Saturday, October 29, 2011
The attack on tenure in Ireland; Timeline
A few prefatory comments before the timeline. First of all, the decision
to end academic tenure is one that states have a right to make.
Secondly, current universities are highly inefficient entities,
particularly in an age when the Internet allows diffusion of information
and facilitation of communication in a way ideally geared to education.
Thirdly, in general academics in Ireland are grossly overpaid,
under-qualified and underworked – like most of the non-frontline civil
service there.
Nevertheless, what we have just witnessed in Ireland is sinister in the extreme, which is why I have remained so exercised about it. If the state wants to curtail academic freedom at its institutions, the loss is the state's; truly independent academies will spring up like mushrooms as talented students refuse to kowtow to nonsense backed by arbitrary authority. For example, globalrevolution.tv operates on a volunteer basis, and has been causative in all recent movements since Tahrir square, and is now livecasting the “occupy” protests. Taking the current MIT opencourseware and other creative commons documents to create “universities” is a piece of cake by comparison.
So why is the Irish situation worth writing about, let alone writing about in the rage which indeed has often possessed me? Essentially, because everything that happened there was illegal, and betokened a very dark future in the relation between power and citizen in Ireland. It is one thing to invoke procedures that have been biased irrevocably in the employer's favour since 1990; it is quite another to issue thousands of illegal employment contracts, keep an illegal statute on the books for over 10% of the lifetime of the Irish “state”, continually to insist in Parliament that the universities can act outside the law, cause the suicide of a foreign guest academic (anticipating Princeton), and in a final, Kafkaeque move, drop documents containing a summons to an illegal meeting into an academic's house late one Friday evening – all of which happened. Let me be clear; Ireland is again a joke internationally in any case; however, it is worth saying that we put up a fierce resistance to an MO that would have been sure to travel
We put this resistance up to the point of repeatedly and successfully suing Ireland's “paper of record”, the Irish Times, which – after for some reason joining the yellow press in a crusade against academic freedom – repeatedly had to issue retractions and apologies. The IT's participation is perhaps the best index of how deep this conspiracy was
Timeline
1985 The NIHE Dublin (NIHED)signs a closed-shop agreement with the FWUI union. Disciplinary procedures are specified, academic freedom is enshrined, and tenure is defined as a job until 65
1980's; Thatcher abolishes tenure, which has never been reinstated in Britain. In U hull, a case involving Edgar Page extends the destruction of tenure to academics tenured BEFORE Thatcher
Mid 1980's; an accreditation commission, headed by Michael Gleeson, repeatedly visits NIHED.
(By) 1987 Gleeson has accepted a job at NIHED
1989 The accreditation commission's report is submitted to Parliament, which transforms NIHED into DCU, with Senator Murphy among others issuing loud misgivings
1990 the effective end of tenure, or any permanency in any job in Ireland, is heralded with new legislation forbidding strikes for “single dismissals” which basically means – no employment rights whatsoever. Employees can summarily be terminated and either take their chances with utterly corrupt EAT and watch their savings be eked away as their erstwhile employer appeals or - if a state employee - can try to get a High Court Injunction. In the latter case, the state employs any one of the 3 gargantuan law firms in Ireland - all in the EU's top 20 biggest law firms as of 2009 - to crush him or her. The Irish state is violently in violation of EU law by not offering resources to the complainant
Mid 1990's An unsolicited tranche of cash in he tens of millions arrives at DCU, ostensibly from chuck Feeney, who claims to have given away all his money beforehand.
1995; DCU publishes a new closed-shop agreement with SIPTU. It abolishes academic freedom and tenure, and specifies that academics cannot work from their homes. In fact, SIPTU to this day has never agreed these terms, and the signatures were copied from the 1985 agreement .At least one of the “signers”, Pat Cullen, had left permanently for Australia 4 years beforehand. Hundreds of workers at DCU have signed these illegal contracts, which would not survive a court challenge
1997; A new universities act is passed by the outgoing government; it enshrines academic freedom and tenure, but refuses to say what the latter is. It also requires new disciplinary procedures to be enacted, subject to union consultation
2000; DCU governing authority refuses to pass a disciplinary procedure presented to it, condemning it as “draconian”
2000; Prondzz#@, third choice for the job and head of hull's law dept, is made President of DCU
2001; DCU's disciplinary procedure is made even more draconian, with a new clause, seized on later by Judge Clarke, allowing summary dismissal of all staff. Tom Mallon BL is called on to declare it kosher (his opinion is appended)
2001 DCU passes this new disciplinary procedure procedure, despite protest by SIPTU
2002 DCU is taken by SIPTU to the labour court and given 3 months to rectify the statute. 10 years after its being passed , it is still on the books
2007 Judge Clarke at Ireland's high court declares the statute incompatible with the 1997 act in its specification of tenure.
2011 Clarke's view is upheld by the supreme court's chosen remedy of full reinstatement with back pay in DCU's appeal, and their statement that there was an ILLEGAL, not an unfair dismissal at play
A personal epilogue
To a large extent, I do not have a dog in this fight. The Irish state has behaved in the antinomial and indeed criminal way that many of us expect it to do. I was fool enough to believe it would honor a contract with me; and in emigrating, I am simply returning to the home of my paternal grandfather. Yet
the experience of someone of non-Irish ethnicity stuffing letters in my door late at night in the name of the Irish state summoning me to an illegally-convened meeting (Edgar Page II) is something I will not tolerate nor easily forget. We have (perhaps overly-liberal) gun laws here in the USA to prevent such behavior by the state.
Finally, lest there be any doubt ; DCU was to be a proof of concept for direct corporate control of universities. In general, they have gotten away with it.
Seán O Nualláin Ph.D. Stanford
29u Deireadh fomhair 2011
Nevertheless, what we have just witnessed in Ireland is sinister in the extreme, which is why I have remained so exercised about it. If the state wants to curtail academic freedom at its institutions, the loss is the state's; truly independent academies will spring up like mushrooms as talented students refuse to kowtow to nonsense backed by arbitrary authority. For example, globalrevolution.tv operates on a volunteer basis, and has been causative in all recent movements since Tahrir square, and is now livecasting the “occupy” protests. Taking the current MIT opencourseware and other creative commons documents to create “universities” is a piece of cake by comparison.
So why is the Irish situation worth writing about, let alone writing about in the rage which indeed has often possessed me? Essentially, because everything that happened there was illegal, and betokened a very dark future in the relation between power and citizen in Ireland. It is one thing to invoke procedures that have been biased irrevocably in the employer's favour since 1990; it is quite another to issue thousands of illegal employment contracts, keep an illegal statute on the books for over 10% of the lifetime of the Irish “state”, continually to insist in Parliament that the universities can act outside the law, cause the suicide of a foreign guest academic (anticipating Princeton), and in a final, Kafkaeque move, drop documents containing a summons to an illegal meeting into an academic's house late one Friday evening – all of which happened. Let me be clear; Ireland is again a joke internationally in any case; however, it is worth saying that we put up a fierce resistance to an MO that would have been sure to travel
We put this resistance up to the point of repeatedly and successfully suing Ireland's “paper of record”, the Irish Times, which – after for some reason joining the yellow press in a crusade against academic freedom – repeatedly had to issue retractions and apologies. The IT's participation is perhaps the best index of how deep this conspiracy was
Timeline
1985 The NIHE Dublin (NIHED)signs a closed-shop agreement with the FWUI union. Disciplinary procedures are specified, academic freedom is enshrined, and tenure is defined as a job until 65
1980's; Thatcher abolishes tenure, which has never been reinstated in Britain. In U hull, a case involving Edgar Page extends the destruction of tenure to academics tenured BEFORE Thatcher
Mid 1980's; an accreditation commission, headed by Michael Gleeson, repeatedly visits NIHED.
(By) 1987 Gleeson has accepted a job at NIHED
1989 The accreditation commission's report is submitted to Parliament, which transforms NIHED into DCU, with Senator Murphy among others issuing loud misgivings
1990 the effective end of tenure, or any permanency in any job in Ireland, is heralded with new legislation forbidding strikes for “single dismissals” which basically means – no employment rights whatsoever. Employees can summarily be terminated and either take their chances with utterly corrupt EAT and watch their savings be eked away as their erstwhile employer appeals or - if a state employee - can try to get a High Court Injunction. In the latter case, the state employs any one of the 3 gargantuan law firms in Ireland - all in the EU's top 20 biggest law firms as of 2009 - to crush him or her. The Irish state is violently in violation of EU law by not offering resources to the complainant
Mid 1990's An unsolicited tranche of cash in he tens of millions arrives at DCU, ostensibly from chuck Feeney, who claims to have given away all his money beforehand.
1995; DCU publishes a new closed-shop agreement with SIPTU. It abolishes academic freedom and tenure, and specifies that academics cannot work from their homes. In fact, SIPTU to this day has never agreed these terms, and the signatures were copied from the 1985 agreement .At least one of the “signers”, Pat Cullen, had left permanently for Australia 4 years beforehand. Hundreds of workers at DCU have signed these illegal contracts, which would not survive a court challenge
1997; A new universities act is passed by the outgoing government; it enshrines academic freedom and tenure, but refuses to say what the latter is. It also requires new disciplinary procedures to be enacted, subject to union consultation
2000; DCU governing authority refuses to pass a disciplinary procedure presented to it, condemning it as “draconian”
2000; Prondzz#@, third choice for the job and head of hull's law dept, is made President of DCU
2001; DCU's disciplinary procedure is made even more draconian, with a new clause, seized on later by Judge Clarke, allowing summary dismissal of all staff. Tom Mallon BL is called on to declare it kosher (his opinion is appended)
2001 DCU passes this new disciplinary procedure procedure, despite protest by SIPTU
2002 DCU is taken by SIPTU to the labour court and given 3 months to rectify the statute. 10 years after its being passed , it is still on the books
2007 Judge Clarke at Ireland's high court declares the statute incompatible with the 1997 act in its specification of tenure.
2011 Clarke's view is upheld by the supreme court's chosen remedy of full reinstatement with back pay in DCU's appeal, and their statement that there was an ILLEGAL, not an unfair dismissal at play
A personal epilogue
To a large extent, I do not have a dog in this fight. The Irish state has behaved in the antinomial and indeed criminal way that many of us expect it to do. I was fool enough to believe it would honor a contract with me; and in emigrating, I am simply returning to the home of my paternal grandfather. Yet
the experience of someone of non-Irish ethnicity stuffing letters in my door late at night in the name of the Irish state summoning me to an illegally-convened meeting (Edgar Page II) is something I will not tolerate nor easily forget. We have (perhaps overly-liberal) gun laws here in the USA to prevent such behavior by the state.
Finally, lest there be any doubt ; DCU was to be a proof of concept for direct corporate control of universities. In general, they have gotten away with it.
Seán O Nualláin Ph.D. Stanford
29u Deireadh fomhair 2011
Tuesday, October 25, 2011
My last copyright assignment
One of the truly appalling consequences of DCU's criminality is that
many of my graduate students were (and are) brilliant - yet none of them
has had an academic career because of DCU's abuse and crime. And so I
hold my nose and submit to conventional academic publishers to keep my
ex-students publishing, including assigning copyright to the
publishers(I can't afford the 3k or so they ask as a price not to do
so). I will never do so again, and will restrict my submissions to
open-source media for reasons hopefully about to be made clear
On this blog, I announced a set of findings about metabolism and consciousness over the past 2 years. Last week, they finally got published in a peer-reviewed and indexed (etc) journal and I included one of the brilliant students shafted by DCU as an author - and deservedly, for he used his expertise in stats (he is a quant) to explode a myth about neural synchrony in epilepsy
The editing process simply added mistakes from my original document, which I append. They somehow assigned me to UCD, which I remember only as a source of hashish in its bar in the late 1970's; the great thing was that the dope expunged the garbage one was exposed to at the lectures. The one thing I share with Iar-Taoiseach Brian
Cowen is a vivid memory of NUI's Dublin drug emporium.
In any case, the paper with Springer's errors after a 3 month delay is at
http://www.springerlink.com/content/x10063878485504n/
and I append the original
Received 12 July 2011
Consciousness is cheap, even if symbols are expensive;
Metabolism and the brain’s dark energy
Seán O Nualláin
Lecturer, Symbolic Systems, Stanford
Ventura Hall, Stanford Ca 94305 USA
***@stanford.edu
and ***
Abstract
Use of symbols, the key to the biosemiotics field as to many others, required bigger brains which implied a promissory note for greater energy consumption; symbols are obviously expensive. A score years before the current estimate of 18-20% for the human brain’s metabolic demand on the organism, it was known that neural tissue is metabolically dear.
This paper first discusses two evolutionary responses to this demand, on both of which there is some consensus. The first, assigning care of altricial infants with burgeoning brains (and in human infants the metabolic demand peaks at 65% of the total) to “allomothers” is not unique to humans. The second, using relatively small neurons as primates do, risks misfires past a certain minimization. Moreover, in apparent paradox, there is an increasing consensus that large “Von Economo” neurons are critical for communication
This paper’s main original contribution is the discussion of two further evolutionary tricks. The first is the use of self-similarity in the cortex, both in structure and process, to allow the cortex readily - and in energetic terms, parsimoniously - to shift between states in a high-dimensional space. This leads to discussion of the kind of formalism appropriate to model these shifts, a formalism which – it is tentatively suggested – may do double duty for the modeling of symbolic thought.
