Research moves on all the time, and new findings confirm and expand on old ones, or sometimes modify or contradict them. Here I aim to provide a guide to some of the most interesting new discoveries and ideas relevant to consciousness and its evolution, and to any particularly interesting new books on the subject. There will also be the occasional thought that has occurred to me since the book went to press.
Curr. Biol. Current Biology
Nat. Neurosci. Nature Neuroscience
Nat. Rev. Neurosci. Nature Reviews: Neuroscience
PNAS Proceedings of the National Academy of Sciences
TICS Trends in Cognitive Sciences
TINS Trends in Neurosciences
It’s known that the report from a sensory area of the neocortex must activate prefrontal cortex if the sensory input is to reach consciousness. However similar levels of input don’t always produce the same result. After studying the success or decay of weak messages a team in Stanislas Dehaene’s laboratory confirmed that the inconsistency is attributable to variation in the pre-stimulus state of the brain. They suggest that sensory consciousness can be explained as the result of a nonlinear ‘ignition’ process in higher brain areas. B. Vugt et al. The threshold for conscious report: Signal loss and response bias in visual and frontal cortex Science Vol. 360 p.537 4 May 2018
The Consciousness Instinct: Unravelling the mystery of how the brain makes the mind Farrar, Strauss & Giroux 2018 is a fascinating book from Michael Gazzaniga, the eminent neuroscientist particularly associated with the study of split brains and divided consciousness. The philosopher Daniel Dennett has written several books on consciousness. He recorded his latest thoughts in From Bacteria to Bach and Back W. W. Norton & Co. 2017. Especially stimulating are his ideas about Richard Dawkin’s ‘memes’ – of which languages, Dennett points out, are outstanding examples. And I must recommend Stanislas Dehaene’s Consciousness and the Brain: Deciphering How the Brain Codes Our Thoughts (Viking Press 2014).
Another famous name in consciousness studies, Christoph Koch, published a paper in Nature: What is consciousness? (Vol. 557 p.59 10 May 2018).
In broad terms there are three different classes of synapse (junction) between neurons. The transmitter that is released may be excitatory – increasing the chances of the next cell firing – or inhibitory, reducing them. Or it may be something more properly called a neuromodulator, which modifies the effect of transmitters. There are quite a few types of transmitter, however, and many types of neuromodulator. There are also several different types of receptor for each of them, so the possible effects are numerous.
The larger and more complicated the brain the longer the chains of excitatory and inhibitory neurons. An excitatory cell may excite an inhibitory one, causing it to inhibit another inhibitory one, thereby relelasing the next neuron in line from inhibition …… clearly the possibilities are enormous.
In brains, especially complex ones, a delicate balance is needed between excitation and inhibition. R. Rubin et al. Balanced excitation and inhibition are required for high-capacity, noise-robust neuronal selectivity PNAS Vol. 114 E9366 31 October 2017 But there is a puzzle here, since some of the synaptic connections of cortical neurons, which have been thought to store the results of learning by growing stronger through use, have turned out to be distinctly volatile. G. Mongillo et al. Intrinsic volatility of synaptic connections – a challenge to the synaptic trace theory of memory Curr.Op. Neurobiol. Vol. 46 p.7 October 2017 However the same authors have found that, although excitatory synapses significantly outnumber inhibitory ones, the latter play a bigger role in storing information. As long as the overall balance between excitation and inhibition is not disrupted the network can operate efficiently, despite the inconsistency of excitatory synapses. G. Mongillo et al. Inhibitory activity defines the realm of excitatory plasticity Nat. Neurosci. Vol. 21 p.1463 October 2018.
