Thinking and sleeping
One of the apparent delights of the English is the profusion of overlapping words and alternative terms, the very reason it is such a challenging language to learn. Every time I use a thesaurus, I have the same two reactions. First, here is another word I can use: thank goodness. Second, who knew there were so many words with similar meanings? All this would be merely interesting, were it not that one current obsession is with artificial intelligence and ‘thinking machines’. This preoccupation raises several confusing questions about what we mean by ‘thinking’ and the words we use to describe cognitive processes.
At one level, thinking about thinking can be fun. When the term ‘knowledge worker’ came into vogue, I used to tease managers as to how they would monitor work performance. Suppose you see someone at his or her desk, eyes closed but awake. You ask: “What are you doing?” They reply: “Thinking!” How would you know if this was true? It was teasing, but it had a serious edge. Knowledge work isn’t physical work, and thinking is an important part of many jobs. What we can’t see is mysterious: we know that we think; we believe that everyone else does it, but we are unable to see it happening, only the outcomes. Scientists become very exciting about various ways to track activity in the brain while someone undertakes a task or thinks, but we don’t ‘see’ the thinking, we just observe electrical activity.
This is rather similar to another process that has intrigued researchers for centuries, and this is sleep. Here is another scenario to test a manager. A member of staff likes to take an afternoon nap. Sleeping on the job? I suspect many managers would like to deduct this time from hours worked. But a recent article draws on several research studies that show too little sleep is harmful to health, increases the risk of making errors, and impairs cognitive ability.[i] We want people to have enough sleep, but we don’t want them to be sleeping when they should be working! Shouldn’t staff be sleeping on their own time? And if they shouldn’t be taking time off to sleep at work, couldn’t they do their thinking at home, too? For many managers, being busy is about being seen to be busy!
Before I get into a large muddle, perhaps I should try to work out what I mean when I write about thinking. As I see it, there are at least two very different activities that come under the umbrella of thinking. There is thinking as a rational process, processing data. This form of thinking allows humans to make sense of, interpret, represent or make models of the world they experience, and to make predictions about action and its consequences. It is an activity that helps us address our needs, objectives, and desires, enabling us to make plans or otherwise attempt to accomplish our goals.
In this definition we are being offered a rather prosaic view of what we do when we think. It is a view that informs much of the work being done to develop artificial intelligence, smart devices and autonomous cars. Computers are already formidable logical processors, never take time off and can do in milliseconds what it would take teams of humans years to accomplish.
Does that make computers intelligent? I don’t think so. However, if a knowledge worker is just a processing machine, then there is every good reason to replace slow human processors with faster and far more reliable computers.
Many people like to draw a distinction between being intelligent and being ‘smart’. Intelligent people are the rational thinkers I have just been describing. They use logical processes. They generate knowledge that can be written down, explained, step by step.
On the other hand, I would argue smart people use a different kind of intelligence, relying on tacit knowledge rather than codified knowledge. It is using what has been learnt from experience, something we might call practical intelligence. For many years, one popular example of this was provided through a study of Xerox maintenance engineers. Each engineer carried a comprehensive handbook, detailing the machine error codes which appeared if a copier stopped running properly, and outlined the corrective steps required to fix the problem. However, in everyday practice, faced with an unresponsive machine these same engineers would often try some other steps to solve the issue, based on ideas they had learnt from experience, or had heard about from a colleague. Sometimes this tacit knowledge would lead to a new section being added to the repair handbook. Sometimes, it wasn’t ‘logical’, but worked. Quite often the approach was individual: “when this particular machine stops producing clear copies, do this!” Some of these solutions were unique and did not work with another copier.
Even though this knowledge wasn’t ‘logical’, it was something that could be learnt and it could be put into practical application through the use of appropriate tasks and processes. However, I would argue there is another form of thinking which is called ‘imagining’. Imagination is not tied to real events in the external world, nor is it the consequence of facts, or any kind of immediate sensory input. Rather it is the capacity to produce images, ideas and sensations in the mind. Imagination creates something new, “in the mind’s eye” as it is often expressed, which is then realised in a creative outcome – which might be physical, a piece of art, a new approach in a chess game; or realised in a non-physical medium, a story for example.
Given this, we can now ask “What does the manager want from a ‘knowledge worker’?” Most of the time I would guess they want a problem solved, a process improved, a logical analysis of a situation allowing a better understanding of what was happening and how it could be changed. However, some of the time the manager might want more. Something creative, innovative: an outcome outside the processes of logic.
This leads us to the second dilemma for the manager. Not only it is hard to determine if a person is thinking, but it can be hard to understand or acknowledge an idea that falls outside of what appears to ‘make sense’. New ideas developed by logical processing are quite different from new ideas generated by imaginative thinking.
I confronted an example of this a few months ago when discussing computer generated music. Mozart was a very imaginative, creative composer. He did things that no-one had done before. Some might say it was incremental innovation, rather than radical, but he certainly composed in ways that were unfamiliar. Today, it is (relatively) easy to compose in a Mozart-like fashion. Given a number of compositional rules and an initial tune, a computer can be programmed to churn out a symphony in his style in no time at all. In the same way, if you are a keen songwriter, once you have a sketchy tune in your head; smart software can convert that into a composition for a rock group or a chamber orchestra. It will sound good.
