpha)$
3714: is $r_\ell(\alpha)$ if $s_\ell \in {\underset\tilde {}\to G_{P_i}},\ell < k$
3715: and is $\emptyset_{P_\alpha}$ if this occurs for no $\ell$. Renaming
3716: $r \in N_i \cap P_{i(*)}$, Dom$(r) \subseteq i(*) \backslash i$ and
3717: $s_\ell \in G^\ell,r_\ell = s_\ell \cup r$ is above $G^{**}$ in
3718: $P_{i(*)}$. Let $\beta_\ell = f_{i,j,1 +\ell}(\alpha)$ for
3719: $\ell \le k$.
3720: We choose by induction on $\ell \le k$ the objects
3721: $Y_\ell,q'_\ell,M_\ell$ such that
3722:
3723: $$
3724: Y_0 = Y
3725: $$
3726:
3727: $$
3728: M_0 = N_1
3729: $$
3730:
3731: $$
3732: N_0 \in M_{\ell +1}
3733: $$
3734:
3735: $$
3736: M_{\ell +1} \in M_\ell \cap {\Cal E}_{\beta_{k-\ell}}
3737: $$
3738:
3739: $$
3740: Y_{\ell +1} \subseteq Y_\ell
3741: $$
3742:
3743: $$
3744: Y_\ell \in D_{\beta_\ell}(M_\ell)
3745: $$
3746:
3747: $$
3748: M_{\ell +1} \in Y_\ell
3749: $$
3750:
3751: $$
3752: q_\ell \le q'_\ell \in P_{i(*)}
3753: $$
3754:
3755: $$
3756: q'_\ell \text{ is } (M_{\ell +1},P_{i(*)}) \text{-generic}
3757: $$
3758:
3759: $$
3760: q'_\ell \text{ forces a value to } {\underset\tilde {}\to G_{P_{i(*)}}}
3761: \cap M_{\ell +1}
3762: $$
3763:
3764: $$
3765: q'_\ell \text{ is } (N_0,P_{i(*)}) \text{-generic}
3766: $$
3767:
3768: $$
3769: q'_\ell \Vdash ``Y_{\ell +1} \subseteq {\Cal M}
3770: [{\underset\tilde {}\to G_{P_{i(*)}}},M_\ell]"
3771: $$
3772:
3773: $$
3774: q'_\ell \restriction i \in G^\ell.
3775: $$
3776: \mn
3777: (In the older version we have to increase $\beta \, \omega$ times if $k$ not
3778: given, arriving to a fixed point.)
3779: \sn
3780: No problem, now apply clause (d) of the definition for $i(*),N_0,M_k$, \nl
3781: $\langle q'_\ell:\ell < k \rangle,Y_k,G^{**}$ and get $G^{***}$ as required.
3782: \mn
3783: \ub{Comment}: The $Y$ transfers information between the various generics.
3784: In \nl
3785: \cite[Ch.V]{Sh:f}, in the first proof after $\omega$ steps we are lost,
3786: \ub{but} having the common tower of models help us.
3787: \bn
3788: \ub{Case 3}: $\delta = \ell g(\bar Q)$ is limit.
3789: \sn
3790: \ub{Part (1)}: Follows from clause (d) in part (2) and part (2) is proved
3791: below.
3792: \bn
3793: \ub{Part (2)}: Let $f_{i,j}$ be fast enough (you can collect the demands
3794: used below). \nl
3795: Clause $(b)'$ is obvious. We first prove clause $(d)'$ and later $(c)'$.
3796: \mn
3797: \ub{Clause $(d)'$}: So again \wilog \, $i < j = \delta$. \nl
3798: Let $N_0,N_1,p,G^*,\alpha,\beta,k < 1 + \kappa$ and $G_\ell,q_\ell$ for
3799: $\ell < k$ be as in the assumption of clause (d) in Definition
3800: \scite{4.1} but for $i < j$.\nl
3801: We for simplicity use $\kappa = \aleph_0$. We choose $\gamma \in N_0,
3802: \alpha < \gamma < \beta$ such that $\gamma$ is larger enough,
3803: in particular: $i \le i' < j' < j \Rightarrow f_{i',j'}(\gamma) = \gamma$. \nl
3804: Choose $\langle i_m:m < \omega \rangle \in N_1$ such that:
3805:
3806: $$
3807: i_0 = i
3808: $$
3809:
3810: $$
3811: i_m < i_{m+1}
3812: $$
3813:
3814: $$
3815: \sup_{m < \omega} i_m = \sup(j \cap N_0).