The second trick is the superimposition on the background “white noise” of neural firing of EEG-detected waves like gamma. The paper describes a method, using the Hilbert transform, of calculating the dips in energy consumption as the brain is transitioned by gamma waves. It is hypothesized that consciousness may be a spandrel, the incidental result of a neurodynamic imperative that the brain enter a maximally sensitive (in sensory terms) “zero power” state a few times a second. If that is the case, then there are obvious benefits for health in meditation, which can be viewed as a state of consciousness extended over time by limiting afferent stimuli.
Keywords
Neurodynamics, thermodynamics, evolution, consciousness, biosemiotics
1 Introduction
We are destined, perhaps doomed, to project models from our current science on the phenomenal space of our minds. Currently, there is immensely exciting analysis going on at UC Berkeley (Freeman et al, 2008)using concepts from chaos theory, fractals, and many-body quantum theory to explain EEG data. This follows an equally exciting 1990's obsession with Bose-Einstein condensates, Renaissance viewsof the mind as a mill, Burton's being satisfied in 17th century “The anatomy of Melancholy” with mere gravity as a dynamo of neural exchange, Descartes and his famous hydraulics, and so on. The UC Berkeley work outlined here stresses the notion of a neural “wave packet, which is discussed below
It is cautiously proposed here that techniques which reveal discontinuities occurring in the order of tenths of seconds may be most fruitful, and recent ECOG work by the author and others will briefly be summarised. Phrenology, by contrast, was and remains an infamous 19th century fad. Bumps for covetousness, licentiousness, knowledge, and so on have a disturbingly medieval resonance, with the witch-hunt for supernumerary nipples to suckle incubi surely only a small distance away, both anatomically and politically, from Gall's crackpot activities. Yet the current obsession in neuroscience with fmri data is not much better. These data manifest statistical assumptions that are normally 10 years' study beyond the knowledge of the functionalist psychologists who presume to interpret them, are obtained in absurdly artificial environments, and - most of all- have a time sensitivity measurable in gross seconds.
Indeed, Vul et al (2009) suggest that many of the results obtained are dubious in their statistical reasoning. Vul et al’s meta-analysis has not been credibly questioned; including the word “voodoo” in the original title to describe the work of other scientists was an unfortunate borrowing from GHW Bush and perhaps diminished the paper’s impact.
It is uncontroversial in the extreme to suggest we need time sensitivity that is one or two orders of magnitude greater than fmri; it does seem to be the case that consciousness can be detected in the tenths of seconds, and that many critical neural events require only hundredths or thousandths of seconds. Specifically, work on microgestures (O Nualláin, 2010) indicates that a facial expression sustained for only 0.04 of a second, well below the sampling rate of consciousness, can affect our evaluation of a person. In turn, the sampling rate of consciousness can be assessed by examining what experimental subjects can actually report; to eschew pseudo-precision, and to anticipate some of the discussion below, it seems to be about a tenth of a second.
EEG and ECOG data (with the latter featuring electrodes attached directly to the cortex), on the other hand, as we have seen, can be sampled at appropriate intervals; indeed, precision well beyond the thousandth of the second level is possible. In our recent work (Freeman et al, 2008; O Nualláin, 2009), we have shown that superimposition of high-frequency gamma waves on the background “dark energy” of the cortex leads to momentary power dips therein, at a rate of several per second. The latter paper postulates that these dips constitute consciousness – perhaps to say “facilitate individual experience of consciousness” may be more precise - which is physically a discrete process and phenomenologically seamless only to a naïve experiencer. These gamma waves are precisely the ones emitted by trained meditators at very high levels.
On the other hand,, issues relating to the trades-off between increased brain size and consequent organismal metabolic demands are among the most fiercely contested in contemporary science ( Isler et al, 2006, 2009 a and b, Isler 2011). Consensus exists on relatively coarse-grained relations, like the positive ones between brain size, the relative altriciality of species, and the degree to which that species tolerates allomothers along with the biological mothers for child-care, a period in which the metabolic demand by the brain is around 65%(Fox, 2011) . Yet even that calculus needs nuance; the necessity of monitoring a much larger body means that a cow with its much larger brain is not necessarily any more “intelligent’ than a rodent. Conversely, humans reflect an evolutionary decision not to increase the size of individual neurons, even though the result may be ion channel misfires.
.
Animals with larger social networks, such as humans, inflict a large metabolic demand on the organism with 18-20% emerging as the consensus figure for humans for this “dark energy”. In this paper, we consider the extent to which brain waves like gamma attenuate the metabolic demand by the brain. It is hypothesized that consciousness itself, as we experience it, may be an artifact of metabolic demand minimization by the brain. While efficient representation algorithms like “sparse coding” free up such energy to some extent, it does seem plausible that synchronized gamma frees energy in excess of normal homeostasis to be used, perhaps, for healing as well as essential maintenance.
We begin with a brief review of the well-travelled evolutionary controversies, but with a twist; information theory predicts that the human brain may have reached certain constraints on the size of individual neurons. Evidence from the Freeman lab suggests that this energy demand may also be attenuated by the brain’s fractal architecture. It may change from one state to another in a high-dimensional space with utmost rapidity. Not incidentally, this fits in with the tensor approach to symbolic brain function pioneered by Smolensky and his colleagues, and we briefly consider the consequences for symbolic function.
The paper goes on to describe the functioning of the Hilbert transform, the data-analysis tool used here. Examples are given of its analysis both of simulated and real data. Then, it is noted that the synchronized gamma that is a signature both of meditation and consciousness has been found to attenuate the brain’s metabolic demand on the organism by a factor of 4. We then discuss the thorny issue of what consciousness actually is. We speculate on some essential maintenance tasks that may become possible due to this access of energy before – in further speculation – remarking that consciousness may after all be a spandrel, the result of a metabolic requirement. Finally, in an appropriately humble spirit, we consider possible consequences for symbolic functioning.
2 Metabolism and evolution
Given that many organisms found stable ecological niches scores of millions of years ago, the quickening of the evolution of symbolic intelligence that homo sapiens sapiens exemplifies with its inevitable metabolic cost requires explanation. One can hardly ask a more critical question for human Biosemiotics. Indeed, it has reasonably been asked whether the human brain is not an optimal vehicle for intelligence; to reach the human level of neurons per body weight, a rodent would have to carry around a 45kg brain, stopping any other salutary process (Fox, 2011) . Conversely , some animals like the koala bear have made an evolutionary decision to keep their brain size extremely small, and indeed reduce metabolic demand even further by sleeping 80% of the time. The route to big brains and symbols is not an inevitable one.
The battleground in evolutionary psychology (Isler et al, 2006, 2009 a and b, Isler 2011) ) is fully as contested as that neural communications theory (ibid. There is consensus in the former field on certain relations, like the positive ones between brain size, the relative altriciality of species, and the degree to which that species tolerates allomothers along with the biological mothers for child-care with marsupials being an exception explicable by a longer nursing phase. ( Isler et al (2006) were unaware of the current 18-20% “dark energy” estimate and simply cite work from a quarter-century before concerning the energetics of neural tissue)
Yet brute brain size does not correlate particularly well with intelligence, as larger mammals need bigger brains to maintain standard homeostatic functions. A better index plots body weight versus brain weight and here humans indeed are outliers with high relative brain size and cattle are outliers with low such.Moreover, the requirement for communicating across a large brain exerts its own discipline. Axonal mass increases faster than gray matter as brain size increases; a strategy that primates seem to have used is to grow with small neurons ( A possible exception involves Von Economo neurons, as noted below). Yet these smaller neurons and axons result in another bottleneck; the axons will become faulty below about 150 nanometers because of consequential misfires in ion channels, just as transistors below 10 nanometers fall prey to faults with the absence or presence of single atoms of boron causing misfires.
Much of the context for production of our experimental work is the well-attested finding that meditators show sustained brain activity in the “Gamma oscillation” category, which goes roughly from 25 to 80 hz (Freeman et al, 2008) with beta approximately from 12.5 up to 25Hz (ibid). Remarkably, there are solid indications that such gamma activity, particularly if synchronized over large areas of the brain, is a signature also of conscious states (O Nuallain, 2009). There is also at least provisional evidence that meditation leads to improved health, and indeed a thickening of cortex (ibid).
The work we did at Walter Freeman's lab at UC Berkeley (Freeman et al, 2008) indicates that there may be a unifying explanation for the beneficial health effects of meditation (or indeed conscious states). Very simply, the brain normally demands 18-20% of the total energy of the human organism (as distinct from 3% in the mouse). Synchronized gamma would briefly attenuate this demand by a factor on 10,000, and do so a few times a second. This frees up energy for many metabolic processes. Put in standard units, an average human consumes 100Watts; of this, the brain consumes 20W. Briefly to attenuate brain demand results in 20W energy being freed up.
It is worth pointing out that there is an assumption here that the extra energy freed up is "usable" for homeostatic functions. For example, the body can autonomously drive up its own heat to fight off infections in fevers. In that case, however, "normal" homeostasis in lost; indeed, the extra energy available to the brain causes hallucinations. This is speculative in our current state of knowledge
Famously, Gary Kasparov using 20w almost beat a computer using several thousand watts (Montague, 2008; of course, a laptop using a few hundred watts could beat most of us while multitasking). The efficiency of human cognition is indeed a marvel; Friston (2010) provides a good summary paper. However, part of the contribution of the present paper is to question and extend some of Friston’s arguments. Confusingly, Friston’s core concept is “free energy”, a term used also in biochemistry, on which he imposes a purely informational meaning. Even granted this latitude, we believe his scenario to evince premature closure, as described below. The brain is far from thermodynamic equilibrium; similarly, it wishes to process high-entropy data and, in order to do so, allows limit cycles. Finally, it is fair to say that Friston inevitably takes his schema into informational territory with remarks about issues like the thermodynamic efficacy of the related notion of “efficient coding” (Friston, 2010, 131)
There is one exception to humans’ preference for smaller neurons; von Economo neurons are inventions of large social animals, and abnormalities in their density in autistic brains relative to controls is attributed to the social deficits. In large social animals the von Economo neurons may regulate that higher social sense required to keep track of up to 150 different individuals in your social group, the Dunbar number (Dunbar, 2003) If that is the case, they are the exception that proves the rule to the generalization that primates use small neurons.
It is speculated in the framework of this paper that consciousness may not be causal. Cognitive resources may be directed to various tasks; conscious moments are replete with retrospective recreations of traces in the brain after these tasks are accomplished. In fact, attempting to become conscious of skilled performance (playing piano, driving) results in a deterioration of that performance. Similarly, establishing and nurturing bonds with the 150 will not necessarily require consciousness.
3 Brain architecture, symbols, and energy consumption
Friston (2010) and Carhart-Harris and Friston (2011) champion two major paradigms in neural processing of information; the “free energy” approach which constrains the new information presented by afferent stimuli to values tolerable in bit terms, and the “default” mode of neural functioning, which they have indeed used to rehabilitate Freud’s psychodynamics. The “default mode” posits brain waves at about 0.1Hz as its signature; “free energy” paradigms depend on a system in thermodynamic equilibrium with its environment. The “default mode” will not be a consideration of ours here, beyond the consensus that the metabolic demand by the brain does not evince any significant upsurge when the brain is “busy” (ie non-default), indicating that symbolic thought seems to exploit an already prepared system.
Our work (Freeman, 2008;Freeman et al, 2008) contrary to Friston’s (2010), proposes that the brain functions far from thermodynamic equilibrium. We suggest that the brain moves in an extremely high-dimensional state space in a manner facilitated by the fact that it is permeated with self-similarity, both spatial and temporal. At the spatial level, self-similarity ensures that, in fractal fashion, each neuron changes in step with the cortex as a whole. In temporal fashion, self-similarity results in the power spectrum of much slower theta waves traversing the brain being identical to, if slower than, much faster gamma waves. Several times a second, the brain enters a limit cycle in which it becomes extremely sensitive to incoming stimuli. Therefore, even without mechanisms like stochastic resonance, we humans can detect stimuli as weak as a few photons, or a few parts per billion of scent. The brain is geared to process novelty as it is to process low entropy information.
Our UC Berkeley work (Freeman et al, 2008) stresses the notion of a neural “wave packet”, which is not to be confused with its quantum mechanics synonym. In essence, a neural wave packet is a trajectory of successive states in the cortex begotten by a stimulus, either incoming or the result of internal processing. The trajectory flows through states which are local basins of attraction; subjectively, the sampling rate of consciousness is such that we consciously apprehend only a very sketchy sample of the perhaps trillions of individual computations that are ongoing.
EEG and ECOG data (with the latter featuring electrodes attached directly to the cortex), can be sampled at appropriate intervals; indeed, precision well beyond the thousandth of the second level is possible. In our recent work (Freeman et al, 2008; O Nualláin, 2009), we have shown that superimposition of high-frequency gamma waves on the background “dark energy” of the cortex leads to momentary power dips therein, at a rate of several per second. The latter paper postulates that these dips constitute consciousness – perhaps to say “facilitate individual experience of consciousness” may be more precise - which is physically a discrete process and phenomenologically seamless only to a naïve experiencer. These gamma waves are precisely the ones emitted by trained meditators at very high levels.
Moreover, it seems to be the case that the meditative state is devoid of cognitive content to a great degree. Finally (O Nualláin, 2009), gamma synchrony, the most attested concomitant of consciousness, may be significant only insofar as gamma is normally localized in specific brain regions; what (experience of) consciousness may consist of, in this scenario, is the result of resonant gamma over the cortex and its effect on the background noise/energy. We describe the methodology used below.