Inhibitory neurons come in several different varieties, each defined by containing some protein that the others don’t. The significance is gradually being unravelled. M. E. Hern & R. A. Nicoll Somatostatin and parvalbumin synapses on to hippocampal pyramidal cells are regulated by distinct mechanisms PNAS Vol. 115 p.589 9 January 2018. Meanwhile a study of cell types in the visual cortex and part of the motor cortex of mice has shown that the excitatory neurons differ more noticeably between the two areas than the inhibitory ones do. B. Tasic et al. Shared and distinct transcriptomic cell types across neocortical areas Nature Vol. 563 p.72 1 November 2018
How learning happens is one of the most interesting questions in neuroscience, and prompts a great deal of research. Much of it focuses on the hippocampus, the seahorse-shaped structure nestling in what’s known as medial temporal cortex – the bit that curls inward under the rest on either side. It was here that it was discovered that neuronal connections strengthen with use and weaken with idleness, something that also happens in other brain areas, with varying timescales and varying degrees of permanence. One of the effects in the hippocampus is that a succession of sensory inputs can be preserved for long enough for large-scale patterns to be perceived. Another, closely related function is preserving the order in which events occur. This has been confirmed by N. C. Heyworth & L. R. Squire The nature of recollection across months and years and after medial temporal lobe damage PNAS Vol. 116 p.4619 5 March 2019. Subjects with damage to the medial temporal lobe were taken on a 25 minute guided walk during which eleven planned events occurred, and their recollections of the walk were compared with those of controls immediately afterward, after one month, or up to two and a half years later. The most outstanding difference was that the memory-impaired patients reported the incidents of the walk in random order, whereas the control group preserved the order.
Another important function to which the hippocampus contributes is recording what is where. C. E. Connor & J. J. Knierim Integration of objects and space in perception and memory Nat. Neurosci. Vol. 20 p.1493 November 2017. This is one of several articles in an issue focussed on spatial cognition.
Subjective experience suggests that the experiences and discoveries we think about later, or dream about in subsequent sleep, are the things more likely to stick in long-term memory. Sure enough, many experiments have shown that the exchanges between hippocampus and neocortex which occur during new experience may be repeated during periods of idleness and in sleep, and do indeed serve to fix the new information in long-term memory.
The replays take the form of condensed, high-frequency bouts of neuronal firing termed ripples, which consolidate the connections initiated earlier. Studies in young rats show how these replays of neuronal activity develop as the brain matures. L. Muessig et al. Co-ordinated emergence of hippocampal replay and theta sequences during post-natal development Curr. Biol. Vol.29 p.834 4 February 2019.
Similar ripples accompany recollection, occurring just before a memory is successfully accessed. A. P. Vaz et al. Coupled ripple oscillations between the medial temporal lobe and neocortex retrieve human memory Science Vol. 363 p.975 1 March 2019. Moreover exceptionally detailed recollection of autobiographical events has been found to correlate with exceptionally strong connectivity between prefrontal cortex and hippocampus. V. Santangelo et al. Enhanced brain activity associated with memory access in highly superior autobiographical memory PNAS Vol. 115 p.7795 24 July 2018.
The hippocampus thus allows us to find patterns in inputs that occur over extended periods of time. The most valuable pattern, for almost any animal, must be the shape of its environment, knowledge which enables it to head as directly as possible for whatever it currently needs. Another area involved in the mapping process is the adjacent entorhinal cortex, where cells fire in such a way that the animal’s position is registered in terms of a grid covering all the area within sensory reach. An important bit of data here, of course, concerns the locomotion that takes the animal from one junction on the grid to the next. A contribution to the understanding of this system comes from M. G. Campbell et al. Principles governing the integration of landmark and self-motion cues in entorhinal cortical codes for navigation Nat. Neurosci. Vol. 21 p.1096 July 2018. Remarkably, it has been possible to create a computer programme which mastered the art of navigating around a modest area using a grid system based on the entorhinal one. A. Banino et al. Vector-based navigation using grid-like representations in artificial agents Nature Vol. 557 p.429 17 May 2018
Alas, in us long-lived humans the area in which hippocampus and entorhinal cortex are found seems to be particularly vulnerable to wearing out. M. Stangl et al. Compromised grid-cell-like representations in old age as a key mechanism to explain age-related navigational deficits Curr. Biol. Vol. 28 p.1108 2 April 2018.
