Chapter 11

UCLA (1993-2007), Memory and Sleep

The aspect of my avian studies that were of the greatest significance, and provided the foundations for my subsequent studies in avian evolution, memory, and the evolution and basic function of sleep, was the finding of relict reproductive behaviors that probably date back to the earliest phases of surface nesting by the small reptilian ancestors of birds. The implications of these findings for memory and sleep emerged only gradually.

Presumptive Relict Reproductive Behavior in Small Parrots

The neural basis for behaviors that were so conservative that they appeared to trace back to cold-blooded ancestors was dealt with adequately in my book on avian behavior and evolution. But, at best, the findings would reach a very limited audience. For that reason I sought to publish a paper summarizing and accounting for the discovered avian behavioral relics. The journal in which this paper was published was essentially determined for me by the course of events after the book appeared. Thus, I had sent a complimentary copy to Professor Theodore H. Bullock at UC San Diego. I had attended Ted's class in Invertebrate Zoology when I was graduate student at UCLA in the late 1940s. He also was one of my main supporters when I was a junior faculty member at UCLA in the late 1950s.

One day shortly after the book appeared, I received a phone call order for the book from the secretary (and wife) of Professor R. Glen Northcutt, also at UC San Diego. Prof. Northcutt had seen the book on Ted's desk and examined it. He decided forthwith, perhaps recommended by Ted, to order it. As it happened, Prof. Northcutt was the editor of the journal, Brain, Behavior and Evolution, a Karger publication. For that reason I submitted my first paper on relict behavior and physiology, "Presumptive relict reproductive behavior in small parrots," to that Journal. Although, I have not located my journal correspondence concerning this paper, suffice it to say that it was accepted for publication, appearing in late 1988 (32:340-352), a year following the appearance of my book.

I summarize the topics of this paper as follows. In the Introduction, I called attention to some known structural relics or "atavisms" and to some abnormal developmental features that lead to them. There followed the mention that evolutionary and behavioral scientists express little optimism that past behaviors will become accessible, quoting Beer's well known expression that, "unfortunately behavior doesn't fossilize." In a following section I called attention to relevant superficial reflexes of man, such as the cutaneous plantar reflex and cremaster reflex, thought to be vestiges of ancient responses, in many cases now held in check by inhibitory influences.

Also mentioned were presumptive relict behaviors in other animals. These matters were followed by a section on the evolutionary conservativeness of neural circuitry and behavior, concluding that much of the central neural circuitry that was used to program behaviors in the ancient ancestors of some vertebrate lineages still may be present in the CNS of modern representatives. The behaviors still may be readily accessible or may be held in check in various ways, with the possibilities enumerated, stressing roles of inhibition, vestigiality and complex neural control.

The paper then proceeds to consider the many relict reproductive behaviors I observed normally, but mostly elicited experimentally, in small parrots, and the proposed bases for their preservation. My findings suggested that the observed sequence of increasingly attentive egg-care responses retraces certain aspects of egg-care during ancestral stages owing, again, to the highly conservative evolution of the CNS. I then proceed to discuss stages in the development and atresia of eggs and follicles in many vertebrates, and how these also may recapitulate ancestral stages and clutch sizes. Again, there is a conservative neural basis for retracing these ancestral ovarian features.

I concluded that studies of relict reproductive responses in 'pure' species and hybrids, and of ovarian development and function, appear to provide very promising avenues for the reconstruction of reproductive behaviors at various ancient stages of vertebrate evolution. Also, that some physiological and behavioral features of reproduction by vertebrate ancestors, that previously were thought to border on being inaccessible, now may be amenable to study.

To my disappointment, this paper produced hardly a ripple of interest among workers in related fields. In the next 2 years I received only 18 reprint requests, only 4 of which were from abroad. Eleven were from individuals in biology and zoology departments and, to my surprise, 7 were from psychology departments. I had little time to reflect on these circumstances, however, as I was already deeply buried in studies for my next paper. Rather than pursue further the specific relict behaviors and physiological aspects and their significance, I had come to concentrate upon the neural basis for conservative behavioral evolution, leading eventually to the topics of memory and sleep..

Conservative Behavioural Evolution, the Neural Substrate

The journal, Animal Behaviour, in which I had not published since 1963 ("The study of social interactions between small animals;" 11:263-273) seemed a logical choice for a paper titled, "Conservative behavioural evolution, the neural substrate." Yet again, I misplaced or accidentally discarded my file for this paper on moving from my old office. It turned out that at the time of submittal, in 1989, Dr. Lee Drickamer was Editor. He had been a doctoral student in Jack King's lab at Michigan State U. when I spent some sabbatical time there in 1966, so the welcome was friendly. Although there were numerous constructive referee comments to take into account in the final version, the paper appeared promptly in 1990 (39:758-767).

Since much material already covered earlier also is treated in this paper, I draw almost entirely on the Abstract, as follows.

The evolution of neural circuitry differs from that of other structures in a significant respect. Natural selection cannot physically eliminate most neuronal pathways for obsolete functions, because the individual neurons of these pathways usually are multifunctional, also being components of pathways for adaptive functions. Selection for loss of obsolete neural functions probably operates largely through mechanisms that suppress or neutralize corresponding neuronal pathways without disabling pathways for adaptive functions.

Accordingly, rather than becoming eliminated physically, obsolete neuronal pathways tend to become inhibited or have their outputs blocked; obsolete pathways not selected against merely cease to be accessed. These mechanisms of loss of neural function provide one of several intrinsically conservative bases for neural ontogeny and evolution. Since the CNS (central nervous system) encodes and programs motor outputs for behavior, the obsolete circuits that are retained include pathways for obsolete behavior. In vertebrates such behavior can sometimes be elicited directly by appropriate stimuli, but when the neural pathways for the obsolete behavior are in an inhibited or blocked condition, they first must be rendered accessible (which can occur adventitiously, pathologically, experimentally, or in hybrids).

The vertebrate CNS also plays a large role in programming the ontogeny and maturation of gametes. Egg formation in many species appears to be highly conservative, extensively retracing ancestral stages. This conservativeness probably derives from influences both of being under neural control and from having a long evolutionary history in which the provisioning of eggs was continually on the increase.

Obsolete pathways and memory traces

The last section in the Animal Behaviour paper was titled as above. In it, I suggested that knowledge of neural pathways for obsolete functions might contribute to our understanding of those for memory traces of other sorts, and vice versa. "A lifetime potential for the activation of uninhibited or disinhibited obsolete pathways for relict responses parallels the often lifetime stability of long-term memory traces....It seems unlikely that long-term memory traces are maintained in suppressed or unrecalled states by more than one type of inhibitory mechanism. Most likely, suppressed long-term memory traces are accessed merely by exceeding heightened excitation thresholds that are imposed by a common means," and similarly for many neural pathways for relict behaviors.

These words were the preamble for my new direction of research into memory, at first, and then, inevitably, also sleep--the two being inextricably intertwined. The response to the Animal Behaviour paper, though not insignificant, again was disappointing. My having elucidated pathways to knowledge of the greatest interest, that previously was believed to be all but unknowable, and having accounted for the existence of such pathways, I expected would have attracted more interest from the scientific community. But interest was, at most, moderate.

The reason for this comparative disregard would appear to be clear. No one else had observed birds so intensively--almost around the clock, and in such great detail. Accordingly, it was not surprising that ornithologists and evolutionists did not put much faith in discoveries of such a revolutionary nature that they had eluded thousands of other workers with captive birds. That remains the situation even today (2004).

Except for its degree of significance, the situation exactly parallels my discovery with Dr. Carl Rischer of "Program clocks in small mammals" (Science 1968;161:1256-1259). There, also, the finding was revolutionary, but no one else had the facilities to repeat it or carry out further work along the same lines. We continued the work at UCLA, finding 'program clocks' in other mammals, as well, even primates (Primates 1974;15:209-217). But this most interesting facet of animal behavior remains not further explored. Instead, studies now concentrate on rhythmicity, itself (but using retina-damaging, ecologically unsuitable bright-dark light regimes).

Returning to the paper in Animal Behaviour, it is of interest to take note of the identity of those who took the greatest interest in it, based on reprint requests. Of the 87 requests received in the year or so after publication (1990-1991), 43 originated in the U.S. and 44 abroad. Twelve were from Germany and 10 from Canada. Judging mostly by department or institute of origin, 46 were from zoologists and 13 from psychologists, 10 from  'biologists' of unknown specialties, 6 from neuroscientists, 4 seemingly from MDs, 2 from physiologists and botanists, and 1 from a bioengineer.

Three years of intensive study

Some 3 years of intensive study in the fields of memory and sleep intervened before my next publication in 1994, after retiring. Little did I suspect either the hurdles to publication that lay ahead or the few outstanding successes. Very early on in my new program of studies, I came to the conclusion that long-term memories were maintained through 'exercise' of their neural circuits through what I called "non-utilitarian" use, mostly during sleep (rapid-eye-movement [REM] sleep in warm-blooded animals [endotherms]), which I eventually concluded accounted largely for the process of dreaming by endotherms.

I referred to this maintenance of long-term memory through non-utilitarian use in endotherms as "dynamic stabilization." In the course of writing my first paper on the topic, titled "Sleep and dynamic stabilization of neural circuitry: A review and synthesis," I discovered that Bernard D. Davis, the well-known Prof. of bacterial physiology at Harvard, had already suggested much the same scenario, in 1985 (Perspectives in Biology and Medicine 28:457-464).

In essence, Davis proposed that maintenance of long-term memories depends on frequent reactivation of their neural pathways, principally during REM sleep. He arrived at that conclusion in the same way I did, reasoning that concentrations of the proteins responsible for very lengthy maintenance of enhanced synaptic efficacy would have to be augmented periodically by synaptic use, as the proteins became degraded.

Accordingly, in about 1992, I wrote to Davis informing him of my ongoing project and views, and including MS pages. Since priority for the theory was his, I invited him to be a coauthor on the paper. Unfortunately, this is another instance in which my correspondence was lost or unintentionally discarded when I moved from my old office, so I do not have the exact wording of Davis' reply. In essence, though, he felt that taking any credit, other than recognition for originating the theory, in my paper would have been an unwarranted imposition, so he declined. I did continue to consult with him, and received valuable comments from him in an exchange of perhaps 5 letters before his untimely death from prostate cancer early in 1994, before my paper appeared.

Sleep and Dynamic Stabilization of Neural Circuitry: A Review and Synthesis

The road to publication of this first paper in the field of sleep and memory in what was essentially a neuroscience journal was by far my longest and most difficult, insofar as journals and referees are concerned, but one of the most rewarding, measured by reader acceptance. The paper eventually appeared in Behavioural Brain Research (1994;63:111-126). My object was to present a brief review of pertinent information concerning memory, sleep, and related fields, and to synthesize this information into a coherent picture, both for sleep researchers and for workers in peripheral, and even remote, fields. The Abstract and Summaries of the principal contents of this paper are as follows.

Abstract. A common mechanism is advanced for the lengthy stabilization of neural circuitry encoding information of both hereditary and experiential origin. Stabilization is proposed to occur through the following means and interrelationships. Synaptic function is intrinsically plastic because of greatly restricted entry of essential, relatively short-lived molecules into synaptic terminals. Alterations that accompany synaptic transmission transiently facilitate this entry ("facilitated entry"). Synaptic efficacy is enhanced as the concentration of these molecules increases following a transmission event but subsequently declines if depletion of the molecules occurs without commensurate replacement.

Accordingly, if lengthy persistence of information encoded by enhancements of synaptic efficacy is to be achieved, the enhancements must be reinforced repeatedly by synaptic transmission ("dynamic stabilization"). Synapses of circuits not in frequent functional use are thought to be dynamically stabilized by spontaneous, internally generated non-utilitarian excitations occurring primarily during rest or sleep. In species with complex, highly developed brains, requirements for dynamic stabilization of infrequently used circuits apparently cannot be met during rest, a restriction that may underlie the origin of sleep. Dynamic stabilization of infrequently used motor circuits of endotherms appears to occur predominantly during REM sleep. [Except for details of the evolution and basic function of sleep, this Abstract essentially captures the themes and germ of my studies of memory and sleep for the next 12 years. Though greatly elaborated and buttressed, it includes Bernard Davis' germinal proposal of 1995.]

The Introduction to the paper elaborated on the general proposals of the Abstract, adding the thesis that REM sleep evolved through modification of a fraction of NREM sleep. A section on the literature on dynamic stabilization and REM sleep reviews the pertinent studies of Roffwarg et al. in 1966, Jouvet in 1975, Steriade in 1978, Davis in 1985, Hobson and Steriade in 1986, Hobson in 1989, Eccles in 1988, and Steriade et al. in 1993. There follows a section on the Properties of Chemical Synapses and the influences of frequent and infrequent transmission and strength of stimulation, touching on LTP (long term potentiation) and its establishment and maintenance. A section on Some Neural Events of Ontogeny treats of the development and maintenance of the vertebrate nervous system, influences of motor activity, spontaneous discharges and phasic movements, stereotyped pathway activations by pontine brainstem neurons during REM sleep, consequences of insufficient excitation, fine tuning and maturing of circuits through remodeling, occurrence of hard-wiring primarily during REM sleep, etc.

A section on the Ontogeny of Sleep touches on development and progressive changes in REM and NREM sleep, and control of the sleep-wake cycle in mammalian fetuses and in the early post-natal years. It is suggested that the gradual decline in total sleep from neonate to adult is a consequence of the magnitude of the decline in requirements for shaping, fine-tuning, and dynamically stabilizing the maturing CNS, exceeding that of the concomitant increase in maintenance requirements.

There follows a treatment of Neural Activity during Adult Sleep, with emphasis on details of EEG rhythms during, and accompanying transitions between, REM and NREM states, and neuronal correlates. It is concluded that, inasmuch as hard-wiring occurs during sleep in the embryo, infant, and child, and is dependent on frequent reactivation of synapses, it very likely is maintained by the same mechanism in the adult, as suggested by Roffwarg et al., Jouvet, and Hobson. It is also concluded that dynamic stabilization during sleep mediates the storage of long-lasting memories, as implied by Steriade, Davis, and Steriade et al.

It follows from the treatment that all hard-wired circuits, both those genetically endowed and those acquired experientially, should be equally susceptible to remodeling. Evidence supporting the occurrence of dynamic stabilization during both sleep states is cited. The Evolutionary Origin of Dynamic Stabilization is treated next, beginning with the realization that dynamic stabilization would not be needed in primitive sessile organisms. At a later stage, jellyfishes, for example, with the development of motor mechanisms and active predation, selective pressures would favor enhancements of synaptic efficacies that persisted for days, possibly even months, but not for years, as their lifetimes are shorter. [At a later date it was discovered that at least one jellyfish, a cubomedusan, sleeps. This phenomenon later became the subject of an Editorial in Medical Hypotheses (see below).]

Finally, in advanced, long-lived animals, lengthy periods, even months or years, might intervene between events requiring specific adaptive responses. The survival of these animals would depend heavily on long-term memories, and it is suggested that the already existing mechanism of achieving hard-wiring, that is, dynamic stabilization, also came to mediate the hard-wiring of experiential memories. A discussion of possible means of selective targeting of circuitry for dynamic stabilization leads to a consideration of the possibility of sleep being a period of increased activation thresholds in some animals.

These considerations provided a clue to the origin of sleep, namely a need for inactivity more profound than mere rest during dynamic stabilization of circuitry, because of its incompatibility with the circuit activities occurring even during rest. Dreaming is seen as an overt manifestation of sleep-state activation of cortical structures that support the conceptual, perceptual, and motor pathways of the waking state, quite in line with the thesis of dynamic stabilization.

A lengthy, in-depth treatment of Facilitated Entry follows. This refers to a relatively non-specific facilitation of the entry into synapses of molecules essential for synaptic function, as achieved by many brain waves during sleep. First considered are the roles of the slender stems of synaptic terminals, which probably 'bottleneck' the entry of essential molecules, and how they might be transiently enlarged. Evidence of great restriction of the entry of these molecules leads to considerations of if, and how, vesicular transport along microtubules could occur. Next to be considered are the sites of synthesis of essential molecules, now known to occur both in the nucleus and at synaptic sites. Lastly, in this section, the possibilities are considered that the facilitation of entry of essential molecules is a highly specific process.

