The preparation of an article about number theoretic aspects of TGD forced to go through various related ideas and led to a considerable integration of the ideas. In this note ideas related directly to consciousness and cognition are discussed.
 Adelic approach strongly suggests the reduction of NMP to number theoretic physics somewhat like second law reduces to probability theory. The dimension of extension rationals characterizing the hierarchy level of physics and defined an observable measured in state function reductions is positive and can only increase in statistical sense. Therefore the maximal value of entanglement negentropy increases as new entangling number theoretic degrees of freedom emerge. h_{eff}/h=n identifiable as factor of Galois group of extension characterizes the number of these degrees of freedom for given spacetime surfaces as number of its sheets.
 State function reduction is hitherto assumed to correspond always to a measurement of density matrix which can be seen as a reaction of subsystem to its environment. This makes perfect sense at spacetime level. Higher level measurements occur however at the level of WCW and correspond to a localization to some sector of WCW determining for instance the quantization axes of various quantum numbers. Even the measurement of h_{eff}/h=n would measure the dimension of Galois group and force a localization to an
extension with Galois group with this dimension. These measurements cannot correspond to measurements of density matrix since different WCW sectors cannot entangle by WCW locality. This finding will be discuss in the following.
Evolution of NMP
The view about Negentropy Maximization Principle (NMP) has coevolved with the notion of self
and I have considered many variants of NMP.
 The original formulation of NMP was in positive energy ontology and made same predictions as standard quantum measurement theory. The new element was that the density matrix of subsystem defines the fundamental observable and the system goes to its eigenstate in state function reduction. As found, the localizations at to WCW sectors define what might be called selfmeasurements and identifiable as active volitions rather than reactions.
 In padic physics one can assign with rational and even algebraic entanglement probabilities number theoretical entanglement negentropy (NEN) satisfying the same basic axioms as the ordinary Shannon entropy but having negative values and therefore having interpretation as information. The definition of padic negentropy (real valued) reads as S_{p}= ∑ P_{k}log(P_{k}_{p}), where  . _{p} denotes padic norm. The news is that N_{p}= S_{p} can be positive and is positive for rational entanglement probabilities. Real entanglement entropy S is always nonnegative.
NMP would force the generation of negentropic entanglement (NE) and stabilize it. NE resources of the Universe  one might call them Akashic records would steadily increase.
 A decisive step of progress was the realization is that NTU forces all states in adelic physics to have entanglement coefficients in some extension of rationals inducing finiteD extension of padic numbers. The same entanglement can be characterized by real entropy S and padic negentropies N_{p}, which can be positive. One can define also total padic negentropy: N= ∑_{p} N_{p} for all p and total negentropy N_{tot}=NS.
For rational entanglement probabilities it is easy to demonstrate that the generalization of adelic theorem holds true: N_{tot}=NS=0. NMP based on N_{tot} rather than N would not say anything about rational entanglement. For extensions of rationals it is easy to find that NS>0 is possible if entanglement probabilities are of form X_{i}/n with X_{i}_{p}=1 and n integer. Should one identify the total negentropy as difference N_{tot}=NS or as N_{tot}=N?
Irrespective of answer, large padic negentropy seems to force large real entropy: this nicely correlates with the paradoxical finding that living systems tend to be entropic although one would expect just the opposite: this relates in very interesting manner to the work of biologists Jeremy England. The negentropy would be cognitive negentropy and not visible for ordinary physics.
 The latest step in the evolution of ideas NMP was the question whether NMP follows from number theory alone just as second law follows form probability theory! This irritates theoretician's ego but is victory for theory. The dimension n of extension is positive integer and cannot but grow in statistical sense in evolution! Since one expects that the maximal value of negentropy (define as NS) must increase with n. Negentropy must increase in long run.
Number theoretic entanglement can be stable
Number theoretical Shannon entropy can serve as a measure for genuine information assignable to a pair of entanglement systems. Entanglement with coefficients in the extension is always negentropic if entanglement negentropy comes from padic sectors only. It can be negentropic if negentropy is defined as the difference of padic negentropy and real entropy.