3816: $$
3817: \mn
3818: Choose $y^* \in N \cap {\Cal H}(\chi_\gamma)$ coding enough. We choose
3819: $M_0,M_1,M_2,M_3,M_4 \in N_1 \cap {\Cal E}_\gamma \cap Y$ such that
3820:
3821: $$
3822: N_0 \in M_0 \in M_1 \in M_2 \in M_3 \in M_4
3823: $$
3824:
3825: $$
3826: Y \cap M_m \in D_\gamma(M_m) \text{ for } m < 5.
3827: $$
3828: \mn
3829: Choose $q'_\ell \in G^\ell \cap M_4$ above $G^\ell \cap M_3$ so \nl
3830: $q'_\ell$ is $(M_0,P_i)$-generic,$(M_1,P_i)$-generic,$(M_2,P_i)$-generic,
3831: $(M_3,P_i)$-generic. \nl
3832: Let $\langle {\Cal I}_m:m < \omega\rangle \in M_0$ list the dense open
3833: subsets of $P_j$ from $N_0$.
3834: \bigskip
3835:
3836: \definition{Explanation} Now we shall use the diagonal argument, choose
3837: ${\underset\tilde {}\to G_P} \cap N_0,p_m \in P_{i_m} \cap N_0,r_m \in
3838: P_{i_n}$ as usual, using in the $m$-th step $f_{i_m,i_{m+1}} +$ relevant
3839: $(d)'$ things. We fulfill the above in $M_4$, so in the end can find a
3840: solution in $N_1$, by using a canonical construction, e.g. each time the
3841: $<^*_\chi$-choice.
3842:
3843: But to carry this we need to have finitely many candidates for
3844: ${\underset\tilde {}\to G_{P_{i_m}}} \cap M_0$ with a common $Y_m$. \nl
3845: (Note: if $Q \times \ldots \times Q$ is proper, like in Ch.XVIII,\S2 we
3846: can get such a common $Y_m$, i.e. the present result is stronger for this
3847: stage.) \nl
3848: To get this in the inductive step, we need in step $m-1$ that for $M_1$ we
3849: just have finitely many candidates for
3850: ${\underset\tilde {}\to G_{P_{i_m}}} \cap M_1$, and in turn to get this in
3851: the step $m-1$ we use that in step $m-2$ for $M_2$ we have:
3852: from every maximal antichain we choose a finite subset. To get this we use
3853: that for $M_3$ we just ask $M_3 [{\underset\tilde {}\to G_{P_{i_{m-3}}}}]
3854: \cap V = M_3$. So along the
3855: way $N_0,M_0,M_1,M_2,M_3$ our induction demands go down, but slowly, so that
3856: in each step $m$, advancing for say $M_0$, we have to preserve less than
3857: really knowing ${\underset\tilde {}\to G_{P_{i_m}}} \cap M_0$, and are
3858: helped by our demand on $M_1$, just like in \cite[Ch.XVIII]{Sh:f}.
3859: So compared to \cite[Ch.V]{Sh:f}, we have a finite tower.
3860:
3861: Returning to the proof we choose by induction on
3862: $m < \omega$ the objects $r_m,G^*_m,p_m,n_m$, \nl
3863: $\langle G^\ell_m:\ell < n_m \rangle,Y_m$ such that
3864: \mr
3865: \item "{$(a)$}" $r_m \in P_{i_m} \cap M_4$
3866: \sn
3867: \item "{$(b)$}" Dom$(r_m) \subseteq [i,i_m)$
3868: \sn
3869: \item "{$(c)$}" $r_{m+1} \restriction i_m = r_m$ \nl
3870: (comment: $r_m$'s act as a generic for $N_0$)
3871: \sn
3872: \item "{$(d)$}" $q'_\ell \cup r_m \in P_{i_m}$ is $(M_0,P_{i_m})$-generic,
3873: $(M_1,P_{i_m})$-generic, $(M_2,P_{i_m})$-generic, $(M_3,P_{i_m})$-generic
3874: \sn
3875: \item "{$(e)$}" if $\ell < k,{\Cal J} \subseteq P_{i_m}$ is dense open and
3876: ${\Cal J} \in M_2$, \ub{then} for some finite \nl
3877: ${\Cal J} \subseteq {\Cal I}
3878: \cap M_2,{\Cal J}$ is predense above $q'_\ell \cup r_m$ \nl
3879: (like in the proof of ${}^\omega \omega$-bounding)
3880: \sn
3881: \item "{$(f)$}" $n_m < \omega$, and \nl
3882: $\ell < n_m \Rightarrow G^\ell_m$ is a subset of $P_{i_m} \cap M_0$ generic
3883: over $M_0$, \nl
3884: and $G^\ell_m \in M_1$
3885: \sn
3886: \item "{$(g)$}" $G^\ell_{m+1} \cap P_{i_m} \in \{G^\ell_m:\ell < n_m\}$
3887: \sn
3888: \item "{$(h)$}" $n_0 =k,G^\ell_0 = G^\ell \cap M_1$
3889: \sn
3890: \item "{$(i)$}" $q_\ell \cup r_m \Vdash ``{\underset\tilde {}\to G_{P_{i,m}}}
3891: \cap M_1 \in \{G^\ell_m:\ell < n_m\}"$
3892: \sn
3893: \item "{$(j)$}" $G^*_m$ is a subset of $P_{i_m} \cap N_0$ generic over
3894: $N_0$
3895: \sn
3896: \item "{$(k)$}" $G^*_m \subseteq G^\ell_m$, so $G^*_m \subseteq G^*_{m+1},
3897: G^*_0 = G^*$
3898: \sn
3899: \item "{$(l)$}" $p_m \in P_j \cap N$
3900: \sn
3901: \item "{$(m)$}" $p_0 = p$
3902: \sn
3903: \item "{$(n)$}" $p_m \restriction i_m \in G^*_m$
3904: \sn
3905: \item "{$(o)$}" $p_m \le p_{m+1} \in {\Cal I}_m$
3906: \sn
3907: \item "{$(p)$}" $Y_m \subseteq {\Cal M}[G^\ell_m,M_0,y^*]$
3908: \sn
3909: \item "{$(q)$}" $Y_m \in D_\gamma(M_0)$.
