In seems that we are conscious of an event a fraction of a second after our brains start to process the perception of the event. This is not surprising – we are conscious of the ‘finished’ perception and ‘finishing’ the perception takes time. The surprise is that we live in the present and not a fraction of a second into the past. How is this done?
There seems to be three possibilities:
-we live a fraction of a second into the past but never realize it.
-we play very weird games with the perception of time and simultaneous events.
-we do not live in the past but in a projection of the past into the future, giving an approximate present.
The last one seems the most useful to us and also may be a good reason to have consciousness. If the present is t0 and at that time the brain is constructing a model of the world as of t-x. This model is then run forward as a simulation through x duration of time so that it represents a simulation of t0. So we would be living in a simulation of the present. This gives a measure of error to prompt corrections of motor action and of perception. That would really be a cool system and well worth the effort of creating a single global model of the world accessible by all the systems of the brain.
It is not take much of a leap to go from thinking of a MPOFBL system (massively parallel overlapping feed-back loops) to thinking of a MPOFFL system (massively parallel overlapping feed-forward loops).
After writing the above but before posting it, I found a blog post on a NYTimes article. See article and blog. They discuss visual illusions that point to a projection into the near future. In particular the article talks about seeing an image before it actually happens.
“In an experiment originated by Dr. Nijhawan, people watch an object pass a flashbulb. The timing is exact: the bulb flashes precisely as the object passes. But people perceive that the object has moved past the bulb before it flashes. Scientists argue that the brain has evolved to see a split second into the future when it perceives motion. Because it takes the brain at least a tenth of a second to model visual information, it is working with old information. By modeling the future during movement, it is “seeing” the present.”
We use a metaphor between brains and computers in a great many situations. This metaphor can be helpful but it can also mislead us. Brains and computers are very different and so any use of the brain-computer metaphor should be justified as useful, not just assumed to be so.
Some people assume that because neurons are either ‘firing’ or not firing, that the brain has a digital mode of operation. This is far from true:
The rate of firing is often the most important aspect of a neuron’s activity and this is a graded activity and not an all or nothing one. A high firing rate gives a strong signal and a low firing rate gives a weak signal. This is not an on-or-off digital signal.
The effect of a firing neuron on another neuron at a synapse can be to excite or inhibit the firing of the second neuron. So even in situations that are not graded, the signal is not binary.
A neuron can have few to many synapses with another neuron and so the effect of one neuron on another is usually graded. This variable number of synapses changes with learning and forming memories.
Other non-neuron cells (some glia cells) do not ‘fire’ but are probably always graded in their effects and take part in signaling in the brain in a not very well understood way. Glia out number the 100 billion neurons by 10 to 1.
The timing of ‘firing’ of neurons can change the nature of the signal. Simultaneous firing is required for some effects. Timing effects may be graded.
Neuron signaling happens in an environment of potent chemicals, electrical fields and magnetic fields. The effect of this environment is graded.
It takes a lot a circuitry to mimic a neuron with all its in-coming synapses (where it is the post-synaptic cell) and all its out-going synapses (where it is the pre-synaptic cell). A neuron is not a simple electronic switch, or even just a logic gate, it is a complex device.
Some people assume that because brains can solve problems and the like, they use algorithms or step-by-step procedures. There is a general idea abroad that a Turing machine is a Turing machine and therefore all types of computing can be translated into all others. I believe that there is a big difference between sequential algorithms and some massively parallel operations. The way our present day computers work is sequential (one small operation at a time) or moderately parallel (from a couple to a couple hundred sequential algorithms working simultaneously and exchanging information at intervals), that is to say they use algorithms.
The brain, on the other hand, incorporates feedback to an almost incomprehensible level. If a nerve track goes from region A to region B there is more than likely another track going from B to A. For example the signals from the retina go to a ‘map’ of the retina in an area called the thalamus. From there nerves take signals to the optic lobe of the neo-cortex where they are mapped again and on to other retinal maps. The neo-cortex neurons send signals back to the thalamus and to the matching part of the other hemisphere. In fact, more or less wherever the neo-cortex sends signals, it also gets signals back from that same place. These loops are basically overlapping parallel feedback loops. Events in such systems cannot be traced sequentially. Signals reverberate until a stable configuration is found (or rattle forever presumably). The brain’s operations are simply not understandable as digital computer programs. The operations are not even that similar to analog computer patch boards. I do not mean that computers cannot be made to mimic some aspects of the brain, but simply that we have to be careful how we interpret such simulations.