Smolensky et al (2011) suggest that the high-dimensional states that characterize the brain’s transitions need higher-order tensor calculus for their modeling. Smolensky was initially stung into action by the goading of Fodor and others who argued that neural network systems lacked “systematicity”, the capacity to handle certain syntactic relations normal in language. Thus, if “Fionn loves Grainne” is a syntactically valid sentence , syntactically (if not pragmatically) the system should be able to determine that “Grainne loves Fionn” is similarly valid. It was the failure of connectionist systems to do this that motivated Smolensky to develop a tensor theory in which neutrally plausible mechanisms could be adduced for each level of the Chomsky hierarchy.
The Chomsky hierarchy famously describes formal automata of ever-increasing power from finite state, through context-free and context-sensitive automata that are historically associated with natural language and finally to type 0 and type I automata that may or may not be recursively enumerable ( O Nuallain 2003, 2008). Finite state automata can most naturally generate sentences resembling a child's pivot grammar (daddy guess, mummy bad); suitably tortured, they can generate natural language. The consensus for natural language is somewhere between context free and context sensitive and below the type 0 grammars.
Chomskyan linguistics has historically seen its task as the formal characterisation of a grammar (syntax alone) or Grammar (incorporating all of language). No complete grammar for any natural language has yet been written; as for Grammar, the failure in this realm is best attested by the primitive natural language systems on the web, despite massive investment by companies like Google, as this is being written in 2011.
Similarly, the papers by William Hoffman and Ken Aizawa in our 1997 collection (O Nualláin et al 1997) propose that Lie groups and the formally equivalent tensors, should be used in cognitive modeling. Lie groups are the brainchild of the Norwegian Marius Sophius Lie whose work on symmetry extends the earlier opera of his brilliant and doomed compatriot Niels Abel and that of Abel's near contemporary, the similarly doomed Evariste Galois. Both Abel and Galois attacked the issue of why radical solutions to equations involving the quintic were impossible. Cardano and other brilliant Italian mathematicians had incidentally invented complex numbers in their successful search for solutions to equations involving the cubic.
The colourful details of the lives of these brilliant mathematicians need not concern us here, nor should their sometimes impenetrable formalisms. What is important is that Galois worked out that groups of permutation of the roots of an equation had properties that could constrain whether the equation in fact was soluble. Lie systematised Galois’ insight in a general theory of groups that turned out to provide a lingua franca with differentiable, locally Euclidean "manifolds" in space.Lie groups dominate current cutting edge thought in quantum theory, a further example of John Wheelers "unreasonable effectiveness of mathematics in the natural sciences"
Tensors, by contrast refer to a meta-formalism in which scalars (like the numbers “one, two”) are tensors of rank zero, vectors are tensors of rank one, and so on. Higher order tensors are possibly necessary for modelling brains function in the high dimensional space opened up by Freeman's neurodynamics. In any case, tensor calculus allows elegant modelling of the "cued reaching movements" described in Friston(2010, 134) . In particular, a single point in four dimensions must be targeted by a nervous system working in a much higher-dimensional space to catch a ball. Tensors, and thus Lie groups, afford a parsimonious way of approaching this problem.
Kime’s article in the same collection (O Nuallain et al, 1997) goes a few stages further - in what now seems a rather prescient move, he argues that the whole enterprise of cognitive science is inappropriately rooted in a belief that the data it deals with are the product of a linear system. Kime argues, following the seminal work of Pellionisz and Llinas (1979), that the space is in fact curves to varying degrees.
While Kime, following Pellionisz and Llinas (ibid), uses their contravariant/covariant terminology, the point is still the same; we may need a completely different paradigm to handle human use of symbols. In fact, the lacunae thrown up by the HGP may similarly thus be explicable. It may be the case that symbolic functioning proceeds over a high-dimensional state-space, with curvature actually changing as the interpretation of the symbol proceeds (O Nualláin et al 2007, p. 257) .
This hypothesis is tentative in the extreme; it suggests that symbolic functioning involves separate phases of “analysis” and “interpretation”. In the former phase, the organism finds out what the general context actually is. Interpretation involves a gradual restriction of this context and a change in the curvature of the system ( O Nualláin 2003, O Nualláin et al 2007). A good analogy is the warping of space-time by a massive object.
What this section has (very) tentatively proposed, therefore, is a different paradigm of brain function due to the attested work of Freeman and his colleagues. It implies a tensorial (or another mathematically equivalent) approach to cognitive function. It eschews Friston’s (2010) “free energy” paradigm as incomplete. Next, let us look at an attested mechanism for reducing metabolic demand.
4 The Hilbert Transform and dips in metabolic demand
An application of the Hilbert transform to both real and simulated ECOG data is described here. Readers with an aversion to technical detail can safely skip to 4.2
The Hilbert transform overcomes limitations in the Fourier Transform partly by repeatedly applying it to a changing signal. The Hilbert transform calculates attributes of a time series, especially the amplitude and instantaneous phase and frequency, before summarizing all of them in an “analytic signal”. Now we can calculate crucial attributes of the cortical activity, both real and simulated, like power. The investigation here confirmed the existence of “lulls” several times a second in the background activity of the brain which may have a central role in sensory processing.
The analytic signal is useful in the area of communications, particularly in bandpass signal processing, where it can be utilised for the minimisation of “aliasing” or “jagging” in graphics applications. The analytic signal has a real part corresponding to the original data, and an imaginary part which contains the Hilbert transform. The imaginary part is a version of the original real sequence with a 90° phase shift. Sines are therefore transformed to cosines and vice versa.
As described in any reliable source like the matlab manual, for a pure sinusoid, the instantaneous amplitude and frequency are constant. The instantaneous phase, however, is a sawtooth, reflecting the way in which the local phase angle varies linearly over a single cycle..
Briefly, we can write
a = Hilbert (b)
where the real part of a is the original real data and the imaginary part is the actual Hilbert transform. a is sometimes called the analytic signal, in reference to the continuous-time analytic signal.
The Hilbert transformed series has the same amplitude and frequency content as the original real data and includes phase information that depends on the phase of the original data. The instantaneous amplitude is the amplitude of the complex Hilbert transform; the instantaneous frequency is the time rate of change of the instantaneous phase angle. What is most critical for the purposes here is that with the Hilbert transform we can calculate the power of all activity in the brain, instant to instant. The reader is referred to Walter Freeman’s (2007) description of the Hilbert in Scholarpedia for further detail
4-1 Methodology
The plan, then, was first to simulate bandpass-filtered brown noise by filtering pseudo-white noise with a filter equivalent to running-summing followed by filtration with a finite-impulse-response bandpass filter such as can be generated in Matlab by fir1. Judicious choice of the order of the filter depends on the digitising step; with our step of 2ms, we found that 500 was an appropriate order, as numerical errors emerged with other choices.
We now choose bottom end and top end of the band of frequencies to be retained by the bandpass filter, expressing these as fractions of the Nyquist frequency. Essentially, the band at 1/80 to 1/40 of the Nyquist frequency was considered theta range; at 1/40 to 1/20 of the Nyquist frequency, alpha range; at 1/20 to 1/10 of the Nyquist frequency, beta range; and finally, at 1/10 to 1/5 of the Nyquist frequency, gamma range.
In each case we proceeded as follows;
1. We computed the corresponding analytic signal by taking the Hilbert transform of the filtered brown noise.
2. We computed the corresponding analytic amplitude (pointwise modulus), analytic phase (pointwise arg), analytic power (pointwise squared modulus), "unwrapped" analytic phase, and analytic frequency (the differenced analytic-phase series divided by the "digitizing step"). We then proceeded to plot analytic power against time, and analytic frequency against time.
For each of the bands we produced an example of the bandpass—filtered brown noise, a longitudinal histogram of the bandpass-filtered brown noise , a corresponding example of the analytic power series, a longitudinal log-log histogram of the analytic power, a corresponding example of the analytic frequency series, and a longitudinal log-log histogram of the analytic frequency. (Naturally, most of these graphs are surplus to the requirements here and only some are shown.) The bandpass simulates local inhibitory feedback; as we hypothesize, a short delay and feedback corresponds to gamma and long such to beta.
In the second stage, we used real data, as described below.
4-2 Results
In our work, the first diagram (fig 1) shows power consumption decreases with gamma first with simulated data, then with real data (fig 2), both of which have appeared in various forms in our previous work. The simulated data looked on neuron firing as random, as white noise with refractory periods. (Without refractory periods, epileptic seizure sets in, with the very clear and tragic consumption of metabolic energy that involves.) We then used the Hilbert transform to calculate how power consumption would vary if the “white noise” signal was convolved with gamma waves.
Several aspects of the results are discussed in O Nualláin (2009). Many immediately striking findings emerge from this use of the Hilbert transform. In the first place, we seem to have “lulls” in the power spectrum of the brain under gamma a few times per second (Freeman historically has argued for 3-7 times per second). In these lulls, the power consumption of the brain decreases by a factor of ten thousand. Let us look first of all at an idealization, in which the power reaches zero in a simulated case;
Figure 1 simulated data
The above is simulated data, derived as described above. We can compare the real data, which also shows “null spikes” occurring 3-7 times a second. Fig 2 uses data acquired, with her consent, from an 8*8 electrode array applied directly to the inferior temporal lobe of a woman undergoing surgery for epilepsy (Incidentally, we also found that epileptic seizures – contra the official wisdom – do NOT involve any special correlation of activity between discrete brain regions). In this case, the “white noise” is actual neural firing as clinically. recorded and the remaining steps to calculate to calculate power are as described above; let us note that energy consumption is power*time, so our power dips attenuate the energy consumption.
Fig 2 real data
As we pointed out (Freeman et al, 2008), the “null spikes” correspond to extremes in the instantaneous frequency, as depicted in figure 3;
Figure 3
Moreover, the power distribution shows that power consumption decreases by an order of 4 during the “null spikes”;
Fig 4
4-3 Seizure and energy
Epileptic seizure is the limit case, wherein the refractory periods in which neurons do not fire are abolished, and the brain’s demand on the organism becomes maximal – and tragically so. Moreover, in our view (Freeman et al, 2008), gamma and beta waves are caused by feedback from inhibitory neurons, and these obviously are drowned out in the case of seizure.
While there does seem to be a marked qualitative difference – and it is worth noting that all of our samples of epileptic patients show this monotonic increase during seizure – it will obviously be interesting to tease these data out more. In any case, it is clear that seizure is not an evolutionarily stable strategy!
So, for example, we can apply the Hilbert transform to the seizure data, and get a sense that gamma would provoke a massive reduction in the power consumption of the brain
This would rather seem to suggest that the absence of gamma and beta is one of the reasons that the brain's consumption of energy becomes so massive in the case of seizure. Refractory periods stabilize the gross activity occurring in the cortices; both these and the salutary effects of inhibition, which incidentally produce both gamma and beta, are now missing
Let it be made clear that the energy that we propose is saved from beta/gamma superposition is many orders of magnitude greater than that from more economical representations like “sparse” coding. If it is correct to interpret our original published work (Freeman et al, 2008) in the evolutionary context, as proposed here, this is an unprecedented energetic difference.
5 And so, what is Consciousness?
It is not controversial to suggest that there is no consensual definition of consciousness. By way of an anti-definition; consciousness reflects the operation of global dissemination and integration of information in the brain. It seems to be the case that it facilitates certain cognitive acts, like processing irregular past tenses; yet these can perhaps be just as easily explained by attention, without any phenomenology
We tend to view ourselves as initiating action having previously decided on it; however, the available data suggests otherwiseIn particular, classic work by Libet ( O Nuallain 2003)suggests that motor neurons show preparation to initiate action tenths of seconds before we are aware that we intend to do so. It is not a stretch to propose that we continually narrate to ourselves a “stream of consciousness” about our actions that is little more than what a sports commentator does about a game he is calling
If we accept that much of our conscious experience is simply an a posteriori commentary, biased to predicate virtue, consistency and potency of ourselves, none of this is surprising. So does consciousness do anything at all? The thesis here; perhaps consciousness does nothing, but is a handmaiden to attention. Alternatively put, we will ourselves to do things, and as an epiphenomenal result become aware of these actions late
I have covered these themes at some length in my books ; “The search for Mind” (1995, 2003), “Being human “ (2002, 2004), “Two sciences of mind” (1997), various papers (Eg 1994) and am going to deal only with two at any length here. The others I respect include L-dice, Global Workspace models, and Damasio's work.
Consciousness has a phenomenology in that humans use consciousness scan internal "space" in the same manner as they scan external objects and relations. In my current work (forthcoming) I point out that as a cognitive science object of study, consciousness is likely to remain at sensorimotor level. Of course as a political phenomenon it is multifarious. Those of us in the West have been encouraged some centuries to view consciousness is everything we are, disregarding the fact that the cells we hold so inviolate are partly the result of social construction as of Manufacturing consent.
