A shorter version of this article was delivered as a lecture to the Philosophy Forum in Melbourne, Australia in 2016.
We feel that our consciousness is the very essence of our existence. The quality of each life is just a matter of the particular content of that consciousness, or in other words, all the things that we are aware of and think about during our life.
But what is consciousness, what is its content, and how did its content arise? The first of these questions, what is consciousness, seems to be inscrutable. Some of the content seems to be hard to describe and some straightforward. And there is a lot of complex argument about all of this.
I will try to untangle the matter by assuming that there are basic elements called data, and that the content of consciousness can, after a series of steps, be ultimately described in terms of data. You might say that the concept of data is just part of the content of our consciousness, so this might be a circular argument. I will come back to that later.
Some of what I say will sound like common knowledge. And some of what I say may challenge your concepts or definitions relating to data, information, meaning, intelligence and consciousness.
The word data is the plural of the Latin word datum, i.e., a datum is one item of data. In English, the word data is used both as a plural noun and as a singular collective noun.
The Latin word datum meant a thing that is given. In modern usage each datum is a representation (or a symbol) of something. What does “being a representation of something” mean? If we want to refer to something, or to find a way to remember something, it is often convenient or efficient to use something simpler, such as a particular mark or a particular sound, etc., to represent it. Written or spoken words represent something, so they are data. To use data it is necessary to remember each representation and what it represents, and to be able to distinguish that from representations of other things. It’s inconvenient when you can’t recall a particular word or remember what a particular symbol means.
Some items of data are constructed by humanity, such as words, numerals and letters of the alphabet, some are produced by devices that record sound, pictures and other aspects of the material world, and some are produced by the brains of individual humans and other species.
Information is a group of data whose members seem to have some relationship to each other. Often the relationship is obvious, e.g., a series of letters that spell a word. Sometimes an analysis may be needed to detect relationships in a collection of data and to show the way the members are related. The term data mining refers to the analysis of huge quantities of data to see what kinds of relationships or associations – that is, what kinds of information – are able to be detected.
There is a saying: Data is not information, information is not knowledge and knowledge is not wisdom. But information is a kind of data, and knowledge is a kind of information, and wisdom is a kind of knowledge. And there would be argument about which pieces of information were knowledge and which were untrue, and about which pieces of information were wisdom and which were fallible. And some generalisations that might not be exclusively true could still contain wisdom. A more justifiable saying would be: Some items of data might not be information, some pieces of information might not …, etc. (However, most people would prefer a presumptuous half-true jingle to something that sounded clunky but was closer to the truth.)
The representations that constitute data and information can be in almost any medium and in almost any form. However, the particular medium and form need to suit the purposes for which the representation is being made.
A particular form of representation occurs in all living organisms. With animals that have neurons, that is, nerves, information is stored as patterns of connections between neurons.
Here is an example of a piece of information occurring in a range of forms and technologies.
The vibrating vocal cords of a person speaking produce speech sounds, which are vibrations in the form of very small compressions and rarefactions in the air. The pattern of vibrations comprises a series of data, whose members are associated with each other as a representation of the speech. So they are a form of information.
These vibrations are propagated to the surrounding area.
A nearby microphone responds to the vibrations and generates an electric current whose variations correspond to the pressure variations of the speech sounds. In other words, the microphone has detected the information. The electric current now contains the information and it is passed to a CD (compact disk) burner.
The CD burner converts the patterns of electric variations into markings under the transparent surface of a plastic disc. This has now stored (or recorded) the speech information onto the disc.
This information stored on the disc can be “read”, that is, a reproduction of the speech sounds can be provided, by a CD player. In this machine a laser beam is shone onto the disc. As the disc rotates, each successive marking on the disc, that is, each stored unit of information, is successively illuminated by the beam. The laser beams are reflected by the markings on the disc. A light-sensitive device converts the patterns of the reflected laser beams into a correspondingly varying electric current.
Then the electric current is fed into a loudspeaker, which converts the electrical symbols into a varying magnetic force that makes a diaphragm vibrate backwards and forwards. This produces sound vibrations in the air.
The ears of any listeners detect the vibrations and convert them into electrochemical symbols in the neurons of the listeners’ brains. The listeners’ brains interpret these symbols into the “message” conveyed by the person who spoke at the beginning of this sequence. If the listeners remember any of what they heard, then the remembered information is now stored in the networks that comprise the short-term memory of their brains.