Finally, computers are general machines built to run programs. The nature of the program, a word processor, a drawing program, a spreadsheet, is not important. A brain on the other hand is an organ in a living organism and so its functions are biological. It does not have a general architecture like a computer but a very specific architecture for achieving very specific ends. It is concerned with keeping its organism alive. Computers are occasionally built for very specific purposes but not on anywhere near the scale of specificity seen in brains.
We must use the convenient metaphor between computers and brains with caution. We would also do well to remember that ‘sufficient quantity has a quality all its own’.
If our brains are to represent the frequency spectrum of light reflected from an object, how can it be done? Would it be better for our survival if objects had numeric frequencies superimposed on them? Would some other representation make it easier to tell if the fruit was ripe or the snake was poisonous? Why not colours? Would some other representation make it easier to remember and recall a place or object? Is there a better way to represent the frequency of air pressure waves then what we call sound?
The qualia that we experience are the forms of the model of reality that our brains create. No doubt the forms that our brains use have evolved to provide the ability to quickly and accurately discriminate between objects and events. And if this is the case, it seems to me that we would probably share the basic nature of qualia with other vertebrates, although details would differ. It certainly seems probably to me that we would share qualia with other humans. When someone asks how we know that my red is the same as your red, the answer is we don’t know; but the best guess by far is that they are the same or similar. Otherwise, we would have to explain how and why they differ. As far as I know, there is no evidence that they differ and no evidence that our sensation of red is developed in response to environmental factors. There is also no evidence (so far at least – and – to my knowledge) that our brain architecture differs in this respect from individual to individual. Why would your short wave length retinal cones follow or produce different development signals then mine do? Obviously the cells are going to follow the same biochemical programs during the development of the brain in you and in me. When this does not happen, the result is learning disabilities and other problem conditions.
I’ve just been reading a ’97 paper by Ramachandran and Hirstein. They place the conscious experience with its qualia in the temporal lobe. They also give three laws of qualia and discuss the relationship between qualia and the experience of ‘self’. If you are at all interested in this subject, the paper is a must to read.
Any hypothesis of the nature of consciousness has to address several facts – at least these facts, as a minimum, and there probably are other constraints as well as these.
There is an intimate connection between memory and consciousness. We remember what we have been conscious of. We know we have been unconscious if there is a discontinuity in our memory of events. Memories ‘package’ experience in a form that closely resembles the conscious experience of events. We feel a flow of future projections to present conscious experience to past memories without any sharp change in the type of awareness.
Consciousness and dreaming have some commonalities but are quite different in their connection to perception and action. Dreams are not about reality and usually not result in action and usually are not remembered.
There is a time delay in the formation of a conscious experience which we do not perceive. It is as if we were automatons who, for no good reason, experience a continuous movie of our immediately past perceptions and actions. We feel our consciousness is a real time event when it is not. The best summary of the experimental evidence for the time delay that I have found so far is Roger Penrose’s review of the evidence.
Consciousness is not a completely faithful model of what is actually happening. Illusions and misinterpretations are frequent and somewhat predictable. We do not even have a completely accurate model of our own minds. For instance: there are blind sight and other knowing-without-awareness-of-knowing; there are phantom limbs and other awareness-of-the-nonexistent.
The way that conscious experience presents itself seems arbitrary and personal. Why is a particular wavelength of light seen as a particular colour? Is my red your red? Qualia (properties of perceptions) are an aspect of consciousness.
Consciousness happens at the same time as some pronounced activities in the brain, particularly reciprocal communication between the cortex and the thalamus, low amplitude but high frequency brain waves, activity in the reticular formation. There is a summary in Scholarpedia but it is fairly technical.
Our consciousness can be affected by various mental illnesses, drugs, stressful situations etc. We can also learn to modify our consciousness (with meditation for example).
We need a picture of consciousness that fits with and helps explain these facts.
When I was young, I have no doubt that I believed in Santa Claus, the Easter Bunny and the Tooth Fairy. I stop believing in them early enough that I do not remember the event – but I do remember the feeling that it was important to continue pretending to believe in them. I have encountered that feeling since when I have given space to some idea that I thought was a harmless little sham, something comforting to say, an old reliable metaphor, a poetic licence.