The degree to which we view our individual consciousness as important varies between political systems, and with the amount of social control that obtains; the speculations are of course beyond the scope of this paper .( In previous work (1994) I discussed this kind of issue in detail ). Similarly I have discussed elsewhere (2006) the "experiential disciplines" which seek to reroute the "stream of consciousness" from Joycean machine, with an often self serving narrative, to being beads on a string with some wholeness
There are numerous phenomena, both in external world and interior space, identified with consciousness. Consciousness has been implicated by researchers like Stapp and Von Neumann in the act of observation in QM, absent some “objective reduction” principle (O Nualláin 2004). It has also been identified by researchers like Hameroff and Penrose as supervening on Bose-Einstein condensates somehow exteding over decimeters in the brain’s febrile environment. (ibid). Quantum coherence in biological systems is indeed being observed, but at a minute fraction of the scale required by Hameroff and Penrose.Yet, there is clear conceptual compatibiilty between gamma synhrony, so-called” zero power” and the desideratum of Quantum coherence, to be explored in a forum other than “Biosemiotics”.
There is also a seductive neo-Hegelian story that sees consciousness as ultimate reality, and the progress of human society as reflecting a path in which consciousness will repossess itself after the supreme act of alienation that comprised creation of the cosmos. Tweak this story to stress economic activity, and revolutionary Utopian socialism emerges as a consequence.
In internal space, the word “consciousness” has sometimes been reserved for moments in which the self can be identified as the locus of observation. This is the beginning restriction for thinkers such as Gurdjieff ; to do so requires prior work in acting authentically in the world, and self-integration. (Ouspensky, 1977; in my 2010 paper for this journal, I have attempted to bring this work into dialogue with contemporary science) Confusingly, Gurdjieff and his ilk identify the human search as being one for “consciousness”. In fact, it is likely that what the neophytes are looking for resembles rather a state of being in which they have command of all their faculties, including professional skills and abilities, and can function in a stable way within the society in a manner that does no violence to their reason. Neither Gurdjieff nor his later followers went even close to providing that kind of nexus. That said, it is indeed possible that the current configuration of our society, with tax being redirected into financial speculation (as distinct from deserving causes) that self-destructs regularly, and the increasingly violent encroachment of many of those self-same interests on both academia and the arts, is not optimal for human development
Conversely, consciousness has been misidentified as mere waking (versus sleep), as attention, and intentionality, the fact that it seems “pointed at' an object in the external world. . Researchers like Jason Brown (O Nualláin, 2003) have correctly pointed out the intricate links with selfhood and will, identifying will as a heightened sense of self with respect to an action (or inaction).
The neo-Piagetian view sees consciousness as constructed from many incidents of subject-object differentiation, a view that I made explicit in this journal (2008,2010), Pace John Ralston Saul and fulminations about “The unconscious civilization”, it can equally be argued that civilization involves the making unconscious/automatic of many processes, the better to indwell on what is important to us, to accord real decisions the greatest number of degrees of freedom as possible while needs lower in Maslow's hierarchy are offshored to the state and civil society. Similarly we are not aware (nor should we be), moment to moment, of how the sewage systems run or – until recently – how whole countries are financed.
Likewise, for each individual, we note skills like cycling a bike becoming unconscious/automatic, and then indwell on avoiding boy racers on the road. What we normally experience as “self” is a dynamic shifting of the subject/object boundary; achieving some stability here results in the person having the option of identifying self with the locus of awareness.
Moreover, it seems to be the case that the meditative state is devoid of cognitive content to a great degree. Finally (O Nualláin, 2009), gamma synchrony, the most attested concomitant of consciousness, may be significant only insofar as gamma is normally localized in specific brain regions; what (experience of) consciousness may consist of, in this scenario, is the result of resonant gamma over the cortex and its effect on the background noise/energy.
Of recent theories, Logie and Pearson's work in my 2000 collection describes consciousness as a mode of access to working memory. Tonoi's recent work(Zimmer, 2010), which sees consciousness as an epiphenomenon of ordered information, has the advantage of being wrong;
"He began to think of consciousness in a different way, as a particularly rich form of information...Consciousness is not simply about quantity of information, he says. Simply combining a lot of photodiodes is not enough to create human consciousness. ...Dr. Tononi argues that his Integrated Information Theory sidesteps a lot of the problems that previous models of consciousness have faced. It neatly explains, for example, why epileptic seizures cause unconsciousness. A seizure forces many neurons to turn on and off together. Their synchrony reduces the number of possible states the brain can be in, lowering its phi. ..."
Finally
“Dr. Tononi sees serious problems in these models. When people lose consciousness from epileptic seizures for instance, their brain waves become more synchronized. If synchronization were the key to consciousness, you would expect the seizures to make people hyperconscious instead of unconscious, he said. “
Jouny et al (2010) would suggest that this is premature closure, with an INCREASE in signal complexity - that is, decline in synchrony - associated with seizure. We (Doris et al, in preparation) looked at ECOG data from the inferior temporal lobe of a patient in the states of sleep, seizure and awake (resting brain) using principal component analysis (PCA) and dynamic time warping (DTW) and found that our results support Jouny, rather than Tononi;
Sleep signal is least complex/disordered under PCA, first component explains 97%, awake is next, with 93% explained by the first component, while seizure has just 63% explained by first component.
This is supported by the DTW result, which indicates that the mean distance is less in sleep than waking than seizure;
> mean(seizure.dtw.mat)
[1] 6.731411
> mean(awake.dtw.mat)
[1] 5.580975
> mean(sleep.dtw.mat)
[1] 4.204780
The PCA for sleep shows that there is effectively only one "signal" across the 64 channels. You can easily construct a linear model (through simple ordinary least squares –ols - regression) of any of the individual channels based on a small number ( < 10) of the other channels, and by doing this you will recover 99% of the channel. Awake is similar but less extreme, there is one dominant signal but there is some differentiation, while seizure still has a dominant signal the PCA shows it is much less influential and there are several other signals present in the telemetry. In info-theory terms, sleep has minimum entropy, awake next, and seizure has highest entropy by a long way. In short, if there is an “Integrated Information Theory” theory of consciousness, it already has one noticeable lacuna. In previous work (2004) I have discussed shortcomings in any informational theory of consciousness, in that they fail to pre-empt the spectre of conscious thermostats – and indeed rocks, to adapt Chalmers (ibid).
The viewpoint here, therefore, is that consciousness manifests itself when the cortex acts under the influence of gamma synchrony. At this point, phenomenology is possible, yet consciousness itself is non-causal in immediate terms. The narratives we weave as a result of the ability to reflect on our experience that consciousness affords us are indeed causal, perhaps the most causal factors in evolutionary history, for good and ill.
6 Comments on the biochemistry of metabolism
Energy released for use for the rest of the organism by states of meditation/consciousness can potentially be used for the following purposes as distinct from the obvious growth, and standard maintenance issues. This requires some preliminary discussion, some of which inevitably – given the immature state of the field – will be speculative
Energy production and consumption each take place on multiple time scales. Energy may be conserved in the form of useful energy reservoirs. Cyclical patterns for consuming energy take place at multiple time scales. Glucose is one store convertible on the time scale of seconds; thru the glycolysis cycle it makes ATP; ATP is the main (but not exclusive) energy currency of the cell on a short time scale and thus is very quickly available to build gradients by driving proton and other motors across membranes and to drive other reactions and motors that do other work, like moving cargo along an axon of a neuron along the microtubules; proton gradients are the ultimate, proximate store instantaneously available like battery charge for conducting nervous energy, for sustaining the charge balance across membrane by moving protons and metal ions particular in the energy conversion factories, the mitochondria.
Astrocytes should be interesting since their calcium waves may be related in periodicity, obviously critical here, and their attested role in neural metabolism. Indeed, to get very speculative indeed, the zero energy points on the Hilbert transformed data may be seen as instants that reveal the collapse of a nonlocal wave function; this could appear in spectrally coupled fields of astrocytes and neurons each set coherently coupled thru their respective gap junction networks.
To summarize, then, here are some effects beyong the glaringly obvious that the extra energy freed up by gamma waves, and incidentally facilitating consciousness, could be used for in the organism;
1. Getting glucose into cells, lessening the need for insulin. Insulin is only a signaling/gating molecule, it doesn’t have inherent toxicity except in excess relative to its native receptors when it starts working too much on it night jobs (cotransporting drugs, metabolites, and so on, and this is what we are trying to avoid). Dogs with their pancreases removed – if fit – do not become diabetic. Thus, conserving energy by managing an appropriate physiological state can reduce the need for insulin.
2. NAD+ and NADH are used as intermediates in producing ATP and the energy saved by gamma may easily be stored in these forms. They contribute to lots of other reactions too, like those involved in biosynthetic machinery making proteins that may or may not be involved in gene expression. They have a role in regulating post translational modification of histones, a process that is intimate with the management of gene expression and of great interest in the study of aging.
3. Gene expression varies with a lot of things. In changing environments it’s primarily controlled by nuclear and other receptors that sense the internal and external environmental changes of the cell by sensing molecules, temperature, etc. In particular, gene expression varies with the ratio of NAD+ to NADH which informs of the metabolic state and health of the cells. This will vary with the energy available .
4. Finally, there is preliminary evidence from the Blascovich lab that conscious experience focussed on an external target can dull pain perception - and indeed increase blood flow
7 Conclusion
The current debate on brain size and evolution is in some ways misguided. Sheer brain size does not guarantee intelligence. It is equally important to have fast communication in the cortex. However, that in turn leads to a set of trades-off between size of neurons, and indeed width of axons
Remarkably, it may be the case that consciousness is an opportunistic consequence of the necessity to curtail energy use by the human brain. The synchronized gamma associated with consciousness decreases by an order of four (if briefly) the metabolic demand by the brain on the organism. Such energy can be used for a variety of essential healing and maintenance functions. In this scenario, meditation is viewed as an extended, content-free, state of consciousness. At the other extreme, epileptic seizure involves massive consumption of energy without gamma to stabilize matters
Humility becomes us at this point. While I have in the past (O Nualláin 2008) proposed a tutorial in biosyntax, it is possible that our models are hopelessly naïve. It is as well to reiterate the ambiguity of the term “free energy” as we head to closure. Informationally, Friston and his school use it to privilege – uncontroversially – an account of neural functioning that stresses that the brain is a machine that works on hypotheses. The argument here, however, contends that this school fails to take into account the fact that the brain is a far-from-equilibrium system, both informationally and thermodynamically.
The dynamical systems notion of the brain here sees it as both protean and deeply self-similar. This led naturally to a discussion about the sort of models appropriate to encompass the symbolic functioning of the brain. It does seem to be the case that tensors and the formally equivalent Lie groups will soon be part of the neuroscientists’ diet. It may even be expedient methodologically to dispense with consciousness outside the social sciences – apart from the sensorimotor area that seems to be yielding its secrets to a cognitive science approach.
Acknowledgements
Many thanks to Marcello Barbieri for his encouragement, and to Tom Doris and Hal Cox for fruitful discussions
References
Carhart-Harris RL, Friston KJ. (2010) “The default-mode, ego-functions and free-energy: a neurobiological account of Freudian ideas.” Brain. 2010 Apr;133(Pt 4):1265-83. Epub 2010 Feb 28.\
Doris T and S. O'Nuallain (in preparation) “Seizure and complexity”
Dunbar R.I.M. The social brain: mind, language, and society in evolutionary perspective. Ann. Rev. Anthropol. 2003;32:163–181. doi:10.1146/annurev.anthro.32.061002.093158
Fox, D. (2011) “The limits of intelligence” Scientific American, July 2011, Pp 36-43
Freeman (2007)Hilbert transform for brain waves
Scholarpedia, 2(1):1338.
Freeman, W., S. O'Nuallain and J Rodriguez(2008) "Simulating cortical background electrocortigram at rest with filtered noise" Journal of integrated neuroscience,7 (3 )Page: 337 - 344 Sept 2008
Freeman, W (2008) “A pseudo-equilibrium thermodynamic model of information processing in nonlinear brain dynamics” Neural Networks
Volume 21, Issues 2-3, March-April 2008, Pages 257-265
Friston, K. (2010) “The free-energy principle”Nat Rev Neurosci. 2010 Feb;11(2):127-38. Epub 2010 Jan 13.
Gazzaniga, M. (1995) Consciousness and the cerebral hemispheres. In Gazzaniga, M. (ed.) (1995) The cognitive neurosciences Pp 1391-1400 Cambridge, Mass; MIT Press
Krishnamurti, J. (1979) Meditations Boston: Shambala
Isler, K. (2011) Energetic trade-offs between brain size and offspring production: marsupials confirm a general mammalian pattern. BioEssays 33(3). DOI 10.1002/bies.201000123.
Isler, K and van Schaik C.P (2006), “.Metabolic costs of brain size evolution”
Biol Lett. 2006 December 22; 2(4): 557–560.
Isler K, van Schaik CP.( 2009a) “The Expensive Brain: a framework for explaining evolutionary changes in brain size.” J Hum Evol. 2009 Oct;57(4):392-400. Epub 2009 Sep 3.
Isler, K., van Schaik, C.P. (2009s) Why are there so few smart mammals (but so many smart birds)? Biology Letters 5:125-129.
Jouny CC, Bergey GK, Franaszczuk PJ. (2010) “Partial seizures are associated with early increases in signal complexity.” Clin Neurophysiol. 2010 Jan;121(1):7-13. Epub 2009 Nov 11.