3910: \ermn
3911: Why is this sufficient? If we can carry out the induction, then without
3912: loss of generality the construction belongs to $N_1$. \nl
3913: So $G^{**} = \{s \in P_j \cap N_0:\dsize \bigvee_{n < \omega} s \le p_m\}$ is
3914: as required, as $q'_\ell =: q_\ell \cup \dbcu_m r_m \in P_j \cap N_1$, and
3915: is above $G^{**}$ and $p \le q'_\ell$.
3916: \enddefinition
3917: \bn
3918: \ub{Induction}:
3919: \sn
3920: \ub{$m=0$}: Trivial.
3921: \bn
3922: \ub{$m+1$}: \ub{Stage A}: Choosing $p_{m+1}$ is trivial, the demands are:
3923: $p_{m+1} \ge p_m,p_{m+1} \restriction i_m \in G^*_m$ and $p_{m + 1} \in
3924: {\Cal I}_m$.
3925: \bn
3926: \ub{Stage B}: Choosing $G^*_{m+1}$: apply the induction hypothesis using
3927: clause $(d)'$ of what we are proving with $i_m,i_{m+1},\gamma,f_{i_m,i_{m+1}}
3928: (\gamma),N_0,M_1$ here standing for $i,j,\alpha,\beta,N_0,N_1$ there.
3929: \bn
3930: \ub{Stage C}: Let $\{H^\ell_m:\ell < n_{m+1}\}$ list the possibilities of
3931: ${\underset\tilde {}\to G_{P_{i_m}}} \cap M_1$
3932: (by clause (e) this exists). \nl
3933: Without loss of generality $H^\ell_m \cap M_0 = G^{h(\ell)}_m$, for some
3934: function $h = h_m:n_{m+1} \rightarrow n_m$. \nl
3935: We choose $s^\ell_m \in P_{i_{m+1}} \cap M_1$, above $G^*_{m+1}$, such that
3936: $s^\ell_m \restriction i_m \in G^{h(\ell)}_m$. Now repeat the argument of
3937: the successor stage of shrinking $Y$ (but now we have a fixed $\gamma$!).
3938:
3939: So we can find $t^\ell_m \in P_{i_{m+1}} \cap M_1$, above $s^\ell_m,
3940: t^\ell_m \restriction i_m \in H^\ell_m,t^\ell_n \Vdash
3941: ``{\underset\tilde {}\to G_{P_{i_{m+1}}}} \cap M_0 =: G^\ell_{m+1}"$ such that
3942:
3943: $$
3944: Y_{m+1} =: \dbca_{i < n_{m+1}} {\Cal M}[G^\ell_{m+1},M_0,y^*]
3945: \in D_\gamma(M_0).
3946: $$
3947: \mn
3948: The rest is as in \S1 so we have finished
3949: proving clause $(d)'$ in the case $\ell g(\bar Q)$ is
3950: a limit ordinal (which is last).
