1: \documentclass[11pt]{article}
2: \textheight=205mm
3: \textwidth=16.5cm
4: \hoffset=-2.2cm
5: \def\baselinestretch{1.0}
6: %\usepackage[textures]{graphicx} % Mac version
7: \usepackage[dvips]{graphicx} % XXX archive version
8: \begin{document}
9:
10: \title{\LARGE \bf PROCESS PHYSICS: INERTIA,
11: GRAVITY\\ and the QUANTUM\footnote{ Contribution to the 3rd Australasian
12: Conference on
13: General Relativity and Gravitation, Perth, Australia, July 2001} }
14: \author{{Reginald T. Cahill}\\
15: {School of Chemistry, Physics and Earth Sciences}\\{ Flinders University
16: }\\ { GPO Box
17: 2100, Adelaide 5001, Australia }\\{(Reg.Cahill@flinders.edu.au)}}
18:
19: \date{}
20: \maketitle
21:
22: \begin{abstract}
23: {\it Process Physics} models reality as self-organising
24: relational or semantic information using a self-referentially limited neural
25: network model. This generalises the traditional non-process syntactical
26: modelling of reality by taking account of the limitations and
27: characteristics of self-referential syntactical information systems,
28: discovered by G\"{o}del and Chaitin, and the analogies with the standard
29: quantum formalism and its limitations. In process physics space and quantum physics
30: are emergent and unified, and time is a distinct non-geometric process. Quantum
31: phenomena are caused by fractal topological defects embedded in and
32: forming a growing three-dimensional fractal process-space. Various
33: features of the emergent physics are briefly discussed including: quantum
34: gravity, quantum field theory, limited causality
35: and the Born quantum measurement metarule, inertia,
36: time-dilation effects, gravity and the equivalence principle, a growing universe with a cosmological
37: constant, black holes and event horizons, and the emergence of classicality.
38: \end{abstract}
39:
40: \vspace{20mm}
41: \hspace{5mm} Key words: process physics, G\"{o}del's theorem, neural network, semantic
42:
43: \hspace{27mm}information, self-referential noise, process-time, process-space,
44:
45: \hspace{27mm}quantum gravity.
46:
47: \newpage
48:
49: \section{Introduction}
50: There is mounting evidence that a unification of gravity and quantum theory has finally
51: been achieved, but only after the realisation that the failure to do so, until recently, arose
52: from deep limitations to the traditional modelling of reality by physicists. From its
53: earliest inception physics has modelled reality using formal or syntactical
54: information systems. These have undefined {\it a priori} entities, such a fields and
55: geometry, together with {\it a priori} laws. However the actual structure has
56: always been a little messier than that: it is in all cases composed of mathematical
57: equations supplying the core structure, together with metarules that make up for
58: deficiencies of the mathematical model, and finally some metaphysical assertions that
59: usually have an ontological flavour.
60:
61: A simple early example is actually Newton's geometrical modelling of the
62: phenomena of time. There the mathematical structure is the one-dimensional
63: continuum or geometrical line, but that fails because it has no matching for the present
64: moment effect or even the distinction between past and future. The geometrical-time metarule
65: here involves the notion that one must actively imagine a point moving along a line at a
66: uniform rate, and in this regard the metarule is certainly not inconsistent with the
67: mathematical model. Finally the metaphysical assertion is that time {\it is} geometrical.
68: Clearly then from the beginning physicists have blurred the three components of modelling.
69:
70: The ongoing failure of physics to fully match all the aspects
71: of the phenomena of time, apart from that of order, arises because physics has always
72: used non-process models, as is the nature of formal or syntactical systems. Such systems
73: do not require any notion of process - they are entirely structural and static. The new
74: process physics
75: \cite{CK97,CK98,CK99,CKK00} overcomes these deficiencies my using a non-geometric
76: process model for time (see \cite{MC} for an early non-technical account), but process physics also argues
77: for the importance of relational or semantic information in modelling reality. Semantic information
78: refers to the notion that reality is a purely informational system where the information is
79: internally meaningful: to be more specific such information has the form of
80: self-organising patterns which also generate their own `rules of interaction'. In this way we
81: see the correctness of Wheeler's insight of `Law without Law'\cite{Wheeler}. Hence the
82: information is `content addressable', rather than is the case in the usual syntactical
83: information modelling where the information is represented by symbols. This symbolic or
84: syntactical mode is only applicable to higher level phenomenological descriptions, and has
85: served physics well.
86:
87: A pure semantic information system must be formed by
88: a subtle bootstrap process. The mathematical model for this has the form of a stochastic
89: neural network (SNN) for the simple reason that neural networks are well known for their
90: pattern or non-symbolic information processing abilities\cite{neural}. The stochastic
91: behaviour is related to the limitations of syntactical systems discovered by
92: G\"{o}del\cite{G} and more recently extended by
93: Chaitin\cite{Chaitin90,Chaitin99,Chaitin01}, but also results in the neural network being
94: innovative in that it creates its own patterns. The neural network is self-referential,
95: and the stochastic input, known as self-referential noise, acts both to limit the depth
96: of the self-referencing and also to generate potential order.
97:
98: Herein is a status report on the ongoing development of process physics beginning with a
99: discussion of the comparison of syntactical and the new semantic information system
100: and their connections with G\"{o}del's incompleteness theorem. Later sections describe
101: the emergent unification of gravitational and quantum phenomena, amounting to a quantum
102: theory of gravity.
103:
104:
105:
106:
107: \section{Syntactical and Semantic Information Systems}
108:
109: In modelling reality with formal or syntactical information systems physicists assume that
110: the full behaviour of a physical system can be compressed into axioms and rules for the
111: manipulation of symbols. However G\"{o}del discovered that self-referential syntactical
112: systems (and these includes basic mathematics) have fundamental
113: limitations which amount to the realisation that not all truths can be compressed into an
114: axiomatic structure, that formal systems are much weaker than previously supposed. In physics
115: such systems have always been used in conjunction with metarules and metaphysical assertions, all
116: being `outside' the formal system and designed to overcome the limitations of the syntax. Fig.1
117: depicts the current understanding of self-referential syntactical systems. Here the key feature is
118: the G\"{o}del boundary demarcating the provable from the unprovable truths of some system.
119: Chaitin has demonstrated that in mathematics the unprovable truths are essentially random in
120: character. This, however, is a structural randomness in the sense that the individual truths do
121: not have any structure to them which could be exploited to condense them down to or be encoded in
122: axioms. This is unlike random physical events which occur in time. Of course syntactical
123: systems are based on the syntax of symbols and this is essentially non-process or
124: non-timelike.
125:
126: \vspace{-10mm}
127:
128:
129: \begin{figure}[h]
130: \hspace{17mm}
131: \setlength{\unitlength}{1.75mm}
132: \hspace{20mm}\begin{picture}(40,25)
133: \thicklines
134: %\put(3.0,21.0){\bf SELF-REFERENTIAL SYNTACTICAL SYSTEM}
135: \qbezier(0,10)(20,30)(40,10)
136: \qbezier(0,0)(20,-20)(40,0)
137: \qbezier(0,0)(-4,5)(0,10)
138: \qbezier(40,10)(44,5)(40,0)
139:
140: \put(19.7,-10){\line(0,1){30}}%\put(20.2,-10){\line(0,1){30}}
141:
142: \qbezier(0,0)(15,5)(0,10)
143:
144: \put(-1,6){{\bf Formal}}
145: \put(-1,3){{\bf Axioms}}
146:
147: \qbezier[60](5,2)(15,-1)(24,0)
148: \put(24,0){\vector(4,1){1}}
149: \put(16,-1){\line(1,2){1.5}}
150: \put(5.0,-3.0){\bf G\"{o}del's Thm}
151: \put(25.5,0.2){\circle*{0.5}}
152: \put(20,-2.2){\bf unprovable truth}
153:
154:
155: \put(13,14){\circle*{0.5}}
156:
157: \put(7,15){\bf provable truth}
158: \qbezier(5,8)(7.5,12.5)(13,14)
159: \put(12,13.71){\vector(2,1){0.1}}
160:
161: \put(22.6,5){\circle*{0.5}}
162: \put(22.6,8){\circle*{0.5}}
163: \put(27.6,10){\circle*{0.5}}
164: \put(25.6,3.5){\circle*{0.5}}
165: \put(24,6){\bf random truths}
166: \put(26,4.0){\bf (Chaitin)}
167:
168:
169: \put(11,-7){\bf THE BOUNDARY}
170:
171: \end{picture}
172: \vspace{15mm}\caption{\small Graphical depiction of the `logic space' of
173: a self-referential syntactical information system, showing the formal system
174: consisting of symbols and rules, and an example of one theorem
175: (a provable truth). Also shown are unprovable truths which in
176: general are random (or unstructured) in character, following the work of Chaitin.