Montague R (2008) Your Brain Is (Almost) Perfect, NY;Perseus
Ó Nualláin Seán (1994): Some Consequences of Current Scientific Treatments of
Consciousness and Selfhood. AI Soc. 8(4): 305-314
O Nualláin, Seán (2010) “Ask not what you can do for yourself: Cartesian chaos, neural dynamics, and immunological cognition” Biosemiotics Vol 3 Issue 1, 2010 DOI 10.1007/s12304-009-9070-4O Nualláin, Seán (2003) "The Search for Mind" ( Third edition); Intellect
O Nualláin, Seán (2004) "Being Human: the Search for Order" (Intellect, 2004) second edition
O Nualláin, Seán (2000) (ed)"Spatial Cognition” (ed.) (Benjamins);
O Nualláin, Seán (1997) (ed)"Two Sciences of Mind" (Benjamins,);
O Nualláin, Seán (2008)“Subjects and Objects” Biosemiotics journal, Volume 2, Pp. 239-251
O Nualláin, Seán (forthcoming ) “Neural correlates of consciousness of what?” New ideas in psychology.
O Nualláin, Seán (2009) “Zero power and selflessness” Cognitive sciences 4(2)
O Nualláin, Seán (2006) “Inner and outer empiricism in consciousness research” in New ideas in psychology Volume 24, Issue 1, April 2006, Pages 30-40
O Nualláin, Seán and R. Strohman ““Genome and natural language” in Witzany (ed.) Proceedings of Biosemiotics 2006. Helsinki; Umweb (2007) Pp. 249-260
O Nualláin, Seán (2008)““Remarks on the foundations of Biology” at “Cosmos and History: special issue on 'What is life?'” Vol 4 Nos 1-2
Ouspensky, P. (1977) In search of the Miraculous. Orlando, Fa: Harcourt
A. Pellionisz and R. Llinás (1979) CEREBELLAR COORDINATION: COVARIANT ANALYSIS AND CONTRAVARIANT SYNTHESIS VIA METRIC TENSOR. A TENSORIAL APPROACH TO THE GEOMETRY OF BRAIN FUNCTION Neuroscience 4:323,1979
Paul Smolensky, Géraldine Legendre(2011) The Harmonic Mind: From Neural Computation to Optimality-Theoretic GrammarVolume I: Cognitive Architecture. MIT Press.
Zimmer, C (2010) “Sizing Up Consciousness by Its Bits” Ny Times, Sept 20, 2010
On this blog, I announced a set of findings about metabolism and consciousness over the past 2 years. Last week, they finally got published in a peer-reviewed and indexed (etc) journal and I included one of the brilliant students shafted by DCU as an author - and deservedly, for he used his expertise in stats (he is a quant) to explode a myth about neural synchrony in epilepsy
The editing process simply added mistakes from my original document, which I append. They somehow assigned me to UCD, which I remember only as a source of hashish in its bar in the late 1970's; the great thing was that the dope expunged the garbage one was exposed to at the lectures. The one thing I share with Iar-Taoiseach Brian
Cowen is a vivid memory of NUI's Dublin drug emporium.
In any case, the paper with Springer's errors after a 3 month delay is at
http://www.springerlink.com/content/x10063878485504n/
and I append the original
Received 12 July 2011
Consciousness is cheap, even if symbols are expensive;
Metabolism and the brain’s dark energy
Seán O Nualláin
Lecturer, Symbolic Systems, Stanford
Ventura Hall, Stanford Ca 94305 USA
***@stanford.edu
and ***
Abstract
Use of symbols, the key to the biosemiotics field as to many others, required bigger brains which implied a promissory note for greater energy consumption; symbols are obviously expensive. A score years before the current estimate of 18-20% for the human brain’s metabolic demand on the organism, it was known that neural tissue is metabolically dear.
This paper first discusses two evolutionary responses to this demand, on both of which there is some consensus. The first, assigning care of altricial infants with burgeoning brains (and in human infants the metabolic demand peaks at 65% of the total) to “allomothers” is not unique to humans. The second, using relatively small neurons as primates do, risks misfires past a certain minimization. Moreover, in apparent paradox, there is an increasing consensus that large “Von Economo” neurons are critical for communication
This paper’s main original contribution is the discussion of two further evolutionary tricks. The first is the use of self-similarity in the cortex, both in structure and process, to allow the cortex readily - and in energetic terms, parsimoniously - to shift between states in a high-dimensional space. This leads to discussion of the kind of formalism appropriate to model these shifts, a formalism which – it is tentatively suggested – may do double duty for the modeling of symbolic thought.
The second trick is the superimposition on the background “white noise” of neural firing of EEG-detected waves like gamma. The paper describes a method, using the Hilbert transform, of calculating the dips in energy consumption as the brain is transitioned by gamma waves. It is hypothesized that consciousness may be a spandrel, the incidental result of a neurodynamic imperative that the brain enter a maximally sensitive (in sensory terms) “zero power” state a few times a second. If that is the case, then there are obvious benefits for health in meditation, which can be viewed as a state of consciousness extended over time by limiting afferent stimuli.
Keywords
Neurodynamics, thermodynamics, evolution, consciousness, biosemiotics
1 Introduction
We are destined, perhaps doomed, to project models from our current science on the phenomenal space of our minds. Currently, there is immensely exciting analysis going on at UC Berkeley (Freeman et al, 2008)using concepts from chaos theory, fractals, and many-body quantum theory to explain EEG data. This follows an equally exciting 1990's obsession with Bose-Einstein condensates, Renaissance viewsof the mind as a mill, Burton's being satisfied in 17th century “The anatomy of Melancholy” with mere gravity as a dynamo of neural exchange, Descartes and his famous hydraulics, and so on. The UC Berkeley work outlined here stresses the notion of a neural “wave packet, which is discussed below
It is cautiously proposed here that techniques which reveal discontinuities occurring in the order of tenths of seconds may be most fruitful, and recent ECOG work by the author and others will briefly be summarised. Phrenology, by contrast, was and remains an infamous 19th century fad. Bumps for covetousness, licentiousness, knowledge, and so on have a disturbingly medieval resonance, with the witch-hunt for supernumerary nipples to suckle incubi surely only a small distance away, both anatomically and politically, from Gall's crackpot activities. Yet the current obsession in neuroscience with fmri data is not much better. These data manifest statistical assumptions that are normally 10 years' study beyond the knowledge of the functionalist psychologists who presume to interpret them, are obtained in absurdly artificial environments, and - most of all- have a time sensitivity measurable in gross seconds.
Indeed, Vul et al (2009) suggest that many of the results obtained are dubious in their statistical reasoning. Vul et al’s meta-analysis has not been credibly questioned; including the word “voodoo” in the original title to describe the work of other scientists was an unfortunate borrowing from GHW Bush and perhaps diminished the paper’s impact.
It is uncontroversial in the extreme to suggest we need time sensitivity that is one or two orders of magnitude greater than fmri; it does seem to be the case that consciousness can be detected in the tenths of seconds, and that many critical neural events require only hundredths or thousandths of seconds. Specifically, work on microgestures (O Nualláin, 2010) indicates that a facial expression sustained for only 0.04 of a second, well below the sampling rate of consciousness, can affect our evaluation of a person. In turn, the sampling rate of consciousness can be assessed by examining what experimental subjects can actually report; to eschew pseudo-precision, and to anticipate some of the discussion below, it seems to be about a tenth of a second.
EEG and ECOG data (with the latter featuring electrodes attached directly to the cortex), on the other hand, as we have seen, can be sampled at appropriate intervals; indeed, precision well beyond the thousandth of the second level is possible. In our recent work (Freeman et al, 2008; O Nualláin, 2009), we have shown that superimposition of high-frequency gamma waves on the background “dark energy” of the cortex leads to momentary power dips therein, at a rate of several per second. The latter paper postulates that these dips constitute consciousness – perhaps to say “facilitate individual experience of consciousness” may be more precise - which is physically a discrete process and phenomenologically seamless only to a naïve experiencer. These gamma waves are precisely the ones emitted by trained meditators at very high levels.
On the other hand,, issues relating to the trades-off between increased brain size and consequent organismal metabolic demands are among the most fiercely contested in contemporary science ( Isler et al, 2006, 2009 a and b, Isler 2011). Consensus exists on relatively coarse-grained relations, like the positive ones between brain size, the relative altriciality of species, and the degree to which that species tolerates allomothers along with the biological mothers for child-care, a period in which the metabolic demand by the brain is around 65%(Fox, 2011) . Yet even that calculus needs nuance; the necessity of monitoring a much larger body means that a cow with its much larger brain is not necessarily any more “intelligent’ than a rodent. Conversely, humans reflect an evolutionary decision not to increase the size of individual neurons, even though the result may be ion channel misfires.
.
Animals with larger social networks, such as humans, inflict a large metabolic demand on the organism with 18-20% emerging as the consensus figure for humans for this “dark energy”. In this paper, we consider the extent to which brain waves like gamma attenuate the metabolic demand by the brain. It is hypothesized that consciousness itself, as we experience it, may be an artifact of metabolic demand minimization by the brain. While efficient representation algorithms like “sparse coding” free up such energy to some extent, it does seem plausible that synchronized gamma frees energy in excess of normal homeostasis to be used, perhaps, for healing as well as essential maintenance.
We begin with a brief review of the well-travelled evolutionary controversies, but with a twist; information theory predicts that the human brain may have reached certain constraints on the size of individual neurons. Evidence from the Freeman lab suggests that this energy demand may also be attenuated by the brain’s fractal architecture. It may change from one state to another in a high-dimensional space with utmost rapidity. Not incidentally, this fits in with the tensor approach to symbolic brain function pioneered by Smolensky and his colleagues, and we briefly consider the consequences for symbolic function.
The paper goes on to describe the functioning of the Hilbert transform, the data-analysis tool used here. Examples are given of its analysis both of simulated and real data. Then, it is noted that the synchronized gamma that is a signature both of meditation and consciousness has been found to attenuate the brain’s metabolic demand on the organism by a factor of 4. We then discuss the thorny issue of what consciousness actually is. We speculate on some essential maintenance tasks that may become possible due to this access of energy before – in further speculation – remarking that consciousness may after all be a spandrel, the result of a metabolic requirement. Finally, in an appropriately humble spirit, we consider possible consequences for symbolic functioning.
2 Metabolism and evolution
Given that many organisms found stable ecological niches scores of millions of years ago, the quickening of the evolution of symbolic intelligence that homo sapiens sapiens exemplifies with its inevitable metabolic cost requires explanation. One can hardly ask a more critical question for human Biosemiotics. Indeed, it has reasonably been asked whether the human brain is not an optimal vehicle for intelligence; to reach the human level of neurons per body weight, a rodent would have to carry around a 45kg brain, stopping any other salutary process (Fox, 2011) . Conversely , some animals like the koala bear have made an evolutionary decision to keep their brain size extremely small, and indeed reduce metabolic demand even further by sleeping 80% of the time. The route to big brains and symbols is not an inevitable one.
The battleground in evolutionary psychology (Isler et al, 2006, 2009 a and b, Isler 2011) ) is fully as contested as that neural communications theory (ibid. There is consensus in the former field on certain relations, like the positive ones between brain size, the relative altriciality of species, and the degree to which that species tolerates allomothers along with the biological mothers for child-care with marsupials being an exception explicable by a longer nursing phase. ( Isler et al (2006) were unaware of the current 18-20% “dark energy” estimate and simply cite work from a quarter-century before concerning the energetics of neural tissue)
Yet brute brain size does not correlate particularly well with intelligence, as larger mammals need bigger brains to maintain standard homeostatic functions. A better index plots body weight versus brain weight and here humans indeed are outliers with high relative brain size and cattle are outliers with low such.Moreover, the requirement for communicating across a large brain exerts its own discipline. Axonal mass increases faster than gray matter as brain size increases; a strategy that primates seem to have used is to grow with small neurons ( A possible exception involves Von Economo neurons, as noted below). Yet these smaller neurons and axons result in another bottleneck; the axons will become faulty below about 150 nanometers because of consequential misfires in ion channels, just as transistors below 10 nanometers fall prey to faults with the absence or presence of single atoms of boron causing misfires.
Much of the context for production of our experimental work is the well-attested finding that meditators show sustained brain activity in the “Gamma oscillation” category, which goes roughly from 25 to 80 hz (Freeman et al, 2008) with beta approximately from 12.5 up to 25Hz (ibid). Remarkably, there are solid indications that such gamma activity, particularly if synchronized over large areas of the brain, is a signature also of conscious states (O Nuallain, 2009). There is also at least provisional evidence that meditation leads to improved health, and indeed a thickening of cortex (ibid).
The work we did at Walter Freeman's lab at UC Berkeley (Freeman et al, 2008) indicates that there may be a unifying explanation for the beneficial health effects of meditation (or indeed conscious states). Very simply, the brain normally demands 18-20% of the total energy of the human organism (as distinct from 3% in the mouse). Synchronized gamma would briefly attenuate this demand by a factor on 10,000, and do so a few times a second. This frees up energy for many metabolic processes. Put in standard units, an average human consumes 100Watts; of this, the brain consumes 20W. Briefly to attenuate brain demand results in 20W energy being freed up.