3951: \bn
3952: \ub{Clause $(c)'$}: Again, \wilog \, $i < j = \delta$. So assume
3953: $f_{i,j}(\alpha) \le \beta,\{i,j,\alpha,\beta\} \in
3954: N^* \in {\Cal E}_\beta,q$ is
3955: $(N^*,P_i)$-generic, $Y^* \in D_\beta (N^*),q \in P_\alpha$ and
3956: $q \Vdash ``Y^* \subseteq {\Cal M}
3957: [{\underset\tilde {}\to G_{P_i}},N^*,y^*]"$ are given. We prove the desired
3958: conclusion by induction on $\alpha$. For each $\alpha$, we would like to
3959: simulate a play of $\Game_{\alpha,\beta}(N^*)$, supplying the challenger
3960: with a strategy. For this we apply the proof of clause $(d)'$. \nl
3961: Choose $N_0,N_1,M_0,\dotsc,M_4,q_0,G_0,G^*$
3962: (and $k=1$) as there for some $\alpha'
3963: < \gamma' < \beta'$ as there such that $\beta < \alpha'$
3964: such that $N^*,q,Y^* \in N_0$ (easy to find).
3965: \bn
3966: \ub{Explanation}: Note: as ${\frak p}$ is simple, we can use
3967: $N \prec ({\Cal H}(\chi),\in),N \cap {\Cal H}
3968: (\chi_\beta) \in {\Cal E}_\beta$ with no mention - there were other places we
3969: could have said so. During the construction this time we demand $p_m \in N^*
3970: \cap P_j$, so a generic for $N_0$ is not necessarily created. But still
3971: $p_m \le p_{m+1},p_m \restriction i_m \in G^*_m$. Now $p_m$ will be played
3972: by the chooser. Now
3973: $g(1 + \alpha)$ will be a fixed point of $f_{i_m,i_m + 1}$. So we can add the
3974: demand $N^*[G^*_m \cap N^*] \cap V = N^*$, i.e. $G^*_m$ is generic over $N^*$
3975: and
3976:
3977: $$
3978: {\Cal M}[G^*_m \cap N^*,N^*,y^*] \in D_{\delta(1 + \alpha)}(N^*).
3979: $$
3980: \mn
3981: The challenger chooses
3982:
3983: $$
3984: X_{m+1} = {\Cal M}[G^*_m \cap N^*,N^*,y^*] \cap \dbcu_{\xi < \gamma
3985: (1+ \alpha)} {\Cal E}_\xi \cap \{M:p_n \in M\} \in N_0.
3986: $$
3987: \mn
3988: Now the chooser chooses $\alpha_m,\beta'_m$ and the challenger chooses
3989: $\beta_n \ge \beta'_m,f_{i_m,i_{m+1}}(\alpha)$ in $N^*_0 \cap j$ and the
3990: chooser chooses $M^*_m$ such that $p_m \in M_{m+1}$.
3991: \bn
3992: \ub{Comment}: The game was defined with this point in mind. \nl
3993: Now the chooser chooses $Y_n \in D_{\beta_n}(M_0),Y_n \subseteq X_n \cap M_0,
3994: Y_n \in N^*_0$ (check definition of game!) \nl
3995: Now we play $Z_n$ for the challenger as follows:
3996:
3997: there is $p_{m+1} \ge p_m,(M^*_{m+1},P_{i_{m+1}})$-generic
3998:
3999: $p_{m+1} \restriction i_m \in G^*_m$
4000:
4001: and $p'_{m+1}$ forces ${\underset\tilde {}\to G_{P_{i_{m+1}}}} \cap N^*$ and
4002:
4003: forces $Z_n = Y_n \cap {\Cal M}[{\underset\tilde {}\to G_{P_{i_{m+1}}}},
4004: M^*_m,y^*] \in D_{\alpha_n}(M^*_m)$. \nl
4005: As the challenger does not have a winning strategy, there is a play he wins,
4006: giving $\dbcu_{n < \omega} G^*_n \cap N^*_0$ with a bound.