177: The G\"{o}delian boundary is the demarcation between provable
178: and unprovable truths.
179: \label{section:figure:Godel}}
180: \end{figure}
181:
182: There is an analogy between the structure of self-referential syntactical
183: information systems and the present structure of quantum theory as depicted in Fig.2.
184: There the formal and hence non-process mathematical structure is capable of producing
185: many provable truths, such as the energy levels of the hydrogen atom, and these are also true in
186: the sense that they agree with reality. But from the beginning of quantum theory the Born
187: measurement metarule was introduced to relate this non-process modelling to the actual randomness
188: of quantum measurement events. The individuality of such random events is not a part of the
189: formal structure of quantum theory. Of course it is well known that the non-process or
190: structural aspects of the probability metarule are consistent with the mathematical formalism, in
191: the form of the usual `conservation of probability' equation and the like. Further, the quantum
192: theory has always been subject to various metaphysical interpretations, although these
193: have never played a key role for practitioners of the theory. This all suggests that
194: perhaps the Born metarule is bridging a G\"{o}del-type boundary, that there is a bigger
195: system required to fully model quantum aspects of reality, and that the boundary is evidence of
196: self-referencing in that system.
197:
198:
199:
200: \vspace{-10mm}
201:
202:
203: \begin{figure}[h]
204: \hspace{17mm}
205: \setlength{\unitlength}{1.75mm}
206: \hspace{20mm}\begin{picture}(40,25)
207: \thicklines
208: \qbezier(0,10)(20,30)(40,10)
209: \qbezier(0,0)(20,-20)(40,0)
210: \qbezier(0,0)(-4,5)(0,10)
211: \qbezier(40,10)(44,5)(40,0)
212:
213: \put(19.7,-10){\line(0,1){30}}%\put(20.2,-10){\line(0,1){30}}
214:
215: \qbezier(0,0)(15,5)(0,10)
216:
217: \put(-1,6){{\bf Formal}}
218: \put(-1,3){{\bf Axioms}}
219:
220: \qbezier[60](5,2)(15,-1)(24,0)
221: \put(24,0){\vector(4,1){1}}
222: %\put(16,-1){\line(1,2){1.5}}
223: \put(8.0,0.0){\bf Born}
224: \put(7.0,-1.5){\bf Metarule}
225: \put(25.5,0.2){\circle*{0.5}}
226: \put(21,-2.2){\bf unprovable truth}
227:
228:
229: \put(13,14){\circle*{0.5}}
230:
231: \put(7,15){\bf provable truth}
232: \put(11,12){\bf (eg $E_n$)}
233: \qbezier(5,8)(7.5,12.5)(13,14)
234: \put(12,13.71){\vector(2,1){0.1}}
235:
236: \put(22.6,5){\circle*{0.5}}
237: \put(22.6,8){\circle*{0.5}}
238: \put(27.6,10){\circle*{0.5}}
239: \put(25.6,3.5){\circle*{0.5}}
240: \put(24,6){\bf random truths}
241:
242:
243:
244: \put(13,-6){\bf A G\"{O}DEL }\put(12,-8){\bf BOUNDARY? }
245:
246: \end{picture}
247: \vspace{15mm}\caption{\small Graphical depiction of the syntactical form of
248: conventional quantum theory. The Born measurement metarule appears to bridge a
249: G\"{o}del-like boundary.
250: \label{section:figure:Quantum}}
251:
252: \end{figure}
253:
254:
255: Together the successes and failures of
256: physics suggest that a generalisation of the traditional use of syntactical
257: information theory is required to model reality, and that this has now been identified as a
258: semantic information system which has the form of a stochastic neural network.
259:
260:
261: \begin{figure}[h]
262: \vspace{-10mm}
263: \setlength{\unitlength}{1.75mm}
264:
265:
266: \hspace{40mm}\begin{picture}(40,25)
267: \thicklines
268: %\put(2.0,21.0){\bf BOOTSTRAPPING a PURE SEMANTIC SYSTEM}
269: \qbezier(0,10)(20,30)(40,10)
270: \qbezier(0,0)(20,-20)(40,0)
271: \qbezier(0,0)(-4,5)(0,10)
272: \qbezier(40,10)(44,5)(40,0)
273:
274: % Boundary lines
275: %\put(19.7,-10){\line(0,1){30}}\put(20.2,-10){\line(0,1){30}}
276:
277: \put(20.1,-10){\line(0,1){14}}
278: \put(20.1,4.45){\line(0,1){5.6}}
279: \put(20.1,10.4){\line(0,1){3.6}}
280: \put(20.1,14.4){\line(0,1){5.6}}
281:
282:
283: \qbezier[70](0,0)(15,5)(0,10)
284: \qbezier(40,0)(25,5)(40,10)
285: \qbezier(37,-3)(18,5)(37,13)
286:
287: \put(-1.5,6.0){{\bf Induced}}
288: \put(-2.0,4.35){{\bf Formal}}
289: \put(-1.5,2.8){{\bf Axioms}}
290:
291: \qbezier[60](5,2)(15,-1)(24,0)
292: \put(24,0){\vector(4,1){1}}
293: \put(16,-1.2){\line(1,2){1.2}}
294: \put(5.5,-3.0){\bf G\"{o}del's Thm}
295: \put(25,0.2){\circle*{0.5}}
296:
297: \qbezier[20](7.4,4.3)(9,3)(14,4)
298: \put(13.28,3.8){\vector(4,1){1}}
299: \put(9,2.5){\line(1,2){1.2}}
300:
301:
302: \put(14,14.25){\oval(2,2)[t,l]}
303: \put(14,14.25){\oval(2,2)[b,l]}
304: \put(14,14.5){\oval(2,1.5)[t,r]}
305: \put(14,14.0){\oval(2,1.5)[b,r]}
306: \put(13.6,14.0){{\bf .}}\put(13.1,13.9){{\bf .}}
307: \put(13.7,14.6){{\bf .}}\put(13.4,13.5){{\bf .}}
308: \put(14.9,14.4){\line(1,0){5.25}}
309: \put(15,14){\line(1,0){5.15}}
310:
311: \put(4,15.5){\bf ensemble truth}
312: \qbezier[70](5,8)(7.5,12.5)(12.8,14)
313: \put(12,13.71){\vector(2,1){0.1}}
314:
315: \put(34,7.5){\vector(-3,1){20}}
316: %\put(34,9.5){\vector(-4,1){20}}
317: %\put(36,2.75){\vector(-2,1){22}}
318:
319: \put(34,6.0){{\bf Stochastic}}
320: \put(34,4.25){{\bf Process}}
321: \put(34.5,2.5){{\bf System}}
322: \put(27.0,5.6){{\bf SOC}}
323: \put(26.7,4.02){{\bf syntax}}
324: \put(28,2.2){{\bf filter}}
325: \put(34,2.5){\vector(-1,-1){5}}\put(28.8,-2.7){\circle*{0.5}}
326: \put(21,-4.3){\bf random (contingent) }
327: \put(25,-5.9){\bf truth}
328:
329: \put(13,-7){\bf INDUCED}
330: \put(13,-9){\bf BOUNDARY}
331:
332: % extra ensembles
333:
334: \put(11,10.25){\oval(2,2)[t,l]}
335: \put(11,10.25){\oval(2,2)[b,l]}
336: \put(11,10.5){\oval(2,1.5)[t,r]}
337: \put(11,10.0){\oval(2,1.5)[b,r]}
338: \put(11.92,10.4){\line(1,0){8.25}}
339: \put(11.92,10){\line(1,0){8.25}}
340: \put(10.9,9.9){{\bf .}}\put(10.2,9.9){{\bf .}}
341: \put(10.5,10.5){{\bf .}}\put(10.5,9.5){{\bf .}}
342:
343:
344:
345: \put(14,4.25){\oval(2,2)[t,l]}
346: \put(14,4.25){\oval(2,2)[b,l]}
347: \put(14,4.5){\oval(2,1.5)[t,r]}
348: \put(14,4.0){\oval(2,1.5)[b,r]}
349: \put(14.1,3.5){{\bf .}}\put(13.1,4.49){{\bf .}}
350: \put(13.6,4.5){{\bf .}}\put(13.85,3.9){{\bf .}}
351: \put(14.9,4.5){\line(1,0){5.2}}
352: \put(14.9,4){\line(1,0){5.25}}
353:
354: % extra random truths
355: \put(25,5){\circle*{0.5}}
356: \put(23,10){\circle*{0.5}}
357: \put(27,12){\circle*{0.5}}
358: \put(23,15){\circle*{0.5}}
359: \put(29,16){\circle*{0.5}}
360: \end{picture}
361: \vspace{15mm}
362: \caption{\small Graphical depiction of the bootstrapping of and the emergent structure of a
363: self-organising pure semantic information system. As a high level effect we see the
364: emergence of an induced formal system, corresponding to the current standard syntactical
365: modelling of reality. There is an emergent G\"{o}del-type boundary which represents the
366: inaccessibility of the random or contingent truths from the induced formal or syntactical
367: system.