In brains, the basic units of data are held in the nuclei of the neurons. A particular kind of molecule is “switched on” by having a small group of atoms attached to it, a bit like the switching on of genes in DNA. The neurons are connected into networks, and the various kinds of networks assemble the data into the various kinds of information that are contained in the brain. The connections between neurons can be long-lasting, brief or fleeting, depending on the nature of the information and what is happening to it.
So in this example of information transfer, the same piece of information has appeared in a range of different media.
Technical imperfections or constrictions in the transfer of the information from one medium to another, and in the storage and reproduction, often reduce and/or “contaminate” the information. This applies to voice, text, pictures, music and other kinds of information.
Also, I might have put a voice recognition system somewhere in the chain of transmission that I just described. This would have converted the spoken words into written words on paper or on a screen. Then an optical character reader could later have read the paper text, and a speech generator could have produced intelligible speech.
The voice of the actual person speaking would have contained more information than in the printed words. The voice would be recognisable as that of a particular person, and there might be recognition of any emotions expressed beyond those implied by the words. These extras would be lost if the system had gone through a printed text stage and then just been read, and also if the text had converted to regenerated speech. That is there would be some loss of meaning.
However, whenever written words are sufficient to convey the desired message, text is more economical than speech in conveying the intended meaning.
And now we come to the next step: what do we mean by the word meaning?
As I implied earlier, the meaning of a single symbol, or any piece of information, is just the identification of the thing it represents. If I write the numeral 5 or say the word cat, you will know the meanings of these two symbols if you can recognise written numbers and speak English. If I write three symbols, the letters c, a, and t, you will recognise some relationships between them. They are all letters of the alphabet and when they are presented in the right sequence they spell the word cat. So they comprise a piece of information, and if you can read English you will know that the meaning of this information is certain kind of animal. If they are presented in a different order, a, c, t, which is a different relationship between these same three symbols, they spell a different word with a different meaning. So the sequential relationship of the items of data can be important in determining the meaning of a piece of information.
The word meaning has several connotations. That means it has more than one meaning. Sometimes meaning could be the equivalent of the word intending. Sometimes it can imply significance. The term the meaning of life could signify some purpose that someone might feel destined to serve. But then, the meaning of life might alternatively be a definition of the characteristics that make living organisms different from inanimate objects. So the meaning of a piece of information can sometimes be its definition, or a description, which itself is a piece of information.
One definition of life is that living organisms have distinctive characteristics: a metabolism, regeneration, a life cycle and interaction with their environment. And what is the meaning of that piece of information, you may ask.
In plainer English it means that living organisms:
take in food and energy and process it
in order to grow from infant to adult and stay alive,
and to produce eggs or babies from within their bodies to bring about the next
and they also interact with their environment
to get the food, etc, that they need,
and to try to keep safe from natural hazards and predators.
And at last, by adding more information, there is an understandable meaning: that life means the capability to perform this collection of processes.
These examples show that for any piece of information to be “meaningful”, we have to know something about what it refers to. Sometimes we might not know enough to understand exactly what is meant. Then we need more information, until we come to whatever is actually being referred to. So the meaning of any piece of information is just the identification of whatever it represents, even when it represents another piece of information.
To one person a collection of data may have no meaning but to another person it is meaningful. Also, different people may put different meanings to a particular piece of information. In both cases this happens because different people know different sorts of things. Or in other words, the relevant information that each already has is different.
That means that what people, or organisms or computers accept as information is dependent on whether they can derive meaning from it. And that depends on what information they already have. So, for example, some noises and continuous noise can contain information. To some people they are just noise. Some people will recognise particular noises and perhaps know their source, so to them the noises are indeed information. This shows that there is an individual personal connotation of the word information, and it conforms, at least partly, to our common concept of information.
Computers store information of various kinds. They use this information to actuate various devices that are attached to the computer. These devices perform electromechanical processes, such as printing documents, or producing sound or moving pictures or monitoring the environment, etc. Each device has a very small range of capabilities. It has to understand the meaning of the information supplied to it by the computer in order to be able to perform its functions correctly. So each has been provided with the relevant information needed to understand the meaning of what the computer sends to it, i.e., what to do. It also must have information in the form of instructions for performing its particular functions, i.e., how to do it. So the meanings of these two pieces of information are what to do and how to do it.
So for each person or machine, meaning is what distinguishes data from information, and from a personal point of view the distinction is arbitrary. And while every possible single item of data would have a meaning, which, by definition, would be whatever it represents, there could be assemblies of data that had no apparent meaning for anyone.