4007: \bn
4008: \ub{Continuation of the Proof}: Choose also $\langle i'_m:m < \omega
4009: \rangle \in N_0$ such that $i_m \in N^*,i_0 = i,i_m < i_{m+1}$,
4010: sup$\{i_m:m < \omega\} = \text{ sup}(N^* \cap j)$ and this time we let
4011: $\langle {\Cal I}'_m:m < \omega \rangle$ list the dense open subsets of
4012: $P_j$ from $N^*$. For $\bold m < \omega$ let ${\Cal T}_{\bold m}$ be the
4013: set of the finite sequences ${\frak x}$ from $M_4$ coding $\langle
4014: r_{{\frak x},m}:m \le \bold m \rangle,\langle G_{{\frak x},m}:m \le
4015: \bold m \rangle,\langle p_{{\frak x},m}:m \le \bold m \rangle,\langle
4016: n_{{\frak x},m}:m \le \bold m \rangle,\langle G^\ell_{{\frak x},m}:\ell \le
4017: n_{{\frak x},m},m \le \bold m \rangle,\langle Y_{{\frak x},m}:m \le \bold m
4018: \rangle$ and also $(X_{{\frak x},m},\alpha_{{\frak x},m},\beta'_{{\frak x},m},
4019: \beta_{{\frak x},m},M_{{\frak x},n},y'_{{\frak x},m},
4020: M'_{{\frak x},m},y_{{\frak x},m})$ for $m \le \bold m$ and
4021: $Z_{{\frak x},m}$ for $m < \bold m$ satisfying:
4022: \sn
4023: clauses (a)-(k),(m),(n),(p),(q) from the proof of (d)$'$ above and
4024: \mr
4025: \item "{$(l)'$}" $p_m \in _j \cap N^*$
4026: \sn
4027: \item "{$(o)'$}" $p_m \le p_{m+1} \in {\Cal I}'_n$
4028: \sn
4029: \item "{$(r)'$}" $r_{{\frak x},m}$ is $(N^*,G_{i_m})$-generic for
4030: $m \le \bold m$
4031: \sn
4032: \item "{$(s)'$}" $\langle (X_{{\frak x},m},\alpha_{{\frak x},m},
4033: \beta'_{{\frak x},m},\beta_{{\frak x},m},M_{{\frak x},m},Y_{{\frak x},m},
4034: M'_{{\frak x},m},Z_{{\frak x},m'}):m \le \bold m \rangle$
4035: belongs to $N$ and is an initial segment of a play of the game
4036: $\Game'_{\alpha,\beta}(N^*,{\frak p})$ or just $\Game'_{\alpha,\beta}(N^*,N,
4037: {\frak p})$, note that in the $\bold m$-th move the challenger has not yet
4038: chose $Z_{{\frak x},\bold m}$, (see clause (e) of Definition \scite{2.2}(1)
4039: \sn
4040: \item "{$(t)'$}" $Z_{{\frak x},m} \subseteq Y_{m+1}$
4041: \sn
4042: \item "{$(u)'$}" $y_{{\frak x},m}$ code $p_m,\langle i_m:m < \omega
4043: \rangle$
4044: \sn
4045: \item "{$(v)'$}" $f_{i_m,i_{m+1}}(\alpha_m) \le \beta'_m$ for $m \le
4046: \bold m$.
4047: \ermn
4048: Let ${\frak x} \triangleleft {\frak y}$ has the natural meaning for ${\frak x}
4049: \in {\Cal T}_{{\bold m}_1},{\frak y} \in {\Cal T}_{{\bold m}_2},
4050: \bold m_1 < \bold m_2$. \nl
4051: Note
4052: \mr
4053: \item "{$\boxtimes_1$}" ${\Cal T}_{\bold m} \subseteq N$ for $\bold m <
4054: \omega$
4055: \sn
4056: \item "{$\boxtimes_2$}" ${\Cal T_0} \ne \emptyset$
4057: \sn
4058: \item "{$\boxtimes_3$}" if $\bold x \in {\Cal T}_{\bold m}$, \ub{then}
4059: considering the game $\Game_{\alpha,\beta}(N^*,{\frak p}),{\frak x}$ is an initial
4060: segment of a play of it (see clause $(s)'$ above).
4061: \ermn
4062: Assume
4063: \mr
4064: \item "{$(*)$}" $M'_{\bold m} \in Y_{{\frak x},\bold m} \cap
4065: {\Cal E}_{\alpha_{\bold m}} \cap (M_{{\frak x},\bold m} \cup
4066: \{M_{{\frak x},\bold m} \cap {\Cal H}(\chi^{\frak p}_{\bold m})\}$
4067: satisfying $y_{\bold m},y'_{\bold m} \in M'_{\bold m}$ and $Z_{\bold m}
4068: \subseteq {\Cal D}_{\alpha_{{\frak x},m}}(M'_m),Z_{\bold m} \subseteq
4069: Y_{\bold m}$ (hence $Z_{\bold m} \subseteq X_{{\frak x},\bold m})$ and
4070: $X_{\bold m +1} \in D_\beta(N^*) \cap X_{{\frak x},\bold m}$ such that
4071: $Z_{\bold m} \subseteq X_{\bold m +1}$ and $\alpha_{\bold m +1} \in
4072: \alpha \cap N^*$ and any $\beta'_{\bold m +1} \in \beta \cap N^* \backslash
4073: p_{i_{\bold m +1},i_{\bold m_r}}(\alpha_{\bold m +1}),y'_{\bold m +1} \in
4074: N \cap {\Cal H}(\chi^{\frak p}_{\alpha_{\bold m +1}})$
4075: and any $\beta_{\bold m}
4076: \in \beta \cap N \backslash \beta'_n \backslash \alpha_n$ and
4077: $M_{\bold m +1} \in X_{\bold m + 1} \cap {\Cal E}_{\beta_{\bold m + 1}},
4078: y_{\bold m + 1} \in M_{\bold m + 1} \cap {\Cal H}(\chi^{\frak p}
4079: _{\alpha_{\bold m}})$ satisfying $y'_{\bold m + 1} \in M_{\bold m + 1}$ and
4080: $Y_{\bold m + 1} \in N \cap D_{X_{\bold m + 1}}(M_{\bold m + 1})$ and any
4081: $M'_{\bold m + 1} \in Y_{\bold m + 1} \cap {\Cal E}_{\alpha_n} \cap
4082: (M_{\bold m + 1} \cup \{M_{\bold m + 1} \cap {\Cal H}(\chi^{\frak p}
4083: _{\alpha_m})\}$
4084: satisfying $y_{\bold m + 1},y'_{\bold m +1} \in M'_{\bold m + 1}$.