Dynamic stabilization is not the only mechanism by which synaptic efficacy is altered. The next section considers the roles of Alterations of Synaptic Structure and Connectivity, including evidence from LTP in the rat hippocampus, rodents raised in complex environments, alterations after training cats and rats, kindling of rats, passive avoidance learning in chicks, etc. The Roles of DNA in Alteration of Synaptic Efficacy by specific means is treated next, including the synthesis of mRNAs and proteins within a specific time window, including targeting to specific synapses.

Differential functions of REM and NREM Sleep is the next topic, with the treatment being largely of a review nature. The fact that REM sleep goes hand in hand with endothermy is emphasized. It is suggested that endothermy imposes more exacting requirements for the suppression of the normal consequences of motor circuit activations than can be met in NREM sleep, and the mechanisms of such suppression during REM sleep are treated. Partial or total exclusion of sensory input, and the mechanism by which it occurs, are discussed. The nature of the fine movements occurring during REM sleep, the reduced ability to thermoregulate, and the absence of piloerection and panting also are mentioned.

The Evolutionary Pathway from NREM to REM Sleep is the next topic, with a discussion of how the evolution of endothermy apparently brought it about. This topic also is that of some later studies (2003-2004), in which a detailed accounting is suggested. It is pointed out at this point in the 1994 paper that, it seems likely that under REM sleep deprivation conditions, redundant mechanisms of the phylogenetically older NREM sleep state can compensate to some degree in the performance of functions normally confined to REM sleep. Even today (2007), this circumstance is not fully appreciated by many sleep researchers, who continue to base their conclusions on all-or-none reasoning.

Many additional correlations between sleep states and motor circuit activations are treated in the last section before the Discussion. The latter concludes with the statement that, a plausible evolutionary pathway for the origin of sleep is suggested by the proposal that lengthy retention of infrequently used neural circuits of animals with complex, highly developed brains depends on the circuits being reinforced by dynamic stabilization without interference from the neural activities that accompany a state of mere rest. I would note that at this time I had not yet hit upon the key interfering neural activity that accompanies "a state of mere rest," namely, the enormous requirements of processing visual input in animals with detailed focal vision. This was to be the theme of most of my findings in future papers. Only in the next paper, in 1996, discussed below, did I still refer merely to "obviating the need for very extensive visual input processing," rather than highlighting detailed focal vision as the key element.

My expectation was that this paper would be greatly welcomed by sleep researchers and workers in all related fields as a useful and fairly comprehensive review and synthesis of many considerations, some largely overlooked, pertaining to sleep, its roles in memory, and its evolution. In this, I was not disappointed. What I did not anticipate was the long road ahead toward getting the paper into print. It was only after submitting to six journals over a 17-month period that this was achieved. In the course of these submissions the paper was successively improved by many helpful comments of referees. Most of these are of a technical nature, so they will not be treated. The paper did, however, cover essentially the same ground enumerated above from its inception.

Submitting to Sleep

With the expectations discussed above, I submitting the paper to the journal that all sleep researchers read, namely, Sleep. Accordingly, on Sept. 21, 1992, I sent the manuscript to Editor-in-Chief Prof. C. Guilleminault. After an acknowledgement on Sept. 24 from the Sleep Editorial Office, Prof. Guilleminault wrote me as follows on Dec. 8. "I am sorry to say that we shall not be able to use your article for publication in Sleep. I enclose the comments of our reviewers, which hopefully will be helpful to you. Thank you for your support of Sleep. Brief overall comments of the reviewers follow.

Reviewer 1. The author has surveyed a broad range of the applicable literature, but what has come forth is hardly a reasoned synthesis.... It needs much sharpening.

Reviewer 2. The author provides a wide-ranging and lucid review in diverse literatures in learning, phylogeny and sleep control. However, the link that the author makes between these literatures is tenuous. Moreover, the predictive value of his theory is unclear..... the theory's primary value is its summary and combination of the literature.

Submitting to Brain, Behavior and Evolution

These comments discouraged me from submitting the paper, in very little modified form, to another journal in the field of sleep. Instead, on January 25, 1993, I submitted to Editor Northcutt and Brain, Behavior and Evolution. On May 21, Editor Northcutt wrote to me that the manuscript would not be accepted. Some comments of the referees follow.

Referee 1.... I fear that Professor Kavanau has misjudged his audience in submitting this review to BBE. In my opinion, the very general treatment of these topics is too non-specific and too hypothetical for the bulk of the readership of this journal, who mostly are neuroanatomists, neuroethologists, and neurophysiologists, many of whom also have a strong background in evolution and systematic biology.....This is not to say that the paper is uninteresting or undeserving of publication. I would suggest that the author consider submitting it to a different journal.....

Referee 2....assertion of possible, very general, cause and effect relationships....untested scenarios about the evolution of sleep....the two ideas that are central to the final conclusions of the paper are not new ones (as the author finally points out, with several appropriate citations)....the only new ideas presented in regards to this is the suggestion of "facilitated entry" of "essential molecules" into synapses as a mechanism of learning....

Submitting to Psychological Reviews

On June 1, 1993, I submitted a slightly modified manuscript to Editor Walter Kintsch (U. of Colorado, Psychology) of the Psychological Review, having decided that the paper might be more appropriately placed in a review journal. On August 31, I received word from Associate Editor John T. Cacioppo (Ohio State U., Psychology) that after extensive reviews by four referees, "I must forgo the opportunity to publish your paper in the Psychological Reviews." My reply to his letter, as follows, treats of the referee comments.

"....It is a new experience for me to have a submitted paper lauded so highly, yet rejected on the basis of critiques, comments and suggestions that could be satisfied to a great degree by addenda and revisions. I have always welcomed the return of a manuscript with reviewer comment sheets for the very great benefits that they usually provide. In this instance, I can give an instant response to two of the main reviewer objections cited in your letter, which appear to be unjustified.

Concerning comments by reviewers B and C, you write, "....some of the ideas (e.g., dynamic stabilization) are not all that new (as you acknowledged)." But how can a review fail to include ideas that are not new? If it is meant that the reviewers received the impression that I was suggesting that the first thesis of the paper (subsumed in the title) was original with me, despite the fact that the paper reviews prior proposals and near proposals of this thesis, the gaining of this impression could be obviated simply by adding the phrase, "for which several precedents exist," on my presentation of the thesis....

Reviewer A writes: "neuronal and molecular portion of this paper is interesting and well presented....creative conception....remainder of the paper would be well worth reviewing again." Reviewer B writes, "....Well written: very clear and highly interesting....Paper should appeal to a wide audience....synthesizes a diverse literature in an imaginative way....provides novel expression of theory....good review of synaptic spine story....bottleneck idea fascinating....concept of facilitated entry of essential molecules novel and interesting." Reviewer C is highly critical and makes many suggestions, but he thinks highly enough of the "ambitious review" to suggest a reorganization. Reviewer D writes, "....an intriguing and thought-provoking integration of very complex processes, functions, and functional states...."

Submitting to Behavioral and Neural Biology

On September 21, 1993, I submitted the paper to James L. McGaugh, Editor of Behavioral and Neural Biology. I received no word, however until Dec. 5, and then without the benefit of any reviewer comment, which was unprecedented in my experience. The 'unacceptance' notice read as follows.

I am returning your paper....As the reviewers judged the paper to constitute no new understanding of the issues addressed. I will not be able to accept the paper for publication in Behavioral and Neural Biology.

I wondered who those reviewers might be (and still do), and did momentarily consider replying with a request for reviewer comments, but thought the better of it. This was essentially the same paper that, not many months later, generated 226 prompt reprint requests from around the world, some from leading researchers, and even a phone call, almost before the ink was dry.

Submitting to Behavioral Neuroscience

Within a very short turnaround time I submitted the paper to Larry R. Squire, Editor of Behavioral Neuroscience on December 14, 1993. The paper was returned essentially by return mail one week later with the word that "...it is not appropriate for Behavioral Neuroscience. The journal publishes almost exclusively original research reports. I would suggest Brain Research Reviews as the journal most suitable for your paper."

Submitting to Behavioural Brain Research

One week later,on December 28, 1993, I submitted the same paper to Ian Steele Russell, Editor-in-Chief of Behavioural Brain Research.This time I met with some success. Editor Russell informed me early in February, as follows.

If, however, you are willing to thoroughly revise your paper to satisfy the points made by each Referee, I would be glad to reconsider your paper for publication in Behavioural Brain Research.

When you send me your final draft, please include a summary sheet of all changes made. This should be a detailed itemization of the changes made to each separate Referee's points. Also please send a copy on floppy diskette.

I am looking forward to receiving your revised manuscript.

With those encouraging words, after about 17 months of effort, including submitting to six journals, I set about to deal with the referees' lengthy but helpful and constructive, though occasionally not entirely valid, comments. In the following, I merely mention the specific matters raised by them and dealt with by me, as a full treatment would consume an entire chapter.

Topics of Reviewer comments. How to test the paper's main tenets experimentally -- why synapses that are not in frequent use should be preferentially excited during sleep -- evidence is not provided that the renewal of proteins depends on synaptic activity -- too frequent citation of monographs and symposia -- matters pertaining to correlations between REM sleep and the degree of ontogenetic development of the CNS -- no real evidence that in REM sleep, 'input to perceptive and cognitive modules is solely internal' -- matters relating to dreams and REM sleep -- thermoregulation and REM sleep -- sleep states during hibernation -- sleep states and body movements in the human fetus -- influences of human exercise on REM and NREM sleep -- insufficient taking into account of recent papers by Horne, Steriade, and others -- REM versus NREM sleep as times of reinforcement of synapses -- taking further account of the slow rhythm of sleep.

Reprint requests

The reader response to this paper was numerous and swift, even a phone call from a sleep researcher who, subsequently, became my main proponent. As to the readership, there were 226 reprint requests, 74 from the U.S. and 152 from abroad. This ratio of roughly 2 from abroad to 1 from the USA held up in the coming years for papers related to sleep. Based upon the institutions of affiliation, there were 46 from institutions of medicine, 40 from fields of psychology, psychobiology and ethology, combined, 26 from physiology and pathology combined, 20 from anatomy and behavior, and neurobiology, combined, 16 from psychiatry, 14 from neuroscience, 12 each from biological sciences, and pharmacology and toxicology, combined, 10 from neurology, 9 from sleep labs, 8 from neurophysiology, 4 from physics and biophysics, combined, 3 from psychophysiology, and 1 each from neuropsychology, molecular biology, and physics and astronomy, combined.

It would appear that I had succeeded in my main objectives of presenting a useful review and synthesis for readers in both closely related and peripheral, even remote, fields. Future years were to show that names of the same workers occurred repeatedly on my reprint requests lists. The record for repeated requests is held by S. Kappen (Michigan) at 9, H. Merica (Geneva) at 8, followed by J. Krieger (Strasbourg), N. McNaughton (Dunedin), and N. Rodrigo (Madrid) at 6. Krieger's name came up again in 2005,. as co-editor of Sleep Medicine Reviews, to which I was invited to contribute a theoretical review ("Evolutionary approaches to understanding sleep," 2005;9:141-152).

Sleep and Memory: Evolutionary Perspectives

Shortly after the above paper appeared, I received a letter (October 18, 1994) from Prof. Peter Shiromani of the Harvard Medical School, and Editor of the Sleep Research Society Bulletin, as follows.

I read your recent review article in Behav. Brain Res. and would like to invite you to contribute your ideas to SRS Bulletin. The theme of our articles for the next issue [April 1, 1995] will be on "Sleep and Memory"....I hope to have 7-8 articles....I very much want you to contribute an article. Please let me know by faxing this letter back with a yes or no.....

I gladly accepted Shiromani's invitation and my article, "Sleep and memory: evolutionary perspectives" duly appeared in the SRS Bulletin (1995;1(3):59-65). Even as I heard from Shiromani, I was working on a new paper on sleep, a sequel to my first, titled, "Memory, sleep, and dynamic stabilization of neural circuitry: evolutionary perspectives." I submitted this for publication only two months later (see below), the topics of which I had also appropriated for the SRS Bulletin.

This paper marked my greater emphasis on evolutionary considerations, long the keynote of my avian studies, but now assuming a major role in my studies of memory and sleep. I present only the first paragraph of the Summary of the SRS Bulletin paper, as its themes reverberate throughout the final pages of this chapter.

Summary. The origin of both sleep and memory appears to be closely tied to the evolution of mechanisms of enhancement and maintenance of synaptic efficacy. The development of use-dependent synaptic plasticity apparently was the first step in the evolution of nervous systems beyond a capacity to respond to environmental stimuli by mere reflexive actions. Each activation of a synapse with use-dependent synaptic plasticity results in a transient enhancement of efficacy. Repetitive activation of synapses, referred to as "Dynamic Stabilization" (DS) induced either in the course of frequent functional use or by spontaneous oscillatory neural activity, is the likely primordial mechanism by which synaptic efficacy enhancements were maintained more lengthily-by serially inducing the individual transient enhancements.

Although I subsequently received many hundreds of reprint requests from researchers in all biological, medical, and even physical science categories, my emphasis on evolutionary perspectives widened the gulf between my work and that of other sleep and memory researchers. This traces largely to the fact that preparation for entry into medical disciplines rarely includes studies in evolution. Although a number of papers by sleep researchers include the word 'evolution' in their titles, they rarely actually treat the evolution of sleep. Words and phrases like, natural selection, selective pressure, adaptedness, environment, phylogeny, ecology, etc., do not, as a rule, appear in such papers. (More on this topic occurs later.)

Mostly, these works are comparative treatments of endpoints in evolution, with efforts to deduce the significance of the endpoints for relatedness, not the pathways by which the endpoints came about. Although evolution typically does not lead to "all or none" results, most sleep researchers think only in those terms. In the fields of memory and sleep, this is most unfortunate. Below I give one example of the kind of reasoning this can lead to, in the form of a critique by a highly regarded, even pioneer, sleep researcher of some work of mine. This appeared in a paper in 1998 titled, "Current perspectives on the function of sleep.

....Would a night of sleep enhance transmission along a defined neural pathway? Would sleep deprivation reduce transmission? Kavanau prepares us for negative results on deprivation studies by noting that they may have "produced little or no evidence of deterioration of mental function" because "A few nights' lack of circuit reinforcement need not impact seriously on brain performance; to a certain degree functional capacities can be extended and maintained by synaptic enhancements during waking" This escape clause seems to protect the theory against functional tests.

The author of this critique apparently believes that nature, via evolution, draws sharp lines between the actions of brains during waking and sleep. The contrary thesis is more valid. In ridiculing the above proposal, he displays a limited appreciation of the underlying forces in play.

An Aside

It was during this period that I noticed a Student's Perspective in the SRS Bulletin [1995;1(1):14] by Gina (1994 SRS Student Representative and Member at Large). At that time, as I learned later, she was completing her degree in neuroscience with Ron Harper, one of UCLA's several well-known sleep researchers.

In 2001, Gina was the senior author on a paper, "Experience-dependent phase-reversal of hippocampal neuron firing during REM sleep," in Brain Research (855:176-180), for which I requested a reprint on April 27, 2001. On May 2, she e-mailed about mailing me the paper, since my printer could not handle a PDF at that time, and mentioned that she was a UCLA graduate. I asked with whom she had obtained her degree. On that same day she wrote,

I want to add that your ideas are very compelling and have influenced my thinking. Is that your rule that you publish alone or would you consider writing a review article with me. I have several ideas that tie in well with yours, and bring in ontogeny as well as phylogeny. In any case, I would love to discuss things with you."

After that unexpected compliment from a fresh, young mind in the field, I replied, "Thanks for your second e-mail, which I didn't see before responding to your first. I have been publishing alone because I am almost 12 years into retirement and my research now is strictly of the library and armchair variety. However, I now have a paper in the works coauthored with Bruce Charlton of Newcastle upon Tyne covering some implications of my studies for the etiology of mental disorders.

Concerning your proposal, anything is possible. First let me know which of my ideas in the literature you find compelling, and send me a complete list of my papers that you have read, so I can complete your file. Also, I have privately published a monograph covering my work through early 1999 containing much additional unpublished material. I have a paper in press and another almost in press (haven't heard about acceptance of the revision yet) that should bring my ideas to the attention of a much wider group of scientists, and, perhaps open some new doors for me.