Another contributor to navigation and the creation of the cognitive map is retrosplenial cortex, which seems to record the large-scale patterns, including those beyond sensory reach, and the overall geometry of a habitat. An addition to this picture comes from A. S. Alexander & D.A. Nitz Spatially periodic activation patterns of retrosplenial cortex encode route sub-spaces and distance travelled Curr. Biol. Vol. 27 p.1551 4 February 2019
Mapping the environment must be an ancient form of learning. There are many other sorts. A recently evolved skill is learning how to use a flexible muscle system to the best advantage, which depends on the evolution of a motor cortex, along with spinal neurons to convey its instructions to muscles. How it happens is becoming clearer. A. J. Peters et al. Reorganisation of corticospinal output during motor learning Nat. Neurosci. Vol. 20 p.113 August 2017
Here too sleep serves to consolidate learning, and even a very brief period of rest can help, it turns out. M. Bonstrup et al. A rapid form of offline consolidation in skill learning Curr. Biol. Vol. 19 p.30219 2019
The primary visual cortex has been most extensively studied in primates and in cats, where neurons favouring lines of a particular orientation cluster together, forming columns. A new study finds that this holds good in carnivores generally and in ungulates – but not in rats. M. Weigand et al. Universal transition from unstructured to structured neural maps PNAS Vol.114 E4057 16 May 2017 This looks like another example of the general principle that the more complicated brains get the more delicately organised they have to be.
A thought about colour vision: the magnificent array of colours we experience, produced by comparing the inputs of just three wavelength-sensitive types of photoreceptors and one less fussy one, is perhaps a particuarly telling example of what the extended processing of sensory input that leads to consciousness can achieve. Some invertebrates have a much wider range of colour-tuned receptors – dragonflies, for instance, and mantis shrimps. At least one variety of the latter has twelve sorts. But the shrimps don’t learn to distinguish among different blends of input. H. H. Thoen et al. A different form of color vision in mantis shrimps Science Vol. 343 p.411 24 January 2014 It seems pretty certain that where there is such a wide range of wavelength-tuned receptors each variety is dedicated to eliciting one particular hard-wired response, comparable to the male stickleback’s automatic display indicating territorial ownership at the sight of any patch of red. The visual systems of some of those brilliantly coloured little reef fish probably work the same way, with receptors pretty precisely matched to the colours displayed by the species. Shoaling with fellow members of the species is managed, I would guess, by a hardwired response to the distinctive colour patterns they display – blue and yellow stripes, red and green ones, or whatever.
It has been becoming ever clearer that activity in sensory pathways throughout the neocortex can be influenced by factors such as attention snd emotion. Now it’s been shown, in the mouse, that even neurons in primary visual cortex registering such basic visual elements as vertical or horizontal lines can fire somewhat differently depending on whether the overall pattern of input is new, or constitutes a view the mouse has seen before. A. B. Saleem et al. Coherent encoding of subjective spatial position in visual cortex and hippocampus Nature Vol. 567 p.158 10 September 2018. This intriguing result was obtained by an ingenious experiment in which the mouse ran on a revolving ball while a changing scene was projected in front of it, allowing tight control of just what the animal sees and where it thinks it’s going. The results of several such virtual reality experiments have been reviewed by Liam Drew in Nature: The mouse in the video game Vol. 567 p.158 4 March 2019. This experimental method has even been extended to study how olfaction guides locomotion. B. A. Radvansky & D. A. Dombeck An olfactory virtual reality system for mice Nature Communications Vol.9 article 839 2018.
Tiny, unnoticed eye movements called microsaccades ensure that fine features of the visual scene are at some point perfectly centred in the receptive fields of appropriate detectors in the visual system. Schelkova and colleagues confirm how helpful this is for distinguishing very small-scale stimulus-patterns. N. Schelkova et al. Task-driven visual exploration at the foveal scale PNAS Vol. 116 p.5811 19 March 2019.
The micromovements of the eye are far too small to be consciously noticed but I nurse a strong suspicion that that they are what make the eyes of mammals and birds seem lively and bright – in contrast to the glassy look of reptilian and fish eyes – and part of what attracts infant attention to eyes. But as far as I know nobody has ever looked for microsaccades in birds, let alone in reptiles, and nobody has checked for unconscious effects on attention.
Mirror neurons, identified first in monkeys then in humans, fire both when an action is performed and when the same action is seen being performed, apparently for the same purpose. They are found in various parts of both motor and parietal cortex.