4085: \ermn
4086: \ub{Then}
4087: \mr
4088: \item "{$(**)$}" \ub{there} is ${\frak y} \in {\Cal T}_{\bold m +1}$ such that ${\frak x}
4089: \triangleleft {\frak y}$ and $(Z_{{\frak y},\bold m},
4090: X_{{\frak y},\bold m + 1}$, $\alpha_{{\frak y},\bold m +1},
4091: \beta'_{{\frak y},m + 1},y'_{{\frak y},\bold m + 1}$, \nl
4092: $\beta_{{\frak y},\bold m + 1},y_{{\frak y},\bold m +1},
4093: M_{{\frak y},\bold m + 1},Y_{{\frak y},\bold m + 1},M'_{{\frak y},\bold m+1})$
4094: is equal to \nl
4095: $(Z_{\bold m},X_{\bold m + 1},\alpha_{\bold m + 1},
4096: \beta'_{\bold m +1}$,
4097: $y'_{\bold m + 1},\beta_{\bold m + 1},y_{\bold m + 1}$,
4098: $M_{\bold m + 1},Y_{\bold m + 1},M'_{\bold m + 1})$. \nl
4099: Why? Because $f_{i_m,i_{m+1}}(\alpha_{\bold m}) \le \beta_{\bold m}$, hence
4100: we know that $P_{i_{\bold m + 1}}/P_{i_{\bold m}}$ is $({\frak p},\alpha_n,
4101: \beta_n)$-proper and let $G_{i_{\bold m}} \subseteq P_{i_m}$ be generic
4102: over $V,r_{\bold m} \in G_{i_{\bold m}}$ to the model
4103: $M'_{{\frak x},\bold m}$ and the set $Y_{{\frak x},m}$.
4104: So now we can describe a strategy for the challenger in the game
4105: $\Game_{\alpha,\beta}(N^*,{\frak p})$ (or $\Game'_{\alpha,\beta}(N^*,N,
4106: {\frak p})$) delaying his choice of $M'_m,Z_m$ to the $(m+1)$-th move, he
4107: just chose on the side ${\frak x}_m \in {\Cal T}_m$ which ``code" they play
4108: so far, and preserve ${\frak x}_m \triangleleft {\frak x}_{m+1}$.
4109: \ermn
4110: By $\boxtimes_3$ this is O.K. - all possible choices of the chooser are
4111: allowed, that is this gives a well defined strategy for the challenger (will
4112: he have some free choice, this does not hurt). But the chosen does not lose
4113: the game, so there is such a choice $\langle {\frak x}_{\bold m}:\bold
4114: m < \omega \rangle$ with $\cup\{(M'_{{\frak x}_{m+1},\bold m}\} \cup
4115: Y_{{\frak x}_{m+1},\bold m}:\bold m < \omega\} \in D_\alpha(N)$ and
4116: $\dbcu_{m < \omega} r_m$ is as required.
4117: \bn
4118: \centerline {$* \qquad * \qquad *$}
4119: \bn
4120: The adaptation for the proof when $\kappa=2$ should be clear.