368: \label{section:figure:Process}}
369: \end{figure}
370:
371:
372:
373:
374:
375: Fig.3 shows a graphical depiction of the bootstrapping of a pure
376: semantic information system, showing the stochastic neural network-like process
377: system from which the semantic system is seeded or bootstrapped. Via a
378: Self-Organised Criticality Filter (SOCF) this seeding system is removed or hidden.
379: From the process system, driven by
380: Self-Referential Noise (SRN), there are emergent
381: truths, some of which are generically true (ensemble truths) while others are
382: purely contingent. The
383: ensemble truths are also reachable from the
384: Induced Formal System as theorems, but from which, because of the non-process nature of the
385: induced formal system, the contingent truths cannot be reached. In this manner there arises a
386: G\"{o}del-type boundary. The existence of the latter leads to induced metarules
387: that enhance the
388: induced formal system, if that is to be used solely in higher order phenomenology.
389:
390:
391:
392:
393: \section{Self-Referentially Limited Neural Networks}
394: \label{sect:NN}
395:
396: \begin{figure}
397: \setlength{\unitlength}{0.5mm}
398: \hspace{20mm}
399: \begin{picture}(150,60)
400: \thicklines
401: \put(10,10){\circle{15}}\put(8,7){1}
402: \put(30,30){\circle{15}}\put(27,27){2}
403: \put(48,0){\circle{15}}\put(45,-3){3}
404: \put(15,15){\vector(1,1){10}}\put(30,12){$B_{23}>0$}
405: \put(5,15){\vector(-1,1){10}}
406: \put(12,3){\vector(1,-4){3}}
407: \put(35,25){\vector(1,-2){9.5}}
408: \put(33,37){\vector(1,2){7}}
409: \put(37,32){\vector(3,1){15}}
410: \put(55,3){\vector(3,1){15}}
411: \put(34,-15){\vector(1,1){9}}
412: \put(95,15){\circle{15}}\put(94,12){i}
413: \put(100,20){\vector(1,1){10}}
414: \put(90,20){\vector(-1,1){10}}
415: \put(25,-22){(a)} \put(115,-22){(b)} \put(210,-22){(c)}
416:
417: \put(145,15){\circle{15}}\put(144,12){i}
418: \put(150,20){\vector(1,1){10}}
419: \put(140,20){\vector(-1,1){10}}
420: \put(95,9){\oval(8,20)[b]}
421: \put(96,-1){\vector(1,0){1.0}}
422: %\put(115,15){\vector(1,0){8}}
423: \put(115,12){$\in$}
424: \end{picture}
425: \hspace{40mm}
426: \setlength{\unitlength}{0.20mm}
427:
428: \hspace{105mm}
429: \begin{picture}(0,50)(40,0) %80
430: \thicklines
431: \put(155,165){\line(3,-5){60}}
432: \put(155,165){\line(-3,-5){60}}
433: \put(115,100){\line(3,-5){42}}
434: \put(195,100){\line(-3,-5){21}}
435: \put(135,160){ \bf $i$}
436: \put(225,160){ \bf $D_0\equiv 1$}
437: \put(225,100){ \bf $D_1=2$}
438: \put(225,60){ \bf $D_2=4$}
439: \put(225,25){ \bf $D_3=1$}
440: \put(155,165){\circle*{5}}
441: \put(115,100){\circle*{5}}
442: \put(195,100){\circle*{5}}
443: \put(95,65){\circle*{5}}
444: \put(135,65){\circle*{5}}
445: \put(175,65){\circle*{5}}
446: \put(215,65){\circle*{5}}
447: \put(155,30){\circle*{5}}
448: \end{picture}
449:
450:
451: \caption{\small (a) Graphical depiction of the neural network with links
452: $B_{ij}\in {\cal R}$ between nodes or pseudo-objects. Arrows indicate sign of
453: $B_{ij}$. (b) Self-links are internal to a node, so $B_{ii}=0$. (c) An $N=8$ spanning
454: tree for a random graph (not shown) with $L=3$. The distance distribution $D_k$ is indicated for
455: node {\it i}.
456: \label{section:figure:neural}}
457: \end{figure}
458:
459:
460: Here we briefly describe a model for a self-referentially limited neural network and in the
461: following section we describe how such a network results in emergent quantum behaviour, and which,
462: increasingly, appears to be a unification of space and quantum phenomena. Process physics is a semantic
463: information system and is devoid of {\it a priori} objects and their laws and so it requires a subtle
464: bootstrap mechanism to set it up. We use a stochastic neural network, Fig.4a, having the structure of
465: real-number valued connections or relational information strengths $B_{ij}$ (considered as forming a
466: square matrix) between pairs of nodes or pseudo-objects
467: $i$ and $j$. In standard
468: neural networks\cite{neural} the network information resides in both link and node
469: variables, with the semantic information residing in attractors of the iterative network.
470: Such systems are also not pure in that there is an assumed underlying and manifest {\it a
471: priori} structure.
472:
473: The nodes and their link variables will be revealed to be themselves sub-networks of informational
474: relations. To avoid explicit self-connections
475: $B_{ii}\neq 0$, which are a part of the sub-network content of
476: $i$, we use antisymmetry $B_{ij}=-B_{ji}$ to conveniently ensure that
477: $B_{ii}=0$, see Fig.4b.
478:
479:
480: At this stage we are using a syntactical system with symbols $B_{ij}$ and, later, rules for the
481: changes in the values of these variables. This system is the syntactical seed for the pure
482: semantic system. Then to ensure that the nodes and links are not remnant {\it a priori}
483: objects the system must generate strongly linked nodes (in the sense that the $B_{ij}$ for
484: these nodes are much larger than the $B_{ij}$ values for non- or weakly-linked nodes) forming
485: a fractal network; then self-consistently the start-up nodes and links may themselves be
486: considered as mere names for sub-networks of relations. For a successful suppression the scheme must
487: display self-organised criticality (SOC)\cite{SOC} which acts as a filter for the start-up syntax. The
488: designation `pure'
489: refers to the notion that all seeding syntax has been removed. SOC is the process
490: where the emergent behaviour displays universal criticality in that the behaviour is
491: independent of the individual start-up syntax; such a start-up syntax then has no
492: ontological significance.
493:
494: To generate a fractal structure we must use a non-linear iterative system for the
495: $B_{ij}$ values. These iterations amount to the necessity to introduce a time-like
496: process. Any system possessing {\it a priori} `objects' can never
497: be fundamental as the explanation of such objects must be outside the system. Hence in
498: process physics the absence of intrinsic undefined objects is linked with the phenomena of
499: time, involving as it does an ordering of `states', the present moment effect, and the
500: distinction between past and present. Conversely in non-process physics the presence of {\it a
501: priori } objects is related to the use of the non-process geometrical model of time, with this modelling
502: and its geometrical-time metarule being an approximate emergent description from process-time. In this
503: way process physics arrives at a new modelling of time, {\it process time}, which is much more complex
504: than that introduced by Galileo, developed by Newton, and reaching its high point with Einstein's
505: spacetime geometrical model.
506:
507: The stochastic neural network so far has been realised with one
508: particular scheme involving a stochastic non-linear matrix iteration, see (1).
509: The matrix inversion $B^{-1}$ then models self-referencing in that it requires all
510: elements of $B$ to compute any one element of $B^{-1}$. As well there is the
511: additive SRN
512: $w_{ij}$ which limits the self-referential information but, significantly, also acts in such
513: a way that the network is innovative in the sense of generating semantic information, that is
514: information which is internally meaningful. The emergent behaviour is believed to be
515: completely generic in that it is not suggested that reality is a computation, rather it
516: appears that reality has the form of an self-referential order-disorder
517: information system. It is important to note that process physics is a non-reductionist
518: modelling of reality; the basic iterator (1) is premised on the general assumption that
519: reality is sufficiently complex that self-referencing occurs, and that this has limitations.
520: Eqn.(1) is then a minimal bootstrapping implementation of these notions. At higher emergent levels
521: this self-referencing manifests itself as {\it interactions} between emergent patterns.
522:
523: To be a successful contender for the Theory of Everything (TOE) process
524: physics must ultimately prove the uniqueness conjecture: that the characteristics (but not
525: the contingent details) of the pure semantic information system are unique. This would
526: involve demonstrating both the effectiveness of the SOC filter and the robustness of the
527: emergent phenomenology, and the complete agreement of the later with observation.