Information can be very complex, particularly large systems of information. And the greater the complexity of the information, the greater is the complexity of its meaning. Examples of very complex systems of information are the various spoken and written languages, and mathematics. These systems of information are important tools for human societies. And similar simpler systems are important for other species. For the successful use of mathematics and languages in a society, it is necessary that the members all put the same or compatible meanings to particular pieces of information. With mathematics this requirement is mostly observed. But with languages there is always a degree of ambiguity of meaning and continual changes in meanings. There is also a continual change in the symbols, that is, in individual words and phrases, etc.
The concept that meaning is just whatever a single datum or larger pieces of information represents is so simple that it might seem to be either wrong or trivial. I think it is neither.
To repeat what has been said earlier, the issue is that any symbol or set of symbols does not become data or information unless it is recognised by some organism or machine that is able to associate it with something. Because different people and machines have different information by which they derive meanings, the relationships between derived meanings and the entities they apply to can be complex. This is the subject of the discipline of Semiotics.
According to the first law of thermodynamics neither mass nor energy can be created or destroyed but each can be converted to the other. So information theory, as a representation of the physical world, says that information cannot be created or destroyed. In this context, information might seem to be different from its every-day tangible counterpart. When a document is burnt, is information destroyed? The answer is no. When the document was printed, all the processes were just operations of the laws of physics. Burning the document was also just the operations of physics. But what about the meaning of the text? As I said earlier, the meaning is just whatever might be recognised by some person or machine that reads it. It is meaningless unless it is being read.
Some physicists have speculated that information is the basic component of matter. The rationale for this is that atoms were once thought to be the fundamental parts of matter, but then it was found that atoms are composed of protons, neutrons and electrons. And later it was discovered that protons and neutrons were composed of quarks. So questions arose as to what quarks and electrons and all the other “fundamental particles” were composed of, and whether there could ever be some kind of fundamental things that were composed only of themselves.
The idea that the individual units of information are the fundamental components of existence is beguiling: the items of information are nothing but themselves, irrespective of what physical forms are used to display them.
But every item of data is meaningless without there being something specific that it represents and some other information that identifies what it represents. Information is essentially non-material, and no amount of units of information could form a quark or an electron. While it must always be carried by some physical entity, information itself has no physical properties. However, this won’t stop some people from wanting to put some mysterious connotation to the idea that everything is composed of information.)
According to Webster’s Dictionary, intelligence is the ability to learn, understand or deal with new situations or the ability to apply knowledge to manipulate one’s environment or to think abstractly. This is meant to imply human intelligence but it could apply also, to a lesser extent, some other animals
Animals, and other species, need many different forms of intelligence in order to survive; obtaining food, avoiding enemies, predators and other dangers, and interacting with their associates and other species. Many species display emotions such as fear, anger, respect, empathy, care, and a sense of fairness. Many species have skills, such as in manipulation of objects, constructing nests and other structures, and finding their way around. Many species have memories that they act upon, and plan for the future. Many species display skills of social manipulation, between their associates and with humans.
I will now suggest what I regard to be a universal concept of intelligence that is derived from what I have said in the earlier sections of this essay. I think it incorporates the dictionary definition and the examples of different forms of intelligence that I have just given.
Intelligence is the act or capability of processing information. To be specific, intelligence relates to one or more of the processes of producing information, and of detecting, recording, recalling, reasoning, analysing, calculating, imagining and deciding about it. The processing can sometimes be very simple and sometimes extremely complex.
Processing of information can occur in both living organisms, and in inanimate objects such as computers, CD burners, burglar alarms and mousetraps. Intelligence can be “built in” (i.e., provided during the assembly of the physical structure) and it can be “learnt”, as an addition to the already available information, and consequently to the intelligence. This applies to both organisms and appropriately designed devices.
Built-in intelligence is sometimes referred to as instinct, particularly with reference to other species that some people don’t think of as being intelligent. All species have built-in intelligence, including people. Sometimes people are said to be acting on instinct when doing something in a way that feels appropriate without having to be thought about. This may be built-in intelligence, but it is usually learnt intelligence, with the information governing its processes having been integrated into the memory from long practice.
A mousetrap that has not been set has only built-in intelligence, which gives it the possibility of being set. The process of setting it to catch a mouse adds the information of putting a spring under tension and being held by a metal tag that can be very easily dislodged. The trap “detects” the mouse when the mouse’s nibbling or its weight moves the tag slightly and dislodges it. This single bit of information is “processed” by the release of a spring that sends a thick wire crashing down onto the mouse. A trap like this has a minimal amount of intelligence.