4121: \bigskip
4122:
4123: \demo{Proof} Similar to the proof of \scite{1.7}, with some changes. We
4124: choose $\langle j_n:n < \omega \rangle \in M_0$ such that $j < j_n < j_{n+1}
4125: < h^*(j,i),h^*(j_n,i_m) < j_{n+1}$
4126: \mr
4127: \item "{$(l)$}" $n_m$ is a power of 2, say $2^{n^*_m}$ and so we can rename
4128: $\{G^\ell_m:\ell < n_m\}$ as $\{G^\eta_m:\eta \in 2^{n^*_m}\}$
4129: \sn
4130: \item "{$(m)$}"
4131: {\roster
4132: \itemitem{ $(\alpha)$ } $M_\eta \in M_1$ for $\eta \in {}^{(n^*_m \ge)}2$
4133: \sn
4134: \itemitem{ $(\beta)$ } $M_{<>} = N_0$
4135: \sn
4136: \itemitem{ $(\gamma)$ } $M_\eta \in M_{\eta \char 94 \langle 0 \rangle} \cap
4137: M_{\eta \char 94 \langle 1 \rangle}$
4138: \sn
4139: \itemitem{ $(\delta)$ } $\eta \triangleleft \nu_1 \in {}^{n^*_m}2,
4140: \eta \triangleleft \nu_2 \in {}^{n^*_m}2 \Rightarrow G^{\nu_1}_m \cap M_\eta
4141: = G^{\nu_2}_m \cap M_\eta$ so we call it $K^\eta_m$
4142: \sn
4143: \itemitem{ $(\varepsilon)$ } $M_{\eta \char 94 \langle 0 \rangle} =
4144: M_{\eta \char 94 \langle 1 \rangle}$ when $\eta \in {}^{n^*_m}2$ call it
4145: $N_\eta$
4146: \sn
4147: \itemitem{ $(\zeta)$ } $N_\eta \in {\Cal E}_{\ell g(\eta)}$ for
4148: $\eta \in {}^{(m^*_m >)}2$
4149: \sn
4150: \itemitem{ $(\eta)$ } $Y_\eta = {\Cal M}[K_{\eta \char 94 \langle 0 \rangle},
4151: N_\eta] \cap {\Cal M}[K_{\eta \char 94 \langle 1 \rangle},N_\eta] \in
4152: D_{j_{\ell g(\eta)}}(N_\eta)$.
4153: \endroster}
4154: \endroster
4155: \enddemo
4156: \bn
4157: \ub{Discussion}: We may be interested in non-proper forcing, say
4158: semi-proper and UP ones (see \cite[Ch.X,XI,XV]{Sh:f} and \cite{Sh:311}).
4159: Here the change from (reasonable) parameter ${\frak p} =
4160: {\frak p}^V$ to ${\frak p}^{V[G]}$ is more serious as $\{N \cap \chi_\alpha:
4161: N \in {\Cal E}^{{\frak p}^{V[G]}}\}$ is in general not equal to
4162: $\{N \cap \chi_\alpha:N \in {\Cal E}^{\frak p}_\alpha\}$. We intend to spell
4163: it out in \cite{Sh:311}.
4164:
4165: We may be interested in combining this work with \cite{Sh:587},
4166: \cite{Sh:F259}. Intend to
4167: deal with it later (see also \cite{Sh:655}).
4168: \newpage
4169:
4170: \head {\S5 Problematic forcing} \endhead \resetall \sectno=5
4171: % \resetall
4172: \bn
4173: \ub{\stag{5.1} Discussion}: 1) In the examples
4174: the ``barely adding reals but not so"
4175: may occur on some stationary ${\Cal E} \subseteq [\chi^*]^{\aleph_0}$ and
4176: otherwise just properness is asked. We do not bother to do this in the
4177: examples. \nl
4178: 2) We may like to put the present lemmas and \cite[Ch.XVIII,\S2]{Sh:f} and
4179: more together. The way is clear, we concentrate on $\kappa \in \aleph_0$.
4180:
4181: So instead of a ``tower" with six levels we have one with $n(*)+1$ levels.
4182: \bigskip
4183:
4184: \definition{\stag{5.2} Definition} 1) We say $\bar Q$ is an
4185: ${\frak p}-NNR^0_{\aleph_0,{\bar Pr},\bar \Xi}$-iteration for ${\frak p}$
4186: is:
4187: \mr
4188: \item "{$(a)$}" $\bar Q$ is a CS iteration of proper notions forcing,
4189: ${\frak p}$-proper
4190: \sn
4191: \item "{$(b)$}" forcing with Lim$(\bar Q) = P_{\ell g(\bar Q)}$ does not
4192: add reals
4193: \sn
4194: \item "{$(c)$}" $\bar Pr = \langle Pr_\ell:\ell < n(*) \rangle,\bar \Xi =
4195: \langle \Xi_\ell:\ell \le n(*) \rangle,\Xi_\ell \subseteq \ell g({\frak p})
4196: \times \ell g({\frak p})$
4197: \sn
4198: \item "{$(d)$}" $Pr_\ell(N,\bar G,P)$ implies: $P$ a forcing notion,
4199: $N$ is a countable elementary submodel of $({\Cal H}(\chi),\in),