528:
529: The stochastic neural network is modelled by the iterative process
530: \begin{equation}
531: B_{ij} \rightarrow B_{ij} -\alpha (B + B^{-1})_{ij} + w_{ij}, \mbox{\ \ } i,j=1,2,...,2M;
532: M
533: \rightarrow
534: \infty,
535: \label{eq:map}\end{equation}
536: where
537: $w_{ij}=-w_{ji}$ are
538: independent random variables for each $ij$ pair and for each iteration and chosen from some probability
539: distribution. Here $\alpha$ is a parameter the precise value of which should not be critical but which
540: influences the self-organisational process.
541: We start the iterator at
542: $B\approx 0$, representing the absence of information.
543: With the noise absent the iterator
544: behaves in a deterministic and reversible manner given by
545: the matrix
546: \begin{equation}
547: B = MDM^{-1}; \mbox{\ \ \ \ }D=\left(\begin{array}{rrrrrr}
548: 0 & +b_1 & 0 & 0\\
549: -b_1 & 0 & 0 & 0 \\
550: 0 & 0 & 0 & +b_2 \\
551: 0 & 0 & -b_2 & 0 & \\
552: &&&&. \\ &&&&&. \\
553: \end{array}\right), \mbox{\ \ \ \ \ } b_1,b_2,... \geq 0,
554: \end{equation}
555: where $M$ is a real orthogonal matrix determined uniquely by the start-up $B$, and each $b_i$
556: evolves according to the iterator
557: $
558: b_i \rightarrow b_i-\alpha(b_i-b_i^{-1}),
559: $
560: which converges to $b_i=1$. This $B$ exhibits no interesting structure. In the presence of the
561: noise the iterator process is non-reversible and non-deterministic. It is also manifestly
562: non-geometric and non-quantum, and so does not assume any of the standard features of syntax based
563: physics models. The dominant mode is the formation of an apparently randomised
564: background (in
565: $B_{ij}$) but, however, it also manifests a self-organising process which results
566: in a growing three-dimensional fractal process-space that competes with this random background -
567: the formation of a `bootstrapped universe'. Here we report on the current status of ongoing
568: work to extract the nature of this `universe'.
569:
570: The emergence of order in this system might
571: appear to violate expectations regarding the 2nd Law of Thermodynamics; however because of the SRN
572: the system behaves as an open system and the growth of order arises from the
573: self-referencing term, $B^{-1}$ in (1), selecting certain implicit order in the SRN. Hence the SRN
574: acts as a source of negentropy\footnote{The term {\it negentropy} was introduced by E.
575: Schr\"{o}dinger\cite{negentropy} in 1945, and since then there has been ongoing discussion of its
576: meaning. In process physics it manifests as the SRN.}
577:
578:
579: This growing three-dimensional fractal process-space is an example of a Prigogine far-from-
580: equilibrium dissipative structure\cite{Prigogine} driven by the SRN.
581: From each iteration the noise term
582: will additively introduce rare large value
583: $w_{ij}$. These $w_{ij}$, which define sets of strongly linked nodes, will persist through more
584: iterations than smaller valued $w_{ij}$ and, as well, they become further linked by the iterator
585: to form a three-dimensional process-space with embedded topological defects. In this way the
586: stochastic neural-network creates stable strange attractors and as well determines their interaction
587: properties. This information is all internal to the system; it is the semantic information within
588: the network.
589:
590: To see the nature of this internally generated information consider a node $i$ involved in one
591: such large
592: $w_{ij}$;
593: it will be connected via other large $w_{ik}$ to a
594: number of other nodes and so on, and this whole set of connected nodes forms a connected random graph unit
595: which we call a gebit as it acts as a small piece or bit of geometry formed from random information links and
596: from which the process-space is self-assembled. The gebits compete for new links and undergo
597: mutations. Indeed, as will become clear, process physics is remarkably analogous in its operation to
598: biological systems. The reason for this is becoming clear: both reality and subsystems of reality must use
599: semantic information processing to maintain existence, and symbol manipulating systems are totally unsuited
600: to this need, and in fact totally contrived.
601:
602: To analyse the connectivity of such
603: gebits assume for simplicity that the large $w_{ij}$ arise with fixed but very small probability $p$,
604: then the geometry of the gebits is revealed by studying the probability distribution for the structure
605: of the random graph units or gebits minimal spanning trees with $D_k$ nodes at $k$ links from node $i$
606: ($D_0
607: \equiv 1$), see Fig.4c, this is given by\cite{Nagels}
608:
609: \begin{equation}{\cal P}[D,L,N] \propto \frac{p^{D_1}}{D_1!D_2!....D_L!}\prod_{i=1}^{L-1}
610: (q^{\sum_{j=0}^{i-1}{D_j}})^{D_{i+1}}(1-q^{D_i})^{D_{i+1}},\end{equation}
611: where $q=1-p$, $N$ is the total number of nodes in the gebit and $L$ is the maximum depth from node $i$.
612: To find the most likely connection pattern we numerically maximise ${\cal P}[D,L,N]$ for fixed $N$ with respect
613: to
614: $L$ and the $D_k$. The resulting $L$ and $\{D_1,D_2,...,D_L\}$ fit very closely to the form $D_k\propto
615: \sin^{d-1}(\pi k/L)$; see Fig.5a for $N=5000$ and $\mbox{Log}_{10}p=-6$. The resultant $d$
616: values for a range of $\mbox{Log}_{10}p$ and $N=5000$ are shown in Fig.5b.
617:
618: \vspace{-35mm}
619: \hspace{-20mm}\includegraphics{figure.eps}
620:
621: \vspace{70mm}
622: \begin{figure}[h]
623: \caption{\small
624: (a) Points show the $D_k$ set and $L=40$ value found by numerically maximising ${\cal P}[D,L,N]$
625: for $\mbox{Log}_{10}p=-6$ for fixed $N=5000$. Curve shows
626: $D_k\propto \sin^{d-1}(\frac{\pi k}{L})$ with best fit $d=3.16$ and $L=40$, showing excellent
627: agreement, and indicating embeddability in an $S^3$ with some topological defects. (b) Dimensionality $d$ of
628: the gebits as a function of the probability $p$. (c) Graphical depiction of the `process space' at one stage of the iterative
629: process-time showing a quantum-foam structure formed from embeddings and linkings of gebits. The
630: linkage connections have the distribution of a 3D space, but the individual gebit components are
631: closed compact spaces and cannot be embedded in a 3D background space. So the drawing is only
632: suggestive. Nevertheless this figure indicates that process physics generates a cellular
633: information system, where the behaviour is determined at all levels by internal information. }
634: \end{figure}
635:
636:
637: This shows, for $p$ below a critical value, that
638: $d=3$, indicating that the connected nodes have a natural embedding in a 3D hypersphere $S^3$;
639: call this a base gebit. Above that value of $p$, the increasing value of $d$
640: indicates the presence of extra links that, while some conform with the embeddability, are in the main defects
641: with respect to the geometry of the $S^3$. These extra links act as topological defects. By themselves these
642: extra links will have the connectivity and embedding geometry of numbers of gebits, but these gebits have a
643: `fuzzy' embedding in the base gebit. This is an indication of fuzzy homotopies (a homotopy is,
644: put simply, an embedding of one space into another).
645:
646: The base gebits $g_1, g_2, ...$ arising from the SRN together with their embedded topological defects have
647: another remarkable property: they are `sticky' with respect to the iterator. Consider the larger valued
648: $B_{ij}$ within a given gebit $g$, they form tree graphs and most tree-graph adjacency matrices are singular
649: (det($g_{tree})=0$). However the presence of other smaller valued $B_{ij}$ and the
650: general
651: background noise ensures that det$(g)$ is small but not exactly zero.
652: Then the
653: $B$ matrix has an inverse with large components that act to cross-link the new and
654: existing gebits. This cross-linking is itself random, due to the presence of background noise, and the above
655: analysis may again be used and we would conclude that the base gebits themselves are formed into a 3D hypersphere
656: with embedded topological defects. The nature of the resulting 3D process-space is suggestively indicated in Fig.5c, and
657: behaves essentially as a quantum foam\cite{foam1}.
658:
659: Over ongoing
660: iterations the existing gebits become cross-linked and eventually lose their ability to undergo further
661: linking; they lose their `stickiness' and decay. The value of the parameter $\alpha$ in (1) must
662: be small enough that the `stickiness' persists over many iterations, that is, it is not quenched
663: too quickly, otherwise the emergent network will not grow. Hence the emergent
664: space is 3D but is continually undergoing replacement of its component gebits; it is an
665: informational process-space, in sharp distinction to the non-process continuum geometrical spaces
666: that have played a dominant role in modelling physical space. If the noise is `turned off' then
667: this emergent dissipative space will decay and cease to exist. We thus see that the nature of
668: space is deeply related to the logic of the limitations of logic, as implemented here as a
669: self-referentially limited neural network.