In information theory, which is a part of physics theory, the “information content” of an object is the number of all the conditions of all the particles of matter that the object is made of. One single unit of data represents one single aspect of one particle in one object. Extending this, every particle of the universe would, theoretically, be represented by the items of data that are required to fully represent its characteristics and its condition. So a change in the conditions of some of the particles means that there is a corresponding change in the relevant information. This means that every possible piece of data is information.
So with a mousetrap, once it is set, among all the information about the conditions of the particles it is composed of there are two pieces of information that are relevant to its function: the unstable equilibrium of the tag, and the condition of tension in the spring, which is held back by the tag. A third piece of information is added when the mouse upsets the trigger’s equilibrium. The spring is released and the laws of physics determine its action of slamming down.
(Just as we would not describe the action of a mousetrap or any other mechanical device in terms of the atoms and molecules it is made of, similarly we would consider only the “macro” elements of information in describing the intelligence of the mousetrap. And we do the same with the intelligence of a computer or a person.)
In more sophisticated arrangements such as computers and brains, much of the data and the forces acting on them are electronic, electromechanical or electrochemical. Every organism, and its functioning, from its beginning to its death, is the product of the processing of the information of its DNA, in conjunction with the information of the rest of its body and of the relevant parts of its environment. This processing occurs at every level of complexity up to and beyond its entire body. All this processing comprises the intelligence of the organism, and it is all just the product of the operation of the laws of nature. So intelligence is just a representation of the laws of nature.
So intelligence is both the handling of information and also the representation of the operation of the laws of nature. And, putting this the other way around, all natural processes are manifestations of intelligence.
Some people will insist that acts of intelligence can occur only in brains. But if you can accept that brains, whose processes are produced by natural forces, are intelligent, why would you not accept also that computers, and dishwashers and mousetraps, and a rushing river that picks up sand and pebbles and deposits them downstream, and the eruption of a volcano, and the melting of a block of ice, and all other natural processes are intelligent?
Analog, Digital and Measurement
It is possible to make precise measurements and representations of very many different things. Voice, music and other sounds consist of vibrations whose various frequencies are within the range of frequencies that humans can hear. Information representing sound vibrations is often depicted as a very jagged type of wave moving across a page or screen. The vibrations of sound are also often represented by the variations of an electric current. In each case, the representation has features that have some similarity with what it represents.
This kind of representation of sound is called analog, which means that the representation has some analogy with what it represents. With an analog representation it is usually possible to understand what is going on. Since, despite quantum theory, the world around us looks to be continuous, so analog representations usually look continuous, as distinct from, for example, looking pixellated.
Many other things can be represented in digital form
Some analog representations do not appear to be continuous, but their pattern can be recognised by “joining the dots” or recognising equivalent sequences.
Another kind of representation is digital. Digital information is a series of distinct elements. The individual elements, or symbols, have no necessary semblance to what is represented. Words and text and numerals are digital. In computing and telecommunications the most common kind of digital representation is composed of only two symbols, usually referred to as 1 and 0. So a passage of sound, or any kind of information, can be represented digitally by a series of 1s and 0s. Computers and brains store their information, and process it, in digital form based on 1s and 0s.
One single digit of data is known as a bit. A bit represents the smallest possible amount of information, the answer to a Yes/No question depending on whether the digit is a 1 or a 0. Digital data and information are measured in bits and multiples of bits, such as kilobits, megabits, gigabits, etc., or in similar units of bytes. (A byte is a sequence of eight bits.) These are precise measurements. The individual symbols of digital representations often require more than one bit of information.
The difference between analogue and digital can be illustrated by comparing an analogue clock whose hands move around its face, with the distance moved representing the elapsed time, and a digital clock that displays numerals, which are arbitrary representations with no intrinsic relationship with the passage of time. The numerals 4,5,6,7,8, etc., bear no analogy with the relative quantities they represent. The information in a table of data, expressed digitally using numerals, can also be expressed in analog format as a graph or a pie chart.
Analog and digital representations can be converted to each other. Each can contain components of the other, as, for example, music notation.
A stream of digital data can be analysed using all of the processes of logic and mathematics, and it can be precisely measured. Analog representation is not well suited to these kinds of manipulation.