When you have time, if you wish, you might give me some idea of what you have in mind ontogenetically. Having taught both vertebrate and invertebrate embryology here at UCLA, studied with Ted Bullock and Paul Weiss, and covered ontogenetic and phylogenetic topics related to memory maintenance in early papers, I have some background in that area. I also have a 2001 book chapter on 'Reinforcement of memory during sleep' that I can send you." Gina replied that same day, as follows.

I worked with Ron Harper in the Brain Research Institute, 7th floor. The project was reflectance imaging of forebrain (hippocampal) activity during spontaneous sleep-wake cycles and respiratory events in cats. I look forward to seeing your forthcoming papers and chapters. I will send you my article this afternoon. It certainly seems like you are bringing forward the field even from retirement!

Though I had expected further word from her, that e-mail was the last of our correspondence at that time. Was she intimidated or merely averse to collaborating with a senior citizen? That question was answered partially a few years later. I had further contact with Gina in 2006. I noticed in her latest publication that she had moved to the Dept. of Anesthesiology at the U. of Mich. That being my Alma Mater, I sent her my recent publications and a short note. In Feb., she replied as follows:

Thank you so much for sending me your articles. I very much enjoy reading them. Thank you for not forgetting me. I am happy here at the U. of Mich. It is nice to see that this place graduates such great scientists as you.

Last August while on vacation I read an article in a non-scientific, non-peer reviewed journal about the box jelly fish, how it is a hunter with many eyes and has been observed to sleep - or at least to spend significant periods of time inactive, unlike more passive jellyfish who do not have the complex visual processing. I got very excited and did think of you and your hypothesis regarding the function of sleep. It is so nice than you were on top of this discovery too!

Of your thoughts about sleep, the only one I need further convincing on is that non-visual animals do not sleep. You may be right, but I need to study it more....

All the best to you, Lee. Again, thank you for not forgetting me. Sincerely yours.

Memory, Sleep, and Dynamic Stabilization of Neural Circuitry: Evolutionary Perspectives

As I mentioned earlier, I was working on the above paper, even as I prepared a contribution based on it to the SRS Bulletin in response to Shiromani's invitation. I submitted a draft of this paper to Prof. Ian Q. Whishaw, Editor of Neuroscience and Biobehavioral Reviews (referred to hereafter as NBR) on December 2, 1994, who acknowledged its receipt two weeks later. Whishaw informed me that he would review the paper himself, and also send it out to two referees.

On the following March 9, Editor Whishaw wrote as follows.

I enclose two reviews of your manuscript, "Memory, sleep, and self-activations in the sleeping brain: Evolutionary perspectives." [the original title] They are favorable but also point out a number of problems that they see in the manuscript. They also make a number of suggestions with respect to both style and content. In addition, I have read the manuscript and I take the liberty of adding a few comments of my own. I hope that these thoughts and suggestions will aid in preparing a revision of the manuscript.

....I hope that you will consider the various comments of the reviewers in the revision of your manuscript. I would be pleased to receive two copies plus a copy on disc.

P.S. I have been wondering about the term dynamic stabilization - where does it come from? Is it possible to have it or a more informative term in the title of the manuscript. I am not sure whether you mean self-activations to mean the same thing.

I returned my revised Ms. on April 3, with the term, "dynamic stabilization," substituted for "self-activations" in the title, and received Whishaw's letter of acceptance, dated April 7. As with the previous paper, I shall merely mention the areas in which the reviewers requested revisions. All 3 reviewers gave lengthy critiques and helpful, constructive suggestions. My reply to Whishaw, with references to revisions, ran to 3-1/2 single-spaced pages.

Whishaw. Hippocampus and hippocampal replay, theta rhythm and possible distraction of the reader from the main point of the paper, excessive length of certain sections of secondary importance, revision of section on Redundancy of Neural Circuitry,

Referee 1. Lack of sufficient references for section on Facilitated Entry, insufficient treatment of whether sleep represents a process of dynamic stabilization, matter of REM versus NREM sleep and ontogeny recapitulating phylogeny, more on significance of low percentage of REM sleep in birds (both referees), matters of closing eyelids, benefits of sleep, and role of alpha rhythm in relaxation, redundant and modulatory circuitry, meaning of continuous use, dream evanescence, question about non-utilitarian reinforcement, conflicts between my views and those Kreuger and coworkers, mechanism by which random firing could stabilize functionally useful circuits.

Referee 2. Significance of low percentage of REM sleep in birds, clarification of a title and two phrases, several critiques on failure to deal with a substantial sleep-learning literature, matter of changes in REM sleep and dynamic stabilization in aging individuals.

Concerning this very lengthy paper (23 pages), itself, as it finally appeared in print, rather than duplicate the lengthy Abstract, I summarize the treatments and give main section titles.

The Introduction defines "non-utilitarian dynamic stabilization (DS) and considers its origin in spontaneously oscillating circuits and brain states, and selective pressures for the origin of such circuitry and brain states. Much additional material summarizes the "Evolutionary Perspectives" paper, as follows.

1.  Introduction
2.  Facilitated Entry: A Possible Primordial Basis for Use-Dependent Synaptic Plasticity
3.  The Hippocampal Formation
4.  Neural Activity in the Sleeping and Waking Brain
5.  Sleep States and Neural Circuitry Consolidation and Reinforcement
6.  The Selective Pressure for the Evolutionary Origin of Sleep
7.  Evolution of REM Sleep
8.  Causation of "Inconsequential Phasic Events" During Adult REM Sleep
9.  Dreaming and Poor Dream Recall
10. Summary of Proposed Evolution of Proposed Neural Adaptation
11. Discussion

Reprint requests

Reader interest in this somewhat more technical paper, as measured by reprint requests, was less than half of that for the BBR paper, but interest abroad was at a greater ratio, the figures being 100 requests from abroad and 35 from the USA.

Memory, Sleep, and the Evolution of Mechanisms of Synaptic Efficacy Maintenance

and

Origin and Evolution of Sleep: Roles of Vision and Endothermy

The publication of these two papers was intertwined in an unusual way, and can best be dealt with together, as events unfolded. Publication of the Memory, Sleep......paper entailed a considerable achievement against what I believed, at the time, were insurmountable odds, as follows.

In 1991, R.R. Llinás and D. Paré authored a Commentary in the journal Neuroscience, titled, "On dreaming and wakefulness." Wishing to reach a general audience in the field of neuroscience, on October 18, 1994, I wrote to Llinás, the Chief Editor of the Journal, as follows.

I am currently preparing a review titled.....I am writing to inquire whether I may be permitted to submit it for consideration for publication as a Commentary in Neuroscience. Because I do not have a long track record in the neurosciences, I have listed my [three] publications in this area below. I also include section headings for the paper to enable you better to judge its suitability.

Perhaps you will remember me from almost 40 years ago when our paths crossed in Paul Weiss' laboratory at the Rockefeller Institute I have been here at UCLA ever since then. [I learned later, after seeing Llinás on TV, that I was mistaken about having crossed paths with him. However, he never commented on my mistaken recall. JLK]

On December 8, I received a reply from Llinás, as follows.

We received the letter outlining your proposed paper as outlined below. The subject is a very interesting one and I would be happy to consider a manuscript for publication as a Commentary in Neuroscience.

The section headings you have suggested are appropriate for a book and the subject matter as proposed is vast. May I suggest that you limit the paper to cover the two points that you consider to be most important? In this way, the readers can take home a clear message. Otherwise, the paper will be either far too general or far too long to be a real Commentary

On January 5, 1995, I replied to Llinás, thanking him for the opportunity and informing him that I had decided on a main thesis of "the evolution of synaptic efficacy." The subthesis was to deal with a postulated accompaniment of one of these mechanisms, the "phasic events that accompany REM sleep." On January 12, I wrote again, submitting an MS. titled, "Evolution of mechanisms of enhancement and maintenance of synaptic efficacy." Llinás received the MS. on Jabuary18, and informed me on the 25th that it had been sent out for review. On March 21, Llinás wrote as follows.

Review has now been completed for your manuscript and the editorial comments are enclosed for your perusal.

Although one reviewer suggested few changes for your manuscript, it has received rather stronger criticisms from the other reviewer [the understatement of the decade - JLK] and we feel that these comments must be carefully considered. If you feel that you can answer these questions satisfactorily, a revised manuscript will be considered by the editors. It would of course be sent to the initial reviewers for their opinions.

Reviewer 1. I was informed that "reviewers 1 had no comment, except that "the manuscript was overall favorable."

The first two paragraphs only of reviewer 2's review are cited. The following 19 paragraphs are in the same vein and tone, and are referred to further in later reference to them, below.

Reviewer 2. This highly speculative manuscript, loaded with poorly supported conjectures and factual errors, is an attempt to explain the evolutionary origin of synaptic plasticity. It should not be published.

The manuscript itself is a litany of loosely connected findings and speculations, ranging from control of respiration to cognition, without a coordinating theme or hypothesis. The "guiding" hypothesis of the author is a neologism termed "dynamic stabilization" (DS; as opposed to rigid stabilization?). The author regards this hypothetical process, introduced in two previous reviews on this very same topic, as an existing, real mechanism without any experimental evidence or clear definition.

Reviewer 2's probable identity was evident to me from his comments. He had requested a reprint of my first sleep paper in BBR. His native language is not English, and I have corrected the grammatical errors in his comments, so as not to appear to be ridiculing him. In the light of his comments, particularly, "It should not be published," and Llinás having informed me that Reviewer 2 would also review any revision, I despaired of getting the paper published as a Commentary in Neuroscience. It was evident from Reviewer 2's comments that he would not recommend publication of a revision. Subsequent events were to prove me wrong, as far as publication and the probable identity of Reviewer 2 are concerned.

In connection with Reviewer 2's comment about the term "dynamic stabilization," recall that on March 9 of the previous year Whishaw had queried me as to its origin. The term is defined in the Introduction to the NBR paper, which begins as follows.

In a recent review and analysis evidence was presented supporting the proposal that the efficacy of synapses in neural circuits encoding inherited information ('phylogenetic memories") and lengthily stored experiential information ("ontogenetic memories") is enhanced, and the enhancements are maintained (reinforced) either by frequent functional use of the circuitry or by synaptic activations induced by spontaneous oscillatory brain activity, mechanisms of reinforcement referred to as dynamic stabilization (DS). Since the spontaneous activations are thought to serve only to enhance and maintain synaptic efficacy, and usually do not trigger the performance of circuit functions (often inhibited by increased thresholds for activation), they are referred to as being "non-utilitarian."

An aside

There was another occasion, later, when I came up against a reviewer who evidently was determined, for whatever reason, to block the publication of a paper of mine, that is, of sufficient interest to comment on it here. On April 24, 1998, I submitted a paper, "Fatal familial insomnia, sleep, and dynamic stabilization of neural circuitry," To R.C. Griggs, Editor of Neurology, a journal strictly in the medical domain. On May 18, Editor Griggs replied as follows.

The two reviewers of your manuscript sent in opposing recommendations. Consequently, I am sending it to a third reviewer for adjudication. That will explain the delay.

On May 29, I received a letter from Associate Editor Karl Kieburtz, as follows.

Your paper has been reviewed and it needs major revision. It is not accepted by Neurology. If you make the revisions, we will reconsider your paper by sending it back to two reviewers. The comments of all three reviewers are enclosed. In addition, all style, format and material adjustments must be made before your revision is reviewed again and approved....Looking forward to receiving your revision.

Concerning the reviews, understandably, MDs specialized in the specific areas dealt with would be able to find shortcomings in a theoretical paper by an armchair researcher, and they certainly did. I will not recount their comments on very technical matters. All three reviewers found the paper to be excessively long. On June 8, I submitted a thoroughly revised Ms., detailed explanations for each change, referring to each reviewer's comments in detail. In my entire career, this had never before failed to lead to an acceptance -- but not this time. It was evident to me from the single reviewer's comments accompanying Grigg's letter of August 3, that I could spend endless hours trying to satisfy them, with little or no assurance of success. Accordingly, on August 13, I requested that Griggs withdraw the Ms. from consideration.

The material of this paper was subsequently incorporated into other papers on sleep pathology, particularly, "Adaptations and pathologies linked to dynamic stabilization of neural circuitry" (NBR, 1999;23:635-648). Ironically, in my just received (03/26/04) invitation from Prof. Michael V. Vitiello, Editor-in-Chief of Sleep Medicine Research, to contribute an article on, "Evolutionary approaches to understanding sleep," he writes: "A focus on sleep per se is certainly appropriate, but if you also could consider certain aspects of sleep disorders such a finer focus would be of particular interest to our readership." Material on the two extreme sleep disorders -- fatal familial insomnia, contrasted with encephalitis lethargica -- the one with no sleep, the other with only sleep, was eventually included in that contribution.

I return now to the matter of the Neuroscience paper and the letter from Llinás on March 21, 1995, with the reviewer's comments that led me to abandon any further effort to get the paper published in Neuroscience. Instead of revising the paper, I submitted it, with the altered title, "Sleep and the evolution of mechanisms of maintenance and enhancement of synaptic efficacy," to Editor D.P. Purpura of Brain Research Reviews on July 5, 1995, receipt of which was acknowledged on July 13. The paper was reviewed by one "referee-editor" as excerpted below. Despite the referee-editor's laudatory, constructive comments, I was informed in a letter of Sept. 15, 1995, that it was not accepted by the Editorial Board because "it has been necessary to accept for publication only those reports considered to be of high priority."

Comments of the referee-editor of Brain Research Reviews: Although quite long, this review reads smoothly and reveals an author with a highly developed synthetic intelligence. He draws on his unusually strong background in biology to elaborate on what the essential differences in the simple nervous systems of cnidarians and platyhelminths and those of complex REM-sleep endowed endotherms, and how REM might have evolved. The phylogenetic aspects of the MS were most interesting to this reviewer

The author has an extraordinary reach, surveying minute details of a field not his own (non-REM and REM physiology). What a prodigious review! It is not perhaps surprising that his grasp often falls short. The uninformed reader is likely to be misinformed at times and subjected to a non-critical acceptance of whatever appears to support the hypothesis. The informed reader wonders about what happened to the negative evidence. This occurs at various points through the review....

The most serious problem for a reviewer in Brain Research Reviews is the under specification of the hypothesis of "dynamic stabilization." As presented it does not appear to be testable....This reviewer found it most frustrating not to be able to understand exactly what was meant at this level....How do we know which circuits are phylogenetic and which ontogenetic? Please help with examples which would also strengthen the argument. As a consequence DS appears something like a slogan without deeper underpinnings. UnfortunateIy, this concept is the central original idea of the review (he appropriately gives credit to earlier avatars of the other hypotheses he avows). Without it, the review becomes a very interesting summary of the literature with a extraordinary depth of knowledge in certain areas, but not quite one that offers neuroscientists clearly testable hypotheses.

I lengthily considered requesting the opportunity to revise the paper and resubmit to BRR, and even composed a letter to that effect. However, I eventually decided to submit a revised preliminary outline, retitled as "Memory, sleep, and the evolution of mechanisms of synaptic efficacy maintenance," to Editor-in-Chief Steven B. Dunnett of the Brain Research Bulletin (BRB). I wrote to him on December 18, 1995, including the same section outline of the review that I originally had sent to Llinás, which he had suggested would require a book-length MS. I explained to Dunnett that the MS. was not yet complete but would be ready in a month or two. Dunnett replied, as follows on January 2, 1996.

Thank you for your letter. I confirm that I am the most appropriate section editor.

We look forward to receiving your review and thank you for considering BRB for this submission.

On June 19, 1996, I wrote to Dunnett again, explaining that I had been too optimistic earlier, but the paper was now (6 months later, rather than a month or two) ready to submit. However, I explained that because it would require 115 double-spaced pages, I was submitting it single-spaced, two-sided MS, with the references in 8-pt type. The submission was meant merely to test the waters as to its appropriateness for BRB. If it were, I would submit a MS in standard format. Receipt of the MS. was acknowledged on June 24, with Dunnett replying on July 10, as follows.

Thank you for submitting your manuscript to BRB. I have now received reports from two referees and I enclose copies of their reports.

In topic and style, this article is highly relevant as a review in BRB. However, I am particularly concerned about the apparent overlap between the present review and that which recently appeared in NBR. We can only consider original works for publication in BRB. There are two options. Either I can ask you to elaborate for me the truly novel features of the present work, although on perusal of the two manuscripts I anticipate that I will take some convincing. Alternatively, I can ask you to revise your manuscript considerably in order to present a new and/or original perspective to the issue. In particular I believe the published review provides a good presentation of your own theoretical framework but gives a much sparser account of other frameworks and theories. I would welcome a revision that provides a comparative review of how your model compares and contrasts with other models and theories. I look forward to receiving your comments.