670:
671:
672: \section{Modelling Gebits and their Topological Defects}
673:
674: We need to extract convenient but approximate syntactical descriptions of the semantic information in the
675: network, and these will have the form of a sequence of mathematical constructions, the first being the Quantum
676: Homotopic Field Theory. Importantly they must all retain explicit manifestations of the SRN. To this end first
677: consider the special case of the iterator when the SRN is frozen at a particular
678: $w$, that is we consider iterations with an artificially fixed SRN term. Then the iterator is equivalent to the
679: minimisation of an `energy' expression (remember that
680: $B$ and $w$ are antisymmetric)
681: \begin{equation}
682: E[B;w]= -\frac{\alpha}{2}\mbox{Tr}[B^2]-\alpha \mbox{TrLn}[B]+\mbox{Tr}[wB].
683: \end{equation}
684: Note that for disconnected gebits $g_1$ and $g_2$ this energy is additive, $E[g_1\oplus g_2]=E[g_1]+E[g_2]$.
685: Now suppose the fixed $w$ has the form of a gebit approximating an $S^3$ network with one embedded topological
686: defect which is itself an $S^3$ network, for simplicity. So we are dissecting the gebit into base gebit, defect gebit
687: and linkings or embeddings between the two. We also ignore the rest of the network, which is permissible if our gebit
688: is disconnected from it. Now if det$(w)$ is not small, then this gebit is non-sticky, and for small $\alpha$, the
689: iterator converges to $B\approx\frac{1}{\alpha}w$, namely an enhancement only of the gebit. However because the
690: gebits are rare constructs they tend to be composed of larger $w_{ij}$ forming tree structures, linked by smaller
691: valued
692: $w_{ij}$. The tree components make det$(w)$ smaller, and then the inverse $B^{-1}$ is activated and generates new
693: links. Hence, in particular, the topological defect relaxes, according to
694: the `energy' measure, with respect to the base gebit. This relaxation is an example of a `non-linear elastic'
695: process\cite{Ogden}. The above gebit has the form of a mapping $\pi: S \rightarrow \Sigma$ from a base space to a
696: target space. Manton\cite{Manton1, Manton2, GP} has constructed the continuum form for the `elastic energy' of such
697: an embedding and for $\pi: S^3 \rightarrow S^3$ it is the Skyrme energy
698: \begin{equation}
699: E[U]=\int \left[ -\frac{1}{2}\mbox{Tr}(\partial_i UU^{-1}\partial_i UU^{-1}) -\frac{1}{16} \mbox{Tr}[\partial_i
700: UU^{-1},\partial_i UU^{-1}]^2\right],
701: \end{equation}
702: where $U(x)$ is an element of $SU(2)$. Via the parametrisation $U(x)=\sigma(x)
703: +i\vec{\pi}(x).\vec{\tau}$, where the
704: $\tau_i$ are Pauli matrices, we have $\sigma(x)^2+\vec{\pi}(x)^2$=1, which parametrises an $S^3$ as
705: a unit hypersphere embedded in $E^4$. Non-trivial minima of
706: $E[U]$ are known as Skyrmions (a form of topological soliton), and have
707: $Z=\pm1,\pm2,...$, where $Z$ is the winding number of the map,
708: \begin{equation}
709: Z=\frac{1}{24\pi^2}\int\sum\epsilon_{ijk}\mbox{Tr}(\partial_i UU^{-1}\partial_j UU^{-1}\partial_k UU^{-1}).
710: \end{equation}
711:
712: The first key to extracting emergent phenomena from the stochastic neural network is the validity of this continuum
713: analogue, namely that $E[B;w]$ and $E[U]$ are describing essentially the same `energy'
714: reduction process. This should be amenable to detailed analysis.
715:
716: This `frozen' SRN analysis of course does not match the time-evolution of the
717: full iterator, (1), for this displays a much richer collection of processes. With ongoing new noise in each
718: iteration and the saturation of the linkage possibilities of the gebits emerging from this noise, there arises a
719: process of ongoing birth, linkaging and then decay of most patterns. The task is then to
720: identify those particular patterns that survive this flux, even though all components of
721: these patterns eventually disappear, and to attempt a description of their modes of
722: behaviour. This brings out the very biological nature of the information processing in
723: the SNN, and which appears to be characteristic of a `pure' semantic information system.
724:
725: In general each gebit, as it emerges from the SRN, has active nodes and embedded topological defects, again with
726: active nodes. Further there will be defects embedded in the defects and so on, and so gebits begin to have the
727: appearance of a fractal defect structure, and all the defects having various classifications and associated winding
728: numbers. The energy analogy above suggests that defects with opposite winding numbers at the same fractal depth
729: may annihilate by drifting together and merging. Furthermore the embedding of the defects is unlikely to be
730: `classical', in the sense of being described by a mapping
731: $\pi(x)$, but rather would be fuzzy, i.e described by some functional, $F[\pi]$, which would correspond to a
732: classical embedding only if $F$ has a supremum at one particular $\pi=\pi_{cl}$. As well these gebits are undergoing
733: linking because their active nodes (see \cite{CK98} for more discussion) activate the $B^{-1}$ new-links
734: process between them, and so by analogy the gebits themselves form larger structures with embedded fuzzy
735: topological defects. This emergent behaviour is suggestive of a quantum space foam, but one containing
736: topological defects which will be preserved by the system, unless annihilation events occur. If these
737: topological defects are sufficiently rich in fractal structure as to be preserved, then their initial
738: formation would have occurred as the process-space relaxed out of its initial, essentially random form.
739: This phase would correspond to the early stages of the Big-Bang. Once the topological defects are trapped
740: in the process-space they are doomed to meander through that space by essentially self-replicating, i.e.
741: continually having their components die away and be replaced by similar components. These residual
742: topological defects are what we call matter. The behaviour of both the process-space and its defects is
743: clearly determined by the same network processes; we have an essential unification of space and matter
744: phenomena. This emergent quantum foam-like behaviour suggests that the full generic description of the
745: network behaviour is via the Quantum Homotopic Field Theory (QHFT) of the next section.
746:
747:
748:
749: \section{Modelling the Emergent Quantum Foam}
750:
751: To construct this QHFT we introduce an appropriate configuration space, namely all the possible homotopic
752: mappings $\pi_{\alpha\beta}: S_\beta \rightarrow S_\alpha$, where the
753: $S_1,S_2,..$, describing `clean' or topological-defect free gebits, are compact spaces of various types. Then QHFT
754: has the form of an iterative functional Schr\"{o}dinger equation for the discrete time-evolution of a wave-functional
755: $\Psi[....,\pi_{\alpha\beta},....;t]$
756: \begin{equation}\Psi[....,\pi_{\alpha\beta},....;t+\Delta
757: t]= \Psi[....,\pi_{\alpha\beta},....;t]
758: -iH\Psi[....,\pi_{\alpha\beta},....;t]\Delta t + \mbox{QSD terms},
759: \end{equation} The time step $\Delta t$ in (7) is relative to the scale of the fractal
760: processes being explicitly described, as we are using a configuration space of mappings between prescribed gebits. At
761: smaller scales we would need a smaller value of $\Delta t$. Clearly this invokes a (finite)
762: renormalisation scheme. We now discuss the form of the hamiltonian and the Quantum State Diffusion (QSD) terms.
763:
764: First (7), without the QSD term, has a form analogous to a `third quantised' system, in
765: conventional terminology\cite{Baby}. These
766: systems were considered as perhaps capable of generating a quantum theory of
767: gravity. The argument here is that this is the emergent behaviour of the SNN, and it does indeed lead to quantum
768: gravity, but with quantum matter as well. More importantly we understand the origin of (7), and it will lead to
769: quantum and then classical gravity, rather than arise from classical gravity via some ad hoc or heuristic
770: quantisation procedure.
771:
772: Depending on the `peaks' of
773: $\Psi$ and the connectivity of the resultant dominant mappings such mappings are to be interpreted as
774: either embeddings or links; Fig.5c then suggests the dominant process-space form within
775: $\Psi$ showing both links and embeddings. The emergent process-space then has the characteristics of a
776: quantum foam. Note that, as indicated in Fig.5c, the original start-up links and nodes are now absent.
777: Contrary to the suggestion in Fig.5c, this process space cannot be embedded in a {\it finite}
778: dimensional geometric space with the emergent metric preserved, as it is composed of nested
779: or fractal finite-dimensional closed spaces.