It is worth pointing out that there is an assumption here that the extra energy freed up is "usable" for homeostatic functions. For example, the body can autonomously drive up its own heat to fight off infections in fevers. In that case, however, "normal" homeostasis in lost; indeed, the extra energy available to the brain causes hallucinations. This is speculative in our current state of knowledge
Famously, Gary Kasparov using 20w almost beat a computer using several thousand watts (Montague, 2008; of course, a laptop using a few hundred watts could beat most of us while multitasking). The efficiency of human cognition is indeed a marvel; Friston (2010) provides a good summary paper. However, part of the contribution of the present paper is to question and extend some of Friston’s arguments. Confusingly, Friston’s core concept is “free energy”, a term used also in biochemistry, on which he imposes a purely informational meaning. Even granted this latitude, we believe his scenario to evince premature closure, as described below. The brain is far from thermodynamic equilibrium; similarly, it wishes to process high-entropy data and, in order to do so, allows limit cycles. Finally, it is fair to say that Friston inevitably takes his schema into informational territory with remarks about issues like the thermodynamic efficacy of the related notion of “efficient coding” (Friston, 2010, 131)
There is one exception to humans’ preference for smaller neurons; von Economo neurons are inventions of large social animals, and abnormalities in their density in autistic brains relative to controls is attributed to the social deficits. In large social animals the von Economo neurons may regulate that higher social sense required to keep track of up to 150 different individuals in your social group, the Dunbar number (Dunbar, 2003) If that is the case, they are the exception that proves the rule to the generalization that primates use small neurons.
It is speculated in the framework of this paper that consciousness may not be causal. Cognitive resources may be directed to various tasks; conscious moments are replete with retrospective recreations of traces in the brain after these tasks are accomplished. In fact, attempting to become conscious of skilled performance (playing piano, driving) results in a deterioration of that performance. Similarly, establishing and nurturing bonds with the 150 will not necessarily require consciousness.
3 Brain architecture, symbols, and energy consumption
Friston (2010) and Carhart-Harris and Friston (2011) champion two major paradigms in neural processing of information; the “free energy” approach which constrains the new information presented by afferent stimuli to values tolerable in bit terms, and the “default” mode of neural functioning, which they have indeed used to rehabilitate Freud’s psychodynamics. The “default mode” posits brain waves at about 0.1Hz as its signature; “free energy” paradigms depend on a system in thermodynamic equilibrium with its environment. The “default mode” will not be a consideration of ours here, beyond the consensus that the metabolic demand by the brain does not evince any significant upsurge when the brain is “busy” (ie non-default), indicating that symbolic thought seems to exploit an already prepared system.
Our work (Freeman, 2008;Freeman et al, 2008) contrary to Friston’s (2010), proposes that the brain functions far from thermodynamic equilibrium. We suggest that the brain moves in an extremely high-dimensional state space in a manner facilitated by the fact that it is permeated with self-similarity, both spatial and temporal. At the spatial level, self-similarity ensures that, in fractal fashion, each neuron changes in step with the cortex as a whole. In temporal fashion, self-similarity results in the power spectrum of much slower theta waves traversing the brain being identical to, if slower than, much faster gamma waves. Several times a second, the brain enters a limit cycle in which it becomes extremely sensitive to incoming stimuli. Therefore, even without mechanisms like stochastic resonance, we humans can detect stimuli as weak as a few photons, or a few parts per billion of scent. The brain is geared to process novelty as it is to process low entropy information.
Our UC Berkeley work (Freeman et al, 2008) stresses the notion of a neural “wave packet”, which is not to be confused with its quantum mechanics synonym. In essence, a neural wave packet is a trajectory of successive states in the cortex begotten by a stimulus, either incoming or the result of internal processing. The trajectory flows through states which are local basins of attraction; subjectively, the sampling rate of consciousness is such that we consciously apprehend only a very sketchy sample of the perhaps trillions of individual computations that are ongoing.
EEG and ECOG data (with the latter featuring electrodes attached directly to the cortex), can be sampled at appropriate intervals; indeed, precision well beyond the thousandth of the second level is possible. In our recent work (Freeman et al, 2008; O Nualláin, 2009), we have shown that superimposition of high-frequency gamma waves on the background “dark energy” of the cortex leads to momentary power dips therein, at a rate of several per second. The latter paper postulates that these dips constitute consciousness – perhaps to say “facilitate individual experience of consciousness” may be more precise - which is physically a discrete process and phenomenologically seamless only to a naïve experiencer. These gamma waves are precisely the ones emitted by trained meditators at very high levels.
Moreover, it seems to be the case that the meditative state is devoid of cognitive content to a great degree. Finally (O Nualláin, 2009), gamma synchrony, the most attested concomitant of consciousness, may be significant only insofar as gamma is normally localized in specific brain regions; what (experience of) consciousness may consist of, in this scenario, is the result of resonant gamma over the cortex and its effect on the background noise/energy. We describe the methodology used below.
Smolensky et al (2011) suggest that the high-dimensional states that characterize the brain’s transitions need higher-order tensor calculus for their modeling. Smolensky was initially stung into action by the goading of Fodor and others who argued that neural network systems lacked “systematicity”, the capacity to handle certain syntactic relations normal in language. Thus, if “Fionn loves Grainne” is a syntactically valid sentence , syntactically (if not pragmatically) the system should be able to determine that “Grainne loves Fionn” is similarly valid. It was the failure of connectionist systems to do this that motivated Smolensky to develop a tensor theory in which neutrally plausible mechanisms could be adduced for each level of the Chomsky hierarchy.
The Chomsky hierarchy famously describes formal automata of ever-increasing power from finite state, through context-free and context-sensitive automata that are historically associated with natural language and finally to type 0 and type I automata that may or may not be recursively enumerable ( O Nuallain 2003, 2008). Finite state automata can most naturally generate sentences resembling a child's pivot grammar (daddy guess, mummy bad); suitably tortured, they can generate natural language. The consensus for natural language is somewhere between context free and context sensitive and below the type 0 grammars.
Chomskyan linguistics has historically seen its task as the formal characterisation of a grammar (syntax alone) or Grammar (incorporating all of language). No complete grammar for any natural language has yet been written; as for Grammar, the failure in this realm is best attested by the primitive natural language systems on the web, despite massive investment by companies like Google, as this is being written in 2011.
Similarly, the papers by William Hoffman and Ken Aizawa in our 1997 collection (O Nualláin et al 1997) propose that Lie groups and the formally equivalent tensors, should be used in cognitive modeling. Lie groups are the brainchild of the Norwegian Marius Sophius Lie whose work on symmetry extends the earlier opera of his brilliant and doomed compatriot Niels Abel and that of Abel's near contemporary, the similarly doomed Evariste Galois. Both Abel and Galois attacked the issue of why radical solutions to equations involving the quintic were impossible. Cardano and other brilliant Italian mathematicians had incidentally invented complex numbers in their successful search for solutions to equations involving the cubic.
The colourful details of the lives of these brilliant mathematicians need not concern us here, nor should their sometimes impenetrable formalisms. What is important is that Galois worked out that groups of permutation of the roots of an equation had properties that could constrain whether the equation in fact was soluble. Lie systematised Galois’ insight in a general theory of groups that turned out to provide a lingua franca with differentiable, locally Euclidean "manifolds" in space.Lie groups dominate current cutting edge thought in quantum theory, a further example of John Wheelers "unreasonable effectiveness of mathematics in the natural sciences"
Tensors, by contrast refer to a meta-formalism in which scalars (like the numbers “one, two”) are tensors of rank zero, vectors are tensors of rank one, and so on. Higher order tensors are possibly necessary for modelling brains function in the high dimensional space opened up by Freeman's neurodynamics. In any case, tensor calculus allows elegant modelling of the "cued reaching movements" described in Friston(2010, 134) . In particular, a single point in four dimensions must be targeted by a nervous system working in a much higher-dimensional space to catch a ball. Tensors, and thus Lie groups, afford a parsimonious way of approaching this problem.
Kime’s article in the same collection (O Nuallain et al, 1997) goes a few stages further - in what now seems a rather prescient move, he argues that the whole enterprise of cognitive science is inappropriately rooted in a belief that the data it deals with are the product of a linear system. Kime argues, following the seminal work of Pellionisz and Llinas (1979), that the space is in fact curves to varying degrees.
While Kime, following Pellionisz and Llinas (ibid), uses their contravariant/covariant terminology, the point is still the same; we may need a completely different paradigm to handle human use of symbols. In fact, the lacunae thrown up by the HGP may similarly thus be explicable. It may be the case that symbolic functioning proceeds over a high-dimensional state-space, with curvature actually changing as the interpretation of the symbol proceeds (O Nualláin et al 2007, p. 257) .
This hypothesis is tentative in the extreme; it suggests that symbolic functioning involves separate phases of “analysis” and “interpretation”. In the former phase, the organism finds out what the general context actually is. Interpretation involves a gradual restriction of this context and a change in the curvature of the system ( O Nualláin 2003, O Nualláin et al 2007). A good analogy is the warping of space-time by a massive object.
What this section has (very) tentatively proposed, therefore, is a different paradigm of brain function due to the attested work of Freeman and his colleagues. It implies a tensorial (or another mathematically equivalent) approach to cognitive function. It eschews Friston’s (2010) “free energy” paradigm as incomplete. Next, let us look at an attested mechanism for reducing metabolic demand.
4 The Hilbert Transform and dips in metabolic demand
An application of the Hilbert transform to both real and simulated ECOG data is described here. Readers with an aversion to technical detail can safely skip to 4.2
The Hilbert transform overcomes limitations in the Fourier Transform partly by repeatedly applying it to a changing signal. The Hilbert transform calculates attributes of a time series, especially the amplitude and instantaneous phase and frequency, before summarizing all of them in an “analytic signal”. Now we can calculate crucial attributes of the cortical activity, both real and simulated, like power. The investigation here confirmed the existence of “lulls” several times a second in the background activity of the brain which may have a central role in sensory processing.
The analytic signal is useful in the area of communications, particularly in bandpass signal processing, where it can be utilised for the minimisation of “aliasing” or “jagging” in graphics applications. The analytic signal has a real part corresponding to the original data, and an imaginary part which contains the Hilbert transform. The imaginary part is a version of the original real sequence with a 90° phase shift. Sines are therefore transformed to cosines and vice versa.
As described in any reliable source like the matlab manual, for a pure sinusoid, the instantaneous amplitude and frequency are constant. The instantaneous phase, however, is a sawtooth, reflecting the way in which the local phase angle varies linearly over a single cycle..
Briefly, we can write
a = Hilbert (b)
where the real part of a is the original real data and the imaginary part is the actual Hilbert transform. a is sometimes called the analytic signal, in reference to the continuous-time analytic signal.
The Hilbert transformed series has the same amplitude and frequency content as the original real data and includes phase information that depends on the phase of the original data. The instantaneous amplitude is the amplitude of the complex Hilbert transform; the instantaneous frequency is the time rate of change of the instantaneous phase angle. What is most critical for the purposes here is that with the Hilbert transform we can calculate the power of all activity in the brain, instant to instant. The reader is referred to Walter Freeman’s (2007) description of the Hilbert in Scholarpedia for further detail
4-1 Methodology
The plan, then, was first to simulate bandpass-filtered brown noise by filtering pseudo-white noise with a filter equivalent to running-summing followed by filtration with a finite-impulse-response bandpass filter such as can be generated in Matlab by fir1. Judicious choice of the order of the filter depends on the digitising step; with our step of 2ms, we found that 500 was an appropriate order, as numerical errors emerged with other choices.
We now choose bottom end and top end of the band of frequencies to be retained by the bandpass filter, expressing these as fractions of the Nyquist frequency. Essentially, the band at 1/80 to 1/40 of the Nyquist frequency was considered theta range; at 1/40 to 1/20 of the Nyquist frequency, alpha range; at 1/20 to 1/10 of the Nyquist frequency, beta range; and finally, at 1/10 to 1/5 of the Nyquist frequency, gamma range.
In each case we proceeded as follows;
1. We computed the corresponding analytic signal by taking the Hilbert transform of the filtered brown noise.
2. We computed the corresponding analytic amplitude (pointwise modulus), analytic phase (pointwise arg), analytic power (pointwise squared modulus), "unwrapped" analytic phase, and analytic frequency (the differenced analytic-phase series divided by the "digitizing step"). We then proceeded to plot analytic power against time, and analytic frequency against time.
For each of the bands we produced an example of the bandpass—filtered brown noise, a longitudinal histogram of the bandpass-filtered brown noise , a corresponding example of the analytic power series, a longitudinal log-log histogram of the analytic power, a corresponding example of the analytic frequency series, and a longitudinal log-log histogram of the analytic frequency. (Naturally, most of these graphs are surplus to the requirements here and only some are shown.) The bandpass simulates local inhibitory feedback; as we hypothesize, a short delay and feedback corresponds to gamma and long such to beta.
In the second stage, we used real data, as described below.
4-2 Results
In our work, the first diagram (fig 1) shows power consumption decreases with gamma first with simulated data, then with real data (fig 2), both of which have appeared in various forms in our previous work. The simulated data looked on neuron firing as random, as white noise with refractory periods. (Without refractory periods, epileptic seizure sets in, with the very clear and tragic consumption of metabolic energy that involves.) We then used the Hilbert transform to calculate how power consumption would vary if the “white noise” signal was convolved with gamma waves.
Several aspects of the results are discussed in O Nualláin (2009). Many immediately striking findings emerge from this use of the Hilbert transform. In the first place, we seem to have “lulls” in the power spectrum of the brain under gamma a few times per second (Freeman historically has argued for 3-7 times per second). In these lulls, the power consumption of the brain decreases by a factor of ten thousand. Let us look first of all at an idealization, in which the power reaches zero in a simulated case;
Figure 1 simulated data
The above is simulated data, derived as described above. We can compare the real data, which also shows “null spikes” occurring 3-7 times a second. Fig 2 uses data acquired, with her consent, from an 8*8 electrode array applied directly to the inferior temporal lobe of a woman undergoing surgery for epilepsy (Incidentally, we also found that epileptic seizures – contra the official wisdom – do NOT involve any special correlation of activity between discrete brain regions). In this case, the “white noise” is actual neural firing as clinically. recorded and the remaining steps to calculate to calculate power are as described above; let us note that energy consumption is power*time, so our power dips attenuate the energy consumption.