The diagonalized density matrix need not belong to the algebraic extension since the probabilities defining its diagonal elements are eigenvalues of the density matrix as roots of N:th order polynomial, which in the generic case requires ndimensional algebraic extension of rationals. One can argue that since diagonalization is not possible, also state function reduction selecting one of the eigenstates is impossible unless a phase transition increasing the dimension of algebraic extension used occurs simultaneously. This kind of NE could give rise to cognitive entanglement.
There is also a special kind of NE, which can result if one requires that density matrix serves a universal observable in state function reduction. The outcome of reduction must be an eigen space of density matrix, which is projector to this subspace acting as identity matrix inside it. This kind NE allows all unitarily related basis as eigenstate basis (unitary transformations must belong to the algebraic extension). This kind of NE could serve as a correlate for "enlightened" states of consciousness. Schrödingers cat is in this kind of state stably in superposition of dead and alive and state basis obtained by unitary rotation from this basis is equally good. One can say that there are no discriminations in this state, and this is what is claimed about "enlightened" states too.
The vision about number theoretical evolution suggests that NMP forces the generation of NE resources as NE assignable to the "passive boundary of CD for which no changes occur during sequence of state function reductions defining self. It would define the unchanging self as negentropy resources, which could be regarded as kind of Akashic records. During the next "reincarnation after the first reduction to opposite boundary of CD the NE associated with the reduced state would serve as new Akashic records for the time reversed self. If NMP reduces to the statistical increase of h_{eff}/h=n the consciousness information contents of the Universe increases in statistical sense. In the best possible world of SNMP it would increase steadily.
Does NMP reduce to number theory?
The heretic question that emerged quite recently is whether NMP is actually needed at all! Is NMP a separate principle or could NMP reduced to mere number theory? Consider first the possibility that NMP is not needed at all as a separate principle.
 The value of h_{eff}/h=n should increase in the evolution by the phase transitions increasing the dimension of the extension of rationals. h_{eff}/h=n has been identified as the number of sheets of some kind of covering space. The Galois group of extension acts on number theoretic discretizations of the monadic surface and the orbit defines a covering space. Suppose n is the number of sheets of this covering and thus the dimension of the Galois group for the extension of rationals or factor of it.
 It has been already noticed that the "big" state function reductions giving rise to death and reincarnation of self could correspond to a measurement of n=h_{eff} implied by the measurement of the extension
of the rationals defining the adeles. The statistical increase of n follows automatically and implies statistical increase of maximal entanglement negentropy. Entanglement negentropy increases in statistical sense.
The resulting world would not be the best possible one unlike for a strong form of NMP demanding that negentropy does increaes in "big" state function reductions. n also decrease temporarily and they seem to be needed. In TGD inspired model of biocatalysis the phase transition reducing the value of n for the magnetic flux tubes connecting reacting biomolecules allows them to find each other in the molecular soup. This would be crucial for understanding processes like DNA replication and transcription.
 State function reduction corresponding to the measurement of density matrix could occur to an eigenstate/eigenspace of density matrix only if the corresponding eigenvalue and eigenstate/eigenspace is expressible using numbers in the extension of rationals defining the adele considered. In the generic case these numbers belong to Ndimensional extension of the original extension. This can make the entanglement stable with respect to state the measurements of density matrix.
A phase transition to an extension of an extension containing these coefficients would be required to make possible reduction. A step in number theoretic evolution would occur. Also an entanglement of measured state pairs with those of measuring system in containing the extension of extension would make possible the reduction. Negentropy could be reduced but higherD extension would provide potential for more negentropic entanglement and NMP would hold true in the statistical sense.
 If one has higherD eigen space of density matrix, padic negentropy is largest for the entire subspace and the sum of real and padic negentropies vanishes for all of them. For negentropy identified as total padic negentropy SNMP would select the entire subspace and NMP would indeed say something explicit about negentropy.
Or is NMP needed as a separate principle?