The evolution of such neurons must underpin the capacity for imitation clearly seen in many mammalian and avian species. And what massive consequences this talent has had. An individual who can imitate can learn what is worth doing without taking the risks involved in experimentation, so knowledge gained by a single adventurous animal can be passed on from generation to generation painlessly. Moreover these new behaviours can become established in a very short time, in contrast to those which emerge from genetic changes to brains or muscular systems. (Indeed the influence may work in the other direction – the learnt new behaviours can create a selective pressure which favours modifications to brain or muscle.)
The capacity for imitation is particularly valuable for infants. In its simplest form it probably involves only the release of an innate action when the infant sees the parent performing it, so that the infant learns to look for food where the parent looks, and what sort of thing prompts flight in the parent. This form seems to be present in birds. I have seen a peahen take a few pecks in a likely piece of earth, which prompted her chick to forage there while she moved on a little. Something similar has been reported in domestic chickens. I imagine it’s a very similar mechanism which often causes a human infant only a few days old to smile back at someone who smiles at them, and infant monkeys to perform similar imitations.
This form of imitation seems not significantly different from the way little fishes in shoals tend to follow any individual which makes a sudden turn, so that the change of direction rapidly sweeps though even a large group. Perhaps we should look for pretty ancient origins for the first mirror neurons. Once they were established, though, they created a situation in which parental care of the young, providing readily available models to imitate, could become increasingly valuable.
The more complex forms of imitation no doubt evolved as muscle systems grew more complex. And as brains grew larger infants could build on the sensations created by automatic imitations to build the ideas which support conscious purpose. Thus humans and no doubt other primate infants gradually develop a conscious sense of how facial expressions coincide with emotion and mood, and how they operate as a means of communication.
All the assumptions about such imitation being confined to social species have been undermined by the observation that Malaysian sun bears are also capable of it. These animals normally lead solitary lives, apart from a period of about three months when a mother is rearing a couple of infants. A team led by Marina Davila-Ross studied some group-housed bears that were being rehabilitated before return to the forest. They observed many bouts of play, during which there were numerous instances of two sorts of open-mouth expression – something that sounds rather like the chimpanzee play-face – being reciprocated. D. Taylor et al. Facial complexixty in sun bears: exact facial mimicry and social sensitivity Scientific Reports 9 article 4961 2019 This report seems, incidentally, to add to the evidence that many large-brained mammals can switch between social and solitary life styles as the circumstances make appropriate. (Coyotes provide a particularly notable example.)
As well as mirror neurons for action there are also emotional mirror neurons – empathy neurons – which are active both when emotion is experienced and when someone else is seen experiencing it. These too must contribute enormously to infant education, demonstrating what is to be feared, what can be approached confidently. They’ve now been identified in rats as well as primates. M. Carillo et al. Emotional mirror neurons in the rat’s anterior cingulate cortex Curr. Biol. Vol. 19 p.30322 2019
There is evidence, incidentally, that motor imitation is a built-in urge that has to be suppressed. By adulthood the instinct is normally completely inhibited and must be released from inhibition when wanted, but damage in a certain brain area can affect the ability to control it. Patients instructed not to copy an experimenter’s actions have difficulty in complying. It’s obvious that emotional mirror neurons work in the same way – it’s all to easy to catch an emotion or mood, and it takes practice and self-control to avoid infection.
On being social
In social animals both sorts of mirror neuron are obviously also useful in adulthood, serving to spread useful information through the group. And if the group must deal with a predator or an attack from a rival group their best chance lies in sharing the same emotion and acting in unison – either all standing up to the threat or all running away. A mixed reaction reduces the chances of survival for everybody. The most successful group leader, meanwhile, is likely to be the animal which can not only defeat all rivals but can also powerfully activate emotional and mirror neurons in others when needed. This ability, which also implies a capacity for attracting attention, is surely the basis of that indefinable quality charisma, and the leader who has it generally makes for a more successful group.
Our capacity for living together in large groups, sharing our ideas, combining our efforts, and submitting to some form of central governance, has clearly contributed greatly to our success as a species. But the very characteristics that made the vast and rapid expansion of our numbers possible carry dangers in the modern world, for what works well in small groups isn’t always beneficial in large societies. Emotional mirror neuerons at play in a crowd can produce mass hysteria. Motor mirror neuerons can prompt people to do things they wouldn’t if they stopped to think about it. Charisma doesn’t necessarily coincide with wisdom. Now that we are organised into very large societies and equipped with the means of destroying ourselves completely (along with many other species) one of the challenges is to keep mirror neurons under control.