It came as a real shock to realize that others believed a myth; it was like finding adults who believed in Santa Claus. The conscious mind as opposed to the subconscious mind was one of those ideas. Although it had never seemed to me that I had any more than one mind, it did seem polite to allow others to separate themselves slightly from their actions. They could appear more modest or more blameless if they were allowed to let their subconscious to take the praise or blame. Sometime in my teens, I began to call people on their ‘my subconscious did it’ remarks. To my surprise they believed they had this other mind inside them that was bent on sabotaging their intentions. The only picture that made sense to me was that I had one mind and it was in no way divided – this mind perceived, thought and acted – and I was conscious of some but not all of this activity.
So I had a mind and it was the only mind I had. If I had to choose between it being a conscious or an unconscious mind, I would have to say unconscious, although the choice is somewhat ridiculous. And I also had consciousness – not a conscious mind just plain consciousness. My consciousness seemed part of my mind but not involved with my perceptions, my thoughts, my decisions and so on. All it did was supply an awareness. I have lived the next 50 years with this self image.
The problem is that other people seem to have a very different self image that they are as sincerely about as I am about mine.
I am encouraged that more and more published material uses the term ‘consciousness’ rather than ‘conscious mind’ as time goes on. Here is John R Searle’s definition which does not imply that the thought process is a conscious process.
What we need at this point in our work is a common sense definition of consciousness and such a definition is not hard to give: ‘Consciousness’ refers to those states of sentience or awareness that typically begin when we wake from a dreamless sleep and continue through the day until we fall asleep again, die, go into a coma or otherwise become ‘unconscious’. Dreams are also a form of consciousness, though in many respects they are quite unlike normal waking states.
I was interested in a post on Mind Hacks about looking at the brain as a Bayesian probability machine. The aspect that got me was the modeling and the importance of prediction.
I was reminded of ideas that I wrote down in the early 80’s.
“We can view ourselves as having a nervous system with many functions, one of the major ones being the function of ‘mind’. The mind function consists of building, maintaining and refining a model of reality; using the model to predict, plan, decide, initiate and control responses to the world; storing an edited form of the model as a memory of experience for comparison and learning. During introspection we only have access to consciousness, the memory being formed, and not the model itself.”
The Mind Hacks post reviews an article from the New Scientist by Gregory T. Huang about the work of Friston and Hinton at UCL:
(Bayesian statistics) “asks the question ‘what is the probability of the belief being true, given the data so far?’. The NewSci article looks at the work neuroscientist Karl Friston, who increasingly believes that from the level of neurons to the level of circuits, the brain operates as if it uses Bayesian statistics. The essential idea is that the brain makes models upon which it bases predictions, and these models and predictions are updated in a Bayesian like-way as new information becomes available.”
Picture the brain’s circuits working to minimize the prediction error of a reality model and doing this by using feedback loops. I envisaged a different type of feedback (something on that in future posts) but the principle is very similar.
“Friston created a computer simulation of the cortex with layers of “neurons” passing signals back and forth. Signals going from higher to lower levels represent the brain’s internal predictions, while signals going the other way represent sensory input. As new information comes in, the higher neurons adjust their predictions according to Bayesian theory. This may seem awfully abstract, but there’s a concrete reason for doing it: it tells Friston what patterns of activity to look for in real brains.”
It is great to look forward to this and similar ideas being experimentally tested!!
Before I start posting, I would like to thank three bloggers. Their combined effect on me this week was to result in my starting this blog.
Greta Christina’s Blog has a posting, Greta’s largely unsolicited advice on blogging, that has giving me courage to jump in the blogging water. Babel’s Dawn by E.B. Bolles has convinced me that big vague questions can be tackled using a blog to create a hypothesis and critically examine it. He is doing this with the evolution of language. I have him to thank for showing me how to be a useful blogger. And, the current piece in Mind Hacks about Bayesian aspects of thought is where I intend to start with my first posting.
Does the world need another blog?Well, I think so or I would not be starting this one. We need to talk about consciousness from a different angle. So… I will put forward various views that I find worthy of discussion and then I will hope that others comment on my pieces and other comments.
Consciousness is a big subject:
What is it?…..Why do we have it? –or- What is its function?…..How did we get it? –or- How did it evolve?…..Where is it? –or- What is its neurobiology?…..What does it feel like?…..Other questions will, no doubt, follow.
It is also approached in different ways – by scientists, philosophers, therapists and artists.
But there is the hope that consciousness is quite simple to understand and with that hope, I blog.