The comments of the two referees were on the whole favorable, but lengthy, so I do not present detailed material from them. Excerpts are as follows.

Referee 1. This extensive review article covers much interesting ground. The author(s) provide a detailed account of the requirement of higher (sleeping and/or hibernating) animals (as opposed to more primitive species) for synchronized oscillatory firing within populations of neurons within their more complex nervous systems. In doing so, they skillfully combine cellular, behavioral, and ontogenetic and evolutionary angles. This synchrony is proposed to maintain synaptic efficacy and functional memory circuitry by 'dynamic stabilization' (DS), a process whose development correlates with the emergence of more complex sensory processing and behavioral repertoires, and the evolution of sleep.

Cellular mechanisms for this process, including LTP and 'facilitated entry', are presented lucidly and put into evolutionary context. The concomitant requirements of the CNS are explained in terms of various mechanisms for reducing (mainly visual) sensory input and motor output, which would interfere with DS. This is correlated with functional and behavioral aspects of REM and Non-REM sleep. Examples are drawn from many taxa with various sleep regimes, including unihemispheric sleep and hibernation.

In summary, this is a very interesting review drawing on many areas of biological science; the work it covers is of undoubted importance. However, it presents similarities to previously published work and the extent of the changes recommended here lead me to invite the author(s) to resubmit the work at the earliest opportunity.

Referee 2. This review is extensive, with much supportive evidence, and tackles a conceptually interesting subject. However, many of the ideas that are proposed and explored here have been aired previously in a not dissimilar manner....The material presented here which does differ from the work cited above-such as the section which examines the nervous system and behavior of 'simple' organisms-differs mainly with regard to specific details, rather than offering anything which is conceptually different. This is reflected also in the discussion parts of the two documents, which appear to cover perspectives akin to one another.

If there are additional, distinct, strands which are presented in this article, then they need to be emphasized far more. Such a wide-ranging review of sleep and synaptic efficacy offers a wealth of possibilities for different aspects to be examined preferentially. For example, there is scope to discuss why non-utilitarian dynamic stabilization, and not some other evolutionary pressure, explains the existence of REM sleep. A more distinctive emphasis is required throughout this paper on memory, sleep, and dynamic stabilization, in comparison with previously published work in this area.

These comments from Dunnett and the two referees, reinforced a decision I had already made. I had decided to change my plans. I would abandon the intention to have the extensive review proposed to both Llinás and Dunnett published by a journal and, instead, publish an even more complete coverage of the topics privately through the UCLA Document Services. I also planned to keep the document up to date with periodic revisions, and send copies of it out when filling other reprint requests. With that purpose in mind, early in July I had the first copy of the review printed by the UCLA Document Services, and I sent it out to a select small group from my reprint request list. One of my reviews was sent to Llinás on July 2, 1996, with a cover letter.

On July 19, I received a letter from Llinás' Journal Assistant, Marcella Villegas, as follows.

I am writing to you concerning a manuscript, referenced below, which you submitted to Neuroscience. Our records show that we have not heard from you regarding this manuscript in over one year.

If you would like this manuscript to remain active, please let us know. If we do not hear from you within one month from the date of this letter we will assume that you wish to withdraw your manuscript. Thank you for your prompt attention to this matter.

I answered Marcella, on July 23, 1996, in her capacity as Journal Assistant to Editor Llinás, as follows. Thank you for your letter of July 19 concerning my MS, "Evolution of mechanisms of enhancement and maintenance of synaptic efficacy." First, my apologies for not replying to the letter of March 21, 1995 from Prof. Llinás who wrote....(see March 21, 1995, letter, above). Though I would have liked nothing better than to submit a revision, I did not feel that there was any chance of satisfying Referee 2, so I took the option of not submitting a revised MS [as noted above, I abandoned the thought of submitting a revision to Neuroscience, and submitted the review to BRR, instead]. I should have informed you of my decision but I failed to do so, being influenced by not wishing to engage in the discussion of the review of Referee 2. At this time, however, it is appropriate that I do so.

This latter referee left me no choice, stating flatly that "[i]t should not be published." In over 50 years of publishing of research findings in professional journals, I have never received a review so unprofessional, unscholarly, biased, and suggestive of animosity, as that of Referee 2. I will not speculate on his motives but the tenor of his review suggests a lack of experience as a referee. His assessments of the paper are diametrically opposed to the very favorable reception my studies have received within the worldwide community of scholars in the areas treated, and certainly are opposed to those of Referee 1, which was said to be "overall favorable."

Referee 2 seems to be indignant that the author of a Commentary that covers a wide range of disciplines does not possess the same high degree of expertise that he, himself, possesses in his narrow area of specialty. While some of his comments are helpful and constructive, I find most of them to be unjustified or entirely erroneous. They also reveal a lack of understanding of some key points in the paper. I will not review his individual comments. I refer only to one revealing statement, namely that "....it's a pity that the author missed the most interesting recent developments." If the paper is as unmeritorious as he would have one believe, why would this be a pity? Would it not simply be yet another example of the paper's shortcomings? [The recent developments were not "missed." They simply were outside the areas agreed on with Prof. Llinás for the Commentary. JLK]

I have been engaged in the revision of the Ms in the months since the time of receipt of Prof. Llinás' letter, and have benefited from those comments of Referee 2 that were constructive. Mostly, I have added much new material and taken into account many recent findings. At the strong urging of colleagues, however, I have preprinted the article and mailed it to my reprint list, which includes Prof. Llinás. I continue to seek an appropriate journal for its publication, which I mentioned to Prof. Llinás in my cover letter of July 2, 1996.

There would be no point in submitting the revised MS to be considered as a possible Commentary to Neuroscience if it is to be reviewed again by Referee 2. However, I would much desire to do so under other circumstances. Perhaps Prof. Llinás would be willing to examine the original MS. and the preprint of the revised MS., and review the comments of Referee 2. If he should concur in my position concerning the latter's review of the original MS., and feels that the revised MS. might make a suitable Commentary, perhaps he would make an exception and not include Referee 2 among the reviewers of the revision. In that event I would be most desirous of formatting the revision for Neuroscience and submitting it for consideration as a Commentary.

To my surprise, I received a phone call from Marcella after she received my letter the following week informing me that Prof. Llinás had agreed to omit Referee 2, whereupon I submitted the revised MS. on August 12. On September 25, 1996 I received Llinás' letter of acceptance for the Commentary, "Memory, sleep, and the evolution of mechanisms of synaptic efficacy maintenance," in Neuroscience (1997;79:7-44).

These extraordinary circumstances account for my decision when, on July 26, I responded to Dunnett as follows. I was pleased to learn that the content and style of the article are highly relevant for BRB. I believe that I can satisfy the requirements of your first alternative proposal to elaborate the truly novel features of the present work, and also revise the MS in accord with the comments and suggestions of the referees. I outline below my proposed revisions to meet your requirements. I hope they will be a good first step, since you feel "it will take some convincing." Almost 3 pages of single-spaced typewritten pages followed. I also informed Dunnett that I had preprinted the original article. On August 19, Dunnett replied as follows.

Thank you for sending me your detailed comments on the referees' reports and your proposal for revision. Although I will, of course, need to consider the revised version itself before making a formal decision, I welcome the proposal you make. This seems an excellent resolution of the issues raised, and I consider it likely that such a revised version will prove acceptable.

On Sept. 2, 1996, I responded to Dunnett, as follows. ....In contemplating and pursuing revisions, I decided to focus the theses of the paper much more sharply than I had indicated in my letter, while, at the same time incorporating the principal suggestion made by you and both of the referees. Accordingly, I have deleted all topics treated previously in my paper in NBR that do not apply directly to the principal theses and main thread of the revision, encompassed in my proposed new title, "Origin and evolution of sleep: Roles of vision and endothermy." This led to a shortening of the body of the paper by over 35%. There followed two single-spaced pages of description and accounting for the changes.

The revision was submitted with the same letter. Of course, my drastic reorientation and shortening of the paper hinged on the expectation that almost all of the omitted material would be appearing in my Commentary in Neuroscience as, indeed, it did. On Sept. 17, 1996, I received the following communication from Dunnett.

Thank you for your letter of 2 Sept. enclosing four copies of your revision of the above manuscript. I am pleased to be able to write and confirm that this revised version has been accepted for publication as a review article in BRB.....

That completes the very unusual story of the publication of these two papers. I should also mention, significantly, that these were my first papers to specifically attribute a major role of detailed focal vision in the origin of sleep. In all previous papers, I had referred merely to roles of sensory input, in general. It remains only to consider the response to these two papers as judged by reprint requests. The latter were losing their significance, with many scientists gaining access to online versions of papers, and downloading copies.

Reprint requests

The paper, "Origin and evolution of sleep: Roles of vision and endothermy," appeared first, in Brain Research Bulletin (1997;42:245-264) in March, 1997. There were 224 requests for reprints of it, of which 161 were from abroad.

The Commentary, "Memory, sleep, and the evolution of mechanisms of synaptic efficacy maintenance," would have reached a somewhat different audience. It appeared later, in Neuroscience (1997;79:7-44) in July, 1997. There were 155 reprint requests for it, of which 103 were from abroad.

Vertebrates that never sleep: Implications for sleep's basic functions

As the above events transpired, I was already at work on a new paper with the above title, which I submitted to Dunnett for BRB with the following cover letter. My expectations were that exposure of sleep researchers to a wide range of little-known sleep occurrences in vertebrates, with clear-cut implications, might broaden perspectives. Though reprint requests were numerous (see below), success was very much limited. It would remain so until some 6 years later, following my Editorial, Sleep researchers need to bring Darwin on board: Elucidating functions of sleep via adaptedness and natural selection, in Medical Hypotheses (2004;62:161-165), when there were active signs of sleep researchers taking notice. The cover letter to Dunnett follows.

I should like to submit the enclosed Ms....for consideration for publication in the BRB. The thesis of the paper is that a consideration of those aspects of the behavior and ecology of non-sleeping vertebrates that are held in common, gives incisive clues to sleep's basic function, as well as seemingly ruling out such functions as rest and rejuvenation as being basic. Along the way, new light is cast on the functions of fish schooling and the ability of birds to migrate thousands of miles for days on end without sleep. The crucial consideration in these phenomena is the achievement of a reduced or nil need to process complex visual information.

As the paper includes review material on the behavior and ecology of non-sleeping vertebrates, almost all fishes, it perhaps could fall into the Review category.

Receipt of the paper was acknowledged, and notice of its being sent to reviewers received, on November 5. On January 22, 1998, Dunnett wrote of its provisional acceptance subject to responding to the comments of two referees, both of which were favorable, with constructive comments. The revised version was accepted on February 11 and appeared in midyear. In addition to some specific suggestions, One referee wrote, "[t]his excellent review thus fills an important gap. It is appropriate for publication in terms of quality, interest and originality." The other referee made numerous helpful comments and queries but gave no general assessment.

Summary of paper

Continuous swimming without a need for rest or sleep is a highly derived condition, dependent on many specializations of morphology, physiology, and modes of life. At one extreme are the large fishes that sometimes do not school as adults, occupy essentially featureless pelagic habitats, and lead a comparatively routine existence. Their memory circuits, of which those in the experiential category would be minimal, are maintained and refreshed by functional and non-utilitarian DS during schooling and swimming at night, with greatly reduced or nil information processing.

At the other extreme are the reef-dwelling fishes that school during the entire day and disperse at night to feed. In most regards, their daily experiences during feeding are no different than those of comparable fishes that rest or sleep. In entailing greatly reduced need for sensory processing, schooling provides them with the essential benefits of restful waking and sleep, that is, it provides favorable conditions for the refreshment of memory circuits that are not in frequent use.

Between these extremes lie the medium to large fishes that characteristically school, lead a comparative routine existence, and occupy chiefly pelagic habitats. Their significant needs to refresh memory circuits, of which those in the experiential category also would be minimal, are met chiefly during daytime schooling and at night.

Reprint requests

Requests for reprints remained at a moderately high level. There were a total of 198, of which 133 were from abroad. Though there were many from the same individuals who had requested reprints of my previous four papers on sleep, the emphasis on fishes brought numerous requests from ichthyologists and other vertebrate zoologists, who also requested related reprints. This was the first paper for which 5-10% of the requests were by e-mail, and the last paper for which I kept detailed records of reprint requests.

A first submission on the topic of dreams: Dreams and dynamic stabilization of memory circuits

My first, very well received, review publication in neuroscience, "Sleep and the dynamic stabilization...." appeared in Behavioural Brain Research in 1994. In early 1998, I was completing a paper dealing primarily with dreaming, "Dreams and dynamic stabilization of memory circuits," which I decided also to submit to BBR, and did so with a covering letter on April 15th.  This was acknowledged and sent out for review on April 24 by Editor-in-Chief Terry E. Robinson. On June 3, Robinson informed me that the paper had been given a low priority rating and was not acceptable for publication. He cordially added, "I am sorry about this negative evaluation and hope it does not dissuade you from submitting future contributions to BBR."

The comments of one reviewer were highly reminiscent of those for my first submission to Llinás for Neuroscience, in which Reviewer 1 had responded that "the manuscript was overall favorable," whereas Reviewer 2 'vented his spleen' on it. While I eventually was successful with a revision in that ten-thousand-to-one circumstance, in which Llinás characterized Reviewer 2's evaluation merely as "rather stronger criticisms," and eventually discounted it entirely, further consideration was out of the question here. Below, I excerpt minimally from BBR's two reviewers. These, by comparison to one another, illustrate how subjective the review process occasionally becomes.

Reviewer 1. The paper is clearly written and advances interesting hypotheses.

Reviewer 2. This is yet another version of speculation regarding sleep, dreaming, memory, temperature, evolution, etc....author has already published 5 review papers on the same topic during the last three years without doing any original research....manuscript is written in a language that is incomprehensible for neuroscientists....author has invented a term "dynamic stabilization" which serves to substitute mechanisms, intermediate variables, etc., and it serves as a mystical tool to explain complex issues....abundant with unjustified claims and flat wrong deductions....neurophysiology section on dream amnesia is a list of unconnected sentences without any message or meaning

Adaptations and pathologies linked to dynamic stabilization of neural circuitry.

My interests increasingly were concentrating on the significance of my proposals concerning the origin, evolution, and function(s) of sleep for sleep and memory disorders and neurological pathologies. On June 22, 1998, I submitted a paper, "Some phenomena linked to the maintenance of memory circuits," to Dr. Ann E. Kelley, Coeditor of Neuroscience and Biobehavioral Reviews. This was my second submission to NBR. I received her reply on August 27, as follows.

Your MS, "Some phenomena linked...." has now been reviewed by two expert colleagues. I am pleased to inform you that these reviews were very positive and your paper will be acceptable pending the revisions that have been detailed by the referees. Both referees were impressed with the timely and thorough nature of your review. However, both also list several areas of concern that if addressed will improve the MS. I urge you to address these issues and suggestions, listing how you made changes in a cover letter.....

Again, I merely list areas of concern of the reviewers.

Reviewer 1, Where are the sensors that direct neuronal activity during sleep... what processes direct this activity....why do perceptual processes have to be blocked....how is it known that the patterns are synaptic enhancers...although there are no sure answers these valid points should be acknowledged and addressed.

Is there evidence that the advanced secondary functions are indeed secondary....would the author care to speculate on various points? Additional excerpted comments of Reviewer 1, which were intended to be in confidentiality to the editor but were included in error (together with his name) with the non-confidential material, were as follows. Of all existing theories of the function of sleep, the "DS" theory is one of the most exciting. Its credibility is supported by myriad animal observations. Importantly, this theory is the only one which has extended to experiments of nature in the human condition (fatal familial insomnia, and delirium with its treatment by electroconvulsive therapy). After minor revision as outlined in the comments to author section, this work will be of intense interest and importance, to your readership and should encourage close cooperation and collaboration between neuroscientists and clinicians in pursuit of one of the most important unanswered questions in neuroscience - "what is the function of sleep?" This theory should be made available to the scientific community for scrutiny, discussion, criticism, and refinement.