780:
781: We now consider the form of the hamiltonian H. The previous section suggested that Manton's non-linear
782: elasticity interpretation of the Skyrme energy is appropriate to the SNN. This then suggests that H is
783: the functional operator
784: \begin{equation}
785: H=\sum_{\alpha\neq\beta}h[\frac{\delta}{\delta
786: \pi_{\alpha\beta}},\pi_{\alpha\beta}]
787: \end{equation}
788: where $h[\frac{\delta}{\delta \pi},\pi]$ is the (quantum) Skyrme Hamiltonian functional operator for the system based
789: on making fuzzy the mappings
790: $\pi: S \rightarrow \Sigma$, by having $h$ act on wave-functionals of the form $\Psi[\pi(x);t]$. Then $H$
791: is the sum of pairwise embedding or homotopy hamiltonians. The corresponding functional Schr\"{o}dinger
792: equation would simply describe the time evolution of quantised Skyrmions with the base space fixed, and
793: $\Sigma \in SU(2)$. There have been very few analyses of even this class of problem, and then the base space
794: is usually taken to be $E^3$. We shall not give the explicit form of $h$ as it is complicated, but wait to present
795: the associated action.
796:
797: In the absence of the QSD terms the time evolution in (7) can be formally written as a functional integral
798: \begin{equation}
799: \Psi[\{\pi_{\alpha\beta}\};t']=\int\prod_{\alpha\neq\beta}{\cal
800: D}\tilde{\pi}_{\alpha\beta}e^{iS[\{\tilde{\pi}_{\alpha\beta}\}]}
801: \Psi[\{\pi_{\alpha\beta}\};t],
802: \end{equation}
803: where, using the continuum $t$ limit notation, the action is a sum of pairwise actions,
804: \begin{equation}
805: S[\{\tilde{\pi}_{\alpha\beta}\}]=\sum_{\alpha\neq\beta}S_{\alpha\beta}[\tilde{\pi}_{\alpha\beta}],
806: \end{equation}
807: \begin{equation}
808: S_{\alpha\beta}[\tilde{\pi}]=\int_t^{t'}dt''\int d^nx\sqrt{ -g} \left[ \frac{1}{2}\mbox{Tr}(\partial_\mu
809: \tilde{U}\tilde{U}^{-1}\partial^\mu
810: \tilde{U}\tilde{U}^{-1}) +\frac{1}{16} \mbox{Tr}[\partial_\mu \tilde{U}\tilde{U}^{-1},\partial^\nu
811: \tilde{U}\tilde{U}^{-1}]^2\right],
812: \end{equation}
813: and the now time-dependent (indicated by the tilde symbol) mappings $\tilde{\pi}$ are parametrised by
814: $\tilde{U}(x,t)$, $\tilde{U}\in S_\alpha$. The metric $g_{\mu\nu}$ is that of the $n$-dimensional base space,
815: $S_\beta$, in
816: $\pi_{\alpha,\beta}: S_\beta
817: \rightarrow S_\alpha$. As usual in the functional integral formalism the functional derivatives in the quantum
818: hamiltonian, in (8), now manifest as the time components $\partial_0$ in (11), so now (11) has the form
819: of a `classical' action, and we see the emergence of `classical' fields, though the emergence of `classical' behaviour
820: is a more complex process. Eqns.(7) or (9) describe an infinite set of quantum skyrme systems, coupled in a
821: pairwise manner. Note that each homotopic mapping appears in both orders; namely $\pi_{\alpha\beta}$ and
822: $\pi_{\beta\alpha}$.
823:
824: The Quantum State Diffusion (QSD)\cite{QSD} terms are non-linear and
825: stochastic,
826: \begin{equation}
827: \mbox{QSD terms} =\sum_\gamma\left(
828: <\!\!L^\dagger_\gamma\!\!>L_\gamma-\frac{1}{2}L^\dagger_\gamma L_\gamma-
829: <\!\!L^\dagger_\gamma\!\!><\!\!L_\gamma\!\!>\right)
830: \Psi\Delta t+\sum_\gamma\left(L_\gamma-<\!\!L_\gamma\!\!> \right)\Psi\Delta \xi_\gamma,
831: \end{equation}
832: which involves summation over the class of Linblad functional operators $L_\gamma$.
833: The QSD terms are up to 5th order in $\Psi$, as in general,
834: \begin{equation}
835: <\!\! A \!\!>_t=\int \prod_{\alpha\neq\beta}{\cal D}\pi_{\alpha\beta}\Psi[\{
836: \pi_{\alpha\beta} \};t]^* A
837: \Psi[\{ \pi_{\alpha\beta}\};t]
838: \end{equation}
839: and where $\Delta \xi_\gamma$
840: are complex statistical variables with means $M (\Delta \xi_\gamma) = 0$, $M( \Delta
841: \xi_\gamma
842: \Delta\xi_{\gamma'})= 0$ and $M(\Delta \xi^*_\gamma\Delta\xi_{\gamma'}) =
843: \delta(\gamma-\gamma')\Delta t$
844:
845:
846: These QSD terms are ultimately responsible for the emergence of
847: classicality via an objectification process, but in particular they produce
848: wave-function(al) collapses during quantum measurements, as the QSD terms tend to `sharpen' the fuzzy homotopies
849: towards classical or sharp homotopies (the forms of the Linblads will be discussed in detail elsewhere). So the
850: QSD terms, as residual SRN effects, lead to
851: the Born quantum measurement random behaviour, but here arising from the process physics, and not being
852: invoked as a metarule.
853: Keeping the QSD terms leads to a functional integral representation for a
854: density matrix formalism in place of (9), and this amounts to a derivation of the decoherence
855: formalism which is usually arrived at by invoking the Born measurement metarule. Here we see that
856: `decoherence' arises from the limitations on self-referencing.
857:
858: In the above we have a deterministic and unitary
859: evolution, tracking and preserving topologically encoded information, together with the stochastic
860: QSD terms, whose form protects that information during localisation events, and which
861: also ensures the full matching in QHFT of process-time to real time: an ordering of
862: events, an intrinsic direction or `arrow' of time and a modelling of the contingent
863: present moment effect. So we see that process physics generates a complete theory of quantum
864: measurements involving the non-local, non-linear and stochastic QSD terms. It does this because
865: it generates both the `objectification' process associated with the classical apparatus and the actual
866: process of (partial) wavefunctional collapse as the quantum modes interact with the measuring
867: apparatus. Indeed many of the mysteries of quantum measurement are resolved when it is realised that
868: it is the measuring apparatus itself that actively provokes the collapse, and it does so because the
869: QSD process is most active when the system deviates strongly from its dominant mode, namely the
870: ongoing relaxation of the system to a 3D process-space. This is essentially the process that
871: Penrose\cite{Penrose} suggested, namely that the quantum measurement process is essentially a
872: manifestation of quantum gravity. The demonstration of the validity of the Penrose argument of course
873: could only come about when quantum gravity was {\it derived} from deeper considerations, and not by
874: some {\it ad hoc} argument such as the {\it quantisation} of Einstein's classical spacetime model.
875:
876:
877: The mappings $\pi_{\alpha\beta}$ are related to group manifold parameter spaces with the group determined by
878: the dynamical stability of the mappings. This symmetry leads to the flavour symmetry of the
879: standard model. Quantum homotopic mappings or skyrmions behave as fermionic or bosonic modes for
880: appropriate winding numbers; so process physics predicts both fermionic and bosonic quantum modes, but
881: with these associated with topologically encoded information and not with objects or `particles'.
882:
883: \section{Quantum Field Theory}
884: The QHFT is a very complex `book-keeping' system for the emergent properties of the neural network,
885: and we now sketch how we may extract a more familiar quantum field theory (QFT) that relates to the
886: standard model of `particle' physics. An effective QHFT should reproduce the emergence
887: of the process-space part of the quantum foam, particularly its 3D aspects. The QSD processes play
888: a key role in this as they tend to enhance classicality. Hence at an appropriate scale QHFT
889: should approximate to a more conventional QFT, namely the emergence of a wave-functional system
890: $\Psi[U(x);t]$ where the configuration space is that of homotopies from a 3-space to $U(x) \in
891: G$, where $G$ is some group manifold space. This $G$ describes `flavour' degrees of freedom.
892: Hence the Schr\"{o}dinger
893: wavefunctional equation for this QFT will have the form
894: \begin{equation}\Psi[U;t+\Delta
895: t]= \Psi[U;t]
896: -iH\Psi[U;t]\Delta t + \mbox{QSD terms},
897: \end{equation}
898: where the general form of $H$ is known, and where a new residual manifestation of the SRN appears
899: as the QSD terms. This system describes skyrmions embedded in a continuum spacetime. It is
900: significant that such Skyrmions are only stable, at least in flat space and for static skyrmions, if
901: that space is 3D. This tends to confirm the observation that 3D space is special for the neural
902: network process system.
