Dear Dorian, did you read Walter Freeman's review of the new Karl Pribram book? It contains an answer to your question: http://www.worldscientific.com/doi/abs/10.1142/S0219635214800028?journalCode=jin

Dear Dorian, did you read Walter Freeman's review of the new Karl Pribram book? It contains an answer to your question: http://www.worldscientific.com/doi/abs/10.1142/S0219635214800028?journalCode=jin

Indeed,  Walter Freeman's review  contains an answer, as well the new book of Karl Pribram.

What does it mean  for  HPB?

It means (in my view) that their concept of brain structure and function is wrong, as well as their concept of scientific progress (that was based on the Genome Project), and then no matter how much money is invested the desired results will not appear. However, they may find interesting technical results - not relevant for a better understanding of the real brain - by accident, as happened with Pasteur...

Do you also make a criticism, based on your theory of brain function, or do you support them?

Based on the current rates of success in understanding how the brain “computes” it may take at least 50 years to realize that we we have followed a wrong path.

By accident, ten years ago we found that action potentials were not stereotyped signals. We tried to tell everyone the issue regarding temporal coding approximation. Unfortunately, in order to preserve the temporal coding theory  it was a concerted effort to hide this information almost everywhere.

And even more, three years ago, from Arizona to Caltech and back to Leicester in UK some pillars of the neuroscience community laughed at me proposing “the concept cell” - a very revolutionary idea at that time http://neuroelectrodynamics.blogspot.com/p/concept-cells.html

Was it so revolutionary?

No, not at all. They repackaged the old grandmother cell concept and they lied to everybody pretending they didn't get the issue regarding temporal patterns and neuroelectrodynamics http://www.ncbi.nlm.nih.gov/pubmed/22262533 .

Now, no one's laughing.

Modha has failed to deliver the “cat” with 1 billion neurons and 10 trillion synapses  - Gerstner’s and Izhikevich spiking neurons do not provide more than previous generations (sigmoidal neurons). Also, Markham's repeated attempts to create the human conscious brain by reverse-engineering are no better than Modha’s projects.

I hope that scientists in Brazil have more honesty, teach less lies, old myths about the real brain http://neuroelectrodynamics.blogspot.com/p/myths-about-brain.html and fund more realistic projects.....

There are a lot of ways to succeed where so many pillars of the neuroscience community have failed.

Dear Dorian, I like the abstract of your paper with Jog and Poznanski. This interactionist model seems to fit well with Freeman's ideas. What is missing is the contribution of glia. I have done theoretical work on neuro-glial interactions and cognitive computation, but was not able to get funding for experimental projects.

The main focus of the paper was on computing forms, to show the presence of  a different form of computation, more powerful and far more efficient than everything we know or have ever studied in computer science - the role of glial cells , astrocytes is included http://www.iospress.nl/book/neuroelectrodynamics/

In fact your funding was blocked by theoretical development, no need  to study anything else beyond temporal patterns or connectivity (synapses). 

Unfortunately good experimental work , data analysis and theoretical work are highly separated today. Many excellent labs that perform experiments are not able develop simple theoretical models, they end up reinforcing obsolete models

Your book looks great, but too expensive! Can you tell us here how you included astrocytes in neuroelectrodynamics? Did you account for possible informational/mental functions of calcium waves? You can find my first speculations on this topic in the link below, and the final results in the paper published in the special issue of Cognitive Computation in which you also participated (edited by Bobbitt, Massad and myself). One of the main problems of Markram seems to be ignoring the role of glial cells in the generation of feelings and in support of conscious mental activity.

http://www.quantumbiosystems.org/admin/files/QBS2%2080-92.pdf

Yes, MacVicar's work regarding voltage-dependent calcium channels in glial cells and also the potassium homeostasis were included.

Clearly, without glial cells the brain would not be able to perform "computations".In fact in order to describe this new form of "computation" we have highlighted that the process of computation in the brain cannot be reduced to "neural" computation. 