Reviewer 2. The author has done an excellent job in citing the noteworthy publications. An MS that deals with....diverse phenomena....requires masterful writing to counter reader-confusion. The author demonstrates that skill by writing an easily readable MS. Since maintenance of memory circuits is provided by a postulated mechanism, it should be mentioned in the title....clarification of experiential versus inherited memories....bilateral functions at higher levels within the avian brain....clarification of abnormal DS....suggest further discussion of visually featureless environments.

I submitted my revisions to Co-Editor Kelley on Sept. 8, 1998, including renaming the MS as suggested by Reviewer 2. I had intended to cease using the term "dynamic stabilization" in favor of two more explicit terms. These were "coordinated reinforcement," which is effected by fast waves (gamma waves), and occurs largely during REM sleep, and "uncoordinated reinforcement," which is effected by slow waves, and occurs largely during NREM sleep. In view of Reviewer 2's suggestion I retained its use for one more title, without being more explicit. But I was more specific about the "phenomena," namely, using "adaptations" and "pathologies." The revision, "Adaptations and pathologies linked to dynamic stabilization of neural circuitry," was accepted for publication on Sept. 14 and appeared in the May issue, 1999 (23:635-648).

The direction my studies had taken is illustrated by the paper's Abstract, as follows:

Brain circuits for infrequently employed memories are reinforced largely during sleep by self-induced, electrical slow waves, a process referred to as "dynamic stabilization" (DS). The essence of waking brain function in the absence of volitional activity is sensory input processing, an enormous amount of which is visual. These two functions: circuit reinforcement by DS and sensory information processing come into conflict when both occur at a high level, a conflict that may have been the selective pressure for sleep's origin. As brain waves are absent at the low temperatures of deep torpor, essential circuitry of hibernating small mammals would lose its competence if the animals did not warm up periodically to temperatures allowing sleep and circuit reinforcement. Blind, cave-dwelling vertebrates require no sleep because their sensory processing does not interfere with DS. Nor does such interference arise in continuously-swimming fishes, whose need to process visual information is reduced greatly by life in visually relatively featureless, pelagic habitats, and by schooling. Dreams are believed to have their origin in DS of memory circuits. They are thought to have illusory content when the circuits are partially degraded (incompetent), with synaptic efficacies weakened through infrequent use. Partially degraded circuits arise normally in the course of synaptic efficacy decay, or pathologically through abnormal regimens of DS. Organic delirium may result from breakdown of normal regimens of DS of circuitry during sleep, leaving many circuits incompetent. Activation of incompetent circuits during wakefulness apparently produces delirium and hallucinations. Some epileptic seizures may be induced by abnormal regimens of DS of motor circuitry. Regimens of remedial DS during seizures induced by electroconvulsive therapy (ECT) apparently produce temporary remission of delirium by restoring functional or 'dedicated' synaptic efficacies in incompetent circuitry. Sparing of sensory circuitry in fatal familial insomnia seemingly owes to supernormal circuit use in the virtual absence of sleep. ECT shocks and cardioverter defibrillation may have analogous remedial influences.

Reinforcement of memory during sleep, Chapter 6 in Toward a Theory of Neuroplasticity, C. A. Shaw & Jill C. McEachern, University of British Columbia, editors

On Dec. 8, 1998, I received an e-mail from the above editors, as follows.

Jill McEachern and I are now in the process of editing a book....which will be published by Taylor and Francis in early 2000. We were very impressed with your review articles on sleep and synaptic stabilization and hope you will agree to reprise these articles for the book.

I accepted the above invitation and submitted the final draft of my chapter in August, 1999. Because of publication delays and needs of all contributors to shorten contributions by 50%, the book did not appear until 2001. Below, I reproduce portions of the Introduction.

Introduction. The concept that brain circuitry benefits from sleep traces back to De Manacéine (1897), who stated with remarkable prescience that dreams "have a direct salutary influence insofar as they serve to exercise regions of the brain which in the waking state remain unemployed." After a lengthy hiatus, Moruzzi (1966) proposed that sleep facilitates recovery processes in circuitry for learned acts, while Roffwarg, Musio and Dement (1966) conjectured that spontaneous repetitive activations of circuitry in the central nervous system during rapid-eye-movement (REM) sleep in the human embryo facilitate circuit development and maintenance. They suggested that such activations during sleep could maintain circuitry throughout life. Since then, many studies have supported the view that the brain's repetitive, self-generated activations during sleep reinforce memory circuitry.

Concepts of memory circuit maintenance were formalized in terms of a paradigm of functional and nonutilitarian reinforcement (also referred to as "dynamic stabilization," reviewed in Kavanau, 1994, 1996). According to this paradigm, "dedicated" or functional synaptic efficacies become established initially by a tailoring of their values in the course of repeated circuit use for specific functions. These dedicated efficacies are maintained in the brain's long-term memory stores by two principal means: first, by continued frequent use during activity ("functional reinforcement"), and second, by self-generate activations induced largely, but not exclusively, by low-frequency oscillation potentials (slow waves) during quiet wakefulness and sleep ("nonutilitarian" reinforcement).

Mental malfunction and memory maintenance mechanisms

Wishing to reach a wider medical audience with my proposals concerning the significance of the origin, evolution, and function(s) of sleep for disorders of sleep and memory and neurological pathologies, on Feb. 3, 1999, I submitted a paper with the above title to Dr. David F. Horrobin, Editor of Medical Hypotheses. On June 4 Horrobin informed me that the paper had been reviewed, and that "This is an unusual and stimulating concept and we will be happy to have it in the journal.' The publication delay for this paper was, however, very lengthy, and it did not appear until the May issue of the following year (2000;54:678-683). Its Summary follows.

Summary. Dreams appear to be generated in the process of reinforcing memory circuits of the brain, as circuits are activated by self-generated electrical waves, with dream contents reflecting information stored in activated circuits. Illusory dreams and other healthy delirious states appear to occur when activated memory circuits are incompetent, containing synapses whose efficacies deviate from their 'dedicated' values. Organic delirium and some other mental disorders may have their basis in brain pathologies that alter reinforcing slow waves, causing synaptic efficacies to depart from dedicated values. Activation of these incompetent circuits leads to recall of faulty memories - a substrate for delirium. In treatment of organic delirium by electroconvulsive therapy (ECT), the electric shock temporarily suppresses abnormal slow wave regimes, allowing remedial reinforcement regimes to resume. These restore dedicated synaptic efficacies, temporarily alleviating the delirium. The action of ECT shocks appears to parallel closely that of cardiac defibrillating shocks. Greater than normal amounts of circuit reinforcement protect sensory circuitry in fatal familial insomnia, and cognitive circuitry in encephalitis lethargica.

In the meantime, a paper by Bruce G. Charlton, MD, of the University of Newcastle upon Tyne, Dept. of Psychology, titled, "The 'anti-delirium' theory of electroconvulsive therapy action," appeared in the same journal. Inasmuch as Charlton's views were consistent with mine, I contacted him and added a note in proof to my paper, as follows.

Based on, and in terms of, symptomatic and diagnostic criteria, B.G. Charlton recently has advanced an 'antidelirium' theory of the action of electroconvulsive therapy (Medical Hypotheses 1999;52:609-611) for which the proposals made here are supportive.

In following years, Bruce and I kept in fairly close contact and collaborated on a paper (see below). Bruce is a very prolific and original; researcher. Among his most recent books are, The Modernization Imperative, Psychiatry and the Human Condition, Palliative Psychopharmacology, and Self-Management of Psychiatric Symptoms.

A submission to Neuropsychiatry, Neuropsychology, and Behavioral Neurology.

I continued to concentrate my efforts in the areas mentioned above, but now also sought to bring them to the attention of clinicians. On March 2, 1999, I submitted an MS to Neuropsychiatry, Neuropsychology, and Behavioral Neurology.. Titled, "Illusory dream genesis: clues to a common path to many mental disorders," the MS was directed to Editor-in-Chief, Michael Alan Taylor, M.D. On May 19, I received his regret that "it is not possible to accept it for publication. He also enclosed reviewer comments "in the hope that you may find them helpful." I include illustrative excerpts from these reviews primarily to show the differing views that exist among specialists as regards the value of theoretical papers. The paper was of the same genre as all my prior and subsequent publications having to do with neuroscience topics.

Referee 1. This is a very interesting speculative, almost fanciful, piece on theories of dreams and delirium. To be publishable, it would have to be identified as such: i.e. opinion, speculation, a "thought piece." Also the absence of any reference to Freud is noteworthy. [Researchers continue to be divided about Freud's contributions to dreams and dreaming. In one camp are those who find some value in his works. In the other, which includes me, are those who find them almost entirely fanciful. JLK]

Referee 2. The author presents an interesting flight of fancy, seeking another explanation for the phenomenon of dreaming. He bases his explanation on a neurophysiologic model of the brain using terms like "self-generated electrical slow waves, " "illusory dreams and other healthy delirious states," and "synapses whose efficacies deviate from their 'dedicated values.'" ....they are not defined in the text with sufficient clarity to be comprehensible to a neurophysiologist....

Sleep, memory maintenance, and mental disorders

I continued in the above-stated vein of "the significance of the origin, evolution, and function(s) of sleep for sleep and memory disorders and neurological pathologies, and on bringing them to the attention of clinicians." On May 25, 1999, I submitted a paper with the above title to The Journal of Neuropsychiatry and Clinical Neurosciences. This was acknowledged with thanks by Editor Stuart C. Yudofsky, M. D. on June 3, informing me that "we will try to provide a decision for you within eight weeks. "On August 17, I received the further communication, "we are unable to accept your manuscript in its present form; however, if you would like to revise it according to the enclosed reviewers' comments, we would reconsider it for possible publication."  The reviewer comments in toto appear below. Bear in mind that, except for a few changes in the Abstract the submitted paper was the same one reviewed above for NNBN and characterized as "almost fanciful" and a "flight of fancy."

Reviewer 1. Excellent article, provocative and heuristic. Some suggestions:

1. The abstract suggests that the article is focused on dream mechanisms rather than memory maintenance and psychopathology. Suggest you reformulate the abstract to reflect this.
2. Not all readers will fully appreciate the concept of synaptic efficacy. You should spend some time expanding on this concept.

Reviewer 2. The audience for this journal will be interested in these concepts but many will not be closely familiar with the underlying work. The manuscript should be revised with this in mind. Some suggestions.

1. Include a clearer roadmap for the readers with clear statement of the purpose of the work and overview of the organization of the paper. Conclude with a summary.
2. More detailed explanation and empiric justification is required for key assumptions such as the generation of dreams and illusory dreams by the process of memory reinforcement.
3. Speculations should be sharpened and narrowed; e.g., it may suffice to talk about neuropsychiatric symptoms, rather than broader comment on disease states running from depression, mania, and delirium to Alzheimer's disease. Some of the evidence discussed relevant to specific disorders (e.g., ECT/delirium) may not reflect current consensus.

The MS was accepted on Nov. 22, 1999, and appeared as a SPECIAL ARTICLE in the Spring, 2000, issue. Its abstract and summary are reproduced below.

Abstract. Memory circuits of the brain are activated by self-generated brain waves, primarily during sleep. These activations refresh the efficacies (strengths) of synapses in affected circuits, maintaining the efficacies at the "dedicated" values that support circuit functions. The neural pathologies underlying many mental disorders appear to exert their deleterious influences by inducing abnormalities in brain waves. The abnormal waves, in turn, fail to sustain dedicated synaptic efficacies in memory circuits, leading to mental malfunction. Dreaming is an "unconscious" awareness of circuit reinforcement during sleep, with dream contents being derived from the activated circuits. When synaptic efficacies are degraded, the dreams are illusory.

Summary. Mentation during sleep states is thought to involve activation of the brain circuits that store and coordinate the use of inherited and experiential information (memories). All brain circuits are susceptible to spontaneous degradation of the efficacies (strengths) of their synapses because of turnover and other depletion of molecules essential for synaptic function. Synaptic efficacies in circuits employed frequently during waking are reinforced and maintained at their functional levels in the course of waking use by virtue of activity-dependent synaptic plasticity. During sleep, circuit activations induced by spontaneous, self-generated, primarily low frequency brain waves refresh synaptic efficacies in circuits employed infrequently during waking, also by virtue of activity-dependent synaptic plasticity. This process is accompanied by dreams whose contents are derived from the reinforced circuitry. In the absence of refreshment during sleep, synaptic efficacies in infrequently used circuits would decay and the circuits would become incompetent, their encoded memories degraded or lost. Maintenance of infrequently used circuitry during sleep is not 100% efficient. Efficacy losses accumulate in the synapses of some circuits, particularly in those for memories of the distant past. When these "incompetent" circuits are activated during sleep, illusory or bizarre dreams are produced. Activation of the circuits during waking can lead to faulty recall, illusions, and hallucinations. The organic pathologies underlying many mental disorders are believed to cause deviations of the brain's self-generated, circuit-reinforcing waves from their normal parameters (such as frequency, amplitude, and spiking properties) and distribution. When this occurs, the normal efficacy-refreshing actions become compromised to varying degrees and circuits become partially degraded. They remain degraded for as long as the waves remain abnormal and the underlying pathologies persist. Activation of these incompetent circuits during waking and sleep leads directly and indirectly to some of the pathologies that accompany many mental disorders. Inasmuch as many brain waves appear to be intrinsic to brain regions, when the shock of electroconvulsive therapy brings abnormal brain waves to a halt, more nearly normal brain waves of sleep spontaneously resume and temporarily gain ascendancy. This action is believed to parallel that by which a defibrillating shock halts fibrillations and allows the normal heartbeat rhythm to resume. Examples are given for the remediative influences of ECT on several mental disorders. Similarly, efficacious psychoactive drugs are thought to bring about remedial modifications of the abnormal brain waves of psychotic patients, leading to more nearly normal refreshment of synaptic efficacies in degraded circuitry. This action is thought to bring about partial to complete, but temporary, restoration of circuit competencies, accompanied by remedial influences on the mental disorders, sleep disorders, seizures, and other dyskinesias.

Efforts to publish semi-popular articles

At about this time I decided to try my hand at semi-popular articles in areas covered by the above JNCN paper. Although these efforts were unsuccessful, at first, as described below, they ultimately led to a most unusual result. This took the form of another technical article in a specialty journal, Neuropsychobiology, plus a following invited, independent Commentary on my article by other researchers.

In one effort, on Dec. 6, 1999, I submitted a paper, "Sleep, dreams, memory failures, and mental disorders," to John Rennie, Editor-in-Chief of Scientific American. Its receipt was acknowledged on Dec. 15, but on Feb. 1, 2000, I was informed that they were "in the midst of developing a large overview report on dreaming....this major treatment of the subject precludes our doing another major article on dreams...." [Such an "overview report" never materialized. JLK]

Accordingly, I submitted the same article to David Schoonmaker, Managing Editor of American Scientist, on Feb. 5, 2000. The reader may recall that in 1994 Schoonmaker had accepted my letter to AmSci on hydrophobic bonding. On May 3, I was informed that, "regrettably, we have decided that your submission does not meet our current editorial needs.

Through these same months I was working on another popular article titled, "Memory failures, illusory dream, and mental disorders." On Dec. 24, 1999, I submitted it to Ellen Goldensohn, Acting Editor-in-Chief of Natural History. Upon learning that the Editors did not find the paper suitable, on January 23, 2000, I sent the same paper to Robert Epstein, Ph.D., Editor-in-Chief of Psychology Today, also without success.

Memory failures, dream illusions, and mental malfunction

Determined not to let all the popularizing effort go to waste, I decided to rewrite the paper in scientific, rather than semi-popular style, adding references. Except for a much greater emphasis on dreams and lesser emphasis on pathologies, the material covered was the same as that in the American journal, JNCN. Accordingly, in about June 2000, I submitted it to the European journal, Neuropsychobiology. The latter is an International Journal of Experimental and Clinical Research, published by Karger, and edited in Basel, Switzerland. Unfortunately, my file for this journal was misplaced or lost on moving to my new office, so I depend only on my memory and the library.

I believe I submitted the paper to the Division of Biological Psychiatry. Receipt was acknowledged, and the paper apparently was then routed between Divisions in search of the appropriate one to review it. Months later, I received word from Prof. W.M. Hermann, Main Editor for the Division of Pharmacoelectroencephalography that the paper had been accepted, but there would be a further publication delay, (galleys arrived in June, 2001) because it was to be accompanied by an invited Commentary, not yet completed. There was no explanation for the accompanying Commentary. The paper appeared in the November, 2001 issue (44:199-211) (additional page charges of over $2,000 were waived). Its Abstract follows. I do not elaborate on the contents of the paper, as this has been done for the above JNCN paper.