All organisms contain large amounts of information in their DNA or RNA within the cells of their bodies. In DNA the information is coded in long sequences of symbols, but in contrast to the 1s and 0s used in electronic devices, uses four symbols. These symbols are A, G, C and T (which are embodied as the chemicals adenine, guanine, adenosine and thymine). Within the environment of the cell, specific combinations of these four symbols are interpreted as instructions to assemble one of 20 amino acids. The sequential production of amino acids assembles specific proteins. DNA has been suggested as an ideal store of information for computers and other devices, where the meaning of the information stored in it would be whatever the user required, in the same way as with the (less efficient) “memory” devices that are currently used.
On the matter of intelligence, we can readily produce measures of the speed of processing, and the capacity of memory of computers. These are usually listed in the specifications of each individual computer. The actual performance of the computer then depends on the particular programs and information that have been loaded onto the computer’s memory.
There is much less certainty when we come to the specifications of brains. We can estimate the number of neurons and of their axons and dendrites, and hence derive the possible number of connections that might be made between neurons. From this we might estimate the potential capacity to contain information. Comparisons could be made between brains of human beings and other species, perhaps giving some indication of potential processing potential. But some people appear to be more intelligent than others without necessarily having more neurons, and the same applies other species. So while in theory we might be able to measure the information content and intelligence of a brain, in practice we have only rough indicators. And we are often amazed at the intelligence of microscopic organisms that appear to have negligible amounts of brain or nerve material.
One indicator of intelligence is the testing of functional skills. Human beings can do many things that other species can’t do. But most other species can do things that human beings can’t do. Individual members of a species can often do things that many other members can’t. So comparative measurements must make an arbitrary valuation of different capabilities. Speed of performing mental or physical tasks is another measure, but this could be as much a function of experience and practice as of intelligence.
With humans, there can be brain conditions such as autism and other congenital handicaps that reduce the range of intellectual abilities but sometimes also produce abnormally high abilities in specialised areas such as, but not confined to, memory and mathematical prowess.
The “standard” measurement of human intelligence is the Intelligence Quotient or IQ, which assesses various aspects of problem-solving and ranks individuals in accordance with population-wide averages. The measure can be affected by knowledge of cultural norms and the amount of teaching each person has had in the solving of the particular types of task. Recent versions of IQ address the gamut of human areas of intelligence, and their results correlate fairly closely with performance speeds. Also, there are aptitude tests for the suitability of a person for a particular task that requires specified abilities.
And now we have reached consciousness.
Consciousness is described as awareness, or feeling, or experiencing or subjectivity. And as I said at the beginning, it seems to be the essence of our being. It may also be the totality of our being
Consciousness may be:
- merely being aware;
- being aware of something;
- being aware of existing as an entity;
- being aware of processes;
- being aware of abstract ideas;
- or feeling emotion.
This means that, even including emotion, consciousness is being aware of some form of information.
All the information that leads to consciousness comes from the brain, representing a great variety of things. These things are:
- sensations, such as feeling pain and itch, heat and cold, etc., seeing light and dark, and colours and shapes, etc., hearing, smelling, tasting, feeling balance, feeling weight, and inputs from the other senses;
- mental processes, such as knowing and doubting, remembering, reasoning, imagining, deciding, calculating, etc.;
- emotions, such as pleasure and displeasure, happiness and sadness, liking and hating, awe and reverence, disgust, eagerness, boredom, despair, etc.
All these things comprise what can become the content of consciousness. While they may appear to be very different in kind, they are continually interacting with each other to create such things as language, all kinds of art, mathematics, and other systems of thinking.
There is another way of classifying consciousness.
The everyday kind is when we are awake and everything is “normal”.
Conditions of “semi-consciousness” occur when someone is “not quite asleep” but not dreaming.
Dreaming is a kind of consciousness at a time when we are asleep and not aware of our surroundings.
Some people can have “controlled dreaming” when they are sufficiently awake to direct what they are dreaming. People in deep meditation claim to reach “higher levels” of consciousness. Certain experiences that are usually regarded to be abnormal (such as hearing voices that other, “normally functioning”, people cannot hear), and states caused by “mind-altering” – or should that be “brain-altering” – drugs and activities, are other aspects of consciousness. All of these together comprise what we consider to be our personal identity.
Consciousness is often thought of as being the same as the mind. But the term unconscious mind suggests that consciousness and mind are not identical and that mind may have wider connotations, involving both consciousness and unconscious intelligence.
The unconscious mind is part of the operation of the brain in conjunction with other organs. And consciousness is intimately related to the brain’s information and intelligence.