Fig 2 real data
As we pointed out (Freeman et al, 2008), the “null spikes” correspond to extremes in the instantaneous frequency, as depicted in figure 3;
Figure 3
Moreover, the power distribution shows that power consumption decreases by an order of 4 during the “null spikes”;
Fig 4
4-3 Seizure and energy
Epileptic seizure is the limit case, wherein the refractory periods in which neurons do not fire are abolished, and the brain’s demand on the organism becomes maximal – and tragically so. Moreover, in our view (Freeman et al, 2008), gamma and beta waves are caused by feedback from inhibitory neurons, and these obviously are drowned out in the case of seizure.
While there does seem to be a marked qualitative difference – and it is worth noting that all of our samples of epileptic patients show this monotonic increase during seizure – it will obviously be interesting to tease these data out more. In any case, it is clear that seizure is not an evolutionarily stable strategy!
So, for example, we can apply the Hilbert transform to the seizure data, and get a sense that gamma would provoke a massive reduction in the power consumption of the brain
This would rather seem to suggest that the absence of gamma and beta is one of the reasons that the brain's consumption of energy becomes so massive in the case of seizure. Refractory periods stabilize the gross activity occurring in the cortices; both these and the salutary effects of inhibition, which incidentally produce both gamma and beta, are now missing
Let it be made clear that the energy that we propose is saved from beta/gamma superposition is many orders of magnitude greater than that from more economical representations like “sparse” coding. If it is correct to interpret our original published work (Freeman et al, 2008) in the evolutionary context, as proposed here, this is an unprecedented energetic difference.
5 And so, what is Consciousness?
It is not controversial to suggest that there is no consensual definition of consciousness. By way of an anti-definition; consciousness reflects the operation of global dissemination and integration of information in the brain. It seems to be the case that it facilitates certain cognitive acts, like processing irregular past tenses; yet these can perhaps be just as easily explained by attention, without any phenomenology
We tend to view ourselves as initiating action having previously decided on it; however, the available data suggests otherwiseIn particular, classic work by Libet ( O Nuallain 2003)suggests that motor neurons show preparation to initiate action tenths of seconds before we are aware that we intend to do so. It is not a stretch to propose that we continually narrate to ourselves a “stream of consciousness” about our actions that is little more than what a sports commentator does about a game he is calling
If we accept that much of our conscious experience is simply an a posteriori commentary, biased to predicate virtue, consistency and potency of ourselves, none of this is surprising. So does consciousness do anything at all? The thesis here; perhaps consciousness does nothing, but is a handmaiden to attention. Alternatively put, we will ourselves to do things, and as an epiphenomenal result become aware of these actions late
I have covered these themes at some length in my books ; “The search for Mind” (1995, 2003), “Being human “ (2002, 2004), “Two sciences of mind” (1997), various papers (Eg 1994) and am going to deal only with two at any length here. The others I respect include L-dice, Global Workspace models, and Damasio's work.
Consciousness has a phenomenology in that humans use consciousness scan internal "space" in the same manner as they scan external objects and relations. In my current work (forthcoming) I point out that as a cognitive science object of study, consciousness is likely to remain at sensorimotor level. Of course as a political phenomenon it is multifarious. Those of us in the West have been encouraged some centuries to view consciousness is everything we are, disregarding the fact that the cells we hold so inviolate are partly the result of social construction as of Manufacturing consent.
The degree to which we view our individual consciousness as important varies between political systems, and with the amount of social control that obtains; the speculations are of course beyond the scope of this paper .( In previous work (1994) I discussed this kind of issue in detail ). Similarly I have discussed elsewhere (2006) the "experiential disciplines" which seek to reroute the "stream of consciousness" from Joycean machine, with an often self serving narrative, to being beads on a string with some wholeness
There are numerous phenomena, both in external world and interior space, identified with consciousness. Consciousness has been implicated by researchers like Stapp and Von Neumann in the act of observation in QM, absent some “objective reduction” principle (O Nualláin 2004). It has also been identified by researchers like Hameroff and Penrose as supervening on Bose-Einstein condensates somehow exteding over decimeters in the brain’s febrile environment. (ibid). Quantum coherence in biological systems is indeed being observed, but at a minute fraction of the scale required by Hameroff and Penrose.Yet, there is clear conceptual compatibiilty between gamma synhrony, so-called” zero power” and the desideratum of Quantum coherence, to be explored in a forum other than “Biosemiotics”.
There is also a seductive neo-Hegelian story that sees consciousness as ultimate reality, and the progress of human society as reflecting a path in which consciousness will repossess itself after the supreme act of alienation that comprised creation of the cosmos. Tweak this story to stress economic activity, and revolutionary Utopian socialism emerges as a consequence.
In internal space, the word “consciousness” has sometimes been reserved for moments in which the self can be identified as the locus of observation. This is the beginning restriction for thinkers such as Gurdjieff ; to do so requires prior work in acting authentically in the world, and self-integration. (Ouspensky, 1977; in my 2010 paper for this journal, I have attempted to bring this work into dialogue with contemporary science) Confusingly, Gurdjieff and his ilk identify the human search as being one for “consciousness”. In fact, it is likely that what the neophytes are looking for resembles rather a state of being in which they have command of all their faculties, including professional skills and abilities, and can function in a stable way within the society in a manner that does no violence to their reason. Neither Gurdjieff nor his later followers went even close to providing that kind of nexus. That said, it is indeed possible that the current configuration of our society, with tax being redirected into financial speculation (as distinct from deserving causes) that self-destructs regularly, and the increasingly violent encroachment of many of those self-same interests on both academia and the arts, is not optimal for human development
Conversely, consciousness has been misidentified as mere waking (versus sleep), as attention, and intentionality, the fact that it seems “pointed at' an object in the external world. . Researchers like Jason Brown (O Nualláin, 2003) have correctly pointed out the intricate links with selfhood and will, identifying will as a heightened sense of self with respect to an action (or inaction).
The neo-Piagetian view sees consciousness as constructed from many incidents of subject-object differentiation, a view that I made explicit in this journal (2008,2010), Pace John Ralston Saul and fulminations about “The unconscious civilization”, it can equally be argued that civilization involves the making unconscious/automatic of many processes, the better to indwell on what is important to us, to accord real decisions the greatest number of degrees of freedom as possible while needs lower in Maslow's hierarchy are offshored to the state and civil society. Similarly we are not aware (nor should we be), moment to moment, of how the sewage systems run or – until recently – how whole countries are financed.
Likewise, for each individual, we note skills like cycling a bike becoming unconscious/automatic, and then indwell on avoiding boy racers on the road. What we normally experience as “self” is a dynamic shifting of the subject/object boundary; achieving some stability here results in the person having the option of identifying self with the locus of awareness.
Moreover, it seems to be the case that the meditative state is devoid of cognitive content to a great degree. Finally (O Nualláin, 2009), gamma synchrony, the most attested concomitant of consciousness, may be significant only insofar as gamma is normally localized in specific brain regions; what (experience of) consciousness may consist of, in this scenario, is the result of resonant gamma over the cortex and its effect on the background noise/energy.
Of recent theories, Logie and Pearson's work in my 2000 collection describes consciousness as a mode of access to working memory. Tonoi's recent work(Zimmer, 2010), which sees consciousness as an epiphenomenon of ordered information, has the advantage of being wrong;
"He began to think of consciousness in a different way, as a particularly rich form of information...Consciousness is not simply about quantity of information, he says. Simply combining a lot of photodiodes is not enough to create human consciousness. ...Dr. Tononi argues that his Integrated Information Theory sidesteps a lot of the problems that previous models of consciousness have faced. It neatly explains, for example, why epileptic seizures cause unconsciousness. A seizure forces many neurons to turn on and off together. Their synchrony reduces the number of possible states the brain can be in, lowering its phi. ..."
Finally
“Dr. Tononi sees serious problems in these models. When people lose consciousness from epileptic seizures for instance, their brain waves become more synchronized. If synchronization were the key to consciousness, you would expect the seizures to make people hyperconscious instead of unconscious, he said. “
Jouny et al (2010) would suggest that this is premature closure, with an INCREASE in signal complexity - that is, decline in synchrony - associated with seizure. We (Doris et al, in preparation) looked at ECOG data from the inferior temporal lobe of a patient in the states of sleep, seizure and awake (resting brain) using principal component analysis (PCA) and dynamic time warping (DTW) and found that our results support Jouny, rather than Tononi;
Sleep signal is least complex/disordered under PCA, first component explains 97%, awake is next, with 93% explained by the first component, while seizure has just 63% explained by first component.
This is supported by the DTW result, which indicates that the mean distance is less in sleep than waking than seizure;
> mean(seizure.dtw.mat)
[1] 6.731411
> mean(awake.dtw.mat)
[1] 5.580975
> mean(sleep.dtw.mat)
[1] 4.204780
The PCA for sleep shows that there is effectively only one "signal" across the 64 channels. You can easily construct a linear model (through simple ordinary least squares –ols - regression) of any of the individual channels based on a small number ( < 10) of the other channels, and by doing this you will recover 99% of the channel. Awake is similar but less extreme, there is one dominant signal but there is some differentiation, while seizure still has a dominant signal the PCA shows it is much less influential and there are several other signals present in the telemetry. In info-theory terms, sleep has minimum entropy, awake next, and seizure has highest entropy by a long way. In short, if there is an “Integrated Information Theory” theory of consciousness, it already has one noticeable lacuna. In previous work (2004) I have discussed shortcomings in any informational theory of consciousness, in that they fail to pre-empt the spectre of conscious thermostats – and indeed rocks, to adapt Chalmers (ibid).
The viewpoint here, therefore, is that consciousness manifests itself when the cortex acts under the influence of gamma synchrony. At this point, phenomenology is possible, yet consciousness itself is non-causal in immediate terms. The narratives we weave as a result of the ability to reflect on our experience that consciousness affords us are indeed causal, perhaps the most causal factors in evolutionary history, for good and ill.
6 Comments on the biochemistry of metabolism
Energy released for use for the rest of the organism by states of meditation/consciousness can potentially be used for the following purposes as distinct from the obvious growth, and standard maintenance issues. This requires some preliminary discussion, some of which inevitably – given the immature state of the field – will be speculative
Energy production and consumption each take place on multiple time scales. Energy may be conserved in the form of useful energy reservoirs. Cyclical patterns for consuming energy take place at multiple time scales. Glucose is one store convertible on the time scale of seconds; thru the glycolysis cycle it makes ATP; ATP is the main (but not exclusive) energy currency of the cell on a short time scale and thus is very quickly available to build gradients by driving proton and other motors across membranes and to drive other reactions and motors that do other work, like moving cargo along an axon of a neuron along the microtubules; proton gradients are the ultimate, proximate store instantaneously available like battery charge for conducting nervous energy, for sustaining the charge balance across membrane by moving protons and metal ions particular in the energy conversion factories, the mitochondria.
Astrocytes should be interesting since their calcium waves may be related in periodicity, obviously critical here, and their attested role in neural metabolism. Indeed, to get very speculative indeed, the zero energy points on the Hilbert transformed data may be seen as instants that reveal the collapse of a nonlocal wave function; this could appear in spectrally coupled fields of astrocytes and neurons each set coherently coupled thru their respective gap junction networks.
To summarize, then, here are some effects beyong the glaringly obvious that the extra energy freed up by gamma waves, and incidentally facilitating consciousness, could be used for in the organism;
1. Getting glucose into cells, lessening the need for insulin. Insulin is only a signaling/gating molecule, it doesn’t have inherent toxicity except in excess relative to its native receptors when it starts working too much on it night jobs (cotransporting drugs, metabolites, and so on, and this is what we are trying to avoid). Dogs with their pancreases removed – if fit – do not become diabetic. Thus, conserving energy by managing an appropriate physiological state can reduce the need for insulin.
2. NAD+ and NADH are used as intermediates in producing ATP and the energy saved by gamma may easily be stored in these forms. They contribute to lots of other reactions too, like those involved in biosynthetic machinery making proteins that may or may not be involved in gene expression. They have a role in regulating post translational modification of histones, a process that is intimate with the management of gene expression and of great interest in the study of aging.
3. Gene expression varies with a lot of things. In changing environments it’s primarily controlled by nuclear and other receptors that sense the internal and external environmental changes of the cell by sensing molecules, temperature, etc. In particular, gene expression varies with the ratio of NAD+ to NADH which informs of the metabolic state and health of the cells. This will vary with the energy available .