Hitherto I have postulated NMP as a separate principle. Strong form of NMP (SNMP) states that Negentropy does not decrease in "big" state function reductions corresponding to death and reincarnations of self.
One can however argue that SNMP is not realistic. SNMP would force the Universe to be the best possible one, and this does not seem to be the case. Also ethically responsible free will would be very restricted since self would be forced always to do the best deed that is increase maximally the negentropy serving as information resources of the Universe. Giving up separate NMP altogether would allow to have also "Good" and "Evil".
This forces to consider what I christened weak form of NMP (WNMP). Instead of maximal dimension corresponding to Ndimensional projector self can choose also lowerdimensional subspaces and 1D subspace corresponds to the vanishing entanglement and negentropy assumed in standard quantum measurement theory. As a matter fact, this can also lead to larger negentropy gain since negentropy depends strongly on what is the large power of p in the dimension of the resulting eigen subspace of density matrix. This could apply also to the purely number theoretical reduction of NMP.
WNMP suggests how to understand the notions of Good and Evil. Various choices in the state function reduction would correspond to Boolean algebra, which suggests an interpretation in terms of what might be called emotional intelligence . Also it turns out that one can understand how padic length scale hypothesis  actually its generalization  emerges from WNMP.
 One can start from ordinary quantum entanglement. It corresponds to a superposition of pairs of states. Second state corresponds to the internal state of the self and second state to a state of external world or biological body of self. In negentropic quantum entanglement each is replaced with a pair of subspaces of state spaces of self and external world. The dimension of the subspace depends on which pair is in question. In state function reduction one of these pairs is selected and deed is done. How to make some of these deeds good and some bad? Recall that WNMP allows only the possibility to generate NNE but does not force it. WNMP would be like God allowing the possibility to do good but not forcing good deeds.
Self can choose any subspace of the subspace defined by k≤ Ndimensional projector and 1D subspace corresponds to the standard quantum measurement. For k=1 the state function reduction leads to vanishing negentropy, and separation of self and the target of the action. Negentropy does not increase in this action and self is isolated from the target: kind of price for sin.
For the maximal dimension of this subspace the negentropy gain is maximal. This deed would be good and by the proposed criterion NE corresponds to conscious experience with positive emotional coloring. Interestingly, there are 2^{k}1 possible choices, which is almost the dimension of Boolean algebra consisting of k independent bits. The excluded option corresponds to 0dimensional subspace  empty set in set theoretic realization of Boolean algebra. This could relate directly to fermionic oscillator operators defining basis of Boolean algebra  here Fock vacuum would be the excluded state. The deed in this sense would be a choice of how loving the attention towards system of external world is.
 A map of different choices of kdimensional subspaces to kfermion states is suggestive. The realization of logic in terms of emotions of different degrees of positivity would be mapped to manyfermion states  perhaps zero energy states with vanishing total fermion number. State function reductions to kdimensional spaces would be mapped to kfermion states: quantum jumps to quantum states!
The problem brings in mind quantum classical correspondence in quantum measurement theory. The direction of the pointer of the measurement apparatus (in very metaphorical sense) corresponds to the outcome of state function reduction, which is now 1D subspace. For ordinary measurement the pointer has k positions. Now it must have 2^{k}1 positions. To the discrete space of k pointer positions one must assign fermionic Clifford algebra of second quantized fermionic oscillator operators. The hierarchy of Planck constants and dark matter suggests the realization. Replace the pointer with its spacetime ksheeted covering and consider zero energy energy states made of pairs of kfermion states at the sheets of the nsheeted covering? Dark matter would be therefore necessary for cognition. The role of fermions would be to "mark" the k spacetime sheets in the covering.
The cautious conclusion is that NMP as a separate principle is not necessary and follows in statistical sense from
the unavoidable increase of n=h_{eff}/h identified as dimension of extension of rationals define the adeles if this extension or at least the dimension of its Galois group is observable.
For details see the chapter Negentropy Maximization Principle or the article Reexamination of the basic notions of TGD inspired theory of consciousness.