The intelligence of the unconscious mind can perform tasks without our feeling or experiencing them. A lot of these are the “housekeeping” functions of regulating the operations of the body. The unconscious mind also directs the very complex muscular processes of the activities that we can do, without having to think about how we are doing them, such as walking, writing, touch typing, and playing sport. These activities involve the coordination of muscles in many parts of the body with the inputs from the eyes, ears and other sensory organs. At the same time the conscious mind is occupied with the less complex objectives of these activities, such as where we are going, what are we writing or typing, or where to hit the tennis ball.
This multitasking of the unconscious mind does not occur when we are consciously multitasking. When we think we are consciously multitasking, the brain is actually switching from one task to the other, doing only one at any moment and interleaving the segments of each. This has been demonstrated using functional magnetic resonance imaging (fMRI) of the brain. (MRI is a procedure for examining internal tissues of the body, and fMRI examines the changes that occur in the tissues – in this case, the neurons – while the person is performing some task.) We can be conscious of only one thing at any moment, and when we are multitasking our attention is continually switching from one to another.
At any one time we are conscious of only a small amount of all the things that we are able to become conscious of. None of us is immediately aware of everything in our memory. But we recall (i.e., become aware of) specific parts of memory, either intentionally or spontaneously. We are often conscious of only a part of the information that our eyes, ears and other sensory organs send to the brain. When we are consciously preoccupied with something else, we can be oblivious of conditions that would otherwise be very painful, such as running barefoot over sharp stones to escape danger. So there appears to be some (unconscious) regulating process that selects the information that is allowed to be represented as consciousness.
All the contents of human consciousness seem to be dependent on the corresponding conditions of the brain. So, since our brains are the builders of all the content of our consciousness, then consciousness, whatever it is and however it occurs, must be just an observer, even though we usually feel that it is the active agent. The processing, that is, the intelligence, takes place in the brain. Since the content of consciousness is a series of representations, it is information. And the number of things that someone can be consciousness of depends on the amount of information contained in their brain. So the concept of expanding the consciousness by learning and discovery refers to the process of accumulation of information in the brain.
The processes by which information and intelligence occur in the brain can be demonstrated, and explained, in physical terms. But there is neither evidence nor any suggested physical process of how consciousness could be produced. The information and the processing in the brain are digital, but consciousness has the flavour of being analogue. That is, brains represent everything as arrangements of connections between brain cells, but what we see and feel, etc., in our consciousness as a result of these connections has a seamless continuity that resembles what we think the world is actually like.
Also, we can measure the information content and the processing capability of computers. We can also measure them in brains, although much less precisely because of practical difficulties. But while we can recognise degrees and intensity of consciousness, for example of pain or of anger, there is no objective way of measuring it.
We are unable to experience other people’s consciousness, and we can only infer that other people actually have consciousness. This inference is made by recognising peoples’ actions and facial expressions, etc. And people say “I feel happy, angry, astonished”, etc., and “my leg hurts”. And since we are so similar to everyone else, and we are conscious of being conscious and of having such feelings, we would need good reason to assume that humans generally were not conscious. However, with other humans we can only assume that what they actually experience is similar to what we experience.
Some dogs act as if they recognise that people are sad or angry, etc., so we may infer that dogs are conscious. And we extend that to some other species. But, given that not all human information and intelligence leads to consciousness, what criteria would allow us to conclude whether, for example, cockroaches, fungi, plants or bacteria, were conscious? It would be hard to know which things they were conscious of and which things they did unconsciously.
And what about computers? Robots, which are controlled by computers, can mimic many of the actions of humans and other organisms – not very elegantly, but it is still early days with the technology. But machine intelligence continues to increase. Robots can speak. So, presumably, a computer that could answer a tricky quiz question could also say “I feel happy, angry, astonished”, etc., and do the various other things that make us think that people are conscious.
We already have had several decades of experience with the Turing test, in which people have conversations using text with unseen computers or other people, and have to decide whether they are conversing with a computer or a person. The people doing the deciding often get it wrong. We have also had several years of movies with some of the characters being human-like robots and depicted as hard to distinguish as robots. What criteria would we need, in the real world, to determine whether a robot, irrespective of its appearance, was conscious? I am unable to answer this question, but there is also an underlying question, how could a computer or any other device obtain consciousness?
To be able to experience pain, computers would first need to get information from appropriate sensory organs. Computers already can have video cameras and microphones, etc., but that is not enough for consciousness. Some other functionality would be needed. But until we knew how consciousness could arise out of any physical arrangement we would be unable e to produce the relevant programs or to predict whether an inanimate device could actually experience anything.