That is quite different from what Arnold Schoenberg did. He developed twelve-tone technique as a method of musical composition. The technique is a means of ensuring that all 12 notes of the chromatic scale in such a way that all 12 notes are thus given more or less equal importance, and the music avoids being set in a particular musical key. That was imaginative thinking, something a computer could never do (unless you wanted to try the ‘monkeys and Shakespeare’ approach: have enough monkeys playing with enough forms of musical composition that that, by chance, one would come up with twelve-tone technique). It wasn’t well received to begin with. But here’s the other side to this: over time, a new technique will become ‘established’. In this case, this compositional form became more popular, and eventually became widely influential on 20th century composers. Many who didn’t originally support this new direction or even actively opposed this technique eventually adopted it in their music.
We can ‘understand’ logical thinking, but how does imaginative thinking take place. One source may be our sub-conscious. A famous example that has been cited over the years was the biochemist Kekulé who was said to have discovered the hexagonal structure of benzene by dreaming about snakes in a circle, the head of one swallowing the tail of the next. Sadly, we’ve had to abandon the story, as it turns out the story is untrue.[ii] Despite that, most people have experienced a time when, on waking or after thinking about another issue, a new idea pops up, as if it had been lurking out of sight (in the sub-conscious).
Can we ‘make’ someone think creatively? Again, I have to say, I don’t think so. However, we can create conditions which allow imaginative thoughts to flow more freely. Indeed there is a whole industry of consultants and specialised companies who run creativity workshops. Techniques like brainstorming, word association and mind-mapping are used to set aside the stranglehold of ‘logical’ thinking. They help in the sense that they offer the participants the opportunity to be freed from saying what is ‘right’ or makes sense. The downside of these techniques is that they often generate a large number of ideas, and then it takes some time to reduce those to a list that offers a new and useful step: there is a lot of noise produced by sometimes hundred or plain silly suggestions.
The man or woman in the office spending time thinking may use some tricks to free their minds, but quite often the core of the process is devoted to waiting for ideas to come up: no wonder our manager was frustrated about the staff member claiming to be thinking – they may have been doing very little at all, just hoping a good idea would emerge.
Where does sleep fit into all this? Over the last sixty years, out knowledge of sleep has grown dramatically. Some of sleep is devoted to what we might call system management. The brain uses significantly less energy during sleep than it does when awake, especially during non-REM (rapid eye movement) phases. In areas with reduced activity, the brain restores its supply of key chemicals, such as the molecule used for short-term memory storage and the human growth hormone. Dreaming seems to be a function of REM sleep, when the brain waves increase in speed and are closer to a waking state, eyes move about, and otherwise the body appears to be in a state akin to paralysis.
Apart from the restorative side to sleep, dreaming has long been a source of fascination. People woken during REM stages of sleep report the images and stories they have experienced. There is no agreement yet as to what dreaming is, although there is a lot of evidence to suggest that one function is to take material in temporary memory, analyze it, and sort it out, before encoding it in more permanent memory. The contents of temporary memory are often jumbled, unrelated, and partly related to experiences of the preceding day. During this time, parts of the brain are operating at a reduced level, included the logical areas in the pre-frontal cortex, reducing the extent to which any sense making has to be processed in a step by step fashion, freeing the brain to entertain and explore quite bizarre and unconstrained ideas.
To the extent this might be true, there is also some evidence that the initial phases of dreaming are followed by a period of time when the pre-frontal cortex returns from a relatively inactive phase to be back online, as it were. During this time, it may be that the brain tries to make sense of what has been experienced, and the dream may be converted into a narrative, however bizarre the story may still appear. If sleep contributes to creative thinking, it might be the activity that goes on during REM sleep that is an important contributor, with the immediate post-REM period trying to ‘make sense’ of what was experienced.
Perhaps our manager has a good reason to be concerned about a staff member sleeping at work. It might lead him or her to come up with new ideas, but while some could be creative, others could be bizarre or worse!
As far as we know, computers don’t dream. They are unlikely to be creative in the way that humans are. Of course, I can already hear those who would point to the story of AlphaGo, the computer program that defeated South Korean Go grandmaster Lee Se-Dol, in a series of five games in 2016 (AlphaGo won 4 out of five games). At the time, boosters of the computer commented on its creative approach, but Lee was far more perceptive. Like a human player, he noted that AlphaGo had difficulty in responding to unexpected plays. Like a human player, it had a store of previous games, and learnt from every new game better and worse moves. But it was still a processing machine, and lost one game because Lee behaved unpredictably! Once only with that specific tactic, because after the game AlphaGo had one new game in its experience bank.
I would like to offer Shakespeare the last word. Hamlet, in his famous soliloquy considers “to sleep perchance to dream”. He is thinking about suicide, but, illogically, he is worried that, even when dead he will continue to face the troubles that haunted him in life. The lines of an imaginative character in a play, written by a creative genius. A logical device could come up with that sentence, but what make it memorable is not the words but the context. How clever of Shakespeare, we might add, to come up with five words centred on the theme of this blog, and to do some 400 years ago!
[i] Joel Frolich, The Wisdom of the Sloth, < https://knowingneurons.com/2018/06/14/sloth-wisdom/>
[ii] <https://www.nytimes.com/1988/08/16/science/the-benzene-ring-dream-analysis.html>