The knowledge stores
As we learn about the world around us, and about our own capacities, the knowledge is stored in the neocortex, neatly classified into different areas, just as different subjects are stored in different sections of a library. But the brain also constructs a complex network of connections between the areas, so that the record of a visual stimulus-pattern is linked to records of the sounds connected with it, of any emotions it might evoke, and so forth. In humans there is also likely to be a link to a word. Registering a stimulus-pattern in any one area can thus prepare the way for related input elsewhere, or for producing the appropriate word. And as we learn to recognise particular stimulus-patterns in ever greater detail the neuronal response becomes ever more finely tuned, the pathway of excitation increasingly refined by new inhibitory connections.
Particularly complex stimulus-patterns carrying several forms of significance activate several areas. Faces are a notable example, analysed separately for identity, expression, orientation, significance – and new face-sensitive areas keep being discovered. Landi and Freiwald have found (in rhesus monkeys) two in which, as blurred faces gradually beame clearer, the response to familiar ones ramped up quite suddenly, parallelling the way a face can suddenly be recognised. The response to unfamilar faces grew more steadily, as in other face-sensitive areas. S. M. Landi & W. A. Freiwald Two areas for familiar face recognition in the primate brain Science Vol. 357 p.591 11 August 2017
The neocortical areas which store knowledge of such important things as faces, bodies and landmarks are pretty consistently sited. How does this orderly arrangement come about? Several explanations have been put forward, not all mutually exclusive. My own proposition is that these categories of stimulus-pattern are the ones for which the infant has genetically determined responses – imitation for faces, bodies and voices, grabbing for grabbable objects, and so on – and the hardwired action is reported to a predestined part of the neocortex, which is thereby sensitised to the sensory input which evoked the action. This hypothesis doesn’t rule out other factors which might be at work, such as the suggestion that the reward infants gain from looking at faces and engaging in social interaction with them may play a role. L. J. Powell et al. Social origins of cortical face areas TICS Vol. 22 p.752 1 September 2018
Certainly the development of these areas is not dependent on the sensory channel through which the input comes. The language areas of congenitally deaf people who communicate by means of visual sign language are similarly situated to those of people who use spoken language. In like manner, it’s now been found, the congenitally blind store knowledge of faces, bodies, objects and scenes in much the same way as the sighted, although their knowledge must be gained entirely through other senses. J. van den Hurk et al. Development of visual category selectivity in ventral visual cortex does not require visual experience PNAS Vol. 114 E4501 30 May 2017 (This fits well with the long-ago discovery that the blind can draw recognisable pictures of things like tables that they know through touch.)
The output from these areas which deal in subjects such as faces is passed on to yet other neocortical areas which record the asociations among inputs from different senses, such as one towards the front of temporal cortex which responds not only to the appearance which defines an individual but also to the name that goes with it, and any other known aspects of the person concerned. Y. Wang et al. Dynamic neural architecture for social knowledge retrieval PNAS Vol. 114 E3305 18 April 2017
Interesting ideas and new discoveries about the past continue to emerge. It seems that the transition from single-celled animals to multicellular ones may have been less complicated and occurred more often than once thought. E. Pennisi The power of many Science Vol. 360 p.1388 29 June 2018
Our knowledgeof the weird and wonderful multicellular animals that had evolved by the time of the Cambrian period will be increased by the discovery of an extensive deposit of remarkably well preserved fossils – including, unusually, some of soft-bodied animals – on a river bank in China’s Hubei province. It will be fascinating to correlate them with the long-known fossil riches of the Burgess Shale in Canada, also from half a billion years ago. And now this Canadian fossil-bearing shale has been found to extend for kilometres beyond the original site, and more exciting finds are appearing. J. Sokol Cracking the Cambrian Science Vol. 632 23 November 2018
Why should there be two locomotory centres in the brain? I hypothesised in the book that they might govern slightly different forms of locomotion, serving different purposes, such as gentle exploration and hasty escape. New research suggests that I was partially right. The different forms are not quite what I guessed at – and there are probably more than two such centres. V. Caggiano et al. Midbrain circuits that set locomotor speed and gait selection Nature Vol. 553 p.455 25 January 2018.