Many people would say that consciousness makes us do, or enables us to do, what would be unlikely or impossible without it. They quote pain that keeps us safe. They quote the other senses that we enjoy and that make life richer. They quote ambition, empathy, love and other emotions. And they quote insight, imagination and inventiveness.
But all of these impulses come from the brain and can be stimulated and suppressed by interfering with the processes of the brain. Examples of such interference are drugs, such as alcohol, hallucinogens, and anaesthetics, and also magnetic and electric stimulation and suppression. The unconscious brain already seems to have everything needed for all the urges that we attribute to consciousness.
Sometimes it is claimed that consciousness is necessary for there to be meaning, so could a presumably unconscious computer derive meaning from a piece of information? I think it could.
Information in the form of computer programs is installed in computers to tell them how to operate. If there is an instruction telling a computer to perform a particular task under particular conditions at a particular time, the computer is able to derive the meaning of the instruction and perform the task that it was meant to do. And the range of tasks that computers can perform is very wide. To derive meaning, a computer – as with a person – just needs to already have the appropriate information. So when you connect a new type of attachment to a computer, you also need to put the relevant information into the computer in the form a program so that the computer will “know the meaning” of what the new application has to do. This is equivalent to teaching a new task to a person. In each case, if there is something wrong with the information, then neither the computer nor the person will get the right meaning of the task.
Many people would say that meaning is much more than that. For humans, meaning often does seem to be a lot more, because we have much bigger brains and a wider range of inputs. But is that a difference of degree or of kind? I think this is only a difference of degree. But I am not saying that the computer feels that it understands the meaning of the instruction.
I would say that, while I have described meaning (and data, information, and intelligence) as all being aspects of material processes, consciousness feels very different from just a material process. For example, we don’t like to accept that our feelings of amazement and awe, when we suddenly come upon a magnificent scene, are nothing more than a rearrangement of the connections between neurons in our brains.
And meaning is an identification of something. Consciousness is more than just an identification.
But many people do think that consciousness is produced by the brain, and they refer to various tests to support this view. Activity within specific areas of the brain can be detected in conjunction with associated kinds of consciousness. Also, there are particular areas in the brain that become active when someone is performing a particular task and thinking about what they are doing, but these areas are not active when the person is performing the same task without thinking about it. And many tests have demonstrated that such things as making decisions, and recognising inputs from the senses, occur in the brain before the person becomes conscious of them.
It is fairly common knowledge that neurotransmitters such as serotonin and dopamine are associated with emotion, and also that the amygdalae are the parts of the brain that are associated with emotion. Some people think this demonstrates that the brain produces consciousness. But neither neurotransmitters nor the other associated regions of the brain produce the conscious feelings associated with emotions. They are just parts of the brain processes that produce the information upon which consciousness depends.
But while all this information that is contained in the brain can be explained by the patterns of connections between neurons, there is as yet no way of explaining the nature of consciousness in terms of neural patterns. There is as yet no way of explaining how any material system could produce consciousness.
I have already described how a computer can use information consisting of sequences of 1s and 0s to activate a range of devices that performed different kinds of function; mechanical devices such as printers that produce documents and pictures, electronic devices that produce moving pictures on a video screen, and loudspeakers that produce music and other sounds. This might be compared with the way the brain uses information, encoded as networks of connections between neurons, to activate the muscles and other parts of the body to perform physical processes, and also produce the various kinds of content of consciousness. The brain and its processes are, of course, vastly more complex and sophisticated than computers. But there is a practical example to compare these processes of a computer with those of the brain and body.
In the brain and body, the motor nerves transmit electrical signals to activate systems of muscles. The muscles then perform manipulations of specific kinds, such as writing, walking or bodily functions. By detecting the relevant electrical signals it is possible to control specially-designed prostheses to perform actions similar to those of a hand or arm or leg. Such prostheses are used by people who have lost or badly damaged the particular body part that the prosthesis replaces. The invention of this kind of prosthesis was possible because we were able to observe and understand the brain-muscle-body process.
We have no consciousness of what is actually happening in most of these muscle processes. When we are conscious of them, this consciousness comes from our sensory organs, an entirely different kind of process. So, what are the processes that provide that other product of the brain, consciousness? There are neural networks that are active when we are conscious of something, and we might expect to see receptor components for them, equivalent to the “peripheral” components of a computer. However, there is no apparent component that is equivalent either to those of a computer or to the cells in the muscles that respond to the activation by motor nerves. The range of different kinds of content in consciousness is much greater than the range of peripheral components of computers, so we should be able to see at least one component somewhere in the brain or other part of the body. Will we eventually find such a component, or is consciousness something other than material?