4. Finally, there is preliminary evidence from the Blascovich lab that conscious experience focussed on an external target can dull pain perception - and indeed increase blood flow
7 Conclusion
The current debate on brain size and evolution is in some ways misguided. Sheer brain size does not guarantee intelligence. It is equally important to have fast communication in the cortex. However, that in turn leads to a set of trades-off between size of neurons, and indeed width of axons
Remarkably, it may be the case that consciousness is an opportunistic consequence of the necessity to curtail energy use by the human brain. The synchronized gamma associated with consciousness decreases by an order of four (if briefly) the metabolic demand by the brain on the organism. Such energy can be used for a variety of essential healing and maintenance functions. In this scenario, meditation is viewed as an extended, content-free, state of consciousness. At the other extreme, epileptic seizure involves massive consumption of energy without gamma to stabilize matters
Humility becomes us at this point. While I have in the past (O Nualláin 2008) proposed a tutorial in biosyntax, it is possible that our models are hopelessly naïve. It is as well to reiterate the ambiguity of the term “free energy” as we head to closure. Informationally, Friston and his school use it to privilege – uncontroversially – an account of neural functioning that stresses that the brain is a machine that works on hypotheses. The argument here, however, contends that this school fails to take into account the fact that the brain is a far-from-equilibrium system, both informationally and thermodynamically.
The dynamical systems notion of the brain here sees it as both protean and deeply self-similar. This led naturally to a discussion about the sort of models appropriate to encompass the symbolic functioning of the brain. It does seem to be the case that tensors and the formally equivalent Lie groups will soon be part of the neuroscientists’ diet. It may even be expedient methodologically to dispense with consciousness outside the social sciences – apart from the sensorimotor area that seems to be yielding its secrets to a cognitive science approach.
Acknowledgements
Many thanks to Marcello Barbieri for his encouragement, and to Tom Doris and Hal Cox for fruitful discussions
References
Carhart-Harris RL, Friston KJ. (2010) “The default-mode, ego-functions and free-energy: a neurobiological account of Freudian ideas.” Brain. 2010 Apr;133(Pt 4):1265-83. Epub 2010 Feb 28.\
Doris T and S. O'Nuallain (in preparation) “Seizure and complexity”
Dunbar R.I.M. The social brain: mind, language, and society in evolutionary perspective. Ann. Rev. Anthropol. 2003;32:163–181. doi:10.1146/annurev.anthro.32.061002.093158
Fox, D. (2011) “The limits of intelligence” Scientific American, July 2011, Pp 36-43
Freeman (2007)Hilbert transform for brain waves
Scholarpedia, 2(1):1338.
Freeman, W., S. O'Nuallain and J Rodriguez(2008) "Simulating cortical background electrocortigram at rest with filtered noise" Journal of integrated neuroscience,7 (3 )Page: 337 - 344 Sept 2008
Freeman, W (2008) “A pseudo-equilibrium thermodynamic model of information processing in nonlinear brain dynamics” Neural Networks
Volume 21, Issues 2-3, March-April 2008, Pages 257-265
Friston, K. (2010) “The free-energy principle”Nat Rev Neurosci. 2010 Feb;11(2):127-38. Epub 2010 Jan 13.
Gazzaniga, M. (1995) Consciousness and the cerebral hemispheres. In Gazzaniga, M. (ed.) (1995) The cognitive neurosciences Pp 1391-1400 Cambridge, Mass; MIT Press
Krishnamurti, J. (1979) Meditations Boston: Shambala
Isler, K. (2011) Energetic trade-offs between brain size and offspring production: marsupials confirm a general mammalian pattern. BioEssays 33(3). DOI 10.1002/bies.201000123.
Isler, K and van Schaik C.P (2006), “.Metabolic costs of brain size evolution”
Biol Lett. 2006 December 22; 2(4): 557–560.
Isler K, van Schaik CP.( 2009a) “The Expensive Brain: a framework for explaining evolutionary changes in brain size.” J Hum Evol. 2009 Oct;57(4):392-400. Epub 2009 Sep 3.
Isler, K., van Schaik, C.P. (2009s) Why are there so few smart mammals (but so many smart birds)? Biology Letters 5:125-129.
Jouny CC, Bergey GK, Franaszczuk PJ. (2010) “Partial seizures are associated with early increases in signal complexity.” Clin Neurophysiol. 2010 Jan;121(1):7-13. Epub 2009 Nov 11.
Montague R (2008) Your Brain Is (Almost) Perfect, NY;Perseus
Ó Nualláin Seán (1994): Some Consequences of Current Scientific Treatments of
Consciousness and Selfhood. AI Soc. 8(4): 305-314
O Nualláin, Seán (2010) “Ask not what you can do for yourself: Cartesian chaos, neural dynamics, and immunological cognition” Biosemiotics Vol 3 Issue 1, 2010 DOI 10.1007/s12304-009-9070-4O Nualláin, Seán (2003) "The Search for Mind" ( Third edition); Intellect
O Nualláin, Seán (2004) "Being Human: the Search for Order" (Intellect, 2004) second edition
O Nualláin, Seán (2000) (ed)"Spatial Cognition” (ed.) (Benjamins);
O Nualláin, Seán (1997) (ed)"Two Sciences of Mind" (Benjamins,);
O Nualláin, Seán (2008)“Subjects and Objects” Biosemiotics journal, Volume 2, Pp. 239-251
O Nualláin, Seán (forthcoming ) “Neural correlates of consciousness of what?” New ideas in psychology.
O Nualláin, Seán (2009) “Zero power and selflessness” Cognitive sciences 4(2)
O Nualláin, Seán (2006) “Inner and outer empiricism in consciousness research” in New ideas in psychology Volume 24, Issue 1, April 2006, Pages 30-40
O Nualláin, Seán and R. Strohman ““Genome and natural language” in Witzany (ed.) Proceedings of Biosemiotics 2006. Helsinki; Umweb (2007) Pp. 249-260
O Nualláin, Seán (2008)““Remarks on the foundations of Biology” at “Cosmos and History: special issue on 'What is life?'” Vol 4 Nos 1-2
Ouspensky, P. (1977) In search of the Miraculous. Orlando, Fa: Harcourt
A. Pellionisz and R. Llinás (1979) CEREBELLAR COORDINATION: COVARIANT ANALYSIS AND CONTRAVARIANT SYNTHESIS VIA METRIC TENSOR. A TENSORIAL APPROACH TO THE GEOMETRY OF BRAIN FUNCTION Neuroscience 4:323,1979
Paul Smolensky, Géraldine Legendre(2011) The Harmonic Mind: From Neural Computation to Optimality-Theoretic GrammarVolume I: Cognitive Architecture. MIT Press.
Zimmer, C (2010) “Sizing Up Consciousness by Its Bits” Ny Times, Sept 20, 2010
Sunday, October 23, 2011
The real cost of these filthy criminals
Today, “60 minutes” ran a feature on a tool for autistic kids that
facilitates communication. This is ethically impeccable. Though the
mantra “Ipad” indicated that some surreptitious branding was going on –
particularly as the rest of the show was an interview with Isaacson
about his Steve Jobs bio. Of course, the technology can be implemented
on any touch-sensitive pad;
http://www.cbsnews.com/video/watch/?id=7385686n
In fact, I used the entire inheritance (10k) I got from my father to set up a lab to do this technology. It was impossible to do at DCU; Clare X, the enablement officer there told me that alan smeaton, the outgoing head of computing there, had embezzled 10k from her budget to set up a wireless network that had no relation to enablement. In fact, all our research at DCU was stolen in any case, with research funds also being diverted, and the gardai at Whitehall refusing to investigate
The real cost to the Irish people of the filthy criminals operative since 1997 at DCU and elsewhere is clearly in the hundreds of billions. I hired Declan Kelly, who had been refused a postgrad at DCU, and another spurned programmer and continued our previous Stanford links; the scum at DCU would have cut them off in a day. I paid Declan from my own pocket to work on this project both during summer 2001 and later at a lab set up in my home county of Clare. Eventually., AARTI came through with some funds to pay for a months salary for him
Here is the Sunday Tribune's contemperaneous report (2001), which we sent to DCU president's office to be received with the usual bovine silence;
http://tribune.maithu.com/archive/article/2001/nov/04/new-hand-held-hope-for-autism-sufferers/
I left a copy of the software on a computer in Ireland when I left in 2003.
Apart from the money that could have been made for Irish people, and the good reputation we would have gotten for the country, many Irish parents of autistic kids would have had enormous respite since 2001 but for the filth who took over a “University” in northside Dublin. Eventually, the first installation was in Inagh, Co Clare; we then ran out of funds, as the funds were otherwise occupied (wireless networks etc)
Forgive me and my employees here in the SF area of the USA if we refuse to touch anything emerging from the dark maw of the so-called “Irish state” which certainly was a mafia state for well over a decade. That includes their pathetic consulate and - get this - "Innovation center" in the Bay Area. We hate this filth wandering around what is a healthy and beautiful culture
Seán O Nualláin Ph.D. Stanford
23u Deireadh fomhair 2011
http://www.cbsnews.com/video/watch/?id=7385686n
In fact, I used the entire inheritance (10k) I got from my father to set up a lab to do this technology. It was impossible to do at DCU; Clare X, the enablement officer there told me that alan smeaton, the outgoing head of computing there, had embezzled 10k from her budget to set up a wireless network that had no relation to enablement. In fact, all our research at DCU was stolen in any case, with research funds also being diverted, and the gardai at Whitehall refusing to investigate
The real cost to the Irish people of the filthy criminals operative since 1997 at DCU and elsewhere is clearly in the hundreds of billions. I hired Declan Kelly, who had been refused a postgrad at DCU, and another spurned programmer and continued our previous Stanford links; the scum at DCU would have cut them off in a day. I paid Declan from my own pocket to work on this project both during summer 2001 and later at a lab set up in my home county of Clare. Eventually., AARTI came through with some funds to pay for a months salary for him
Here is the Sunday Tribune's contemperaneous report (2001), which we sent to DCU president's office to be received with the usual bovine silence;
http://tribune.maithu.com/archive/article/2001/nov/04/new-hand-held-hope-for-autism-sufferers/
I left a copy of the software on a computer in Ireland when I left in 2003.
Apart from the money that could have been made for Irish people, and the good reputation we would have gotten for the country, many Irish parents of autistic kids would have had enormous respite since 2001 but for the filth who took over a “University” in northside Dublin. Eventually, the first installation was in Inagh, Co Clare; we then ran out of funds, as the funds were otherwise occupied (wireless networks etc)
Forgive me and my employees here in the SF area of the USA if we refuse to touch anything emerging from the dark maw of the so-called “Irish state” which certainly was a mafia state for well over a decade. That includes their pathetic consulate and - get this - "Innovation center" in the Bay Area. We hate this filth wandering around what is a healthy and beautiful culture
Seán O Nualláin Ph.D. Stanford
23u Deireadh fomhair 2011
Saturday, October 8, 2011
In the end, there may be very few universities
There are now 130,000 + students enrolled for this course, making its enrollment about an order of magnitude higher than Stanford's entire resident student body;
http://www.nytimes.com/2011/10/03/opinion/the-university-of-wherever.html?_r=1
The issue, of course arises; why hire very expensive and often poorly-trained management and teaching staff at taxpayers' expense to provide an inferior version of such courses on expensive city real estate, particularly as much of the core content of many such courses has been available under an irrevocable creative commons license for some time, due to MIT's open courseware initiative;
http://ocw.mit.edu/index.htm
Alternatively put, it has been clear for some time that countries like Ireland may not need banks, given the (admittedly morally ambivalent) existence of as Euro zone; that a national airline may similarly be an ego trip; why have universities? There are positive arguments in the context of”student moral development” and “national spirit”; but these are precisely what successive governments have explained to us, at great cost, that Irish universities no longer do;
http://9thlevel.ie/2011/07/21/dangerous-legal-vacuum-at-the-irish-universities/
I have few, if any, illusions about the significance of this blog, or indeed of the very costly struggle that both Paul Cahill and I underwent to ensure that some form of procedure other than summary dismissal obtained at Irish universities – and by extension in the public service in general. One would expect trash like the Sunday Independent and the Daily Mail to join the war against academic freedom on behalf of their corporate masters. What was surprising was that the "paper of record", the Irish Times, joined with great gusto in the fray. All are being sued; the Irish Times has already printed several retractions and one apology, and the Sunday Independent has yet again altered its website's record of events - I had a similar experience with them in 2003, resulting in them removing a webpage. That is not enough.
It also is as well to confirm that IBEC, acting for DCU in 2003 on foot of their appeal to the Fleming judgement, did indeed use a little-known provision in a 1990 IR act that forbade strikes for “single dismissal” and threatened SIPTU with a High Court injunction. SIPTU, of course, then went over to management's side in an egregious act that ended with them losing their “closed shop” status at DCU
What I have learned is that, absent an alternative narrative and claim to sovereignty and jurisdiction such as that provided by the republican movement, the Irish state made it clear that it was perfectly happy to operate in a lawless – indeed in a downright antinomial environment. Who better to call the shots in such an environment than a scion of a WW2 forced labour enterprise?
Time was when we used to sit down as Irish people and work out our differences. The major significance of the past decade is that we have flushed out a certain segment of Irish society, who now are administering a bankrupt state for a colonial body. In the meantime, I am open to privatization of the universities (with the exception of TCD) precisely to bring them under the rule of law. There is no way that a private enterprise would have gotten away with what DCU, among others, has gotten away with over the past decade
I am fully aware that the putative purchasers might decide to shrink the campus size 98%, eventually turning them into server farms with a skeleton administrative staff. It is possible that there will be no need for universities in Ireland at all, and all courses will be run remotely from a few sites in the USA and China. Yet it is clear that, as previously happened in health, the Irish state has proved unable to achieve even the moral stature - let alone technical ability - of the catholic church, against which it now rails so petulantly, in its running of key institutions.
Seán O Nualláin Ph.D. Stanford
8u Deireadh fomhair 2011