903: Again, in the absence of the QSD terms, we may express (15) in terms of the functional integral
904: \begin{equation}
905: \Psi[U;t']=\int{\cal
906: D}\tilde{U}e^{iS[\tilde{U}]}
907: \Psi[U;t].
908: \end{equation}
909: To gain some insight into the phenomena present in (14) or (15), it is convenient to use the
910: fact that functional integrals of this Skyrmionic form my be written in terms of Grassmann-variable
911: functional integrals, but only by introducing a fictitious `metacolour' degree of freedom and
912: associated coloured fictitious vector bosons. This is essentially the reverse of the Functional Integral Calculus
913: (FIC) hadronisation technique in the Global Colour Model (GCM) of QCD\cite{GCM}. The action for the Grassmann and
914: vector boson part of the system is of the form (written for flat spacetime)
915: \begin{equation}
916: S[\overline{p},p,A^a_\mu]=\int d^4x\left( \overline{p}\gamma^\mu(i\partial_\mu
917: +g\frac{\lambda^a}{2}A^a_\mu)p-\frac{1}{4}F^a_{\mu\nu}(A)F^{a\mu\nu}(A)
918: \right),
919: \end{equation}
920: where the Grassmann variables $p_{f c}(x)$ and $\overline{p}_{f c}(x)$ have
921: flavour and metacolour labels. The Skyrmions are then the low energy Nambu-Goldstone
922: modes of this Grassmann system; other emergent modes are of higher energy and can be
923: ignored. These coloured and flavoured but fictitious fermionic fields
924: $\overline{p}$ and $p$ correspond to the proposed preon system\cite{Preons1,Preons2}. As
925: they are purely fictitious, in the sense that there are no excitations in the system corresponding
926: to them, the metacolour degree of freedom must be hidden or confined. Then while the QHFT and the QFT
927: represent an induced syntax for the semantic information, the preons may be considered as an
928: induced `alphabet' for that syntax. The advantage of introducing this preon alphabet is that we can
929: more easily determine the states of the system by using the more familiar language of fermions and
930: bosons, rather than working with the skyrmionic system, so long as only colour singlet states
931: are finally permitted. However it is important to note that (16) and the action in (15) are certainly not
932: the final forms. Further analysis will be required to fully extract the induced
933: actions for the emergent QFT.
934:
935: \section{Inertia and Gravity}
936: Process physics predicts that the neural network behaviour will be charactersied by a growing
937: 3-dimensional process-space having, at a large scale, the form of an $S^3$ hypersphere, which is
938: one of the forms allowed by Einstein's syntactical modelling. It is possible to give the dominant
939: rate of growth of this hypersphere. However first, from random graph theory\cite{randomg}, we
940: expect more than one such spatial system, with
941: each having the character of a growing hypersphere, and all embedded in the random background
942: discussed previously. This background has no metric structure, and so these various hyperspheres
943: have no distance measure over them. We have then a multi-world universe (our `universe' being
944: merely one of these `worlds'). Being process spaces they compete for new gebits, and so long as we
945: avoid a saturation case, each will grow according to
946: \begin{equation}
947: \frac{dN_i}{dt}=aN_i-bN_i \mbox{\ \ \ \ \ } a>0, b>0,
948: \end{equation}
949: where the
950: last term describes the decay of gebits at a rate $b$, while the first describes growth of the
951: $i$-th `world', this being proportional to the size (as measured by its gebit content number)
952: $N_i(t)$ as success in randomly attaching new gebits is proportional to the number of gebits
953: present (the `stickiness' effect), so long as we ignore the topological defects (quantum `matter')
954: as these have a different stickiness, and also affect the decay rate, and so slow down the
955: expansion. Thus
956: $N_i(t)$ will show exponential growth, as appears to be the case as indicated by recent observations
957: of very distant supernovae counts\cite{cosmo}. Hence process physics predicts a
958: positive cosmological constant.
959:
960: One striking outcome of process physics is a possible explanation for the phenomenon of gravity.
961: First note that matter is merely topological defects embedded in the process space, and we expect
962: such defects to have a larger than usual gebit connectivity; indeed matter is a violation of the
963: 3-D connectivity of this space, and it is for this reason we needed to introduce fields to emulate
964: this extra non-spatial connectivity. One consequence of this is that in the region of these
965: matter fields the gebits decay faster, they are less sticky because of the extra connectivity.
966: Hence in this region, compared to other nearby matter-free regions the gebits are being `turned
967: over' more frequently but at the same time are less effective in attracting new gebits. Overall
968: this suggests that matter-occupying-regions act as net sinks for gebits, and there will be a
969: trend for the neighbouring process-space to undergo a diffusion/relaxation process in which
970: the space effectively moves towards the matter: matter acts as a sink for space, and never as a
971: source. Such a process would clearly correspond to gravity. As the effect is described by
972: a net diffusion/relaxation of space which acts equally on all surrounding matter, the
973: in-fall mechanism is independent of the nature of the surrounding matter. This is nothing
974: more than Einstein's Equivalence Principle. As well if the in-fall rate exceeds the rate at which
975: `motion' through the process-space is possible then an event horizon appears, and this is clearly the
976: black hole scenario. Such an event horizon is sufficient condition for the occurrence of Hawking
977: radiation.
978:
979: Finally we mention one long standing unsolved problem in physics, namely an understanding of
980: inertia. This is the effect where objects continue in uniform motion unless acted upon by `forces',
981: and was first analysed by Galileo. However there has never been an explanation for this effect;
982: in Newtonian physics it was built into the syntactical description rather than being a prediction
983: of that modelling. Of course current physics is essentially a static modelling of reality, with
984: motion indirectly accessed via the geometrical-time metarule, and so the failure to explain motion
985: is not unexpected. However process physics offers a simple explanation.
986:
987: The argument for inertia follows from a simple self-consistency argument. Suppose a
988: topological defect, or indeed a vast bound collection of such defects, is indeed `in motion'. This
989: implies that the gebits are being preferentially replaced in the direction of that `motion', for
990: motion as such is a self-replication process; there is no mechanism in process physics for a fixed
991: pattern to say `slide' {\it through} the process-space. Rather motion is self-replication of the
992: gebit connectivity patterns in a set direction. Since the newest gebits, and hence the stickiest
993: gebits, in each topological defect, are on the side corresponding to the direction of motion, the
994: gebits on that side are preferentially replaced. Hence the motion is self-consistent and
995: self-perpetuating.
996:
997: An additional effect expected in process physics is that such motion results in a time dilation
998: effect; the self-replication effect is to be considered as partly the self-replication associated
999: with any internal oscillations and partly self-replication associated with `motion'. This `competition for
1000: resources' results in the slowing down of internal oscillations, an idea discussed by
1001: Toffoli\cite{Toffoli}. We then see that many of the effects essentially assumed by the formalism of general
1002: relativity appear to be emergent phenomena within process physics.
1003:
1004: However there is one novel effect that is of some significance. The process-space appears to
1005: represent a preferred frame of reference, but one that is well hidden by the time-dilation effects.
1006: Hence we would not expect a Michelson-Morley type experiment to reveal this frame. However using
1007: the arguments of Hardy\cite{Hardy} we expect that
1008: the action of the QSD wavefunctional `collapse' processes would reveal the proper frame
1009: through a multi-component EPR experiment, as the non-local QSD collapse occurs in a truly
1010: simultaneous manner; essentially it exposes the underlying global iterations.
1011:
1012:
1013:
1014:
1015:
1016: \section{Conclusions}
1017: Sketched out here is the radical proposal that to comprehend reality we need a system richer than
1018: mere syntax to capture the notion that reality is at all levels about, what may be
1019: called, internal, relational or semantic information, and not, as in the case of syntax,
1020: information that is essentially accessible to or characterisable by observers. This
1021: necessitates an evolution in modelling reality from a non-process physics to a process
1022: physics. Such a development has been long anticipated outside of physics where it is
1023: known as {\it Process Philosophy}\cite{ProcessPhilosophy}.
1024:
1025: To
1026: realise such a system one representation has been proposed and studied, namely that of a
1027: self-referentially limited neural-network model, for neural networks are powerful examples of
1028: non-symbolic information processing. This neural-network model is manifestly free of any notions
1029: of geometry, quantum phenomena or even `laws of physics'. Nevertheless the arguments presented here
1030: strongly suggest that such phenomena are emergent in the neural network but only because the
1031: self-referential noise acts as a source of negentropy or order. The model has been developed to the
1032: stage where various phenomena have been identified and appropriate induced syntactical descriptions
1033: have been suggested. These correspond essentially to the concepts of current physics which, over
1034: the years, have been arrived at via increasingly more abstract non-process syntactical modelling. One
1035: important addition being the ever-present QSD terms which, as it happens, ensure that the phenomena of
1036: time fully matches our experiences of time, and which also plays various other key roles. This new
1037: process-physics is inherently non-reductionist as it explicitly assumes that reality is sufficiently
1038: complex that it is self-referential, and which may be accessed by using a subtle bootstrap approach.