The book was written before 2009, it was hard to find a publisher since at that time the spike timing dogma was at its peak. And I completely agree, for consciousness the quantum hypothesis needs to be seriously considered, nice paper.

Some paragraphs from http://www.iospress.nl/book/neuroelectrodynamics/ are included 

.....Other cellular structures including glia also regulate this extracellular flux (MacVicar, 1984; Volterra and Steinhäuser, 2004). As an example theastrocytes provide potassium homeostasis by removing the excess of ions (Nielsen et al., 1997). Since glial cells outnumber neurons (Steinhaeuser and Gallo 1996) their role in charge motion (or flux), homeostasis and computation cannot be ignored. Theexistence of such refined regulatory mechanisms at the brain scale proves that the dynamics of electrical charges is an essential component in this process and is not solely the attribute of neurons.

.....Since the homeostatic processes including ion regulation in the brain is provided by glial cells, a deeper understanding of their role in assisting and maintaining normal ionic environment is essential and cannot be ignored. Therefore, the term ‘neural computation’ doesn’t seem to fully describe the complexity and dynamics of electric charges in brain. This is the main reason why we prefer to avoid the term “neural computation” and we think that ‘brain computations’ or ‘brain language’ as a metaphoric expression are more appropriate to describe computational processes in the brain.

.....The thermodynamic price is required to create a non-even distribution

of particles in chambers and some energy resources are needed to make this process physically realizable. Such mechanism of computation is similar to neuronal spike phenomena where sodium and potassium ions are driven in or out of the neuron and the ionic homeostasis in the extracellular space is maintained by glial cells/astrocytes. This framework can be extended to any particle, molecule that crosses the membrane barriers.

.....

If one imagines charges as particles, neurons can be reduced to these

chambers where the bath may represent the extracellular space. Glial cells

/astrocytes are responsible for homeostatic processes and regulate the

required concentration of charges needed to restart and perform the

computational process.

Hi Dorian,

What about population coding? BTW I did take a look at the price of your book .. it's too expensive. Maybe a paperback version at half price might help sales $174 is steep.

I know, probably the second edition will be better and less expensive since we are no longer close to the peak of the spike timing dogma.

The basic idea of a “network” model is that neurons can cooperate in neuronal coding (Hebb 1949 ) so the term “population coding” just repackages the network idea, no fundamental change.

As long as spikes from each neuron are treated as identical, many scientists are “throwing away the baby with the bathwater” . The simplified model simulated on digital computers  cannot reveal what the real brain does (e.g. semantics) http://neuroelectrodynamics.blogspot.com/p/from-spike-timing-dogma-to.html .

So changing the name from “network” to “populations” or neuronal “ensembles” and adding mathematically well-formulated solutions (e.g probabilistic models, Bayesian “population” codes) does not solve the main issue.

The main problem is that mental states and consciousness cannot be instantiated using algorithms  on digital computers as Markram wants. No theoretical framework or concrete evidence to support his claims, so the scam part and their protest http://neurofuture.eu is justified

A separation of the brain in networks (population of neurons) is there just for our convenience. The whole brain is the “computing machine” and to succeed to understand how mental states and consciousness are instantiated and to treat brain disorders (e.g. autism) we need a different approach.

Is Sejnowski right?

https://www.facebook.com/photo.php?v=10152595712149411&fref=nf

I apologize for my shamelessness, but you can not talk about brain information processing without defining what actually is information. And then recognizing that there are two kinds of information - Physical Information and Semantic Information. You can go to my web-site (www.vidia-mant.info) and learn more about the issues. Again, I apologize that I can not take part in an alchemy inspired discourse.

Excellent remark, in fact that's the important part .

We have clearly highlighted this difference. That's  the fundamental issue for temporal patterns, no information about semantics.

The firing rate does not tell WHAT information was “read” or “written”. In fact the firing rate doesn't tell the meaning -semantics,  in two different  experiments http://neuroelectrodynamics.blogspot.com/p/spike-directivity.html

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3636996/

Repackaging the same old concept over and over becomes  the issue. Anything that we write today needs at least two, three decades to make a real impact  to change the way we move forward.