There is a point of view that no “peripheral” components are necessary for consciousness: that consciousness is just the observation or review of the information or processing of the brain by other parts of the brain. But this says nothing about how this “observation” provides or discriminates between such things as itchiness, ecstasy, the smell of smoke and the feeling of uncertainty. Synaesthesia and the association of emotions with body feelings do not provide an answer. Associations of this kind are just aspects of the information in the brain, as are the associations of words with the concepts they represent. To me, this point of view is like expecting a computer to perform all of its functions without having any kind of monitor or printer or any other peripheral device.
The US philosopher Douglas Hofstadter in his book I Am A Strange Loop presents a case that self-awareness, and therefore consciousness, is the brain’s representation to itself of its own processes, i.e., by high-level networks repeatedly connecting back to each other. He illustrates this by showing what happens when a video camera is pointed at the screen that is displaying the picture from camera while he moves the camera about. But while this produces some interesting patterns on the screen, and while it is a self-referencing intelligent system, Hofstadter doesn’t claim that it has actually produced consciousness. And he doesn’t say why any such process would be expected to produce consciousness. But then, in the same book he claims that the flushing system of a toilet, with not a skerrick of self-representation, is conscious. He gives no suggestion as to how the toilet system obtains its consciousness. And to the best of my knowledge no one else has.
As with many other, mainly atheist, philosophers, Hofstadter has confused intelligence with consciousness.
Well, that was the chain of development from data to consciousness. There is, of course, much more that goes beyond this discussion, such as the uses and complexities of language and other structures built out of information. But I have been looking at the matter from a perspective analogous to that of atoms and molecules.
However I think that there are still two issues to be resolved:
- is consciousness necessary before we could have had the initial concept of data, which would mean that all this has not been a well-founded structure but just a circular argument?
- and is my argument unduly biased towards materialism?
So, was consciousness necessary before we could have had the initial concept of data? I think the issue is that data is a representation of something, not the thing itself. The processes that led to the development of brains and sensory organs did not rely on the presence of consciousness or data. The brains and sensory organisms created the representations.
When we are seeing something, our brains are processing data that were not consciously assigned. Such data arise when light enters the eye and causes electrical impulses in the optic nerve. The electrical pulses are data, representing the light. The processing of a series of such data in the brain results in a conscious picture. This picture that we are conscious of is a representation of the things that reflected or emitted the light that entered our eyes.
All of our sensory inputs, and our memories of them, and our thoughts about them, are also representations in our brains. None of these representations – these pieces of information or collections of data – were consciously assigned. The associations between specific kinds of phenomena and their representations within the brain and other parts of the body are the results of long processes of evolution.
Can there be such a thing as consciousness without any content? I think there cannot. The absence of conscious content is unconsciousness.
Since everything that is in our consciousness depends on the information contained in our brains, our concept of data depends on the information already in the brain. And this information is derived from evolved structures in the brain and the various inputs that human brains have collectively gathered and processed over the ages.
And since consciousness is the outcome of this representative chain starting with data, it is also the outcome of the more tangible chain starting with actual phenomena in the outside world. Material phenomena created our brains and our sensory organs. Material phenomena stimulate our sensory organs, which send information to our brains and this is then sometimes experienced as consciousness.
So the sequence from data to consciousness is not self-referencing.
As for the second issue, the point of view of this essay has indeed been mainly materialistic. It regards the intelligence of computers and mousetraps as being the same kind of phenomenon as human intelligence. It is close to being deterministic, it throws doubt on free will, and it could make us feel that we are just machines. But I think it is logical and based on the available evidence.
But the thrust of this essay has not been entirely materialistic. I have also said that there is as yet neither evidence nor any viable suggested physical process of how consciousness could be produced, and also that many parts of consciousness feel very unlike the tangible realities of the material world. This does not mean that consciousness could not be the result of a material process. But I think that the idea that consciousness is non-material still has to be a possibility. However, I have no idea of what such a non-material entity would be like or how it might produce consciousness.
So what has been outlined throughout this essay is probably the closest we can get with present-day science to account for what I have called the essence of our existence.
Yet I wonder whether the unconscious mind is also an essential part, perhaps the greater part.