4200: \bar G = \langle G_m:m < k \rangle,k < \omega,G_m \subseteq N \cap P$ is
4201: generic over $N$
4202: \sn
4203: \item "{$(e)$}" if $\ell = 0,Pr_\ell(N,\bar G,P)$, then
4204: $\dsize \bigwedge_\ell G_\ell = G_0$
4205: \sn
4206: \item "{$(f)$}" if $\ell = n(*)-1$, \ub{then} $Pr_\ell(N,\bar G,P)$, iff
4207: the demand in clause (d) holds
4208: \sn
4209: \item "{$(g)$}" if Pr$_\ell(N,\bar G,P)$ and
4210: $\bar G \triangleleft \bar G',\ell g(\bar G')$
4211: finite and Rang$(\bar G) = \text{ Rang}(\bar G')$ \ub{then} $Pr_\ell(N,
4212: \bar G',P)$
4213: \sn
4214: \item "{$(h)$}" $Pr_0(N,\bar G,P)$ holds iff (the condition in clause (c)
4215: and) $\dsize \bigwedge_\ell G_\ell = G_0$
4216: \sn
4217: \item "{$(i)$}" if $\ell < n(*),i < j \le \ell g(\bar Q),N_0 \prec N_1
4218: \prec ({\Cal H}(\chi),\in),{\frak p} \in N_0,N_0,N_1$ are countable and for
4219: some $(\alpha,\beta) \in \Xi_\ell \cap N_0$ we have $N_0 \cap {\Cal H}
4220: (\chi^{\frak p}_\alpha) \in {\Cal E}_\alpha,N_1 \cap {\Cal H}(\chi^{\frak p}
4221: _\beta) \in {\Cal E}_\beta,Pr_\ell(N_0,\bar G,P_i),Pr_{\ell +1}(N_1,\bar H,
4222: P_i),\ell g(\bar G) = \ell g(\bar H),G_k \subseteq H_k,\bar p = \langle p_k:
4223: k < \ell g(\bar G) \rangle,p_\ell \in N_0 \cap P_j,p_k \restriction i \in
4224: G_k,\ell = 0 \Rightarrow p_k = p_0$, \ub{then} we can find $\bar G^+ =
4225: \langle G^+_k:k < \ell g(\bar G) \rangle \in N_1$ such that $Pr(N_0,\bar G^+,
4226: P_j)$ and $p_k \in G^+_k$
4227: \sn
4228: \item "{$(j)$}" assume $Pr_\ell(N,\bar G,P_i),i_n < i_{n+1},i_n \in N,i_0=i,
4229: \sup_{n < \omega} i_n = \sup(N \cap j),(\alpha,\beta) \in N \cap \Xi_\ell,N'=
4230: N \cap {\Cal H}(\chi^{\frak p}_\alpha) \in {\Cal E}_\alpha$, \ub{then} in
4231: the following game the $Pr^+_\ell$-player has a winning strategy (or just
4232: does not lose).
4233: \ermn
4234: Before the $n$-th move $\bar G^n$ such that $Pr_\ell(N,\bar G^n,P_{i_n})$ is
4235: chosen $\ell g(\bar G^n) = \ell g(\bar Q)$ with $\bar G^0 = \bar G$. In the
4236: $n$-th move the challenger chooses $\bar p^n = \langle p^n_k:k < \ell g
4237: (\bar G) \rangle,p^n_k \in N_0 \cap P_{in+1},P^n_k \restriction i_n \in
4238: G^n_k,\ell = 0 \Rightarrow p^n_k = p^n_0$ and then the chooser chooses
4239: $\bar G^{n+1}$ is above such that $p^n_k \in G^{n+1}_k$. In the end of the
4240: play the chooser wins if $Pr_\ell(N,\bar G^*,P_j)$ where $G^*_k = \{p \in
4241: N \cap P_j:\text{for every } n,p \restriction i_n \in G^n_k\}$.
4242: \enddefinition
4243: \bn
4244: \ub{\stag{5.3} Comment}: 0) Of course,
4245: we can also use a reasonable parameter for induction. \nl
4246: 1) In \cite[Ch.XVIII,\S2]{Sh:f} for $NNR^2_\kappa$
4247: we can use weaken clause (5):
4248: $t$ a tree with 2 branches if $\bar Q \in N \prec ({\Cal H}(\chi),\in)$, if
4249: $i^* < \alpha,P_{i^*,i} =: \{(q,p_0,p_1) \in P_i:p_0 \restriction i^* =
4250: p_1 \restriction i^*\}$, then for $i < j \le \delta,i$ non-limit we have:
4251: $P_{i^*,i}/P_{i^*,j}$ is ${\Cal E}_2$-proper.
4252:
4253: In light of the previous proofs this is straight. \nl
4254: 2) We can replace $D_\alpha(N)$ by $D_{\alpha,i,j}(N)$ for $i < j \le
4255: \delta^*,\delta^*$ the supremum of the length of the iterations we consider.
4256:
4257: %% you may want to move the following lines up a bit
4258: \newpage
4259:
4260: \shlhetal
4261:
4262: REFERENCES.
4263: \bibliographystyle{lit-plain}
4264: \bibliography{lista,listb,listx,listf,liste}
4265:
4266:
4267: \enddocument%%
4268:
4269:
4270:
4271: \bye
4272:
4273: