The brain is a rather complicated physical system in constant interaction with the external world or environment. Very generically and in grosso modo the brain functions as follows:

(i) Imagine that the brain is in some state | A), when some external stimulus is applied, for some given period of time, then

(ii) after the removal of the external stimulus, the brain is in some state | B) , which in principle should have in some way coded (or recorded) the "message" that was carried by the external stimulus, in such away that

(iii) "later" it is possible to retrieve (or recall) the message directly from the state | B), keeping in mind that

(iv) it is possible that the brain has not necessarily gone directly from | A) to | B), but many intermediate steps may have occurred: | A) -> | A1) -> | A2) ... -> | B) i.e., the information (or message) has been processed in the brain before it was recorded.

There are some fundamental properties that characterize successful brain function, namely: long-term stability and non-locality, as strongly suggested by the plethora of experimental data.

While the need for long-term stability is rather obvious, nonlocality, i.e., coherent neuronic activity at spatially remote cortical locations, makes the classical treatment of the brain function rather questionable. At the same time, non-locality is strongly suggestive of quantum treatment [11, 12, 13].

Since we are concerned here clearly with macroscopic states, and at the same time we need to invoke quantum treatment, we have to look at the so-called Macroscopic Quantum States (MQS), which are abundant in the quantum world. Superconductivity, superfluidity, magnetization, etc are typical examples of MQS with very specific characteristics:

(i) For special "structures" and "conditions",

(ii) a critical degree of coherence may beachieved that leads to an

(iii) ordered state, that is highly stable.

Consider for example Magnetization: the special "structures" are the Weiss regions, small regions in a ferromagnet within which all electron spins are polarized in a specific direction. Though, because there are many small regions and polarizations, on the average there is no magnetization visible in the ferromagnet. If we now apply a sufficiently strong magnetic field B or we decrease suffficiently the temperature (below the P. Curie point), i.e., the special "conditions", the ferromagnet exhibits magnetization because now all electron spins in the whole macroscopic crystal, are polarized in the same direction, strongly correlated with each other, thus leading to a highly stable macroscopic coherent (or quantum) state, the ordered state.

In a more physical language, the transition from an unordered state (e.g., many Weiss-regions) to an ordered state (e.g., magnetization) is called a phase transition.

The value(s) of the crucial parameter(s) (e.g., the magnetic field B or temperature T) at the transition point characterize the phase transition and define the critical point (e.g., Curie temperature).

It should be apparent that an ordered state contains more information (e.g., all electron spins polarized in the same direction) than the unordered state (electron spins randomly polarized).

On the other hand, the unordered state is more symmetric (randomly distributed electron spins are rotationally invariant, i.e., there is no preferred direction), while the ordered state exhibits less symmetry (polarized electron spins have chosen spontaneously a specific direction, thus breaking the rotational symmetry).

Thus, ordered states are the net result of spontaneous symmetry breaking that triggers the phase transition.

There are certain characteristics of phase transitions very useful for our subsequent discussions

(i) Universality: many, qualitatively and quantitatively different, systems can be described by the same phase transition.

(ii) Attractor: by varying suitably the system parameters, they can be brought close to their critical values, so as to cause a phase transition. It is not necessary to be infinitesimally close to the critical point. The critical point acts as an attractor for anything in its environment. In other words, we don't really need a fine-tuning of our system parameters to reach an ordered state.

(iii) Evolution equations: All the basic properties of phase transitions (including (i) and (ii) above) can be encoded in a set of evolution equations called renormalization group equations (RGEs). They describe deviations (and approach) from (to) criticality, as well as other characteristics of phase transitions [14].

Macroscopic coherent (or quantum) states, or ordered states have some highly exclusive characteristics:

(i) Long-range/term stability: highly stable, long-range correlations between the fundamental elements are maintained by wave-like, self-propagating excitation loops (e.g.: phonons, spin-waves, magnons, etc.) that regulate the behavior of the "other" fundamental elements and feedback to the original fundamental element that caused the "disturbance". I will call this the R+F property of MQS.

David Chalmers criticises models like Nanopoulos's here because, in spite of the exquisite theoretical constructs, still there is no logical link to qualia. The only way to make this link is to posit it. To say, that qualia is a fundamental potential of the quantum wave properties of matter that is actualized by back-action. Without this explicit postulate, Nanopoulos model is incomplete and fails. With it, it is plausible.

(ii) Nonlocality: clearly MQS, as its very nature indicates may go beyond microscopic locality.

(iii) Emergence: MQS have new properties that are not present at the fundamental elements level. The new properties characterize states at a hierarchical level above the level where the fundamental interactions among the fundamental constituents apply.

For example, superconductivity is a new property/phenomenon, i.e., emerging from a collective treatment of electrons under special circumstances, while of course each electron follows at the fundamental level the laws of quantum electrodynamics.

Let us use now the physical language of MQS and phase transitions to describe by analogy, for the time being, the basic functions of the Brain:

(I) Uncoded Brain: random signals, unattended perception are some of the characteristics of this case. It corresponds to the random polarizations in the many, small Weiss regions of the ferromagnet.

(II) Learning: An external stimulus is applied, say for a few seconds, that "straightens out" or "puts an order" to the random neuronic signals so that they are able to represent some coherent piece of information.

It corresponds, in the case of the ferromagnet, to applying for some time an external magnetic field B or lowering the temperature below the Curie point. They cause the breaking of the multi-domain small structures with their random polarizations,and thus they lead to the ordered state, where all electron spins, throughout the whole ferromagnet, are strongly correlated to all point in the same direction.

We are talking about a phase transition or, in the spirit of the previous discussion, a spontaneous breaking of some symmetry.

Clearly, it depends on the nature of the external stimulus with which specific fundamental elements will interact and set them "straight", so that a corresponding MQS, or ordered state, is created.

Realistically, in order to be able to encode all qualitatively different signals and create a coherent unitary sense of self, a tremendous number of qualitatively different ordered states is needed, i.e., practically an infinite number of qualitatively different spontaneously broken symmetries.

Furthermore, these symmetries should be accompanied by a set of selection rules, thus providing a physical filter against undesirable, irrelevant "stray" signals. A very tall order indeed, if one recalls the fact that, until now, the only "known" (observable) spontaneously broken symmetry, at the fundamental level, is the one describing the electroweak interactions. Just one, which is kind of short with respect to the desirable infinity of spontaneously broken symmetries! We will see later how string theory takes care of this problem.

(III) Coded Brain or Memory: the resulting, highly stable, coherent "firing" of a bunch of involved neurons, not necessarily localized, corresponding, in the case of the ferromagnet, to the stability and macroscopic nature (including nonlocality) of the emerging magnetization (ordered) state. Such a kind of naturally organized, coherent neuron firing, not necessarily localized, may provide the solution to the so-called "binding problem". More later.

(IV) Recall Process: In this picture, a replication weak signal, sufficiently resembling the learning signal, may excite momentarily the ordered state, but, thanks to its R+F property, it will relax back to its previous form.

It is this, ordered-state -> excitation -> ordered-state process that make us aware of recalling something, i.e., we "feel" it!

It corresponds in the case of the ferromagnet, to apply a weak magnetic field Bo, not necessarily exactly parallel to the original B, which will force the electron spins to oscillate, momentarily, before they relax back to their equilibrium, i.e., we recover the ordered state, thanks of course to the R+F property of MQS.

It should be stressed that it is not necessary for the replication signal to be exactly identical to the learning signal in order to recall full information, thanks to the attractor property of the phase transitions, discussed above.

In phase-transition language, the recall memory process corresponds to the act of an irrelevant operator.

It should not escape our attention that, in the framework of phase transitions, R+F and attractor properties facilitate tremendously the retrieving of information, without the need of complete identity of the replication and learning signal. Otherwise, it would take extraneous fine-tuning, which here translates to very long time periods, in order to retrieve information. Imagine what would happen if we need to see all the details of a fast approaching, hungry lion, including say the length and shape of its claws, before we run up a tree! Not very practical indeed.

The above presented generic picture for the brain function may sound plausible and promising. But, is there any "experimental" evidence for its support? The answer is yes.

The main observational tool is the Electro-Encephalo-Gram (EEG). It is usually (7) assumed that the EEG waveforms emerge from the summation of local neuron firings, but things are a bit more complicated.

One would expect that asynchronous firing of randomly distributed neurons would produce a zero net effect on the scalp electrodes. By studying electric potentials evoked during sensory stimulation and during learning trials, E. R. John has been able to show that these evoked potentials arise from the firing of large and disperse neural groups and that they are radically different from those obtained by the spontaneous random cortical activity [15].

Temporal rearrangement within the neural groups characterizes the externally evoked potentials.

Furthermore, Sayers et. al.[16], presented independent evidence strengthening the temporal rearrangement case, by studying EEG phase coherence. Frequency com- ponents of the EEG spectrum obtained during spontaneous cortical activity show a random configuration of phase relations, which shifts to a distinct pattern of phase coherence immediately following sensory stimulation.

Amazingly enough, imposing the phase characteristics of the evoked potential on the spontaneous waveform, we can reproduce the characteristic shape of the observed evoked waveform [16].

These findings support E. R. John's [15] case for temporal rearrangement, while at the same time it falsifies the kind of classical expectation that the EEG arises from the summation of neural firings, which would imply that just the amplitude characteristics is the only difference between spontaneous and evoked waveforms.

Clearly, it seems that the external stimulus does not just add energy to the brain, but it organizes it in a coherent way, in a similar fashion that an external field B acts on a ferromagnet!

It seems that the analogy between brain function and critical phenomena dynamics may be quite useful and fruitful.

In the unified approach suggested here (see Eq. (2)) the "effective" mental world (W2) is actively interacting with the emerging MQS, and thus through the R+F property of the MQS and the subsequent triggering by W2 of the collapse of the MQS, it provides the solution to the age-old problem of how intentional/emotional acts are strongly correlated to body acts, as explained in the Introduction.

It should be stressed that emergence here has a multi-valued meaning: it encompasses the natural (Darwinian [17]) evolution and selection, the development of brain in specific subjects and eventually the "conscious" moment under consideration.