Abstract. Dreams are widely believed to be produced as the brain's memory circuits are reinforced during sleep by self-generated brain waves. Reinforcement maintains synaptic strengths in the 'dedicated' ranges that support circuit functions. Without these activations, turnover of molecules essential for synaptic function would lead to deleterious alterations in, and eventual loss of, memories. The pathological processes underlying many mental disorders appear to exert their deleterious influences by inducing abnormalities in brain waves, largely in slow waves of less than about 14 cycles/s. The pathologically altered slow waves, in turn, cause long-lasting weakening or dysfunction of synapses of affected circuits, frequently resulting in mental disorders and deviant sleep. These abnormalities can be remediated for varying periods by therapies that restore normal brain waves. With many trillions of synapses between billions of neurons in enormously complex circuits needing reinforcement during sleep, the process is susceptible to failures. As a result, some synapses 'normally' weaken or become dysfunctional, accumulating to the greatest extent in old memory circuits. Activation of resulting incompetent circuits during waking may lead to hallucinations and delusions; activation during sleep may lead to dreams with bizarre, incoherent or impossible contents.

Following the paper, there appeared the Commentary, "Memory formation in sleep: giving a wave to dreams" (Neuropsychobiology 2001;44:212-214). This was authored by Profs. J. Born and U. Wagner, highly regarded sleep researchers at the Dept. of Clinical Neuroendocrinology at the Univ. of Lübeck, Germany. My guess is that the reviewers of the paper regarded it as sufficiently meritorious to suggest to the Editor that it be accompanied by a Commentary. This practice, of course, is not at all customary. I was greatly encouraged on learning of it, expecting perhaps some highly laudatory comments. Though the Commentary was favorable, on the whole, the authors gave no explanation for why Main Editor Hermann commissioned it. It is too technical to consider in detail, so I give only excerpts, as follows.

...Kavanau's review represents a timely and very stimulating contribution to the ongoing discussion on the central question: Why do we sleep? Its appealing quality clearly derives from the attempt to integrate very different perspectives on the function of sleep. In a very creative manner, Kavanau puts together a mosaic of sleep functions, referring to vastly different areas of sleep research, such as evolutionary biology, developmental psychology and electrophysiology, in order to infer eventually an essential role of sleep for the pathogenesis of mental disorders. This approach implies a high level of abstraction necessarily neglecting a fine-grained analysis of special EEG waves or of certain psychopathological phenomena....taking up a fundamental evolutionary perspective elaborated in previous reviews, Kavanau starts from the assumption that sleep has evolved from increasing processing demands encountered by higher organisms, which requires processes of memory consolidation to be postponed into a period of inactivity in which the respective neuronal networks are not preoccupied with acute processing of sensory inputs and motor outputs....This is a bold explanation for the development of mental dysfunction which, of course is in need of careful consideration....However, the essential issue remains to be explained, i.e., that some disorders are characterized by changes in mood, others by changes in thought, and again others by changes in motor behavior, but this issue is not addressed by Kavanau....Compared with other concepts on the function of brain waves as proposed, for example, by Sejnowski and Destexhe, Kavanau's propositions are on a much more general level, leaving open an exact specification of EEG frequency ranges and of the temporal order of events that might be relevant for the consolidation process....While his brain wave concept is not well refined and does not allow for specific hypotheses to be derived and tested experimentally, Kavanau's suggestions attract attention to the phenomenon of EEG slow waves, which currently appear most relevant for the memory functions of sleep....In this regard, Kavanau's concept stimulates investigation relating alteration in the sleep EEG in selected groups of psychiatric patients to signs of impaired long-term memory formation....this approach may not necessarily be successful.

A colleague's reaction to the Commentary, with which I agree, is given below. The only sense in which the paper was a review was that most of the topics discussed were my previous proposals.

....thanks very much for the excellent article (plus not so excellent commentary - which uses the rather dismissive term 'review' for a paper that (to me) is clearly theoretical biology - the review element is subordinate to the engagement with ideas).

At a later date I communicated with Profs. Born and Wagner, informing them that my paper, the subject of their Commentary, actually originated as an intended popular article. Failing that, I had converted it to a technical offering by rewording it and adding references. I sent them a reprint of the corresponding unwatered-down technical article, "Dream contents and failing memories," (see below) published the following year in Archives Italiennes de Biology.

Brain-processing limitations and selective pressures for sleep, fish schooling and avian flocking

To bring to biologists the implications of my sleep studies for other biological phenomena, on Feb. 15, 2001, I submitted an MS of the above title to Animal Behaviour, where I had not published since 1990. I also lack the file for this paper. After some revisions it was accepted on May 16, and appeared in the Dec. issue (62:1219-1224). I quote only one paragraph from the Discussion.

Although there have been impediments to elucidating the basic function of sleep in sleeping animals, compelling, but little pursued, clues to its function exist in nonsleeping vertebrates. These clues hinge on drawing comparisons with their close relatives that sleep. Contrasting lifestyles of two groups of relatively small, nocturnal reef-dwelling fish provide the most convincing evidence for analogizing restful waking and sleep with inactive schooling. Although these lifestyles have been known to ichthyologists for upwards of 40 years, the analogous features of the 'inactivity' periods attracted little attention. Pursuing their implications, it becomes evident that both lifestyles share the key feature of greatly reduced or minimal needs for complex visual input and analysis during 'inactivity'. These minimal needs allow unimpeded formation and reinforcement of memory circuits.

An attempt to publish a corrective note in the journal Sleep

In the year 2000, Dr. J. H. Benington published a paper in Sleep (23:959-966), titled, "Sleep homeostasis and the function of sleep" I had submitted my very first paper on sleep to Sleep in 1994. In the paper, Benington critiqued four current hypotheses concerning "the function of sleep." Inasmuch as his article contained much misinformation concerning my proposals, I submitted a corrective note titled, "On the function of sleep," to the Editor-in-Chief (who will remain unnamed) of Sleep on July 2, 2001. This led to a unique scientific journal experience for me. Even the highly regarded Editor's letter of receipt was unique in its professed, but unfulfilled, commitments. The Editor acknowledged receipt of the MS in July, including the following statements.

The focus of my editorship is to support science of the highest caliber, and to ensure prompt and thorough peer review of all submissions. It is through scientific integrity that the field of sleep has progressed so far in so short a time, and it will be through continued commitment to strengthening and solidifying our scientific base that sleep research and sleep medicine will prosper in the 21st century.

Your manuscript will be sent to reviewers, and you will receive further communication from us within 90 days.

The Summary of my note read as follows:

In a recent article (1) Benington dealt with sleep homeostasis and the function of sleep, critiquing "Four current hypotheses suggesting that the function of sleep is to assist in the process of activity-dependent synaptic reorganization...." In this note I call attention to misinformation in that  article concerning my proposals for the function of sleep and for a major brain activity thought to occur during sleep, and also review the proposals.....

I quote further from my note as follows:

The second misrepresentation of my hypotheses involves a finer distinction, wherein I do not equate the neural activities that occur during sleep to the function of sleep. Recalling Benington's earlier cited statement: "J. Lee Kavanau (1994): the function of sleep is to reinforce activity-dependent enhancements of synaptic activity by spontaneous....." Nowhere in the three cited papers nor elsewhere do I make this assertion. Rather than regarding DS [dynamic stabilization] as the function of sleep, I consider it to be merely a major activity that occurs during sleep. I postulate that the basic or primal function of sleep (the functions that provided the selective pressure for sleep's origin) is to provide a vigilance state in which DS of memory circuits and some other neural activities (e.g., consolidation of short-term memories) can occur without interference from, or interfering with, crucial moment-to-moment wakeful functions of the brain.

Submission of this note to Sleep was my first opportunity since 1995, when I reviewed my work in the Sleep Research Society Bulletin, to bring my then current proposals primarily before sleep researchers rather than primarily before neuroscientists. But this opportunity was seemingly intentionally denied me by Sleep's Editor as recounted below.

Having received no further word from Sleep four months and 8 days after submission of my Note, I e-mailed T. L. Meyer, the production Editor, on Nov. 21, asking about the current status of my MS #925. Receiving no reply, I e-mailed again a few days later. On Dec. 4, I received the following reply from Dawn Pfannkuch, the Manuscript and Subscription Coordinator.

Thank you for your emails and I apologize for the delay in my response. We were close to receiving a decision and I hoped to have a decision before I emailed you back and then your message got buried. A decision of "Reject" has been made on your manuscript and a decision letter and reviewer comments will be sent to you shortly. Again, I apologize in my lateness in responding and wish you a good week.

I thanked Dawn promptly, but still receiving no word I inquired again on Jan. 2, 2002. Her reply on Jan. 4 was as follows.

Thank you for your emails and I must apologize for my late response. The Holidays have been very busy and kept me away from my office quite a bit. I have now checked on the comments for your MS #925 and it appears that Dr. (the Editor) has yet to send these out and I have been trying to reach him in this regards with no luck. I will continue to try to get these comments sent to you and apologize greatly for this delay. I will e-mail you as soon as they are mailed. Thank you again for your emails and your patience.

Having received no further word from Dawn by Feb. 6, I informed her and inquired yet again. No further word was ever received. Ironically, the Editor of Sleep was the Director of a Research Center at Henry Ford Hospital, in my home town, and a Clinical Prof. at the U. of Mich., my Alma Mater. He had coauthored at least one paper with Benington. He apparently intentionally withheld reviewer comments from me. This was the only instance in which I encountered an editor whose conduct of his office appears to have been unprofessional.

Overtones of this behavior possibly account partially for a letter I received on March 13, 2003 from David P. White, MD, Editor-in-Chief of the Associated Professional Sleep Societies, LLC, and of the journal Sleep. He thanked me for previously submitting an MS to Sleep, and informing me as follows. To create an efficient process for manuscript submission and review that allows ease for all involved parties, the Societies were lunching a Manuscript Central, a Web-based system for manuscript submission and peer review. The former Editor-in-Chief of Sleep was no longer on the Board of Directors of the Societies.

Delirium and psychotic symptoms - an integrative model

I began corresponding with Bruce Charlton of Univ. of Newcastle in the early weeks of 2000, shortly after his paper in Medical Hypotheses (1999;52:609-611) came to my attention. After many exchanges of views we decided to publish a joint paper with the above title, with Bruce as senior author. Some unavoidable delays intervened before the paper was submitted in May, 2001. It appeared in the Jan. 2002 issue of the same Journal. Its Summary appears below.

Summary. Delirium may be a common cause of psychotic symptoms such as hallucinations, bizarre delusions and thought-disorder, even in conditions such as schizophrenia, mania and depression, where delirium has traditionally been excluded by definition. This situation is a consequence of the insensitivity of current clinical criteria for the diagnosis of delirium, which recognize only the most severe forms of functional brain impairment (including disorientation and clouding of consciousness). Serial electroencephalograms (EEGs) are the most sensitive method for detecting delirium, and until such studies are performed, the true incidence of delirium in psychotic patients will not be known. The suggested causal mechanism of delirium in psychosis is sleep disruption. Sleep is essential for maintenance of memory circuits, which otherwise suffer progressive synaptic weakening due to molecular turnover. When sleep is disrupted, memory circuits deteriorate, and subsequent activation of incompetent circuits can generate psychotic symptoms. Induction of physiologically normal sleep would therefore be expected to produce significant clinical improvement in patients with psychotic symptoms. Furthermore, the 'anti-delirium' action of electroconvulsive therapy may account for its effectiveness in alleviating a wide range of psychiatric and neurological pathologies.

As the senior author almost all correspondence and reprint requests, of which there were considerable (mostly via e-mail) went to Bruce. At last mention on Feb. 25, 2005, the paper on Bruce's website (www.hedweb.com/bgcharlton) was averaging about 20 hits per day.

Dream contents and failing memories

The paper of the above title is a more detailed and technical treatment of dreams and dreaming than given by me previously. The MS was submitted in late 2001 to Chief Editor O. Pompeiano of the quarterly Archives Italiennes de Biologie, published by the Univ. of Pisa, Italy. The paper was sent to referees and their comments were taken into account in a revision, but the file has been misplaced. I received page proofs on Feb. 8, 2002 and the paper appeared in the first issue of the year (140:109-127). My different approaches and points of view are illustrated by reproducing both the Introduction and the Summary, as follows.

Introduction. Dream theorists have proposed several explanations for the distorted and illusory (implausible, bizarre, or impossible) contents of dreams during REM sleep, when most dreams occur. In the activation-synthesis hypothesis of Hobson and McCarley they are thought to owe to the brain's generation of internal signals in an erroneous and unpredictable manner because of aminergic demodulation of cortical networks. In another explanation, dream illusions are believed to owe to an absence of fixation of temporal and spatial associations because of the lack of hippocampal involvement in image construction. In a third, the brain is said to lack the normal restraints for coordinating circuit activations, because external sensory input is absent. While in a fourth, the cause is thought to be impaired binding of circuits in the absence of real world feedback.

The proposals, however, make no allowance for the fact that the great majority of dream contents are authentic, that is, that they consist of thoughts and perceptions that are actually, or at least plausibly, within the realm of one's experiences. Additionally, they assume that basic neural processes operate flawlessly in producing illusory dreams, that is, that the disruptive alterations responsible for deviations from authentic dreams occur at higher neural levels than the strengths of synapses and competency of memory circuits. The goal of this paper is to present a novel view that accounts for the same phenomena, anchored in neurological relations that include the most recent experimental findings, and takes account of the fallibility of living processes. While not excluding the possible occurrence of disruptive alterations at higher levels, the basic neural processes are not assumed to operate flawlessly. Rather, it is proposed that there is a normal presence, and increase in numbers with age, of synapses with defective strengths (also referred to as "weights"), leading to incompetent circuits and faulty memories. When these circuits contribute to dreams, which occurs frequently, illusory and/or distorted contents result.

Summary. Mentation during sleep states is thought to originate in an activation of brain circuits that encode inherited and experiential memories. Spontaneous degradation of the strengths of synapses occurs in all brain circuits because of "turnover" of molecules essential for synaptic function. In circuits employed frequently during waking, synaptic strengths are refreshed and maintained in their dedicated or functional ranges largely through use, by virtue of activity-dependent synaptic plasticity. In circuits employed infrequently during waking, synaptic strengths are refreshed largely during sleep, by circuit activations induced by spontaneous, self-generated, largely low-frequency brain waves, also by virtue of activity-dependent synaptic plasticity. The outputs of circuits activated during sleep do not necessarily rise to the level of 'unconscious' awareness. Such an absence of awareness of the outputs of individual circuits, that is, an absence of dreaming, is proposed to be the primitive condition in animals that sleep. On the other hand, temporal binding of these outputs is accompanied by the thoughts and perceptions of dreams, which is proposed to be the advanced condition. Linking or serial ordering of otherwise 'static' thoughts and perceptions gives rise to continuous, often narrative and veridical, dreams. In all cases, dream contents are derived from the memories-not necessarily veridical--encoded in the reinforced circuitry. In the absence of synaptic strength refreshments during sleep, synaptic strengths in infrequently used circuits would weaken and the circuits would become incompetent, with their encoded memories degraded or lost.

Maintenance of synaptic strengths in infrequently used circuitry during sleep apparently does not always achieve perfection. Weakened synapses begin to occur in circuits in appreciable numbers in children after the age of about 5 years. When these 'incompetent' circuits (with weakened synapses) are activated during sleep, there are minimal influences on dream contents, namely, distortions that make some objects, such as animals, faces, and scenes, unrecognizable. As weakened synapses increase in numbers with age, the numbers of distorted objects in dreams also increase. In adults, people in as many as 80% of dreams may be unrecognizable. Besides the normal weakening of synaptic strengths, some synapses become defective, in consequence of deleterious, adventitious, exogenous influences, for example, radiation. As these faulty synapses accumulate in old memories, activation of circuits incorporating them during sleep leads to dreams with incoherent, bizarre, or impossible contents. The infrequent activation of such old, incompetent circuits in some waking conditions leads to false memories, delusions, or hallucinations.

Are birds half asleep on long-distance flights?

In late Feb., 2001, Wm James Davis, PhD, Editor of the Interpretive Birding Bulletin, invited me to contribute a Commentary (of 750-1,000 words), which received the above title. He had read my Commentary, "Brain-Processing Limitations....," in Animal Behaviour (2001;62:119-1224), and wanted to bring my proposals to the attention of his readers. A brief, very superficial treatment appeared in the June, 2002 issue of the Bulletin.

REM and NREM sleep as natural accompaniments of the evolution of warm-bloodedness

At about this time I decided to undertake a detailed analysis of REM sleep's origin. I asked the question, "how could a portion of a uniform NREM condition of primitive sleep have become specialized to yield REM sleep?" The answer was that it was not possible, that is, it is not possible for evolutionary forces acting on a uniform state to lead to a new state that is not also uniform. Some non-uniformity must exist in the initial state if a non-uniform state is to evolve from it. In other words, primitive sleep could, itself, not have been uniform.

This led me to conclude that an existing non-uniformity of primitive sleep must have been produced by environmental influences. Since, even today, sleep and body temperature are intimately interrelated, my attention became focused on temperature differences during sleep sessions as probably being the sought-after environmental influence. This was the main thesis of  this paper (see Abstract).

On July 9, 2002, I submitted a paper of the above title to Editor Ann E. Kelley of Neuroscience and Biobehavioral Reviews, my third submission to that Journal. It was acknowledged shortly thereafter. On Sept. 20, I received the following communication from Editor Kelley.

We now have received two reviews of your paper entitled, "REM and NREM sleep...." I am pleased to inform you that your paper is potentially acceptable for publication pending minor revisions. Both reviewers were quite complimentary and copies of both their comments are enclosed. Please respond appropriately to their suggestions and send two copies of a revised manuscript as soon as possible. Include....

Excerpted referee comments are as follows.

Referee 1. ....suggests to emphasize further that sleep's function of memory maintenance applies to long-term memory, not short-term consolidation....additional technical comments given with references.

Referee 2. ....well-written MS, carefully provides the sequences through careful argument that finally ends with the advent of REM and NREM states as we presently understand them....uses considerable literature from a number of fields, to exemplify conditions where the REM state (i.e., the presence of fast waves) is absent....provides a great deal of insight into the evolution of the REM state of sleep....discourage use of acronyms CR and UR, since already in use for conditioned and unconditioned responses....several clarifications needed.

The revised MS was returned on Oct. 2, accepted on Oct. 11 and appeared in the Dec. issue, close to a record time-to-print for me. The paper's Abstract follows.

Abstract. Divergence of primitive sleep into REM and NREM states is thought to have occurred in the nocturnal Triassic ancestors of mammals as a natural accompaniment of the evolution of warm-bloodedness. As ambient temperatures during twilight portions of primitive sleep traversed these evolving ancestors' core temperature, mechanisms of thermoregulatory control that employ muscle contractions became superfluous.

The resulting loss of need for such contractions during twilight sleep led to muscle atonia. With muscle tone absent, selection favored the persistence of the fast waves of nocturnal activity during twilight sleep. Stimulations by these waves reinforce motor circuits at the increasing temperatures of evolving warm-bloodedness without leading to sleep-disturbing muscle contractions.

By these and related interlinked adaptations, twilight sleep evolved into REM sleep. The daytime period of sleep became NREM sleep. The evolution of NREM and REM sleep following this scenario has implications for sleep's maintenance processes for long-term memories. During NREM sleep, there is an unsynchronized, uncoordinated stimulation and reinforcement of individual distributed component circuits of consolidated memories by slow wave potentials, a process termed 'uncoordinated reinforcement'.

The corresponding process during REM sleep is the coordinated stimulation and reinforcement of these circuits by fast wave potentials. This action temporally binds the individual component circuit outputs into fully formed memories, a process termed 'coordinated reinforcement'. Sequential uncoordinated and coordinated reinforcement, that is, NREM followed by REM sleep, emerges as the most effective mechanism of long-term memory maintenance in vertebrates.

With the evolution of this two-stage mechanism of long-term memory maintenance, it became adaptive to partition sleep into several NREM-REM cycles, thereby achieving a more lengthy application of the cooperative sequential actions.

It was in this paper that I emphasized that dynamic stabilization consists of two phases: "uncoordinated reinforcement," effected by slow waves, predominates during NREM sleep, and primarily reinforces individual component circuits of memories; "coordinated reinforcement" effected by fast waves, predominates during REM sleep, and primarily links component circuits of memories into the fully-formed memories. As emphasized in a later paper (see below; "Evolutionary approaches to understanding sleep"), both types of reinforcement, and both types of brain waves perform both types of actions, merely to different degrees at different times. This reflects the circumstance that natural selection typically is not an "all-or-none" process, a relationship too little appreciated among non-evolutionists.

It was the appearance of this paper that led Editor Mark Mattson, on Aug. 20, 2003, to invite me to contribute the lead chapter to the book, "Sleep and Aging" (see below). In March, 2005, I learned through correspondence that C. J. Bae and N. Foldvary-Schaefer had used this article as their main reference for 'memory reinforcement' in their chapter on "Normal Human Sleep" in the book, Clinical Sleep Disorders, edited by Carney, Berry and Geyer, Lippincott, Williams and Wilkins, 2005.

New clues to the origin and functions of REM and NREM sleep

During recent years I had noticed that reviews and other contributions by leading sleep researchers were appearing regularly in Sleep and Hypnosis, a new journal published quarterly in Istanbul, edited by M. Y. Agargun, M.D. All was clarified when Editor Agargun, on Oct. 15, 2002, invited me to contribute an article (eventually an Editorial) to the Journal on my current work. Upon acceptance, he also invited me to join the Advisory Board. An Editorial with the above title appeared shortly thereafter (2002;4:85-92). The content of this paper was merely a brief recap of the above paper in NBR.

Sleep researchers need to bring Darwin on board: elucidating functions of sleep via adaptedness and natural selection

On April 1, 2003, Founding Editor David Horrobin of Medical Hypotheses passed away and Bruce Charlton of Univ. of Newcastle succeeded him. Bruce invited me to contribute an Editorial on my studies in sleep to The Journal, and also to join the Editorial Board. An Editorial with the above title appeared in the following February issue (2004;62:161-165). Its Summary and concluding paragraph follow. In them and the remainder of the paper I exercised editorial license, that is, I freely voiced my views on the shortcomings I perceive in many current approaches to understanding the origin and function(s) of sleep.

Summary. The development of neural multifunctionality--given brain regions carrying out more than one function -- conferred great efficiency on brain function at early stages of evolution. This applied to animals that led relatively simple lives with few needs for long-term memories, such as many lower invertebrates-many molluscs, echinoderms, worms, etc. As more complex lifestyles and detailed focal vision evolved, needs for self-initiated and reflexive activities increased in frequency, and recognition of many locales, conspecifics, and other forms of life became essential. These developments were accompanied by greatly expanded needs for neural processing supporting sensory and motor activities, and establishing and storing long-term memories. Since these categories of neural processing occur in largely overlapping brain regions, brain functioning would have become increasingly maladaptive, had the evolution of these more complex lifestyles not been accompanied by compensating adaptations that obviated these potentially conflicting brain activities. These adaptations consisted of: first, restful waking; second, primitive sleep; and finally, fully developed sleep, with its specialized rapid-eye-movement (REM) and non-rapid-eye-movement (NREM) states, that contribute to the maintenance of great efficiency of brain function. The only animals with detailed focal vision that can achieve highly efficient brain function without sleep, are those in which demands on memory processing are greatly reduced in consequence of routine, monotonous, almost purely reflexive lifestyles, with few needs for acquiring experiential long-term memories. The best known animals in this non-sleeping category are tunas and many sharks.

On evolutionary approaches to sleep. In this editorial, I have sought to illustrate the utility of an evolutionary approach-always seeking and mindful of differential adaptedness and underlying selective pressures -- to guide research into the origin and functions of sleep. It is of paramount importance in such pursuits to recognize that physiological and behavioral adaptive responses to many types of selective pressures are basically highly conservative, usually occur gradually and continuously, and typically are restricted to modifying or co-opting existing substrates, rarely effecting new mechanisms. It is equally important to be mindful of the often-overlooked circumstance that, though counterexamples unequivocally invalidate mathematical and logical theorems, they do not enjoy the same power of discrimination in the much more complex biological realm. Otherwise, since biological counterexamples are by no means rare (for example, sighted fishes that do not sleep), one may be too easily diverted from pursuing valid lines of analysis. Evolutionary processes sometimes achieve adaptations that, on the face of it, appear to be unattainable.

More than the usual number of e-mail reprint requests for this paper were forthcoming, as well as two invitations to contribute articles. Editor Charlton informed me that In the first five months of the article's appearance, there were approximately 800 downloads of it from The Journal's website.

Evolutionary approaches to understanding sleep

In the month following the appearance of the above MeHy Editorial (on March 26, 2004) I received an invitation to contribute a Theoretical Review of the above title to Sleep Medicine Reviews. The invitation was from Editor-in-Chief (for the Americas) Michael Vitiello, PhD. Interestingly, the Editor-in-Chief (for Europe and the rest of the world) is Jean Krieger, MD, who had requested reprints of most of my sleep research papers going back to 1994. I was advised that:

A focus on sleep per se is certainly appropriate, but if you could also consider certain aspects of sleep disorders such a finer focus would be of particular interest to our readership. I would hope that you could provide such a review by October 1st,  although a somewhat earlier or later submission date could certainly be arranged if that would better suit your schedule. All invited reviews are evaluated by two outside referees and the Coeditors-in-Chiefs may require revision before publication.

I accepted this invitation and submitted my MS on the following June 21, including an analysis of sleep's two most extreme sleep disorders, namely, fatal familial insomnia and encephalitis lethargica. This was patterned after my treatment, "Adaptations and pathologies...." in NBR (1999;23:635-648), which was ideally suited for the SMR readership. On the following Sept. 27, Editor-in-Chief Vitiello informed me that the MS had been reviewed by two referees and was acceptable for publication in SMR pending adequate revision. I completed and returned my revision on Oct. 17, it was accepted on Oct. 29, and appeared in May (2005;9:141-152); online beginning Feb. 8).

However, the review process was not routine, nor without interest. To begin with, the referees had not been informed that the article was intended to be a theoretical review. Instead, they critiqued it as a clinical review, which is measured against quite different criteria. At any rate, the paper benefited greatly from constructive comments and issues raised by the reviewers.

I reproduce the paper's Summary in the following, with the addition of part of a section of the paper which needs particular emphasis.

Summary. A major controversy over REM sleep's role in memory processing may owe to inadequate allowances for the highly conservative nature of evolutionary adaptations. The controversy hinges on whether NREM sleep, alone, retains primitive memory processing capabilities.

The selective pressure for primitive sleep, is thought to have been the need to obviate conflicts between enormous neural processing requirements of complex visual analysis and split-second control of movements, on the one hand, and memory processing, on the other. The most efficient memory processing during mammalian and avian sleep appears to be a two-step process: synapses in individual component circuits of events are reinforced primarily by slow brain waves during NREM sleep, with the reinforced components temporally bound by fast waves, and manifested as dreams, during REM sleep.

This dual action could account for partitioning of sleep periods into multiple NREM-REM cycles. It is proposed that in the absence of REM sleep, all needed memory processing can be accomplished by NREM sleep, alone, though less efficiently.

Many symptoms of fatal familial insomnia are attributed to subnormal nightly reinforcement of brain circuitry because of almost total loss of sleep, and compensatory responses thereto during waking. During this disorder, sensory circuitry seemingly is spared by virtue of its supernormal reinforcement during almost continuous waking. Contrariwise, sparing of an adult's "higher faculties" in encephalitis lethargica appears to owe to supernormal circuit reinforcement during almost continuous sleep.

Conservativeness and predominantly 'non-all-or-none' characteristics of evolutionary processes. In essence, then, the most efficient memory processing in mammals and birds is proposed to be a two-step process: individual component circuits are reinforced primarily by slow waves during NREM sleep, while the reinforced circuits are temporally bound by fast waves. acting over the entire cortical mantle and other regions (see [25]), during REM sleep. In the absence of REM sleep, the entire process can be accomplished by the primitive mechanisms of NREM sleep, alone, merely less efficiently, a relationship that renders it unlikely that deprivation studies would lead to insight into REM function.

Following these proposals, the basic function of REM sleep and fast waves, that "proverbial riddle wrapped in a mystery inside an enigma," would appear to be to increase the efficiency of memory processing during sleep -- of which dreams would simply be the overt manifestation.

These proposed actions of REM and NREM sleep are consonant with the conservativeness, and predominantly 'non-all-or-none' characteristics, of evolutionary processes. To a degree, the new state of REM sleep possesses some slow waves, carried over from the more primitive NREM state, together with their abilities to carry out the main function of NREM sleep (uncoordinated reinforcement), though less efficiently. But, it carries out the 'new' function (coordinated reinforcement), highly efficiently.

Likewise, NREM sleep acquired small amounts of the fast waves of REM sleep, and their capability for coordinated reinforcement. But it also retains a less efficient, primitive capability for coordinated reinforcement by slow waves, which becomes the sole mechanism for subserving that function when REM sleep and fast waves are totally lacking. NREM sleep remains highly efficient in uncoordinated reinforcement. The latter is thought to be a gradual process, probably occurring in many subthreshold steps, and with 'dedicated' value's of synaptic strengths covering a range, the "synaptic modification range"

Wozu eigentlich schlafen? (Why do we sleep?)

In the month following the appearance of the above MeHy Editorial (on March 22, 2004) I also received an invitation to contribute an article on the question, "why do we sleep?" to a new German magazine on medicine and culture, Leib und Leben (Body and Life). The invitation was from Martin Lindner, MD, a reporter on science and society for a major German and Swiss print publications. The invitation included:

I am preparing a special issue on sleep....Would you write such an article for our magazine? The text could be based on your editorial, however, it should be somewhat more popular, encompass some more examples in animals for your principal hypothesis and refer, where possible, to humans as well.

I was most pleased to receive and accept this invitation, which would have been only my second opportunity to present my proposals concerning the evolution and basic function of sleep in popular form, though not in English. Recall, my first opportunity led to a brief, very superficial treatment, "Are birds half asleep...." in the Interpretive Birding Bulletin in June, 2002. The article appeared in the Sept.-Oct., 2004 issue Leib und Leben (pp. 50-52). Suffice it to say that the Editor cut out the thread and much of the substance of the article, leaving it twelve unconnected paragraphs, and a total disappointment to me. He gave me no warning that the text would be altered, replying, "thanks a lot for your article which I like very much. I think it will fit nicely into our sleep issue." 

Evolutionary Aspects of sleep and its REM and NREM states.

On Aug. 20, 2003, Mark P. Mattson, PhD, editor of the series, Advances in Cell Aging and Gerontology invited me to contribute the lead chapter to the forthcoming volume Sleep and Aging on the topic of "Evolutionary Aspects of Sleep," due by Feb. 16, 2004. I submitted my MS on Jan. 26, and it was duly accepted. Being somewhat delayed in publication, the book did not appear until mid 2005. Although more complete in some respects the Chapter contains no significant new information, and only a listing of Contents is given below.

Contents

1. Introduction
2. Evolutionary origin of sleep

2.1. Schwartz's admonition and Rauschecker's "Fundamental Dogma"
2.2. Restful waking, detailed focal vision, and conflicting brain activities
2.3. Primitive sleep obviated potentially conflicting brain activities
2.4. Vertebrates that never sleep
2.5. Need for sleep in congenitally and adventitiously blind mammals
2.6. Earliest sleep
2.7. Complex retinal processing

3. Warm-bloodedness and selective pressures for REM and NREM sleep

3.1. Historical
3.2. Sleep was non-uniform before the origin and REM and NREM sleep
3.3. Presence and absence of thermoregulation during sleep states
3.4. Implications for the evolution of sleep states
3.5. Influences of ambient temperature
3.6. Actions of brain waves during waking
3.7. Actions of brain waves during sleep
3.8. Long-term memory maintenance
3.9. Basic links between muscle-controlling and visual circuits
3.10. Absence or lesser amounts of REM sleep in marine mammals and birds
3.11. Origin of fast waves of REM sleep

4. Evolutionary origin of sleep within a brain-wave paradigm
5. Sequential cycling of NREM and REM sleep
6. Overview

Is sleep's 'supreme mystery' unraveling? An evolutionary analysis of sleep encounters no mystery; nor does life' earliest sleep, recently discovered in jellyfish

In Dec., 2004, after the comparatively great success of my "Sleep researchers need to bring Darwin on board"-Editorial in MeHy, Editor Charlton invited me to contribute another. I replied, "I'm flattered that you believe I have enough wisdom in reserve to make up another successful sleep editorial. Offhand, I don't think I do." But only a few weeks later, I learned of an article in the Medical Journal of Australia (2004;181:707) by Seymour, Carrette, and Sutherland, titled "Do box jellyfish sleep at night?"

Using radio telemetry with tagged animals Seymour et al. obtained evidence that under certain conditions box jellyfish sleep 15 hr per day. Although, as yet unconfirmed, these findings in animals with only two germ layers cast new light on the basis for sleep. Although box jellyfish have abundant focal vision, including eight eyes of complex camera-type, their vision may not be detailed. Moreover, their nervous system is comparatively simple. This raises the possibility that the basis for sleep is more general than heretofore believed. Accordingly, I submitted another editorial treating these matters. Its thrust is elucidated by pertinent excerpts from the Summary and Overview, as follows.

Summary. Biotelemetry has revealed daily 15-hr behavioral sleep periods in a cubomedusan jellyfish, Chironex fleckeri. Its sleep is expected to be phylogenetically most primitive, since jellyfish possess only two germ layers. They belong to the phylum Cnidaria, the 'simplest' multicellular organisms with an organized nervous system.

 Cubomedusae have a complex visual system with 24 eyes of four different types, each type specialized for a different task. Input to these eyes during visually-guided fast-swimming predation requires enormous amounts of neural processing, possibly nearly saturating the capacity of their comparatively simple nervous system.

These heavy neural demands may account for the need for fifteen hours of sleep. C. fleckeri is the only animal known for which sleep may be either present or absent, dependent on lifestyle. Limited knowledge of behavior of some other cubomedusae suggests that they also possess this faculty. The finding of sleep in C. fleckeri supports current proposals of sleep's origin and basic function.

Evolutionary analyses link sleep to a conflict produced by excessive processing demands on multifunctional neural circuitry for detailed focal vision by complex lensed eyes. The conflict arises between the enormous demands of complex visual analysis and needs for split-second control of actions, on the one hand, and non-urgent processing of memories of ongoing and stored events, on the other. Conflict is resolved by deferring the non-urgent processing to periods of sleep. Without sleep, selection would favor the evolution of circuitry 'dedicated' to single or but few tasks, with corresponding lesser efficiency.

Overview. Behavior of medusae seemingly provides a window into the basis for sleep's earliest function and evolution, as proposed above. Possessing the first organized nervous system, following those proposals, an implication is that neural multifunctionality exists in cnidarians, and played a role in the evolution of sleep.

Since both sleeping and non-sleeping medusae exist, the presence of complex lensed eyes, the more complex lifestyles, and the greater needs for neural processing in those that sleep, become implicated in the need for sleep....

In essence, the condition determining the need for sleep in cubomedusae appears to be the degree to which the demanding neural processing needs for incoming 'visual' information and visually guided actions saturates their processing capacity. Accordingly, it can be suggested that, even the neural processing needs for vision with complex lensed eyes may be too restrictive a condition on the mode of light reception, as regards the need for a second vigilance state.

Taking the broadest view in unraveling the "supreme mystery," one can suspect that circumstances in cubomedusae constitute extremes of those that lead to the need for a second vigilance state in other medusae. Even lesser neural processing demands that monopolize the capacity for light-sensing and responding in nervous systems of lesser capacity, might lead to a requirement for rest or sleep.

Supporting this view, the following observations confirm rest -- possibly sleep -- in medusae of Aurelia aurita, in the class, Scyphozoa, which have only two tiny, non-lensed ocelli, of different composition on each of their eight rhopalia. "In the dark, the medusae were quiescent, with bell contraction rates reduced or absent....on still nights medusae covered the surface....animals were almost motionless..."

On Oct. 6, 2006, recommended by my former student, Dr. Donald Perry, who also was contributing, I received the following invitation from Dr. H. James Birx, Editor of the SAGE Encyclopedia of Time:

Please consider this invitation to write the entry or entries listed below for the forthcoming SAGE Reference publication: "Encyclopedia of Time." Below you will find an introduction to the project and instructions for responding to this invitation. I hope that you will be able to join me in the creation of this authoritative reference text.

The entries in question were to be on "Sleep" and "Dreams." I accepted his invitation and submitted the entries shortly before July 9, 2007, the date on which they were accepted. The entries are now in press. In these essentially popular entries for an encyclopedia, in which the Editor desired that the role of time be emphasized, I took somewhat greater speculative liberties than I would have for a regular Journal article. I reproduce the first two paragraphs of each entry below:

Sleep (2,657 words). From time immemorial the ultimate function of sleep has occupied the attention of scientists, philosophers, and physicians, though little basic understanding emerged. In recent times, J.A. Hobson, a leading sleep researcher from Harvard, succinctly observed that sleep is of the brain, by the brain, and for the brain. Recent approaches from the perspectives of efficiency of brain function and natural selection are unraveling sleep's longstanding mysteries, particularly how sleep is "for the brain."

One obtains the greatest insight into the rationale of a biological act or state through elucidating its evolution, namely, why, when, and how it came into existence? With this inherent advantage of an evolutionary perspective, sleep's ultimate function now presents little more mystery than many other phenomena of life that increasingly are yielding their secrets. In fact, it is now evident that time, in the sense of simultaneity, played a pivotal role in sleep's origin.

Dreams (2,775 words). Time plays a pivotal role in dreams. Much of the fantasy, incoherence, inaccuracies, and ambiguities of dream contents owes largely to continuous weakening of the strengths of connections ("synapses") between nerve cells of the brain with time. Considerations here relate mainly to this weakening influence. Emphasis is on the dreams of children, which are expressed in their 'purest' form, minimally encumbered by the complicating influences that accumulate with age.

Dreams are the accompaniments or by-products of certain essential activities of the brain during sleep. They primarily occur as neocortical circuits (the most recently evolved brain circuits) become activated by spontaneous, self-generated, complex electrical oscillations. Superficially, these oscillations are expressed as the slow and fast scalp-waves of EEGs (electroencephalograms). A major role of dreams during sleep is the processing of phylogenetic and experiential memories, that is, inherited memories and memories of past waking events.

On  March 9, 2007, there appeared in Science, in the NEWS FOCUS section, an article titled, "Hunting for Meaning after Midnight," in which science writer Greg Miller summarized sleep researchers current views and disagreements on the function of sleep. Since, like all other material that had appeared in Science over the years, Miller's write-up was entirely medically oriented, I decided that this was an excellent opportunity to get something into print in Science presenting the evolutionist's point of view on sleep's origin and ultimate function.

I had recently been in touch with Etta Kavanagh, Science's Letters Editor, in another connection. I decided to send her a Letter to the Editor. However, she was not aware of my studies of sleep, so she would not be likely to publish a letter from me critical of numerous familiar-to-her, medically-oriented sleep researchers. What was needed was an introductory private letter to her explaining the situation, and including my credentials as a sleep researcher. For that reason I also included the following private letter to her on March 30, 2007, together with my bibliography in the field of sleep:

My regards from California. I hope it is permissible to contact you directly and include correspondence when submitting a Letter to the Editor to Science. I feel I should inform you of the background of the submission.

I wish to bring to your attention that the recent NEWS FOCUS, "Hunting for Meaning After Midnight" (9 March), presents entirely one-sided views on topics relating to the function of sleep, namely, only those of medically-oriented sleep researchers who disagree, basically, among themselves.

Since evolutionary perspectives on the function of sleep (my area of expertise) are entirely lacking in this piece, readers of Science receive a decidedly lop-sided presentation of current views in the field. One seeks in vain for the words "Darwin," "selective pressure," "adaptation," etc., among the works of the featured sleep researchers. In fact, their use of the word, "evolution," goes no further than recognizing evolution's existence. This, despite the fact that the greatest insight into the rationale for the existence of any biological act or state is obtained through probing and elucidating its evolution, namely, why, how, and when it came into existence.

For those reasons, I am submitting the attached Letter to the Editor to Science. In presenting current evolutionary perspectives, and contrasting them with those of medical researchers, it could provide some balance.

Concerning the matter of whether my views on sleep carry the authority to back up the contents of my Letter, I attach my Sleep Bibliography. It consists of 21 contributions of the last 13 years in 16 different journals, books, and magazines. Ten of these contributions were invited by the editors, including 2 book chapters (12,20), 3 reviews (2,5,19), 3 editorials (16,17,21), and 2 popular articles (15,18).

I hope the Letter is in order and will be judged to be suitable. Thank you for your consideration.

It was not until two full months later, on May 31 that I heard from Etta through her assistant, Brian White, as follows:

On 30 March, you submitted a Letter to the Editor of  Science in response to the NEWS FOCUS article titled, "Hunting for Meaning After Midnight." We are unable to print it, as we have very limited space for letters, but we would like to run it as an E-letter online. If this is acceptable to you, please let me know.

Although greatly disappointed, I replied in the affirmative. On July 6, the following E-letter appeared online at sciencemag.org/cgi/eletters/315/5817/1360:

Function of sleep, and why experts disagree

The recent NEWS FOCUS (9 March), "Hunting for Meaning After Midnight," summarizes recent views of medically-oriented sleep researchers, including their disagreements. Largely at the root of the disagreements is their equating of the ultimate function of sleep, that is, the function responsible for sleep's origin, to neural events that occur during sleep.

Thus, each of the researchers studying neural events of sleep disagrees with most of the others. Each proposes that his or her particular event is the function of sleep. They make no distinction between whether the proximate or ultimate function is meant, but the ultimate function can be assumed, because the long-term goal of most sleep researchers is to solve the "supreme mystery" -- the enigma of sleep's origin. Proposed functions include: active learning, identifying rules of cause and effect, finding missing connections, consolidating memories, strengthening memories, and improving waking performance. One or another combination of such actions would comprise what the evolutionist regards as the proximate functions of sleep.

The key to elucidating sleep's ultimate function concerns why a second vigilance state became necessary to accommodate certain activities of the brain. Why couldn't these activities, like the others, simply have been accommodated during waking? I suggest that probing the "supreme mystery" from Darwinian evolutionary perspectives would be most productive. Thus, an evolutionary analysis might lead expeditiously to sleep's ultimate function, proceeding as follows.

The evolution of sleep made it possible, for the first time, to defer some brain activities to a second vigilance state. This new flexibility probably achieved or maintained maximal efficiency of brain operation. One would expect sleep to have evolved in animals whose waking brains could no longer efficiently support both activities critical for survival and large, increasing amounts of non-urgent activities. At that time, some non-urgent activities, of which memory processing would have been a leading candidate, would have been deferred to the new sleep state.

It remains to identify the critical activities. Animals that move very slowly -- mussels, sea anemones, starfish, many worms, etc., need no sleep (1). However, all fast-moving animals [with rare, reconcilable exceptions (2)] need sleep. Accordingly, one critical activity of the waking brain during natural selection for sleep probably supported rapid movements. Since rapidly-moving animals have complex eyes or eyespots, another critical activity during selection for sleep probably supported vision (1, 2). Accordingly, when non-urgent neural activities began to conflict with the brain's waking processing for fast movements and vision, the conflict probably was avoided by deferring the conflicting, non-urgent activities to the new sleep state.

Accordingly, the ultimate function of sleep probably is to enable brains to operate highly efficiently at all times. During sleep, brains can engage lengthily in non-urgent functional activities without danger of conflicting with critical activities of the waking brain.

J. Lee Kavanau

References

1.  J. L. Kavanau, Neurosci. 79, 7 (1997).

2.  J. L. Kavanau, Sleep Med. Rev. 9, 141 (2005).

I have not received any correspondence in connection with this E-letter

In the next event of note, on June 5, 2007, I received an invitation from Prof. S.R. Pandi-Perumal of Columbia U. I was familiar with his name from his requests for my reprints over the years. He invited me to contribute a chapter to a forthcoming volume titled, "Sleep and Circadian Visual Neuroscience," to be published by Cambridge University Press in 2008, and edited by Per O. Lundmark (Buskerud U. College, Norway), Pandi-Perumal, D.P. Cardinali (U. of Buenos Aires), and G.C. Brainard (Jefferson Medical College).

I emailed my acceptance to Lundmark on June 5, 2007, as requested by Pandi-Perumal, suggesting the title, "Vision and the Evolution and Ultimate Function of Sleep." Having received no reply or acknowledgement over one month later, I followed up with a query and duplicate of my acceptance email. Having received no reply to that email, either, I simply let the matter stand. I was fully occupied with the Sleep and Dreams entries mentioned above, another chapter discussed below, and my autobiography, so I welcomed the delay or possible collapse of plans for the subject book.

On July 12, 2007, while the above events were transpiring, I received another invitation to contribute a chapter to yet another book, also contracted to be published by the Cambridge University Press. The invitation came from Co-Editor, Prof. P. McNamara of the Boston U. School of Medicine for a book titled, "Evolution of Sleep: Phylogenetic and Functional Perspectives." Co-Editors are R. Barton (Durham U.) and C. Nunn (Max Planck Institute, Leipzig). The invitation read:

Your distinguished work in the field of sleep has prompted us to contact you about a book....We are writing to ask if you would be interested in providing a chapter in this book on some aspect of sleep evolution. We would be particularly grateful if you could review what is known about sleep or lack of it in fish. We know that this may not be your primary interest but there are few scientists who have written on this as well as you have.

They were referring to my Review Article, "Vertebrates That Never Sleep.....," in Brain Research Bulletin, (1998;46:269-279) in which I treated non-sleeping blind cave fishes, sharks, and teleosts (see above). At that time, however (July, 1996 to Dec., 1998), I also was privately publishing a monograph titled, "Memory, Sleep, and the Evolution of Mechanisms of Synaptic Efficacy Maintenance." In this, I kept my Neuroscience Commentary (1997;79:7-44) of the same title up to date, and was mailing it to my reprint mailing list. It included much more detail about sleeping and non-sleeping fishes.

Accordingly, the McNamara invitation provided the opportunity to publish some of this unpublished, privately circulated material. So I was delighted to accept it. Additionally, the habits of many schooling and other non-sleeping fishes, as compared to those of fishes that sleep, provide direct evidence of sleep's functions, both ultimate and proximate. Since this direct evidence strengthens the case deduced indirectly for these functions derived from studies of terrestrial animals, the opportunity also was provided to strengthen the indirect case. On July 20, I accepted the invitation. On Dec. 10, I submitted my preliminary version titled, "Schooling by Continuously Active Fishes: Clues to Sleep's Ultimate Function," the Abstract of which follows below:

Abstract. Aquatic habitats were the last to be probed in seeking sleep's ultimate function, namely, enabling the brain to operate with high efficiency at all times. Yet these domains were the cradle of sleep, many millions of years before sleep evolved in terrestrial animals. Behaviors in aquatic habitats are found to offer compelling clues to sleep's origins. Overtly, the sleep of most fishes is essentially indistinguishable from that of terrestrial vertebrates. Yet, by virtue of the rich variety and great permissiveness of aquatic habitats, some fishes appear never to have encountered selective pressures for sleep. It is remarkable that three continuously active states of perpetual vigilance exist in these fishes, in which they achieve comparable, and even greater, benefits than accrue to animals that sleep.

"Schooling" (swimming synchronously in polarized groups) by many of these fishes plays the major role in the lack of a need for sleep. Through schooling, they can achieve sleep's benefits without losing consciousness or closing or occluding their eyes, otherwise usually one of sleep's key accompaniments. Since the evident benefits of schooling by some fishes overlap the obscure benefits of sleep by other fishes, the evident benefits give clues to the obscure ones. These clues support views on the ultimate function of sleep in terrestrials. Even some continuously active but non-schooling, fishes (some "pelagic cruisers") also achieve highly efficient brain operation at all times, illustrating the exceptional permissiveness of pelagic environments (open oceans).

After reviewing circumstances relating to the evolution of sleep in terrestrial animals, relevant topics in the morphology, ecology and behavior of teleost fishes are treated, including both their sleep states and continuously awake states.