Other interesting papers on motor subjects are M. N. Economo et al. Distinct descending motor cortex pathways and their role in movement Nature Vol. 563 p.79 1 November 2018 and Z. Gao et al. A cortico-cerebellar loop for motor planning Nature Vol. 563 p.113 1 November 2018
Modern lifestyles make considerable demands on attentional systems, and the effects are receiving some interest from neuroscientists, as in M. R. Uncapher & A. D. Wagner Minds and brains of media multitaskers: Current findings and future directions PNAS Vol. 115 p.9889 2 Occtober 2010 The effect of multimedia use in early years is also increasingly studied, as in D. A. Christakis et al. How early media exposure may affect cognitive function: A review of results from observations in humans and experiments in mice PNAS Vol. 115 p.9851 2 October 2018
The thalamus is a central structure in the brain, part of the diencephalon (the second of the major divisions of the brain, which are based on the way the swelling at the front end of the embryonic spinal cord is initially divided). In primates and many other species the thalalmus is thoroughly buried under the vastly expanded telencephalon, and was consequently very difficult to get at with recording needles until extremely fine ones were developed. For a long time it was known principally for relaying visual, auditory and somatosensory information to the necortex for further analysis, and for its contribution to shutting off access from sensory systems to the neocortex during sleep. Now a great deal more is being discovered, though the overall picture is still far from clear. A valuable exposition of the new evidence is L. Acsady The Thalamic Paradox Nat. Neurosci. Vol. 20 p.901 27 June 2017.
Here are some other recent papers: S. S. Bolkan et al. Thalamic projections sustain prefrontal activity during working memory performance Nat. Neurosci. Vol. 20 p.987 July 2017; M. M. Hakassa & S. Kastner Thalamic functions in distributed cognitve control Nat. Neurosci. Vol. 20 p.1669 December 2017; G. Pergola et al. The regulatory role of the human mediodorsal thalamus TICS Vol. 22 p.1011 November 2018.
The mammalian thalamus has many detectable specialised divisions. Those of such amphibians as have been studied are much simpler. I’ve suggested that the structure began as a sort of map, in a self-centred reference frame, of the animal’s current situation, one that came into existence at a period when all actions were pretty much hardwired. Actions would be reported to the map, and in a way that indicated the position of the stimulus that prompted them – the position in relation to the animal that is. Probably just enough visual information came from the retina to the proto-thalamus to anchor the map to the external world. (The visual data which guided the action would be dealt with by the visual centre in the lower brain which in mammals has been so greatly overshadowed by the visual cortex.) The complex functions of the mammalian thalamus could have evolved, I postulate, from this sort of beginning.
One of the most exciting aspects of neurobiology is the increasing understanding of mental disorders, and the promise of reducing the impact they have. In more general biology, meanwhile, a hot topic is the internal microbiome – the vast array of micro-organisms that live within and on us in profitable symbiosis. Skolnick and Greig have brought the two subjects together, suggesting that some symbiotic microbes, established in us for untold generations but sometimes lost in these days of liberal antibiotic consumption, may be essential for mental health. They cite two that contribute to the synthesis of neurotransmitters and one that assists in the excretion of mercury and possibly other heavy metals. These pathways, they say, ‘have the potential to impact systems relelvant to a wide range of neurodevelopmental and psychiatric conditions including autism, depression, anxiety and schizophrenia.’ S. D. Skolnick & N. H. Greig Potential neuropsychiatric consequences of dysbiosis TINS Vol. 42 p.151 March 2019
A suggestive new idea about schizophrenia is to be found in L. Barkovitch et al. Disruption of conscious access in schizophrenia TINS Vol. 21 p878 November 2017. And it looks as if pyschopathy might be explained at least partially by a deficiency of emotional mirror neurons. L. A. Drayton et al. Psychopaths fail to automatically take the perspective of others PNAS Vol. 115 p.3302 27 March 2018 Nevertheless they are not incapable of considering the welfare of others, it was established. Hope for the addicted, meanwhile, is held out by M. Diana et al. Rehabilitating the addicted brain with trancranial magnetic stimulation Nat. Rev. Neurosci. Vol. 18 p.685 November 2017