1039: Clearly we see the beginnings of a unification of physics that leads to quantum gravity and classicality
1040: and the emergence of syntax and its associated logic of named objects. The confirmation of process
1041: physics will involve developing more powerful techniques to facilitate the extraction of the induced
1042: syntax for the emergent phenomena. This is a novel and subtle problem.
1043:
1044: Process physics also implies that the geometrical construct of spacetime is merely an induced formal system or syntax
1045: and does not itself have any ontological status. Nevertheless this syntax is enormously useful for
1046: analysing certain technical aspects of reality, so long as we can be sure that the syntax alone does not
1047: introduce spurious problems. One well known spurious problem are the divergences in quantum field theory
1048: caused by the continuum nature of spacetime. As well it is obvious that we can never discover deeper
1049: physics by {\it quantising} classical systems.
1050:
1051: The author thanks the organisers of this conference for assembling an interesting collection of
1052: contributions, and in particular expresses his regards to Dr Peter Szekeres on his retirement
1053: after a distinguished career in general relativity studies.
1054:
1055:
1056: \newpage
1057: \begin{thebibliography}{99}
1058:
1059:
1060: \bibitem{CK97} Cahill, R.T. and Klinger, C.M. (1997). {\it Bootstrap Universe from Self-Referential
1061: Noise}, preprint gr-qc/9708013.
1062:
1063: \bibitem{CK98} Cahill, R.T. and Klinger, C.M. (2000). {\it Self-Referential Noise
1064: and the Synthesis of Three-Dimensional Space}, {\it Gen. Rel. and Grav.} {\bf 32}(3),
1065: 529, gr-qc/9812083.
1066:
1067: \bibitem{CK99} Cahill, R.T. and Klinger, C.M. (2000). {\it Self-Referential Noise
1068: as a Fundamental Aspect of Reality},
1069: Proc.~2nd Int.~Conf.~on Unsolved Problems of Noise and Fluctuations
1070: (UPoN'99), eds.~D.~Abbott and L.~Kish, Adelaide, Australia, 11-15th July
1071: 1999, {\bf Vol.~511,} p.~43 (American Institute of Physics, New York), gr-qc/9905082.
1072:
1073: \bibitem{CKK00} Cahill, R.T., Klinger, C.M. and Kitto, K. (2000). {\it Process Physics:
1074: Modelling Reality as Self-Organising Information}, {\it The Physicist} {\bf 37}(6), 191, gr-qc/0009023.
1075:
1076: \bibitem{MC} Chown, M. (2000). {\it Random Reality}, {\it New Scientist}, Feb 26, {\bf 165}, No
1077: 2227, 24-28.
1078:
1079:
1080: \bibitem{Wheeler} Wheeler, J.M. (1983). {\it Law without Law}, in {\it Quantum Theory of Measurement},
1081: Wheeler, J.A. and Zurek, W.H., eds. (Princeton Univ. Press).
1082:
1083: \bibitem{neural} M\"{u}ller, B., Reinhardt, J.
1084: and Strickland, M.T. (1995). {\it Neural Networks - An Introduction} 2nd ed. (Springer).
1085:
1086: \bibitem{G} Nagel, E. and Newman, J.R. (1958). {\it G\"{o}del's Proof} (New York University Press).
1087:
1088: \bibitem{Chaitin90} Chaitin, G.J. (1990). {\it Information, Randomness and Incompleteness}, 2nd
1089: ed. (World Scientific).
1090:
1091: \bibitem{Chaitin99} Chaitin, G.J. (1999). {\it The Unknowable} (Springer-Verlag).
1092:
1093: \bibitem{Chaitin01} Chaitin, G.J. (2001). {\it Exploring Randomness} (Springer-Verlag).
1094:
1095: \bibitem{SOC} Bak, P., Tang, C. and Wiesenfeld, K. (1987). {\it Phys. Rev. Lett.} {\bf 59}, 381; (1988)
1096: {\it Phys. Rev. A} {\bf 38}, 364.
1097:
1098: \bibitem{negentropy} Schr\"{o}dinger, E. (1945). {\it What is Life?} (Cambridge Univ. Press).
1099:
1100: \bibitem{Prigogine} Nicholis, G. and Prigogine, I. (1997).
1101: {\it Self-Organization in Non-Equilibrium Systems: From
1102: Dissipative Structures to Order Through Fluctuations}
1103: (J. Wiley \& Sons, New York).
1104:
1105: \bibitem{Nagels} Nagels, G. (1985). {\it Gen. Rel. and Grav.} {\bf 17}, 545.
1106:
1107: \bibitem{foam1} Wheeler, J.A. (1964). {\it Relativity, Groups and Topology}, ed. by B.S. Dewitt and C.M. Dewitt
1108: (Gorden and Breach, New York).
1109:
1110: \bibitem{Ogden} Ogden, R.W. (1984). {\it Non-Linear Elastic Deformations} (Halstead Press,
1111: New York).
1112:
1113: \bibitem{Manton1} Manton, N.S. and Ruback, P. J. (1986). {\it Skyrmions in Flat Space and
1114: Curved Space}, {\it Phys. Lett. B} {\bf 181}, 137.
1115:
1116: \bibitem{Manton2} Manton, N.S. (1987). {\it Geometry of Skyrmions}, {\it Comm. Math. Phys.} {\bf
1117: 111}, 469.
1118:
1119: \bibitem{GP} Gisiger, T. and Paranjape, M.B. (1998). {\it Recent Mathematical Developments in
1120: the Skyrme Model}, {\it Physics Reports} {\bf 36}, 109.
1121:
1122: \bibitem{Baby} Coleman, S., Hartle, J.B., Piran, T. and Weinberg, S. eds. (1991). {\it Quantum Cosmology and
1123: Baby Universes} (World Scientfic, Singapore).
1124:
1125: \bibitem{QSD} Percival, I.C. (1998). {\it Quantum State Diffusion} (Cambridge Univ. Press).
1126:
1127:
1128: \bibitem{Penrose} Penrose, R. (1989). {\it The Emperors New Mind: Concerning Computers, Minds and the
1129: Laws of Physics} (Oxford University Press, Oxford).
1130:
1131: \bibitem{GCM} Cahill, R.T. (1989). {\it Aust. J. Phys.} {\bf 42}, 171; Cahill, R.T. (1992) {\it Nucl. Phys. A} {\bf
1132: 543}, 63; Cahill, R.T. and Gunner, S.M. (1998). {\it Fizika B} {\bf 7}, 171.
1133: %,hep-ph/9812491.
1134:
1135: \bibitem{Preons1} Marshak, R.E. (1993). {\it Conceptual Foundations of Modern Particle Physics}
1136: (World Scientific).
1137:
1138: \bibitem{Preons2} Dugne, J-J., Fredriksson, S., Hansson, J. and Predazzi, E. (2000). {\it Preon
1139: Trinity: A New Model of Leptons and Quarks},
1140: 2nd International Conference Physics Beyond The Standard Model: Beyond The Desert 99: Accelerator, Nonaccelerator And
1141: Space Approaches ,Proceedings, edited by H.V. Klapdor-Kleingrothaus, I.V. Krivosheina. Bristol,
1142: IOP, 1236.%, hep-ph/9909569.
1143:
1144: \bibitem{randomg} Bollab\'{a}s, B. (1985). {\it Random Graphs} (Academic Press, London).
1145:
1146: \bibitem{cosmo} Riess, A. {\it et al.} (1998). {\it Astron. J.}, {\bf 116}, 1009(1998);
1147: Perlmutter, S. (1999). {\it et al.}, {\it Astrophys. J.}, {\bf 517},565.
1148:
1149:
1150: \bibitem{Toffoli} Toffoli, T. (1990), in {\it Complexity, Entropy and the Physics of Information}, p
1151: 301, Zurek W.H., ed. (Addison-Wesley).
1152:
1153: \bibitem{Hardy} Hardy, L. (1992). {\it Quantum Mechanics, Local Realistic Theories and Lorentz-Invariant
1154: Realistic Theories}, {\it Phys. Rev. Lett.}, {\bf 68}, 2981.
1155:
1156:
1157:
1158:
1159: \bibitem{ProcessPhilosophy} Browning, D. and
1160: Myers, W.T., eds. (1998). {\it Philosophers of Process}, 2nd ed. (Fordham Univ. Press).
1161:
1162:
1163:
1164: \end{thebibliography}
1165:
1166: \end{document}
1167: